diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 487a3ecd0385dfeb173192a250696c1371fa84f2..41e0512ee8312d893dae32a940a63de47dcbf125 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": "20260429T180023Z", - "generated_at": "2026-04-29T18:07:38Z", + "snapshot_id": "20260430T000042Z", + "generated_at": "2026-04-30T00:10:16Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1149,152 +1149,74 @@ ] }, { - "cluster_id": "cluster-43656-4", - "summary": "Cluster of 1 issues and 3 PRs centered on issue #43824.", + "cluster_id": "cluster-43240-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43240.", "status": "open", - "confidence": 0.8, - "canonical_issue_number": 43824, - "canonical_pr_number": 43656, + "confidence": 0.75, + "canonical_issue_number": 43240, + "canonical_pr_number": 43251, "issue_numbers": [ - 43824 + 43240 ], "pr_numbers": [ - 43656, - 43836, - 43842 + 43251, + 43254 ], "evidence_types": [ "closing_reference", - "shared_issue_target", - "soft_similarity" + "shared_issue_target" ], "pr_comparisons": [ { - "left_pr_number": 43656, - "right_pr_number": 43836, - "code_similarity": 0.36, - "size_similarity": 0.176, + "left_pr_number": 43251, + "right_pr_number": 43254, + "code_similarity": 0.794, + "size_similarity": 0.64, "file_overlap": 1.0, - "area_overlap": 0.051, - "patch_similarity": 0.048, + "area_overlap": 0.667, + "patch_similarity": 0.882, "shared_filenames": [ - "src/transformers/cli/serve.py" + "src/transformers/loss/loss_utils.py" ], "shared_file_areas": [ { - "filename": "src/transformers/cli/serve.py", + "filename": "src/transformers/loss/loss_utils.py", "left_ranges": [ - [ - 11, - 18 - ], [ 30, - 36 - ], - [ - 315, - 323 - ], - [ - 665, - 671 - ], - [ - 931, - 937 - ], - [ - 1843, - 1849 - ], - [ - 1868, - 1874 + 50 ] ], "right_ranges": [ - [ - 11, - 18 - ], - [ - 359, - 429 - ], - [ - 584, - 590 - ], - [ - 1892, - 1910 - ], - [ - 1917, - 1923 - ] - ] - } - ] - }, - { - "left_pr_number": 43656, - "right_pr_number": 43842, - "code_similarity": 0.405, - "size_similarity": 0.5, - "file_overlap": 1.0, - "area_overlap": 0.0, - "patch_similarity": 0.036, - "shared_filenames": [ - "src/transformers/cli/serve.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/cli/serve.py", - "left_ranges": [ - [ - 11, - 18 - ], [ 30, - 36 - ], - [ - 315, - 323 - ], - [ - 665, - 671 - ], - [ - 931, - 937 - ], - [ - 1843, - 1849 - ], - [ - 1868, - 1874 - ] - ], - "right_ranges": [ - [ - 54, - 61 - ], - [ - 587, - 593 + 43 ] ] } ] - }, + } + ] + }, + { + "cluster_id": "cluster-43824-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43824.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43824, + "canonical_pr_number": 43836, + "issue_numbers": [ + 43824 + ], + "pr_numbers": [ + 43836, + 43842 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ { "left_pr_number": 43836, "right_pr_number": 43842, @@ -1346,56 +1268,6 @@ } ] }, - { - "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.", @@ -1727,7 +1599,7 @@ "cluster_id": "cluster-45561-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", "status": "open", - "confidence": 0.8, + "confidence": 0.55, "canonical_issue_number": 45561, "canonical_pr_number": 45645, "issue_numbers": [ @@ -1738,9 +1610,7 @@ 45645 ], "evidence_types": [ - "closing_reference", - "shared_issue_target", - "soft_similarity" + "closing_reference" ], "pr_comparisons": [ { @@ -2103,16 +1973,6 @@ "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-43656-4", - "canonical_pr_number": 43656, - "duplicate_pr_numbers": [ - 43836, - 43842 - ], - "target_issue_number": 43824, - "reason": "PRs in cluster-43656-4 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target, soft_similarity evidence." - }, { "cluster_id": "cluster-43698-3", "canonical_pr_number": 43779, @@ -2122,6 +1982,15 @@ "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-43824-3", + "canonical_pr_number": 43836, + "duplicate_pr_numbers": [ + 43842 + ], + "target_issue_number": 43824, + "reason": "PRs in cluster-43824-3 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target evidence." + }, { "cluster_id": "cluster-43979-11", "canonical_pr_number": 44007, @@ -2164,7 +2033,7 @@ 45639 ], "target_issue_number": 45561, - "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference, shared_issue_target, soft_similarity evidence." + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference evidence." } ], "best_issue": { diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index ee72a1afd19be8aff3fda0fd12321c85e3585758..f9edecb733ece301ebfd6e9db2e7a9d00e0970e4 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": "20260429T180023Z", - "generated_at": "2026-04-29T18:07:38Z", + "snapshot_id": "20260430T000042Z", + "generated_at": "2026-04-30T00:10:16Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -26,12 +26,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "bcb9be742c1b65129046184bfc5f17388b62d222e450f95a680819a69a023c43", + "prepared_review_unit_hash": "653a36cc0fa13814dd047e575350566e83973ed7e845de8e6418d49b2d89e3c5", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2479, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -39,7 +39,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2479, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -48,24 +48,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues do not look like duplicates. Two are about model loading/error handling, but they target different scenarios; the third is unrelated tokenization behavior. No PRs are present.", - "confidence": 0.86, - "canonical_issue_reason": "issue #16618 is the broadest and most central of the set, with the most discussion around a general model-loading failure.", + "summary": "These issues are only loosely related by Hugging Face model/tokenizer usage, but they describe different problems: tokenization mismatch, private-model auth/download messaging, and a failure loading bert-base-uncased. I would not merge any of the soft pairs as duplicates.", + "confidence": 0.82, + "canonical_issue_reason": "issue:16618 is the broadest and most user-visible report in the set, but the three issues do not appear to be duplicates of one another.", "canonical_pr_reason": null, - "best_issue_reason": "#16618 is the best anchor only by breadth and activity; it is still not clearly the same bug as the other issues.", + "best_issue_reason": "issue:16618 is the best representative only in the sense that it is the clearest general model-loading failure; the others are narrower and about different code paths.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tokenization mismatch vs. model loading failure are different problem areas and do not indicate the same underlying bug." + "reason": "Tokenization behavior in Tapas versus TensorFlow code is a different underlying bug than failing to load bert-base-uncased." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "#15980 is specifically about a private model download/authentication error, while #16618 is a general 'can't load bert-base-uncased' report; same subsystem, but not the same concrete failure." + "reason": "A bad error message for unauthenticated private-model downloads is an auth/download UX issue, not the same bug as loading a public model failing." } ] }, @@ -87,7 +87,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:24643", "issue:30277", @@ -173,8 +173,8 @@ "issue:34928|issue:41720", "issue:43716|issue:45685", "issue:43541|issue:43716", - "issue:43828|issue:44292", "issue:45507|issue:45685", + "issue:43828|issue:44292", "issue:43716|issue:44292", "issue:44292|issue:44918", "issue:41762|issue:43638", @@ -187,8 +187,8 @@ "issue:30277|issue:36331", "issue:43039|issue:44805", "issue:34928|issue:36010", - "issue:41762|issue:43596", "issue:43901|issue:44512", + "issue:41762|issue:43596", "issue:39290|issue:41720", "issue:43866|issue:44863", "issue:43039|issue:44560", @@ -206,437 +206,326 @@ "issue:44841|issue:45084", "issue:41093|issue:41762" ], - "prepared_review_unit_hash": "983d91afd356fa80b76e00cb493d3751127a36c1d1b3d67c66fd0dcf08fb5d31", + "prepared_review_unit_hash": "f2639e21844f4bc54c73177296a8ed4ce4b51c9e01eafc039cae3debc36da180", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 67, - "serialized_chars": 47121, - "estimated_input_tokens": 11781, - "estimated_eval_tokens": 23818 + "serialized_chars": 47120, + "estimated_input_tokens": 11780, + "estimated_eval_tokens": 23816 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous. The only nearby subcluster is the DeepSpeed ZeRO-3/Bert index-error reports, but the available titles still don\u2019t prove a single shared root cause. I\u2019d keep one broad zero3/Bert index-error issue as the closest canonical anchor and reject the rest as non-duplicates.", - "confidence": 0.57, - "canonical_issue_reason": "issue:44805 is the broadest of the repeated ZeRO-3 index/mask failure reports, so it is the best umbrella candidate for that small subcluster.", + "summary": "This cluster is mostly heterogeneous, but there is a small duplicate set around DeepSpeed ZeRO-3 loading failures with `index 0 is out of bounds for dimension 0 with size 0`. Most other pairs are only loosely related by broad subsystem or error family and should stay separate.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 43596 is the cleanest representative of the shared ZeRO-3 loader failure; 43638 and 41762 look like model-specific variants of the same underlying `index 0` out-of-bounds bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44805 is the most general and reusable issue among the similar zero3/Bert index-error reports, making it the best representative if this subcluster is revisited.", + "best_issue_reason": "43596 is the best issue anchor because it captures the common DeepSpeed ZeRO-3 initialization failure in a generic way, making it the most suitable duplicate target for the accepted variants.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different Qwen3.5 failures: missing config field vs streaming chat/completions error." + "reason": "Different Qwen3.5 failures: missing config attribute vs streaming chat/completions inference error." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Unrelated hardware/runtime problems on different backends (AMD GPU vs MPS int histogram)." + "reason": "Different backend/model problems: AMD GPU support for GPT-OSS vs MPS int histogram error in MoE." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different model path and error type: image preprocessor dtype mismatch vs AMD GPU runtime issue." + "reason": "Different model families and failure modes: Mistral-3 dtype mismatch vs GPT-OSS on AMD GPUs." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Both are model/runtime failures, but not the same bug or code path." + "reason": "Both are dtype/backend complaints, but they affect different models and code paths." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Different models and failures; no shared concrete bug." + "reason": "AMD GPU runtime issue vs GraniteMoEHybrid calling an invalid method; not the same bug." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Different Qwen NVFP4 loading issue vs AMD GPU execution failure." + "reason": "Qwen3 NVFP4 runtime failure is unrelated to the GPT-OSS AMD GPU issue." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "Different subsystems and symptoms: ZeRO3 collective mismatch vs ONNX export error." + "reason": "DeepSpeed collective mismatch is unrelated to ModernBERT ONNX export." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "Different model/export issue vs FineGrainedFP8 failure." + "reason": "ONNX export error and FineGrainedFP8Config failure are different model/tooling issues." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "ONNX export bug and auto device-map cuda assert are not the same code-path problem." + "reason": "ModernBERT export to ONNX is unrelated to Qwen3 auto-device-map CUDA assert." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are v5 docs cleanup, but they concern different removed references and likely different files." + "reason": "Both are stale docs problems, but they reference different removed features and do not look like the same underlying change." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export error is unrelated to the GenerationMixin import breakage." + "reason": "ONNX export error vs `GenerationMixin` import breakage are separate problems." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Related ZeRO-3/index-error family, but the reported failures differ enough to avoid merging confidently." + "reason": "Both involve shape/index errors under ZeRO-3, but the specific failure messages and contexts differ enough to avoid deduping." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Same general area, but one is an empty-tensor index error and the other is a mask/tensor shape mismatch." + "reason": "ZeRO-3 BertModel index-out-of-bounds is not the same as the mask-shape mismatch report." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Different issue classes: ZeRO-3 indexing failure vs GPT-OSS tensor-parallel bug." - }, - { - "left": "issue:44805", - "right": "issue:44918", - "accept": false, - "reason": "Different training stack problem; no clear shared bug." + "reason": "Mask/tensor shape mismatch and GPT-OSS tensor-parallel failure are unrelated." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Tokenizer-mapping logic bug is unrelated to the ZeRO-3 index/mask error." + "reason": "Tokenizer mapping bug is unrelated to the mask/tensor shape mismatch issue." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "Bark voice preset issue is unrelated to ModernBERT ONNX export." + "reason": "BarkProcessor voice preset and ModernBERT ONNX export are different subsystems and failures." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both mention DeepSpeed, but the failures and models are different." + "reason": "Collective mismatch during ZeRO-3 is a different failure from the ZeRO-3 index-out-of-bounds loader bug." }, { "left": "issue:43596", "right": "issue:43638", - "accept": false, - "reason": "Very similar family, but the model/setup details differ and the titles do not identify the same exact bug." + "accept": true, + "reason": "Same ZeRO-3 loader/index-out-of-bounds bug on BertModel; the wording differs but the failure is the same." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Different model/output bugs; no shared underlying change." + "reason": "Incorrect detokenization and LoRA save/reload mismatch are distinct model-output issues." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy call vs GPT-OSS tensor-parallel failure are unrelated." + "reason": "Liger Kernel cross-entropy path and GPT-OSS tensor parallel issue are not the same bug." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Qwen NVFP4 execution failure is unrelated to the LoRA merge/save issue." + "reason": "Qwen NVFP4 runtime failure is unrelated to LoRA merge/save inconsistency." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpointing/FSDP recompute mismatch is a different training issue." + "reason": "Activation-checkpointing/FSDP recomputation mismatch is a different shape bug than the ZeRO-3 loader issue." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Different DeepSpeed symptoms and different model paths." + "reason": "DeepSpeed training `'weight' must be 2-D` is unrelated to the collective mismatch report." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "ZeRO-3 Bert index error and GPT-OSS TP failure are not the same bug." + "reason": "ZeRO-3 index-out-of-bounds and GPT-OSS tensor-parallel failure do not share the same code path." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "Activation checkpointing bug is unrelated to compute_loss signature breakage." + "reason": "FSDP/activation checkpointing recomputation mismatch is unrelated to the `compute_loss` signature change." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "Different subsystems and failures." + "reason": "Different training/load failures: ZeRO-3 BertModel loading vs GPT-OSS tensor-parallel behavior." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Different model families and unrelated runtime errors." + "reason": "Mistral-3 dtype mismatch and GraniteMoEHybrid invalid method are unrelated." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3 vLLM config issue is unrelated to the Mistral tokenizer regex crash." + "reason": "Gemma3 sliding-window attribute error is unrelated to the Mistral tokenizer patch crash." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Autocast dtype mismatch is not the same as GPT-OSS tensor-parallel failure." - }, - { - "left": "issue:44292", - "right": "issue:44918", - "accept": false, - "reason": "Different Qwen-related failures, but not the same underlying bug." - }, - { - "left": "issue:34928", - "right": "issue:41720", - "accept": false, - "reason": "FSDP/activation-checkpointing mismatch is unrelated to auto device mapping cuda asserts." - }, - { - "left": "issue:43716", - "right": "issue:45685", - "accept": false, - "reason": "Mistral-3 dtype mismatch and MPS histogram error are unrelated." - }, - { - "left": "issue:43541", - "right": "issue:43716", - "accept": false, - "reason": "Different precision/tracing failures; no same code path." - }, - { - "left": "issue:43828", - "right": "issue:44292", - "accept": false, - "reason": "Different model families and different failure modes." - }, - { - "left": "issue:45507", - "right": "issue:45685", - "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid is unrelated to MPS histogram support." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to GPT-OSS tensor-parallel failure." }, { - "left": "issue:43716", - "right": "issue:44292", + "left": "issue:44918", + "right": "issue:45161", "accept": false, - "reason": "Different models and error paths." + "reason": "Qwen3.5 embedding unpacking with TRL is unrelated to GPT-OSS MoE tensor-parallel behavior." }, { - "left": "issue:44292", - "right": "issue:44918", + "left": "issue:43596", + "right": "issue:45161", "accept": false, - "reason": "NVFP4 runtime error vs TRL SFT unpacking failure are unrelated." + "reason": "ZeRO-3 loader bug and GPT-OSS tensor-parallel bug are different code paths." }, { "left": "issue:41762", "right": "issue:43638", - "accept": false, - "reason": "Both are zero3 loading/indexing issues, but the model and exact failure differ." + "accept": true, + "reason": "Same ZeRO-3 `index 0` out-of-bounds loader failure, just on a different model family." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger Kernel cross-entropy issue is unrelated to TRL embedding unpacking." + "reason": "Liger Kernel cross-entropy call and TRL Qwen3.5 embedding unpacking are unrelated." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions, but they refer to different removed references and likely different edits." - }, - { - "left": "issue:43716", - "right": "issue:44560", - "accept": false, - "reason": "Dtype mismatch and StopIteration in video handling are unrelated." - }, - { - "left": "issue:43828", - "right": "issue:45685", - "accept": false, - "reason": "Autocast dtype mismatch is unrelated to MPS integer histogram support." - }, - { - "left": "issue:44918", - "right": "issue:45685", - "accept": false, - "reason": "TRL SFT embedding issue is unrelated to the MPS histogram backend error." - }, - { - "left": "issue:43039", - "right": "issue:43828", - "accept": false, - "reason": "Different models and different runtime failures." - }, - { - "left": "issue:30277", - "right": "issue:36331", - "accept": false, - "reason": "DeepSpeed collective mismatch is unrelated to the CustomTrainer signature break." - }, - { - "left": "issue:43039", - "right": "issue:44805", - "accept": false, - "reason": "Liger Kernel cross-entropy call is unrelated to ZeRO-3 index/mask errors." - }, - { - "left": "issue:34928", - "right": "issue:36010", - "accept": false, - "reason": "Different stack layers and failure types." - }, - { - "left": "issue:41762", - "right": "issue:43596", - "accept": false, - "reason": "Both are ZeRO-3 related, but the loading failures are not the same concrete bug." - }, - { - "left": "issue:43901", - "right": "issue:44512", - "accept": false, - "reason": "Different documentation regressions touching different APIs." - }, - { - "left": "issue:39290", - "right": "issue:41720", - "accept": false, - "reason": "Gemma3/vLLM config mismatch and Qwen3 auto-device-map assert are unrelated." + "reason": "Both are docs issues, but they concern different removed v5 features and look like separate cleanup tasks." }, { "left": "issue:43866", - "right": "issue:44863", - "accept": false, - "reason": "Corrupted checkpoint and checkpoint-loading implementation bug are not the same issue." - }, - { - "left": "issue:43039", - "right": "issue:44560", - "accept": false, - "reason": "Liger Kernel cross-entropy call is unrelated to Qwen3-vl video StopIteration." - }, - { - "left": "issue:43854", - "right": "issue:43866", + "right": "issue:43854", "accept": false, - "reason": "Model-load failure in tests vs corrupted checkpoint are different problems." + "reason": "Checkpoint corruption and unit-test load failure for GLM-4.7-Flash are different issues." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are index-related, but the reported shapes and setup do not match the same bug." + "reason": "Mask/index length mismatch and ZeRO-3 loader index-out-of-bounds are different failure modes." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Gemma3 sliding-window config issue is unrelated to Qwen3-MoE sliding_window handling." + "reason": "Gemma3 sliding-window attribute error is unrelated to Qwen3-MoE sliding_window behavior." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch version compatibility and pydantic PretrainedConfig breakage are unrelated." + "reason": "Torch version incompatibility and pydantic `PretrainedConfig` field breakage are unrelated." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Same model family, but different surfaces and no evidence of the same root cause." + "reason": "Qwen3.5 chat crash and Qwen3.5 streaming inference failure are separate runtime problems." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Tied lm_head bug is unrelated to checkpoint corruption." + "reason": "Mistral3 tied-weight bug is unrelated to Ovis2 checkpoint corruption." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper runtime issue is unrelated to the pydantic model field breakage." + "reason": "Whisper run failure is unrelated to the pydantic config-model issue." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 streaming failure and MPS histogram backend error are unrelated." + "reason": "Qwen3.5 streaming failure and MoE MPS histogram error are unrelated." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Grouped_mm tracing failure is unrelated to the MPS backend limitation." + "reason": "Grouped_mm tracing error and MPS histogram-not-implemented error are different execution paths." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both involve dtype/precision behavior, but not the same concrete bug or code path." + "reason": "Both are numerical/runtime issues in MoE-ish models, but they are not the same bug or fix." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "GenerationMixin import breakage is unrelated to Gemma3/vLLM config issues." + "reason": "GenerationMixin import breakage and Gemma3 sliding_window attribute error are unrelated." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Different model/processor issue vs template compilation error." + "reason": "Voxtral processor failure and template compilation error are different issues." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are shape/index errors, but the model and failure mode differ." + "reason": "Mask-shape mismatch and ZeRO-3 index-out-of-bounds are different bugs." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the ZeRO-3 duplicate cluster around 43596/43638/41762 is supported by the titles. The negative soft-edge verdicts are mostly conservative and do not overclaim shared root causes beyond what the packet shows." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -651,7 +540,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:39186", "issue:39290", @@ -737,238 +626,238 @@ "issue:43828|issue:44928", "issue:43866|issue:44403" ], - "prepared_review_unit_hash": "3ed520963970f2f98588f432d5845e236a8fc903b8735c745a6e103894b2683c", + "prepared_review_unit_hash": "95e5979fce976694d7de52a9d67a135591463c353438dfa4e55d43a090c767c7", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37183, - "estimated_input_tokens": 9296, - "estimated_eval_tokens": 18848 + "serialized_chars": 37179, + "estimated_input_tokens": 9295, + "estimated_eval_tokens": 18846 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issue set. I only see two clear duplicate groups: the non-persistent buffer corruption regression, and the local custom-module cache/path collision. The rest are only broad model-family or error-type similarities, not the same concrete bug.", - "confidence": 0.81, - "canonical_issue_reason": "issue:43950 is the best anchor for the non-persistent-buffer regression: it states the concrete corruption behavior, the 5.x regression scope, and it sits between the shorter follow-up reports (43644, 44534).", + "summary": "The cluster is mostly a grab-bag of unrelated Transformer/vLLM issues. The clearest duplicate set is the non-persistent buffer corruption regression in v5.x. A second likely duplicate pair is the SigLIP2 bad-output reports. Most other pairs only share a subsystem or vague symptom class and should stay separate.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43950 is the clearest and broadest statement of the non-persistent-buffer regression, with an explicit `from_pretrained()` corruption description that subsumes the narrower \"junk\" reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43950 is the strongest representative issue in this set overall because it most clearly names the root regression and matches the accepted duplicate chain around buffer corruption.", + "best_issue_reason": "issue:43950 is the best representative issue for triage because it most precisely names the regression, its scope, and the loading path involved.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different model/loading problems; one is a PaddleOCR-VL vLLM load failure, the other is GGUF gpt-oss support." + "reason": "Different models and failures: PaddleOCR-VL load failure in vLLM vs GGUF/gpt-oss architecture support." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Both mention config-field regressions, but the titles point to different symptoms and scopes; not enough to treat as the same bug." + "reason": "Both involve config-related v5 breakage, but one is a missing StableLmConfig field and the other is a pydantic/PretrainedConfig field regression; not the same bug." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Unrelated regressions: int4 OOM vs tied-weights handling." + "reason": "One is int4 quantization memory growth/OOM, the other is tied-weights behavior on translation models; different code paths." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Both touch MoE models, but the concrete failures differ: torch dynamo grouped_mm tracing vs tensor-parallel support." + "reason": "Mixtral torch-dynamo grouped_mm tracing failure is unrelated to GPT-OSS tensor-parallel support." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Different code paths and symptoms: autocast dtype mismatch vs streaming chat/completions failure." + "reason": "Both are runtime failures, but one is an autocast dtype mismatch in Phi-tiny-MoE and the other is a streaming chat/completions failure in Qwen3.5." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "init_empty_weights argument handling is a different failure mode from the pydantic PretrainedConfig field regression." + "reason": "Unexpected `_is_hf_initialized` during init_empty_weights is a different loading-path bug from the pydantic PretrainedConfig field issue." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "Different models and failures; gpt-oss GGUF support is unrelated to Qwen3.5 streaming inference." + "reason": "GPT-OSS GGUF support and Qwen3.5 streaming API failure are unrelated." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is not the same bug as the PretrainedConfig/pydantic field issue." + "reason": "Qwen3-MoE sliding_window handling and pydantic config-field breakage are different regressions." }, { "left": "issue:39290", "right": "issue:43742", "accept": false, - "reason": "Both are load-time errors, but they involve different configs/models and distinct missing-key/attribute failures." + "reason": "Gemma3TextConfig sliding-window attribute error and MobileLLM key error are different model-load failures." }, { "left": "issue:43054", "right": "issue:43994", - "accept": false, - "reason": "Same model family, but one is a comparative embedding-quality complaint and the other is a pipeline/AutoModel nonsense-output report; not clearly the same defect." + "accept": true, + "reason": "Both report SigLIP2 producing clearly wrong outputs; one frames it as much worse text embeddings, the other as nonsensical AutoModel/pipeline results, consistent with the same underlying model wiring bug." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Different IndexError shapes and different contexts; not a duplicate bug." + "reason": "Mask-shape mismatch during indexing and a zero-sized-tensor IndexError under deepspeed zero3 are distinct symptoms and likely distinct causes." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "These are unrelated runtime failures; grouped_mm tracing is not the same as missing Float8 storage support." + "reason": "Grouped_mm tracing failure is unrelated to missing Float8 storage type support." }, { "left": "issue:45632", "right": "issue:45698", - "accept": true, - "reason": "Both point to the same dynamic-module loading/cache collision causing the wrong custom module to be loaded after save/load." + "accept": false, + "reason": "Both concern remote custom-module loading, but one is specifically a trust_remote_code cache-path collision and the other is a wrong-module load after save_pretrained; not enough evidence they are the same bug." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "Perplexity on gpt-oss-20b and a GLM5 issue are unrelated." + "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-quality reports with no shared failure mode." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Different bugs: image-preprocessor/model dtype mismatch vs SFT unpacking failure for Qwen3.5 embeddings." + "reason": "Mistral-3 image-preprocessor dtype mismatch and Qwen3.5 embedding unpacking failure in TRL are unrelated." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Both mention tied weights, but they are different concrete issues and model families; too broad to merge." + "reason": "Both mention tied weights, but one is a Mistral3 lm_head tie bug and the other is a broader translation-model v5 issue; not the same concrete defect." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model initialization regression is unrelated to flash-attn fallback gating." + "reason": "Custom-model notebook initialization and flash-attn2 fallback gating are unrelated issues." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache linear_attention crash is a different code-path problem from tied-weights handling." + "reason": "PagedAttentionCache invalid group type on Qwen3.5 is not a tied-weights regression." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint-loading failure is unrelated to translation-model tied-weights issues." + "reason": "NemotronH checkpoint loading failure is unrelated to tied-weight handling." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "DSA indexer ReLU discussion and removed run-command docs are unrelated." + "reason": "DSA indexer/ReLU discussion and a docs typo about a removed run command are unrelated." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "FineGrainedFP8 Qwen3Moe failure and flash-attn fallback gating are different issues." + "reason": "Qwen3Moe FP8 config failure and flash-attn2 fallback blocking are different problems." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is not the same bug as tied-weights handling." + "reason": "Quantization/offloading behavior is unrelated to tied weights." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP 2-D weight error is unrelated to Qwen3.5 embedding unpacking in TRL." + "reason": "FSDP 2-D weight runtime error is unrelated to Qwen3.5 input-embedding unpacking in TRL." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "A typo in cache-strategy docs and a removed run-command docs issue are different documentation nits." + "reason": "Cache-strategy typo and removed-doc-command mention are both documentation-level but not the same issue." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Different model families and failures; not the same underlying bug." + "reason": "TRL embedding unpacking failure and GraniteMoEHybrid invalid method call are different model/integration bugs." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Detokenization regression and NVFP4 runtime error are unrelated." + "reason": "DeepSeek detokenization regression and Qwen-3-8B-NVFP4 runtime failure are unrelated." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Same regression: non-persistent buffers are being corrupted/fill with junk in v5." + "reason": "Same regression: v5 corrupts non-persistent buffers during load; 44534 is a shorter restatement of the same bug." }, { "left": "issue:43644", "right": "issue:43950", "accept": true, - "reason": "Same underlying non-persistent-buffer corruption regression; 43950 is just the more explicit report." + "reason": "Same non-persistent-buffer corruption regression; 43950 is the more explicit issue, while 43644 is the broader symptom description." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "Unrelated model/runtime failures." + "reason": "Qwen-3-8B-NVFP4 runtime failure and GraniteMoEHybrid invalid method are unrelated." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "SigLIP2 doc example errors are not the same as the separate SigLIP2 embedding-quality complaint." + "reason": "SigLIP2 doc-example errors/quantization failure are not the same as the broader bad-embedding quality report." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Bad AutoTokenizer error messaging and missing all_tied_weights_keys are different issues." + "reason": "Bad AutoTokenizer error message for Voxtral and missing `all_tied_weights_keys` on MolmoForCausalLM are unrelated." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Autocast dtype mismatch and RLHF gradient explosion from 3D position_ids are different bugs." + "reason": "Both involve numerics, but autocast dtype mismatch is not the same as Qwen3.5 RLHF NaNs from 3D position_ids/SDPA fallback." }, { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "A corrupted checkpoint and generic loading noise are not the same defect." + "reason": "Corrupted Ovis2 checkpoint and generic loading noise are not the same defect." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The two accepted duplicate groups are well supported by the titles: the non-persistent-buffer regression around 43950/43644/44534, and the custom-module cache/path collision around 45632/45698. The rejected soft-edge pairs are mostly justified by concrete symptom/model differences rather than vague similarity. No obvious overstatement." + "feedback": "The summary is well grounded in the packet and stays mostly conservative. The canonical issue choice for the non-persistent-buffer regression is supported by the titles and pair evidence, and the only accepted soft-edge duplicate outside that cluster is reasonably framed as likely rather than certain. No mergeability issue applies here since these are issue pairs, not PR pairs." }, "evaluator_used": true, "retried": false, @@ -987,7 +876,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36331", "issue:38175", @@ -1040,8 +929,8 @@ "soft_pairs": [ "issue:36331|issue:41762", "issue:41553|issue:43881", - "issue:43054|issue:45701", "issue:43872|issue:44589", + "issue:43054|issue:45701", "issue:38175|issue:43054", "issue:42947|issue:44387", "issue:43054|issue:43493", @@ -1065,8 +954,8 @@ "issue:43792|issue:44220", "issue:43404|issue:43950", "issue:43257|issue:43866", - "issue:43742|issue:45399", "issue:41553|issue:44991", + "issue:43742|issue:45399", "issue:42915|issue:45020", "issue:43742|issue:45020", "issue:43940|issue:45313", @@ -1075,20 +964,20 @@ "issue:43854|issue:44863", "issue:42915|issue:43872" ], - "prepared_review_unit_hash": "90ca327ed1faedac4e7c9c75d789da3d2326ffc75f0c2f0f5ba8b887cdd2a1fb", + "prepared_review_unit_hash": "d6738f153ada2df681fb22c1db23dfee47938210a92023b71667259e6473cc42", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 36, - "serialized_chars": 37865, + "serialized_chars": 37867, "estimated_input_tokens": 9467, "estimated_eval_tokens": 19190 }, @@ -1097,234 +986,222 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are distinct model-loading, tokenizer, or docs regressions rather than duplicates. The only plausible duplicate pair is the Qwen3/Qwen3.5 DeepSpeed ZeRO-3 weight-loading failure; the rest should stay separate.", - "confidence": 0.58, - "canonical_issue_reason": "issue:43940 is the best canonical issue because it is the clearest, earliest report of the Qwen3 ZeRO-3 missing-weights failure, which also appears in issue:45313.", + "summary": "This cluster is mostly heterogeneous. I found a few true duplicate pairs around the same underlying regression/theme (SigLIP2 quality, Whisper fbank extraction, a docs cleanup, and one `_is_hf_initialized` loading regression), but most pairs are distinct issues in different models or subsystems.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45020 is the best umbrella candidate: it is broad, open, and captures the recurring recent-transformers / remote-code model-loading breakage pattern better than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43940 is the strongest representative of the only real duplicate cluster and names the concrete failure mode more clearly than the follow-up report.", + "best_issue_reason": "Issue 45020 is the most suitable global anchor for this mixed cluster because it is the broadest, still-active report and can absorb the repeated loading/regression theme better than the more symptom-specific issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Training API signature error vs DeepSpeed load-time IndexError; different code paths and symptoms." + "reason": "Different failures: training-time `compute_loss` signature mismatch vs DeepSpeed ZeRO-3 loading crash." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Voxtral AutoTokenizer error-message problem vs GLM-4V loading failure; unrelated models and bugs." + "reason": "Both are load failures, but for unrelated models and different error paths." }, { - "left": "issue:43054", - "right": "issue:45701", + "left": "issue:43872", + "right": "issue:44589", "accept": false, - "reason": "SigLIP2 embedding quality regression vs generic tokenization changes; not the same underlying bug." + "reason": "Distinct incompatibilities: `_is_hf_initialized`/Int8Params vs missing Float8 storage." }, { - "left": "issue:43872", - "right": "issue:44589", + "left": "issue:43054", + "right": "issue:45701", "accept": false, - "reason": "bitsandbytes Int8Params constructor incompatibility vs missing Float8 storage type; different dependency breakages." + "reason": "SigLIP2 embedding regression vs a generic tokenization-change complaint; not the same bug." }, { "left": "issue:38175", "right": "issue:43054", - "accept": false, - "reason": "siglip2 zero-probability output issue vs worse SigLIP2 text embeddings; similar model family but different failure modes." + "accept": true, + "reason": "Both report SigLIP2 output quality/behavior regressions and look like symptoms of the same implementation mismatch." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "PEFT LoRA gradient-checkpointing ineffective vs int4 quantization OOM; different mechanisms and fixes." + "reason": "LoRA gradient checkpointing vs int4 quantization memory growth are different issues." }, { "left": "issue:43054", "right": "issue:43493", - "accept": false, - "reason": "Embedding-quality discrepancy vs HF/JAX implementation mismatch; both SigLIP2-related but not the same concrete bug." + "accept": true, + "reason": "43493 describes the likely root-cause SigLIP2 implementation discrepancy that matches 43054's degraded embeddings." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "GLM-4V loading failure vs Molmo missing all_tied_weights_keys; different model implementations." + "reason": "Different models and different missing-attribute/load failures." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise vs inability to load a specific model; not the same defect." + "reason": "Generic loading noise vs a specific model load failure for ScandiBERT." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Detokenization bug in DeepSeek Coder vs flash_attn import KeyError on Python 3.13; unrelated subsystems." + "reason": "Tokenizer detokenization bug vs Python 3.13 flash_attn import error." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Missing fields from AutoConfig with trust_remote_code vs broad remote_code breakage; related area, but not one concrete bug." + "reason": "`AutoConfig` missing fields under `trust_remote_code` is narrower than the broader remote-code breakage report; not clearly the same underlying bug." }, { "left": "issue:43975", "right": "issue:45237", "accept": false, - "reason": "Tokenizer/detokenization regression vs AMD GPU incompatibility for GPT-OSS; unrelated." + "reason": "Unrelated problems: detokenization vs AMD GPU support." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Generic transformer loading noise vs specific failure to load cjvt/sleng-bert; same theme, different bug." + "reason": "Different model-loading complaints with no clear shared root cause." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Checkpointing issue in LoRA training vs Qwen3 MoE memory inefficiency; different performance bugs." + "reason": "Checkpointing effectiveness and MoE memory usage are separate training issues." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight_only loading error vs GLM5 issue; different models and loaders." + "reason": "Different models and different failure modes (`weight_only=True` vs GLM5 loading)." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "Qwen3 MoE LoRA compatibility change across versions vs warning about tie_word_embeddings; not the same bug." + "reason": "LoRA compatibility regression vs a tie-word-embeddings warning; not the same concrete bug." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Gemma3 ZeRO-3 load IndexError vs Qwen3 MoE FineGrainedFP8Config failure; both distributed-training related but different code paths." + "reason": "Gemma3 ZeRO-3 load failure vs Qwen3 MoE FP8 failure are unrelated." }, { "left": "issue:43975", "right": "issue:44451", "accept": false, - "reason": "Tokenizer detokenization regression vs model loading failure for ScandiBERT; unrelated." + "reason": "Different tokenizer/model-specific failures." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Broad remote_code regressions vs Kimi-K2.5 tokenizer codec/fix_mistral_regex regression; separate tokenizer bug." + "reason": "Broad remote-code regressions vs a Kimi-K2.5 tokenizer regression; not the same issue." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "add-new-model-like mapping bug vs remote_code breakage; different functionality." + "reason": "Model-addition helper failure vs general remote-code loading regressions." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Qwen2.5-VL import/export issue vs bitsandbytes constructor incompatibility; unrelated." + "reason": "Importing a missing class and an `_is_hf_initialized`/bitsandbytes TypeError are different bugs." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading failure for a specific model vs flash-attn fallback logic blocked by checks; different problems." + "reason": "Tokenizer load failure vs flash-attn fallback gating; unrelated code paths." }, { "left": "issue:43827", "right": "issue:44509", - "accept": false, - "reason": "Both are docs-related, but they cover different pipeline-task references and likely the same documentation cleanup only broadly." + "accept": true, + "reason": "Both are the same docs regression: stale references to removed pipeline tasks after v5 cleanup." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Mistral tokenizer regex crash vs flash-attn fallback gating; different subsystems and fixes." + "reason": "Mistral regex patch crash vs flash-attn fallback checks are unrelated." }, { "left": "issue:43792", "right": "issue:44220", - "accept": false, - "reason": "Whisper runtime failure vs a specific fbank extraction issue; could be related, but not enough to call the same bug." + "accept": true, + "reason": "Both point to Whisper audio feature extraction failing; `_torch_extract_fbank_features` is the concrete path behind the general can't-run report." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Weight tying bug in Mistral3 vs non-persistent buffer corruption on from_pretrained; distinct state-loading issues." + "reason": "Untied lm_head weights vs non-persistent buffer corruption are different model-state bugs." }, { "left": "issue:43257", "right": "issue:43866", "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed vs Ovis2 checkpoint corruption; different models and failure modes." + "reason": "Qwen3 MoE weight-conversion issue vs corrupted Ovis2 checkpoint." }, { - "left": "issue:43742", - "right": "issue:45399", + "left": "issue:41553", + "right": "issue:44991", "accept": false, - "reason": "MobileLLM key error vs flash-attn fallback blockage; unrelated." + "reason": "Voxtral error messaging vs tokenizer load failure for a different model." }, { - "left": "issue:41553", - "right": "issue:44991", + "left": "issue:43742", + "right": "issue:45399", "accept": false, - "reason": "AutoTokenizer Voxtral error messaging vs tokenizer load failure for EMBEDDIA/est-roberta; different models and failures." + "reason": "MobileLLM key error and flash-attn fallback gating do not share a concrete root cause." }, { "left": "issue:42915", "right": "issue:45020", "accept": false, - "reason": "Qwen3 MoE FP8 training failure vs broad remote_code regressions; not the same concrete code path." + "reason": "Qwen3 MoE FP8 failure is too specific to be the same bug as broad remote-code breakage." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "MobileLLM loading key error vs remote_code compatibility regressions; unrelated." + "reason": "Model-specific key error vs generic remote-code breakage." }, { "left": "issue:43940", "right": "issue:45313", - "accept": true, - "reason": "Both report the same ZeRO-3 missing-weights load failure in the Qwen3 family, just on different variants (Qwen3-Next vs Qwen3.5)." + "accept": false, + "reason": "Both are Qwen DeepSpeed ZeRO-3 load failures, but for different model generations and with different reported symptoms." }, { "left": "issue:43872", "right": "issue:44291", - "accept": false, - "reason": "bitsandbytes Int8Params compatibility issue vs init_empty_weights/_is_hf_initialized TypeError; different constructor/API breaks." + "accept": true, + "reason": "Same underlying `_is_hf_initialized` regression surfacing during model loading / parameter construction." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Bad AutoTokenizer error for Voxtral vs inability to load a specific BERT model; different reports." - }, - { - "left": "issue:43854", - "right": "issue:44863", - "accept": false, - "reason": "GLM-4.7-Flash unit-test load failure vs NemotronH checkpoint loading failure; different models." - }, - { - "left": "issue:42915", - "right": "issue:43872", - "accept": false, - "reason": "Qwen3 MoE FP8 training issue vs bitsandbytes Int8Params signature incompatibility; unrelated." + "reason": "Different model-load issues with no shared concrete failure path." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and stays conservative overall. The canonical choice for issue:43940 is supported by the titles and creation dates, and the only accepted duplicate pair (43940/45313) looks like the same ZeRO-3 missing-weights bug across Qwen3 variants. The other soft-edge rejections are appropriately cautious given the limited evidence." + "feedback": "The cluster summary is grounded in the packet and stays conservative overall. The accepted soft pairs are all reasonably supported by the issue titles/descriptions, and the rejected pairs do not overreach by forcing unrelated model/subsystem failures together. The choice of 45020 as a broad anchor is also defensible for this mixed set." }, "evaluator_used": true, "retried": false, @@ -1343,7 +1220,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:24643", "issue:30064", @@ -1427,226 +1304,226 @@ "issue:38175|issue:39692", "issue:42491|issue:43720" ], - "prepared_review_unit_hash": "5d046ebe4b5b48f39ab5f7a5edb173b58276503d66005f44c1c8b30b4b826349", + "prepared_review_unit_hash": "1e67f73b9376a7253e4c21b957f48177514053709b6302000a34634762090488", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37221, - "estimated_input_tokens": 9306, - "estimated_eval_tokens": 18868 + "serialized_chars": 37219, + "estimated_input_tokens": 9305, + "estimated_eval_tokens": 18866 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous transformers issues with only a few true duplicate pairs. The strongest matches are the non-persistent buffer regression, the SigLIP2 bad-output report, and the custom-model-initialization regression; the rest are same-subsystem or same-model-family reports but not the same concrete bug.", - "confidence": 0.88, - "canonical_issue_reason": "Issue 43644 is the clearest representative of a real duplicate family: it states the non-persistent buffer regression directly, and it has an essentially identical companion in 44534.", + "summary": "This set is mostly unrelated issue reports. Only a few pairs look like true duplicates: repeated Transformers v5 non-persistent-buffer regression, identical mask-shape IndexError reports, the same SigLIP2 output bug, and the custom-model initialization regression.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 43644 is the clearest canonical representative: it states the non-persistent-buffer regression directly and is the strongest match for its near-identical duplicate (44534).", "canonical_pr_reason": null, - "best_issue_reason": "43644 is the best issue to anchor the cluster because it is specific, broadly worded, and maps cleanly to an exact duplicate-style report.", + "best_issue_reason": "Issue 43644 is the best issue to anchor the cluster because it is specific, broadly descriptive, and maps cleanly to a near-exact duplicate report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "Both involve Qwen/attention/video-style failures, but they describe different code paths and different symptoms; not the same bug." + "reason": "Different bugs: one is a Qwen3.5 training/SDPA gradient-collapse report, the other is an s_aux=None flash-attention crash." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Both mention DeepSpeed/Qwen3, but one is weight conversion and the other is gradient averaging; separate issues." + "reason": "Both involve DeepSpeed, but the failures are different code paths: MoE weight conversion vs gradient averaging with GAS." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Unexpected compute_loss kwarg and tensor mask shape mismatch are unrelated failures." + "reason": "Unrelated errors: custom Trainer loss-signature incompatibility vs a tensor/mask shape mismatch." }, { "left": "issue:41093", "right": "issue:44805", - "accept": false, - "reason": "The error text is similar, but there is no evidence these are the same underlying shape-mismatch bug." + "accept": true, + "reason": "Same underlying mask/index shape mismatch IndexError; the second report is essentially the same bug with genericized shapes." }, { "left": "issue:43872", "right": "issue:45538", "accept": false, - "reason": "Bitsandbytes parameter construction and CLIPTokenizer max length are unrelated." + "reason": "Different subsystems and failures: bitsandbytes Int8Params constructor incompatibility vs tokenizer model_max_length behavior." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both concern video inputs, but the models and failure modes differ; not enough to treat as one bug." + "reason": "Both mention video/Qwen models, but the concrete failures and model paths differ; not enough evidence of one shared bug." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Qwen3.5 cache/attention crash and NemotronH checkpoint loading are unrelated." + "reason": "Different models and errors: PagedAttentionCache linear_attention crash vs NemotronH checkpoint-loading failure." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "DeepSpeed training weight-dimension error and void segmentation-map processing are unrelated." + "reason": "Completely different areas: DeepSpeed training weight shape error vs image-processor segmentation-map handling." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF gpt-oss support and Phi MoE autocast dtype mismatch are different problems." + "reason": "No shared bug: GGUF/gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Both are audio-related, but one is batched inference leakage and the other is empty-transcription handling." + "reason": "Different workloads and root causes: AudioFlamingo batched-track leakage vs Whisper empty-transcription generation failure." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "These are different model-support/regression reports with no shared concrete code path." + "reason": "Both are Qwen/GGUF-adjacent, but one is support for gpt-oss architecture and the other is a Qwen3-MoE sliding_window issue." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Both mention attention/SDPA, but the model paths and root causes are different." + "reason": "Both are runtime/attention bugs, but the model families and concrete failure modes differ (Bamba compile/SDPA vs Qwen3.5 flash-attn position_ids)." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Video StopIteration and MPS histogram integer failure are unrelated." + "reason": "Unrelated bugs: Qwen3-VL video StopIteration vs an MPS histogram limitation for MoE integers." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Both report the same Transformers 5.0 custom model initialization regression; the Jupyter detail is just context." + "reason": "Same Transformers 5 custom-model initialization regression; one report is just a Jupyter-specific manifestation of the same bug." }, { "left": "issue:43950", "right": "issue:45325", "accept": false, - "reason": "Non-persistent buffer corruption and Qwen2.5-VL rope scaling are separate regressions." + "reason": "Distinct regressions: non-persistent buffer corruption vs Qwen2.5-VL rope index scaling for still images." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Missing import for Qwen2.5-VL and meta-device loading failures are different issues." + "reason": "Different problems: missing Qwen2_5_VL import symbol vs meta-device model-loading breakage." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "FineGrainedFP8Config failure and meta-device loading breakage are unrelated." + "reason": "Different failures: Qwen3-MoE with FineGrainedFP8Config vs general torch.device('meta') loading regressions." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "GGUF architecture support and missing Float8 storage are unrelated compatibility issues." + "reason": "No common bug: gpt-oss support vs missing Float8 storage type during load/serialization." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Both involve Qwen video/embedding workflows, but one is a video StopIteration bug and the other is TRL input-embedding unpacking." + "reason": "Both mention Qwen3.5/video, but one is a video-input StopIteration regression and the other is a TRL embedding unpacking failure." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "3d_parallel.py failure and GGUF gpt-oss support are unrelated." + "reason": "Different issues: 3d_parallel.py execution failure vs GGUF gpt-oss architecture support." }, { "left": "issue:38175", "right": "issue:43994", "accept": true, - "reason": "Same SigLIP2 model and same bad-output symptom family; these look like duplicate reports of the same regression." + "reason": "Same SigLIP2 bug report: zero probabilities / nonsensical outputs for google/siglip2-base-patch16-224." }, { "left": "issue:43646", "right": "issue:43950", "accept": false, - "reason": "Custom model initialization and non-persistent buffer corruption are different regressions." + "reason": "Both are Transformers 5 regressions, but one is custom model initialization and the other is non-persistent buffer corruption." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import failure and Qwen3Moe FP8 failure are unrelated." + "reason": "Unrelated: GenerationMixin import regression vs Qwen3-MoE FP8 configuration failure." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Video StopIteration and mask/tensor shape mismatch are different bugs." + "reason": "Different bugs: Qwen3-VL video StopIteration vs the mask/index shape mismatch IndexError." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Strict config loading for granite_speech and glm4v config validation are separate config issues." + "reason": "Different config-validation problems affecting different models." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "Bert zero3 index error and TOKENIZER_MAPPING_NAMES failure are unrelated." + "reason": "Different bugs: DeepSpeed/zero3 Bert out-of-bounds index vs tokenizer-mapping add-new-model-like failure." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Both are loading/initialization errors, but they affect different subsystems and code paths." + "reason": "Different bugs: DeepSpeed zero3 Bert index-out-of-bounds vs tokenizer mapping name handling." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Dequantized model offload failure and mmap OOM on Strix Halo are not the same bug." + "reason": "Different topics: dequantized-model loading with device_map=auto vs mmap/OOM on Strix Halo." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "These are the same non-persistent buffer regression, phrased almost identically." + "reason": "Near-exact duplicate: both report the Transformers v5 non-persistent-buffer junk/corruption regression." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "The docs-example issues mention SigLIP2, but one is output correctness and the other is documentation/model-processor mismatch with extra quantization problems." + "reason": "Related model family, but different bugs: unexpected zero probabilities vs documentation example errors." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "LoRA/Qwen3-MoE compatibility across versions and BitNet packed-weight loading are unrelated." + "reason": "Different issues: Qwen3 MoE LoRA compatibility across versions vs BitNet packed-weight unpacking during accelerate loading." } ] }, "evaluator_result": { "accept": true, - "feedback": "The cluster summary is grounded in the packet, and the soft-edge verdicts stay conservative. The accepted duplicate pairs are well supported by the issue titles and descriptions, especially 43644/44534. No obvious overstatement of evidence." + "feedback": "Grounded overall: the canonical choice (43644) is well supported by the near-exact duplicate 44534, and the other accepted pairs are plausible from the titles. The summary stays mostly conservative. The only pair that feels slightly less certain is 38175/43994, which reads more like a closely related SigLIP2 output issue than a guaranteed exact duplicate, but that does not undermine the cluster as a whole." }, "evaluator_used": true, "retried": false, @@ -1665,7 +1542,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36010", "issue:39404", @@ -1673,6 +1550,7 @@ "issue:41720", "issue:42491", "issue:42915", + "issue:42947", "issue:43257", "issue:43335", "issue:43404", @@ -1691,17 +1569,14 @@ "issue:44155", "issue:44164", "issue:44291", - "issue:44292", "issue:44387", "issue:44402", "issue:44451", "issue:44488", "issue:44661", "issue:44683", - "issue:44743", "issue:44756", "issue:44910", - "issue:44912", "issue:44960", "issue:44977", "issue:44991", @@ -1714,7 +1589,8 @@ "issue:45362", "issue:45406", "issue:45588", - "issue:45663" + "issue:45663", + "issue:45704" ], "soft_pairs": [ "issue:45313|issue:45357", @@ -1724,266 +1600,259 @@ "issue:43708|issue:44075", "issue:42491|issue:45276", "issue:41720|issue:42915", - "issue:43856|issue:43873", "issue:43257|issue:43854", + "issue:43856|issue:43873", "issue:39692|issue:43994", "issue:44488|issue:44991", "issue:42915|issue:44756", "issue:42491|issue:43931", "issue:45362|issue:45406", "issue:43644|issue:43645", - "issue:43856|issue:44387", "issue:44661|issue:44683", + "issue:43856|issue:44387", "issue:44910|issue:45588", "issue:42491|issue:45292", - "issue:45276|issue:45335", "issue:43404|issue:43454", + "issue:45276|issue:45335", "issue:36010|issue:39404", "issue:44164|issue:44291", "issue:44402|issue:44488", "issue:41720|issue:44155", "issue:39692|issue:44960", "issue:43335|issue:45663", - "issue:43873|issue:44387", "issue:43425|issue:43957", + "issue:43873|issue:44387", + "issue:42947|issue:45704", "issue:44387|issue:44756", "issue:43854|issue:45216", - "issue:44291|issue:44756", - "issue:44292|issue:44912", - "issue:44743|issue:44960" + "issue:44291|issue:44756" ], - "prepared_review_unit_hash": "f761079c61ccee9c6750281613b329ea075d6005d02038ef8c76da33e9559efc", + "prepared_review_unit_hash": "fdffd8c81c6072f8a365a9b343bb55d04b9bc936e8c7b932017e249082684122", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37851, - "estimated_input_tokens": 9463, - "estimated_eval_tokens": 19182 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 37003, + "estimated_input_tokens": 9251, + "estimated_eval_tokens": 18758 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are largely unrelated Transformers bugs and regressions. The soft pairs mostly share broad model-family or subsystem keywords, but they do not describe the same concrete failure mode, so none should be merged as duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "No true canonical duplicate stands out; issue 43873 is the broadest open item with inbound references, but it is still just one specific quantization/offloading bug rather than the center of a real duplicate cluster.", + "summary": "The cluster is mostly a grab bag of unrelated Transformers regression reports (model loading, quantization, generation, tokenizer, and embedding-resize bugs). There isn\u2019t a strong duplicate core, and the soft pairs mostly share superficial wording rather than the same concrete code path.", + "confidence": 0.46, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43873 is the most generally framed and still-active issue in the set, so it is the best practical anchor if one must pick one, but the cluster is too heterogeneous to treat it as a genuine duplicate group.", + "best_issue_reason": "Issue 45292 is the closest reusable representative of the small resize_token_embeddings subgroup, but the overall cluster is too heterogeneous for a single strong canonical issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Different Qwen3.5 failures: ZeRO-3 weight loading for language_model vs incorrect visual-encoder keys in save_pretrained." + "reason": "Both are Qwen3.5 save/load regressions, but one is ZeRO-3 language_model weight loading and the other is incorrect visual-encoder keys; different code paths." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Both are Qwen-related, but one is a GLM5 issue and the other is a Qwen3.5 flash-attention generation bug; different code paths and symptoms." + "reason": "Different models and failures: GLM5 unspecified load issue vs Qwen3.5 flash-attention generation failure." }, { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Both are model-load regressions, but they involve different BERT models and no shared concrete bug is shown." + "reason": "Both are model-loading failures, but they concern different BERT models and likely different tokenizer/model-specific breakage." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different models and different loading failures: GLM-4.7-Flash unit-test loading vs Qwen3-VL weight-shape mismatch." + "reason": "Different model families and symptoms: GLM-4.7-Flash test loading vs Qwen3-VL weight-shape mismatch." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Unrelated bugs: checkpoint resume/max_steps calculation vs SGD optimizer arguments being ignored." + "reason": "One is Trainer checkpoint resumption math, the other is SGD argument handling; unrelated bugs." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Different problems on different model families: Qwen3-MoE LoRA compatibility vs Gemma4 resize_token_embeddings not updating outputs." + "reason": "LoRA/Qwen3 MoE hf4.x-to-hf5.x compatibility is unrelated to Gemma4 token-embedding resize propagation." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Both touch Qwen3/Qwen3-MoE, but one is device-mapping cudaErrorAssert and the other is FineGrainedFP8Config failure; not the same bug." + "reason": "Auto device-mapping CUDA assert and FineGrainedFP8Config failure are different Qwen3/MoE problems." }, { - "left": "issue:43856", - "right": "issue:43873", + "left": "issue:43257", + "right": "issue:43854", "accept": false, - "reason": "Both mention Qwen3 MoE and quantization/offloading, but one is inefficient training memory and the other is offloading behavior with quantization." + "reason": "Qwen3 MoE weight-conversion under accelerate+deepspeed is unrelated to GLM-4.7-Flash loading in unit tests." }, { - "left": "issue:43257", - "right": "issue:43854", + "left": "issue:43856", + "right": "issue:43873", "accept": false, - "reason": "Different model-loading issues in different systems: Qwen3 MoE accelerate+deepspeed conversion vs GLM-4.7-Flash loading in tests." + "reason": "Both mention memory/quantization, but one is training-memory inefficiency and the other is offloading behavior; not the same bug." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Both involve SigLIP2, but one is a broken documentation example and the other is nonsensical AutoModel/pipeline inference results." + "reason": "Same model family (SigLIP2) but different failures: docs example errors vs nonsensical outputs from AutoModel/pipeline." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/load regressions, but they affect different models and no shared root cause is established." + "reason": "Both are tokenizer/loading regressions, but they affect different models and are not clearly the same underlying tokenizer bug." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Different issues: Qwen3-MoE FP8 loading failure vs mmap-related OOM on Strix Halo." + "reason": "FineGrainedFP8Config failure and Strix Halo mmap/OOM are unrelated." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different models and failure modes: LoRA compatibility across versions vs Qwen3-VL weight-shape mismatch." + "reason": "Different regressions: Qwen3 MoE LoRA compatibility vs Qwen3-VL shape mismatch on load." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Different crashes in different components: Qwen3.5 chat generation vs Gemma4Processor missing _tokenizer in transformers serve." + "reason": "Qwen3.5 chat crash and Gemma4Processor missing _tokenizer are unrelated model/serving bugs." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Both are Transformers 5.0 regressions, but one is junk non-persistent buffers and the other is custom-model initialization in Jupyter notebooks." + "reason": "Both are Transformers 5.0.0 regressions, but one concerns junk-filled buffers and the other Jupyter custom-model initialization." }, { - "left": "issue:43856", - "right": "issue:44387", + "left": "issue:44661", + "right": "issue:44683", "accept": false, - "reason": "Both concern memory/quantization, but one is Qwen3 MoE training inefficiency and the other is increased CUDA reserved memory under int4 leading to OOM." + "reason": "Tokenizer-mapping model creation and compiled flex_attention on torch>=2.9 are unrelated." }, { - "left": "issue:44661", - "right": "issue:44683", + "left": "issue:43856", + "right": "issue:44387", "accept": false, - "reason": "Completely different bugs: tokenizer mapping logic vs compiled flex_attention on newer torch." + "reason": "Training memory inefficiency vs int4 quantization reserved-memory growth; similar area, different concrete bug." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both are flash-attention crashes, but one is a 3D position_ids packing bug and the other is an AttributeError on s_aux=None." + "reason": "Flash-attention packed-sequence misinterpretation and s_aux=None AttributeError are different failure modes." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "One is a Qwen3-MoE LoRA/version-compatibility problem; the other is a generic resize_token_embeddings output_embeddings issue." + "reason": "Qwen3 MoE checkpoint compatibility is unrelated to a generic resize_token_embeddings/output_embeddings bug." }, { - "left": "issue:45276", - "right": "issue:45335", + "left": "issue:43404", + "right": "issue:43454", "accept": false, - "reason": "Similar API name, but different models and different affected components: Gemma4 embed_tokens_per_layer/output_embeddings vs t5gemma decoder.embed_tokens." + "reason": "Both are weight-tying issues, but they affect different model classes and likely different implementation paths." }, { - "left": "issue:43404", - "right": "issue:43454", + "left": "issue:45276", + "right": "issue:45335", "accept": false, - "reason": "Both are lm_head tying bugs, but they affect different model classes and are not clearly the same underlying defect." + "reason": "Both say resize_token_embeddings fails to update model-specific embeddings, but the affected architectures differ and there\u2019s not enough evidence this is one shared code-path bug." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "An import-path regression for GenerationMixin is unrelated to Whisper pipeline return_language behavior." + "reason": "ImportError for GenerationMixin and Whisper return_language regression are unrelated." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Different serialization/loading problems: extra_state handling vs init_empty_weights getting an unexpected _is_hf_initialized argument." + "reason": "save/from_pretrained extra_state handling is unrelated to init_empty_weights/_is_hf_initialized errors." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "A tokenizer vocab-size inconsistency is not the same as a separate model-loading failure for a different checkpoint." + "reason": "Vocabulary-size mismatch and tokenizer/model loading failure are different issues." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Qwen3 auto device mapping on A800 and AudioFlamingo3 batched embedding/token leak are unrelated bugs." + "reason": "CUDA assert during Qwen3 auto device mapping is unrelated to AudioFlamingo3 track leakage in batched inference." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Different problems in different model families: SigLIP2 doc/inference issues vs GLM5/Qwen3.5 flash-attention generation." + "reason": "SigLIP2 documentation/runtime issues and GLM5 loading are unrelated." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "Both concern sparse/MoE-style internals, but one is SwitchTransformersConfig layer creation and the other is Gemma-4 FSDP2 shared_kv_states rebuilding." + "reason": "SwitchTransformers sparse-layer config and Gemma-4 FSDP2 shared_kv_states KeyError are different model internals." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 incompatibility and meta-device loading breakage are different compatibility regressions." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Related to quantization and memory, but one is offloading behavior and the other is increased reserved memory causing OOM." + "reason": "Offloading with quantization and increased CUDA reserved memory under int4 quantization overlap conceptually but are not the same concrete bug." }, { - "left": "issue:43425", - "right": "issue:43957", + "left": "issue:42947", + "right": "issue:45704", "accept": false, - "reason": "Torch-version incompatibility is too broad; the second issue is specifically about meta-device loading breaking some models on Transformers 5." + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to apex.FusedRMSNorm memory leak." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Both are memory-related, but one is an int4 quantization reserved-memory regression and the other is mmap-induced OOM on Strix Halo." + "reason": "Both mention memory pressure, but one is int4 quantization reserved-memory growth and the other is mmap/OOM on Strix Halo." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different Qwen3.5-related problems: model loading in tests vs incorrect checkpoint saving with save_pretrained." + "reason": "GLM-4.7-Flash loading in tests and Qwen3.5 save_pretrained regression are unrelated." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "Init-empty-weights loading error and mmap OOM are unrelated failure modes." - }, - { - "left": "issue:44292", - "right": "issue:44912", - "accept": false, - "reason": "Both are quantization issues, but they involve different model families and different quantization formats/failure modes." - }, - { - "left": "issue:44743", - "right": "issue:44960", - "accept": false, - "reason": "A recurrent-state reset bug in modular_qwen3_5.py is unrelated to GLM5/Qwen3.5 flash-attention generation." + "reason": "Unexpected _is_hf_initialized argument during meta init and Strix Halo mmap/OOM are unrelated." } ] }, @@ -2005,7 +1874,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:39404", "issue:42947", @@ -2013,7 +1882,6 @@ "issue:43299", "issue:43381", "issue:43425", - "issue:43596", "issue:43611", "issue:43645", "issue:43646", @@ -2029,6 +1897,7 @@ "issue:44079", "issue:44164", "issue:44265", + "issue:44292", "issue:44402", "issue:44451", "issue:44479", @@ -2036,15 +1905,15 @@ "issue:44530", "issue:44589", "issue:44671", - "issue:44683", + "issue:44743", "issue:44756", "issue:44805", "issue:44863", "issue:44877", "issue:44898", "issue:44912", - "issue:44918", "issue:44928", + "issue:44960", "issue:44991", "issue:45084", "issue:45276", @@ -2057,293 +1926,293 @@ "issue:45701" ], "soft_pairs": [ + "issue:44292|issue:44912", + "issue:44743|issue:44960", "issue:45684|issue:45698", "issue:45276|issue:45292", - "issue:42947|issue:43381", "issue:43957|issue:44756", + "issue:42947|issue:43381", "issue:43381|issue:44928", "issue:44164|issue:44756", "issue:43646|issue:43708", "issue:45588|issue:45663", "issue:44479|issue:44877", - "issue:45292|issue:45335", "issue:44402|issue:44451", + "issue:45292|issue:45335", "issue:43645|issue:44164", "issue:43257|issue:44863", "issue:43828|issue:43957", "issue:43425|issue:43828", - "issue:39404|issue:45701", "issue:44589|issue:44756", + "issue:39404|issue:45701", "issue:44479|issue:45663", "issue:43906|issue:44530", "issue:43931|issue:44863", "issue:43611|issue:43645", "issue:43901|issue:44509", - "issue:44530|issue:45084", "issue:43425|issue:43716", + "issue:44530|issue:45084", "issue:39404|issue:44912", "issue:43299|issue:43931", "issue:44863|issue:45663", - "issue:43257|issue:45663", "issue:44479|issue:44805", + "issue:43257|issue:45663", "issue:44265|issue:44898", "issue:43994|issue:44079", "issue:44671|issue:44991", "issue:43856|issue:44756", - "issue:39404|issue:44671", - "issue:43425|issue:44918", - "issue:43596|issue:44683" + "issue:39404|issue:44671" ], - "prepared_review_unit_hash": "2e70d7ae42e4972ab2acedc27ddd9c76faa532c17c0d5c35eb6749676cc19eaa", + "prepared_review_unit_hash": "6fed7dd65fa1b0106113d7e902a78e19e2e61f384ba4df15b1f2a201728e334a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 36, - "serialized_chars": 38479, - "estimated_input_tokens": 9620, - "estimated_eval_tokens": 19496 + "serialized_chars": 38395, + "estimated_input_tokens": 9599, + "estimated_eval_tokens": 19454 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are clearly separate regressions despite superficial overlap around Transformers v5, loading, or model-init behavior. The only strong duplicate family is the resize_token_embeddings/output_embeddings reports, where a generic issue can cover the model-specific variants.", - "confidence": 0.73, - "canonical_issue_reason": "issue:45292 is the broadest and most direct statement of the shared bug: resize_token_embeddings not propagating to output embeddings. It can serve as the umbrella for the Gemma4 and T5Gemma-specific reports.", + "summary": "The cluster is mostly heterogeneous Transformers regressions and model-specific loading bugs. The only clear duplicate chain is around `resize_token_embeddings` not propagating to model-specific embedding/output modules (`output_embeddings`, `embed_tokens_per_layer`, `decoder.embed_tokens`).", + "confidence": 0.58, + "canonical_issue_reason": "issue:45292 is the broadest, most reusable statement of the shared `resize_token_embeddings` regression, and it best covers the more specific variants in issue:45276 and issue:45335.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45292 best represents the duplicate set because it states the core failure without narrowing to one architecture.", + "best_issue_reason": "issue:45292 is the strongest representative issue because it names the shared API-level failure rather than one specific architecture, while the rest of the cluster is largely unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are Qwen/quantization-related, but they describe different loading failures and code paths." + }, + { + "left": "issue:44743", + "right": "issue:44960", + "accept": false, + "reason": "Different models and different bugs: cache state reset vs a GLM5 issue." + }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Different save/load problems: one is read-only permissions during save_pretrained, the other loads the wrong custom module after save_pretrained." + "reason": "Both involve save/load custom modules, but one is a permissions propagation bug and the other is loading the wrong module." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same underlying bug family: resize_token_embeddings fails to update output embeddings; 45276 is a Gemma4-specific variant." + "reason": "Same underlying `resize_token_embeddings` propagation bug; both report output/embedding modules not being updated." }, { - "left": "issue:42947", - "right": "issue:43381", + "left": "issue:43957", + "right": "issue:44756", "accept": false, - "reason": "Both mention gradient checkpointing, but one is about PEFT/LoRA ineffectiveness and the other about eval-mode restrictions; different code paths." + "reason": "`torch.device(\"meta\")` loading failures are unrelated to the Strix Halo mmap OOM issue." }, { - "left": "issue:43957", - "right": "issue:44756", + "left": "issue:42947", + "right": "issue:43381", "accept": false, - "reason": "Unrelated failures: meta-device model loading vs disabling mmap on Strix Halo to avoid OOM." + "reason": "Both mention gradient checkpointing, but one is LoRA ineffectiveness and the other is eval-mode incompatibility." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to Qwen3.5 RLHF NaN explosions from 3D position_ids and SDPA fallback." + "reason": "Completely different failures: eval-mode checkpointing vs RLHF NaN/SDPA fallback." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "save/from_pretrained extra_state handling is a different bug from Strix Halo mmap/OOM behavior." + "reason": "`extra_state` save/load handling is unrelated to the mmap/OOM issue." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization in notebooks is unrelated to Trainer resume_from_checkpoint max_steps calculation." + "reason": "Custom model initialization and resume-from-checkpoint step \u0917\u0923culation are different bugs." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Different failures: flash_attention s_aux=None crash versus Gemma4 FSDP2 shared_kv_states rebuilding." + "reason": "Both are attention/Gemma-related, but they describe different failures in different code paths." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Separate regressions: video input handling for Qwen VL models vs strict config blocking granite_speech loading." - }, - { - "left": "issue:45292", - "right": "issue:45335", - "accept": true, - "reason": "Same core resize_token_embeddings propagation bug, expressed generically in 45292 and specifically for t5gemma.decoder.embed_tokens in 45335." + "reason": "Video-input regression for Qwen models is unrelated to strict config loading for granite_speech." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "Different tokenizer/loading failures affecting different models and symptoms." + "reason": "Tokenizer vocab-size mismatch vs a specific model loading failure; not the same bug." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same `resize_token_embeddings` regression, with `decoder.embed_tokens` as a model-specific manifestation of the broader output-embedding bug." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Jupyter custom model initialization is not the same as extra_state save/load handling." + "reason": "Notebook custom-model initialization is unrelated to `extra_state` serialization handling." }, { "left": "issue:43257", "right": "issue:44863", "accept": false, - "reason": "Qwen3 MoE weight conversion/loading is unrelated to NemotronH checkpoint loading." + "reason": "Different model families and different loading failures." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Different issues: autocast dtype mismatch for Phi-tiny-MoE versus meta-device loading failures." + "reason": "Both are dtype/device-related, but one is autocast mismatch and the other is meta-device loading." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "Torch version incompatibility is too broad and unrelated to a specific autocast dtype mismatch bug." + "reason": "Torch version incompatibility is not the same as a specific autocast dtype mismatch." }, { - "left": "issue:39404", - "right": "issue:45701", + "left": "issue:44589", + "right": "issue:44756", "accept": false, - "reason": "Whisper pipeline return_language regression is unrelated to tokenizer behavior changing by transformers version." + "reason": "Float8 storage lookup failure and mmap OOM are unrelated." }, { - "left": "issue:44589", - "right": "issue:44756", + "left": "issue:39404", + "right": "issue:45701", "accept": false, - "reason": "Float8 storage lookup error is unrelated to mmap/OOM mitigation." + "reason": "Whisper pipeline `return_language` regression is unrelated to a tokenizer-version behavior change." }, { "left": "issue:44479", "right": "issue:45663", "accept": false, - "reason": "Video regression in Qwen VL models does not match the Gemma4 FSDP2 KeyError." + "reason": "Qwen video regression and Gemma-4 FSDP2 KeyError are different bugs." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "An isolated reproduction of another issue is not the same as the Qwen3.5 PagedAttentionCache crash." + "reason": "Isolated reproduction of an old issue vs a Qwen3.5 PagedAttentionCache crash; no shared bug." }, { "left": "issue:43931", "right": "issue:44863", "accept": false, - "reason": "Both are loading failures, but one is Qwen3-VL shape mismatch and the other is NemotronH checkpoint incompatibility." + "reason": "Different model-loading failures with no clear shared root cause." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Both involve custom models in v5, but one is about base_model_prefix loading and the other about notebook initialization." + "reason": "Both are v5 custom-model loading regressions, but the symptoms and likely code paths differ." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs complaints, but they cover different docs and different removed/changed APIs." + "reason": "Both are docs-related, but they cover different pipeline-task deprecations." }, { - "left": "issue:44530", - "right": "issue:45084", + "left": "issue:43425", + "right": "issue:43716", "accept": false, - "reason": "PagedAttentionCache linear_attention crash is unrelated to the non-template-node compilation TypeError." + "reason": "Torch incompatibility and a Mistral-3 dtype mismatch are unrelated." }, { - "left": "issue:43425", - "right": "issue:43716", + "left": "issue:44530", + "right": "issue:45084", "accept": false, - "reason": "Torch compatibility and Mistral-3 dtype mismatch are separate problems." + "reason": "Different error classes and unrelated subsystems." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper pipeline return_language is unrelated to MXFP4 quantization fallback on git-oss-20b." + "reason": "Whisper pipeline behavior and Qwen quantization fallback are unrelated." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Both concern Qwen3-VL loading, but the symptoms differ enough to indicate separate bugs." + "reason": "Both touch Qwen3-VL loading, but the reported failures differ enough that they do not clearly share one bug." }, { "left": "issue:44863", "right": "issue:45663", "accept": false, - "reason": "NemotronH checkpoint loading and Gemma4 FSDP2 shared_kv_states are unrelated." + "reason": "NemotronH checkpoint loading and Gemma-4 FSDP2 attention state are different problems." }, { - "left": "issue:43257", - "right": "issue:45663", + "left": "issue:44479", + "right": "issue:44805", "accept": false, - "reason": "Qwen3 MoE conversion/loading is a different code path from Gemma4 FSDP2 attention state rebuilding." + "reason": "Video regression vs tensor mask shape mismatch: different code paths." }, { - "left": "issue:44479", - "right": "issue:44805", + "left": "issue:43257", + "right": "issue:45663", "accept": false, - "reason": "Video-input regression and a mask shape IndexError are not the same underlying bug." + "reason": "Qwen3 MoE conversion/loading and Gemma-4 FSDP2 attention-state rebuilding are unrelated." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export with torch_compilable_check is unrelated to Perceiver interpolation-position-encoding failure." + "reason": "`torch.export`/`torch_compilable_check` failure is unrelated to Perceiver interpolation behavior." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Siglip2 pipeline nonsense outputs and ModelOutput key assignment are different issues." + "reason": "SigLIP2 inference mismatch and `ModelOutput` key assignment are different bugs." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT MLM prediction regression is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + "reason": "CamemBERT prediction drift and tokenizer loading for est-roberta are unrelated." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Qwen3 MoE training memory inefficiency is not the same as Strix Halo mmap OOM handling." + "reason": "Qwen3 MoE training memory usage is unrelated to Strix Halo mmap OOM." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper pipeline return_language and CamemBERT masked LM predictions are unrelated model-specific regressions." - }, - { - "left": "issue:43425", - "right": "issue:44918", - "accept": false, - "reason": "Torch 2.10 compatibility and Qwen3.5 SFT embedding unpacking are separate bugs." - }, - { - "left": "issue:43596", - "right": "issue:44683", - "accept": false, - "reason": "Zero3/BertModel indexing failure is unrelated to compiled flex_attention on torch >= 2.9." + "reason": "Whisper pipeline behavior and CamemBERT MLM predictions are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet, and the only accepted duplicate family\u2014resize_token_embeddings not propagating to output/output-related embeddings\u2014looks supported by the issue titles. The other soft-edge verdicts are conservative and the justifications avoid overclaiming, so I accept." + "feedback": "Grounded and suitably conservative overall. The resize_token_embeddings issues are correctly identified as the only clear duplicate chain, and the false-pair rejections do not overclaim shared root cause. Canonical issue choice is well supported by the packet." }, "evaluator_used": true, "retried": false, @@ -2362,7 +2231,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:32090", "issue:34567", @@ -2382,6 +2251,7 @@ "issue:43482", "issue:43541", "issue:43582", + "issue:43596", "issue:43611", "issue:43632", "issue:43643", @@ -2399,32 +2269,33 @@ "issue:44479", "issue:44488", "issue:44530", - "issue:44534", "issue:44589", - "issue:44756", + "issue:44683", "issue:44811", "issue:44849", + "issue:44918", "issue:44987", "issue:44991", "issue:45081", "issue:45084", "issue:45325", - "issue:45507", "issue:45584", "issue:45685" ], "soft_pairs": [ - "issue:40990|issue:42947", + "issue:43425|issue:44918", + "issue:43596|issue:44683", "issue:44589|issue:45084", - "issue:36032|issue:43452", + "issue:40990|issue:42947", "issue:43425|issue:45685", - "issue:32090|issue:39404", + "issue:36032|issue:43452", "issue:44991|issue:45081", + "issue:32090|issue:39404", "issue:43883|issue:44488", "issue:43957|issue:44164", "issue:43582|issue:43632", - "issue:44164|issue:44479", "issue:44488|issue:45081", + "issue:44164|issue:44479", "issue:40990|issue:43421", "issue:43611|issue:43644", "issue:44488|issue:44987", @@ -2433,7 +2304,6 @@ "issue:44811|issue:45584", "issue:34567|issue:40444", "issue:43716|issue:45084", - "issue:42947|issue:44756", "issue:43644|issue:43646", "issue:43643|issue:44530", "issue:43452|issue:43482", @@ -2444,260 +2314,243 @@ "issue:42915|issue:44530", "issue:44849|issue:45325", "issue:43404|issue:44466", - "issue:36032|issue:43645", "issue:44987|issue:44991", - "issue:43262|issue:43377", - "issue:43646|issue:44534", - "issue:45084|issue:45507" + "issue:36032|issue:43645", + "issue:43262|issue:43377" ], - "prepared_review_unit_hash": "4076e73bf023e7ecb1d1a5fa9729112f86b2aed86ad7ddb6671e913a753e17b5", + "prepared_review_unit_hash": "077a914ff50fa0d219fba6960cbee3004fa9af326550782ac7c9c7e17edb3374", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 38207, - "estimated_input_tokens": 9552, - "estimated_eval_tokens": 19360 + "soft_pair_count": 34, + "serialized_chars": 37985, + "estimated_input_tokens": 9497, + "estimated_eval_tokens": 19250 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous set of unrelated Transformers issues. The only clearly duplicate-like subcluster is the tied `lm_head` / serialization regression around Mistral3, where one report is specific and the other is broader.", - "confidence": 0.74, - "canonical_issue_reason": "No single issue cleanly represents the whole cluster, but 44466 is the broadest representative of the only plausible duplicate subcluster (tied `lm_head.weight` / serialization behavior).", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers regression reports, with a loose theme around v5 model/tokenizer loading failures. I did not find any soft-edge pairs that clearly describe the same underlying bug/change.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43611 is the broadest and most representative loading-regression report in the set; several nearby issues are also Transformers v5 init/load failures, but this one is the cleanest umbrella candidate.", "canonical_pr_reason": null, - "best_issue_reason": "44466 is the best issue representative because it describes the broader regression and can subsume the more specific tied-weight report.", + "best_issue_reason": "issue:43611 is the best single representative issue for the cluster because it captures the general v5 loading regression theme without being overly model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:40990", - "right": "issue:42947", + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch version compatibility and Qwen3.5/TRL embedding unpacking are different problems." + }, + { + "left": "issue:43596", + "right": "issue:44683", "accept": false, - "reason": "Different problems: perplexity/regression on one model vs gradient-checkpointing behavior with PEFT LoRA." + "reason": "DeepSpeed ZeRO-3/BertModel index failure is unrelated to compiled flex_attention on torch>=2.9." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Unrelated failures: float8 storage lookup vs template compilation error." + "reason": "Float8 storage lookup and template-node compilation are unrelated error classes." }, { - "left": "issue:36032", - "right": "issue:43452", + "left": "issue:40990", + "right": "issue:42947", "accept": false, - "reason": "Both involve loading, but the tokenizer-method conflict and GGUF loading break are different code paths." + "reason": "Perplexity regression and LoRA gradient-checkpointing behavior are not the same bug." }, { "left": "issue:43425", "right": "issue:45685", "accept": false, - "reason": "Different platforms and errors: torch version incompatibility vs an MPS histogram kernel limitation." + "reason": "Torch 2.10 incompatibility and an MPS histogram int error are unrelated." }, { - "left": "issue:32090", - "right": "issue:39404", + "left": "issue:36032", + "right": "issue:43452", "accept": false, - "reason": "Unrelated: Trainer broadcast TypeError vs Whisper pipeline return_language regression." + "reason": "T5 tokenizer method conflict and GGUF loading breakage are different code paths." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Both tokenizer-load failures, but model-specific causes and fixes appear different." + "reason": "Different tokenizer-loading failures for different models; no clear shared root cause." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper return_language pipeline behavior." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Different models and different failure modes; no shared concrete bug is evident." + "reason": "Missing all_tied_weights_keys and failing to load a specific BERT model are unrelated." }, { "left": "issue:43957", "right": "issue:44164", "accept": false, - "reason": "One is meta-device loading, the other is save/from_pretrained extra_state handling." + "reason": "Meta-device loading regression and extra_state save/load handling are distinct issues." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "Different regressions: Apple Silicon allocator warmup vs `_is_hf_initialized` flag behavior." + "reason": "Apple Silicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." }, { - "left": "issue:44164", - "right": "issue:44479", + "left": "issue:44488", + "right": "issue:45081", "accept": false, - "reason": "Save/load extra_state handling is unrelated to the Qwen video-input regression." + "reason": "Both are tokenizer/model loading problems, but for different models and different failure points." }, { - "left": "issue:44488", - "right": "issue:45081", + "left": "issue:44164", + "right": "issue:44479", "accept": false, - "reason": "Different model-loading issues; no evidence they share the same underlying tokenizer bug." + "reason": "extra_state serialization and Qwen video-input regression are unrelated." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity issue vs runtime post-processor update behavior; not the same bug." + "reason": "Perplexity on GPT-OSS and runtime special-token post-processing updates are unrelated." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization." + "reason": "Base_model_prefix loading regression and junk-filled non-persistent buffers are different bugs." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading complaints, but for different models and likely different root causes." + "reason": "Separate model-loading failures for different models; not the same concrete issue." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Related audio area, but the reported breakages are not the same concrete issue." + "reason": "Unsupported center argument in LasrFeatureExtractor and _torch_extract_fbank_features are related thematically but not clearly the same bug." }, { "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Different tokenizer/load failures on unrelated models." + "reason": "Missing tied-weights attribute and tokenizer loading failure are unrelated." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Both Whisper-related, but batch_decode skip_special_tokens and empty-transcription generation are distinct behaviors." + "reason": "Both involve Whisper special tokens, but the symptoms and code paths differ enough that they are not clearly the same bug." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "TrainerState token counting vs multimodal IterableDataset failures are unrelated." + "reason": "TrainerState token accounting and Qwen2.5-VL iterable-dataset multimodal failure are unrelated." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Different subsystems: image preprocessor dtype mismatch vs template compilation error." - }, - { - "left": "issue:42947", - "right": "issue:44756", - "accept": false, - "reason": "LoRA/gradient checkpointing and mmap/OOM on Strix Halo are unrelated." + "reason": "Mistral-3 dtype mismatch and non-template-node compilation are unrelated." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Both are v5 init regressions, but one is about junk buffers and the other about custom model initialization." + "reason": "Non-persistent buffer serialization and custom model initialization regressions are different failures." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "Different areas: AutoConfig trust_remote_code fields vs PagedAttentionCache group-type error." + "reason": "AutoConfig trust_remote_code field loss and PagedAttentionCache linear_attention crash are unrelated." }, { "left": "issue:43452", "right": "issue:43482", "accept": false, - "reason": "Both involve GGUF loading, but one is a broad from_pretrained bug and the other is a model-specific failure." + "reason": "Both mention GGUF loading, but one is a broad from_pretrained break and the other is a specific model report without enough evidence of the same code-path." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Separate model-loading failures with no shared code-path evidence." + "reason": "Loading ScandiBERT and loading physical-intelligence/fast are separate model-specific failures." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "Tokenizer add_special_tokens conflict vs FineGrainedFP8Config failure are unrelated." + "reason": "T5 tokenizer conflict and Qwen3Moe FP8 failure are unrelated." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "Different loading paths: device_map/offload vs GGUF model loading." + "reason": "Dequantized model offload failure and Qwen2.5-GGUF loading failure are distinct regressions." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Torch dynamo grouped_mm tracing error is unrelated to template-node compilation." + "reason": "Torch dynamo grouped_mm tracing failure is unrelated to non-template-node compilation." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Both Qwen-family/MoE-ish, but the FP8 config failure and PagedAttentionCache group-type crash are different bugs." + "reason": "Qwen3Moe FP8 config failure and Qwen3.5 PagedAttentionCache crash are different bugs." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Different Qwen3.5/VL bugs: hidden-state output handling vs RoPE temporal indexing." + "reason": "Qwen3.5 hidden-states behavior and Qwen2.5-VL rope-index scaling are unrelated." }, { "left": "issue:43404", "right": "issue:44466", - "accept": true, - "reason": "Both point to the same tied-weights regression around `lm_head.weight`: one says Mistral3ForConditionalGeneration leaves it untied, the other reports device-dependent serialization of the same weight." - }, - { - "left": "issue:36032", - "right": "issue:43645", "accept": false, - "reason": "Tokenizer-load method conflict vs custom model initialization in notebooks are different regressions." + "reason": "Mistral3 weight-tying bug and device-dependent serialization inconsistency are related in theme but not the same concrete failure." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are load failures on different models, but there is no concrete shared bug." + "reason": "Different model-loading regressions affecting different repositories/models." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio processor sampling-rate default vs MIMI encoder padding-mask/batched-output mismatch are unrelated." - }, - { - "left": "issue:43646", - "right": "issue:44534", - "accept": false, - "reason": "Custom model initialization and non-persistent buffer junk are different symptoms, not the same concrete bug." - }, - { - "left": "issue:45084", - "right": "issue:45507", - "accept": false, - "reason": "Template compilation error and GraniteMoEHybrid invalid-method call are unrelated." + "reason": "Audio chat-template sampling-rate default and MIMI padding-mask batching bug are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The tied `lm_head.weight` / serialization pair (43404, 44466) is a plausible duplicate-like subcluster, and the other pairwise rejections are supported by the issue titles. No clear overstatement of evidence." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -2712,7 +2565,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36032", "issue:40444", @@ -2732,8 +2585,8 @@ "issue:43611", "issue:43632", "issue:43645", + "issue:43646", "issue:43742", - "issue:43818", "issue:43819", "issue:43856", "issue:43994", @@ -2747,7 +2600,6 @@ "issue:44479", "issue:44534", "issue:44568", - "issue:44589", "issue:44610", "issue:44683", "issue:44756", @@ -2760,10 +2612,13 @@ "issue:45216", "issue:45325", "issue:45357", + "issue:45507", "issue:45542", "issue:45663" ], "soft_pairs": [ + "issue:43646|issue:44534", + "issue:45084|issue:45507", "issue:43994|issue:44610", "issue:43645|issue:44534", "issue:41553|issue:43582", @@ -2773,242 +2628,233 @@ "issue:43377|issue:43819", "issue:44451|issue:45081", "issue:44220|issue:44683", - "issue:44265|issue:44610", "issue:43404|issue:45127", - "issue:43632|issue:44756", - "issue:42915|issue:45084", + "issue:44265|issue:44610", "issue:44987|issue:45081", + "issue:42915|issue:45084", + "issue:43632|issue:44756", "issue:44479|issue:45325", "issue:44062|issue:44987", "issue:43296|issue:43531", "issue:42915|issue:44291", "issue:43611|issue:44534", - "issue:43317|issue:43856", "issue:36032|issue:43742", "issue:40444|issue:45325", + "issue:43317|issue:43856", "issue:40990|issue:44811", "issue:44079|issue:45663", - "issue:42175|issue:45542", "issue:43421|issue:44568", - "issue:43531|issue:44877", - "issue:42915|issue:44589", - "issue:43632|issue:44534", - "issue:43818|issue:45357" + "issue:42175|issue:45542", + "issue:43531|issue:44877" ], - "prepared_review_unit_hash": "3979c6e46cce1fec9b47a9818c76e911ba5bcb1ef0bc4d4f8d9c5f85edea646a", + "prepared_review_unit_hash": "14df59b0869d2fee19904f1fb1d582b0ca6707931208f8491b30f2e972f4be4e", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36770, - "estimated_input_tokens": 9193, - "estimated_eval_tokens": 18642 + "soft_pair_count": 29, + "serialized_chars": 36385, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated model-loading, tokenizer, multimodal, and backend issues. Only one soft pair looks like a likely duplicate/same fix candidate; the rest are too model- or subsystem-specific to merge safely.", - "confidence": 0.67, - "canonical_issue_reason": "issue:43611 is the broadest, most central v5 loading regression here, and is the best umbrella representative for the cluster\u2019s repeated model-load breakages.", + "summary": "This cluster is mostly heterogeneous. I only see two clear duplicate subgroups: the Qwen3.5 save_pretrained checkpoint-key regression, and the Qwen2.5-VL / Qwen3-VL temporal-position-id video regression. The rest are broad-similarity matches across different bugs and should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45216 is the clearest representative: a concrete Qwen3.5 save_pretrained regression, and issue:45357 reads like a narrower follow-up on the same incorrect visual-encoder key handling.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43611 is the best single issue to anchor triage because it is general, high-signal, and closer to the shared theme than the narrower model-specific reports.", + "best_issue_reason": "issue:45216 is the best issue to anchor the cluster because it is concrete, regression-scoped, and has the strongest duplicate evidence among the items.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is custom-model init and the other is buffer initialization; different code paths and symptoms." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Unrelated failures: template compilation versus an invalid method call in GraniteMoEHybrid." + }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Different models and symptoms: Siglip2/pipeline output quality vs OmDet processor image-size mismatch." + "reason": "Different models and different bug types: bad outputs/perplexity versus a processor image-size mismatch." }, { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Notebook custom-model initialization bug vs non-persistent buffer junk; not the same code path." + "reason": "Notebook custom-model initialization and non-persistent buffer junk are separate regression mechanisms." }, { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "Tokenizer-loading error message issue vs Apple Silicon warmup TypeError; unrelated." + "reason": "Bad AutoTokenizer error message for Voxtral is unrelated to Apple Silicon allocator warmup TypeError." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Llava-Next load crash vs Qwen3OmniMoe video-processing ValueError; different models and failures." + "reason": "Both are multimodal model-loading issues, but they affect different models and different failure modes." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Generic loading noise vs Mistral tokenizer regex patch crash; no shared underlying bug." + "reason": "Generic loading noise versus a specific Mistral tokenizer regex crash; not the same bug." }, { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Both report the same Qwen3.5 save_pretrained serialization regression, with incorrect checkpoint/visual-encoder key handling." + "reason": "Both report the same Qwen3.5 save_pretrained regression producing incorrect checkpoint keys, with 45357 looking like a later, narrower manifestation." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "MIMI padding-mask batching bug vs DAC STE mismatch; different components and failure modes." + "reason": "Audio batch-vs-single mismatch from padding-mask handling is not the same as DAC from_latents / STE mismatch." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT model loading failure vs Mistral tokenizer regex patch crash; unrelated." + "reason": "ScandiBERT load failure and Mistral regex patch crash are unrelated tokenizer/model-loading problems." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio feature extraction bug vs torch>=2.9 compiled flex_attention failure; different stacks." + "reason": "Fbank extraction and compiled flex_attention are unrelated subsystems." }, { - "left": "issue:44265", - "right": "issue:44610", + "left": "issue:43404", + "right": "issue:45127", "accept": false, - "reason": "torch.export/torch_compilable_check failure vs OmDet input-size mismatch; not the same issue." + "reason": "One is a missing lm_head tie in Mistral3; the other is LoRA merge collapse with extended vocab. Similar theme, different bug." }, { - "left": "issue:43404", - "right": "issue:45127", + "left": "issue:44265", + "right": "issue:44610", "accept": false, - "reason": "Missing lm_head weight tying vs LoRA merge collapse with extended vocab; distinct bugs." + "reason": "torch.export / torch_compilable_check is unrelated to the OmDet-Turbo processor size issue." }, { - "left": "issue:43632", - "right": "issue:44756", + "left": "issue:44987", + "right": "issue:45081", "accept": false, - "reason": "_is_hf_initialized regression vs mmap OOM workaround on Strix Halo; unrelated." + "reason": "Loading physical-intelligence/fast and the Mistral tokenizer backend crash are different failure reports." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "Qwen3MoE FP8 config failure vs template-node compilation error; no common underlying bug." + "reason": "Qwen3Moe FP8 config failure is unrelated to template-node compilation." }, { - "left": "issue:44987", - "right": "issue:45081", + "left": "issue:43632", + "right": "issue:44756", "accept": false, - "reason": "Physical-intelligence/fast loading failure vs Mistral tokenizer backend_attribute crash; different failures." + "reason": "_is_hf_initialized / model init behavior is unrelated to mmap OOM avoidance." }, { "left": "issue:44479", "right": "issue:45325", - "accept": false, - "reason": "Both involve Qwen VL/video-position logic, but one is a broad video-input regression and the other is a still-image rope-index bug; not clearly the same concrete defect." + "accept": true, + "reason": "Both point to the same Qwen VL temporal-position-id / rope-index regression affecting video or still-image paths in 5.3.0." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "AddedToken argument error vs model loading failure; unrelated." + "reason": "AddedToken keyword conflict is unrelated to loading physical-intelligence/fast." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM vs Qwen3-MoE sliding_window behavior; different model paths." + "reason": "PaddleOCR-VL loading and Qwen3-MoE sliding_window behavior are different issues." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "FP8 config failure vs init_empty_weights/_is_hf_initialized constructor issue; unrelated." + "reason": "FP8 Qwen3Moe failure and unexpected _is_hf_initialized in init_empty_weights are separate Transformer v5 regressions." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "base_model_prefix loading regression vs non-persistent buffer initialization junk; different bugs." - }, - { - "left": "issue:43317", - "right": "issue:43856", - "accept": false, - "reason": "Dequantized model offload/load failure vs Qwen3 MoE training memory usage; not the same defect." + "reason": "base_model_prefix loading breakage and non-persistent buffer junk are distinct initialization bugs." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "T5Tokenizer add_special_tokens conflict vs MobileLLM-125M key error; unrelated." + "reason": "T5Tokenizer add_special_tokens conflict and MobileLLM-125M key error are unrelated tokenizer/model-loading bugs." }, { "left": "issue:40444", "right": "issue:45325", "accept": false, - "reason": "IterableDataset multi-image training failure vs rope-index scaling for still images; different symptoms and code paths." + "reason": "Both touch Qwen2.5-VL, but one is multi-image IterableDataset training and the other is temporal position-id scaling." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "device_map/offload loading failure and Qwen3 MoE memory usage are not the same code-path issue." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Perplexity issue on gpt-oss-20b vs Whisper batch_decode skip_special_tokens handling; unrelated." + "reason": "High perplexity on gpt-oss and Whisper batch_decode skip_special_tokens behavior are unrelated." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment bug vs Gemma4 FSDP2 shared_kv_states issue; different subsystems." - }, - { - "left": "issue:42175", - "right": "issue:45542", - "accept": false, - "reason": "Backend packaging/install issue vs tensorboard-without-TF runtime error; related theme, but not the same underlying bug." + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states rebuilding are different bugs." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Runtime special-token post-processor sync issue vs model-specific add_special_tokens BOS/EOS regression; too different." + "reason": "Runtime special-token post-processor updates and mdeberta add_special_tokens BOS/EOS regression are similar area, but not clearly the same underlying bug." }, { - "left": "issue:43531", - "right": "issue:44877", - "accept": false, - "reason": "Qwen3-MoE sliding_window bug vs granite_speech strict-config load failure; unrelated." - }, - { - "left": "issue:42915", - "right": "issue:44589", - "accept": false, - "reason": "Qwen3MoE FineGrainedFP8Config failure vs Float8 storage lookup error; different problems." - }, - { - "left": "issue:43632", - "right": "issue:44534", + "left": "issue:42175", + "right": "issue:45542", "accept": false, - "reason": "_is_hf_initialized regression vs non-persistent buffer junk; not the same issue." + "reason": "Both involve TensorFlow optional deps, but one is install extras packaging and the other is a tensorboard-without-TF runtime backend error." }, { - "left": "issue:43818", - "right": "issue:45357", + "left": "issue:43531", + "right": "issue:44877", "accept": false, - "reason": "Video-LLaVA temporal-attention/weight-sharing problem vs Qwen3.5 save_pretrained key serialization bug; unrelated." + "reason": "Qwen3-MoE sliding_window and granite_speech config strictness are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The lone accepted soft pair (issue:45216 / issue:45357) is plausibly the same save_pretrained regression, and the rest of the rejections do not overclaim duplication. The chosen anchor issue:43611 is a reasonable broad representative for the load-regression cluster." + "feedback": "The summary is grounded in the packet and stays mostly conservative. The two accepted pairs are plausibly the same bug family based on the titles, and the other soft-edge rejections avoid over-claiming duplicate status across clearly different models or failure modes." }, "evaluator_used": true, "retried": false, @@ -3027,7 +2873,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:33666", "issue:35707", @@ -3050,10 +2896,9 @@ "issue:43646", "issue:43749", "issue:43792", - "issue:43828", + "issue:43818", "issue:43881", "issue:43883", - "issue:43957", "issue:43994", "issue:44079", "issue:44265", @@ -3073,18 +2918,22 @@ "issue:45072", "issue:45081", "issue:45325", + "issue:45357", "issue:45381", "issue:45479", "issue:45491", "issue:45538" ], "soft_pairs": [ - "issue:36032|issue:41720", + "issue:42915|issue:44589", + "issue:43818|issue:45357", + "issue:43632|issue:44534", "issue:43425|issue:44403", + "issue:36032|issue:41720", "issue:44683|issue:45020", "issue:44829|issue:45479", - "issue:33666|issue:40444", "issue:43482|issue:45538", + "issue:33666|issue:40444", "issue:44403|issue:44991", "issue:43792|issue:44589", "issue:44079|issue:44534", @@ -3092,222 +2941,231 @@ "issue:35707|issue:40444", "issue:43482|issue:43525", "issue:43749|issue:44079", - "issue:44291|issue:44589", "issue:43881|issue:44488", + "issue:44291|issue:44589", "issue:43644|issue:44079", "issue:40444|issue:45381", "issue:45325|issue:45381", "issue:44265|issue:45072", - "issue:43632|issue:45072", "issue:41720|issue:45081", + "issue:43632|issue:45072", "issue:43611|issue:43646", "issue:43531|issue:44291", - "issue:43994|issue:44898", "issue:42175|issue:43296", + "issue:43994|issue:44898", "issue:43454|issue:44466", - "issue:44387|issue:44617", "issue:42915|issue:43482", "issue:43526|issue:43883", - "issue:43828|issue:44683", - "issue:43525|issue:44589", - "issue:43792|issue:43957" + "issue:44387|issue:44617" ], - "prepared_review_unit_hash": "c02fb4469075169de6b190374cfa975df7ea85a17e05c5eb6f75d80770fec19f", + "prepared_review_unit_hash": "2d109976159c33823f486d687d02b80cfbf395531e73e16c6ea3ded976b51dde", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37339, - "estimated_input_tokens": 9335, - "estimated_eval_tokens": 18926 + "serialized_chars": 37355, + "estimated_input_tokens": 9339, + "estimated_eval_tokens": 18934 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly separate model- or feature-specific regressions rather than true duplicates. The only clear duplicate-style soft edge is the Qwen2.5-VL position-id/rope-index pair; the rest are too broad or too different to merge safely.", - "confidence": 0.7, - "canonical_issue_reason": "issue:45020 is the broadest umbrella report here: recent Transformers versions breaking models using `remote_code`. It best represents the shared v5-regression theme across the cluster.", + "summary": "This cluster is mostly a grab-bag of recent Transformers v5 regressions, with a few genuine near-duplicates around model initialization and Qwen2.5-VL position-id handling. Most soft pairs are only superficially similar and should stay separate.", + "confidence": 0.71, + "canonical_issue_reason": "issue:45020 is the broadest umbrella in the set: it captures the general class of recent Transformers-version regressions breaking model loading/behavior, which best represents the shared triage theme across many of these reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the strongest global representative because it captures the widest cross-model compatibility regression among these issues, whereas most others are narrower symptoms.", + "best_issue_reason": "issue:45020 is the best single triage anchor because it is open, broad, and most likely to accumulate links from the surrounding v5 regression reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:36032", - "right": "issue:41720", + "left": "issue:42915", + "right": "issue:44589", "accept": false, - "reason": "Unrelated failures: T5 tokenizer init error vs Qwen3 auto device-mapping CUDA assert." + "reason": "Both involve fp8/floats, but they report different failures and likely different code paths (Qwen3MoE FineGrainedFP8Config vs missing Float8 storage object)." }, { - "left": "issue:43425", - "right": "issue:44403", + "left": "issue:43818", + "right": "issue:45357", "accept": false, - "reason": "One is a Torch compatibility complaint; the other is a logging/noise issue when loading models." + "reason": "Different models and failures: Video-LLaVA tower/weight-sharing bug vs Qwen3.5 save_pretrained visual-key regression." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but one is about `_is_hf_initialized` while the other is about non-persistent buffers; not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Unrelated issues: Torch version compatibility vs noisy transformer loading output." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer name conflict and Qwen3 auto device-map CUDA assert are unrelated." }, { "left": "issue:44683", "right": "issue:45020", "accept": false, - "reason": "Different bug classes: compiled flex_attention on torch>=2.9 vs remote_code loading regressions." + "reason": "torch.compile/flex_attention failure is a different problem from remote_code model-loading regressions." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Different training bugs: flash-attention-3 degenerate training vs single-label zero-loss classification behavior." + "reason": "Different sequence-classification failures with different triggers; not the same underlying bug." }, { - "left": "issue:33666", - "right": "issue:40444", + "left": "issue:43482", + "right": "issue:45538", "accept": false, - "reason": "Both are Qwen2-VL-ish training issues, but one is multi-GPU training and the other is iterable-dataset/multi-image handling." + "reason": "Qwen2.5-GGUF loading under v5 and CLIPTokenizer model_max_length are unrelated tokenizer/loading behaviors." }, { - "left": "issue:43482", - "right": "issue:45538", + "left": "issue:33666", + "right": "issue:40444", "accept": false, - "reason": "Qwen2.5-GGUF loading failure is unrelated to CLIPTokenizer model_max_length behavior." + "reason": "Both are Qwen2-VL fine-tuning issues, but one is multi-GPU training and the other is IterableDataset/multi-image batching." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Generic loading noise is not the same bug as tokenizer loading failing for a specific model." + "reason": "General loading noise vs a specific tokenizer-loading regression for EMBEDDIA/est-roberta." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper runtime failure and Float8 storage lookup errors are different failure modes." + "reason": "Whisper runtime failure and Float8 storage lookup failure are different issues." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "ModelOutput key assignment and non-persistent buffer serialization are separate internals bugs." + "reason": "ModelOutput key assignment and non-persistent buffer junk are distinct internal-data regressions." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both involve batching/padding, but the MIMI padding-mask issue and Gemma3 sliding-window NaNs are distinct model-path failures." + "reason": "Both mention batched/masked behavior, but the affected models and concrete bugs are different." }, { "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 iterable dataset failures." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL multi-image IterableDataset finetuning." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "Qwen2.5-GGUF loading and missing pad_token_id on Llama4Config are unrelated." + "reason": "GGUF loading under v5 and missing pad_token_id on Llama4Config are different failures." }, { "left": "issue:43749", "right": "issue:44079", "accept": false, - "reason": "FSDP CPU RAM efficient loading and ModelOutput key assignment do not share the same code-path bug." + "reason": "FSDP CPU RAM efficient loading and ModelOutput key handling are not the same bug." }, { - "left": "issue:44291", - "right": "issue:44589", + "left": "issue:43881", + "right": "issue:44488", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized and Float8 storage lookup are separate Transformer v5 regressions." + "reason": "Different model-loading failures with no clear shared code path." }, { - "left": "issue:43881", - "right": "issue:44488", + "left": "issue:44291", + "right": "issue:44589", "accept": false, - "reason": "Different model loading failures for different models; no shared underlying bug is evident." + "reason": "Unexpected `_is_hf_initialized` argument and missing Float8 storage object are different regressions." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Non-persistent buffer junk values and ModelOutput key handling are unrelated." + "reason": "Both are v5 internal-behavior reports, but one is about buffers and the other about ModelOutput keys." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen2.5-VL reports, but one is dataset/multi-image training and the other is video position-id handling." + "reason": "Multi-image iterable finetuning and Qwen2.5-VL video vision_position_ids are related only at a high level." }, { "left": "issue:45325", "right": "issue:45381", "accept": true, - "reason": "Both report Qwen2.5-VL position-id/rope-index regressions in transformers 5.3.0 and likely share the same visual temporal position code path." + "reason": "Same Qwen2.5-VL temporal/vision position-id bug: both reports point to incorrect position_id computation in v5.3.0." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export with torch_compilable_check is a different issue from dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "torch.export/torch_compilable_check failure and bfloat16 dtype mismatch are unrelated." }, { - "left": "issue:43632", - "right": "issue:45072", + "left": "issue:41720", + "right": "issue:45081", "accept": false, - "reason": "_is_hf_initialized breakage and bfloat16 dtype mismatch are unrelated regression classes." + "reason": "Qwen3 auto device mapping CUDA assert is unrelated to the Mistral regex patch crash." }, { - "left": "issue:41720", - "right": "issue:45081", + "left": "issue:43632", + "right": "issue:45072", "accept": false, - "reason": "Qwen3 device-mapping CUDA assert and Mistral tokenizer regex patch crash are unrelated." + "reason": "`_is_hf_initialized` breakage and bfloat16 dtype mismatches do not share the same concrete bug." }, { "left": "issue:43611", "right": "issue:43646", - "accept": false, - "reason": "Both are v5 init/load regressions, but one is about `base_model_prefix` loading and the other about custom model initialization; not enough evidence of the same bug." + "accept": true, + "reason": "Both describe the same v5 model-initialization/loading regression affecting custom models/base_model_prefix behavior." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior and init_empty_weights argument handling are different problems." - }, - { - "left": "issue:43994", - "right": "issue:44898", - "accept": false, - "reason": "Different vision-model inference bugs: SigLIP2 pipeline outputs vs Perceiver interpolation/resolution failure." + "reason": "Qwen3-MoE sliding-window issues and init_empty_weights `_is_hf_initialized` errors are different code paths." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Package/backend dependency issue is unrelated to PaddleOCR-VL loading in vLLM." + "reason": "Backend packaging omission and PaddleOCR-VL loading failure are unrelated." }, { - "left": "issue:43454", - "right": "issue:44466", + "left": "issue:43994", + "right": "issue:44898", "accept": false, - "reason": "Both touch tied lm_head weights, but one is failure to tie weights causing bad generation and the other is inconsistent serialization by device; not the same concrete bug." + "reason": "Different vision-model failures: SIGLIP2 nonsensical outputs vs Perceiver non-default-resolution classification failure." }, { - "left": "issue:44387", - "right": "issue:44617", + "left": "issue:43454", + "right": "issue:44466", "accept": false, - "reason": "CUDA memory growth under int4 quantization is not the same as a specific Sam3Video OOM report." + "reason": "AyaVision lm_head tying failure and device-dependent serialization of tied weights are distinct problems." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen2.5-GGUF loading failure are unrelated." + "reason": "Qwen3MoE FP8 failure and Qwen2.5-GGUF v5 loading failure are unrelated." }, { "left": "issue:43526", @@ -3316,28 +3174,16 @@ "reason": "BeitImageProcessorFast label reduction bug and Molmo tied-weights attribute error are unrelated." }, { - "left": "issue:43828", - "right": "issue:44683", - "accept": false, - "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to compiled flex_attention breakage on newer torch." - }, - { - "left": "issue:43525", - "right": "issue:44589", - "accept": false, - "reason": "Missing pad_token_id on Llama4Config is unrelated to the Float8 storage object error." - }, - { - "left": "issue:43792", - "right": "issue:43957", + "left": "issue:44387", + "right": "issue:44617", "accept": false, - "reason": "Whisper model runtime failure and meta-device loading regressions are different code paths." + "reason": "Int4 memory regression and Sam3Video OOM are both memory-related but not the same underlying bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The single accepted soft edge (issue:45325 vs issue:45381) is plausibly the same Qwen2.5-VL position-id/rope-index regression, and the remaining pairwise rejections are cautious rather than overstated. The chosen representative issue:45020 is also supported by the packet as the broadest cross-model remote_code regression." + "feedback": "The summary stays grounded in the packet: it correctly treats the set as mostly unrelated v5 regressions with only a couple of plausible near-duplicates. The accepted soft pairs are reasonably conservative from the titles alone, and the chosen umbrella issue (45020) is a defensible broad anchor for the cluster." }, "evaluator_used": true, "retried": false, @@ -3356,7 +3202,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:34634", "issue:35707", @@ -3371,6 +3217,7 @@ "issue:43377", "issue:43404", "issue:43454", + "issue:43525", "issue:43526", "issue:43550", "issue:43606", @@ -3381,17 +3228,18 @@ "issue:43749", "issue:43792", "issue:43819", + "issue:43828", "issue:43873", "issue:43881", + "issue:43957", "issue:44186", "issue:44265", "issue:44291", - "issue:44423", "issue:44466", "issue:44534", + "issue:44589", "issue:44610", "issue:44683", - "issue:44756", "issue:44877", "issue:44898", "issue:44977", @@ -3399,7 +3247,6 @@ "issue:45072", "issue:45325", "issue:45357", - "issue:45406", "issue:45563", "issue:45584", "issue:45632", @@ -3407,16 +3254,19 @@ "issue:45701" ], "soft_pairs": [ - "issue:43701|issue:45325", + "issue:43828|issue:44683", + "issue:43525|issue:44589", + "issue:43792|issue:43957", "issue:43881|issue:44991", + "issue:43701|issue:45325", "issue:43262|issue:43746", "issue:43317|issue:43873", "issue:43632|issue:45563", "issue:40990|issue:45584", - "issue:34634|issue:35707", "issue:44291|issue:44877", - "issue:43873|issue:45701", + "issue:34634|issue:35707", "issue:43606|issue:44265", + "issue:43873|issue:45701", "issue:44186|issue:44265", "issue:44610|issue:45325", "issue:35707|issue:41720", @@ -3425,9 +3275,9 @@ "issue:43377|issue:44610", "issue:43377|issue:44898", "issue:43454|issue:45072", + "issue:44466|issue:45357", "issue:43749|issue:44534", "issue:39692|issue:40444", - "issue:44466|issue:45357", "issue:43792|issue:44683", "issue:43454|issue:43720", "issue:43296|issue:44991", @@ -3436,251 +3286,244 @@ "issue:45632|issue:45684", "issue:43526|issue:43881", "issue:43404|issue:45357", - "issue:41720|issue:44977", - "issue:44423|issue:45406", - "issue:40990|issue:45701", - "issue:44186|issue:45072", - "issue:43873|issue:44756" + "issue:41720|issue:44977" ], - "prepared_review_unit_hash": "0b75973bddd05926c3ff13f36fd7aeb520b2744466ded8df31128d4937bdc069", + "prepared_review_unit_hash": "a3c08204e9219a529d9a164bf9eb1e0aa8cfcd127865a3c6135ed9e14cf2f362", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 34, - "serialized_chars": 37393, - "estimated_input_tokens": 9349, - "estimated_eval_tokens": 18954 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37631, + "estimated_input_tokens": 9408, + "estimated_eval_tokens": 19072 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: it mixes unrelated model-specific regressions, tokenizer/processor bugs, offloading/loading failures, and a few serialization/config issues. There are no PRs in the cluster. The best anchor issue is the FSDP CPU RAM efficient loading report because it has the most discussion/inbound references and sits closest to the broader loader/offload failures, but the overall duplicate density is low.", - "confidence": 0.56, - "canonical_issue_reason": "Issue 43749 is the strongest anchor: it has the most activity/inbound references and describes a core loading-path regression that is nearer to several offloading/memory-related reports than the model-specific bugs.", + "summary": "This cluster is mostly a set of unrelated model-loading / processor / generation bugs. The only plausible duplicate pair is the two offloading+quantization loading failures; the rest are distinct code paths and should stay separate.", + "confidence": 0.63, + "canonical_issue_reason": "issue:43873 is the broadest and most central report in the cluster: it describes offloading not working with quantization, which subsumes the more specific device_map/offload loading failure.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43749 is the most representative operational regression in this set; it is broad enough to serve as a cluster anchor, even though most other items are not true duplicates.", + "best_issue_reason": "issue:43873 is the best representative issue because it covers the shared offloading/quantization failure mode and is phrased as the umbrella bug rather than one narrow instance.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43701", - "right": "issue:45325", + "left": "issue:43828", + "right": "issue:44683", + "accept": false, + "reason": "Different failures: autocast dtype mismatch vs compiled flex_attention failure on torch>=2.9." + }, + { + "left": "issue:43525", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated symptoms and code paths: missing pad_token_id vs missing Float8 storage object." + }, + { + "left": "issue:43792", + "right": "issue:43957", "accept": false, - "reason": "Both are generation/runtime regressions, but one is a checkpoint key mismatch and the other is a rope-index scaling bug in Qwen2.5-VL; different code paths." + "reason": "Different problem areas: Whisper runtime failure vs meta-device loading regression." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "Unrelated failures: one is a model loading failure for glm-4v-9b, the other is a tokenizer loading regression for a specific RoBERTa model." + "reason": "GLM-4V loading failure and tokenizer loading for a different model are not the same bug." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen2.5-VL rope/position-id scaling." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting is unrelated to GraniteSpeech PEFT checkpoint loading." + "reason": "Audio chat-template sampling-rate default bug is unrelated to GraniteSpeech PEFT checkpoint loading." }, { "left": "issue:43317", "right": "issue:43873", - "accept": false, - "reason": "Both mention offloading/quantization, but the titles point to different failures and not the same concrete bug." + "accept": true, + "reason": "Both report the same offloading/quantization loading path failing under device_map/CPU-GPU offload; one is a specific instance of the broader bug." }, { "left": "issue:43632", "right": "issue:45563", "accept": false, - "reason": "Transformers v5 initialization flag breakage is unrelated to a stale warning in paged generate()." + "reason": "Global v5 initialization breakage is unrelated to a paged generate warning about num_return_sequences." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Different functionality and symptoms: perplexity on WikiText-2 vs Whisper empty-transcription generation failure." - }, - { - "left": "issue:34634", - "right": "issue:35707", - "accept": false, - "reason": "Bark voice_preset failure and progressive generation with inputs_embeds/past_key_values are separate issues." + "reason": "High perplexity on GPT-OSS and Whisper empty-transcription generation are unrelated model behaviors." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "One is an unexpected _is_hf_initialized argument in init_empty_weights; the other is strict config loading for granite_speech." + "reason": "init_empty_weights/_is_hf_initialized argument breakage is not the same as strict config blocking granite_speech loading." }, { - "left": "issue:43873", - "right": "issue:45701", + "left": "issue:34634", + "right": "issue:35707", "accept": false, - "reason": "Offloading/quantization behavior is unrelated to a version-dependent tokenization change." + "reason": "Bark voice_preset issue and progressive generation with inputs_embeds/past_key_values are unrelated." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU-offload device mismatch for Bark is unrelated to torch.export failures with torch_compilable_check." + "reason": "CPU offload device mismatch for Bark is unrelated to torch.export failing on torch_compilable_check." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offloading failure is unrelated to a tokenization version-change regression." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer crashes on NER/padding are unrelated to export-time torch_compilable_check failures." + "reason": "LayoutLMv2 tokenizer crash is unrelated to torch.export compatibility." }, { "left": "issue:44610", "right": "issue:45325", "accept": false, - "reason": "Both are multimodal-model issues, but one is a processor image-size mismatch and the other is a temporal position-id scaling bug." + "reason": "Processor image-size mismatch is unrelated to Qwen2.5-VL temporal position-id scaling." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation with embeddings/cache and Qwen3 auto device-map cuda asserts are different bugs." + "reason": "Progressive generation with cached embeddings is unrelated to Qwen3 auto device mapping cudaErrorAssert." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "CPU offload device mismatch for Bark and DAC.from_latents/STE mismatch are unrelated." + "reason": "Both touch audio-ish models, but one is CPU offload device mismatch and the other is DAC latent consistency; not the same bug." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "Weight tying/serialization for AyaVision is unrelated to Bamba-9B-v2 torch.compile SDPA failure." + "reason": "lm_head weight tying failure is unrelated to torch.compile+SDPA behavior." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI batching/padding-mask divergence and OmDet processor image resolution mismatch are different code paths." + "reason": "Batched-vs-single encoder output mismatch is unrelated to OmDet-Turbo processor resolution mismatch." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Batched-vs-single MIMI outputs do not match a Perceiver interpolation bug." + "reason": "Missing padding mask support in MIMI encoder is not the same as Perceiver non-default resolution handling." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "A weight-tying bug is not the same as dtype mismatches in inference wrappers." + "reason": "Weight-tying/serialization issue is unrelated to bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "General lm_head serialization inconsistency is unrelated to Qwen3.5 visual encoder key saving." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "FSDP CPU RAM efficient loading and non-persistent buffer junk in v5 are both loader-related but not the same underlying bug." + "reason": "FSDP CPU RAM efficient loading breakage and non-persistent buffer junk are different initialization/loading bugs." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "Documentation example errors for SigLIP2 are unrelated to Qwen2.5-VL finetuning with multi-image iterable datasets." - }, - { - "left": "issue:44466", - "right": "issue:45357", - "accept": false, - "reason": "Both involve save_pretrained serialization, but one concerns lm_head.weight tying/serialization and the other visual encoder key names; different fixes." + "reason": "SigLIP2 doc example errors are unrelated to Qwen2.5-VL finetuning with multi-image iterable datasets." }, { "left": "issue:43792", "right": "issue:44683", "accept": false, - "reason": "Whisper runtime failure and compiled flex_attention failures on torch>=2.9 are unrelated." + "reason": "Whisper runtime failure is unrelated to compiled flex_attention on torch>=2.9." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "AyaVision lm_head tying and BitNet packed-weight unpacking are distinct loading/serialization bugs." + "reason": "lm_head weight tying bug is unrelated to packed-weight unpacking during accelerate loading." }, { "left": "issue:43296", "right": "issue:44991", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM and tokenizer loading regression for est-roberta are unrelated." + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to est-roberta tokenizer loading." }, { "left": "issue:43296", "right": "issue:44877", "accept": false, - "reason": "PaddleOCR-VL loading in vLLM is unrelated to strict config loading for granite_speech." + "reason": "vLLM/PaddleOCR-VL loading failure is unrelated to strict granite_speech config validation." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Qwen3 auto device-map cudaErrorAssert and Llava Next segmentation fault are different model-loading crashes." + "reason": "Qwen3 auto device mapping cuda assert and Llava Next segfault are different model-specific loading crashes." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "One is a trust_remote_code cache-path collision, the other is read-only permission propagation during save_pretrained; different filesystem bugs." + "reason": "Cache-path collisions for trust_remote_code and permission propagation in save_pretrained are unrelated." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BeitImageProcessorFast label reduction and glm-4v-9b model loading failure are unrelated." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to GLM-4V loading failure." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "lm_head weight tying in Mistral3ForConditionalGeneration is not the same bug as incorrect visual encoder keys in Qwen3.5 save_pretrained." + "reason": "lm_head tying in Mistral3 is unrelated to incorrect visual encoder keys in Qwen3.5 save_pretrained." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Auto device-mapping crash and flash-attention generation failure are distinct Qwen3.5 problems." - }, - { - "left": "issue:44423", - "right": "issue:45406", - "accept": false, - "reason": "Continuous-batching multimodal crash with a string-to-device error is unrelated to Gemma4Processor missing _tokenizer." - }, - { - "left": "issue:40990", - "right": "issue:45701", - "accept": false, - "reason": "Perplexity regression on a language model is unrelated to version-dependent tokenization changes." - }, - { - "left": "issue:44186", - "right": "issue:45072", - "accept": false, - "reason": "NER/padding tokenizer crashes and dtype mismatches in bfloat16 inference are different bugs." - }, - { - "left": "issue:43873", - "right": "issue:44756", - "accept": false, - "reason": "Quantized offloading problems and the Strix Halo mmap/OOM request are not the same underlying issue." + "reason": "Auto device mapping crash on A800 and flash-attention generation failure are different Qwen3.5 issues." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted soft pair (43317/43873) is a reasonable conservative duplicate call based on the titles. The remaining false verdicts are generally cautious enough given the available evidence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3695,14 +3538,11 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:30064", - "issue:33357", - "issue:34634", - "issue:35707", "issue:36010", "issue:36331", + "issue:40990", "issue:41720", "issue:42617", "issue:42915", @@ -3721,37 +3561,44 @@ "issue:43606", "issue:43708", "issue:43720", - "issue:43746", "issue:43782", "issue:43873", "issue:44186", "issue:44292", "issue:44361", + "issue:44387", + "issue:44423", "issue:44479", "issue:44492", "issue:44509", - "issue:44514", - "issue:44545", "issue:44610", "issue:44671", "issue:44743", + "issue:44756", "issue:44877", "issue:44898", "issue:44913", "issue:44918", "issue:44991", "issue:45059", + "issue:45072", "issue:45161", "issue:45406", "issue:45464", - "issue:45701" + "issue:45701", + "issue:45704" ], "soft_pairs": [ + "issue:44423|issue:45406", + "issue:40990|issue:45701", + "issue:44186|issue:45072", + "issue:43873|issue:44756", "issue:43577|issue:43708", + "issue:44387|issue:45704", "issue:42617|issue:42915", "issue:44877|issue:44991", - "issue:43450|issue:45701", "issue:44479|issue:44918", + "issue:43450|issue:45701", "issue:44610|issue:44898", "issue:43550|issue:43720", "issue:36331|issue:42915", @@ -3763,267 +3610,259 @@ "issue:43295|issue:45701", "issue:43550|issue:43606", "issue:41720|issue:42617", - "issue:41720|issue:45161", "issue:43531|issue:43572", + "issue:41720|issue:45161", "issue:44292|issue:44479", "issue:43388|issue:45059", "issue:44186|issue:44361", - "issue:42915|issue:43322", "issue:43540|issue:43782", + "issue:42915|issue:43322", "issue:41720|issue:43873", "issue:36010|issue:42617", "issue:45406|issue:45464", "issue:36331|issue:41720", "issue:43421|issue:45701", - "issue:33357|issue:35707", - "issue:44492|issue:44509", - "issue:44514|issue:44545", - "issue:30064|issue:34634", - "issue:41720|issue:43746", - "issue:43606|issue:43720" + "issue:44492|issue:44509" ], - "prepared_review_unit_hash": "c48353bc824f4ef298c15d072acfce77eb662ea3805b4ed48bf307b658b01525", + "prepared_review_unit_hash": "c35f09a929e6485b695a29fdfa30baf06a5f865f65f120ce847b2292999aa646", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 34, - "serialized_chars": 37733, - "estimated_input_tokens": 9434, - "estimated_eval_tokens": 19124 + "serialized_chars": 37212, + "estimated_input_tokens": 9303, + "estimated_eval_tokens": 18862 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly unrelated issues with one clear duplicate pair: the two Qwen2_5_VLProcessor.apply_chat_template batched-input/padding=False reports. All other soft pairs look like broad thematic similarity at best, not the same bug.", - "confidence": 0.96, - "canonical_issue_reason": "issue:44545 is the best canonical issue: it is the same bug as issue:44514, but has more inbound references/discussion and the same concrete failure mode (batched Qwen2_5_VLProcessor.apply_chat_template with padding=False).", + "summary": "This cluster is largely over-grouped by semantic similarity: it spans unrelated tokenizer, multimodal/serve, config, trainer, offloading, and memory regressions. None of the soft pairs look like true duplicates, so I would not merge any of them.", + "confidence": 0.9, + "canonical_issue_reason": "No strong canonical issue emerges because the cluster is heterogeneous. If one issue must anchor it, 43873 is the broadest infrastructure-level report (offloading/quantization), but it does not \u05d1\u05d0\u05de\u05ea unify the rest.