diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 646b8617d44d183d6f13296bc93dd4955b1745ca..30c3eed06d5a76b1796f58cf9a81bbe837204f3b 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": "20260610T060026Z", - "generated_at": "2026-06-10T06:04:33Z", + "snapshot_id": "20260610T120053Z", + "generated_at": "2026-06-10T12:12:01Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1737,58 +1737,6 @@ } ] }, - { - "cluster_id": "cluster-44001-3", - "summary": "Cluster of 3 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 44002, - "issue_numbers": [], - "pr_numbers": [ - 44001, - 44002, - 44004 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 44001, - "right_pr_number": 44002, - "code_similarity": 0.175, - "size_similarity": 0.595, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.375, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 44001, - "right_pr_number": 44004, - "code_similarity": 0.087, - "size_similarity": 0.237, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.262, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 44002, - "right_pr_number": 44004, - "code_similarity": 0.125, - "size_similarity": 0.398, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.304, - "shared_filenames": [], - "shared_file_areas": [] - } - ] - }, { "cluster_id": "cluster-44018-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", @@ -2120,16 +2068,6 @@ "target_issue_number": 43979, "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." }, - { - "cluster_id": "cluster-44001-3", - "canonical_pr_number": 44002, - "duplicate_pr_numbers": [ - 44001, - 44004 - ], - "target_issue_number": null, - "reason": "PRs in cluster-44001-3 are treated as duplicates because they share soft_similarity evidence." - }, { "cluster_id": "cluster-44018-2", "canonical_pr_number": 44068, diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 595ebf92e2f6c29fd814d9c0059f37e65cbd9a25..57680c707ddb183bf3d810d3584519a7878f92b7 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": "20260610T060026Z", - "generated_at": "2026-06-10T06:04:33Z", + "snapshot_id": "20260610T120053Z", + "generated_at": "2026-06-10T12:12:01Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -13,8 +13,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -117,8 +117,8 @@ "issue:43716|issue:45685", "issue:44918|issue:45161", "issue:44292|issue:44918", - "issue:43716|issue:44292", "issue:45507|issue:45685", + "issue:43716|issue:44292", "issue:43039|issue:44918", "issue:41762|issue:43638", "issue:43827|issue:44512", @@ -132,358 +132,346 @@ "issue:43866|issue:46402", "issue:45800|issue:46304" ], - "prepared_review_unit_hash": "8d7fe39bedbf018d0ab112f2f49c002ee1124c5b279f044536ffd44769d159b9", + "prepared_review_unit_hash": "a370e3b41d1e9d19bffbc775c368791147f3d06736cc5b802b108fbe72314601", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 63, - "serialized_chars": 46518, - "estimated_input_tokens": 11630, - "estimated_eval_tokens": 23516 + "serialized_chars": 46528, + "estimated_input_tokens": 11632, + "estimated_eval_tokens": 23520 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous. The DeepSpeed ZeRO-3 empty-index failures are genuine duplicates, but the Qwen3.5MoE rotary/offload pair does not have enough body-level overlap to treat as the same bug, so it should be rejected.", - "confidence": 0.87, - "canonical_issue_reason": "Issue 43596 is the best canonical issue: it describes the shared DeepSpeed ZeRO-3 index-out-of-bounds/loading failure in the most generic form and fits both 43638 and 41762 well.", + "summary": "Mostly heterogeneous. The packet supports two likely duplicate issue pairs: the ZeRO-3/BertModel index-out-of-bounds reports and the torch 2.4.1 incompatibility reports. The docs pair is related but too broad to call a true duplicate from this packet alone.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43596 is the clearest representative of the shared ZeRO-3/BertModel index-out-of-bounds failure; issue:43638 looks like the same underlying bug with a slightly different model/setup description.", "canonical_pr_reason": null, - "best_issue_reason": "43596 is the strongest cluster representative because it captures the common ZeRO-3 empty-dimension/index-0 failure without overfitting to a specific model.", + "best_issue_reason": "issue:43596 is the best global issue canonical because it states the concrete failure mode most cleanly and matches the strongest duplicate pair.", "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 failures: a missing config attribute vs a streaming chat/completions bug." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Different problems and platforms: AMD GPU runtime failure vs MPS histogram op not implemented." + "reason": "Both are runtime/backend issues, but they involve different models and different failure modes." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Unrelated model-specific dtype mismatch vs AMD GPU failure." + "reason": "One is an image dtype mismatch; the other is an AMD GPU support problem for GPT-OSS." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Different bug classes: autocast dtype mismatch vs AMD GPU incompatibility." + "reason": "Autocast dtype mismatch for Phi-tiny-MoE is unrelated to the GPT-OSS AMD runtime issue." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Different models and symptoms; no shared concrete failure path." + "reason": "Different models and different breakage: AMD GPU support vs a broken official Nemotron example." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Both mention GPT-OSS/MoE, but one is AMD GPU support and the other is an invalid method call." + "reason": "These describe different model/runtime problems and do not point to the same code path." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch vs ModernBERT ONNX export are unrelated." + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ONNX export errors." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "Different failures: ONNX export vs Qwen3Moe FP8 configuration error." + "reason": "Different models and different export/runtime failures; not the same bug." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "ONNX export error and Qwen3 auto-device-map cuda assert are unrelated." + "reason": "ONNX export error is unrelated to Qwen3 auto device mapping CUDA asserts." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Different documentation/example problems; not the same underlying change." + "reason": "Table QA crash and outdated RAG examples are different issues." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are v5 docs cleanup, but they reference different removed items." + "reason": "Both are stale v5 docs issues, but they reference different removed artifacts and are too broad to treat as the same duplicate." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "Different failures: export-to-ONNX vs GenerationMixin import breakage." + "reason": "ONNX export failure and GenerationMixin import failure are unrelated." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "Different training issues: SFT embedding unpacking vs resume-training with trust_remote_code." + "reason": "TRL SFT embedding unpacking and resume-training/trust-remote-code behavior are different problems." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Generic mask-shape IndexError vs Gemma4 device-map mismatch are not the same bug." + "reason": "Different symptoms and subsystems: a mask-shape IndexError vs device-map forward mismatch." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Different failure modes: ZeRO-3 index error vs mask/tensor shape mismatch." + "reason": "Both mention indexing errors, but the reported failures and triggers are different enough to keep separate." }, { "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "Only broad training/loading overlap; no evidence of the same underlying bug." + "reason": "ZeRO-3 Bert indexing error is unrelated to model resume with trust_remote_code." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "ZeRO-3 empty-dimension index error is distinct from the mask/tensor shape mismatch." + "reason": "Same broad area, but not the same concrete bug: BertModel ZeRO-3 init error vs a mask/tensor shape mismatch." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Different validation failures: generic mask shape mismatch vs video feature/token mismatch." + "reason": "Different multimodal/input-shape issues with no shared failure path." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Different errors and models; no shared concrete code path." + "reason": "Mask-shape IndexError is not the same as GPT-OSS tensor-parallel failure." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Different resume-training and Nemotron-example issues." + "reason": "Resume-training issue and Nemotron example breakage are unrelated." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video chat-template StopIteration vs resume-training failure are unrelated." + "reason": "Video StopIteration in chat templating is unrelated to resume-training behavior." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "Same broad training area only; no evidence of the same underlying bug." + "reason": "Same ZeRO-3 theme does not make these the same bug; the underlying failures differ." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice_preset issue is unrelated to ModernBERT ONNX export." + "reason": "BarkProcessor voice_preset and ModernBERT ONNX export are different model-specific issues." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both are DeepSpeed-related, but collective mismatch and ZeRO-3 index error are different bugs." + "reason": "Both use DeepSpeed, but the collective mismatch and Bert zero-sized index error are different code paths." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same DeepSpeed ZeRO-3 index-out-of-bounds failure during model loading; only the model name differs." + "reason": "Both report the same ZeRO-3/BertModel index-out-of-bounds failure during model init/load." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross_entropy routing vs GPT-OSS tensor-parallel failure are different issues." + "reason": "Liger Kernel cross_entropy routing and GPT-OSS tensor-parallel failure are unrelated." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "DeepSeek detokenization bug vs LoRA save/reload mismatch are unrelated." + "reason": "Detokenization bug and LoRA merge/save mismatch are different model behaviors." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Different bugs: Qwen-3 NVFP4 runtime error vs LoRA save/reload output mismatch." + "reason": "Qwen NVFP4 runtime error and LoRA save/reload inconsistency are unrelated." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpointing/FSDP recompute mismatch is unrelated to the mask-shape IndexError." + "reason": "FSDP checkpointing mismatch and mask/index errors are not the same bug." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Both mention DeepSpeed, but the weight-shape training error and collective mismatch are different failures." + "reason": "DeepSpeed training with 2-D weight error and collective mismatch are distinct failure modes." }, { "left": "issue:45859", "right": "issue:45902", "accept": false, - "reason": "The titles suggest related Qwen3.5MoE rotary/offload areas, but that is not enough to confirm the same concrete bug without stronger body-level overlap." + "reason": "Same module, but CPU offload incompatibility and uninitialized-memory-after-to_empty are different bugs." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "Different model families and failure modes; no shared concrete bug." + "reason": "Bert ZeRO-3 indexing error is unrelated to GPT-OSS TP failure." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "Checkpointing size mismatch vs Trainer compute_loss signature change are unrelated." + "reason": "Activation-checkpointing/FSDP mismatch is unrelated to the CustomTrainer compute_loss signature change." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "ZeRO-3 BERT loading error and GPT-OSS tensor-parallel issue are different." + "reason": "ZeRO-3 Bert loading error and GPT-OSS tensor-parallel failure are unrelated." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Different model-specific problems: dtype mismatch in Mistral-3 vs invalid method call in GraniteMoEHybrid." + "reason": "Different model-specific issues with no shared bug." }, { "left": "issue:45874", "right": "issue:46294", "accept": false, - "reason": "Both are multimodal issues, but reverse-embedding lookup and video token/feature mismatch are not the same bug." + "reason": "Reverse embedding lookup cost and video token mismatch are different multimodal issues." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Different Qwen-3 NVFP4 vs GPT-OSS AMD runtime failures." + "reason": "Qwen NVFP4 runtime error and GPT-OSS AMD GPU failure are unrelated." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "vLLM config attribute error for Gemma3 is unrelated to the Mistral tokenizer regex crash." + "reason": "Gemma3 config/vLLM attribute error is unrelated to the Mistral regex patch crash." }, { "left": "issue:45800", "right": "issue:46515", - "accept": false, - "reason": "Both are torch compatibility complaints, but they do not identify the same concrete incompatibility." + "accept": true, + "reason": "Both report the same torch 2.4.1 incompatibility regression; the titles differ only in wording/detail." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Different Gemma4 bugs: device-map mismatch vs expensive reverse embedding lookup." + "reason": "Device-map forward mismatch and reverse embedding lookup slowdown are different Gemma4 problems." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Generic torch 2.10 incompatibility vs Qwen-3 NVFP4 runtime error are unrelated." + "reason": "Torch version incompatibility and Qwen NVFP4 runtime failure are not the same issue." }, { "left": "issue:34928", "right": "issue:41720", "accept": false, - "reason": "Activation checkpointing/FSDP issue and Qwen3 auto-device-map cuda assert are different failures." - }, - { - "left": "issue:43827", - "right": "issue:44512", - "accept": false, - "reason": "Different v5 docs removals: pipeline tasks vs transformers run command." + "reason": "FSDP activation checkpointing and Qwen3 auto device mapping CUDA assert are unrelated." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Autocast dtype mismatch and MPS integer histogram limitation are unrelated." + "reason": "Autocast dtype mismatch and MPS histogram integer error are different backend-specific bugs." }, { "left": "issue:44918", "right": "issue:45685", "accept": false, - "reason": "TRL SFT unpacking failure is unrelated to the MPS histogram error." + "reason": "TRL SFT embedding issue and MPS histogram failure do not share a code path." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Liger Kernel cross_entropy issue is not the same as the autocast dtype mismatch." + "reason": "Liger Kernel cross_entropy usage is unrelated to a Phi MoE autocast mismatch." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "DeepSpeed collective mismatch and CustomTrainer compute_loss signature breakage are unrelated." + "reason": "DeepSpeed collective mismatch and compute_loss signature incompatibility are unrelated." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "FSDP checkpointing mismatch vs GenerationMixin import breakage are unrelated." + "reason": "Activation checkpointing/FSDP mismatch and GenerationMixin import failure are unrelated." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger Kernel cross_entropy issue is not the same as the mask-shape IndexError." + "reason": "Liger Kernel cross_entropy routing and an index/mask shape error are unrelated." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "GPT-OSS AMD GPU runtime failure and outdated RAG examples are unrelated." + "reason": "GPT-OSS AMD runtime issue and outdated RAG examples are unrelated." }, { "left": "issue:43866", "right": "issue:46402", "accept": false, - "reason": "Checkpoint corruption vs Sam3Model weights not loading are different problems." + "reason": "Corrupted checkpoint and missing Sam3Model text encoder weights are different load-time problems." }, { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Both are compatibility regressions, but they involve different torch/version breakages." - }, - { - "left": "issue:41762", - "right": "issue:43638", - "accept": true, - "reason": "Same DeepSpeed ZeRO-3 empty-dimension/index-0 failure during model loading; Gemma3 and Bert are just different reproductions." + "reason": "Torch 2.4.1 compatibility and USE_DISTRIBUTED=0 support regression are different compatibility issues." } ] }, "evaluator_result": { "accept": false, - "feedback": "Mostly grounded, but one soft-edge acceptance is too strong for the packet. The accepted pair issue:45859 vs issue:45902 only shows related Qwen3.5MoE rotary/offload wording in the titles; that is not enough to safely call them the same underlying bug without more body-level overlap. The DeepSpeed ZeRO-3 duplicates (issue:43596/43638/41762) are well supported, and the canonical issue choice is reasonable, but the 45859/45902 acceptance overstates the evidence." + "feedback": "Mostly grounded, but not conservative enough overall. The ZeRO-3/BertModel pair (43596/43638) and torch 2.4.1 pair (45800/46515) look supportable from the titles alone, but the docs pair (44509/44512) is too broad to call a true duplicate from this packet: both are stale-v5 documentation issues, yet they reference different removed artifacts. The summary also overstates the result by saying there are 'true duplicate pairs' rather than 'likely related' pairs." }, "evaluator_used": true, "retried": true, @@ -499,8 +487,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -572,8 +560,8 @@ "issue:43872|issue:46304", "issue:39290|issue:43531", "issue:41093|issue:43638", - "issue:43854|issue:43866", "issue:43792|issue:45901", + "issue:43854|issue:43866", "issue:43425|issue:45070", "issue:43596|issue:45800", "issue:43975|issue:45901", @@ -581,10 +569,10 @@ "issue:45362|issue:45464", "issue:44292|issue:45901", "issue:39290|issue:45823", - "issue:43404|issue:43866", "issue:45464|issue:45685", "issue:43541|issue:45685", "issue:43541|issue:43828", + "issue:43404|issue:43866", "issue:44841|issue:45084", "issue:36010|issue:39290", "issue:41093|issue:41762", @@ -596,14 +584,14 @@ "issue:43541|issue:45161", "issue:43828|issue:45464" ], - "prepared_review_unit_hash": "32f433ed1338de1beec756155d0a254c883c043abe38223c734be0a60298f4c5", + "prepared_review_unit_hash": "019191a6dede7d5eec4b1c15214c43516e47c0a245ca83adc75399689cf3eade", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, @@ -618,246 +606,252 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous; only a couple of soft pairs look like true duplicates. The strongest duplicate family is the v5 AutoTokenizer regression, and there is also a plausible shared DeepSpeed ZeRO-3 load-time index error. Most other pairs share only broad subsystem or symptom similarity, so I rejected them.", - "confidence": 0.68, - "canonical_issue_reason": "issue:45920 is the best canonical issue because it is the broadest open report in the tokenizer-regression family, covering multiple affected model families and likely subsuming narrower duplicates like issue:45812.", + "summary": "The cluster is mostly heterogeneous. Only two soft edges look like true duplicates: the ZeRO-3 load-time IndexError pair (41762/43596) and the broad AutoTokenizer wrong-token-ID regression pair (45812/45920). The mask-shape IndexError pairs are rejected because the titles point to different failure shapes and code paths, not the same underlying bug.", + "confidence": 0.88, + "canonical_issue_reason": "45920 is the best canonical issue because it captures the broad multi-model AutoTokenizer v5 wrong-token-ID regression and subsumes the narrower Granite-specific report in 45812.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best overall issue to anchor the cluster: it is open, clearly describes the underlying regression, and has the widest scope among the near-duplicate candidates.", + "best_issue_reason": "45920 is the strongest overall representative issue: it is the broadest tokenizer regression report and matches the accepted duplicate pair against 45812.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "Both are model-loading failures, but they point to different causes and subsystems; not the same concrete bug." + "reason": "Different bug classes: vLLM model load/support failure vs a PyTorch distributed build/runtime compatibility break." }, { "left": "issue:41762", "right": "issue:43596", "accept": true, - "reason": "Both report the same ZeRO-3 load-time \"index 0 out of bounds\" failure; the model differs, but the code path and likely fix overlap." + "reason": "Both describe the same ZeRO-3 load-time IndexError on zero-sized tensors; only the model/context differs." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Different failures: missing Gemma3 config attribute for vLLM vs Qwen3 auto device mapping CUDA assert." + "reason": "Both are model-loading failures, but the titles indicate different underlying bugs and error paths." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Unrelated errors: float8 storage type lookup vs USE_DISTRIBUTED=0 incompatibility." + "reason": "Float8 storage lookup failure is unrelated to the distributed build/runtime incompatibility." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs-related, but they concern different documentation problems and not the same bug." + "reason": "Both are docs-related, but they describe different documentation regressions." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Different model/input paths: Gemma4 device-mapping mismatch vs Qwen3.5 video feature/token mismatch." + "reason": "Different model families and different symptoms: device mismatch vs video/chat-template feature mismatch." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "Unrelated failures across different tasks and models; not the same underlying issue." + "reason": "Completely different subsystems and error surfaces; not the same bug." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Different code paths: Voxtral processor failure vs Gemma4 PLE device mismatch." + "reason": "Processor failure and device-map mismatch are different code paths and different problems." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger cross-entropy dispatch and Qwen3.5 video StopIteration are unrelated." + "reason": "Liger-kernel cross-entropy behavior is unrelated to the video StopIteration error." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Different concerns: detokenization regression vs outdated RAG examples." + "reason": "Tokenizer detokenization regression and outdated RAG examples are unrelated." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "NemotronH checkpoint loading and RAG docs are unrelated." + "reason": "Checkpoint-loading failure and outdated documentation are unrelated." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Remote-code breakage and USE_DISTRIBUTED=0 compatibility are separate regressions." + "reason": "Remote-code breakage is a different compatibility problem than the USE_DISTRIBUTED=0 issue." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Same tokenizer regression: wrong token IDs in v5 across multiple model families; 45920 is the broader superset report." + "reason": "Same AutoTokenizer v5 wrong-token-ID regression; 45920 is the broader multi-model report." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Corrupted checkpoint vs model-class loading failure; not the same bug." + "reason": "Corrupted checkpoint vs implementation load failure are not the same underlying bug." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "Qwen-3 NVFP4 runtime error and RAG docs are unrelated." + "reason": "Model runtime failure and RAG docs are unrelated." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes constructor mismatch and distributed-build incompatibility are unrelated." + "reason": "bitsandbytes constructor incompatibility is unrelated to the PyTorch distributed build/runtime issue." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding_window, but they are different model/config failures and not clearly the same bug." + "reason": "Both mention sliding_window, but one is a missing config attribute and the other is a different Qwen3-MoE sliding_window bug." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are index errors with ZeRO-3, but the reported symptoms and contexts differ enough to treat as separate." + "reason": "Both are IndexErrors, but one is a mask-shape mismatch and the other is a zero-sized tensor access under ZeRO-3." }, { - "left": "issue:43854", - "right": "issue:43866", + "left": "issue:43792", + "right": "issue:45901", "accept": false, - "reason": "Different model-loading problems: unit-test load failure vs corrupted checkpoint." + "reason": "Whisper runtime failure and table QA crash are unrelated." }, { - "left": "issue:43792", - "right": "issue:45901", + "left": "issue:43854", + "right": "issue:43866", "accept": false, - "reason": "Whisper model failure and table-question-answering crash are unrelated." + "reason": "Different model-loading failures with different causes." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch compatibility and pydantic PretrainedConfig regression are unrelated." + "reason": "Torch version incompatibility and a pydantic field regression are unrelated." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "ZeRO-3 Bert loading issue and torch/transformers version incompatibility are unrelated." + "reason": "ZeRO-3 loading IndexError and torch/transformers version compatibility are different issues." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Detokenization regression and table-question-answering crash are unrelated." + "reason": "Detokenization regression and table QA crash are unrelated." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper loading failure and pydantic config regression are unrelated." + "reason": "Different subsystems and different errors." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Both involve Qwen3.5 chat, but one is a crash and the other is a streaming API failure; not enough evidence of the same bug." + "reason": "Both are Qwen3.5-related, but they describe different failures and different interfaces." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "Different model/runtime failures; no shared concrete code path." + "reason": "Different model/runtime problem versus table QA crash." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Different regressions: Gemma3 config attribute missing vs Gemma4 device mismatch." + "reason": "Different models and different failure modes." }, { - "left": "issue:43404", - "right": "issue:43866", + "left": "issue:45464", + "right": "issue:45685", "accept": false, - "reason": "Tied-weight bug in Mistral3 and corrupted Ovis2 checkpoint are unrelated." + "reason": "Qwen3.5 streaming failure is unrelated to the MPS unsupported-op error." }, { - "left": "issue:45464", + "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 streaming inference failure and MPS Int histogram error are unrelated." + "reason": "Different model/math-path error versus MPS unsupported-op error." }, { "left": "issue:43541", - "right": "issue:45685", + "right": "issue:43828", + "accept": false, + "reason": "Both involve MoE/precision/tracing themes, but the concrete failures and code paths differ." + }, + { + "left": "issue:43404", + "right": "issue:43866", "accept": false, - "reason": "Torch dynamo grouped_mm tracing error and MPS histogram failure are unrelated." + "reason": "Weight-tying bug and corrupted checkpoint are unrelated." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Both are indexing-related errors, but they arise in different trainer/model paths and do not look like the same bug." + "reason": "Similar-looking IndexErrors, but the titles point to different operations and different failure points." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "GenerationMixin import error and Gemma3 sliding_window_pattern error are unrelated." + "reason": "GenerationMixin import error is unrelated to the Gemma3 sliding_window_pattern failure." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are index errors, but one is a mask-shape mismatch and the other is a ZeRO-3 empty-tensor load failure." + "reason": "Both are IndexErrors, but one is a mask-shape mismatch and the other is a ZeRO-3 zero-sized tensor access." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch/version compatibility issue and Gemma4 device mismatch are unrelated." + "reason": "Torch/transformers version compatibility is unrelated to Gemma4 device placement." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different unsupported-model/loading issues; not the same underlying change." + "reason": "PaddleOCR-VL load failure and gpt-oss GGUF support are different model-support issues." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and PretrainedConfig pydantic breakage are different config regressions." + "reason": "Missing pad_token_idx in a config and a pydantic-field regression are unrelated." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Different failure modes and contexts; no clear shared code path." + "reason": "Different stack traces and different failing operations; not the same bug." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Int4 OOM and tied-weights translation-model issues are unrelated." + "reason": "int4 OOM regression and tied-weights translation-model issue are unrelated." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Grouped_mm tracing failure and GPT-OSS TP-only failure are unrelated." + "reason": "Different MoE-related failures with different symptoms and code paths." }, { "left": "issue:43828", @@ -868,11 +862,11 @@ ] }, "evaluator_result": { - "accept": true, - "feedback": "The cluster summary is grounded in the packet: issue:45920 is a reasonable canonical anchor, and the only accepted soft pairs are conservative and supported by strong title-level overlap (especially the AutoTokenizer v5 regression family). The rejected pairs also stay cautious where the packet only shows broad symptom similarity. No overstatement stands out." + "accept": false, + "feedback": "Most of the pairwise calls are conservative and grounded in the titles, and the tokenizer anchor on 45920 looks reasonable. However, the summary overstates the evidence by saying there are genuine duplicate mask-shape IndexError pairs, even though the corresponding verdicts are both rejects. Please align the summary with the actual soft-edge outcomes and avoid calling those cases duplicates unless you can justify it from the packet." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -885,8 +879,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -972,226 +966,226 @@ "issue:43644|issue:43950", "issue:36331|issue:41762" ], - "prepared_review_unit_hash": "e923f41a238c0b2aa14d0a15ac368d442d7f01b3d9b409ab63b75d3efad16e80", + "prepared_review_unit_hash": "80bed78dbc4b86e7f9d2a05dc6818c90ecf30c49e8f735370d050e6af059dc47", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36944, - "estimated_input_tokens": 9236, - "estimated_eval_tokens": 18728 + "serialized_chars": 36940, + "estimated_input_tokens": 9235, + "estimated_eval_tokens": 18726 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issues with a few real duplicate families: non-persistent-buffer regression, Qwen3.5 video-input regression, and input-validation error-message bugs. Most other soft pairs are only broadly related by model family or subsystem and should remain separate.", - "confidence": 0.78, - "canonical_issue_reason": "43644 is the broadest/earliest report in the strongest duplicate family here; 43950 and 44534 look like the same non-persistent-buffer corruption regression.", + "summary": "This cluster is mostly a noisy mix of distinct Transformers issues that only share broad wording (v5 regressions, model-loading bugs, tokenizer bugs, etc.). I found one clear duplicate family around non-persistent buffer corruption and one likely duplicate around input validation error messaging; the rest are too different to merge.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 43950 is the clearest representative of the strongest duplicate family here: it states the non-persistent-buffer corruption regression explicitly and matches the near-identical v5 reports in the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "43950 is the clearest reproduction-focused report of the non-persistent-buffer regression, so it is the best issue representative among the matched set.", + "best_issue_reason": "43950 is the best issue anchor because it names the root bug most precisely and is the most actionable summary of the duplicate buffer-corruption reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "Different regressions: init_empty_weights/_is_hf_initialized vs. PretrainedConfig/pydantic field handling." + "reason": "Different bugs: unexpected init argument handling vs pydantic field compatibility." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "Unrelated bugs: GGUF gpt-oss support vs. Qwen3.5 streaming chat/completions failure." + "reason": "Unrelated symptoms and code paths: GGUF architecture support vs streaming chat/completions failure." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Different Qwen3-MoE sliding_window bug vs. a v5 pydantic/config regression." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to pydantic config regression." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Both are multimodal input issues, but one is image preprocessor dtype mismatch and the other is video token/feature mismatch; not the same bug." + "reason": "Different multimodal issues: image preprocessor dtype mismatch vs video feature/token count mismatch." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote-code breakage and torch/transformers version incompatibility are different failure modes." + "reason": "Remote-code breakage and torch/transformers incompatibility are separate root causes." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "Different code paths: torch dynamo grouped_mm tracing vs. float8 storage lookup." + "reason": "Grouped_mm tracing failure is not the same as missing Float8 storage type support." }, { "left": "issue:44479", "right": "issue:46294", - "accept": true, - "reason": "Same Qwen3.5/Qwen3.5-MoE video-input regression; 46294 is a specific manifestation of the broader 44479 bug." + "accept": false, + "reason": "Both are video-related, but one is a v5.3 regression across models and the other is a specific chat-template mismatch." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are IndexErrors, but the reported shapes/code paths differ and do not point to the same defect." + "reason": "Both are shape/index errors, but they involve different model setups and failure modes." }, { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Same model family, but one is a subjective quality complaint and the other is an AutoModel/pipeline misbehavior; not clearly the same bug." + "reason": "Different SigLIP2 problems: embedding quality complaint vs incorrect pipeline/AutoModel output." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Different multimodal/model-init problems: unwrapping text_config vs. uninitialized rotary embedding memory." + "reason": "Composite-config unwrapping bug and uninitialized rotary embedding memory are distinct issues." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "Different failures: FSDP weight-shape error vs. TRL SFT embedding unpacking issue." + "reason": "FSDP weight-dimension error is unrelated to TRL SFT embedding unpacking." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Separate multimodal/training issues with different models and symptoms." + "reason": "Mistral-3 preprocessing dtype issue and Qwen3.5 SFT trainer embedding issue are different." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "Different model/regression areas: GPT-OSS perplexity vs. GLM5 behavior." + "reason": "High perplexity on GPT-OSS and GLM5 are unrelated model-specific reports." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model init regression is unrelated to flash-attn fallback gating." + "reason": "Jupyter custom-model initialization regression is not the same as flash-attn fallback gating." }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Both involve remote-code/custom-module loading, but the cache-path collision and post-save_pretrained module mismatch are not the same concrete bug." + "reason": "Both involve custom-code loading, but one is a local leaf-dir cache collision and the other is wrong module selection after save/load." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Both mention tied weights, but these are different model families and different weight-tying regressions." + "reason": "Both mention tied weights, but they are model-family-specific reports without clear evidence of the same bug." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure is unrelated to flash-attn fallback checks." + "reason": "Qwen3MoE FP8 failure is unrelated to flash-attn2 fallback blocking." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache/linear_attention crash is a different subsystem from tied-weights regressions." + "reason": "PagedAttentionCache linear_attention crash is not the same as tied-weight regressions." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is not the same bug as tied-weight handling." + "reason": "Quantization offloading behavior is unrelated to tied-weight issues." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A DSA indexer implementation issue is unrelated to a docs typo." + "reason": "DSA indexer discussion and documentation wording are unrelated." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading failure is distinct from tied-weight issues." + "reason": "Checkpoint-loading failure for NemotronH is not a tied-weight regression." }, { "left": "issue:45183", "right": "issue:46463", "accept": true, - "reason": "Same input-validation/error-message bug; 46463 is the T5-specific instance of 45183's broader XOR validation issue." + "reason": "Same validation code path: 46463 looks like a concrete T5-family instance of the broader input_ids/inputs_embeds XOR error-message bug." }, { "left": "issue:46489", "right": "issue:46491", "accept": false, - "reason": "Different tokenizer regressions for different models and different symptoms." + "reason": "Different tokenizer regressions affecting different model families and round-trip behaviors." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Unpacking Qwen3.5 embeddings in TRL is unrelated to GraniteMoEHybrid invalid method calls." + "reason": "TRL embedding unpacking and GraniteMoEHybrid invalid-method calls are unrelated." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Docs typo vs. removed run command documentation mismatch; not the same bug." + "reason": "A typo in docs is not the same as a removed command mention." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Different tokenization/runtime failures in different models." + "reason": "DeepSeek-Coder detokenization and Qwen-3-8B-NVFP4 runtime failure are different problems." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "Qwen NVFP4 runtime error is unrelated to GraniteMoEHybrid method dispatch." + "reason": "NVFP4 runtime failure and GraniteMoEHybrid method selection are unrelated." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Near-identical reports of the same v5 non-persistent-buffer corruption regression." + "reason": "Near-identical regression: v5 corrupts/fills non-persistent buffers with junk; these are the same underlying bug." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Both mention SigLIP2, but one is a docs/example mismatch and the other is an embedding-quality complaint; not the same concrete bug." + "reason": "SigLIP2 doc-example errors and text-embedding quality complaints are not the same bug." }, { "left": "issue:43644", "right": "issue:43950", "accept": true, - "reason": "Same non-persistent-buffer bug; 43644 is the broader report and 43950 describes the same corruption more explicitly." + "reason": "Same non-persistent-buffer corruption regression in Transformers v5, just described from different angles." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Different TypeErrors/IndexErrors with unrelated training vs. ZeRO-3 loading code paths." + "reason": "Unexpected trainer kwarg and DeepSpeed zero-3 index error are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The accepted soft pairs are limited to clearly overlapping bug families (non-persistent-buffer corruption, Qwen3.5 video-input regression, and XOR/error-message validation), while the rejected pairs are appropriately cautious about shared model families or subsystems without stronger evidence of the same defect." + "feedback": "The summary is grounded in the packet and stays conservative: it identifies the non-persistent-buffer corruption family and the input-validation error-message family without overextending beyond the evidence. The accepted soft-edge pairings also look defensible as same-bug families, and the other rejections are suitably cautious." }, "evaluator_used": true, "retried": false, @@ -1207,8 +1201,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -1271,8 +1265,8 @@ "issue:38175|issue:43054", "issue:41553|issue:43883", "issue:44910|issue:45859", - "issue:45663|issue:45902", "issue:42947|issue:44387", + "issue:45663|issue:45902", "issue:43054|issue:43493", "issue:41553|issue:43881", "issue:44403|issue:44451", @@ -1284,250 +1278,250 @@ "issue:41762|issue:42915", "issue:45685|issue:45823", "issue:43782|issue:44960", - "issue:43881|issue:43883", "issue:43824|issue:43872", "issue:44991|issue:45399", + "issue:43881|issue:43883", "issue:45020|issue:45823", "issue:44661|issue:45020", - "issue:42491|issue:44368", "issue:43828|issue:45823", "issue:45081|issue:45399", + "issue:42491|issue:44368", "issue:45715|issue:46315", "issue:45488|issue:46489", "issue:45020|issue:45356" ], - "prepared_review_unit_hash": "560d85c3b4cd279cc2ba2c685fe54cca85fc347e032c260dc00c5d72df54bae1", + "prepared_review_unit_hash": "e5cdb599d7e2ea15e317bb8ff8309e35275c065f421906120607b3edb620de1c", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37995, - "estimated_input_tokens": 9499, - "estimated_eval_tokens": 19254 + "serialized_chars": 38002, + "estimated_input_tokens": 9501, + "estimated_eval_tokens": 19258 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose mix of separate Transformers regressions, with the strongest cohesive subgroup around tokenizer/output regressions in v5. Most soft pairs are not true duplicates; one pair is plausibly the same tokenizer bug family.", - "confidence": 0.62, - "canonical_issue_reason": "issue:45920 is the broadest and most representative report of the shared tokenizer regression class: incorrect token IDs across multiple models in v5. It captures the common failure mode better than the narrower follow-up reports.", + "summary": "Most items are distinct bugs. One soft duplicate looks plausible in the non-persistent-buffer/materialization path: a broad `from_pretrained()` buffer corruption regression and a model-specific `inv_freq` uninitialized-memory failure. The strongest issue candidate overall is the broad tokenizer regression affecting multiple model families.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45920 is the best umbrella issue: it describes a current v5 tokenizer regression with wrong token IDs across multiple model families, which is broader and more representative than the more model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best issue anchor for the cluster because it is general, open, and centered on the core shared code-path problem affecting multiple model families.", + "best_issue_reason": "Issue 45920 is the most actionable global representative because it names the shared failure mode, cites several affected models, and matches the recurring tokenizer-breakage theme in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Different failures: one is a dtype mismatch under autocast, the other is a Qwen3.5 attention/position_ids NaN/illegal-memory issue." + "reason": "Both involve model/runtime failures, but the concrete bugs differ: one is an autocast dtype mismatch, the other is a flash-attention packed-sequence/position_ids crash." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Both are loading/compatibility errors, but they involve different low-level incompatibilities and different failure signatures." + "reason": "Both are quantization/storage compatibility errors, but they point to different failure modes and different tensor/storage classes." }, { "left": "issue:45920", "right": "issue:46489", - "accept": true, - "reason": "Both describe v5 tokenizer output/token-id regressions; 46489 reads like a specific DeepSeek-Coder v1 manifestation of the broader issue in 45920." + "accept": false, + "reason": "Both are tokenizer regressions, but 45920 is a broad multi-model wrong-ID bug while 46489 is a specific DeepSeek-Coder v1 issue tied to a particular PR gap." }, { "left": "issue:43950", "right": "issue:45902", - "accept": false, - "reason": "Both involve buffer/materialization problems, but one is a broad from_pretrained non-persistent buffer regression and the other is a specific Qwen3_5Moe inv_freq materialization bug." + "accept": true, + "reason": "These look like the same underlying non-persistent-buffer/materialization regression: 43950 is the generic `from_pretrained()` buffer corruption report, and 45902 is a concrete Qwen3_5Moe manifestation reading uninitialized buffer memory after meta materialization." }, { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "Unrelated topics: a corrupted checkpoint versus an issue about loading noise." + "reason": "One reports a corrupted checkpoint, the other just noisy loading output; they do not describe the same bug." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "Different problem classes: model-quality regression for SigLIP2 versus a tokenizer-version change issue." + "reason": "SigLIP2 embedding quality and a general tokenization-version change are not the same failure mode." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Same model family, but different symptom and no clear evidence of the same underlying defect." + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is degraded text embeddings; too different to merge." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Different model-loading errors with no shared concrete code path." + "reason": "Different model-loading problems for different models and code paths." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Both are Qwen3.5-related, but one is flash-attention packed-sequence misinterpretation and the other is CPU-offload incompatibility." + "reason": "Both touch Qwen3.5 MoE, but one is a flash-attention memory access crash and the other is CPU-offload incompatibility in rotary embeddings." }, { - "left": "issue:45663", - "right": "issue:45902", + "left": "issue:42947", + "right": "issue:44387", "accept": false, - "reason": "Different models and different failure mechanisms; no clear duplicate evidence." + "reason": "Gradient checkpointing/LoRA and int4 quantization OOM are separate training-memory regressions." }, { - "left": "issue:42947", - "right": "issue:44387", + "left": "issue:45663", + "right": "issue:45902", "accept": false, - "reason": "Both mention training/memory concerns, but they are distinct bugs with different triggers and code paths." + "reason": "FSDP2 shared_kv_state rebuilding and non-persistent buffer materialization are unrelated code paths." }, { "left": "issue:43054", "right": "issue:43493", "accept": false, - "reason": "Related SigLIP2 topic, but one is a symptom report and the other is an implementation-discrepancy report; not enough to treat as the same concrete bug." + "reason": "Both concern SigLIP2, but one is relative embedding quality and the other is an implementation discrepancy versus JAX; not clearly the same bug." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Different model-specific loading failures." + "reason": "Bad error message for Voxtral AutoTokenizer and glm-4v-9b loading failure are distinct issues." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "A generic loading-noise complaint is not the same bug as a specific model load failure." + "reason": "Generic loading noise versus a specific model load failure." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "Completely different runtime/training behaviors: DeepSpeed weight shape error versus early-stopping metric handling." + "reason": "DeepSpeed training weight-shape error and early-stopping metric detection are unrelated." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Tokenizer detokenization bug versus import-time flash-attn lookup on Python 3.13; unrelated." + "reason": "Tokenizer detokenization and a flash-attn import error are unrelated." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Different severity and likely different root cause; one is generic noise, the other is a concrete tokenizer load failure." + "reason": "Noise on load versus a specific tokenizer load failure for a different model." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Both concern training efficiency, but they are not the same concrete bug or code path." + "reason": "These are both training/memory-related but cover different regressions and model families." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Related to remote-code loading, but 45020 is a broad umbrella issue and 43643 is a specific missing-fields symptom; too little evidence of one exact duplicate." + "reason": "Missing fields in `AutoConfig` with `trust_remote_code=True` is not the same as a broad remote-code compatibility regression." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Different models and different DeepSpeed/quantization failures." + "reason": "DeepSpeed ZeRO-3 loading crash and FineGrainedFP8Config failure are separate model-initialization bugs." }, { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "MPS histogram support error is unrelated to a device mismatch in Gemma4 forward." + "reason": "MPS histogram integer support and Gemma4 device placement mismatch are different backend issues." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Different model-loading failures with no shared concrete defect." - }, - { - "left": "issue:43881", - "right": "issue:43883", - "accept": false, - "reason": "Different model families and different attribute/loading errors." + "reason": "Qwen3VL weight-only loading and GLM5 are unrelated." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "ImportError for a missing class versus a bitsandbytes storage-parameter incompatibility." + "reason": "ImportError for a model class and bitsandbytes storage incompatibility are unrelated." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression versus flash-attn fallback logic; unrelated." + "reason": "Tokenizer loading for a specific model and flash-attn fallback gating are unrelated." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different models and different failure modes during loading." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Broad remote-code breakage versus a Gemma4 device placement bug." + "reason": "Remote-code breakage and Gemma4 device mismatch are separate regressions." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping-name edge case versus broad remote-code loading regressions; not the same issue." - }, - { - "left": "issue:42491", - "right": "issue:44368", - "accept": false, - "reason": "Different Qwen-related problems: LoRA compatibility across versions versus a warning about tied embeddings." + "reason": "Model-registration logic for tokenizer mappings is not the same as a remote-code loading regression." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Different training/device issues with no shared bug signature." + "reason": "Dtype mismatch in autocast and device mismatch in Gemma4 are different runtime errors." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Mistral regex crash versus flash-attn fallback gating; unrelated." + "reason": "Tokenizer regex patch crash and flash-attn fallback checks are unrelated." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA/qwen3_moe hf4.x vs hf5.x compatibility and a warning about `tie_word_embeddings` are not the same issue." }, { "left": "issue:45715", "right": "issue:46315", "accept": false, - "reason": "Both are tokenizer performance regressions, but they affect different conversion functions and likely different loops." + "reason": "Both are tokenizer performance regressions, but one is `convert_ids_to_tokens` and the other is `convert_tokens_to_ids`; different code paths." }, { "left": "issue:45488", "right": "issue:46489", "accept": false, - "reason": "Both are DeepSeek-tokenizer regressions, but the titles point to different concrete defects and models; not enough evidence they are the same bug." + "reason": "Both involve DeepSeek tokenizer breakage, but 45488 is a LlamaTokenizer pre-tokenizer override bug while 46489 is a DeepSeek-Coder v1 regression tied to a specific fix gap." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Remote-code loading breakage and a Kimi-K2.5 tokenizer codec regression are different issues." + "reason": "Generic remote-code breakage and Kimi-K2.5 tokenizer codec handling are different problems." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately conservative. The soft-edge calls mostly separate distinct failure modes, and the one accepted tokenizer pair (45920/46489) is plausibly the same v5 tokenizer-regression family without overstating certainty." + "feedback": "The summary is grounded in the packet and stays mostly conservative. The chosen umbrella issue (45920) is well supported by its title as a broad v5 tokenizer regression affecting multiple model families, and the single accepted soft pair (43950/45902) is plausibly the same non-persistent-buffer/materialization path. The rejected pairs are also framed cautiously without overstating equivalence." }, "evaluator_used": true, "retried": false, @@ -1543,8 +1537,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -1578,6 +1572,7 @@ "issue:44509", "issue:44530", "issue:44560", + "issue:44589", "issue:44805", "issue:44863", "issue:44928", @@ -1589,7 +1584,6 @@ "issue:45584", "issue:45588", "issue:45663", - "issue:45685", "issue:45823", "issue:45834", "issue:45902", @@ -1599,15 +1593,15 @@ ], "soft_pairs": [ "issue:43742|issue:45399", - "issue:43404|issue:43950", "issue:43827|issue:44509", + "issue:43404|issue:43950", "issue:43872|issue:44291", "issue:42915|issue:43872", "issue:43792|issue:44220", "issue:42947|issue:45923", - "issue:41553|issue:44991", - "issue:43257|issue:45305", "issue:36331|issue:41093", + "issue:43257|issue:45305", + "issue:41553|issue:44991", "issue:43940|issue:45313", "issue:43854|issue:44863", "issue:44928|issue:45588", @@ -1616,226 +1610,226 @@ "issue:41553|issue:44488", "issue:44863|issue:46402", "issue:24643|issue:30064", - "issue:44530|issue:44863", "issue:43596|issue:46515", + "issue:44530|issue:44863", "issue:43366|issue:43828", "issue:45325|issue:45902", "issue:43638|issue:46515", - "issue:45663|issue:45823", "issue:43366|issue:43531", + "issue:45663|issue:45823", "issue:44155|issue:45584", "issue:43716|issue:44560", "issue:43950|issue:45834", - "issue:44560|issue:45685" + "issue:43366|issue:44589" ], - "prepared_review_unit_hash": "18330106eea73c9789dc3d7254766aaeceefcb61e2b54471e7276b39548b34a4", + "prepared_review_unit_hash": "f9a54e435c702397705f9c31d69855ff3b4cf86b6a3ad617d2f7f7a0b1e4dbc4", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36680, - "estimated_input_tokens": 9170, - "estimated_eval_tokens": 18596 + "serialized_chars": 36669, + "estimated_input_tokens": 9168, + "estimated_eval_tokens": 18592 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mix of unrelated issues with a few clear duplicate subclusters. The strongest matches are the docs cleanup around removed pipeline tasks, the identical mask-shape IndexError reports, the BertModel/ZeRO-3 out-of-bounds loading reports, and the `_is_hf_initialized` loading regression; most other pairs are only broadly similar and should stay separate.", - "confidence": 0.84, - "canonical_issue_reason": "No single canonical issue fits the whole set; the clearest duplicate subcluster is 41093/44805, where 41093 is the best representative because it is the more generic formulation of the same mask-shape mismatch.", + "summary": "Mostly heterogeneous issue cluster with two clear duplicate subpairs: docs references to removed v5 pipeline tasks, and the `_is_hf_initialized` constructor regression. The rest look like related-at-a-distance model-loading or training failures, but not the same underlying bug.", + "confidence": 0.64, + "canonical_issue_reason": "Issue 43872 is the strongest concrete bug report here: it captures the `_is_hf_initialized` loading regression clearly, and 44291 is essentially a narrower reproducer of the same constructor failure.", "canonical_pr_reason": null, - "best_issue_reason": "41093 is the best issue to anchor the strongest duplicate pair: its wording is generic, and 44805 appears to be the same mask/tensor shape mismatch with different concrete sizes.", + "best_issue_reason": "43872 is the cleanest representative issue in the set: specific failure, clear symptom, and the best duplicate anchor among the items.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43742", "right": "issue:45399", "accept": false, - "reason": "Different failures: MobileLLM tokenizer/key loading vs flash-attn2 fallback gating. Same broad loading area, but not the same bug." - }, - { - "left": "issue:43404", - "right": "issue:43950", - "accept": false, - "reason": "Both concern model loading/state, but one is untied lm_head weights in Mistral3 and the other is non-persistent buffer corruption in from_pretrained(). Different code-path bug." + "reason": "Different problems: tokenizer/key loading failure vs flash-attn fallback gating; same broad loading area but not the same bug." }, { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "Both report obsolete docs still referencing removed pipeline tasks after v5; this looks like the same documentation cleanup." + "reason": "Both report the same docs regression: v5 docs still mention removed pipeline tasks / pipeline() references." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Different code paths: tied lm_head weight bug vs silent corruption of non-persistent buffers during from_pretrained()." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Both describe the same `_is_hf_initialized` unexpected-argument TypeError during model loading; one shows it via bitsandbytes Int8Params, the other via init_empty_weights." + "reason": "Same underlying `_is_hf_initialized` constructor/loading regression, just with different reproductions and wording." }, { "left": "issue:42915", "right": "issue:43872", "accept": false, - "reason": "Different regressions: Qwen3 MoE with FineGrainedFP8Config vs a bitsandbytes `_is_hf_initialized` compatibility issue." + "reason": "Both mention loading failures, but one is a Qwen3-MoE + FP8 config issue and the other is a constructor kwarg regression; not the same bug." }, { "left": "issue:43792", "right": "issue:44220", "accept": false, - "reason": "Whisper runtime failure and `_torch_extract_fbank_features()` are related to audio preprocessing, but the reported failures are not clearly the same underlying bug." + "reason": "Related audio area, but a general model run failure is not the same as a specific fbank extraction issue." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Both mention gradient checkpointing, but one is PEFT/LoRA ineffectiveness and the other is a missing supports_gradient_checkpointing flag on a trust_remote_code variant." + "reason": "Different training behaviors: LoRA gradient checkpointing ineffectiveness vs missing supports_gradient_checkpointing flag." }, { - "left": "issue:41553", - "right": "issue:44991", + "left": "issue:36331", + "right": "issue:41093", "accept": false, - "reason": "Both are tokenizer-loading problems, but they affect different models and different failure modes; not enough evidence of a shared root cause." + "reason": "Unexpected trainer kwarg vs tensor-mask shape mismatch; unrelated failures." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Different DeepSpeed-related bugs: Qwen3 MoE weight conversion with accelerate+deepspeed vs gradient averaging with GAS/model_accepts_loss_kwargs." + "reason": "Different DeepSpeed issues: weight conversion/loading vs gradient averaging under GAS." }, { - "left": "issue:36331", - "right": "issue:41093", + "left": "issue:41553", + "right": "issue:44991", "accept": false, - "reason": "Unexpected compute_loss kwarg and mask-shape IndexError are unrelated failures." + "reason": "Different tokenizer/model load failures on different models; too generic to be the same bug." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are ZeRO-3 weight-loading failures in Qwen-family models, but the affected models and missing-weight patterns differ enough that this is not clearly one bug." + "reason": "Both are ZeRO-3 weight-loading failures, but the affected model paths and symptoms differ enough that they do not clearly collapse to one bug." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Both are model-loading issues, but GLM-4.7-Flash test loading and NemotronH checkpoint loading are distinct problems." + "reason": "Different model-loading failures: unit-test load issue for GLM-4.7-Flash vs NemotronH checkpoint load failure." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "Qwen3.5 RLHF NaNs from position_ids/SDPA fallback and a flash_attention.py s_aux=None crash are unrelated." + "reason": "Unrelated failures: RLHF gradient explosion from position_ids/SDPA behavior vs flash_attention crash on s_aux=None." }, { "left": "issue:41093", "right": "issue:44805", - "accept": true, - "reason": "These are effectively the same mask/tensor shape mismatch IndexError, with only the concrete sizes changed." + "accept": false, + "reason": "Both are generic mask/tensor index errors, but there is no evidence they share the same root cause or code path." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both concern video handling in Qwen VL models, but one is a broad v5.3.0 regression and the other is a specific StopIteration in Qwen3-vl-embedding; not enough to merge." + "reason": "Different regressions in the video stack: one is a broad v5 input regression across multiple models, the other is a StopIteration in Qwen3-VL embedding video handling." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Different tokenizer-loading failures for different models; same general area, not the same bug." + "reason": "One is a bad error message for Voxtral tokenizer loading; the other is a tokenizer load failure for a different model." }, { "left": "issue:44863", "right": "issue:46402", "accept": false, - "reason": "Different model families and failure modes: NemotronH checkpoint loading vs Sam3Model text encoder weights not loading." + "reason": "Both are model-load related, but NemotronH checkpoint loading and Sam3 text-encoder weights not loading are distinct failures." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "DeepSpeed training weight-dimension error and segmentation-map processing failure are unrelated." + "reason": "Completely unrelated: DeepSpeed training RuntimeError vs image processor void segmentation-map handling." }, { - "left": "issue:44530", - "right": "issue:44863", + "left": "issue:43596", + "right": "issue:46515", "accept": false, - "reason": "PagedAttentionCache / linear_attention crash and NemotronH checkpoint loading are different code paths." + "reason": "Different failures: ZeRO-3/Bert index error vs torch version incompatibility." }, { - "left": "issue:43596", - "right": "issue:46515", + "left": "issue:44530", + "right": "issue:44863", "accept": false, - "reason": "Zero-3 BertModel index-out-of-bounds and torch 2.4.1 incompatibility are not the same underlying issue." + "reason": "PagedAttention invalid group type and NemotronH checkpoint loading are different model/runtime issues." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF gpt-oss support and Phi-tiny-MoE autocast dtype mismatch are unrelated." + "reason": "GGUF support request vs autocast dtype mismatch; unrelated." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Different bugs in different model components: Qwen2.5-VL rope scaling vs Qwen3.5-MoE rotary embedding materialization." + "reason": "Both touch rotary/position handling in Qwen models, but the symptoms and failure points differ and do not look like the same bug." }, { "left": "issue:43638", "right": "issue:46515", "accept": false, - "reason": "The BertModel zero-3 loading bug is unrelated to the torch version incompatibility report." + "reason": "Non-pretrained Bert ZeRO-3 index error is unrelated to the torch 2.4.1 incompatibility report." }, { - "left": "issue:45663", - "right": "issue:45823", + "left": "issue:43366", + "right": "issue:43531", "accept": false, - "reason": "Both are Gemma4 issues, but one is a shared_kv_states/FSDP2 KeyError and the other is a device mismatch in PLE during forward; distinct bugs." + "reason": "GGUF architecture support and Qwen3-MoE sliding_window behavior are unrelated." }, { - "left": "issue:43366", - "right": "issue:43531", + "left": "issue:45663", + "right": "issue:45823", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window behavior are unrelated." + "reason": "Gemma4 FSDP2 shared_kv_states bug vs PLE device mismatch; different forward-path defects." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Audio-flamingo embedding/token leakage across tracks and Whisper empty-transcription failure are different bugs." + "reason": "AudioFlamingo3 track leakage vs Whisper empty transcription after align_special_tokens; not the same underlying issue." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Both involve multimodal/video models, but the reported dtype mismatch and StopIteration are different failures." + "reason": "Different video/model issues: dtype mismatch in image preprocessing vs StopIteration in Qwen3-VL video embedding." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer corruption and Kosmos2.5 OCR index errors are unrelated." + "reason": "Silent buffer corruption during from_pretrained() is unrelated to OCR long-input index errors." }, { - "left": "issue:44560", - "right": "issue:45685", + "left": "issue:43366", + "right": "issue:44589", "accept": false, - "reason": "Qwen3-vl video StopIteration and an MPS histogram int-typing error are different code paths." + "reason": "GGUF gpt-oss support request and Float8 storage lookup failure are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The accepted pairs are well-supported by the titles/bodies (docs cleanup, the `_is_hf_initialized` loading TypeError, and the mask-shape IndexError), and the rejected pairs do not overclaim shared root cause. The note that the cluster is mostly unrelated with a few clear duplicate subclusters is appropriately cautious." + "feedback": "The summary is grounded and conservative. The two accepted soft pairs are well-supported by the packet: 43827/44509 is the same v5 docs regression, and 43872/44291 is the same `_is_hf_initialized` loading/construction bug. The other rejections stay appropriately cautious and do not overclaim duplication from broad thematic similarity alone." }, "evaluator_used": true, "retried": false, @@ -1851,8 +1845,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -1880,7 +1874,6 @@ "issue:44488", "issue:44534", "issue:44560", - "issue:44589", "issue:44661", "issue:44683", "issue:44756", @@ -1896,6 +1889,7 @@ "issue:45313", "issue:45357", "issue:45507", + "issue:45685", "issue:45812", "issue:45823", "issue:45834", @@ -1908,9 +1902,9 @@ "soft_pairs": [ "issue:44683|issue:46515", "issue:43824|issue:43957", - "issue:43366|issue:44589", - "issue:43645|issue:43646", + "issue:44560|issue:45685", "issue:42915|issue:43957", + "issue:43645|issue:43646", "issue:43550|issue:44910", "issue:44560|issue:44918", "issue:42617|issue:43366", @@ -1926,262 +1920,262 @@ "issue:43596|issue:44661", "issue:44661|issue:44805", "issue:43317|issue:44756", - "issue:43644|issue:44534", "issue:38175|issue:39692", + "issue:43644|issue:44534", "issue:42491|issue:43720", "issue:41720|issue:42915", "issue:46459|issue:46512", "issue:42915|issue:44756", - "issue:45313|issue:45357", "issue:44451|issue:44488", - "issue:44960|issue:44977", "issue:42491|issue:45276", + "issue:44960|issue:44977", + "issue:45313|issue:45357", "issue:45507|issue:46468", - "issue:43644|issue:43645", "issue:44488|issue:44991", + "issue:43644|issue:43645", "issue:39692|issue:43994" ], - "prepared_review_unit_hash": "9d28ad13f10cb2bff3ad856a4d1542e67b14a3c419bb1d68bf80d8f9c7939a21", + "prepared_review_unit_hash": "616cbca3151856ec32c02a2c9fc6a12e4acb39b7f577c08ccb04e6b394fdc1e4", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37628, - "estimated_input_tokens": 9407, - "estimated_eval_tokens": 19070 + "serialized_chars": 37648, + "estimated_input_tokens": 9412, + "estimated_eval_tokens": 19080 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated issues; only two pairs look like true duplicates: the non-persistent-buffer regression and the custom-model initialization regression. The SigLIP2 and several Qwen/Granite/tensorizer pairs are similar in topic but not the same underlying bug.", - "confidence": 0.87, - "canonical_issue_reason": "No single canonical issue fits the whole set because the items span many unrelated bugs. Among the clear duplicates, 44534 is the best anchor for the non-persistent-buffer regression, and 43646 is the best anchor for the custom-model initialization regression.", + "summary": "Mostly a heterogeneous issue cluster with a few tight duplicate pairs. The strongest matches are the custom-model init reports (43645/43646), the non-persistent-buffer regression (43644/44534), the SigLIP2 bad-output reports (38175/43994), and the stricter-config-loading failures (44877/45030).", + "confidence": 0.83, + "canonical_issue_reason": "issue:43646 is the best canonical issue because it is the broadest, clearest statement of the custom-model initialization regression, and it cleanly subsumes the narrower Jupyter-specific report in 43645.", "canonical_pr_reason": null, - "best_issue_reason": "44534 is the strongest global issue anchor: it is a clean, version-agnostic description of an exact duplicate of 43644, while the other candidates are either narrower variants or separate bugs.", + "best_issue_reason": "issue:43646 is the best issue anchor for this cluster because it captures a concrete, general bug rather than a niche manifestation, while still representing one of the few clearly duplicate threads present.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44683", "right": "issue:46515", "accept": false, - "reason": "Both mention torch incompatibility, but they describe different version-specific failures and different code paths." + "reason": "Different failures: compiled flex_attention on torch 2.9 vs a torch 2.4.1 incompatibility report." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Different problems: one is an import error for a missing class, the other is a meta-device model loading issue." + "reason": "A missing import for a specific class and a meta-device loading regression are different code paths." }, { - "left": "issue:43366", - "right": "issue:44589", + "left": "issue:44560", + "right": "issue:45685", "accept": false, - "reason": "Unrelated: GGUF gpt-oss support versus a Float8 storage lookup error." - }, - { - "left": "issue:43645", - "right": "issue:43646", - "accept": true, - "reason": "Same underlying bug: Transformers v5 breaks custom model initialization; 43645 is just a Jupyter-specific manifestation." + "reason": "Different model families and different failure modes; too little evidence of one underlying bug." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Both are model-loading failures, but with different triggers and unrelated failure modes." + "reason": "Qwen3Moe FineGrainedFP8 loading and generic device='meta' loading breakage are not the same issue." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same underlying custom-model initialization regression; 43645 is the Jupyter-specific manifestation of 43646." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Both touch attention/SDPA, but the model, symptom, and root cause are different." + "reason": "Bamba torch.compile/SDPA failure and Qwen3.5 flash-attention packed-sequence crash are different bugs." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Both involve Qwen3.x, but one is a video StopIteration bug and the other is trainer-side embedding unpacking." + "reason": "Qwen3.5 video StopIteration and TRL embedding unpacking are different symptoms and likely different code paths." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "No shared underlying bug; a script execution failure is unrelated to GGUF architecture support." + "reason": "3D parallel execution failure is unrelated to GGUF gpt-oss architecture support." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and are not the same fix." + "reason": "Both are tokenizer regressions, but on different model families with different described failures." }, { "left": "issue:45823", "right": "issue:46515", "accept": false, - "reason": "Different issues: Gemma4 device-mismatch versus a generic torch version incompatibility." + "reason": "Gemma4 device mismatch and a torch 2.4.1 incompatibility are unrelated." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "Importing GenerationMixin and Qwen3Moe FP8 loading are unrelated." + "reason": "GenerationMixin import failure and Qwen3Moe FP8 loading are different problems." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "Different subsystems: tokenizer-mapping logic versus Gemma4 device placement." + "reason": "Tokenizer-mapping edge case and Gemma4 PLE device mismatch are not the same underlying bug." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Different failures: video StopIteration versus tensor mask shape mismatch." + "reason": "Different symptoms: Qwen3 video StopIteration versus a tensor/mask shape mismatch index error." }, { "left": "issue:38175", "right": "issue:43994", - "accept": false, - "reason": "Same model family, but the reports describe different symptoms and no clear shared root cause." + "accept": true, + "reason": "Same SigLIP2 bad-output regression; both describe near-zero/nonsensical predictions for the same model family." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "Unrelated regressions: the _is_hf_initialized flag versus a long-ocr index error." + "reason": "HF v5 initialization flag breakage and Kosmos2.5 long-OCR indexing are unrelated." }, { "left": "issue:44877", "right": "issue:45030", - "accept": false, - "reason": "Both are config-related, but they concern different models and different validation changes." + "accept": true, + "reason": "Both report loading failures caused by stricter config validation in v5." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "Deepspeed ZeRO-3 initialization and TOKENIZER_MAPPING_NAMES behavior are unrelated." + "reason": "DeepSpeed zero3 indexing failure and tokenizer-mapping logic are different initialization bugs." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Different bugs: ZeRO-3 with BertModel versus add-new-model-like logic." + "reason": "Zero3 empty-tensor indexing and add-new-model-like tokenizer mapping are unrelated." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "No overlap in the underlying failure mode." + "reason": "Tokenizer mapping edge case does not match the separate shape-mismatch indexing error." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Different causes: dequantized model offload failure versus mmap OOM avoidance." - }, - { - "left": "issue:43644", - "right": "issue:44534", - "accept": true, - "reason": "Exact same regression about non-persistent buffers being filled with junk; 44534 is the broader wording." + "reason": "Dequantized-model offload loading and Strix Halo mmap OOM are different load-path problems." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "The docs/example bug is not the same as the zero-probabilities inference bug." + "reason": "Related SigLIP2 area, but docs-example errors and bad model outputs are not the same concrete bug." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same non-persistent-buffer junk regression in Transformers v5." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Different model families and different loading/training paths." + "reason": "Qwen3 MoE LoRA compatibility and BitNet packed-weight loading are separate issues." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Both are Qwen-related, but one is an auto-device-map CUDA assert and the other is an FP8 config failure." + "reason": "Auto device-mapping CUDA assert and FineGrainedFP8Config failure do not share the same underlying cause." }, { "left": "issue:46459", "right": "issue:46512", "accept": false, - "reason": "Both concern dtype, but one is checkpoint loading and the other is from_config dtype propagation." + "reason": "Closely related dtype-handling themes, but one is about ignored load dtype and the other about stale config.dtype after from_config." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "No shared code-path problem." + "reason": "FP8 loading and Strix Halo mmap/OOM are unrelated." }, { - "left": "issue:45313", - "right": "issue:45357", + "left": "issue:44451", + "right": "issue:44488", "accept": false, - "reason": "Different regressions: ZeRO-3 weight loading versus incorrect visual-encoder key saving." + "reason": "Both are model-load failures, but the repo-specific evidence is too thin to treat them as one bug." }, { - "left": "issue:44451", - "right": "issue:44488", + "left": "issue:42491", + "right": "issue:45276", "accept": false, - "reason": "Similar loading complaints, but different models and no evidence of the same underlying bug." + "reason": "LoRA training compatibility and Gemma4 embedding-resize behavior are unrelated." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Completely different models and failure modes." + "reason": "GLM5 and Qwen3.5 flash-attention failures are different models and failure modes." }, { - "left": "issue:42491", - "right": "issue:45276", + "left": "issue:45313", + "right": "issue:45357", "accept": false, - "reason": "LoRA compatibility on qwen3_moe is unrelated to Gemma4 token-embedding resizing." + "reason": "ZeRO-3 weight loading and save_pretrained key serialization are different directions of failure." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "GraniteMoEHybrid invalid method calls are unrelated to the Nemotron example breakage." + "reason": "GraniteMoEHybrid invalid-method calls and the Nemotron example breakage are not the same bug." }, { - "left": "issue:43644", - "right": "issue:43645", + "left": "issue:44488", + "right": "issue:44991", "accept": false, - "reason": "Both are Transformers v5 regressions, but they are distinct bugs: buffers versus custom-model initialization." + "reason": "Both are tokenizer/load regressions, but they target different repos and lack evidence of a shared root cause." }, { - "left": "issue:44488", - "right": "issue:44991", + "left": "issue:43644", + "right": "issue:43645", "accept": false, - "reason": "Different tokenizer-loading failures on different models; not the same bug." + "reason": "These are related v5 regressions, but non-persistent buffer corruption and notebook custom-model init breakage are not the same concrete bug." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "The documentation example error and the model-output regression are separate issues." + "reason": "SigLIP2 docs-example problems and wrong model outputs are related at a high level, but not the same defect." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats the cluster as mostly unrelated, and the two accepted duplicate pairs (43644/44534 and 43645/43646) look supportable from the titles. The other soft-edge rejections are cautious and do not overclaim equivalence." + "accept": false, + "feedback": "Reject: the high-level summary overstates the evidence by calling the Qwen3.5 packed-sequence pair a true duplicate, but that pair was explicitly rejected in the soft verdicts. The rest is mostly conservative, and the 43645/43646 custom-model init linkage is grounded, but the summary is not fully supported by the packet." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2194,8 +2188,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -2230,7 +2224,6 @@ "issue:44910", "issue:44912", "issue:44960", - "issue:45216", "issue:45276", "issue:45292", "issue:45335", @@ -2249,11 +2242,11 @@ "issue:46468" ], "soft_pairs": [ - "issue:43708|issue:44075", "issue:43856|issue:43873", + "issue:43708|issue:44075", "issue:43854|issue:43931", - "issue:43257|issue:43854", "issue:43856|issue:44387", + "issue:43257|issue:43854", "issue:42491|issue:45292", "issue:44960|issue:46050", "issue:45488|issue:45812", @@ -2263,8 +2256,8 @@ "issue:44164|issue:44291", "issue:44910|issue:45588", "issue:43404|issue:43454", - "issue:41720|issue:44155", "issue:43425|issue:43957", + "issue:41720|issue:44155", "issue:44402|issue:44488", "issue:39692|issue:44960", "issue:42947|issue:45704", @@ -2273,260 +2266,256 @@ "issue:45663|issue:45759", "issue:44291|issue:44756", "issue:45663|issue:45859", + "issue:44387|issue:44756", "issue:44292|issue:44912", "issue:44155|issue:45823", - "issue:44387|issue:44756", "issue:43957|issue:44756", - "issue:43335|issue:45663", "issue:42947|issue:43381", + "issue:43335|issue:45663", "issue:45276|issue:45292", "issue:44863|issue:46468", "issue:45685|issue:46468", - "issue:43854|issue:45216", "issue:44164|issue:44756" ], - "prepared_review_unit_hash": "237dd767b48f6171ef65ad0924a4cdd35d4162aa8f12b63b613628e440776d93", + "prepared_review_unit_hash": "8f552334bfc14c7c490ab16da6d3760170a74f26ed847634510591eb2f039ca6", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 38268, - "estimated_input_tokens": 9567, - "estimated_eval_tokens": 19390 - }, + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37366, + "estimated_input_tokens": 9342, + "estimated_eval_tokens": 18940 + }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated Transformers issue cluster. The only clear duplicate-like pair in this packet is the Gemma4 resize_token_embeddings/output_embeddings report, where one issue is Gemma4-specific and the other is the broader phrasing of the same symptom.", - "confidence": 0.9, - "canonical_issue_reason": "Issue 45276 is the better canonical issue because its title is more specific: it names Gemma4 and includes both the embed_tokens_per_layer and output_embeddings behavior, while 45292 uses the broader output_embeddings phrasing.", + "summary": "This cluster is heterogeneous: most items are unrelated bugs across model loading, quantization/offloading, tokenizer regressions, checkpointing, and model-specific architecture issues. I found no strong duplicate chain; all soft edges are too weak to merge strictly.", + "confidence": 0.94, + "canonical_issue_reason": "No canonical issue: the cluster does not form a single duplicate family; the titles span many unrelated subsystems and model families.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45276 is the strongest issue candidate in this set because it is the most specific description of the resize_token_embeddings regression and covers the Gemma4-specific scope shown in the packet.", + "best_issue_reason": "No clear best representative issue for deduplication; none of the issues is a convincing target for the rest of the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43708", - "right": "issue:44075", + "left": "issue:43856", + "right": "issue:43873", "accept": false, - "reason": "Different bugs: checkpoint resume step calculation vs optimizer SGD args not being applied." + "reason": "Both mention memory/quantization, but one is Qwen3 MoE training memory use and the other is quantized offloading behavior; different bugs." }, { - "left": "issue:43856", - "right": "issue:43873", + "left": "issue:43708", + "right": "issue:44075", "accept": false, - "reason": "Both are memory-related, but they describe different model/code-path failures." + "reason": "Trainer resume/max_steps calculation and SGD arguments being ignored are unrelated code paths." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different model-loading failures for different models." + "reason": "Both are load failures, but for different models and different failure modes." }, { - "left": "issue:43257", - "right": "issue:43854", + "left": "issue:43856", + "right": "issue:44387", "accept": false, - "reason": "Different model families and different loading/conversion problems." + "reason": "Both involve quantization-related memory pressure, but the reported problems and contexts differ materially." }, { - "left": "issue:43856", - "right": "issue:44387", + "left": "issue:43257", + "right": "issue:43854", "accept": false, - "reason": "Both involve memory pressure, but not the same underlying bug." + "reason": "Qwen3 MoE weight conversion with accelerate/deepspeed is unrelated to GLM-4.7-Flash loading failures." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Different issues: Qwen3 MoE LoRA compatibility vs resize_token_embeddings/output_embeddings behavior." + "reason": "LoRA model compatibility on hf5.x is unrelated to resize_token_embeddings not updating output embeddings." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "Unrelated model/loading topics." + "reason": "GLM5 model support and sinq save/load corruption are separate issues." }, { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but the affected models/APIs differ." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different underlying mechanisms." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different model-specific loading/compatibility issues." + "reason": "A LoRA compatibility regression and a Qwen3-VL shape-mismatch load error are not the same bug." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Different frontends and model classes." + "reason": "Qwen3.5 chat crashing and Gemma4Processor missing _tokenizer are unrelated model/serving bugs." }, { "left": "issue:45276", "right": "issue:45335", "accept": false, - "reason": "Same API name, but the titles point to different model-specific resize bugs." + "reason": "Both are resize_token_embeddings issues, but they affect different model families and different embedding components." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Save/from_pretrained extra_state handling vs init_empty_weights argument handling are different problems." + "reason": "Extra_state save/load handling and init_empty_weights _is_hf_initialized are distinct model-loading internals." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Different flash-attention failures with different triggers." + "reason": "Both concern flash attention, but one is a Qwen3.5 position_ids crash and the other is an s_aux=None AttributeError." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both mention lm_head tying, but the reports are model-specific and not the same concrete bug." + "reason": "Both are weight-tying bugs, but they occur in different model classes and likely different code paths." }, { - "left": "issue:41720", - "right": "issue:44155", + "left": "issue:43425", + "right": "issue:43957", "accept": false, - "reason": "Unrelated issues: Qwen3 auto device mapping crash vs AudioFlamingo3 batched inference leak." + "reason": "Torch 2.10 incompatibility and meta-device loading regressions are unrelated failures." }, { - "left": "issue:43425", - "right": "issue:43957", + "left": "issue:41720", + "right": "issue:44155", "accept": false, - "reason": "Torch version incompatibility and meta-device loading failure are different problems." + "reason": "Qwen3 auto device mapping CUDA asserts and AudioFlamingo3 track leakage are unrelated." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Both are tokenizer-related load problems, but the packet does not show a shared root cause." + "reason": "Tokenizer vocab mismatch and inability to load a different model are not the same underlying bug." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Documentation/example issue versus GLM5 loading issue." + "reason": "SigLIP2 documentation errors and GLM5 model support are unrelated." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/LoRA is unrelated to T5 apex RMSNorm memory leak." + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to apex RMSNorm memory leak behavior." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Related theme, but offloading failure and int4 CUDA memory regression are different bugs." + "reason": "Both touch quantization/memory, but one is offloading behavior and the other is reserved-memory/OOM growth." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "ImportError for GenerationMixin and Whisper return_language pipeline regression are unrelated." + "reason": "An import-path regression for GenerationMixin and Whisper return_language pipeline behavior are different issues." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Different model families and different failure points." + "reason": "Gemma-4 FSDP2 shared_kv_states and Qwen multimodal config unwrapping are separate model-architecture problems." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "Init-empty-weights loading error vs mmap OOM are different code paths." + "reason": "init_empty_weights argument handling and mmap OOM on Strix Halo are unrelated." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Different bugs in Gemma4 training and Qwen3 MoE offload." + "reason": "Gemma-4 FSDP2 KeyError and Qwen3_5MoE CPU offload incompatibility are different code paths." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention OOM, but the causes are different: int4 reserved-memory growth vs mmap behavior on Strix Halo." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Both concern quantized loading, but they target different models/schemes." + "reason": "Both are quantization-load regressions, but with different formats, models, and failure modes." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "Different runtime problems: token leakage vs device mismatch." + "reason": "AudioFlamingo3 batched inference leakage and Gemma4 device-mismatch during forward are unrelated." }, { - "left": "issue:44387", + "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Both can cause memory issues, but the root causes differ." + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are separate problems." }, { - "left": "issue:43957", - "right": "issue:44756", + "left": "issue:42947", + "right": "issue:43381", "accept": false, - "reason": "Meta-device loading regression and mmap OOM are different code paths." + "reason": "Both mention gradient checkpointing, but one is ineffective under PEFT LoRA and the other is disallowed in eval mode." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "Different config bug versus Gemma4 FSDP2 attention failure." + "reason": "SwitchTransformers sparse-layer config bug and Gemma-4 FSDP2 attention state bug are unrelated." }, { "left": "issue:45276", "right": "issue:45292", - "accept": true, - "reason": "The titles indicate the same resize_token_embeddings regression, with 45276 being the Gemma4-specific version and 45292 the broader wording." + "accept": false, + "reason": "Closely related surface area, but Gemma4 versus t5gemma suggests separate resize paths rather than one concrete duplicate bug." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "Related by model family at most, but not the same bug." + "reason": "NemotronH checkpoint-loading failures and a broken official example are not clearly the same bug." }, { "left": "issue:45685", "right": "issue:46468", "accept": false, - "reason": "MPS MoE histogram error is unrelated to the Nemotron example failure." - }, - { - "left": "issue:43854", - "right": "issue:45216", - "accept": false, - "reason": "Different models and different loading/saving regressions." + "reason": "An MPS histogram op issue and a Nemotron example failure are unrelated." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "Save/from_pretrained extra_state handling is unrelated to Strix Halo mmap OOM." + "reason": "save/from_pretrained extra_state handling and mmap OOM are unrelated." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The overall clustering judgment is plausible, but the canonical-issue rationale overstates what is supported by the packet. From the provided metadata/titles we can infer that 45276 is Gemma4-specific and 45292 is the broader resize_token_embeddings/output_embeddings report, but we cannot verify claims like \"detailed reproduction\" or that it \"fully describes\" the failure. Please keep the justification limited to what is directly grounded in the packet." - }, - "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -2538,14 +2527,13 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ "issue:39404", "issue:42796", - "issue:43299", "issue:43381", "issue:43425", "issue:43611", @@ -2554,9 +2542,9 @@ "issue:43708", "issue:43716", "issue:43828", + "issue:43854", "issue:43901", "issue:43906", - "issue:43931", "issue:43957", "issue:43994", "issue:44079", @@ -2577,6 +2565,7 @@ "issue:44918", "issue:44928", "issue:45084", + "issue:45216", "issue:45292", "issue:45335", "issue:45588", @@ -2594,6 +2583,7 @@ "soft_pairs": [ "issue:42796|issue:46326", "issue:44479|issue:44877", + "issue:43854|issue:45216", "issue:43381|issue:44928", "issue:45684|issue:45698", "issue:45588|issue:45663", @@ -2613,237 +2603,236 @@ "issue:43425|issue:43716", "issue:44530|issue:45084", "issue:43906|issue:44530", - "issue:44265|issue:44898", "issue:44479|issue:44805", + "issue:44265|issue:44898", "issue:44589|issue:45084", "issue:39404|issue:44912", "issue:43901|issue:44509", "issue:43425|issue:44918", "issue:43906|issue:46326", "issue:46050|issue:46326", - "issue:43994|issue:44079", - "issue:43299|issue:43931" + "issue:43994|issue:44079" ], - "prepared_review_unit_hash": "a46f4aa2f503d1024b7b5f15e1727f512982ee01ca25de8f4c4c599402f77224", + "prepared_review_unit_hash": "0898fc6b154ced86ec1140da651ea4f714278b23d9a715725ebc01ab2b5ce87b", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36484, - "estimated_input_tokens": 9121, - "estimated_eval_tokens": 18498 + "serialized_chars": 36519, + "estimated_input_tokens": 9130, + "estimated_eval_tokens": 18516 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated Transformers regression reports grouped by broad semantic similarity. The only clear duplicate-like pair is the resize_token_embeddings/output-embedding bug; the Qwen3-VL MoE loading pair also looks like the same underlying model-loading failure.", - "confidence": 0.74, - "canonical_issue_reason": "issue:45698 is the most active and currently open report, with multiple inbound references and a broad, user-facing custom-model save/load failure that best represents the cluster\u2019s serialization-related regressions.", + "summary": "The set is mostly a collection of unrelated issue threads that only share broad Transformers subsystems. I only see one strong duplicate-like pair: the `resize_token_embeddings` regression affecting output embeddings, with one issue phrased generally and the other as a T5Gemma-specific manifestation.", + "confidence": 0.77, + "canonical_issue_reason": "Issue 45292 is the broadest formulation of the `resize_token_embeddings` bug; 45335 looks like the same underlying propagation problem reported on a specific model path (`decoder.embed_tokens`).", "canonical_pr_reason": null, - "best_issue_reason": "issue:45698 is the strongest representative issue here because it is active, well-scoped, and central to a core save/load code path with clear external impact.", + "best_issue_reason": "Issue 45292 is the best representative because it describes the core failure mode in a general way and can subsume the model-specific variant.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42796", "right": "issue:46326", "accept": false, - "reason": "Both involve tokenizers, but one is a BPE-marker handling bug and the other is continue_final_message behavior for mistral_common; different failure modes." + "reason": "Both are tokenizer-related, but one is about BPE marker handling and the other about `continue_final_message`; different code paths and symptoms." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Different subsystems: Qwen video-input regression vs. granite_speech config loading/strict config handling." + "reason": "Different models and failures: video input regression vs strict config loading for `granite_speech`." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different bug reports on different models: GLM-4.7-Flash unit-test loading vs Qwen3.5 save_pretrained regression." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "One is eval-mode gradient checkpointing, the other is Qwen3.5 RLHF numerical instability from 3D position_ids/SDPA fallback; not the same bug." + "reason": "Unrelated training-time problems: gradient checkpointing in eval mode vs Qwen3.5 RLHF NaN explosion from position_ids/SDPA behavior." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Both touch custom model save/load, but one is read-only file permission propagation and the other is wrong module resolution after save_pretrained." + "reason": "Both involve custom-model save/load, but one is permissions propagation and the other is wrong module resolution after save; not the same bug." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Different code paths: flash_attention sink-less crash vs. Gemma4 FSDP2 shared_kv_states KeyError." + "reason": "Different failures in different areas: flash-attention crash on `s_aux=None` vs Gemma-4 FSDP2 shared_kv_states KeyError." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization in notebooks is unrelated to Trainer checkpoint-resume max_steps calculation." + "reason": "Custom model initialization in notebooks is unrelated to `resume_from_checkpoint` max_steps calculation." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same underlying resize_token_embeddings bug: the resized weights are not propagated to the output/decoder embeddings, with t5gemma as a model-specific manifestation." + "reason": "Same underlying `resize_token_embeddings` propagation bug; the second issue is a model-specific instance of the same failure to update output embeddings." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Notebook custom-model initialization breakage is not the same as save/from_pretrained extra_state handling." + "reason": "Both mention custom models, but one is notebook initialization and the other is save/from_pretrained handling of `extra_state`." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Both are dtype/device-related, but one is autocast mismatch on Phi-tiny-MoE and the other is meta-device loading failures on different models." + "reason": "Different dtype/device issues on different models and code paths: autocast mismatch vs meta-device loading." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "Tokenizer vocab-size mismatch for pad/unk tokens is not the same as a general inability to load ScandiBERT." + "reason": "Both are model/tokenizer loading complaints, but the reported failures and models are different." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "Torch 2.10 incompatibility is broader/platform-level, while the other is a model-specific autocast dtype mismatch." + "reason": "Torch version incompatibility is not the same as a Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:46315", "right": "issue:46491", "accept": false, - "reason": "Performance regression in convert_tokens_to_ids is unrelated to CodeLlama leading-space round-trip behavior." + "reason": "Tokenizer regression in performance vs a round-trip spacing bug; different symptoms and likely different fixes." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Granite token IDs are wrong, while the other issue is about convert_tokens_to_ids slowness; different symptoms and fixes." + "reason": "Granite token IDs regression and slow `convert_tokens_to_ids` are separate tokenizer issues." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to Strix Halo mmap OOM behavior." + "reason": "Storage-object lookup failure is unrelated to Strix Halo mmap OOM avoidance." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 and Gemma4 PLE device-map mismatch are different code paths." + "reason": "Compiled flex_attention on torch>=2.9 and Gemma4 device-mismatch under `device_map='auto'` are distinct bugs." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language regression is unrelated to tokenization changes in transformers version." + "reason": "Whisper pipeline `return_language` regression is unrelated to general tokenization changes across versions." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "base_model_prefix loading regression is not the same as notebook-defined custom model initialization failure." + "reason": "Both are Transformers 5 migration issues, but one is `base_model_prefix` loading and the other is notebook initialization of custom models." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "Torch version incompatibility is not the same as Gemma4 device placement mismatch." + "reason": "Torch 2.10 incompatibility and Gemma4 forward device mismatch are different concrete problems." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "A general torch incompatibility is not the same as a Mistral-3 image-preprocessor dtype mismatch." + "reason": "Torch version compatibility and Mistral-3 image preprocessor/model dtype mismatch are not the same bug." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "PagedAttentionCache linear_attention crash and compile non-template nodes are different failures." + "reason": "PagedAttentionCache linear_attention crash and compile failure for non-template nodes are different failures." }, { "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": "An isolated reproduction of a prior issue is not the same as the Qwen3.5 PagedAttentionCache crash." }, { - "left": "issue:44265", - "right": "issue:44898", + "left": "issue:44479", + "right": "issue:44805", "accept": false, - "reason": "torch.export.export with torch_compilable_check is unrelated to Perceiver non-default resolution interpolation failure." + "reason": "Video-input regression and a mask-shape IndexError are different failure modes." }, { - "left": "issue:44479", - "right": "issue:44805", + "left": "issue:44265", + "right": "issue:44898", "accept": false, - "reason": "Video-input regression in Qwen models and an index/mask shape mismatch are not the same bug." + "reason": "`torch.export` failing with `torch_compilable_check` is unrelated to Perceiver image classification failing at non-default resolution." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Float8 storage lookup and non-template node compilation are unrelated." + "reason": "Float8 storage lookup and non-template-node compile failure are separate issues." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper pipeline return_language and MXFP4 quantization fallback are unrelated regressions." + "reason": "Whisper pipeline language return regression is unrelated to MXFP4 quantization fallback on git-oss-20b." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Outdated return_all_scores docs are separate from docs for removed pipeline tasks." + "reason": "Both are docs mismatches, but they concern different pipeline APIs and different removed/deprecated behaviors." }, { "left": "issue:43425", "right": "issue:44918", "accept": false, - "reason": "Torch 2.10 compatibility and Qwen3.5 embedding unpacking in TRL SFT trainer are unrelated." + "reason": "Torch compatibility and Qwen3.5 input-embedding unpacking in TRL SFT are different bugs." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "An isolated reproduction of a prior issue is not the same as the mistral_common continue_final_message tokenizer bug." + "reason": "Isolated reproduction of another issue is not the same as the mistral_common `continue_final_message` tokenizer bug." }, { "left": "issue:46050", "right": "issue:46326", "accept": false, - "reason": "Quantized-model save/load breakage is unrelated to continue_final_message tokenizer behavior." + "reason": "Quantized save/load corruption and tokenizer message-finalization behavior are unrelated." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "SigLIP2 nonsense outputs and ModelOutput key assignment with None are different failure modes." - }, - { - "left": "issue:43299", - "right": "issue:43931", - "accept": true, - "reason": "Both describe the same Qwen3-VL MoE loading regression: a broad dev0 breakage showing up as specific weight-shape mismatch on a Qwen3-VL-30B-A3B-Instruct checkpoint." + "reason": "Siglip2 wrong outputs via AutoModel/pipeline and `ModelOutput` key assignment are different code paths." } ] }, "evaluator_result": { "accept": true, - "feedback": "The clustering judgment is grounded in the packet. The two accepted pairs are conservative and well-supported by the issue titles and descriptions: resize_token_embeddings/output_embeddings, and the Qwen3-VL model-loading shape-mismatch reports. The rejected pairs generally involve clearly different subsystems or failure modes. Canonical issue selection is also defensible from the packet metadata (open, active, multiple inbound refs)." + "feedback": "The summary is grounded in the packet, and the soft-edge decisions are generally conservative. The only accepted pair, `issue:45292` / `issue:45335`, is well-supported by the titles as the same `resize_token_embeddings` propagation problem affecting output embeddings / `decoder.embed_tokens`. I don\u2019t see any soft-edge verdict that overstates duplicate evidence." }, "evaluator_used": true, "retried": false, @@ -2859,8 +2848,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -2873,6 +2862,7 @@ "issue:42175", "issue:42796", "issue:42915", + "issue:43299", "issue:43404", "issue:43421", "issue:43452", @@ -2886,7 +2876,7 @@ "issue:43646", "issue:43716", "issue:43856", - "issue:43883", + "issue:43931", "issue:44164", "issue:44451", "issue:44462", @@ -2914,9 +2904,10 @@ ], "soft_pairs": [ "issue:36032|issue:43452", + "issue:43299|issue:43931", "issue:44671|issue:44991", - "issue:44164|issue:45759", "issue:43596|issue:44683", + "issue:44164|issue:45759", "issue:44991|issue:45081", "issue:43856|issue:44756", "issue:44164|issue:44479", @@ -2926,239 +2917,238 @@ "issue:44488|issue:45081", "issue:44661|issue:44683", "issue:44488|issue:44987", - "issue:44737|issue:45859", "issue:32090|issue:39404", "issue:43611|issue:43644", - "issue:43582|issue:43632", + "issue:44737|issue:45859", "issue:46285|issue:46402", "issue:43716|issue:45084", + "issue:43582|issue:43632", "issue:36032|issue:42915", "issue:40990|issue:43421", "issue:43643|issue:44530", - "issue:34567|issue:40444", "issue:44451|issue:44987", + "issue:34567|issue:40444", "issue:42175|issue:46178", "issue:43541|issue:45084", - "issue:44811|issue:45584", "issue:42915|issue:44530", + "issue:44811|issue:45584", "issue:43644|issue:43646", - "issue:40444|issue:46326", - "issue:43883|issue:44488" + "issue:40444|issue:46326" ], - "prepared_review_unit_hash": "7f26d4145cb70e3fed33c9a57ccb81c6b94bb08f831c47185f41749e8ec6be83", + "prepared_review_unit_hash": "0d9bb9f820131d2b6c733a503af612a85eb33142b2c6ed6b2402708a80af24af", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36474, - "estimated_input_tokens": 9119, - "estimated_eval_tokens": 18494 + "serialized_chars": 37029, + "estimated_input_tokens": 9258, + "estimated_eval_tokens": 18772 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is loosely centered on Transformers v5/v5.x regression reports, especially model/tokenizer loading and initialization issues, but many pairs are still unrelated concrete bugs. I did not find any PRs in the cluster.", - "confidence": 0.43, - "canonical_issue_reason": "Issue 43611 is the broadest representative of the dominant theme: v5 breaking model loading/initialization behavior across models.", + "summary": "Loose cluster of unrelated closed issues; the soft links are mostly superficial topical overlaps (tokenizers, loading, Qwen, Whisper, v5 regressions) rather than the same bug/change.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out: the items span many unrelated subsystems and failure modes, and none of the soft pairs look like true duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43611 is the best single anchor for the cluster because it describes the most general and recurring failure mode seen across the set.", + "best_issue_reason": "No good global best issue for dedup triage here; the cluster is too heterogeneous to anchor on one representative bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Different subsystems and failures: T5 tokenizer method conflict vs gguf loading path breakage." + "reason": "Both involve tokenizer/model loading, but one is a T5Tokenizer method-name conflict and the other is gguf_file / AutoTokenizer-from-pretrained handling." }, { - "left": "issue:44671", - "right": "issue:44991", + "left": "issue:43299", + "right": "issue:43931", "accept": false, - "reason": "Both are tokenizer/model regressions, but they affect different models and different concrete code paths." + "reason": "Different Qwen model families and different failures: FineGrainedFP8Config loading vs Qwen3-VL weight-shape mismatch." }, { - "left": "issue:44164", - "right": "issue:45759", + "left": "issue:44671", + "right": "issue:44991", "accept": false, - "reason": "Unrelated: extra_state save/load handling vs composite Qwen config unwrapping." + "reason": "CamemBERT masked-LM prediction regression and EMBEDDIA tokenizer loading are unrelated bugs." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "Different problem domains: DeepSpeed zero3/BertModel init vs torch 2.9 flex_attention compilation." + "reason": "Deepspeed ZeRO-3/BertModel index error and compiled flex_attention on torch>=2.9 are separate code paths." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "Extra_state save/load handling is distinct from AutoModelForCausalLM.from_config failing to unwrap text_config." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Both touch tokenizers, but one is a model-specific tokenizer load failure and the other is a Mistral regex patch crash." + "reason": "Both touch tokenizer loading, but the causes differ: EMBEDDIA tokenizer load failure vs Mistral regex patch hitting a backend_tokenizer attribute error." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Different runtime environments and causes: Qwen3 MoE training memory usage vs Strix Halo mmap/OOM behavior." + "reason": "Qwen3 MoE memory inefficiency and disabling mmap on Strix Halo to avoid OOM are not the same bug." }, { "left": "issue:44164", "right": "issue:44479", "accept": false, - "reason": "Different bugs: checkpoint extra_state handling vs Whisper video-input regression." + "reason": "save/from_pretrained extra_state support and Qwen video-input regression are unrelated." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Tokenizer BPE marker handling is unrelated to Mistral3 lm_head weight tying." + "reason": "Tokenizer BPE-marker handling and Mistral3 lm_head weight tying are different issues." }, { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer-related, but one ignores tokenizer.json and the other is a DeepSeek-Coder tokenization output regression." + "reason": "One is AutoTokenizer ignoring tokenizer.json, the other is a DeepSeek-Coder v1 tokenization regression attributed to a prior fix gap." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper pipeline return_language regression is unrelated to CamemBERT masked-LM prediction changes." + "reason": "Whisper return_language pipeline behavior and CamemBERT masked-LM predictions do not share the same underlying change." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Both are model load issues, but they involve different models and different failure mechanisms." + "reason": "Loading cjvt/sleng-bert and Mistral regex patching fail in different places and for different reasons." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Model registration logic vs compiled flex_attention are unrelated code paths." + "reason": "Model-like registration in TOKENIZER_MAPPING_NAMES and compiled flex_attention on torch>=2.9 are unrelated." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are loading failures, but they concern different repositories and likely different root causes." - }, - { - "left": "issue:44737", - "right": "issue:45859", - "accept": false, - "reason": "Different performance/device issues: XLNet CPU arange placement vs Qwen3.5 MoE CPU offload incompatibility." + "reason": "Both are load failures, but they affect different models and mechanisms: cjvt/sleng-bert vs physical-intelligence/fast." }, { "left": "issue:32090", "right": "issue:39404", "accept": false, - "reason": "Trainer GPU broadcast TypeError is unrelated to Whisper pipeline return_language behavior." + "reason": "Trainer GPU broadcast NoneType and Whisper return_language pipeline failure are unrelated." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; not the same bug." + "reason": "base_model_prefix loading breakage and non-persistent buffers being filled with junk are separate v5 regressions." }, { - "left": "issue:43582", - "right": "issue:43632", + "left": "issue:44737", + "right": "issue:45859", "accept": false, - "reason": "AppleSilicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." + "reason": "XLNet CPU device placement and Qwen3_5Moe CPU offload incompatibility are different model-specific device bugs." }, { "left": "issue:46285", "right": "issue:46402", "accept": false, - "reason": "Different model-structure regressions: CLIP text-model structure mismatch vs Sam3 text encoder weights loading." + "reason": "CLIPTextModel structure inconsistency and Sam3 text encoder weights not loading are distinct load-path problems." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Mistral dtype mismatch in multimodal preprocessing is unrelated to template compilation errors." + "reason": "Mistral-3 dtype mismatch and a template compilation error are unrelated failures." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "AppleSilicon caching_allocator_warmup TypeError and _is_hf_initialized flag breakage are not the same issue." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "T5 tokenizer load conflict and Qwen3Moe FP8 failure are unrelated." + "reason": "T5Tokenizer add_special_tokens conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity anomaly on gpt-oss is unrelated to runtime post-processor updates in TokenizersBackend." + "reason": "High perplexity on gpt-oss-20b and TokenizersBackend not updating its post-processor are separate bugs." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "AutoConfig trust_remote_code field loss is unrelated to PagedAttentionCache group-type crashes." + "reason": "trust_remote_code AutoConfig missing fields and PagedAttentionCache linear_attention crashes are different subsystems." }, { - "left": "issue:34567", - "right": "issue:40444", + "left": "issue:44451", + "right": "issue:44987", "accept": false, - "reason": "TrainerState token counting and multimodal IterableDataset failures are unrelated." + "reason": "Different model loads fail for different reasons; there is no shared underlying defect." }, { - "left": "issue:44451", - "right": "issue:44987", + "left": "issue:34567", + "right": "issue:40444", "accept": false, - "reason": "Both are load failures, but they concern different models and no shared concrete bug is evident." + "reason": "TrainerState token counting and Qwen2.5-VL iterable multi-image finetuning failures are unrelated." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "Package backend omission and BACKENDS_MAPPING validation errors are different backend-selection issues." + "reason": "Missing tensorflow in a torch extra and an unknown tensorflow_text backend mapping error are distinct packaging/backend issues." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Mixtral torch-dynamo grouped_mm tracing failure is unrelated to non-template-node compilation errors." + "reason": "Mixtral grouped_mm tracing failure and non-template-node compilation error are unrelated." }, { - "left": "issue:44811", - "right": "issue:45584", + "left": "issue:42915", + "right": "issue:44530", "accept": false, - "reason": "Both involve Whisper special-token handling, but one is batch_decode skipping and the other is empty-transcription generation after align_special_tokens." + "reason": "Qwen3Moe FP8 config failure and Qwen3.5 paged-attention linear_attention crash are different model/path problems." }, { - "left": "issue:42915", - "right": "issue:44530", + "left": "issue:44811", + "right": "issue:45584", "accept": false, - "reason": "Qwen3Moe FP8 training and Qwen3.5 PagedAttentionCache crashes are different failures." + "reason": "Whisper batch_decode ignoring skip_special_tokens and empty-transcription failure after align_special_tokens are related only by model family, not the same bug." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Both are v5 regressions, but one concerns non-persistent buffers and the other custom model initialization; too broad to merge." + "reason": "Non-persistent buffer junk filling and custom model initialization breakage are separate v5 issues." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Multimodal fine-tuning with multiple images and Mistral-common final-message handling are unrelated." - }, - { - "left": "issue:43883", - "right": "issue:44488", - "accept": false, - "reason": "Molmo tied-weights attribute error is unrelated to tokenizer loading for sleng-bert." + "reason": "Qwen2.5-VL multi-image iterable training and mistral_common continue_final_message behavior are unrelated." } ] }, @@ -3177,8 +3167,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -3213,6 +3203,7 @@ "issue:44464", "issue:44466", "issue:44479", + "issue:44488", "issue:44534", "issue:44589", "issue:44610", @@ -3232,8 +3223,9 @@ ], "soft_pairs": [ "issue:44206|issue:44220", - "issue:43452|issue:43482", + "issue:43883|issue:44488", "issue:36032|issue:43645", + "issue:43452|issue:43482", "issue:43317|issue:43482", "issue:45084|issue:45507", "issue:44987|issue:44991", @@ -3242,264 +3234,270 @@ "issue:43262|issue:43377", "issue:43746|issue:46326", "issue:43645|issue:44534", - "issue:43883|issue:44991", "issue:43646|issue:44534", "issue:43404|issue:44466", "issue:44403|issue:45081", + "issue:43883|issue:44991", "issue:42915|issue:45084", "issue:43994|issue:44610", "issue:43377|issue:43819", - "issue:43322|issue:43540", "issue:28530|issue:34567", + "issue:43322|issue:43540", "issue:43089|issue:45987", - "issue:44451|issue:45081", "issue:44987|issue:45081", + "issue:44451|issue:45081", "issue:44265|issue:44610", "issue:42915|issue:44291", "issue:44220|issue:44683", - "issue:45216|issue:45357", "issue:36032|issue:43742", - "issue:44991|issue:46315", "issue:44062|issue:44987", + "issue:45216|issue:45357", + "issue:44991|issue:46315", "issue:42915|issue:44589", "issue:44479|issue:45325", "issue:36032|issue:41720" ], - "prepared_review_unit_hash": "f11cd03c0331b2e50a4492ae8723ef1df516087ab15faea527d37322672ac38a", + "prepared_review_unit_hash": "49b91f710add350202ffaf5279d57c2a47e5dde37732246d1963a97765dbdcb9", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 33, - "serialized_chars": 37183, - "estimated_input_tokens": 9296, - "estimated_eval_tokens": 18848 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 38040, + "estimated_input_tokens": 9510, + "estimated_eval_tokens": 19276 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous cluster. The only clearly supported duplicate subgroup is the Qwen3.5 `save_pretrained` regression pair (45216/45357); the other soft links are too broad or point to different root causes.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 45216 is the best canonical anchor: it reports the core Qwen3.5 `save_pretrained` regression, and 45357 is a later follow-up showing the same incorrect visual encoder keys persist.", + "summary": "The set is mostly heterogeneous with one clear duplicate pair around Qwen3.5 save_pretrained serialization; the other soft matches are mostly loose topical similarity, not the same bug.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45216 is the best canonical issue because it states the broader Qwen3.5 save_pretrained regression, while issue:45357 is a narrower follow-up about the same incorrect checkpoint serialization path.", "canonical_pr_reason": null, - "best_issue_reason": "45216 is the strongest representative issue in the cluster because it captures the main Qwen3.5 checkpoint-saving defect and has a closely matching follow-up in 45357.", + "best_issue_reason": "issue:45216 is the strongest global anchor in this set: it is the most general and representative report for the only true duplicate pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Both are audio-related, but they describe different failures: unsupported `center` handling vs fbank feature extraction." + "reason": "Both touch audio feature extraction, but one is a LasrFeatureExtractor center-arg regression and the other is a different fbank helper issue." }, { - "left": "issue:43452", - "right": "issue:43482", + "left": "issue:43883", + "right": "issue:44488", "accept": false, - "reason": "Both involve GGUF/model loading, but the titles point to different failures and not the same concrete bug." + "reason": "Both are loading failures, but they involve different models and different failure modes." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Tokenizer constructor conflict and custom model initialization in notebooks are unrelated regressions." + "reason": "T5 tokenizer method-name conflict is unrelated to custom model initialization breaking in notebooks." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "GGUF file handling problems are involved in both, but the concrete failures and models differ." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "Dequantized offload loading failure and Qwen2.5-GGUF loading failure are different code paths and error modes." + "reason": "Different code paths: device_map/offload dequantized model loading versus GGUF loading under v5." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Template compilation error and GraniteMoEHybrid invalid method call are unrelated bugs." + "reason": "Unrelated failures: template compilation error versus GraniteMoE hybrid invalid method call." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are loading regressions, but for different models and likely different root causes." + "reason": "Both are load failures in v5, but they are distinct model/tokenizer regressions." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Qwen3.5 hidden-states output bug and Qwen2.5-VL temporal position-id scaling bug are different defects." + "reason": "Qwen3.5 output_hidden_states bug is not the same as the Qwen2.5-VL rope-index scaling regression." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Chunked generation inconsistency and `get_seq_length()` returning the wrong type are not the same bug." + "reason": "Chunked generation inconsistency and StaticCache seq_length type mismatch are separate cache/generation bugs." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Sampling-rate defaulting and missing padding-mask support are distinct audio issues." + "reason": "Audio chat-template sampling-rate defaulting and MIMI padding-mask behavior are different bugs." }, { "left": "issue:43746", "right": "issue:46326", "accept": false, - "reason": "PEFT local-checkpoint loading and `continue_final_message` behavior are unrelated." + "reason": "Local PEFT checkpoint loading and mistral_common continue_final_message are unrelated tokenizer/model-loading paths." }, { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization in notebooks is unrelated to non-persistent buffer serialization junk." - }, - { - "left": "issue:43883", - "right": "issue:44991", - "accept": false, - "reason": "Missing tied-weights attribute and tokenizer loading failure are different problems." + "reason": "Custom model init in notebooks is unrelated to non-persistent buffers being serialized with junk." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization regression and buffer serialization corruption are unrelated." + "reason": "General custom model initialization breakage does not match the buffer serialization bug." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "Both mention tied weights, but one is a model-tying bug and the other is device-dependent serialization." + "reason": "Both mention tied lm_head weights, but one is about model tying and the other about save_pretrained serialization behavior." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Generic loading noise is not the same as the Mistral regex crash." + "reason": "Noise when loading a transformer is not the same as the Mistral regex crash." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Different model-loading problems with different root causes." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "FP8 model-loading failure and template compilation error are unrelated." + "reason": "FineGrainedFP8Config failure is unrelated to the non-template-node compile error." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Bad vision-model outputs and processor size mismatch are different regressions." + "reason": "Nonsensical SigLIP outputs and OmDet-Turbo processor input-size mismatch are different bugs." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "Missing padding-mask support and missing STE affect different audio mechanisms." + "reason": "MIMI batching/padding issues are unrelated to DAC.from_latents missing STE behavior." }, { - "left": "issue:43322", - "right": "issue:43540", + "left": "issue:28530", + "right": "issue:34567", "accept": false, - "reason": "Llava Next segfault and Qwen3OmniMoe video ValueError are different failures." + "reason": "Early-stopping metric selection and num_input_tokens_seen updating are unrelated Trainer issues." }, { - "left": "issue:28530", - "right": "issue:34567", + "left": "issue:43322", + "right": "issue:43540", "accept": false, - "reason": "Early-stopping metric handling and `num_input_tokens_seen` updates are unrelated trainer issues." + "reason": "Segfault when loading Llava Next and video ValueError in Qwen3OmniMoe are different model-specific failures." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation performance overhead is not the same defect as the cache sequence-length contract violation." + "reason": "Generation sync overhead is a performance issue; StaticCache.get_seq_length is a contract/type bug." }, { - "left": "issue:44451", + "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT loading failure and Mistral regex patch crash are different tokenizer/model issues." + "reason": "Model loading failure and Mistral regex patch crash are unrelated." }, { - "left": "issue:44987", + "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "Tokenizer/model loading regression and tokenizer regex crash are different bugs." + "reason": "ScandiBERT loading failure is not the same as the Mistral tokenizer regex crash." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "torch.export failure with `torch_compilable_check` is unrelated to OmDet-Turbo preprocessing size." + "reason": "torch.export failure with torch_compilable_check is unrelated to the OmDet-Turbo processor size regression." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "FP8 config loading and `_is_hf_initialized` argument errors are different initialization failures." + "reason": "FP8 model loading and init_empty_weights argument handling are separate issues." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio fbank extraction and compiled flex_attention on torch>=2.9 are unrelated." - }, - { - "left": "issue:45216", - "right": "issue:45357", - "accept": true, - "reason": "Both describe the same Qwen3.5 `save_pretrained` regression producing incorrect checkpoint contents; 45357 is a direct follow-up." + "reason": "fbank feature extraction and compiled flex_attention on torch>=2.9 are different subsystems and bugs." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "Tokenizer constructor conflict and MobileLLM key error are unrelated." + "reason": "Tokenizer special-token conflict is unrelated to the MobileLLM loading key error." }, { - "left": "issue:44991", - "right": "issue:46315", + "left": "issue:44062", + "right": "issue:44987", "accept": false, - "reason": "Tokenizer load failure and tokenizer conversion slowdown are different bugs." + "reason": "AddedToken keyword conflict and a model-loading regression are not the same bug." }, { - "left": "issue:44062", - "right": "issue:44987", + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same Qwen3.5 save_pretrained regression; 45357 is a narrower symptom about incorrect visual encoder keys." + }, + { + "left": "issue:44991", + "right": "issue:46315", "accept": false, - "reason": "AddedToken constructor conflict is not the same as a model loading failure." + "reason": "Tokenizer loading failure and convert_tokens_to_ids slowdown are different problems." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "Both involve float8/fp8, but they are different code paths and failures." + "reason": "FP8 loading failure and missing Float8 storage lookup are not the same code-path bug." }, { "left": "issue:44479", "right": "issue:45325", "accept": false, - "reason": "Video-input regression across several Qwen models and still-image temporal position-id scaling are not the same root bug." + "reason": "Both are Qwen video/rope regressions, but they describe different concrete failures and should not be merged as one bug." }, { "left": "issue:44534", - "right": "issue:44683", + "right": "issue:45081", "accept": false, - "reason": "Non-persistent buffer junk and compiled flex_attention failure are unrelated regressions." + "reason": "Non-persistent buffer junk and Mistral regex crash are unrelated." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The Qwen3.5 save_pretrained pair is well grounded, but the other accepted soft-edge links are too strong for the packet evidence. `43452` vs `43482` are both GGUF/loading issues, yet the titles do not support the same root bug. `44479` vs `45325` are both Qwen/VL-related, but one is a broad video-input regression and the other is a still-image temporal position-id scaling bug; that is not enough to call them the same defect. The summary also overstates the cluster structure by describing these as clear duplicate subgroups. Conservative links would be to reject these softer matches and keep only the clearly supported Qwen3.5 pair." + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted soft pair is well supported by the Qwen3.5 `save_pretrained` titles, and the remaining soft matches are appropriately rejected as broader topical similarity rather than the same bug." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -3512,8 +3510,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -3569,250 +3567,250 @@ "soft_pairs": [ "issue:43531|issue:44877", "issue:43632|issue:44756", + "issue:43296|issue:43531", + "issue:43425|issue:44403", "issue:43611|issue:44534", "issue:40444|issue:45325", "issue:42796|issue:43746", - "issue:43425|issue:44403", - "issue:43296|issue:43531", "issue:45812|issue:46491", "issue:43792|issue:44589", "issue:44403|issue:44991", - "issue:43632|issue:44534", "issue:33666|issue:40444", - "issue:40990|issue:44811", + "issue:43632|issue:44534", "issue:43381|issue:45923", + "issue:40990|issue:44811", "issue:44079|issue:45663", "issue:44291|issue:44589", "issue:43421|issue:44568", - "issue:41553|issue:43582", - "issue:35707|issue:40444", "issue:43482|issue:45538", + "issue:35707|issue:40444", "issue:43482|issue:43525", "issue:43317|issue:43856", "issue:42175|issue:45542", - "issue:44829|issue:45479", + "issue:41553|issue:43582", "issue:43818|issue:46402", - "issue:43818|issue:45357", "issue:44079|issue:44534", "issue:44462|issue:45812", "issue:43531|issue:44291", + "issue:43818|issue:45357", + "issue:44829|issue:45479", "issue:43377|issue:45491", "issue:41720|issue:45081" ], - "prepared_review_unit_hash": "c4d421295e85b599a102ccbca40a5c53cb2646799151664a6c97f73e030f6c2a", + "prepared_review_unit_hash": "f2094fe4cf51ec3ea651727d46e47a2394b83f99d052f602047b0640c07364e8", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37140, - "estimated_input_tokens": 9285, - "estimated_eval_tokens": 18826 + "serialized_chars": 37145, + "estimated_input_tokens": 9287, + "estimated_eval_tokens": 18830 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose bag of unrelated Transformers issues, with only a mild tokenizer/v5 regression theme. I do not see any true duplicate PRs, and none of the soft issue pairs look like the same underlying bug or change.", - "confidence": 0.29, - "canonical_issue_reason": "Issue 45812 is the broadest, most central tokenizer regression in the set: it describes a silent v4\u2192v5 break affecting all Granite models and has inbound references, making it the best single umbrella issue among otherwise disparate reports.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers/model-loading/tokenizer issues. There is a loose theme around v5 regressions and multimodal/model-loading bugs, but none of the soft pairs look like true duplicates or mergeable PR-equivalents.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43611 is the broadest, most representative loading-regression report in the set: it targets a core Transformers v5 model-loading path and is thematically closest to several other failures here.", "canonical_pr_reason": null, - "best_issue_reason": "45812 is the strongest standalone issue to anchor the cluster because it is user-impactful, cross-cutting, and closer to a root-cause tokenizer regression than the more model-specific reports.", + "best_issue_reason": "issue:43611 is the best single issue anchor because it is general, user-facing, and central to the cluster's repeated v5 loading regression theme.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Different problems: Qwen3-MoE sliding-window behavior vs GraniteSpeech config strictness." + "reason": "Different models and different failure modes; same broad area only