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44545 is the strongest representative of the only true duplicate in this cluster, with the clearest description of the concrete crash path.", + "best_issue_reason": "Issue 43873 is the best available anchor because it is open, broader than the others, and sits in a recurring system area (offloading/quantization). Still, it is only a weak representative for this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both involve serve/multimodal crashes, but they hit different code paths and different processor errors; not the same bug." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "One is high perplexity on a specific model/dataset, the other is tokenization changing across versions; related area, different underlying issue." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash vs dtype mismatch in a different model path; no shared concrete failure." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both mention offloading/memory, but one is quantization offload behavior and the other is mmap OOM on Strix Halo; different root causes." + }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Different problems: dtype propagation in Blip2 loading vs Trainer checkpoint resume step \u0917\u0923culation." + "reason": "BLIP2 dtype loading bug and Trainer checkpoint max_steps regression are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Both are memory-related, but one is int4 CUDA reserved memory and the other is an apex RMSNorm leak in T5; not the same bug." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "Both are model/runtime failures, but one is 3D parallel script startup and the other is Qwen3MoE FP8 loading." + "reason": "3d_parallel.py failure and Qwen3Moe FP8 failure are different model/runtime problems." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Both concern config/tokenizer loading regressions, but they involve different models and different failure paths." + "reason": "Strict config loading and tokenizer loading failures are separate API regressions." }, { - "left": "issue:43450", - "right": "issue:45701", + "left": "issue:44479", + "right": "issue:44918", "accept": false, - "reason": "Video processor batching bug vs version-dependent tokenization changes; not the same underlying defect." + "reason": "Video-input regression in inference vs Qwen3.5 embedding unpacking in TRL training; different paths and symptoms." }, { - "left": "issue:44479", - "right": "issue:44918", + "left": "issue:43450", + "right": "issue:45701", "accept": false, - "reason": "Both mention Qwen multimodal workflows, but one is a video-input regression and the other is SFT/input-embeds packing." + "reason": "Video batch-shape bug and tokenization-version change are unrelated." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both involve image size handling, but they affect different models and different components (processor output vs model interpolation handling)." + "reason": "Both are image preprocessing mismatches, but the models, expected sizes, and failure modes differ." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "Unrelated failures: torch.compile+SDPA on Bamba vs accelerate loading of packed BitNet weights." + "reason": "torch.compile/SDPA failure and packed-weight loading failure are distinct bugs." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Trainer API signature change vs Qwen3MoE FP8 loading; no shared code path." + "reason": "CustomTrainer kwargs regression and Qwen3Moe FP8 loading failure are unrelated." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Unexpected trainer keyword argument vs 3d_parallel runtime failure; unrelated." + "reason": "Trainer loss-signature change and 3d_parallel.py failure are different issues." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Tokenizer crash on NER/padding vs image resolution mismatch in OmDet-Turbo; different subsystems." + "reason": "Both are input-processing bugs, but they affect different processors and failure modes." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Both are config/API consistency issues, but they concern different config fields and models." + "reason": "Config API discrepancy and rotary_pct reload loss are not the same underlying defect." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "Masked LM prediction regression vs Qwen3.5 recurrent cache-state bug; unrelated." + "reason": "CamemBERT masked-LM regression and Qwen3.5 recurrent-state reset bug are unrelated model-specific issues." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Both mention loading/quantization, but one is FP8 config failure and the other is dequantized model offload loading." + "reason": "FineGrainedFP8 loading failure and dequantized model offload loading failure are different code paths." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "Both are version regressions, but one is processor/tokenizer API breakage and the other is tokenization differences." + "reason": "Processor/tokenizer regression and tokenization-version change are not the same bug." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Different models and different device/offload failures; not the same bug." + "reason": "Compile+SDPA issue and CPU offload device mismatch are unrelated." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert vs a separate 3d_parallel execution issue." + "reason": "Auto device-map cudaErrorAssert and 3d_parallel.py failure are different runtime problems." }, { - "left": "issue:41720", - "right": "issue:45161", + "left": "issue:43531", + "right": "issue:43572", "accept": false, - "reason": "Both involve model parallelism, but one is a device-map crash and the other is TP not working for GPT-OSS MoE." + "reason": "Sliding-window behavior in Qwen3-MoE and missing pad_token_idx in StableLmConfig are unrelated." }, { - "left": "issue:43531", - "right": "issue:43572", + "left": "issue:41720", + "right": "issue:45161", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE vs missing pad_token_idx in StableLmConfig; unrelated." + "reason": "Both touch distributed inference on MoE models, but the concrete failures and affected paths differ." }, { "left": "issue:44292", "right": "issue:44479", "accept": false, - "reason": "Different model families and different failures: NVFP4 execution vs video-input regression." + "reason": "NVFP4 model loading/running error and video-input regression are different issues." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric gathering tuple-label bug vs SAM3 text/bounding-box behavior; unrelated." + "reason": "Metric-gather label truncation and SAM3 PCS text/box behavior are unrelated." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Two tokenizer bugs, but on different tokenizers and different crash mechanisms." + "reason": "Two tokenizer failures, but on different tokenizers and different failure mechanisms." }, { - "left": "issue:42915", - "right": "issue:43322", + "left": "issue:43540", + "right": "issue:43782", "accept": false, - "reason": "Qwen3MoE FP8 loading vs Llava Next segmentation fault; no shared underlying bug." + "reason": "Video processing failure and from_pretrained weight_only error are not the same bug." }, { - "left": "issue:43540", - "right": "issue:43782", + "left": "issue:42915", + "right": "issue:43322", "accept": false, - "reason": "Both involve Qwen multimodal models, but one is video-input processing and the other is from_pretrained weight_only loading." + "reason": "Qwen3Moe FP8 failure and Llava Next segmentation fault are unrelated model-loading problems." }, { "left": "issue:41720", "right": "issue:43873", "accept": false, - "reason": "Both relate to offloading/quantization, but the concrete failures are different and not mergeable as one fix." + "reason": "Auto device mapping crash and quantization offloading issue are both infra-related but not the same defect." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "GenerationMixin import regression vs 3d_parallel script failure; unrelated." + "reason": "ImportError for GenerationMixin and 3d_parallel.py runtime failure have no common code path." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Different APIs and models: Gemma4Processor _tokenizer attribute crash vs Qwen3.5 streaming inference failure." + "reason": "Serve crashes in different multimodal flows: missing _tokenizer vs streaming inference failure." }, { "left": "issue:36331", "right": "issue:41720", "accept": false, - "reason": "Trainer loss-signature incompatibility vs Qwen3 auto device map CUDA assert; unrelated." + "reason": "Trainer compute_loss signature regression and device-map cudaErrorAssert are unrelated." }, { "left": "issue:43421", "right": "issue:45701", "accept": false, - "reason": "Runtime post-processor update bug vs general version-dependent tokenization changes; too broad to be same bug." - }, - { - "left": "issue:33357", - "right": "issue:35707", - "accept": false, - "reason": "MacOS bus error in CLIP model loading vs progressive generation with inputs_embeds/past_key_values; unrelated." + "reason": "Runtime special-token post-processing update bug and tokenization-version change are different issues." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A typo in cache strategy docs/code vs documentation for removed pipeline tasks; not the same issue." - }, - { - "left": "issue:44514", - "right": "issue:44545", - "accept": true, - "reason": "Exact same bug report: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding=False." - }, - { - "left": "issue:30064", - "right": "issue:34634", - "accept": false, - "reason": "ImageProcessor void segmentation maps vs BarkProcessor voice_preset; unrelated." - }, - { - "left": "issue:41720", - "right": "issue:43746", - "accept": false, - "reason": "Qwen3 auto device map CUDA assert vs GraniteSpeech PEFT/local checkpoint loading; different code paths." - }, - { - "left": "issue:43606", - "right": "issue:43720", - "accept": false, - "reason": "CPU-offload device mismatch in Bark-small vs packed-weight unpacking in BitNet; unrelated." + "reason": "A cache-strategy typo and stale docs about removed pipeline tasks are separate documentation/config issues." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet: only issue:44514/issue:44545 is a clearly supported duplicate pair, and the canonical choice of issue:44545 is consistent with its slightly higher discussion/inbound reference count and the same failure mode. The other soft-edge rejections are conservative and do not overclaim mergeability or sameness based on the titles/body summaries provided." - }, - "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 }, @@ -4038,13 +3877,14 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ + "issue:30064", + "issue:33357", "issue:34567", "issue:34634", "issue:35707", "issue:39692", - "issue:40444", "issue:41720", "issue:42886", "issue:43381", @@ -4052,11 +3892,9 @@ "issue:43525", "issue:43531", "issue:43550", - "issue:43596", "issue:43606", - "issue:43618", "issue:43653", - "issue:43697", + "issue:43720", "issue:43746", "issue:43782", "issue:43844", @@ -4066,13 +3904,13 @@ "issue:44062", "issue:44077", "issue:44186", - "issue:44220", "issue:44368", "issue:44451", + "issue:44514", + "issue:44545", "issue:44589", "issue:44610", "issue:44625", - "issue:44661", "issue:44671", "issue:44843", "issue:44898", @@ -4083,13 +3921,19 @@ "issue:45042", "issue:45072", "issue:45081", - "issue:45183", "issue:45356", "issue:45357", "issue:45381", - "issue:45479" + "issue:45479", + "issue:45704" ], "soft_pairs": [ + "issue:33357|issue:35707", + "issue:44514|issue:44545", + "issue:30064|issue:34634", + "issue:41720|issue:43746", + "issue:43606|issue:43720", + "issue:43873|issue:45704", "issue:43550|issue:43653", "issue:44625|issue:45479", "issue:44368|issue:45357", @@ -4097,252 +3941,235 @@ "issue:43606|issue:43653", "issue:34567|issue:35707", "issue:42886|issue:44843", - "issue:34634|issue:41720", "issue:44062|issue:44451", - "issue:43782|issue:45381", + "issue:34634|issue:41720", "issue:44451|issue:44991", + "issue:43782|issue:45381", "issue:43421|issue:43653", "issue:44898|issue:45072", "issue:44077|issue:44625", "issue:44987|issue:45042", - "issue:43994|issue:44671", "issue:44918|issue:44936", "issue:43531|issue:44589", "issue:43421|issue:44186", + "issue:43994|issue:44671", "issue:43381|issue:43844", "issue:39692|issue:44610", "issue:45081|issue:45356", - "issue:43525|issue:43531", - "issue:43596|issue:44220", - "issue:43618|issue:43697", - "issue:43746|issue:45381", - "issue:45183|issue:45479", - "issue:39692|issue:43873", - "issue:43746|issue:43873", - "issue:40444|issue:43746", - "issue:44220|issue:44661" + "issue:43525|issue:43531" ], - "prepared_review_unit_hash": "bff0e04eb6149b4023c5ecd6f58b071e26ee5ad3bdb423aaebad7e97ff712da4", + "prepared_review_unit_hash": "1e084c2a9d3d5df268b261583074a68fdc8e5977a8f7b9ad5604c697d2a083d3", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37250, - "estimated_input_tokens": 9313, - "estimated_eval_tokens": 18882 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36086, + "estimated_input_tokens": 9022, + "estimated_eval_tokens": 18300 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are distinct bugs across different models/subsystems. The only strong thematic overlap is around tokenizer/offline-regression reports, but the candidate pairs still look like separate code paths or model-specific failures, so I would not merge them as duplicates.", - "confidence": 0.59, - "canonical_issue_reason": "Issue 44843 is the most concrete, self-contained bug report in the set: it names the exact failing path (`AutoTokenizer.from_pretrained` / `_patch_mistral_regex`) and the offline-mode regression clearly.", + "summary": "Mostly heterogeneous cluster with a few true duplicate/near-duplicate issue pairs. The only strong acceptances are the Qwen2.5-VL batched chat-template crash pair and the HF_HUB_OFFLINE/Mistral-regex tokenizer-loading regression pair; the rest are related-at-a-distance but not the same concrete bug.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue cleanly represents the whole cluster because the items span unrelated bugs. If forced to anchor the genuine duplicate pair, issue 44514 is the clearest representative: it is a near-verbatim match to 44545 with the same Qwen2_5_VLProcessor.apply_chat_template batched-padding crash.", "canonical_pr_reason": null, - "best_issue_reason": "Best representative of the cluster because it has a precise root cause, clear repro surface, and broader triage value for tokenizer-loading regressions.", + "best_issue_reason": "Issue 44514 is the best issue anchor because it is the clearest exact-match duplicate in the set and describes a concrete, narrowly scoped crash path that is duplicated by 44545.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "Different subsystems and failure modes: MacOS CLIP bus error vs progressive generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Near-verbatim duplicate: both report the same Qwen2_5_VLProcessor.apply_chat_template batched-input crash when padding=False." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Unrelated bugs: void segmentation-map processing vs BarkProcessor voice_preset handling." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different models and code paths: Qwen3 auto device mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve loading/offload, but the concrete failures differ: Bark CPU offload device mismatch vs BitNet packed-weights handling during accelerate loading." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Not the same bug: quantization/offloading behavior vs a T5 apex.FusedRMSNorm memory leak." + }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "Different bugs: torch.compile+SDPA model failure vs tokenizer special-token registration/decoding." + "reason": "Different root causes: torch.compile/SDPA on Bamba vs tokenizer special-token registration for BigBird." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Both involve classification config fields, but one is Qwen3.5 text-config propagation and the other is a general zero-loss bug for num_labels=1." + "reason": "Related around num_labels, but one is config propagation and the other is degenerate single-label loss; not the same concrete bug." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "Unrelated failures: tie_word_embeddings warning during LoRA fine-tuning vs incorrect visual encoder keys on save_pretrained." + "reason": "Same model family, but different issues: LoRA warning about tie_word_embeddings vs incorrect visual-encoder keys saved in Qwen3.5." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Related theme, but different underlying problems: runtime post-processor update vs save/load loss of custom token IDs." + "reason": "Both are tokenizer/config related, but runtime post-processor refresh is not the same as save/load loss of custom token IDs causing IndexError." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same underlying offline tokenizer regression: unconditional _patch_mistral_regex/model_info() behavior breaks HF_HUB_OFFLINE cache loading." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Device mismatch with CPU offload is unrelated to BigBirdTokenizer special-token decode behavior." + "reason": "Different bugs in different models: Bark offload device mismatch vs BigBird mask-token special registration." }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "Trainer token accounting bug vs progressive generation with inputs_embeds/past_key_values are separate code paths." + "reason": "TrainerState token-count update bug is unrelated to progressive generation with cached past_key_values." }, { - "left": "issue:42886", - "right": "issue:44843", + "left": "issue:44062", + "right": "issue:44451", "accept": false, - "reason": "Both mention HF_HUB_OFFLINE and tokenizer loading, but 44843 is specifically the Mistral regex patch path while 42886 is broader and not clearly the same bug." + "reason": "Both are loading-related, but the concrete failure signatures differ and do not show the same root cause." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "BarkProcessor voice_preset and Qwen3 auto device mapping CUDA assert are unrelated." + "reason": "BarkProcessor voice_preset failure and Qwen3 auto device mapping CUDA assert are unrelated." }, { - "left": "issue:44062", - "right": "issue:44451", + "left": "issue:44451", + "right": "issue:44991", "accept": false, - "reason": "AddedToken constructor error is not the same as a generic ScandiBERT loading failure." + "reason": "Both concern model/tokenizer loading, but the affected models and failure mechanisms are different and not shown to share a code path." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Qwen3VL weight_only load error and Qwen2.5-VL video vision_position_ids mismatch are different issues." - }, - { - "left": "issue:44451", - "right": "issue:44991", - "accept": false, - "reason": "Different tokenizer-loading failures for different models; no shared concrete code path is evident." + "reason": "Different Qwen multimodal bugs: weight_only loading error vs incorrect video vision_position_ids." }, { "left": "issue:43421", "right": "issue:43653", "accept": false, - "reason": "Both concern token/special-token behavior, but one is post-processor refresh and the other is mask-token registration/decoding." + "reason": "Same broad tokenizer area, but not the same concrete bug: runtime special-token post-processing vs empty decode from mask-token registration." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Perceiver non-default resolution failure is unrelated to SwitchTransformers/TimmWrapperModel bfloat16 dtype mismatch." + "reason": "Unrelated image-resolution issue vs dtype mismatch in inference." }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "PatchTSMixer post_init policy and Qwen3.5 num_labels propagation are unrelated." + "reason": "No overlap in concrete code path: patchtsmixer post_init policy vs Qwen3.5 num_labels propagation." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Loading physical-intelligence/fast under newer transformers is unrelated to PIL backend processors requiring torchvision." - }, - { - "left": "issue:43994", - "right": "issue:44671", - "accept": false, - "reason": "SigLIP2 pipeline nonsense and CamemBERT MLM predictions are different model-specific regressions." + "reason": "Different problems: loading a specific model vs PIL image processors incorrectly requiring torchvision." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "TRL SFT unpacking issue is not the same as evaluate-after-train trainer lifecycle failure." + "reason": "Different trainer failures: unpacking Qwen3.5 input embeddings in TRL SFT vs evaluate-after-train behavior." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Qwen3-MoE sliding_window bug and Float8 storage lookup failure are unrelated." + "reason": "Sliding-window model issue is unrelated to missing Float8 storage type errors." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Runtime special-token post-processor updates and LayoutLMv2Tokenizer NER/padding crashes are different tokenizer bugs." + "reason": "Different tokenizer bugs: runtime special-token post-processing vs LayoutLMv2 NER padding/truncation crashes." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and behaviors: SigLIP2 nonsensical pipeline outputs vs CamemBERT masked LM regression." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction is not the same as ZeRO-3 gradient growth." + "reason": "Eval-mode gradient checkpointing limitation is unrelated to ZeRO-3 gradient growth." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Both are doc/processor-model mismatch flavored, but they concern different models and failure modes." + "reason": "Documentation example/model-processor mismatch is unrelated to OmDet-Turbo processor resolution mismatch." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Mistral regex AttributeError and Kimi-K2.5 codec handling regression are distinct tokenizer bugs." + "reason": "Related Mistral-regex family, but the evidence suggests distinct bugs: one is an AttributeError during tokenizer loading, the other is a Kimi-K2.5 codec-handling regression with a misleading warning." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Llama4Config pad_token_id attribute error and Qwen3-MoE sliding_window behavior are unrelated." - }, - { - "left": "issue:43596", - "right": "issue:44220", - "accept": false, - "reason": "DeepSpeed zero3/BertModel index error is unrelated to `_torch_extract_fbank_features()`." - }, - { - "left": "issue:43618", - "right": "issue:43697", - "accept": false, - "reason": "CLIPOutput attentions assignment and RTDetrV2 differing outputs are different model regressions." - }, - { - "left": "issue:43746", - "right": "issue:45381", - "accept": false, - "reason": "PEFT/local checkpoint loading for GraniteSpeech is unrelated to Qwen2.5-VL video position IDs." - }, - { - "left": "issue:45183", - "right": "issue:45479", - "accept": false, - "reason": "Input validation error messaging is not the same as the single-label zero-loss bug." - }, - { - "left": "issue:39692", - "right": "issue:43873", - "accept": false, - "reason": "SigLIP2 example issues and quantization offloading behavior are different failures." - }, - { - "left": "issue:43746", - "right": "issue:43873", - "accept": false, - "reason": "Local checkpoint loading with adapters and quantization/offloading issues are separate." - }, - { - "left": "issue:40444", - "right": "issue:43746", - "accept": false, - "reason": "Qwen2.5-VL iterable dataset multi-image finetuning failure is unrelated to GraniteSpeech PEFT checkpoint loading." - }, - { - "left": "issue:44220", - "right": "issue:44661", - "accept": false, - "reason": "Audio feature extraction issue and add-new-model-like failure in TOKENIZER_MAPPING_NAMES are unrelated." + "reason": "Unrelated config attribute error vs sliding_window bug in Qwen3-MoE." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level summary is mostly grounded, and the 44514\u219444545 and 42886\u219444843 edges look supportable from the titles. However, the accept on 45081\u219445356 is a bit too strong for the evidence provided: both touch _patch_mistral_regex/fix_mistral_regex, but one is an AttributeError crash in Mistral tokenizer loading while the other is a Kimi-K2.5 regression about codec handling and a misleading warning. That reads as related family, not clearly the same bug/near-duplicate. Because the soft-edge verdicts should stay conservative, I\u2019d reject this output." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -4357,61 +4184,70 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:30064", "issue:36010", + "issue:39692", + "issue:40444", "issue:41720", "issue:43322", "issue:43335", + "issue:43381", "issue:43421", "issue:43531", "issue:43540", "issue:43550", "issue:43577", - "issue:43582", + "issue:43596", + "issue:43618", "issue:43643", "issue:43644", "issue:43653", + "issue:43697", "issue:43716", "issue:43720", + "issue:43746", "issue:43749", "issue:43756", - "issue:43782", "issue:43818", "issue:43819", - "issue:43825", - "issue:43827", + "issue:43873", "issue:43927", - "issue:43994", "issue:44008", "issue:44062", "issue:44186", - "issue:44442", + "issue:44220", "issue:44448", - "issue:44485", "issue:44488", - "issue:44509", - "issue:44561", + "issue:44661", "issue:44671", "issue:44811", - "issue:44821", "issue:44843", "issue:44877", "issue:44898", - "issue:44977", - "issue:44991", - "issue:45020", "issue:45072", "issue:45081", - "issue:45084", + "issue:45183", "issue:45381", - "issue:45685" + "issue:45479", + "issue:45685", + "issue:45704" ], "soft_pairs": [ + "issue:43596|issue:44220", + "issue:43618|issue:43697", + "issue:41720|issue:45704", + "issue:43746|issue:45381", + "issue:43381|issue:45704", + "issue:45183|issue:45479", + "issue:39692|issue:43873", + "issue:44220|issue:44661", + "issue:40444|issue:43746", + "issue:43746|issue:43873", "issue:43540|issue:44008", - "issue:43335|issue:43756", "issue:43716|issue:44008", + "issue:43335|issue:43756", "issue:43720|issue:43819", "issue:43653|issue:43720", "issue:30064|issue:36010", @@ -4419,235 +4255,218 @@ "issue:44843|issue:45081", "issue:43421|issue:44811", "issue:43322|issue:45685", - "issue:44186|issue:44898", "issue:43577|issue:43818", + "issue:44186|issue:44898", "issue:43550|issue:45072", "issue:43531|issue:43643", "issue:43643|issue:44877", "issue:43550|issue:43927", "issue:43540|issue:45381", "issue:44448|issue:44671", - "issue:43644|issue:43749", - "issue:43782|issue:44977", - "issue:43825|issue:43994", - "issue:43653|issue:43819", - "issue:44561|issue:45020", - "issue:43577|issue:43582", - "issue:43825|issue:43827", - "issue:43550|issue:43819", - "issue:44821|issue:45084", - "issue:44186|issue:44442", - "issue:41720|issue:44991", - "issue:44485|issue:44509" + "issue:43644|issue:43749" ], - "prepared_review_unit_hash": "709b3f65a91c62cd7f378cc62aec5b45734ceeb505d802c6e5c0e104625a44f3", + "prepared_review_unit_hash": "40e4647c1813ba9acdc604dd1240f56463c6a7045c037da2257b84896c3b3f64", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36824, - "estimated_input_tokens": 9206, - "estimated_eval_tokens": 18668 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36147, + "estimated_input_tokens": 9037, + "estimated_eval_tokens": 18330 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: the soft pairs span unrelated model-, tokenizer-, config-, and runtime-specific bugs. I would not merge any of the suggested soft edges; the closest thematic overlap is the recent-version remote_code regression group, but even that looks broader than a single duplicate bug.", - "confidence": 0.78, - "canonical_issue_reason": "Issue 45020 is the broadest umbrella here (recent transformers versions breaking remote_code models) and is the most plausible representative of the shared v5/remote_code regression theme.", + "summary": "These are mostly unrelated issue reports spanning tokenizer/config loading, multimodal preprocessing, dtype/device-map bugs, model-specific regressions, and a few generic runtime failures. The soft-similarity pairs share keywords or subsystem overlap, but they do not describe the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43749 has the broadest impact and strongest signal in this set (multiple discussion threads and inbound references) around a concrete broken loading path, making it the best representative issue if one must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45020 best captures the widest recurring theme in the set; the rest are narrower, model-specific failures or documentation issues rather than one duplicate bug.", + "best_issue_reason": "issue:43749 is the most central/visible issue in the cluster and clearly describes a specific user-facing regression in loading behavior.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43540", - "right": "issue:44008", + "left": "issue:43596", + "right": "issue:44220", "accept": false, - "reason": "Different models and failure modes: Qwen3OmniMoe video processing ValueError vs Gemma3n audio_mel_mask AttributeError." + "reason": "Different failures: BertModel + ZeRO-3 initialization vs audio feature extraction helper error." }, { - "left": "issue:43335", - "right": "issue:43756", + "left": "issue:43618", + "right": "issue:43697", "accept": false, - "reason": "Unrelated model internals: sparse-layer config bug in SwitchTransformers vs RoPE-layer count issue in Smollm3." + "reason": "CLIP attentions regression is unrelated to RTDetrV2 output drift." }, { - "left": "issue:43716", - "right": "issue:44008", + "left": "issue:41720", + "right": "issue:45704", "accept": false, - "reason": "Different multimodal paths: image-preprocessor dtype mismatch vs audio tensor field collision." + "reason": "Qwen3 auto device mapping CUDA assert is unrelated to T5/apex RMSNorm memory leak." }, { - "left": "issue:43720", - "right": "issue:43819", + "left": "issue:43746", + "right": "issue:45381", "accept": false, - "reason": "Packed-weight loading/unpacking bug vs DAC from_latents forward mismatch; different code paths and symptoms." + "reason": "GraniteSpeech PEFT local checkpoint loading differs from Qwen2.5-VL video position-id bug." }, { - "left": "issue:43653", - "right": "issue:43720", + "left": "issue:43381", + "right": "issue:45704", "accept": false, - "reason": "Tokenizer special-token registration bug vs BitNet accelerate loading issue; not the same underlying defect." + "reason": "Gradient checkpointing eval-mode restriction is not the same as T5 fused RMSNorm leak." }, { - "left": "issue:30064", - "right": "issue:36010", + "left": "issue:45183", + "right": "issue:45479", "accept": false, - "reason": "ImageProcessor segmentation-map failure and GenerationMixin import error are unrelated." + "reason": "Generic input validation error messaging is unrelated to single-label classification zero-loss behavior." }, { - "left": "issue:44062", - "right": "issue:44488", + "left": "issue:39692", + "right": "issue:43873", "accept": false, - "reason": "AddedToken kwarg duplication vs tokenizer load failure for sleng-bert; different tokenizer bugs." + "reason": "SigLIP2 docs/model mismatch and quantization issue is not the same as quantization offloading behavior." }, { - "left": "issue:44843", - "right": "issue:45081", - "accept": false, - "reason": "Both involve _patch_mistral_regex, but one is an offline model_info call and the other is a backend_tokenizer AttributeError; separate failures." - }, - { - "left": "issue:43421", - "right": "issue:44811", + "left": "issue:44220", + "right": "issue:44661", "accept": false, - "reason": "Runtime post-processor update behavior vs Whisper batch_decode skip_special_tokens handling; different APIs and fixes." + "reason": "Audio feature extraction bug is unrelated to add-new-model-like failure in tokenizer mapping names." }, { - "left": "issue:43322", - "right": "issue:45685", + "left": "issue:40444", + "right": "issue:43746", "accept": false, - "reason": "Llava Next segfault and MPS histogram Int limitation are unrelated." + "reason": "Qwen2.5-VL iterable dataset/multi-image finetuning failure is unrelated to GraniteSpeech local checkpoint loading." }, { - "left": "issue:44186", - "right": "issue:44898", + "left": "issue:43746", + "right": "issue:43873", "accept": false, - "reason": "LayoutLMv2 tokenizer crash on NER/padding vs Perceiver image-size failure; different components." + "reason": "PEFT checkpoint loading and quantization offloading are different code paths and different symptoms." }, { - "left": "issue:43577", - "right": "issue:43818", + "left": "issue:43540", + "right": "issue:44008", "accept": false, - "reason": "BLIP2 dtype propagation bug vs Video-LLaVA temporal-attention/weight-sharing issue; not the same bug." + "reason": "Qwen3OmniMoe video processing ValueError is unrelated to Gemma3n variable-name collision in forward()." }, { - "left": "issue:43550", - "right": "issue:45072", + "left": "issue:43716", + "right": "issue:44008", "accept": false, - "reason": "torch.compile+SDPA failure in Bamba vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel." + "reason": "Mistral-3 image preprocessor dtype mismatch is not the same as Gemma3n audio mask attribute error." }, { - "left": "issue:43531", - "right": "issue:43643", + "left": "issue:43335", + "right": "issue:43756", "accept": false, - "reason": "Qwen3-MoE sliding_window bug and trust_remote_code config field loss are unrelated." + "reason": "Sparse-layer creation bug in SwitchTransformers is distinct from Smollm3 RoPE-layer mismatch." }, { - "left": "issue:43643", - "right": "issue:44877", + "left": "issue:43720", + "right": "issue:43819", "accept": false, - "reason": "Both are config-loading problems, but one is missing fields under trust_remote_code and the other is strict config rejection; different bugs." + "reason": "BitNet packed-weight loading bug is unrelated to DAC.from_latents / missing STE mismatch." }, { - "left": "issue:43550", - "right": "issue:43927", + "left": "issue:43653", + "right": "issue:43720", "accept": false, - "reason": "Compile/SDPA runtime error vs DiaConfig custom token-ID persistence issue; unrelated." + "reason": "BigBirdTokenizer special-token decode bug is unrelated to BitNet packed-weight unpacking during accelerate loading." }, { - "left": "issue:43540", - "right": "issue:45381", + "left": "issue:30064", + "right": "issue:36010", "accept": false, - "reason": "Different video-model bugs on different models: Qwen3OmniMoe processing error vs Qwen2.5-VL vision_position_ids mismatch." + "reason": "Void segmentation map processing and GenerationMixin import failure are completely different issues." }, { - "left": "issue:44448", - "right": "issue:44671", + "left": "issue:44062", + "right": "issue:44488", "accept": false, - "reason": "Both are output regressions, but for different models and different symptoms; not one duplicate fix." + "reason": "AddedToken special-argument conflict is unrelated to loading cjvt/sleng-bert." }, { - "left": "issue:43644", - "right": "issue:43749", + "left": "issue:44843", + "right": "issue:45081", "accept": false, - "reason": "Non-persistent buffer initialization bug vs FSDP CPU RAM efficient loading issue; different mechanisms." + "reason": "Offline model_info patch issue and tokenizer backend_tokenizer crash are different Mistral-regex failures." }, { - "left": "issue:43782", - "right": "issue:44977", + "left": "issue:43421", + "right": "issue:44811", "accept": false, - "reason": "Qwen3VL from_pretrained weight_only load error vs Qwen3.5 flash-attention generation issue; distinct code paths." + "reason": "Runtime post-processor update for special tokens is not the same as Whisper batch_decode skip_special_tokens handling." }, { - "left": "issue:43825", - "right": "issue:43994", + "left": "issue:43322", + "right": "issue:45685", "accept": false, - "reason": "Pipeline error-message regression vs SigLIP2 nonsensical outputs; unrelated." + "reason": "Llava Next segmentation fault and MPS histogram Int error are unrelated runtime failures." }, { - "left": "issue:44561", - "right": "issue:45020", + "left": "issue:43577", + "right": "issue:43818", "accept": false, - "reason": "Related theme, but 44561 is a specific API-removal break while 45020 is a broad umbrella for remote_code regressions; not clearly the same underlying bug." + "reason": "BLIP2 dtype propagation bug is unrelated to Video-LLaVA temporal-attention/weight-sharing defect." }, { - "left": "issue:43577", - "right": "issue:43582", + "left": "issue:44186", + "right": "issue:44898", "accept": false, - "reason": "BLIP2 dtype propagation bug and Apple Silicon caching_allocator_warmup TypeError are unrelated." + "reason": "LayoutLMv2Tokenizer NER/padding crash is not Perceiver non-default resolution failure." }, { - "left": "issue:43825", - "right": "issue:43827", + "left": "issue:43550", + "right": "issue:45072", "accept": false, - "reason": "Runtime pipeline error message and docs still referencing removed tasks are different artifacts, not one duplicate bug." + "reason": "Bamba SDPA torch.compile failure is a different model/path than SwitchTransformers and TimmWrapperModel bfloat16 mismatches." }, { - "left": "issue:43550", - "right": "issue:43819", + "left": "issue:43531", + "right": "issue:43643", "accept": false, - "reason": "Bamba torch.compile/SDPA issue and DAC from_latents/STE mismatch are unrelated." + "reason": "Qwen3-MoE sliding_window bug is unrelated to AutoConfig trust_remote_code missing fields." }, { - "left": "issue:44821", - "right": "issue:45084", + "left": "issue:43643", + "right": "issue:44877", "accept": false, - "reason": "AutoImageProcessor URL loading and template-node compile failure are different subsystems." + "reason": "Missing fields from AutoConfig with trust_remote_code is different from strict config blocking granite_speech loading." }, { - "left": "issue:44186", - "right": "issue:44442", + "left": "issue:43550", + "right": "issue:43927", "accept": false, - "reason": "Tokenizer crash in LayoutLMv2 vs AutoTokenizer failing FastSpeech2ConformerTokenizer; different models and failure modes." + "reason": "Bamba SDPA compile bug is unrelated to DiaConfig losing custom token IDs on save/load." }, { - "left": "issue:41720", - "right": "issue:44991", + "left": "issue:43540", + "right": "issue:45381", "accept": false, - "reason": "Qwen3 auto device-map cuda assert and est-roberta tokenizer load failure are unrelated." + "reason": "Video-input processing failure in Qwen3OmniMoe is distinct from Qwen2.5-VL video vision_position_ids mismatch." }, { - "left": "issue:44485", - "right": "issue:44509", + "left": "issue:44448", + "right": "issue:44671", "accept": false, - "reason": "GLM-5 RoPE implementation discussion vs docs still mentioning removed pipeline tasks; not the same bug or change." + "reason": "Pegasus v4/v5 output drift is unrelated to CamemBERT masked-LM prediction regression." }, { - "left": "issue:43653", - "right": "issue:43819", + "left": "issue:43644", + "right": "issue:43749", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug and DAC latent/forward mismatch are unrelated." + "reason": "Non-persistent buffer junk in Transformers v5 is not the same as FSDP CPU RAM efficient loading breakage." } ] }, @@ -4669,24 +4488,20 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:30064", "issue:34634", "issue:36010", "issue:36331", - "issue:38175", + "issue:41720", "issue:42175", - "issue:43257", "issue:43262", - "issue:43329", - "issue:43452", - "issue:43454", "issue:43526", + "issue:43550", "issue:43577", "issue:43582", "issue:43606", - "issue:43638", "issue:43650", "issue:43653", "issue:43697", @@ -4694,264 +4509,271 @@ "issue:43749", "issue:43782", "issue:43819", - "issue:43854", + "issue:43825", + "issue:43827", "issue:43906", + "issue:43994", "issue:44060", + "issue:44186", "issue:44291", "issue:44368", + "issue:44442", + "issue:44485", + "issue:44509", "issue:44554", - "issue:44568", + "issue:44561", "issue:44610", "issue:44671", "issue:44805", + "issue:44821", "issue:44857", - "issue:44912", "issue:44936", "issue:44977", "issue:44991", + "issue:45020", "issue:45072", "issue:45081", "issue:45084", "issue:45325", "issue:45406", - "issue:45488", - "issue:45538", - "issue:45600", "issue:45701" ], "soft_pairs": [ + "issue:43782|issue:44977", + "issue:44561|issue:45020", + "issue:43653|issue:43819", + "issue:43825|issue:43994", + "issue:43577|issue:43582", + "issue:43550|issue:43819", + "issue:43825|issue:43827", + "issue:44821|issue:45084", + "issue:41720|issue:44991", + "issue:44485|issue:44509", + "issue:44186|issue:44442", "issue:44671|issue:45701", "issue:30064|issue:36331", "issue:44805|issue:44936", "issue:43606|issue:45072", - "issue:43650|issue:43906", "issue:42175|issue:45081", - "issue:34634|issue:36010", - "issue:44060|issue:44368", - "issue:43782|issue:45325", + "issue:43650|issue:43906", "issue:45081|issue:45406", + "issue:43782|issue:45325", + "issue:44060|issue:44368", + "issue:34634|issue:36010", "issue:44610|issue:45072", "issue:43749|issue:44291", "issue:43720|issue:45072", "issue:43526|issue:43697", "issue:43582|issue:44857", "issue:43526|issue:44554", - "issue:43262|issue:43819", - "issue:43329|issue:45600", - "issue:45084|issue:45538", - "issue:43653|issue:44912", - "issue:34634|issue:36331", - "issue:44568|issue:45488", - "issue:43452|issue:44991", - "issue:43257|issue:43329", - "issue:43638|issue:44936", - "issue:38175|issue:44610", - "issue:43454|issue:43577", - "issue:43854|issue:44977" + "issue:43262|issue:43819" ], - "prepared_review_unit_hash": "1de67c9371c00e0e8925233bbbf1c9902455b1541657389e7094be6ceb2220b5", + "prepared_review_unit_hash": "399d3c76bd6649d7fcbde980289f5e4987a051d2a38adbdb09275ec2c53d1368", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35962, - "estimated_input_tokens": 8991, - "estimated_eval_tokens": 18238 + "serialized_chars": 35710, + "estimated_input_tokens": 8928, + "estimated_eval_tokens": 18112 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a loose collection of unrelated Transformers bug reports. The soft pairs mostly share broad themes like loading, tokenization, dtype, or trainer regressions, but none look like the same concrete underlying bug/change, so no soft edge should be accepted.", + "summary": "Mostly unrelated issues; the only clear duplicate-style match is the v5 pipeline-removal docs/error-message cleanup pair.", "confidence": 0.88, - "canonical_issue_reason": null, + "canonical_issue_reason": "issue:43825 is the best representative of the one real cluster theme: it is the concrete runtime-facing v5 pipeline-removal mismatch, and issue:43827 looks like the docs-only sibling of the same cleanup.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45072 is the most central concrete report among this set: it is a specific bfloat16 dtype-mismatch regression and appears closest to several other loose inference/loading failures, making it the best representative if one issue must anchor the cluster.", + "best_issue_reason": "issue:43825 is the most actionable and specific item in the set, with the clearest duplicate relationship and strongest chance to anchor a merged fix.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44671", - "right": "issue:45701", + "left": "issue:43782", + "right": "issue:44977", "accept": false, - "reason": "Both involve tokenization/model-version behavior, but one is a CamemBERT masked-LM regression and the other is a generic version-related tokenization change; too broad to be the same bug." + "reason": "Different models and failures: a Qwen3VL weight_only load error versus a Qwen3.5 flash-attention generation bug." }, { - "left": "issue:30064", - "right": "issue:36331", + "left": "issue:44561", + "right": "issue:45020", "accept": false, - "reason": "Image segmentation-map processing and a CustomTrainer compute_loss signature error are unrelated code paths." + "reason": "Both mention remote_code breakage, but one is a specific v5 API removal and the other is a broad umbrella report; not the same concrete bug." }, { - "left": "issue:44805", - "right": "issue:44936", + "left": "issue:43653", + "right": "issue:43819", "accept": false, - "reason": "Mask/tensor shape mismatch in indexing vs trainer evaluate-after-train failure; different failure modes and subsystems." + "reason": "Unrelated tokenizer special-token handling versus DAC latent reconstruction mismatch." }, { - "left": "issue:43606", - "right": "issue:45072", + "left": "issue:43825", + "right": "issue:43994", "accept": false, - "reason": "Both mention inference/offload issues, but Bark CPU offload device mismatch and SwitchTransformers/TimmWrapper bfloat16 dtype mismatches are different concrete bugs." + "reason": "One is a v5 pipeline-removal error-message/docs mismatch; the other is a model output quality regression." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:43577", + "right": "issue:43582", "accept": false, - "reason": "Generic contentless issue title versus an isolated reproduction of a specific upstream bug; no shared underlying defect is evident." + "reason": "Different subsystems and symptoms: dtype propagation in BLIP2 loading versus an Apple Silicon allocator TypeError." }, { - "left": "issue:42175", - "right": "issue:45081", + "left": "issue:43550", + "right": "issue:43819", "accept": false, - "reason": "Packaging/backend dependency issue and a Mistral tokenizer regex crash are different problems." + "reason": "Bamba torch.compile/SDPA failure is unrelated to the DAC forward/from_latents mismatch." }, { - "left": "issue:34634", - "right": "issue:36010", - "accept": false, - "reason": "BarkProcessor voice_preset behavior and GenerationMixin import failure do not describe the same bug." + "left": "issue:43825", + "right": "issue:43827", + "accept": true, + "reason": "Same underlying v5 pipeline-removal cleanup: runtime/error-message wording and docs still referencing removed pipeline tasks." }, { - "left": "issue:44060", - "right": "issue:44368", + "left": "issue:44821", + "right": "issue:45084", "accept": false, - "reason": "Both concern Qwen3/Qwen3.5 weight-tie warnings, but one is an incorrect tied-weights mapping and the other is a user-facing warning in fine-tuning; not the same concrete failure." + "reason": "AutoImageProcessor URL loading and template compilation errors are unrelated code paths." }, { - "left": "issue:43782", - "right": "issue:45325", + "left": "issue:41720", + "right": "issue:44991", "accept": false, - "reason": "Qwen3VL weight_only loading error and Qwen2.5-VL rope-index scaling bug are unrelated model-path problems." + "reason": "Different regressions: Qwen3 auto device mapping CUDA assert versus tokenizer loading for EMBEDDIA/est-roberta." }, { - "left": "issue:45081", - "right": "issue:45406", + "left": "issue:44485", + "right": "issue:44509", "accept": false, - "reason": "Tokenizer regex patch crash and Gemma4Processor missing _tokenizer are separate tokenizer implementation issues." + "reason": "GLM-5 RoPE implementation discussion is unrelated to docs still mentioning removed pipeline tasks." }, { - "left": "issue:44610", - "right": "issue:45072", + "left": "issue:44186", + "right": "issue:44442", "accept": false, - "reason": "OmDet-Turbo input-size mismatch and bfloat16 dtype mismatches affect different model paths and symptom types." + "reason": "LayoutLMv2 NER/padding crash and FastSpeech2Conformer tokenizer loading failure are different tokenizer issues." }, { - "left": "issue:43749", - "right": "issue:44291", + "left": "issue:44671", + "right": "issue:45701", "accept": false, - "reason": "FSDP CPU RAM-efficient loading and init_empty_weights/_is_hf_initialized argument handling are related only at a very high level; not the same bug." + "reason": "CamemBERT masked-LM regression is not the same as a generic version-dependent tokenization change report." }, { - "left": "issue:43720", - "right": "issue:45072", + "left": "issue:30064", + "right": "issue:36331", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading is a different failure than generic bfloat16 inference dtype mismatches." + "reason": "Image processor void segmentation maps and CustomTrainer.compute_loss kwarg handling are unrelated." }, { - "left": "issue:43526", - "right": "issue:43697", + "left": "issue:44805", + "right": "issue:44936", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug and RTDetrV2 output divergence are distinct model/processor behaviors." + "reason": "Mask-shape IndexError and trainer train/evaluate lifecycle failure are different bugs." }, { - "left": "issue:43582", - "right": "issue:44857", + "left": "issue:43606", + "right": "issue:45072", "accept": false, - "reason": "Apple Silicon TypeError in caching_allocator_warmup and CUDA float16 AMP loss crash are different runtime environments and code paths." + "reason": "CPU offload device mismatch is unrelated to bfloat16 dtype mismatches." }, { - "left": "issue:43526", - "right": "issue:44554", + "left": "issue:42175", + "right": "issue:45081", "accept": false, - "reason": "A processor label-reduction bug and an MPS attention correctness issue are unrelated." + "reason": "Packaging/back-end installation and Mistral regex patch crashes are unrelated." }, { - "left": "issue:43262", - "right": "issue:43819", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "Audio sampling-rate defaulting in apply_chat_template and DAC from_latents vs forward mismatch are unrelated multimodal bugs." + "reason": "A malformed/empty issue title is not the same as an isolated reproduction of another issue." }, { - "left": "issue:43329", - "right": "issue:45600", + "left": "issue:45081", + "right": "issue:45406", "accept": false, - "reason": "Undefined video-token helper names in multimodal token counting and stale auto_mappings references to removed configs are separate maintenance bugs." + "reason": "Both are tokenizer/processor loading issues, but they involve different missing attributes and different models." }, { - "left": "issue:45084", - "right": "issue:45538", + "left": "issue:43782", + "right": "issue:45325", "accept": false, - "reason": "A template compilation TypeError and a CLIPTokenizer max-length constant issue are not the same problem." + "reason": "Qwen3VL weight_only loading and Qwen2.5-VL rope-index scaling are different failure modes." }, { - "left": "issue:43653", - "right": "issue:44912", + "left": "issue:44060", + "right": "issue:44368", "accept": false, - "reason": "BigBirdTokenizer special-token registration and MXFP4 quantization fallback are different tokenizer vs quantization issues." + "reason": "Both are about Qwen weight-tying warnings, but they describe different warning sources and different model behaviors." }, { "left": "issue:34634", - "right": "issue:36331", + "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated." + "reason": "BarkProcessor voice_preset and GenerationMixin import errors are unrelated." }, { - "left": "issue:44568", - "right": "issue:45488", + "left": "issue:44610", + "right": "issue:45072", "accept": false, - "reason": "mdeberta-v3-base add_special_tokens regression and LlamaTokenizer pre-tokenizer override are both tokenizer regressions, but they affect different tokenizer classes and mechanisms." + "reason": "Processor image-size mismatch and dtype mismatch in inference are different bugs." }, { - "left": "issue:43452", - "right": "issue:44991", + "left": "issue:43749", + "right": "issue:44291", "accept": false, - "reason": "Both are tokenizer-loading regressions, but one is specifically about gguf_file support and the other about a model-specific load failure; insufficient evidence they are the same bug." + "reason": "FSDP CPU RAM efficient loading and init_empty_weights argument handling are separate loading-path issues." }, { - "left": "issue:43257", - "right": "issue:43329", + "left": "issue:43720", + "right": "issue:45072", "accept": false, - "reason": "Qwen3 MOE weight conversion during accelerate+deepspeed loading and multimodal token counting video-branch errors are unrelated." + "reason": "BitNet packed-weight unpacking under accelerate loading is unrelated to bfloat16 dtype mismatches." }, { - "left": "issue:43638", - "right": "issue:44936", + "left": "issue:43526", + "right": "issue:43697", "accept": false, - "reason": "Zero3/non-pretrained Bert indexing failure and trainer evaluate-after-train failure are different training bugs." + "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output differences are different image-processing/model regressions." }, { - "left": "issue:38175", - "right": "issue:44610", + "left": "issue:43582", + "right": "issue:44857", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and OmDet-Turbo input-size mismatch are different model-specific inference issues." + "reason": "AppleSilicon caching allocator TypeError and CUDA AMP crash are unrelated platform-specific failures." }, { - "left": "issue:43454", - "right": "issue:43577", + "left": "issue:43526", + "right": "issue:44554", "accept": false, - "reason": "AyaVision lm_head tying failure and BLIP2 dtype propagation bug are distinct model-configuration problems." + "reason": "Label-array reduction bug and MPS attention correctness issue are unrelated." }, { - "left": "issue:43854", - "right": "issue:44977", + "left": "issue:43262", + "right": "issue:43819", "accept": false, - "reason": "GLM-4.7-Flash unit-test loading failure and Qwen3.5 flash-attention generation bug are different model-path regressions." + "reason": "Audio chat-template sampling-rate defaulting and DAC latent reconstruction mismatch are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted soft edge, issue:43825 \u2194 issue:43827, is well-supported by the packet as the same v5 pipeline-removal cleanup theme (runtime wording plus docs references). The remaining rejections stay cautious and do not overclaim duplication beyond the evidence shown." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -4966,41 +4788,38 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:33357", "issue:34567", + "issue:34634", "issue:36010", "issue:36331", + "issue:38175", "issue:39692", "issue:40990", - "issue:41720", - "issue:42915", "issue:43257", "issue:43262", - "issue:43317", "issue:43329", "issue:43450", "issue:43452", + "issue:43454", "issue:43531", - "issue:43540", "issue:43550", "issue:43577", - "issue:43582", + "issue:43638", "issue:43645", + "issue:43653", "issue:43696", "issue:43697", "issue:43720", - "issue:43742", "issue:43746", "issue:43749", "issue:43756", "issue:43819", + "issue:43854", "issue:43927", "issue:43931", - "issue:44291", "issue:44479", - "issue:44560", "issue:44568", "issue:44589", "issue:44610", @@ -5008,23 +4827,37 @@ "issue:44811", "issue:44849", "issue:44912", - "issue:44987", + "issue:44936", + "issue:44977", + "issue:44991", "issue:45042", - "issue:45072", + "issue:45084", "issue:45356", "issue:45357", - "issue:45381", + "issue:45488", "issue:45538", + "issue:45600", "issue:45701" ], "soft_pairs": [ + "issue:43329|issue:45600", + "issue:45084|issue:45538", + "issue:34634|issue:36331", + "issue:43653|issue:44912", + "issue:44568|issue:45488", + "issue:43452|issue:44991", + "issue:43638|issue:44936", + "issue:43257|issue:43329", + "issue:38175|issue:44610", + "issue:43454|issue:43577", + "issue:43854|issue:44977", "issue:43746|issue:44811", "issue:43257|issue:43696", "issue:44589|issue:45042", "issue:39692|issue:44912", "issue:36010|issue:43645", - "issue:43819|issue:43927", "issue:44568|issue:45356", + "issue:43819|issue:43927", "issue:43720|issue:44610", "issue:43749|issue:44756", "issue:43550|issue:44912", @@ -5034,242 +4867,237 @@ "issue:36010|issue:36331", "issue:43257|issue:43931", "issue:34567|issue:40990", - "issue:43257|issue:43756", - "issue:43329|issue:43450", "issue:44849|issue:45701", + "issue:43257|issue:43756", "issue:43697|issue:45357", - "issue:44987|issue:45538", - "issue:44479|issue:45381", - "issue:43577|issue:45072", - "issue:41720|issue:43540", - "issue:43452|issue:43582", - "issue:33357|issue:34567", - "issue:43927|issue:44912", - "issue:43540|issue:44560", - "issue:42915|issue:43742", - "issue:43317|issue:45538", - "issue:44291|issue:45042" + "issue:43329|issue:43450" ], - "prepared_review_unit_hash": "50130eef93c4b5c3abefe19aecad963e153d4addf98c61a4bf19bc27988c850b", + "prepared_review_unit_hash": "76ed711b7e3ba35c25b66980d46795fb5625e2281d66bb64986cda7b27a72fd9", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37101, - "estimated_input_tokens": 9276, - "estimated_eval_tokens": 18808 + "serialized_chars": 37130, + "estimated_input_tokens": 9283, + "estimated_eval_tokens": 18822 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are clearly about unrelated bugs in different models, loaders, tokenizers, or processors. Only one soft edge looks like the same underlying regression family; the rest are too broad or too model-specific to merge.", - "confidence": 0.84, - "canonical_issue_reason": "issue:44479 is the strongest canonical issue here because it captures a broad, version-specific video regression across multiple Qwen model families and subsumes the narrower Qwen2.5-VL report.", + "summary": "Most candidate pairs are clearly different bugs despite some shared subsystems. The only plausible duplicate is the Qwen/VL video-input regression pair, where one issue gives the concrete `_get_num_multimodal_tokens` video-branch failure and the other reports the broader regression across Qwen VL variants.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43329 is the most concrete root-cause report: it names the broken video branch in `_get_num_multimodal_tokens` and the undefined helpers, which makes it the best anchor for the shared Qwen video regression.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best representative anchor: it describes a concrete regression, names the affected model families, and is the broader of the closely related video-input reports.", + "best_issue_reason": "issue:44479 is the clearest user-facing umbrella: it reports the regression across multiple Qwen VL variants, so it's the best triage entry point.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43746", - "right": "issue:44811", + "left": "issue:43329", + "right": "issue:45600", "accept": false, - "reason": "Different bugs: GraniteSpeech PEFT/local checkpoint loading vs Whisper batch_decode skip_special_tokens handling." + "reason": "Different bugs: multimodal video token counting vs stale auto_mappings config references." }, { - "left": "issue:43257", - "right": "issue:43696", + "left": "issue:45084", + "right": "issue:45538", "accept": false, - "reason": "Different failure modes: Qwen3 MoE weight conversion/loading vs GPT-oss OOM." + "reason": "Unrelated failures: template compilation error vs tokenizer max-length constant issue." }, { - "left": "issue:44589", - "right": "issue:45042", + "left": "issue:34634", + "right": "issue:36331", "accept": false, - "reason": "Unrelated: float8 storage lookup error vs PIL image processor torchvision dependency regression." + "reason": "Different code paths: BarkProcessor voice preset vs Trainer custom compute_loss signature." }, { - "left": "issue:39692", + "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "Docs/example issues for SigLIP2 vs a quantization/loading fallback bug for gpt-oss-20b." + "reason": "Tokenizer special-token registration bug vs MXFP4 quantization/loading fallback bug." }, { - "left": "issue:36010", - "right": "issue:43645", + "left": "issue:44568", + "right": "issue:45488", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to the Jupyter custom-model initialization regression." + "reason": "Both are tokenizer-related, but they affect different models and different mechanisms." }, { - "left": "issue:43819", - "right": "issue:43927", + "left": "issue:43452", + "right": "issue:44991", "accept": false, - "reason": "Different model behaviors: DAC latent reconstruction mismatch vs DiaConfig token-id persistence loss." + "reason": "Different loading bugs: gguf_file regression vs a specific tokenizer-loading failure for est-roberta." }, { - "left": "issue:44568", - "right": "issue:45356", + "left": "issue:43638", + "right": "issue:44936", "accept": false, - "reason": "Both are tokenizer-related, but they describe different regressions in different tokenizers and code paths." + "reason": "Different training/evaluation failures with different triggers and symptoms." }, { - "left": "issue:43720", - "right": "issue:44610", + "left": "issue:43257", + "right": "issue:43329", "accept": false, - "reason": "Packed-weight loading for BitNet is unrelated to OmDet-Turbo processor image-size mismatch." + "reason": "Different issues: Qwen3 MoE weight conversion/loading vs multimodal video token counting." }, { - "left": "issue:43749", - "right": "issue:44756", + "left": "issue:38175", + "right": "issue:44610", "accept": false, - "reason": "FSDP RAM-efficient loading and Strix Halo mmap OOM are both loading/memory topics, but not the same bug." + "reason": "Both are vision-model mismatches, but they are different models and different concrete bugs." }, { - "left": "issue:43550", - "right": "issue:44912", + "left": "issue:43454", + "right": "issue:43577", "accept": false, - "reason": "Different problems: torch.compile/SDPA failure vs MXFP4 quantization fallback." + "reason": "Different model-init bugs: lm_head tying vs dtype propagation to qformer." }, { - "left": "issue:43329", - "right": "issue:44479", + "left": "issue:43854", + "right": "issue:44977", "accept": false, - "reason": "Both involve video paths, but one is an undefined-helper bug in token counting and the other is a broader video-input regression." + "reason": "Different model-specific flash-attention/load problems; same general area is not enough." }, { - "left": "issue:43262", + "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting is unrelated to Whisper decode special-token handling." + "reason": "Local checkpoint/PEFT loading bug vs processor batch_decode skip_special_tokens bug." }, { - "left": "issue:43531", - "right": "issue:45042", + "left": "issue:43257", + "right": "issue:43696", "accept": false, - "reason": "Sliding-window behavior for Qwen3-MoE is unrelated to PIL backend torchvision requirements." + "reason": "Different failures: MoE weight conversion/loading vs CUDA OOM." }, { - "left": "issue:36010", - "right": "issue:36331", + "left": "issue:44589", + "right": "issue:45042", "accept": false, - "reason": "Import path breakage for GenerationMixin is unrelated to compute_loss signature incompatibility." + "reason": "Unrelated issues: Float8 storage lookup vs image processor torchvision dependency." }, { - "left": "issue:43257", - "right": "issue:43931", + "left": "issue:39692", + "right": "issue:44912", "accept": false, - "reason": "Qwen3 MoE weight-conversion/loading and Qwen3-VL weight-shape mismatch are different model-loading failures." + "reason": "Documentation example/model-processor mismatch vs quantization fallback on a different model." }, { - "left": "issue:34567", - "right": "issue:40990", + "left": "issue:36010", + "right": "issue:43645", "accept": false, - "reason": "Trainer token-count tracking and GPT-oss perplexity are unrelated training issues." + "reason": "Import-path regression and Jupyter custom-model initialization are different bugs." }, { - "left": "issue:43257", - "right": "issue:43756", + "left": "issue:44568", + "right": "issue:45356", "accept": false, - "reason": "Different models and symptoms: Qwen3 MoE loading vs Smollm3 RoPE-layer drop regression." + "reason": "Different tokenizer regressions affecting different models and different failure modes." }, { - "left": "issue:43329", - "right": "issue:43450", + "left": "issue:43819", + "right": "issue:43927", "accept": false, - "reason": "Different video bugs: undefined helper usage in multimodal token counting vs batched video processor shape output." + "reason": "Different model behavior: DAC latent reconstruction mismatch vs Dia config token ID persistence." }, { - "left": "issue:44849", - "right": "issue:45701", + "left": "issue:43720", + "right": "issue:44610", "accept": false, - "reason": "Qwen3.5 hidden-state behavior is unrelated to a version-dependent tokenization change." + "reason": "Different subsystems: BitNet packed-weight loading vs OmDet processor/model image-size mismatch." }, { - "left": "issue:43697", - "right": "issue:45357", + "left": "issue:43749", + "right": "issue:44756", "accept": false, - "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained key mapping are separate regressions." + "reason": "Both concern loading/memory, but one is FSDP CPU RAM-efficient loading and the other is mmap on Strix Halo." }, { - "left": "issue:44987", - "right": "issue:45538", + "left": "issue:43550", + "right": "issue:44912", "accept": false, - "reason": "Physical-intelligence/fast loading failure is unrelated to CLIP tokenizer max-length handling." + "reason": "Different bugs: torch.compile+SDPA failure vs quantization fallback." }, { - "left": "issue:44479", - "right": "issue:45381", + "left": "issue:43329", + "right": "issue:44479", "accept": true, - "reason": "Same underlying video regression family: both are transformers==5.3.0 video-input bugs in the Qwen2.5-VL/Qwen3-VL path, with 45381 describing the narrower vision_position_ids symptom." + "reason": "Same underlying video-input regression in the Qwen/VL multimodal token path; 43329 identifies the concrete failing branch that plausibly explains the broader 44479 report." }, { - "left": "issue:43577", - "right": "issue:45072", + "left": "issue:43262", + "right": "issue:44811", "accept": false, - "reason": "BLIP2 dtype propagation and SwitchTransformers/TimmWrapperModel bfloat16 mismatches are distinct dtype bugs." + "reason": "Audio chat-template sampling-rate default vs Whisper batch_decode special-token handling are different bugs." }, { - "left": "issue:41720", - "right": "issue:43540", + "left": "issue:43531", + "right": "issue:45042", "accept": false, - "reason": "Both are Qwen-related runtime errors, but one is auto-device-map CUDA assert and the other is a video-input ValueError in a different model." + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs PIL image-processor torchvision dependency." }, { - "left": "issue:43452", - "right": "issue:43582", + "left": "issue:36010", + "right": "issue:36331", "accept": false, - "reason": "gguf_file tokenizer/model loading and AppleSilicon caching_allocator_warmup TypeError are unrelated." + "reason": "Different regression categories: import error vs custom-trainer method signature breakage." }, { - "left": "issue:33357", - "right": "issue:34567", + "left": "issue:43257", + "right": "issue:43931", "accept": false, - "reason": "MacOS CLIP bus error and TrainerState token counting are unrelated." + "reason": "Different load issues on different Qwen model families; not the same concrete bug." }, { - "left": "issue:43927", - "right": "issue:44912", + "left": "issue:34567", + "right": "issue:40990", "accept": false, - "reason": "DiaConfig token ID persistence and gpt-oss quantization fallback are unrelated failures." + "reason": "Trainer token-count tracking bug vs gpt-oss perplexity issue are unrelated." }, { - "left": "issue:43540", - "right": "issue:44560", + "left": "issue:44849", + "right": "issue:45701", "accept": false, - "reason": "Both are video errors, but they hit different models and have different symptoms (ValueError vs StopIteration)." + "reason": "Qwen3.5 hidden-state bug vs generic tokenization-version change are not the same issue." }, { - "left": "issue:42915", - "right": "issue:43742", + "left": "issue:43257", + "right": "issue:43756", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and MobileLLM key error are different loading bugs." + "reason": "Different Qwen3 MoE loading bug vs Smollm3 RoPE-layer mismatch." }, { - "left": "issue:43317", - "right": "issue:45538", + "left": "issue:43697", + "right": "issue:45357", "accept": false, - "reason": "Dequantized model offload loading and CLIPTokenizer max-length are unrelated." + "reason": "Different model regressions: RTDetrV2 output changes vs Qwen3.5 save_pretrained visual-key regression." }, { - "left": "issue:44291", - "right": "issue:45042", + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Different subsystems and failure modes; no shared concrete code path." + }, + { + "left": "issue:43746", + "right": "issue:44811", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized argument handling is unrelated to PIL backend torchvision requirements." + "reason": "Local checkpoint loading for GraniteSpeech vs Whisper batch_decode are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall: the anchor choice is supported by the titles, and the lone accepted soft edge (44479 \u2194 45381) is a conservative merge of closely related video-regression reports in the same Qwen/5.3.0 path. The rejected pairs are framed cautiously as distinct model or code-path bugs, which fits the packet." + "feedback": "The clustering is grounded in the packet and the soft-edge calls are generally conservative. The Qwen video pair (issue:43329 / issue:44479) is a plausible duplicate/umbrella match, and the other rejections are supported by clearly different titles and bug surfaces." }, "evaluator_used": true, "retried": false, @@ -5288,47 +5116,48 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:28282", "issue:29127", + "issue:33357", "issue:34567", "issue:34689", "issue:36010", - "issue:36331", "issue:38175", + "issue:41720", "issue:42175", "issue:42915", + "issue:43317", "issue:43452", "issue:43493", "issue:43526", "issue:43540", + "issue:43577", "issue:43582", "issue:43643", - "issue:43688", "issue:43742", "issue:43749", "issue:43756", - "issue:43782", "issue:43824", "issue:43906", + "issue:43927", "issue:43994", "issue:44062", "issue:44164", "issue:44291", "issue:44479", - "issue:44484", "issue:44488", - "issue:44561", - "issue:44704", + "issue:44560", "issue:44737", - "issue:44792", "issue:44821", "issue:44898", + "issue:44912", "issue:44938", - "issue:44964", + "issue:44987", "issue:44991", - "issue:45081", + "issue:45042", + "issue:45072", "issue:45137", "issue:45325", "issue:45375", @@ -5336,255 +5165,258 @@ "issue:45405", "issue:45479", "issue:45538", - "issue:45588", - "issue:45600" + "issue:45588" ], "soft_pairs": [ + "issue:44987|issue:45538", + "issue:44479|issue:45381", + "issue:41720|issue:43540", + "issue:43577|issue:45072", + "issue:43452|issue:43582", + "issue:43540|issue:44560", + "issue:33357|issue:34567", + "issue:43317|issue:45538", "issue:42175|issue:43824", "issue:42175|issue:44938", + "issue:44291|issue:45042", + "issue:43927|issue:44912", + "issue:42915|issue:43742", "issue:42915|issue:45538", - "issue:42915|issue:44821", "issue:44062|issue:44991", + "issue:42915|issue:44821", "issue:43540|issue:44479", "issue:43526|issue:45479", "issue:42175|issue:44488", - "issue:45137|issue:45405", "issue:43906|issue:45381", "issue:38175|issue:44898", - "issue:44737|issue:45588", + "issue:45137|issue:45405", "issue:43749|issue:44164", - "issue:28282|issue:34689", + "issue:44737|issue:45588", "issue:43756|issue:45325", + "issue:28282|issue:34689", "issue:36010|issue:42175", + "issue:43643|issue:45375", "issue:28282|issue:29127", "issue:43493|issue:43994", - "issue:43582|issue:43782", - "issue:43643|issue:45375", - "issue:43742|issue:44291", - "issue:44062|issue:45081", - "issue:34567|issue:36331", - "issue:43452|issue:44488", - "issue:44484|issue:44704", - "issue:44291|issue:44821", - "issue:44792|issue:44964", - "issue:43643|issue:44561", - "issue:45405|issue:45600", - "issue:43688|issue:43749" + "issue:43742|issue:44291" ], - "prepared_review_unit_hash": "bea8d40233b09927f9399bcd2c2f8a3a0549eb652ba0522462a66f8213c7e545", + "prepared_review_unit_hash": "b9a0d3129b7c4e783b4b63fd7c49fa59213918e74dfbe945a81f3385a38e9c7d", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36831, - "estimated_input_tokens": 9208, - "estimated_eval_tokens": 18672 + "soft_pair_count": 31, + "serialized_chars": 37116, + "estimated_input_tokens": 9279, + "estimated_eval_tokens": 18814 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft-edge pair looks like a true duplicate: the titles share generic terms like import/load/error, but the underlying bugs span unrelated models, trainers, tokenizers, config handling, and runtime backends. No PRs are present in the cluster.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue fits this cluster; it is heterogeneous rather than one duplicate bug thread.", + "summary": "This cluster is still mostly heterogeneous. With only titles/metadata, the Qwen video-related issues and the config-field-loss issues are better treated as theme-level similarities than confirmed duplicates; no soft edge is strong enough to merge conservatively.", + "confidence": 0.36, + "canonical_issue_reason": "issue:44479 is the most representative anchor for the Qwen video-input theme, but the packet does not justify treating it as the same underlying bug as another issue; it should be used as a cautious thematic anchor, not a proven duplicate root.", "canonical_pr_reason": null, - "best_issue_reason": "No clear best issue representative either; the closest items are still about different code paths and should remain separate.", + "best_issue_reason": "issue:44479 is the best global issue representative because it is a concrete, specific regression report with a clear version scope and it sits in the strongest apparent subtheme, but the evidence here is not enough to collapse the subtheme into a confirmed duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42175", - "right": "issue:43824", + "left": "issue:44987", + "right": "issue:45538", "accept": false, - "reason": "Both are import/backend failures, but one is TensorFlow packaging/backend presence and the other is a missing Qwen2.5-VL symbol; different bugs." + "reason": "Different bugs: a model-loading failure for one model vs a tokenizer max-length complaint for CLIPTokenizer." }, { - "left": "issue:42175", - "right": "issue:44938", + "left": "issue:44479", + "right": "issue:45381", "accept": false, - "reason": "TensorFlow backend packaging issue vs Python 3.14 load failure; no shared code-path bug." + "reason": "Both are Qwen video-related regressions, but the packet only gives titles and does not establish the same underlying bug or a shared fix path." }, { - "left": "issue:42915", - "right": "issue:45538", + "left": "issue:41720", + "right": "issue:43540", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and CLIPTokenizer model_max_length behavior are unrelated." + "reason": "Both involve Qwen and runtime failures, but the symptoms and code paths differ materially: device-mapping cuda assert vs video-input ValueError." }, { - "left": "issue:42915", - "right": "issue:44821", + "left": "issue:43577", + "right": "issue:45072", "accept": false, - "reason": "FP8 config bug for Qwen3Moe vs AutoImageProcessor URL loading; different subsystems." + "reason": "Both mention dtype problems, but they affect different models and behaviors; not the same concrete bug." }, { - "left": "issue:44062", - "right": "issue:44991", + "left": "issue:43452", + "right": "issue:43582", "accept": false, - "reason": "AddedToken keyword collision and tokenizer loading regression for est-roberta are distinct failures." + "reason": "Tokenizer/model loading for gguf_file is unrelated to the Apple Silicon allocator warmup TypeError." }, { "left": "issue:43540", - "right": "issue:44479", + "right": "issue:44560", "accept": false, - "reason": "Both involve video inputs, but the affected models and failure modes differ (Qwen3OmniMoe TypeError vs Qwen VL position-id regression)." + "reason": "Both are video-input issues, but the titles indicate different Qwen model families and different failures, so this is not enough to merge." }, { - "left": "issue:43526", - "right": "issue:45479", + "left": "issue:33357", + "right": "issue:34567", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to sequence-classification zero-loss behavior." + "reason": "Completely different problems: MacOS CLIP bus error vs TrainerState token counter not updating." }, { - "left": "issue:42175", - "right": "issue:44488", + "left": "issue:43317", + "right": "issue:45538", "accept": false, - "reason": "Backend install/dependency issue vs a specific model loading failure; not the same bug." + "reason": "Dequantized model offload/load failure is unrelated to CLIPTokenizer max-length behavior." }, { - "left": "issue:45137", - "right": "issue:45405", + "left": "issue:42175", + "right": "issue:43824", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error and PEFT version bump are unrelated." + "reason": "Packaging/backend availability issue vs missing Qwen class import; different subsystems and failure modes." }, { - "left": "issue:43906", - "right": "issue:45381", + "left": "issue:42175", + "right": "issue:44938", "accept": false, - "reason": "Isolated reproduction of one earlier issue vs a Qwen2.5-VL video position_ids regression; not the same defect." + "reason": "Packaging/backend availability issue vs Python 3.14 load failure; not the same bug." }, { - "left": "issue:38175", - "right": "issue:44898", + "left": "issue:44291", + "right": "issue:45042", "accept": false, - "reason": "SigLIP2 zero-probabilities behavior and Perceiver non-default-resolution failure are different model bugs." + "reason": "Different code paths: init_empty_weights argument handling vs PIL image processor torchvision dependency regression." }, { - "left": "issue:44737", - "right": "issue:45588", + "left": "issue:43927", + "right": "issue:44912", "accept": false, - "reason": "XLNet CPU arange bug and flash_attention s_aux=None crash are unrelated code paths." + "reason": "Config save/load token-ID loss is unrelated to MXFP4 quantization falling back to bf16." }, { - "left": "issue:43749", - "right": "issue:44164", + "left": "issue:42915", + "right": "issue:43742", "accept": false, - "reason": "FSDP CPU RAM efficient loading and save/from_pretrained extra_state handling are distinct loading/saving issues." + "reason": "Qwen3Moe FP8 loading failure vs MobileLLM key error; unrelated model-loading bugs." }, { - "left": "issue:28282", - "right": "issue:34689", + "left": "issue:42915", + "right": "issue:45538", "accept": false, - "reason": "Missing PyTorch import error for AutoModel and Llama 3.2 vision loading regression are different problems." + "reason": "Different areas entirely: Qwen3Moe FP8 loading vs CLIPTokenizer max-length." }, { - "left": "issue:43756", - "right": "issue:45325", + "left": "issue:44062", + "right": "issue:44991", "accept": false, - "reason": "Smollm3 RoPE layer mismatch and Qwen2.5-VL still-image temporal scaling are unrelated." + "reason": "Both are tokenizer-loading failures, but the concrete errors and affected models differ too much to treat as one bug." }, { - "left": "issue:36010", - "right": "issue:42175", + "left": "issue:42915", + "right": "issue:44821", "accept": false, - "reason": "GenerationMixin import failure vs TensorFlow backend missing from a torch install extra; no common bug." + "reason": "FP8 config failure vs AutoImageProcessor URL loading; unrelated." }, { - "left": "issue:28282", - "right": "issue:29127", + "left": "issue:43540", + "right": "issue:44479", "accept": false, - "reason": "Generic PyTorch import error and LayoutLMv3 box-content error are unrelated." + "reason": "Related Qwen/video theme, but different model families and different error types; not enough evidence for the same underlying bug." }, { - "left": "issue:43493", - "right": "issue:43994", + "left": "issue:43526", + "right": "issue:45479", "accept": false, - "reason": "Both concern SigLIP2, but one is JAX/HF parity and the other is bad outputs through AutoModel/pipeline; insufficient evidence of the same underlying bug." + "reason": "ImageProcessorFast label reduction bug vs sequence-classification zero-loss bug; different functionality." }, { - "left": "issue:43582", - "right": "issue:43782", + "left": "issue:42175", + "right": "issue:44488", "accept": false, - "reason": "Apple Silicon TypeError in caching_allocator_warmup and Qwen3VL weight_only=True loading error are unrelated." + "reason": "Packaging/backend availability issue vs a specific model not loading; not the same root cause." }, { - "left": "issue:43643", - "right": "issue:45375", + "left": "issue:43906", + "right": "issue:45381", "accept": false, - "reason": "Both involve config fields and trust_remote_code-style loading, but they target different models and different missing-field causes." + "reason": "An isolated reproduction reference vs a specific Qwen2.5-VL video-position regression; no direct match." }, { - "left": "issue:43742", - "right": "issue:44291", + "left": "issue:38175", + "right": "issue:44898", "accept": false, - "reason": "MobileLLM-125M key error and init_empty_weights unexpected argument are different loading failures." + "reason": "SigLIP2 zero-probabilities vs Perceiver interpolation failure are different model bugs." }, { - "left": "issue:44062", - "right": "issue:45081", + "left": "issue:45137", + "right": "issue:45405", "accept": false, - "reason": "AddedToken special-argument collision and Mistral regex patch backend_tokenizer AttributeError are unrelated." + "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT version bump." }, { - "left": "issue:34567", - "right": "issue:36331", + "left": "issue:43749", + "right": "issue:44164", "accept": false, - "reason": "TrainerState token counter not updating and CustomTrainer.compute_loss signature mismatch are different trainer API issues." + "reason": "FSDP CPU RAM efficient loading regression vs save/from_pretrained extra_state handling; different code paths." }, { - "left": "issue:43452", - "right": "issue:44488", + "left": "issue:44737", + "right": "issue:45588", "accept": false, - "reason": "GGUF file loading breakage and cjvt/sleng-bert model loading failure are unrelated." + "reason": "XLNet CPU placement bug vs flash-attention crash on sink-less models; unrelated." }, { - "left": "issue:44484", - "right": "issue:44704", + "left": "issue:43756", + "right": "issue:45325", "accept": false, - "reason": "save_pretrained shard-size question and AutoProcessor kwargs propagation bug do not describe the same change or bug." + "reason": "Different rope/position-id problems in different models; not the same underlying bug." }, { - "left": "issue:44291", - "right": "issue:44821", + "left": "issue:28282", + "right": "issue:34689", "accept": false, - "reason": "init_empty_weights loading error and AutoImageProcessor URL loading are unrelated." + "reason": "PyTorch-missing ImportError vs Llama 3.2 Vision loading regression; unrelated." }, { - "left": "issue:44792", - "right": "issue:44964", + "left": "issue:36010", + "right": "issue:42175", "accept": false, - "reason": "Janus test failure and Phi-4 multimodal load failure are different issues." + "reason": "GenerationMixin import failure is unrelated to the tensorflow backend packaging issue." }, { "left": "issue:43643", - "right": "issue:44561", + "right": "issue:45375", "accept": false, - "reason": "Both touch trust_remote_code, but one is missing returned config fields and the other is removal of is_torch_fx_available; different root causes." + "reason": "Both concern config fields, but the packet does not establish the same preservation bug; one is trust_remote_code missing fields, the other is a specific config field dropped by strict parsing." }, { - "left": "issue:45405", - "right": "issue:45600", + "left": "issue:28282", + "right": "issue:29127", "accept": false, - "reason": "PEFT version pinning and stale auto_mappings references to removed configs are unrelated." + "reason": "ImportError due to missing PyTorch vs LayoutLMv3 error-message clarity; unrelated." }, { - "left": "issue:43688", - "right": "issue:43749", + "left": "issue:43493", + "right": "issue:43994", "accept": false, - "reason": "Auxiliary-loss normalization and FSDP CPU RAM efficient loading are different parts of the stack." + "reason": "Both are SigLIP2-related, but the packet does not prove the same underlying defect from titles alone." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The summary and verdicts are a bit too strong for the packet. In particular, the accepted pair issue:44479 \u2194 issue:45381 is not grounded enough: the packet only gives titles, and while both are Qwen video-input regressions, it does not justify claiming they are the same bug or that they \u201ccould be fixed together.\u201d That makes the canonical anchor/duplicate-subcluster claim too assertive. The 43643 \u2194 45375 accept is also somewhat speculative, but the qwen accept is enough to reject this as not conservative enough." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -5600,47 +5432,45 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:24643", - "issue:28282", - "issue:29127", - "issue:32090", - "issue:38175", - "issue:41720", + "issue:34567", + "issue:36331", "issue:41762", "issue:43295", "issue:43452", "issue:43525", "issue:43531", "issue:43577", + "issue:43582", "issue:43643", - "issue:43646", "issue:43653", "issue:43688", "issue:43697", "issue:43720", "issue:43742", "issue:43749", + "issue:43782", "issue:43819", - "issue:43906", - "issue:44361", - "issue:44442", - "issue:44464", + "issue:44062", + "issue:44291", "issue:44466", - "issue:44479", + "issue:44484", + "issue:44488", "issue:44514", + "issue:44561", "issue:44568", "issue:44589", "issue:44625", - "issue:44743", + "issue:44704", + "issue:44792", "issue:44811", "issue:44821", "issue:44843", "issue:44877", "issue:44898", "issue:44936", - "issue:44991", + "issue:44964", "issue:45005", "issue:45042", "issue:45081", @@ -5648,249 +5478,245 @@ "issue:45290", "issue:45356", "issue:45381", + "issue:45405", "issue:45440", - "issue:45701" + "issue:45600" ], "soft_pairs": [ + "issue:43582|issue:43782", + "issue:44062|issue:45081", + "issue:34567|issue:36331", + "issue:43452|issue:44488", + "issue:44484|issue:44704", + "issue:44291|issue:44821", + "issue:44792|issue:44964", + "issue:43643|issue:44561", + "issue:45405|issue:45600", "issue:43697|issue:45216", + "issue:43688|issue:43749", "issue:45042|issue:45081", "issue:44514|issue:45381", "issue:43742|issue:44589", "issue:43531|issue:44821", "issue:44514|issue:45290", + "issue:43452|issue:44843", "issue:43452|issue:43577", "issue:44625|issue:44877", - "issue:43452|issue:44843", "issue:43688|issue:45440", - "issue:43720|issue:44898", "issue:41762|issue:44936", "issue:43653|issue:44568", "issue:43295|issue:45356", + "issue:43720|issue:44898", "issue:43525|issue:43643", "issue:43819|issue:44811", - "issue:44466|issue:45005", - "issue:24643|issue:29127", - "issue:44464|issue:44743", - "issue:41720|issue:43906", - "issue:38175|issue:43295", - "issue:43577|issue:43697", - "issue:44843|issue:44991", - "issue:43646|issue:43749", - "issue:43688|issue:45701", - "issue:44361|issue:44442", - "issue:43653|issue:44361", - "issue:28282|issue:32090", - "issue:44479|issue:44936" + "issue:44466|issue:45005" ], - "prepared_review_unit_hash": "3574c585373c677bfda8be1b94aa152f611f8441caa32b3f31030dd426a22458", + "prepared_review_unit_hash": "fc6e53ca6d87b7252597aeaa4d6fac1ae11c452984b95774da4542a30c206606", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36701, - "estimated_input_tokens": 9176, - "estimated_eval_tokens": 18608 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35532, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a grab bag of unrelated regression reports. There are a few loose theme overlaps (tokenizer/loading, chat templates, tied weights), but none of the soft-edge pairs look like the same concrete bug/change, so all are rejected.", - "confidence": 0.88, - "canonical_issue_reason": "issue:44466 is the most concrete, reproducible report in the set, centered on a specific tied-weights serialization regression with a clear symptom and version boundary.", + "summary": "These issues are mostly unrelated despite broad overlap around model loading, tokenizers, and training APIs. The only strong duplicate-looking pair is the v5 tied-weights serialization regression.", + "confidence": 0.86, + "canonical_issue_reason": "Use issue 44466 as the canonical issue: it states the broader tied-weights/lm_head serialization regression clearly, while 45005 is a narrower translation-model manifestation of the same problem.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45005 is the broadest tied-weights regression report and the best umbrella target among the listed issues, even though the overall cluster is mixed.", + "best_issue_reason": "Issue 44466 is the best representative because it describes the underlying bug in a general, reproducible way and covers the broader failure mode.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43582", + "right": "issue:43782", + "accept": false, + "reason": "Different failures: Apple Silicon allocator warmup TypeError vs Qwen3-VL weight_only loading error. No shared code path." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Both touch tokenizers, but one is an AddedToken constructor bug and the other is a Mistral regex/backend_tokenizer crash. Different root causes." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "Trainer state counter not updating and custom compute_loss signature break are separate trainer API issues." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading problems, but one is about gguf_file handling and the other is a specific model/config load failure. Not the same bug." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "save_pretrained shard-size behavior and AutoProcessor cached_file kwargs forwarding are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized argument issue is unrelated to AutoImageProcessor URL loading." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal model failures with different entry points; not enough evidence of one shared bug." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "One is missing fields in AutoConfig with trust_remote_code; the other is a missing helper in v5 breaking remote-code models. Different fixes." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "Dependency-version bump and removed-config auto_mappings CI failure are unrelated maintenance issues." + }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "Different models and failure modes: RTDetrV2 output drift vs Qwen3.5 save_pretrained checkpoint regression." + "reason": "RTDetr output drift and Qwen3.5 save_pretrained checkpoint corruption are different regressions." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Auxiliary-loss normalization and FSDP CPU RAM-efficient loading breakage are different training-path bugs." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "Unrelated code paths: PIL image processors requiring torchvision vs _patch_mistral_regex tokenizer crash." + "reason": "Image processor torchvision dependency regression and Mistral regex crash are unrelated subsystems." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Both touch Qwen2.5-VL/video paths, but one is a batched apply_chat_template crash and the other is wrong vision_position_ids." + "reason": "Chat-template batching crash and Qwen2.5-VL video position-id bug are different problems." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "A key error while loading MobileLLM is not the same bug as the Float8 storage lookup failure." + "reason": "Key error when loading a model and missing Float8 storage lookup are different deserialization/load failures." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor URL loading are different subsystems and symptoms." + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to loading AutoImageProcessor from URL." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Same API surface, but different triggers: batched padding=False crash vs assistant messages with tool calls and empty content." + "reason": "Both involve apply_chat_template, but one is batching/padding and the other is assistant tool-call messages with empty content. Not clearly the same underlying bug." }, { "left": "issue:43452", - "right": "issue:43577", + "right": "issue:44843", "accept": false, - "reason": "gguf_file tokenizer/model loading breakage is unrelated to BLIP2 dtype propagation staying float32." + "reason": "gguf_file loading regression and offline model_info() call in _patch_mistral_regex are separate tokenizer-loading issues." }, { - "left": "issue:44625", - "right": "issue:44877", + "left": "issue:43452", + "right": "issue:43577", "accept": false, - "reason": "num_labels propagation in Qwen3.5 is not the same as strict config rejection for granite_speech." + "reason": "gguf_file loading breakage and Blip2 dtype propagation are unrelated." }, { - "left": "issue:43452", - "right": "issue:44843", + "left": "issue:44625", + "right": "issue:44877", "accept": false, - "reason": "gguf_file loading failure and offline mistral regex hub access are different code paths." + "reason": "Qwen3.5 num_labels propagation and granite_speech strict-config loading are different config issues." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Aux-loss normalization is a training-loss bug; DeepseekV3MoE divergence is a model-implementation mismatch." - }, - { - "left": "issue:43720", - "right": "issue:44898", - "accept": false, - "reason": "BitNet packed-weight loading and Perceiver non-default-resolution failure are unrelated." + "reason": "Aux-loss normalization bug and DeepSeekV3MoE divergence from remote implementation are not the same concrete defect." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "DeepSpeed ZeRO-3 loading crash and trainer.train()/evaluate() lifecycle failure are different problems." + "reason": "Gemma3 ZeRO-3 load IndexError and trainer.evaluate() after train() failure are unrelated." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "BigBird special-token registration bug and mdeberta add_special_tokens regression affect different tokenizers and behavior." + "reason": "Both concern special tokens, but one is BigBird mask-token registration and the other is mdeberta add_special_tokens behavior. Different tokenizer-specific bugs." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Processor/tokenizer regression in v4.57.5 is not the same as the Kimi-K2.5 codec handling and warning regression." + "reason": "Processor.tokenizer/images regression and Kimi-K2.5 codec/fix_mistral_regex warning regression are different loading/tokenizer problems." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific regressions." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Missing pad_token_id on Llama4Config and missing fields from trust_remote_code AutoConfig are different config-loading issues." + "reason": "Missing pad_token_id on a config and missing fields from trust_remote_code AutoConfig are distinct config-shape problems." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC.from_latents mismatch and Whisper batch_decode skip_special_tokens handling are unrelated behaviors." + "reason": "DAC.from_latents vs forward mismatch and Whisper batch_decode skip_special_tokens behavior are unrelated." }, { "left": "issue:44466", "right": "issue:45005", - "accept": false, - "reason": "Both mention tied weights, but 44466 is a specific device-dependent serialization bug while 45005 is a broad translation-model report; not enough evidence of one concrete duplicate." - }, - { - "left": "issue:24643", - "right": "issue:29127", - "accept": false, - "reason": "DeepSpeed runtime weight-shape error and LayoutLMv3 error-message clarity are not the same underlying bug." - }, - { - "left": "issue:44464", - "right": "issue:44743", - "accept": false, - "reason": "Chunked generation inconsistencies with compiled forward and recurrent-state resets under cache are distinct model-execution bugs." - }, - { - "left": "issue:41720", - "right": "issue:43906", - "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and an isolated reproduction of another issue are not the same concrete failure." + "accept": true, + "reason": "Both describe the same v5 tied-weights serialization bug around lm_head.weight/save_pretrained; 45005 is just a translation-model-specific instance." }, { - "left": "issue:38175", - "right": "issue:43295", + "left": "issue:43589", + "right": "issue:44561", "accept": false, - "reason": "Siglip2 zero probabilities and processor.tokenizer regression are unrelated." - }, - { - "left": "issue:43577", - "right": "issue:43697", - "accept": false, - "reason": "BLIP2 dtype remaining float32 is unrelated to RTDetrV2 output differences." - }, - { - "left": "issue:44843", - "right": "issue:44991", - "accept": false, - "reason": "Offline model_info() access in _patch_mistral_regex is not the same as failing to load a specific tokenizer model." - }, - { - "left": "issue:43646", - "right": "issue:43749", - "accept": false, - "reason": "Custom model initialization breakage in Transformers 5.0 and FSDP CPU RAM efficient loading are different regression surfaces." - }, - { - "left": "issue:43688", - "right": "issue:45701", - "accept": false, - "reason": "Auxiliary-loss normalization is unrelated to a general tokenization version-change report." - }, - { - "left": "issue:44361", - "right": "issue:44442", - "accept": false, - "reason": "Both are tokenizer-related, but MLuke task AttributeError and FastSpeech2Conformer tokenizer loading failure are different bugs." - }, - { - "left": "issue:43653", - "right": "issue:44361", - "accept": false, - "reason": "BigBirdTokenizer special-token registration and MLuke task AttributeError do not share the same code-path problem." - }, - { - "left": "issue:28282", - "right": "issue:32090", - "accept": false, - "reason": "Missing PyTorch ImportError and unsupported NoneType broadcast in Trainer are unrelated errors." - }, - { - "left": "issue:44479", - "right": "issue:44936", - "accept": false, - "reason": "Qwen VL video-input regression and trainer.evaluate() after train are different issues in different layers." + "reason": "Not enough relation; Float8 storage lookup failure is unrelated to the removal of is_torch_fx_available." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is well grounded in the packet and stays mostly conservative. The only accepted pair, 44466/45005, is plausibly the same tied-weights serialization regression. The rejected pairs are generally framed cautiously as distinct issues without overstating shared evidence. The summary matches the evidence available." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -5905,303 +5731,318 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ + "issue:24643", "issue:28282", + "issue:29127", "issue:30064", "issue:32090", "issue:33357", "issue:33666", "issue:34567", "issue:35141", - "issue:36010", + "issue:38175", "issue:41720", "issue:43122", - "issue:43262", - "issue:43278", + "issue:43295", "issue:43329", "issue:43334", "issue:43452", - "issue:43493", "issue:43504", - "issue:43525", "issue:43531", - "issue:43540", "issue:43550", "issue:43572", "issue:43577", + "issue:43646", "issue:43653", "issue:43688", - "issue:43716", + "issue:43697", "issue:43720", "issue:43742", - "issue:43756", - "issue:43825", + "issue:43749", "issue:43906", "issue:43927", "issue:44361", + "issue:44442", "issue:44451", "issue:44464", "issue:44479", "issue:44554", "issue:44610", "issue:44625", - "issue:44821", - "issue:44849", - "issue:44913", + "issue:44743", + "issue:44843", + "issue:44936", "issue:44977", - "issue:45084", + "issue:44991", "issue:45325", "issue:45356", - "issue:45357", - "issue:45588" + "issue:45588", + "issue:45701", + "issue:45704" ], "soft_pairs": [ - "issue:28282|issue:33357", - "issue:28282|issue:35141", + "issue:24643|issue:29127", + "issue:44464|issue:44743", + "issue:41720|issue:43906", + "issue:43577|issue:43697", + "issue:38175|issue:43295", + "issue:43688|issue:45704", + "issue:44843|issue:44991", + "issue:43646|issue:43749", + "issue:44361|issue:44442", + "issue:43688|issue:45701", + "issue:43653|issue:44361", + "issue:44479|issue:44936", + "issue:28282|issue:32090", "issue:44977|issue:45356", + "issue:28282|issue:33357", "issue:30064|issue:33666", + "issue:28282|issue:35141", "issue:43122|issue:43906", "issue:43577|issue:45588", "issue:43577|issue:44554", "issue:43653|issue:43927", - "issue:43329|issue:45325", "issue:43531|issue:43742", + "issue:43329|issue:45325", "issue:44464|issue:44977", "issue:44479|issue:44625", - "issue:43550|issue:44361", "issue:43452|issue:44451", - "issue:32090|issue:34567", + "issue:43550|issue:44361", "issue:43504|issue:43720", "issue:43504|issue:43550", - "issue:32090|issue:33666", + "issue:32090|issue:34567", "issue:43504|issue:44610", - "issue:43278|issue:43688", - "issue:41720|issue:44464", "issue:43334|issue:43572", - "issue:43525|issue:43572", - "issue:43493|issue:43825", - "issue:43756|issue:44913", - "issue:44849|issue:45357", - "issue:43742|issue:44821", - "issue:43262|issue:45084", - "issue:33666|issue:36010", - "issue:43540|issue:43716" + "issue:32090|issue:33666" ], - "prepared_review_unit_hash": "994aaaf004583af9310d268e60be0e86afc62c3d887acdbcd996a33ab4901879", + "prepared_review_unit_hash": "200cd44a223ffee01aff40799117fdc4c3e680322467224adcbf40c34acdcdfb", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36825, - "estimated_input_tokens": 9207, - "estimated_eval_tokens": 18670 + "soft_pair_count": 33, + "serialized_chars": 37743, + "estimated_input_tokens": 9436, + "estimated_eval_tokens": 19128 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft pairs mostly share broad wording (loading, config, generation, multimodal, tokenizer) but not the same concrete code-path bug. I reject all soft-edge merges. The best issue anchor is the pretrained-loading regression in issue 43504 because it is concrete and one of the few nodes with multiple soft links.", - "confidence": 0.28, - "canonical_issue_reason": "Issue 43504 is the most central concrete bug here: a specific pretrained-model loading regression, and it connects to several other loading-related soft candidates, making it the best cluster anchor even though the overall cluster is mixed.", + "summary": "The set is highly heterogeneous: most items are unrelated model-, tokenizer-, training-, or version-specific regressions. There is no strong duplicate cluster here, so the best we can do is pick a broad regression ticket as a loose representative.", + "confidence": 0.18, + "canonical_issue_reason": "issue:43646 is the broadest umbrella in the group: a Transformers 5.0 initialization regression that matches the general version-breakage theme better than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "43504 is the strongest representative of the set because it is a narrowly described functional failure rather than a broad symptom or version-only report.", + "best_issue_reason": "issue:43646 is the most suitable anchor only because it is a concrete, broadly worded regression ticket; it is still not a true duplicate of most other issues in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:28282", - "right": "issue:33357", + "left": "issue:24643", + "right": "issue:29127", "accept": false, - "reason": "Different failures: missing PyTorch import vs MacOS bus error with a community CLIP model." + "reason": "Different bugs: DeepSpeed training weight shape error vs. LayoutLMv3 box-content validation/error messaging." }, { - "left": "issue:28282", - "right": "issue:35141", + "left": "issue:44464", + "right": "issue:44743", "accept": false, - "reason": "Unrelated bugs: import-time backend requirement vs embedding reinitialization after resize." + "reason": "Both touch generation, but one is chunked/compiled-forward inconsistency and the other is Qwen3.5 cache state reset; not the same concrete bug." }, { - "left": "issue:44977", - "right": "issue:45356", + "left": "issue:41720", + "right": "issue:43906", "accept": false, - "reason": "Both are tokenizer/model regressions, but they affect different models and different code paths (flash-attention generation vs codec handling)." + "reason": "Unrelated failures: Qwen3 auto device mapping cuda assert vs. an isolated reproduction of a different issue." }, { - "left": "issue:30064", - "right": "issue:33666", + "left": "issue:43577", + "right": "issue:43697", "accept": false, - "reason": "Image processor segmentation-map handling is unrelated to Qwen2-VL multi-GPU training." + "reason": "Different model paths and symptoms: Blip2 dtype loading bug vs. RTDetrV2 output divergence in v5." }, { - "left": "issue:43122", - "right": "issue:43906", + "left": "issue:38175", + "right": "issue:43295", "accept": false, - "reason": "A tokenizer output change across versions is not the same as an isolated reproduction of a separate issue without clearer linkage." + "reason": "Different regressions: SigLIP2 zero probabilities vs. processor/tokenizer regression for custom model code." }, { - "left": "issue:43577", - "right": "issue:45588", + "left": "issue:43688", + "right": "issue:45704", "accept": false, - "reason": "Different subsystems: dtype propagation in Blip2 loading vs flash_attention handling of s_aux=None." + "reason": "Different subsystems entirely: auxiliary loss normalization vs. T5 Apex RMSNorm memory leak." }, { - "left": "issue:43577", - "right": "issue:44554", + "left": "issue:44843", + "right": "issue:44991", "accept": false, - "reason": "Different bug classes: model dtype loading vs MPS attention correctness with mismatched head dimensions." + "reason": "Both are tokenizer-loading regressions, but one is an offline model_info call in regex patching and the other is a specific tokenizer loading failure; root causes differ." }, { - "left": "issue:43653", - "right": "issue:43927", + "left": "issue:43646", + "right": "issue:43749", "accept": false, - "reason": "Tokenizer special-token registration bug is not the same as config custom-token-ID loss causing generation IndexError." + "reason": "Different issues: custom model initialization breakage vs. FSDP CPU RAM efficient loading regression." }, { - "left": "issue:43329", - "right": "issue:45325", + "left": "issue:44361", + "right": "issue:44442", "accept": false, - "reason": "Video token counting bug is separate from Qwen2.5-VL rope-index temporal-position scaling." + "reason": "Both are tokenizer/model loading failures, but they affect different tokenizer classes and failure modes." }, { - "left": "issue:43531", - "right": "issue:43742", + "left": "issue:43688", + "right": "issue:45701", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to a MobileLLM-125M load-time KeyError." + "reason": "One is auxiliary-loss normalization; the other is a broad tokenization change report. Not the same bug." }, { - "left": "issue:44464", - "right": "issue:44977", + "left": "issue:43653", + "right": "issue:44361", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is a different path from Qwen3.5 flash-attention generation failure." + "reason": "Different tokenizer defects: BigBird special-token registration vs. MLuke tokenizer AttributeError." }, { "left": "issue:44479", - "right": "issue:44625", + "right": "issue:44936", "accept": false, - "reason": "Video-input regression in multiple Qwen models is not the same as num_labels propagation in Qwen3.5 config." + "reason": "Video-input regression for Qwen VL models vs. trainer evaluate-after-train failure; unrelated code paths." }, { - "left": "issue:43550", - "right": "issue:44361", + "left": "issue:28282", + "right": "issue:32090", "accept": false, - "reason": "Model compile/SDPA failure is unrelated to tokenizer task AttributeError." + "reason": "Different layers: missing PyTorch import error vs. Trainer broadcast TypeError with NoneType." }, { - "left": "issue:43452", - "right": "issue:44451", + "left": "issue:44977", + "right": "issue:45356", "accept": false, - "reason": "Both are loading failures, but one is gguf/Auto* handling and the other is a specific model load failure; not the same bug." + "reason": "Different regressions: flash-attention generation issue vs. tokenizer codec/regression and warning behavior." }, { - "left": "issue:32090", - "right": "issue:34567", + "left": "issue:28282", + "right": "issue:33357", "accept": false, - "reason": "Trainer broadcast NoneType error is unrelated to TrainerState token-count tracking." + "reason": "Unrelated: environment import error vs. MacOS bus error with a community CLIP model." }, { - "left": "issue:43504", - "right": "issue:43720", + "left": "issue:30064", + "right": "issue:33666", "accept": false, - "reason": "Both involve loading models, but one is a legacy-field preset issue and the other is packed-weight unpacking during accelerate loading." + "reason": "Different areas: void segmentation map processing vs. Qwen2-VL multi-GPU training." }, { - "left": "issue:43504", - "right": "issue:43550", + "left": "issue:28282", + "right": "issue:35141", "accept": false, - "reason": "Different concrete problems: pretrained preset loading vs torch.compile with SDPA." + "reason": "Different bugs: missing PyTorch dependency vs. output embedding reinitialization after resizing embeddings." }, { - "left": "issue:32090", - "right": "issue:33666", + "left": "issue:43122", + "right": "issue:43906", "accept": false, - "reason": "Trainer distributed broadcast error is unrelated to Qwen2-VL multi-GPU training." + "reason": "Tokenization/version-change report vs. isolated reproduction of another issue; not the same defect." }, { - "left": "issue:43504", - "right": "issue:44610", + "left": "issue:43577", + "right": "issue:45588", "accept": false, - "reason": "Model preset loading failure is not the same as processor/model input-size mismatch." + "reason": "Different bugs: dtype not propagated in Blip2 vs. flash_attention crash on s_aux=None." }, { - "left": "issue:43278", - "right": "issue:43688", + "left": "issue:43577", + "right": "issue:44554", "accept": false, - "reason": "Embedding dtype mismatch is unrelated to auxiliary-loss normalization." + "reason": "Different issues: Blip2 dtype loading vs. MPS attention correctness when head dims differ." }, { - "left": "issue:41720", - "right": "issue:44464", + "left": "issue:43653", + "right": "issue:43927", "accept": false, - "reason": "Auto device mapping CUDA assert is a different failure mode from compiled-forward chunked generation inconsistency." + "reason": "Different bugs: BigBird special-token handling vs. DiaConfig custom token IDs lost on save/load." }, { - "left": "issue:43334", - "right": "issue:43572", + "left": "issue:43531", + "right": "issue:43742", "accept": false, - "reason": "Both mention missing pad token fields, but they are different model configs and likely separate regressions." + "reason": "Different failures: Qwen3-MoE sliding_window handling vs. MobileLLM-125M key error on load." }, { - "left": "issue:43525", - "right": "issue:43572", + "left": "issue:43329", + "right": "issue:45325", "accept": false, - "reason": "Same symptom family, but different config classes and model families; not enough to treat as one bug." + "reason": "Both involve multimodal/video code, but one is undefined helper usage in token counting and the other is temporal position scaling in rope index." }, { - "left": "issue:43493", - "right": "issue:43825", + "left": "issue:44464", + "right": "issue:44977", "accept": false, - "reason": "SigLIP2 implementation discrepancy is unrelated to the pipeline translation-support error message." + "reason": "Different generation regressions: compiled-forward chunking vs. flash-attention behavior in Qwen3.5." }, { - "left": "issue:43756", - "right": "issue:44913", + "left": "issue:44479", + "right": "issue:44625", "accept": false, - "reason": "RoPE-layer dropping in Smollm3 is a different issue from rotary_pct not persisting on reload." + "reason": "Different bugs: video regression in multiple VLMs vs. num_labels propagation in Qwen3.5 config." }, { - "left": "issue:44849", - "right": "issue:45357", + "left": "issue:43452", + "right": "issue:44451", "accept": false, - "reason": "Both are Qwen3.5 issues, but one concerns hidden states and the other save_pretrained visual encoder keys; different code paths." + "reason": "Different issues: gguf_file breaks two from_pretrained entry points vs. a specific tokenizer load failure for a model repo." }, { - "left": "issue:43742", - "right": "issue:44821", + "left": "issue:43550", + "right": "issue:44361", "accept": false, - "reason": "A MobileLLM loading KeyError is unrelated to loading AutoImageProcessor from a URL." + "reason": "Different code paths: torch.compile/SDPA failure vs. tokenizer AttributeError." }, { - "left": "issue:43262", - "right": "issue:45084", + "left": "issue:43504", + "right": "issue:43720", "accept": false, - "reason": "Audio chat-template defaults are unrelated to template-node compilation errors." + "reason": "Different bugs: legacy field during Beit loading vs. unpacking packed weights during accelerate loading." }, { - "left": "issue:33666", - "right": "issue:36010", + "left": "issue:43504", + "right": "issue:43550", "accept": false, - "reason": "Multi-GPU training for Qwen2-VL is unrelated to the GenerationMixin import error." + "reason": "Different model-loading and runtime-compile bugs; no shared concrete code path." }, { - "left": "issue:43540", - "right": "issue:43716", + "left": "issue:32090", + "right": "issue:34567", "accept": false, - "reason": "Video-processing ValueError in Qwen3OmniMoe is unrelated to dtype mismatch between image preprocessor and model." + "reason": "Different Trainer problems: NoneType broadcast failure vs. num_input_tokens_seen not updating." }, { - "left": "issue:44554", - "right": "issue:44625", + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different bugs: Beit legacy field load failure vs. OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both are config-field regressions, but they affect different models and different missing fields." + }, + { + "left": "issue:32090", + "right": "issue:33666", "accept": false, - "reason": "MPS attention correctness and Qwen3.5 num_labels propagation are different bugs with no shared code path." + "reason": "Different training issues: broadcast TypeError vs. Qwen2-VL multi-GPU training." } ] }, @@ -6223,299 +6064,297 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:33357", "issue:33666", + "issue:36010", "issue:36331", "issue:39692", + "issue:41720", "issue:43257", + "issue:43262", "issue:43278", "issue:43295", "issue:43299", "issue:43493", - "issue:43504", + "issue:43525", "issue:43540", "issue:43572", - "issue:43575", "issue:43606", "issue:43646", - "issue:43653", "issue:43688", - "issue:43701", + "issue:43716", "issue:43720", + "issue:43742", "issue:43756", "issue:43824", + "issue:43825", "issue:43844", - "issue:43873", "issue:43927", - "issue:43994", - "issue:44112", - "issue:44291", "issue:44361", "issue:44423", "issue:44464", "issue:44466", "issue:44479", "issue:44488", - "issue:44493", - "issue:44521", "issue:44589", "issue:44625", "issue:44734", - "issue:44792", + "issue:44821", "issue:44843", + "issue:44849", + "issue:44913", "issue:44964", "issue:44977", + "issue:45084", "issue:45200", "issue:45229", - "issue:45290", + "issue:45357", "issue:45405", "issue:45520", "issue:45584" ], "soft_pairs": [ - "issue:43688|issue:43873", + "issue:43525|issue:43572", + "issue:43278|issue:43688", + "issue:41720|issue:44464", + "issue:44849|issue:45357", + "issue:43756|issue:44913", + "issue:43493|issue:43825", + "issue:43742|issue:44821", + "issue:43540|issue:43716", + "issue:43262|issue:45084", + "issue:33666|issue:36010", "issue:43720|issue:43927", "issue:43606|issue:43927", "issue:43257|issue:43299", "issue:43278|issue:43844", "issue:44488|issue:44843", - "issue:39692|issue:43493", - "issue:44964|issue:45405", "issue:43493|issue:43844", + "issue:44964|issue:45405", + "issue:39692|issue:43493", "issue:43720|issue:44361", - "issue:33666|issue:36331", "issue:43540|issue:44977", "issue:43756|issue:44479", - "issue:44625|issue:45584", + "issue:33666|issue:36331", "issue:45200|issue:45229", + "issue:44625|issue:45584", "issue:43824|issue:44589", "issue:43646|issue:44466", - "issue:44589|issue:45520", "issue:44423|issue:44734", + "issue:44589|issue:45520", "issue:39692|issue:43295", - "issue:33357|issue:33666", - "issue:43572|issue:44291", - "issue:43504|issue:43606", - "issue:43575|issue:45405", - "issue:43504|issue:43653", - "issue:44521|issue:45290", - "issue:43927|issue:44361", - "issue:43994|issue:44493", - "issue:44792|issue:45405", - "issue:44464|issue:45584", - "issue:43278|issue:43701", - "issue:43606|issue:44112" + "issue:33357|issue:33666" ], - "prepared_review_unit_hash": "d9eed9410e348464f4f63149208d639a77b261dcdf37a94c9d5d161186ec0866", + "prepared_review_unit_hash": "388a2b484695b446f08731c54da1c72e968fac28f580465bfd4c7b2ff9b9af53", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37396, - "estimated_input_tokens": 9349, - "estimated_eval_tokens": 18954 + "soft_pair_count": 30, + "serialized_chars": 36733, + "estimated_input_tokens": 9184, + "estimated_eval_tokens": 18624 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items do not form a single duplicate cluster; they span many unrelated regressions across multimodal loading, trainer APIs, quantization/offload, configs, docs, and CI. I rejected every soft edge because none clearly describe the same underlying bug/change.", - "confidence": 0.98, - "canonical_issue_reason": "No single canonical issue: the set is heterogeneous and does not collapse to one duplicate family.", + "summary": "This cluster is not a duplicate set; it mixes many unrelated issues across model families, configs, training, serving, and serialization. None of the soft edges are strong enough to merge.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue fits the cluster: the items describe distinct bugs and regressions in different subsystems and model families, so there is no clear duplicate anchor.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong global representative because the collection mixes unrelated model families and failure modes.", + "best_issue_reason": "No issue is a good global representative because the cluster is heterogeneous rather than one underlying bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43688", - "right": "issue:43873", + "left": "issue:43525", + "right": "issue:43572", "accept": false, - "reason": "Aux-loss normalization in OLMoE/GPT Oss is unrelated to a quantization/offloading bug." + "reason": "Both involve missing padding-related config fields after v5, but they affect different model classes and different attributes (`pad_token_id` vs `pad_token_idx`)." }, { - "left": "issue:43720", - "right": "issue:43927", + "left": "issue:43278", + "right": "issue:43688", "accept": false, - "reason": "BitNet packed-weight loading/unpacking is a different bug from DiaConfig token-ID persistence." + "reason": "Different failure modes: dtype drift in evaluate vs auxiliary-loss normalization in specific MoE models." }, { - "left": "issue:43606", - "right": "issue:43927", + "left": "issue:41720", + "right": "issue:44464", "accept": false, - "reason": "CPU offload device mismatch for bark-small is unrelated to DiaConfig save/load corruption." + "reason": "Unrelated code paths: auto device mapping CUDA assert versus compiled-forward chunked generation inconsistency." }, { - "left": "issue:43257", - "right": "issue:43299", + "left": "issue:44849", + "right": "issue:45357", "accept": false, - "reason": "Both mention Qwen MoE loading, but they target different model variants and different failure modes." + "reason": "Both are Qwen3.5 regressions, but one is about hidden states output and the other about save_pretrained visual encoder keys." }, { - "left": "issue:43278", - "right": "issue:43844", + "left": "issue:43756", + "right": "issue:44913", "accept": false, - "reason": "Embedding dtype drift between train/eval is not the same as ZeRO-3 gradient inflation." + "reason": "Both mention rotary/RoPE settings, but one is a SmolLM3 architecture issue and the other is a GPTNeoX config reload bug." }, { - "left": "issue:44488", - "right": "issue:44843", + "left": "issue:43493", + "right": "issue:43825", "accept": false, - "reason": "A model load failure for sleng-bert is unrelated to the offline-mode model_info call bug." + "reason": "SigLIP2 implementation mismatch and pipeline translation-task messaging are unrelated bugs." }, { - "left": "issue:39692", - "right": "issue:43493", + "left": "issue:43742", + "right": "issue:44821", "accept": false, - "reason": "Broken SigLIP2 docs example is distinct from a core implementation mismatch with JAX." + "reason": "Different loading failures: a specific model key error versus AutoImageProcessor URL loading support." }, { - "left": "issue:44964", - "right": "issue:45405", + "left": "issue:43540", + "right": "issue:43716", "accept": false, - "reason": "Phi-4 multimodal load regression is unrelated to an unreleased PEFT minimum-version bump." + "reason": "Both are multimodal input issues, but one is video processing in Qwen3OmniMoe and the other is image-preprocessor dtype mismatch in Mistral-3." }, { - "left": "issue:43493", - "right": "issue:43844", + "left": "issue:43262", + "right": "issue:45084", "accept": false, - "reason": "SigLIP2 implementation parity issues are unrelated to ZeRO-3 gradient behavior." + "reason": "Audio chat-template sampling rate behavior is unrelated to template compilation errors." }, { - "left": "issue:43720", - "right": "issue:44361", + "left": "issue:33666", + "right": "issue:36010", "accept": false, - "reason": "BitNet accelerate-loading failure and MLukeTokenizer AttributeError are different code paths and bugs." + "reason": "Multi-GPU training for Qwen2-VL is unrelated to the GenerationMixin import regression." }, { - "left": "issue:33666", - "right": "issue:36331", + "left": "issue:43720", + "right": "issue:43927", "accept": false, - "reason": "Qwen2-VL multi-GPU training is unrelated to a custom trainer compute_loss signature error." + "reason": "Packed-weight unpacking during accelerate loading is a different issue from losing custom token IDs on save/load." }, { - "left": "issue:43540", - "right": "issue:44977", + "left": "issue:43606", + "right": "issue:43927", "accept": false, - "reason": "Qwen3OmniMoe video-processing ValueError is not the same as Qwen3.5 flash-attention generation trouble." + "reason": "CPU offload device mismatch is unrelated to DiaConfig token ID persistence." }, { - "left": "issue:43756", - "right": "issue:44479", + "left": "issue:43257", + "right": "issue:43299", "accept": false, - "reason": "Smollm3 RoPE-layer dropping and Qwen VL video regression are unrelated model-specific bugs." + "reason": "Both concern Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is a broader v5.0 dev loading break." }, { - "left": "issue:44625", - "right": "issue:45584", + "left": "issue:43278", + "right": "issue:43844", "accept": false, - "reason": "Qwen3.5 num_labels propagation is unrelated to Whisper empty-transcription handling." + "reason": "Different numerical/training problems: evaluation dtype change versus ZeRO-3 gradient growth." }, { - "left": "issue:43824", - "right": "issue:44589", + "left": "issue:44488", + "right": "issue:44843", "accept": false, - "reason": "Missing Qwen2_5_VL import export is unrelated to the Float8 storage lookup TypeError." + "reason": "Loading a specific model is not the same underlying bug as unconditional Hub metadata access breaking offline mode." }, { - "left": "issue:43646", - "right": "issue:44466", + "left": "issue:43493", + "right": "issue:43844", "accept": false, - "reason": "Custom model initialization breakage is different from inconsistent lm_head serialization." + "reason": "Implementation mismatch in SigLIP2 is unrelated to DeepSpeed gradient instability." }, { - "left": "issue:44589", - "right": "issue:45520", + "left": "issue:44964", + "right": "issue:45405", "accept": false, - "reason": "Float8 storage resolution failure is unrelated to the flash_attn import KeyError on Python 3.13." + "reason": "Model loading failure and a version requirement bump are not the same bug." }, { - "left": "issue:44423", - "right": "issue:44734", + "left": "issue:39692", + "right": "issue:43493", "accept": false, - "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing are separate serve bugs." + "reason": "Doc-example errors are distinct from a model implementation discrepancy with the JAX reference." }, { - "left": "issue:39692", - "right": "issue:43295", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "Docs/example errors do not match the processor.tokenizer regression in custom model code." + "reason": "Different loader failures in different components: packed BitNet weights versus MLukeTokenizer attribute handling." }, { - "left": "issue:33357", - "right": "issue:33666", + "left": "issue:43540", + "right": "issue:44977", "accept": false, - "reason": "MacOS bus error on CLIP is unrelated to Qwen2-VL multi-GPU training." + "reason": "Both are Qwen3.5-family issues, but one is video-input validation and the other is flash-attention generation behavior." }, { - "left": "issue:43572", - "right": "issue:44291", + "left": "issue:43756", + "right": "issue:44479", "accept": false, - "reason": "StableLmConfig pad_token_idx loss is a different issue from init_empty_weights argument incompatibility." + "reason": "RoPE-layer documentation/model-structure issue is unrelated to the Qwen video-input regression." }, { - "left": "issue:43504", - "right": "issue:43606", + "left": "issue:33666", + "right": "issue:36331", "accept": false, - "reason": "Beit preset load failure and bark-small CPU offload mismatch are unrelated." + "reason": "Training setup for Qwen2-VL is unrelated to a Trainer API signature mismatch." }, { - "left": "issue:43504", - "right": "issue:43653", + "left": "issue:45200", + "right": "issue:45229", "accept": false, - "reason": "Legacy-field loading for Beit is unrelated to BigBirdTokenizer special-token registration." + "reason": "Defaulting mm_token_type_ids to zeros and multi-GPU inference OOM are unrelated." }, { - "left": "issue:44521", - "right": "issue:45290", + "left": "issue:44625", + "right": "issue:45584", "accept": false, - "reason": "All-zero assistant masks for multimodal inputs is not the same bug as tool-call assistant message crashes." + "reason": "Config propagation for Qwen3.5 labels is unrelated to Whisper empty-transcription generation failure." }, { - "left": "issue:43927", - "right": "issue:44361", + "left": "issue:43824", + "right": "issue:44589", "accept": false, - "reason": "DiaConfig save/load token-ID corruption and MLukeTokenizer AttributeError are different failures." + "reason": "Importing a missing class and failing to find Float8 storage are different version/runtime issues." }, { - "left": "issue:43994", - "right": "issue:44493", + "left": "issue:43646", + "right": "issue:44466", "accept": false, - "reason": "SigLIP2 nonsense outputs are unrelated to widespread position-id key warnings." + "reason": "Both involve v5 behavior, but custom model initialization and lm_head serialization are different code paths and bugs." }, { - "left": "issue:44792", - "right": "issue:45405", + "left": "issue:44423", + "right": "issue:44734", "accept": false, - "reason": "Janus image-generation test failure is unrelated to the PEFT version bump." + "reason": "Serving crashes in different endpoints and failure modes: multimodal `.to()` handling versus KV-cache continuation indexing." }, { - "left": "issue:44464", - "right": "issue:45584", + "left": "issue:44589", + "right": "issue:45520", "accept": false, - "reason": "Chunked-generation compiled-forward inconsistency is unrelated to Whisper empty-transcription handling." + "reason": "Float8 storage lookup and Python 3.13 flash_attn import key errors are unrelated." }, { - "left": "issue:43278", - "right": "issue:43701", + "left": "issue:39692", + "right": "issue:43295", "accept": false, - "reason": "Embedding dtype drift between train/eval is unrelated to resume_from_checkpoint key mismatch." + "reason": "A documentation example bug is not the same as a regression in custom model code using processor.tokenizer." }, { - "left": "issue:43606", - "right": "issue:44112", + "left": "issue:33357", + "right": "issue:33666", "accept": false, - "reason": "bark-small CPU offload mismatch and GraniteSpeech stale device-override CI failure are different bugs." + "reason": "MacOS bus error with CLIP is unrelated to Qwen2-VL multi-GPU training." } ] }, @@ -6537,7 +6376,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:15354", "issue:29942", @@ -6546,273 +6385,266 @@ "issue:38175", "issue:39401", "issue:43232", - "issue:43295", + "issue:43278", "issue:43334", - "issue:43388", - "issue:43493", + "issue:43504", "issue:43525", - "issue:43526", "issue:43540", "issue:43572", + "issue:43575", "issue:43582", + "issue:43606", + "issue:43653", "issue:43701", - "issue:43824", "issue:43844", - "issue:43873", + "issue:43927", "issue:43957", - "issue:43975", + "issue:43994", + "issue:44112", + "issue:44291", "issue:44315", + "issue:44361", "issue:44464", - "issue:44466", + "issue:44493", + "issue:44521", "issue:44554", "issue:44568", - "issue:44610", "issue:44792", "issue:44849", - "issue:44898", - "issue:44936", "issue:44964", "issue:44977", - "issue:44987", "issue:45072", "issue:45084", - "issue:45127", "issue:45198", "issue:45200", + "issue:45290", "issue:45325", "issue:45356", - "issue:45412", + "issue:45405", "issue:45446", "issue:45520", - "issue:45563", + "issue:45584", "issue:45701" ], "soft_pairs": [ - "issue:29942|issue:30333", + "issue:43572|issue:44291", + "issue:43504|issue:43606", + "issue:44521|issue:45290", + "issue:43504|issue:43653", + "issue:43927|issue:44361", + "issue:43994|issue:44493", + "issue:43575|issue:45405", + "issue:44792|issue:45405", + "issue:44464|issue:45584", + "issue:43278|issue:43701", + "issue:43606|issue:44112", "issue:44568|issue:45701", + "issue:29942|issue:30333", "issue:44554|issue:45446", "issue:15354|issue:38175", "issue:43844|issue:44964", "issue:43540|issue:45200", "issue:45084|issue:45520", - "issue:43582|issue:44977", "issue:39401|issue:44568", + "issue:43582|issue:44977", "issue:43232|issue:44849", "issue:45072|issue:45198", "issue:37428|issue:43957", "issue:44315|issue:44849", "issue:44554|issue:45356", "issue:43334|issue:43525", - "issue:43295|issue:44610", - "issue:45200|issue:45325", - "issue:44466|issue:45127", - "issue:43493|issue:43701", - "issue:44898|issue:45412", - "issue:44792|issue:45563", - "issue:43873|issue:44464", - "issue:43388|issue:43526", - "issue:37428|issue:43824", - "issue:43572|issue:44987", - "issue:43844|issue:44936", - "issue:43572|issue:43957", - "issue:43824|issue:43975" + "issue:45200|issue:45325" ], - "prepared_review_unit_hash": "9ba2e1813d090e51d681d6203b380c350687afd6af2f55902ee3317f5d436028", + "prepared_review_unit_hash": "99f87a597a6426f04244d45185d76d12399c100e1f875f2ba7b30eeb753f6e09", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35700, - "estimated_input_tokens": 8925, - "estimated_eval_tokens": 18106 + "soft_pair_count": 27, + "serialized_chars": 35289, + "estimated_input_tokens": 8823, + "estimated_eval_tokens": 17902 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is very heterogeneous: most items are unrelated model-, generation-, loading-, or infrastructure-specific bugs. There is no strong duplicate set here; at most there is a loose tokenization-regression theme around a few issues.", - "confidence": 0.21, - "canonical_issue_reason": "issue:45701 is the broadest, most generic tokenizer-regression report in the set, so it is the least arbitrary representative of this weakly related cluster.", + "summary": "This is a heterogeneous grab-bag of unrelated Transformers issues; the soft pairs mostly share only broad keywords like tokenization, loading, or generation, not the same underlying bug. No pair looks safe to merge as a duplicate, and there is no strong single canonical issue for the cluster.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45701 best fits the only recurring theme (tokenization/version regressions), but the cluster overall does not look like true duplicates.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:29942", - "right": "issue:30333", + "left": "issue:43572", + "right": "issue:44291", "accept": false, - "reason": "Flash Attention 2 test failures and MLflow job-status reporting are unrelated subsystems and bugs." + "reason": "Both are Transformers 5.x regressions, but one is StableLmConfig padding metadata and the other is init_empty_weights handling; different code paths and fixes." }, { - "left": "issue:44568", - "right": "issue:45701", + "left": "issue:43504", + "right": "issue:43606", "accept": false, - "reason": "Both concern tokenization, but one is a model-specific special-tokens bug and the other is a broad version/regression report; not the same concrete issue." + "reason": "Different bug classes: BEiT pretrained loading with a legacy field vs bark-small CPU offload device mismatch." }, { - "left": "issue:44554", - "right": "issue:45446", + "left": "issue:44521", + "right": "issue:45290", "accept": false, - "reason": "MPS attention correctness and a flex_attention import version check are different code paths." + "reason": "Both touch apply_chat_template, but one is wrong assistant mask generation for multimodal inputs and the other is a crash on tool-call messages with no content." }, { - "left": "issue:15354", - "right": "issue:38175", + "left": "issue:43504", + "right": "issue:43653", "accept": false, - "reason": "TorchScript GeneratorExp export failure is unrelated to SigLIP2 zero-probability outputs." + "reason": "Unrelated failures: model loading legacy-field cleanup vs BigBirdTokenizer special-token registration/empty decode." }, { - "left": "issue:43844", - "right": "issue:44964", + "left": "issue:43927", + "right": "issue:44361", "accept": false, - "reason": "DeepSpeed ZeRO-3 gradient growth and Phi-4 multimodal loading failure are different problems." + "reason": "Different components and symptoms: DiaConfig losing custom token IDs vs MLukeTokenizer attribute errors on tasks." }, { - "left": "issue:43540", - "right": "issue:45200", + "left": "issue:43994", + "right": "issue:44493", "accept": false, - "reason": "Video-input processing in Qwen3OmniMoe and Gemma 4 multimodal token-type IDs are distinct issues." + "reason": "Different regressions: Siglip2 producing bad outputs vs a position-id key warning/error affecting many models." }, { - "left": "issue:45084", - "right": "issue:45520", + "left": "issue:43575", + "right": "issue:45405", "accept": false, - "reason": "Template-node compilation error and flash_attn import handling on Python 3.13 are unrelated." + "reason": "No shared underlying bug; one is tensor-parallel OOM, the other is a dependency version bump for PEFT." }, { - "left": "issue:43582", - "right": "issue:44977", + "left": "issue:44792", + "right": "issue:45405", "accept": false, - "reason": "Apple Silicon allocator TypeError and Qwen3.5 flash-attention generation bug do not match." + "reason": "A failing janus test is unrelated to an unreleased PEFT minimum-version change." }, { - "left": "issue:39401", - "right": "issue:44568", + "left": "issue:44464", + "right": "issue:45584", "accept": false, - "reason": "Both are tokenizer bugs, but offset_mapping and missing BOS/EOS special-token insertion are not the same failure." + "reason": "Different generation-path bugs in different models: compiled-forward chunking vs Whisper empty transcription after token alignment." }, { - "left": "issue:43232", - "right": "issue:44849", + "left": "issue:43278", + "right": "issue:43701", "accept": false, - "reason": "Generation kwargs sync_gpus bug and Qwen3.5 hidden-state output bug are separate generation-path issues." + "reason": "One is a BF16-to-FP32 dtype change during eval, the other a resume_from_checkpoint key mismatch; no common fix." }, { - "left": "issue:45072", - "right": "issue:45198", + "left": "issue:43606", + "right": "issue:44112", "accept": false, - "reason": "dtype mismatch inference bugs and Wav2Vec2 save/tokenization failures are unrelated." + "reason": "Runtime CPU offload device mismatch is not the same as a stale CI test override in GraniteSpeech." }, { - "left": "issue:37428", - "right": "issue:43957", + "left": "issue:44568", + "right": "issue:45701", "accept": false, - "reason": "Flash-attention import error and meta-device model loading breakage are different root causes." + "reason": "The second issue is a broad version-dependent tokenization complaint, not evidence of the same specific BOS/EOS bug." }, { - "left": "issue:44315", - "right": "issue:44849", + "left": "issue:29942", + "right": "issue:30333", "accept": false, - "reason": "Liger Kernel application with model_init and Qwen3.5 hidden-states behavior are unrelated." + "reason": "Completely unrelated: Flash Attention 2 test failures vs MLFlow reporting a job as still running after failure." }, { "left": "issue:44554", - "right": "issue:45356", - "accept": false, - "reason": "MPS attention correctness and Kimi-K2.5 tokenizer codec handling are not the same bug." - }, - { - "left": "issue:43334", - "right": "issue:43525", + "right": "issue:45446", "accept": false, - "reason": "Both report missing pad_token_id, but they involve different model families and config paths; too weak to treat as one duplicate." + "reason": "Distinct areas: MPS attention correctness with mismatched head dims vs a PyTorch version check in flex_attention import logic." }, { - "left": "issue:43295", - "right": "issue:44610", + "left": "issue:15354", + "right": "issue:38175", "accept": false, - "reason": "Processor/tokenizer regression and OmDet-Turbo image size mismatch are different issues." + "reason": "TorchScript GeneratorExp export failure and Siglip2 zero-probability outputs are unrelated bugs." }, { - "left": "issue:45200", - "right": "issue:45325", + "left": "issue:43844", + "right": "issue:44964", "accept": false, - "reason": "Gemma 4 token-type defaults and Qwen2.5-VL rope-index scaling are separate multimodal bugs." + "reason": "Different problems: ZeRO-3 gradient growth during training vs failure loading Phi-4 multimodal models." }, { - "left": "issue:44466", - "right": "issue:45127", + "left": "issue:43540", + "right": "issue:45200", "accept": false, - "reason": "Serialization of lm_head.weight and LoRA merge collapse with extended vocabulary are not the same concrete defect." + "reason": "Different multimodal input bugs: Qwen3OmniMoe video processing vs Gemma 4 text-only mm_token_type_ids defaults." }, { - "left": "issue:43493", - "right": "issue:43701", + "left": "issue:45084", + "right": "issue:45520", "accept": false, - "reason": "SigLIP2 implementation discrepancy and checkpoint key mismatch are unrelated." + "reason": "No shared code path: template compilation TypeError vs flash_attn import KeyError on Python 3.13." }, { - "left": "issue:44898", - "right": "issue:45412", + "left": "issue:39401", + "right": "issue:44568", "accept": false, - "reason": "Perceiver interpolation failure and RT-DETR memory release behavior are different classes of bugs." + "reason": "Both are tokenizer issues, but one is offset_mapping for Qwen3 and the other is add_special_tokens not adding BOS/EOS for mdeberta-v3." }, { - "left": "issue:44792", - "right": "issue:45563", + "left": "issue:43582", + "right": "issue:44977", "accept": false, - "reason": "Janus image-generation test failure and paged-generate warning behavior are not the same underlying problem." + "reason": "Different runtime/model issues: AppleSilicon allocator warmup TypeError vs Qwen3.5 flash-attention generation regression." }, { - "left": "issue:43873", - "right": "issue:44464", + "left": "issue:43232", + "right": "issue:44849", "accept": false, - "reason": "Quantization offloading and chunked generation with compiled forward are different subsystems and failures." + "reason": "Separate generation bugs: sync_gpus/_update_model_kwargs_for_generation vs Qwen3.5 output_hidden_states behavior." }, { - "left": "issue:43388", - "right": "issue:43526", + "left": "issue:45072", + "right": "issue:45198", "accept": false, - "reason": "Both involve labels, but one is trainer metric gathering and the other is BeitImageProcessorFast reduction logic." + "reason": "Different subsystems and failures: bfloat16 dtype mismatches vs Wav2Vec2 save_pretrained/tokenization problems." }, { "left": "issue:37428", - "right": "issue:43824", + "right": "issue:43957", "accept": false, - "reason": "Flash-attention import error and Qwen2.5-VL import failure are distinct missing-symbol problems." + "reason": "Import-time flash-attn symbol breakage is unrelated to meta-device model loading regressions." }, { - "left": "issue:43572", - "right": "issue:44987", + "left": "issue:44315", + "right": "issue:44849", "accept": false, - "reason": "StableLmConfig pad_token_idx regression and loading physical-intelligence/fast are different issues." + "reason": "Liger Kernel not being applied in model_init is unrelated to Qwen3.5 hidden-state generation behavior." }, { - "left": "issue:43844", - "right": "issue:44936", + "left": "issue:44554", + "right": "issue:45356", "accept": false, - "reason": "ZeRO-3 gradient growth and trainer.evaluate after train are unrelated training bugs." + "reason": "Different issues: MPS attention correctness vs Kimi-K2.5 tokenizer codec handling/regression." }, { - "left": "issue:43572", - "right": "issue:43957", + "left": "issue:43334", + "right": "issue:43525", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and meta-device loading breakage are not the same bug." + "reason": "Both mention missing pad_token_id, but they are model-specific config errors in different classes and are not the same underlying bug." }, { - "left": "issue:43824", - "right": "issue:43975", + "left": "issue:45200", + "right": "issue:45325", "accept": false, - "reason": "Import failure for Qwen2.5-VL and DeepSeek detokenization regression are unrelated." + "reason": "Unrelated multimodal bugs: Gemma 4 token-type IDs vs Qwen2.5-VL rope-index scaling for still images." } ] }, @@ -6834,63 +6666,75 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:29942", - "issue:32090", "issue:37428", "issue:38175", "issue:42915", - "issue:43262", "issue:43278", "issue:43295", - "issue:43425", + "issue:43388", "issue:43450", "issue:43493", + "issue:43526", + "issue:43572", "issue:43575", - "issue:43606", "issue:43653", + "issue:43701", "issue:43819", "issue:43824", - "issue:43825", "issue:43827", + "issue:43844", "issue:43856", "issue:43873", "issue:43901", + "issue:43957", + "issue:43975", "issue:43994", "issue:44112", "issue:44206", "issue:44360", "issue:44442", - "issue:44448", "issue:44462", + "issue:44464", + "issue:44466", "issue:44514", "issue:44521", "issue:44561", "issue:44610", - "issue:44743", "issue:44779", - "issue:44857", + "issue:44792", + "issue:44898", + "issue:44936", + "issue:44987", "issue:45083", "issue:45127", "issue:45198", - "issue:45200", "issue:45245", "issue:45290", - "issue:45310", "issue:45381", "issue:45412", - "issue:45478", - "issue:45584", - "issue:45588", - "issue:45593" + "issue:45563", + "issue:45584" ], "soft_pairs": [ + "issue:43295|issue:44610", + "issue:44466|issue:45127", + "issue:44898|issue:45412", + "issue:44792|issue:45563", + "issue:43493|issue:43701", + "issue:37428|issue:43824", + "issue:43388|issue:43526", + "issue:43873|issue:44464", + "issue:43572|issue:44987", + "issue:43844|issue:44936", + "issue:43572|issue:43957", + "issue:43824|issue:43975", "issue:44514|issue:44521", - "issue:43295|issue:43450", "issue:44360|issue:44561", - "issue:43295|issue:43873", + "issue:43295|issue:43450", "issue:38175|issue:45412", + "issue:43295|issue:43873", "issue:42915|issue:43824", "issue:43295|issue:43994", "issue:44442|issue:45198", @@ -6899,245 +6743,240 @@ "issue:43278|issue:43856", "issue:44206|issue:45584", "issue:43653|issue:44112", - "issue:43450|issue:43493", - "issue:45127|issue:45245", "issue:44462|issue:44561", + "issue:45127|issue:45245", + "issue:43450|issue:43493", "issue:44561|issue:44779", "issue:43819|issue:45083", - "issue:43450|issue:43575", "issue:45290|issue:45381", - "issue:44610|issue:44857", - "issue:43819|issue:44112", - "issue:37428|issue:43425", - "issue:44857|issue:45593", - "issue:43262|issue:43450", - "issue:43278|issue:43825", - "issue:45200|issue:45588", - "issue:43606|issue:45083", - "issue:44448|issue:44743", - "issue:29942|issue:32090", - "issue:45310|issue:45478" + "issue:43450|issue:43575" ], - "prepared_review_unit_hash": "a2d0cb6ce830e1143dc638c181f50ead3405d3d0a2c60756c568ba022d970d45", + "prepared_review_unit_hash": "0ca9fbe24d0caa4a60284d1e14a35eb9fae0e40983dd8c9fc8936ac8f471f5c6", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36937, - "estimated_input_tokens": 9235, - "estimated_eval_tokens": 18726 + "soft_pair_count": 32, + "serialized_chars": 37338, + "estimated_input_tokens": 9335, + "estimated_eval_tokens": 18926 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issue cluster with two real duplicate-style matches: a Qwen3.5 MoE from_pretrained regression pair and a SigLIP2 implementation-discrepancy pair. The remaining soft edges span unrelated models, processors, docs, and runtime failures.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43493 is the best canonical anchor because it states the broader SigLIP2 root-cause discrepancy, while issue:38175 is a symptom report of that same model bug.", + "summary": "This is a mostly heterogeneous cluster of regressions, API breakages, model-specific bugs, and docs/test noise. Only the two SigLIP2 reports look like the same underlying issue; the rest are not true duplicates.", + "confidence": 0.66, + "canonical_issue_reason": "issue:43493 is the clearest root-cause report in the only true duplicate pair: it describes the SigLIP2 HF-vs-JAX implementation discrepancy directly, while issue:38175 reads more like a symptom report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43493 is the strongest overall issue to keep as the representative anchor: it is root-cause oriented, model-specific, and more reusable than the narrower symptom reports.", + "best_issue_reason": "issue:43295 is the most central issue in the overall cluster, with the broadest regression framing and the strongest soft-link connectivity to other nearby API-breakage reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44514", - "right": "issue:44521", + "left": "issue:43295", + "right": "issue:44610", "accept": false, - "reason": "Both are apply_chat_template multimodal regressions, but one is a batched padding crash and the other is all-zero assistant masks; different failure modes." + "reason": "Processor/tokenizer regression vs. a feature-extractor input-size mismatch; related area, but different bug and code path." }, { - "left": "issue:43295", - "right": "issue:43450", + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Inconsistent lm_head serialization/tied-weights behavior vs. LoRA-merge collapse with extended vocabulary; not the same failure." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver interpolation failure vs. RT-DETR memory leak; unrelated symptoms and code paths." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "janus image-generation test failure vs. stale paged-generate warning; different issue types." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 output discrepancy vs. resume_from_checkpoint key mismatch; distinct regressions." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and different modules/models." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both mention label handling, but one is gather_for_metrics truncation and the other is BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:43873", + "right": "issue:44464", + "accept": false, + "reason": "Quantization/offloading problem vs. compiled-forward chunked generation inconsistency; not the same bug." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in config vs. model-loading failure for a specific checkpoint; unrelated." + }, + { + "left": "issue:43844", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO-3 gradient explosion vs. trainer.evaluate() after train; different training lifecycle issues." + }, + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "Config field regression vs. meta-device model-loading breakage; separate problems." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "Qwen2.5-VL import failure vs. DeepSeek detokenization regression; different model families and failures." + }, + { + "left": "issue:44514", + "right": "issue:44521", "accept": false, - "reason": "One is a processor.tokenizer / image-tokenizer regression, the other is batched video shape output; unrelated code paths." + "reason": "Both concern apply_chat_template, but one crashes on batched padding=False input while the other yields empty assistant masks; related but not the same concrete bug." }, { "left": "issue:44360", "right": "issue:44561", "accept": false, - "reason": "A DSA indexer ReLU discussion and a removed FX helper breaking trust_remote_code models are unrelated." + "reason": "DSA/ReLU discussion vs. removal of is_torch_fx_available breaking trust_remote_code models; unrelated." }, { "left": "issue:43295", - "right": "issue:43873", + "right": "issue:43450", "accept": false, - "reason": "Processor.tokenizer regression versus quantization/offload behavior; no shared underlying bug." + "reason": "Processor.tokenizer/image handling regression vs. batched video-shape bug; both in processors, but different modalities and code paths." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 zero-probability outputs and RT-DETR memory release issues are unrelated." + "reason": "SigLIP2 zero-probability outputs vs. RT-DETR GC/memory leak; unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Custom processor regression vs. quantization/offloading bug; different subsystems." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3 MoE FP8 failure and Qwen2.5-VL import error are different models and different bugs." + "reason": "Qwen3MoE FP8 config failure vs. Qwen2.5-VL import failure; different models and symptoms." }, { "left": "issue:43295", "right": "issue:43994", "accept": false, - "reason": "Both mention SigLIP2-related models, but one is a processor/tokenizer regression and the other is a model output-quality issue; not the same bug." + "reason": "Processor/tokenizer regression vs. SigLIP2 nonsensical outputs; different affected areas." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "FastSpeech2 tokenizer loading and Wav2Vec2 save/tokenization failures are separate tokenizer problems." + "reason": "AutoTokenizer loading failure for one tokenizer vs. save_pretrained/tokenization failure for another audio model; not a duplicate." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Different docs issues: v5 pipeline removals versus TextClassificationPipeline return_all_scores wording." + "reason": "Both are docs inconsistencies, but they refer to different deprecated arguments and different pipeline docs." }, { "left": "issue:38175", "right": "issue:43493", "accept": true, - "reason": "Both report the same SigLIP2 implementation mismatch: one as zero probabilities, the other as an HF-vs-JAX discrepancy." + "reason": "Both point to the same SigLIP2 model-implementation bug; the zero-probability report is consistent with the broader HF-vs-JAX discrepancy." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype drift in eval and Qwen3 MoE training memory use are unrelated." + "reason": "Embedding dtype drift in eval vs. memory inefficiency in Qwen3 MoE training; distinct issues." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "LasrFeatureExtractor center-arg crash and Whisper empty-transcription failure are different audio bugs." + "reason": "Feature extractor center-arg crash vs. Whisper empty-transcription generation failure; different code paths." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "BigBirdTokenizer special-token registration and GraniteSpeech CI device-override failure are unrelated." + "reason": "BigBirdTokenizer special-token registration bug vs. a GraniteSpeech CI device-override test; unrelated." }, { - "left": "issue:43450", - "right": "issue:43493", + "left": "issue:44462", + "right": "issue:44561", "accept": false, - "reason": "Video batching shape bug and SigLIP2 JAX discrepancy are different code paths." + "reason": "AutoTokenizer ignoring tokenizer.json vs. removal of is_torch_fx_available breaking remote code; separate regressions." }, { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "LoRA tie_word_embeddings collapse and category-cardinality overflow are unrelated." + "reason": "LoRA merge / tied-weights collapse vs. category-count limit runtime error; unrelated." }, { - "left": "issue:44462", - "right": "issue:44561", + "left": "issue:43450", + "right": "issue:43493", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and trust_remote_code FX-helper removal are different regressions." + "reason": "Video processor batched-shape bug vs. SigLIP2 implementation discrepancy; different subsystems." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "Shared v5 timeframe only; one is a removed-helper import error, the other is Deepseek tokenizer output mismatch." + "reason": "Remote-code breakage from removed helper vs. DeepSeek tokenizer regression; distinct issues." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "DAC.from_latents mismatch and qwen3_omni_moe length-helper behavior are unrelated." - }, - { - "left": "issue:43450", - "right": "issue:43575", - "accept": false, - "reason": "Video batch shape bug and Qwen2-57B-A14B tensor-parallel OOM are different issues." + "reason": "DAC forward/from_latents mismatch vs. Qwen3 Omni MoE feature-length helper behavior; not the same bug." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat-template tool-call crash and Qwen2.5-VL vision_position_ids regression are separate bugs." - }, - { - "left": "issue:44610", - "right": "issue:44857", - "accept": false, - "reason": "OmDet input-size mismatch and LwDetr AMP crash are different detection-model problems." - }, - { - "left": "issue:43819", - "right": "issue:44112", - "accept": false, - "reason": "DAC latents mismatch and GraniteSpeech CI device-override failure are unrelated." - }, - { - "left": "issue:37428", - "right": "issue:43425", - "accept": false, - "reason": "Flash-attention import error and Torch 2.10 incompatibility are adjacent compatibility topics, not the same underlying bug." - }, - { - "left": "issue:44857", - "right": "issue:45593", - "accept": false, - "reason": "D-FINE auxiliary-loss bug and LwDetr AMP crash are different detection-model issues." + "reason": "Chat-template tool-call crash vs. Qwen2.5-VL video vision_position_ids bug; separate multimodal regressions." }, { - "left": "issue:43262", - "right": "issue:43450", - "accept": false, - "reason": "Audio chat_template sampling-rate default and video batching shape bug are unrelated." - }, - { - "left": "issue:43278", - "right": "issue:43825", - "accept": false, - "reason": "BF16/FP32 dtype shift and pipeline docs wording are unrelated." - }, - { - "left": "issue:45200", - "right": "issue:45588", + "left": "issue:43450", + "right": "issue:43575", "accept": false, - "reason": "Gemma 4 mm_token_type_ids defaults and flash_attention s_aux=None crash are different model/runtime bugs." - }, - { - "left": "issue:43606", - "right": "issue:45083", - "accept": false, - "reason": "CPU-offload device mismatch and qwen3_omni_moe output-length helper behavior are unrelated." - }, - { - "left": "issue:44448", - "right": "issue:44743", - "accept": false, - "reason": "Pegasus v4/v5 output drift and Qwen3.5 recurrent-state reset are different regressions." - }, - { - "left": "issue:29942", - "right": "issue:32090", - "accept": false, - "reason": "Flash Attention 2 test failures and Trainer _gpu_broadcast_one NoneType error are unrelated." - }, - { - "left": "issue:45310", - "right": "issue:45478", - "accept": true, - "reason": "Same Qwen3.5 MoE from_pretrained regression; 45478 is the later duplicate with the same failure mode." + "reason": "Batched video processor shape issue vs. tensor-parallel OOM on model load; unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is conservative and grounded in the packet. The two accepted soft edges look like genuine duplicates or near-duplicates: the identical Qwen3.5 MoE from_pretrained regression pair and the SigLIP2 discrepancy pair. The remaining verdicts are mostly cautious and do not overclaim shared root causes." + "feedback": "The cluster summary is grounded in the packet, and the soft-edge calls are generally conservative. The only accepted duplicate pair, issue:38175 \u2194 issue:43493, is well supported by the SigLIP2 titles, while the rest are reasonably treated as related but distinct." }, "evaluator_used": true, "retried": false, @@ -7156,57 +6995,68 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:29127", + "issue:29942", "issue:32090", - "issue:34689", - "issue:35141", "issue:37428", "issue:38175", "issue:39692", "issue:40444", - "issue:43066", + "issue:43262", + "issue:43278", "issue:43335", + "issue:43425", + "issue:43450", "issue:43504", "issue:43526", "issue:43550", "issue:43606", "issue:43611", "issue:43653", - "issue:43701", "issue:43720", "issue:43723", "issue:43819", - "issue:43927", + "issue:43825", "issue:43931", "issue:44112", "issue:44292", "issue:44351", "issue:44448", - "issue:44462", "issue:44493", "issue:44554", "issue:44568", + "issue:44610", + "issue:44743", "issue:44779", "issue:44855", "issue:44857", "issue:44977", - "issue:44987", "issue:45072", "issue:45083", "issue:45200", "issue:45229", - "issue:45372", + "issue:45310", "issue:45381", + "issue:45478", "issue:45491", "issue:45520", - "issue:45563", "issue:45588", - "issue:45593", - "issue:45698" + "issue:45593" ], "soft_pairs": [ + "issue:37428|issue:43425", + "issue:43819|issue:44112", + "issue:44610|issue:44857", + "issue:44857|issue:45593", + "issue:43278|issue:43825", + "issue:45200|issue:45588", + "issue:43262|issue:43450", + "issue:43606|issue:45083", + "issue:45310|issue:45478", + "issue:44448|issue:44743", + "issue:29942|issue:32090", "issue:37428|issue:44351", "issue:43335|issue:43550", "issue:39692|issue:43526", @@ -7216,264 +7066,263 @@ "issue:43335|issue:43653", "issue:29127|issue:32090", "issue:44292|issue:45520", - "issue:44448|issue:44779", "issue:43335|issue:45083", + "issue:44448|issue:44779", "issue:44855|issue:45083", "issue:39692|issue:44493", "issue:43335|issue:43504", - "issue:43550|issue:44112", "issue:38175|issue:44493", + "issue:43550|issue:44112", "issue:44568|issue:45491", - "issue:44977|issue:45229", "issue:43723|issue:44554", - "issue:43335|issue:45072", + "issue:44977|issue:45229", "issue:40444|issue:43526", + "issue:43335|issue:45072", "issue:43335|issue:45593", - "issue:43720|issue:44112", - "issue:44448|issue:44977", - "issue:43066|issue:44462", - "issue:34689|issue:35141", - "issue:44977|issue:45563", - "issue:45372|issue:45588", - "issue:43927|issue:45698", - "issue:43606|issue:44857", - "issue:44857|issue:45072", - "issue:43701|issue:45083", - "issue:37428|issue:44987" + "issue:43720|issue:44112" ], - "prepared_review_unit_hash": "10d638bb137d79c2a385cda8f50a62382c271f5e63a46d69b24a44f5726e2b69", + "prepared_review_unit_hash": "ecdd49ce1a0682c54be375b38bfddb1d2ef97ef70827be4aff934b013ad283b6", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 37215, - "estimated_input_tokens": 9304, - "estimated_eval_tokens": 18864 + "soft_pair_count": 34, + "serialized_chars": 37450, + "estimated_input_tokens": 9363, + "estimated_eval_tokens": 18982 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous and does not contain any true duplicate pairs. issue:43335 is the best cluster anchor only because it is the most connected node in the weak similarity graph, not because it matches the others concretely.", - "confidence": 0.38, - "canonical_issue_reason": "issue:43335 has the most soft links to other items, so it is the least-bad representative of this noisy cluster.", + "summary": "This cluster is mostly a heterogeneous bundle of unrelated issues. The only clear duplicate pair is the repeated Qwen3.5 MoE `from_pretrained` failure reported in issues 45310 and 45478; the rest are different bugs or regressions and should not be merged together.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45310 is the best canonical issue because it is the earlier report of the same Qwen3.5 MoE `from_pretrained` error, and issue:45478 is a near-identical follow-up duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43335 is the most central item here and the most plausible anchor for triage, though the cluster is not a real duplicate set.", + "best_issue_reason": "issue:45310 is the strongest representative issue in this set, but only for the 45310/45478 duplicate pair; it should not be treated as a canonical for the whole cluster since the remaining items are unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:37428", - "right": "issue:44351", + "right": "issue:43425", "accept": false, - "reason": "Different missing import targets and different failure surfaces; both are import errors, but not the same bug." + "reason": "Different problems: one is a missing `_flash_supports_window_size` import, the other is a broad Torch 2.10 incompatibility report." }, { - "left": "issue:43335", - "right": "issue:43550", + "left": "issue:43819", + "right": "issue:44112", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug vs Bamba torch.compile/SDPA failure; different model and code path." + "reason": "Unrelated failures: DAC latent reconstruction mismatch vs a GraniteSpeech CI device-override test issue." }, { - "left": "issue:39692", - "right": "issue:43526", + "left": "issue:44610", + "right": "issue:44857", "accept": false, - "reason": "Docs example/model-processor mismatch and quantization failure vs BeitImageProcessorFast reduce_labels bug." + "reason": "Different code paths: OmDet-Turbo processor input sizing vs LwDetrImageLoss AMP/CUDA crash." }, { - "left": "issue:43611", - "right": "issue:43931", + "left": "issue:44857", + "right": "issue:45593", "accept": false, - "reason": "base_model_prefix loading regression vs Qwen3-VL weight-shape mismatch; unrelated load-time failures." + "reason": "One is an AMP/CUDA crash, the other is missing auxiliary losses when denoising is off; not the same bug." }, { - "left": "issue:43819", - "right": "issue:44493", + "left": "issue:43278", + "right": "issue:43825", "accept": false, - "reason": "DAC.from_latents vs forward mismatch is unrelated to position-id key warnings across models." + "reason": "Different symptoms and components: dtype mismatch during evaluate vs a pipeline error message about unsupported translation tasks." }, { "left": "issue:45200", - "right": "issue:45381", + "right": "issue:45588", "accept": false, - "reason": "Gemma 4 mm_token_type_ids defaulting issue vs Qwen2.5-VL video vision_position_ids bug; different models and inputs." + "reason": "Different issues in multimodal/flash-attention handling; no shared concrete failing code path." }, { - "left": "issue:43335", - "right": "issue:43653", + "left": "issue:43262", + "right": "issue:43450", "accept": false, - "reason": "Sparse-layer creation in SwitchTransformers is unrelated to BigBirdTokenizer special-token registration." + "reason": "Audio chat-template sampling-rate default vs batched video processor output shapes are unrelated." }, { - "left": "issue:29127", - "right": "issue:32090", + "left": "issue:43606", + "right": "issue:45083", "accept": false, - "reason": "LayoutLMv3 error-message clarity issue vs Trainer NoneType broadcast TypeError; different problems." + "reason": "CPU offload device mismatch in bark-small is not the same as the qwen3_omni_moe feature-length helper behavior." }, { - "left": "issue:44292", - "right": "issue:45520", - "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure vs Python 3.13 flash_attn import KeyError; different code paths." + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same underlying bug: both report `transformers>=5.4.0` Qwen3.5 MoE `from_pretrained` loading failure with essentially identical title/body." }, { "left": "issue:44448", - "right": "issue:44779", + "right": "issue:44743", "accept": false, - "reason": "Pegasus v4/v5 output drift vs Deepseek tokenizer regression; both are version regressions but not the same bug." + "reason": "Different regressions: Pegasus v4/v5 output drift vs recurrent-state reset in modular Qwen3.5 code with cache." }, { - "left": "issue:43335", - "right": "issue:45083", + "left": "issue:29942", + "right": "issue:32090", "accept": false, - "reason": "SwitchTransformers config bug vs qwen3_omni_moe feature-length helper behavior; unrelated." + "reason": "Flash Attention 2 test failures and a Trainer `_gpu_broadcast_one` NoneType error are unrelated." }, { - "left": "issue:44855", - "right": "issue:45083", + "left": "issue:37428", + "right": "issue:44351", "accept": false, - "reason": "Python 3.13 DebertaV2Model import failure vs qwen3_omni_moe helper mismatch; unrelated." + "reason": "Both are import errors, but for different symbols (`_flash_supports_window_size` vs `HybridCache`) and likely different fixes." }, { - "left": "issue:39692", - "right": "issue:44493", + "left": "issue:43335", + "right": "issue:43550", "accept": false, - "reason": "SigLIP2 docs/example problems vs generic position-id key warnings; different symptoms and targets." + "reason": "SwitchTransformers sparse-layer configuration bug is unrelated to Bamba-9B-v2 torch.compile/SDPA failure." }, { - "left": "issue:43335", - "right": "issue:43504", + "left": "issue:39692", + "right": "issue:43526", "accept": false, - "reason": "SwitchTransformers sparse-layer creation vs BEiT legacy-field model loading issue; different models and fixes." + "reason": "SigLIP2 docs/model mismatch and BeitImageProcessorFast `reduce_labels` returning one label are different bugs." }, { - "left": "issue:43550", - "right": "issue:44112", + "left": "issue:43611", + "right": "issue:43931", "accept": false, - "reason": "Bamba SDPA/torch.compile bug vs GraniteSpeech stale CI test; not the same underlying issue." + "reason": "Base-model-prefix loading breakage and Qwen3-VL weight-shape mismatch are distinct load failures." }, { - "left": "issue:38175", + "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 zero-probabilities issue vs position-id key warnings; unrelated bugs." + "reason": "DAC forward/latent mismatch is unrelated to the generic `UNEXPECTED Key with position id` issue." }, { - "left": "issue:44568", - "right": "issue:45491", + "left": "issue:45200", + "right": "issue:45381", "accept": false, - "reason": "Tokenizer add_special_tokens regression vs Gemma3 NaNs from sliding-window attention; different subsystems." + "reason": "Different multimodal bugs: missing default `mm_token_type_ids` vs wrong `vision_position_ids` for Qwen2.5-VL video input." }, { - "left": "issue:44977", - "right": "issue:45229", + "left": "issue:43335", + "right": "issue:43653", "accept": false, - "reason": "Flash-attention generation issue vs multi-GPU CUDA OOM; different failure modes." + "reason": "SwitchTransformers config bug and BigBirdTokenizer mask-token registration are unrelated." }, { - "left": "issue:43723", - "right": "issue:44554", + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error-message clarity issue is not the same as the Trainer `_gpu_broadcast_one` NoneType crash." + }, + { + "left": "issue:44292", + "right": "issue:45520", "accept": false, - "reason": "AutoTokenizer.from_pretrained loading problem vs MPS attention correctness issue; unrelated." + "reason": "Qwen-3-8B-NVFP4 runtime error and Python 3.13 `flash_attn` import KeyError are different failures." }, { "left": "issue:43335", - "right": "issue:45072", + "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers config bug vs dtype mismatch in Bfloat16 inference; different code paths." + "reason": "Sparse-layer config bug in SwitchTransformers is unrelated to the qwen3_omni_moe feature-extraction helper behavior." }, { - "left": "issue:40444", - "right": "issue:43526", + "left": "issue:44448", + "right": "issue:44779", "accept": false, - "reason": "Qwen2.5-VL iterable-dataset/multi-image finetuning failure vs BeitImageProcessorFast reduce_labels bug." + "reason": "Pegasus output drift between v4/v5 is unrelated to DeepSeek tokenizer changes in v5." }, { - "left": "issue:43335", - "right": "issue:45593", + "left": "issue:44855", + "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers sparse-layer creation vs D-FINE auxiliary-loss handling; unrelated." + "reason": "Python 3.13 `IndentationError` from `@torch.jit.script` parsing is a different problem than qwen3_omni_moe feature-length behavior." }, { - "left": "issue:43720", - "right": "issue:44112", + "left": "issue:39692", + "right": "issue:44493", "accept": false, - "reason": "BitNet packed-weight loading issue vs GraniteSpeech stale device override test; different problems." + "reason": "SigLIP2 documentation errors do not match the generic position-id key warning issue." }, { - "left": "issue:44448", - "right": "issue:44977", + "left": "issue:43335", + "right": "issue:43504", "accept": false, - "reason": "Pegasus output change vs Qwen3.5 flash-attention generation bug; both affect outputs but not the same bug." + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to BEiT pretrained preset loading with a legacy field." }, { - "left": "issue:43066", - "right": "issue:44462", + "left": "issue:38175", + "right": "issue:44493", "accept": false, - "reason": "Wrong tokenizer decoder type vs AutoTokenizer ignoring tokenizer.json; related subsystem, but different concrete bugs." + "reason": "Zero probabilities in SigLIP2 and unexpected position-id keys are different symptoms with different roots." }, { - "left": "issue:34689", - "right": "issue:35141", + "left": "issue:43550", + "right": "issue:44112", "accept": false, - "reason": "Llama 3.2 Vision loading breakage vs embedding reinitialization after resize; unrelated model-loading issues." + "reason": "Bamba torch.compile/SDPA issue is unrelated to GraniteSpeech CI stale device override behavior." }, { - "left": "issue:44977", - "right": "issue:45563", + "left": "issue:44568", + "right": "issue:45491", "accept": false, - "reason": "Flash-attention generation bug vs stale warning for num_return_sequences; different severity and behavior." + "reason": "Tokenizer special-token behavior and Gemma3 NaN embeddings under mixed-length batching are different bugs." }, { - "left": "issue:45372", - "right": "issue:45588", + "left": "issue:43723", + "right": "issue:44554", "accept": false, - "reason": "Gemma 4 processor import dependency issue vs s_aux=None crash in flash_attention integration; unrelated." + "reason": "Tokenizer loading in v5 and MPS attention correctness with mismatched head dims are unrelated." }, { - "left": "issue:43927", - "right": "issue:45698", + "left": "issue:44977", + "right": "issue:45229", "accept": false, - "reason": "One issue loses custom token IDs on save/load, the other loads the wrong custom module; both serialization-related but not the same bug." + "reason": "Qwen3.5 flash-attention generation issue and Gemma4 multi-GPU OOM are different failure modes." }, { - "left": "issue:43606", - "right": "issue:44857", + "left": "issue:40444", + "right": "issue:43526", "accept": false, - "reason": "CPU offload device mismatch vs float16 AMP CUDA loss crash; different precision/device paths." + "reason": "Qwen2.5-VL iterable-dataset multi-image finetuning failure is unrelated to BeitImageProcessorFast `reduce_labels`." }, { - "left": "issue:44857", + "left": "issue:43335", "right": "issue:45072", "accept": false, - "reason": "float16 AMP crash in a loss vs bfloat16 inference dtype mismatch; same broad theme, but not the same concrete bug." + "reason": "SwitchTransformers sparse-layer bug is not the same as the bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel inference." }, { - "left": "issue:43701", - "right": "issue:45083", + "left": "issue:43335", + "right": "issue:45593", "accept": false, - "reason": "resume_from_checkpoint key mismatch vs qwen3_omni_moe helper behavior; unrelated." + "reason": "SwitchTransformers config bug and D-FINE auxiliary-loss behavior are unrelated." }, { - "left": "issue:37428", - "right": "issue:44987", + "left": "issue:43720", + "right": "issue:44112", "accept": false, - "reason": "flash-support import error vs model-loading failure for physical-intelligence/fast; too little concrete overlap." + "reason": "BitNet packed-weight unpacking during accelerate loading is not the same as the GraniteSpeech CI device override failure." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is well grounded in the packet and stays conservative. The only clearly supported duplicate pair is issue:45310/45478, and the summary correctly treats the rest as heterogeneous rather than trying to force cluster-wide canonicality. The soft-edge verdicts are appropriately cautious and do not overstate evidence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -7488,43 +7337,41 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:25251", "issue:29942", - "issue:30064", "issue:32090", "issue:33357", + "issue:34689", "issue:35141", + "issue:37428", "issue:41720", - "issue:42915", - "issue:42994", "issue:43065", - "issue:43257", + "issue:43066", "issue:43278", "issue:43329", - "issue:43381", "issue:43412", "issue:43421", "issue:43550", + "issue:43606", "issue:43653", - "issue:43756", + "issue:43701", "issue:43873", + "issue:43927", "issue:43976", "issue:43994", "issue:44186", - "issue:44206", - "issue:44265", - "issue:44361", "issue:44448", "issue:44451", + "issue:44462", "issue:44466", - "issue:44479", "issue:44493", "issue:44743", "issue:44857", - "issue:44871", "issue:44977", + "issue:44987", + "issue:45072", "issue:45083", "issue:45198", "issue:45229", @@ -7537,255 +7384,257 @@ "issue:45491", "issue:45520", "issue:45563", - "issue:45588" + "issue:45588", + "issue:45698" ], "soft_pairs": [ + "issue:43066|issue:44462", + "issue:44448|issue:44977", + "issue:34689|issue:35141", + "issue:44977|issue:45563", + "issue:43927|issue:45698", + "issue:45372|issue:45588", + "issue:37428|issue:44987", + "issue:43606|issue:44857", + "issue:44857|issue:45072", + "issue:43701|issue:45083", "issue:43976|issue:45440", + "issue:45245|issue:45563", "issue:43412|issue:45588", "issue:43550|issue:45083", - "issue:45245|issue:45563", "issue:43873|issue:44493", "issue:43873|issue:45229", "issue:29942|issue:35141", "issue:43278|issue:43329", - "issue:32090|issue:33357", - "issue:45372|issue:45406", "issue:45372|issue:45459", + "issue:45372|issue:45406", + "issue:32090|issue:33357", "issue:41720|issue:45229", "issue:43065|issue:43994", - "issue:44448|issue:44466", "issue:43065|issue:45412", - "issue:25251|issue:32090", + "issue:44448|issue:44466", "issue:44451|issue:45520", + "issue:25251|issue:32090", "issue:43653|issue:44743", - "issue:44857|issue:45588", "issue:43421|issue:45491", + "issue:44857|issue:45588", "issue:44186|issue:45198", - "issue:43278|issue:43381", - "issue:43994|issue:45412", - "issue:42915|issue:42994", - "issue:44361|issue:45198", - "issue:43756|issue:45440", - "issue:44871|issue:44977", - "issue:44265|issue:45198", - "issue:44206|issue:44479", - "issue:43257|issue:43278", - "issue:29942|issue:30064" + "issue:43994|issue:45412" ], - "prepared_review_unit_hash": "576d37d188d189b03e5ee707d3a66e65a0eb22c3278a78adad8c3fa37bd6c2d8", + "prepared_review_unit_hash": "4df963f96764a8402df8c6e6deba2fcd9b56703c392328e8ba276a92cc3278ca", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37074, - "estimated_input_tokens": 9269, - "estimated_eval_tokens": 18794 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36809, + "estimated_input_tokens": 9203, + "estimated_eval_tokens": 18662 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft links are mostly superficial similarity between unrelated failure modes, and none of the candidate pairs look like true duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the items span unrelated bugs across tokenization, model loading, quantization, export, attention, and runtime/device-mapping paths, with no confirmed duplicate pair.", + "summary": "This cluster is heterogeneous: the items span unrelated bugs in tokenization, model loading, generation, offloading, dtype handling, and export. No soft pair looks like the same underlying issue, so all candidate duplicate links should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue stands out because these reports are not one duplicate family; they cover different subsystems and failure modes, so consolidating them into a single issue would be misleading.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue stands out; any pick would be arbitrary because the cluster is not a duplicate set and the soft links are false positives.", + "best_issue_reason": "No single issue is a good global representative for this cluster; the most similar titles still describe different bugs, not one shared code-path regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43976", - "right": "issue:45440", + "left": "issue:43066", + "right": "issue:44462", "accept": false, - "reason": "Different problems: Python-version compatibility vs Native DeepSeekV3 divergence." + "reason": "Different tokenizer bugs: decoder type handling in v5 vs ignoring tokenizer.json on load." }, { - "left": "issue:43412", - "right": "issue:45588", + "left": "issue:44448", + "right": "issue:44977", "accept": false, - "reason": "Different code paths: Gemma3 export/loading failure vs flash_attention aux-state crash." + "reason": "Different models and symptoms: Pegasus output drift vs Qwen3.5 flash-attention generation failure." }, { - "left": "issue:43550", - "right": "issue:45083", + "left": "issue:34689", + "right": "issue:35141", "accept": false, - "reason": "Bamba torch.compile/SDPA regression is unrelated to qwen3_omni_moe feature-length calculation." + "reason": "One is a model-loading regression for Llama 3.2 Vision; the other is post_init reinitializing resized embeddings." }, { - "left": "issue:45245", + "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Categories-limit RuntimeError is unrelated to the stale paged-generate warning." + "reason": "Flash-attention generation bug vs a stale warning in paged generate(); not the same code path." }, { - "left": "issue:43873", - "right": "issue:44493", + "left": "issue:43927", + "right": "issue:45698", "accept": false, - "reason": "Quantization/offloading behavior vs position-id key mismatch; not the same bug." + "reason": "Both involve save/load behavior, but one is custom token IDs and the other is loading the wrong custom module." }, { - "left": "issue:43873", - "right": "issue:45229", + "left": "issue:45372", + "right": "issue:45588", "accept": false, - "reason": "Offloading/quantization issue is unrelated to Gemma4 multi-GPU OOM." + "reason": "Gemma4 processor import failure vs flash_attention.py crashing on s_aux=None; unrelated modules." }, { - "left": "issue:29942", - "right": "issue:35141", + "left": "issue:37428", + "right": "issue:44987", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to token-embedding reinitialization in post_init." + "reason": "Missing flash-support helper import vs a specific model load failure for physical-intelligence/fast." }, { - "left": "issue:43278", - "right": "issue:43329", + "left": "issue:43606", + "right": "issue:44857", "accept": false, - "reason": "Embedding dtype drift in eval is unrelated to the multimodal token-counting video-branch bug." + "reason": "CPU offload device mismatch is unrelated to the float16 AMP CUDA crash in LwDetrImageLoss." }, { - "left": "issue:32090", - "right": "issue:33357", + "left": "issue:44857", + "right": "issue:45072", "accept": false, - "reason": "Trainer `_gpu_broadcast_one` NoneType error is unrelated to the MacOS bus error in CLIP." + "reason": "Both mention dtype issues, but they affect different models and different failure modes." }, { - "left": "issue:45372", - "right": "issue:45406", + "left": "issue:43701", + "right": "issue:45083", "accept": false, - "reason": "Both mention Gemma4, but one is processor import/loading and the other is missing `_tokenizer` in serve." + "reason": "Checkpoint key mismatch is unrelated to the qwen3_omni_moe helper-function bug." }, { - "left": "issue:45372", - "right": "issue:45459", + "left": "issue:43976", + "right": "issue:45440", "accept": false, - "reason": "Gemma4 processor import failure is unrelated to protobuf exception masking tokenizer errors." + "reason": "Python version compatibility regression vs native DeepseekV3MoE divergence; different problem classes." }, { - "left": "issue:41720", - "right": "issue:45229", + "left": "issue:45245", + "right": "issue:45563", "accept": false, - "reason": "Qwen3 auto-device cuda assert is unrelated to Gemma4 CUDA OOM." + "reason": "Category-count limit error is not the same as the stale num_return_sequences warning." }, { - "left": "issue:43065", - "right": "issue:43994", + "left": "issue:43412", + "right": "issue:45588", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to Siglip2 nonsensical outputs." + "reason": "Export/tracing guard and erfinv support issue vs an attention-path AttributeError; unrelated." }, { - "left": "issue:44448", - "right": "issue:44466", + "left": "issue:43550", + "right": "issue:45083", "accept": false, - "reason": "Pegasus output drift across v4/v5 is a different bug from `lm_head.weight` serialization/tied-weights inconsistency." + "reason": "torch.compile/SDPA failure in Bamba is unrelated to qwen3_omni_moe length computation." }, { - "left": "issue:43065", - "right": "issue:45412", + "left": "issue:43873", + "right": "issue:44493", "accept": false, - "reason": "Sam3PixelDecoder dummy layer issue is unrelated to RT-DETR memory-release behavior." + "reason": "Quantization/offloading behavior is a different bug from unexpected position-id keys." }, { - "left": "issue:25251", - "right": "issue:32090", + "left": "issue:43873", + "right": "issue:45229", "accept": false, - "reason": "Pipeline `top_k` nesting bug is unrelated to Trainer `_gpu_broadcast_one` NoneType error." + "reason": "Offloading/quantization issue vs Gemma4 multi-GPU OOM; not the same defect." }, { - "left": "issue:44451", - "right": "issue:45520", + "left": "issue:29942", + "right": "issue:35141", "accept": false, - "reason": "ScandiBERT loading failure is unrelated to the Python 3.13 `flash_attn` import KeyError." + "reason": "Flash Attention 2 test failures are unrelated to embedding reinitialization after resize." }, { - "left": "issue:43653", - "right": "issue:44743", + "left": "issue:43278", + "right": "issue:43329", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug is unrelated to qwen3_5 recurrent-state reset." + "reason": "Embedding dtype drift in eval vs undefined variables in video multimodal token counting." }, { - "left": "issue:44857", - "right": "issue:45588", + "left": "issue:45372", + "right": "issue:45459", "accept": false, - "reason": "LwDetrImageLoss AMP crash is unrelated to the flash_attention `s_aux=None` crash." + "reason": "Gemma4 processor import breakage is unrelated to tokenizer error masking when protobuf is absent." }, { - "left": "issue:43421", - "right": "issue:45491", + "left": "issue:45372", + "right": "issue:45406", "accept": false, - "reason": "Runtime post-processor update bug is unrelated to Gemma3 NaN embeddings with mixed-length batches." + "reason": "ImportError for ReasoningEffort is a different failure than Gemma4Processor missing _tokenizer in serve." }, { - "left": "issue:44186", - "right": "issue:45198", + "left": "issue:32090", + "right": "issue:33357", "accept": false, - "reason": "LayoutLMv2 NER padding/truncation crash is unrelated to Wav2Vec2 save/tokenization failures." + "reason": "Trainer _gpu_broadcast_one NoneType error vs MacOS bus error in CLIP loading; different code paths." }, { - "left": "issue:43278", - "right": "issue:43381", + "left": "issue:41720", + "right": "issue:45229", "accept": false, - "reason": "Embedding dtype change in eval is unrelated to gradient checkpointing being disallowed in eval mode." + "reason": "Qwen3 auto device-mapping cudaErrorAssert is not the same as Gemma4 multi-GPU OOM." }, { - "left": "issue:43994", - "right": "issue:45412", + "left": "issue:43065", + "right": "issue:43994", "accept": false, - "reason": "Siglip2 incorrect outputs are unrelated to RT-DETR memory not being released." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to ScandiBERT loading failure." }, { - "left": "issue:42915", - "right": "issue:42994", + "left": "issue:43065", + "right": "issue:45412", "accept": false, - "reason": "MoE weight conversion during accelerate+deepspeed loading is unrelated to quantized model saving." + "reason": "A placeholder Conv2d bug is unrelated to RT-DETR memory not being released." }, { - "left": "issue:44361", - "right": "issue:45198", + "left": "issue:44448", + "right": "issue:44466", "accept": false, - "reason": "MLukeTokenizer AttributeError is unrelated to Wav2Vec2 save_pretrained/tokenization failures." + "reason": "Output drift between v4/v5 Pegasus is not the same as lm_head.weight serialization depending on device." }, { - "left": "issue:43756", - "right": "issue:45440", + "left": "issue:44451", + "right": "issue:45520", "accept": false, - "reason": "Smollm3 RoPE-layer omission is unrelated to DeepSeekV3 implementation divergence." + "reason": "ScandiBERT loading failure is unrelated to the Python 3.13 flash_attn KeyError." }, { - "left": "issue:44871", - "right": "issue:44977", + "left": "issue:25251", + "right": "issue:32090", "accept": false, - "reason": "Gemma3 eos_token_id mismatch is unrelated to Qwen3.5 flash-attention generation issues." + "reason": "Pipeline top_k nesting issue is unrelated to the Trainer broadcast TypeError." }, { - "left": "issue:44265", - "right": "issue:45198", + "left": "issue:43653", + "right": "issue:44743", "accept": false, - "reason": "torch.export.export with `torch_compilable_check` is unrelated to Wav2Vec2 save/tokenization failure." + "reason": "BigBirdTokenizer special-token decoding bug is unrelated to qwen3_5 recurrent-state reset behavior." }, { - "left": "issue:44206", - "right": "issue:44479", + "left": "issue:43421", + "right": "issue:45491", "accept": false, - "reason": "LasrFeatureExtractor `center` regression is unrelated to the Qwen VL video-input regression." + "reason": "Runtime post-processor update bug is unrelated to Gemma3 NaN embeddings with sliding-window attention." }, { - "left": "issue:43257", - "right": "issue:43278", + "left": "issue:44186", + "right": "issue:45198", "accept": false, - "reason": "Qwen3 MoE conversion/loading bug is unrelated to the embedding dtype change in evaluation." + "reason": "LayoutLMv2 tokenizer padding/truncation crash and Wav2Vec2 save_pretrained/tokenization failure are different bugs." }, { - "left": "issue:29942", - "right": "issue:30064", + "left": "issue:43994", + "right": "issue:45412", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to void segmentation-map processing." + "reason": "ScandiBERT loading failure is unrelated to RT-DETR memory-leak behavior." } ] }, @@ -7807,10 +7656,11 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:29127", "issue:29942", + "issue:30064", "issue:32090", "issue:33290", "issue:33357", @@ -7819,58 +7669,66 @@ "issue:36010", "issue:39692", "issue:42915", + "issue:42994", "issue:43232", - "issue:43262", + "issue:43257", + "issue:43278", "issue:43334", "issue:43381", "issue:43388", - "issue:43441", - "issue:43452", "issue:43526", "issue:43531", "issue:43653", "issue:43701", - "issue:43723", + "issue:43756", "issue:43761", "issue:43824", "issue:43931", "issue:44079", "issue:44186", "issue:44206", + "issue:44265", "issue:44279", + "issue:44361", "issue:44442", "issue:44464", + "issue:44479", "issue:44610", - "issue:44743", - "issue:44811", - "issue:44857", + "issue:44871", "issue:44977", - "issue:45003", "issue:45072", "issue:45103", + "issue:45198", "issue:45229", "issue:45245", "issue:45372", "issue:45412", "issue:45440", - "issue:45479", "issue:45491", - "issue:45588", "issue:45593" ], "soft_pairs": [ + "issue:42915|issue:42994", + "issue:43278|issue:43381", + "issue:43756|issue:45440", + "issue:44361|issue:45198", + "issue:44206|issue:44479", + "issue:44265|issue:45198", + "issue:44871|issue:44977", + "issue:43257|issue:43278", "issue:44610|issue:45412", + "issue:45440|issue:45593", + "issue:29942|issue:30064", "issue:43824|issue:44279", "issue:45103|issue:45372", - "issue:45440|issue:45593", "issue:33357|issue:39692", "issue:29127|issue:34567", "issue:32090|issue:33290", - "issue:45072|issue:45229", "issue:43334|issue:43531", - "issue:43381|issue:43701", - "issue:43653|issue:44442", + "issue:45072|issue:45229", "issue:43526|issue:44186", + "issue:43653|issue:44442", + "issue:43381|issue:43701", "issue:44464|issue:45491", "issue:44977|issue:45245", "issue:29942|issue:36010", @@ -7878,219 +7736,210 @@ "issue:42915|issue:43931", "issue:29127|issue:35141", "issue:43232|issue:43388", - "issue:43761|issue:44079", - "issue:43262|issue:44206", - "issue:43526|issue:44857", - "issue:43441|issue:45588", - "issue:44442|issue:45072", - "issue:43723|issue:45479", - "issue:45003|issue:45440", - "issue:43526|issue:44610", - "issue:44743|issue:44811", - "issue:43452|issue:45588" + "issue:43761|issue:44079" ], - "prepared_review_unit_hash": "f0a039896315f1523875d3ecddd89027e86accb92d71c1c996a318e6a882f8fe", + "prepared_review_unit_hash": "e41c9e1678e739bca6841a1f292400d3f14262fbbd77952f8a6d385b819a2c5b", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36520, - "estimated_input_tokens": 9130, - "estimated_eval_tokens": 18516 + "serialized_chars": 36448, + "estimated_input_tokens": 9112, + "estimated_eval_tokens": 18480 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft edges link unrelated bugs across different models, utilities, and failure modes. I reject all pairings as duplicates and do not see a single coherent canonical issue/PR for the set.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "This cluster looks noisy and mostly heterogeneous: tokenizers, processors, generation, quantization, evaluation, and trainer bugs are all mixed together. None of the soft-edge pairs look like true duplicates, so all soft edges are rejected.", + "confidence": 0.24, + "canonical_issue_reason": "Issue 36010 is the broadest, most core-library regression here (a central import failure in the generation API) and is clearer as a representative issue than the many model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 36010 has the strongest combination of breadth, clarity, and likely downstream impact, making it the best single issue representative of this set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both are quantization-related, but one is an FP8 config failure and the other is quantized-model saving; different bugs." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "BF16/FP32 eval dtype drift and a gradient-checkpointing eval restriction are separate code-path issues." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Both concern model parity, but they affect different models and different mechanisms, so not the same bug." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer failures in different models; too broad to be duplicates." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Processor center-arg crash and video-input regression are unrelated processor bugs." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to save_pretrained/tokenization failure." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Eos-token config mismatch and flash-attention generation failure are distinct bugs." + }, + { + "left": "issue:43257", + "right": "issue:43278", + "accept": false, + "reason": "Qwen3 MOE weight conversion and embedding dtype changes are different issues." + }, { "left": "issue:44610", "right": "issue:45412", "accept": false, - "reason": "Both mention memory-related failure, but one is a processor/model input-size mismatch and the other is a model deletion/GC leak." + "reason": "Input-size mismatch and a memory-leak/deallocation problem are not the same underlying bug." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepSeek implementation divergence and missing auxiliary losses are separate model/training issues." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Failing Flash Attention tests and void segmentation map processing target different failures." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "One is a missing Qwen2.5-VL import, the other is a vague dependency problem; no shared code-path or concrete fix." + "reason": "A specific Qwen2.5-VL import error is not the same as a generic dependency issue." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "Auto-docstring annotation handling and a mistral_common import error are unrelated failures." - }, - { - "left": "issue:45440", - "right": "issue:45593", - "accept": false, - "reason": "Different models and bugs: DeepseekV3MoE implementation divergence vs D-FINE auxiliary-loss behavior." + "reason": "Auto-docstring annotation handling and a missing ReasoningEffort import are unrelated." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "Both are multimodal issues, but one is a MacOS bus error on a CLIP model and the other is a SigLIP2 docs-example mismatch/quantization failure." + "reason": "MacOS bus error loading CLIP and SigLIP2 docs/example errors are different failure modes." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 box validation messaging and TrainerState token-count tracking are unrelated." + "reason": "LayoutLMv3 error messaging and TrainerState token counting are unrelated." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "NoneType broadcast in Trainer and Adafactor OOM in DeepSpeed are different trainer/backend problems." + "reason": "NoneType broadcast errors and deepspeed OOM with Adafactor are different trainer/optimizer problems." }, { - "left": "issue:45072", - "right": "issue:45229", + "left": "issue:43334", + "right": "issue:43531", "accept": false, - "reason": "Dtype mismatch in specific models is not the same issue as Gemma4 multi-GPU OOM." + "reason": "Missing pad_token_id on Qwen3-VL and sliding-window behavior in Qwen3-MoE are separate model bugs." }, { - "left": "issue:43334", - "right": "issue:43531", + "left": "issue:45072", + "right": "issue:45229", "accept": false, - "reason": "Qwen3-VL load-time pad_token_id attribute error and Qwen3-MoE sliding-window behavior are separate code paths." + "reason": "bfloat16 dtype mismatches and multi-GPU CUDA OOM are not the same underlying issue." }, { - "left": "issue:43381", - "right": "issue:43701", + "left": "issue:43526", + "right": "issue:44186", "accept": false, - "reason": "Gradient checkpointing in eval mode and resume-from-checkpoint key mismatch are distinct training issues." + "reason": "BeitImageProcessorFast label reduction and LayoutLMv2Tokenizer NER/padding crashes are unrelated." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer special-token registration and FastSpeech2ConformerTokenizer loading failure are different tokenizer bugs." + "reason": "BigBirdTokenizer special-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." }, { - "left": "issue:43526", - "right": "issue:44186", + "left": "issue:43381", + "right": "issue:43701", "accept": false, - "reason": "BEiT reduce_labels behavior and LayoutLMv2Tokenizer NER/padding crashes are unrelated processor/tokenizer bugs." + "reason": "Eval-mode checkpointing limitation and resume_from_checkpoint key mismatch do not share a concrete code path." }, { "left": "issue:44464", "right": "issue:45491", "accept": false, - "reason": "Chunked generation with compiled forward and Gemma3 NaN embeddings from sliding-window attention are different failures." + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaN embeddings are distinct generation/attention issues." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Flash-attention generation behavior and the 2^24 categories limit are entirely unrelated." + "reason": "Flash-attention generation breakage and a category-count runtime limit are unrelated." }, { "left": "issue:29942", "right": "issue:36010", "accept": false, - "reason": "Flash Attention 2 test failures do not match a GenerationMixin import error." + "reason": "Flash Attention test failures and GenerationMixin import failure are different regressions." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "LayoutLMv3 error messaging and a CLIP bus error on MacOS are unrelated issues." + "reason": "LayoutLMv3 error clarity and CLIP bus error on macOS are not duplicates." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "Qwen3Moe FP8 failure and Qwen3-VL weight-shape mismatch are different model-loading problems." + "reason": "FP8 config failure in Qwen3MoE and Qwen3-VL weight-shape mismatch are separate model-loading bugs." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "Different subsystems: LayoutLMv3 box handling vs embedding reinitialization after resize." + "reason": "LayoutLMv3 error messaging and embedding reinitialization after resize are unrelated." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation kwargs syncing and metric-gathering label truncation are separate bugs." + "reason": "Generation kwargs update after sync_gpus and metric-gather label dropping are different trainer internals." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "CLIPVisionModel missing hidden_states and ModelOutput key assignment are both output-related, but not the same underlying defect." - }, - { - "left": "issue:43262", - "right": "issue:44206", - "accept": false, - "reason": "Audio chat-template sampling-rate default and a feature extractor center-arg crash are unrelated." - }, - { - "left": "issue:43526", - "right": "issue:44857", - "accept": false, - "reason": "BEiT label reduction and LwDetrImageLoss AMP/CUDA crash have no shared code path." - }, - { - "left": "issue:43441", - "right": "issue:45588", - "accept": false, - "reason": "Both involve flash attention, but they are different failures in different models/code paths; too broad to merge." - }, - { - "left": "issue:44442", - "right": "issue:45072", - "accept": false, - "reason": "Tokenizer loading failure and bfloat16 dtype mismatch are unrelated." - }, - { - "left": "issue:43723", - "right": "issue:45479", - "accept": false, - "reason": "AutoTokenizer loading in v5 and zero-loss single-label classification are unrelated bugs." - }, - { - "left": "issue:45003", - "right": "issue:45440", - "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils and DeepseekV3MoE divergence are different layers of the stack." - }, - { - "left": "issue:43526", - "right": "issue:44610", - "accept": false, - "reason": "BEiT reduce_labels and OmDet-Turbo input-size mismatch are different processor bugs." - }, - { - "left": "issue:44743", - "right": "issue:44811", - "accept": false, - "reason": "Recurrent-state reset with cache and Whisper batch_decode skip_special_tokens are unrelated behaviors." - }, - { - "left": "issue:43452", - "right": "issue:45588", - "accept": false, - "reason": "gguf_file breaking tokenizer/model loading and a flash-attention AttributeError are not the same issue." + "reason": "CLIPVisionModel hidden_states handling and ModelOutput key assignment are separate issues." } ] }, @@ -8112,28 +7961,27 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:33290", "issue:34567", "issue:34689", - "issue:35141", - "issue:36010", "issue:39692", "issue:41628", "issue:42907", + "issue:43262", "issue:43316", - "issue:43381", + "issue:43441", + "issue:43452", "issue:43493", "issue:43525", + "issue:43526", "issue:43575", "issue:43611", - "issue:43643", "issue:43646", "issue:43673", "issue:43701", - "issue:43720", - "issue:43749", + "issue:43723", "issue:43761", "issue:43827", "issue:43873", @@ -8141,41 +7989,51 @@ "issue:43994", "issue:44077", "issue:44188", - "issue:44190", "issue:44206", "issue:44387", "issue:44442", "issue:44479", - "issue:44496", + "issue:44610", "issue:44743", - "issue:44871", - "issue:44877", + "issue:44811", + "issue:44857", "issue:44898", - "issue:44913", "issue:44977", + "issue:45003", + "issue:45072", "issue:45216", "issue:45290", "issue:45372", + "issue:45440", + "issue:45479", "issue:45491", "issue:45584", "issue:45588", "issue:45600", - "issue:45656", "issue:45698" ], "soft_pairs": [ + "issue:43262|issue:44206", + "issue:43441|issue:45588", + "issue:43526|issue:44857", + "issue:45003|issue:45440", + "issue:43723|issue:45479", + "issue:44442|issue:45072", + "issue:43526|issue:44610", + "issue:44743|issue:44811", + "issue:43452|issue:45588", "issue:43316|issue:43525", - "issue:43646|issue:44077", "issue:43761|issue:44188", - "issue:45588|issue:45600", "issue:44479|issue:45290", - "issue:43575|issue:43701", + "issue:43646|issue:44077", + "issue:45588|issue:45600", "issue:44977|issue:45584", "issue:43931|issue:45216", + "issue:43575|issue:43701", "issue:42907|issue:45698", - "issue:43575|issue:44387", "issue:39692|issue:45491", "issue:43994|issue:44206", + "issue:43575|issue:44387", "issue:43873|issue:45588", "issue:34689|issue:43611", "issue:43673|issue:44743", @@ -8183,233 +8041,210 @@ "issue:33290|issue:34567", "issue:43493|issue:43827", "issue:41628|issue:45372", - "issue:43381|issue:43873", - "issue:44077|issue:44479", - "issue:33290|issue:34689", - "issue:33290|issue:35141", - "issue:43643|issue:44913", - "issue:44190|issue:44496", - "issue:43749|issue:43931", - "issue:43720|issue:44442", - "issue:43873|issue:44871", - "issue:33290|issue:45656", - "issue:34689|issue:36010", - "issue:43316|issue:44877" + "issue:44077|issue:44479" ], - "prepared_review_unit_hash": "e029fc329dda6967290b1a9e5fb145e3de1927be47b30c36a79ef5c93f622f18", + "prepared_review_unit_hash": "2ad312bdab7ffaa700a8c2c08312469345e5b334135c687cf615c8118e53d7ba", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37063, - "estimated_input_tokens": 9266, - "estimated_eval_tokens": 18788 + "soft_pair_count": 29, + "serialized_chars": 36464, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous and the soft pairs are mostly high-similarity false positives rather than true duplicates. No PRs are present. If an issue must be used as the cluster anchor, issue 43749 is the most broadly scoped and active loading/regression report.", - "confidence": 0.74, - "canonical_issue_reason": "Issue 43749 is the most central-looking anchor here: it is broad, active, open, and has inbound references, so it best fits a cluster representative even though the overall set is not a true duplicate cluster.", + "summary": "This is a noisy similarity cluster, not a duplicate set: every soft pair describes a different bug, regression, or docs issue, so all soft edges should be rejected.", + "confidence": 0.91, + "canonical_issue_reason": "No canonical issue stands out; the items span unrelated bugs across audio, vision, tokenizer loading, quantization, checkpointing, and model internals, so they should not be collapsed into one duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43749 is the best issue to keep as the representative because it has the strongest cross-thread activity and the broadest regression/loading scope among the listed issues.", + "best_issue_reason": "issue:45584 is the strongest standalone report: it is open, concrete, and describes a specific Whisper generation regression after `align_special_tokens` with a clear failure mode.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43316", - "right": "issue:43525", - "accept": false, - "reason": "Both are config/API complaints, but they concern different models and different missing fields; not the same bug." - }, - { - "left": "issue:43646", - "right": "issue:44077", - "accept": false, - "reason": "Both mention initialization/post_init, but one is a broad Transformers 5.0 custom init regression while the other is a model-specific restriction." - }, - { - "left": "issue:43761", - "right": "issue:44188", + "left": "issue:43262", + "right": "issue:44206", "accept": false, - "reason": "Unrelated code paths: CLIP vision hidden_states regression vs attention-kernel divergence under torch.compile." + "reason": "Different subsystems and failures: audio chat-template sampling-rate default vs a feature-extractor `center`-argument crash." }, { - "left": "issue:45588", - "right": "issue:45600", + "left": "issue:43441", + "right": "issue:45588", "accept": false, - "reason": "Different failures entirely: flash_attention s_aux handling vs stale auto_mappings references to removed configs." + "reason": "Both mention flash attention, but they are different bugs in different code paths/models: Ministral-3 FA failure vs `s_aux=None` AttributeError in flash_attention integration." }, { - "left": "issue:44479", - "right": "issue:45290", + "left": "issue:43526", + "right": "issue:44857", "accept": false, - "reason": "Different subsystems and symptoms: video input regression for Qwen VL models vs chat-template tool-call handling." + "reason": "Unrelated bugs: `reduce_labels` returning one label in BEiT fast processor vs LwDetrImageLoss crashing under float16 AMP on CUDA." }, { - "left": "issue:43575", - "right": "issue:43701", + "left": "issue:45003", + "right": "issue:45440", "accept": false, - "reason": "OOM on tensor-parallel model load is not the same as a resume_from_checkpoint key mismatch." + "reason": "These are different model/framework issues: unsafe `sys.modules` access in modeling_utils vs DeepSeekV3MoE divergence from the remote implementation." }, { - "left": "issue:44977", - "right": "issue:45584", + "left": "issue:43723", + "right": "issue:45479", "accept": false, - "reason": "Both are generation bugs, but they affect different models and different failure modes." + "reason": "Tokenizer loading regression vs a sequence-classification loss bug; no shared code-path." }, { - "left": "issue:43931", - "right": "issue:45216", + "left": "issue:44442", + "right": "issue:45072", "accept": false, - "reason": "Both involve Qwen variants, but one is a weight-shape load error and the other is a bad save_pretrained regression." + "reason": "FastSpeech2Conformer tokenizer loading failure vs bfloat16 dtype mismatches in inference; different problems." }, { - "left": "issue:42907", - "right": "issue:45698", + "left": "issue:43526", + "right": "issue:44610", "accept": false, - "reason": "Both touch save/load, but one is about saving dequantized models and the other about loading the wrong custom module." + "reason": "BEiT label reduction bug and OmDet-Turbo processor image-size mismatch are unrelated." }, { - "left": "issue:43575", - "right": "issue:44387", + "left": "issue:44743", + "right": "issue:44811", "accept": false, - "reason": "Both are OOM reports, but the trigger is different: tensor-parallel loading vs int4 quantization memory growth." + "reason": "Recurrent state reset with cache vs Whisper `batch_decode()` ignoring `skip_special_tokens`; distinct behaviors." }, { - "left": "issue:39692", - "right": "issue:45491", + "left": "issue:43452", + "right": "issue:45588", "accept": false, - "reason": "Documentation/example errors for SigLIP2 are unrelated to Gemma3 NaN embeddings on GPU." + "reason": "GGUF tokenizer/model loading breakage is not the same as the flash-attention `s_aux` AttributeError." }, { - "left": "issue:43994", - "right": "issue:44206", + "left": "issue:43316", + "right": "issue:43525", "accept": false, - "reason": "Different root causes: nonsensical SigLIP2 outputs vs a LasrFeatureExtractor argument incompatibility." + "reason": "API discrepancy in `Gemma3TextConfig` vs missing `pad_token_id` on `Llama4Config`; different config regressions." }, { - "left": "issue:43873", - "right": "issue:45588", + "left": "issue:43761", + "right": "issue:44188", "accept": false, - "reason": "Quantization/offloading behavior and a flash_attention AttributeError are not the same code-path problem." + "reason": "CLIPVisionModel hidden-states regression and torch.compile attention-kernel divergence are separate bugs." }, { - "left": "issue:34689", - "right": "issue:43611", + "left": "issue:44479", + "right": "issue:45290", "accept": false, - "reason": "Both are model-loading regressions, but they describe different failures and different model families." + "reason": "Qwen video-input regression vs chat-template tool-call crash; different input pipelines and failure modes." }, { - "left": "issue:43673", - "right": "issue:44743", + "left": "issue:43646", + "right": "issue:44077", "accept": false, - "reason": "Generation cache issues, but one is a v5 chunked_prefill cache import/regression and the other is Qwen3.5 recurrent-state resetting." + "reason": "General custom model initialization regression is not the same as forbidding optional `post_init` in patchtsmixer." }, { - "left": "issue:44442", - "right": "issue:44898", + "left": "issue:45588", + "right": "issue:45600", "accept": false, - "reason": "Tokenizer-loading failure vs image classification resizing failure; no common underlying bug." + "reason": "Flash-attention `AttributeError` and stale `auto_mappings.py` references are unrelated CI/runtime issues." }, { - "left": "issue:33290", - "right": "issue:34567", + "left": "issue:44977", + "right": "issue:45584", "accept": false, - "reason": "Unrelated training issues: deepspeed/adafactor OOM vs TrainerState token counter not updating." + "reason": "Qwen3.5 flash-attention generation bug vs Whisper empty-transcription failure; different models and code paths." }, { - "left": "issue:43493", - "right": "issue:43827", + "left": "issue:43931", + "right": "issue:45216", "accept": false, - "reason": "SigLIP2 implementation mismatch and pipeline-docs breakage are different issues." + "reason": "Model weight-shape mismatch on load vs incorrect checkpoint saving in Qwen3.5; related theme, but not the same underlying bug." }, { - "left": "issue:41628", - "right": "issue:45372", + "left": "issue:43575", + "right": "issue:43701", "accept": false, - "reason": "Both are import errors, but they involve different symbols, packages, and loading paths." + "reason": "TP load OOM and resume-from-checkpoint key mismatch are distinct training-time issues." }, { - "left": "issue:43381", - "right": "issue:43873", + "left": "issue:42907", + "right": "issue:45698", "accept": false, - "reason": "Gradient-checkpointing eval-mode behavior is unrelated to quantization offloading." + "reason": "Saving dequantized Ministral/Devstral models is not the same as loading the wrong custom module after `save_pretrained`." }, { - "left": "issue:44077", - "right": "issue:44479", + "left": "issue:39692", + "right": "issue:45491", "accept": false, - "reason": "A post_init constraint on patchtsmixer is unrelated to Qwen VL video-input regression." + "reason": "SigLIP2 docs-example errors and Gemma3 NaN embeddings are unrelated model bugs." }, { - "left": "issue:33290", - "right": "issue:34689", + "left": "issue:43994", + "right": "issue:44206", "accept": false, - "reason": "OOM during deepspeed optimizer use is not the same as Llama model loading failure." + "reason": "SigLIP2 nonsensical outputs vs a feature extractor crash from unsupported `center`; different layers of the stack." }, { - "left": "issue:33290", - "right": "issue:35141", + "left": "issue:43575", + "right": "issue:44387", "accept": false, - "reason": "Adafactor/deepspeed OOM and embedding reinitialization after resize are different bugs." + "reason": "Both are OOMs, but one is TP model loading and the other is int4 quantization reserved-memory growth; not the same concrete bug." }, { - "left": "issue:43643", - "right": "issue:44913", + "left": "issue:43873", + "right": "issue:45588", "accept": false, - "reason": "trust_remote_code config field loss and GPTNeoX rotary_pct reload drift are unrelated config issues." + "reason": "Quantization/offloading behavior and a flash-attention AttributeError are different problems." }, { - "left": "issue:44190", - "right": "issue:44496", + "left": "issue:34689", + "right": "issue:43611", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to unrecognized-model config loading." + "reason": "Llama 3.2 vision loading breakage vs base_model_prefix handling regression; too different to merge." }, { - "left": "issue:43749", - "right": "issue:43931", + "left": "issue:43673", + "right": "issue:44743", "accept": false, - "reason": "Both are load-time problems, but one is FSDP CPU-efficient loading and the other is a Qwen3-VL weight-shape mismatch." + "reason": "Both involve cache/state, but one is missing GenerationMixin cache in v5 chunked_prefill and the other is recurrent-state reset; separate bugs." }, { - "left": "issue:43720", - "right": "issue:44442", + "left": "issue:44442", + "right": "issue:44898", "accept": false, - "reason": "BitNet accelerate-loading unpacking is unrelated to tokenizer loading." + "reason": "Tokenizer loading failure vs Perceiver non-default resolution failure; no shared underlying defect." }, { - "left": "issue:43873", - "right": "issue:44871", + "left": "issue:33290", + "right": "issue:34567", "accept": false, - "reason": "Quantization/offloading behavior and Gemma-3 eos_token_id config mismatch are different issues." + "reason": "AdaFactor/DeepSpeed OOM is unrelated to `num_input_tokens_seen` not updating." }, { - "left": "issue:33290", - "right": "issue:45656", + "left": "issue:43493", + "right": "issue:43827", "accept": false, - "reason": "Different DeepSpeed symptoms: optimizer OOM vs optimizer.step being called twice." + "reason": "SigLIP2 HF-vs-JAX implementation discrepancy is unrelated to docs still referencing `pipeline()`." }, { - "left": "issue:34689", - "right": "issue:36010", + "left": "issue:41628", + "right": "issue:45372", "accept": false, - "reason": "Different import/load failures: Llama model loading vs GenerationMixin import path breakage." + "reason": "Missing `AutoImageProcessor` import and `ReasoningEffort` import failure in Gemma 4 processor loading are different import regressions." }, { - "left": "issue:43316", - "right": "issue:44877", + "left": "issue:44077", + "right": "issue:44479", "accept": false, - "reason": "Config API discrepancy for Gemma3TextConfig is unrelated to strict config loading for granite_speech." + "reason": "`post_init` policy for patchtsmixer has nothing to do with Qwen video-input regression." } ] }, @@ -8431,12 +8266,16 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:16998", "issue:22355", "issue:25251", "issue:28282", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:36010", "issue:41950", "issue:42175", "issue:43065", @@ -8446,261 +8285,263 @@ "issue:43388", "issue:43404", "issue:43441", - "issue:43452", "issue:43475", "issue:43540", - "issue:43550", - "issue:43606", - "issue:43646", + "issue:43643", "issue:43688", - "issue:43701", "issue:43716", + "issue:43720", "issue:43742", + "issue:43749", "issue:43761", "issue:43825", "issue:43856", + "issue:43873", "issue:43931", - "issue:44060", - "issue:44077", "issue:44162", - "issue:44242", + "issue:44190", + "issue:44442", "issue:44479", "issue:44496", "issue:44514", - "issue:44534", "issue:44617", - "issue:44734", "issue:44792", "issue:44855", "issue:44871", + "issue:44877", "issue:44913", - "issue:44936", "issue:45127", "issue:45137", "issue:45200", - "issue:45406", "issue:45584", - "issue:45698" + "issue:45656" ], "soft_pairs": [ + "issue:43381|issue:43873", + "issue:33290|issue:34689", + "issue:33290|issue:35141", + "issue:43643|issue:44913", + "issue:44190|issue:44496", + "issue:43749|issue:43931", + "issue:43720|issue:44442", + "issue:43873|issue:44871", + "issue:33290|issue:45656", + "issue:43316|issue:44877", "issue:44162|issue:44792", + "issue:34689|issue:36010", "issue:44855|issue:45137", - "issue:43381|issue:43856", "issue:43540|issue:44514", + "issue:43381|issue:43856", "issue:42175|issue:43316", "issue:43404|issue:43441", - "issue:43388|issue:43688", "issue:41950|issue:43475", "issue:43316|issue:43716", - "issue:43381|issue:43742", + "issue:43388|issue:43688", "issue:16998|issue:22355", + "issue:43381|issue:43742", "issue:25251|issue:28282", "issue:43295|issue:43825", + "issue:43761|issue:44479", "issue:43065|issue:44617", "issue:44871|issue:45584", - "issue:43761|issue:44479", - "issue:45127|issue:45200", - "issue:44077|issue:44534", - "issue:44936|issue:45698", - "issue:43295|issue:44871", - "issue:44734|issue:45406", - "issue:43550|issue:43701", - "issue:43606|issue:43701", - "issue:43646|issue:44913", - "issue:43452|issue:44496", - "issue:43761|issue:44242", - "issue:43931|issue:44060" + "issue:45127|issue:45200" ], - "prepared_review_unit_hash": "49c8292659c5a3e5985f932830cd32be43bd8129667f34ce9d1368ecbdb6d8b3", + "prepared_review_unit_hash": "ec5d8858a0c1fde3cd0c7dea8af68b64fa03749b52702516ef51a2b44399736a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35708, - "estimated_input_tokens": 8927, - "estimated_eval_tokens": 18110 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35447, + "estimated_input_tokens": 8862, + "estimated_eval_tokens": 17980 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is a loose collection of unrelated Transformers issues spanning model loading, multimodal processors, training/runtime regressions, and packaging/import problems. I did not find any soft pair that looks like the same underlying bug or change.", - "confidence": 0.91, - "canonical_issue_reason": "No true duplicate center stands out because the issues are largely unrelated. If one issue must serve as the anchor, issue 44534 is the broadest framework-level regression report and is more representative than the model-specific edge cases.", + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs across model loading, configs, pipelines, distributed training, tokenizer behavior, and model-specific regressions. None of the soft pairs look like true duplicates, so all candidate merges should be rejected.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the cluster is a grab-bag of unrelated regressions and feature/behavior bugs, so nothing cleanly represents the whole set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44534 is the best overall representative because it describes a general v5 regression with potentially wide impact, whereas most other items are narrow, model-specific, or packaging/import quirks.", + "best_issue_reason": "No issue cleanly subsumes the others; the closest candidates are still model- or subsystem-specific and not representative of the cluster as a whole.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44162", - "right": "issue:44792", + "left": "issue:43381", + "right": "issue:43873", "accept": false, - "reason": "Different models and failure modes: ESM2 workflow breakage vs a janus image-generation test failure." + "reason": "Gradient checkpointing in eval mode and offloading with quantization are different training-path bugs." }, { - "left": "issue:44855", - "right": "issue:45137", + "left": "issue:33290", + "right": "issue:34689", "accept": false, - "reason": "Import-time IndentationError in DebertaV2Model is unrelated to the DeepSpeed ZeRO3 deque pop error." + "reason": "AdaFactor/deepspeed OOM is unrelated to the Llama 3.2 Vision model-loading regression." }, { - "left": "issue:43381", - "right": "issue:43856", + "left": "issue:33290", + "right": "issue:35141", "accept": false, - "reason": "Gradient checkpointing in eval mode is a runtime-mode bug; Qwen3 MoE memory usage is a training efficiency issue." + "reason": "Optimizer OOM and embedding reinitialization after resize are distinct bugs in different code paths." }, { - "left": "issue:43540", - "right": "issue:44514", + "left": "issue:43643", + "right": "issue:44913", "accept": false, - "reason": "Both are multimodal, but one is a video-input validation failure in Qwen3OmniMoe and the other is a batched apply_chat_template padding bug in Qwen2_5_VL." + "reason": "Remote-code config field loss and GPTNeoX rotary_pct reload mismatch are separate config-handling issues." }, { - "left": "issue:42175", - "right": "issue:43316", + "left": "issue:44190", + "right": "issue:44496", "accept": false, - "reason": "Packaging/dependency backend omission is unrelated to Gemma3TextConfig API inconsistency." + "reason": "Local dataset loading failure in an example script is unrelated to an unrecognized model/config error." }, { - "left": "issue:43404", - "right": "issue:43441", + "left": "issue:43749", + "right": "issue:43931", "accept": false, - "reason": "Mistral3 weight tying and Ministral-3 FlashAttention failure are different concrete code-path problems." + "reason": "FSDP CPU RAM efficient loading and Qwen3-VL weight-shape mismatch are different loading failures." }, { - "left": "issue:43388", - "right": "issue:43688", + "left": "issue:43720", + "right": "issue:44442", "accept": false, - "reason": "Metric gathering drops tuple labels; auxiliary-loss normalization is a separate model-loss bug." + "reason": "BitNet packed-weight loading and FastSpeech2Conformer tokenizer loading are unrelated code paths." }, { - "left": "issue:41950", - "right": "issue:43475", + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading and Gemma eos_token_id inconsistency are unrelated bugs." + }, + { + "left": "issue:33290", + "right": "issue:45656", "accept": false, - "reason": "Video pipeline processor lookup and Sam3Video output attribute access are unrelated issues in different components." + "reason": "Both involve deepspeed/optimizer behavior, but OOM from AdaFactor is not the same as double optimizer-step invocation." }, { "left": "issue:43316", - "right": "issue:43716", + "right": "issue:44877", "accept": false, - "reason": "Config API discrepancy is not the same bug as an image preprocessor/model dtype mismatch." + "reason": "Gemma3TextConfig API mismatch and strict granite_speech config loading are different config-schema problems." }, { - "left": "issue:43381", - "right": "issue:43742", + "left": "issue:44162", + "right": "issue:44792", "accept": false, - "reason": "Gradient checkpointing eval-mode behavior and MobileLLM key loading errors have no shared root cause." + "reason": "ESM2 breakage and a janus image-generation test failure are unrelated model-specific issues." }, { - "left": "issue:16998", - "right": "issue:22355", + "left": "issue:34689", + "right": "issue:36010", "accept": false, - "reason": "A model_max_length question for DeBERTa-V3 is unrelated to a missing transformers.onnx module import." + "reason": "Llama 3.2 Vision loading regression and missing GenerationMixin import are separate import/load failures." }, { - "left": "issue:25251", - "right": "issue:28282", + "left": "issue:44855", + "right": "issue:45137", "accept": false, - "reason": "Pipeline top_k output shape change is unrelated to the PyTorch-missing ImportError." + "reason": "TorchScript parsing on Python 3.13 and DeepSpeed ZeRO3 deque errors do not share an underlying bug." }, { - "left": "issue:43295", - "right": "issue:43825", + "left": "issue:43540", + "right": "issue:44514", "accept": false, - "reason": "Processor/tokenizer access regression is a different bug from an inaccurate pipeline error message about translation support." + "reason": "Qwen3OmniMoe video-input processing and Qwen2_5_VL batched chat-template padding are different failures." }, { - "left": "issue:43065", - "right": "issue:44617", + "left": "issue:43381", + "right": "issue:43856", "accept": false, - "reason": "A dummy Conv2d implementation in Sam3PixelDecoder is not the same as a Sam3Video CUDA OOM." + "reason": "Gradient checkpointing eval-mode behavior and Qwen3 MoE memory usage are not the same bug." }, { - "left": "issue:44871", - "right": "issue:45584", + "left": "issue:42175", + "right": "issue:43316", "accept": false, - "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure after align_special_tokens are different tokenizer issues." + "reason": "Missing TensorFlow in a pip extra and a Gemma3TextConfig API discrepancy are unrelated." }, { - "left": "issue:43761", - "right": "issue:44479", + "left": "issue:43404", + "right": "issue:43441", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and Qwen video-input regression affect different models and code paths." + "reason": "Weight tying in Mistral3ForConditionalGeneration and FlashAttention failure in Ministral-3 are different concrete problems." }, { - "left": "issue:45127", - "right": "issue:45200", + "left": "issue:41950", + "right": "issue:43475", "accept": false, - "reason": "LoRA merge with extended vocabulary causing collapse is unrelated to Gemma 4 mm_token_type_ids defaults." + "reason": "Video-classification pipeline processor selection and Sam3VisionEncoderOutput missing attribute are unrelated." }, { - "left": "issue:44077", - "right": "issue:44534", + "left": "issue:43316", + "right": "issue:43716", "accept": false, - "reason": "PatchTSMixer post_init policy and non-persistent buffer junk are separate v5 cleanup/regression topics, not the same bug." + "reason": "Gemma3TextConfig API mismatch and Mistral-3 image-preprocessor dtype mismatch are different model-specific bugs." }, { - "left": "issue:44936", - "right": "issue:45698", + "left": "issue:43388", + "right": "issue:43688", "accept": false, - "reason": "trainer.evaluate() after train() and wrong custom module loading after save_pretrained are distinct training vs serialization problems." + "reason": "Last-batch label truncation in gather_for_metrics and auxiliary-loss normalization are separate training issues." }, { - "left": "issue:43295", - "right": "issue:44871", + "left": "issue:16998", + "right": "issue:22355", "accept": false, - "reason": "Processor/tokenizer access regression is unrelated to Gemma-3 eos_token_id inconsistency." + "reason": "DeBERTa-V3 model_max_length discussion and missing transformers.onnx module are unrelated." }, { - "left": "issue:44734", - "right": "issue:45406", + "left": "issue:43381", + "right": "issue:43742", "accept": false, - "reason": "Serve KV-cache continuation indexing bug and Gemma4Processor missing _tokenizer are different server failures." + "reason": "Gradient checkpointing eval-mode limitation is unrelated to MobileLLM-125M loading key errors." }, { - "left": "issue:43550", - "right": "issue:43701", + "left": "issue:25251", + "right": "issue:28282", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba is unrelated to a resume_from_checkpoint key mismatch." + "reason": "Pipeline top_k nesting behavior and AutoModel PyTorch import errors are unrelated." }, { - "left": "issue:43606", - "right": "issue:43701", + "left": "issue:43295", + "right": "issue:43825", "accept": false, - "reason": "CPU offload device mismatch is a different root cause from checkpoint key naming/loading." + "reason": "Processor/tokenizer regression and a v5 translation error-message issue are different API regressions." }, { - "left": "issue:43646", - "right": "issue:44913", + "left": "issue:43761", + "right": "issue:44479", "accept": false, - "reason": "Custom model initialization breakage in v5 is not the same as rotary_pct silently resetting on reload." + "reason": "CLIPVisionModel hidden_states regression and Qwen video-input regression do not share the same failure mode." }, { - "left": "issue:43452", - "right": "issue:44496", + "left": "issue:43065", + "right": "issue:44617", "accept": false, - "reason": "GGUF loader argument handling and missing model_type in config.json are both loading-related but not the same concrete defect." + "reason": "Sam3PixelDecoder dummy Conv2d and Sam3Video CUDA OOM are different Sam3 issues." }, { - "left": "issue:43761", - "right": "issue:44242", + "left": "issue:44871", + "right": "issue:45584", "accept": false, - "reason": "CLIPVisionModel hidden_states regression is unrelated to missing load-balancing loss when output_router_logits=False." + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure are unrelated tokenization/generation bugs." }, { - "left": "issue:43931", - "right": "issue:44060", + "left": "issue:45127", + "right": "issue:45200", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch on load and Qwen3-Next tied-weights warning are different weight/config issues in different models." + "reason": "LoRA merge collapse with tied embeddings and Gemma 4 mm_token_type_ids defaults are different model/config problems." } ] }, @@ -8722,43 +8563,43 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:32090", "issue:35141", "issue:41720", "issue:42617", "issue:43065", - "issue:43335", - "issue:43388", + "issue:43295", + "issue:43452", "issue:43550", "issue:43575", - "issue:43644", - "issue:43650", + "issue:43606", + "issue:43646", + "issue:43701", "issue:43742", "issue:43761", "issue:43827", "issue:43906", - "issue:43976", + "issue:43931", "issue:43994", + "issue:44060", "issue:44077", - "issue:44079", "issue:44190", - "issue:44220", + "issue:44242", "issue:44464", "issue:44479", + "issue:44496", "issue:44521", "issue:44534", "issue:44610", "issue:44617", "issue:44734", - "issue:44743", "issue:44792", - "issue:44855", "issue:44871", "issue:44913", "issue:44933", - "issue:44938", + "issue:44936", "issue:44964", "issue:44977", "issue:45127", @@ -8767,252 +8608,246 @@ "issue:45357", "issue:45362", "issue:45381", - "issue:45542", + "issue:45406", "issue:45563", "issue:45584", - "issue:45588", - "issue:45600" + "issue:45600", + "issue:45698" ], "soft_pairs": [ + "issue:44077|issue:44534", + "issue:44734|issue:45406", + "issue:44936|issue:45698", + "issue:43295|issue:44871", + "issue:43550|issue:43701", + "issue:43606|issue:43701", + "issue:43646|issue:44913", + "issue:43452|issue:44496", + "issue:43931|issue:44060", + "issue:43761|issue:44242", "issue:43575|issue:43742", "issue:44617|issue:45584", - "issue:42617|issue:43575", "issue:45563|issue:45584", + "issue:42617|issue:43575", "issue:43906|issue:45290", - "issue:32090|issue:35141", "issue:44521|issue:45381", - "issue:41720|issue:43575", "issue:44479|issue:44734", "issue:44933|issue:45362", - "issue:44464|issue:44871", + "issue:32090|issue:35141", + "issue:41720|issue:43575", "issue:43065|issue:44610", - "issue:43827|issue:43994", + "issue:44464|issue:44871", "issue:44464|issue:45563", + "issue:43827|issue:43994", "issue:44792|issue:44977", "issue:44534|issue:44913", "issue:41720|issue:45245", "issue:44190|issue:44964", "issue:44792|issue:45600", - "issue:44938|issue:45542", - "issue:43388|issue:44079", - "issue:45127|issue:45357", - "issue:43650|issue:44220", - "issue:44855|issue:45588", - "issue:43761|issue:44743", - "issue:43335|issue:44855", - "issue:43976|issue:44855", - "issue:44077|issue:44079", - "issue:43644|issue:44913", - "issue:43550|issue:44190" + "issue:45127|issue:45357" ], - "prepared_review_unit_hash": "54f36fec3d66d4c9620de38880d697f87a83b829be04ad064e4803965333a645", + "prepared_review_unit_hash": "bb6fd536b1973c01c5da1e46bddb08436a119d15305a9c1ad3cb15853acf843c", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36588, - "estimated_input_tokens": 9147, - "estimated_eval_tokens": 18550 + "serialized_chars": 36570, + "estimated_input_tokens": 9143, + "estimated_eval_tokens": 18542 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: the soft matches mostly share only broad subsystem keywords (generation, multimodal, loading, config) but not the same concrete bug. I would keep all soft edges rejected. If a single representative issue is needed, issue:44521 is the strongest canonical-style report because it is specific, active, and already externally referenced.", - "confidence": 0.89, - "canonical_issue_reason": "If forced to pick one representative issue, issue:44521 is the best canonical candidate: it describes a concrete multimodal chat_template bug, has inbound references, and is narrower/more actionable than the others.", + "summary": "This cluster is mostly heterogeneous: many issues mention loading, initialization, or multimodal regressions, but they point to different code paths and symptoms. I found no true duplicate pair among the soft edges, so all soft links should stay rejected.", + "confidence": 0.96, + "canonical_issue_reason": "No exact canonical duplicate emerges; issue 43646 is the broadest representative of the initialization/loading regression subset, but it is not a duplicate of the others.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:44521 is the highest-signal standalone bug report in the set: specific failure mode, clear scope, and existing references suggest it is the most suitable anchor issue.", + "best_issue_reason": "Issue 43646 is the best single representative for the cluster only because it is broadly about model initialization/loading regressions; however, the cluster is too mixed for a real canonical duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43575", - "right": "issue:43742", + "left": "issue:44077", + "right": "issue:44534", "accept": false, - "reason": "Different models and failure modes: tensor-parallel OOM vs a loading KeyError." + "reason": "Both relate to v5/post-init behavior, but one is about patchtsmixer allowing optional post_init and the other about junk non-persistent buffers; different bugs." }, { - "left": "issue:44617", - "right": "issue:45584", + "left": "issue:44734", + "right": "issue:45406", "accept": false, - "reason": "Unrelated bugs: SAM3 video OOM versus Whisper empty-transcription generation failure." + "reason": "Distinct crashes in different areas: KV-cache continuation indexing in serve vs Gemma4Processor missing _tokenizer." }, { - "left": "issue:42617", - "right": "issue:43575", + "left": "issue:44936", + "right": "issue:45698", "accept": false, - "reason": "A generic 3d_parallel.py execution problem is not the same as Qwen loading OOM." + "reason": "Trainer lifecycle failure after train() is unrelated to from_pretrained loading the wrong custom module after save_pretrained." }, { - "left": "issue:45563", - "right": "issue:45584", + "left": "issue:43295", + "right": "issue:44871", "accept": false, - "reason": "A stale warning about num_return_sequences is unrelated to Whisper crashing on empty transcription." + "reason": "Multimodal processor/tokenizer regression and eos_token_id mismatch are different model/config issues." }, { - "left": "issue:43906", - "right": "issue:45290", + "left": "issue:43550", + "right": "issue:43701", "accept": false, - "reason": "The first is only a reproduction of another issue; the second is a specific apply_chat_template crash, and there is no clear evidence they are the same bug." + "reason": "torch.compile+SDPA failure for Bamba is unrelated to resume_from_checkpoint key mismatch." }, { - "left": "issue:32090", - "right": "issue:35141", + "left": "issue:43606", + "right": "issue:43701", "accept": false, - "reason": "Trainer broadcast TypeError and embedding reinitialization in post_init are different code paths." + "reason": "CPU offload device mismatch is a different failure mode from checkpoint key mismatch." }, { - "left": "issue:44521", - "right": "issue:45381", + "left": "issue:43646", + "right": "issue:44913", "accept": false, - "reason": "Both are multimodal, but assistant_mask generation and wrong vision_position_ids are distinct defects." + "reason": "Both involve model config/loading, but one is custom model initialization and the other is rotary_pct not persisting on reload." }, { - "left": "issue:41720", - "right": "issue:43575", + "left": "issue:43452", + "right": "issue:44496", "accept": false, - "reason": "A device-mapping cudaErrorAssert is not the same as a tensor-parallel OOM." + "reason": "Both are loading-related, but gguf_file breaks a specific AutoTokenizer/AutoModel path while the other lacks a model_type in config.json." }, { - "left": "issue:44479", - "right": "issue:44734", + "left": "issue:43931", + "right": "issue:44060", "accept": false, - "reason": "Video-input regression for Qwen models and KV-cache continuation indexing are different concrete bugs." + "reason": "Weight shape mismatch in Qwen3-VL is not the same bug as the incorrect tied-weights warning in Qwen3-Next." }, { - "left": "issue:44933", - "right": "issue:45362", + "left": "issue:43761", + "right": "issue:44242", "accept": false, - "reason": "A missing import and a Qwen3.5 chat crash are not the same underlying issue." + "reason": "Hidden_states=None in CLIPVisionModel.forward is unrelated to missing load-balancing loss when output_router_logits=False." }, { - "left": "issue:44464", - "right": "issue:44871", + "left": "issue:43575", + "right": "issue:43742", "accept": false, - "reason": "Chunked generation inconsistency and eos_token_id config mismatch are unrelated." + "reason": "TP OOM on a 57B model and a key error loading MobileLLM-125M are unrelated load failures." }, { - "left": "issue:43065", - "right": "issue:44610", + "left": "issue:44617", + "right": "issue:45584", "accept": false, - "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to OmDet-Turbo input-size mismatch." + "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after align_special_tokens." }, { - "left": "issue:43827", - "right": "issue:43994", + "left": "issue:45563", + "right": "issue:45584", "accept": false, - "reason": "Docs still referencing pipeline() is not the same as a bad pipeline result for SigLIP2." + "reason": "A stale warning for num_return_sequences is not the same as an empty-transcription generation crash." }, { - "left": "issue:44464", - "right": "issue:45563", + "left": "issue:42617", + "right": "issue:43575", "accept": false, - "reason": "Compiled-forward generation inconsistency and a stale warning are unrelated." + "reason": "Failed 3d_parallel.py execution and tp OOM are different problems with different code paths." }, { - "left": "issue:44792", - "right": "issue:44977", + "left": "issue:43906", + "right": "issue:45290", "accept": false, - "reason": "Different models and symptoms: Janus image-generation test failure versus Qwen3.5 flash-attention generation bug." + "reason": "An isolated reproduction of an older issue is not the same as the assistant tool-call crash in apply_chat_template(tokenize=True)." }, { - "left": "issue:44534", - "right": "issue:44913", + "left": "issue:44521", + "right": "issue:45381", "accept": false, - "reason": "Non-persistent buffers being junk-filled is not the same as rotary_pct not persisting on reload." + "reason": "Both are multimodal regressions, but assistant_masks being all zero is different from wrong vision_position_ids for video inputs." }, { - "left": "issue:41720", - "right": "issue:45245", - "accept": false, - "reason": "Device-mapping cuda asserts and category-count limit errors are unrelated." - }, - { - "left": "issue:44190", - "right": "issue:44964", + "left": "issue:44479", + "right": "issue:44734", "accept": false, - "reason": "Local dataset loading in a no-trainer script is unrelated to loading a specific multimodal model." + "reason": "Video-input regression in Qwen VL models is unrelated to serve KV-cache continuation tensor indexing." }, { - "left": "issue:44792", - "right": "issue:45600", + "left": "issue:44933", + "right": "issue:45362", "accept": false, - "reason": "Janus image-generation test breakage and stale auto_mappings config references are different failures." + "reason": "Missing import from image_utils and Qwen3.5 chat crash are unrelated." }, { - "left": "issue:44938", - "right": "issue:45542", + "left": "issue:32090", + "right": "issue:35141", "accept": false, - "reason": "Python 3.14 import/load failure and tensorboard-without-TF backend error are different environment issues." + "reason": "Trainer GPU broadcast TypeError and embedding reinitialization in post_init are different bugs." }, { - "left": "issue:43388", - "right": "issue:44079", + "left": "issue:41720", + "right": "issue:43575", "accept": false, - "reason": "Metric gathering dropping tuple labels is unrelated to ModelOutput key assignment when a previous key was None." + "reason": "Auto device mapping cudaErrorAssert and TP OOM are both resource-related but not the same underlying issue." }, { - "left": "issue:45127", - "right": "issue:45357", + "left": "issue:43065", + "right": "issue:44610", "accept": false, - "reason": "LoRA merge/vocab collapse and incorrect visual-encoder save_pretrained keys are separate bugs." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to processor output size mismatch for OmDet-Turbo." }, { - "left": "issue:43650", - "right": "issue:44220", + "left": "issue:44464", + "right": "issue:44871", "accept": false, - "reason": "A vague request for data and an fbank feature extraction issue do not describe the same bug." + "reason": "Compiled-forward chunked generation inconsistency is not the same as an eos_token_id configuration mismatch." }, { - "left": "issue:44855", - "right": "issue:45588", + "left": "issue:43827", + "right": "issue:43994", "accept": false, - "reason": "A Python 3.13 IndentationError in DebertaV2Model is unrelated to flash_attention.py failing on s_aux=None." + "reason": "Docs still referencing pipeline() is a documentation cleanup issue, not the same as nonsensical model outputs." }, { - "left": "issue:43761", - "right": "issue:44743", + "left": "issue:44792", + "right": "issue:44977", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and Qwen recurrent-state reset are different model-path bugs." + "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation bug are different models and failure modes." }, { - "left": "issue:43335", - "right": "issue:44855", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "SwitchTransformers sparse-layer config and a DebertaV2 import parsing error are unrelated." + "reason": "Non-persistent buffer junk and rotary_pct reset on reload both touch persistence, but they are not the same concrete bug." }, { - "left": "issue:43976", - "right": "issue:44855", + "left": "issue:41720", + "right": "issue:45245", "accept": false, - "reason": "A broad Python-version compatibility complaint is not the same as the DebertaV2 JIT parsing failure." + "reason": "cudaErrorAssert on A800 is unrelated to the category-count limit error." }, { - "left": "issue:44077", - "right": "issue:44079", + "left": "issue:44190", + "right": "issue:44964", "accept": false, - "reason": "An optional post_init constraint issue is unrelated to ModelOutput None-key assignment." + "reason": "Local dataset loading in a training script is unrelated to loading a specific multimodal model." }, { - "left": "issue:43644", - "right": "issue:44913", + "left": "issue:44792", + "right": "issue:45600", "accept": false, - "reason": "Non-persistent buffer junk values and rotary_pct reload behavior are different configuration/state bugs." + "reason": "A failing janus generation test is unrelated to auto_mappings.py referencing removed Sam3LiteText configs." }, { - "left": "issue:43550", - "right": "issue:44190", + "left": "issue:45127", + "right": "issue:45357", "accept": false, - "reason": "A torch.compile SDPA model failure is unrelated to loading a local dataset in a training script." + "reason": "LoRA merge collapse with extended vocab and incorrect visual encoder keys saved are different save/merge regressions." } ] }, @@ -9034,7 +8869,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36246", "issue:38175", @@ -9043,302 +8878,300 @@ "issue:41628", "issue:41950", "issue:42175", - "issue:42907", "issue:43065", "issue:43116", "issue:43295", - "issue:43299", + "issue:43335", + "issue:43388", "issue:43404", - "issue:43441", "issue:43479", - "issue:43531", "issue:43540", - "issue:43643", - "issue:43673", + "issue:43550", + "issue:43644", + "issue:43650", + "issue:43761", "issue:43976", "issue:44016", - "issue:44062", + "issue:44077", + "issue:44079", "issue:44112", "issue:44186", "issue:44190", "issue:44220", - "issue:44246", "issue:44279", - "issue:44336", "issue:44466", - "issue:44496", "issue:44561", "issue:44683", "issue:44734", + "issue:44743", + "issue:44855", "issue:44913", - "issue:44964", - "issue:44987", - "issue:45003", + "issue:44938", "issue:45216", "issue:45245", "issue:45276", - "issue:45335", "issue:45375", "issue:45406", "issue:45446", + "issue:45542", "issue:45584", + "issue:45588", "issue:45600", "issue:45632" ], "soft_pairs": [ + "issue:44938|issue:45542", + "issue:43388|issue:44079", + "issue:43761|issue:44743", + "issue:44855|issue:45588", + "issue:43335|issue:44855", + "issue:43976|issue:44855", + "issue:44077|issue:44079", + "issue:43550|issue:44190", + "issue:43644|issue:44913", "issue:43540|issue:44734", + "issue:43650|issue:44220", "issue:44112|issue:45600", - "issue:45276|issue:45584", "issue:43116|issue:44190", - "issue:41950|issue:44016", + "issue:45276|issue:45584", "issue:39692|issue:44190", + "issue:41950|issue:44016", "issue:43479|issue:45375", - "issue:43295|issue:45276", "issue:44734|issue:45632", - "issue:45245|issue:45584", "issue:36246|issue:41628", "issue:43404|issue:45216", + "issue:43295|issue:45276", + "issue:45245|issue:45584", "issue:36246|issue:38617", "issue:44016|issue:44220", "issue:44561|issue:45632", + "issue:44186|issue:44734", "issue:44683|issue:45446", - "issue:42175|issue:44279", "issue:38175|issue:43065", - "issue:44186|issue:44734", + "issue:42175|issue:44279", "issue:45406|issue:45632", - "issue:44466|issue:44913", - "issue:44246|issue:45003", - "issue:43643|issue:44496", - "issue:42907|issue:43441", - "issue:44016|issue:44062", - "issue:44964|issue:45600", - "issue:43295|issue:45335", - "issue:43299|issue:45216", - "issue:43976|issue:44987", - "issue:43673|issue:44336", - "issue:44112|issue:44964", - "issue:43531|issue:45600" + "issue:44466|issue:44913" ], - "prepared_review_unit_hash": "4a521767983f6749dbb4722576c787278359467298185ed42c8b018cd90cdc40", + "prepared_review_unit_hash": "0336dbca82552ae41b3801f1a5fca9f4cc0ef0a2e5ef320b8141f4e0b9514e0c", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37342, - "estimated_input_tokens": 9336, - "estimated_eval_tokens": 18928 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36640, + "estimated_input_tokens": 9160, + "estimated_eval_tokens": 18576 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of unrelated Transformers issues: import/export regressions, multimodal model loading problems, pipeline/example bugs, config/serialization regressions, and CI/test flakes. I don\u2019t see a tight duplicate set, so I would not accept any soft-edge merges.", - "confidence": 0.78, - "canonical_issue_reason": "issue:44964 is the broadest active user-facing regression in the set and is closest to the recurring latest-transformers multimodal loading failures, but the overall cluster is still heterogeneous.", + "summary": "The soft links are overwhelmingly false positives: they span unrelated import errors, config/serialization regressions, model-specific bugs, pipeline issues, and docs/example failures. None of the candidate pairs look like the same underlying defect or change.", + "confidence": 0.93, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44964 is the most representative anchor among these issues because it has a concrete reproduction, active discussion, and the widest apparent applicability within the group.", + "best_issue_reason": "issue:41628 is the strongest standalone candidate to surface as a representative user-facing regression: it is a concrete, high-activity import failure with broad impact. That said, the set does not form a true duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43540", - "right": "issue:44734", + "left": "issue:44938", + "right": "issue:45542", "accept": false, - "reason": "Different bugs: Qwen3OmniMoe video-input handling vs transformers serve KV-cache continuation indexing." + "reason": "Unrelated failures: Python import/load issue vs missing TensorFlow backend when only tensorboard is installed." }, { - "left": "issue:44112", - "right": "issue:45600", + "left": "issue:43388", + "right": "issue:44079", "accept": false, - "reason": "Both touch CI-related failures, but one is a GraniteSpeech test flake and the other is an auto_mappings reference cleanup issue." + "reason": "Both involve internals around output handling, but one is metric-gathering label truncation and the other is ModelOutput key assignment; different code paths." }, { - "left": "issue:45276", - "right": "issue:45584", + "left": "issue:43761", + "right": "issue:44743", "accept": false, - "reason": "Unrelated symptoms: Gemma4 resize_token_embeddings propagation vs Whisper empty-transcription generation failure." + "reason": "Different models and bugs: CLIPVisionModel hidden_states regression vs Qwen recurrent-state reset with cache." }, { - "left": "issue:43116", - "right": "issue:44190", + "left": "issue:44855", + "right": "issue:45588", "accept": false, - "reason": "Different example scripts and different failure modes: multi-label classification output vs local dataset loading." + "reason": "Completely different problems: Python 3.13 JIT parse/indentation issue vs flash-attention crashing on s_aux=None." }, { - "left": "issue:41950", - "right": "issue:44016", + "left": "issue:43335", + "right": "issue:44855", "accept": false, - "reason": "Pipeline processor lookup bug vs notebook syntax error; no shared underlying code path." + "reason": "SwitchTransformers config construction bug is unrelated to the DebertaV2 Python 3.13 import/parser issue." }, { - "left": "issue:39692", - "right": "issue:44190", + "left": "issue:43976", + "right": "issue:44855", "accept": false, - "reason": "Both are example-script issues, but one is a SigLIP2 docs/model mismatch and the other is dataset loading for image classification." + "reason": "A packaging/Python-version compatibility report does not match the specific DebertaV2 @torch.jit.script parsing failure." }, { - "left": "issue:43479", - "right": "issue:45375", + "left": "issue:44077", + "right": "issue:44079", "accept": false, - "reason": "Both are config-shape problems in multimodal models, but they concern different models and different config initialization behaviors." + "reason": "Both are model-output/config validation themed, but one is patchtsmixer post_init policy and the other is None-key assignment in ModelOutput." }, { - "left": "issue:43295", - "right": "issue:45276", + "left": "issue:43550", + "right": "issue:44190", "accept": false, - "reason": "Processor/tokenizer regression is unrelated to Gemma4 token embedding resizing behavior." + "reason": "Torch.compile+SDPA failure in Bamba is unrelated to the image-classification script failing to load a local dataset." }, { - "left": "issue:44734", - "right": "issue:45632", + "left": "issue:43644", + "right": "issue:44913", "accept": false, - "reason": "Serve KV-cache indexing crash is unrelated to trust_remote_code cache-path collisions." + "reason": "Both are persistence-related, but one is buffer initialization junk and the other is GPTNeoX rotary_pct not reloading; not the same bug." }, { - "left": "issue:45245", - "right": "issue:45584", + "left": "issue:43540", + "right": "issue:44734", "accept": false, - "reason": "No overlap: category-count limit in embeddings vs Whisper generation after special-token alignment." + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to serve KV-cache continuation tensor indexing." }, { - "left": "issue:36246", - "right": "issue:41628", + "left": "issue:43650", + "right": "issue:44220", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different import surfaces." + "reason": "A noisy generic issue title does not match the concrete audio feature-extraction bug." }, { - "left": "issue:43404", - "right": "issue:45216", + "left": "issue:44112", + "right": "issue:45600", "accept": false, - "reason": "Different regressions: tied lm_head weights in Mistral3 vs incorrect checkpoint serialization in Qwen3.5." + "reason": "CI failures in different areas: GraniteSpeech stale device override test vs auto_mappings referencing removed Sam3LiteText configs." }, { - "left": "issue:36246", - "right": "issue:38617", + "left": "issue:43116", + "right": "issue:44190", "accept": false, - "reason": "Different missing names in different modules; no clear shared root cause." + "reason": "Both mention example scripts, but one is multi-label classification output behavior and the other is local dataset loading for image classification." }, { - "left": "issue:44016", - "right": "issue:44220", + "left": "issue:45276", + "right": "issue:45584", "accept": false, - "reason": "Notebook syntax error is unrelated to the audio feature-extraction failure." + "reason": "Gemma4 embedding resize behavior and Whisper empty-transcription handling are unrelated." }, { - "left": "issue:44561", - "right": "issue:45632", + "left": "issue:39692", + "right": "issue:44190", "accept": false, - "reason": "One is a removed-helper breakage for trust_remote_code models; the other is a local cache-path collision." + "reason": "Both touch examples/docs, but one is SigLIP2 documentation errors and the other is a local dataset loading problem." }, { - "left": "issue:44683", - "right": "issue:45446", + "left": "issue:41950", + "right": "issue:44016", "accept": false, - "reason": "Both are flex_attention-related, but one is a compiled runtime failure and the other is a version-check/import issue." + "reason": "Video-classification processor mismatch is unrelated to a notebook syntax error." }, { - "left": "issue:42175", - "right": "issue:44279", + "left": "issue:43479", + "right": "issue:45375", "accept": false, - "reason": "General dependency packaging complaint, but not the same backend-selection problem." + "reason": "Both are config-field issues, but they affect different multimodal model families and different missing fields." }, { - "left": "issue:38175", - "right": "issue:43065", + "left": "issue:44734", + "right": "issue:45632", "accept": false, - "reason": "Unrelated model/runtime behavior: SigLIP2 zero probabilities vs a dummy Conv2d in Sam3PixelDecoder." + "reason": "Serve KV-cache indexing and trust_remote_code cache-path collisions are different subsystems and defects." }, { - "left": "issue:44186", - "right": "issue:44734", + "left": "issue:36246", + "right": "issue:41628", "accept": false, - "reason": "Tokenizer crashes on NER/padding are unrelated to serve KV-cache continuation indexing." + "reason": "Both are import errors, but one is a specific missing Qwen2_5_VLImageProcessor symbol and the other is top-level AutoImageProcessor availability." }, { - "left": "issue:45406", - "right": "issue:45632", + "left": "issue:43404", + "right": "issue:45216", "accept": false, - "reason": "Gemma4Processor missing _tokenizer is a processor API regression, not a cache-path collision." + "reason": "Tied lm_head weights in Mistral3 versus incorrect Qwen3.5 checkpoint saving are different model/regression bugs." }, { - "left": "issue:44466", - "right": "issue:44913", + "left": "issue:43295", + "right": "issue:45276", "accept": false, - "reason": "Both involve config/serialization persistence, but one is tied lm_head serialization and the other is GPTNeoX rotary_pct reload handling." + "reason": "Processor/tokenizer regression and Gemma4 resize_token_embeddings behavior do not share the same concrete code path." }, { - "left": "issue:44246", - "right": "issue:45003", + "left": "issue:45245", + "right": "issue:45584", "accept": false, - "reason": "Import latency and unsafe sys.modules access are related at most indirectly; not the same bug." + "reason": "Category-cardinality runtime error is unrelated to Whisper's empty-transcription failure." }, { - "left": "issue:43643", - "right": "issue:44496", + "left": "issue:36246", + "right": "issue:38617", "accept": false, - "reason": "Both are config-loading issues, but one drops fields under trust_remote_code and the other lacks model_type for an unrecognized model." + "reason": "Different missing imports in different modules; no shared underlying code change is evident." }, { - "left": "issue:42907", - "right": "issue:43441", + "left": "issue:44016", + "right": "issue:44220", "accept": false, - "reason": "Different Ministral failures: dequantized save_pretrained vs FlashAttention runtime on v5 RC." + "reason": "Notebook syntax error and audio feature extraction bug are unrelated." }, { - "left": "issue:44016", - "right": "issue:44062", + "left": "issue:44561", + "right": "issue:45632", "accept": false, - "reason": "Notebook syntax error is unrelated to AddedToken special-argument duplication." + "reason": "Removal of a helper breaking trust_remote_code models is unrelated to local cache-path collisions." }, { - "left": "issue:44964", - "right": "issue:45600", + "left": "issue:44186", + "right": "issue:44734", "accept": false, - "reason": "Phi-4 multimodal loading regression is unrelated to auto_mappings referring to removed Sam3LiteText configs." + "reason": "Tokenizer crashes on NER padding/truncation are not the same as serve KV-cache continuation indexing failures." }, { - "left": "issue:43295", - "right": "issue:45335", + "left": "issue:44683", + "right": "issue:45446", "accept": false, - "reason": "Different resize_token_embeddings regressions with different affected code paths and models." + "reason": "Both mention flex_attention/PyTorch versions, but one is compiled flex_attention failing on new torch and the other is an incorrect version gate for AuxRequest import." }, { - "left": "issue:43299", - "right": "issue:45216", + "left": "issue:38175", + "right": "issue:43065", "accept": false, - "reason": "Both are Qwen regressions, but one breaks model loading and the other breaks save_pretrained checkpoint correctness." + "reason": "SigLIP2 zero-probabilities and a dummy Conv2d in Sam3PixelDecoder are unrelated model bugs." }, { - "left": "issue:43976", - "right": "issue:44987", + "left": "issue:42175", + "right": "issue:44279", "accept": false, - "reason": "Python-version support problem vs a model-loading failure on a specific package version; not the same issue." + "reason": "A specific missing TensorFlow dependency from a torch install is not the same as a vague transformers dependency issue." }, { - "left": "issue:43673", - "right": "issue:44336", + "left": "issue:45406", + "right": "issue:45632", "accept": false, - "reason": "Generation cache regression and ANSI terminal formatting are unrelated." + "reason": "Gemma4Processor _tokenizer crash and trust_remote_code cache collision are unrelated." }, { - "left": "issue:44112", - "right": "issue:44964", + "left": "issue:45588", + "right": "issue:45600", "accept": false, - "reason": "A CI test flake in GraniteSpeech is not the same bug as Phi-4 multimodal model loading." + "reason": "Flash-attention s_aux=None crash and removed Sam3LiteText config references breaking CI are different failures." }, { - "left": "issue:43531", - "right": "issue:45600", + "left": "issue:44466", + "right": "issue:44913", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to stale auto_mappings config references." + "reason": "Both concern state persistence, but one is lm_head.weight serialization and the other is GPTNeoX rotary_pct reload behavior." } ] }, @@ -9360,43 +9193,42 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:38175", - "issue:39692", "issue:41950", + "issue:42907", "issue:43295", "issue:43299", "issue:43381", - "issue:43450", + "issue:43441", "issue:43479", "issue:43525", "issue:43526", + "issue:43531", "issue:43577", - "issue:43650", + "issue:43643", + "issue:43673", "issue:43824", "issue:43854", "issue:43901", "issue:43976", "issue:44016", + "issue:44062", "issue:44075", + "issue:44112", "issue:44242", - "issue:44261", - "issue:44263", + "issue:44246", "issue:44279", - "issue:44291", - "issue:44297", + "issue:44336", "issue:44351", "issue:44368", "issue:44462", "issue:44496", "issue:44521", - "issue:44617", - "issue:44683", - "issue:44779", - "issue:44877", "issue:44933", "issue:44964", + "issue:44987", + "issue:45003", "issue:45020", "issue:45042", "issue:45081", @@ -9407,257 +9239,243 @@ "issue:45335", "issue:45357", "issue:45405", - "issue:45542", "issue:45584", - "issue:45593" + "issue:45593", + "issue:45600" ], "soft_pairs": [ + "issue:44246|issue:45003", + "issue:43643|issue:44496", + "issue:42907|issue:43441", + "issue:44016|issue:44062", + "issue:44964|issue:45600", + "issue:43299|issue:45216", + "issue:43295|issue:45335", + "issue:43976|issue:44987", + "issue:43673|issue:44336", + "issue:44112|issue:44964", "issue:43824|issue:44933", + "issue:43531|issue:45600", + "issue:43976|issue:44279", "issue:44496|issue:45042", "issue:43854|issue:44964", - "issue:43976|issue:44279", "issue:43526|issue:43901", "issue:45335|issue:45584", "issue:43299|issue:45357", "issue:44075|issue:44368", "issue:44521|issue:45325", - "issue:43381|issue:45593", "issue:43577|issue:45216", + "issue:43381|issue:45593", "issue:44351|issue:45278", "issue:43976|issue:45405", + "issue:41950|issue:45020", "issue:43901|issue:44496", "issue:44242|issue:45200", - "issue:41950|issue:45020", "issue:43479|issue:43525", - "issue:44462|issue:45081", - "issue:41950|issue:43650", - "issue:44496|issue:44877", - "issue:44683|issue:45542", - "issue:44297|issue:44462", - "issue:44279|issue:45405", - "issue:43299|issue:43854", - "issue:44462|issue:44779", - "issue:39692|issue:43901", - "issue:43295|issue:44779", - "issue:38175|issue:44617", - "issue:44016|issue:44291", - "issue:43450|issue:44279", - "issue:44261|issue:44263" + "issue:44462|issue:45081" ], - "prepared_review_unit_hash": "04725aa6b5ec28fc0178e814ed68a65abbc80ab8b98ce0e27fd6c4e36afe14c2", + "prepared_review_unit_hash": "c7ea46c1a54a2e80ddce8ffefe3f500e239089a6f2e3ef54520abd7bf7889388", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36976, - "estimated_input_tokens": 9244, - "estimated_eval_tokens": 18744 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35784, + "estimated_input_tokens": 8946, + "estimated_eval_tokens": 18148 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of distinct Transformers regressions, model-loading failures, tokenizer/processor bugs, and documentation issues. The soft pairs share broad themes, but they do not look like the same underlying bug or change.", - "confidence": 0.44, - "canonical_issue_reason": "issue:45020 is the best umbrella candidate: it is an open, broad regression report about recent Transformers versions breaking remote-code model loading, which matches the cluster\u2019s recurring version/regression-loading theme better than the more model-specific issues.", + "summary": "This cluster is mostly a collection of related-but-distinct Transformers regression reports, many around model loading, config handling, tokenizers, and multimodal support. There are no clear exact duplicates among the soft pairs, so all candidate edges are rejected. The broadest representative issue is the remote_code/model-loading regression report.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45020 is the broadest, most representative open report in the set: it describes a recent-version regression affecting model loading with remote_code, which is a common umbrella for several adjacent loading/config failures here.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the most central and reusable issue for triage because it captures the broad regression pattern seen across several loading/import failures without being as narrowly tied to one model family.", + "best_issue_reason": "issue:45020 best captures the shared theme of recent-version loading regressions without being too narrowly tied to one model family or one specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43824", - "right": "issue:44933", - "accept": false, - "reason": "Both are import-related, but one is a missing Qwen2.5-VL export and the other is an invalid image_utils import; different symbols and failure paths." - }, - { - "left": "issue:44496", - "right": "issue:45042", + "left": "issue:44246", + "right": "issue:45003", "accept": false, - "reason": "One is model/config recognition failure; the other is PIL image processors incorrectly requiring torchvision. Same broad area, different bug." + "reason": "Import slowdown and unsafe sys.modules access are related to import-time behavior, but they are not the same concrete bug or fix path." }, { - "left": "issue:43854", - "right": "issue:44964", + "left": "issue:43643", + "right": "issue:44496", "accept": false, - "reason": "Both are model-loading problems, but for different models and likely different root causes; not the same concrete bug." + "reason": "Both involve config/model loading, but one is missing fields with trust_remote_code and the other is an unrecognized model / missing model_type case." }, { - "left": "issue:43976", - "right": "issue:44279", + "left": "issue:42907", + "right": "issue:43441", "accept": false, - "reason": "Python/version compatibility issue versus a dependency issue report; too generic and not the same underlying defect." + "reason": "Same model family, different failures: saving dequantized models vs FlashAttention load/runtime behavior." }, { - "left": "issue:43526", - "right": "issue:43901", + "left": "issue:44016", + "right": "issue:44062", "accept": false, - "reason": "Processor bug versus documentation/API mismatch; these are unrelated despite both involving image classification tooling." + "reason": "A notebook syntax error and a tokenizer AddedToken argument error are unrelated." }, { - "left": "issue:45335", - "right": "issue:45584", + "left": "issue:44964", + "right": "issue:45600", "accept": false, - "reason": "Resize-token-embeddings behavior and Whisper empty-transcription generation are different code paths and symptoms." + "reason": "Phi-4 multimodal loading and stale auto_mappings references to Sam3LiteText are different breakages." }, { "left": "issue:43299", - "right": "issue:45357", - "accept": false, - "reason": "Same model family, but one is loading failure and the other is save_pretrained serialization of visual encoder keys; different defects." - }, - { - "left": "issue:44075", - "right": "issue:44368", + "right": "issue:45216", "accept": false, - "reason": "Optimizer arguments not used is unrelated to a tied-embeddings warning in Qwen fine-tuning." + "reason": "Both are Qwen-related, but one is a model-loading failure and the other is an incorrect save_pretrained checkpoint issue." }, { - "left": "issue:44521", - "right": "issue:45325", + "left": "issue:43295", + "right": "issue:45335", "accept": false, - "reason": "Chat-template mask generation and Qwen2.5-VL rope-index scaling are separate multimodal bugs." + "reason": "Processor/tokenizer regression in multimodal code and a t5gemma embedding-resize bug are different code paths." }, { - "left": "issue:43381", - "right": "issue:45593", + "left": "issue:43976", + "right": "issue:44987", "accept": false, - "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated training-time issues." + "reason": "A Python version compatibility issue is not the same bug as a model-loading failure for one specific repository." }, { - "left": "issue:43577", - "right": "issue:45216", + "left": "issue:43673", + "right": "issue:44336", "accept": false, - "reason": "Blip2 dtype propagation and Qwen3.5 save_pretrained checkpoint correctness are different model-specific regressions." + "reason": "Generation cache behavior and ANSI output formatting are unrelated." }, { - "left": "issue:44351", - "right": "issue:45278", + "left": "issue:44112", + "right": "issue:44964", "accept": false, - "reason": "One missing export versus a broad import-error report; related theme, but not the same concrete failure." + "reason": "A CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal loading." }, { - "left": "issue:43976", - "right": "issue:45405", + "left": "issue:43824", + "right": "issue:44933", "accept": false, - "reason": "Runtime compatibility issue versus an unreleased dependency-version bump; these do not describe the same bug." + "reason": "Two import errors, but for different symbols and likely different causes; not the same fix." }, { - "left": "issue:43901", - "right": "issue:44496", + "left": "issue:43531", + "right": "issue:45600", "accept": false, - "reason": "Docs mismatch about return_all_scores is unrelated to a model/config loading failure." + "reason": "Qwen3-MoE sliding-window behavior and removed config references are unrelated." }, { - "left": "issue:44242", - "right": "issue:45200", + "left": "issue:43976", + "right": "issue:44279", "accept": false, - "reason": "Load-balancing loss routing bug and mm_token_type_ids defaulting bug affect different models and logic." + "reason": "A Python support/version issue and a vague dependency problem are too different to treat as the same bug." }, { - "left": "issue:41950", - "right": "issue:45020", + "left": "issue:44496", + "right": "issue:45042", "accept": false, - "reason": "Video-classification pipeline processor lookup and remote_code model loading are different regression types." + "reason": "Model config recognition and PIL/torchvision dependency handling are different issues." }, { - "left": "issue:43479", - "right": "issue:43525", + "left": "issue:43854", + "right": "issue:44964", "accept": false, - "reason": "Different config initialization bugs for different model classes; not the same underlying issue." + "reason": "Both are model-loading failures, but for different models and symptom types." }, { - "left": "issue:44462", - "right": "issue:45081", + "left": "issue:43526", + "right": "issue:43901", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashing are separate tokenizer-loading defects." + "reason": "A BeitImageProcessorFast bug and a docs/behavior mismatch in TextClassificationPipeline do not share a concrete fix path." }, { - "left": "issue:41950", - "right": "issue:43650", + "left": "issue:45335", + "right": "issue:45584", "accept": false, - "reason": "A substantive pipeline bug versus a non-informative placeholder issue; no shared bug." + "reason": "Embedding resize behavior and Whisper empty-transcription generation are unrelated." }, { - "left": "issue:44496", - "right": "issue:44877", + "left": "issue:43299", + "right": "issue:45357", "accept": false, - "reason": "Both concern config/model loading, but one is unknown model recognition and the other is strict config blocking granite_speech; different checks." + "reason": "Both concern Qwen3.5, but one is load-time breakage and the other is incorrect save_pretrained output keys." }, { - "left": "issue:44683", - "right": "issue:45542", + "left": "issue:44075", + "right": "issue:44368", "accept": false, - "reason": "Compiled flex_attention on torch 2.9 and tensorboard-without-TensorFlow are unrelated dependency/runtime issues." + "reason": "Optimizer argument handling and a warning about tie_word_embeddings are unrelated." }, { - "left": "issue:44297", - "right": "issue:44462", + "left": "issue:44521", + "right": "issue:45325", "accept": false, - "reason": "Tokenizer save metadata mismatch and tokenizer.json loading preference are separate serialization/loading problems." + "reason": "Both are multimodal, but assistant mask generation and rope index scaling are different concrete bugs." }, { - "left": "issue:44279", - "right": "issue:45405", + "left": "issue:43577", + "right": "issue:45216", "accept": false, - "reason": "General dependency issue and unreleased PEFT version bump are not the same underlying defect." + "reason": "BLIP2 dtype propagation and Qwen3.5 checkpoint saving are unrelated." }, { - "left": "issue:43299", - "right": "issue:43854", + "left": "issue:43381", + "right": "issue:45593", "accept": false, - "reason": "Different model families and different loading failures; too broad to merge." + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary loss behavior are unrelated." }, { - "left": "issue:44462", - "right": "issue:44779", + "left": "issue:44351", + "right": "issue:45278", "accept": false, - "reason": "Both touch tokenizers, but one ignores a repository tokenizer.json while the other returns incorrect DeepSeek tokenization; not the same bug." + "reason": "The latter is a broad umbrella for many import errors, but the pair does not identify the same concrete import failure." }, { - "left": "issue:39692", - "right": "issue:43901", + "left": "issue:43976", + "right": "issue:45405", "accept": false, - "reason": "Documentation problems, but one is a SigLIP2 example with multiple runtime errors and the other is a stale pipeline doc note." + "reason": "Python support compatibility and an unreleased PEFT minimum version are different release-management issues." }, { - "left": "issue:43295", - "right": "issue:44779", + "left": "issue:41950", + "right": "issue:45020", "accept": false, - "reason": "Processor/tokenizer API regression versus incorrect DeepSeek tokenization; different failure modes." + "reason": "Both mention model-loading/pipeline behavior, but image processor lookup in a video pipeline is a different bug from remote_code model loading regressions." }, { - "left": "issue:38175", - "right": "issue:44617", + "left": "issue:43901", + "right": "issue:44496", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and Sam3Video CUDA OOM are unrelated model/runtime issues." + "reason": "A documentation/behavior mismatch and an unrecognized model/config issue are not the same underlying bug." }, { - "left": "issue:44016", - "right": "issue:44291", + "left": "issue:44242", + "right": "issue:45200", "accept": false, - "reason": "Notebook syntax error and init_empty_weights argument handling are completely different defects." + "reason": "Load-balancing loss computation and mm_token_type_ids defaults are separate model-training concerns." }, { - "left": "issue:43450", - "right": "issue:44279", + "left": "issue:43479", + "right": "issue:43525", "accept": false, - "reason": "Batched video processor shape bug is unrelated to a generic dependency issue." + "reason": "Phi4MultimodalConfig default initialization and Llama4Config pad_token_id absence are unrelated config bugs." }, { - "left": "issue:44261", - "right": "issue:44263", + "left": "issue:44462", + "right": "issue:45081", "accept": false, - "reason": "Precision/config issue in MLA q_a_layernorm versus a torch.split return-value issue in GlmMoeDsaIndexer; different code paths." + "reason": "Tokenizer.json precedence and a Mistral regex patch crash are different tokenizer-loading problems." } ] }, @@ -9679,275 +9497,290 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:41628", + "issue:38175", + "issue:39692", + "issue:41950", "issue:42175", "issue:42994", "issue:43065", "issue:43122", "issue:43262", + "issue:43295", + "issue:43299", "issue:43317", - "issue:43352", "issue:43441", - "issue:43475", - "issue:43525", + "issue:43450", "issue:43650", "issue:43653", "issue:43673", "issue:43723", "issue:43761", "issue:43784", - "issue:43824", - "issue:43881", + "issue:43854", "issue:43901", "issue:43976", - "issue:43994", "issue:44016", "issue:44162", - "issue:44188", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44291", "issue:44297", - "issue:44336", - "issue:44361", + "issue:44462", "issue:44466", + "issue:44496", "issue:44521", - "issue:44589", + "issue:44617", "issue:44623", "issue:44671", "issue:44683", "issue:44749", "issue:44779", + "issue:44877", "issue:44898", "issue:44987", - "issue:45250", "issue:45276", "issue:45290", - "issue:45362", "issue:45397", + "issue:45405", "issue:45542", - "issue:45561", - "issue:45563", "issue:45593" ], "soft_pairs": [ + "issue:44683|issue:45542", + "issue:44496|issue:44877", + "issue:44297|issue:44462", + "issue:44279|issue:45405", + "issue:43299|issue:43854", + "issue:41950|issue:43650", + "issue:44462|issue:44779", + "issue:39692|issue:43901", + "issue:43295|issue:44779", + "issue:38175|issue:44617", + "issue:44016|issue:44291", + "issue:44261|issue:44263", "issue:43723|issue:44162", + "issue:43450|issue:44279", "issue:42994|issue:43317", "issue:43065|issue:44898", "issue:43441|issue:44683", - "issue:44466|issue:45276", "issue:43650|issue:45290", + "issue:44466|issue:45276", "issue:43262|issue:44521", "issue:44623|issue:45397", "issue:44623|issue:44749", - "issue:43784|issue:45542", "issue:43673|issue:43761", + "issue:43784|issue:45542", "issue:44671|issue:44779", - "issue:43122|issue:43650", - "issue:43901|issue:43994", - "issue:43653|issue:45593", "issue:42175|issue:43976", + "issue:43122|issue:43650", "issue:44987|issue:45542", - "issue:44361|issue:45593", - "issue:43824|issue:43976", - "issue:44297|issue:45397", - "issue:43475|issue:44016", - "issue:43525|issue:43881", - "issue:45561|issue:45563", - "issue:43352|issue:45250", - "issue:44589|issue:45362", - "issue:44188|issue:44336", - "issue:41628|issue:43976" + "issue:43653|issue:45593" ], - "prepared_review_unit_hash": "e992243d7a35d6de35e2d869b22d648f6c0ea7792a032d4629175324e23739ba", + "prepared_review_unit_hash": "d9725c8c877b54420f59f9685a27a17f9861fc869b61c2800a999a94cdf5205a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 34989, - "estimated_input_tokens": 8748, - "estimated_eval_tokens": 17752 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36273, + "estimated_input_tokens": 9069, + "estimated_eval_tokens": 18394 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is heterogeneous: the soft pairs span different models, subsystems, and failure modes, so none look like true duplicates. If one issue must stand in as a representative, 44521 is the strongest standalone report because it is detailed, reproducible, open, and already referenced by others.", - "confidence": 0.71, - "canonical_issue_reason": "No single canonical duplicate emerges from this mix of unrelated bugs. Issue 44521 is the best representative only as a practical stand-in: it is a concrete, reproducible regression with multiple inbound references.", + "summary": "This cluster is noisy: most items are unrelated v5 regressions across tokenizer loading, processors, quantization, vision, audio, and docs. None of the suggested soft edges look like true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue is strong here; the set mixes several different bug classes and model families rather than one underlying defect.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44521 is the strongest standalone bug report in the set: open, well-described, reproducible, and externally referenced.", + "best_issue_reason": "issue:43723 is the broadest representative of the recurring tokenizer-loading regression theme in v5, but it still does not unify the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43723", - "right": "issue:44162", + "left": "issue:44683", + "right": "issue:45542", "accept": false, - "reason": "Both are loading regressions, but one is AutoTokenizer in v5 and the other is an ESM2/model-specific failure; different code paths." + "reason": "Unrelated domains: flex_attention on torch>=2.9 vs TensorFlow backend/package installation." }, { - "left": "issue:42994", - "right": "issue:43317", + "left": "issue:44496", + "right": "issue:44877", "accept": false, - "reason": "Quantized save failure vs dequantized offload load failure; opposite ends of the workflow and different root causes." + "reason": "Both involve model/config loading, but the failures are different: unrecognized model type vs strict config rejection for granite_speech." }, { - "left": "issue:43065", - "right": "issue:44898", + "left": "issue:44297", + "right": "issue:44462", "accept": false, - "reason": "Different vision model bugs with no shared concrete code path." + "reason": "Both touch tokenizer serialization/loading, but one is save_pretrained tokenizer_class mismatch and the other is AutoTokenizer ignoring tokenizer.json." }, { - "left": "issue:43441", - "right": "issue:44683", + "left": "issue:44279", + "right": "issue:45405", "accept": false, - "reason": "Both mention FlashAttention, but one is model support gating and the other is a torch>=2.9 flex_attention compilation bug." + "reason": "Dependency complaints, but about different packages and failures; not the same bug." }, { - "left": "issue:44466", - "right": "issue:45276", + "left": "issue:43299", + "right": "issue:43854", "accept": false, - "reason": "Serialization of tied weights vs resize_token_embeddings propagation; related theme, different bug." + "reason": "Different models and different loading failures; only a broad 'model loading broke in v5' theme is shared." }, { - "left": "issue:43650", - "right": "issue:45290", + "left": "issue:41950", + "right": "issue:43650", "accept": false, - "reason": "Unrelated issue titles and failure modes; no shared underlying bug." + "reason": "Completely unrelated: video-classification processor bug vs an unrelated placeholder issue." }, { - "left": "issue:43262", - "right": "issue:44521", + "left": "issue:44462", + "right": "issue:44779", "accept": false, - "reason": "Both touch apply_chat_template, but one is audio sampling-rate defaulting and the other is multimodal assistant-mask generation." + "reason": "Tokenizer-related, but one is repository tokenizer.json precedence and the other is Deepseek tokenization correctness; not the same concrete path." }, { - "left": "issue:44623", - "right": "issue:45397", + "left": "issue:39692", + "right": "issue:43901", "accept": false, - "reason": "save_pretrained missing files vs from_pretrained with zero3 loading; different directions and failure mechanisms." + "reason": "Both are docs-related, but one is SigLIP2 example errors and the other is a TextClassificationPipeline docs mismatch." }, { - "left": "issue:44623", - "right": "issue:44749", + "left": "issue:43295", + "right": "issue:44779", "accept": false, - "reason": "Processor serialization bug vs a data-filtering performance regression; unrelated." + "reason": "Different regressions: processor.tokenizer/image passing breakage vs Deepseek tokenizer output changes." }, { - "left": "issue:43784", - "right": "issue:45542", + "left": "issue:38175", + "right": "issue:44617", "accept": false, - "reason": "Both are backend/import errors, but one is missing nn on import and the other is TensorFlow backend detection without TensorFlow." + "reason": "Different subsystems and symptoms: zero probabilities on SigLIP2 vs CUDA OOM in Sam3Video." }, { - "left": "issue:43673", - "right": "issue:43761", + "left": "issue:44016", + "right": "issue:44291", "accept": false, - "reason": "Generation cache/chunked_prefill bug vs CLIP hidden_states regression; different subsystems." + "reason": "Notebook syntax error vs init_empty_weights argument/type error; unrelated code paths." }, { - "left": "issue:44671", - "right": "issue:44779", + "left": "issue:44261", + "right": "issue:44263", "accept": false, - "reason": "CamemBERT MLM output bug vs Deepseek tokenizer regression; different model families." + "reason": "Both are model-specific correctness bugs, but they affect different components and failure modes." }, { - "left": "issue:43122", - "right": "issue:43650", + "left": "issue:43723", + "right": "issue:44162", "accept": false, - "reason": "Tokenizer regression vs an unrelated data request; no overlap." + "reason": "Tokenizer loading regression vs ESM2 model breakage; too broad to be duplicates." }, { - "left": "issue:43901", - "right": "issue:43994", + "left": "issue:43450", + "right": "issue:44279", "accept": false, - "reason": "Documentation mismatch vs a model/pipeline inference bug; not the same issue." + "reason": "Video processor batching shape bug vs dependency issue; unrelated." }, { - "left": "issue:43653", - "right": "issue:45593", + "left": "issue:42994", + "right": "issue:43317", "accept": false, - "reason": "BigBird tokenizer special-token bug vs D-FINE auxiliary-loss behavior; unrelated." + "reason": "Both involve quantization/dequantization, but one is saving and the other is device_map offload loading; not the same bug." }, { - "left": "issue:42175", - "right": "issue:43976", + "left": "issue:43065", + "right": "issue:44898", "accept": false, - "reason": "Packaging/backend dependency issue vs Python version compatibility problem; distinct root causes." + "reason": "Different model components: dummy Conv2d in Sam3PixelDecoder vs Perceiver interpolation failure." }, { - "left": "issue:44987", - "right": "issue:45542", + "left": "issue:43441", + "right": "issue:44683", "accept": false, - "reason": "A specific model load failure vs a missing TensorFlow backend error; different bugs." + "reason": "Both mention attention kernels, but the concrete failures are different code paths and different triggering conditions." }, { - "left": "issue:44361", - "right": "issue:45593", + "left": "issue:43650", + "right": "issue:45290", "accept": false, - "reason": "Tokenizer AttributeError on tasks vs detection training-loss behavior; unrelated." + "reason": "Completely unrelated issues." }, { - "left": "issue:43824", - "right": "issue:43976", + "left": "issue:44466", + "right": "issue:45276", "accept": false, - "reason": "Specific model import failure vs Python-version compatibility regression; different failure modes." + "reason": "Both touch Gemma 4 embeddings, but one is serialization of lm_head.weight and the other is resize_token_embeddings not propagating to outputs." }, { - "left": "issue:44297", + "left": "issue:43262", + "right": "issue:44521", + "accept": false, + "reason": "Both use apply_chat_template, but audio sampling-rate defaults and multimodal assistant masks are different bugs." + }, + { + "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Tokenizer serialization metadata bug vs model loading under zero3; not the same code path." + "reason": "processor.save_pretrained missing files vs gemma-4 zero3 from_pretrained; save/load failures but not the same defect." }, { - "left": "issue:43475", - "right": "issue:44016", + "left": "issue:44623", + "right": "issue:44749", "accept": false, - "reason": "Model output attribute bug vs a notebook syntax error; unrelated." + "reason": "Missing processor files vs performance regression in filtering after upgrade; unrelated." }, { - "left": "issue:43525", - "right": "issue:43881", + "left": "issue:43673", + "right": "issue:43761", "accept": false, - "reason": "Missing config attribute in one model vs load failure in another model; different issues." + "reason": "Generation cache/chunked_prefill issue vs CLIPVisionModel hidden_states regression; different models and APIs." }, { - "left": "issue:45561", - "right": "issue:45563", + "left": "issue:43784", + "right": "issue:45542", "accept": false, - "reason": "pytest-xdist file race vs a stale generation warning; unrelated." + "reason": "Import NameError in sentence-transformers integration vs TensorFlow backend installation issue." }, { - "left": "issue:43352", - "right": "issue:45250", + "left": "issue:44671", + "right": "issue:44779", "accept": false, - "reason": "Specific model unsupported by Flash Attention 2.0 vs a generic Flash Attention issue; not enough to be the same bug." + "reason": "Both are v5 regressions affecting NLP outputs, but one is CamemBERT masked LM predictions and the other is Deepseek tokenizer behavior." }, { - "left": "issue:44589", - "right": "issue:45362", + "left": "issue:42175", + "right": "issue:43976", "accept": false, - "reason": "Storage deserialization error vs chat crash on Qwen3.5; different failure mechanisms." + "reason": "Backend packaging issue vs Python-version compatibility issue; not duplicates." }, { - "left": "issue:44188", - "right": "issue:44336", + "left": "issue:43122", + "right": "issue:43650", "accept": false, - "reason": "torch.compile attention-kernel divergence vs ANSI code emission in loading reports; unrelated." + "reason": "Tokenizer behavior regression vs unrelated placeholder issue." }, { - "left": "issue:41628", - "right": "issue:43976", + "left": "issue:44987", + "right": "issue:45542", + "accept": false, + "reason": "Model loading failure vs TensorFlow installation/backend issue; unrelated." + }, + { + "left": "issue:43653", + "right": "issue:45593", "accept": false, - "reason": "AutoImageProcessor import error vs a Python version compatibility bug; different underlying problems." + "reason": "BigBirdTokenizer special-token bug vs D-FINE auxiliary-loss behavior; unrelated." } ] }, @@ -9969,290 +9802,297 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36246", "issue:41628", "issue:41950", "issue:42491", "issue:42757", - "issue:42886", - "issue:43066", "issue:43295", "issue:43299", - "issue:43441", + "issue:43352", + "issue:43475", + "issue:43525", "issue:43526", "issue:43531", - "issue:43540", - "issue:43596", "issue:43673", + "issue:43824", + "issue:43881", "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", "issue:44162", + "issue:44188", "issue:44220", "issue:44242", "issue:44246", "issue:44295", + "issue:44297", "issue:44336", "issue:44351", + "issue:44361", "issue:44373", "issue:44462", "issue:44485", "issue:44496", "issue:44561", + "issue:44589", "issue:44617", "issue:44661", - "issue:44749", - "issue:44779", "issue:44857", - "issue:44871", - "issue:44877", - "issue:44987", "issue:45003", "issue:45200", - "issue:45229", - "issue:45292", + "issue:45250", "issue:45356", "issue:45362", - "issue:45375", "issue:45397", - "issue:45459", "issue:45542", "issue:45561", + "issue:45563", "issue:45593" ], "soft_pairs": [ + "issue:43901|issue:43994", + "issue:44361|issue:45593", + "issue:43824|issue:43976", + "issue:44297|issue:45397", + "issue:43475|issue:44016", + "issue:43525|issue:43881", + "issue:45561|issue:45563", + "issue:43352|issue:45250", + "issue:44589|issue:45362", + "issue:44188|issue:44336", + "issue:41628|issue:43976", "issue:43295|issue:44617", "issue:42491|issue:43299", + "issue:44336|issue:44373", "issue:41950|issue:44220", "issue:43901|issue:44246", - "issue:44336|issue:44373", "issue:44242|issue:45593", "issue:44295|issue:45397", "issue:44462|issue:45356", "issue:44661|issue:45542", - "issue:44246|issue:44496", "issue:45003|issue:45561", "issue:43531|issue:45362", + "issue:44246|issue:44496", "issue:36246|issue:42757", "issue:36246|issue:44351", "issue:43673|issue:44561", - "issue:43526|issue:45003", "issue:44336|issue:44485", + "issue:43526|issue:45003", "issue:44162|issue:45200", - "issue:44242|issue:44857", - "issue:44246|issue:44871", - "issue:36246|issue:43596", - "issue:44295|issue:44749", - "issue:43441|issue:43540", - "issue:43066|issue:44779", - "issue:44877|issue:45375", - "issue:44987|issue:45459", - "issue:44779|issue:45292", - "issue:42886|issue:45229", - "issue:41628|issue:41950" + "issue:44242|issue:44857" ], - "prepared_review_unit_hash": "6807feb3aee73e8946152fe25254ed5747d0ff7d3f1ed83663e1d5d0cdadcfc0", + "prepared_review_unit_hash": "351a01c8be83cb7059d32306ef1c7ae50421a07d66f58637ff8e69f87de47b17", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36119, - "estimated_input_tokens": 9030, - "estimated_eval_tokens": 18316 + "soft_pair_count": 30, + "serialized_chars": 36322, + "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 broadly heterogeneous: most items are unrelated Transformer regressions, import errors, tokenizer/config issues, or model-specific bugs. None of the soft pairs are close enough to confidently merge as duplicates, so I rejected all soft edges and treated issue:36246 as the best representative of the small import-error subgroup.", - "confidence": 0.58, - "canonical_issue_reason": "issue:36246 is the clearest representative of the recurring concrete import-error pattern in this set, and it is closer to the other symbol/import regressions than the more generic or model-specific reports.", - "canonical_pr_reason": "No PRs were provided in this cluster.", - "best_issue_reason": "issue:36246 is the best overall fit for the cluster\u2019s duplicate-like theme because it is a specific, user-facing import failure and sits closest to the small group of similar import regressions.", - "best_pr_reason": "No PRs were provided in this cluster.", + "summary": "This is a heterogeneous cluster of unrelated issues: import/export regressions, model-loading failures, pipeline behavior bugs, documentation problems, and testing/runtime edge cases. None of the proposed soft pairs looks like the same underlying bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 45003 is the broadest open bug in the set and has a concrete implementation focus in core code (`modeling_utils` / `sys.modules`). The other issues are mostly narrow, model-specific, or unrelated regressions, so 45003 is the least arbitrary cluster anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "45003 is the best representative issue because it is open, implementation-oriented, and broader than the many one-off model/import/doc issues. It is still not a true duplicate of the others.", + "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Docs wording about `return_all_scores` is unrelated to a SigLIP2 model/pipeline correctness problem." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer AttributeError on tasks and D-FINE auxiliary-loss behavior affect different code paths and model families." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Missing `Qwen2_5_VLForConditionalGeneration` export is a specific symbol issue, not the Python-version compatibility problem described by 43976." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer config mismatch in `save_pretrained` is unrelated to Gemma-4 `from_pretrained`/zero3 loading." + }, + { + "left": "issue:43475", + "right": "issue:44016", + "accept": false, + "reason": "A SAM3 vision encoder attribute error is unrelated to a notebook syntax error." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Missing `pad_token_id` on `Llama4Config` is a different loading bug than the GLM-4V-9B failure." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "Pytest xdist file-race behavior is unrelated to a stale `num_return_sequences` warning in paged generation." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "One is a model-specific unsupported feature error; the other is just the generic Flash Attention 2.0 topic." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage deserialization failure is unrelated to the Qwen3.5 chat crash." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI-code output in `loading_report`." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "`AutoImageProcessor` import failure is a different missing-symbol issue than the Python-version breakage in 43976." + }, { "left": "issue:43295", "right": "issue:44617", "accept": false, - "reason": "Different subsystems and symptoms: processor/tokenizer regression vs CUDA OOM in a video model." + "reason": "Processor/tokenizer regression in custom model code is unrelated to SAM3 video CUDA OOM." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Same broad Qwen3/MoE family, but the reported breakage differs enough that they do not clearly look like one underlying bug." + "reason": "Both mention Qwen3 MoE, but one is a LoRA compatibility problem and the other is a model-loading regression; they are not clearly the same fix." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI terminal output and a wrong docstring for `position_ids` are unrelated." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification image-processor lookup vs audio feature extraction are unrelated code paths." + "reason": "A video-classification pipeline bug and `_torch_extract_fbank_features()` are different pipeline/code-path issues." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Docs mismatch for return_all_scores vs intermittent import latency; no shared bug." - }, - { - "left": "issue:44336", - "right": "issue:44373", - "accept": false, - "reason": "Both are doc/reporting issues, but they concern unrelated files and behaviors." + "reason": "A pipeline docs mismatch is unrelated to intermittent import-time slowness." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Both mention missing loss terms, but they are different models and different conditions." + "reason": "Missing load-balancing loss with `output_router_logits=False` is a different training-loss bug than D-FINE auxiliary-loss handling." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "Buffer/position_ids error vs Gemma-4 zero3 loading problem; not the same failure mode." + "reason": "Reading `position_ids` after buffer registration is unrelated to Gemma-4 zero3 loading." }, { "left": "issue:44462", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but for different models and different tokenizer-codec behavior." + "reason": "Ignoring `tokenizer.json` is a tokenizer-loading bug, while 45356 is a codec/regex regression for Kimi-K2.5." }, { "left": "issue:44661", "right": "issue:45542", "accept": false, - "reason": "Model-mapping registration bug vs TensorBoard/TensorFlow backend setup issue." - }, - { - "left": "issue:44246", - "right": "issue:44496", - "accept": false, - "reason": "Import slowdown is unrelated to unrecognized-model config loading." + "reason": "`TOKENIZER_MAPPING_NAMES` handling and undefined TensorFlow backend detection are unrelated." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "Unsafe sys.modules access and pytest-xdist file races are unrelated." + "reason": "Unsafe `sys.modules` access in core modeling code is unrelated to the xdist captured-info race." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Different Qwen model variants and different runtime failures." + "reason": "Qwen3-MoE sliding-window behavior and Qwen3.5 chat crashes are different model/runtime bugs." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import slowness and unrecognized-model config loading are unrelated." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they involve different packages, symbols, and root causes." + "reason": "Different missing import symbols (`Qwen2_5_VLImageProcessor` vs `is_offline_mode`) with no shared code-path shown." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Different missing symbols from different modules; only the surface symptom matches." + "reason": "Missing `Qwen2_5_VLImageProcessor` and missing `HybridCache` are separate export/import failures." }, { "left": "issue:43673", "right": "issue:44561", "accept": false, - "reason": "Both are v5 regressions, but the cache/chunked_prefill issue is distinct from removal of torch-fx helpers." + "reason": "Chunked-prefill cache handling and removal of `is_torch_fx_available` affect different runtime paths." }, { - "left": "issue:43526", - "right": "issue:45003", + "left": "issue:44336", + "right": "issue:44485", "accept": false, - "reason": "Label-reduction behavior in an image processor is unrelated to sys.modules access." + "reason": "ANSI code emission is unrelated to GLM-5 RoPE implementation." }, { - "left": "issue:44336", - "right": "issue:44485", + "left": "issue:43526", + "right": "issue:45003", "accept": false, - "reason": "ANSI escape handling in reporting is unrelated to GLM-5 RoPE implementation." + "reason": "A BEiT image-processor label bug is unrelated to core modeling sys.modules access." }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 model breakage and Gemma-4 token-type defaults are different model-specific problems." + "reason": "ESM2 breakage and Gemma-4 `mm_token_type_ids` defaults are different model-specific regressions." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Missing auxiliary loss vs AMP/CUDA crash are different bug classes." - }, - { - "left": "issue:44246", - "right": "issue:44871", - "accept": false, - "reason": "Import latency is unrelated to Gemma-3 eos_token_id configuration." - }, - { - "left": "issue:36246", - "right": "issue:43596", - "accept": false, - "reason": "ImportError for a missing processor class vs a deepspeed/BertModel IndexError are unrelated." - }, - { - "left": "issue:44295", - "right": "issue:44749", - "accept": false, - "reason": "Position_ids buffer handling and post-upgrade slowdown are unrelated." - }, - { - "left": "issue:43441", - "right": "issue:43540", - "accept": false, - "reason": "FlashAttention failure in Ministral-3 vs video-input validation in Qwen3OmniMoe are unrelated." - }, - { - "left": "issue:43066", - "right": "issue:44779", - "accept": false, - "reason": "Tokenizer decoder-type mismatch vs DeepSeek tokenizer incorrect results; not the same root cause." - }, - { - "left": "issue:44877", - "right": "issue:45375", - "accept": false, - "reason": "Both involve config strictness, but they affect different model configs and different missing-field behavior." - }, - { - "left": "issue:44987", - "right": "issue:45459", - "accept": false, - "reason": "Model loading on transformers>=5.1.0 and hidden tokenizer errors from missing protobuf are separate issues." - }, - { - "left": "issue:44779", - "right": "issue:45292", - "accept": false, - "reason": "Tokenizer output regression vs resize_token_embeddings not updating output embeddings are unrelated." - }, - { - "left": "issue:42886", - "right": "issue:45229", - "accept": false, - "reason": "Offline cache loading and multi-GPU CUDA OOM are different failure modes." - }, - { - "left": "issue:41628", - "right": "issue:41950", - "accept": false, - "reason": "Missing AutoImageProcessor import and video pipeline image-processor lookup are related only broadly, not the same bug." + "reason": "Load-balancing loss omission is unrelated to LwDetrImageLoss AMP crashes." } ] }, @@ -10274,295 +10114,290 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:28282", - "issue:30990", - "issue:42673", + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42886", "issue:43065", + "issue:43066", "issue:43232", "issue:43257", "issue:43262", "issue:43408", + "issue:43441", "issue:43475", "issue:43540", - "issue:43646", + "issue:43596", "issue:43653", "issue:43673", - "issue:43704", "issue:43901", - "issue:43931", "issue:44060", "issue:44077", "issue:44162", "issue:44246", - "issue:44297", + "issue:44295", "issue:44336", - "issue:44361", "issue:44368", "issue:44466", "issue:44496", - "issue:44623", "issue:44734", "issue:44749", - "issue:44869", - "issue:44933", + "issue:44779", + "issue:44871", + "issue:44877", "issue:44977", + "issue:44987", "issue:44995", "issue:45003", "issue:45137", "issue:45200", - "issue:45278", + "issue:45229", "issue:45292", "issue:45335", - "issue:45356", + "issue:45375", "issue:45397", "issue:45459", "issue:45563", "issue:45584", "issue:45588", - "issue:45600", "issue:45656", "issue:45701" ], "soft_pairs": [ + "issue:44246|issue:44871", + "issue:36246|issue:43596", + "issue:44295|issue:44749", + "issue:43066|issue:44779", + "issue:43441|issue:43540", + "issue:44877|issue:45375", + "issue:44987|issue:45459", + "issue:42886|issue:45229", + "issue:41628|issue:41950", + "issue:44779|issue:45292", "issue:45292|issue:45584", - "issue:45292|issue:45701", "issue:43065|issue:43408", + "issue:43232|issue:44336", + "issue:45292|issue:45701", "issue:43901|issue:45003", "issue:44995|issue:45200", - "issue:43232|issue:44336", "issue:43257|issue:45137", - "issue:44496|issue:45003", "issue:44060|issue:44077", - "issue:43653|issue:44162", + "issue:44496|issue:45003", "issue:43475|issue:43540", + "issue:43653|issue:44162", "issue:44995|issue:45588", - "issue:43262|issue:44246", "issue:44368|issue:45563", - "issue:44466|issue:45335", + "issue:43262|issue:44246", "issue:44368|issue:44977", "issue:43673|issue:44734", - "issue:45397|issue:45656", - "issue:44933|issue:45459", - "issue:44869|issue:45356", - "issue:45137|issue:45600", - "issue:43931|issue:44368", - "issue:43673|issue:45278", - "issue:43232|issue:43673", - "issue:44297|issue:44623", - "issue:43646|issue:44162", - "issue:44361|issue:44749", - "issue:44623|issue:45656", - "issue:28282|issue:30990", - "issue:42673|issue:43704" + "issue:44466|issue:45335", + "issue:45397|issue:45656" ], - "prepared_review_unit_hash": "6612f2d1bbb3029566ad87ce03bfb80f9571cf59310771857b177d4c67470ec2", + "prepared_review_unit_hash": "fe0fdc116d6a004fdd9dc12508a956159014ee18625f52f0c6f1a6f1e8dc466a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36680, - "estimated_input_tokens": 9170, - "estimated_eval_tokens": 18596 + "soft_pair_count": 28, + "serialized_chars": 36192, + "estimated_input_tokens": 9048, + "estimated_eval_tokens": 18352 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Only one soft edge looks like a true duplicate: the two Qwen3ForCausalLM VRAM leak reports. The rest share broad areas (generation, tokenizers, DeepSpeed, SAM3, etc.) but describe different concrete bugs or regressions.", - "confidence": 0.96, - "canonical_issue_reason": "issue:42673 is the earlier, more detailed report of the exact same Qwen3ForCausalLM VRAM leak, and it already has more discussion/context than issue:43704.", + "summary": "The cluster is mostly heterogeneous, with only loose thematic overlap around Transformers v5 regressions, tokenizers, and embedding/tied-weight behavior. I do not see any soft pair that is clearly the same underlying bug or change, so I rejected all candidate duplicates.", + "confidence": 0.39, + "canonical_issue_reason": "Issue 45292 is the clearest standalone bug report: it names a concrete, reproducible API regression in `resize_token_embeddings` and is phrased broadly enough to serve as a duplicate target for closely related embedding-resize reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42673 is the best representative issue for the accepted duplicate pair: it is the canonical bug report, older, and has the fuller discussion trail.", + "best_issue_reason": "Issue 45292 is the best issue candidate in this set because it describes a specific code-path failure cleanly and is more reusable as a triage target than the more model-specific or symptom-heavy reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45292", - "right": "issue:45584", + "left": "issue:44246", + "right": "issue:44871", "accept": false, - "reason": "Resize-token-embeddings bug vs Whisper empty-transcription generation failure; different code paths and symptoms." + "reason": "Import performance and Gemma EOS configuration are unrelated bugs." }, { - "left": "issue:45292", - "right": "issue:45701", + "left": "issue:36246", + "right": "issue:43596", "accept": false, - "reason": "Embedding resize/serialization behavior vs version-dependent tokenization changes; not the same bug." + "reason": "Different failures: an import error for a vision image processor vs a DeepSpeed/Bert forward crash." }, { - "left": "issue:43065", - "right": "issue:43408", + "left": "issue:44295", + "right": "issue:44749", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d implementation detail and the other is a model-type warning; different underlying issues." + "reason": "Registering `position_ids` as a buffer is unrelated to the reported tokenizer slowdown after upgrade." }, { - "left": "issue:43901", - "right": "issue:45003", + "left": "issue:43066", + "right": "issue:44779", "accept": false, - "reason": "Docs mismatch for return_all_scores vs unsafe sys.modules access in modeling_utils; unrelated." + "reason": "Both are tokenizer-related, but one is a wrong decoder-type warning and the other is incorrect Deepseek tokenization; not the same concrete bug." }, { - "left": "issue:44995", - "right": "issue:45200", + "left": "issue:43441", + "right": "issue:43540", "accept": false, - "reason": "Stale second-forward cache crash vs missing default mm_token_type_ids; different failures." + "reason": "FlashAttention failure for Ministral-3 is unrelated to Qwen3OmniMoe video input processing." }, { - "left": "issue:43232", - "right": "issue:44336", + "left": "issue:44877", + "right": "issue:45375", "accept": false, - "reason": "Generation kwargs handling after sync_gpus vs ANSI codes in loading_report; unrelated." + "reason": "Both involve config handling, but they describe different model fields and opposite failure modes, so they are not the same bug." }, { - "left": "issue:43257", - "right": "issue:45137", + "left": "issue:44987", + "right": "issue:45459", "accept": false, - "reason": "Both involve DeepSpeed, but one is MOE weight conversion during loading and the other is a ZeRO3 deque underflow; different code paths." + "reason": "Model-loading failure for `physical-intelligence/fast` is unrelated to protobuf-masked tokenizer errors." }, { - "left": "issue:44496", - "right": "issue:45003", + "left": "issue:42886", + "right": "issue:45229", "accept": false, - "reason": "Unrecognized model/config loading error vs unsafe sys.modules access; not the same issue." + "reason": "Offline cache loading and Gemma4 CUDA OOM are unrelated." }, { - "left": "issue:44060", - "right": "issue:44077", + "left": "issue:41628", + "right": "issue:41950", "accept": false, - "reason": "Incorrect tied-weights warning vs optional post_init support policy change; different bugs." + "reason": "Missing `AutoImageProcessor` import and a video-classification pipeline misdetecting processors are different issues." }, { - "left": "issue:43653", - "right": "issue:44162", + "left": "issue:44779", + "right": "issue:45292", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug vs ESM2 being broken; unrelated." + "reason": "Tokenizer correctness regression and embedding resize not updating output embeddings are different code paths." }, { - "left": "issue:43475", - "right": "issue:43540", + "left": "issue:45292", + "right": "issue:45584", "accept": false, - "reason": "SAM3 VisionEncoderOutput missing an attribute vs Qwen3OmniMoe video-input ValueError; different problems." + "reason": "Embedding resize behavior is unrelated to Whisper empty-transcription generation failure." }, { - "left": "issue:44995", - "right": "issue:45588", + "left": "issue:43065", + "right": "issue:43408", "accept": false, - "reason": "Stale cache crash vs flash-attention AttributeError on s_aux=None; different failures." + "reason": "SAM3 dummy Conv2d implementation and SAM3 model-type warning are not the same bug." }, { - "left": "issue:43262", - "right": "issue:44246", + "left": "issue:43232", + "right": "issue:44336", "accept": false, - "reason": "Audio processor sampling-rate default bug vs import-time slowdown; unrelated." + "reason": "Generation kwargs/sync-gpus logic has nothing to do with ANSI code emission in loading reports." }, { - "left": "issue:44368", - "right": "issue:45563", + "left": "issue:45292", + "right": "issue:45701", "accept": false, - "reason": "Tied-embeddings warning text vs paged-generate stale num_return_sequences warning; different warnings." + "reason": "Embedding resizing and version-dependent tokenization changes are different regressions." }, { - "left": "issue:44466", - "right": "issue:45335", + "left": "issue:43901", + "right": "issue:45003", "accept": false, - "reason": "Both concern embedding weights, but one is device-dependent serialization and the other is resize_token_embeddings not updating decoder.embed_tokens; different bugs." + "reason": "A docs mismatch for `return_all_scores` is unrelated to unsafe `sys.modules` access." }, { - "left": "issue:44933", - "right": "issue:45459", + "left": "issue:44995", + "right": "issue:45200", "accept": false, - "reason": "Missing image_utils import vs protobuf exception masking tokenizer errors; unrelated." + "reason": "A second-forward cache crash in GlmMoeDsa is unrelated to Gemma4 token type defaults." }, { - "left": "issue:44869", - "right": "issue:45356", + "left": "issue:43257", + "right": "issue:45137", "accept": false, - "reason": "Whisper timestamp decode crash on trailing replacement character vs Kimi-K2.5 codec-handling regression; similar area, different bug." + "reason": "Both touch DeepSpeed, but one is MOE weight conversion and the other is a ZeRO3 deque error; not the same bug." }, { - "left": "issue:45137", - "right": "issue:45600", + "left": "issue:44060", + "right": "issue:44077", "accept": false, - "reason": "ZeRO3 deque pop error vs stale auto_mappings references to removed configs; unrelated." + "reason": "Tied-weights warning in Qwen3-Next and `post_init` allowance in patchtsmixer are unrelated." }, { - "left": "issue:43931", - "right": "issue:44368", + "left": "issue:44496", + "right": "issue:45003", "accept": false, - "reason": "Model weight-shape mismatch vs tied-weights warning; not the same issue." + "reason": "Model config loading failure and unsafe `sys.modules` access are not the same underlying issue." }, { - "left": "issue:43673", - "right": "issue:45278", + "left": "issue:43475", + "right": "issue:43540", "accept": false, - "reason": "Generation cache missing in chunked prefill vs broad import errors after upgrade; different scope and symptoms." + "reason": "Missing `fpn_position_embeddings` in SAM3 output is unrelated to Qwen3OmniMoe video processing." }, { - "left": "issue:43232", - "right": "issue:43673", + "left": "issue:43653", + "right": "issue:44162", "accept": false, - "reason": "Both are generation-related, but one is sync_gpus kwargs handling and the other is missing GenerationMixin cache in chunked prefill; different code paths." + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are different tokenizer/model problems." }, { - "left": "issue:44297", - "right": "issue:44623", + "left": "issue:44995", + "right": "issue:45588", "accept": false, - "reason": "tokenizer.save_pretrained metadata mismatch vs processor.save_pretrained missing files; related API surface, different bug." + "reason": "Stale indexer cache crash and FlashAttention `s_aux=None` crash are different failures." }, { - "left": "issue:43646", - "right": "issue:44162", + "left": "issue:44368", + "right": "issue:45563", "accept": false, - "reason": "Transformers 5.0.0 custom model init regression vs ESM2 breakage; unrelated." + "reason": "A tied-word-embeddings warning and a stale `num_return_sequences` warning are not the same bug." }, { - "left": "issue:44361", - "right": "issue:44749", + "left": "issue:43262", + "right": "issue:44246", "accept": false, - "reason": "MLukeTokenizer AttributeError vs a performance regression after version upgrade; different problems." + "reason": "Audio chat-template sampling-rate defaults and import latency are unrelated." }, { - "left": "issue:44623", - "right": "issue:45656", + "left": "issue:44368", + "right": "issue:44977", "accept": false, - "reason": "Missing processor files on save vs optimizer.step being called twice under DeepSpeed; unrelated." + "reason": "A tied-embedding warning is not the same as a generation failure with FlashAttention." }, { - "left": "issue:28282", - "right": "issue:30990", + "left": "issue:43673", + "right": "issue:44734", "accept": false, - "reason": "Missing PyTorch import error vs Sentence Transformers loading hang; different failures." + "reason": "Generation cache missing during chunked prefill and a KV-cache indexing crash in transformers serve are different code paths." }, { - "left": "issue:42673", - "right": "issue:43704", - "accept": true, - "reason": "Same exact Qwen3ForCausalLM VRAM leak report in multiple dataloader threads; same underlying bug and same fix candidate." + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both concern weights/embeddings, but one is serialization inconsistency and the other is resize propagation; they are not clearly the same bug." }, { "left": "issue:45397", "right": "issue:45656", "accept": false, - "reason": "Gemma-4 Zero3 from_pretrained issue vs double optimizer-step under DeepSpeed; different bugs." + "reason": "Gemma4 Zero3 loading and duplicate optimizer stepping under DeepSpeed are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The accepted duplicate pair is well grounded: the two Qwen3ForCausalLM issues have identical titles, overlapping symptoms, and the older thread is the better canonical representative. The remaining soft-edge verdicts stay conservative and are supported by the packet-level titles/descriptions; none overstate duplication or mergeability." - }, - "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 }, @@ -10577,66 +10412,78 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ + "issue:28282", "issue:30990", - "issue:35141", - "issue:36010", "issue:41628", "issue:42222", "issue:42371", + "issue:42673", "issue:42831", - "issue:42898", "issue:42994", "issue:43232", - "issue:43377", "issue:43475", - "issue:43479", "issue:43482", - "issue:43504", "issue:43525", - "issue:43526", "issue:43643", + "issue:43646", "issue:43673", + "issue:43704", "issue:43873", + "issue:43931", "issue:44038", - "issue:44079", - "issue:44164", + "issue:44162", "issue:44188", "issue:44222", - "issue:44242", - "issue:44351", - "issue:44496", - "issue:44568", + "issue:44297", + "issue:44361", + "issue:44368", "issue:44623", "issue:44661", "issue:44749", - "issue:44792", + "issue:44869", "issue:44933", "issue:44945", "issue:45081", "issue:45092", "issue:45103", + "issue:45137", "issue:45200", "issue:45230", + "issue:45278", "issue:45310", + "issue:45356", "issue:45362", "issue:45375", "issue:45381", "issue:45397", + "issue:45459", "issue:45478", - "issue:45561", + "issue:45600", "issue:45656" ], "soft_pairs": [ - "issue:44749|issue:45310", - "issue:42994|issue:43482", - "issue:42371|issue:45092", - "issue:44749|issue:45656", + "issue:44933|issue:45459", + "issue:44869|issue:45356", + "issue:43931|issue:44368", + "issue:45137|issue:45600", + "issue:43232|issue:43673", + "issue:43673|issue:45278", + "issue:44297|issue:44623", + "issue:43646|issue:44162", + "issue:44361|issue:44749", + "issue:44623|issue:45656", + "issue:28282|issue:30990", + "issue:44749|issue:45310", + "issue:42994|issue:43482", + "issue:42371|issue:45092", "issue:44038|issue:44623", + "issue:44749|issue:45656", + "issue:42673|issue:43704", "issue:43475|issue:43643", - "issue:45375|issue:45478", "issue:45230|issue:45397", + "issue:45375|issue:45478", "issue:43475|issue:45381", "issue:41628|issue:44933", "issue:45310|issue:45375", @@ -10644,258 +10491,235 @@ "issue:42994|issue:43525", "issue:43475|issue:44661", "issue:45362|issue:45375", - "issue:43873|issue:44945", "issue:42831|issue:44222", + "issue:43873|issue:44945", "issue:43232|issue:44188", - "issue:42222|issue:45200", - "issue:41628|issue:44351", - "issue:44222|issue:45310", - "issue:42994|issue:44164", - "issue:44038|issue:45310", - "issue:43526|issue:45561", - "issue:36010|issue:41628", - "issue:44496|issue:45561", - "issue:43673|issue:44188", - "issue:43504|issue:45561", - "issue:43377|issue:44242", - "issue:43232|issue:44792", - "issue:42898|issue:44568", - "issue:30990|issue:35141", - "issue:43479|issue:44079" + "issue:42222|issue:45200" ], - "prepared_review_unit_hash": "c60c1db150050724f354a4526abc11ed255e47693d0cd356da872a8667e9b13f", + "prepared_review_unit_hash": "52199f130e253451c8936cfbb0c44da0f5db32a34afedf23058c0eb30f7158e2", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37335, - "estimated_input_tokens": 9334, - "estimated_eval_tokens": 18924 + "soft_pair_count": 31, + "serialized_chars": 36717, + "estimated_input_tokens": 9180, + "estimated_eval_tokens": 18616 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is a set of mostly unrelated Transformers issues spanning imports, model configs, generation, quantization, and tokenizer behavior. None of the soft pairs look like true duplicates of the same underlying bug/change; several are only loosely related by model family or subsystem.", - "confidence": 0.81, - "canonical_issue_reason": "issue:45310 is the clearest concrete bug report in the set and a plausible representative of the Qwen3.5-Moe from_pretrained failure family; it is more specific than the vague or broader tickets.", + "summary": "This cluster is mostly heterogeneous Transformers breakages that share broad upgrade/import/model-loading themes, not one single bug. The clear duplicate pair in the soft-edge list is the repeated Qwen3ForCausalLM VRAM leak report (42673/43704), and the packet also contains another exact-title duplicate pair, 45310/45478, for the Qwen3.5 MoE from_pretrained error. Outside those, the issues are mostly only loosely related by subsystem or model family.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is still the best umbrella issue for the cluster because it is open and broadly representative of the upgrade/import-regression theme, but it should be treated as an umbrella, not as evidence that the whole set is duplicate-heavy.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the strongest single issue to anchor triage because it has a concrete, reproducible model-loading failure and a narrow title that maps to a specific code path.", + "best_issue_reason": "issue:45278 is the strongest global issue candidate because it is open and broad enough to anchor triage across several upgrade-related reports, while the more specific items are better treated as separate bugs or duplicate subclusters.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44749", - "right": "issue:45310", + "left": "issue:44933", + "right": "issue:45459", "accept": false, - "reason": "Both mention post-upgrade regressions, but one is a slow data-filtering/performance issue and the other is a Qwen3.5-Moe from_pretrained error; different bugs and code paths." + "reason": "Both touch tokenizer/import failures, but the concrete paths differ: image_utils import vs. protobuf-related tokenizer error masking." }, { - "left": "issue:42994", - "right": "issue:43482", + "left": "issue:44869", + "right": "issue:45356", "accept": false, - "reason": "Saving a quantized model and loading a Qwen2.5-GGUF model are opposite operations with different failure modes; not the same bug." + "reason": "Different model areas and symptoms: Whisper timestamp decoding crash vs. Kimi tokenizer codec/regression handling." }, { - "left": "issue:42371", - "right": "issue:45092", + "left": "issue:43931", + "right": "issue:44368", "accept": false, - "reason": "TF32 settings guidance and InternVL2 meta-initialization incompatibility are unrelated issues." + "reason": "Loading-time shape mismatch for a VL model and a tie_word_embeddings warning are not the same bug." }, { - "left": "issue:44749", - "right": "issue:45656", + "left": "issue:45137", + "right": "issue:45600", "accept": false, - "reason": "A tokenizer/data-processing slowdown is unrelated to a DeepSpeed optimizer step being called twice." + "reason": "DeepSpeed ZeRO3 deque error and stale auto_mappings config references are distinct code-path failures." }, { - "left": "issue:44038", - "right": "issue:44623", + "left": "issue:43232", + "right": "issue:43673", "accept": false, - "reason": "Qwen3-VL-Moe loading problems and processor.save_pretrained missing files are different failure points and not the same defect." + "reason": "Both are generation-related, but one is a sync_gpus kwargs update issue and the other is a missing cache path during chunked prefill." }, { - "left": "issue:43475", - "right": "issue:43643", + "left": "issue:43673", + "right": "issue:45278", "accept": false, - "reason": "A missing vision output attribute in SAM3 and missing config fields with trust_remote_code=True are different object shapes/code paths." + "reason": "A specific generation cache bug is not the same concrete bug as a broad report about many import errors after upgrading." }, { - "left": "issue:45375", - "right": "issue:45478", + "left": "issue:44297", + "right": "issue:44623", "accept": false, - "reason": "Both are Qwen3.5-Moe related, but one is a missing config field silently dropped while the other is a from_pretrained loading error; related family, not the same bug." + "reason": "Tokenizer metadata saving and processor file saving are different save_pretrained failure modes." }, { - "left": "issue:45230", - "right": "issue:45397", + "left": "issue:43646", + "right": "issue:44162", "accept": false, - "reason": "The first report is too vague to establish identity, and the second is a Gemma-4 ZeRO-3 from_pretrained issue; no clear duplicate evidence." + "reason": "trust_remote_code field loss and ESM2 being broken are unrelated bugs." }, { - "left": "issue:43475", - "right": "issue:45381", + "left": "issue:44361", + "right": "issue:44749", "accept": false, - "reason": "SAM3 vision output attribute loss and Qwen2.5-VL video position-id errors are different model-specific bugs." + "reason": "Tokenizer AttributeError on tasks and a performance regression after upgrading are not the same underlying issue." }, { - "left": "issue:41628", - "right": "issue:44933", + "left": "issue:44623", + "right": "issue:45656", "accept": false, - "reason": "Both are import/export problems, but they concern different symbols and APIs; not the same missing-export bug." + "reason": "Processor save_pretrained file loss and deepspeed optimizer double-step are unrelated." }, { - "left": "issue:45310", - "right": "issue:45375", + "left": "issue:28282", + "right": "issue:30990", "accept": false, - "reason": "One is a from_pretrained failure and the other is a missing config field in a vision config; same model family but not the same underlying bug." + "reason": "Missing PyTorch ImportError and Sentence Transformers getting stuck loading are different problems." }, { - "left": "issue:45081", - "right": "issue:45103", + "left": "issue:44749", + "right": "issue:45310", "accept": false, - "reason": "A tokenizer regex patch crash and an auto-docstring annotation crash are unrelated implementation bugs." + "reason": "A data filtering slowdown and a Qwen3.5 MoE from_pretrained failure are not the same bug." }, { "left": "issue:42994", - "right": "issue:43525", + "right": "issue:43482", "accept": false, - "reason": "Quantized model save failure and missing pad_token_id on Llama4Config are separate config/save-path issues." + "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are separate serialization/loading issues." }, { - "left": "issue:43475", - "right": "issue:44661", + "left": "issue:42371", + "right": "issue:45092", "accept": false, - "reason": "SAM3 output attribute handling and add-new-model-like tokenizer mapping failures do not share the same code path." + "reason": "TF32 API guidance and InternVL2 meta-initialization incompatibility are unrelated." }, { - "left": "issue:45362", - "right": "issue:45375", + "left": "issue:44038", + "right": "issue:44623", "accept": false, - "reason": "A chat crash for Qwen3.5-35B is different from a config field being dropped in Qwen3_5MoeVisionConfig." + "reason": "Qwen3-VL-Moe/transformers-v5 incompatibility and processor.save_pretrained missing files are different failures." }, { - "left": "issue:43873", - "right": "issue:44945", + "left": "issue:44749", + "right": "issue:45656", "accept": false, - "reason": "Quantization offloading and pipeline-parallel LLM output correctness are different runtime problems." + "reason": "Upgrade-related slowdown and deepspeed optimizer step duplication are different failure modes." }, { - "left": "issue:42831", - "right": "issue:44222", - "accept": false, - "reason": "FineGrainedFP8 accuracy drift and FP8 save_pretrained for MoE are distinct FP8-related issues, not one bug." + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same symptom: Qwen3ForCausalLM VRAM leak in multiple dataloader threads, so these are the same underlying bug." }, { - "left": "issue:43232", - "right": "issue:44188", + "left": "issue:43475", + "right": "issue:43643", "accept": false, - "reason": "Generation kwargs sync issues and attention-kernel divergence under torch.compile are unrelated generation internals." + "reason": "A missing attribute on Sam3VisionEncoderOutput is not the same as missing fields in AutoConfig under trust_remote_code." }, { - "left": "issue:42222", - "right": "issue:45200", + "left": "issue:45230", + "right": "issue:45397", "accept": false, - "reason": "Broken VitPose model loading and Gemma 4 mm_token_type_ids defaults are different multimodal-model bugs." + "reason": "Generic bug report vs. a specific Gemma-4 ZeRO3 from_pretrained issue; no shared concrete code path." }, { - "left": "issue:41628", - "right": "issue:44351", + "left": "issue:45375", + "right": "issue:45478", "accept": false, - "reason": "Both are import failures, but they involve different missing names (AutoImageProcessor vs HybridCache) and different fixes." + "reason": "Both involve Qwen3.5 MoE/config loading, but one is a specific missing-field report and the other is a broader from_pretrained error; not enough evidence they are the same bug." }, { - "left": "issue:44222", - "right": "issue:45310", + "left": "issue:43475", + "right": "issue:45381", "accept": false, - "reason": "Saving FP8 MoE weights and a Qwen3.5-Moe from_pretrained error are different failure modes." + "reason": "SAM 3 video missing attribute and qwen2.5-vl video position-id regression are different model bugs." }, { - "left": "issue:42994", - "right": "issue:44164", + "left": "issue:41628", + "right": "issue:44933", "accept": false, - "reason": "Quantized save failure and extra_state handling in save/from_pretrained are both persistence-related but not the same underlying defect." + "reason": "Two import errors, but from different modules and different root causes; too broad to merge." }, { - "left": "issue:44038", - "right": "issue:45310", + "left": "issue:45310", + "right": "issue:45375", "accept": false, - "reason": "Qwen3-VL-Moe loading and Qwen3.5-Moe loading are different model families and different failure reports." + "reason": "The former is a model-loading error report; the latter is a specific config-field omission. They may be related, but not clearly the same concrete bug." }, { - "left": "issue:43526", - "right": "issue:45561", + "left": "issue:45081", + "right": "issue:45103", "accept": false, - "reason": "A BeitImageProcessorFast label-reduction bug is unrelated to a pytest-xdist race in testing utilities." + "reason": "Tokenizer regex patch crash and auto-docstring annotation handling crash are unrelated." }, { - "left": "issue:36010", - "right": "issue:41628", + "left": "issue:42994", + "right": "issue:43525", "accept": false, - "reason": "Both are import errors, but for different public symbols and likely different resolution paths." + "reason": "Quantized save_pretrained failure and Llama4Config missing pad_token_id are distinct issues." }, { - "left": "issue:44496", - "right": "issue:45561", + "left": "issue:43475", + "right": "issue:44661", "accept": false, - "reason": "An unrecognized model/config loading issue is unrelated to a captured_info.txt race in test workers." + "reason": "SAM 3 output attribute failure and add-new-model-like tokenizer mapping failure are different code paths." }, { - "left": "issue:43673", - "right": "issue:44188", + "left": "issue:45362", + "right": "issue:45375", "accept": false, - "reason": "Missing GenerationMixin cache during chunked prefill and attention-kernel branching divergence are different generation bugs." + "reason": "Qwen3.5-35B chat crash and a Qwen3_5MoeVisionConfig field being dropped are not the same bug." }, { - "left": "issue:43504", - "right": "issue:45561", + "left": "issue:42831", + "right": "issue:44222", "accept": false, - "reason": "A Beit pretrained-model preset loading bug and a testing utility race are unrelated." + "reason": "FineGrainedFP8 accuracy issues and FP8 save_pretrained issues are related only at a high level, not the same bug." }, { - "left": "issue:43377", - "right": "issue:44242", + "left": "issue:43873", + "right": "issue:44945", "accept": false, - "reason": "MIMI encoder batch-vs-single output mismatch from padding-mask handling is unrelated to MoE router loss logic." + "reason": "Quantization offloading and incorrect output under pipeline parallelism are different problems." }, { "left": "issue:43232", - "right": "issue:44792", - "accept": false, - "reason": "Generation kwargs synchronization and a janus image-generation test failure are not the same issue." - }, - { - "left": "issue:42898", - "right": "issue:44568", - "accept": false, - "reason": "Tokenizer clean-up-space behavior and add_special_tokens BOS/EOS behavior are separate tokenizer regressions." - }, - { - "left": "issue:30990", - "right": "issue:35141", + "right": "issue:44188", "accept": false, - "reason": "A Sentence Transformers loading hang and token-embedding resizing reinitialization are unrelated model-init issues." + "reason": "Generation kwargs sync bug and diverging attention kernels under torch.compile are unrelated." }, { - "left": "issue:43479", - "right": "issue:44079", + "left": "issue:42222", + "right": "issue:45200", "accept": false, - "reason": "Default multimodal config initialization and ModelOutput key assignment with None are different object-construction bugs." + "reason": "VitPose model breakage and Gemma 4 mm_token_type_ids defaults are different model-specific issues." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The summary is not fully grounded in the packet: it says the only clear duplicate pair is the Qwen3ForCausalLM VRAM-leak report, but the packet also contains two issues with the exact same title, `issue:45310` and `issue:45478` (`[BUG] transformers>=5.4.0, Qwen3.5 Moe from_pretrained error`). That omission makes the duplicate/umbrella assessment too strong. Several soft-edge false verdicts are conservative, but the high-level summary overstates the uniqueness of the VRAM-leak pair." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -10910,47 +10734,48 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ - "issue:22355", - "issue:28282", "issue:30333", + "issue:30990", "issue:33290", - "issue:36331", + "issue:35141", + "issue:36010", "issue:41628", "issue:42175", - "issue:42757", - "issue:42831", "issue:42898", + "issue:42994", + "issue:43232", "issue:43278", - "issue:43295", + "issue:43377", "issue:43479", - "issue:43644", + "issue:43504", + "issue:43526", + "issue:43673", "issue:43723", "issue:43784", "issue:43824", "issue:43874", "issue:43976", + "issue:44038", "issue:44060", + "issue:44079", + "issue:44164", "issue:44188", - "issue:44206", "issue:44222", - "issue:44263", - "issue:44295", - "issue:44297", - "issue:44360", + "issue:44242", + "issue:44351", "issue:44373", - "issue:44442", "issue:44448", - "issue:44484", + "issue:44496", "issue:44568", "issue:44623", "issue:44734", "issue:44779", + "issue:44792", "issue:44869", "issue:44933", "issue:44964", - "issue:44977", "issue:45092", "issue:45103", "issue:45125", @@ -10958,9 +10783,23 @@ "issue:45310", "issue:45356", "issue:45459", - "issue:45656" + "issue:45561" ], "soft_pairs": [ + "issue:41628|issue:44351", + "issue:44222|issue:45310", + "issue:42994|issue:44164", + "issue:44038|issue:45310", + "issue:43526|issue:45561", + "issue:43673|issue:44188", + "issue:44496|issue:45561", + "issue:36010|issue:41628", + "issue:43504|issue:45561", + "issue:43377|issue:44242", + "issue:43232|issue:44792", + "issue:42898|issue:44568", + "issue:30990|issue:35141", + "issue:43479|issue:44079", "issue:43479|issue:44869", "issue:43976|issue:44933", "issue:42175|issue:44933", @@ -10969,259 +10808,212 @@ "issue:44623|issue:45310", "issue:43784|issue:43824", "issue:43723|issue:44568", + "issue:44060|issue:45125", "issue:43278|issue:45137", "issue:42898|issue:45356", - "issue:44060|issue:45125", "issue:44734|issue:45103", - "issue:42898|issue:44779", "issue:43874|issue:45092", + "issue:42898|issue:44779", "issue:30333|issue:33290", "issue:42898|issue:44448", - "issue:41628|issue:45459", - "issue:44295|issue:45656", - "issue:42898|issue:43295", - "issue:44222|issue:44297", - "issue:44297|issue:45656", - "issue:43479|issue:43644", - "issue:44442|issue:45459", - "issue:42831|issue:44623", - "issue:44222|issue:44623", - "issue:44206|issue:45103", - "issue:22355|issue:28282", - "issue:41628|issue:42757", - "issue:44263|issue:44360", - "issue:44297|issue:45310", - "issue:30333|issue:36331", - "issue:44222|issue:44484", - "issue:43723|issue:44977", - "issue:44263|issue:45092" + "issue:41628|issue:45459" ], - "prepared_review_unit_hash": "456bf06986ef0157c2922ccc662d6f2fb39e4a2c28faf97514a1acba23b93f97", + "prepared_review_unit_hash": "78fc8464d3e7c20c947f8558813743b43c6421db49c7e89129c3d34454a4986b", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 34, - "serialized_chars": 37248, - "estimated_input_tokens": 9312, - "estimated_eval_tokens": 18880 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36970, + "estimated_input_tokens": 9243, + "estimated_eval_tokens": 18742 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft pairs mostly connect unrelated import errors, tokenizer regressions, multimodal/model-loading bugs, and DeepSpeed/training issues. I found no true duplicate pairs to accept, and there is no single canonical issue/PR for the cluster.", - "confidence": 0.93, - "canonical_issue_reason": "No clear canonical issue: the items do not share one underlying bug or code path, so a single anchor would be misleading.", + "summary": "These are mostly separate Transformers regression reports spanning import/export issues, tokenizer behavior changes, model-specific load/save bugs, and generation/runtime edge cases. The soft pairs share broad themes, but none look like the same concrete underlying defect.", + "confidence": 0.86, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single best issue for duplicate triage; the set is too mixed to represent one problem. If forced, an open model-loading regression like issue:44964 is more actionable, but it is not a cluster center.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43479", - "right": "issue:44869", - "accept": false, - "reason": "Different bugs: Phi-4 multimodal config init vs Whisper trailing-replacement decode crash." - }, - { - "left": "issue:43976", - "right": "issue:44933", - "accept": false, - "reason": "Unrelated failures: Python version/support issue vs a missing image_utils import." - }, - { - "left": "issue:42175", - "right": "issue:44933", - "accept": false, - "reason": "Different subsystems and symptoms: backend extras install vs image_utils import regression." - }, - { - "left": "issue:44964", - "right": "issue:45137", - "accept": false, - "reason": "Model loading incompatibility vs DeepSpeed ZeRO3 deque error; no shared code path." - }, - { - "left": "issue:44188", - "right": "issue:44373", + "left": "issue:41628", + "right": "issue:44351", "accept": false, - "reason": "Attention-kernel divergence vs a docstring mismatch; not the same bug." + "reason": "Both are import failures, but they concern different missing symbols (`AutoImageProcessor` vs `HybridCache`) and likely different export paths." }, { - "left": "issue:44623", + "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Processor save_pretrained file omission vs Qwen3.5 MoE from_pretrained error; different serialization/loading issues." + "reason": "One is FP8 `save_pretrained` for MoE models, the other is a `from_pretrained` load error for Qwen3.5 MoE; different phases and failure modes." }, { - "left": "issue:43784", - "right": "issue:43824", + "left": "issue:42994", + "right": "issue:44164", "accept": false, - "reason": "Different import problems with different packages/models and root causes." + "reason": "Both involve saving/loading, but one is quantized model saving and the other is `extra_state` handling in save/from_pretrained plumbing." }, { - "left": "issue:43723", - "right": "issue:44568", + "left": "issue:44038", + "right": "issue:45310", "accept": false, - "reason": "Both are tokenizer regressions, but on different models and behaviors; not one concrete defect." + "reason": "Both mention Qwen/MoE, but the model families and concrete breakage differ (`Qwen3-VL-Moe` vs `Qwen3.5 Moe` from_pretrained error)." }, { - "left": "issue:43278", - "right": "issue:45137", + "left": "issue:43526", + "right": "issue:45561", "accept": false, - "reason": "BF16/FP32 evaluation dtype drift is unrelated to a ZeRO3 deque underflow." + "reason": "These are unrelated: one is a BEiT label reduction bug, the other is a pytest-xdist race in captured test output." }, { - "left": "issue:42898", - "right": "issue:45356", + "left": "issue:43673", + "right": "issue:44188", "accept": false, - "reason": "Both mention tokenizer regressions, but the model-specific codec regression is distinct from clean_up_tokenization_spaces changes." + "reason": "Both touch generation/runtime behavior, but one is missing cache support in chunked prefill while the other is attention-kernel divergence under `torch.compile`." }, { - "left": "issue:44060", - "right": "issue:45125", + "left": "issue:44496", + "right": "issue:45561", "accept": false, - "reason": "Different Qwen family issues: tied-weights warning vs missing _tp_plan for tensor parallelism." + "reason": "An unrecognized model config error is unrelated to a test-worker race on `captured_info.txt`." }, { - "left": "issue:44734", - "right": "issue:45103", + "left": "issue:36010", + "right": "issue:41628", "accept": false, - "reason": "Serving KV-cache tensor indexing crash is unrelated to auto-docstring annotation parsing." + "reason": "Both are import errors, but they are for different APIs and modules, so they are not the same bug." }, { - "left": "issue:42898", - "right": "issue:44779", + "left": "issue:43504", + "right": "issue:45561", "accept": false, - "reason": "Different tokenizer regressions affecting different models and code paths." + "reason": "A legacy-field load failure in BEiT segmentation is unrelated to the xdist file race." }, { - "left": "issue:43874", - "right": "issue:45092", + "left": "issue:43377", + "right": "issue:44242", "accept": false, - "reason": "Both are multimodal-related, but one is a missing image-patch method and the other is meta-init compatibility for old remote-code checkpoints." + "reason": "Different model behaviors: missing padding-mask support in MIMI encoder vs gating/load-balancing loss omission when router logits are off." }, { - "left": "issue:30333", - "right": "issue:33290", + "left": "issue:43232", + "right": "issue:44792", "accept": false, - "reason": "MLflow reporting on failed jobs is unrelated to an Adafactor OOM in DeepSpeed." + "reason": "One is a generation kwargs update bug after `sync_gpus`; the other is a Janus image-generation test failure." }, { "left": "issue:42898", - "right": "issue:44448", - "accept": false, - "reason": "Tokenizer cleanup behavior vs model output divergence between v4 and v5 are not the same bug." - }, - { - "left": "issue:41628", - "right": "issue:45459", + "right": "issue:44568", "accept": false, - "reason": "Different import errors from different dependencies and failure modes." + "reason": "Both are tokenizer regressions in v5, but one is `clean_up_tokenization_spaces` behavior and the other is `add_special_tokens` not inserting BOS/EOS." }, { - "left": "issue:44295", - "right": "issue:45656", + "left": "issue:30990", + "right": "issue:35141", "accept": false, - "reason": "Position_ids buffer access error is unrelated to DeepSpeed double optimizer stepping." + "reason": "A Sentence Transformers loading hang and a token-embedding reinitialization bug are unrelated code paths." }, { - "left": "issue:42898", - "right": "issue:43295", + "left": "issue:43479", + "right": "issue:44079", "accept": false, - "reason": "Both involve processors/tokenizers, but the breakages are different API regressions." + "reason": "Both mention `None` handling, but one is multimodal config initialization and the other is `ModelOutput` key assignment logic." }, { - "left": "issue:44222", - "right": "issue:44297", + "left": "issue:43479", + "right": "issue:44869", "accept": false, - "reason": "FP8 MoE save_pretrained and tokenizer_config save_pretrained are unrelated serialization issues." + "reason": "Different subsystems: multimodal config defaults vs Whisper word-timestamp decoding crash." }, { - "left": "issue:44297", - "right": "issue:45656", + "left": "issue:43976", + "right": "issue:44933", "accept": false, - "reason": "Tokenizer metadata mismatch is unrelated to optimizer stepping twice." + "reason": "A Python-version compatibility issue is unrelated to a missing `image_utils` import." }, { - "left": "issue:43479", - "right": "issue:43644", + "left": "issue:42175", + "right": "issue:44933", "accept": false, - "reason": "Config-default initialization bug vs non-persistent buffer junk values; different code paths." + "reason": "Package extras/backend composition and a missing import are distinct problems." }, { - "left": "issue:44442", - "right": "issue:45459", + "left": "issue:44964", + "right": "issue:45137", "accept": false, - "reason": "Tokenizer class loading failure is not the same as protobuf-masked tokenizer decode errors." + "reason": "Loading a specific Phi-4 multimodal checkpoint and a DeepSpeed ZeRO3 deque underflow are unrelated." }, { - "left": "issue:42831", - "right": "issue:44623", + "left": "issue:44188", + "right": "issue:44373", "accept": false, - "reason": "FP8 accuracy issue is unrelated to missing files in processor.save_pretrained." + "reason": "A runtime attention-kernel branching bug is not the same as a docstring mismatch." }, { - "left": "issue:44222", - "right": "issue:44623", + "left": "issue:44623", + "right": "issue:45310", "accept": false, - "reason": "Both mention saving, but one is MoE FP8 model persistence and the other is processor file output." + "reason": "Processor `save_pretrained` missing files and Qwen3.5 MoE `from_pretrained` failure are different serialization directions and targets." }, { - "left": "issue:44206", - "right": "issue:45103", + "left": "issue:43784", + "right": "issue:43824", "accept": false, - "reason": "Feature-extractor crash from unsupported argument is unrelated to auto-docstring annotation handling." + "reason": "Both are import errors, but one is `nn` missing during sentence-transformers import and the other is a missing Qwen2.5-VL class export." }, { - "left": "issue:22355", - "right": "issue:28282", + "left": "issue:43723", + "right": "issue:44568", "accept": false, - "reason": "Different import/dependency problems: missing transformers.onnx module vs missing PyTorch backend." + "reason": "Both involve tokenizers, but one is tokenizer loading via `AutoTokenizer.from_pretrained` and the other is special-token insertion semantics." }, { - "left": "issue:41628", - "right": "issue:42757", + "left": "issue:44060", + "right": "issue:45125", "accept": false, - "reason": "Different missing symbols from different libraries; not duplicates." + "reason": "Both are Qwen3-era model issues, but one is an incorrect tied-weights warning and the other is a missing `_tp_plan` for tensor parallelism." }, { - "left": "issue:44263", - "right": "issue:44360", + "left": "issue:43278", + "right": "issue:45137", "accept": false, - "reason": "Same general component name, but one is a split/unpacking issue and the other is a missing ReLU; not the same defect." + "reason": "An embedding dtype drift between train/eval is unrelated to a ZeRO3 deque pop error." }, { - "left": "issue:44297", - "right": "issue:45310", + "left": "issue:42898", + "right": "issue:45356", "accept": false, - "reason": "Tokenizer save metadata mismatch and Qwen3.5 MoE from_pretrained failure are different issues." + "reason": "Different tokenizer regressions: cleanup-space behavior versus Kimi codec/regex handling." }, { - "left": "issue:44222", - "right": "issue:44484", + "left": "issue:30333", + "right": "issue:33290", "accept": false, - "reason": "FP8 MoE save_pretrained bug is unrelated to the 50GB shard-size question." + "reason": "MLFlow reporting on failed training jobs and a deepspeed OOM are unrelated." }, { - "left": "issue:43723", - "right": "issue:44977", + "left": "issue:42898", + "right": "issue:44448", "accept": false, - "reason": "Tokenizer loading regression vs flash-attention generation failure are different code paths." + "reason": "One is a tokenizer cleanup option change, the other is different model outputs for Pegasus between v4 and v5." }, { - "left": "issue:44263", - "right": "issue:45092", + "left": "issue:41628", + "right": "issue:45459", "accept": false, - "reason": "DSA indexer split behavior is unrelated to InternVL2 remote-code meta-init incompatibility." + "reason": "A missing exported class and tokenizer error masking via protobuf exception handling are different import/error-handling bugs." } ] }, @@ -11243,48 +11035,47 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ + "issue:22355", "issue:28282", "issue:30333", "issue:30990", - "issue:33357", "issue:34689", "issue:35141", + "issue:36331", "issue:38175", "issue:41628", - "issue:42371", "issue:42757", "issue:42831", "issue:42886", - "issue:42907", + "issue:42898", "issue:42994", - "issue:43066", - "issue:43122", - "issue:43232", + "issue:43295", "issue:43479", - "issue:43576", + "issue:43644", "issue:43723", "issue:43756", - "issue:43784", "issue:43824", "issue:44038", "issue:44060", - "issue:44117", + "issue:44206", + "issue:44222", "issue:44261", "issue:44263", "issue:44295", + "issue:44297", + "issue:44360", "issue:44373", "issue:44442", "issue:44483", + "issue:44484", "issue:44485", "issue:44514", - "issue:44521", "issue:44623", - "issue:44704", "issue:44871", "issue:44913", - "issue:44945", + "issue:44977", "issue:44991", "issue:45092", "issue:45103", @@ -11292,9 +11083,27 @@ "issue:45310", "issue:45357", "issue:45406", - "issue:45593" + "issue:45459", + "issue:45656" ], "soft_pairs": [ + "issue:44295|issue:45656", + "issue:42898|issue:43295", + "issue:44222|issue:44297", + "issue:44297|issue:45656", + "issue:43479|issue:43644", + "issue:44442|issue:45459", + "issue:42831|issue:44623", + "issue:44222|issue:44623", + "issue:44206|issue:45103", + "issue:41628|issue:42757", + "issue:22355|issue:28282", + "issue:44263|issue:44360", + "issue:44297|issue:45310", + "issue:44222|issue:44484", + "issue:30333|issue:36331", + "issue:44263|issue:45092", + "issue:43723|issue:44977", "issue:44483|issue:45406", "issue:30333|issue:34689", "issue:28282|issue:30333", @@ -11302,231 +11111,271 @@ "issue:44295|issue:44623", "issue:44373|issue:44485", "issue:44261|issue:45092", - "issue:44514|issue:45103", "issue:43723|issue:44871", + "issue:44514|issue:45103", "issue:43723|issue:44991", - "issue:43479|issue:44913", "issue:45092|issue:45357", "issue:43723|issue:44442", + "issue:43479|issue:44913", "issue:30990|issue:38175", "issue:42886|issue:42994", "issue:41628|issue:43824", "issue:42831|issue:45310", "issue:42831|issue:44038", - "issue:42757|issue:45278", "issue:44263|issue:44373", - "issue:43756|issue:44060", - "issue:43232|issue:45593", - "issue:43576|issue:43784", - "issue:43122|issue:44521", - "issue:42886|issue:43066", - "issue:42371|issue:44263", - "issue:44117|issue:44704", - "issue:30990|issue:33357", - "issue:42907|issue:44945" + "issue:42757|issue:45278", + "issue:43756|issue:44060" ], - "prepared_review_unit_hash": "d70436ebcc528f04e3ab7ea0afa69e365c39d22ac42cc717972ffe6957fe6f33", + "prepared_review_unit_hash": "42d2da78a2984a97224d7c2df82c299fe0dfebdea278f4dcf5524e6672417374", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36113, - "estimated_input_tokens": 9029, - "estimated_eval_tokens": 18314 + "soft_pair_count": 38, + "serialized_chars": 38851, + "estimated_input_tokens": 9713, + "estimated_eval_tokens": 19682 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of unrelated Transformers regressions. The soft matches are largely superficial similarities around loading/import/config codepaths, but none look like the same underlying bug or change. No pull requests are present.", - "confidence": 0.91, - "canonical_issue_reason": "issue:43723 is the broadest, most user-facing loading regression in the set (AutoTokenizer.from_pretrained in v5), but the items do not form a true duplicate family.", + "summary": "This cluster is highly heterogeneous and mostly contains separate bug reports around imports, tokenizers, config serialization, and model-specific regressions. No soft-edge pair looks like the same underlying bug/change, so all candidate links should be rejected.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: it describes widespread import errors after a Transformers version upgrade, which is the closest fit to the cluster\u2019s repeated import/load-regression theme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43723 is the best representative only in a loose sense because several other items are nearby tokenizer/import failures; however, the cluster is too heterogeneous for a strong canonical.", + "best_issue_reason": "issue:45278 is the best representative issue because it is the most general, user-facing, and cluster-like report; it can plausibly anchor the broader import/error regressions in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44483", - "right": "issue:45406", + "left": "issue:44295", + "right": "issue:45656", "accept": false, - "reason": "Different surfaces: chat-completions API rejection vs Gemma4Processor missing _tokenizer." + "reason": "Different bugs: position_ids buffer access vs optimizer step being called twice under DeepSpeed." }, { - "left": "issue:30333", - "right": "issue:34689", + "left": "issue:42898", + "right": "issue:43295", "accept": false, - "reason": "MLflow job reporting vs Llama 3.2 vision model-loading failure; unrelated bugs." + "reason": "Tokenizer cleanup behavior change is unrelated to processor.tokenizer/image-passing regression." }, { - "left": "issue:28282", - "right": "issue:30333", + "left": "issue:44222", + "right": "issue:44297", "accept": false, - "reason": "PyTorch import dependency error vs MLflow reporting on failed training jobs." + "reason": "Both involve save_pretrained, but one is FP8 MoE saving and the other is tokenizer metadata mismatch." }, { - "left": "issue:30333", - "right": "issue:35141", + "left": "issue:44297", + "right": "issue:45656", "accept": false, - "reason": "Job-status reporting bug vs embedding resize/post_init weight reinitialization." + "reason": "Tokenizer save metadata bug is unrelated to DeepSpeed optimizer-step duplication." }, { - "left": "issue:44295", - "right": "issue:44623", + "left": "issue:43479", + "right": "issue:43644", "accept": false, - "reason": "position_ids buffer access issue vs processor.save_pretrained missing files; different codepaths." + "reason": "Multimodal config defaults and non-persistent buffer junk are different initialization/storage problems." }, { - "left": "issue:44373", - "right": "issue:44485", + "left": "issue:44442", + "right": "issue:45459", "accept": false, - "reason": "Docstring mistake vs GLM-5 RoPE implementation bug; not the same issue." + "reason": "AutoTokenizer failing on a specific tokenizer is different from protobuf exception masking during tokenizer decode." }, { - "left": "issue:44261", - "right": "issue:45092", + "left": "issue:42831", + "right": "issue:44623", "accept": false, - "reason": "Precision/config note for MLA q_a_layernorm vs meta-init incompatibility for old InternVL2 checkpoints." + "reason": "FineGrainedFP8 accuracy and processor.save_pretrained missing files are unrelated." }, { - "left": "issue:44514", + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both touch saving, but FP8 MoE save failure is not the same as missing processor files." + }, + { + "left": "issue:44206", "right": "issue:45103", "accept": false, - "reason": "Multimodal chat template crash vs auto_docstring AttributeError from future annotations." + "reason": "Unsupported feature argument in LasrFeatureExtractor is unrelated to auto-docstring annotation parsing." }, { - "left": "issue:43723", - "right": "issue:44871", + "left": "issue:41628", + "right": "issue:42757", "accept": false, - "reason": "Generic tokenizer loading regression vs Gemma-3 eos_token_id config mismatch." + "reason": "Both are import errors, but they are for different symbols and different packages." }, { - "left": "issue:43723", - "right": "issue:44991", + "left": "issue:22355", + "right": "issue:28282", "accept": false, - "reason": "Both are tokenizer-loading failures, but for different models and failure modes; no shared root cause shown." + "reason": "Missing transformers.onnx module and missing PyTorch are separate dependency/import failures." }, { - "left": "issue:43479", - "right": "issue:44913", + "left": "issue:44263", + "right": "issue:44360", "accept": false, - "reason": "Config defaulting bug in Phi4Multimodal vs rotary_pct not persisting in GPTNeoXConfig." + "reason": "Same broad DSA area, but torch.split return handling is not the same as a missing ReLU in the indexer." }, { - "left": "issue:45092", - "right": "issue:45357", + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer class metadata mismatch is unrelated to Qwen3.5 from_pretrained loading error." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save failure is not the same issue as max_shard_size default behavior." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting on failed jobs is unrelated to the trainer compute_loss signature change." + }, + { + "left": "issue:44263", + "right": "issue:45092", "accept": false, - "reason": "Remote-code/meta-init incompatibility vs incorrect save_pretrained visual encoder keys." + "reason": "DSA indexer split semantics and remote-code/meta-init incompatibility are different code paths." }, { "left": "issue:43723", - "right": "issue:44442", + "right": "issue:44977", "accept": false, - "reason": "Different tokenizer loading failures for different tokenizer classes; not enough to unify." + "reason": "Tokenizer loading regression and flash-attention generation failure are different symptoms and fixes." }, { - "left": "issue:30990", - "right": "issue:38175", + "left": "issue:44483", + "right": "issue:45406", "accept": false, - "reason": "Loading hangs in Sentence Transformers vs unexpected zero probabilities in SigLIP2 inference." + "reason": "Chat-completions request rejection is unrelated to Gemma4Processor missing _tokenizer." }, { - "left": "issue:42886", - "right": "issue:42994", + "left": "issue:30333", + "right": "issue:34689", "accept": false, - "reason": "Offline cache tokenization failure vs quantized model saving failure." + "reason": "MLflow job reporting and Llama model loading regression are unrelated." }, { - "left": "issue:41628", - "right": "issue:43824", + "left": "issue:28282", + "right": "issue:30333", "accept": false, - "reason": "Missing AutoImageProcessor export vs missing Qwen2_5_VLForConditionalGeneration export; separate import bugs." + "reason": "Missing PyTorch import error is unrelated to MLflow not reporting failed training jobs." }, { - "left": "issue:42831", - "right": "issue:45310", + "left": "issue:30333", + "right": "issue:35141", "accept": false, - "reason": "FineGrainedFP8 accuracy issue vs Qwen3.5 Moe from_pretrained error." + "reason": "Training-job reporting and embedding reinitialization after resize are unrelated bugs." }, { - "left": "issue:42831", - "right": "issue:44038", + "left": "issue:44295", + "right": "issue:44623", "accept": false, - "reason": "FP8 accuracy regression vs Qwen3-VL-Moe bug; different problems." + "reason": "Position_ids buffer read error is unrelated to processor save_pretrained file omission." }, { - "left": "issue:42757", - "right": "issue:45278", + "left": "issue:44373", + "right": "issue:44485", "accept": false, - "reason": "Specific huggingface_hub import error vs broad Transformers v5 import failures; too vague to merge." + "reason": "Wrong docstring for position_ids is documentation-only and not the same as GLM-5 RoPE behavior." }, { - "left": "issue:44263", - "right": "issue:44373", + "left": "issue:44261", + "right": "issue:45092", "accept": false, - "reason": "torch.split behavior bug vs docstring correction for position_ids." + "reason": "MLA layernorm epsilon precision issue is unrelated to InternVL2 meta-initialization incompatibility." }, { - "left": "issue:43756", - "right": "issue:44060", + "left": "issue:43723", + "right": "issue:44871", "accept": false, - "reason": "Smollm3 RoPE layer dropping vs Qwen3-Next tied-weights warning; unrelated model issues." + "reason": "Tokenizer loading regression and eos_token_id configuration mismatch are different model/tokenizer issues." }, { - "left": "issue:43232", - "right": "issue:45593", + "left": "issue:44514", + "right": "issue:45103", "accept": false, - "reason": "Generation kwargs sync bug vs D-FINE auxiliary-loss issue." + "reason": "Batched chat-template crash is unrelated to annotation-handling in auto-docstring." }, { - "left": "issue:43576", - "right": "issue:43784", + "left": "issue:43723", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-load failures, but they concern different models and likely different failure modes." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Remote-code meta-init incompatibility is different from incorrect visual-encoder keys on save_pretrained." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Config defaults for multimodal models and GPTNeoX rotary_pct reload behavior are separate config bugs." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "SentenceTransformers loading stall and SigLIP2 zero probabilities are unrelated model-specific problems." + }, + { + "left": "issue:42886", + "right": "issue:42994", "accept": false, - "reason": "transformers env CLI regression vs sentence-transformers import NameError." + "reason": "Offline tokenizer cache loading and quantized model saving are different subsystems." }, { - "left": "issue:43122", - "right": "issue:44521", + "left": "issue:41628", + "right": "issue:43824", "accept": false, - "reason": "Tokenization changes across versions vs multimodal assistant_masks all-zero bug." + "reason": "Both are import errors, but they concern different classes and likely different missing exports." }, { - "left": "issue:42886", - "right": "issue:43066", + "left": "issue:42831", + "right": "issue:45310", "accept": false, - "reason": "Offline tokenizer cache failure vs wrong tokenizer decoder type in v5." + "reason": "FineGrainedFP8 accuracy issue is unrelated to Qwen3.5 MoE from_pretrained errors." }, { - "left": "issue:42371", - "right": "issue:44263", + "left": "issue:42831", + "right": "issue:44038", "accept": false, - "reason": "TF32 API migration warning vs GlmMoeDsaIndexer torch.split issue." + "reason": "FP8 accuracy problems and Qwen3-VL-Moe loading bugs are not the same underlying issue." }, { - "left": "issue:44117", - "right": "issue:44704", + "left": "issue:44263", + "right": "issue:44373", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES/from_pretrained assumption bug vs AutoProcessor cached_file kwargs propagation bug." + "reason": "A torch.split return-value bug is unrelated to a wrong docstring for position_ids." }, { - "left": "issue:30990", - "right": "issue:33357", + "left": "issue:42757", + "right": "issue:45278", "accept": false, - "reason": "Loading hang in Sentence Transformers vs MacOS bus error on CLIP model." + "reason": "Missing is_offline_mode from huggingface_hub is a specific dependency mismatch, not the same as broad import errors after upgrading Transformers." }, { - "left": "issue:42907", - "right": "issue:44945", + "left": "issue:43756", + "right": "issue:44060", "accept": false, - "reason": "Quantized model save failure vs incorrect output under pipeline parallelism." + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are distinct architecture/configuration bugs." } ] }, @@ -11548,35 +11397,36 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:29127", "issue:30990", + "issue:33357", "issue:36296", "issue:39401", "issue:41628", "issue:42371", - "issue:42757", "issue:42831", + "issue:42886", "issue:42907", + "issue:43066", "issue:43116", "issue:43122", "issue:43232", - "issue:43278", "issue:43408", "issue:43452", "issue:43475", + "issue:43576", "issue:43784", - "issue:43824", "issue:43937", "issue:43976", - "issue:43992", "issue:44038", + "issue:44117", "issue:44222", "issue:44261", + "issue:44263", "issue:44295", "issue:44297", - "issue:44351", "issue:44360", "issue:44373", "issue:44485", @@ -11590,21 +11440,27 @@ "issue:44829", "issue:44945", "issue:44964", - "issue:44977", "issue:45200", - "issue:45216", "issue:45245", "issue:45278", "issue:45290", - "issue:45478", + "issue:45593", "issue:45600" ], "soft_pairs": [ + "issue:43232|issue:45593", + "issue:43576|issue:43784", + "issue:43122|issue:44521", + "issue:42886|issue:43066", + "issue:42371|issue:44263", + "issue:44117|issue:44704", + "issue:30990|issue:33357", "issue:43976|issue:44945", + "issue:42907|issue:44945", "issue:44704|issue:45290", "issue:43122|issue:44514", - "issue:43232|issue:44743", "issue:44295|issue:44297", + "issue:43232|issue:44743", "issue:39401|issue:44488", "issue:43408|issue:43475", "issue:43232|issue:45245", @@ -11613,252 +11469,245 @@ "issue:43784|issue:45600", "issue:44360|issue:44485", "issue:39401|issue:44521", - "issue:44625|issue:44829", "issue:44261|issue:44373", - "issue:29127|issue:30990", - "issue:43122|issue:44779", + "issue:44625|issue:44829", "issue:43116|issue:45200", + "issue:43122|issue:44779", + "issue:29127|issue:30990", "issue:42371|issue:42831", "issue:44038|issue:44297", "issue:42907|issue:44964", "issue:44038|issue:44295", "issue:36296|issue:43116", "issue:42907|issue:43452", - "issue:44038|issue:44222", - "issue:43824|issue:44351", - "issue:45216|issue:45478", - "issue:43992|issue:44704", - "issue:42757|issue:44351", - "issue:42907|issue:44222", - "issue:43122|issue:44977", - "issue:43278|issue:44945" + "issue:44038|issue:44222" ], - "prepared_review_unit_hash": "7c880a32e575767c192f79e8912a9c7b6fbde3c73b3be576db97e46f2c6cecd0", + "prepared_review_unit_hash": "1883f0137650b61085dcd3b6734d1605d3dea7cab5b786fc87016bf0209c2729", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 36881, - "estimated_input_tokens": 9221, - "estimated_eval_tokens": 18698 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36560, + "estimated_input_tokens": 9140, + "estimated_eval_tokens": 18536 }, "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 regressions, import errors, tokenizer issues, and model-specific bugs. None of the soft pairs are close enough to treat as the same underlying fix, so I would keep them separate.", - "confidence": 0.81, - "canonical_issue_reason": "issue:43824 is the best representative anchor because it is a concrete import regression with the strongest cross-link signal in the set; the rest of the items are too heterogeneous to form a single true duplicate cluster.", + "summary": "No duplicate cluster here: the soft pairs mostly connect unrelated regressions, API changes, and model-specific bugs rather than the same underlying issue.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue; the cluster is heterogeneous and the soft pairs are all false positives.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43824 is the most suitable issue to anchor triage: specific, high-signal, and broadly relevant compared with the more niche or ambiguous reports.", + "best_issue_reason": "issue:45278 is the broadest user-facing report in the set (multiple import errors after upgrading), so it is the best umbrella issue if one must be chosen.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Different bugs: generation kwargs after sync_gpus vs D-FINE auxiliary-loss handling." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "CLI env command breakage is unrelated to a sentence-transformers import NameError." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer output regression vs multimodal assistant_mask generation; different code paths." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong decoder type are separate issues." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 settings warning is unrelated to torch.split return-value handling." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Both involve loading logic, but one is mapping-name None handling and the other is cached_file kwargs propagation." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Sentence-Transformers loading hang and Mac bus error on CLIP are different failures." + }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Both mention generation/runtime problems, but one is a Python version compatibility issue and the other is incorrect pipeline-parallel LLM output; different bugs." + "reason": "Python version support breakage is not the same as pipeline parallelism producing wrong outputs." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Dequantized save failure and incorrect PP outputs are unrelated." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "Both involve processor/chat-template code paths, but one is about kwargs not reaching cached_file and the other is a crash on assistant tool-call messages." + "reason": "Processor cached_file kwargs and chat-template crashes on tool calls are different APIs." }, { "left": "issue:43122", "right": "issue:44514", "accept": false, - "reason": "Tokenizer behavior is involved in both, but one is a version-to-version tokenization change and the other is a batched apply_chat_template padding crash." + "reason": "Tokenization regression and batched apply_chat_template crash are distinct bugs." }, { - "left": "issue:43232", - "right": "issue:44743", + "left": "issue:44295", + "right": "issue:44297", "accept": false, - "reason": "Both touch generation/cache internals, but the reported failures are different: sync_gpus kwarg handling versus recurrent state reset in modular_qwen3_5." + "reason": "Position_ids buffer access and tokenizer_class mismatch on save_pretrained are different serialization problems." }, { - "left": "issue:44295", - "right": "issue:44297", + "left": "issue:43232", + "right": "issue:44743", "accept": false, - "reason": "These are unrelated save/load issues: one concerns reading a registered position_ids buffer, the other a tokenizer_class mismatch in tokenizer_config.json." + "reason": "Generation update after sync_gpus is unrelated to Qwen3.5 recurrent-state reset with cache." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Both are model/tokenizer complaints, but one is an offset_mapping bug and the other is a load failure for a specific tokenizer/model artifact." + "reason": "Qwen3 offset_mapping error and sleng-bert load failure are unrelated tokenizer/model-loading issues." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "Same SAM3 Video family, but one is a config/model-type warning and the other is a missing attribute on Sam3VisionEncoderOutput; not the same concrete bug." + "reason": "Model-type warning and missing attribute in SAM 3 Video are not the same bug." }, { "left": "issue:43232", "right": "issue:45245", "accept": false, - "reason": "Generation kwarg handling and a category-count runtime limit are unrelated problems." + "reason": "Generation kwargs bug and category-count runtime error are unrelated." }, { "left": "issue:42907", "right": "issue:43937", "accept": false, - "reason": "Different model-family errors: dequantized save failure versus invalid GenerationConfig for GLM-5." + "reason": "Model save failure for dequantized Ministrals/Devstrals does not match invalid GenerationConfig." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "Both are import-related, but one is a single missing AutoImageProcessor export and the other is a broad upgrade breakage report with many possible causes." + "reason": "Both mention imports, but AutoImageProcessor missing is not the same concrete regression as the broad import-error report." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Sentence-transformers import NameError and SAM3 auto-mapping CI breakage are distinct failures." + "reason": "Sentence-transformers import NameError and auto_mappings CI break from removed configs are different import failures." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "A missing ReLU in the DSA indexer and a GLM-5 RoPE implementation issue are unrelated." + "reason": "DSA indexer ReLU discussion and GLM-5 RoPE implementation are unrelated model internals." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Both involve Qwen tokenizer/chat-template behavior, but one is offset_mapping and the other is assistant_masks being all zero for multimodal inputs." + "reason": "Wrong offset_mapping and zero assistant_masks are both tokenization-adjacent but not the same underlying bug." }, { - "left": "issue:44625", - "right": "issue:44829", + "left": "issue:44261", + "right": "issue:44373", "accept": false, - "reason": "Qwen3.5 num_labels propagation and flash_attention_3 degenerate training are different model behavior bugs." + "reason": "Numerical precision issue in MLA q_a_layernorm is unrelated to a docstring correction." }, { - "left": "issue:44261", - "right": "issue:44373", + "left": "issue:44625", + "right": "issue:44829", "accept": false, - "reason": "One is a missing config.rms_norm_eps causing precision mismatch; the other is a wrong docstring for position_ids." + "reason": "num_labels propagation in Qwen3.5 is unrelated to flash_attention_3 degenerate training." }, { - "left": "issue:29127", - "right": "issue:30990", + "left": "issue:43116", + "right": "issue:45200", "accept": false, - "reason": "LayoutLMv3 error-message clarity and Sentence Transformers loading hang are unrelated." + "reason": "Example-script multi-label bug and Gemma 4 mm_token_type_ids defaulting are different training/fine-tuning issues." }, { "left": "issue:43122", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions, but one is a generic tokenization change across versions and the other is a Deepseek tokenizer correctness bug." + "reason": "Both are tokenizer regressions, but they involve different models and different failure modes." }, { - "left": "issue:43116", - "right": "issue:45200", + "left": "issue:29127", + "right": "issue:30990", "accept": false, - "reason": "Multi-label classification empty results and Gemma 4 mm_token_type_ids defaults are different example/model configuration issues." + "reason": "LayoutLMv3 error-message clarity and Sentence-Transformers loading hang are unrelated." }, { "left": "issue:42371", "right": "issue:42831", "accept": false, - "reason": "TF32 API guidance and FineGrainedFP8 accuracy regression are not the same bug." + "reason": "TF32 behavior settings and FineGrainedFP8 accuracy are different numerical-stability topics." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "Qwen3-VL-Moe loading/version issue and tokenizer save metadata mismatch are unrelated." + "reason": "Qwen3-VL-Moe v5 bug and tokenizer_class mismatch on save are unrelated." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Different affected models and failure modes: save_pretrained on dequantized models versus loading Phi-4 multimodal instruct." + "reason": "Saving dequantized Ministrals/Devstrals and loading Phi-4 multimodal are different model-loading failures." }, { "left": "issue:44038", - "right": "issue:44222", - "accept": false, - "reason": "Both mention MoE/quantization themes, but one is a load regression and the other is an FP8 save_pretrained problem." - }, - { - "left": "issue:43824", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but for different symbols and different upstream breakages; not a single duplicate fix." - }, - { - "left": "issue:45216", - "right": "issue:45478", - "accept": false, - "reason": "Both are Qwen3.5 regressions, but one is checkpoint saving and the other is from_pretrained loading; too different to merge as one bug." - }, - { - "left": "issue:43992", - "right": "issue:44704", + "right": "issue:44295", "accept": false, - "reason": "UMT5Encoder missing embed_tokens.weight and AutoProcessor kwargs forwarding are separate issues." + "reason": "Qwen3-VL-Moe regression and position_ids buffer read error do not share the same code path." }, { - "left": "issue:42757", - "right": "issue:44351", + "left": "issue:36296", + "right": "issue:43116", "accept": false, - "reason": "These are different missing-import complaints from different modules and symbols." + "reason": "Tensor-parallel training bug and empty multi-label classification results are unrelated." }, { "left": "issue:42907", - "right": "issue:44222", - "accept": false, - "reason": "Both concern saving models, but one is dequantized Ministral/Devstral saving and the other is FP8 MoE saving; different code paths." - }, - { - "left": "issue:43122", - "right": "issue:44977", - "accept": false, - "reason": "Both are tokenizer/generation-adjacent Qwen3.5 reports, but one is a tokenization version change and the other is flash-attention generation failure." - }, - { - "left": "issue:43278", - "right": "issue:44945", + "right": "issue:43452", "accept": false, - "reason": "Embedding dtype drift during evaluation and incorrect output under pipeline parallelism are unrelated behaviors." + "reason": "GGUF from_pretrained breakage and failing to save dequantized models are different format-handling bugs." }, { "left": "issue:44038", - "right": "issue:45290", - "accept": false, - "reason": "Qwen3-VL-Moe loading/version regression and chat-template crash on tool-call messages are different failures." - }, - { - "left": "issue:43116", - "right": "issue:44521", - "accept": false, - "reason": "Classification example script behavior and multimodal assistant_masks generation are unrelated." - }, - { - "left": "issue:43408", - "right": "issue:44261", - "accept": false, - "reason": "A SAM3 model-type warning and an MLA q_a_layernorm precision/config issue do not share the same underlying bug." - }, - { - "left": "issue:44295", - "right": "issue:44360", + "right": "issue:44222", "accept": false, - "reason": "A buffer access problem and a missing ReLU in the DSA indexer are unrelated." + "reason": "Both mention MoE, but one is a v5 model bug and the other is FP8 save_pretrained behavior." } ] }, @@ -11880,51 +11729,50 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:22355", "issue:29127", "issue:33453", "issue:38617", "issue:39401", - "issue:41628", "issue:42222", "issue:42371", "issue:42673", + "issue:42757", + "issue:42907", "issue:43064", "issue:43116", - "issue:43421", + "issue:43122", + "issue:43278", "issue:43441", - "issue:43450", "issue:43493", "issue:43673", "issue:43784", + "issue:43824", "issue:43825", "issue:43827", "issue:43874", - "issue:43906", "issue:43931", "issue:43937", "issue:43976", - "issue:43986", + "issue:43992", "issue:44038", "issue:44075", "issue:44117", "issue:44222", - "issue:44291", + "issue:44351", "issue:44485", + "issue:44704", "issue:44779", - "issue:44829", "issue:44936", "issue:44945", "issue:44977", - "issue:45071", "issue:45072", "issue:45083", "issue:45216", "issue:45278", "issue:45341", - "issue:45405", "issue:45446", "issue:45478", "issue:45561", @@ -11932,19 +11780,26 @@ "issue:45588" ], "soft_pairs": [ + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:43992|issue:44704", + "issue:42757|issue:44351", + "issue:42907|issue:44222", + "issue:43122|issue:44977", + "issue:43278|issue:44945", "issue:43784|issue:45446", + "issue:43937|issue:44485", + "issue:38617|issue:45278", "issue:42222|issue:43441", "issue:42673|issue:45072", - "issue:38617|issue:45278", - "issue:43937|issue:44485", "issue:43064|issue:43673", "issue:22355|issue:29127", "issue:42371|issue:43874", "issue:43116|issue:43493", "issue:44075|issue:44936", - "issue:43976|issue:45446", "issue:44038|issue:45478", "issue:44222|issue:45478", + "issue:43976|issue:45446", "issue:43116|issue:43825", "issue:43937|issue:45563", "issue:33453|issue:39401", @@ -11954,211 +11809,216 @@ "issue:44117|issue:45588", "issue:43827|issue:45446", "issue:43937|issue:44977", - "issue:42673|issue:43906", - "issue:45341|issue:45561", - "issue:45405|issue:45446", - "issue:44291|issue:45071", - "issue:43450|issue:43986", - "issue:39401|issue:43421", - "issue:44829|issue:45216", - "issue:38617|issue:41628" + "issue:45341|issue:45561" ], - "prepared_review_unit_hash": "be34952fbb8ab53def645b932d2d395d2e9e21eb0be92fa7228340ff3fc07d44", + "prepared_review_unit_hash": "77e24539e91bbae2c7d72f8b591ea983e10a0309f6a8c13ebacfbfe924774449", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36460, - "estimated_input_tokens": 9115, - "estimated_eval_tokens": 18486 + "serialized_chars": 35672, + "estimated_input_tokens": 8918, + "estimated_eval_tokens": 18092 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is thematically related by Transformers v5 regressions and import/model-loading failures, but the items are mostly distinct bugs rather than true duplicates. The broadest umbrella issue is the generic import-regression report, while the rest are specific model- or feature-level failures.", - "confidence": 0.82, - "canonical_issue_reason": "Issue 45278 is the broadest representative of the cluster because it describes multiple import failures after upgrading to v5.5.0, which is the closest umbrella among these largely unrelated regressions.", + "summary": "This cluster is not a true duplicate set; it mixes unrelated Transformers v5 regressions, model-specific bugs, import failures, and docs/API issues. The closest cluster anchor is the broad import-regression report, but most pairs should stay separate.", + "confidence": 0.92, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: it describes multiple import errors after upgrading from 4.57.0 to 5.5.0, which makes it a better anchor than the narrower model- or API-specific issues.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45278 is the best single issue to represent the cluster because it is the most general and central v5 import-regression report; the other items are narrower, unrelated concrete failures.", + "best_issue_reason": "issue:45278 is the best representative issue for this loose group because it captures the widest user-facing regression pattern, while the other issues are mostly specific to one model, one symbol, or one code path.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import failures, but they are different missing symbols (`Qwen2_5_VLForConditionalGeneration` vs `HybridCache`) with different likely causes and fixes." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5/Moe save-or-load regressions, but one is about incorrect saved checkpoints and the other is a from_pretrained load error; not the same concrete bug." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "`embed_tokens.weight` missing in `UMT5Encoder.from_pretrained` is a model-loading state issue, while `AutoProcessor.from_pretrained` not passing kwargs is a processor/file-resolution bug." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "These are distinct import errors from different modules and symbols; they do not point to one shared underlying regression." + }, + { + "left": "issue:42907", + "right": "issue:44222", + "accept": false, + "reason": "One is about saving dequantized Ministral/Devstral models, the other about FP8 save_pretrained behavior for moe models; same broad save path, but not the same bug." + }, + { + "left": "issue:43122", + "right": "issue:44977", + "accept": false, + "reason": "Both mention tokenization differences, but one is a version-to-version tokenizer behavior change and the other is a Qwen3.5 flash-attention generation problem." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype mismatch in evaluate and incorrect output with pipeline parallelism are separate training/eval runtime issues with different code paths." + }, { "left": "issue:43784", "right": "issue:45446", "accept": false, - "reason": "Both are import/version-related failures, but they affect different symbols and code paths; not the same bug." + "reason": "One is a missing `nn` import in sentence-transformers integration, the other is an incorrect PyTorch version gate for AuxRequest in flex_attention; unrelated failures." }, { - "left": "issue:42222", - "right": "issue:43441", + "left": "issue:43937", + "right": "issue:44485", "accept": false, - "reason": "Different model families and different failure modes (VitPose breakage vs FlashAttention on Ministral-3)." + "reason": "A generation-config validation error for GLM-5 is not the same as a RoPE implementation discussion/bug; they affect different parts of the model stack." }, { - "left": "issue:42673", - "right": "issue:45072", + "left": "issue:38617", + "right": "issue:45278", "accept": false, - "reason": "Unrelated subsystems and symptoms: VRAM leak in dataloader threads vs dtype mismatch in bfloat16 inference." + "reason": "Both are import-related, but one is a missing `layer_type_validation` symbol and the other is a broad upgrade report with multiple import errors; not one concrete bug." }, { - "left": "issue:38617", - "right": "issue:45278", + "left": "issue:42222", + "right": "issue:43441", "accept": false, - "reason": "45278 is an umbrella report about many import errors; 38617 is one specific missing import. Related theme, but not the same underlying bug." + "reason": "Vitpose model breakage and Ministral-3 flash-attention failures are different model-specific regressions with different execution paths." }, { - "left": "issue:43937", - "right": "issue:44485", + "left": "issue:42673", + "right": "issue:45072", "accept": false, - "reason": "Same model prefix, but one is a GenerationConfig validation error and the other is a RoPE implementation discussion." + "reason": "A VRAM leak in multi-threaded Qwen3ForCausalLM use is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel bfloat16 inference." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Different training/loading paths and different symptoms; not plausibly one merged fix." + "reason": "FSDP2/PEFT optimizer-state corruption and missing GenerationMixin cache during chunked prefill are different subsystems and failure modes." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "No shared concrete bug: missing module import vs LayoutLMv3 error-message clarity." + "reason": "`transformers.onnx` import failure and LayoutLMv3 box-validation messaging are entirely unrelated." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 settings guidance and a missing image-patch helper are unrelated issues." + "reason": "TF32 behavior settings and a missing `get_number_of_image_patches` method in Glm46VImageProcessorFast are unrelated issues." }, { "left": "issue:43116", "right": "issue:43493", "accept": false, - "reason": "Different models and different correctness bugs; only a broad NLP/modeling similarity." + "reason": "Multi-label classification example output emptiness and SigLIP2 HF-vs-JAX discrepancy are different problems in different model/example paths." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "Optimizer argument handling and trainer post-train evaluation failure are distinct code paths." - }, - { - "left": "issue:43976", - "right": "issue:45446", - "accept": false, - "reason": "Both mention version compatibility, but they concern different dependencies and different checks." + "reason": "SGD argument handling and `trainer.evaluate()` failing after `trainer.train()` are separate trainer/optimizer behaviors." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Both are Qwen3 MoE-related, but they describe different failures (v5.0/Qwen3-VL-Moe vs from_pretrained error on Qwen3.5 Moe)." + "reason": "Qwen3-VL-Moe support in Transformers 5.0 and Qwen3.5 Moe from_pretrained errors are related by model family, but not the same concrete bug or fix." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "Save-pretrained FP8 behavior and from_pretrained loading failure are different concrete bugs." + "reason": "FP8 save_pretrained for moe and Qwen3.5 Moe from_pretrained load errors affect different code paths; they are not mergeable into one PR." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python 3.9/3.10 compatibility claims for Transformers 5.1.0 and a flex_attention import/version check bug are unrelated." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "Example-script multi-label output bug and pipeline doc/error-message mismatch are unrelated." + "reason": "A classification example script returning empty multi-label results is not the same as a pipeline() error message incorrectly advertising translation support." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "Different generation issues: invalid GenerationConfig vs stale warning for num_return_sequences." + "reason": "GLM-5 generation-config validation and a stale warning in paged generate() are different generation-path issues." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different tokenizers and different behaviors." + "reason": "Tokenizer loading regression and Qwen3 offset_mapping errors are both tokenizer-related but refer to different regressions and likely different fixes." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Feature-extraction helper behavior and PyTorch version-check logic are unrelated." + "reason": "A helper-function behavior issue in qwen3_omni_moe is unrelated to the AuxRequest import/version-check bug in flex_attention." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Different tokenizer models and different incorrect outputs; not the same underlying defect." + "reason": "Both are tokenizer bugs, but one is Qwen3 offset_mapping and the other is Deepseek incorrect tokenization; different models and likely different root causes." }, { "left": "issue:43931", "right": "issue:44945", "accept": false, - "reason": "Weight-shape loading mismatch and pipeline-parallel output corruption are distinct failure modes." + "reason": "Qwen3-VL-30B weight-shape mismatch and incorrect output under pipeline parallelism are different model-loading vs runtime-parallelism problems." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "Tokenizer mapping fallback bug and flash_attention s_aux handling are unrelated code paths." + "reason": "TOKENIZER_MAPPING_NAMES returning None and a flash_attention crash on `s_aux=None` are unrelated implementation bugs." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs referencing removed pipeline APIs and a PyTorch version check bug are not the same issue." + "reason": "Docs still referencing pipeline() after removals and an incorrect PyTorch version check are not the same underlying bug or change." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "Different models and different symptoms; both are generation-related but not the same bug." - }, - { - "left": "issue:42673", - "right": "issue:43906", - "accept": false, - "reason": "43906 is an isolated reproduction of another issue; nothing indicates it is the same VRAM leak." + "reason": "GLM-5 generation config validation and Qwen3.5 flash-attention generation failure are separate model-specific generation bugs." }, { "left": "issue:45341", "right": "issue:45561", "accept": false, - "reason": "Both touch testing_utils.py, but one is a generic bug report and the other is a specific xdist race; insufficient evidence they are the same bug." - }, - { - "left": "issue:45405", - "right": "issue:45446", - "accept": false, - "reason": "PEFT minimum-version packaging issue and a PyTorch import-version check are unrelated." - }, - { - "left": "issue:44291", - "right": "issue:45071", - "accept": false, - "reason": "Both are v5 regressions, but they affect different objects and code paths: init_empty_weights loading vs PretrainedConfig type checking." - }, - { - "left": "issue:43450", - "right": "issue:43986", - "accept": false, - "reason": "Batched video processor shape bug and AutoProcessor crash without torchvision are distinct failures." - }, - { - "left": "issue:38617", - "right": "issue:41628", - "accept": false, - "reason": "Both are import errors, but they involve different missing exports and different call sites." + "reason": "A testing_utils bug and an xdist race on captured_info.txt are both test-infra related, but they are not the same concrete race or code path." } ] }, @@ -12180,9 +12040,10 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:38617", + "issue:39401", "issue:41628", "issue:42175", "issue:42222", @@ -12194,37 +12055,36 @@ "issue:43097", "issue:43122", "issue:43408", - "issue:43441", - "issue:43452", + "issue:43421", + "issue:43450", "issue:43577", "issue:43582", "issue:43673", "issue:43742", "issue:43874", - "issue:43931", + "issue:43906", "issue:43937", "issue:43986", - "issue:44038", "issue:44075", "issue:44117", "issue:44220", "issue:44261", - "issue:44263", + "issue:44291", "issue:44410", "issue:44483", "issue:44484", - "issue:44485", "issue:44493", "issue:44496", "issue:44829", "issue:44933", - "issue:44987", - "issue:45003", "issue:45071", "issue:45081", + "issue:45216", "issue:45278", "issue:45310", "issue:45341", + "issue:45405", + "issue:45446", "issue:45459", "issue:45464", "issue:45478", @@ -12232,280 +12092,245 @@ "issue:45656" ], "soft_pairs": [ + "issue:44291|issue:45071", + "issue:45405|issue:45446", + "issue:42673|issue:43906", + "issue:43450|issue:43986", + "issue:39401|issue:43421", + "issue:44829|issue:45216", + "issue:38617|issue:41628", "issue:44410|issue:45310", "issue:44117|issue:45081", "issue:44484|issue:45310", "issue:42371|issue:44261", - "issue:43408|issue:43742", "issue:42222|issue:45542", - "issue:42673|issue:43582", + "issue:43408|issue:43742", "issue:43937|issue:44496", + "issue:42673|issue:43582", "issue:42673|issue:43577", "issue:42175|issue:42222", "issue:38617|issue:44220", "issue:44933|issue:45278", "issue:44483|issue:45464", - "issue:43012|issue:43577", - "issue:42617|issue:45341", "issue:44075|issue:45656", - "issue:43066|issue:43122", + "issue:42617|issue:45341", + "issue:43012|issue:43577", "issue:43673|issue:45341", + "issue:43066|issue:43122", "issue:44484|issue:45478", "issue:44410|issue:45478", "issue:43874|issue:44261", "issue:43097|issue:44493", - "issue:42222|issue:45459", - "issue:43441|issue:44829", - "issue:43012|issue:43408", - "issue:44263|issue:44485", - "issue:41628|issue:45071", - "issue:45071|issue:45542", - "issue:43986|issue:44484", - "issue:43931|issue:44410", - "issue:45003|issue:45341", - "issue:43986|issue:45341", - "issue:43931|issue:44038", - "issue:44117|issue:44987", - "issue:43097|issue:43452" + "issue:42222|issue:45459" ], - "prepared_review_unit_hash": "9cf55afc2d22aa12ebbb3030f7629cf459c83e6baa0f0cab4afaca34de0ff21d", + "prepared_review_unit_hash": "a76d19cbc6d7ee461bdbb3f6a22fbfc57fe17ed5a35cb3d3d8b41ec8962191c6", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 37837, - "estimated_input_tokens": 9460, - "estimated_eval_tokens": 19176 + "soft_pair_count": 30, + "serialized_chars": 36526, + "estimated_input_tokens": 9132, + "estimated_eval_tokens": 18520 }, "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 v5 regressions and upgrade breakages (imports, model loading, tokenizer behavior, API changes). The soft pairs are usually only related by broad subsystem/version, not the same underlying bug, so I would not merge them as duplicates.", + "summary": "The set is highly heterogeneous: it mixes import-regression reports, tokenizer behavior bugs, model/config loading errors, backend/version checks, and unrelated feature/optimizer issues. No soft pair looks like a true duplicate; the apparent overlaps are only topical, not the same underlying bug.", "confidence": 0.77, - "canonical_issue_reason": "issue:45278 is the broadest representative of the cluster\u2019s recurring post-upgrade import/regression theme, though it is not a true duplicate of the other items.", + "canonical_issue_reason": "issue:45278 is the broadest umbrella among the import-regression reports (multiple import errors after upgrading), so it is the best representative issue if one must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best umbrella issue for triage because it captures the widest version-upgrade import breakage pattern seen across the set.", + "best_issue_reason": "issue:45278 has the widest scope and is most likely to subsume nearby import-failure reports, making it the best global issue representative in this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Different failures: unexpected `_is_hf_initialized` argument during init_empty_weights vs PreTrainedConfig type-checking regression." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "Unrelated backend/version problems: PEFT minimum version bump vs PyTorch version gating for flex_attention AuxRequest." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "Both mention Qwen3-related behavior, but one is a VRAM leak in dataloader threads and the other is a reproduction of a separate prior issue." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Different subsystems: batched video processor output shape vs AutoProcessor crash when torchvision is missing." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Tokenizer offset-mapping bug is not the same as runtime post-processor updates after special token modification." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Different concrete problems: flash_attention_3 causes degenerate training, while Qwen3.5 save_pretrained produces an incorrect checkpoint." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols and code paths." + }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Different concrete failures: missing projections in qwen3next vs a Qwen3.5-MoE from_pretrained loading error." + "reason": "Different model families and failures: missing projections in qwen3next vs Qwen3.5 MoE from_pretrained error." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Both involve tokenizer internals, but one is TOKENIZER_MAPPING_NAMES returning None and the other is a Mistral regex patch crash; not the same bug." + "reason": "Tokenizer mapping None handling is separate from the Mistral regex patch crash in backend_tokenizer access." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "save_pretrained shard-size behavior is unrelated to Qwen3.5-MoE loading failure." + "reason": "One is a save_pretrained shard-size question; the other is a Qwen3.5 MoE loading regression." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API guidance and MLA rms_norm_eps precision mismatch are separate issues." - }, - { - "left": "issue:43408", - "right": "issue:43742", - "accept": false, - "reason": "Model-type warning for sam3_tracker/sam3_video is not the same as a KeyError loading MobileLLM-125M." + "reason": "TF32 control API concern is unrelated to MLA q_a_layernorm missing rms_norm_eps precision mismatch." }, { "left": "issue:42222", "right": "issue:45542", "accept": false, - "reason": "VitPose model breakage and missing TensorFlow backend are unrelated." + "reason": "Different domains: vitpose model breakage vs missing TensorFlow backend when only tensorboard is installed." }, { - "left": "issue:42673", - "right": "issue:43582", + "left": "issue:43408", + "right": "issue:43742", "accept": false, - "reason": "VRAM leakage across dataloader threads is a different code path from Apple Silicon caching_allocator_warmup TypeError." + "reason": "Different warnings/errors: sam3_video vs sam3_tracker type mismatch is not the same as MobileLLM loading key error." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "GenerationConfig validation for GLM-5 is not the same as unrecognized model_type / config.json loading failure." + "reason": "GenerationConfig validation for GLM-5 is unrelated to unrecognized model/config loading for Olmo-Hybrid-Instruct." + }, + { + "left": "issue:42673", + "right": "issue:43582", + "accept": false, + "reason": "Both are runtime issues, but VRAM leak in threads is not the same as Apple Silicon warmup TypeError." }, { "left": "issue:42673", "right": "issue:43577", "accept": false, - "reason": "VRAM leak in multithreaded dataloading is unrelated to dtype remaining float32 on BLIP2 load." + "reason": "VRAM leak in threaded use is different from BLIP2 dtype fields staying float32 after loading." }, { "left": "issue:42175", "right": "issue:42222", "accept": false, - "reason": "Package extra dependency resolution is not the same bug as broken VitPose model loading." + "reason": "TensorFlow backend packaging issue is unrelated to vitpose model breakage." }, { "left": "issue:38617", "right": "issue:44220", "accept": false, - "reason": "ImportError from configuration_utils and _torch_extract_fbank_features are separate failures." + "reason": "Different import/feature failures: missing layer_type_validation symbol vs audio fbank extraction issue." }, { "left": "issue:44933", "right": "issue:45278", "accept": false, - "reason": "A specific nonexistent import from image_utils is broader-related only in that both are import problems after upgrade." + "reason": "Same broad import-error theme, but 44933 is a specific missing import from image_utils while 45278 is a broader version-upgrade import breakage report." }, { "left": "issue:44483", "right": "issue:45464", "accept": false, - "reason": "Both mention chat/completions, but one is a general request rejection and the other is a streaming inference failure on a specific model." + "reason": "Chat/completions API request rejection is not the same as streaming inference failure on Qwen3.5-0.8B." }, { - "left": "issue:43012", - "right": "issue:43577", + "left": "issue:44075", + "right": "issue:45656", "accept": false, - "reason": "A compile-time precision warning is not the same as BLIP2 components staying float32 on load." + "reason": "SGD optimizer args ignored and optimizer.step called twice are different optimizer bugs." }, { "left": "issue:42617", "right": "issue:45341", "accept": false, - "reason": "Running 3d_parallel.py and a bug in testing_utils.py are unrelated." + "reason": "3d_parallel.py runtime failure is unrelated to a testing_utils bug." }, { - "left": "issue:44075", - "right": "issue:45656", + "left": "issue:43012", + "right": "issue:43577", "accept": false, - "reason": "SGD argument handling and deepspeed causing double optimizer steps are different optimizer bugs." + "reason": "PyTorch warning on compiling bfloat16 models is different from BLIP2 dtype fields remaining float32." }, { - "left": "issue:43066", - "right": "issue:43122", + "left": "issue:43673", + "right": "issue:45341", "accept": false, - "reason": "Tokenizer decoder type regression and different tokenization output are related only at a very high level; not the same concrete defect." + "reason": "Generation cache missing in v5.0.0 during chunked_prefill is unrelated to a testing_utils bug." }, { - "left": "issue:43673", - "right": "issue:45341", + "left": "issue:43066", + "right": "issue:43122", "accept": false, - "reason": "Generation cache/chunked_prefill behavior is unrelated to testing_utils." + "reason": "Both are tokenization-related, but one is decoder-type selection in v5 and the other is a concrete tokenization difference across versions." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "save_pretrained shard sizing is unrelated to Qwen3.5-MoE from_pretrained errors." + "reason": "Different Qwen3.5 MoE loading bug vs unrelated save_pretrained max_shard_size question." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Both are Qwen-family issues, but the missing projections bug and the MoE loading bug are different code-path problems." + "reason": "Missing qkv/gate projections in qwen3next is not the same as the Qwen3.5 MoE from_pretrained error." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Missing get_number_of_image_patches in GLM46V image processor is unrelated to MLA layernorm epsilon precision handling." + "reason": "Missing get_number_of_image_patches in GLM46V is separate from MLA q_a_layernorm rms_norm_eps precision error." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removed tie_embeddings_and_encoder_decoder API and unexpected position-id keys are separate v5 regressions." + "reason": "Removed tie_embeddings_and_encoder_decoder API concern is different from unexpected position-id keys in model loading." }, { "left": "issue:42222", "right": "issue:45459", "accept": false, - "reason": "VitPose model breakage and protobuf-hiding tokenizer errors are unrelated." - }, - { - "left": "issue:43441", - "right": "issue:44829", - "accept": false, - "reason": "FlashAttention load failure for Ministral-3 is not the same as degenerate training with flash_attention_3." - }, - { - "left": "issue:43012", - "right": "issue:43408", - "accept": false, - "reason": "bfloat16 precision warning during compile and sam3 tracker/video model-type warning are unrelated." - }, - { - "left": "issue:44263", - "right": "issue:44485", - "accept": false, - "reason": "A torch.split return-value issue in GlmMoeDsaIndexer is not the RoPE implementation bug." - }, - { - "left": "issue:41628", - "right": "issue:45071", - "accept": false, - "reason": "AutoImageProcessor import failure is not the same as PreTrainedConfig type-checking breakage." - }, - { - "left": "issue:45071", - "right": "issue:45542", - "accept": false, - "reason": "Config type checking regression and missing TensorFlow backend are different failures." - }, - { - "left": "issue:43986", - "right": "issue:44484", - "accept": false, - "reason": "AutoProcessor/torchvision video-model loading crash is unrelated to max_shard_size behavior." - }, - { - "left": "issue:43931", - "right": "issue:44410", - "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and qwen3next missing projections are different model-loading problems." - }, - { - "left": "issue:45003", - "right": "issue:45341", - "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to the testing_utils bug." - }, - { - "left": "issue:43986", - "right": "issue:45341", - "accept": false, - "reason": "Video-model AutoProcessor crash and testing_utils are separate issues." - }, - { - "left": "issue:43931", - "right": "issue:44038", - "accept": false, - "reason": "Qwen3-VL shape mismatch and Qwen3-VL-Moe bug are related by model family only, not the same defect." - }, - { - "left": "issue:44117", - "right": "issue:44987", - "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and loading physical-intelligence/fast failures are not the same code-path problem." - }, - { - "left": "issue:43097", - "right": "issue:43452", - "accept": false, - "reason": "Removal of tie_embeddings_and_encoder_decoder and gguf_file breaking from_pretrained are distinct regressions." + "reason": "Vitpose model breakage is unrelated to tokenizer error masking when protobuf is absent." } ] }, @@ -12527,30 +12352,28 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:38617", + "issue:41628", "issue:42757", - "issue:42886", - "issue:42907", "issue:43012", "issue:43064", "issue:43066", "issue:43097", - "issue:43208", "issue:43335", "issue:43352", - "issue:43381", "issue:43408", + "issue:43441", + "issue:43452", "issue:43502", "issue:43531", - "issue:43550", "issue:43824", "issue:43825", "issue:43873", "issue:43874", - "issue:43901", "issue:43906", + "issue:43931", "issue:43976", "issue:43986", "issue:43992", @@ -12569,18 +12392,32 @@ "issue:44655", "issue:44704", "issue:44829", + "issue:44987", + "issue:45003", "issue:45071", "issue:45103", "issue:45245", "issue:45341", "issue:45372", "issue:45375", - "issue:45479", + "issue:45542", "issue:45588" ], "soft_pairs": [ - "issue:43408|issue:44368", + "issue:43441|issue:44829", + "issue:44263|issue:44485", + "issue:43012|issue:43408", + "issue:41628|issue:45071", + "issue:45071|issue:45542", + "issue:43931|issue:44410", + "issue:43986|issue:44484", + "issue:45003|issue:45341", + "issue:43931|issue:44038", + "issue:43986|issue:45341", + "issue:44117|issue:44987", + "issue:43097|issue:43452", "issue:43335|issue:45071", + "issue:43408|issue:44368", "issue:43874|issue:44263", "issue:43976|issue:45071", "issue:44829|issue:45245", @@ -12599,246 +12436,255 @@ "issue:43502|issue:43992", "issue:43824|issue:43986", "issue:43097|issue:43873", - "issue:43825|issue:44655", "issue:43066|issue:43906", "issue:43352|issue:44380", - "issue:43992|issue:44484", - "issue:43097|issue:43901", - "issue:43012|issue:44368", - "issue:43012|issue:43550", - "issue:42886|issue:43381", - "issue:44380|issue:45479", - "issue:42907|issue:44655", - "issue:43208|issue:43550", - "issue:43352|issue:43976", - "issue:43064|issue:44380" + "issue:43825|issue:44655", + "issue:43992|issue:44484" ], - "prepared_review_unit_hash": "0580941e963835b72009d708b2a34d9ce123b8422d8de1dc8086282e3f64d353", + "prepared_review_unit_hash": "1d71b0352666eec2bc7df93a43f1f5f30eb5ef9d00768f5377f00f311c8cb35a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37630, - "estimated_input_tokens": 9408, - "estimated_eval_tokens": 19072 + "soft_pair_count": 36, + "serialized_chars": 38501, + "estimated_input_tokens": 9626, + "estimated_eval_tokens": 19508 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly separate Transformers regressions across different subsystems and models; none of the soft pairs look like the same underlying bug/change closely enough to merge. The closest cluster center is a broad v5 API regression issue, but overall duplicate confidence is low.", - "confidence": 0.31, - "canonical_issue_reason": "issue:43097 is the most central of the set: it concerns a broad v5 core-API regression, has the most soft links, and is more representative than the narrower import/attention/model-specific bugs.", + "summary": "This cluster is mostly a loose title-similarity grouping, not a true duplicate set. The paired issues span unrelated import errors, model-specific loading bugs, training regressions, and warning/compatibility problems, so all soft edges should be rejected.", + "confidence": 0.41, + "canonical_issue_reason": "If a single issue must anchor the cluster, issue 43502 is the broadest core-loading regression with the most discussion, but the set is not actually one duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": "Best overall issue candidate for the cluster because it is the broadest and most connected issue; not a true duplicate hub, but the closest representative.", + "best_issue_reason": "Issue 43502 is the best overall representative only because it is a high-signal, broadly relevant loading bug with substantial user impact and discussion; it is still not a strong canonical for the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43408", - "right": "issue:44368", + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both mention FlashAttention, but one is a model-specific FA2 failure and the other is a degenerate training bug with FA3; different code paths and symptoms." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "Unrelated GLM issues: one is a torch.split/indexer bug, the other is a RoPE implementation discussion." + }, + { + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "Both are warnings, but about different subsystems and causes: bfloat16 precision warning vs model-type mismatch warning." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "Different regressions: missing AutoImageProcessor import vs PretrainedConfig type-checking breakage." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Type checking in PretrainedConfig is unrelated to the TensorBoard/TF backend detection bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Same broad family name, but one is a Qwen3-VL load shape mismatch and the other is missing projections in qwen3next; not the same bug." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "Video-model AutoProcessor dependency crash is unrelated to the max_shard_size save_pretrained question." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils does not match a small testing_utils bug." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Both are Qwen-VL loading problems, but they affect different model variants and likely different failure modes." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "A torchvision dependency crash is unrelated to a testing utility bug." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping returning None is a distinct loading-path bug from the physical-intelligence/fast model load failure." + }, + { + "left": "issue:43097", + "right": "issue:43452", "accept": false, - "reason": "Both are warning-style reports, but they concern different model families and different warnings; not the same bug." + "reason": "Removed tie_embeddings API and gguf_file breakage are different compatibility issues in different code paths." }, { "left": "issue:43335", "right": "issue:45071", "accept": false, - "reason": "One is a SwitchTransformersConfig edge case, the other is a general PretrainedConfig type-checking regression; distinct failures." + "reason": "SwitchTransformers sparse-layer config bug is unrelated to PretrainedConfig type checking." + }, + { + "left": "issue:43408", + "right": "issue:44368", + "accept": false, + "reason": "Both are warnings, but they concern different configuration fields and different models." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different GLM-family bugs in different components; the symptoms and code paths do not match." + "reason": "Missing image-patch helper in a fast image processor is unrelated to a torch.split return-value bug." }, { "left": "issue:43976", "right": "issue:45071", "accept": false, - "reason": "Python version compatibility failure versus config type-checking breakage; unrelated regressions." + "reason": "Python version compatibility for Transformers 5.1.0 is unrelated to the type-checking regression." }, { "left": "issue:44829", "right": "issue:45245", "accept": false, - "reason": "Different errors in different paths: FlashAttention training degeneration vs category cardinality limit." + "reason": "FlashAttention training degeneration and a category-count limit error are unrelated." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Both involve loading processors/models, but one is a missing torchvision dependency and the other is a mistral_common import break; not the same underlying issue." + "reason": "Different missing-dependency/import failures in different processor paths." }, { "left": "issue:43531", "right": "issue:45341", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to a testing utility bug." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are ImportError reports, but for different missing symbols from different modules and likely different fixes." + "reason": "Both are ImportError reports, but for different missing symbols from different packages and likely different fixes." }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "Different training-stack problems: distributed optimizer-state corruption versus Liger Kernel not applying with model_init." + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to Liger Kernel not being applied under model_init." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "Different GLM-related issues with different root causes; one is missing eps config, the other is RoPE implementation." + "reason": "A missing rms_norm_eps precision issue is not the same as a GLM RoPE implementation discussion." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Both are Qwen-family config/model bugs, but they affect different architectures and missing fields/projections." + "reason": "Missing qkv/gate projections in qwen3next is unrelated to a missing deepstack_visual_indexes config field." }, { "left": "issue:44038", "right": "issue:44410", "accept": false, - "reason": "Related model family, but one is a broader Qwen3-VL-Moe issue and the other is a specific missing projection bug in qwen3next." + "reason": "Different Qwen model families and different architecture bugs; not a single underlying defect." }, { "left": "issue:38617", "right": "issue:44351", "accept": false, - "reason": "Both are import failures, but they concern different API names and different breakpoints." + "reason": "Different missing imports from different modules/symbols; too broad to be one duplicate." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Tokenizer construction bug versus tokenizer mapping assumption bug; different code paths." + "reason": "Tokenizer AddedToken argument duplication is a different bug from TOKENIZER_MAPPING_NAMES sometimes returning None." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "AutoProcessor kwargs forwarding and auto_docstring annotation handling are unrelated." + "reason": "AutoProcessor kwargs forwarding and auto_docstring annotation handling are unrelated code paths." }, { "left": "issue:43097", "right": "issue:45588", "accept": false, - "reason": "One is a removed config attribute warning/regression; the other is a flash_attention integration crash on s_aux=None. Different problems." + "reason": "A removed embedding-tie API and a flash_attention.py AttributeError with s_aux=None are unrelated." }, { "left": "issue:43986", "right": "issue:44117", "accept": false, - "reason": "Processor loading with missing torchvision versus tokenizer mapping None-handling; unrelated." + "reason": "Video-model loading without torchvision is unrelated to tokenizer mapping assumptions." }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "Local-files-only network leakage and missing embed_tokens.weight during load are different loading bugs." + "reason": "local_files_only network leakage and missing embed_tokens.weight are different loading bugs." }, { "left": "issue:43824", "right": "issue:43986", "accept": false, - "reason": "Top-level model import failure versus AutoProcessor/torchvision crash; separate issues." + "reason": "A missing Qwen2.5-VL import and a torchvision-related video processor crash are distinct issues." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "Removed tie_embeddings config behavior and quantization offloading problems are distinct." - }, - { - "left": "issue:43825", - "right": "issue:44655", - "accept": false, - "reason": "Pipeline translation-task messaging and pipeline save_pretrained behavior are different features." + "reason": "Removed tie_embeddings behavior is unrelated to quantization offloading not working." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue do not match as the same bug." + "reason": "The tokenizer decoder type bug is not the same as a generic isolated reproduction of another issue." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Different attention-backend issues on different models; not a single fixable code-path problem." - }, - { - "left": "issue:43992", - "right": "issue:44484", - "accept": false, - "reason": "Missing encoder weights versus shard-size defaults are unrelated save/load behaviors." - }, - { - "left": "issue:43097", - "right": "issue:43901", - "accept": false, - "reason": "Config removal warning and stale pipeline docs about return_all_scores are different issues." - }, - { - "left": "issue:43012", - "right": "issue:44368", - "accept": false, - "reason": "Different warning semantics in different model contexts; not the same defect." - }, - { - "left": "issue:43012", - "right": "issue:43550", - "accept": false, - "reason": "Bfloat16 precision warning and torch.compile SDPA failure are unrelated." - }, - { - "left": "issue:42886", - "right": "issue:43381", - "accept": false, - "reason": "Offline cache loading bug versus gradient-checkpointing eval-mode restriction; distinct topics." - }, - { - "left": "issue:44380", - "right": "issue:45479", - "accept": false, - "reason": "Attention scaling ignored under backends and zero-loss sequence-classification degeneration are unrelated." + "reason": "Nemotron Flash Attention 2 support and GPT2 attention scaling under SDPA/FA are different attention bugs." }, { - "left": "issue:42907", + "left": "issue:43825", "right": "issue:44655", "accept": false, - "reason": "Saving dequantized models and saving Pipeline objects are different save paths." - }, - { - "left": "issue:43208", - "right": "issue:43550", - "accept": false, - "reason": "xLSTM training blockers and Bamba torch.compile/SDPA failure are different model-specific bugs." - }, - { - "left": "issue:43352", - "right": "issue:43976", - "accept": false, - "reason": "Flash Attention support warning and Python-version compatibility failure are unrelated." + "reason": "pipeline() error messaging and Pipeline save_pretrained support are different pipeline-related problems." }, { - "left": "issue:43064", - "right": "issue:44380", + "left": "issue:43992", + "right": "issue:44484", "accept": false, - "reason": "Distributed optimizer-state corruption and GPT2 attention scaling under SDPA/FlashAttention are separate code paths." + "reason": "Missing encoder weights on load is unrelated to the default max_shard_size discussion." } ] }, @@ -12860,24 +12706,28 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:36010", "issue:36683", "issue:38617", "issue:42757", + "issue:42886", + "issue:42907", "issue:42994", - "issue:43010", "issue:43012", "issue:43064", "issue:43066", + "issue:43097", "issue:43208", "issue:43299", "issue:43352", + "issue:43381", "issue:43475", "issue:43502", "issue:43525", "issue:43526", + "issue:43550", "issue:43582", "issue:43618", "issue:43756", @@ -12885,16 +12735,14 @@ "issue:43824", "issue:43867", "issue:43901", - "issue:43931", "issue:43937", "issue:43976", "issue:44038", "issue:44077", - "issue:44188", "issue:44220", "issue:44230", - "issue:44242", "issue:44263", + "issue:44368", "issue:44380", "issue:44483", "issue:44485", @@ -12904,13 +12752,21 @@ "issue:44683", "issue:44938", "issue:45216", - "issue:45278", "issue:45290", "issue:45381", "issue:45440", - "issue:45593" + "issue:45479" ], "soft_pairs": [ + "issue:43097|issue:43901", + "issue:43012|issue:44368", + "issue:43012|issue:43550", + "issue:42886|issue:43381", + "issue:42907|issue:44655", + "issue:44380|issue:45479", + "issue:43208|issue:43550", + "issue:43352|issue:43976", + "issue:43064|issue:44380", "issue:43618|issue:44380", "issue:43208|issue:43526", "issue:44483|issue:44514", @@ -12926,258 +12782,273 @@ "issue:43756|issue:44485", "issue:43901|issue:44655", "issue:36683|issue:43475", - "issue:44220|issue:44655", "issue:42994|issue:44655", "issue:43867|issue:45216", + "issue:44220|issue:44655", "issue:43352|issue:43937", "issue:44263|issue:45440", - "issue:44077|issue:44380", "issue:36683|issue:44077", "issue:43066|issue:45381", + "issue:44077|issue:44380", "issue:43976|issue:44938", "issue:36683|issue:43525", "issue:36010|issue:42757", - "issue:43064|issue:44492", - "issue:44188|issue:44380", - "issue:43352|issue:45593", - "issue:43618|issue:45216", - "issue:36683|issue:44242", - "issue:43010|issue:45278", - "issue:43352|issue:43931" + "issue:43064|issue:44492" ], - "prepared_review_unit_hash": "0f410c4c1563418dd8798a8da3031683b72c047b14329408ef0083499e0a3781", + "prepared_review_unit_hash": "59dca2591857cd72ea2ef0f71608f3b3f18c88d248c43ba1e7121764ba4789e7", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 36610, - "estimated_input_tokens": 9153, - "estimated_eval_tokens": 18562 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38172, + "estimated_input_tokens": 9543, + "estimated_eval_tokens": 19342 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose grab-bag of Transformers regressions and import/model-loading issues rather than one duplicate set. The closest umbrella is the broad import-error report after upgrading to 5.5.0, but most pairings are different concrete bugs.", - "confidence": 0.42, - "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: a generic post-upgrade import-failure report that can plausibly subsume several other import-related regressions. The rest are mostly model-specific or unrelated bugs.", + "summary": "This cluster is mostly a loose similarity set of unrelated regression reports. The only plausible duplicate pair is the CLIP v5 output regression (missing attentions vs missing hidden_states), which looks like the same forward-output code path.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43618 is the best canonical issue because it captures the shared CLIP v5 output regression at a broader level; issue:43761 appears to be a specific symptom of the same model-forward bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best representative because it is the most general, upgrade-wide import regression and sits closest to the repeated import-error theme in the cluster.", + "best_issue_reason": "issue:43618 is the strongest issue to keep as the representative because it is concrete, regression-oriented, and broad enough to subsume the related CLIP hidden-state failure.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43097", + "right": "issue:43901", + "accept": false, + "reason": "Different problems: API removal/regression vs stale docs text. Same theme of API changes, but not the same bug." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Both mention warnings, but one is a PyTorch compile precision warning and the other is a model config warning about tied embeddings." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Unrelated failure modes: bfloat16 compile warning vs Bamba torch.compile/SDPA runtime bug." + }, + { + "left": "issue:42886", + "right": "issue:43381", + "accept": false, + "reason": "Different subsystems and symptoms: offline cache loading vs gradient checkpointing in eval mode." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Both involve saving, but one is dequantized model save failure and the other is saving Pipeline objects; not the same code path." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Different model families and bugs: attention scaling under SDPA/FlashAttention vs degenerate single-label classification loss." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba compile/SDPA issues are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron Flash Attention support and Python-version compatibility are unrelated." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT2 attention scaling under backend changes." + }, { "left": "issue:43618", "right": "issue:44380", "accept": false, - "reason": "Both touch attention code, but one is a CLIP output regression and the other is GPT-2 attention scaling under SDPA/FlashAttention; different models and different failure modes." + "reason": "Both are regressions, but in different model paths: CLIP output fields vs GPT2 attention scaling." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "Completely different subsystems: xLSTM training bugs vs BeitImageProcessorFast label reduction." + "reason": "Different xLSTM training bugs vs BeitImageProcessorFast label reduction bug." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "One is an API request rejection in /v1/chat/completions, the other is a chat-template batching crash; related area, but not the same bug." + "reason": "Different API surfaces: chat/completions request validation vs processor chat-template batching crash." }, { "left": "issue:36683", "right": "issue:43582", "accept": false, - "reason": "Unrelated errors: Gemma3Config missing vocab_size vs Apple Silicon TypeError in caching_allocator_warmup." + "reason": "Gemma3Config attribute error and Apple Silicon allocator TypeError are unrelated." }, { "left": "issue:43618", "right": "issue:43761", - "accept": false, - "reason": "Both are CLIP regressions, but one concerns attentions and the other hidden_states; too different to treat as the same concrete bug." + "accept": true, + "reason": "Both are CLIP v5 regressions in model outputs: one loses attentions, the other loses hidden_states. They plausibly come from the same forward/output assembly bug and could be fixed together." }, { "left": "issue:43012", "right": "issue:44683", "accept": false, - "reason": "Both are compile-related, but one is a warning and the other is a hard flex_attention failure on newer torch; different code paths." + "reason": "Precision warning during compile is unrelated to compiled flex_attention failure on torch >= 2.9." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but from different libraries and symbols; not the same underlying breakage." + "reason": "Both are ImportErrors, but for completely different symbols and packages." }, { "left": "issue:44038", "right": "issue:44230", "accept": false, - "reason": "Both mention Qwen3-VL/MoE, but one is model loading regression and the other is fp8 inference support; not mergeable as one fix." + "reason": "Related model family names, but different issues: general Qwen3-VL-Moe breakage vs FP8 inference support." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "Both involve chat/completions handling, but one is endpoint acceptance and the other is apply_chat_template with tool-call assistant messages." + "reason": "OpenAI-style chat endpoint request rejection and assistant tool-call chat-template crash are different code paths." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Both are load/save-ish failures, but one is Qwen3-VL-Moe loading and the other is a sorted state_dict load error; different triggers." + "reason": "Both concern model loading/saving, but the described failures are not the same concrete bug." }, { "left": "issue:44483", "right": "issue:45381", "accept": false, - "reason": "Different issues in different layers: request validation vs Qwen2.5-VL video vision_position_ids." + "reason": "Chat completion API validation and Qwen2.5-VL video position-id corruption are unrelated." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "ImportError from transformers.configuration_utils vs API calls happening despite local_files_only=True; unrelated." + "reason": "Importing layer_type_validation and ignoring local_files_only are unrelated bugs." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Smollm3 RoPE-layer count bug vs GLM-5 RoPE implementation discussion; same broad topic, but not the same specific defect." + "reason": "Both mention RoPE, but they refer to different models and different implementation problems." }, { "left": "issue:43901", "right": "issue:44655", "accept": false, - "reason": "Docs mismatch for TextClassificationPipeline vs inability to save Pipeline objects; documentation and serialization are distinct." + "reason": "Docs wording about return_all_scores is unrelated to pipeline serialization." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Gemma3Config missing vocab_size is unrelated to SAM 3 Video missing fpn_position_embeddings." - }, - { - "left": "issue:44220", - "right": "issue:44655", - "accept": false, - "reason": "Audio feature extraction bug vs pipeline save_pretrained failure; no shared code path." + "reason": "Different missing-attribute bugs on different model/config objects." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Quantized model saving failure and pipeline save_pretrained failure are both save-related, but they are different save flows and likely different fixes." + "reason": "Quantized model saving and pipeline saving are different save paths and symptoms." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Both concern checkpoint/loading, but one is load failure from state_dict ordering and the other is incorrect Qwen3.5 save_pretrained output." + "reason": "Both are checkpoint/load regressions, but there is not enough evidence they share the same model or code path." + }, + { + "left": "issue:44220", + "right": "issue:44655", + "accept": false, + "reason": "Audio feature extraction bug vs pipeline save bug are unrelated." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Nemotron Flash Attention 2 support vs GLM-5 GenerationConfig validation; unrelated bugs." + "reason": "Flash Attention support issue for Nemotron is unrelated to GLM-5 generation config validation." }, { "left": "issue:44263", "right": "issue:45440", "accept": false, - "reason": "Both are MoE-related, but they affect different model implementations (GLM MoE DSA indexer vs DeepseekV3MoE divergence)." - }, - { - "left": "issue:44077", - "right": "issue:44380", - "accept": false, - "reason": "PatchTSMixer post_init validation vs GPT-2 attention scaling under backend-specific attention; unrelated." + "reason": "Different MoE-related failures in different models and code paths." }, { "left": "issue:36683", "right": "issue:44077", "accept": false, - "reason": "Gemma3 config attribute error vs PatchTSMixer config validation change; different bugs." + "reason": "Gemma3 config attribute missing and patchtsmixer post_init policy are unrelated." }, { "left": "issue:43066", "right": "issue:45381", "accept": false, - "reason": "Tokenizer decoder type regression vs Qwen2.5-VL video position ids; different components." + "reason": "Tokenizer decoder type mismatch and Qwen2.5-VL video position-id bug are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44380", + "accept": false, + "reason": "Patchtsmixer post_init validation and GPT2 attention scaling are unrelated." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Both are Python compatibility complaints, but they target different Python versions and likely different root causes." + "reason": "Python version support problem at 5.1.0 is not the same as a separate Python 3.14 load failure." }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Different config attribute missing errors for different models; not the same underlying issue." + "reason": "Different missing config attributes for different models." }, { "left": "issue:36010", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they involve different packages and symbols, so they are not duplicates." - }, - { - "left": "issue:43064", - "right": "issue:44492", - "accept": false, - "reason": "Distributed optimizer-state bug vs a cache-strategy typo; no shared bug." - }, - { - "left": "issue:44188", - "right": "issue:44380", - "accept": false, - "reason": "Both mention attention/torch.compile, but one is kernel divergence from branching and the other is ignored scaling config; different defects." - }, - { - "left": "issue:43352", - "right": "issue:45593", - "accept": false, - "reason": "Nemotron Flash Attention support and D-FINE auxiliary-loss handling are unrelated." - }, - { - "left": "issue:43618", - "right": "issue:45216", - "accept": false, - "reason": "CLIP attentions regression vs Qwen3.5 checkpoint corruption; different models and code paths." - }, - { - "left": "issue:36683", - "right": "issue:44242", - "accept": false, - "reason": "Gemma3Config missing vocab_size vs missing load-balancing loss when output_router_logits=False; unrelated." + "reason": "Both are import failures from different modules, with different missing symbols." }, - { - "left": "issue:43010", - "right": "issue:45278", - "accept": false, - "reason": "Cache/layer update should use no_grad is unrelated to the broad import errors in 45278." - }, - { - "left": "issue:43352", - "right": "issue:43931", + { + "left": "issue:43064", + "right": "issue:44492", "accept": false, - "reason": "Different model families and failures: Flash Attention support vs Qwen3-VL weight-shape mismatch." + "reason": "FSDP optimizer-state corruption and a cache-strategy typo are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet and is appropriately cautious. The only accepted soft pair, issue:43618 vs issue:43761, is a reasonable conservative merge candidate because both are CLIP v5 output-regression reports affecting forward output fields. The other pair verdicts also read as appropriately negative and do not overclaim shared bugs." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -13192,38 +13063,35 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:28282", - "issue:30990", - "issue:31795", "issue:33453", "issue:36296", "issue:36683", "issue:41669", - "issue:42548", "issue:42890", "issue:43010", "issue:43064", - "issue:43065", - "issue:43232", + "issue:43352", "issue:43502", "issue:43575", "issue:43618", "issue:43723", "issue:43867", + "issue:43931", "issue:43976", "issue:43992", "issue:44062", "issue:44075", "issue:44077", "issue:44112", - "issue:44222", + "issue:44188", + "issue:44242", "issue:44246", - "issue:44351", "issue:44355", "issue:44360", - "issue:44387", + "issue:44380", "issue:44484", "issue:44556", "issue:44855", @@ -13231,7 +13099,8 @@ "issue:44938", "issue:44964", "issue:45071", - "issue:45125", + "issue:45216", + "issue:45278", "issue:45310", "issue:45341", "issue:45357", @@ -13241,9 +13110,16 @@ "issue:45478", "issue:45561", "issue:45588", + "issue:45593", "issue:45600" ], "soft_pairs": [ + "issue:44188|issue:44380", + "issue:43352|issue:45593", + "issue:43618|issue:45216", + "issue:36683|issue:44242", + "issue:43010|issue:45278", + "issue:43352|issue:43931", "issue:43992|issue:45310", "issue:41669|issue:44246", "issue:43867|issue:45375", @@ -13268,243 +13144,232 @@ "issue:43502|issue:44062", "issue:43723|issue:44556", "issue:42890|issue:45341", - "issue:43010|issue:44075", - "issue:44222|issue:45125", - "issue:43010|issue:43065", - "issue:44387|issue:44556", - "issue:42548|issue:44351", - "issue:43010|issue:43232", - "issue:45125|issue:45478", - "issue:45125|issue:45375", - "issue:30990|issue:31795" + "issue:43010|issue:44075" ], - "prepared_review_unit_hash": "96a4abaaa9d7821020b68f204114acf29c3d211f3e9ff3e5286f74df18628687", + "prepared_review_unit_hash": "f4e2b9a8861790d29d0332552e4b6cb6f41a6e83118b979017fad71eb566a11d", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37073, - "estimated_input_tokens": 9269, - "estimated_eval_tokens": 18794 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36065, + "estimated_input_tokens": 9017, + "estimated_eval_tokens": 18290 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly a loose family of Transformers loading/import/runtime regressions, but the proposed soft pairs usually point to different models, environments, or code paths rather than the same underlying bug. I would not merge any of the soft edges.", - "confidence": 0.58, - "canonical_issue_reason": "Issue 43723 is the best representative of the cluster\u2019s main theme: a concrete, user-facing loading regression in a core public API (`AutoTokenizer.from_pretrained`).", + "summary": "This cluster is mostly noisy and contains many unrelated issues. The only soft pair that looks like the same underlying problem is the import-time slowdown pair; the rest are different bugs or features in the same broad subsystem but not true duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:44246 is the broader user-facing report of slow `import transformers`; issue:41669 reads as a specific root-cause/optimization ticket (`import *` in models) that plausibly addresses that slowdown.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43723 is the strongest single exemplar of the broader import/load regression cluster because it is specific, practical, and centrally about model/tokenizer loading behavior.", + "best_issue_reason": "issue:44246 is the best anchor because it states the symptom clearly, is broader than the implementation-specific follow-up, and has the strongest duplicate relationship among the soft pairs.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44188", + "right": "issue:44380", + "accept": false, + "reason": "Both are attention-related, but one is about `torch.compile` branching causing kernel divergence and the other about GPT-2 attention scaling being ignored under SDPA/FlashAttention; different failure modes and fixes." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Flash Attention 2 support gating for Nemotron is unrelated to D-FINE auxiliary-loss behavior when denoising is off." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "`CLIPOutput.attentions` disappearing and Qwen3.5 save_pretrained checkpoint corruption are different code paths and different bug classes." + }, + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "Missing `vocab_size` on `Gemma3Config` and router-logits load-balancing-loss gating are unrelated model/config bugs." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "`@torch.no_grad` on cache/layer update and broad import errors after upgrading versions are not the same underlying issue." + }, + { + "left": "issue:43352", + "right": "issue:43931", + "accept": false, + "reason": "Flash Attention 2 unsupported vs Qwen3-VL weight-shape mismatch are distinct loading/runtime failures." + }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "Different models and symptoms: UMT5Encoder missing `embed_tokens.weight` vs Qwen3.5 MoE `from_pretrained` error." + "reason": "Missing `embed_tokens.weight` in UMT5Encoder and Qwen3.5 MoE `from_pretrained` errors are unrelated model-loading regressions." }, { "left": "issue:41669", "right": "issue:44246", - "accept": false, - "reason": "Both mention import cost, but one is a code cleanup/perf issue and the other is an intermittent slow-import report; not the same concrete bug." + "accept": true, + "reason": "Both point to import-time slowdown in `transformers`; 41669 provides a likely concrete root cause (`import *` in models), while 44246 is the general symptom report." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "State-dict sorting load failure and a missing config field in Qwen3.5-VL are unrelated failures." + "reason": "A sorted-state_dict load failure and a missing `deepstack_visual_indexes` config field are different problems." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "Gemma3 config attribute error and FSDP2/PEFT optimizer-state corruption are different layers of the stack." + "reason": "Gemma3 config missing `vocab_size` is unrelated to FSDP2/PEFT optimizer state corruption on nonzero ranks." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Compiled-file import errors and a TorchScript parser/IndentationError are different mechanisms." + "reason": "Generic compiled-file errors and a Python 3.13 `@torch.jit.script` parsing indentation issue are not the same bug." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "Test flakiness from missing `set_seed()` is unrelated to xdist file races in testing utils." + "reason": "Missing `set_seed()` in SamHQ integration tests and an xdist race on `captured_info.txt` are different test flakiness causes." }, { "left": "issue:44938", "right": "issue:45071", "accept": false, - "reason": "Python 3.14 import/load failure and `PretrainedConfig` type-checking breakage are different regressions." + "reason": "Python 3.14 import/load failure and `PretrainedConfig` type-checking regression are different compatibility issues." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Loading a sorted state dict and saving incorrect visual encoder keys are opposite directions of the serialization path." + "reason": "Sorted state_dict loading errors and Qwen3.5 visual-encoder save_pretrained corruption are different save/load paths." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Same broad model family, but the concrete failures differ: missing encoder weights vs generic Qwen3.5 MoE load error." + "reason": "UMT5Encoder missing weights and Qwen3.5 MoE `from_pretrained` errors are unrelated model-loading bugs." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "`@torch.no_grad` decoration on cache updates is unrelated to a Python-version compatibility issue." + "reason": "`torch.no_grad` decoration for cache updates has nothing to do with Python version support for Transformers 5.1.0." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "A missing ReLU in a DSA indexer is not the same bug as DeepSeekV3 implementation divergence." + "reason": "A ReLU missing from the DSA indexer and DeepSeekV3MoE parity issues are different algorithmic concerns." }, { "left": "issue:44938", "right": "issue:45381", "accept": false, - "reason": "Python 3.14 loading failure and Qwen2.5-VL video position IDs are unrelated." + "reason": "Python 3.14 load failures and Qwen2.5-VL video position-id bugs are unrelated." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIP attentions assignment and patchtsmixer `post_init` allowance are unrelated model-API issues." + "reason": "Missing CLIP attentions and `patchtsmixer` post_init policy are separate API/architecture issues." }, { "left": "issue:43867", "right": "issue:44484", "accept": false, - "reason": "Load failure on sorted state dicts is unrelated to the default `max_shard_size` question in `save_pretrained()`." + "reason": "Model loading with sorted state_dict and the 50GB shard-size question are unrelated save/load behaviors." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Both involve tensor parallelism, but one is a training bug and the other is an OOM during loading; not the same code path." + "reason": "Tensor-parallel training bug and TP OOM when loading a large Qwen model are both TP-related but not the same concrete failure." }, { "left": "issue:43618", "right": "issue:44857", "accept": false, - "reason": "Different model subsystems and failure modes: CLIP output fields vs LwDETR AMP/CUDA crash." + "reason": "Missing CLIP attentions and LwDETR float16 AMP crashes are different bugs in different models." }, { "left": "issue:43618", "right": "issue:45588", "accept": false, - "reason": "Attention output regression and flash-attention crash on `s_aux=None` are not the same issue." + "reason": "Missing CLIP attentions and a flash-attention crash on `s_aux=None` are unrelated failures." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "PyTorch-missing import error and tokenizer-loading regression are different problems." + "reason": "PyTorch-not-found import error and tokenizer-loading regression are unrelated setup/loading issues." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "Test seed flakiness is unrelated to a specific multimodal model load failure." + "reason": "SamHQ test flakiness and inability to load Phi-4 multimodal are unrelated." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "A flaky integration test and a stale-device-override CI failure are different test issues." + "reason": "Seed-related integration-test flakiness and GraniteSpeech stale device override CI failure are different test issues." }, { "left": "issue:42890", "right": "issue:45600", "accept": false, - "reason": "Missing `set_seed()` in tests is unrelated to removed `Sam3LiteText` configs breaking auto-mappings." + "reason": "SamHQ seeding flakiness and `auto_mappings.py` referencing removed configs are unrelated." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "`local_files_only=True` network leakage is unrelated to `AddedToken()` keyword duplication." + "reason": "Unexpected network access with `local_files_only=True` and `AddedToken(..., special=...)` argument collision are unrelated tokenizer/load issues." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Both concern loading, but one is tokenizer loading in v5 and the other is checkpoint reload compatibility after version upgrade; too broad to treat as the same bug." + "reason": "Tokenizer loading regression in v5 and checkpoint reload failure after upgrading versions are different failures." }, { "left": "issue:42890", "right": "issue:45341", "accept": false, - "reason": "A seed-related test failure is unrelated to a generic testing_utils bug." + "reason": "Seed-related test flakiness and a bug in `testing_utils.py` are not the same underlying problem." }, { "left": "issue:43010", "right": "issue:44075", "accept": false, - "reason": "`no_grad` decoration on cache/layer updates and SGD argument handling are unrelated." - }, - { - "left": "issue:44222", - "right": "issue:45125", - "accept": false, - "reason": "FP8 save-pretrained behavior and missing `_tp_plan` for tensor parallelism are different MoE issues." - }, - { - "left": "issue:43010", - "right": "issue:43065", - "accept": false, - "reason": "Cache update decoration and a dummy `Conv2d` in Sam3PixelDecoder are unrelated." - }, - { - "left": "issue:44387", - "right": "issue:44556", - "accept": false, - "reason": "Int4 CUDA memory growth and v4.57-to-v5 reload failure are different regressions." - }, - { - "left": "issue:42548", - "right": "issue:44351", - "accept": false, - "reason": "Both are import-name failures, but for different exported symbols and likely different root causes." - }, - { - "left": "issue:43010", - "right": "issue:43232", - "accept": false, - "reason": "`no_grad` decoration and generation kwargs syncing after `sync_gpus` are different code paths." - }, - { - "left": "issue:45125", - "right": "issue:45478", - "accept": false, - "reason": "Missing tensor-parallel plan and a generic Qwen3.5 MoE `from_pretrained` error are related in topic but not clearly the same concrete bug." - }, - { - "left": "issue:30990", - "right": "issue:31795", - "accept": false, - "reason": "Sentence Transformers loading stall and forward-doc confusion are unrelated." + "reason": "Cache update no-grad decoration and ignored SGD args are unrelated training/runtime bugs." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the soft-edge decisions are generally conservative. The only accepted pair (41669/44246) is plausibly the same import-slowdown problem, while the rejected pairs are supported by clearly different titles and bug classes." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -13519,31 +13384,31 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:30333", + "issue:30990", "issue:31795", "issue:33453", "issue:34689", "issue:36296", "issue:38617", "issue:41628", - "issue:41669", - "issue:42371", "issue:42548", "issue:42757", "issue:42831", "issue:43010", + "issue:43012", + "issue:43065", + "issue:43232", "issue:43299", "issue:43502", - "issue:43526", "issue:43576", "issue:43673", "issue:43824", "issue:43827", "issue:43874", "issue:43901", - "issue:44038", "issue:44222", "issue:44230", "issue:44261", @@ -13553,26 +13418,33 @@ "issue:44355", "issue:44360", "issue:44368", - "issue:44393", + "issue:44387", "issue:44485", "issue:44492", "issue:44556", - "issue:44623", "issue:44655", "issue:44749", "issue:44908", "issue:44961", "issue:45003", "issue:45071", - "issue:45092", "issue:45125", - "issue:45230", + "issue:45375", "issue:45468", + "issue:45478", "issue:45542" ], "soft_pairs": [ + "issue:44222|issue:45125", + "issue:43010|issue:43065", + "issue:44387|issue:44556", + "issue:42548|issue:44351", + "issue:45125|issue:45478", + "issue:43010|issue:43232", + "issue:45125|issue:45375", "issue:44355|issue:44749", "issue:43502|issue:44485", + "issue:30990|issue:31795", "issue:30333|issue:33453", "issue:44315|issue:44360", "issue:43010|issue:44368", @@ -13580,8 +13452,8 @@ "issue:43502|issue:44961", "issue:43299|issue:44556", "issue:44355|issue:44655", - "issue:30333|issue:36296", "issue:42831|issue:43874", + "issue:30333|issue:36296", "issue:44222|issue:44230", "issue:44355|issue:45071", "issue:33453|issue:34689", @@ -13599,287 +13471,286 @@ "issue:41628|issue:42548", "issue:44485|issue:44961", "issue:36296|issue:43901", - "issue:36296|issue:43526", + "issue:36296|issue:43012", "issue:41628|issue:44355", - "issue:44273|issue:45003", "issue:44261|issue:44360", - "issue:44273|issue:45542", - "issue:44623|issue:45230", - "issue:43576|issue:44556", - "issue:41669|issue:43673", - "issue:44038|issue:45125", - "issue:44393|issue:45071", - "issue:44556|issue:45092", - "issue:42371|issue:44492" + "issue:44273|issue:45003", + "issue:44273|issue:45542" ], - "prepared_review_unit_hash": "43c672f933ba029ccef1e3530b471ef999b68937be56f43d9afaeebfc7d30a64", + "prepared_review_unit_hash": "d2cb30de7a95e77cdd642607d3dceacfdc42957774af529ccea9462eecf2a20a", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 40, - "serialized_chars": 39050, - "estimated_input_tokens": 9763, - "estimated_eval_tokens": 19782 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 41, + "serialized_chars": 38808, + "estimated_input_tokens": 9702, + "estimated_eval_tokens": 19660 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All soft pairs look like false positives rather than true duplicates; the set is mostly a grab-bag of unrelated Transformers regressions and docs/import issues. No PRs were provided.", - "confidence": 0.74, - "canonical_issue_reason": "issue:43576 is the most central concrete bug in the set, with the most soft similarity links and a clearly user-facing v5 regression.", + "summary": "Most soft pairs are only loosely similar by topic (imports, loading, v5 regressions, FP8/MoE) but describe different concrete bugs or code paths, so they should not be deduplicated.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43576 is the strongest hub among the issue candidates: it is a concrete, user-visible v5 regression and has the most nearby import/CLI-related soft matches, even though the matches are still distinct bugs.", "canonical_pr_reason": null, - "best_issue_reason": "Best overall issue anchor: specific, concrete, and the most connected candidate among the listed issues.", + "best_issue_reason": "issue:43576 is the best single issue representative here because it has clear repro impact and the broadest duplicate-like neighborhood in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "Both mention FP8/MoE, but one is save_pretrained and the other tensor-parallel model loading." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "No shared code path: cache/no_grad decoration vs a dummy Conv2d in Sam3PixelDecoder." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Different failures: int4 memory growth vs checkpoint reload compatibility after version upgrade." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different export regressions." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Same model family, but tensor-parallel support is not the same as a from_pretrained error." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Different generation internals: no_grad cache updates vs sync_gpus/_update_model_kwargs behavior." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Different model components: tensor parallel plan vs a missing config field silently dropped by strict parsing." + }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Completely different bug: compiled-file errors vs data-processing slowdown after upgrade." + "reason": "Unrelated issues: compiled Python file errors vs a post-upgrade data filtering slowdown." }, { "left": "issue:43502", "right": "issue:44485", "accept": false, - "reason": "Unrelated: local_files_only network leak vs GLM-5 RoPE implementation." + "reason": "Local-files-only network leakage is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Loading hang in Sentence Transformers is not the same as documentation confusion for model.forward arguments." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "Different subsystems: MLflow training-job reporting vs tokenizer loading regression." + "reason": "MLflow reporting on failed jobs is unrelated to tokenizer loading regression." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger kernel/model_init behavior is unrelated to DSA indexer ReLU." + "reason": "Liger Kernel application in model_init and a DSA indexer ReLU bug are different subsystems." }, { "left": "issue:43010", "right": "issue:44368", "accept": false, - "reason": "Cache no_grad decoration and tie_word_embeddings warning are separate issues." + "reason": "Cache update no_grad behavior is unrelated to a tie_word_embeddings warning." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "local_files_only API leakage is unrelated to FP8 save_pretrained behavior." + "reason": "Different bugs: unwanted API requests vs FP8 save_pretrained behavior." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "local_files_only networking bug is unrelated to 'racoon'." + "reason": "No meaningful overlap beyond being generic issues." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Qwen3VL loading regression and v4.57->v5 reload incompatibility are different model-loading bugs." + "reason": "Qwen3VL MoE loading failure and checkpoint reload breakage are different model/version regressions." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "Compiled-file errors and pipeline save_pretrained are different code paths." + "reason": "Compiled-file runtime errors and Pipeline save_pretrained failure are not the same bug." }, { - "left": "issue:30333", - "right": "issue:36296", + "left": "issue:42831", + "right": "issue:43874", "accept": false, - "reason": "Training-job MLflow reporting and tensor-parallel training bug differ." + "reason": "FP8 accuracy drift and missing image-patch API in GLM46V are separate issues." }, { - "left": "issue:42831", - "right": "issue:43874", + "left": "issue:30333", + "right": "issue:36296", "accept": false, - "reason": "FineGrainedFP8 accuracy and GLM46V patch-count AttributeError are unrelated." + "reason": "MLflow reporting on failure is unrelated to tensor-parallel training bugs." }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "FP8 save_pretrained and FP8 inference support are distinct issues." + "reason": "Both are FP8/MoE, but save_pretrained support and inference support are different change requests." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled-file errors and PretrainedConfig type-checking regression are different." + "reason": "Compiled-file errors are unrelated to PretrainedConfig type-checking regressions." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer-loading regression and Llama 3.2 Vision loading break are not the same defect." + "reason": "Tokenizer loading regression and Llama 3.2 Vision model loading breakage are different failure modes." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Hugging Face hub import and transformers env command are unrelated." + "reason": "Missing is_offline_mode import and broken `transformers env` may both be v5-related, but they are distinct bugs." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs still referencing pipeline() is unrelated to 'racoon'." + "reason": "Docs still mentioning pipeline() is unrelated to the unrelated racoon issue." }, { "left": "issue:33453", "right": "issue:36296", "accept": false, - "reason": "Tokenizer-loading regression and tensor-parallel training bug are different failures." + "reason": "Tokenizer loading regression and tensor-parallel training bug are not the same underlying defect." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "CLI env breakage and type-checking regression are unrelated." + "reason": "A broken CLI env command and config type-checking are different v5 regressions." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "CLI env breakage and Qwen2_5_VL import failure are different issues." + "reason": "Broken env command and missing Qwen2_5_VL export are different import/export issues." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Chunked-prefill cache bug and cache strategy typo are not the same bug." + "reason": "Chunked prefill cache handling and a cache-strategy typo are not the same bug." }, { "left": "issue:38617", "right": "issue:43576", "accept": false, - "reason": "layer_type_validation import error and env command breakage differ." + "reason": "Missing layer_type_validation export and a broken env CLI command are different import regressions." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "'racoon' and Gemma4 audio positional encoding are unrelated." + "reason": "No meaningful overlap; racoon is unrelated to Gemma-4 audio positional encoding." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Docs confusion and import error are different issues." + "reason": "Documentation confusion and an import error for layer_type_validation are unrelated." }, { "left": "issue:43576", "right": "issue:44351", "accept": false, - "reason": "env command breakage and HybridCache import failure are not the same." + "reason": "Broken env CLI and missing HybridCache import are distinct missing-symbol bugs." }, { "left": "issue:44908", "right": "issue:44961", "accept": false, - "reason": "Scheduler kwargs bug and 'racoon' are unrelated." + "reason": "Scheduler kwargs handling has nothing to do with the racoon issue." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "AutoImageProcessor import and PreTrainedModel import are different missing symbols." + "reason": "Both are import errors, but for different classes and likely different export paths." }, { "left": "issue:44485", "right": "issue:44961", "accept": false, - "reason": "RoPE implementation and 'racoon' are unrelated." + "reason": "GLM-5 RoPE implementation and racoon are unrelated." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor-parallel training bug and text-classification docs are unrelated." + "reason": "Tensor-parallel training and TextClassificationPipeline docs are unrelated." }, { "left": "issue:36296", - "right": "issue:43526", + "right": "issue:43012", "accept": false, - "reason": "Tensor-parallel training bug and BeitImageProcessorFast label reduction are unrelated." + "reason": "Tensor-parallel training bugs do not match bfloat16 compile warnings." }, { "left": "issue:41628", "right": "issue:44355", "accept": false, - "reason": "AutoImageProcessor import failure and compiled-file errors are unrelated." - }, - { - "left": "issue:44273", - "right": "issue:45003", - "accept": false, - "reason": "Lazy loading and sys.modules access may be related, but the titles indicate different bugs." + "reason": "AutoImageProcessor import failure and compiled Python file errors are unrelated." }, { "left": "issue:44261", "right": "issue:44360", "accept": false, - "reason": "MLA layernorm epsilon bug and DSA indexer ReLU are unrelated." + "reason": "MLA layer norm epsilon precision issue is not the same as the DSA indexer ReLU bug." }, { "left": "issue:44273", - "right": "issue:45542", - "accept": false, - "reason": "Lazy loading bug and TensorBoard-without-TF backend error are different." - }, - { - "left": "issue:44623", - "right": "issue:45230", - "accept": false, - "reason": "processor.save_pretrained missing files and generic bug report are not the same." - }, - { - "left": "issue:43576", - "right": "issue:44556", - "accept": false, - "reason": "env command breakage and checkpoint reload incompatibility are distinct regressions." - }, - { - "left": "issue:41669", - "right": "issue:43673", - "accept": false, - "reason": "Import-star cleanup and chunked-prefill cache bug are unrelated." - }, - { - "left": "issue:44038", - "right": "issue:45125", - "accept": false, - "reason": "Qwen3-VL-Moe bug and missing _tp_plan for Qwen3_5Moe are different." - }, - { - "left": "issue:44393", - "right": "issue:45071", - "accept": false, - "reason": "Qwen3-VL bbox output issue and PretrainedConfig type checking are unrelated." - }, - { - "left": "issue:44556", - "right": "issue:45092", + "right": "issue:45003", "accept": false, - "reason": "v4.57 checkpoint reload and InternVL2 meta-init incompatibility are different model-loading issues." + "reason": "Lazy loading failure and sys.modules access concerns may be adjacent, but not the same concrete bug." }, { - "left": "issue:42371", - "right": "issue:44492", + "left": "issue:44273", + "right": "issue:45542", "accept": false, - "reason": "TF32 settings guidance and cache strategy typo are unrelated." + "reason": "Lazy loading and a TensorFlow backend import error are unrelated." } ] }, @@ -13901,7 +13772,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:31795", "issue:33453", @@ -13910,18 +13781,18 @@ "issue:41669", "issue:42371", "issue:42890", - "issue:42913", - "issue:43066", "issue:43352", "issue:43408", "issue:43502", "issue:43519", "issue:43576", "issue:43606", + "issue:43673", "issue:43704", "issue:43723", "issue:43827", "issue:43976", + "issue:44038", "issue:44077", "issue:44246", "issue:44261", @@ -13935,6 +13806,7 @@ "issue:44393", "issue:44485", "issue:44492", + "issue:44556", "issue:44623", "issue:44655", "issue:44704", @@ -13942,17 +13814,23 @@ "issue:44861", "issue:44908", "issue:44945", - "issue:44998", "issue:45071", "issue:45092", + "issue:45125", "issue:45230", - "issue:45290", "issue:45310", "issue:45341", "issue:45468", "issue:45478" ], "soft_pairs": [ + "issue:44623|issue:45230", + "issue:43576|issue:44556", + "issue:41669|issue:43673", + "issue:44038|issue:45125", + "issue:44393|issue:45071", + "issue:44556|issue:45092", + "issue:42371|issue:44492", "issue:42890|issue:44908", "issue:44704|issue:44908", "issue:44246|issue:44273", @@ -13974,13 +13852,13 @@ "issue:44077|issue:44861", "issue:41669|issue:44908", "issue:43723|issue:44393", - "issue:44655|issue:44861", "issue:43576|issue:44273", + "issue:44655|issue:44861", "issue:44360|issue:45468", "issue:44273|issue:44908", "issue:45230|issue:45310", - "issue:41669|issue:43704", "issue:43519|issue:44485", + "issue:41669|issue:43704", "issue:43519|issue:44908", "issue:31795|issue:34689", "issue:31795|issue:37428", @@ -13994,311 +13872,330 @@ "issue:44263|issue:45468", "issue:44261|issue:45468", "issue:43606|issue:45468", - "issue:43704|issue:45341", - "issue:42913|issue:43066", - "issue:44908|issue:44998", - "issue:44861|issue:45290", - "issue:44998|issue:45341", - "issue:44492|issue:44998" + "issue:43704|issue:45341" ], - "prepared_review_unit_hash": "a6ac5773ca405ce4a194331d723b621c9d193eef421d4eb9fcfd824cfc4123a7", + "prepared_review_unit_hash": "0d59a3380e2ef8ba456de821000717d91f42a8ad9b61bf84d3c9571a5b697ea4", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 47, - "serialized_chars": 41144, - "estimated_input_tokens": 10286, - "estimated_eval_tokens": 20828 + "soft_pair_count": 49, + "serialized_chars": 41786, + "estimated_input_tokens": 10447, + "estimated_eval_tokens": 21150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: most items are unrelated issues spanning tokenizer loading, model-loading regressions, import performance, docs, and assorted model-specific bugs. I did not find any soft pair that clearly represents the same underlying bug/change, so all soft edges are rejected.", - "confidence": 0.94, - "canonical_issue_reason": "No single true canonical duplicate exists in this set; issue 44273 is the broadest ongoing loading-related report, but the cluster is not actually one bug.", + "summary": "This cluster is mostly a loose set of unrelated Transformers issues with a weak subtheme around v5 loading/tokenizer/config regressions. The soft edges are driven by broad lexical similarity, but the underlying bugs and code paths are usually different.", + "confidence": 0.69, + "canonical_issue_reason": "Issue 44556 is the clearest concrete anchor in the loading/regression subtheme: it describes a specific checkpoint reload failure after upgrading to v5.2/v5.3, which is more representative than generic or overly narrow reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue 44273 is the best representative placeholder because it is broad, open, and closest to the few loading/import-related themes in the set.", + "best_issue_reason": "44556 is the most actionable and broadly representative issue among the candidates, since it is a concrete v5 compatibility/loading failure rather than a generic placeholder or a narrowly model-specific report.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Processor save_pretrained missing-files bug vs a generic placeholder report; not the same issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Both are v5 regressions, but one is the env command and the other is checkpoint reload after upgrade; different failures." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-time file-generation/performance issue vs chunked_prefill cache regression; unrelated code paths." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Same model family, but Qwen3-VL-Moe breakage and missing _tp_plan are different bugs." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "2D bbox output quality bug vs PretrainedConfig type-checking regression; not the same defect." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both concern loading old checkpoints in v5, but the specific incompatibility and failure mode differ." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 settings warning vs cache-strategy typo; unrelated." + }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Different problems: SAM-HQ test seed flakiness vs scheduler kwarg handling." + "reason": "Flaky tests due to missing seeds vs scheduler kwargs propagation; different subsystems." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "Processor cached_file kwarg forwarding is unrelated to inverse_sqrt scheduler kwargs." + "reason": "AutoProcessor kwargs forwarding vs inverse_sqrt scheduler kwargs handling; unrelated." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Both touch import/lazy loading performance, but the reported bugs are not concretely the same." + "reason": "Import slowdowns and lazy-loading malfunction are related only loosely; not clearly the same bug." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model-type warning vs Liger Kernel/model_init behavior; unrelated code paths." + "reason": "Model type warning vs Liger Kernel not applied with model_init; different behaviors." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version compatibility issue vs Qwen3-VL bbox output bug." + "reason": "Python-version compatibility regression vs Qwen3-VL output bug; unrelated." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention support error vs Qwen3-VL output/formatting issue." + "reason": "Flash Attention support limitation vs Qwen3-VL bounding-box issue; not the same." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "Broken env CLI in v5 is unrelated to scheduler kwargs." + "reason": "Env command regression vs scheduler kwargs bug; no shared code-path bug." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler behavior and config type-checking are different bugs." + "reason": "Scheduler kwargs bug vs PretrainedConfig typing regression; different defects." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer save_pretrained mismatch is not the same as the generic bug report." + "reason": "Tokenizer class mismatch vs generic placeholder bug report; not enough overlap." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "position_ids buffer access bug is unrelated to the generic report." + "reason": "Position_ids buffer read error vs placeholder report; unrelated." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-star performance cleanup vs missing HybridCache import; different symptoms." + "reason": "Import * performance issue vs HybridCache import failure; different problems." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import performance issue vs a testing_utils bug." + "reason": "Import-time slowdown vs a bug in testing_utils; unrelated." }, { "left": "issue:44492", "right": "issue:44945", "accept": false, - "reason": "Cache-strategy typo and pipeline-parallelism output error are unrelated." + "reason": "Cache-strategy typo vs incorrect LLM output under pipeline parallelism; different issues." }, { "left": "issue:41669", "right": "issue:44492", "accept": false, - "reason": "Import-star cleanup is not the same as a cache strategy typo." + "reason": "Import * performance regression vs cache-strategy typo; not the same bug." }, { "left": "issue:44393", "right": "issue:45478", "accept": false, - "reason": "Different Qwen versions and different failure modes." + "reason": "Different Qwen3.5/Qwen3-VL failures and different symptoms; not duplicates." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "model_init/Liger Kernel behavior does not match flash_attention_3 training degeneration." + "reason": "Liger Kernel application bug vs flash_attention_3 training degeneration; distinct code paths." }, { "left": "issue:43827", "right": "issue:44908", "accept": false, - "reason": "Docs referencing pipeline() are unrelated to scheduler kwargs." + "reason": "Docs still referencing pipeline() vs scheduler kwargs handling; unrelated." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion vs tokenizer loading regression; not the same bug." + "reason": "Documentation confusion in model.forward vs tokenizer loading regression; different issues." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "post_init acceptance rules vs tied-weight-key AttributeError are different issues." + "reason": "Optional post_init policy vs _get_tied_weight_keys AttributeError; not the same bug." }, { "left": "issue:41669", "right": "issue:44908", "accept": false, - "reason": "Import-time slowdown and scheduler kwargs are unrelated." + "reason": "Import performance issue vs scheduler kwargs bug; unrelated." }, { "left": "issue:43723", "right": "issue:44393", "accept": false, - "reason": "Tokenizer loading in v5 vs Qwen3-VL bounding-box output bug." + "reason": "Tokenizer loading failure vs Qwen3-VL bounding-box bug; different subsystems." }, { - "left": "issue:44655", - "right": "issue:44861", + "left": "issue:43576", + "right": "issue:44273", "accept": false, - "reason": "save_pretrained pipeline support and tied-weight-key crash are different code paths." + "reason": "Env command regression vs lazy-loading issue; not the same underlying failure." }, { - "left": "issue:43576", - "right": "issue:44273", + "left": "issue:44655", + "right": "issue:44861", "accept": false, - "reason": "Broken CLI command and lazy loading bug are not the same." + "reason": "Pipeline save_pretrained problem vs tied-weights AttributeError; different code paths." }, { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "DSA indexer activation bug vs Gemma-4 audio positional encoding issue." + "reason": "DSA indexer math complaint vs Gemma-4 audio positional-encoding bug; unrelated." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy loading failure and inverse_sqrt scheduler kwargs are unrelated." + "reason": "Lazy loading malfunction vs scheduler kwargs bug; different issues." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "Generic bug report vs specific Qwen3.5 MoE from_pretrained error." + "reason": "Generic bug report vs specific Qwen3.5 Moe from_pretrained error; no concrete duplicate evidence." }, { - "left": "issue:41669", - "right": "issue:43704", + "left": "issue:43519", + "right": "issue:44485", "accept": false, - "reason": "Import-star performance issue vs VRAM leak in dataloader threads." + "reason": "Qwen3VL timestamp calculation vs GLM-5 RoPE implementation; different model/math paths." }, { - "left": "issue:43519", - "right": "issue:44485", + "left": "issue:41669", + "right": "issue:43704", "accept": false, - "reason": "Timestamp calculation in Qwen3VL processor vs GLM-5 RoPE implementation." + "reason": "Import * performance regression vs VRAM leak in dataloader threads; unrelated." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor timestamp math is unrelated to scheduler kwargs." + "reason": "Processor timestamp math vs scheduler kwargs bug; not the same issue." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Model.forward docs confusion vs Llama model loading breakage." + "reason": "Documentation issue vs model loading regression; different bug classes." }, { "left": "issue:31795", "right": "issue:37428", "accept": false, - "reason": "Documentation issue vs flash-attention import error." + "reason": "Docs confusion vs flash-attention import error; unrelated." }, { "left": "issue:44315", "right": "issue:45092", "accept": false, - "reason": "Liger Kernel model_init behavior is not the same as remote-code/meta-init incompatibility." + "reason": "Liger Kernel/model_init issue vs legacy checkpoint meta-init incompatibility; different failures." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak in threads vs RoPE implementation bug." + "reason": "VRAM leak in dataloader threads vs GLM-5 RoPE implementation; unrelated." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "GLM-5 RoPE implementation vs Gemma-4 audio positional encoding; different models and bugs." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 API guidance and VRAM leak are unrelated." + "reason": "TF32 behavior warning vs VRAM leak; unrelated." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "local_files_only network requests vs Qwen3VL timestamp calculation." + "reason": "API requests despite local_files_only vs Qwen3VL timestamp bug; different subsystems." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp math vs processor save_pretrained missing files; not the same issue." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Timestamp bug and VRAM leak are different failures." + "reason": "Qwen3VL timestamp bug vs multi-thread VRAM leak; unrelated." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "torch.split return values bug vs Gemma-4 audio positional encoding." + "reason": "torch.split return-value issue vs Gemma-4 audio positional encoding bug; different defects." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "MLA layernorm epsilon precision issue vs Gemma-4 audio positional encoding." + "reason": "MLA q_a_layernorm precision/config issue vs Gemma-4 audio positional encoding; unrelated." }, { "left": "issue:43606", "right": "issue:45468", "accept": false, - "reason": "CPU offload device mismatch vs Gemma-4 audio positional encoding." + "reason": "CPU offload device-mismatch CI bug vs Gemma-4 audio positional encoding; different code paths." }, { "left": "issue:43704", "right": "issue:45341", "accept": false, - "reason": "VRAM leak and testing_utils bug are unrelated." - }, - { - "left": "issue:42913", - "right": "issue:43066", - "accept": false, - "reason": "Both are tokenizer-v5 related, but the concrete failures differ and are not clearly the same bug." - }, - { - "left": "issue:44908", - "right": "issue:44998", - "accept": false, - "reason": "Scheduler kwargs bug and the unrelated placeholder issue have no overlap." - }, - { - "left": "issue:44861", - "right": "issue:45290", - "accept": false, - "reason": "Tied-weight-key AttributeError and chat-template tool-call crash are different paths." - }, - { - "left": "issue:44998", - "right": "issue:45341", - "accept": false, - "reason": "Unemployment placeholder issue is unrelated to testing_utils." - }, - { - "left": "issue:44492", - "right": "issue:44998", - "accept": false, - "reason": "Cache-strategy typo and unrelated placeholder issue are not duplicates." + "reason": "VRAM leak in dataloader threads vs testing_utils bug; unrelated." } ] }, @@ -14320,7 +14217,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-380", + "cluster_id": "cluster-15354-381", "nodes": [ "issue:30333", "issue:31515", @@ -14328,18 +14225,29 @@ "issue:33453", "issue:36296", "issue:38617", + "issue:42913", + "issue:43066", "issue:43298", "issue:43525", "issue:43572", "issue:43618", "issue:44230", + "issue:44492", "issue:44556", "issue:44623", "issue:44861", + "issue:44908", "issue:44998", - "issue:45310" + "issue:45290", + "issue:45310", + "issue:45341" ], "soft_pairs": [ + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44998|issue:45341", + "issue:44861|issue:45290", + "issue:44492|issue:44998", "issue:43618|issue:44998", "issue:44556|issue:44998", "issue:43525|issue:44861", @@ -14353,106 +14261,136 @@ "issue:31515|issue:33453", "issue:43298|issue:44230" ], - "prepared_review_unit_hash": "2d4500210bd28743176ced9eb3e4b6a90a7cd3088a796cb20024674e50eb3226", + "prepared_review_unit_hash": "69926dc5a0c0196e2e3ca9c4aa37810c3a1b0dd39e8a144a55cfba67930f1085", "component_budget": { - "node_count": 380, - "item_count": 380, - "soft_pair_count": 1423, - "serialized_chars": 733782, - "estimated_input_tokens": 183446, - "estimated_eval_tokens": 367148 + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 }, "budget": { - "node_count": 16, - "item_count": 16, - "soft_pair_count": 12, - "serialized_chars": 12581, - "estimated_input_tokens": 3146, - "estimated_eval_tokens": 6548 + "node_count": 22, + "item_count": 22, + "soft_pair_count": 17, + "serialized_chars": 17392, + "estimated_input_tokens": 4348, + "estimated_eval_tokens": 8952 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft edges mostly reflect generic transformer loading/config similarity, not the same bug. No PRs are present. Issue 31515 is the closest hub because it is the most broadly connected loading/checkpoint-related report, but it is not a true duplicate canonical for the whole set.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 31515 is the broadest and most connected item in the set, centered on from_pretrained/checkpoint loading behavior, so it is the closest representative among these mostly unrelated issues.", + "summary": "These items are mostly unrelated single-issue reports rather than duplicates. The soft pairs are mostly broad thematic matches (e.g. tokenizer/v5 regressions, loading issues) but not the same concrete bug or code path, so none should be merged.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 44556 is the clearest, most actionable regression report in the set: it describes a specific checkpoint reload failure after a v4.57\u2192v5.x upgrade. It is more concrete and triage-friendly than the other broad or ambiguous reports.", "canonical_pr_reason": null, - "best_issue_reason": "31515 is the best issue candidate only as a weak hub for the loading/checkpoint regressions; however, the cluster overall is too mixed to treat it as a real duplicate family.", + "best_issue_reason": "Issue 44556 is the strongest representative issue because it is specific, reproducible, and describes a likely regression with a clear upgrade path and impact.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both are tokenizer-related v5 complaints, but one is a broad behavior difference and the other is a specific wrong decoder-type report; they do not describe the same concrete defect." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Inverse-sqrt scheduler kwarg handling and an unrelated unemployment/off-topic issue are clearly different." + }, + { + "left": "issue:44998", + "right": "issue:45341", + "accept": false, + "reason": "Completely unrelated: one is off-topic, the other is a testing_utils bug." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Different code paths: tied-weights key extraction crash versus chat template handling for tool-call assistant messages." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "A typo in cache strategies is unrelated to the off-topic unemployment issue." + }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "Unrelated: CLIP attentions regression vs an unrelated miscellaneous issue; no shared bug or code path." + "reason": "CLIPOutput attentions regression and an off-topic issue are unrelated." }, { "left": "issue:44556", "right": "issue:44998", "accept": false, - "reason": "Versioned checkpoint reload failure is unrelated to the nonsensical 'Unemployment' issue." + "reason": "Checkpoint reload regression after version upgrade is unrelated to the off-topic unemployment issue." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Missing pad_token_id on Llama4Config is a config attribute issue; _get_tied_weight_keys crash is a separate helper bug." + "reason": "Missing pad_token_id in Llama4Config is a different bug from the tied-weights AttributeError." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "StableLmConfig pad_token_idx regression and tied-weights list handling are different failures in different code paths." + "reason": "Missing pad_token_idx in StableLmConfig is a separate config regression, not the tied-weights crash." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Checkpoint loading slowness is not the same as an ImportError for layer_type_validation." + "reason": "Slow checkpoint loading and an ImportError for layer_type_validation are unrelated problems." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 inference support for Qwen models is unrelated to processor.save_pretrained missing files." + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5 MoE is not the same as missing files from processor.save_pretrained." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "from_pretrained checkpoint performance and tensor-parallel training bug are distinct problems." + "reason": "Checkpoint loading performance and a tensor-parallel training bug are different issues." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLFlow reporting on failed training jobs is unrelated to checkpoint loading slowdown." + "reason": "MLflow reporting on failed training jobs is unrelated to checkpoint loading speed." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Checkpoint loading regression and documentation confusion about input_ids/past_key_values are not the same bug." + "reason": "Checkpoint loading performance and confusing model.forward docs are not the same underlying bug." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "Qwen3.5 MoE FP8 support and a from_pretrained error after v5.4 are not clearly the same concrete defect." + "reason": "Both mention Qwen3-family models, but one is an FP8 support request and the other is a concrete from_pretrained error; too different to merge as duplicates." }, { "left": "issue:31515", "right": "issue:33453", "accept": false, - "reason": "Both mention loading, but tokenizer loading regression and slow checkpoint loading are different symptoms and likely different code paths." + "reason": "A slow from_pretrained checkpoint load and a tokenizer-loading regression are not the same defect." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Continuous batching audio-model support and Qwen FP8 inference support are separate feature/bug areas." + "reason": "Continuous batching lacking audio-model support is a different capability gap from FP8 Qwen3-VL/Qwen3.5 MoE support." } ] }, @@ -14482,12 +14420,12 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "dd70c2886ee85a370cb5808444c1e2cc5382568bb9c929148a51cd56306a9972", + "prepared_review_unit_hash": "7296b965239bcf252d33470058bf50138ae7b73dd6f6b05d4c9e923658578584", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1590, + "serialized_chars": 1589, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -14495,7 +14433,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1590, + "serialized_chars": 1589, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -14504,18 +14442,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two distinct feature-request issues in the same broad model-architecture support area, but they are not duplicates: one asks for Flash Attention 2 support, the other for GGUF support. The soft similarity is too weak to merge them.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 26350 is the stronger anchor for the cluster because it has much higher discussion activity and is the more established request, but it does not cover the separate GGUF request in 33260.", + "summary": "These two issues are about different community-contribution feature requests: Flash Attention 2 support for more architectures vs. GGUF support for more architectures. They are not duplicates and should not be merged.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "26350 is the best representative issue only by activity and prominence; it is not a substitute for 33260 because the underlying feature requests are different.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Reject: Flash Attention 2 support and GGUF support are different architectural features with different implementation paths; they are related only at a very broad subsystem level." + "reason": "Different underlying feature requests with different targets (Flash Attention 2 vs GGUF). They do not describe the same bug or change, so this is not a duplicate pair." } ] }, @@ -14552,70 +14490,70 @@ "issue:43834|issue:44365", "issue:44365|issue:44370" ], - "prepared_review_unit_hash": "6bb35bf04769b438ae5f78efaf8c437da1cc911ca40cac128b0a9e8a843ae5f9", + "prepared_review_unit_hash": "03f7a243c5ca3148b4ff924254d232dc6f8cfaf1b0e702cb0914d602ef327482", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4453, - "estimated_input_tokens": 1114, - "estimated_eval_tokens": 2484 + "serialized_chars": 4452, + "estimated_input_tokens": 1113, + "estimated_eval_tokens": 2482 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4453, - "estimated_input_tokens": 1114, - "estimated_eval_tokens": 2484 + "serialized_chars": 4452, + "estimated_input_tokens": 1113, + "estimated_eval_tokens": 2482 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are repeated docs-translation tracking issues using the same template, but they appear to be separate language-specific requests rather than duplicates of the same underlying task. No PRs are present.", - "confidence": 0.92, - "canonical_issue_reason": "issue:27088 is the only concrete, language-specific report (Turkish) rather than a generic placeholder template, so it is the most defensible anchor if one must be chosen.", + "summary": "This cluster is dominated by generic i18n translation-template issues, with one concrete Turkish docs translation issue. The concrete Turkish issue is the best canonical issue; the placeholder/template issues are too underspecified to treat as exact duplicates of each other or of the Turkish one.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 27088 is the only concrete, language-specific report ('Turkish') rather than a placeholder template, so it best represents the real actionable issue in the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:27088 is the most specific and actionable issue in the set; the others are generic template variants without a concrete language named in the title.", + "best_issue_reason": "Issue 27088 is the strongest issue to keep as the representative: it names a specific language, is not a generic template, and is the most actionable artifact here.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Both are docs-translation requests, but there is no evidence they target the same language or the same underlying task; they look like separate localization issues." + "reason": "Different concrete language targets: one is Turkish, the other is a generic placeholder. Similar topic, but not the same underlying issue." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Same template, but 27088 is specifically Turkish while 44365 is just a generic translation placeholder; not the same concrete change." + "reason": "The Turkish-specific report is not the same as a generic placeholder translation issue; no evidence they refer to the same language or same task." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "These appear to be different language translation requests, not duplicates of one another." + "reason": "One issue is explicitly Turkish while the other is only a template stub. Too little evidence that they describe the same change or bug." }, { "left": "issue:43834", "right": "issue:44370", "accept": false, - "reason": "High textual similarity comes from the shared template, but there is no sign they are the same language or same translation effort." + "reason": "Both are template-like translation issues, but there is no concrete language or scope match to prove they are the same underlying request." }, { "left": "issue:43834", "right": "issue:44365", "accept": false, - "reason": "Template duplicates only; the issues are not clearly the same underlying language-specific request." + "reason": "Near-identical template text alone is not enough; these could be separate language translation requests, so do not merge." }, { "left": "issue:44365", "right": "issue:44370", "accept": false, - "reason": "Both are generic translation placeholders, but nothing shows they refer to the same language or same work item." + "reason": "Same generic template wording, but no concrete evidence that they target the same language or identical work item." } ] }, @@ -14684,12 +14622,12 @@ "issue:43837|issue:45027", "issue:43837|issue:45177" ], - "prepared_review_unit_hash": "51bcff846754e7c55b7b9100a4978f132756e0708f45cb86af0767eafc3649fb", + "prepared_review_unit_hash": "a63d57906e1cae48af1797b2aed0699a2adf9e7e258d981d62645c813b5d635b", "component_budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 28, - "serialized_chars": 17646, + "serialized_chars": 17648, "estimated_input_tokens": 4412, "estimated_eval_tokens": 9080 }, @@ -14697,7 +14635,7 @@ "node_count": 14, "item_count": 14, "soft_pair_count": 28, - "serialized_chars": 17646, + "serialized_chars": 17648, "estimated_input_tokens": 4412, "estimated_eval_tokens": 9080 }, @@ -14706,180 +14644,180 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All items are distinct model-request issues, not duplicates. They reference different model families and/or different modalities (text, audio, OCR, segmentation), so the cluster should not be merged.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue: the requests are for different models/features (for example EuroBert, NeoBERT, MiniCPM3, DEIMv2, Qwen3-TTS/ASR, LFM2.5 Audio, Voxtral, DeepSeek-OCR2), so they are not the same underlying bug/change.", + "summary": "These are all separate model-support requests across different model families and modalities; despite broad topical similarity, none look like true duplicates or the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a safe global canonical; the similarities are only at the level of \"add model support\" and do not indicate duplicate reports.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config vs DEIMv2 are different model requests with no shared code-path problem." + "reason": "EuroBert config support and DEIMv2 model support are unrelated requests for different architectures." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert model config and Qwen3-TTS support are unrelated models/modalities." + "reason": "EuroBert config support is unrelated to Qwen3-TTS audio support." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are different model families and tasks." + "reason": "NeoBERT support and Qwen3-ASR support target different models and different inference paths." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT support and the weight-sparse transformer request are not the same change." + "reason": "NeoBERT model addition is unrelated to the OpenAI weight-sparse transformer request." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT and LFM2.5 Audio target different architectures and modalities." + "reason": "NeoBERT support and LFM2.5 Audio support are different model families with different requirements." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are distinct model additions, not duplicate reports." + "reason": "NeoBERT and ModernVBERT are separate model requests, not the same concrete bug or change." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS are unrelated model-support requests." + "reason": "NeoBERT and Voxtral TTS support are unrelated model additions." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture and Qwen3-ASR support are different concrete model integrations." + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are distinct feature requests." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 support is not the same as the sparse-transformer request." + "reason": "MiniCPM3 support is unrelated to the weight-sparse transformer request." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different models; no duplicate bug/change is evident." + "reason": "MiniCPM3 and ModernVBERT are different model families and not duplicates." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 and Voxtral TTS are separate model requests." + "reason": "MiniCPM3 support is unrelated to Voxtral-4B-TTS support." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are unrelated model additions." + "reason": "DEIMv2 model support and Qwen3-TTS support are separate requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT and Qwen3-ASR are different models and tasks." + "reason": "ModernVBERT support and Qwen3-ASR support are different models and different tasks." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are distinct model-support requests." + "reason": "ModernVBERT and Qwen3-TTS are unrelated model-support requests." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "A specific ModernVBERT request is not the same as the generic BERT-like RoPE proposal." + "reason": "ModernVBERT models and general BERT-like RoPE support overlap conceptually, but they are not the same concrete request or code change." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "Generic BERT-like RoPE support and Voxtral TTS are unrelated." + "reason": "General BERT-like RoPE support is unrelated to Voxtral TTS support." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are different models/tasks; same vendor name is not enough to merge." + "reason": "Both are Qwen3 audio requests, but TTS and ASR are different model targets and would not be merged into one fix." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and Voxtral TTS are different model families." + "reason": "Qwen3-TTS and Voxtral-4B-TTS are different TTS model additions." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and DeepSeek-OCR2 target different modalities and code paths." + "reason": "Qwen3-TTS support is unrelated to DeepSeek-OCR2 support." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS support is unrelated to the sparse-transformer request." + "reason": "Qwen3-TTS support and weight-sparse transformer support are unrelated." }, { "left": "issue:43671", "right": "issue:43909", "accept": false, - "reason": "Both are audio-related, but Qwen3-TTS and LFM2.5 Audio are different models and integrations." + "reason": "Qwen3-TTS and LFM2.5 Audio are separate audio-model support requests." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are different model-support requests." + "reason": "Qwen3-ASR and LFM2.5 Audio are different models and different integration paths." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR support and EUPE are unrelated model additions." + "reason": "Qwen3-ASR support is unrelated to EUPE model support." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR and the weight-sparse transformer request do not describe the same change." + "reason": "Qwen3-ASR support and the weight-sparse transformer request are unrelated." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio and SAM3-LiteText are different modalities and code paths." + "reason": "LFM2.5 Audio support and SAM3-LiteText are different model requests." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Weight-sparse transformer support and EUPE are unrelated model requests." + "reason": "The weight-sparse transformer request and EUPE are different proposed models, not the same change." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral TTS are different speech-model integrations." + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different speech-model support requests." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR and DeepSeek-OCR2 are different tasks and model families." + "reason": "Qwen3-ASR support and DeepSeek-OCR2 support are unrelated." } ] }, @@ -14920,7 +14858,7 @@ "issue:43668|issue:44031", "issue:43668|issue:43813" ], - "prepared_review_unit_hash": "6abaf3305ff7c06f0ce8b58e5850662fe0072d72ac6168feb61498499adad766", + "prepared_review_unit_hash": "be9be9e3985fe90301d4d44c167d875dd8408d5cc3c6b96540eda1fdfb961d15", "component_budget": { "node_count": 6, "item_count": 6, @@ -14942,60 +14880,60 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This is not a true duplicate cluster: the issues span unrelated areas (model norm behavior, Mask2Former loss plumbing, config type hints, a quantization typo, and tokenizer warnings). The soft similarities are mostly generic bug-report phrasing rather than the same underlying defect.", - "confidence": 0.16, - "canonical_issue_reason": "If one issue must represent the set, issue 44031 is the most broadly scoped and most discussed, but it is still not a duplicate of the others.", + "summary": "These issues are not duplicates of one another; they span unrelated model, config, API, typo, and tokenizer-regression topics. The cluster looks like a loose soft-similarity grouping rather than a true duplicate set.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43240 is the most broadly phrased and externally referenced item in the set, so it is the best representative of the cluster even though the cluster itself does not contain a true duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44031 is the best single representative only because it has the broadest user impact and highest discussion activity; it is not substantively the same bug as the other issues.", + "best_issue_reason": "issue:43240 has the strongest centrality signals (inbound references/discussion) and the most general bug framing, making it the best issue candidate for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Different problems in different model/config code paths: a Llama norm-design question vs a ModernBERT config type-hint issue." + "reason": "Different model components and problem types: Llama4TextL2Norm vs ModernBERTConfig type hint. One is about normalization semantics, the other about an annotation typo." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Mask2Former ignore_value handling and loss-kwargs plumbing are separate defects, even though both affect training behavior." + "reason": "Mask2Former ignore_value handling and loss kwargs forwarding are separate code paths and bugs; they are not the same underlying defect." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Unrelated subsystems: a model loss/ignore_value bug vs a tokenizer regex-warning regression." + "reason": "Unrelated subsystems: Mask2Former loss/ignore_value versus tokenizer regex warning regression." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "A model-specific ignore_value issue is not the same as a quantization attribute typo." + "reason": "Model loss-handling bug vs a literal typo in a quantization-related name; no shared code-path." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "These concern different components and bug classes: model behavior vs config annotation correctness." + "reason": "Mask2Former runtime behavior bug and ModernBERT config typing issue are unrelated." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss-function kwargs forwarding and a config type-hint error are unrelated." + "reason": "Loss kwargs not being forwarded is an execution bug; incorrect norm_eps type hint is a configuration annotation issue." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "A ModernBERT config typing issue does not match the tokenizer regex warning regression." + "reason": "ModernBERTConfig type hint mismatch and tokenizer regex warning regression are different components and symptoms." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Both are small correctness issues, but they affect different symbols and code paths; not the same underlying bug." + "reason": "Type-hint typo in ModernBERTConfig and a quantization-operation naming typo are distinct, unrelated fixes." } ] }, @@ -15025,7 +14963,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "538c6cdfaf2182fb2744206c2d402b8340a5f9cc4e2cb263852c4f88c5e99649", + "prepared_review_unit_hash": "f4d78dc97b5407eab1e1590a0f09e0d1085dc140440f218726e372fe02bd8564", "component_budget": { "node_count": 2, "item_count": 2, @@ -15047,18 +14985,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one is about block decoding, the other is a request for DeepSeek v3.2 model support. They do not describe the same underlying bug or change set.", - "confidence": 0.97, + "summary": "These two issues do not look like duplicates: one is about block decoding, the other is about adding DeepSeek v3.2 support. They target different features and likely different code paths.", + "confidence": 0.96, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No single canonical issue stands out because the reports are unrelated; if forced, neither is a good representative of the other.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different problem domains and requested outcomes; no shared code path or concrete change target." + "reason": "Different requests with no clear shared underlying bug or change: block decoding vs model/version support." } ] }, @@ -15088,12 +15026,12 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "bb6004f7bde93d08eb84a4aaeb4ba431b730c23af94b8f9bb9dd7350d4e76477", + "prepared_review_unit_hash": "29be77c41d2f9e905ca615515af09c801401d926b6cc4b2a58c5678e89ac985f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -15101,7 +15039,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -15110,18 +15048,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are related at a high level to Trainer extensibility, but they target different behaviors: one is about `remove_unused_column` customizability, the other about FLOPS estimation for custom trainers. They should not be deduplicated.", - "confidence": 0.9, - "canonical_issue_reason": "No canonical duplicate: the issues describe different problems in Trainer, not the same underlying bug or change.", + "summary": "These are related by theme (custom trainers in `Trainer`), but they describe different problems: one is about `remove_unused_column` customizability, the other about FLOPS estimation for custom trainers. I would not merge them as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "No clear canonical duplicate; if one must be chosen as the cluster representative, #43598 is a slightly broader Trainer customizability issue.", "canonical_pr_reason": null, - "best_issue_reason": "`issue:43598` is the better representative of the cluster\u2019s broader Trainer customizability theme, while `issue:43600` is a separate FLOPS-estimation concern.", + "best_issue_reason": "#43598 is the better representative because it frames a general Trainer behavior/customizability concern rather than a narrower FLOPS-estimation bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Different code paths and user-facing problems: `remove_unused_column` behavior vs FLOPS estimation for custom trainers." + "reason": "Same general area, but different underlying bugs: one concerns `remove_unused_column` behavior, the other FLOPS counting for custom trainers. They would not be fixed by one change." } ] }, @@ -15151,12 +15089,12 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "f22e18399701643dc129502cf2e7cb08d61721af184e5322e5b2dc66ef7b2a1e", + "prepared_review_unit_hash": "5ee5a8ba56314a13f59cd9f056d9c9cfe948b5ae25ea05e4767a9da7a4e5b7c2", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -15164,7 +15102,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1611, + "serialized_chars": 1612, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -15173,18 +15111,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are about different Qwen model variants and different AutoModel targets, so they do not look like duplicates. Issue 43684 is a broader registration request for Qwen3-Omni, while 44405 is a specific request for sequence classification support for Qwen3.5.", - "confidence": 0.96, - "canonical_issue_reason": "Issue 43684 is the broader and more general model-registration request, so it is the better representative if one issue must be chosen.", + "summary": "The two issues are about different Qwen model families and different auto-model capabilities: one asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, while the other asks for AutoModelForSequenceClassification support for Qwen3.5. They are related only at a broad subsystem level, not the same underlying bug/change.", + "confidence": 0.95, + "canonical_issue_reason": "If one issue must represent the cluster, 43684 is the broader model-registration request and is more general than the Qwen3.5 sequence-classification support request.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the most central/general issue in the pair because it targets AutoModel and AutoModelForConditionalGeneration registration rather than a single task head.", + "best_issue_reason": "43684 is the better representative because it targets core auto-model registration, which is broader and more central than a single-task head enablement.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different model families (Qwen3-Omni vs Qwen3.5) and different capabilities (model registration vs sequence classification support); they do not appear to describe the same underlying bug or change." + "reason": "Different model families (Qwen3-Omni vs Qwen3.5) and different requested capabilities (registration/conditional generation vs sequence classification); not the same bug or change." } ] }, @@ -15221,12 +15159,12 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "fe5b9a6f72664c2d56975d736d9be7abcf29be0f3a7c3ef66ab70328d8e07e57", + "prepared_review_unit_hash": "1730361b56ed37366fec8c84f511d6f3bab2e07c14a2eac01b0d4d377c8c82ab", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4262, + "serialized_chars": 4264, "estimated_input_tokens": 1066, "estimated_eval_tokens": 2388 }, @@ -15234,7 +15172,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4262, + "serialized_chars": 4264, "estimated_input_tokens": 1066, "estimated_eval_tokens": 2388 }, @@ -15243,54 +15181,54 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Issue #43835 looks unrelated. The three \"huggingface \" issues (#43845, #43846, #43859) are strongly similar by title and timing, with #43845 the best canonical representative, but the soft edges should remain tentative because the packet only provides soft_similarity evidence and deterministic_accept is false for every pair.", - "confidence": 0.7, - "canonical_issue_reason": "#43845 is the earliest of the near-identical \"huggingface \" issues, so it is the best canonical representative if this cluster is treated as one report.", + "summary": "The three \"huggingface\" issues are plausibly related and 43845 is the best representative, but the packet only gives soft_similarity for every pair and no deterministic_accept, so the duplicate links should stay tentative. \"Fett\" appears separate.", + "confidence": 0.63, + "canonical_issue_reason": "issue:43845 is the best representative of the huggingface subgroup because it is the earliest of the near-identical placeholder-title reports and aligns most closely with issue:43846, but the evidence is still only soft similarity.", "canonical_pr_reason": null, - "best_issue_reason": "#43845 has the clearest cluster signal: same title as #43846/#43859 and the earliest timestamp, while #43835 appears different.", + "best_issue_reason": "issue:43845 is the strongest cluster representative: it is earliest among the huggingface reports, while issue:43859 is shorter and the only one with discussion. Even so, the available evidence does not justify treating the links as confirmed duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles and no shared concrete bug details or targets; only loose similarity, so this is not enough to treat them as duplicates." + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no shared concrete bug details or targets; only loose similarity, so this is not enough to treat them as duplicates." + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles and no shared concrete bug details or targets; only loose similarity, so this is not enough to treat them as duplicates." + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." }, { "left": "issue:43845", "right": "issue:43846", "accept": false, - "reason": "Very similar titles and close timestamps, but the packet only gives soft_similarity and deterministic_accept is false, so this should stay tentative rather than accepted." + "reason": "Same placeholder title, but the packet provides only soft_similarity and no deterministic_accept or concrete shared targets/files, so this is not strong enough to accept as the same issue." }, { "left": "issue:43845", "right": "issue:43859", "accept": false, - "reason": "Very similar titles, but only soft_similarity evidence is present and deterministic_accept is false; tentative duplicate candidate, not a confirmed match." + "reason": "The match is only superficial: same placeholder title, but no concrete issue target, filename, or diff evidence, and deterministic_accept is false." }, { "left": "issue:43846", "right": "issue:43859", "accept": false, - "reason": "Very similar titles, but only soft_similarity evidence is present and deterministic_accept is false; tentative duplicate candidate, not a confirmed match." + "reason": "The match is only superficial: same placeholder title, but no concrete issue target, filename, or diff evidence, and deterministic_accept is false." } ] }, "evaluator_result": { "accept": false, - "feedback": "The cluster summary is mostly grounded, but the soft-edge verdicts are not conservative enough. For 43845/43846/43859 the packet only provides soft_similarity evidence and all three pairs have deterministic_accept=false, so marking them as accepted duplicates overstates the evidence. The 43835 rejections are fine, but the positive duplicate claims should be downgraded to tentative rather than accepted." + "feedback": "The cluster summary is mostly grounded, but the soft-edge verdicts are too confident. In the packet, all pairwise links are only soft_similarity, and deterministic_accept is false for every pair. Accepting issue:43845 with issue:43859 and issue:43846 with issue:43859 overstates the evidence; those should be treated more conservatively. The representative choice for the huggingface duplicates is otherwise reasonable, and Fett does look separate." }, "evaluator_used": true, "retried": true, @@ -15328,88 +15266,88 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "7abfb22130c5be7a554f542dc1d9141ed283bed23984322b74e705c9db054332", + "prepared_review_unit_hash": "55c9e3606b1d35cc759370c42b473c916265c7d1cc6f092823bcac48bcf1a107", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5785, - "estimated_input_tokens": 1447, - "estimated_eval_tokens": 3150 + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5785, - "estimated_input_tokens": 1447, - "estimated_eval_tokens": 3150 + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly unrelated placeholder/spam issues. There is no strong evidence that any pair reports the same underlying bug or change; the only non-generic item is a documentation suggestion, which still doesn\u2019t match the others.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 44418 is the only one with a substantive, descriptive title; the others are generic placeholders or unrelated spam-like text. If a representative must be chosen, it is the least ambiguous.", + "summary": "This cluster is dominated by low-content spam/generic issue stubs, with no actionable bug or change details to justify deduping on substance.", + "confidence": 0.28, + "canonical_issue_reason": "Issue 44418 is the least spam-like representative: it is the only non- title and is still just a generic placeholder, so it\u2019s the best available cluster anchor.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44418 is the best standalone issue in the set because it is the only clearly described topic. It is not a duplicate of the others, but it is the most meaningful anchor for the cluster.", + "best_issue_reason": "No issue here contains enough concrete problem description to be a true canonical duplicate target; 44418 is only the least bad representative, not a substantive one.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are placeholder '' issues with no evidence of a shared underlying bug/change; same-looking titles are not enough." + "reason": "Same '' title, but there is no actual bug/change context to show they describe the same underlying issue." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are placeholder '' issues and there is no code-path, target, or description evidence tying them together." + "reason": "Both are spam placeholders with no concrete problem statement; identical titles alone are not enough to merge them as duplicates." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is generic spam, the other is a documentation suggestion; they do not appear to describe the same issue." + "reason": "Titles are unrelated generic placeholders, and neither issue has enough substance to establish a shared underlying bug." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "A spam placeholder and an unrelated 'Check out \"Google LLC\"' issue do not look like the same underlying report." + "reason": "'' and 'Check out \"Google LLC\"' do not indicate the same bug or change; this looks like unrelated spam." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "A documentation enhancement request and a spam placeholder are not the same underlying change or bug." + "reason": "Generic documentation-sounding text versus a spam placeholder; no concrete common issue is described." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "The titles point to unrelated content; no shared concrete problem is visible." + "reason": "No shared actionable content; these are just low-information placeholder issues." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Identical spam titles alone are insufficient to treat them as duplicates of the same underlying bug/change." + "reason": "Identical spam titles, but there is no evidence of a real shared bug/change to deduplicate." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Both are unrelated spam-like submissions with no shared concrete issue or code path." + "reason": "A spam placeholder and a promotional-sounding title are not the same underlying issue." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "No evidence suggests these describe the same bug/change; they are just unrelated spam-like issues." + "reason": "No substantive overlap beyond both looking like spam; not a valid duplicate match." } ] }, @@ -15439,12 +15377,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "9092c16ce8635c47d275044ca294d081aadeadb5ef5a6b19627b414abf02028a", + "prepared_review_unit_hash": "d2bb998fa78245e61231f6e0459f0c4a1fb58531e8791f9f966fbecb99a5ab5e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -15452,7 +15390,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1645, + "serialized_chars": 1646, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -15461,24 +15399,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster has two closed issues with identical titles and similar metadata, so they are plausible duplicates, but the evidence is limited because we do not have body text or stronger linkage. This should be treated as a likely duplicate pair rather than a near-certain match.", - "confidence": 0.82, - "canonical_issue_reason": "#45145 is a reasonable canonical issue because it has the inbound reference and was created earlier, so it provides slightly more cluster context.", + "summary": "The pair looks strongly related, but the available evidence is not enough to call them definite duplicates with high confidence. They share the exact title and similar body size, yet we do not see body-content overlap, explicit links, or scope details to prove they are the same underlying issue.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 45145 is the earlier report and has an inbound reference, so it is the best tentative anchor if these are later found to be duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best issue representative for the same reason: earlier creation plus the inbound reference make it the better anchor for the duplicate pair.", + "best_issue_reason": "Issue 45145 is the best representative of the pair because it is earlier and already has more visible downstream attention, though the duplicate call is not definitive.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": false, - "reason": "The titles match and the soft similarity is notable, but with no body text and only moderate similarity evidence, this is not strong enough to confidently treat as the same underlying issue." + "reason": "The evidence is too thin to confirm the same underlying issue: identical titles and similar body lengths alone do not establish duplicate scope, and there is no explicit cross-linking or content-level proof." } ] }, "evaluator_result": { "accept": false, - "feedback": "The cluster is plausibly a duplicate pair, but the summary overstates the evidence: we only know the titles match and there is soft similarity, not that the bodies are \"near-identical\" or \"highly similar.\" The soft-edge accept is also a bit strong given the moderate jaccard score and lack of body text. The canonical issue choice is reasonable, but the overall writeup should stay more conservative." + "feedback": "The evidence supports strong similarity, but not a high-confidence duplicate conclusion. The packet only shows matching titles, close body lengths, and a soft-similarity signal; it does not establish identical scope or that 45145 is definitively the canonical anchor. The soft-edge verdict is a bit too assertive given the available evidence, and the confidence of 0.98 is overstated." }, "evaluator_used": true, "retried": true, @@ -15559,30 +15497,30 @@ "soft_pairs": [ "issue:45672|issue:45674" ], - "prepared_review_unit_hash": "b713f8d4d9d9e002214f50108d6170567e434ea4bb94834287c86f46b3736c7a", + "prepared_review_unit_hash": "32fdae562cd9cbc5b133f37e1d8671c7f7ef673a5d70a63e234a12a6cdc8d70f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1669, - "estimated_input_tokens": 418, - "estimated_eval_tokens": 1092 + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1669, - "estimated_input_tokens": 418, - "estimated_eval_tokens": 1092 + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both 4-bit inference issues, but they describe different bugs in different code paths: one is a Gemma4/audio-module uint8 `torch.finfo()` failure, the other is a BitsAndBytesConfig `llm_int8_skip_modules` / `lm_head` exclusion regression.", - "confidence": 0.96, + "summary": "These are related to 4-bit inference/bitsandbytes behavior, but they describe different failure modes in different code paths, so they should not be clustered as duplicates.", + "confidence": 0.94, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, @@ -15592,7 +15530,7 @@ "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Different failure modes and fixes: `torch.finfo()` on uint8 weights in audio modules vs. `llm_int8_skip_modules` clearing the default `lm_head` exclusion. Same broad area, but not the same underlying bug." + "reason": "Different underlying bugs: one is a Gemma4 audio-module uint8/torch.finfo TypeError during NF4 inference, the other is a BitsAndBytesConfig skip-modules regression that removes the default lm_head exclusion and triggers an AssertionError. Same broad area, but not the same fix." } ] }, @@ -15649,151 +15587,151 @@ "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "fe7bf5cf6e8094faf1722b79c920793bf48b4711d32804baded89020744cbb1e", + "prepared_review_unit_hash": "0e21cce5ec3a1d2e2c4d1312b93711d4fa9328fab7211a2910a281bacdf88f06", "component_budget": { "node_count": 12, "item_count": 12, "soft_pair_count": 18, - "serialized_chars": 32877, - "estimated_input_tokens": 8220, - "estimated_eval_tokens": 16696 + "serialized_chars": 32874, + "estimated_input_tokens": 8219, + "estimated_eval_tokens": 16694 }, "budget": { "node_count": 12, "item_count": 12, "soft_pair_count": 18, - "serialized_chars": 32877, - "estimated_input_tokens": 8220, - "estimated_eval_tokens": 16696 + "serialized_chars": 32874, + "estimated_input_tokens": 8219, + "estimated_eval_tokens": 16694 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The only clear duplicate pair is PR #43251/#43254: both modify `src/transformers/loss/loss_utils.py` in the same function to pass supported kwargs through to `nn.functional.cross_entropy` (including `weight` and `label_smoothing`). The rest of the soft pairs are still distinct changes across unrelated models, tests, or refactors and remain reject-worthy.", - "confidence": 0.99, + "summary": "This cluster is mostly a grab bag of unrelated PRs; the only true duplicate is the two fixed_cross_entropy kwargs patches (43251 and 43254).", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43251 is a strong canonical representative for the loss fix because it makes the same `fixed_cross_entropy` change as #43254 in the same file/function and both PRs explicitly target issue #43240. Its patch is a straightforward representative of that concrete bug fix.", + "canonical_pr_reason": "PR 43251 is the best canonical representative for the only real duplicate pair: it has the explicit issue link, targets the same loss_utils bug, and cleanly extends fixed_cross_entropy with weight/label_smoothing handling.", "best_issue_reason": null, - "best_pr_reason": "PR #43251 is the best overall PR candidate for this cluster because it cleanly represents the shared `fixed_cross_entropy` kwargs fix in `src/transformers/loss/loss_utils.py` and carries the same explicit issue target (#43240) as the duplicate PR.", + "best_pr_reason": "43251 is the strongest standalone PR in the set because it fixes a concrete bug in a single code path and is clearly tied to the tracked issue; 43254 is a near-duplicate alternative.", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "VideoPrism model/docs addition vs an Executorch dynamic-shape export test; they are different changes and do not fix the same code path." + "reason": "VideoPrism model addition vs an executorch dynamic-shapes test; unrelated changes." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Mistral tokenizer method additions vs adding a new MobileLLM model; no shared underlying bug or change." + "reason": "Mistral tokenizer helper methods vs adding MobileLLM; different subsystems and goals." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Tokenizer API patch vs Executorch export test; unrelated functionality." + "reason": "Tokenizer method additions are unrelated to executorch export testing." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness changes vs a new MobileLLM model implementation; only broad subsystem overlap." + "reason": "Processor robustness changes vs adding a new model; not the same underlying change." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Cross-entropy kwargs fix vs new MobileLLM model work; completely different code paths." + "reason": "Loss-function kwargs fix vs MobileLLM model implementation; unrelated." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Loss-function fix vs Mistral4 model/test changes; same general training area, but not the same bug or patch." + "reason": "Both touch code quality in different contexts, but one fixes cross-entropy kwargs while the other addresses Mistral4 test/runtime issues; not the same bug." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Executorch dynamic-shape test vs Mistral4 refactor/tests; different concrete changes." + "reason": "Executorch dynamic-shape export test vs Mistral4 test fixes; different concerns." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Dynamic-shape export test vs MobileLLM model addition; no duplicate evidence beyond broad framework overlap." + "reason": "Executorch exportability test vs MobileLLM model addition; unrelated." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New MobileLLM model vs `cls.config` resolution-order fix; unrelated code paths." + "reason": "New MobileLLM model support vs config_class resolution ordering fix; different code paths." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "New model implementation vs Falcon output-collection refactor; superficially related at most." + "reason": "Model addition vs Falcon refactor; not the same underlying change." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor vs model-creation skill/docs draft; not the same underlying code change." + "reason": "Falcon output-interface refactor vs a model-creation skill/documentation PR; unrelated." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon interface refactor vs Mistral4 model/test adjustments; different models and goals." + "reason": "Falcon refactor vs Mistral4 tests; different models and different problems." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test/model work vs `layrnorm` rename cleanup; distinct changes." + "reason": "Mistral4 test fixes vs a CLIP-like layrnorm typo correction; unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 work vs class config resolution fix; unrelated code paths." + "reason": "Mistral4 tests vs config resolution fix; no shared bug or change." }, { "left": "pull_request:43823", "right": "pull_request:44827", "accept": false, - "reason": "MobileLLM model addition vs Mistral4 fixes; broad model overlap only." + "reason": "MobileLLM model addition vs Mistral4 tests; unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 model/test changes vs model-creation skill draft; not the same underlying change." + "reason": "Mistral4 tests vs model-creation skill/docs; unrelated." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "MobileLLM implementation vs model-creation skill/docs; one is code, the other is tooling/docs." + "reason": "MobileLLM model implementation vs a model-creation skill draft; different artifact types and purposes." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs patch the same function in the same file to extend `fixed_cross_entropy` with the same new `cross_entropy` kwargs support, including `weight` and `label_smoothing`; they are the same underlying bug fix and could plausibly be merged into one PR." + "reason": "Both patches modify the same fixed_cross_entropy function in the same file to pass supported cross-entropy kwargs (weight/label_smoothing); they are the same underlying fix and could be merged into one PR." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The duplicate-pair verdict for #43251/#43254 is grounded and conservative, but the PR-ranking rationale overstates the evidence: both PRs have an explicit issue link to #43240, so #43251 is not \"the only one with a direct issue link.\" Please avoid that claim and keep the canonical/summary language strictly tied to the packet. All other soft-edge rejections look conservative." + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted pair is well supported by the diffs: 43251 and 43254 are near-identical fixes to the same function in the same file. The other soft-edge rejections are conservative and do not overstate evidence. Minor note: phrases like \"only true duplicate\" are a bit strong, but not enough to invalidate the review." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -15817,7 +15755,7 @@ "soft_pairs": [ "pull_request:40546|pull_request:42781" ], - "prepared_review_unit_hash": "90c03fdb1240328059365b98069966f9df99aa43577e4b36bb3552ddf92b09c7", + "prepared_review_unit_hash": "caa0e7bef8faea9dac1b85be81a6d1783b3bc60403bd297258c98d620359889c", "component_budget": { "node_count": 2, "item_count": 2, @@ -15839,18 +15777,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are VibeVoice-related, but they are not duplicates: #40546 is the initial core VibeVoice implementation, while #42781 is a later draft adding VibeVoice Realtime plus an acoustic tokenizer and related generation changes.", + "summary": "These are related VibeVoice implementation PRs, but they are not duplicates: the first adds the base VibeVoice model, while the second adds VibeVoice Realtime and an acoustic tokenizer with extra generation/model changes. They overlap in shared package/auto-registration files, but they address different deliverables and code paths.", "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "#40546 is the best representative PR for the cluster because it is the foundational end-to-end VibeVoice implementation (model code, auto mappings, processing, docs, and integration points).", + "canonical_pr_reason": "PR 40546 is the cleaner canonical PR for the core VibeVoice model implementation; it is the earlier, broader base integration and has substantial review activity.", "best_issue_reason": null, - "best_pr_reason": "#40546 is broader and more central to the base feature; #42781 looks like a follow-on extension for realtime/acoustic-tokenizer support rather than the same underlying change.", + "best_pr_reason": "PR 40546 is the best representative of the underlying VibeVoice model work because it introduces the foundational model/docs/auto mappings, whereas PR 42781 is a later extension focused on realtime/tokenizer support.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Reject: although they share VibeVoice files, #40546 implements the base model and integration, while #42781 adds a separate realtime/acoustic-tokenizer feature set. They are related work but not the same concrete bug/change and would not plausibly merge into one PR." + "reason": "Reject: despite heavy filename overlap, 42781 is a separate realtime/tokenizer extension with additional model and generation changes, not the same concrete bug fix or mergeable duplicate of 40546." } ] }, @@ -15880,7 +15818,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "ec59e90a908096743c0414660f696803abddfaa280dbe4af0047773bcc329535", + "prepared_review_unit_hash": "38f21745407f14b5136192ef8c8a8cd5bc260286e9bdc61f93701f765d589fd9", "component_budget": { "node_count": 2, "item_count": 2, @@ -15902,7 +15840,7 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster does not appear to be a duplicate set: one PR adds TimesFM covariate forecasting, while the other adds DINOv3 image-classification support. They touch different models, different code paths, and different user-facing features.", + "summary": "The two PRs are unrelated: one adds TimesFM covariate forecasting support, the other adds DINOv3 image classification support. They do not fix the same code-path problem and should not be deduplicated.", "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, @@ -15913,7 +15851,7 @@ "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different underlying changes in unrelated subsystems: TimesFM forecasting with covariates vs DINOv3 image classification. They do not fix the same bug or implement the same feature and would not be merged into one PR." + "reason": "Different models and changes: TimesFM covariate forecasting vs DINOv3 image classification. No shared bug or mergeable single fix." } ] }, @@ -16006,12 +15944,12 @@ "soft_pairs": [ "pull_request:41797|pull_request:45186" ], - "prepared_review_unit_hash": "0af3d4b90eb598c0ee5e8481b384ce86d232fc9c8482be17b39d780ca5b4dc3e", + "prepared_review_unit_hash": "6bc8242cf798dd2a779ef0d830fa98ea2a72e60f3d2b6ab3edfd21b9334fe638", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6722, + "serialized_chars": 6723, "estimated_input_tokens": 1681, "estimated_eval_tokens": 3618 }, @@ -16019,7 +15957,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6722, + "serialized_chars": 6723, "estimated_input_tokens": 1681, "estimated_eval_tokens": 3618 }, @@ -16028,18 +15966,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are unrelated model-addition pull requests. They share standard Transformers boilerplate and registry/docs updates, but they implement different models (DeepSeek OCR vs Isaac) and do not fix the same code-path problem.", - "confidence": 0.96, + "summary": "Both items are model-addition PRs that touch shared registration/docs infrastructure, but they introduce different models (DeepSeekOCR vs Isaac) with different implementations and tests. They are not the same underlying change, so the soft duplicate edge should be rejected.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 41797 is the better representative because it has more review activity and broader integration changes, but it is not a duplicate of 45186.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 41797 is the stronger standalone representative of the cluster due to higher discussion/review activity; 45186 is a separate, unrelated model addition.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "Different model integrations with different filenames, configs, and test suites; shared registry/docs boilerplate is not enough to treat them as the same change." + "reason": "Different model integrations with distinct model directories, configs, and test suites; shared auto/docs plumbing is not enough to treat them as the same PR or merge into one change." } ] }, @@ -16069,7 +16007,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "c7dec2125d7ce5a0d73948eba3c5fb4ea7d38dd1e8d8cc261fc123e521471f23", + "prepared_review_unit_hash": "1b9bf525443a07db81b3c6525437715628283fbb00744db00aee32e1d0cf61e5", "component_budget": { "node_count": 2, "item_count": 2, @@ -16091,18 +16029,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both workflow dependency-upgrade PRs, but they change different actions and fix different compatibility concerns, so they should not be merged as duplicates.", - "confidence": 0.95, + "summary": "Two separate maintenance PRs touching overlapping GitHub Actions workflows, but they update different actions and target different upgrade goals. They should not be treated as duplicates.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR 42975 is the better canonical representative: it is broader, explicitly about Node 24 compatibility, and updates several core workflow actions across the repo.", "best_issue_reason": null, - "best_pr_reason": "Neither PR is a duplicate of the other. PR 42975 is the broader GitHub Actions runtime compatibility update; PR 42976 is specifically about upgrading docker/build-push-action in Docker image workflows.", + "best_pr_reason": "PR 42975 is the most representative item in the cluster because it covers the larger workflow-action upgrade effort; PR 42976 is a narrower docker-action version bump.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Different underlying changes: 42975 upgrades checkout/cache/upload-artifact and several workflow actions for Node 24 compatibility, while 42976 upgrades docker/build-push-action in Docker image workflows. They do not appear to fix the same concrete bug or change the same code path." + "reason": "Overlap in workflow filenames is not enough: 42975 upgrades checkout/cache/upload-artifact for Node 24 compatibility, while 42976 only bumps docker/build-push-action. Different actions, different concrete change, so not the same underlying PR." } ] }, @@ -16146,12 +16084,12 @@ "pull_request:45490|pull_request:45630", "pull_request:45626|pull_request:45630" ], - "prepared_review_unit_hash": "1a368cc8b66c85b21900f89b99ae221721d9991202474745ff3dcdf603f8dc8a", + "prepared_review_unit_hash": "014709e08748e947ca9b15ee3418d60cec68e8754e980c3c69045833c036ad3a", "component_budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 10, - "serialized_chars": 23911, + "serialized_chars": 23910, "estimated_input_tokens": 5978, "estimated_eval_tokens": 12212 }, @@ -16159,7 +16097,7 @@ "node_count": 7, "item_count": 7, "soft_pair_count": 10, - "serialized_chars": 23911, + "serialized_chars": 23910, "estimated_input_tokens": 5978, "estimated_eval_tokens": 12212 }, @@ -16168,8 +16106,8 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is not a duplicate set: it mixes several unrelated model-addition PRs (ViT NEPA, Molmo, Molmo2, CTSM, Kimi2-6, PP-FormulaNet) plus a distributed pipeline-parallel feature PR. The overlap is mostly shared Transformers boilerplate (docs, auto mappings, repo checks), not the same underlying change or bug.", - "confidence": 0.97, + "summary": "This cluster is a loose collection of unrelated pull requests: several separate model additions (ViT NEPA, Molmo, Molmo2, CTSM, PP-FormulaNet, Kimi2-6) plus one distributed loading feature PR. They share repo areas like docs and auto mappings, but not the same underlying change.", + "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, @@ -16179,61 +16117,61 @@ "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Both are model-introduction PRs, but for different models (ViT NEPA vs Molmo) with different codepaths and files; shared docs/auto-registration boilerplate is not enough to treat them as the same change." + "reason": "Different model additions: ViT NEPA vs Molmo. Shared docs/auto-registration changes are generic scaffolding, not the same bug or feature." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "Different model additions (ViT NEPA vs Molmo2). They share generic registration/documentation edits, but the model implementations and targets are unrelated." + "reason": "Different model additions: ViT NEPA vs Molmo2. They are separate model integrations, not one mergeable change." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "ViT NEPA and CTSM are distinct model additions with different architectures and files; the overlap is only standard Transformers model-scaffolding." + "reason": "Different model additions: ViT NEPA vs CTSM. No shared concrete code-path problem." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo and Molmo2 are separate model families with different model code and docs; this looks like related naming, not the same underlying PR." + "reason": "Molmo and Molmo2 are related by name, but they are distinct model implementations with different files, APIs, and tests; not the same underlying PR." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are unrelated model additions; shared auto/model registration and docs are broad boilerplate, not duplicate evidence." + "reason": "Molmo vs CTSM are unrelated model integrations with different architectures and code paths." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM are different model ports with different implementations; similarity is limited to generic model onboarding changes." + "reason": "Molmo2 vs CTSM are separate model additions; shared registry/docs edits are not enough to merge them." }, { "left": "pull_request:44601", "right": "pull_request:45626", "accept": false, - "reason": "Pipeline-parallel loading support is a distributed systems feature, while PP-FormulaNet is a model addition. They do not fix the same code-path problem." + "reason": "One is native pipeline-parallel loading support, the other is PP-FormulaNet model support. Different feature areas and code paths." }, { "left": "pull_request:45490", "right": "pull_request:45626", "accept": false, - "reason": "CTSM and PP-FormulaNet are separate model addition PRs for different models; shared auto-mapping and docs updates are routine and do not indicate duplication." + "reason": "CTSM and PP-FormulaNet are different model additions; same broad subsystem, but not the same change." }, { "left": "pull_request:45490", "right": "pull_request:45630", "accept": false, - "reason": "CTSM vs Kimi2-6 are unrelated model additions. The shared files are just standard registration/docs, not the same concrete change." + "reason": "CTSM and Kimi2-6 are separate model integrations with distinct model files and tests." }, { "left": "pull_request:45626", "right": "pull_request:45630", "accept": false, - "reason": "PP-FormulaNet and Kimi2-6 are different model support PRs with distinct implementations and model-specific files; not mergeable as one PR for the same bug/change." + "reason": "PP-FormulaNet and Kimi2-6 are different model additions; similar documentation/auto-registration edits do not indicate duplicate content." } ] }, @@ -16266,52 +16204,52 @@ "pull_request:43291|pull_request:44815", "pull_request:43665|pull_request:44815" ], - "prepared_review_unit_hash": "2ce7c2f6954cac3624c8f1a2d3a0983e353e4be187998ed1a04f921d31af6e76", + "prepared_review_unit_hash": "de3d173608954f2699c6d6a6040b302db186163d732f2f12b4c7f2a7d24aebfb", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7093, - "estimated_input_tokens": 1774, - "estimated_eval_tokens": 3804 + "serialized_chars": 7092, + "estimated_input_tokens": 1773, + "estimated_eval_tokens": 3802 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7093, - "estimated_input_tokens": 1774, - "estimated_eval_tokens": 3804 + "serialized_chars": 7092, + "estimated_input_tokens": 1773, + "estimated_eval_tokens": 3802 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are not duplicates: one is a Whisper tokenizer/test fix, one is a test-only CLIP/ViT change, and one is a substantive FP8 dequant/loading fix. The cluster is too mixed to merge as a single change.", - "confidence": 0.92, + "summary": "These three PRs are not duplicates of the same change. PR 44815 is a substantive FP8/dequantization fix, PR 43291 is a Whisper tokenizer/test fix, and PR 43665 is a test-only placeholder change across CLIP/ViT. They share only broad similarity as small repo maintenance PRs.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44815 is the most substantive and self-contained code fix in the set, addressing FP8 dequant/model loading behavior across core loading and quantization code with tests.", + "canonical_pr_reason": "PR 44815 is the clearest canonical PR because it contains a concrete, coherent code-path fix in model loading and FP8 integration, with meaningful implementation changes and related tests.", "best_issue_reason": null, - "best_pr_reason": "PR 44815 best represents the cluster because it is the only clear product fix; the other PRs are unrelated test-only or Whisper-specific changes.", + "best_pr_reason": "PR 44815 is the best representative of the cluster because it is the most substantive and self-contained bug fix; the others are either unrelated test changes or a separate Whisper-specific tokenizer/test issue.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Different subsystems and changes: Whisper tokenizer/test code versus CLIP/ViT test edits. No shared concrete bug or mergeable change." + "reason": "Different targets and different problems: Whisper tokenizer/test behavior versus CLIP/ViT test assertions. No shared code path or mergeable unified fix." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper decoding/tokenization fix is unrelated to FP8 dequant/model-loading logic." + "reason": "Whisper decoding/tests are unrelated to FP8 dequant/model-loading logic. They do not appear to fix the same underlying bug or change." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Test-only CLIP/ViT edits are unrelated to the FP8 dequant/loading fix; they do not target the same code-path problem." + "reason": "One is test-only changes for CLIP/ViT; the other is a concrete FP8 dequantization fix. Same broad repository area, but not the same bug or change." } ] }, @@ -16344,7 +16282,7 @@ "pull_request:43488|pull_request:45476", "pull_request:43532|pull_request:45476" ], - "prepared_review_unit_hash": "169a4ba8c6795b061efa9d41d4298b003d8de2e7b435146d44ea51972cce38f2", + "prepared_review_unit_hash": "4f2afa3ba9085205d6993f828b6b32c56156c501a0409469f2c60b65a339e59a", "component_budget": { "node_count": 3, "item_count": 3, @@ -16366,30 +16304,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three separate synthetic \u201c[don't merge]\u201d PRs used to exercise repo/CI behavior. They overlap only superficially (mainly via shared test files) and do not fix the same concrete bug or implement the same change.", - "confidence": 0.93, + "summary": "These are three unrelated synthetic '[don't merge]' PRs; the apparent overlap is incidental, mostly a shared vit test filename, not the same underlying bug or change.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43532 is the most representative cluster member: it has the broadest CI/workflow-oriented diff and the most discussion/review activity, while the others are narrower bot-test variants.", + "canonical_pr_reason": "No true canonical duplicate exists here. PR #43532 is the most central by shared filenames and discussion activity, but it still changes unrelated workflow/test scaffolding rather than the same concrete fix as the others.", "best_issue_reason": null, - "best_pr_reason": "PR 43532 is the best single PR to anchor the cluster because it is the broadest and most central of the synthetic CI/test changes, even though the three PRs are not true duplicates.", + "best_pr_reason": "Not a good duplicate representative overall; if one must be picked, #43532 is the least-bad cluster center, but it is not actually a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "Different underlying changes: 43488 is a setup/version/dependency tweak plus a test edit, while 43532 rewrites workflow/comment CI paths and related utilities. Shared vit test filename is incidental." + "reason": "Different changes: setup.py/dependency/version tweaks versus workflow/testing utility edits. Shared vit test file is incidental, not the same bug or fix." }, { "left": "pull_request:43488", "right": "pull_request:45476", "accept": false, - "reason": "Different targets and code paths: 43488 touches packaging/versioning and a test, whereas 45476 adds a CI workflow caller and makes unrelated CLIP/Chinese-CLIP edits. Not the same fix." + "reason": "Unrelated code paths: packaging/version/deps changes versus CI workflow plus CLIP/tokenization and Chinese-CLIP test edits. Not mergeable as one fix." }, { "left": "pull_request:43532", "right": "pull_request:45476", "accept": false, - "reason": "Both are CI/test-oriented bot PRs, but they modify different workflows and different model files; they do not appear to be one mergeable fix for the same concrete problem." + "reason": "Both touch tests/workflows, but they address different areas and no single concrete problem is shared. The overlap is too broad and incidental to treat as one PR." } ] }, @@ -16419,7 +16357,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "d1927a427bd74233de9754680cf0b6aed0a6aa0a468d3bc834533967e7994aed", + "prepared_review_unit_hash": "4dfe0f50a9782a3941dd045116d4b6435a5268d773c763f37e811d89deb67dcb", "component_budget": { "node_count": 2, "item_count": 2, @@ -16441,18 +16379,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two distinct feature PRs: one adds a Promptable Concept Segmentation pipeline, the other adds a Promptable Visual Segmentation pipeline. They share scaffolding and documentation patterns, but target different models, tests, and pipeline code paths, so they should not be deduped.", - "confidence": 0.97, + "summary": "These are two different PRs for two different segmentation pipelines: one for Promptable Concept Segmentation and one for Promptable Visual Segmentation. They share scaffolding and docs/pipeline infrastructure, but they target different models and code paths, so they should not be deduplicated.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: 43612 and 43613 implement different segmentation features with different target model integrations and test coverage.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "43613 is the broader/more central user-facing pipeline addition, but it is still not a duplicate of 43612.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different underlying change: concept segmentation vs visual segmentation. The overlap is mostly shared docs/boilerplate and common pipeline scaffolding, not the same concrete bug or mergeable PR." + "reason": "Different underlying features: PR 43612 adds Promptable Concept Segmentation around SAM3, while PR 43613 adds Promptable Visual Segmentation across EdgeTAM/SAM/SAM2/SAM3-tracker. Similar framework changes do not make them the same fix/change." } ] }, @@ -16482,7 +16420,7 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "da1cab2ac0268d6826c6608a49c75ebcc18cf6fb7bd52935b5ab811f1e036bd8", + "prepared_review_unit_hash": "8d57bb2670f7c55b9ffc5336198ed8d163eb53b1071f9559bf4ca70fc329eb7a", "component_budget": { "node_count": 2, "item_count": 2, @@ -16504,18 +16442,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs both touch `trainer.py` and expose new customization hooks, but they address different trainer behaviors: gradient-accumulation loss scaling vs. signature-column filtering. They should not be clustered as duplicates.", - "confidence": 0.95, + "summary": "Two open PRs in `trainer.py` touch custom-trainer extensibility, but they change different code paths: one adds a loss-scaling hook for gradient accumulation, the other extracts signature-column filtering into an overrideable helper. They are related, not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43651 is the better canonical representative only if one PR must be chosen, because it introduces a concrete behavior change around loss scaling in `training_step`; PR 43663 is a separate refactor/customization point for column filtering.", + "canonical_pr_reason": "PR #43651 is the stronger canonical representative because it introduces a concrete behavioral hook in the training loss path and changes the actual scaling logic, making it the more substantive and user-facing change.", "best_issue_reason": null, - "best_pr_reason": "PR 43651 is the strongest standalone PR in the cluster: it has a focused code-path change with a clear user-facing effect on training loss scaling. PR 43663 is unrelated in behavior and should not be treated as the same change.", + "best_pr_reason": "PR #43651 is the best single representative of the cluster because it targets a specific training bug/behavior with a clear functional outcome, whereas PR #43663 is mainly an internal refactor to expose override points for column filtering.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Both are Trainer-extension PRs in the same file, but they fix different problems and touch different code paths. One controls gradient-accumulation loss scaling; the other refactors signature-column filtering. Not the same underlying change." + "reason": "Both PRs edit `trainer.py` and improve custom trainer extensibility, but they solve different problems: #43651 changes gradient-accumulation loss scaling, while #43663 changes signature-column filtering. They do not appear to be the same underlying bug or mergeable into one PR." } ] }, @@ -16547,7 +16485,7 @@ "pull_request:43656|pull_request:43842", "pull_request:43836|pull_request:43842" ], - "prepared_review_unit_hash": "a64e962a8cf435d17bcbbdebcf17406b707dd686133f8ba1dd24ce17ceba496a", + "prepared_review_unit_hash": "7fcee83d8ac3d649e08cd8f329e5a97b91decb58f81b8d25e699463944f13649", "component_budget": { "node_count": 3, "item_count": 3, @@ -16569,30 +16507,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three PRs touch `src/transformers/cli/serve.py` and are about import-time/type-annotation NameErrors in the CLI, but they are not all equally focused. PR 43842 is the cleanest targeted fix for the `TypeAdapter`/pydantic annotation problem; 43656 is a broader annotation cleanup in the same file, and 43836 is a more expansive variant of the same underlying serve-cli typing issue.", - "confidence": 0.79, + "summary": "Three open PRs all touch `src/transformers/cli/serve.py` and two of them target the same tracked issue. The strongest duplicate pair is the two issue-43824 fixes; the 43656 PR looks related by topic but is not clearly the same concrete change.", + "confidence": 0.77, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43842 is the most directly scoped fix: it explicitly addresses the `TypeAdapter` NameError when pydantic is not installed, uses a TYPE_CHECKING import plus a quoted annotation, and has the clearest title/issue linkage.", + "canonical_pr_reason": "PR #43842 is the best canonical representative: it directly addresses the TypeAdapter import/runtime NameError with a narrow, explicit fix and a clear issue link. It is more focused than #43836 and easier to treat as the primary duplicate target.", "best_issue_reason": null, - "best_pr_reason": "PR 43842 is the best overall PR in the cluster because it is the most specific, minimal, and clearly aligned with the underlying runtime NameError in `serve.py`.", + "best_pr_reason": "PR #43842 best matches the underlying bug report and contains the most precise code-path fix (guarding the `TypeAdapter` import with `TYPE_CHECKING` and stringizing the annotation).", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", - "accept": true, - "reason": "Both PRs appear to fix the same `TypeAdapter`/annotation NameError in the CLI serve path; 43656 is broader, but it is plausibly the same underlying bug and could be consolidated with 43842." + "accept": false, + "reason": "Both mention TypeAdapter NameError in the CLI, but they target different issues and the diffs suggest different fixes; not enough evidence that they are the same concrete change." }, { "left": "pull_request:43836", "right": "pull_request:43842", "accept": true, - "reason": "Both changes target the same `serve.py` import-time typing issue around `TypeAdapter` and pydantic availability, so they look like the same underlying bug rather than separate changes." + "reason": "Same file, same issue target, and both are alternative fixes for the same import-time `TypeAdapter` NameError in `serve.py`; they are plausibly mergeable into one PR." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays mostly conservative: it distinguishes the narrow TypeAdapter/pydantic fix in 43842 from the broader serve.py annotation changes in 43656 and 43836. The soft-edge acceptances are also defensible given the shared file and closely related typing bug surface, with no obvious overstatement." + "feedback": "The summary is grounded in the packet: all three PRs touch `src/transformers/cli/serve.py`, 43836 and 43842 both target issue 43824, and 43656 targets a different issue (43576). The conservative non-match verdict for 43656 vs 43842 is justified. The 43836 vs 43842 pairing is also reasonable as a mergeable same-bugfix candidate, since both address the same TypeAdapter NameError in the same file and the diffs look compatible rather than contradictory." }, "evaluator_used": true, "retried": false, @@ -16782,12 +16720,12 @@ "pull_request:44044|pull_request:44154", "pull_request:44024|pull_request:44076" ], - "prepared_review_unit_hash": "29166f7a1bec9d6a1636a94187f7fc336e5b4cecf57177892f8b828fac2ac039", + "prepared_review_unit_hash": "f9393d5a028c4cadc735050563d400d5f0d3c36796367e5e53d87dad9e2a2da2", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84503, + "serialized_chars": 84501, "estimated_input_tokens": 21126, "estimated_eval_tokens": 42508 }, @@ -16795,7 +16733,7 @@ "node_count": 28, "item_count": 28, "soft_pair_count": 72, - "serialized_chars": 76935, + "serialized_chars": 76933, "estimated_input_tokens": 19234, "estimated_eval_tokens": 38724 }, @@ -16804,312 +16742,354 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of model-specific PRs for the same broad output-tracing/decorator refactor theme, but only a few pairs are true duplicates at the concrete code-path level. The clear duplicate subgroup is the GPT-J/CodeGen refactor, plus a separate GPT-Neo duplicate pair; the rest are just similar standardized-output migrations across different models.", - "confidence": 0.61, - "canonical_issue_reason": "No standalone issue artifact is present in the cluster; several PRs reference tracking issue #43979, but it is not one of the listed items.", - "canonical_pr_reason": "PR 44066 is the strongest representative of the duplicate GPT-J/CodeGen subgroup: it has an explicit issue link, covers the same concrete output-tracing refactor across both model files, and is broader than the narrower follow-up PRs.", - "best_issue_reason": "No issue item is available to serve as a cluster canonical.", - "best_pr_reason": "PR 44066 is the best overall representative among the listed PRs because it most clearly captures the concrete GPT-J/CodeGen output-tracing change rather than just the generic refactor pattern.", + "summary": "This cluster is mostly a set of model-specific standardized output-tracing refactors, not one single duplicate set. The only clear duplicate subgroup is the GPT-J/CodeGen work, with an earlier GPT-Neo pair also looking like a rebased duplicate; the rest are different models and should stay separate.", + "confidence": 0.7, + "canonical_issue_reason": null, + "canonical_pr_reason": "44722 is the best representative for the duplicated GPT-J/CodeGen subgroup: it is the latest and most complete refactor on the same concrete output-tracing code paths, with the same standardized-decorator direction as the earlier GPT-J variants.", + "best_issue_reason": null, + "best_pr_reason": "44722 is the strongest overall PR in the cluster because it cleanly targets the same output-tracing refactor, covers the shared GPT-J/CodeGen implementation, and appears to subsume earlier GPT-J-only variants.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model code paths (CVT/FNet vs GPT-J); same refactor theme only." + "reason": "Different model implementations (CVT/FNet vs GPT-J); same broad theme, but not the same code path or change." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Unrelated models and implementations; only the standardized output-tracing pattern matches." + "reason": "Both are output-tracing refactors, but for unrelated models (CVT/FNet vs DeBERTa-v2)." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Same GPT-J output-tracing refactor family, with overlapping GPT-J file changes and the same tracking issue." + "reason": "Both change GPT-J output tracing in the same file; 44085 reads like a narrower/rebased variant of the same fix." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "Both are output-tracing migrations, but they target different vision models and different concrete implementations." + "reason": "ResNet/RT-DETR ResNet vs EfficientNet: same refactor pattern, but not the same model or bug." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet vs GPT-J/CodeGen; not the same bug or change." + "reason": "EfficientNet is unrelated to GPT-J/CodeGen; same umbrella refactor idea only." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "Different model families and code paths; only the refactor style is shared." + "reason": "GPT-J/CodeGen vs EfficientNet are different code paths and different model families." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 vs DeBERTa V2; same broad migration pattern but not the same concrete change." + "reason": "MobileNetV2 and DeBERTa-v2 are unrelated implementations despite similar output-tracing cleanup." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen vs MGP-STR; unrelated implementations." + "reason": "GPT-J/CodeGen vs MGP-STR are unrelated model changes." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo output-tracing refactor in the same model/file set, so these are duplicate candidates." + "reason": "Same GPT-Neo output-tracing refactor on the same model file; 44068 looks like a later/reworked version of 44018." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen vs MPT; different concrete code paths." + "reason": "GPT-J/CodeGen vs MPT are different model code paths." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen vs GPT-Neo; same refactor theme but different models." + "reason": "GPT-J/CodeGen and GPT-Neo are different model implementations; not the same fix." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen; not the same underlying model change." + "reason": "GPT-Neo vs GPT-J/CodeGen: same standardized-output theme, but different code paths." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet refactor vs GPT-J refactor; unrelated code paths." + "reason": "ResNet is unrelated to GPT-J/CodeGen; different model family and code path." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR vs GPT-J/CodeGen; only the decorator migration pattern overlaps." + "reason": "MGP-STR and GPT-J/CodeGen are unrelated implementations." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet vs GPT-J; no shared concrete bug/change." + "reason": "ResNet vs GPT-J: different model and different underlying change." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet are different model implementations." + "reason": "GPT-Neo and ResNet are unrelated model-specific refactors." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet vs MPT; unrelated changes." + "reason": "ResNet vs MPT are different code paths; not a duplicate." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert/XLM; same style of refactor, different code paths." + "reason": "GPT-Neo vs Flaubert are different models and not the same bug/change." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen; similar output-tracing work but not the same change." + "reason": "GPT-Neo and GPT-J/CodeGen are separate model implementations." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert/XLM; different models." + "reason": "GPT-Neo and Flaubert are unrelated model changes." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT vs GPT-J/CodeGen; not a duplicate." + "reason": "MPT and GPT-J/CodeGen are different model families and code paths." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo vs MGP-STR; unrelated models." + "reason": "GPT-Neo vs MGP-STR: same pattern, different implementation." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo vs GPT-J; different concrete refactors." + "reason": "GPT-Neo and GPT-J are different models; not mergeable as one fix." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo vs MGP-STR; only broad pattern matches." + "reason": "GPT-Neo vs MGP-STR are unrelated model-specific refactors." }, { "left": "pull_request:44071", "right": "pull_request:44085", "accept": false, - "reason": "MPT vs GPT-J/CodeGen; different model code paths." + "reason": "MPT vs GPT-J are different model code paths." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-2 vs MGP-STR; not the same underlying bug/change." + "reason": "GPT2 and MGP-STR are unrelated changes." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J vs MGP-STR; unrelated implementations." + "reason": "GPT-J and MGP-STR are different model implementations." }, { "left": "pull_request:44059", "right": "pull_request:44722", "accept": false, - "reason": "GPT-2 vs GPT-J/CodeGen; similar output-tracing style but different models." + "reason": "GPT2 and GPT-J/CodeGen are different model families; same refactor style only." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT vs Flaubert/XLM; different model families." + "reason": "ViLT and Flaubert/XLM are different model architectures and code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "Segformer and GPT-J/CodeGen are different models; not the same change." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "Segformer and LongT5 are unrelated model-specific refactors." }, { "left": "pull_request:43997", "right": "pull_request:44017", "accept": false, - "reason": "RegNet vs SegFormer; both are refactors of output tracing, but not the same concrete fix." + "reason": "RegNet and Segformer are different code paths." }, { "left": "pull_request:43997", "right": "pull_request:44056", "accept": false, - "reason": "RegNet vs MPNet; unrelated model code paths." + "reason": "RegNet and MPNet are unrelated model implementations." }, { "left": "pull_request:44071", "right": "pull_request:44076", "accept": false, - "reason": "MPT vs ImageGPT; not a duplicate." + "reason": "MPT and ImageGPT are different models; same refactor pattern only." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet vs Flaubert/XLM; same migration style only." + "reason": "TextNet and Flaubert/XLM are unrelated model changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are different model families." }, { "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "MPNet vs LongT5; different concrete implementations." + "reason": "MPNet and LongT5 are unrelated code paths." }, { - "left": "pull_request:44073", - "right": "pull_request:44101", + "left": "pull_request:43997", + "right": "pull_request:44161", "accept": false, - "reason": "VisualBERT vs Flaubert/XLM; different models." + "reason": "RegNet and LongT5 are different model implementations." }, { - "left": "pull_request:44059", - "right": "pull_request:44068", + "left": "pull_request:44074", + "right": "pull_request:44161", "accept": false, - "reason": "GPT-2 vs GPT-Neo; both touch tracing, but not the same code-path problem." + "reason": "TextNet and LongT5 are unrelated changes." }, { - "left": "pull_request:44056", + "left": "pull_request:43997", + "right": "pull_request:44071", + "accept": false, + "reason": "RegNet and MPT are different model families." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "MPNet vs VisualBERT; different refactors." + "reason": "SqueezeBert and VisualBert are different code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are unrelated model-specific refactors." }, { "left": "pull_request:43996", "right": "pull_request:44161", "accept": false, - "reason": "CVT/FNet vs LongT5; unrelated model code." + "reason": "CVT/FNet and LongT5 are different model families." }, { "left": "pull_request:43996", "right": "pull_request:44017", "accept": false, - "reason": "CVT/FNet vs SegFormer; only broad theme overlaps." + "reason": "CVT/FNet and Segformer are separate model changes." }, { "left": "pull_request:44010", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBERT vs DeBERTa V2; different model families." + "reason": "SqueezeBert and DeBERTa-v2 are unrelated implementations." }, { "left": "pull_request:44044", "right": "pull_request:44074", "accept": false, - "reason": "DeBERTa V2 vs TextNet; same style of refactor, not the same fix." + "reason": "DeBERTa-v2 and TextNet are different code paths." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor, same issue target, and overlapping files." + "reason": "Same GPT-J/CodeGen standardized output-tracing refactor on the same files; 44722 is the later, fuller version." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J output-tracing change; the GPT-J file overlap indicates the same underlying refactor." + "reason": "Both cover the same GPT-J output-tracing change on the same code path; 44722 is broader because it also includes CodeGen." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J vs SpeechT5; not the same concrete bug/change." + "reason": "GPT-J and SpeechT5 are different model families." }, { "left": "pull_request:43996", "right": "pull_request:44007", "accept": false, - "reason": "CVT/FNet vs RegNet/ResNet/RT-DETR-ResNet; only a broad output-tracing migration theme is shared." + "reason": "CVT/FNet and ResNet/RT-DETR ResNet are unrelated model changes." }, { "left": "pull_request:43996", "right": "pull_request:44129", "accept": false, - "reason": "CVT/FNet vs SpeechT5; different model code paths." + "reason": "CVT/FNet and SpeechT5 are different model implementations." }, { "left": "pull_request:44044", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa V2 vs SpeechT5; not a duplicate." + "reason": "DeBERTa-v2 and SpeechT5 are unrelated code paths." }, { "left": "pull_request:44044", "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa V2 vs ViTS; same migration pattern only." + "reason": "DeBERTa-v2 and ViTS are different model families." }, { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "FocalNet vs ImageGPT; different model implementations." + "reason": "FocalNet and ImageGPT are different models; same output-handling theme only." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and stays mostly conservative. The accepted soft edges are supported by shared model/file overlap and the same issue target or same concrete GPT-J/GPT-Neo refactor family. The rejected edges are generally justified as same broad output-tracing pattern but different model code paths. No obvious overstatement beyond the evidence in the packet." + "feedback": "The cluster summary is mostly grounded in the packet: it correctly separates the shared-model duplicates from the broader model-specific refactors, and the accepted soft pairs are limited to cases that look mergeable within the same code path (GPT-J/CodeGen and GPT-Neo). The wording stays mostly cautious, with only mild inference about rebasing/later variants." }, "evaluator_used": true, "retried": false, @@ -17144,12 +17124,12 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "f56cd10a74d1a3e4ae2b169152c5f6627919f82cc82f8f828dacb9c6d73a4104", + "prepared_review_unit_hash": "1804fb911aff06d9dd743cca001df0c77a2f50fae53a4a53f2cdce2f99ce49fc", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84503, + "serialized_chars": 84501, "estimated_input_tokens": 21126, "estimated_eval_tokens": 42508 }, @@ -17166,42 +17146,42 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All items are model-specific refactors to standardized output tracing/capture in different architectures, so they are related in theme but not duplicates of the same underlying change.", - "confidence": 0.93, + "summary": "These PRs are all in the same broad migration theme\u2014standardized output tracing / capture_outputs adoption across different model implementations\u2014but they do not look like duplicate fixes or mergeable alternative implementations of the same concrete change. Each PR targets a different model code path, so the soft edges should be rejected.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "If a representative PR is needed, 44073 is the clearest example of the shared refactor pattern, but it only applies to VisualBert and does not subsume the others.", + "canonical_pr_reason": "No issue is present. As a representative PR for this cluster, PR #44010 is the clearest and most self-contained example of the standardized output-collection refactor, with explicit capture_outputs/can_return_tuple integration in one model file.", "best_issue_reason": null, - "best_pr_reason": "44073 is the best single representative of the cluster because its diff cleanly shows the standardized output-tracing migration with the new decorators, but it is not a duplicate anchor for the other model-specific PRs.", + "best_pr_reason": "PR #44010 is a strong cluster representative because its title and diff directly reflect the shared refactor pattern, and the change is compact and focused on a single model implementation.", "soft_edge_verdicts": [ { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "Same broad refactor theme, but different models and files; these are separate model-specific changes, not one mergeable code-path fix." + "reason": "Both are output-tracing refactors, but for different models (VisualBert vs TextNet). They do not fix the same underlying bug or change the same code path." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Both adopt output tracing helpers, but MPNet and VisualBert are unrelated implementations with distinct code paths." + "reason": "Same broad refactor style, but unrelated model implementations (MPNet vs VisualBert) and different internal tracing hooks. Not the same concrete change." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "Shared refactor style only; MPNet and TextNet change different model code and are not the same underlying bug or change." + "reason": "Both migrate to capture_outputs, but they touch different architectures and model-specific tracing logic. Too broad to treat as duplicates." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "Different architectures and forward paths; this is a parallel cleanup, not a duplicate fix." + "reason": "RegNet and SqueezeBert are different code paths and the diffs are model-specific refactors, not alternative fixes for one shared defect." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "Both mention standardized output handling, but SqueezeBert and SpeechT5 are separate model-specific refactors with different semantics." + "reason": "Although both involve output tracing, SqueezeBert and SpeechT5 have distinct implementations and semantics; these are separate model-specific migrations, not one mergeable PR." } ] }, @@ -17284,292 +17264,292 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "fa60728e768668aee18bc2bac807f54f14b2b5441a8ba446a033ae3e52277b53", + "prepared_review_unit_hash": "4e0b0001e227255a5c7243202731f77ec05a5fd1105866914319761a00da76a4", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41328, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41324, + "estimated_input_tokens": 10331, + "estimated_eval_tokens": 20918 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41328, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41324, + "estimated_input_tokens": 10331, + "estimated_eval_tokens": 20918 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a set of thematically similar but separate PRs: each refactors output tracing for a different model family/code path. They are not duplicates of one another, though they share the same general design direction. No issue is present; PR 44003 is the best representative because it is the broadest concrete refactor in the set.", - "confidence": 0.96, + "summary": "This is a cluster of model-specific output-tracing refactors that share a common pattern but are not duplicate fixes or mergeable changes to the same code path.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44003 is the most representative umbrella change: it updates two related model files and introduces the output-capturing plumbing, making it the broadest concrete example of the cluster's theme.", + "canonical_pr_reason": "No issue is present; if one PR must represent the cluster, 44025 is a good canonical example because it is a clean, self-contained tracing refactor across two closely related Depth Anything files.", "best_issue_reason": null, - "best_pr_reason": "44003 best fits the cluster as a canonical representative because it is more substantial than the single-model refactors and sits closest to the shared output-tracing refactor pattern.", + "best_pr_reason": "44025 is the strongest representative PR: focused, easy to understand, and directly shows the output-tracing conversion pattern without depending on unrelated model code.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different model families and different forward paths; same refactor theme is not enough to treat them as the same change." + "reason": "Different model implementations and code paths; same refactor theme only." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Unrelated model implementations; both are output-tracing cleanups, but they fix different code paths." + "reason": "Separate model families; similar output-tracing cleanup, not the same change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Separate model-specific refactors with no shared concrete bug or code path." + "reason": "Different model code paths; only the refactor style overlaps." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different model stacks; one is timm_backbone, the other is mamba/falcon_mamba, so these are not the same underlying change." + "reason": "Unrelated model families; not the same underlying bug or patch." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Both touch tracing utilities, but they apply to different architectures and independent forward logic." + "reason": "Different architectures and files; shared tracing refactor pattern only." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "MobileNetV1 and UnivNet are unrelated model paths; similar refactor pattern only." + "reason": "Different models; both alter output handling but not the same concrete fix." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different models and different output contracts; not the same bug or change." + "reason": "Different model implementations; too broad to be duplicates." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Both are refactors for tracing, but they affect distinct model families and code paths." + "reason": "Different model families and files; not one mergeable fix." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "No shared concrete fix; these are separate model-specific output-tracing updates." + "reason": "Separate encoder/decoder code path versus MobileNet; same refactor theme only." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Different encoder-decoder model code, not one mergeable fix for the same problem." + "reason": "Different model-specific refactors, not the same underlying issue." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Vision-text dual encoder and UnivNet are unrelated implementations; the similarity is only in the refactor style." + "reason": "Distinct model code paths; similar tracing refactor intent only." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Distinct model code paths with no evidence of the same underlying bug." + "reason": "Different model families and implementations." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and different output semantics; not duplicate changes." + "reason": "Unrelated models; no shared concrete bug or change." }, { "left": "pull_request:44000", "right": "pull_request:44026", "accept": false, - "reason": "These affect unrelated model architectures, so they should not be merged as one PR." + "reason": "Different model implementations; same broad refactor pattern only." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "Unrelated model implementations; same tracing refactor theme is too broad." + "reason": "Different model-specific changes; not mergeable as one PR." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and forward methods; not the same concrete issue." + "reason": "Separate models and code paths; not duplicates." }, { "left": "pull_request:44001", "right": "pull_request:44026", "accept": false, - "reason": "Separate model-specific refactors with no shared fix target." + "reason": "Different model families; same style of refactor, not same fix." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "Different architectures and code paths; no evidence of a common underlying bug." + "reason": "Unrelated models and code paths; only the output-tracing motif overlaps." }, { "left": "pull_request:44002", "right": "pull_request:44025", "accept": false, - "reason": "Both are output-tracing refactors, but for different models and different forward behavior." + "reason": "Different model implementations; not the same concrete change." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Unrelated model families; this is not one concrete fix spanning both PRs." + "reason": "Different model families and code paths." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different model code paths and different tracing updates; not duplicates." + "reason": "Separate model-specific refactors; not duplicate fixes." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Vision-text dual encoder and mamba/falcon_mamba are unrelated; same broad subsystem is insufficient." + "reason": "Different architectures and files; too broad to merge as one PR." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "Different model families and different refactors, even though both mention output handling." + "reason": "Different model code paths; only broad tracing refactor similarity." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Mamba/falcon_mamba output tracing is unrelated to CodeGen's cache/output signature change." + "reason": "Different model families and concrete implementation details." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Separate model-specific changes; no shared concrete bug or mergeable fix." + "reason": "Different models and forward paths; not the same bug or change." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "Unrelated models and different output contracts; not the same underlying change." + "reason": "Separate model code paths; overlap is only in tracing cleanup." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "These target different model implementations and should remain separate." + "reason": "Different model implementations; not mergeable into one PR." }, { "left": "pull_request:44000", "right": "pull_request:44004", "accept": false, - "reason": "Different code paths and different model families; only the output-refactor theme overlaps." + "reason": "Different models and changes; same general refactor theme only." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Depth Anything and VisionEncoderDecoder are unrelated models; no common concrete bug." + "reason": "Different model families; no shared concrete bug." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Different encoder/decoder stacks and separate tracing updates." + "reason": "Different encoder/decoder stacks; not the same fix." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Depth-anything depth estimation and SuperPoint feature extraction are distinct code paths." + "reason": "Distinct model implementations and output paths." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "Separate model families; the shared output-tracing wording does not indicate a duplicate." + "reason": "Different model architectures; same refactor motif only." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Different models and different forward signatures, so they are not the same PR in disguise." + "reason": "Separate model-specific tracing changes; not duplicates." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "VisionEncoderDecoder and RWKV are unrelated; no same underlying fix." + "reason": "Different model code paths and files; not the same underlying issue." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "Different model stacks and different output handling changes." + "reason": "Different encoder/decoder models; too broad to merge." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "Separate model-specific refactors with no shared bug target." + "reason": "Distinct model implementations; only similar refactor style." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "DPR and VisionEncoderDecoder are different code paths; not mergeable as one duplicate change." + "reason": "Different model families and output handling code." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "Speech/Vision encoder-decoder code is unrelated to SuperPoint's model forward path." + "reason": "Different model-specific changes; not the same patch." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Different model families and different tracing mechanics; not the same issue." + "reason": "Separate model families and code paths." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Separate model implementations; the common theme is too broad to call them duplicates." + "reason": "Different models; overlap is only in tracing refactor intent." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "SuperPoint and RWKV are unrelated architectures, so these are distinct changes." + "reason": "Different architectures and files; not the same fix." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "Different model code paths and different output semantics; not duplicate PRs." + "reason": "Distinct model implementations; not mergeable as one PR." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "RWKV and DPR are unrelated models; same tracing/return-shape theme is insufficient for a duplicate." + "reason": "Different model families and code paths; same theme only." } ] }, @@ -17602,52 +17582,52 @@ "pull_request:44662|pull_request:45075", "pull_request:45075|pull_request:45401" ], - "prepared_review_unit_hash": "0631109f60244d3876ba91d78598f72b9e7c08aad17176061ff29bb128588a41", + "prepared_review_unit_hash": "2d0f0fc5315955fd6aa62002914147a970807a370e18e8c6a1ffb515d2a9293d", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10914, - "estimated_input_tokens": 2729, - "estimated_eval_tokens": 5714 + "serialized_chars": 10912, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10914, - "estimated_input_tokens": 2729, - "estimated_eval_tokens": 5714 + "serialized_chars": 10912, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three PRs are unrelated model-addition implementations for different models (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). They share boilerplate infrastructure changes, but not the same underlying bug or change, so none should be deduplicated.", + "summary": "These are three separate model-addition PRs that share Transformers integration boilerplate but implement different model families and code paths, so they should not be treated as duplicates.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR for duplicate-triage purposes: each PR introduces a different model and distinct codepaths, so they should remain separate.", + "canonical_pr_reason": "PR 45075 is the most complete standalone submission in the cluster: it includes model, processing, image-processing, conversion, auto-mapping, docs, and tests. Even so, it is specific to DeepSeek-OCR-2 and not a duplicate of the other model PRs.", "best_issue_reason": null, - "best_pr_reason": "PR 45075 is the strongest standalone representative only in the sense that it is a complete model-addition PR with conversion code, docs, and tests; however, it is still not duplicate-equivalent to the others.", + "best_pr_reason": "PR 45075 is the strongest representative PR here because it has the broadest end-to-end implementation and validation coverage, but it still addresses a different model than the others.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different models and different implementations: PenguinVL vs Voxtral TTS. Shared auto/docs files are generic boilerplate, not the same underlying change." + "reason": "PenguinVL and Voxtral TTS are unrelated model additions for different tasks (vision-language vs TTS); they only share standard integration files, not the same underlying change." }, { "left": "pull_request:44662", "right": "pull_request:45075", "accept": false, - "reason": "Different model additions (PenguinVL vs DeepSeek-OCR-2) with separate configs, processors, and model code; only generic framework files overlap." + "reason": "PenguinVL and DeepSeek-OCR-2 are different model implementations with distinct architectures and targets; shared boilerplate files do not make them the same change." }, { "left": "pull_request:45075", "right": "pull_request:45401", "accept": false, - "reason": "DeepSeek-OCR-2 and Voxtral TTS are distinct model integrations with different codepaths and assets; overlap is limited to shared registration/docs plumbing." + "reason": "DeepSeek-OCR-2 and Voxtral TTS modify different concrete model code paths and would not plausibly be merged into one PR; they are separate additions, not duplicates." } ] }, @@ -17677,7 +17657,7 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "78113aea60c0df532f485743f9954f0dea987e00f69c1b46516821e68e9f5bcf", + "prepared_review_unit_hash": "10791feb36235b332c347e0259d1d07798c1ed67a48d1abde9a2c6601f689eaa", "component_budget": { "node_count": 2, "item_count": 2, @@ -17699,18 +17679,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related by the broad theme of fusion, but they implement different layers of the stack: one adds a generic module-fusion API, the other wires n-to-1 kernel fusion through `KernelConfig` and model loading. They do not look like the same underlying change.", - "confidence": 0.89, + "summary": "These are related by the general theme of fusion, but they are not duplicates: PR 44979 adds a generic Module Fusion API, while PR 45363 implements a specific n-to-1 KernelConfig-based kernel fusion path with integration changes and an example.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45363 is the better representative because it delivers the more concrete end-user feature and touches the actual loading/integration path for kernel fusion; PR 44979 is a lower-level API scaffold for module fusion.", + "canonical_pr_reason": "PR 44979 is the more foundational/general artifact: it introduces the standalone module_fusion API surface and tests, whereas PR 45363 is a separate, concrete application of fusion built around KernelConfig and runtime integrations.", "best_issue_reason": null, - "best_pr_reason": "PR 45363 is the strongest standalone PR in this cluster since it implements a specific kernel-fusion workflow end to end, whereas PR 44979 is a foundational API addition that is not the same code-path change.", + "best_pr_reason": "PR 45363 is the more complete end-user-facing implementation, but it is still distinct from 44979; if one needs a representative PR for the broader fusion work, 44979 is the cleaner canonical choice.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Both are about fusion, but 44979 introduces a generic module-fusion API and 45363 implements n-to-1 kernel fusion via `KernelConfig`; they target different mechanisms and would not plausibly be merged as one PR." + "reason": "Different underlying changes. 44979 adds a generic module fusion API in a new module; 45363 wires up n-to-1 kernel fusion through KernelConfig and integration/modeling changes. They are adjacent, but not the same code-path fix and would not be merged as one PR." } ] }, @@ -17740,12 +17720,12 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "9a6e7a29d63945f326f4807b641b922079a0ba74b1aa8dcbb1fbe0b4b4032cd7", + "prepared_review_unit_hash": "6d955ea1fdf1cd9524b976f7b1bd68da43b1c1e5a7521e1552eeae35f878d4a6", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -17753,7 +17733,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -17762,18 +17742,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs fix the same `_patch_mistral_regex` AttributeError in `src/transformers/tokenization_utils_tokenizers.py` for issue #45081; PR 45317 is the stronger representative because it includes test coverage and a fuller fix context.", - "confidence": 0.97, - "canonical_issue_reason": "No issue artifact is present in this cluster; both PRs explicitly target issue #45081.", - "canonical_pr_reason": "PR 45317 is the better canonical PR because it appears to be the more complete fix, adds tests, and still addresses the same concrete AttributeError in `_patch_mistral_regex`.", + "summary": "Both pull requests address the same AttributeError in `_patch_mistral_regex` by changing the same pre-tokenizer access in `tokenization_utils_tokenizers.py`; PR 45317 is the more complete version because it also adds tests.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the better canonical PR because it implements the same fix as 45086 and adds test coverage, making it the more complete representation of the change.", "best_issue_reason": null, - "best_pr_reason": "PR 45317 is the best overall representative of the cluster because it fixes the same bug as 45086 but with added test coverage and clearer validation of the change.", + "best_pr_reason": "PR 45317 is the best overall PR because it fixes the same concrete bug as 45086 and includes an added test file, indicating a more finalized patch.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying bug and same target issue (#45081), same code path, and the later PR is a superset with tests; these are duplicates/overlapping fixes." + "reason": "Same underlying bug fix in the same code path, same target issue, and the second PR is essentially the first fix plus tests; they could plausibly be merged into one PR." } ] }, @@ -17803,12 +17783,12 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "87c7f4ae45a2627f5440557195465b60c2b382ef6592783890d3d95e8e1d3bee", + "prepared_review_unit_hash": "cfb8b7461ef565b40654a7cdf0a087eee24689338dc543f480a26bca8841f7ab", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6066, + "serialized_chars": 6065, "estimated_input_tokens": 1517, "estimated_eval_tokens": 3290 }, @@ -17816,7 +17796,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6066, + "serialized_chars": 6065, "estimated_input_tokens": 1517, "estimated_eval_tokens": 3290 }, @@ -17825,18 +17805,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs touch the same NemotronH area but are not the same change: one is a narrow refactor/conversion-mapping update, while the other splits NemotronH into dense/sparse variants with docs and tests. They should not be merged as duplicates.", + "summary": "These are related NemotronH refactors, but they are not the same change: #45115 is a narrow inheritance/mapping cleanup, while #45569 is a broader model split introducing NemotronHDense and NemotronHSparse plus docs/tests. They overlap on shared core files, but they do not look mergeable into one PR without collapsing distinct goals.", "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the better representative PR for this cluster because it implements the broader, concrete model-structure change (NemotronH dispatcher plus dense/sparse variants) and includes docs/test updates, whereas #45115 is a narrower inheritance/refactoring/conversion-mapping change.", + "canonical_pr_reason": "#45569 is the broader, more authoritative architectural change: it introduces the new NemotronHDense/NemotronHSparse structure, updates auto mappings, docs, and tests, and reads like the main follow-up model refactor rather than just a conversion-mapping cleanup.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 is the strongest standalone PR candidate: it has the larger, more complete implementation footprint and directly addresses the model split across code, docs, and tests. PR #45115 appears to be an earlier, narrower refactor in the same subsystem.", + "best_pr_reason": "#45569 is the best representative of the cluster because it covers the larger, user-facing model refactor and likely subsumes the narrower NemotronH inheritance cleanup in scope, even though the two PRs are not duplicates.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Not the same underlying change. #45115 focuses on NemotronH inheritance/refactoring and conversion key renames, while #45569 introduces dedicated NemotronH dense/sparse model variants and related docs/tests. They overlap in subsystem, not in concrete fix." + "reason": "Both touch NemotronH internals, but #45115 is focused on conversion mapping and inheritance cleanup, while #45569 adds separate dense/sparse model packages, docs, and tests. They address related architecture work, not the same concrete code-path fix/change." } ] }, @@ -17866,12 +17846,12 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "6a6d7d25afcb2192525944e38142d9fc83a2a30e56d44a1a9ac7c7e37f40aa9e", + "prepared_review_unit_hash": "5986cbc6945d4170982623bf046289a50c3fc4820d898aeac2f04160437713b9", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5398, + "serialized_chars": 5397, "estimated_input_tokens": 1350, "estimated_eval_tokens": 2956 }, @@ -17879,7 +17859,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5398, + "serialized_chars": 5397, "estimated_input_tokens": 1350, "estimated_eval_tokens": 2956 }, @@ -17888,18 +17868,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs in the same heterogeneity effort, but they are not duplicates: one introduces per-layer config plumbing, while the other extends model/cache/masking behavior to use heterogeneous configs.", - "confidence": 0.92, + "summary": "These are related but not duplicates: one PR adds heterogeneous per-layer configuration, the other adds heterogeneous modeling/cache support that builds on that config machinery.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45332 is the broader downstream change: it builds heterogeneous model support across cache/modeling paths and appears to depend on the per-layer config plumbing added separately.", + "canonical_pr_reason": "45332 is the broader umbrella PR: it covers heterogeneous modeling behavior and includes the new per-layer config plumbing it depends on, whereas 45333 is the narrower config-only slice.", "best_issue_reason": null, - "best_pr_reason": "PR #45332 is the best representative because it spans the actual runtime behavior changes, not just the config data model.", + "best_pr_reason": "45332 is the best representative of the cluster because it spans the larger end-to-end change and is the more central change path for heterogeneous model support.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related but distinct scopes: #45333 adds heterogeneous config serialization/representation, while #45332 changes model/cache behavior to consume that config. They share infrastructure, but they are not the same concrete bug/change and would not be merged as one PR." + "reason": "Related scope, but not the same concrete code change: 45333 adds config serialization/per-layer config support, while 45332 adds modeling/cache behavior for heterogeneous layers. They are complementary, not duplicative, and would not plausibly be merged as one PR without collapsing two distinct changes." } ] }, @@ -17929,7 +17909,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "4c04bf8c8866cc0de53a932cd8b3784b3b7349d4d0d04d581a6fce7e4ed58f2a", + "prepared_review_unit_hash": "298519b048c5901fc6a3fb69bd24e622f1d22a0776d739189a1a947a53945ed4", "component_budget": { "node_count": 2, "item_count": 2, @@ -17951,18 +17931,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are typing-focused, but they address different scopes: #45415 is a broad type-checking sweep across many Transformers files, while #45425 adds/adjusts typing protocols and helpers for specific model/PEFT-related paths. They are related, but not the same underlying change.", - "confidence": 0.82, + "summary": "Both items are typing-related pull requests, but they address different scopes: one is a broad repo-wide type-checking sweep, the other adds/adjusts specific typing protocols and related model utilities. They are not the same underlying change.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45415 is the broader, more representative typing cleanup across the codebase and is the better anchor for this cluster.", + "canonical_pr_reason": "PR 45425 is the better canonical representative because it is the more focused, concrete typing change centered on `_typing.py` and related model utility compatibility, whereas 45415 is a broad type-checking cleanup across many files.", "best_issue_reason": null, - "best_pr_reason": "PR #45415 is the strongest cluster representative because it spans many affected modules and appears to be the main type-checking pass rather than a narrow follow-up.", + "best_pr_reason": "PR 45425 is the most suitable overall because it targets a specific code-path/typing compatibility problem and is easier to reason about as a standalone change; 45415 is a wider type-checking sweep rather than the same fix.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Reject: both are typing-related, but #45415 is a broad repo-wide type-checking sweep and #45425 is a narrower typing/protocol update for modeling_utils and PEFT-like interfaces. They do not appear to fix the same concrete code-path problem and would not plausibly merge as one PR." + "reason": "Reject: 45415 is a broad typing sweep with many unrelated file edits and pyproject typing-ignore changes, while 45425 is a targeted typing-protocol/update PR. They share a theme but not the same concrete bug or mergeable change." } ] }, @@ -17992,7 +17972,7 @@ "soft_pairs": [ "pull_request:45487|pull_request:45615" ], - "prepared_review_unit_hash": "42d131bd162c030fab2d14338a06f5b5a1867a4446e12a624575166f4c4d4aaf", + "prepared_review_unit_hash": "a671d878d8083c012e599652ea63b278cdaf38554fbd255017e00bc2d723da64", "component_budget": { "node_count": 2, "item_count": 2, @@ -18014,18 +17994,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two pull requests are not duplicates: one fixes a model-parallel device placement bug in AltCLIP/ChineseCLIP code paths, while the other only updates Qianfan OCR XPU test expectations. They touch different models, different files, and different kinds of changes.", + "summary": "These are unrelated pull requests: one fixes a real model-parallel device placement bug in AltCLIP/ChineseCLIP code, while the other only adds XPU test expectations for qianfan_ocr.", "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45487 is the better canonical PR because it contains the actual code fix for a concrete runtime bug, while #45615 is only a test expectation update for a different model and hardware path.", + "canonical_pr_reason": "PR 45487 is the stronger canonical PR because it contains the substantive code fix: moving buffered token type ids to the correct device in model forward paths, with corresponding model coverage across several affected architectures.", "best_issue_reason": null, - "best_pr_reason": "PR #45487 is the strongest representative of the cluster\u2019s substantive work since it addresses a real implementation issue; #45615 is not a good representative because it is test-only and unrelated in underlying behavior.", + "best_pr_reason": "PR 45487 is the best overall representative of the cluster because it addresses an actual runtime bug in model code, whereas PR 45615 is limited to test data updates and is not the same kind of change.", "soft_edge_verdicts": [ { "left": "pull_request:45487", "right": "pull_request:45615", "accept": false, - "reason": "Different models and different changes: #45487 fixes device handling in AltCLIP/ChineseCLIP implementation code, while #45615 only adds XPU test expectations for Qianfan OCR. They do not appear to fix the same concrete bug or change." + "reason": "Reject: they do not fix the same underlying bug or change. 45487 patches device placement in model forward code for parallel execution; 45615 only adds XPU-specific expected outputs in tests for a different model." } ] }, @@ -18055,7 +18035,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "332037c32dd45aae99430e2709229819b4e9d019d35805cfdfc91f94c7f74ad1", + "prepared_review_unit_hash": "63e01767042f4ae9ab394af5a1beea79583758b7b6f16ac2fe35b8b1340898ac", "component_budget": { "node_count": 2, "item_count": 2, @@ -18077,18 +18057,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both expert-parallel loading-related PRs, but they fix different concrete bugs: one targets config serialization and a has_ep helper for Accelerate/ZeRO-3 loading, while the other prevents expert tensors from being overwritten during FSDP2/distributed loading. Not duplicates.", - "confidence": 0.95, + "summary": "These are related EP/distributed-loading PRs, but they fix different failure modes and touch different code paths: one is about accelerate-launch/ZeRO-3 config handling, the other about FSDP2 expert tensors being overwritten during load. They should not be deduplicated.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "#45662 is the more concrete expert-parallel loading fix, centered on preserving loaded expert parameters during distributed post-processing. #45548 is a separate compatibility patch around config serialization and an EP detection helper.", + "canonical_pr_reason": "No canonical PR for duplication: the PRs target different EP bugs and are not merge-equivalent.", "best_issue_reason": null, - "best_pr_reason": "#45662 is the best representative of the cluster because it addresses a specific runtime corruption bug in the loading path, with a clearly defined code-path and fix.", + "best_pr_reason": "45662 is the stronger standalone PR because it addresses a specific, concrete expert-overwrite bug with a direct code-path fix.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Both concern EP loading, but they fix different problems and touch different mechanisms: #45548 adds config cleanup and a has_ep property for Accelerate/ZeRO-3 loading, while #45662 fixes distributed expert parameters being overwritten via post-shard/broadcast handling. Not the same underlying change." + "reason": "Different underlying bugs. 45548 strips distributed_config and adds has_ep for accelerate-launch/ZeRO-3 loading, while 45662 post-wraps sharded params to prevent expert tensors being overwritten under FSDP2. Shared EP plumbing is not enough to treat them as the same change." } ] }, @@ -18118,12 +18098,12 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "c1c991545d92adab2602115e6794bf26917422e663ef07403f13d7b0d8b4cbce", + "prepared_review_unit_hash": "13152e7f6d461b0ba41b83b184fbf1299fcd29dbd99ff0ed17d45f315a85b29b", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4589, + "serialized_chars": 4590, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -18131,7 +18111,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4589, + "serialized_chars": 4590, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -18140,18 +18120,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs target the same xdist/captured_info collision problem; #45645 appears to subsume #45639 by adding the xdist-safe logging fix plus CI log collection updates.", + "summary": "Both PRs target the same xdist/captured_info logging area, but PR 45645 is broader and not a strict duplicate of 45639; it adds CI log aggregation and other artifact-handling changes beyond the core testing_utils fix.", "confidence": 0.84, "canonical_issue_reason": null, - "canonical_pr_reason": "#45645 is the more complete fix: it includes the xdist-safe captured_info handling and the downstream CI workflow changes needed to preserve and aggregate the logs.", + "canonical_pr_reason": "PR 45645 is the more representative change in this cluster because it contains the core xdist-safe captured_info fix plus the CI workflow changes needed to consume multiple per-worker files.", "best_issue_reason": null, - "best_pr_reason": "#45645 is the best representative because it covers the shared bug and the surrounding CI handling, while #45639 is a narrower subset of the same fix.", + "best_pr_reason": "PR 45645 is the best overall representative because it addresses the underlying collision problem and also updates the downstream workflow to preserve debug output end-to-end.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", - "accept": true, - "reason": "Same underlying xdist captured_info collision bug; #45645 extends the same code-path fix into workflow/log aggregation, so these could plausibly be merged as one change." + "accept": false, + "reason": "Related but not the same change set: 45639 only makes the testing_utils log file xdist-safe, while 45645 also changes the CI workflow (and related artifact handling) to read multiple captured_info files. They share a subsystem and issue target, but they are not strict duplicates and would not be merged as one PR without losing scope." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index 2456a5ae5412f803369bad828e282560788132af..0f38dcf545f37518eb2edb4513dbfafd045e07e8 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260429t180023z", + "analysis_id": "hybrid-model-20260430t000042z", "archived_artifacts": { - "hybrid": "snapshots/20260429T180023Z/analysis-runs/hybrid-model-20260429t180023z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260429T180023Z/analysis-runs/hybrid-model-20260429t180023z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-04-29T18:07:39Z", + "published_at": "2026-04-30T00:10:18Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260429T180023Z", + "snapshot_id": "20260430T000042Z", "variant": "hybrid" } diff --git a/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..41e0512ee8312d893dae32a940a63de47dcbf125 --- /dev/null +++ b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json @@ -0,0 +1,2051 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260430T000042Z", + "generated_at": "2026-04-30T00:10:16Z", + "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-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-43824-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43824.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43824, + "canonical_pr_number": 43836, + "issue_numbers": [ + 43824 + ], + "pr_numbers": [ + 43836, + 43842 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 43836, + "right_pr_number": 43842, + "code_similarity": 0.332, + "size_similarity": 0.088, + "file_overlap": 1.0, + "area_overlap": 0.033, + "patch_similarity": 0.017, + "shared_filenames": [ + "src/transformers/cli/serve.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/serve.py", + "left_ranges": [ + [ + 11, + 18 + ], + [ + 359, + 429 + ], + [ + 584, + 590 + ], + [ + 1892, + 1910 + ], + [ + 1917, + 1923 + ] + ], + "right_ranges": [ + [ + 54, + 61 + ], + [ + 587, + 593 + ] + ] + } + ] + } + ] + }, + { + "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": 0.854, + "size_similarity": 0.901, + "file_overlap": 0.917, + "area_overlap": 0.842, + "patch_similarity": 0.693, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "docs/source/en/model_doc/minicpm3.md", + "src/transformers/models/__init__.py", + "src/transformers/models/auto/auto_mappings.py", + "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": [ + [ + 711, + 718 + ] + ] + }, + { + "filename": "docs/source/en/model_doc/minicpm3.md", + "left_ranges": [ + [ + 1, + 67 + ] + ], + "right_ranges": [ + [ + 1, + 45 + ] + ] + }, + { + "filename": "src/transformers/models/__init__.py", + "left_ranges": [ + [ + 250, + 256 + ] + ], + "right_ranges": [ + [ + 248, + 254 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 331, + 337 + ] + ], + "right_ranges": [ + [ + 329, + 335 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 285, + 291 + ], + [ + 700, + 706 + ], + [ + 1298, + 1304 + ] + ], + "right_ranges": [ + [ + 283, + 289 + ], + [ + 696, + 702 + ], + [ + 1292, + 1298 + ] + ] + }, + { + "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-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-45561-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 45561, + "canonical_pr_number": 45645, + "issue_numbers": [ + 45561 + ], + "pr_numbers": [ + 45639, + 45645 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 45639, + "right_pr_number": 45645, + "code_similarity": 0.532, + "size_similarity": 0.619, + "file_overlap": 0.5, + "area_overlap": 0.543, + "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": [ + [ + 3529, + 3536 + ], + [ + 3753, + 3787 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 86 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + } + ] + } + ] + }, + { + "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-43824-3", + "canonical_pr_number": 43836, + "duplicate_pr_numbers": [ + 43842 + ], + "target_issue_number": 43824, + "reason": "PRs in cluster-43824-3 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43979-11", + "canonical_pr_number": 44007, + "duplicate_pr_numbers": [ + 43996, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "target_issue_number": 43979, + "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45561-3", + "canonical_pr_number": 45645, + "duplicate_pr_numbers": [ + 45639 + ], + "target_issue_number": 45561, + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference evidence." + } + ], + "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/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..f9edecb733ece301ebfd6e9db2e7a9d00e0970e4 --- /dev/null +++ b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,18146 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260430T000042Z", + "generated_at": "2026-04-30T00:10:16Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 14, + "review_unit_index": 1, + "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": "653a36cc0fa13814dd047e575350566e83973ed7e845de8e6418d49b2d89e3c5", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues are only loosely related by Hugging Face model/tokenizer usage, but they describe different problems: tokenization mismatch, private-model auth/download messaging, and a failure loading bert-base-uncased. I would not merge any of the soft pairs as duplicates.", + "confidence": 0.82, + "canonical_issue_reason": "issue:16618 is the broadest and most user-visible report in the set, but the three issues do not appear to be duplicates of one another.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:16618 is the best representative only in the sense that it is the clearest general model-loading failure; the others are narrower and about different code paths.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tokenization behavior in Tapas versus TensorFlow code is a different underlying bug than failing to load bert-base-uncased." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "A bad error message for unauthenticated private-model downloads is an auth/download UX issue, not the same bug as loading a public model failing." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41093", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44661", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45070", + "issue:45081", + "issue:45084", + "issue:45161", + "issue:45237", + "issue:45362", + "issue:45464", + "issue:45507", + "issue:45685" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:45507", + "issue:44292|issue:45237", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:44509|issue:44512", + "issue:35545|issue:36010", + "issue:43638|issue:44805", + "issue:43596|issue:44805", + "issue:44805|issue:45161", + "issue:44805|issue:44918", + "issue:44661|issue:44805", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43975|issue:44457", + "issue:43039|issue:45161", + "issue:44292|issue:44457", + "issue:34928|issue:44805", + "issue:24643|issue:30277", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:39290|issue:45081", + "issue:43828|issue:45161", + "issue:43425|issue:44292", + "issue:44918|issue:45161", + "issue:34928|issue:41720", + "issue:43716|issue:45685", + "issue:43541|issue:43716", + "issue:45507|issue:45685", + "issue:43828|issue:44292", + "issue:43716|issue:44292", + "issue:44292|issue:44918", + "issue:41762|issue:43638", + "issue:43039|issue:44918", + "issue:43827|issue:44512", + "issue:43716|issue:44560", + "issue:43828|issue:45685", + "issue:44918|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:43039|issue:44805", + "issue:34928|issue:36010", + "issue:43901|issue:44512", + "issue:41762|issue:43596", + "issue:39290|issue:41720", + "issue:43866|issue:44863", + "issue:43039|issue:44560", + "issue:43854|issue:43866", + "issue:41093|issue:43638", + "issue:39290|issue:43531", + "issue:43425|issue:45070", + "issue:45362|issue:45464", + "issue:43404|issue:43866", + "issue:43792|issue:45070", + "issue:45464|issue:45685", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:36010|issue:39290", + "issue:44841|issue:45084", + "issue:41093|issue:41762" + ], + "prepared_review_unit_hash": "f2639e21844f4bc54c73177296a8ed4ce4b51c9e01eafc039cae3debc36da180", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 67, + "serialized_chars": 47120, + "estimated_input_tokens": 11780, + "estimated_eval_tokens": 23816 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous, but there is a small duplicate set around DeepSpeed ZeRO-3 loading failures with `index 0 is out of bounds for dimension 0 with size 0`. Most other pairs are only loosely related by broad subsystem or error family and should stay separate.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 43596 is the cleanest representative of the shared ZeRO-3 loader failure; 43638 and 41762 look like model-specific variants of the same underlying `index 0` out-of-bounds bug.", + "canonical_pr_reason": null, + "best_issue_reason": "43596 is the best issue anchor because it captures the common DeepSpeed ZeRO-3 initialization failure in a generic way, making it the most suitable duplicate target for the accepted variants.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different Qwen3.5 failures: missing config attribute vs streaming chat/completions inference error." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different backend/model problems: AMD GPU support for GPT-OSS vs MPS int histogram error in MoE." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different model families and failure modes: Mistral-3 dtype mismatch vs GPT-OSS on AMD GPUs." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Both are dtype/backend complaints, but they affect different models and code paths." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "AMD GPU runtime issue vs GraniteMoEHybrid calling an invalid method; not the same bug." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Qwen3 NVFP4 runtime failure is unrelated to the GPT-OSS AMD GPU issue." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collective mismatch is unrelated to ModernBERT ONNX export." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export error and FineGrainedFP8Config failure are different model/tooling issues." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "ModernBERT export to ONNX is unrelated to Qwen3 auto-device-map CUDA assert." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are stale docs problems, but they reference different removed features and do not look like the same underlying change." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ONNX export error vs `GenerationMixin` import breakage are separate problems." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both involve shape/index errors under ZeRO-3, but the specific failure messages and contexts differ enough to avoid deduping." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "ZeRO-3 BertModel index-out-of-bounds is not the same as the mask-shape mismatch report." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Mask/tensor shape mismatch and GPT-OSS tensor-parallel failure are unrelated." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer mapping bug is unrelated to the mask/tensor shape mismatch issue." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "BarkProcessor voice preset and ModernBERT ONNX export are different subsystems and failures." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Collective mismatch during ZeRO-3 is a different failure from the ZeRO-3 index-out-of-bounds loader bug." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 loader/index-out-of-bounds bug on BertModel; the wording differs but the failure is the same." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Incorrect detokenization and LoRA save/reload mismatch are distinct model-output issues." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross-entropy path and GPT-OSS tensor parallel issue are not the same bug." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen NVFP4 runtime failure is unrelated to LoRA merge/save inconsistency." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpointing/FSDP recomputation mismatch is a different shape bug than the ZeRO-3 loader issue." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "DeepSpeed training `'weight' must be 2-D` is unrelated to the collective mismatch report." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 index-out-of-bounds and GPT-OSS tensor-parallel failure do not share the same code path." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "FSDP/activation checkpointing recomputation mismatch is unrelated to the `compute_loss` signature change." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different training/load failures: ZeRO-3 BertModel loading vs GPT-OSS tensor-parallel behavior." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Mistral-3 dtype mismatch and GraniteMoEHybrid invalid method are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3 sliding-window attribute error is unrelated to the Mistral tokenizer patch crash." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "Qwen3.5 embedding unpacking with TRL is unrelated to GPT-OSS MoE tensor-parallel behavior." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 loader bug and GPT-OSS tensor-parallel bug are different code paths." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3 `index 0` out-of-bounds loader failure, just on a different model family." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger Kernel cross-entropy call and TRL Qwen3.5 embedding unpacking are unrelated." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs issues, but they concern different removed v5 features and look like separate cleanup tasks." + }, + { + "left": "issue:43866", + "right": "issue:43854", + "accept": false, + "reason": "Checkpoint corruption and unit-test load failure for GLM-4.7-Flash are different issues." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Mask/index length mismatch and ZeRO-3 loader index-out-of-bounds are different failure modes." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Gemma3 sliding-window attribute error is unrelated to Qwen3-MoE sliding_window behavior." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch version incompatibility and pydantic `PretrainedConfig` field breakage are unrelated." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Qwen3.5 chat crash and Qwen3.5 streaming inference failure are separate runtime problems." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Mistral3 tied-weight bug is unrelated to Ovis2 checkpoint corruption." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper run failure is unrelated to the pydantic config-model issue." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Qwen3.5 streaming failure and MoE MPS histogram error are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Grouped_mm tracing error and MPS histogram-not-implemented error are different execution paths." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Both are numerical/runtime issues in MoE-ish models, but they are not the same bug or fix." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "GenerationMixin import breakage and Gemma3 sliding_window attribute error are unrelated." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Voxtral processor failure and template compilation error are different issues." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Mask-shape mismatch and ZeRO-3 index-out-of-bounds are different bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the ZeRO-3 duplicate cluster around 43596/43638/41762 is supported by the titles. The negative soft-edge verdicts are mostly conservative and do not overclaim shared root causes beyond what the packet shows." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 2, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:39186", + "issue:39290", + "issue:39692", + "issue:40990", + "issue:41093", + "issue:41553", + "issue:42915", + "issue:43054", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43644", + "issue:43645", + "issue:43716", + "issue:43742", + "issue:43828", + "issue:43866", + "issue:43873", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:43994", + "issue:44291", + "issue:44292", + "issue:44360", + "issue:44387", + "issue:44403", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44918", + "issue:44928", + "issue:44960", + "issue:45005", + "issue:45070", + "issue:45161", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45632", + "issue:45698" + ], + "soft_pairs": [ + "issue:43296|issue:43366", + "issue:43572|issue:45070", + "issue:44387|issue:45005", + "issue:43541|issue:45161", + "issue:43828|issue:45464", + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:43531|issue:45070", + "issue:39290|issue:43742", + "issue:43054|issue:43994", + "issue:41093|issue:43596", + "issue:43541|issue:44589", + "issue:45632|issue:45698", + "issue:40990|issue:44960", + "issue:43716|issue:44918", + "issue:43404|issue:45005", + "issue:43645|issue:45399", + "issue:44530|issue:45005", + "issue:44863|issue:45005", + "issue:44360|issue:44512", + "issue:42915|issue:45399", + "issue:43873|issue:45005", + "issue:39186|issue:44918", + "issue:44492|issue:44512", + "issue:44918|issue:45507", + "issue:43975|issue:44292", + "issue:43950|issue:44534", + "issue:43644|issue:43950", + "issue:44292|issue:45507", + "issue:39692|issue:43054", + "issue:41553|issue:43883", + "issue:43828|issue:44928", + "issue:43866|issue:44403" + ], + "prepared_review_unit_hash": "95e5979fce976694d7de52a9d67a135591463c353438dfa4e55d43a090c767c7", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37179, + "estimated_input_tokens": 9295, + "estimated_eval_tokens": 18846 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a grab-bag of unrelated Transformer/vLLM issues. The clearest duplicate set is the non-persistent buffer corruption regression in v5.x. A second likely duplicate pair is the SigLIP2 bad-output reports. Most other pairs only share a subsystem or vague symptom class and should stay separate.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43950 is the clearest and broadest statement of the non-persistent-buffer regression, with an explicit `from_pretrained()` corruption description that subsumes the narrower \"junk\" reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43950 is the best representative issue for triage because it most precisely names the regression, its scope, and the loading path involved.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "Different models and failures: PaddleOCR-VL load failure in vLLM vs GGUF/gpt-oss architecture support." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Both involve config-related v5 breakage, but one is a missing StableLmConfig field and the other is a pydantic/PretrainedConfig field regression; not the same bug." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "One is int4 quantization memory growth/OOM, the other is tied-weights behavior on translation models; different code paths." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Mixtral torch-dynamo grouped_mm tracing failure is unrelated to GPT-OSS tensor-parallel support." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Both are runtime failures, but one is an autocast dtype mismatch in Phi-tiny-MoE and the other is a streaming chat/completions failure in Qwen3.5." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Unexpected `_is_hf_initialized` during init_empty_weights is a different loading-path bug from the pydantic PretrainedConfig field issue." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GPT-OSS GGUF support and Qwen3.5 streaming API failure are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window handling and pydantic config-field breakage are different regressions." + }, + { + "left": "issue:39290", + "right": "issue:43742", + "accept": false, + "reason": "Gemma3TextConfig sliding-window attribute error and MobileLLM key error are different model-load failures." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": true, + "reason": "Both report SigLIP2 producing clearly wrong outputs; one frames it as much worse text embeddings, the other as nonsensical AutoModel/pipeline results, consistent with the same underlying model wiring bug." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Mask-shape mismatch during indexing and a zero-sized-tensor IndexError under deepspeed zero3 are distinct symptoms and likely distinct causes." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Grouped_mm tracing failure is unrelated to missing Float8 storage type support." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both concern remote custom-module loading, but one is specifically a trust_remote_code cache-path collision and the other is a wrong-module load after save_pretrained; not enough evidence they are the same bug." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-quality reports with no shared failure mode." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 image-preprocessor dtype mismatch and Qwen3.5 embedding unpacking failure in TRL are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a Mistral3 lm_head tie bug and the other is a broader translation-model v5 issue; not the same concrete defect." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Custom-model notebook initialization and flash-attn2 fallback gating are unrelated issues." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache invalid group type on Qwen3.5 is not a tied-weights regression." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading failure is unrelated to tied-weight handling." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA indexer/ReLU discussion and a docs typo about a removed run command are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and flash-attn2 fallback blocking are different problems." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to tied weights." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP 2-D weight runtime error is unrelated to Qwen3.5 input-embedding unpacking in TRL." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "Cache-strategy typo and removed-doc-command mention are both documentation-level but not the same issue." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "TRL embedding unpacking failure and GraniteMoEHybrid invalid method call are different model/integration bugs." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "DeepSeek detokenization regression and Qwen-3-8B-NVFP4 runtime failure are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same regression: v5 corrupts non-persistent buffers during load; 44534 is a shorter restatement of the same bug." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Same non-persistent-buffer corruption regression; 43950 is the more explicit issue, while 43644 is the broader symptom description." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime failure and GraniteMoEHybrid invalid method are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "SigLIP2 doc-example errors/quantization failure are not the same as the broader bad-embedding quality report." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad AutoTokenizer error message for Voxtral and missing `all_tied_weights_keys` on MolmoForCausalLM are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Both involve numerics, but autocast dtype mismatch is not the same as Qwen3.5 RLHF NaNs from 3D position_ids/SDPA fallback." + }, + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "Corrupted Ovis2 checkpoint and generic loading noise are not the same defect." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays mostly conservative. The canonical issue choice for the non-persistent-buffer regression is supported by the titles and pair evidence, and the only accepted soft-edge duplicate outside that cluster is reasonably framed as likely rather than certain. No mergeability issue applies here since these are issue pairs, not PR pairs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 3, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36331", + "issue:38175", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43257", + "issue:43404", + "issue:43493", + "issue:43643", + "issue:43742", + "issue:43782", + "issue:43792", + "issue:43824", + "issue:43827", + "issue:43854", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43940", + "issue:43950", + "issue:43975", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44509", + "issue:44589", + "issue:44661", + "issue:44863", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45237", + "issue:45313", + "issue:45356", + "issue:45399", + "issue:45520", + "issue:45701" + ], + "soft_pairs": [ + "issue:36331|issue:41762", + "issue:41553|issue:43881", + "issue:43872|issue:44589", + "issue:43054|issue:45701", + "issue:38175|issue:43054", + "issue:42947|issue:44387", + "issue:43054|issue:43493", + "issue:43881|issue:43883", + "issue:44403|issue:44451", + "issue:43975|issue:45520", + "issue:43643|issue:45020", + "issue:43975|issue:45237", + "issue:44403|issue:44488", + "issue:42947|issue:43856", + "issue:43782|issue:44960", + "issue:42491|issue:44368", + "issue:41762|issue:42915", + "issue:43975|issue:44451", + "issue:45020|issue:45356", + "issue:44661|issue:45020", + "issue:43824|issue:43872", + "issue:44991|issue:45399", + "issue:43827|issue:44509", + "issue:45081|issue:45399", + "issue:43792|issue:44220", + "issue:43404|issue:43950", + "issue:43257|issue:43866", + "issue:41553|issue:44991", + "issue:43742|issue:45399", + "issue:42915|issue:45020", + "issue:43742|issue:45020", + "issue:43940|issue:45313", + "issue:43872|issue:44291", + "issue:41553|issue:44488", + "issue:43854|issue:44863", + "issue:42915|issue:43872" + ], + "prepared_review_unit_hash": "d6738f153ada2df681fb22c1db23dfee47938210a92023b71667259e6473cc42", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 36, + "serialized_chars": 37867, + "estimated_input_tokens": 9467, + "estimated_eval_tokens": 19190 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. I found a few true duplicate pairs around the same underlying regression/theme (SigLIP2 quality, Whisper fbank extraction, a docs cleanup, and one `_is_hf_initialized` loading regression), but most pairs are distinct issues in different models or subsystems.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45020 is the best umbrella candidate: it is broad, open, and captures the recurring recent-transformers / remote-code model-loading breakage pattern better than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45020 is the most suitable global anchor for this mixed cluster because it is the broadest, still-active report and can absorb the repeated loading/regression theme better than the more symptom-specific issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Different failures: training-time `compute_loss` signature mismatch vs DeepSpeed ZeRO-3 loading crash." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Both are load failures, but for unrelated models and different error paths." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Distinct incompatibilities: `_is_hf_initialized`/Int8Params vs missing Float8 storage." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "SigLIP2 embedding regression vs a generic tokenization-change complaint; not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": true, + "reason": "Both report SigLIP2 output quality/behavior regressions and look like symptoms of the same implementation mismatch." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "LoRA gradient checkpointing vs int4 quantization memory growth are different issues." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": true, + "reason": "43493 describes the likely root-cause SigLIP2 implementation discrepancy that matches 43054's degraded embeddings." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different models and different missing-attribute/load failures." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic loading noise vs a specific model load failure for ScandiBERT." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer detokenization bug vs Python 3.13 flash_attn import error." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "`AutoConfig` missing fields under `trust_remote_code` is narrower than the broader remote-code breakage report; not clearly the same underlying bug." + }, + { + "left": "issue:43975", + "right": "issue:45237", + "accept": false, + "reason": "Unrelated problems: detokenization vs AMD GPU support." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Different model-loading complaints with no clear shared root cause." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Checkpointing effectiveness and MoE memory usage are separate training issues." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Different models and different failure modes (`weight_only=True` vs GLM5 loading)." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA compatibility regression vs a tie-word-embeddings warning; not the same concrete bug." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Gemma3 ZeRO-3 load failure vs Qwen3 MoE FP8 failure are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:44451", + "accept": false, + "reason": "Different tokenizer/model-specific failures." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Broad remote-code regressions vs a Kimi-K2.5 tokenizer regression; not the same issue." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Model-addition helper failure vs general remote-code loading regressions." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Importing a missing class and an `_is_hf_initialized`/bitsandbytes TypeError are different bugs." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer load failure vs flash-attn fallback gating; unrelated code paths." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both are the same docs regression: stale references to removed pipeline tasks after v5 cleanup." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Mistral regex patch crash vs flash-attn fallback checks are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": true, + "reason": "Both point to Whisper audio feature extraction failing; `_torch_extract_fbank_features` is the concrete path behind the general can't-run report." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Untied lm_head weights vs non-persistent buffer corruption are different model-state bugs." + }, + { + "left": "issue:43257", + "right": "issue:43866", + "accept": false, + "reason": "Qwen3 MoE weight-conversion issue vs corrupted Ovis2 checkpoint." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Voxtral error messaging vs tokenizer load failure for a different model." + }, + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "MobileLLM key error and flash-attn fallback gating do not share a concrete root cause." + }, + { + "left": "issue:42915", + "right": "issue:45020", + "accept": false, + "reason": "Qwen3 MoE FP8 failure is too specific to be the same bug as broad remote-code breakage." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "Model-specific key error vs generic remote-code breakage." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are Qwen DeepSpeed ZeRO-3 load failures, but for different model generations and with different reported symptoms." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same underlying `_is_hf_initialized` regression surfacing during model loading / parameter construction." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Different model-load issues with no shared concrete failure path." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet and stays conservative overall. The accepted soft pairs are all reasonably supported by the issue titles/descriptions, and the rejected pairs do not overreach by forcing unrelated model/subsystem failures together. The choice of 45020 as a broad anchor is also defensible for this mixed set." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 4, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:41093", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43720", + "issue:43824", + "issue:43828", + "issue:43872", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44479", + "issue:44530", + "issue:44534", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:45030", + "issue:45305", + "issue:45325", + "issue:45538", + "issue:45584", + "issue:45588", + "issue:45685" + ], + "soft_pairs": [ + "issue:44928|issue:45588", + "issue:43257|issue:45305", + "issue:36331|issue:41093", + "issue:41093|issue:44805", + "issue:43872|issue:45538", + "issue:44479|issue:44560", + "issue:44530|issue:44863", + "issue:24643|issue:30064", + "issue:43366|issue:43828", + "issue:44155|issue:45584", + "issue:43366|issue:43531", + "issue:43550|issue:44910", + "issue:44560|issue:45685", + "issue:43645|issue:43646", + "issue:43950|issue:45325", + "issue:43824|issue:43957", + "issue:42915|issue:43957", + "issue:43366|issue:44589", + "issue:44560|issue:44918", + "issue:42617|issue:43366", + "issue:38175|issue:43994", + "issue:43646|issue:43950", + "issue:36010|issue:42915", + "issue:44560|issue:44805", + "issue:44877|issue:45030", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:43317|issue:44756", + "issue:43644|issue:44534", + "issue:38175|issue:39692", + "issue:42491|issue:43720" + ], + "prepared_review_unit_hash": "1e67f73b9376a7253e4c21b957f48177514053709b6302000a34634762090488", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37219, + "estimated_input_tokens": 9305, + "estimated_eval_tokens": 18866 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly unrelated issue reports. Only a few pairs look like true duplicates: repeated Transformers v5 non-persistent-buffer regression, identical mask-shape IndexError reports, the same SigLIP2 output bug, and the custom-model initialization regression.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 43644 is the clearest canonical representative: it states the non-persistent-buffer regression directly and is the strongest match for its near-identical duplicate (44534).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43644 is the best issue to anchor the cluster because it is specific, broadly descriptive, and maps cleanly to a near-exact duplicate report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Different bugs: one is a Qwen3.5 training/SDPA gradient-collapse report, the other is an s_aux=None flash-attention crash." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Both involve DeepSpeed, but the failures are different code paths: MoE weight conversion vs gradient averaging with GAS." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unrelated errors: custom Trainer loss-signature incompatibility vs a tensor/mask shape mismatch." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": true, + "reason": "Same underlying mask/index shape mismatch IndexError; the second report is essentially the same bug with genericized shapes." + }, + { + "left": "issue:43872", + "right": "issue:45538", + "accept": false, + "reason": "Different subsystems and failures: bitsandbytes Int8Params constructor incompatibility vs tokenizer model_max_length behavior." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both mention video/Qwen models, but the concrete failures and model paths differ; not enough evidence of one shared bug." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different models and errors: PagedAttentionCache linear_attention crash vs NemotronH checkpoint-loading failure." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Completely different areas: DeepSpeed training weight shape error vs image-processor segmentation-map handling." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "No shared bug: GGUF/gpt-oss support request vs Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Different workloads and root causes: AudioFlamingo batched-track leakage vs Whisper empty-transcription generation failure." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Both are Qwen/GGUF-adjacent, but one is support for gpt-oss architecture and the other is a Qwen3-MoE sliding_window issue." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Both are runtime/attention bugs, but the model families and concrete failure modes differ (Bamba compile/SDPA vs Qwen3.5 flash-attn position_ids)." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Unrelated bugs: Qwen3-VL video StopIteration vs an MPS histogram limitation for MoE integers." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same Transformers 5 custom-model initialization regression; one report is just a Jupyter-specific manifestation of the same bug." + }, + { + "left": "issue:43950", + "right": "issue:45325", + "accept": false, + "reason": "Distinct regressions: non-persistent buffer corruption vs Qwen2.5-VL rope index scaling for still images." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Different problems: missing Qwen2_5_VL import symbol vs meta-device model-loading breakage." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Different failures: Qwen3-MoE with FineGrainedFP8Config vs general torch.device('meta') loading regressions." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "No common bug: gpt-oss support vs missing Float8 storage type during load/serialization." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Both mention Qwen3.5/video, but one is a video-input StopIteration regression and the other is a TRL embedding unpacking failure." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "Different issues: 3d_parallel.py execution failure vs GGUF gpt-oss architecture support." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 bug report: zero probabilities / nonsensical outputs for google/siglip2-base-patch16-224." + }, + { + "left": "issue:43646", + "right": "issue:43950", + "accept": false, + "reason": "Both are Transformers 5 regressions, but one is custom model initialization and the other is non-persistent buffer corruption." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "Unrelated: GenerationMixin import regression vs Qwen3-MoE FP8 configuration failure." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different bugs: Qwen3-VL video StopIteration vs the mask/index shape mismatch IndexError." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Different config-validation problems affecting different models." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Different bugs: DeepSpeed/zero3 Bert out-of-bounds index vs tokenizer-mapping add-new-model-like failure." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different bugs: DeepSpeed zero3 Bert index-out-of-bounds vs tokenizer mapping name handling." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Different topics: dequantized-model loading with device_map=auto vs mmap/OOM on Strix Halo." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Near-exact duplicate: both report the Transformers v5 non-persistent-buffer junk/corruption regression." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Related model family, but different bugs: unexpected zero probabilities vs documentation example errors." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Different issues: Qwen3 MoE LoRA compatibility across versions vs BitNet packed-weight unpacking during accelerate loading." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded overall: the canonical choice (43644) is well supported by the near-exact duplicate 44534, and the other accepted pairs are plausible from the titles. The summary stays mostly conservative. The only pair that feels slightly less certain is 38175/43994, which reads more like a closely related SigLIP2 output issue than a guaranteed exact duplicate, but that does not undermine the cluster as a whole." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 5, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43335", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44488", + "issue:44661", + "issue:44683", + "issue:44756", + "issue:44910", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45313", + "issue:45335", + "issue:45357", + "issue:45362", + "issue:45406", + "issue:45588", + "issue:45663", + "issue:45704" + ], + "soft_pairs": [ + "issue:45313|issue:45357", + "issue:44960|issue:44977", + "issue:44451|issue:44488", + "issue:43854|issue:43931", + "issue:43708|issue:44075", + "issue:42491|issue:45276", + "issue:41720|issue:42915", + "issue:43257|issue:43854", + "issue:43856|issue:43873", + "issue:39692|issue:43994", + "issue:44488|issue:44991", + "issue:42915|issue:44756", + "issue:42491|issue:43931", + "issue:45362|issue:45406", + "issue:43644|issue:43645", + "issue:44661|issue:44683", + "issue:43856|issue:44387", + "issue:44910|issue:45588", + "issue:42491|issue:45292", + "issue:43404|issue:43454", + "issue:45276|issue:45335", + "issue:36010|issue:39404", + "issue:44164|issue:44291", + "issue:44402|issue:44488", + "issue:41720|issue:44155", + "issue:39692|issue:44960", + "issue:43335|issue:45663", + "issue:43425|issue:43957", + "issue:43873|issue:44387", + "issue:42947|issue:45704", + "issue:44387|issue:44756", + "issue:43854|issue:45216", + "issue:44291|issue:44756" + ], + "prepared_review_unit_hash": "fdffd8c81c6072f8a365a9b343bb55d04b9bc936e8c7b932017e249082684122", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 37003, + "estimated_input_tokens": 9251, + "estimated_eval_tokens": 18758 + }, + "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 regression reports (model loading, quantization, generation, tokenizer, and embedding-resize bugs). There isn\u2019t a strong duplicate core, and the soft pairs mostly share superficial wording rather than the same concrete code path.", + "confidence": 0.46, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the closest reusable representative of the small resize_token_embeddings subgroup, but the overall cluster is too heterogeneous for a single strong canonical issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5 save/load regressions, but one is ZeRO-3 language_model weight loading and the other is incorrect visual-encoder keys; different code paths." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failures: GLM5 unspecified load issue vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-loading failures, but they concern different BERT models and likely different tokenizer/model-specific breakage." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different model families and symptoms: GLM-4.7-Flash test loading vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "One is Trainer checkpoint resumption math, the other is SGD argument handling; unrelated bugs." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "LoRA/Qwen3 MoE hf4.x-to-hf5.x compatibility is unrelated to Gemma4 token-embedding resize propagation." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Auto device-mapping CUDA assert and FineGrainedFP8Config failure are different Qwen3/MoE problems." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Qwen3 MoE weight-conversion under accelerate+deepspeed is unrelated to GLM-4.7-Flash loading in unit tests." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/quantization, but one is training-memory inefficiency and the other is offloading behavior; not the same bug." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family (SigLIP2) but different failures: docs example errors vs nonsensical outputs from AutoModel/pipeline." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/loading regressions, but they affect different models and are not clearly the same underlying tokenizer bug." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "FineGrainedFP8Config failure and Strix Halo mmap/OOM are unrelated." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "Different regressions: Qwen3 MoE LoRA compatibility vs Qwen3-VL shape mismatch on load." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Qwen3.5 chat crash and Gemma4Processor missing _tokenizer are unrelated model/serving bugs." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers 5.0.0 regressions, but one concerns junk-filled buffers and the other Jupyter custom-model initialization." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Tokenizer-mapping model creation and compiled flex_attention on torch>=2.9 are unrelated." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Training memory inefficiency vs int4 quantization reserved-memory growth; similar area, different concrete bug." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Flash-attention packed-sequence misinterpretation and s_aux=None AttributeError are different failure modes." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Qwen3 MoE checkpoint compatibility is unrelated to a generic resize_token_embeddings/output_embeddings bug." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both are weight-tying issues, but they affect different model classes and likely different implementation paths." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Both say resize_token_embeddings fails to update model-specific embeddings, but the affected architectures differ and there\u2019s not enough evidence this is one shared code-path bug." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin and Whisper return_language regression are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "save/from_pretrained extra_state handling is unrelated to init_empty_weights/_is_hf_initialized errors." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Vocabulary-size mismatch and tokenizer/model loading failure are different issues." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "CUDA assert during Qwen3 auto device mapping is unrelated to AudioFlamingo3 track leakage in batched inference." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "SigLIP2 documentation/runtime issues and GLM5 loading are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "SwitchTransformers sparse-layer config and Gemma-4 FSDP2 shared_kv_states KeyError are different model internals." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 incompatibility and meta-device loading breakage are different compatibility regressions." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Offloading with quantization and increased CUDA reserved memory under int4 quantization overlap conceptually but are not the same concrete bug." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to apex.FusedRMSNorm memory leak." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention memory pressure, but one is int4 quantization reserved-memory growth and the other is mmap/OOM on Strix Halo." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "GLM-4.7-Flash loading in tests and Qwen3.5 save_pretrained regression are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument during meta init and Strix Halo mmap/OOM are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 6, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:39404", + "issue:42947", + "issue:43257", + "issue:43299", + "issue:43381", + "issue:43425", + "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:44292", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44671", + "issue:44743", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44898", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45084", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45701" + ], + "soft_pairs": [ + "issue:44292|issue:44912", + "issue:44743|issue:44960", + "issue:45684|issue:45698", + "issue:45276|issue:45292", + "issue:43957|issue:44756", + "issue:42947|issue:43381", + "issue:43381|issue:44928", + "issue:44164|issue:44756", + "issue:43646|issue:43708", + "issue:45588|issue:45663", + "issue:44479|issue:44877", + "issue:44402|issue:44451", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:43257|issue:44863", + "issue:43828|issue:43957", + "issue:43425|issue:43828", + "issue:44589|issue:44756", + "issue:39404|issue:45701", + "issue:44479|issue:45663", + "issue:43906|issue:44530", + "issue:43931|issue:44863", + "issue:43611|issue:43645", + "issue:43901|issue:44509", + "issue:43425|issue:43716", + "issue:44530|issue:45084", + "issue:39404|issue:44912", + "issue:43299|issue:43931", + "issue:44863|issue:45663", + "issue:44479|issue:44805", + "issue:43257|issue:45663", + "issue:44265|issue:44898", + "issue:43994|issue:44079", + "issue:44671|issue:44991", + "issue:43856|issue:44756", + "issue:39404|issue:44671" + ], + "prepared_review_unit_hash": "6fed7dd65fa1b0106113d7e902a78e19e2e61f384ba4df15b1f2a201728e334a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38395, + "estimated_input_tokens": 9599, + "estimated_eval_tokens": 19454 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous Transformers regressions and model-specific loading bugs. The only clear duplicate chain is around `resize_token_embeddings` not propagating to model-specific embedding/output modules (`output_embeddings`, `embed_tokens_per_layer`, `decoder.embed_tokens`).", + "confidence": 0.58, + "canonical_issue_reason": "issue:45292 is the broadest, most reusable statement of the shared `resize_token_embeddings` regression, and it best covers the more specific variants in issue:45276 and issue:45335.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45292 is the strongest representative issue because it names the shared API-level failure rather than one specific architecture, while the rest of the cluster is largely unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are Qwen/quantization-related, but they describe different loading failures and code paths." + }, + { + "left": "issue:44743", + "right": "issue:44960", + "accept": false, + "reason": "Different models and different bugs: cache state reset vs a GLM5 issue." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load custom modules, but one is a permissions propagation bug and the other is loading the wrong module." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "Same underlying `resize_token_embeddings` propagation bug; both report output/embedding modules not being updated." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "`torch.device(\"meta\")` loading failures are unrelated to the Strix Halo mmap OOM issue." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both mention gradient checkpointing, but one is LoRA ineffectiveness and the other is eval-mode incompatibility." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Completely different failures: eval-mode checkpointing vs RLHF NaN/SDPA fallback." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "`extra_state` save/load handling is unrelated to the mmap/OOM issue." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization and resume-from-checkpoint step \u0917\u0923culation are different bugs." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Both are attention/Gemma-related, but they describe different failures in different code paths." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video-input regression for Qwen models is unrelated to strict config loading for granite_speech." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Tokenizer vocab-size mismatch vs a specific model loading failure; not the same bug." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same `resize_token_embeddings` regression, with `decoder.embed_tokens` as a model-specific manifestation of the broader output-embedding bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Notebook custom-model initialization is unrelated to `extra_state` serialization handling." + }, + { + "left": "issue:43257", + "right": "issue:44863", + "accept": false, + "reason": "Different model families and different loading failures." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Both are dtype/device-related, but one is autocast mismatch and the other is meta-device loading." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch version incompatibility is not the same as a specific autocast dtype mismatch." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup failure and mmap OOM are unrelated." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline `return_language` regression is unrelated to a tokenizer-version behavior change." + }, + { + "left": "issue:44479", + "right": "issue:45663", + "accept": false, + "reason": "Qwen video regression and Gemma-4 FSDP2 KeyError are different bugs." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Isolated reproduction of an old issue vs a Qwen3.5 PagedAttentionCache crash; no shared bug." + }, + { + "left": "issue:43931", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures with no clear shared root cause." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both are v5 custom-model loading regressions, but the symptoms and likely code paths differ." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but they cover different pipeline-task deprecations." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch incompatibility and a Mistral-3 dtype mismatch are unrelated." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "Different error classes and unrelated subsystems." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline behavior and Qwen quantization fallback are unrelated." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Both touch Qwen3-VL loading, but the reported failures differ enough that they do not clearly share one bug." + }, + { + "left": "issue:44863", + "right": "issue:45663", + "accept": false, + "reason": "NemotronH checkpoint loading and Gemma-4 FSDP2 attention state are different problems." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video regression vs tensor mask shape mismatch: different code paths." + }, + { + "left": "issue:43257", + "right": "issue:45663", + "accept": false, + "reason": "Qwen3 MoE conversion/loading and Gemma-4 FSDP2 attention-state rebuilding are unrelated." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "`torch.export`/`torch_compilable_check` failure is unrelated to Perceiver interpolation behavior." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "SigLIP2 inference mismatch and `ModelOutput` key assignment are different bugs." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT prediction drift and tokenizer loading for est-roberta are unrelated." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory usage is unrelated to Strix Halo mmap OOM." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper pipeline behavior and CamemBERT MLM predictions are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and suitably conservative overall. The resize_token_embeddings issues are correctly identified as the only clear duplicate chain, and the false-pair rejections do not overclaim shared root cause. Canonical issue choice is well supported by the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 7, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:32090", + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42915", + "issue:42947", + "issue:43262", + "issue:43317", + "issue:43377", + "issue:43404", + "issue:43421", + "issue:43425", + "issue:43452", + "issue:43482", + "issue:43541", + "issue:43582", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43883", + "issue:43957", + "issue:44164", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44530", + "issue:44589", + "issue:44683", + "issue:44811", + "issue:44849", + "issue:44918", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45584", + "issue:45685" + ], + "soft_pairs": [ + "issue:43425|issue:44918", + "issue:43596|issue:44683", + "issue:44589|issue:45084", + "issue:40990|issue:42947", + "issue:43425|issue:45685", + "issue:36032|issue:43452", + "issue:44991|issue:45081", + "issue:32090|issue:39404", + "issue:43883|issue:44488", + "issue:43957|issue:44164", + "issue:43582|issue:43632", + "issue:44488|issue:45081", + "issue:44164|issue:44479", + "issue:40990|issue:43421", + "issue:43611|issue:43644", + "issue:44488|issue:44987", + "issue:44206|issue:44220", + "issue:43883|issue:44991", + "issue:44811|issue:45584", + "issue:34567|issue:40444", + "issue:43716|issue:45084", + "issue:43644|issue:43646", + "issue:43643|issue:44530", + "issue:43452|issue:43482", + "issue:44451|issue:44987", + "issue:36032|issue:42915", + "issue:43317|issue:43482", + "issue:43541|issue:45084", + "issue:42915|issue:44530", + "issue:44849|issue:45325", + "issue:43404|issue:44466", + "issue:44987|issue:44991", + "issue:36032|issue:43645", + "issue:43262|issue:43377" + ], + "prepared_review_unit_hash": "077a914ff50fa0d219fba6960cbee3004fa9af326550782ac7c9c7e17edb3374", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37985, + "estimated_input_tokens": 9497, + "estimated_eval_tokens": 19250 + }, + "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 regression reports, with a loose theme around v5 model/tokenizer loading failures. I did not find any soft-edge pairs that clearly describe the same underlying bug/change.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43611 is the broadest and most representative loading-regression report in the set; several nearby issues are also Transformers v5 init/load failures, but this one is the cleanest umbrella candidate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single representative issue for the cluster because it captures the general v5 loading regression theme without being overly model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch version compatibility and Qwen3.5/TRL embedding unpacking are different problems." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "DeepSpeed ZeRO-3/BertModel index failure is unrelated to compiled flex_attention on torch>=2.9." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup and template-node compilation are unrelated error classes." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "Perplexity regression and LoRA gradient-checkpointing behavior are not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:45685", + "accept": false, + "reason": "Torch 2.10 incompatibility and an MPS histogram int error are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "T5 tokenizer method conflict and GGUF loading breakage are different code paths." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer-loading failures for different models; no clear shared root cause." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper return_language pipeline behavior." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Missing all_tied_weights_keys and failing to load a specific BERT model are unrelated." + }, + { + "left": "issue:43957", + "right": "issue:44164", + "accept": false, + "reason": "Meta-device loading regression and extra_state save/load handling are distinct issues." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer/model loading problems, but for different models and different failure points." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "extra_state serialization and Qwen video-input regression are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity on GPT-OSS and runtime special-token post-processing updates are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Base_model_prefix loading regression and junk-filled non-persistent buffers are different bugs." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Separate model-loading failures for different models; not the same concrete issue." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Unsupported center argument in LasrFeatureExtractor and _torch_extract_fbank_features are related thematically but not clearly the same bug." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Missing tied-weights attribute and tokenizer loading failure are unrelated." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both involve Whisper special tokens, but the symptoms and code paths differ enough that they are not clearly the same bug." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "TrainerState token accounting and Qwen2.5-VL iterable-dataset multimodal failure are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Mistral-3 dtype mismatch and non-template-node compilation are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Non-persistent buffer serialization and custom model initialization regressions are different failures." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss and PagedAttentionCache linear_attention crash are unrelated." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "Both mention GGUF loading, but one is a broad from_pretrained break and the other is a specific model report without enough evidence of the same code-path." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Loading ScandiBERT and loading physical-intelligence/fast are separate model-specific failures." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "T5 tokenizer conflict and Qwen3Moe FP8 failure are unrelated." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Dequantized model offload failure and Qwen2.5-GGUF loading failure are distinct regressions." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Torch dynamo grouped_mm tracing failure is unrelated to non-template-node compilation." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Qwen3.5 PagedAttentionCache crash are different bugs." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Qwen3.5 hidden-states behavior and Qwen2.5-VL rope-index scaling are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Mistral3 weight-tying bug and device-dependent serialization inconsistency are related in theme but not the same concrete failure." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Different model-loading regressions affecting different repositories/models." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio chat-template sampling-rate default and MIMI padding-mask batching bug are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 8, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36032", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:42175", + "issue:42915", + "issue:43296", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43819", + "issue:43856", + "issue:43994", + "issue:44062", + "issue:44079", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44479", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44877", + "issue:44987", + "issue:45081", + "issue:45084", + "issue:45127", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45507", + "issue:45542", + "issue:45663" + ], + "soft_pairs": [ + "issue:43646|issue:44534", + "issue:45084|issue:45507", + "issue:43994|issue:44610", + "issue:43645|issue:44534", + "issue:41553|issue:43582", + "issue:43322|issue:43540", + "issue:44403|issue:45081", + "issue:45216|issue:45357", + "issue:43377|issue:43819", + "issue:44451|issue:45081", + "issue:44220|issue:44683", + "issue:43404|issue:45127", + "issue:44265|issue:44610", + "issue:44987|issue:45081", + "issue:42915|issue:45084", + "issue:43632|issue:44756", + "issue:44479|issue:45325", + "issue:44062|issue:44987", + "issue:43296|issue:43531", + "issue:42915|issue:44291", + "issue:43611|issue:44534", + "issue:36032|issue:43742", + "issue:40444|issue:45325", + "issue:43317|issue:43856", + "issue:40990|issue:44811", + "issue:44079|issue:45663", + "issue:43421|issue:44568", + "issue:42175|issue:45542", + "issue:43531|issue:44877" + ], + "prepared_review_unit_hash": "14df59b0869d2fee19904f1fb1d582b0ca6707931208f8491b30f2e972f4be4e", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36385, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. I only see two clear duplicate subgroups: the Qwen3.5 save_pretrained checkpoint-key regression, and the Qwen2.5-VL / Qwen3-VL temporal-position-id video regression. The rest are broad-similarity matches across different bugs and should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45216 is the clearest representative: a concrete Qwen3.5 save_pretrained regression, and issue:45357 reads like a narrower follow-up on the same incorrect visual-encoder key handling.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the best issue to anchor the cluster because it is concrete, regression-scoped, and has the strongest duplicate evidence among the items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is custom-model init and the other is buffer initialization; different code paths and symptoms." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Unrelated failures: template compilation versus an invalid method call in GraniteMoEHybrid." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Different models and different bug types: bad outputs/perplexity versus a processor image-size mismatch." + }, + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Notebook custom-model initialization and non-persistent buffer junk are separate regression mechanisms." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Bad AutoTokenizer error message for Voxtral is unrelated to Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Both are multimodal model-loading issues, but they affect different models and different failure modes." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Generic loading noise versus a specific Mistral tokenizer regex crash; not the same bug." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Both report the same Qwen3.5 save_pretrained regression producing incorrect checkpoint keys, with 45357 looking like a later, narrower manifestation." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Audio batch-vs-single mismatch from padding-mask handling is not the same as DAC from_latents / STE mismatch." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "ScandiBERT load failure and Mistral regex patch crash are unrelated tokenizer/model-loading problems." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Fbank extraction and compiled flex_attention are unrelated subsystems." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "One is a missing lm_head tie in Mistral3; the other is LoRA merge collapse with extended vocab. Similar theme, different bug." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export / torch_compilable_check is unrelated to the OmDet-Turbo processor size issue." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Loading physical-intelligence/fast and the Mistral tokenizer backend crash are different failure reports." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to template-node compilation." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "_is_hf_initialized / model init behavior is unrelated to mmap OOM avoidance." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": true, + "reason": "Both point to the same Qwen VL temporal-position-id / rope-index regression affecting video or still-image paths in 5.3.0." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken keyword conflict is unrelated to loading physical-intelligence/fast." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL loading and Qwen3-MoE sliding_window behavior are different issues." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "FP8 Qwen3Moe failure and unexpected _is_hf_initialized in init_empty_weights are separate Transformer v5 regressions." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "base_model_prefix loading breakage and non-persistent buffer junk are distinct initialization bugs." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "T5Tokenizer add_special_tokens conflict and MobileLLM-125M key error are unrelated tokenizer/model-loading bugs." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Both touch Qwen2.5-VL, but one is multi-image IterableDataset training and the other is temporal position-id scaling." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "device_map/offload loading failure and Qwen3 MoE memory usage are not the same code-path issue." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "High perplexity on gpt-oss and Whisper batch_decode skip_special_tokens behavior are unrelated." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states rebuilding are different bugs." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Runtime special-token post-processor updates and mdeberta add_special_tokens BOS/EOS regression are similar area, but not clearly the same underlying bug." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both involve TensorFlow optional deps, but one is install extras packaging and the other is a tensorboard-without-TF runtime backend error." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window and granite_speech config strictness are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The two accepted pairs are plausibly the same bug family based on the titles, and the other soft-edge rejections avoid over-claiming duplicate status across clearly different models or failure modes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 9, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:36032", + "issue:40444", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43296", + "issue:43377", + "issue:43425", + "issue:43454", + "issue:43482", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43749", + "issue:43792", + "issue:43818", + "issue:43881", + "issue:43883", + "issue:43994", + "issue:44079", + "issue:44265", + "issue:44291", + "issue:44387", + "issue:44403", + "issue:44466", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44617", + "issue:44683", + "issue:44829", + "issue:44898", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45491", + "issue:45538" + ], + "soft_pairs": [ + "issue:42915|issue:44589", + "issue:43818|issue:45357", + "issue:43632|issue:44534", + "issue:43425|issue:44403", + "issue:36032|issue:41720", + "issue:44683|issue:45020", + "issue:44829|issue:45479", + "issue:43482|issue:45538", + "issue:33666|issue:40444", + "issue:44403|issue:44991", + "issue:43792|issue:44589", + "issue:44079|issue:44534", + "issue:43377|issue:45491", + "issue:35707|issue:40444", + "issue:43482|issue:43525", + "issue:43749|issue:44079", + "issue:43881|issue:44488", + "issue:44291|issue:44589", + "issue:43644|issue:44079", + "issue:40444|issue:45381", + "issue:45325|issue:45381", + "issue:44265|issue:45072", + "issue:41720|issue:45081", + "issue:43632|issue:45072", + "issue:43611|issue:43646", + "issue:43531|issue:44291", + "issue:42175|issue:43296", + "issue:43994|issue:44898", + "issue:43454|issue:44466", + "issue:42915|issue:43482", + "issue:43526|issue:43883", + "issue:44387|issue:44617" + ], + "prepared_review_unit_hash": "2d109976159c33823f486d687d02b80cfbf395531e73e16c6ea3ded976b51dde", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37355, + "estimated_input_tokens": 9339, + "estimated_eval_tokens": 18934 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of recent Transformers v5 regressions, with a few genuine near-duplicates around model initialization and Qwen2.5-VL position-id handling. Most soft pairs are only superficially similar and should stay separate.", + "confidence": 0.71, + "canonical_issue_reason": "issue:45020 is the broadest umbrella in the set: it captures the general class of recent Transformers-version regressions breaking model loading/behavior, which best represents the shared triage theme across many of these reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the best single triage anchor because it is open, broad, and most likely to accumulate links from the surrounding v5 regression reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Both involve fp8/floats, but they report different failures and likely different code paths (Qwen3MoE FineGrainedFP8Config vs missing Float8 storage object)." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Different models and failures: Video-LLaVA tower/weight-sharing bug vs Qwen3.5 save_pretrained visual-key regression." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but one is about `_is_hf_initialized` while the other is about non-persistent buffers; not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Unrelated issues: Torch version compatibility vs noisy transformer loading output." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer name conflict and Qwen3 auto device-map CUDA assert are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45020", + "accept": false, + "reason": "torch.compile/flex_attention failure is a different problem from remote_code model-loading regressions." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Different sequence-classification failures with different triggers; not the same underlying bug." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Qwen2.5-GGUF loading under v5 and CLIPTokenizer model_max_length are unrelated tokenizer/loading behaviors." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both are Qwen2-VL fine-tuning issues, but one is multi-GPU training and the other is IterableDataset/multi-image batching." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "General loading noise vs a specific tokenizer-loading regression for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper runtime failure and Float8 storage lookup failure are different issues." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key assignment and non-persistent buffer junk are distinct internal-data regressions." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both mention batched/masked behavior, but the affected models and concrete bugs are different." + }, + { + "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 IterableDataset finetuning." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "GGUF loading under v5 and missing pad_token_id on Llama4Config are different failures." + }, + { + "left": "issue:43749", + "right": "issue:44079", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and ModelOutput key handling are not the same bug." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Different model-loading failures with no clear shared code path." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Unexpected `_is_hf_initialized` argument and missing Float8 storage object are different regressions." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Both are v5 internal-behavior reports, but one is about buffers and the other about ModelOutput keys." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Multi-image iterable finetuning and Qwen2.5-VL video vision_position_ids are related only at a high level." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": true, + "reason": "Same Qwen2.5-VL temporal/vision position-id bug: both reports point to incorrect position_id computation in v5.3.0." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export/torch_compilable_check failure and bfloat16 dtype mismatch are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert is unrelated to the Mistral regex patch crash." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "`_is_hf_initialized` breakage and bfloat16 dtype mismatches do not share the same concrete bug." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": true, + "reason": "Both describe the same v5 model-initialization/loading regression affecting custom models/base_model_prefix behavior." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding-window issues and init_empty_weights `_is_hf_initialized` errors are different code paths." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Backend packaging omission and PaddleOCR-VL loading failure are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Different vision-model failures: SIGLIP2 nonsensical outputs vs Perceiver non-default-resolution classification failure." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "AyaVision lm_head tying failure and device-dependent serialization of tied weights are distinct problems." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Qwen3MoE FP8 failure and Qwen2.5-GGUF v5 loading failure are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and Molmo tied-weights attribute error are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Int4 memory regression and Sam3Video OOM are both memory-related but not the same underlying bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet: it correctly treats the set as mostly unrelated v5 regressions with only a couple of plausible near-duplicates. The accepted soft pairs are reasonably conservative from the titles alone, and the chosen umbrella issue (45020) is a defensible broad anchor for the cluster." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 10, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:43262", + "issue:43296", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43454", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43632", + "issue:43701", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43792", + "issue:43819", + "issue:43828", + "issue:43873", + "issue:43881", + "issue:43957", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45325", + "issue:45357", + "issue:45563", + "issue:45584", + "issue:45632", + "issue:45684", + "issue:45701" + ], + "soft_pairs": [ + "issue:43828|issue:44683", + "issue:43525|issue:44589", + "issue:43792|issue:43957", + "issue:43881|issue:44991", + "issue:43701|issue:45325", + "issue:43262|issue:43746", + "issue:43317|issue:43873", + "issue:43632|issue:45563", + "issue:40990|issue:45584", + "issue:44291|issue:44877", + "issue:34634|issue:35707", + "issue:43606|issue:44265", + "issue:43873|issue:45701", + "issue:44186|issue:44265", + "issue:44610|issue:45325", + "issue:35707|issue:41720", + "issue:43606|issue:43819", + "issue:43454|issue:43550", + "issue:43377|issue:44610", + "issue:43377|issue:44898", + "issue:43454|issue:45072", + "issue:44466|issue:45357", + "issue:43749|issue:44534", + "issue:39692|issue:40444", + "issue:43792|issue:44683", + "issue:43454|issue:43720", + "issue:43296|issue:44991", + "issue:43296|issue:44877", + "issue:41720|issue:43322", + "issue:45632|issue:45684", + "issue:43526|issue:43881", + "issue:43404|issue:45357", + "issue:41720|issue:44977" + ], + "prepared_review_unit_hash": "a3c08204e9219a529d9a164bf9eb1e0aa8cfcd127865a3c6135ed9e14cf2f362", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37631, + "estimated_input_tokens": 9408, + "estimated_eval_tokens": 19072 + }, + "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 / processor / generation bugs. The only plausible duplicate pair is the two offloading+quantization loading failures; the rest are distinct code paths and should stay separate.", + "confidence": 0.63, + "canonical_issue_reason": "issue:43873 is the broadest and most central report in the cluster: it describes offloading not working with quantization, which subsumes the more specific device_map/offload loading failure.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best representative issue because it covers the shared offloading/quantization failure mode and is phrased as the umbrella bug rather than one narrow instance.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43828", + "right": "issue:44683", + "accept": false, + "reason": "Different failures: autocast dtype mismatch vs compiled flex_attention failure on torch>=2.9." + }, + { + "left": "issue:43525", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated symptoms and code paths: missing pad_token_id vs missing Float8 storage object." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "Different problem areas: Whisper runtime failure vs meta-device loading regression." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "GLM-4V loading failure and tokenizer loading for a different model are not the same bug." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Resume-from-checkpoint key mismatch is unrelated to Qwen2.5-VL rope/position-id scaling." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate default bug is unrelated to GraniteSpeech PEFT checkpoint loading." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": true, + "reason": "Both report the same offloading/quantization loading path failing under device_map/CPU-GPU offload; one is a specific instance of the broader bug." + }, + { + "left": "issue:43632", + "right": "issue:45563", + "accept": false, + "reason": "Global v5 initialization breakage is unrelated to a paged generate warning about num_return_sequences." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on GPT-OSS and Whisper empty-transcription generation are unrelated model behaviors." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized argument breakage is not the same as strict config blocking granite_speech loading." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "Bark voice_preset issue and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch for Bark is unrelated to torch.export failing on torch_compilable_check." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offloading failure is unrelated to a tokenization version-change regression." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "LayoutLMv2 tokenizer crash is unrelated to torch.export compatibility." + }, + { + "left": "issue:44610", + "right": "issue:45325", + "accept": false, + "reason": "Processor image-size mismatch is unrelated to Qwen2.5-VL temporal position-id scaling." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with cached embeddings is unrelated to Qwen3 auto device mapping cudaErrorAssert." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Both touch audio-ish models, but one is CPU offload device mismatch and the other is DAC latent consistency; not the same bug." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "lm_head weight tying failure is unrelated to torch.compile+SDPA behavior." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Batched-vs-single encoder output mismatch is unrelated to OmDet-Turbo processor resolution mismatch." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Missing padding mask support in MIMI encoder is not the same as Perceiver non-default resolution handling." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Weight-tying/serialization issue is unrelated to bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "General lm_head serialization inconsistency is unrelated to Qwen3.5 visual encoder key saving." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM efficient loading breakage and non-persistent buffer junk are different initialization/loading bugs." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "SigLIP2 doc example errors are unrelated to Qwen2.5-VL finetuning with multi-image iterable datasets." + }, + { + "left": "issue:43792", + "right": "issue:44683", + "accept": false, + "reason": "Whisper runtime failure is unrelated to compiled flex_attention on torch>=2.9." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "lm_head weight tying bug is unrelated to packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43296", + "right": "issue:44991", + "accept": false, + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to est-roberta tokenizer loading." + }, + { + "left": "issue:43296", + "right": "issue:44877", + "accept": false, + "reason": "vLLM/PaddleOCR-VL loading failure is unrelated to strict granite_speech config validation." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 auto device mapping cuda assert and Llava Next segfault are different model-specific loading crashes." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Cache-path collisions for trust_remote_code and permission propagation in save_pretrained are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to GLM-4V loading failure." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "lm_head tying in Mistral3 is unrelated to incorrect visual encoder keys in Qwen3.5 save_pretrained." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Auto device mapping crash on A800 and flash-attention generation failure are different Qwen3.5 issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted soft pair (43317/43873) is a reasonable conservative duplicate call based on the titles. The remaining false verdicts are generally cautious enough given the available evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 11, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36010", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43316", + "issue:43317", + "issue:43322", + "issue:43388", + "issue:43421", + "issue:43450", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43708", + "issue:43720", + "issue:43782", + "issue:43873", + "issue:44186", + "issue:44292", + "issue:44361", + "issue:44387", + "issue:44423", + "issue:44479", + "issue:44492", + "issue:44509", + "issue:44610", + "issue:44671", + "issue:44743", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44918", + "issue:44991", + "issue:45059", + "issue:45072", + "issue:45161", + "issue:45406", + "issue:45464", + "issue:45701", + "issue:45704" + ], + "soft_pairs": [ + "issue:44423|issue:45406", + "issue:40990|issue:45701", + "issue:44186|issue:45072", + "issue:43873|issue:44756", + "issue:43577|issue:43708", + "issue:44387|issue:45704", + "issue:42617|issue:42915", + "issue:44877|issue:44991", + "issue:44479|issue:44918", + "issue:43450|issue:45701", + "issue:44610|issue:44898", + "issue:43550|issue:43720", + "issue:36331|issue:42915", + "issue:36331|issue:42617", + "issue:44186|issue:44610", + "issue:43316|issue:44913", + "issue:44671|issue:44743", + "issue:42915|issue:43317", + "issue:43295|issue:45701", + "issue:43550|issue:43606", + "issue:41720|issue:42617", + "issue:43531|issue:43572", + "issue:41720|issue:45161", + "issue:44292|issue:44479", + "issue:43388|issue:45059", + "issue:44186|issue:44361", + "issue:43540|issue:43782", + "issue:42915|issue:43322", + "issue:41720|issue:43873", + "issue:36010|issue:42617", + "issue:45406|issue:45464", + "issue:36331|issue:41720", + "issue:43421|issue:45701", + "issue:44492|issue:44509" + ], + "prepared_review_unit_hash": "c35f09a929e6485b695a29fdfa30baf06a5f865f65f120ce847b2292999aa646", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37212, + "estimated_input_tokens": 9303, + "estimated_eval_tokens": 18862 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is largely over-grouped by semantic similarity: it spans unrelated tokenizer, multimodal/serve, config, trainer, offloading, and memory regressions. None of the soft pairs look like true duplicates, so I would not merge any of them.", + "confidence": 0.9, + "canonical_issue_reason": "No strong canonical issue emerges because the cluster is heterogeneous. If one issue must anchor it, 43873 is the broadest infrastructure-level report (offloading/quantization), but it does not \u05d1\u05d0\u05de\u05ea unify the rest.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the best available anchor because it is open, broader than the others, and sits in a recurring system area (offloading/quantization). Still, it is only a weak representative for this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both involve serve/multimodal crashes, but they hit different code paths and different processor errors; not the same bug." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "One is high perplexity on a specific model/dataset, the other is tokenization changing across versions; related area, different underlying issue." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash vs dtype mismatch in a different model path; no shared concrete failure." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both mention offloading/memory, but one is quantization offload behavior and the other is mmap OOM on Strix Halo; different root causes." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "BLIP2 dtype loading bug and Trainer checkpoint max_steps regression are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Both are memory-related, but one is int4 CUDA reserved memory and the other is an apex RMSNorm leak in T5; not the same bug." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel.py failure and Qwen3Moe FP8 failure are different model/runtime problems." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config loading and tokenizer loading failures are separate API regressions." + }, + { + "left": "issue:44479", + "right": "issue:44918", + "accept": false, + "reason": "Video-input regression in inference vs Qwen3.5 embedding unpacking in TRL training; different paths and symptoms." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video batch-shape bug and tokenization-version change are unrelated." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Both are image preprocessing mismatches, but the models, expected sizes, and failure modes differ." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile/SDPA failure and packed-weight loading failure are distinct bugs." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "CustomTrainer kwargs regression and Qwen3Moe FP8 loading failure are unrelated." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer loss-signature change and 3d_parallel.py failure are different issues." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Both are input-processing bugs, but they affect different processors and failure modes." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Config API discrepancy and rotary_pct reload loss are not the same underlying defect." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM regression and Qwen3.5 recurrent-state reset bug are unrelated model-specific issues." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "FineGrainedFP8 loading failure and dequantized model offload loading failure are different code paths." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Processor/tokenizer regression and tokenization-version change are not the same bug." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Compile+SDPA issue and CPU offload device mismatch are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Auto device-map cudaErrorAssert and 3d_parallel.py failure are different runtime problems." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and missing pad_token_idx in StableLmConfig are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both touch distributed inference on MoE models, but the concrete failures and affected paths differ." + }, + { + "left": "issue:44292", + "right": "issue:44479", + "accept": false, + "reason": "NVFP4 model loading/running error and video-input regression are different issues." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Metric-gather label truncation and SAM3 PCS text/box behavior are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Two tokenizer failures, but on different tokenizers and different failure mechanisms." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Video processing failure and from_pretrained weight_only error are not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3Moe FP8 failure and Llava Next segmentation fault are unrelated model-loading problems." + }, + { + "left": "issue:41720", + "right": "issue:43873", + "accept": false, + "reason": "Auto device mapping crash and quantization offloading issue are both infra-related but not the same defect." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "ImportError for GenerationMixin and 3d_parallel.py runtime failure have no common code path." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Serve crashes in different multimodal flows: missing _tokenizer vs streaming inference failure." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Trainer compute_loss signature regression and device-map cudaErrorAssert are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:45701", + "accept": false, + "reason": "Runtime special-token post-processing update bug and tokenization-version change are different issues." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A cache-strategy typo and stale docs about removed pipeline tasks are separate documentation/config issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 12, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30064", + "issue:33357", + "issue:34567", + "issue:34634", + "issue:35707", + "issue:39692", + "issue:41720", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43525", + "issue:43531", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:43994", + "issue:44062", + "issue:44077", + "issue:44186", + "issue:44368", + "issue:44451", + "issue:44514", + "issue:44545", + "issue:44589", + "issue:44610", + "issue:44625", + "issue:44671", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45704" + ], + "soft_pairs": [ + "issue:33357|issue:35707", + "issue:44514|issue:44545", + "issue:30064|issue:34634", + "issue:41720|issue:43746", + "issue:43606|issue:43720", + "issue:43873|issue:45704", + "issue:43550|issue:43653", + "issue:44625|issue:45479", + "issue:44368|issue:45357", + "issue:43421|issue:43927", + "issue:43606|issue:43653", + "issue:34567|issue:35707", + "issue:42886|issue:44843", + "issue:44062|issue:44451", + "issue:34634|issue:41720", + "issue:44451|issue:44991", + "issue:43782|issue:45381", + "issue:43421|issue:43653", + "issue:44898|issue:45072", + "issue:44077|issue:44625", + "issue:44987|issue:45042", + "issue:44918|issue:44936", + "issue:43531|issue:44589", + "issue:43421|issue:44186", + "issue:43994|issue:44671", + "issue:43381|issue:43844", + "issue:39692|issue:44610", + "issue:45081|issue:45356", + "issue:43525|issue:43531" + ], + "prepared_review_unit_hash": "1e084c2a9d3d5df268b261583074a68fdc8e5977a8f7b9ad5604c697d2a083d3", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36086, + "estimated_input_tokens": 9022, + "estimated_eval_tokens": 18300 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous cluster with a few true duplicate/near-duplicate issue pairs. The only strong acceptances are the Qwen2.5-VL batched chat-template crash pair and the HF_HUB_OFFLINE/Mistral-regex tokenizer-loading regression pair; the rest are related-at-a-distance but not the same concrete bug.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue cleanly represents the whole cluster because the items span unrelated bugs. If forced to anchor the genuine duplicate pair, issue 44514 is the clearest representative: it is a near-verbatim match to 44545 with the same Qwen2_5_VLProcessor.apply_chat_template batched-padding crash.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44514 is the best issue anchor because it is the clearest exact-match duplicate in the set and describes a concrete, narrowly scoped crash path that is duplicated by 44545.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "Different subsystems and failure modes: MacOS CLIP bus error vs progressive generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Near-verbatim duplicate: both report the same Qwen2_5_VLProcessor.apply_chat_template batched-input crash when padding=False." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Unrelated bugs: void segmentation-map processing vs BarkProcessor voice_preset handling." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different models and code paths: Qwen3 auto device mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve loading/offload, but the concrete failures differ: Bark CPU offload device mismatch vs BitNet packed-weights handling during accelerate loading." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Not the same bug: quantization/offloading behavior vs a T5 apex.FusedRMSNorm memory leak." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "Different root causes: torch.compile/SDPA on Bamba vs tokenizer special-token registration for BigBird." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Related around num_labels, but one is config propagation and the other is degenerate single-label loss; not the same concrete bug." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "Same model family, but different issues: LoRA warning about tie_word_embeddings vs incorrect visual-encoder keys saved in Qwen3.5." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Both are tokenizer/config related, but runtime post-processor refresh is not the same as save/load loss of custom token IDs causing IndexError." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same underlying offline tokenizer regression: unconditional _patch_mistral_regex/model_info() behavior breaks HF_HUB_OFFLINE cache loading." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Different bugs in different models: Bark offload device mismatch vs BigBird mask-token special registration." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token-count update bug is unrelated to progressive generation with cached past_key_values." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "Both are loading-related, but the concrete failure signatures differ and do not show the same root cause." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset failure and Qwen3 auto device mapping CUDA assert are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both concern model/tokenizer loading, but the affected models and failure mechanisms are different and not shown to share a code path." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen multimodal bugs: weight_only loading error vs incorrect video vision_position_ids." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Same broad tokenizer area, but not the same concrete bug: runtime special-token post-processing vs empty decode from mask-token registration." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Unrelated image-resolution issue vs dtype mismatch in inference." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "No overlap in concrete code path: patchtsmixer post_init policy vs Qwen3.5 num_labels propagation." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Different problems: loading a specific model vs PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Different trainer failures: unpacking Qwen3.5 input embeddings in TRL SFT vs evaluate-after-train behavior." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Sliding-window model issue is unrelated to missing Float8 storage type errors." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Different tokenizer bugs: runtime special-token post-processing vs LayoutLMv2 NER padding/truncation crashes." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and behaviors: SigLIP2 nonsensical pipeline outputs vs CamemBERT masked LM regression." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Eval-mode gradient checkpointing limitation is unrelated to ZeRO-3 gradient growth." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Documentation example/model-processor mismatch is unrelated to OmDet-Turbo processor resolution mismatch." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Related Mistral-regex family, but the evidence suggests distinct bugs: one is an AttributeError during tokenizer loading, the other is a Kimi-K2.5 codec-handling regression with a misleading warning." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Unrelated config attribute error vs sliding_window bug in Qwen3-MoE." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level summary is mostly grounded, and the 44514\u219444545 and 42886\u219444843 edges look supportable from the titles. However, the accept on 45081\u219445356 is a bit too strong for the evidence provided: both touch _patch_mistral_regex/fix_mistral_regex, but one is an AttributeError crash in Mistral tokenizer loading while the other is a Kimi-K2.5 regression about codec handling and a misleading warning. That reads as related family, not clearly the same bug/near-duplicate. Because the soft-edge verdicts should stay conservative, I\u2019d reject this output." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 13, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:40444", + "issue:41720", + "issue:43322", + "issue:43335", + "issue:43381", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43577", + "issue:43596", + "issue:43618", + "issue:43643", + "issue:43644", + "issue:43653", + "issue:43697", + "issue:43716", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43756", + "issue:43818", + "issue:43819", + "issue:43873", + "issue:43927", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44448", + "issue:44488", + "issue:44661", + "issue:44671", + "issue:44811", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45081", + "issue:45183", + "issue:45381", + "issue:45479", + "issue:45685", + "issue:45704" + ], + "soft_pairs": [ + "issue:43596|issue:44220", + "issue:43618|issue:43697", + "issue:41720|issue:45704", + "issue:43746|issue:45381", + "issue:43381|issue:45704", + "issue:45183|issue:45479", + "issue:39692|issue:43873", + "issue:44220|issue:44661", + "issue:40444|issue:43746", + "issue:43746|issue:43873", + "issue:43540|issue:44008", + "issue:43716|issue:44008", + "issue:43335|issue:43756", + "issue:43720|issue:43819", + "issue:43653|issue:43720", + "issue:30064|issue:36010", + "issue:44062|issue:44488", + "issue:44843|issue:45081", + "issue:43421|issue:44811", + "issue:43322|issue:45685", + "issue:43577|issue:43818", + "issue:44186|issue:44898", + "issue:43550|issue:45072", + "issue:43531|issue:43643", + "issue:43643|issue:44877", + "issue:43550|issue:43927", + "issue:43540|issue:45381", + "issue:44448|issue:44671", + "issue:43644|issue:43749" + ], + "prepared_review_unit_hash": "40e4647c1813ba9acdc604dd1240f56463c6a7045c037da2257b84896c3b3f64", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36147, + "estimated_input_tokens": 9037, + "estimated_eval_tokens": 18330 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly unrelated issue reports spanning tokenizer/config loading, multimodal preprocessing, dtype/device-map bugs, model-specific regressions, and a few generic runtime failures. The soft-similarity pairs share keywords or subsystem overlap, but they do not describe the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43749 has the broadest impact and strongest signal in this set (multiple discussion threads and inbound references) around a concrete broken loading path, making it the best representative issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43749 is the most central/visible issue in the cluster and clearly describes a specific user-facing regression in loading behavior.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Different failures: BertModel + ZeRO-3 initialization vs audio feature extraction helper error." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "CLIP attentions regression is unrelated to RTDetrV2 output drift." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert is unrelated to T5/apex RMSNorm memory leak." + }, + { + "left": "issue:43746", + "right": "issue:45381", + "accept": false, + "reason": "GraniteSpeech PEFT local checkpoint loading differs from Qwen2.5-VL video position-id bug." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction is not the same as T5 fused RMSNorm leak." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Generic input validation error messaging is unrelated to single-label classification zero-loss behavior." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and quantization issue is not the same as quantization offloading behavior." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction bug is unrelated to add-new-model-like failure in tokenizer mapping names." + }, + { + "left": "issue:40444", + "right": "issue:43746", + "accept": false, + "reason": "Qwen2.5-VL iterable dataset/multi-image finetuning failure is unrelated to GraniteSpeech local checkpoint loading." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "PEFT checkpoint loading and quantization offloading are different code paths and different symptoms." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Qwen3OmniMoe video processing ValueError is unrelated to Gemma3n variable-name collision in forward()." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Mistral-3 image preprocessor dtype mismatch is not the same as Gemma3n audio mask attribute error." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "Sparse-layer creation bug in SwitchTransformers is distinct from Smollm3 RoPE-layer mismatch." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "BitNet packed-weight loading bug is unrelated to DAC.from_latents / missing STE mismatch." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBirdTokenizer special-token decode bug is unrelated to BitNet packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Void segmentation map processing and GenerationMixin import failure are completely different issues." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken special-argument conflict is unrelated to loading cjvt/sleng-bert." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Offline model_info patch issue and tokenizer backend_tokenizer crash are different Mistral-regex failures." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Runtime post-processor update for special tokens is not the same as Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next segmentation fault and MPS histogram Int error are unrelated runtime failures." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "BLIP2 dtype propagation bug is unrelated to Video-LLaVA temporal-attention/weight-sharing defect." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crash is not Perceiver non-default resolution failure." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Bamba SDPA torch.compile failure is a different model/path than SwitchTransformers and TimmWrapperModel bfloat16 mismatches." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug is unrelated to AutoConfig trust_remote_code missing fields." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Missing fields from AutoConfig with trust_remote_code is different from strict config blocking granite_speech loading." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "Bamba SDPA compile bug is unrelated to DiaConfig losing custom token IDs on save/load." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Video-input processing failure in Qwen3OmniMoe is distinct from Qwen2.5-VL video vision_position_ids mismatch." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Pegasus v4/v5 output drift is unrelated to CamemBERT masked-LM prediction regression." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer junk in Transformers v5 is not the same as FSDP CPU RAM efficient loading breakage." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 14, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30064", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42175", + "issue:43262", + "issue:43526", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43650", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:43994", + "issue:44060", + "issue:44186", + "issue:44291", + "issue:44368", + "issue:44442", + "issue:44485", + "issue:44509", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44821", + "issue:44857", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45406", + "issue:45701" + ], + "soft_pairs": [ + "issue:43782|issue:44977", + "issue:44561|issue:45020", + "issue:43653|issue:43819", + "issue:43825|issue:43994", + "issue:43577|issue:43582", + "issue:43550|issue:43819", + "issue:43825|issue:43827", + "issue:44821|issue:45084", + "issue:41720|issue:44991", + "issue:44485|issue:44509", + "issue:44186|issue:44442", + "issue:44671|issue:45701", + "issue:30064|issue:36331", + "issue:44805|issue:44936", + "issue:43606|issue:45072", + "issue:42175|issue:45081", + "issue:43650|issue:43906", + "issue:45081|issue:45406", + "issue:43782|issue:45325", + "issue:44060|issue:44368", + "issue:34634|issue:36010", + "issue:44610|issue:45072", + "issue:43749|issue:44291", + "issue:43720|issue:45072", + "issue:43526|issue:43697", + "issue:43582|issue:44857", + "issue:43526|issue:44554", + "issue:43262|issue:43819" + ], + "prepared_review_unit_hash": "399d3c76bd6649d7fcbde980289f5e4987a051d2a38adbdb09275ec2c53d1368", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35710, + "estimated_input_tokens": 8928, + "estimated_eval_tokens": 18112 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issues; the only clear duplicate-style match is the v5 pipeline-removal docs/error-message cleanup pair.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43825 is the best representative of the one real cluster theme: it is the concrete runtime-facing v5 pipeline-removal mismatch, and issue:43827 looks like the docs-only sibling of the same cleanup.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43825 is the most actionable and specific item in the set, with the clearest duplicate relationship and strongest chance to anchor a merged fix.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failures: a Qwen3VL weight_only load error versus a Qwen3.5 flash-attention generation bug." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Both mention remote_code breakage, but one is a specific v5 API removal and the other is a broad umbrella report; not the same concrete bug." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "Unrelated tokenizer special-token handling versus DAC latent reconstruction mismatch." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "One is a v5 pipeline-removal error-message/docs mismatch; the other is a model output quality regression." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Different subsystems and symptoms: dtype propagation in BLIP2 loading versus an Apple Silicon allocator TypeError." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure is unrelated to the DAC forward/from_latents mismatch." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": true, + "reason": "Same underlying v5 pipeline-removal cleanup: runtime/error-message wording and docs still referencing removed pipeline tasks." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor URL loading and template compilation errors are unrelated code paths." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "Different regressions: Qwen3 auto device mapping CUDA assert versus tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE implementation discussion is unrelated to docs still mentioning removed pipeline tasks." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 NER/padding crash and FastSpeech2Conformer tokenizer loading failure are different tokenizer issues." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM regression is not the same as a generic version-dependent tokenization change report." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image processor void segmentation maps and CustomTrainer.compute_loss kwarg handling are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask-shape IndexError and trainer train/evaluate lifecycle failure are different bugs." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to bfloat16 dtype mismatches." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "Packaging/back-end installation and Mistral regex patch crashes are unrelated." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "A malformed/empty issue title is not the same as an isolated reproduction of another issue." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Both are tokenizer/processor loading issues, but they involve different missing attributes and different models." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Qwen3VL weight_only loading and Qwen2.5-VL rope-index scaling are different failure modes." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both are about Qwen weight-tying warnings, but they describe different warning sources and different model behaviors." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset and GenerationMixin import errors are unrelated." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor image-size mismatch and dtype mismatch in inference are different bugs." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and init_empty_weights argument handling are separate loading-path issues." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight unpacking under accelerate loading is unrelated to bfloat16 dtype mismatches." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output differences are different image-processing/model regressions." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "AppleSilicon caching allocator TypeError and CUDA AMP crash are unrelated platform-specific failures." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Label-array reduction bug and MPS attention correctness issue are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and DAC latent reconstruction mismatch are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted soft edge, issue:43825 \u2194 issue:43827, is well-supported by the packet as the same v5 pipeline-removal cleanup theme (runtime wording plus docs references). The remaining rejections stay cautious and do not overclaim duplication beyond the evidence shown." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 15, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:34567", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43450", + "issue:43452", + "issue:43454", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43638", + "issue:43645", + "issue:43653", + "issue:43696", + "issue:43697", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43756", + "issue:43819", + "issue:43854", + "issue:43927", + "issue:43931", + "issue:44479", + "issue:44568", + "issue:44589", + "issue:44610", + "issue:44756", + "issue:44811", + "issue:44849", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45084", + "issue:45356", + "issue:45357", + "issue:45488", + "issue:45538", + "issue:45600", + "issue:45701" + ], + "soft_pairs": [ + "issue:43329|issue:45600", + "issue:45084|issue:45538", + "issue:34634|issue:36331", + "issue:43653|issue:44912", + "issue:44568|issue:45488", + "issue:43452|issue:44991", + "issue:43638|issue:44936", + "issue:43257|issue:43329", + "issue:38175|issue:44610", + "issue:43454|issue:43577", + "issue:43854|issue:44977", + "issue:43746|issue:44811", + "issue:43257|issue:43696", + "issue:44589|issue:45042", + "issue:39692|issue:44912", + "issue:36010|issue:43645", + "issue:44568|issue:45356", + "issue:43819|issue:43927", + "issue:43720|issue:44610", + "issue:43749|issue:44756", + "issue:43550|issue:44912", + "issue:43329|issue:44479", + "issue:43262|issue:44811", + "issue:43531|issue:45042", + "issue:36010|issue:36331", + "issue:43257|issue:43931", + "issue:34567|issue:40990", + "issue:44849|issue:45701", + "issue:43257|issue:43756", + "issue:43697|issue:45357", + "issue:43329|issue:43450" + ], + "prepared_review_unit_hash": "76ed711b7e3ba35c25b66980d46795fb5625e2281d66bb64986cda7b27a72fd9", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37130, + "estimated_input_tokens": 9283, + "estimated_eval_tokens": 18822 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most candidate pairs are clearly different bugs despite some shared subsystems. The only plausible duplicate is the Qwen/VL video-input regression pair, where one issue gives the concrete `_get_num_multimodal_tokens` video-branch failure and the other reports the broader regression across Qwen VL variants.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43329 is the most concrete root-cause report: it names the broken video branch in `_get_num_multimodal_tokens` and the undefined helpers, which makes it the best anchor for the shared Qwen video regression.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the clearest user-facing umbrella: it reports the regression across multiple Qwen VL variants, so it's the best triage entry point.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Different bugs: multimodal video token counting vs stale auto_mappings config references." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Unrelated failures: template compilation error vs tokenizer max-length constant issue." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Different code paths: BarkProcessor voice preset vs Trainer custom compute_loss signature." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration bug vs MXFP4 quantization/loading fallback bug." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer-related, but they affect different models and different mechanisms." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Different loading bugs: gguf_file regression vs a specific tokenizer-loading failure for est-roberta." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Different training/evaluation failures with different triggers and symptoms." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Different issues: Qwen3 MoE weight conversion/loading vs multimodal video token counting." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Both are vision-model mismatches, but they are different models and different concrete bugs." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "Different model-init bugs: lm_head tying vs dtype propagation to qformer." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Different model-specific flash-attention/load problems; same general area is not enough." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "Local checkpoint/PEFT loading bug vs processor batch_decode skip_special_tokens bug." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Different failures: MoE weight conversion/loading vs CUDA OOM." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Unrelated issues: Float8 storage lookup vs image processor torchvision dependency." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Documentation example/model-processor mismatch vs quantization fallback on a different model." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "Import-path regression and Jupyter custom-model initialization are different bugs." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Different tokenizer regressions affecting different models and different failure modes." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "Different model behavior: DAC latent reconstruction mismatch vs Dia config token ID persistence." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Different subsystems: BitNet packed-weight loading vs OmDet processor/model image-size mismatch." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "Both concern loading/memory, but one is FSDP CPU RAM-efficient loading and the other is mmap on Strix Halo." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "Different bugs: torch.compile+SDPA failure vs quantization fallback." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": true, + "reason": "Same underlying video-input regression in the Qwen/VL multimodal token path; 43329 identifies the concrete failing branch that plausibly explains the broader 44479 report." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat-template sampling-rate default vs Whisper batch_decode special-token handling are different bugs." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs PIL image-processor torchvision dependency." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "Different regression categories: import error vs custom-trainer method signature breakage." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Different load issues on different Qwen model families; not the same concrete bug." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "Trainer token-count tracking bug vs gpt-oss perplexity issue are unrelated." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-state bug vs generic tokenization-version change are not the same issue." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Different Qwen3 MoE loading bug vs Smollm3 RoPE-layer mismatch." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different model regressions: RTDetrV2 output changes vs Qwen3.5 save_pretrained visual-key regression." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Different subsystems and failure modes; no shared concrete code path." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "Local checkpoint loading for GraniteSpeech vs Whisper batch_decode are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering is grounded in the packet and the soft-edge calls are generally conservative. The Qwen video pair (issue:43329 / issue:44479) is a plausible duplicate/umbrella match, and the other rejections are supported by clearly different titles and bug surfaces." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 16, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:36010", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43317", + "issue:43452", + "issue:43493", + "issue:43526", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43643", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43824", + "issue:43906", + "issue:43927", + "issue:43994", + "issue:44062", + "issue:44164", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44737", + "issue:44821", + "issue:44898", + "issue:44912", + "issue:44938", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45137", + "issue:45325", + "issue:45375", + "issue:45381", + "issue:45405", + "issue:45479", + "issue:45538", + "issue:45588" + ], + "soft_pairs": [ + "issue:44987|issue:45538", + "issue:44479|issue:45381", + "issue:41720|issue:43540", + "issue:43577|issue:45072", + "issue:43452|issue:43582", + "issue:43540|issue:44560", + "issue:33357|issue:34567", + "issue:43317|issue:45538", + "issue:42175|issue:43824", + "issue:42175|issue:44938", + "issue:44291|issue:45042", + "issue:43927|issue:44912", + "issue:42915|issue:43742", + "issue:42915|issue:45538", + "issue:44062|issue:44991", + "issue:42915|issue:44821", + "issue:43540|issue:44479", + "issue:43526|issue:45479", + "issue:42175|issue:44488", + "issue:43906|issue:45381", + "issue:38175|issue:44898", + "issue:45137|issue:45405", + "issue:43749|issue:44164", + "issue:44737|issue:45588", + "issue:43756|issue:45325", + "issue:28282|issue:34689", + "issue:36010|issue:42175", + "issue:43643|issue:45375", + "issue:28282|issue:29127", + "issue:43493|issue:43994", + "issue:43742|issue:44291" + ], + "prepared_review_unit_hash": "b9a0d3129b7c4e783b4b63fd7c49fa59213918e74dfbe945a81f3385a38e9c7d", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37116, + "estimated_input_tokens": 9279, + "estimated_eval_tokens": 18814 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is still mostly heterogeneous. With only titles/metadata, the Qwen video-related issues and the config-field-loss issues are better treated as theme-level similarities than confirmed duplicates; no soft edge is strong enough to merge conservatively.", + "confidence": 0.36, + "canonical_issue_reason": "issue:44479 is the most representative anchor for the Qwen video-input theme, but the packet does not justify treating it as the same underlying bug as another issue; it should be used as a cautious thematic anchor, not a proven duplicate root.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best global issue representative because it is a concrete, specific regression report with a clear version scope and it sits in the strongest apparent subtheme, but the evidence here is not enough to collapse the subtheme into a confirmed duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Different bugs: a model-loading failure for one model vs a tokenizer max-length complaint for CLIPTokenizer." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen video-related regressions, but the packet only gives titles and does not establish the same underlying bug or a shared fix path." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both involve Qwen and runtime failures, but the symptoms and code paths differ materially: device-mapping cuda assert vs video-input ValueError." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype problems, but they affect different models and behaviors; not the same concrete bug." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "Tokenizer/model loading for gguf_file is unrelated to the Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-input issues, but the titles indicate different Qwen model families and different failures, so this is not enough to merge." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "Completely different problems: MacOS CLIP bus error vs TrainerState token counter not updating." + }, + { + "left": "issue:43317", + "right": "issue:45538", + "accept": false, + "reason": "Dequantized model offload/load failure is unrelated to CLIPTokenizer max-length behavior." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Packaging/backend availability issue vs missing Qwen class import; different subsystems and failure modes." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Packaging/backend availability issue vs Python 3.14 load failure; not the same bug." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Different code paths: init_empty_weights argument handling vs PIL image processor torchvision dependency regression." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Config save/load token-ID loss is unrelated to MXFP4 quantization falling back to bf16." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3Moe FP8 loading failure vs MobileLLM key error; unrelated model-loading bugs." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Different areas entirely: Qwen3Moe FP8 loading vs CLIPTokenizer max-length." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading failures, but the concrete errors and affected models differ too much to treat as one bug." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "FP8 config failure vs AutoImageProcessor URL loading; unrelated." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Related Qwen/video theme, but different model families and different error types; not enough evidence for the same underlying bug." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "ImageProcessorFast label reduction bug vs sequence-classification zero-loss bug; different functionality." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Packaging/backend availability issue vs a specific model not loading; not the same root cause." + }, + { + "left": "issue:43906", + "right": "issue:45381", + "accept": false, + "reason": "An isolated reproduction reference vs a specific Qwen2.5-VL video-position regression; no direct match." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "SigLIP2 zero-probabilities vs Perceiver interpolation failure are different model bugs." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT version bump." + }, + { + "left": "issue:43749", + "right": "issue:44164", + "accept": false, + "reason": "FSDP CPU RAM efficient loading regression vs save/from_pretrained extra_state handling; different code paths." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU placement bug vs flash-attention crash on sink-less models; unrelated." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different rope/position-id problems in different models; not the same underlying bug." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "PyTorch-missing ImportError vs Llama 3.2 Vision loading regression; unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42175", + "accept": false, + "reason": "GenerationMixin import failure is unrelated to the tensorflow backend packaging issue." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Both concern config fields, but the packet does not establish the same preservation bug; one is trust_remote_code missing fields, the other is a specific config field dropped by strict parsing." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "ImportError due to missing PyTorch vs LayoutLMv3 error-message clarity; unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": false, + "reason": "Both are SigLIP2-related, but the packet does not prove the same underlying defect from titles alone." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary and verdicts are a bit too strong for the packet. In particular, the accepted pair issue:44479 \u2194 issue:45381 is not grounded enough: the packet only gives titles, and while both are Qwen video-input regressions, it does not justify claiming they are the same bug or that they \u201ccould be fixed together.\u201d That makes the canonical anchor/duplicate-subcluster claim too assertive. The 43643 \u2194 45375 accept is also somewhat speculative, but the qwen accept is enough to reject this as not conservative enough." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 17, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:34567", + "issue:36331", + "issue:41762", + "issue:43295", + "issue:43452", + "issue:43525", + "issue:43531", + "issue:43577", + "issue:43582", + "issue:43643", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:44062", + "issue:44291", + "issue:44466", + "issue:44484", + "issue:44488", + "issue:44514", + "issue:44561", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44704", + "issue:44792", + "issue:44811", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:44936", + "issue:44964", + "issue:45005", + "issue:45042", + "issue:45081", + "issue:45216", + "issue:45290", + "issue:45356", + "issue:45381", + "issue:45405", + "issue:45440", + "issue:45600" + ], + "soft_pairs": [ + "issue:43582|issue:43782", + "issue:44062|issue:45081", + "issue:34567|issue:36331", + "issue:43452|issue:44488", + "issue:44484|issue:44704", + "issue:44291|issue:44821", + "issue:44792|issue:44964", + "issue:43643|issue:44561", + "issue:45405|issue:45600", + "issue:43697|issue:45216", + "issue:43688|issue:43749", + "issue:45042|issue:45081", + "issue:44514|issue:45381", + "issue:43742|issue:44589", + "issue:43531|issue:44821", + "issue:44514|issue:45290", + "issue:43452|issue:44843", + "issue:43452|issue:43577", + "issue:44625|issue:44877", + "issue:43688|issue:45440", + "issue:41762|issue:44936", + "issue:43653|issue:44568", + "issue:43295|issue:45356", + "issue:43720|issue:44898", + "issue:43525|issue:43643", + "issue:43819|issue:44811", + "issue:44466|issue:45005" + ], + "prepared_review_unit_hash": "fc6e53ca6d87b7252597aeaa4d6fac1ae11c452984b95774da4542a30c206606", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35532, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated despite broad overlap around model loading, tokenizers, and training APIs. The only strong duplicate-looking pair is the v5 tied-weights serialization regression.", + "confidence": 0.86, + "canonical_issue_reason": "Use issue 44466 as the canonical issue: it states the broader tied-weights/lm_head serialization regression clearly, while 45005 is a narrower translation-model manifestation of the same problem.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44466 is the best representative because it describes the underlying bug in a general, reproducible way and covers the broader failure mode.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43582", + "right": "issue:43782", + "accept": false, + "reason": "Different failures: Apple Silicon allocator warmup TypeError vs Qwen3-VL weight_only loading error. No shared code path." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Both touch tokenizers, but one is an AddedToken constructor bug and the other is a Mistral regex/backend_tokenizer crash. Different root causes." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "Trainer state counter not updating and custom compute_loss signature break are separate trainer API issues." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading problems, but one is about gguf_file handling and the other is a specific model/config load failure. Not the same bug." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "save_pretrained shard-size behavior and AutoProcessor cached_file kwargs forwarding are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized argument issue is unrelated to AutoImageProcessor URL loading." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal model failures with different entry points; not enough evidence of one shared bug." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "One is missing fields in AutoConfig with trust_remote_code; the other is a missing helper in v5 breaking remote-code models. Different fixes." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "Dependency-version bump and removed-config auto_mappings CI failure are unrelated maintenance issues." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "RTDetr output drift and Qwen3.5 save_pretrained checkpoint corruption are different regressions." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Auxiliary-loss normalization and FSDP CPU RAM-efficient loading breakage are different training-path bugs." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "Image processor torchvision dependency regression and Mistral regex crash are unrelated subsystems." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template batching crash and Qwen2.5-VL video position-id bug are different problems." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Key error when loading a model and missing Float8 storage lookup are different deserialization/load failures." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to loading AutoImageProcessor from URL." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both involve apply_chat_template, but one is batching/padding and the other is assistant tool-call messages with empty content. Not clearly the same underlying bug." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "gguf_file loading regression and offline model_info() call in _patch_mistral_regex are separate tokenizer-loading issues." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading breakage and Blip2 dtype propagation are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and granite_speech strict-config loading are different config issues." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Aux-loss normalization bug and DeepSeekV3MoE divergence from remote implementation are not the same concrete defect." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 load IndexError and trainer.evaluate() after train() failure are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Both concern special tokens, but one is BigBird mask-token registration and the other is mdeberta add_special_tokens behavior. Different tokenizer-specific bugs." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor.tokenizer/images regression and Kimi-K2.5 codec/fix_mistral_regex warning regression are different loading/tokenizer problems." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight loading and Perceiver interpolation failure are unrelated model-specific regressions." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Missing pad_token_id on a config and missing fields from trust_remote_code AutoConfig are distinct config-shape problems." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC.from_latents vs forward mismatch and Whisper batch_decode skip_special_tokens behavior are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": true, + "reason": "Both describe the same v5 tied-weights serialization bug around lm_head.weight/save_pretrained; 45005 is just a translation-model-specific instance." + }, + { + "left": "issue:43589", + "right": "issue:44561", + "accept": false, + "reason": "Not enough relation; Float8 storage lookup failure is unrelated to the removal of is_torch_fx_available." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is well grounded in the packet and stays mostly conservative. The only accepted pair, 44466/45005, is plausibly the same tied-weights serialization regression. The rejected pairs are generally framed cautiously as distinct issues without overstating shared evidence. The summary matches the evidence available." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 18, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:30064", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:34567", + "issue:35141", + "issue:38175", + "issue:41720", + "issue:43122", + "issue:43295", + "issue:43329", + "issue:43334", + "issue:43452", + "issue:43504", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43906", + "issue:43927", + "issue:44361", + "issue:44442", + "issue:44451", + "issue:44464", + "issue:44479", + "issue:44554", + "issue:44610", + "issue:44625", + "issue:44743", + "issue:44843", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45325", + "issue:45356", + "issue:45588", + "issue:45701", + "issue:45704" + ], + "soft_pairs": [ + "issue:24643|issue:29127", + "issue:44464|issue:44743", + "issue:41720|issue:43906", + "issue:43577|issue:43697", + "issue:38175|issue:43295", + "issue:43688|issue:45704", + "issue:44843|issue:44991", + "issue:43646|issue:43749", + "issue:44361|issue:44442", + "issue:43688|issue:45701", + "issue:43653|issue:44361", + "issue:44479|issue:44936", + "issue:28282|issue:32090", + "issue:44977|issue:45356", + "issue:28282|issue:33357", + "issue:30064|issue:33666", + "issue:28282|issue:35141", + "issue:43122|issue:43906", + "issue:43577|issue:45588", + "issue:43577|issue:44554", + "issue:43653|issue:43927", + "issue:43531|issue:43742", + "issue:43329|issue:45325", + "issue:44464|issue:44977", + "issue:44479|issue:44625", + "issue:43452|issue:44451", + "issue:43550|issue:44361", + "issue:43504|issue:43720", + "issue:43504|issue:43550", + "issue:32090|issue:34567", + "issue:43504|issue:44610", + "issue:43334|issue:43572", + "issue:32090|issue:33666" + ], + "prepared_review_unit_hash": "200cd44a223ffee01aff40799117fdc4c3e680322467224adcbf40c34acdcdfb", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37743, + "estimated_input_tokens": 9436, + "estimated_eval_tokens": 19128 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is highly heterogeneous: most items are unrelated model-, tokenizer-, training-, or version-specific regressions. There is no strong duplicate cluster here, so the best we can do is pick a broad regression ticket as a loose representative.", + "confidence": 0.18, + "canonical_issue_reason": "issue:43646 is the broadest umbrella in the group: a Transformers 5.0 initialization regression that matches the general version-breakage theme better than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the most suitable anchor only because it is a concrete, broadly worded regression ticket; it is still not a true duplicate of most other issues in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "Different bugs: DeepSpeed training weight shape error vs. LayoutLMv3 box-content validation/error messaging." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Both touch generation, but one is chunked/compiled-forward inconsistency and the other is Qwen3.5 cache state reset; not the same concrete bug." + }, + { + "left": "issue:41720", + "right": "issue:43906", + "accept": false, + "reason": "Unrelated failures: Qwen3 auto device mapping cuda assert vs. an isolated reproduction of a different issue." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "Different model paths and symptoms: Blip2 dtype loading bug vs. RTDetrV2 output divergence in v5." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Different regressions: SigLIP2 zero probabilities vs. processor/tokenizer regression for custom model code." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Different subsystems entirely: auxiliary loss normalization vs. T5 Apex RMSNorm memory leak." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading regressions, but one is an offline model_info call in regex patching and the other is a specific tokenizer loading failure; root causes differ." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Different issues: custom model initialization breakage vs. FSDP CPU RAM efficient loading regression." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both are tokenizer/model loading failures, but they affect different tokenizer classes and failure modes." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "One is auxiliary-loss normalization; the other is a broad tokenization change report. Not the same bug." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer defects: BigBird special-token registration vs. MLuke tokenizer AttributeError." + }, + { + "left": "issue:44479", + "right": "issue:44936", + "accept": false, + "reason": "Video-input regression for Qwen VL models vs. trainer evaluate-after-train failure; unrelated code paths." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Different layers: missing PyTorch import error vs. Trainer broadcast TypeError with NoneType." + }, + { + "left": "issue:44977", + "right": "issue:45356", + "accept": false, + "reason": "Different regressions: flash-attention generation issue vs. tokenizer codec/regression and warning behavior." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "Unrelated: environment import error vs. MacOS bus error with a community CLIP model." + }, + { + "left": "issue:30064", + "right": "issue:33666", + "accept": false, + "reason": "Different areas: void segmentation map processing vs. Qwen2-VL multi-GPU training." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "Different bugs: missing PyTorch dependency vs. output embedding reinitialization after resizing embeddings." + }, + { + "left": "issue:43122", + "right": "issue:43906", + "accept": false, + "reason": "Tokenization/version-change report vs. isolated reproduction of another issue; not the same defect." + }, + { + "left": "issue:43577", + "right": "issue:45588", + "accept": false, + "reason": "Different bugs: dtype not propagated in Blip2 vs. flash_attention crash on s_aux=None." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Different issues: Blip2 dtype loading vs. MPS attention correctness when head dims differ." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "Different bugs: BigBird special-token handling vs. DiaConfig custom token IDs lost on save/load." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Different failures: Qwen3-MoE sliding_window handling vs. MobileLLM-125M key error on load." + }, + { + "left": "issue:43329", + "right": "issue:45325", + "accept": false, + "reason": "Both involve multimodal/video code, but one is undefined helper usage in token counting and the other is temporal position scaling in rope index." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Different generation regressions: compiled-forward chunking vs. flash-attention behavior in Qwen3.5." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Different bugs: video regression in multiple VLMs vs. num_labels propagation in Qwen3.5 config." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "Different issues: gguf_file breaks two from_pretrained entry points vs. a specific tokenizer load failure for a model repo." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Different code paths: torch.compile/SDPA failure vs. tokenizer AttributeError." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Different bugs: legacy field during Beit loading vs. unpacking packed weights during accelerate loading." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Different model-loading and runtime-compile bugs; no shared concrete code path." + }, + { + "left": "issue:32090", + "right": "issue:34567", + "accept": false, + "reason": "Different Trainer problems: NoneType broadcast failure vs. num_input_tokens_seen not updating." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different bugs: Beit legacy field load failure vs. OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both are config-field regressions, but they affect different models and different missing fields." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Different training issues: broadcast TypeError vs. Qwen2-VL multi-GPU training." + } + ] + }, + "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": 14, + "review_unit_index": 19, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:33357", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43257", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43299", + "issue:43493", + "issue:43525", + "issue:43540", + "issue:43572", + "issue:43606", + "issue:43646", + "issue:43688", + "issue:43716", + "issue:43720", + "issue:43742", + "issue:43756", + "issue:43824", + "issue:43825", + "issue:43844", + "issue:43927", + "issue:44361", + "issue:44423", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44589", + "issue:44625", + "issue:44734", + "issue:44821", + "issue:44843", + "issue:44849", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45084", + "issue:45200", + "issue:45229", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45584" + ], + "soft_pairs": [ + "issue:43525|issue:43572", + "issue:43278|issue:43688", + "issue:41720|issue:44464", + "issue:44849|issue:45357", + "issue:43756|issue:44913", + "issue:43493|issue:43825", + "issue:43742|issue:44821", + "issue:43540|issue:43716", + "issue:43262|issue:45084", + "issue:33666|issue:36010", + "issue:43720|issue:43927", + "issue:43606|issue:43927", + "issue:43257|issue:43299", + "issue:43278|issue:43844", + "issue:44488|issue:44843", + "issue:43493|issue:43844", + "issue:44964|issue:45405", + "issue:39692|issue:43493", + "issue:43720|issue:44361", + "issue:43540|issue:44977", + "issue:43756|issue:44479", + "issue:33666|issue:36331", + "issue:45200|issue:45229", + "issue:44625|issue:45584", + "issue:43824|issue:44589", + "issue:43646|issue:44466", + "issue:44423|issue:44734", + "issue:44589|issue:45520", + "issue:39692|issue:43295", + "issue:33357|issue:33666" + ], + "prepared_review_unit_hash": "388a2b484695b446f08731c54da1c72e968fac28f580465bfd4c7b2ff9b9af53", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36733, + "estimated_input_tokens": 9184, + "estimated_eval_tokens": 18624 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a duplicate set; it mixes many unrelated issues across model families, configs, training, serving, and serialization. None of the soft edges are strong enough to merge.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue fits the cluster: the items describe distinct bugs and regressions in different subsystems and model families, so there is no clear duplicate anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a good global representative because the cluster is heterogeneous rather than one underlying bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Both involve missing padding-related config fields after v5, but they affect different model classes and different attributes (`pad_token_id` vs `pad_token_idx`)." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Different failure modes: dtype drift in evaluate vs auxiliary-loss normalization in specific MoE models." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Unrelated code paths: auto device mapping CUDA assert versus compiled-forward chunked generation inconsistency." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5 regressions, but one is about hidden states output and the other about save_pretrained visual encoder keys." + }, + { + "left": "issue:43756", + "right": "issue:44913", + "accept": false, + "reason": "Both mention rotary/RoPE settings, but one is a SmolLM3 architecture issue and the other is a GPTNeoX config reload bug." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation mismatch and pipeline translation-task messaging are unrelated bugs." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "Different loading failures: a specific model key error versus AutoImageProcessor URL loading support." + }, + { + "left": "issue:43540", + "right": "issue:43716", + "accept": false, + "reason": "Both are multimodal input issues, but one is video processing in Qwen3OmniMoe and the other is image-preprocessor dtype mismatch in Mistral-3." + }, + { + "left": "issue:43262", + "right": "issue:45084", + "accept": false, + "reason": "Audio chat-template sampling rate behavior is unrelated to template compilation errors." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU training for Qwen2-VL is unrelated to the GenerationMixin import regression." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Packed-weight unpacking during accelerate loading is a different issue from losing custom token IDs on save/load." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to DiaConfig token ID persistence." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Both concern Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is a broader v5.0 dev loading break." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Different numerical/training problems: evaluation dtype change versus ZeRO-3 gradient growth." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Loading a specific model is not the same underlying bug as unconditional Hub metadata access breaking offline mode." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "Implementation mismatch in SigLIP2 is unrelated to DeepSpeed gradient instability." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model loading failure and a version requirement bump are not the same bug." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Doc-example errors are distinct from a model implementation discrepancy with the JAX reference." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Different loader failures in different components: packed BitNet weights versus MLukeTokenizer attribute handling." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen3.5-family issues, but one is video-input validation and the other is flash-attention generation behavior." + }, + { + "left": "issue:43756", + "right": "issue:44479", + "accept": false, + "reason": "RoPE-layer documentation/model-structure issue is unrelated to the Qwen video-input regression." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Training setup for Qwen2-VL is unrelated to a Trainer API signature mismatch." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Defaulting mm_token_type_ids to zeros and multi-GPU inference OOM are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Config propagation for Qwen3.5 labels is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing a missing class and failing to find Float8 storage are different version/runtime issues." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Both involve v5 behavior, but custom model initialization and lm_head serialization are different code paths and bugs." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Serving crashes in different endpoints and failure modes: multimodal `.to()` handling versus KV-cache continuation indexing." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage lookup and Python 3.13 flash_attn import key errors are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "A documentation example bug is not the same as a regression in custom model code using processor.tokenizer." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS bus error with CLIP is unrelated to Qwen2-VL multi-GPU training." + } + ] + }, + "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": 14, + "review_unit_index": 20, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:15354", + "issue:29942", + "issue:30333", + "issue:37428", + "issue:38175", + "issue:39401", + "issue:43232", + "issue:43278", + "issue:43334", + "issue:43504", + "issue:43525", + "issue:43540", + "issue:43572", + "issue:43575", + "issue:43582", + "issue:43606", + "issue:43653", + "issue:43701", + "issue:43844", + "issue:43927", + "issue:43957", + "issue:43994", + "issue:44112", + "issue:44291", + "issue:44315", + "issue:44361", + "issue:44464", + "issue:44493", + "issue:44521", + "issue:44554", + "issue:44568", + "issue:44792", + "issue:44849", + "issue:44964", + "issue:44977", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45325", + "issue:45356", + "issue:45405", + "issue:45446", + "issue:45520", + "issue:45584", + "issue:45701" + ], + "soft_pairs": [ + "issue:43572|issue:44291", + "issue:43504|issue:43606", + "issue:44521|issue:45290", + "issue:43504|issue:43653", + "issue:43927|issue:44361", + "issue:43994|issue:44493", + "issue:43575|issue:45405", + "issue:44792|issue:45405", + "issue:44464|issue:45584", + "issue:43278|issue:43701", + "issue:43606|issue:44112", + "issue:44568|issue:45701", + "issue:29942|issue:30333", + "issue:44554|issue:45446", + "issue:15354|issue:38175", + "issue:43844|issue:44964", + "issue:43540|issue:45200", + "issue:45084|issue:45520", + "issue:39401|issue:44568", + "issue:43582|issue:44977", + "issue:43232|issue:44849", + "issue:45072|issue:45198", + "issue:37428|issue:43957", + "issue:44315|issue:44849", + "issue:44554|issue:45356", + "issue:43334|issue:43525", + "issue:45200|issue:45325" + ], + "prepared_review_unit_hash": "99f87a597a6426f04244d45185d76d12399c100e1f875f2ba7b30eeb753f6e09", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35289, + "estimated_input_tokens": 8823, + "estimated_eval_tokens": 17902 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a heterogeneous grab-bag of unrelated Transformers issues; the soft pairs mostly share only broad keywords like tokenization, loading, or generation, not the same underlying bug. No pair looks safe to merge as a duplicate, and there is no strong single canonical issue for the cluster.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Both are Transformers 5.x regressions, but one is StableLmConfig padding metadata and the other is init_empty_weights handling; different code paths and fixes." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Different bug classes: BEiT pretrained loading with a legacy field vs bark-small CPU offload device mismatch." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both touch apply_chat_template, but one is wrong assistant mask generation for multimodal inputs and the other is a crash on tool-call messages with no content." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Unrelated failures: model loading legacy-field cleanup vs BigBirdTokenizer special-token registration/empty decode." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "Different components and symptoms: DiaConfig losing custom token IDs vs MLukeTokenizer attribute errors on tasks." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "Different regressions: Siglip2 producing bad outputs vs a position-id key warning/error affecting many models." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "No shared underlying bug; one is tensor-parallel OOM, the other is a dependency version bump for PEFT." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A failing janus test is unrelated to an unreleased PEFT minimum-version change." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Different generation-path bugs in different models: compiled-forward chunking vs Whisper empty transcription after token alignment." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "One is a BF16-to-FP32 dtype change during eval, the other a resume_from_checkpoint key mismatch; no common fix." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Runtime CPU offload device mismatch is not the same as a stale CI test override in GraniteSpeech." + }, + { + "left": "issue:44568", + "right": "issue:45701", + "accept": false, + "reason": "The second issue is a broad version-dependent tokenization complaint, not evidence of the same specific BOS/EOS bug." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Completely unrelated: Flash Attention 2 test failures vs MLFlow reporting a job as still running after failure." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "Distinct areas: MPS attention correctness with mismatched head dims vs a PyTorch version check in flex_attention import logic." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure and Siglip2 zero-probability outputs are unrelated bugs." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "Different problems: ZeRO-3 gradient growth during training vs failure loading Phi-4 multimodal models." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Different multimodal input bugs: Qwen3OmniMoe video processing vs Gemma 4 text-only mm_token_type_ids defaults." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "No shared code path: template compilation TypeError vs flash_attn import KeyError on Python 3.13." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer issues, but one is offset_mapping for Qwen3 and the other is add_special_tokens not adding BOS/EOS for mdeberta-v3." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "Different runtime/model issues: AppleSilicon allocator warmup TypeError vs Qwen3.5 flash-attention generation regression." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Separate generation bugs: sync_gpus/_update_model_kwargs_for_generation vs Qwen3.5 output_hidden_states behavior." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "Different subsystems and failures: bfloat16 dtype mismatches vs Wav2Vec2 save_pretrained/tokenization problems." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Import-time flash-attn symbol breakage is unrelated to meta-device model loading regressions." + }, + { + "left": "issue:44315", + "right": "issue:44849", + "accept": false, + "reason": "Liger Kernel not being applied in model_init is unrelated to Qwen3.5 hidden-state generation behavior." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "Different issues: MPS attention correctness vs Kimi-K2.5 tokenizer codec handling/regression." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention missing pad_token_id, but they are model-specific config errors in different classes and are not the same underlying bug." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Unrelated multimodal bugs: Gemma 4 token-type IDs vs Qwen2.5-VL rope-index scaling for still images." + } + ] + }, + "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": 14, + "review_unit_index": 21, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:37428", + "issue:38175", + "issue:42915", + "issue:43278", + "issue:43295", + "issue:43388", + "issue:43450", + "issue:43493", + "issue:43526", + "issue:43572", + "issue:43575", + "issue:43653", + "issue:43701", + "issue:43819", + "issue:43824", + "issue:43827", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43901", + "issue:43957", + "issue:43975", + "issue:43994", + "issue:44112", + "issue:44206", + "issue:44360", + "issue:44442", + "issue:44462", + "issue:44464", + "issue:44466", + "issue:44514", + "issue:44521", + "issue:44561", + "issue:44610", + "issue:44779", + "issue:44792", + "issue:44898", + "issue:44936", + "issue:44987", + "issue:45083", + "issue:45127", + "issue:45198", + "issue:45245", + "issue:45290", + "issue:45381", + "issue:45412", + "issue:45563", + "issue:45584" + ], + "soft_pairs": [ + "issue:43295|issue:44610", + "issue:44466|issue:45127", + "issue:44898|issue:45412", + "issue:44792|issue:45563", + "issue:43493|issue:43701", + "issue:37428|issue:43824", + "issue:43388|issue:43526", + "issue:43873|issue:44464", + "issue:43572|issue:44987", + "issue:43844|issue:44936", + "issue:43572|issue:43957", + "issue:43824|issue:43975", + "issue:44514|issue:44521", + "issue:44360|issue:44561", + "issue:43295|issue:43450", + "issue:38175|issue:45412", + "issue:43295|issue:43873", + "issue:42915|issue:43824", + "issue:43295|issue:43994", + "issue:44442|issue:45198", + "issue:43827|issue:43901", + "issue:38175|issue:43493", + "issue:43278|issue:43856", + "issue:44206|issue:45584", + "issue:43653|issue:44112", + "issue:44462|issue:44561", + "issue:45127|issue:45245", + "issue:43450|issue:43493", + "issue:44561|issue:44779", + "issue:43819|issue:45083", + "issue:45290|issue:45381", + "issue:43450|issue:43575" + ], + "prepared_review_unit_hash": "0ca9fbe24d0caa4a60284d1e14a35eb9fae0e40983dd8c9fc8936ac8f471f5c6", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37338, + "estimated_input_tokens": 9335, + "estimated_eval_tokens": 18926 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a mostly heterogeneous cluster of regressions, API breakages, model-specific bugs, and docs/test noise. Only the two SigLIP2 reports look like the same underlying issue; the rest are not true duplicates.", + "confidence": 0.66, + "canonical_issue_reason": "issue:43493 is the clearest root-cause report in the only true duplicate pair: it describes the SigLIP2 HF-vs-JAX implementation discrepancy directly, while issue:38175 reads more like a symptom report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43295 is the most central issue in the overall cluster, with the broadest regression framing and the strongest soft-link connectivity to other nearby API-breakage reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Processor/tokenizer regression vs. a feature-extractor input-size mismatch; related area, but different bug and code path." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Inconsistent lm_head serialization/tied-weights behavior vs. LoRA-merge collapse with extended vocabulary; not the same failure." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver interpolation failure vs. RT-DETR memory leak; unrelated symptoms and code paths." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "janus image-generation test failure vs. stale paged-generate warning; different issue types." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 output discrepancy vs. resume_from_checkpoint key mismatch; distinct regressions." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and different modules/models." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both mention label handling, but one is gather_for_metrics truncation and the other is BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:43873", + "right": "issue:44464", + "accept": false, + "reason": "Quantization/offloading problem vs. compiled-forward chunked generation inconsistency; not the same bug." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in config vs. model-loading failure for a specific checkpoint; unrelated." + }, + { + "left": "issue:43844", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO-3 gradient explosion vs. trainer.evaluate() after train; different training lifecycle issues." + }, + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "Config field regression vs. meta-device model-loading breakage; separate problems." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "Qwen2.5-VL import failure vs. DeepSeek detokenization regression; different model families and failures." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both concern apply_chat_template, but one crashes on batched padding=False input while the other yields empty assistant masks; related but not the same concrete bug." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA/ReLU discussion vs. removal of is_torch_fx_available breaking trust_remote_code models; unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer/image handling regression vs. batched video-shape bug; both in processors, but different modalities and code paths." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 zero-probability outputs vs. RT-DETR GC/memory leak; unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Custom processor regression vs. quantization/offloading bug; different subsystems." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3MoE FP8 config failure vs. Qwen2.5-VL import failure; different models and symptoms." + }, + { + "left": "issue:43295", + "right": "issue:43994", + "accept": false, + "reason": "Processor/tokenizer regression vs. SigLIP2 nonsensical outputs; different affected areas." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "AutoTokenizer loading failure for one tokenizer vs. save_pretrained/tokenization failure for another audio model; not a duplicate." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs inconsistencies, but they refer to different deprecated arguments and different pipeline docs." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": true, + "reason": "Both point to the same SigLIP2 model-implementation bug; the zero-probability report is consistent with the broader HF-vs-JAX discrepancy." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Embedding dtype drift in eval vs. memory inefficiency in Qwen3 MoE training; distinct issues." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Feature extractor center-arg crash vs. Whisper empty-transcription generation failure; different code paths." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug vs. a GraniteSpeech CI device-override test; unrelated." + }, + { + "left": "issue:44462", + "right": "issue:44561", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json vs. removal of is_torch_fx_available breaking remote code; separate regressions." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge / tied-weights collapse vs. category-count limit runtime error; unrelated." + }, + { + "left": "issue:43450", + "right": "issue:43493", + "accept": false, + "reason": "Video processor batched-shape bug vs. SigLIP2 implementation discrepancy; different subsystems." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Remote-code breakage from removed helper vs. DeepSeek tokenizer regression; distinct issues." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "DAC forward/from_latents mismatch vs. Qwen3 Omni MoE feature-length helper behavior; not the same bug." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template tool-call crash vs. Qwen2.5-VL video vision_position_ids bug; separate multimodal regressions." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Batched video processor shape issue vs. tensor-parallel OOM on model load; unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet, and the soft-edge calls are generally conservative. The only accepted duplicate pair, issue:38175 \u2194 issue:43493, is well supported by the SigLIP2 titles, while the rest are reasonably treated as related but distinct." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 22, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:32090", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:40444", + "issue:43262", + "issue:43278", + "issue:43335", + "issue:43425", + "issue:43450", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43720", + "issue:43723", + "issue:43819", + "issue:43825", + "issue:43931", + "issue:44112", + "issue:44292", + "issue:44351", + "issue:44448", + "issue:44493", + "issue:44554", + "issue:44568", + "issue:44610", + "issue:44743", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45200", + "issue:45229", + "issue:45310", + "issue:45381", + "issue:45478", + "issue:45491", + "issue:45520", + "issue:45588", + "issue:45593" + ], + "soft_pairs": [ + "issue:37428|issue:43425", + "issue:43819|issue:44112", + "issue:44610|issue:44857", + "issue:44857|issue:45593", + "issue:43278|issue:43825", + "issue:45200|issue:45588", + "issue:43262|issue:43450", + "issue:43606|issue:45083", + "issue:45310|issue:45478", + "issue:44448|issue:44743", + "issue:29942|issue:32090", + "issue:37428|issue:44351", + "issue:43335|issue:43550", + "issue:39692|issue:43526", + "issue:43611|issue:43931", + "issue:43819|issue:44493", + "issue:45200|issue:45381", + "issue:43335|issue:43653", + "issue:29127|issue:32090", + "issue:44292|issue:45520", + "issue:43335|issue:45083", + "issue:44448|issue:44779", + "issue:44855|issue:45083", + "issue:39692|issue:44493", + "issue:43335|issue:43504", + "issue:38175|issue:44493", + "issue:43550|issue:44112", + "issue:44568|issue:45491", + "issue:43723|issue:44554", + "issue:44977|issue:45229", + "issue:40444|issue:43526", + "issue:43335|issue:45072", + "issue:43335|issue:45593", + "issue:43720|issue:44112" + ], + "prepared_review_unit_hash": "ecdd49ce1a0682c54be375b38bfddb1d2ef97ef70827be4aff934b013ad283b6", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37450, + "estimated_input_tokens": 9363, + "estimated_eval_tokens": 18982 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a heterogeneous bundle of unrelated issues. The only clear duplicate pair is the repeated Qwen3.5 MoE `from_pretrained` failure reported in issues 45310 and 45478; the rest are different bugs or regressions and should not be merged together.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45310 is the best canonical issue because it is the earlier report of the same Qwen3.5 MoE `from_pretrained` error, and issue:45478 is a near-identical follow-up duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the strongest representative issue in this set, but only for the 45310/45478 duplicate pair; it should not be treated as a canonical for the whole cluster since the remaining items are unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Different problems: one is a missing `_flash_supports_window_size` import, the other is a broad Torch 2.10 incompatibility report." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "Unrelated failures: DAC latent reconstruction mismatch vs a GraniteSpeech CI device-override test issue." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different code paths: OmDet-Turbo processor input sizing vs LwDetrImageLoss AMP/CUDA crash." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "One is an AMP/CUDA crash, the other is missing auxiliary losses when denoising is off; not the same bug." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Different symptoms and components: dtype mismatch during evaluate vs a pipeline error message about unsupported translation tasks." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Different issues in multimodal/flash-attention handling; no shared concrete failing code path." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate default vs batched video processor output shapes are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch in bark-small is not the same as the qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same underlying bug: both report `transformers>=5.4.0` Qwen3.5 MoE `from_pretrained` loading failure with essentially identical title/body." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Different regressions: Pegasus v4/v5 output drift vs recurrent-state reset in modular Qwen3.5 code with cache." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 test failures and a Trainer `_gpu_broadcast_one` NoneType error are unrelated." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols (`_flash_supports_window_size` vs `HybridCache`) and likely different fixes." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer configuration bug is unrelated to Bamba-9B-v2 torch.compile/SDPA failure." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and BeitImageProcessorFast `reduce_labels` returning one label are different bugs." + }, + { + "left": "issue:43611", + "right": "issue:43931", + "accept": false, + "reason": "Base-model-prefix loading breakage and Qwen3-VL weight-shape mismatch are distinct load failures." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC forward/latent mismatch is unrelated to the generic `UNEXPECTED Key with position id` issue." + }, + { + "left": "issue:45200", + "right": "issue:45381", + "accept": false, + "reason": "Different multimodal bugs: missing default `mm_token_type_ids` vs wrong `vision_position_ids` for Qwen2.5-VL video input." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "SwitchTransformers config bug and BigBirdTokenizer mask-token registration are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error-message clarity issue is not the same as the Trainer `_gpu_broadcast_one` NoneType crash." + }, + { + "left": "issue:44292", + "right": "issue:45520", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime error and Python 3.13 `flash_attn` import KeyError are different failures." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "Sparse-layer config bug in SwitchTransformers is unrelated to the qwen3_omni_moe feature-extraction helper behavior." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Pegasus output drift between v4/v5 is unrelated to DeepSeek tokenizer changes in v5." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 `IndentationError` from `@torch.jit.script` parsing is a different problem than qwen3_omni_moe feature-length behavior." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 documentation errors do not match the generic position-id key warning issue." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to BEiT pretrained preset loading with a legacy field." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Zero probabilities in SigLIP2 and unexpected position-id keys are different symptoms with different roots." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Bamba torch.compile/SDPA issue is unrelated to GraniteSpeech CI stale device override behavior." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer special-token behavior and Gemma3 NaN embeddings under mixed-length batching are different bugs." + }, + { + "left": "issue:43723", + "right": "issue:44554", + "accept": false, + "reason": "Tokenizer loading in v5 and MPS attention correctness with mismatched head dims are unrelated." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation issue and Gemma4 multi-GPU OOM are different failure modes." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Qwen2.5-VL iterable-dataset multi-image finetuning failure is unrelated to BeitImageProcessorFast `reduce_labels`." + }, + { + "left": "issue:43335", + "right": "issue:45072", + "accept": false, + "reason": "SwitchTransformers sparse-layer bug is not the same as the bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel inference." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "SwitchTransformers config bug and D-FINE auxiliary-loss behavior are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading is not the same as the GraniteSpeech CI device override failure." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is well grounded in the packet and stays conservative. The only clearly supported duplicate pair is issue:45310/45478, and the summary correctly treats the rest as heterogeneous rather than trying to force cluster-wide canonicality. The soft-edge verdicts are appropriately cautious and do not overstate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 23, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:25251", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:41720", + "issue:43065", + "issue:43066", + "issue:43278", + "issue:43329", + "issue:43412", + "issue:43421", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43701", + "issue:43873", + "issue:43927", + "issue:43976", + "issue:43994", + "issue:44186", + "issue:44448", + "issue:44451", + "issue:44462", + "issue:44466", + "issue:44493", + "issue:44743", + "issue:44857", + "issue:44977", + "issue:44987", + "issue:45072", + "issue:45083", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45520", + "issue:45563", + "issue:45588", + "issue:45698" + ], + "soft_pairs": [ + "issue:43066|issue:44462", + "issue:44448|issue:44977", + "issue:34689|issue:35141", + "issue:44977|issue:45563", + "issue:43927|issue:45698", + "issue:45372|issue:45588", + "issue:37428|issue:44987", + "issue:43606|issue:44857", + "issue:44857|issue:45072", + "issue:43701|issue:45083", + "issue:43976|issue:45440", + "issue:45245|issue:45563", + "issue:43412|issue:45588", + "issue:43550|issue:45083", + "issue:43873|issue:44493", + "issue:43873|issue:45229", + "issue:29942|issue:35141", + "issue:43278|issue:43329", + "issue:45372|issue:45459", + "issue:45372|issue:45406", + "issue:32090|issue:33357", + "issue:41720|issue:45229", + "issue:43065|issue:43994", + "issue:43065|issue:45412", + "issue:44448|issue:44466", + "issue:44451|issue:45520", + "issue:25251|issue:32090", + "issue:43653|issue:44743", + "issue:43421|issue:45491", + "issue:44857|issue:45588", + "issue:44186|issue:45198", + "issue:43994|issue:45412" + ], + "prepared_review_unit_hash": "4df963f96764a8402df8c6e6deba2fcd9b56703c392328e8ba276a92cc3278ca", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36809, + "estimated_input_tokens": 9203, + "estimated_eval_tokens": 18662 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the items span unrelated bugs in tokenization, model loading, generation, offloading, dtype handling, and export. No soft pair looks like the same underlying issue, so all candidate duplicate links should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue stands out because these reports are not one duplicate family; they cover different subsystems and failure modes, so consolidating them into a single issue would be misleading.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a good global representative for this cluster; the most similar titles still describe different bugs, not one shared code-path regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Different tokenizer bugs: decoder type handling in v5 vs ignoring tokenizer.json on load." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and symptoms: Pegasus output drift vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "One is a model-loading regression for Llama 3.2 Vision; the other is post_init reinitializing resized embeddings." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation bug vs a stale warning in paged generate(); not the same code path." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load behavior, but one is custom token IDs and the other is loading the wrong custom module." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 processor import failure vs flash_attention.py crashing on s_aux=None; unrelated modules." + }, + { + "left": "issue:37428", + "right": "issue:44987", + "accept": false, + "reason": "Missing flash-support helper import vs a specific model load failure for physical-intelligence/fast." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to the float16 AMP CUDA crash in LwDetrImageLoss." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype issues, but they affect different models and different failure modes." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch is unrelated to the qwen3_omni_moe helper-function bug." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python version compatibility regression vs native DeepseekV3MoE divergence; different problem classes." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-count limit error is not the same as the stale num_return_sequences warning." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Export/tracing guard and erfinv support issue vs an attention-path AttributeError; unrelated." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to qwen3_omni_moe length computation." + }, + { + "left": "issue:43873", + "right": "issue:44493", + "accept": false, + "reason": "Quantization/offloading behavior is a different bug from unexpected position-id keys." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Offloading/quantization issue vs Gemma4 multi-GPU OOM; not the same defect." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to embedding reinitialization after resize." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift in eval vs undefined variables in video multimodal token counting." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma4 processor import breakage is unrelated to tokenizer error masking when protobuf is absent." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "ImportError for ReasoningEffort is a different failure than Gemma4Processor missing _tokenizer in serve." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Trainer _gpu_broadcast_one NoneType error vs MacOS bus error in CLIP loading; different code paths." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3 auto device-mapping cudaErrorAssert is not the same as Gemma4 multi-GPU OOM." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to ScandiBERT loading failure." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "A placeholder Conv2d bug is unrelated to RT-DETR memory not being released." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Output drift between v4/v5 Pegasus is not the same as lm_head.weight serialization depending on device." + }, + { + "left": "issue:44451", + "right": "issue:45520", + "accept": false, + "reason": "ScandiBERT loading failure is unrelated to the Python 3.13 flash_attn KeyError." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting issue is unrelated to the Trainer broadcast TypeError." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "BigBirdTokenizer special-token decoding bug is unrelated to qwen3_5 recurrent-state reset behavior." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime post-processor update bug is unrelated to Gemma3 NaN embeddings with sliding-window attention." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "LayoutLMv2 tokenizer padding/truncation crash and Wav2Vec2 save_pretrained/tokenization failure are different bugs." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "ScandiBERT loading failure is unrelated to RT-DETR memory-leak behavior." + } + ] + }, + "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": 14, + "review_unit_index": 24, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:30064", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34567", + "issue:35141", + "issue:36010", + "issue:39692", + "issue:42915", + "issue:42994", + "issue:43232", + "issue:43257", + "issue:43278", + "issue:43334", + "issue:43381", + "issue:43388", + "issue:43526", + "issue:43531", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43931", + "issue:44079", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44279", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44479", + "issue:44610", + "issue:44871", + "issue:44977", + "issue:45072", + "issue:45103", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45412", + "issue:45440", + "issue:45491", + "issue:45593" + ], + "soft_pairs": [ + "issue:42915|issue:42994", + "issue:43278|issue:43381", + "issue:43756|issue:45440", + "issue:44361|issue:45198", + "issue:44206|issue:44479", + "issue:44265|issue:45198", + "issue:44871|issue:44977", + "issue:43257|issue:43278", + "issue:44610|issue:45412", + "issue:45440|issue:45593", + "issue:29942|issue:30064", + "issue:43824|issue:44279", + "issue:45103|issue:45372", + "issue:33357|issue:39692", + "issue:29127|issue:34567", + "issue:32090|issue:33290", + "issue:43334|issue:43531", + "issue:45072|issue:45229", + "issue:43526|issue:44186", + "issue:43653|issue:44442", + "issue:43381|issue:43701", + "issue:44464|issue:45491", + "issue:44977|issue:45245", + "issue:29942|issue:36010", + "issue:29127|issue:33357", + "issue:42915|issue:43931", + "issue:29127|issue:35141", + "issue:43232|issue:43388", + "issue:43761|issue:44079" + ], + "prepared_review_unit_hash": "e41c9e1678e739bca6841a1f292400d3f14262fbbd77952f8a6d385b819a2c5b", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36448, + "estimated_input_tokens": 9112, + "estimated_eval_tokens": 18480 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster looks noisy and mostly heterogeneous: tokenizers, processors, generation, quantization, evaluation, and trainer bugs are all mixed together. None of the soft-edge pairs look like true duplicates, so all soft edges are rejected.", + "confidence": 0.24, + "canonical_issue_reason": "Issue 36010 is the broadest, most core-library regression here (a central import failure in the generation API) and is clearer as a representative issue than the many model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 36010 has the strongest combination of breadth, clarity, and likely downstream impact, making it the best single issue representative of this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both are quantization-related, but one is an FP8 config failure and the other is quantized-model saving; different bugs." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "BF16/FP32 eval dtype drift and a gradient-checkpointing eval restriction are separate code-path issues." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Both concern model parity, but they affect different models and different mechanisms, so not the same bug." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer failures in different models; too broad to be duplicates." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Processor center-arg crash and video-input regression are unrelated processor bugs." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to save_pretrained/tokenization failure." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Eos-token config mismatch and flash-attention generation failure are distinct bugs." + }, + { + "left": "issue:43257", + "right": "issue:43278", + "accept": false, + "reason": "Qwen3 MOE weight conversion and embedding dtype changes are different issues." + }, + { + "left": "issue:44610", + "right": "issue:45412", + "accept": false, + "reason": "Input-size mismatch and a memory-leak/deallocation problem are not the same underlying bug." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepSeek implementation divergence and missing auxiliary losses are separate model/training issues." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Failing Flash Attention tests and void segmentation map processing target different failures." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "A specific Qwen2.5-VL import error is not the same as a generic dependency issue." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Auto-docstring annotation handling and a missing ReasoningEffort import are unrelated." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error loading CLIP and SigLIP2 docs/example errors are different failure modes." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error messaging and TrainerState token counting are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "NoneType broadcast errors and deepspeed OOM with Adafactor are different trainer/optimizer problems." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id on Qwen3-VL and sliding-window behavior in Qwen3-MoE are separate model bugs." + }, + { + "left": "issue:45072", + "right": "issue:45229", + "accept": false, + "reason": "bfloat16 dtype mismatches and multi-GPU CUDA OOM are not the same underlying issue." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and LayoutLMv2Tokenizer NER/padding crashes are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Eval-mode checkpointing limitation and resume_from_checkpoint key mismatch do not share a concrete code path." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaN embeddings are distinct generation/attention issues." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation breakage and a category-count runtime limit are unrelated." + }, + { + "left": "issue:29942", + "right": "issue:36010", + "accept": false, + "reason": "Flash Attention test failures and GenerationMixin import failure are different regressions." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error clarity and CLIP bus error on macOS are not duplicates." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "FP8 config failure in Qwen3MoE and Qwen3-VL weight-shape mismatch are separate model-loading bugs." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "LayoutLMv3 error messaging and embedding reinitialization after resize are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and metric-gather label dropping are different trainer internals." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVisionModel hidden_states handling and ModelOutput key assignment are separate issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 25, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:33290", + "issue:34567", + "issue:34689", + "issue:39692", + "issue:41628", + "issue:42907", + "issue:43262", + "issue:43316", + "issue:43441", + "issue:43452", + "issue:43493", + "issue:43525", + "issue:43526", + "issue:43575", + "issue:43611", + "issue:43646", + "issue:43673", + "issue:43701", + "issue:43723", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44077", + "issue:44188", + "issue:44206", + "issue:44387", + "issue:44442", + "issue:44479", + "issue:44610", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44898", + "issue:44977", + "issue:45003", + "issue:45072", + "issue:45216", + "issue:45290", + "issue:45372", + "issue:45440", + "issue:45479", + "issue:45491", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45698" + ], + "soft_pairs": [ + "issue:43262|issue:44206", + "issue:43441|issue:45588", + "issue:43526|issue:44857", + "issue:45003|issue:45440", + "issue:43723|issue:45479", + "issue:44442|issue:45072", + "issue:43526|issue:44610", + "issue:44743|issue:44811", + "issue:43452|issue:45588", + "issue:43316|issue:43525", + "issue:43761|issue:44188", + "issue:44479|issue:45290", + "issue:43646|issue:44077", + "issue:45588|issue:45600", + "issue:44977|issue:45584", + "issue:43931|issue:45216", + "issue:43575|issue:43701", + "issue:42907|issue:45698", + "issue:39692|issue:45491", + "issue:43994|issue:44206", + "issue:43575|issue:44387", + "issue:43873|issue:45588", + "issue:34689|issue:43611", + "issue:43673|issue:44743", + "issue:44442|issue:44898", + "issue:33290|issue:34567", + "issue:43493|issue:43827", + "issue:41628|issue:45372", + "issue:44077|issue:44479" + ], + "prepared_review_unit_hash": "2ad312bdab7ffaa700a8c2c08312469345e5b334135c687cf615c8118e53d7ba", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36464, + "estimated_input_tokens": 9116, + "estimated_eval_tokens": 18488 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a noisy similarity cluster, not a duplicate set: every soft pair describes a different bug, regression, or docs issue, so all soft edges should be rejected.", + "confidence": 0.91, + "canonical_issue_reason": "No canonical issue stands out; the items span unrelated bugs across audio, vision, tokenizer loading, quantization, checkpointing, and model internals, so they should not be collapsed into one duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45584 is the strongest standalone report: it is open, concrete, and describes a specific Whisper generation regression after `align_special_tokens` with a clear failure mode.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Different subsystems and failures: audio chat-template sampling-rate default vs a feature-extractor `center`-argument crash." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both mention flash attention, but they are different bugs in different code paths/models: Ministral-3 FA failure vs `s_aux=None` AttributeError in flash_attention integration." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "Unrelated bugs: `reduce_labels` returning one label in BEiT fast processor vs LwDetrImageLoss crashing under float16 AMP on CUDA." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "These are different model/framework issues: unsafe `sys.modules` access in modeling_utils vs DeepSeekV3MoE divergence from the remote implementation." + }, + { + "left": "issue:43723", + "right": "issue:45479", + "accept": false, + "reason": "Tokenizer loading regression vs a sequence-classification loss bug; no shared code-path." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "FastSpeech2Conformer tokenizer loading failure vs bfloat16 dtype mismatches in inference; different problems." + }, + { + "left": "issue:43526", + "right": "issue:44610", + "accept": false, + "reason": "BEiT label reduction bug and OmDet-Turbo processor image-size mismatch are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Recurrent state reset with cache vs Whisper `batch_decode()` ignoring `skip_special_tokens`; distinct behaviors." + }, + { + "left": "issue:43452", + "right": "issue:45588", + "accept": false, + "reason": "GGUF tokenizer/model loading breakage is not the same as the flash-attention `s_aux` AttributeError." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "API discrepancy in `Gemma3TextConfig` vs missing `pad_token_id` on `Llama4Config`; different config regressions." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden-states regression and torch.compile attention-kernel divergence are separate bugs." + }, + { + "left": "issue:44479", + "right": "issue:45290", + "accept": false, + "reason": "Qwen video-input regression vs chat-template tool-call crash; different input pipelines and failure modes." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "General custom model initialization regression is not the same as forbidding optional `post_init` in patchtsmixer." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "Flash-attention `AttributeError` and stale `auto_mappings.py` references are unrelated CI/runtime issues." + }, + { + "left": "issue:44977", + "right": "issue:45584", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug vs Whisper empty-transcription failure; different models and code paths." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Model weight-shape mismatch on load vs incorrect checkpoint saving in Qwen3.5; related theme, but not the same underlying bug." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "TP load OOM and resume-from-checkpoint key mismatch are distinct training-time issues." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Saving dequantized Ministral/Devstral models is not the same as loading the wrong custom module after `save_pretrained`." + }, + { + "left": "issue:39692", + "right": "issue:45491", + "accept": false, + "reason": "SigLIP2 docs-example errors and Gemma3 NaN embeddings are unrelated model bugs." + }, + { + "left": "issue:43994", + "right": "issue:44206", + "accept": false, + "reason": "SigLIP2 nonsensical outputs vs a feature extractor crash from unsupported `center`; different layers of the stack." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both are OOMs, but one is TP model loading and the other is int4 quantization reserved-memory growth; not the same concrete bug." + }, + { + "left": "issue:43873", + "right": "issue:45588", + "accept": false, + "reason": "Quantization/offloading behavior and a flash-attention AttributeError are different problems." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Llama 3.2 vision loading breakage vs base_model_prefix handling regression; too different to merge." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Both involve cache/state, but one is missing GenerationMixin cache in v5 chunked_prefill and the other is recurrent-state reset; separate bugs." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure vs Perceiver non-default resolution failure; no shared underlying defect." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "AdaFactor/DeepSpeed OOM is unrelated to `num_input_tokens_seen` not updating." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 HF-vs-JAX implementation discrepancy is unrelated to docs still referencing `pipeline()`." + }, + { + "left": "issue:41628", + "right": "issue:45372", + "accept": false, + "reason": "Missing `AutoImageProcessor` import and `ReasoningEffort` import failure in Gemma 4 processor loading are different import regressions." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "`post_init` policy for patchtsmixer has nothing to do with Qwen video-input regression." + } + ] + }, + "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": 14, + "review_unit_index": 26, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:25251", + "issue:28282", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:36010", + "issue:41950", + "issue:42175", + "issue:43065", + "issue:43295", + "issue:43316", + "issue:43381", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43475", + "issue:43540", + "issue:43643", + "issue:43688", + "issue:43716", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43761", + "issue:43825", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:44162", + "issue:44190", + "issue:44442", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44617", + "issue:44792", + "issue:44855", + "issue:44871", + "issue:44877", + "issue:44913", + "issue:45127", + "issue:45137", + "issue:45200", + "issue:45584", + "issue:45656" + ], + "soft_pairs": [ + "issue:43381|issue:43873", + "issue:33290|issue:34689", + "issue:33290|issue:35141", + "issue:43643|issue:44913", + "issue:44190|issue:44496", + "issue:43749|issue:43931", + "issue:43720|issue:44442", + "issue:43873|issue:44871", + "issue:33290|issue:45656", + "issue:43316|issue:44877", + "issue:44162|issue:44792", + "issue:34689|issue:36010", + "issue:44855|issue:45137", + "issue:43540|issue:44514", + "issue:43381|issue:43856", + "issue:42175|issue:43316", + "issue:43404|issue:43441", + "issue:41950|issue:43475", + "issue:43316|issue:43716", + "issue:43388|issue:43688", + "issue:16998|issue:22355", + "issue:43381|issue:43742", + "issue:25251|issue:28282", + "issue:43295|issue:43825", + "issue:43761|issue:44479", + "issue:43065|issue:44617", + "issue:44871|issue:45584", + "issue:45127|issue:45200" + ], + "prepared_review_unit_hash": "ec5d8858a0c1fde3cd0c7dea8af68b64fa03749b52702516ef51a2b44399736a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35447, + "estimated_input_tokens": 8862, + "estimated_eval_tokens": 17980 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs across model loading, configs, pipelines, distributed training, tokenizer behavior, and model-specific regressions. None of the soft pairs look like true duplicates, so all candidate merges should be rejected.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue: the cluster is a grab-bag of unrelated regressions and feature/behavior bugs, so nothing cleanly represents the whole set.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue cleanly subsumes the others; the closest candidates are still model- or subsystem-specific and not representative of the cluster as a whole.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and offloading with quantization are different training-path bugs." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "AdaFactor/deepspeed OOM is unrelated to the Llama 3.2 Vision model-loading regression." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Optimizer OOM and embedding reinitialization after resize are distinct bugs in different code paths." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "Remote-code config field loss and GPTNeoX rotary_pct reload mismatch are separate config-handling issues." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading failure in an example script is unrelated to an unrecognized model/config error." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and Qwen3-VL weight-shape mismatch are different loading failures." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight loading and FastSpeech2Conformer tokenizer loading are unrelated code paths." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading and Gemma eos_token_id inconsistency are unrelated bugs." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Both involve deepspeed/optimizer behavior, but OOM from AdaFactor is not the same as double optimizer-step invocation." + }, + { + "left": "issue:43316", + "right": "issue:44877", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and strict granite_speech config loading are different config-schema problems." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage and a janus image-generation test failure are unrelated model-specific issues." + }, + { + "left": "issue:34689", + "right": "issue:36010", + "accept": false, + "reason": "Llama 3.2 Vision loading regression and missing GenerationMixin import are separate import/load failures." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "TorchScript parsing on Python 3.13 and DeepSpeed ZeRO3 deque errors do not share an underlying bug." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Qwen3OmniMoe video-input processing and Qwen2_5_VL batched chat-template padding are different failures." + }, + { + "left": "issue:43381", + "right": "issue:43856", + "accept": false, + "reason": "Gradient checkpointing eval-mode behavior and Qwen3 MoE memory usage are not the same bug." + }, + { + "left": "issue:42175", + "right": "issue:43316", + "accept": false, + "reason": "Missing TensorFlow in a pip extra and a Gemma3TextConfig API discrepancy are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Weight tying in Mistral3ForConditionalGeneration and FlashAttention failure in Ministral-3 are different concrete problems." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Video-classification pipeline processor selection and Sam3VisionEncoderOutput missing attribute are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and Mistral-3 image-preprocessor dtype mismatch are different model-specific bugs." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Last-batch label truncation in gather_for_metrics and auxiliary-loss normalization are separate training issues." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "DeBERTa-V3 model_max_length discussion and missing transformers.onnx module are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:43742", + "accept": false, + "reason": "Gradient checkpointing eval-mode limitation is unrelated to MobileLLM-125M loading key errors." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline top_k nesting behavior and AutoModel PyTorch import errors are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "Processor/tokenizer regression and a v5 translation error-message issue are different API regressions." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and Qwen video-input regression do not share the same failure mode." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Sam3PixelDecoder dummy Conv2d and Sam3Video CUDA OOM are different Sam3 issues." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure are unrelated tokenization/generation bugs." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA merge collapse with tied embeddings and Gemma 4 mm_token_type_ids defaults are different model/config problems." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 27, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:32090", + "issue:35141", + "issue:41720", + "issue:42617", + "issue:43065", + "issue:43295", + "issue:43452", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43701", + "issue:43742", + "issue:43761", + "issue:43827", + "issue:43906", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44521", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:44933", + "issue:44936", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45245", + "issue:45290", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45406", + "issue:45563", + "issue:45584", + "issue:45600", + "issue:45698" + ], + "soft_pairs": [ + "issue:44077|issue:44534", + "issue:44734|issue:45406", + "issue:44936|issue:45698", + "issue:43295|issue:44871", + "issue:43550|issue:43701", + "issue:43606|issue:43701", + "issue:43646|issue:44913", + "issue:43452|issue:44496", + "issue:43931|issue:44060", + "issue:43761|issue:44242", + "issue:43575|issue:43742", + "issue:44617|issue:45584", + "issue:45563|issue:45584", + "issue:42617|issue:43575", + "issue:43906|issue:45290", + "issue:44521|issue:45381", + "issue:44479|issue:44734", + "issue:44933|issue:45362", + "issue:32090|issue:35141", + "issue:41720|issue:43575", + "issue:43065|issue:44610", + "issue:44464|issue:44871", + "issue:44464|issue:45563", + "issue:43827|issue:43994", + "issue:44792|issue:44977", + "issue:44534|issue:44913", + "issue:41720|issue:45245", + "issue:44190|issue:44964", + "issue:44792|issue:45600", + "issue:45127|issue:45357" + ], + "prepared_review_unit_hash": "bb6fd536b1973c01c5da1e46bddb08436a119d15305a9c1ad3cb15853acf843c", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36570, + "estimated_input_tokens": 9143, + "estimated_eval_tokens": 18542 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: many issues mention loading, initialization, or multimodal regressions, but they point to different code paths and symptoms. I found no true duplicate pair among the soft edges, so all soft links should stay rejected.", + "confidence": 0.96, + "canonical_issue_reason": "No exact canonical duplicate emerges; issue 43646 is the broadest representative of the initialization/loading regression subset, but it is not a duplicate of the others.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43646 is the best single representative for the cluster only because it is broadly about model initialization/loading regressions; however, the cluster is too mixed for a real canonical duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Both relate to v5/post-init behavior, but one is about patchtsmixer allowing optional post_init and the other about junk non-persistent buffers; different bugs." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "Distinct crashes in different areas: KV-cache continuation indexing in serve vs Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:44936", + "right": "issue:45698", + "accept": false, + "reason": "Trainer lifecycle failure after train() is unrelated to from_pretrained loading the wrong custom module after save_pretrained." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Multimodal processor/tokenizer regression and eos_token_id mismatch are different model/config issues." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile+SDPA failure for Bamba is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch is a different failure mode from checkpoint key mismatch." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Both involve model config/loading, but one is custom model initialization and the other is rotary_pct not persisting on reload." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Both are loading-related, but gguf_file breaks a specific AutoTokenizer/AutoModel path while the other lacks a model_type in config.json." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Weight shape mismatch in Qwen3-VL is not the same bug as the incorrect tied-weights warning in Qwen3-Next." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "Hidden_states=None in CLIPVisionModel.forward is unrelated to missing load-balancing loss when output_router_logits=False." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "TP OOM on a 57B model and a key error loading MobileLLM-125M are unrelated load failures." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after align_special_tokens." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale warning for num_return_sequences is not the same as an empty-transcription generation crash." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "Failed 3d_parallel.py execution and tp OOM are different problems with different code paths." + }, + { + "left": "issue:43906", + "right": "issue:45290", + "accept": false, + "reason": "An isolated reproduction of an older issue is not the same as the assistant tool-call crash in apply_chat_template(tokenize=True)." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal regressions, but assistant_masks being all zero is different from wrong vision_position_ids for video inputs." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Video-input regression in Qwen VL models is unrelated to serve KV-cache continuation tensor indexing." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "Missing import from image_utils and Qwen3.5 chat crash are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer GPU broadcast TypeError and embedding reinitialization in post_init are different bugs." + }, + { + "left": "issue:41720", + "right": "issue:43575", + "accept": false, + "reason": "Auto device mapping cudaErrorAssert and TP OOM are both resource-related but not the same underlying issue." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to processor output size mismatch for OmDet-Turbo." + }, + { + "left": "issue:44464", + "right": "issue:44871", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency is not the same as an eos_token_id configuration mismatch." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing pipeline() is a documentation cleanup issue, not the same as nonsensical model outputs." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation bug are different models and failure modes." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk and rotary_pct reset on reload both touch persistence, but they are not the same concrete bug." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "cudaErrorAssert on A800 is unrelated to the category-count limit error." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading in a training script is unrelated to loading a specific multimodal model." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "A failing janus generation test is unrelated to auto_mappings.py referencing removed Sam3LiteText configs." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge collapse with extended vocab and incorrect visual encoder keys saved are different save/merge regressions." + } + ] + }, + "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": 14, + "review_unit_index": 28, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36246", + "issue:38175", + "issue:38617", + "issue:39692", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:43065", + "issue:43116", + "issue:43295", + "issue:43335", + "issue:43388", + "issue:43404", + "issue:43479", + "issue:43540", + "issue:43550", + "issue:43644", + "issue:43650", + "issue:43761", + "issue:43976", + "issue:44016", + "issue:44077", + "issue:44079", + "issue:44112", + "issue:44186", + "issue:44190", + "issue:44220", + "issue:44279", + "issue:44466", + "issue:44561", + "issue:44683", + "issue:44734", + "issue:44743", + "issue:44855", + "issue:44913", + "issue:44938", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45375", + "issue:45406", + "issue:45446", + "issue:45542", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45632" + ], + "soft_pairs": [ + "issue:44938|issue:45542", + "issue:43388|issue:44079", + "issue:43761|issue:44743", + "issue:44855|issue:45588", + "issue:43335|issue:44855", + "issue:43976|issue:44855", + "issue:44077|issue:44079", + "issue:43550|issue:44190", + "issue:43644|issue:44913", + "issue:43540|issue:44734", + "issue:43650|issue:44220", + "issue:44112|issue:45600", + "issue:43116|issue:44190", + "issue:45276|issue:45584", + "issue:39692|issue:44190", + "issue:41950|issue:44016", + "issue:43479|issue:45375", + "issue:44734|issue:45632", + "issue:36246|issue:41628", + "issue:43404|issue:45216", + "issue:43295|issue:45276", + "issue:45245|issue:45584", + "issue:36246|issue:38617", + "issue:44016|issue:44220", + "issue:44561|issue:45632", + "issue:44186|issue:44734", + "issue:44683|issue:45446", + "issue:38175|issue:43065", + "issue:42175|issue:44279", + "issue:45406|issue:45632", + "issue:44466|issue:44913" + ], + "prepared_review_unit_hash": "0336dbca82552ae41b3801f1a5fca9f4cc0ef0a2e5ef320b8141f4e0b9514e0c", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36640, + "estimated_input_tokens": 9160, + "estimated_eval_tokens": 18576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The soft links are overwhelmingly false positives: they span unrelated import errors, config/serialization regressions, model-specific bugs, pipeline issues, and docs/example failures. None of the candidate pairs look like the same underlying defect or change.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:41628 is the strongest standalone candidate to surface as a representative user-facing regression: it is a concrete, high-activity import failure with broad impact. That said, the set does not form a true duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Unrelated failures: Python import/load issue vs missing TensorFlow backend when only tensorboard is installed." + }, + { + "left": "issue:43388", + "right": "issue:44079", + "accept": false, + "reason": "Both involve internals around output handling, but one is metric-gathering label truncation and the other is ModelOutput key assignment; different code paths." + }, + { + "left": "issue:43761", + "right": "issue:44743", + "accept": false, + "reason": "Different models and bugs: CLIPVisionModel hidden_states regression vs Qwen recurrent-state reset with cache." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Completely different problems: Python 3.13 JIT parse/indentation issue vs flash-attention crashing on s_aux=None." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "SwitchTransformers config construction bug is unrelated to the DebertaV2 Python 3.13 import/parser issue." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "A packaging/Python-version compatibility report does not match the specific DebertaV2 @torch.jit.script parsing failure." + }, + { + "left": "issue:44077", + "right": "issue:44079", + "accept": false, + "reason": "Both are model-output/config validation themed, but one is patchtsmixer post_init policy and the other is None-key assignment in ModelOutput." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "Torch.compile+SDPA failure in Bamba is unrelated to the image-classification script failing to load a local dataset." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Both are persistence-related, but one is buffer initialization junk and the other is GPTNeoX rotary_pct not reloading; not the same bug." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Video-input ValueError in Qwen3OmniMoe is unrelated to serve KV-cache continuation tensor indexing." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "A noisy generic issue title does not match the concrete audio feature-extraction bug." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "CI failures in different areas: GraniteSpeech stale device override test vs auto_mappings referencing removed Sam3LiteText configs." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Both mention example scripts, but one is multi-label classification output behavior and the other is local dataset loading for image classification." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Gemma4 embedding resize behavior and Whisper empty-transcription handling are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "Both touch examples/docs, but one is SigLIP2 documentation errors and the other is a local dataset loading problem." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification processor mismatch is unrelated to a notebook syntax error." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both are config-field issues, but they affect different multimodal model families and different missing fields." + }, + { + "left": "issue:44734", + "right": "issue:45632", + "accept": false, + "reason": "Serve KV-cache indexing and trust_remote_code cache-path collisions are different subsystems and defects." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but one is a specific missing Qwen2_5_VLImageProcessor symbol and the other is top-level AutoImageProcessor availability." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 versus incorrect Qwen3.5 checkpoint saving are different model/regression bugs." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Processor/tokenizer regression and Gemma4 resize_token_embeddings behavior do not share the same concrete code path." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Category-cardinality runtime error is unrelated to Whisper's empty-transcription failure." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Different missing imports in different modules; no shared underlying code change is evident." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error and audio feature extraction bug are unrelated." + }, + { + "left": "issue:44561", + "right": "issue:45632", + "accept": false, + "reason": "Removal of a helper breaking trust_remote_code models is unrelated to local cache-path collisions." + }, + { + "left": "issue:44186", + "right": "issue:44734", + "accept": false, + "reason": "Tokenizer crashes on NER padding/truncation are not the same as serve KV-cache continuation indexing failures." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both mention flex_attention/PyTorch versions, but one is compiled flex_attention failing on new torch and the other is an incorrect version gate for AuxRequest import." + }, + { + "left": "issue:38175", + "right": "issue:43065", + "accept": false, + "reason": "SigLIP2 zero-probabilities and a dummy Conv2d in Sam3PixelDecoder are unrelated model bugs." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "A specific missing TensorFlow dependency from a torch install is not the same as a vague transformers dependency issue." + }, + { + "left": "issue:45406", + "right": "issue:45632", + "accept": false, + "reason": "Gemma4Processor _tokenizer crash and trust_remote_code cache collision are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "Flash-attention s_aux=None crash and removed Sam3LiteText config references breaking CI are different failures." + }, + { + "left": "issue:44466", + "right": "issue:44913", + "accept": false, + "reason": "Both concern state persistence, but one is lm_head.weight serialization and the other is GPTNeoX rotary_pct reload behavior." + } + ] + }, + "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": 14, + "review_unit_index": 29, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:41950", + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43577", + "issue:43643", + "issue:43673", + "issue:43824", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44112", + "issue:44242", + "issue:44246", + "issue:44279", + "issue:44336", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44933", + "issue:44964", + "issue:44987", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45081", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45325", + "issue:45335", + "issue:45357", + "issue:45405", + "issue:45584", + "issue:45593", + "issue:45600" + ], + "soft_pairs": [ + "issue:44246|issue:45003", + "issue:43643|issue:44496", + "issue:42907|issue:43441", + "issue:44016|issue:44062", + "issue:44964|issue:45600", + "issue:43299|issue:45216", + "issue:43295|issue:45335", + "issue:43976|issue:44987", + "issue:43673|issue:44336", + "issue:44112|issue:44964", + "issue:43824|issue:44933", + "issue:43531|issue:45600", + "issue:43976|issue:44279", + "issue:44496|issue:45042", + "issue:43854|issue:44964", + "issue:43526|issue:43901", + "issue:45335|issue:45584", + "issue:43299|issue:45357", + "issue:44075|issue:44368", + "issue:44521|issue:45325", + "issue:43577|issue:45216", + "issue:43381|issue:45593", + "issue:44351|issue:45278", + "issue:43976|issue:45405", + "issue:41950|issue:45020", + "issue:43901|issue:44496", + "issue:44242|issue:45200", + "issue:43479|issue:43525", + "issue:44462|issue:45081" + ], + "prepared_review_unit_hash": "c7ea46c1a54a2e80ddce8ffefe3f500e239089a6f2e3ef54520abd7bf7889388", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35784, + "estimated_input_tokens": 8946, + "estimated_eval_tokens": 18148 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a collection of related-but-distinct Transformers regression reports, many around model loading, config handling, tokenizers, and multimodal support. There are no clear exact duplicates among the soft pairs, so all candidate edges are rejected. The broadest representative issue is the remote_code/model-loading regression report.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45020 is the broadest, most representative open report in the set: it describes a recent-version regression affecting model loading with remote_code, which is a common umbrella for several adjacent loading/config failures here.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 best captures the shared theme of recent-version loading regressions without being too narrowly tied to one model family or one specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import slowdown and unsafe sys.modules access are related to import-time behavior, but they are not the same concrete bug or fix path." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Both involve config/model loading, but one is missing fields with trust_remote_code and the other is an unrecognized model / missing model_type case." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Same model family, different failures: saving dequantized models vs FlashAttention load/runtime behavior." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "A notebook syntax error and a tokenizer AddedToken argument error are unrelated." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Phi-4 multimodal loading and stale auto_mappings references to Sam3LiteText are different breakages." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both are Qwen-related, but one is a model-loading failure and the other is an incorrect save_pretrained checkpoint issue." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Processor/tokenizer regression in multimodal code and a t5gemma embedding-resize bug are different code paths." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "A Python version compatibility issue is not the same bug as a model-loading failure for one specific repository." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Generation cache behavior and ANSI output formatting are unrelated." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "A CI test flake in GraniteSpeech is unrelated to Phi-4 multimodal loading." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Two import errors, but for different symbols and likely different causes; not the same fix." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and removed config references are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "A Python support/version issue and a vague dependency problem are too different to treat as the same bug." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Model config recognition and PIL/torchvision dependency handling are different issues." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Both are model-loading failures, but for different models and symptom types." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "A BeitImageProcessorFast bug and a docs/behavior mismatch in TextClassificationPipeline do not share a concrete fix path." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize behavior and Whisper empty-transcription generation are unrelated." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both concern Qwen3.5, but one is load-time breakage and the other is incorrect save_pretrained output keys." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer argument handling and a warning about tie_word_embeddings are unrelated." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal, but assistant mask generation and rope index scaling are different concrete bugs." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "BLIP2 dtype propagation and Qwen3.5 checkpoint saving are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary loss behavior are unrelated." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "The latter is a broad umbrella for many import errors, but the pair does not identify the same concrete import failure." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python support compatibility and an unreleased PEFT minimum version are different release-management issues." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Both mention model-loading/pipeline behavior, but image processor lookup in a video pipeline is a different bug from remote_code model loading regressions." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "A documentation/behavior mismatch and an unrecognized model/config issue are not the same underlying bug." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Load-balancing loss computation and mm_token_type_ids defaults are separate model-training concerns." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Phi4MultimodalConfig default initialization and Llama4Config pad_token_id absence are unrelated config bugs." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer.json precedence and a Mistral regex patch crash are different tokenizer-loading problems." + } + ] + }, + "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": 14, + "review_unit_index": 30, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41950", + "issue:42175", + "issue:42994", + "issue:43065", + "issue:43122", + "issue:43262", + "issue:43295", + "issue:43299", + "issue:43317", + "issue:43441", + "issue:43450", + "issue:43650", + "issue:43653", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44162", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44291", + "issue:44297", + "issue:44462", + "issue:44466", + "issue:44496", + "issue:44521", + "issue:44617", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:44877", + "issue:44898", + "issue:44987", + "issue:45276", + "issue:45290", + "issue:45397", + "issue:45405", + "issue:45542", + "issue:45593" + ], + "soft_pairs": [ + "issue:44683|issue:45542", + "issue:44496|issue:44877", + "issue:44297|issue:44462", + "issue:44279|issue:45405", + "issue:43299|issue:43854", + "issue:41950|issue:43650", + "issue:44462|issue:44779", + "issue:39692|issue:43901", + "issue:43295|issue:44779", + "issue:38175|issue:44617", + "issue:44016|issue:44291", + "issue:44261|issue:44263", + "issue:43723|issue:44162", + "issue:43450|issue:44279", + "issue:42994|issue:43317", + "issue:43065|issue:44898", + "issue:43441|issue:44683", + "issue:43650|issue:45290", + "issue:44466|issue:45276", + "issue:43262|issue:44521", + "issue:44623|issue:45397", + "issue:44623|issue:44749", + "issue:43673|issue:43761", + "issue:43784|issue:45542", + "issue:44671|issue:44779", + "issue:42175|issue:43976", + "issue:43122|issue:43650", + "issue:44987|issue:45542", + "issue:43653|issue:45593" + ], + "prepared_review_unit_hash": "d9725c8c877b54420f59f9685a27a17f9861fc869b61c2800a999a94cdf5205a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36273, + "estimated_input_tokens": 9069, + "estimated_eval_tokens": 18394 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is noisy: most items are unrelated v5 regressions across tokenizer loading, processors, quantization, vision, audio, and docs. None of the suggested soft edges look like true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical issue is strong here; the set mixes several different bug classes and model families rather than one underlying defect.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43723 is the broadest representative of the recurring tokenizer-loading regression theme in v5, but it still does not unify the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Unrelated domains: flex_attention on torch>=2.9 vs TensorFlow backend/package installation." + }, + { + "left": "issue:44496", + "right": "issue:44877", + "accept": false, + "reason": "Both involve model/config loading, but the failures are different: unrecognized model type vs strict config rejection for granite_speech." + }, + { + "left": "issue:44297", + "right": "issue:44462", + "accept": false, + "reason": "Both touch tokenizer serialization/loading, but one is save_pretrained tokenizer_class mismatch and the other is AutoTokenizer ignoring tokenizer.json." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Dependency complaints, but about different packages and failures; not the same bug." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different models and different loading failures; only a broad 'model loading broke in v5' theme is shared." + }, + { + "left": "issue:41950", + "right": "issue:43650", + "accept": false, + "reason": "Completely unrelated: video-classification processor bug vs an unrelated placeholder issue." + }, + { + "left": "issue:44462", + "right": "issue:44779", + "accept": false, + "reason": "Tokenizer-related, but one is repository tokenizer.json precedence and the other is Deepseek tokenization correctness; not the same concrete path." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs-related, but one is SigLIP2 example errors and the other is a TextClassificationPipeline docs mismatch." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Different regressions: processor.tokenizer/image passing breakage vs Deepseek tokenizer output changes." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "Different subsystems and symptoms: zero probabilities on SigLIP2 vs CUDA OOM in Sam3Video." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error vs init_empty_weights argument/type error; unrelated code paths." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Both are model-specific correctness bugs, but they affect different components and failure modes." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "Tokenizer loading regression vs ESM2 model breakage; too broad to be duplicates." + }, + { + "left": "issue:43450", + "right": "issue:44279", + "accept": false, + "reason": "Video processor batching shape bug vs dependency issue; unrelated." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Both involve quantization/dequantization, but one is saving and the other is device_map offload loading; not the same bug." + }, + { + "left": "issue:43065", + "right": "issue:44898", + "accept": false, + "reason": "Different model components: dummy Conv2d in Sam3PixelDecoder vs Perceiver interpolation failure." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "Both mention attention kernels, but the concrete failures are different code paths and different triggering conditions." + }, + { + "left": "issue:43650", + "right": "issue:45290", + "accept": false, + "reason": "Completely unrelated issues." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Both touch Gemma 4 embeddings, but one is serialization of lm_head.weight and the other is resize_token_embeddings not propagating to outputs." + }, + { + "left": "issue:43262", + "right": "issue:44521", + "accept": false, + "reason": "Both use apply_chat_template, but audio sampling-rate defaults and multimodal assistant masks are different bugs." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "processor.save_pretrained missing files vs gemma-4 zero3 from_pretrained; save/load failures but not the same defect." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "Missing processor files vs performance regression in filtering after upgrade; unrelated." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Generation cache/chunked_prefill issue vs CLIPVisionModel hidden_states regression; different models and APIs." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Import NameError in sentence-transformers integration vs TensorFlow backend installation issue." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "Both are v5 regressions affecting NLP outputs, but one is CamemBERT masked LM predictions and the other is Deepseek tokenizer behavior." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Backend packaging issue vs Python-version compatibility issue; not duplicates." + }, + { + "left": "issue:43122", + "right": "issue:43650", + "accept": false, + "reason": "Tokenizer behavior regression vs unrelated placeholder issue." + }, + { + "left": "issue:44987", + "right": "issue:45542", + "accept": false, + "reason": "Model loading failure vs TensorFlow installation/backend issue; unrelated." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "BigBirdTokenizer special-token bug vs D-FINE auxiliary-loss behavior; unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 31, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:43295", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43673", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", + "issue:44162", + "issue:44188", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44297", + "issue:44336", + "issue:44351", + "issue:44361", + "issue:44373", + "issue:44462", + "issue:44485", + "issue:44496", + "issue:44561", + "issue:44589", + "issue:44617", + "issue:44661", + "issue:44857", + "issue:45003", + "issue:45200", + "issue:45250", + "issue:45356", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45561", + "issue:45563", + "issue:45593" + ], + "soft_pairs": [ + "issue:43901|issue:43994", + "issue:44361|issue:45593", + "issue:43824|issue:43976", + "issue:44297|issue:45397", + "issue:43475|issue:44016", + "issue:43525|issue:43881", + "issue:45561|issue:45563", + "issue:43352|issue:45250", + "issue:44589|issue:45362", + "issue:44188|issue:44336", + "issue:41628|issue:43976", + "issue:43295|issue:44617", + "issue:42491|issue:43299", + "issue:44336|issue:44373", + "issue:41950|issue:44220", + "issue:43901|issue:44246", + "issue:44242|issue:45593", + "issue:44295|issue:45397", + "issue:44462|issue:45356", + "issue:44661|issue:45542", + "issue:45003|issue:45561", + "issue:43531|issue:45362", + "issue:44246|issue:44496", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:43673|issue:44561", + "issue:44336|issue:44485", + "issue:43526|issue:45003", + "issue:44162|issue:45200", + "issue:44242|issue:44857" + ], + "prepared_review_unit_hash": "351a01c8be83cb7059d32306ef1c7ae50421a07d66f58637ff8e69f87de47b17", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36322, + "estimated_input_tokens": 9081, + "estimated_eval_tokens": 18418 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a heterogeneous cluster of unrelated issues: import/export regressions, model-loading failures, pipeline behavior bugs, documentation problems, and testing/runtime edge cases. None of the proposed soft pairs looks like the same underlying bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 45003 is the broadest open bug in the set and has a concrete implementation focus in core code (`modeling_utils` / `sys.modules`). The other issues are mostly narrow, model-specific, or unrelated regressions, so 45003 is the least arbitrary cluster anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "45003 is the best representative issue because it is open, implementation-oriented, and broader than the many one-off model/import/doc issues. It is still not a true duplicate of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Docs wording about `return_all_scores` is unrelated to a SigLIP2 model/pipeline correctness problem." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer AttributeError on tasks and D-FINE auxiliary-loss behavior affect different code paths and model families." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Missing `Qwen2_5_VLForConditionalGeneration` export is a specific symbol issue, not the Python-version compatibility problem described by 43976." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer config mismatch in `save_pretrained` is unrelated to Gemma-4 `from_pretrained`/zero3 loading." + }, + { + "left": "issue:43475", + "right": "issue:44016", + "accept": false, + "reason": "A SAM3 vision encoder attribute error is unrelated to a notebook syntax error." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Missing `pad_token_id` on `Llama4Config` is a different loading bug than the GLM-4V-9B failure." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "Pytest xdist file-race behavior is unrelated to a stale `num_return_sequences` warning in paged generation." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "One is a model-specific unsupported feature error; the other is just the generic Flash Attention 2.0 topic." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage deserialization failure is unrelated to the Qwen3.5 chat crash." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI-code output in `loading_report`." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "`AutoImageProcessor` import failure is a different missing-symbol issue than the Python-version breakage in 43976." + }, + { + "left": "issue:43295", + "right": "issue:44617", + "accept": false, + "reason": "Processor/tokenizer regression in custom model code is unrelated to SAM3 video CUDA OOM." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both mention Qwen3 MoE, but one is a LoRA compatibility problem and the other is a model-loading regression; they are not clearly the same fix." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI terminal output and a wrong docstring for `position_ids` are unrelated." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "A video-classification pipeline bug and `_torch_extract_fbank_features()` are different pipeline/code-path issues." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "A pipeline docs mismatch is unrelated to intermittent import-time slowness." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Missing load-balancing loss with `output_router_logits=False` is a different training-loss bug than D-FINE auxiliary-loss handling." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "Reading `position_ids` after buffer registration is unrelated to Gemma-4 zero3 loading." + }, + { + "left": "issue:44462", + "right": "issue:45356", + "accept": false, + "reason": "Ignoring `tokenizer.json` is a tokenizer-loading bug, while 45356 is a codec/regex regression for Kimi-K2.5." + }, + { + "left": "issue:44661", + "right": "issue:45542", + "accept": false, + "reason": "`TOKENIZER_MAPPING_NAMES` handling and undefined TensorFlow backend detection are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe `sys.modules` access in core modeling code is unrelated to the xdist captured-info race." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and Qwen3.5 chat crashes are different model/runtime bugs." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import slowness and unrecognized-model config loading are unrelated." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Different missing import symbols (`Qwen2_5_VLImageProcessor` vs `is_offline_mode`) with no shared code-path shown." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Missing `Qwen2_5_VLImageProcessor` and missing `HybridCache` are separate export/import failures." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Chunked-prefill cache handling and removal of `is_torch_fx_available` affect different runtime paths." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI code emission is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "A BEiT image-processor label bug is unrelated to core modeling sys.modules access." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma-4 `mm_token_type_ids` defaults are different model-specific regressions." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Load-balancing loss omission is unrelated to LwDetrImageLoss AMP crashes." + } + ] + }, + "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": 14, + "review_unit_index": 32, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42886", + "issue:43065", + "issue:43066", + "issue:43232", + "issue:43257", + "issue:43262", + "issue:43408", + "issue:43441", + "issue:43475", + "issue:43540", + "issue:43596", + "issue:43653", + "issue:43673", + "issue:43901", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44368", + "issue:44466", + "issue:44496", + "issue:44734", + "issue:44749", + "issue:44779", + "issue:44871", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45459", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45656", + "issue:45701" + ], + "soft_pairs": [ + "issue:44246|issue:44871", + "issue:36246|issue:43596", + "issue:44295|issue:44749", + "issue:43066|issue:44779", + "issue:43441|issue:43540", + "issue:44877|issue:45375", + "issue:44987|issue:45459", + "issue:42886|issue:45229", + "issue:41628|issue:41950", + "issue:44779|issue:45292", + "issue:45292|issue:45584", + "issue:43065|issue:43408", + "issue:43232|issue:44336", + "issue:45292|issue:45701", + "issue:43901|issue:45003", + "issue:44995|issue:45200", + "issue:43257|issue:45137", + "issue:44060|issue:44077", + "issue:44496|issue:45003", + "issue:43475|issue:43540", + "issue:43653|issue:44162", + "issue:44995|issue:45588", + "issue:44368|issue:45563", + "issue:43262|issue:44246", + "issue:44368|issue:44977", + "issue:43673|issue:44734", + "issue:44466|issue:45335", + "issue:45397|issue:45656" + ], + "prepared_review_unit_hash": "fe0fdc116d6a004fdd9dc12508a956159014ee18625f52f0c6f1a6f1e8dc466a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36192, + "estimated_input_tokens": 9048, + "estimated_eval_tokens": 18352 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous, with only loose thematic overlap around Transformers v5 regressions, tokenizers, and embedding/tied-weight behavior. I do not see any soft pair that is clearly the same underlying bug or change, so I rejected all candidate duplicates.", + "confidence": 0.39, + "canonical_issue_reason": "Issue 45292 is the clearest standalone bug report: it names a concrete, reproducible API regression in `resize_token_embeddings` and is phrased broadly enough to serve as a duplicate target for closely related embedding-resize reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the best issue candidate in this set because it describes a specific code-path failure cleanly and is more reusable as a triage target than the more model-specific or symptom-heavy reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import performance and Gemma EOS configuration are unrelated bugs." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Different failures: an import error for a vision image processor vs a DeepSpeed/Bert forward crash." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Registering `position_ids` as a buffer is unrelated to the reported tokenizer slowdown after upgrade." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer-related, but one is a wrong decoder-type warning and the other is incorrect Deepseek tokenization; not the same concrete bug." + }, + { + "left": "issue:43441", + "right": "issue:43540", + "accept": false, + "reason": "FlashAttention failure for Ministral-3 is unrelated to Qwen3OmniMoe video input processing." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both involve config handling, but they describe different model fields and opposite failure modes, so they are not the same bug." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model-loading failure for `physical-intelligence/fast` is unrelated to protobuf-masked tokenizer errors." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline cache loading and Gemma4 CUDA OOM are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:41950", + "accept": false, + "reason": "Missing `AutoImageProcessor` import and a video-classification pipeline misdetecting processors are different issues." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Tokenizer correctness regression and embedding resize not updating output embeddings are different code paths." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize behavior is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 dummy Conv2d implementation and SAM3 model-type warning are not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:44336", + "accept": false, + "reason": "Generation kwargs/sync-gpus logic has nothing to do with ANSI code emission in loading reports." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Embedding resizing and version-dependent tokenization changes are different regressions." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "A docs mismatch for `return_all_scores` is unrelated to unsafe `sys.modules` access." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "A second-forward cache crash in GlmMoeDsa is unrelated to Gemma4 token type defaults." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both touch DeepSpeed, but one is MOE weight conversion and the other is a ZeRO3 deque error; not the same bug." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Tied-weights warning in Qwen3-Next and `post_init` allowance in patchtsmixer are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Model config loading failure and unsafe `sys.modules` access are not the same underlying issue." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "Missing `fpn_position_embeddings` in SAM3 output is unrelated to Qwen3OmniMoe video processing." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are different tokenizer/model problems." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Stale indexer cache crash and FlashAttention `s_aux=None` crash are different failures." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A tied-word-embeddings warning and a stale `num_return_sequences` warning are not the same bug." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio chat-template sampling-rate defaults and import latency are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "A tied-embedding warning is not the same as a generation failure with FlashAttention." + }, + { + "left": "issue:43673", + "right": "issue:44734", + "accept": false, + "reason": "Generation cache missing during chunked prefill and a KV-cache indexing crash in transformers serve are different code paths." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both concern weights/embeddings, but one is serialization inconsistency and the other is resize propagation; they are not clearly the same bug." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma4 Zero3 loading and duplicate optimizer stepping under DeepSpeed are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 33, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43704", + "issue:43873", + "issue:43931", + "issue:44038", + "issue:44162", + "issue:44188", + "issue:44222", + "issue:44297", + "issue:44361", + "issue:44368", + "issue:44623", + "issue:44661", + "issue:44749", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45200", + "issue:45230", + "issue:45278", + "issue:45310", + "issue:45356", + "issue:45362", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45459", + "issue:45478", + "issue:45600", + "issue:45656" + ], + "soft_pairs": [ + "issue:44933|issue:45459", + "issue:44869|issue:45356", + "issue:43931|issue:44368", + "issue:45137|issue:45600", + "issue:43232|issue:43673", + "issue:43673|issue:45278", + "issue:44297|issue:44623", + "issue:43646|issue:44162", + "issue:44361|issue:44749", + "issue:44623|issue:45656", + "issue:28282|issue:30990", + "issue:44749|issue:45310", + "issue:42994|issue:43482", + "issue:42371|issue:45092", + "issue:44038|issue:44623", + "issue:44749|issue:45656", + "issue:42673|issue:43704", + "issue:43475|issue:43643", + "issue:45230|issue:45397", + "issue:45375|issue:45478", + "issue:43475|issue:45381", + "issue:41628|issue:44933", + "issue:45310|issue:45375", + "issue:45081|issue:45103", + "issue:42994|issue:43525", + "issue:43475|issue:44661", + "issue:45362|issue:45375", + "issue:42831|issue:44222", + "issue:43873|issue:44945", + "issue:43232|issue:44188", + "issue:42222|issue:45200" + ], + "prepared_review_unit_hash": "52199f130e253451c8936cfbb0c44da0f5db32a34afedf23058c0eb30f7158e2", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36717, + "estimated_input_tokens": 9180, + "estimated_eval_tokens": 18616 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous Transformers breakages that share broad upgrade/import/model-loading themes, not one single bug. The clear duplicate pair in the soft-edge list is the repeated Qwen3ForCausalLM VRAM leak report (42673/43704), and the packet also contains another exact-title duplicate pair, 45310/45478, for the Qwen3.5 MoE from_pretrained error. Outside those, the issues are mostly only loosely related by subsystem or model family.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is still the best umbrella issue for the cluster because it is open and broadly representative of the upgrade/import-regression theme, but it should be treated as an umbrella, not as evidence that the whole set is duplicate-heavy.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the strongest global issue candidate because it is open and broad enough to anchor triage across several upgrade-related reports, while the more specific items are better treated as separate bugs or duplicate subclusters.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44933", + "right": "issue:45459", + "accept": false, + "reason": "Both touch tokenizer/import failures, but the concrete paths differ: image_utils import vs. protobuf-related tokenizer error masking." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Different model areas and symptoms: Whisper timestamp decoding crash vs. Kimi tokenizer codec/regression handling." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Loading-time shape mismatch for a VL model and a tie_word_embeddings warning are not the same bug." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error and stale auto_mappings config references are distinct code-path failures." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both are generation-related, but one is a sync_gpus kwargs update issue and the other is a missing cache path during chunked prefill." + }, + { + "left": "issue:43673", + "right": "issue:45278", + "accept": false, + "reason": "A specific generation cache bug is not the same concrete bug as a broad report about many import errors after upgrading." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Tokenizer metadata saving and processor file saving are different save_pretrained failure modes." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "trust_remote_code field loss and ESM2 being broken are unrelated bugs." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Tokenizer AttributeError on tasks and a performance regression after upgrading are not the same underlying issue." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Processor save_pretrained file loss and deepspeed optimizer double-step are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "Missing PyTorch ImportError and Sentence Transformers getting stuck loading are different problems." + }, + { + "left": "issue:44749", + "right": "issue:45310", + "accept": false, + "reason": "A data filtering slowdown and a Qwen3.5 MoE from_pretrained failure are not the same bug." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are separate serialization/loading issues." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API guidance and InternVL2 meta-initialization incompatibility are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe/transformers-v5 incompatibility and processor.save_pretrained missing files are different failures." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Upgrade-related slowdown and deepspeed optimizer step duplication are different failure modes." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same symptom: Qwen3ForCausalLM VRAM leak in multiple dataloader threads, so these are the same underlying bug." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "A missing attribute on Sam3VisionEncoderOutput is not the same as missing fields in AutoConfig under trust_remote_code." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Generic bug report vs. a specific Gemma-4 ZeRO3 from_pretrained issue; no shared concrete code path." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 MoE/config loading, but one is a specific missing-field report and the other is a broader from_pretrained error; not enough evidence they are the same bug." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "SAM 3 video missing attribute and qwen2.5-vl video position-id regression are different model bugs." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Two import errors, but from different modules and different root causes; too broad to merge." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "The former is a model-loading error report; the latter is a specific config-field omission. They may be related, but not clearly the same concrete bug." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Tokenizer regex patch crash and auto-docstring annotation handling crash are unrelated." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized save_pretrained failure and Llama4Config missing pad_token_id are distinct issues." + }, + { + "left": "issue:43475", + "right": "issue:44661", + "accept": false, + "reason": "SAM 3 output attribute failure and add-new-model-like tokenizer mapping failure are different code paths." + }, + { + "left": "issue:45362", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3.5-35B chat crash and a Qwen3_5MoeVisionConfig field being dropped are not the same bug." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "FineGrainedFP8 accuracy issues and FP8 save_pretrained issues are related only at a high level, not the same bug." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading and incorrect output under pipeline parallelism are different problems." + }, + { + "left": "issue:43232", + "right": "issue:44188", + "accept": false, + "reason": "Generation kwargs sync bug and diverging attention kernels under torch.compile are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "VitPose model breakage and Gemma 4 mm_token_type_ids defaults are different model-specific issues." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is not fully grounded in the packet: it says the only clear duplicate pair is the Qwen3ForCausalLM VRAM-leak report, but the packet also contains two issues with the exact same title, `issue:45310` and `issue:45478` (`[BUG] transformers>=5.4.0, Qwen3.5 Moe from_pretrained error`). That omission makes the duplicate/umbrella assessment too strong. Several soft-edge false verdicts are conservative, but the high-level summary overstates the uniqueness of the VRAM-leak pair." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 34, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:36010", + "issue:41628", + "issue:42175", + "issue:42898", + "issue:42994", + "issue:43232", + "issue:43278", + "issue:43377", + "issue:43479", + "issue:43504", + "issue:43526", + "issue:43673", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43976", + "issue:44038", + "issue:44060", + "issue:44079", + "issue:44164", + "issue:44188", + "issue:44222", + "issue:44242", + "issue:44351", + "issue:44373", + "issue:44448", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:44869", + "issue:44933", + "issue:44964", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45137", + "issue:45310", + "issue:45356", + "issue:45459", + "issue:45561" + ], + "soft_pairs": [ + "issue:41628|issue:44351", + "issue:44222|issue:45310", + "issue:42994|issue:44164", + "issue:44038|issue:45310", + "issue:43526|issue:45561", + "issue:43673|issue:44188", + "issue:44496|issue:45561", + "issue:36010|issue:41628", + "issue:43504|issue:45561", + "issue:43377|issue:44242", + "issue:43232|issue:44792", + "issue:42898|issue:44568", + "issue:30990|issue:35141", + "issue:43479|issue:44079", + "issue:43479|issue:44869", + "issue:43976|issue:44933", + "issue:42175|issue:44933", + "issue:44964|issue:45137", + "issue:44188|issue:44373", + "issue:44623|issue:45310", + "issue:43784|issue:43824", + "issue:43723|issue:44568", + "issue:44060|issue:45125", + "issue:43278|issue:45137", + "issue:42898|issue:45356", + "issue:44734|issue:45103", + "issue:43874|issue:45092", + "issue:42898|issue:44779", + "issue:30333|issue:33290", + "issue:42898|issue:44448", + "issue:41628|issue:45459" + ], + "prepared_review_unit_hash": "78fc8464d3e7c20c947f8558813743b43c6421db49c7e89129c3d34454a4986b", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36970, + "estimated_input_tokens": 9243, + "estimated_eval_tokens": 18742 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly separate Transformers regression reports spanning import/export issues, tokenizer behavior changes, model-specific load/save bugs, and generation/runtime edge cases. The soft pairs share broad themes, but none look like the same concrete underlying defect.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import failures, but they concern different missing symbols (`AutoImageProcessor` vs `HybridCache`) and likely different export paths." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "One is FP8 `save_pretrained` for MoE models, the other is a `from_pretrained` load error for Qwen3.5 MoE; different phases and failure modes." + }, + { + "left": "issue:42994", + "right": "issue:44164", + "accept": false, + "reason": "Both involve saving/loading, but one is quantized model saving and the other is `extra_state` handling in save/from_pretrained plumbing." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen/MoE, but the model families and concrete breakage differ (`Qwen3-VL-Moe` vs `Qwen3.5 Moe` from_pretrained error)." + }, + { + "left": "issue:43526", + "right": "issue:45561", + "accept": false, + "reason": "These are unrelated: one is a BEiT label reduction bug, the other is a pytest-xdist race in captured test output." + }, + { + "left": "issue:43673", + "right": "issue:44188", + "accept": false, + "reason": "Both touch generation/runtime behavior, but one is missing cache support in chunked prefill while the other is attention-kernel divergence under `torch.compile`." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "An unrecognized model config error is unrelated to a test-worker race on `captured_info.txt`." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they are for different APIs and modules, so they are not the same bug." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "A legacy-field load failure in BEiT segmentation is unrelated to the xdist file race." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Different model behaviors: missing padding-mask support in MIMI encoder vs gating/load-balancing loss omission when router logits are off." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "One is a generation kwargs update bug after `sync_gpus`; the other is a Janus image-generation test failure." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is `clean_up_tokenization_spaces` behavior and the other is `add_special_tokens` not inserting BOS/EOS." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "A Sentence Transformers loading hang and a token-embedding reinitialization bug are unrelated code paths." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Both mention `None` handling, but one is multimodal config initialization and the other is `ModelOutput` key assignment logic." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Different subsystems: multimodal config defaults vs Whisper word-timestamp decoding crash." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "A Python-version compatibility issue is unrelated to a missing `image_utils` import." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Package extras/backend composition and a missing import are distinct problems." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Loading a specific Phi-4 multimodal checkpoint and a DeepSpeed ZeRO3 deque underflow are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "A runtime attention-kernel branching bug is not the same as a docstring mismatch." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor `save_pretrained` missing files and Qwen3.5 MoE `from_pretrained` failure are different serialization directions and targets." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but one is `nn` missing during sentence-transformers import and the other is a missing Qwen2.5-VL class export." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Both involve tokenizers, but one is tokenizer loading via `AutoTokenizer.from_pretrained` and the other is special-token insertion semantics." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Both are Qwen3-era model issues, but one is an incorrect tied-weights warning and the other is a missing `_tp_plan` for tensor parallelism." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "An embedding dtype drift between train/eval is unrelated to a ZeRO3 deque pop error." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Different tokenizer regressions: cleanup-space behavior versus Kimi codec/regex handling." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLFlow reporting on failed training jobs and a deepspeed OOM are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "One is a tokenizer cleanup option change, the other is different model outputs for Pegasus between v4 and v5." + }, + { + "left": "issue:41628", + "right": "issue:45459", + "accept": false, + "reason": "A missing exported class and tokenizer error masking via protobuf exception handling are different import/error-handling 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": 14, + "review_unit_index": 35, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30333", + "issue:30990", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:38175", + "issue:41628", + "issue:42757", + "issue:42831", + "issue:42886", + "issue:42898", + "issue:42994", + "issue:43295", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43756", + "issue:43824", + "issue:44038", + "issue:44060", + "issue:44206", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44442", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44514", + "issue:44623", + "issue:44871", + "issue:44913", + "issue:44977", + "issue:44991", + "issue:45092", + "issue:45103", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45459", + "issue:45656" + ], + "soft_pairs": [ + "issue:44295|issue:45656", + "issue:42898|issue:43295", + "issue:44222|issue:44297", + "issue:44297|issue:45656", + "issue:43479|issue:43644", + "issue:44442|issue:45459", + "issue:42831|issue:44623", + "issue:44222|issue:44623", + "issue:44206|issue:45103", + "issue:41628|issue:42757", + "issue:22355|issue:28282", + "issue:44263|issue:44360", + "issue:44297|issue:45310", + "issue:44222|issue:44484", + "issue:30333|issue:36331", + "issue:44263|issue:45092", + "issue:43723|issue:44977", + "issue:44483|issue:45406", + "issue:30333|issue:34689", + "issue:28282|issue:30333", + "issue:30333|issue:35141", + "issue:44295|issue:44623", + "issue:44373|issue:44485", + "issue:44261|issue:45092", + "issue:43723|issue:44871", + "issue:44514|issue:45103", + "issue:43723|issue:44991", + "issue:45092|issue:45357", + "issue:43723|issue:44442", + "issue:43479|issue:44913", + "issue:30990|issue:38175", + "issue:42886|issue:42994", + "issue:41628|issue:43824", + "issue:42831|issue:45310", + "issue:42831|issue:44038", + "issue:44263|issue:44373", + "issue:42757|issue:45278", + "issue:43756|issue:44060" + ], + "prepared_review_unit_hash": "42d2da78a2984a97224d7c2df82c299fe0dfebdea278f4dcf5524e6672417374", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38851, + "estimated_input_tokens": 9713, + "estimated_eval_tokens": 19682 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous and mostly contains separate bug reports around imports, tokenizers, config serialization, and model-specific regressions. No soft-edge pair looks like the same underlying bug/change, so all candidate links should be rejected.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: it describes widespread import errors after a Transformers version upgrade, which is the closest fit to the cluster\u2019s repeated import/load-regression theme.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative issue because it is the most general, user-facing, and cluster-like report; it can plausibly anchor the broader import/error regressions in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different bugs: position_ids buffer access vs optimizer step being called twice under DeepSpeed." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Tokenizer cleanup behavior change is unrelated to processor.tokenizer/image-passing regression." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8 MoE saving and the other is tokenizer metadata mismatch." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata bug is unrelated to DeepSpeed optimizer-step duplication." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Multimodal config defaults and non-persistent buffer junk are different initialization/storage problems." + }, + { + "left": "issue:44442", + "right": "issue:45459", + "accept": false, + "reason": "AutoTokenizer failing on a specific tokenizer is different from protobuf exception masking during tokenizer decode." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FineGrainedFP8 accuracy and processor.save_pretrained missing files are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both touch saving, but FP8 MoE save failure is not the same as missing processor files." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "Unsupported feature argument in LasrFeatureExtractor is unrelated to auto-docstring annotation parsing." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they are for different symbols and different packages." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Missing transformers.onnx module and missing PyTorch are separate dependency/import failures." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Same broad DSA area, but torch.split return handling is not the same as a missing ReLU in the indexer." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer class metadata mismatch is unrelated to Qwen3.5 from_pretrained loading error." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save failure is not the same issue as max_shard_size default behavior." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting on failed jobs is unrelated to the trainer compute_loss signature change." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "DSA indexer split semantics and remote-code/meta-init incompatibility are different code paths." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading regression and flash-attention generation failure are different symptoms and fixes." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat-completions request rejection is unrelated to Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "MLflow job reporting and Llama model loading regression are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:30333", + "accept": false, + "reason": "Missing PyTorch import error is unrelated to MLflow not reporting failed training jobs." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "Training-job reporting and embedding reinitialization after resize are unrelated bugs." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Position_ids buffer read error is unrelated to processor save_pretrained file omission." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Wrong docstring for position_ids is documentation-only and not the same as GLM-5 RoPE behavior." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "MLA layernorm epsilon precision issue is unrelated to InternVL2 meta-initialization incompatibility." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer loading regression and eos_token_id configuration mismatch are different model/tokenizer issues." + }, + { + "left": "issue:44514", + "right": "issue:45103", + "accept": false, + "reason": "Batched chat-template crash is unrelated to annotation-handling in auto-docstring." + }, + { + "left": "issue:43723", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-load failures, but they concern different models and likely different failure modes." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Remote-code meta-init incompatibility is different from incorrect visual-encoder keys on save_pretrained." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Config defaults for multimodal models and GPTNeoX rotary_pct reload behavior are separate config bugs." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "SentenceTransformers loading stall and SigLIP2 zero probabilities are unrelated model-specific problems." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Offline tokenizer cache loading and quantized model saving are different subsystems." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different classes and likely different missing exports." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy issue is unrelated to Qwen3.5 MoE from_pretrained errors." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FP8 accuracy problems and Qwen3-VL-Moe loading bugs are not the same underlying issue." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "A torch.split return-value bug is unrelated to a wrong docstring for position_ids." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Missing is_offline_mode from huggingface_hub is a specific dependency mismatch, not the same as broad import errors after upgrading Transformers." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are distinct architecture/configuration 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": 14, + "review_unit_index": 36, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:29127", + "issue:30990", + "issue:33357", + "issue:36296", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43576", + "issue:43784", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44117", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44485", + "issue:44488", + "issue:44514", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44743", + "issue:44779", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45200", + "issue:45245", + "issue:45278", + "issue:45290", + "issue:45593", + "issue:45600" + ], + "soft_pairs": [ + "issue:43232|issue:45593", + "issue:43576|issue:43784", + "issue:43122|issue:44521", + "issue:42886|issue:43066", + "issue:42371|issue:44263", + "issue:44117|issue:44704", + "issue:30990|issue:33357", + "issue:43976|issue:44945", + "issue:42907|issue:44945", + "issue:44704|issue:45290", + "issue:43122|issue:44514", + "issue:44295|issue:44297", + "issue:43232|issue:44743", + "issue:39401|issue:44488", + "issue:43408|issue:43475", + "issue:43232|issue:45245", + "issue:42907|issue:43937", + "issue:41628|issue:45278", + "issue:43784|issue:45600", + "issue:44360|issue:44485", + "issue:39401|issue:44521", + "issue:44261|issue:44373", + "issue:44625|issue:44829", + "issue:43116|issue:45200", + "issue:43122|issue:44779", + "issue:29127|issue:30990", + "issue:42371|issue:42831", + "issue:44038|issue:44297", + "issue:42907|issue:44964", + "issue:44038|issue:44295", + "issue:36296|issue:43116", + "issue:42907|issue:43452", + "issue:44038|issue:44222" + ], + "prepared_review_unit_hash": "1883f0137650b61085dcd3b6734d1605d3dea7cab5b786fc87016bf0209c2729", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36560, + "estimated_input_tokens": 9140, + "estimated_eval_tokens": 18536 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No duplicate cluster here: the soft pairs mostly connect unrelated regressions, API changes, and model-specific bugs rather than the same underlying issue.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue; the cluster is heterogeneous and the soft pairs are all false positives.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the broadest user-facing report in the set (multiple import errors after upgrading), so it is the best umbrella issue if one must be chosen.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Different bugs: generation kwargs after sync_gpus vs D-FINE auxiliary-loss handling." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "CLI env command breakage is unrelated to a sentence-transformers import NameError." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer output regression vs multimodal assistant_mask generation; different code paths." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong decoder type are separate issues." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 settings warning is unrelated to torch.split return-value handling." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Both involve loading logic, but one is mapping-name None handling and the other is cached_file kwargs propagation." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Sentence-Transformers loading hang and Mac bus error on CLIP are different failures." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version support breakage is not the same as pipeline parallelism producing wrong outputs." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Dequantized save failure and incorrect PP outputs are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "Processor cached_file kwargs and chat-template crashes on tool calls are different APIs." + }, + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Tokenization regression and batched apply_chat_template crash are distinct bugs." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Position_ids buffer access and tokenizer_class mismatch on save_pretrained are different serialization problems." + }, + { + "left": "issue:43232", + "right": "issue:44743", + "accept": false, + "reason": "Generation update after sync_gpus is unrelated to Qwen3.5 recurrent-state reset with cache." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Qwen3 offset_mapping error and sleng-bert load failure are unrelated tokenizer/model-loading issues." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Model-type warning and missing attribute in SAM 3 Video are not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:45245", + "accept": false, + "reason": "Generation kwargs bug and category-count runtime error are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Model save failure for dequantized Ministrals/Devstrals does not match invalid GenerationConfig." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Both mention imports, but AutoImageProcessor missing is not the same concrete regression as the broad import-error report." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Sentence-transformers import NameError and auto_mappings CI break from removed configs are different import failures." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "DSA indexer ReLU discussion and GLM-5 RoPE implementation are unrelated model internals." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Wrong offset_mapping and zero assistant_masks are both tokenization-adjacent but not the same underlying bug." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "Numerical precision issue in MLA q_a_layernorm is unrelated to a docstring correction." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is unrelated to flash_attention_3 degenerate training." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example-script multi-label bug and Gemma 4 mm_token_type_ids defaulting are different training/fine-tuning issues." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they involve different models and different failure modes." + }, + { + "left": "issue:29127", + "right": "issue:30990", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and Sentence-Transformers loading hang are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 behavior settings and FineGrainedFP8 accuracy are different numerical-stability topics." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe v5 bug and tokenizer_class mismatch on save are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Saving dequantized Ministrals/Devstrals and loading Phi-4 multimodal are different model-loading failures." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe regression and position_ids buffer read error do not share the same code path." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor-parallel training bug and empty multi-label classification results are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "GGUF from_pretrained breakage and failing to save dequantized models are different format-handling bugs." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Both mention MoE, but one is a v5 model bug and the other is FP8 save_pretrained behavior." + } + ] + }, + "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": 14, + "review_unit_index": 37, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:33453", + "issue:38617", + "issue:39401", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42757", + "issue:42907", + "issue:43064", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43441", + "issue:43493", + "issue:43673", + "issue:43784", + "issue:43824", + "issue:43825", + "issue:43827", + "issue:43874", + "issue:43931", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44222", + "issue:44351", + "issue:44485", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:44945", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45216", + "issue:45278", + "issue:45341", + "issue:45446", + "issue:45478", + "issue:45561", + "issue:45563", + "issue:45588" + ], + "soft_pairs": [ + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:43992|issue:44704", + "issue:42757|issue:44351", + "issue:42907|issue:44222", + "issue:43122|issue:44977", + "issue:43278|issue:44945", + "issue:43784|issue:45446", + "issue:43937|issue:44485", + "issue:38617|issue:45278", + "issue:42222|issue:43441", + "issue:42673|issue:45072", + "issue:43064|issue:43673", + "issue:22355|issue:29127", + "issue:42371|issue:43874", + "issue:43116|issue:43493", + "issue:44075|issue:44936", + "issue:44038|issue:45478", + "issue:44222|issue:45478", + "issue:43976|issue:45446", + "issue:43116|issue:43825", + "issue:43937|issue:45563", + "issue:33453|issue:39401", + "issue:45083|issue:45446", + "issue:39401|issue:44779", + "issue:43931|issue:44945", + "issue:44117|issue:45588", + "issue:43827|issue:45446", + "issue:43937|issue:44977", + "issue:45341|issue:45561" + ], + "prepared_review_unit_hash": "77e24539e91bbae2c7d72f8b591ea983e10a0309f6a8c13ebacfbfe924774449", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35672, + "estimated_input_tokens": 8918, + "estimated_eval_tokens": 18092 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a true duplicate set; it mixes unrelated Transformers v5 regressions, model-specific bugs, import failures, and docs/API issues. The closest cluster anchor is the broad import-regression report, but most pairs should stay separate.", + "confidence": 0.92, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: it describes multiple import errors after upgrading from 4.57.0 to 5.5.0, which makes it a better anchor than the narrower model- or API-specific issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative issue for this loose group because it captures the widest user-facing regression pattern, while the other issues are mostly specific to one model, one symbol, or one code path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import failures, but they are different missing symbols (`Qwen2_5_VLForConditionalGeneration` vs `HybridCache`) with different likely causes and fixes." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5/Moe save-or-load regressions, but one is about incorrect saved checkpoints and the other is a from_pretrained load error; not the same concrete bug." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "`embed_tokens.weight` missing in `UMT5Encoder.from_pretrained` is a model-loading state issue, while `AutoProcessor.from_pretrained` not passing kwargs is a processor/file-resolution bug." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "These are distinct import errors from different modules and symbols; they do not point to one shared underlying regression." + }, + { + "left": "issue:42907", + "right": "issue:44222", + "accept": false, + "reason": "One is about saving dequantized Ministral/Devstral models, the other about FP8 save_pretrained behavior for moe models; same broad save path, but not the same bug." + }, + { + "left": "issue:43122", + "right": "issue:44977", + "accept": false, + "reason": "Both mention tokenization differences, but one is a version-to-version tokenizer behavior change and the other is a Qwen3.5 flash-attention generation problem." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype mismatch in evaluate and incorrect output with pipeline parallelism are separate training/eval runtime issues with different code paths." + }, + { + "left": "issue:43784", + "right": "issue:45446", + "accept": false, + "reason": "One is a missing `nn` import in sentence-transformers integration, the other is an incorrect PyTorch version gate for AuxRequest in flex_attention; unrelated failures." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "A generation-config validation error for GLM-5 is not the same as a RoPE implementation discussion/bug; they affect different parts of the model stack." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but one is a missing `layer_type_validation` symbol and the other is a broad upgrade report with multiple import errors; not one concrete bug." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Vitpose model breakage and Ministral-3 flash-attention failures are different model-specific regressions with different execution paths." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "A VRAM leak in multi-threaded Qwen3ForCausalLM use is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel bfloat16 inference." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption and missing GenerationMixin cache during chunked prefill are different subsystems and failure modes." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "`transformers.onnx` import failure and LayoutLMv3 box-validation messaging are entirely unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 behavior settings and a missing `get_number_of_image_patches` method in Glm46VImageProcessorFast are unrelated issues." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Multi-label classification example output emptiness and SigLIP2 HF-vs-JAX discrepancy are different problems in different model/example paths." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "SGD argument handling and `trainer.evaluate()` failing after `trainer.train()` are separate trainer/optimizer behaviors." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3-VL-Moe support in Transformers 5.0 and Qwen3.5 Moe from_pretrained errors are related by model family, but not the same concrete bug or fix." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "FP8 save_pretrained for moe and Qwen3.5 Moe from_pretrained load errors affect different code paths; they are not mergeable into one PR." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python 3.9/3.10 compatibility claims for Transformers 5.1.0 and a flex_attention import/version check bug are unrelated." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "A classification example script returning empty multi-label results is not the same as a pipeline() error message incorrectly advertising translation support." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GLM-5 generation-config validation and a stale warning in paged generate() are different generation-path issues." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression and Qwen3 offset_mapping errors are both tokenizer-related but refer to different regressions and likely different fixes." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "A helper-function behavior issue in qwen3_omni_moe is unrelated to the AuxRequest import/version-check bug in flex_attention." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer bugs, but one is Qwen3 offset_mapping and the other is Deepseek incorrect tokenization; different models and likely different root causes." + }, + { + "left": "issue:43931", + "right": "issue:44945", + "accept": false, + "reason": "Qwen3-VL-30B weight-shape mismatch and incorrect output under pipeline parallelism are different model-loading vs runtime-parallelism problems." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and a flash_attention crash on `s_aux=None` are unrelated implementation bugs." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() after removals and an incorrect PyTorch version check are not the same underlying bug or change." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "GLM-5 generation config validation and Qwen3.5 flash-attention generation failure are separate model-specific generation bugs." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "A testing_utils bug and an xdist race on captured_info.txt are both test-infra related, but they are not the same concrete race or code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 38, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42617", + "issue:42673", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43408", + "issue:43421", + "issue:43450", + "issue:43577", + "issue:43582", + "issue:43673", + "issue:43742", + "issue:43874", + "issue:43906", + "issue:43937", + "issue:43986", + "issue:44075", + "issue:44117", + "issue:44220", + "issue:44261", + "issue:44291", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44493", + "issue:44496", + "issue:44829", + "issue:44933", + "issue:45071", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45405", + "issue:45446", + "issue:45459", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45656" + ], + "soft_pairs": [ + "issue:44291|issue:45071", + "issue:45405|issue:45446", + "issue:42673|issue:43906", + "issue:43450|issue:43986", + "issue:39401|issue:43421", + "issue:44829|issue:45216", + "issue:38617|issue:41628", + "issue:44410|issue:45310", + "issue:44117|issue:45081", + "issue:44484|issue:45310", + "issue:42371|issue:44261", + "issue:42222|issue:45542", + "issue:43408|issue:43742", + "issue:43937|issue:44496", + "issue:42673|issue:43582", + "issue:42673|issue:43577", + "issue:42175|issue:42222", + "issue:38617|issue:44220", + "issue:44933|issue:45278", + "issue:44483|issue:45464", + "issue:44075|issue:45656", + "issue:42617|issue:45341", + "issue:43012|issue:43577", + "issue:43673|issue:45341", + "issue:43066|issue:43122", + "issue:44484|issue:45478", + "issue:44410|issue:45478", + "issue:43874|issue:44261", + "issue:43097|issue:44493", + "issue:42222|issue:45459" + ], + "prepared_review_unit_hash": "a76d19cbc6d7ee461bdbb3f6a22fbfc57fe17ed5a35cb3d3d8b41ec8962191c6", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36526, + "estimated_input_tokens": 9132, + "estimated_eval_tokens": 18520 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is highly heterogeneous: it mixes import-regression reports, tokenizer behavior bugs, model/config loading errors, backend/version checks, and unrelated feature/optimizer issues. No soft pair looks like a true duplicate; the apparent overlaps are only topical, not the same underlying bug.", + "confidence": 0.77, + "canonical_issue_reason": "issue:45278 is the broadest umbrella among the import-regression reports (multiple import errors after upgrading), so it is the best representative issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 has the widest scope and is most likely to subsume nearby import-failure reports, making it the best global issue representative in this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Different failures: unexpected `_is_hf_initialized` argument during init_empty_weights vs PreTrainedConfig type-checking regression." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "Unrelated backend/version problems: PEFT minimum version bump vs PyTorch version gating for flex_attention AuxRequest." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "Both mention Qwen3-related behavior, but one is a VRAM leak in dataloader threads and the other is a reproduction of a separate prior issue." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Different subsystems: batched video processor output shape vs AutoProcessor crash when torchvision is missing." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Tokenizer offset-mapping bug is not the same as runtime post-processor updates after special token modification." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Different concrete problems: flash_attention_3 causes degenerate training, while Qwen3.5 save_pretrained produces an incorrect checkpoint." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols and code paths." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Different model families and failures: missing projections in qwen3next vs Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer mapping None handling is separate from the Mistral regex patch crash in backend_tokenizer access." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "One is a save_pretrained shard-size question; the other is a Qwen3.5 MoE loading regression." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 control API concern is unrelated to MLA q_a_layernorm missing rms_norm_eps precision mismatch." + }, + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "Different domains: vitpose model breakage vs missing TensorFlow backend when only tensorboard is installed." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Different warnings/errors: sam3_video vs sam3_tracker type mismatch is not the same as MobileLLM loading key error." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "GenerationConfig validation for GLM-5 is unrelated to unrecognized model/config loading for Olmo-Hybrid-Instruct." + }, + { + "left": "issue:42673", + "right": "issue:43582", + "accept": false, + "reason": "Both are runtime issues, but VRAM leak in threads is not the same as Apple Silicon warmup TypeError." + }, + { + "left": "issue:42673", + "right": "issue:43577", + "accept": false, + "reason": "VRAM leak in threaded use is different from BLIP2 dtype fields staying float32 after loading." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "TensorFlow backend packaging issue is unrelated to vitpose model breakage." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "Different import/feature failures: missing layer_type_validation symbol vs audio fbank extraction issue." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "Same broad import-error theme, but 44933 is a specific missing import from image_utils while 45278 is a broader version-upgrade import breakage report." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Chat/completions API request rejection is not the same as streaming inference failure on Qwen3.5-0.8B." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "SGD optimizer args ignored and optimizer.step called twice are different optimizer bugs." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "3d_parallel.py runtime failure is unrelated to a testing_utils bug." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "PyTorch warning on compiling bfloat16 models is different from BLIP2 dtype fields remaining float32." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache missing in v5.0.0 during chunked_prefill is unrelated to a testing_utils bug." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both are tokenization-related, but one is decoder-type selection in v5 and the other is a concrete tokenization difference across versions." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen3.5 MoE loading bug vs unrelated save_pretrained max_shard_size question." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Missing qkv/gate projections in qwen3next is not the same as the Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing get_number_of_image_patches in GLM46V is separate from MLA q_a_layernorm rms_norm_eps precision error." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder API concern is different from unexpected position-id keys in model loading." + }, + { + "left": "issue:42222", + "right": "issue:45459", + "accept": false, + "reason": "Vitpose model breakage is unrelated to tokenizer error masking when protobuf is absent." + } + ] + }, + "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": 14, + "review_unit_index": 39, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:38617", + "issue:41628", + "issue:42757", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43335", + "issue:43352", + "issue:43408", + "issue:43441", + "issue:43452", + "issue:43502", + "issue:43531", + "issue:43824", + "issue:43825", + "issue:43873", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44351", + "issue:44368", + "issue:44380", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44704", + "issue:44829", + "issue:44987", + "issue:45003", + "issue:45071", + "issue:45103", + "issue:45245", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45542", + "issue:45588" + ], + "soft_pairs": [ + "issue:43441|issue:44829", + "issue:44263|issue:44485", + "issue:43012|issue:43408", + "issue:41628|issue:45071", + "issue:45071|issue:45542", + "issue:43931|issue:44410", + "issue:43986|issue:44484", + "issue:45003|issue:45341", + "issue:43931|issue:44038", + "issue:43986|issue:45341", + "issue:44117|issue:44987", + "issue:43097|issue:43452", + "issue:43335|issue:45071", + "issue:43408|issue:44368", + "issue:43874|issue:44263", + "issue:43976|issue:45071", + "issue:44829|issue:45245", + "issue:43986|issue:45372", + "issue:43531|issue:45341", + "issue:38617|issue:42757", + "issue:43064|issue:44315", + "issue:44261|issue:44485", + "issue:44410|issue:45375", + "issue:44038|issue:44410", + "issue:38617|issue:44351", + "issue:44062|issue:44117", + "issue:44704|issue:45103", + "issue:43097|issue:45588", + "issue:43986|issue:44117", + "issue:43502|issue:43992", + "issue:43824|issue:43986", + "issue:43097|issue:43873", + "issue:43066|issue:43906", + "issue:43352|issue:44380", + "issue:43825|issue:44655", + "issue:43992|issue:44484" + ], + "prepared_review_unit_hash": "1d71b0352666eec2bc7df93a43f1f5f30eb5ef9d00768f5377f00f311c8cb35a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38501, + "estimated_input_tokens": 9626, + "estimated_eval_tokens": 19508 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose title-similarity grouping, not a true duplicate set. The paired issues span unrelated import errors, model-specific loading bugs, training regressions, and warning/compatibility problems, so all soft edges should be rejected.", + "confidence": 0.41, + "canonical_issue_reason": "If a single issue must anchor the cluster, issue 43502 is the broadest core-loading regression with the most discussion, but the set is not actually one duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43502 is the best overall representative only because it is a high-signal, broadly relevant loading bug with substantial user impact and discussion; it is still not a strong canonical for the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both mention FlashAttention, but one is a model-specific FA2 failure and the other is a degenerate training bug with FA3; different code paths and symptoms." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "Unrelated GLM issues: one is a torch.split/indexer bug, the other is a RoPE implementation discussion." + }, + { + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "Both are warnings, but about different subsystems and causes: bfloat16 precision warning vs model-type mismatch warning." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "Different regressions: missing AutoImageProcessor import vs PretrainedConfig type-checking breakage." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Type checking in PretrainedConfig is unrelated to the TensorBoard/TF backend detection bug." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Same broad family name, but one is a Qwen3-VL load shape mismatch and the other is missing projections in qwen3next; not the same bug." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "Video-model AutoProcessor dependency crash is unrelated to the max_shard_size save_pretrained question." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils does not match a small testing_utils bug." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Both are Qwen-VL loading problems, but they affect different model variants and likely different failure modes." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "A torchvision dependency crash is unrelated to a testing utility bug." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping returning None is a distinct loading-path bug from the physical-intelligence/fast model load failure." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Removed tie_embeddings API and gguf_file breakage are different compatibility issues in different code paths." + }, + { + "left": "issue:43335", + "right": "issue:45071", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug is unrelated to PretrainedConfig type checking." + }, + { + "left": "issue:43408", + "right": "issue:44368", + "accept": false, + "reason": "Both are warnings, but they concern different configuration fields and different models." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Missing image-patch helper in a fast image processor is unrelated to a torch.split return-value bug." + }, + { + "left": "issue:43976", + "right": "issue:45071", + "accept": false, + "reason": "Python version compatibility for Transformers 5.1.0 is unrelated to the type-checking regression." + }, + { + "left": "issue:44829", + "right": "issue:45245", + "accept": false, + "reason": "FlashAttention training degeneration and a category-count limit error are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Different missing-dependency/import failures in different processor paths." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to a testing utility bug." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are ImportError reports, but for different missing symbols from different packages and likely different fixes." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to Liger Kernel not being applied under model_init." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "A missing rms_norm_eps precision issue is not the same as a GLM RoPE implementation discussion." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Missing qkv/gate projections in qwen3next is unrelated to a missing deepstack_visual_indexes config field." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Different Qwen model families and different architecture bugs; not a single underlying defect." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different modules/symbols; too broad to be one duplicate." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "Tokenizer AddedToken argument duplication is a different bug from TOKENIZER_MAPPING_NAMES sometimes returning None." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and auto_docstring annotation handling are unrelated code paths." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "A removed embedding-tie API and a flash_attention.py AttributeError with s_aux=None are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Video-model loading without torchvision is unrelated to tokenizer mapping assumptions." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "local_files_only network leakage and missing embed_tokens.weight are different loading bugs." + }, + { + "left": "issue:43824", + "right": "issue:43986", + "accept": false, + "reason": "A missing Qwen2.5-VL import and a torchvision-related video processor crash are distinct issues." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Removed tie_embeddings behavior is unrelated to quantization offloading not working." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "The tokenizer decoder type bug is not the same as a generic isolated reproduction of another issue." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Nemotron Flash Attention 2 support and GPT2 attention scaling under SDPA/FA are different attention bugs." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "pipeline() error messaging and Pipeline save_pretrained support are different pipeline-related problems." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing encoder weights on load is unrelated to the default max_shard_size discussion." + } + ] + }, + "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": 14, + "review_unit_index": 40, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:36010", + "issue:36683", + "issue:38617", + "issue:42757", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43381", + "issue:43475", + "issue:43502", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43582", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43867", + "issue:43901", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44220", + "issue:44230", + "issue:44263", + "issue:44368", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44492", + "issue:44514", + "issue:44655", + "issue:44683", + "issue:44938", + "issue:45216", + "issue:45290", + "issue:45381", + "issue:45440", + "issue:45479" + ], + "soft_pairs": [ + "issue:43097|issue:43901", + "issue:43012|issue:44368", + "issue:43012|issue:43550", + "issue:42886|issue:43381", + "issue:42907|issue:44655", + "issue:44380|issue:45479", + "issue:43208|issue:43550", + "issue:43352|issue:43976", + "issue:43064|issue:44380", + "issue:43618|issue:44380", + "issue:43208|issue:43526", + "issue:44483|issue:44514", + "issue:36683|issue:43582", + "issue:43618|issue:43761", + "issue:43012|issue:44683", + "issue:42757|issue:43824", + "issue:44038|issue:44230", + "issue:44483|issue:45290", + "issue:43299|issue:43867", + "issue:44483|issue:45381", + "issue:38617|issue:43502", + "issue:43756|issue:44485", + "issue:43901|issue:44655", + "issue:36683|issue:43475", + "issue:42994|issue:44655", + "issue:43867|issue:45216", + "issue:44220|issue:44655", + "issue:43352|issue:43937", + "issue:44263|issue:45440", + "issue:36683|issue:44077", + "issue:43066|issue:45381", + "issue:44077|issue:44380", + "issue:43976|issue:44938", + "issue:36683|issue:43525", + "issue:36010|issue:42757", + "issue:43064|issue:44492" + ], + "prepared_review_unit_hash": "59dca2591857cd72ea2ef0f71608f3b3f18c88d248c43ba1e7121764ba4789e7", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38172, + "estimated_input_tokens": 9543, + "estimated_eval_tokens": 19342 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose similarity set of unrelated regression reports. The only plausible duplicate pair is the CLIP v5 output regression (missing attentions vs missing hidden_states), which looks like the same forward-output code path.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43618 is the best canonical issue because it captures the shared CLIP v5 output regression at a broader level; issue:43761 appears to be a specific symptom of the same model-forward bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43618 is the strongest issue to keep as the representative because it is concrete, regression-oriented, and broad enough to subsume the related CLIP hidden-state failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43097", + "right": "issue:43901", + "accept": false, + "reason": "Different problems: API removal/regression vs stale docs text. Same theme of API changes, but not the same bug." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Both mention warnings, but one is a PyTorch compile precision warning and the other is a model config warning about tied embeddings." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Unrelated failure modes: bfloat16 compile warning vs Bamba torch.compile/SDPA runtime bug." + }, + { + "left": "issue:42886", + "right": "issue:43381", + "accept": false, + "reason": "Different subsystems and symptoms: offline cache loading vs gradient checkpointing in eval mode." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Both involve saving, but one is dequantized model save failure and the other is saving Pipeline objects; not the same code path." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Different model families and bugs: attention scaling under SDPA/FlashAttention vs degenerate single-label classification loss." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba compile/SDPA issues are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron Flash Attention support and Python-version compatibility are unrelated." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is unrelated to GPT2 attention scaling under backend changes." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "Both are regressions, but in different model paths: CLIP output fields vs GPT2 attention scaling." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "Different xLSTM training bugs vs BeitImageProcessorFast label reduction bug." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Different API surfaces: chat/completions request validation vs processor chat-template batching crash." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config attribute error and Apple Silicon allocator TypeError are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": true, + "reason": "Both are CLIP v5 regressions in model outputs: one loses attentions, the other loses hidden_states. They plausibly come from the same forward/output assembly bug and could be fixed together." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "Precision warning during compile is unrelated to compiled flex_attention failure on torch >= 2.9." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are ImportErrors, but for completely different symbols and packages." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Related model family names, but different issues: general Qwen3-VL-Moe breakage vs FP8 inference support." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "OpenAI-style chat endpoint request rejection and assistant tool-call chat-template crash are different code paths." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Both concern model loading/saving, but the described failures are not the same concrete bug." + }, + { + "left": "issue:44483", + "right": "issue:45381", + "accept": false, + "reason": "Chat completion API validation and Qwen2.5-VL video position-id corruption are unrelated." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Importing layer_type_validation and ignoring local_files_only are unrelated bugs." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Both mention RoPE, but they refer to different models and different implementation problems." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "Docs wording about return_all_scores is unrelated to pipeline serialization." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Different missing-attribute bugs on different model/config objects." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model saving and pipeline saving are different save paths and symptoms." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "Both are checkpoint/load regressions, but there is not enough evidence they share the same model or code path." + }, + { + "left": "issue:44220", + "right": "issue:44655", + "accept": false, + "reason": "Audio feature extraction bug vs pipeline save bug are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Flash Attention support issue for Nemotron is unrelated to GLM-5 generation config validation." + }, + { + "left": "issue:44263", + "right": "issue:45440", + "accept": false, + "reason": "Different MoE-related failures in different models and code paths." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3 config attribute missing and patchtsmixer post_init policy are unrelated." + }, + { + "left": "issue:43066", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer decoder type mismatch and Qwen2.5-VL video position-id bug are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44380", + "accept": false, + "reason": "Patchtsmixer post_init validation and GPT2 attention scaling are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Python version support problem at 5.1.0 is not the same as a separate Python 3.14 load failure." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Different missing config attributes for different models." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Both are import failures from different modules, with different missing symbols." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "FSDP optimizer-state corruption and a cache-strategy typo are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet and is appropriately cautious. The only accepted soft pair, issue:43618 vs issue:43761, is a reasonable conservative merge candidate because both are CLIP v5 output-regression reports affecting forward output fields. The other pair verdicts also read as appropriately negative and do not overclaim shared bugs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 41, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42890", + "issue:43010", + "issue:43064", + "issue:43352", + "issue:43502", + "issue:43575", + "issue:43618", + "issue:43723", + "issue:43867", + "issue:43931", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44075", + "issue:44077", + "issue:44112", + "issue:44188", + "issue:44242", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44380", + "issue:44484", + "issue:44556", + "issue:44855", + "issue:44857", + "issue:44938", + "issue:44964", + "issue:45071", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45357", + "issue:45375", + "issue:45381", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45588", + "issue:45593", + "issue:45600" + ], + "soft_pairs": [ + "issue:44188|issue:44380", + "issue:43352|issue:45593", + "issue:43618|issue:45216", + "issue:36683|issue:44242", + "issue:43010|issue:45278", + "issue:43352|issue:43931", + "issue:43992|issue:45310", + "issue:41669|issue:44246", + "issue:43867|issue:45375", + "issue:36683|issue:43064", + "issue:44355|issue:44855", + "issue:42890|issue:45561", + "issue:44938|issue:45071", + "issue:43867|issue:45357", + "issue:43992|issue:45478", + "issue:43010|issue:43976", + "issue:44360|issue:45440", + "issue:44938|issue:45381", + "issue:43618|issue:44077", + "issue:43867|issue:44484", + "issue:36296|issue:43575", + "issue:43618|issue:44857", + "issue:43618|issue:45588", + "issue:28282|issue:33453", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:43502|issue:44062", + "issue:43723|issue:44556", + "issue:42890|issue:45341", + "issue:43010|issue:44075" + ], + "prepared_review_unit_hash": "f4e2b9a8861790d29d0332552e4b6cb6f41a6e83118b979017fad71eb566a11d", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36065, + "estimated_input_tokens": 9017, + "estimated_eval_tokens": 18290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly noisy and contains many unrelated issues. The only soft pair that looks like the same underlying problem is the import-time slowdown pair; the rest are different bugs or features in the same broad subsystem but not true duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:44246 is the broader user-facing report of slow `import transformers`; issue:41669 reads as a specific root-cause/optimization ticket (`import *` in models) that plausibly addresses that slowdown.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44246 is the best anchor because it states the symptom clearly, is broader than the implementation-specific follow-up, and has the strongest duplicate relationship among the soft pairs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44188", + "right": "issue:44380", + "accept": false, + "reason": "Both are attention-related, but one is about `torch.compile` branching causing kernel divergence and the other about GPT-2 attention scaling being ignored under SDPA/FlashAttention; different failure modes and fixes." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Flash Attention 2 support gating for Nemotron is unrelated to D-FINE auxiliary-loss behavior when denoising is off." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "`CLIPOutput.attentions` disappearing and Qwen3.5 save_pretrained checkpoint corruption are different code paths and different bug classes." + }, + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "Missing `vocab_size` on `Gemma3Config` and router-logits load-balancing-loss gating are unrelated model/config bugs." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "`@torch.no_grad` on cache/layer update and broad import errors after upgrading versions are not the same underlying issue." + }, + { + "left": "issue:43352", + "right": "issue:43931", + "accept": false, + "reason": "Flash Attention 2 unsupported vs Qwen3-VL weight-shape mismatch are distinct loading/runtime failures." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Missing `embed_tokens.weight` in UMT5Encoder and Qwen3.5 MoE `from_pretrained` errors are unrelated model-loading regressions." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": true, + "reason": "Both point to import-time slowdown in `transformers`; 41669 provides a likely concrete root cause (`import *` in models), while 44246 is the general symptom report." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "A sorted-state_dict load failure and a missing `deepstack_visual_indexes` config field are different problems." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Gemma3 config missing `vocab_size` is unrelated to FSDP2/PEFT optimizer state corruption on nonzero ranks." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Generic compiled-file errors and a Python 3.13 `@torch.jit.script` parsing indentation issue are not the same bug." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Missing `set_seed()` in SamHQ integration tests and an xdist race on `captured_info.txt` are different test flakiness causes." + }, + { + "left": "issue:44938", + "right": "issue:45071", + "accept": false, + "reason": "Python 3.14 import/load failure and `PretrainedConfig` type-checking regression are different compatibility issues." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Sorted state_dict loading errors and Qwen3.5 visual-encoder save_pretrained corruption are different save/load paths." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "UMT5Encoder missing weights and Qwen3.5 MoE `from_pretrained` errors are unrelated model-loading bugs." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "`torch.no_grad` decoration for cache updates has nothing to do with Python version support for Transformers 5.1.0." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "A ReLU missing from the DSA indexer and DeepSeekV3MoE parity issues are different algorithmic concerns." + }, + { + "left": "issue:44938", + "right": "issue:45381", + "accept": false, + "reason": "Python 3.14 load failures and Qwen2.5-VL video position-id bugs are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "Missing CLIP attentions and `patchtsmixer` post_init policy are separate API/architecture issues." + }, + { + "left": "issue:43867", + "right": "issue:44484", + "accept": false, + "reason": "Model loading with sorted state_dict and the 50GB shard-size question are unrelated save/load behaviors." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training bug and TP OOM when loading a large Qwen model are both TP-related but not the same concrete failure." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "Missing CLIP attentions and LwDETR float16 AMP crashes are different bugs in different models." + }, + { + "left": "issue:43618", + "right": "issue:45588", + "accept": false, + "reason": "Missing CLIP attentions and a flash-attention crash on `s_aux=None` are unrelated failures." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "PyTorch-not-found import error and tokenizer-loading regression are unrelated setup/loading issues." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "SamHQ test flakiness and inability to load Phi-4 multimodal are unrelated." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Seed-related integration-test flakiness and GraniteSpeech stale device override CI failure are different test issues." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "SamHQ seeding flakiness and `auto_mappings.py` referencing removed configs are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Unexpected network access with `local_files_only=True` and `AddedToken(..., special=...)` argument collision are unrelated tokenizer/load issues." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading regression in v5 and checkpoint reload failure after upgrading versions are different failures." + }, + { + "left": "issue:42890", + "right": "issue:45341", + "accept": false, + "reason": "Seed-related test flakiness and a bug in `testing_utils.py` are not the same underlying problem." + }, + { + "left": "issue:43010", + "right": "issue:44075", + "accept": false, + "reason": "Cache update no-grad decoration and ignored SGD args are unrelated training/runtime bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the soft-edge decisions are generally conservative. The only accepted pair (41669/44246) is plausibly the same import-slowdown problem, while the rejected pairs are supported by clearly different titles and bug classes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 42, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41628", + "issue:42548", + "issue:42757", + "issue:42831", + "issue:43010", + "issue:43012", + "issue:43065", + "issue:43232", + "issue:43299", + "issue:43502", + "issue:43576", + "issue:43673", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43901", + "issue:44222", + "issue:44230", + "issue:44261", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44368", + "issue:44387", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44655", + "issue:44749", + "issue:44908", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45125", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:45542" + ], + "soft_pairs": [ + "issue:44222|issue:45125", + "issue:43010|issue:43065", + "issue:44387|issue:44556", + "issue:42548|issue:44351", + "issue:45125|issue:45478", + "issue:43010|issue:43232", + "issue:45125|issue:45375", + "issue:44355|issue:44749", + "issue:43502|issue:44485", + "issue:30990|issue:31795", + "issue:30333|issue:33453", + "issue:44315|issue:44360", + "issue:43010|issue:44368", + "issue:43502|issue:44222", + "issue:43502|issue:44961", + "issue:43299|issue:44556", + "issue:44355|issue:44655", + "issue:42831|issue:43874", + "issue:30333|issue:36296", + "issue:44222|issue:44230", + "issue:44355|issue:45071", + "issue:33453|issue:34689", + "issue:42757|issue:43576", + "issue:43827|issue:44961", + "issue:33453|issue:36296", + "issue:43576|issue:45071", + "issue:43576|issue:43824", + "issue:43673|issue:44492", + "issue:38617|issue:43576", + "issue:44961|issue:45468", + "issue:31795|issue:38617", + "issue:43576|issue:44351", + "issue:44908|issue:44961", + "issue:41628|issue:42548", + "issue:44485|issue:44961", + "issue:36296|issue:43901", + "issue:36296|issue:43012", + "issue:41628|issue:44355", + "issue:44261|issue:44360", + "issue:44273|issue:45003", + "issue:44273|issue:45542" + ], + "prepared_review_unit_hash": "d2cb30de7a95e77cdd642607d3dceacfdc42957774af529ccea9462eecf2a20a", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 41, + "serialized_chars": 38808, + "estimated_input_tokens": 9702, + "estimated_eval_tokens": 19660 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most soft pairs are only loosely similar by topic (imports, loading, v5 regressions, FP8/MoE) but describe different concrete bugs or code paths, so they should not be deduplicated.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43576 is the strongest hub among the issue candidates: it is a concrete, user-visible v5 regression and has the most nearby import/CLI-related soft matches, even though the matches are still distinct bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43576 is the best single issue representative here because it has clear repro impact and the broadest duplicate-like neighborhood in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "Both mention FP8/MoE, but one is save_pretrained and the other tensor-parallel model loading." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "No shared code path: cache/no_grad decoration vs a dummy Conv2d in Sam3PixelDecoder." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Different failures: int4 memory growth vs checkpoint reload compatibility after version upgrade." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different export regressions." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Same model family, but tensor-parallel support is not the same as a from_pretrained error." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Different generation internals: no_grad cache updates vs sync_gpus/_update_model_kwargs behavior." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Different model components: tensor parallel plan vs a missing config field silently dropped by strict parsing." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated issues: compiled Python file errors vs a post-upgrade data filtering slowdown." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Local-files-only network leakage is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Loading hang in Sentence Transformers is not the same as documentation confusion for model.forward arguments." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLflow reporting on failed jobs is unrelated to tokenizer loading regression." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel application in model_init and a DSA indexer ReLU bug are different subsystems." + }, + { + "left": "issue:43010", + "right": "issue:44368", + "accept": false, + "reason": "Cache update no_grad behavior is unrelated to a tie_word_embeddings warning." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Different bugs: unwanted API requests vs FP8 save_pretrained behavior." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No meaningful overlap beyond being generic issues." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Qwen3VL MoE loading failure and checkpoint reload breakage are different model/version regressions." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Compiled-file runtime errors and Pipeline save_pretrained failure are not the same bug." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "FP8 accuracy drift and missing image-patch API in GLM46V are separate issues." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow reporting on failure is unrelated to tensor-parallel training bugs." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both are FP8/MoE, but save_pretrained support and inference support are different change requests." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file errors are unrelated to PretrainedConfig type-checking regressions." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer loading regression and Llama 3.2 Vision model loading breakage are different failure modes." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing is_offline_mode import and broken `transformers env` may both be v5-related, but they are distinct bugs." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still mentioning pipeline() is unrelated to the unrelated racoon issue." + }, + { + "left": "issue:33453", + "right": "issue:36296", + "accept": false, + "reason": "Tokenizer loading regression and tensor-parallel training bug are not the same underlying defect." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "A broken CLI env command and config type-checking are different v5 regressions." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "Broken env command and missing Qwen2_5_VL export are different import/export issues." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache handling and a cache-strategy typo are not the same bug." + }, + { + "left": "issue:38617", + "right": "issue:43576", + "accept": false, + "reason": "Missing layer_type_validation export and a broken env CLI command are different import regressions." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "No meaningful overlap; racoon is unrelated to Gemma-4 audio positional encoding." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and an import error for layer_type_validation are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44351", + "accept": false, + "reason": "Broken env CLI and missing HybridCache import are distinct missing-symbol bugs." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs handling has nothing to do with the racoon issue." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Both are import errors, but for different classes and likely different export paths." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "GLM-5 RoPE implementation and racoon are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor-parallel training and TextClassificationPipeline docs are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43012", + "accept": false, + "reason": "Tensor-parallel training bugs do not match bfloat16 compile warnings." + }, + { + "left": "issue:41628", + "right": "issue:44355", + "accept": false, + "reason": "AutoImageProcessor import failure and compiled Python file errors are unrelated." + }, + { + "left": "issue:44261", + "right": "issue:44360", + "accept": false, + "reason": "MLA layer norm epsilon precision issue is not the same as the DSA indexer ReLU bug." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Lazy loading failure and sys.modules access concerns may be adjacent, but not the same concrete bug." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading and a TensorFlow backend import error are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 43, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:31795", + "issue:33453", + "issue:34689", + "issue:37428", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43502", + "issue:43519", + "issue:43576", + "issue:43606", + "issue:43673", + "issue:43704", + "issue:43723", + "issue:43827", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44655", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44945", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45478" + ], + "soft_pairs": [ + "issue:44623|issue:45230", + "issue:43576|issue:44556", + "issue:41669|issue:43673", + "issue:44038|issue:45125", + "issue:44393|issue:45071", + "issue:44556|issue:45092", + "issue:42371|issue:44492", + "issue:42890|issue:44908", + "issue:44704|issue:44908", + "issue:44246|issue:44273", + "issue:43408|issue:44315", + "issue:43976|issue:44393", + "issue:43352|issue:44393", + "issue:43576|issue:44908", + "issue:44908|issue:45071", + "issue:44297|issue:45230", + "issue:44295|issue:45230", + "issue:41669|issue:44351", + "issue:41669|issue:45341", + "issue:44492|issue:44945", + "issue:41669|issue:44492", + "issue:44393|issue:45478", + "issue:44315|issue:44829", + "issue:43827|issue:44908", + "issue:31795|issue:33453", + "issue:44077|issue:44861", + "issue:41669|issue:44908", + "issue:43723|issue:44393", + "issue:43576|issue:44273", + "issue:44655|issue:44861", + "issue:44360|issue:45468", + "issue:44273|issue:44908", + "issue:45230|issue:45310", + "issue:43519|issue:44485", + "issue:41669|issue:43704", + "issue:43519|issue:44908", + "issue:31795|issue:34689", + "issue:31795|issue:37428", + "issue:44315|issue:45092", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:42371|issue:43704", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:43519|issue:43704", + "issue:44263|issue:45468", + "issue:44261|issue:45468", + "issue:43606|issue:45468", + "issue:43704|issue:45341" + ], + "prepared_review_unit_hash": "0d59a3380e2ef8ba456de821000717d91f42a8ad9b61bf84d3c9571a5b697ea4", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 49, + "serialized_chars": 41786, + "estimated_input_tokens": 10447, + "estimated_eval_tokens": 21150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose set of unrelated Transformers issues with a weak subtheme around v5 loading/tokenizer/config regressions. The soft edges are driven by broad lexical similarity, but the underlying bugs and code paths are usually different.", + "confidence": 0.69, + "canonical_issue_reason": "Issue 44556 is the clearest concrete anchor in the loading/regression subtheme: it describes a specific checkpoint reload failure after upgrading to v5.2/v5.3, which is more representative than generic or overly narrow reports.", + "canonical_pr_reason": null, + "best_issue_reason": "44556 is the most actionable and broadly representative issue among the candidates, since it is a concrete v5 compatibility/loading failure rather than a generic placeholder or a narrowly model-specific report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Processor save_pretrained missing-files bug vs a generic placeholder report; not the same issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Both are v5 regressions, but one is the env command and the other is checkpoint reload after upgrade; different failures." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-time file-generation/performance issue vs chunked_prefill cache regression; unrelated code paths." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Same model family, but Qwen3-VL-Moe breakage and missing _tp_plan are different bugs." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "2D bbox output quality bug vs PretrainedConfig type-checking regression; not the same defect." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both concern loading old checkpoints in v5, but the specific incompatibility and failure mode differ." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 settings warning vs cache-strategy typo; unrelated." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Flaky tests due to missing seeds vs scheduler kwargs propagation; different subsystems." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor kwargs forwarding vs inverse_sqrt scheduler kwargs handling; unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Import slowdowns and lazy-loading malfunction are related only loosely; not clearly the same bug." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Model type warning vs Liger Kernel not applied with model_init; different behaviors." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python-version compatibility regression vs Qwen3-VL output bug; unrelated." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention support limitation vs Qwen3-VL bounding-box issue; not the same." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Env command regression vs scheduler kwargs bug; no shared code-path bug." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug vs PretrainedConfig typing regression; different defects." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer class mismatch vs generic placeholder bug report; not enough overlap." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "Position_ids buffer read error vs placeholder report; unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import * performance issue vs HybridCache import failure; different problems." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import-time slowdown vs a bug in testing_utils; unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44945", + "accept": false, + "reason": "Cache-strategy typo vs incorrect LLM output under pipeline parallelism; different issues." + }, + { + "left": "issue:41669", + "right": "issue:44492", + "accept": false, + "reason": "Import * performance regression vs cache-strategy typo; not the same bug." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen3.5/Qwen3-VL failures and different symptoms; not duplicates." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger Kernel application bug vs flash_attention_3 training degeneration; distinct code paths." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Docs still referencing pipeline() vs scheduler kwargs handling; unrelated." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion in model.forward vs tokenizer loading regression; different issues." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init policy vs _get_tied_weight_keys AttributeError; not the same bug." + }, + { + "left": "issue:41669", + "right": "issue:44908", + "accept": false, + "reason": "Import performance issue vs scheduler kwargs bug; unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44393", + "accept": false, + "reason": "Tokenizer loading failure vs Qwen3-VL bounding-box bug; different subsystems." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Env command regression vs lazy-loading issue; not the same underlying failure." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline save_pretrained problem vs tied-weights AttributeError; different code paths." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer math complaint vs Gemma-4 audio positional-encoding bug; unrelated." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading malfunction vs scheduler kwargs bug; different issues." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report vs specific Qwen3.5 Moe from_pretrained error; no concrete duplicate evidence." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Qwen3VL timestamp calculation vs GLM-5 RoPE implementation; different model/math paths." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import * performance regression vs VRAM leak in dataloader threads; unrelated." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp math vs scheduler kwargs bug; not the same issue." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation issue vs model loading regression; different bug classes." + }, + { + "left": "issue:31795", + "right": "issue:37428", + "accept": false, + "reason": "Docs confusion vs flash-attention import error; unrelated." + }, + { + "left": "issue:44315", + "right": "issue:45092", + "accept": false, + "reason": "Liger Kernel/model_init issue vs legacy checkpoint meta-init incompatibility; different failures." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in dataloader threads vs GLM-5 RoPE implementation; unrelated." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "GLM-5 RoPE implementation vs Gemma-4 audio positional encoding; different models and bugs." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 behavior warning vs VRAM leak; unrelated." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "API requests despite local_files_only vs Qwen3VL timestamp bug; different subsystems." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp math vs processor save_pretrained missing files; not the same issue." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Qwen3VL timestamp bug vs multi-thread VRAM leak; unrelated." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return-value issue vs Gemma-4 audio positional encoding bug; different defects." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "MLA q_a_layernorm precision/config issue vs Gemma-4 audio positional encoding; unrelated." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device-mismatch CI bug vs Gemma-4 audio positional encoding; different code paths." + }, + { + "left": "issue:43704", + "right": "issue:45341", + "accept": false, + "reason": "VRAM leak in dataloader threads vs testing_utils bug; unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 14, + "review_unit_index": 44, + "review_unit_count": 44, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-381", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:36296", + "issue:38617", + "issue:42913", + "issue:43066", + "issue:43298", + "issue:43525", + "issue:43572", + "issue:43618", + "issue:44230", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45290", + "issue:45310", + "issue:45341" + ], + "soft_pairs": [ + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44998|issue:45341", + "issue:44861|issue:45290", + "issue:44492|issue:44998", + "issue:43618|issue:44998", + "issue:44556|issue:44998", + "issue:43525|issue:44861", + "issue:43572|issue:44861", + "issue:31515|issue:38617", + "issue:44230|issue:44623", + "issue:31515|issue:36296", + "issue:30333|issue:31515", + "issue:31515|issue:31795", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230" + ], + "prepared_review_unit_hash": "69926dc5a0c0196e2e3ca9c4aa37810c3a1b0dd39e8a144a55cfba67930f1085", + "component_budget": { + "node_count": 381, + "item_count": 381, + "soft_pair_count": 1427, + "serialized_chars": 735811, + "estimated_input_tokens": 183953, + "estimated_eval_tokens": 368162 + }, + "budget": { + "node_count": 22, + "item_count": 22, + "soft_pair_count": 17, + "serialized_chars": 17392, + "estimated_input_tokens": 4348, + "estimated_eval_tokens": 8952 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated single-issue reports rather than duplicates. The soft pairs are mostly broad thematic matches (e.g. tokenizer/v5 regressions, loading issues) but not the same concrete bug or code path, so none should be merged.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 44556 is the clearest, most actionable regression report in the set: it describes a specific checkpoint reload failure after a v4.57\u2192v5.x upgrade. It is more concrete and triage-friendly than the other broad or ambiguous reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44556 is the strongest representative issue because it is specific, reproducible, and describes a likely regression with a clear upgrade path and impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both are tokenizer-related v5 complaints, but one is a broad behavior difference and the other is a specific wrong decoder-type report; they do not describe the same concrete defect." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Inverse-sqrt scheduler kwarg handling and an unrelated unemployment/off-topic issue are clearly different." + }, + { + "left": "issue:44998", + "right": "issue:45341", + "accept": false, + "reason": "Completely unrelated: one is off-topic, the other is a testing_utils bug." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Different code paths: tied-weights key extraction crash versus chat template handling for tool-call assistant messages." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "A typo in cache strategies is unrelated to the off-topic unemployment issue." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attentions regression and an off-topic issue are unrelated." + }, + { + "left": "issue:44556", + "right": "issue:44998", + "accept": false, + "reason": "Checkpoint reload regression after version upgrade is unrelated to the off-topic unemployment issue." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is a different bug from the tied-weights AttributeError." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is a separate config regression, not the tied-weights crash." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow checkpoint loading and an ImportError for layer_type_validation are unrelated problems." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 inference support for Qwen3-VL/Qwen3.5 MoE is not the same as missing files from processor.save_pretrained." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint loading performance and a tensor-parallel training bug are different issues." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLflow reporting on failed training jobs is unrelated to checkpoint loading speed." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Checkpoint loading performance and confusing model.forward docs are not the same underlying bug." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen3-family models, but one is an FP8 support request and the other is a concrete from_pretrained error; too different to merge as duplicates." + }, + { + "left": "issue:31515", + "right": "issue:33453", + "accept": false, + "reason": "A slow from_pretrained checkpoint load and a tokenizer-loading regression are not the same defect." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching lacking audio-model support is a different capability gap from FP8 Qwen3-VL/Qwen3.5 MoE support." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "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": "7296b965239bcf252d33470058bf50138ae7b73dd6f6b05d4c9e923658578584", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are about different community-contribution feature requests: Flash Attention 2 support for more architectures vs. GGUF support for more architectures. They are not duplicates and should not be merged.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different underlying feature requests with different targets (Flash Attention 2 vs GGUF). They do not describe the same bug or change, so this is not a duplicate pair." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-4", + "nodes": [ + "issue:27088", + "issue:43834", + "issue:44365", + "issue:44370" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:44365|issue:44370" + ], + "prepared_review_unit_hash": "03f7a243c5ca3148b4ff924254d232dc6f8cfaf1b0e702cb0914d602ef327482", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4452, + "estimated_input_tokens": 1113, + "estimated_eval_tokens": 2482 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4452, + "estimated_input_tokens": 1113, + "estimated_eval_tokens": 2482 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is dominated by generic i18n translation-template issues, with one concrete Turkish docs translation issue. The concrete Turkish issue is the best canonical issue; the placeholder/template issues are too underspecified to treat as exact duplicates of each other or of the Turkish one.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 27088 is the only concrete, language-specific report ('Turkish') rather than a placeholder template, so it best represents the real actionable issue in the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 27088 is the strongest issue to keep as the representative: it names a specific language, is not a generic template, and is the most actionable artifact here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Different concrete language targets: one is Turkish, the other is a generic placeholder. Similar topic, but not the same underlying issue." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "The Turkish-specific report is not the same as a generic placeholder translation issue; no evidence they refer to the same language or same task." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "One issue is explicitly Turkish while the other is only a template stub. Too little evidence that they describe the same change or bug." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "Both are template-like translation issues, but there is no concrete language or scope match to prove they are the same underlying request." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "Near-identical template text alone is not enough; these could be separate language translation requests, so do not merge." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "Same generic template wording, but no concrete evidence that they target the same language or identical work item." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-14", + "nodes": [ + "issue:36817", + "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" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:37015|issue:42503", + "issue:37015|issue:45027", + "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:43671|issue:43909", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43837|issue:44121", + "issue:43909|issue:44205", + "issue:44121|issue:45322", + "issue:43837|issue:45027", + "issue:43837|issue:45177" + ], + "prepared_review_unit_hash": "a63d57906e1cae48af1797b2aed0699a2adf9e7e258d981d62645c813b5d635b", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 28, + "serialized_chars": 17648, + "estimated_input_tokens": 4412, + "estimated_eval_tokens": 9080 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 28, + "serialized_chars": 17648, + "estimated_input_tokens": 4412, + "estimated_eval_tokens": 9080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are all separate model-support requests across different model families and modalities; despite broad topical similarity, none look like true duplicates or the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "EuroBert config support and DEIMv2 model support are unrelated requests for different architectures." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config support is unrelated to Qwen3-TTS audio support." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT support and Qwen3-ASR support target different models and different inference paths." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT model addition is unrelated to the OpenAI weight-sparse transformer request." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT support and LFM2.5 Audio support are different model families with different requirements." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are separate model requests, not the same concrete bug or change." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS support are unrelated model additions." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are distinct feature requests." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 support is unrelated to the weight-sparse transformer request." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 and ModernVBERT are different model families and not duplicates." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 support is unrelated to Voxtral-4B-TTS support." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 model support and Qwen3-TTS support are separate requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT support and Qwen3-ASR support are different models and different tasks." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT and Qwen3-TTS are unrelated model-support requests." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT models and general BERT-like RoPE support overlap conceptually, but they are not the same concrete request or code change." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "General BERT-like RoPE support is unrelated to Voxtral TTS support." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Both are Qwen3 audio requests, but TTS and ASR are different model targets and would not be merged into one fix." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral-4B-TTS are different TTS model additions." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS support is unrelated to DeepSeek-OCR2 support." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support and weight-sparse transformer support are unrelated." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-TTS and LFM2.5 Audio are separate audio-model support requests." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are different models and different integration paths." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR support is unrelated to EUPE model support." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support and the weight-sparse transformer request are unrelated." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio support and SAM3-LiteText are different model requests." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "The weight-sparse transformer request and EUPE are different proposed models, not the same change." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different speech-model support requests." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR support and DeepSeek-OCR2 support 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": 6, + "component_count": 14, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-37934-6", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031" + ], + "soft_pairs": [ + "issue:37934|issue:43668", + "issue:42936|issue:43240", + "issue:42936|issue:44031", + "issue:42936|issue:43813", + "issue:42936|issue:43668", + "issue:43240|issue:43668", + "issue:43668|issue:44031", + "issue:43668|issue:43813" + ], + "prepared_review_unit_hash": "be9be9e3985fe90301d4d44c167d875dd8408d5cc3c6b96540eda1fdfb961d15", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 8, + "serialized_chars": 6290, + "estimated_input_tokens": 1573, + "estimated_eval_tokens": 3402 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 8, + "serialized_chars": 6290, + "estimated_input_tokens": 1573, + "estimated_eval_tokens": 3402 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues are not duplicates of one another; they span unrelated model, config, API, typo, and tokenizer-regression topics. The cluster looks like a loose soft-similarity grouping rather than a true duplicate set.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43240 is the most broadly phrased and externally referenced item in the set, so it is the best representative of the cluster even though the cluster itself does not contain a true duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43240 has the strongest centrality signals (inbound references/discussion) and the most general bug framing, making it the best issue candidate for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Different model components and problem types: Llama4TextL2Norm vs ModernBERTConfig type hint. One is about normalization semantics, the other about an annotation typo." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Mask2Former ignore_value handling and loss kwargs forwarding are separate code paths and bugs; they are not the same underlying defect." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Unrelated subsystems: Mask2Former loss/ignore_value versus tokenizer regex warning regression." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Model loss-handling bug vs a literal typo in a quantization-related name; no shared code-path." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former runtime behavior bug and ModernBERT config typing issue are unrelated." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Loss kwargs not being forwarded is an execution bug; incorrect norm_eps type hint is a configuration annotation issue." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "ModernBERTConfig type hint mismatch and tokenizer regex warning regression are different components and symptoms." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Type-hint typo in ModernBERTConfig and a quantization-operation naming typo are distinct, unrelated fixes." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "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": "f4d78dc97b5407eab1e1590a0f09e0d1085dc140440f218726e372fe02bd8564", + "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 do not look like duplicates: one is about block decoding, the other is about adding DeepSeek v3.2 support. They target different features and likely different code paths.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single canonical issue stands out because the reports are unrelated; if forced, neither is a good representative of the other.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different requests with no clear shared underlying bug or change: block decoding vs model/version support." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "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": "29be77c41d2f9e905ca615515af09c801401d926b6cc4b2a58c5678e89ac985f", + "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 by theme (custom trainers in `Trainer`), but they describe different problems: one is about `remove_unused_column` customizability, the other about FLOPS estimation for custom trainers. I would not merge them as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "No clear canonical duplicate; if one must be chosen as the cluster representative, #43598 is a slightly broader Trainer customizability issue.", + "canonical_pr_reason": null, + "best_issue_reason": "#43598 is the better representative because it frames a general Trainer behavior/customizability concern rather than a narrower FLOPS-estimation bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Same general area, but different underlying bugs: one concerns `remove_unused_column` behavior, the other FLOPS counting for custom trainers. They would not be fixed by one change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 9, + "component_count": 14, + "review_unit_index": 1, + "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": "5ee5a8ba56314a13f59cd9f056d9c9cfe948b5ae25ea05e4767a9da7a4e5b7c2", + "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": "The two issues are about different Qwen model families and different auto-model capabilities: one asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, while the other asks for AutoModelForSequenceClassification support for Qwen3.5. They are related only at a broad subsystem level, not the same underlying bug/change.", + "confidence": 0.95, + "canonical_issue_reason": "If one issue must represent the cluster, 43684 is the broader model-registration request and is more general than the Qwen3.5 sequence-classification support request.", + "canonical_pr_reason": null, + "best_issue_reason": "43684 is the better representative because it targets core auto-model registration, which is broader and more central than a single-task head enablement.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different model families (Qwen3-Omni vs Qwen3.5) and different requested capabilities (registration/conditional generation vs sequence classification); not the same bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 14, + "review_unit_index": 1, + "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": "1730361b56ed37366fec8c84f511d6f3bab2e07c14a2eac01b0d4d377c8c82ab", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The three \"huggingface\" issues are plausibly related and 43845 is the best representative, but the packet only gives soft_similarity for every pair and no deterministic_accept, so the duplicate links should stay tentative. \"Fett\" appears separate.", + "confidence": 0.63, + "canonical_issue_reason": "issue:43845 is the best representative of the huggingface subgroup because it is the earliest of the near-identical placeholder-title reports and aligns most closely with issue:43846, but the evidence is still only soft similarity.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43845 is the strongest cluster representative: it is earliest among the huggingface reports, while issue:43859 is shorter and the only one with discussion. Even so, the available evidence does not justify treating the links as confirmed duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different titles and no shared concrete details, filenames, or targets; this does not look like the same underlying issue." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": false, + "reason": "Same placeholder title, but the packet provides only soft_similarity and no deterministic_accept or concrete shared targets/files, so this is not strong enough to accept as the same issue." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": false, + "reason": "The match is only superficial: same placeholder title, but no concrete issue target, filename, or diff evidence, and deterministic_accept is false." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": false, + "reason": "The match is only superficial: same placeholder title, but no concrete issue target, filename, or diff evidence, and deterministic_accept is false." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster summary is mostly grounded, but the soft-edge verdicts are too confident. In the packet, all pairwise links are only soft_similarity, and deterministic_accept is false for every pair. Accepting issue:43845 with issue:43859 and issue:43846 with issue:43859 overstates the evidence; those should be treated more conservatively. The representative choice for the huggingface duplicates is otherwise reasonable, and Fett does look separate." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 11, + "component_count": 14, + "review_unit_index": 1, + "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": "55c9e3606b1d35cc759370c42b473c916265c7d1cc6f092823bcac48bcf1a107", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is dominated by low-content spam/generic issue stubs, with no actionable bug or change details to justify deduping on substance.", + "confidence": 0.28, + "canonical_issue_reason": "Issue 44418 is the least spam-like representative: it is the only non- title and is still just a generic placeholder, so it\u2019s the best available cluster anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue here contains enough concrete problem description to be a true canonical duplicate target; 44418 is only the least bad representative, not a substantive one.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Same '' title, but there is no actual bug/change context to show they describe the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Both are spam placeholders with no concrete problem statement; identical titles alone are not enough to merge them as duplicates." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "Titles are unrelated generic placeholders, and neither issue has enough substance to establish a shared underlying bug." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "'' and 'Check out \"Google LLC\"' do not indicate the same bug or change; this looks like unrelated spam." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "Generic documentation-sounding text versus a spam placeholder; no concrete common issue is described." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "No shared actionable content; these are just low-information placeholder issues." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Identical spam titles, but there is no evidence of a real shared bug/change to deduplicate." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "A spam placeholder and a promotional-sounding title are not the same underlying issue." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "No substantive overlap beyond both looking like spam; not a valid duplicate match." + } + ] + }, + "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": 14, + "review_unit_index": 1, + "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": "d2bb998fa78245e61231f6e0459f0c4a1fb58531e8791f9f966fbecb99a5ab5e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1646, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1646, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The pair looks strongly related, but the available evidence is not enough to call them definite duplicates with high confidence. They share the exact title and similar body size, yet we do not see body-content overlap, explicit links, or scope details to prove they are the same underlying issue.", + "confidence": 0.76, + "canonical_issue_reason": "Issue 45145 is the earlier report and has an inbound reference, so it is the best tentative anchor if these are later found to be duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45145 is the best representative of the pair because it is earlier and already has more visible downstream attention, though the duplicate call is not definitive.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": false, + "reason": "The evidence is too thin to confirm the same underlying issue: identical titles and similar body lengths alone do not establish duplicate scope, and there is no explicit cross-linking or content-level proof." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The evidence supports strong similarity, but not a high-confidence duplicate conclusion. The packet only shows matching titles, close body lengths, and a soft-similarity signal; it does not establish identical scope or that 45145 is definitively the canonical anchor. The soft-edge verdict is a bit too assertive given the available evidence, and the confidence of 0.98 is overstated." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 14, + "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": 14, + "component_count": 14, + "review_unit_index": 1, + "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": "32fdae562cd9cbc5b133f37e1d8671c7f7ef673a5d70a63e234a12a6cdc8d70f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1668, + "estimated_input_tokens": 417, + "estimated_eval_tokens": 1090 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related to 4-bit inference/bitsandbytes behavior, but they describe different failure modes in different code paths, so they should not be clustered as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Different underlying bugs: one is a Gemma4 audio-module uint8/torch.finfo TypeError during NF4 inference, the other is a BitsAndBytesConfig skip-modules regression that removes the default lm_head exclusion and triggers an AssertionError. Same broad area, but not 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": 1, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-39895-12", + "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" + ], + "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:45655", + "pull_request:43823|pull_request:44827", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "0e21cce5ec3a1d2e2c4d1312b93711d4fa9328fab7211a2910a281bacdf88f06", + "component_budget": { + "node_count": 12, + "item_count": 12, + "soft_pair_count": 18, + "serialized_chars": 32874, + "estimated_input_tokens": 8219, + "estimated_eval_tokens": 16694 + }, + "budget": { + "node_count": 12, + "item_count": 12, + "soft_pair_count": 18, + "serialized_chars": 32874, + "estimated_input_tokens": 8219, + "estimated_eval_tokens": 16694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated PRs; the only true duplicate is the two fixed_cross_entropy kwargs patches (43251 and 43254).", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43251 is the best canonical representative for the only real duplicate pair: it has the explicit issue link, targets the same loss_utils bug, and cleanly extends fixed_cross_entropy with weight/label_smoothing handling.", + "best_issue_reason": null, + "best_pr_reason": "43251 is the strongest standalone PR in the set because it fixes a concrete bug in a single code path and is clearly tied to the tracked issue; 43254 is a near-duplicate alternative.", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "VideoPrism model addition vs an executorch dynamic-shapes test; unrelated changes." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Mistral tokenizer helper methods vs adding MobileLLM; different subsystems and goals." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Tokenizer method additions are unrelated to executorch export testing." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor robustness changes vs adding a new model; not the same underlying change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Loss-function kwargs fix vs MobileLLM model implementation; unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Both touch code quality in different contexts, but one fixes cross-entropy kwargs while the other addresses Mistral4 test/runtime issues; not the same bug." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Executorch dynamic-shape export test vs Mistral4 test fixes; different concerns." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Executorch exportability test vs MobileLLM model addition; unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New MobileLLM model support vs config_class resolution ordering fix; different code paths." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "Model addition vs Falcon refactor; not the same underlying change." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon output-interface refactor vs a model-creation skill/documentation PR; unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon refactor vs Mistral4 tests; different models and different problems." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test fixes vs a CLIP-like layrnorm typo correction; unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 tests vs config resolution fix; no shared bug or change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:44827", + "accept": false, + "reason": "MobileLLM model addition vs Mistral4 tests; unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Mistral4 tests vs model-creation skill/docs; unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model implementation vs a model-creation skill draft; different artifact types and purposes." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both patches modify the same fixed_cross_entropy function in the same file to pass supported cross-entropy kwargs (weight/label_smoothing); they are the same underlying fix and could be merged into one PR." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted pair is well supported by the diffs: 43251 and 43254 are near-identical fixes to the same function in the same file. The other soft-edge rejections are conservative and do not overstate evidence. Minor note: phrases like \"only true duplicate\" are a bit strong, but not enough to invalidate the review." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-2", + "nodes": [ + "pull_request:40546", + "pull_request:42781" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781" + ], + "prepared_review_unit_hash": "caa0e7bef8faea9dac1b85be81a6d1783b3bc60403bd297258c98d620359889c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related VibeVoice implementation PRs, but they are not duplicates: the first adds the base VibeVoice model, while the second adds VibeVoice Realtime and an acoustic tokenizer with extra generation/model changes. They overlap in shared package/auto-registration files, but they address different deliverables and code paths.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 40546 is the cleaner canonical PR for the core VibeVoice model implementation; it is the earlier, broader base integration and has substantial review activity.", + "best_issue_reason": null, + "best_pr_reason": "PR 40546 is the best representative of the underlying VibeVoice model work because it introduces the foundational model/docs/auto mappings, whereas PR 42781 is a later extension focused on realtime/tokenizer support.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "Reject: despite heavy filename overlap, 42781 is a separate realtime/tokenizer extension with additional model and generation changes, not the same concrete bug fix or mergeable duplicate of 40546." + } + ] + }, + "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": 24, + "review_unit_index": 1, + "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": "38f21745407f14b5136192ef8c8a8cd5bc260286e9bdc61f93701f765d589fd9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are unrelated: one adds TimesFM covariate forecasting support, the other adds DINOv3 image classification support. They do not fix the same code-path problem and should not be deduplicated.", + "confidence": 0.98, + "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 models and changes: TimesFM covariate forecasting vs DINOv3 image classification. No shared bug or mergeable single 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": 4, + "component_count": 24, + "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": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41797-2", + "nodes": [ + "pull_request:41797", + "pull_request:45186" + ], + "soft_pairs": [ + "pull_request:41797|pull_request:45186" + ], + "prepared_review_unit_hash": "6bc8242cf798dd2a779ef0d830fa98ea2a72e60f3d2b6ab3edfd21b9334fe638", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6723, + "estimated_input_tokens": 1681, + "estimated_eval_tokens": 3618 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6723, + "estimated_input_tokens": 1681, + "estimated_eval_tokens": 3618 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are model-addition PRs that touch shared registration/docs infrastructure, but they introduce different models (DeepSeekOCR vs Isaac) with different implementations and tests. They are not the same underlying change, so the soft duplicate edge should be rejected.", + "confidence": 0.98, + "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 distinct model directories, configs, and test suites; shared auto/docs plumbing is not enough to treat them as the same PR or merge into one change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 24, + "review_unit_index": 1, + "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": "1b9bf525443a07db81b3c6525437715628283fbb00744db00aee32e1d0cf61e5", + "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": "Two separate maintenance PRs touching overlapping GitHub Actions workflows, but they update different actions and target different upgrade goals. They should not be treated as duplicates.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the better canonical representative: it is broader, explicitly about Node 24 compatibility, and updates several core workflow actions across the repo.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the most representative item in the cluster because it covers the larger workflow-action upgrade effort; PR 42976 is a narrower docker-action version bump.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Overlap in workflow filenames is not enough: 42975 upgrades checkout/cache/upload-artifact for Node 24 compatibility, while 42976 only bumps docker/build-push-action. Different actions, different concrete change, so not the same underlying PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 7, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42978-7", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:43451", + "pull_request:44601", + "pull_request:45490", + "pull_request:45626", + "pull_request:45630" + ], + "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", + "pull_request:44601|pull_request:45626", + "pull_request:45490|pull_request:45626", + "pull_request:45490|pull_request:45630", + "pull_request:45626|pull_request:45630" + ], + "prepared_review_unit_hash": "014709e08748e947ca9b15ee3418d60cec68e8754e980c3c69045833c036ad3a", + "component_budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 10, + "serialized_chars": 23910, + "estimated_input_tokens": 5978, + "estimated_eval_tokens": 12212 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 10, + "serialized_chars": 23910, + "estimated_input_tokens": 5978, + "estimated_eval_tokens": 12212 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a loose collection of unrelated pull requests: several separate model additions (ViT NEPA, Molmo, Molmo2, CTSM, PP-FormulaNet, Kimi2-6) plus one distributed loading feature PR. They share repo areas like docs and auto mappings, but not the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model additions: ViT NEPA vs Molmo. Shared docs/auto-registration changes are generic scaffolding, not the same bug or feature." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "Different model additions: ViT NEPA vs Molmo2. They are separate model integrations, not one mergeable change." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "Different model additions: ViT NEPA vs CTSM. No shared concrete code-path problem." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are related by name, but they are distinct model implementations with different files, APIs, and tests; not the same underlying PR." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo vs CTSM are unrelated model integrations with different architectures and code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 vs CTSM are separate model additions; shared registry/docs edits are not enough to merge them." + }, + { + "left": "pull_request:44601", + "right": "pull_request:45626", + "accept": false, + "reason": "One is native pipeline-parallel loading support, the other is PP-FormulaNet model support. Different feature areas and code paths." + }, + { + "left": "pull_request:45490", + "right": "pull_request:45626", + "accept": false, + "reason": "CTSM and PP-FormulaNet are different model additions; same broad subsystem, but not the same change." + }, + { + "left": "pull_request:45490", + "right": "pull_request:45630", + "accept": false, + "reason": "CTSM and Kimi2-6 are separate model integrations with distinct model files and tests." + }, + { + "left": "pull_request:45626", + "right": "pull_request:45630", + "accept": false, + "reason": "PP-FormulaNet and Kimi2-6 are different model additions; similar documentation/auto-registration edits 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": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-3", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43665|pull_request:44815" + ], + "prepared_review_unit_hash": "de3d173608954f2699c6d6a6040b302db186163d732f2f12b4c7f2a7d24aebfb", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7092, + "estimated_input_tokens": 1773, + "estimated_eval_tokens": 3802 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7092, + "estimated_input_tokens": 1773, + "estimated_eval_tokens": 3802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three PRs are not duplicates of the same change. PR 44815 is a substantive FP8/dequantization fix, PR 43291 is a Whisper tokenizer/test fix, and PR 43665 is a test-only placeholder change across CLIP/ViT. They share only broad similarity as small repo maintenance PRs.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44815 is the clearest canonical PR because it contains a concrete, coherent code-path fix in model loading and FP8 integration, with meaningful implementation changes and related tests.", + "best_issue_reason": null, + "best_pr_reason": "PR 44815 is the best representative of the cluster because it is the most substantive and self-contained bug fix; the others are either unrelated test changes or a separate Whisper-specific tokenizer/test issue.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Different targets and different problems: Whisper tokenizer/test behavior versus CLIP/ViT test assertions. No shared code path or mergeable unified fix." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decoding/tests are unrelated to FP8 dequant/model-loading logic. They do not appear to fix the same underlying bug or change." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "One is test-only changes for CLIP/ViT; the other is a concrete FP8 dequantization fix. Same broad repository area, but not the same bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 9, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43488-3", + "nodes": [ + "pull_request:43488", + "pull_request:43532", + "pull_request:45476" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:43532", + "pull_request:43488|pull_request:45476", + "pull_request:43532|pull_request:45476" + ], + "prepared_review_unit_hash": "4f2afa3ba9085205d6993f828b6b32c56156c501a0409469f2c60b65a339e59a", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7469, + "estimated_input_tokens": 1868, + "estimated_eval_tokens": 3992 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7469, + "estimated_input_tokens": 1868, + "estimated_eval_tokens": 3992 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three unrelated synthetic '[don't merge]' PRs; the apparent overlap is incidental, mostly a shared vit test filename, not the same underlying bug or change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true canonical duplicate exists here. PR #43532 is the most central by shared filenames and discussion activity, but it still changes unrelated workflow/test scaffolding rather than the same concrete fix as the others.", + "best_issue_reason": null, + "best_pr_reason": "Not a good duplicate representative overall; if one must be picked, #43532 is the least-bad cluster center, but it is not actually a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Different changes: setup.py/dependency/version tweaks versus workflow/testing utility edits. Shared vit test file is incidental, not the same bug or fix." + }, + { + "left": "pull_request:43488", + "right": "pull_request:45476", + "accept": false, + "reason": "Unrelated code paths: packaging/version/deps changes versus CI workflow plus CLIP/tokenization and Chinese-CLIP test edits. Not mergeable as one fix." + }, + { + "left": "pull_request:43532", + "right": "pull_request:45476", + "accept": false, + "reason": "Both touch tests/workflows, but they address different areas and no single concrete problem is shared. The overlap is too broad and incidental to treat 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": 10, + "component_count": 24, + "review_unit_index": 1, + "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": "4dfe0f50a9782a3941dd045116d4b6435a5268d773c763f37e811d89deb67dcb", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two different PRs for two different segmentation pipelines: one for Promptable Concept Segmentation and one for Promptable Visual Segmentation. They share scaffolding and docs/pipeline infrastructure, but they target different models and code paths, so they should not be deduplicated.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different underlying features: PR 43612 adds Promptable Concept Segmentation around SAM3, while PR 43613 adds Promptable Visual Segmentation across EdgeTAM/SAM/SAM2/SAM3-tracker. Similar framework changes do not make them the same 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": 11, + "component_count": 24, + "review_unit_index": 1, + "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": "8d57bb2670f7c55b9ffc5336198ed8d163eb53b1071f9559bf4ca70fc329eb7a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs in `trainer.py` touch custom-trainer extensibility, but they change different code paths: one adds a loss-scaling hook for gradient accumulation, the other extracts signature-column filtering into an overrideable helper. They are related, not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43651 is the stronger canonical representative because it introduces a concrete behavioral hook in the training loss path and changes the actual scaling logic, making it the more substantive and user-facing change.", + "best_issue_reason": null, + "best_pr_reason": "PR #43651 is the best single representative of the cluster because it targets a specific training bug/behavior with a clear functional outcome, whereas PR #43663 is mainly an internal refactor to expose override points for column filtering.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Both PRs edit `trainer.py` and improve custom trainer extensibility, but they solve different problems: #43651 changes gradient-accumulation loss scaling, while #43663 changes signature-column filtering. They do not appear to be the same underlying bug or mergeable 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": 12, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43656-3", + "nodes": [ + "pull_request:43656", + "pull_request:43836", + "pull_request:43842" + ], + "soft_pairs": [ + "pull_request:43656|pull_request:43842", + "pull_request:43836|pull_request:43842" + ], + "prepared_review_unit_hash": "7fcee83d8ac3d649e08cd8f329e5a97b91decb58f81b8d25e699463944f13649", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three open PRs all touch `src/transformers/cli/serve.py` and two of them target the same tracked issue. The strongest duplicate pair is the two issue-43824 fixes; the 43656 PR looks related by topic but is not clearly the same concrete change.", + "confidence": 0.77, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43842 is the best canonical representative: it directly addresses the TypeAdapter import/runtime NameError with a narrow, explicit fix and a clear issue link. It is more focused than #43836 and easier to treat as the primary duplicate target.", + "best_issue_reason": null, + "best_pr_reason": "PR #43842 best matches the underlying bug report and contains the most precise code-path fix (guarding the `TypeAdapter` import with `TYPE_CHECKING` and stringizing the annotation).", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Both mention TypeAdapter NameError in the CLI, but they target different issues and the diffs suggest different fixes; not enough evidence that they are the same concrete change." + }, + { + "left": "pull_request:43836", + "right": "pull_request:43842", + "accept": true, + "reason": "Same file, same issue target, and both are alternative fixes for the same import-time `TypeAdapter` NameError in `serve.py`; they are plausibly mergeable into one PR." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: all three PRs touch `src/transformers/cli/serve.py`, 43836 and 43842 both target issue 43824, and 43656 targets a different issue (43576). The conservative non-match verdict for 43656 vs 43842 is justified. The 43836 vs 43842 pairing is also reasonable as a mergeable same-bugfix candidate, since both address the same TypeAdapter NameError in the same file and the diffs look compatible rather than contradictory." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 24, + "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": 24, + "review_unit_index": 1, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43996", + "pull_request:43997", + "pull_request:44007", + "pull_request:44010", + "pull_request:44013", + "pull_request:44017", + "pull_request:44018", + "pull_request:44019", + "pull_request:44024", + "pull_request:44044", + "pull_request:44056", + "pull_request:44059", + "pull_request:44066", + "pull_request:44068", + "pull_request:44071", + "pull_request:44072", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44085", + "pull_request:44086", + "pull_request:44098", + "pull_request:44101", + "pull_request:44116", + "pull_request:44129", + "pull_request:44154", + "pull_request:44161", + "pull_request:44722" + ], + "soft_pairs": [ + "pull_request:43996|pull_request:44085", + "pull_request:43996|pull_request:44044", + "pull_request:44066|pull_request:44085", + "pull_request:44007|pull_request:44072", + "pull_request:44072|pull_request:44722", + "pull_request:44066|pull_request:44072", + "pull_request:44013|pull_request:44044", + "pull_request:44066|pull_request:44086", + "pull_request:44018|pull_request:44068", + "pull_request:44066|pull_request:44071", + "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:44086", + "pull_request:44018|pull_request:44085", + "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:44722", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:44071|pull_request:44086", + "pull_request:43997|pull_request:44073", + "pull_request:44017|pull_request:44071", + "pull_request:44059|pull_request:44071", + "pull_request:44073|pull_request:44101", + "pull_request:44059|pull_request:44068", + "pull_request:44018|pull_request:44059", + "pull_request:44056|pull_request:44101", + "pull_request:44073|pull_request:44161", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44017", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:43997|pull_request:44018", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request: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:44010|pull_request:44018", + "pull_request:44071|pull_request:44129", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154", + "pull_request:44024|pull_request:44076" + ], + "prepared_review_unit_hash": "f9393d5a028c4cadc735050563d400d5f0d3c36796367e5e53d87dad9e2a2da2", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 + }, + "budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 72, + "serialized_chars": 76933, + "estimated_input_tokens": 19234, + "estimated_eval_tokens": 38724 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of model-specific standardized output-tracing refactors, not one single duplicate set. The only clear duplicate subgroup is the GPT-J/CodeGen work, with an earlier GPT-Neo pair also looking like a rebased duplicate; the rest are different models and should stay separate.", + "confidence": 0.7, + "canonical_issue_reason": null, + "canonical_pr_reason": "44722 is the best representative for the duplicated GPT-J/CodeGen subgroup: it is the latest and most complete refactor on the same concrete output-tracing code paths, with the same standardized-decorator direction as the earlier GPT-J variants.", + "best_issue_reason": null, + "best_pr_reason": "44722 is the strongest overall PR in the cluster because it cleanly targets the same output-tracing refactor, covers the shared GPT-J/CodeGen implementation, and appears to subsume earlier GPT-J-only variants.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model implementations (CVT/FNet vs GPT-J); same broad theme, but not the same code path or change." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "Both are output-tracing refactors, but for unrelated models (CVT/FNet vs DeBERTa-v2)." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Both change GPT-J output tracing in the same file; 44085 reads like a narrower/rebased variant of the same fix." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "ResNet/RT-DETR ResNet vs EfficientNet: same refactor pattern, but not the same model or bug." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet is unrelated to GPT-J/CodeGen; same umbrella refactor idea only." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen vs EfficientNet are different code paths and different model families." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa-v2 are unrelated implementations despite similar output-tracing cleanup." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen vs MGP-STR are unrelated model changes." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo output-tracing refactor on the same model file; 44068 looks like a later/reworked version of 44018." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen vs MPT are different model code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen and GPT-Neo are different model implementations; not the same fix." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo vs GPT-J/CodeGen: same standardized-output theme, but different code paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet is unrelated to GPT-J/CodeGen; different model family and code path." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are unrelated implementations." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet vs GPT-J: different model and different underlying change." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are unrelated model-specific refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet vs MPT are different code paths; not a duplicate." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo vs Flaubert are different models and not the same bug/change." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are separate model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated model changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are different model families and code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo vs MGP-STR: same pattern, different implementation." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo and GPT-J are different models; not mergeable as one fix." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo vs MGP-STR are unrelated model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT vs GPT-J are different model code paths." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT2 and MGP-STR are unrelated changes." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are different model implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT2 and GPT-J/CodeGen are different model families; same refactor style only." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert/XLM are different model architectures and code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "Segformer and GPT-J/CodeGen are different models; not the same change." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "Segformer and LongT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and Segformer are different code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated model implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are different models; same refactor pattern only." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert/XLM are unrelated model changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are different model families." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are unrelated code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are different model implementations." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet and LongT5 are unrelated changes." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44071", + "accept": false, + "reason": "RegNet and MPT are different model families." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are unrelated model implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBert are different code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet and LongT5 are different model families." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "CVT/FNet and Segformer are separate model changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBert and DeBERTa-v2 are unrelated implementations." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa-v2 and TextNet are different code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen standardized output-tracing refactor on the same files; 44722 is the later, fuller version." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both cover the same GPT-J output-tracing change on the same code path; 44722 is broader because it also includes CodeGen." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are different model families." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "CVT/FNet and ResNet/RT-DETR ResNet are unrelated model changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet and SpeechT5 are different model implementations." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa-v2 and SpeechT5 are unrelated code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa-v2 and ViTS are different model families." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are different models; same output-handling theme only." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is mostly grounded in the packet: it correctly separates the shared-model duplicates from the broader model-specific refactors, and the accepted soft pairs are limited to cases that look mergeable within the same code path (GPT-J/CodeGen and GPT-Neo). The wording stays mostly cautious, with only mild inference about rebasing/later variants." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 24, + "review_unit_index": 2, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43997", + "pull_request:44010", + "pull_request:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44129" + ], + "soft_pairs": [ + "pull_request:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "1804fb911aff06d9dd743cca001df0c77a2f50fae53a4a53f2cdce2f99ce49fc", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 5, + "serialized_chars": 13068, + "estimated_input_tokens": 3267, + "estimated_eval_tokens": 6790 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are all in the same broad migration theme\u2014standardized output tracing / capture_outputs adoption across different model implementations\u2014but they do not look like duplicate fixes or mergeable alternative implementations of the same concrete change. Each PR targets a different model code path, so the soft edges should be rejected.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "No issue is present. As a representative PR for this cluster, PR #44010 is the clearest and most self-contained example of the standardized output-collection refactor, with explicit capture_outputs/can_return_tuple integration in one model file.", + "best_issue_reason": null, + "best_pr_reason": "PR #44010 is a strong cluster representative because its title and diff directly reflect the shared refactor pattern, and the change is compact and focused on a single model implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "Both are output-tracing refactors, but for different models (VisualBert vs TextNet). They do not fix the same underlying bug or change the same code path." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "Same broad refactor style, but unrelated model implementations (MPNet vs VisualBert) and different internal tracing hooks. Not the same concrete change." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "Both migrate to capture_outputs, but they touch different architectures and model-specific tracing logic. Too broad to treat as duplicates." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are different code paths and the diffs are model-specific refactors, not alternative fixes for one shared defect." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "Although both involve output tracing, SqueezeBert and SpeechT5 have distinct implementations and semantics; these are separate model-specific migrations, not one mergeable PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-13", + "nodes": [ + "pull_request:43998", + "pull_request:43999", + "pull_request:44000", + "pull_request:44001", + "pull_request:44002", + "pull_request:44003", + "pull_request:44004", + "pull_request:44025", + "pull_request:44026", + "pull_request:44027", + "pull_request:44028", + "pull_request:44029", + "pull_request:44030" + ], + "soft_pairs": [ + "pull_request:43998|pull_request:44002", + "pull_request:43998|pull_request:44000", + "pull_request:43998|pull_request:44001", + "pull_request:43998|pull_request:44003", + "pull_request:43998|pull_request:44029", + "pull_request:43999|pull_request:44001", + "pull_request:43999|pull_request:44002", + "pull_request:43999|pull_request:44025", + "pull_request:43999|pull_request:44026", + "pull_request:43999|pull_request:44027", + "pull_request:44000|pull_request:44001", + "pull_request:44000|pull_request:44002", + "pull_request:44000|pull_request: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" + ], + "prepared_review_unit_hash": "4e0b0001e227255a5c7243202731f77ec05a5fd1105866914319761a00da76a4", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41324, + "estimated_input_tokens": 10331, + "estimated_eval_tokens": 20918 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41324, + "estimated_input_tokens": 10331, + "estimated_eval_tokens": 20918 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This is a cluster of model-specific output-tracing refactors that share a common pattern but are not duplicate fixes or mergeable changes to the same code path.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No issue is present; if one PR must represent the cluster, 44025 is a good canonical example because it is a clean, self-contained tracing refactor across two closely related Depth Anything files.", + "best_issue_reason": null, + "best_pr_reason": "44025 is the strongest representative PR: focused, easy to understand, and directly shows the output-tracing conversion pattern without depending on unrelated model code.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations and code paths; same refactor theme only." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Separate model families; similar output-tracing cleanup, not the same change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different model code paths; only the refactor style overlaps." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Unrelated model families; not the same underlying bug or patch." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures and files; shared tracing refactor pattern only." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "Different models; both alter output handling but not the same concrete fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations; too broad to be duplicates." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and files; not one mergeable fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "Separate encoder/decoder code path versus MobileNet; same refactor theme only." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model-specific refactors, not the same underlying issue." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Distinct model code paths; similar tracing refactor intent only." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and implementations." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Unrelated models; no shared concrete bug or change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model implementations; same broad refactor pattern only." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model-specific changes; not mergeable as one PR." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Separate models and code paths; not duplicates." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model families; same style of refactor, not same fix." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "Unrelated models and code paths; only the output-tracing motif overlaps." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model implementations; not the same concrete change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and code paths." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Separate model-specific refactors; not duplicate fixes." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different architectures and files; too broad to merge as one PR." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model code paths; only broad tracing refactor similarity." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families and concrete implementation details." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different models and forward paths; not the same bug or change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Separate model code paths; overlap is only in tracing cleanup." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model implementations; not mergeable into one PR." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different models and changes; same general refactor theme only." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model families; no shared concrete bug." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder/decoder stacks; not the same fix." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Distinct model implementations and output paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model architectures; same refactor motif only." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Separate model-specific tracing changes; not duplicates." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model code paths and files; not the same underlying issue." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder/decoder models; too broad to merge." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "Distinct model implementations; only similar refactor style." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and output handling code." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model-specific changes; not the same patch." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Separate model families and code paths." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Different models; overlap is only in tracing refactor intent." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures and files; not the same fix." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "Distinct model implementations; not mergeable as one PR." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and code paths; same theme only." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 16, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44662-3", + "nodes": [ + "pull_request:44662", + "pull_request:45075", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401", + "pull_request:44662|pull_request:45075", + "pull_request:45075|pull_request:45401" + ], + "prepared_review_unit_hash": "2d0f0fc5315955fd6aa62002914147a970807a370e18e8c6a1ffb515d2a9293d", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10912, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10912, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three separate model-addition PRs that share Transformers integration boilerplate but implement different model families and code paths, so they should not be treated as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45075 is the most complete standalone submission in the cluster: it includes model, processing, image-processing, conversion, auto-mapping, docs, and tests. Even so, it is specific to DeepSeek-OCR-2 and not a duplicate of the other model PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 45075 is the strongest representative PR here because it has the broadest end-to-end implementation and validation coverage, but it still addresses a different model than the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "PenguinVL and Voxtral TTS are unrelated model additions for different tasks (vision-language vs TTS); they only share standard integration files, not the same underlying change." + }, + { + "left": "pull_request:44662", + "right": "pull_request:45075", + "accept": false, + "reason": "PenguinVL and DeepSeek-OCR-2 are different model implementations with distinct architectures and targets; shared boilerplate files do not make them the same change." + }, + { + "left": "pull_request:45075", + "right": "pull_request:45401", + "accept": false, + "reason": "DeepSeek-OCR-2 and Voxtral TTS modify different concrete model code paths and would not plausibly be merged into one PR; they are separate additions, not duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 17, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44979-2", + "nodes": [ + "pull_request:44979", + "pull_request:45363" + ], + "soft_pairs": [ + "pull_request:44979|pull_request:45363" + ], + "prepared_review_unit_hash": "10791feb36235b332c347e0259d1d07798c1ed67a48d1abde9a2c6601f689eaa", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related by the general theme of fusion, but they are not duplicates: PR 44979 adds a generic Module Fusion API, while PR 45363 implements a specific n-to-1 KernelConfig-based kernel fusion path with integration changes and an example.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44979 is the more foundational/general artifact: it introduces the standalone module_fusion API surface and tests, whereas PR 45363 is a separate, concrete application of fusion built around KernelConfig and runtime integrations.", + "best_issue_reason": null, + "best_pr_reason": "PR 45363 is the more complete end-user-facing implementation, but it is still distinct from 44979; if one needs a representative PR for the broader fusion work, 44979 is the cleaner canonical choice.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Different underlying changes. 44979 adds a generic module fusion API in a new module; 45363 wires up n-to-1 kernel fusion through KernelConfig and integration/modeling changes. They are adjacent, but not the same code-path fix and would not 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": 18, + "component_count": 24, + "review_unit_index": 1, + "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": "6d955ea1fdf1cd9524b976f7b1bd68da43b1c1e5a7521e1552eeae35f878d4a6", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests address the same AttributeError in `_patch_mistral_regex` by changing the same pre-tokenizer access in `tokenization_utils_tokenizers.py`; PR 45317 is the more complete version because it also adds tests.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the better canonical PR because it implements the same fix as 45086 and adds test coverage, making it the more complete representation of the change.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the best overall PR because it fixes the same concrete bug as 45086 and includes an added test file, indicating a more finalized patch.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying bug fix in the same code path, same target issue, and the second PR is essentially the first fix plus tests; they could plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 24, + "review_unit_index": 1, + "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": "cfb8b7461ef565b40654a7cdf0a087eee24689338dc543f480a26bca8841f7ab", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6065, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6065, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related NemotronH refactors, but they are not the same change: #45115 is a narrow inheritance/mapping cleanup, while #45569 is a broader model split introducing NemotronHDense and NemotronHSparse plus docs/tests. They overlap on shared core files, but they do not look mergeable into one PR without collapsing distinct goals.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45569 is the broader, more authoritative architectural change: it introduces the new NemotronHDense/NemotronHSparse structure, updates auto mappings, docs, and tests, and reads like the main follow-up model refactor rather than just a conversion-mapping cleanup.", + "best_issue_reason": null, + "best_pr_reason": "#45569 is the best representative of the cluster because it covers the larger, user-facing model refactor and likely subsumes the narrower NemotronH inheritance cleanup in scope, even though the two PRs are not duplicates.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Both touch NemotronH internals, but #45115 is focused on conversion mapping and inheritance cleanup, while #45569 adds separate dense/sparse model packages, docs, and tests. They address related architecture work, not the same concrete code-path 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": 20, + "component_count": 24, + "review_unit_index": 1, + "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": "5986cbc6945d4170982623bf046289a50c3fc4820d898aeac2f04160437713b9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5397, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5397, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related but not duplicates: one PR adds heterogeneous per-layer configuration, the other adds heterogeneous modeling/cache support that builds on that config machinery.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "45332 is the broader umbrella PR: it covers heterogeneous modeling behavior and includes the new per-layer config plumbing it depends on, whereas 45333 is the narrower config-only slice.", + "best_issue_reason": null, + "best_pr_reason": "45332 is the best representative of the cluster because it spans the larger end-to-end change and is the more central change path for heterogeneous model support.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related scope, but not the same concrete code change: 45333 adds config serialization/per-layer config support, while 45332 adds modeling/cache behavior for heterogeneous layers. They are complementary, not duplicative, and would not plausibly be merged as one PR without collapsing two distinct changes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 24, + "review_unit_index": 1, + "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": "298519b048c5901fc6a3fb69bd24e622f1d22a0776d739189a1a947a53945ed4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are typing-related pull requests, but they address different scopes: one is a broad repo-wide type-checking sweep, the other adds/adjusts specific typing protocols and related model utilities. They are not the same underlying change.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45425 is the better canonical representative because it is the more focused, concrete typing change centered on `_typing.py` and related model utility compatibility, whereas 45415 is a broad type-checking cleanup across many files.", + "best_issue_reason": null, + "best_pr_reason": "PR 45425 is the most suitable overall because it targets a specific code-path/typing compatibility problem and is easier to reason about as a standalone change; 45415 is a wider type-checking sweep rather than the same fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Reject: 45415 is a broad typing sweep with many unrelated file edits and pyproject typing-ignore changes, while 45425 is a targeted typing-protocol/update PR. They share a theme but not the same concrete bug or mergeable 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": 22, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45487-2", + "nodes": [ + "pull_request:45487", + "pull_request:45615" + ], + "soft_pairs": [ + "pull_request:45487|pull_request:45615" + ], + "prepared_review_unit_hash": "a671d878d8083c012e599652ea63b278cdaf38554fbd255017e00bc2d723da64", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5160, + "estimated_input_tokens": 1290, + "estimated_eval_tokens": 2836 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are unrelated pull requests: one fixes a real model-parallel device placement bug in AltCLIP/ChineseCLIP code, while the other only adds XPU test expectations for qianfan_ocr.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45487 is the stronger canonical PR because it contains the substantive code fix: moving buffered token type ids to the correct device in model forward paths, with corresponding model coverage across several affected architectures.", + "best_issue_reason": null, + "best_pr_reason": "PR 45487 is the best overall representative of the cluster because it addresses an actual runtime bug in model code, whereas PR 45615 is limited to test data updates and is not the same kind of change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45487", + "right": "pull_request:45615", + "accept": false, + "reason": "Reject: they do not fix the same underlying bug or change. 45487 patches device placement in model forward code for parallel execution; 45615 only adds XPU-specific expected outputs in tests for a different model." + } + ] + }, + "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": 24, + "review_unit_index": 1, + "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": "63e01767042f4ae9ab394af5a1beea79583758b7b6f16ac2fe35b8b1340898ac", + "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": "These are related EP/distributed-loading PRs, but they fix different failure modes and touch different code paths: one is about accelerate-launch/ZeRO-3 config handling, the other about FSDP2 expert tensors being overwritten during load. They should not be deduplicated.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR for duplication: the PRs target different EP bugs and are not merge-equivalent.", + "best_issue_reason": null, + "best_pr_reason": "45662 is the stronger standalone PR because it addresses a specific, concrete expert-overwrite bug with a direct code-path fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Different underlying bugs. 45548 strips distributed_config and adds has_ep for accelerate-launch/ZeRO-3 loading, while 45662 post-wraps sharded params to prevent expert tensors being overwritten under FSDP2. Shared EP plumbing is not enough to treat them 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": 24, + "component_count": 24, + "review_unit_index": 1, + "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": "13152e7f6d461b0ba41b83b184fbf1299fcd29dbd99ff0ed17d45f315a85b29b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4590, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4590, + "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 target the same xdist/captured_info logging area, but PR 45645 is broader and not a strict duplicate of 45639; it adds CI log aggregation and other artifact-handling changes beyond the core testing_utils fix.", + "confidence": 0.84, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the more representative change in this cluster because it contains the core xdist-safe captured_info fix plus the CI workflow changes needed to consume multiple per-worker files.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the best overall representative because it addresses the underlying collision problem and also updates the downstream workflow to preserve debug output end-to-end.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": false, + "reason": "Related but not the same change set: 45639 only makes the testing_utils log file xdist-safe, while 45645 also changes the CI workflow (and related artifact handling) to read multiple captured_info files. They share a subsystem and issue target, but they are not strict duplicates and would not be merged as one PR without losing scope." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/manifest.json b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..8e0a2f6cfdf94cfd524411242df6ab525fc4bdd1 --- /dev/null +++ b/snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260430t000042z", + "artifacts": { + "hybrid": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-04-30T00:10:18Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260430T000042Z", + "variant": "hybrid" +} diff --git a/snapshots/20260430T000042Z/manifest.json b/snapshots/20260430T000042Z/manifest.json index a89f50b08be9e4b86589c554c538bd8db922037b..7a6242e08a43c45e51a68b5449ba407c9d9ae615 100644 --- a/snapshots/20260430T000042Z/manifest.json +++ b/snapshots/20260430T000042Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 131 }, "extracted_at": "2026-04-30T00:00:42Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260430t000042z", + "runs": { + "hybrid-model-20260430t000042z": { + "analysis_id": "hybrid-model-20260430t000042z", + "artifacts": { + "hybrid": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260430T000042Z/analysis-runs/hybrid-model-20260430t000042z/manifest.json", + "model": null, + "published_at": "2026-04-30T00:10:18Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260430T000042Z", "watermark": {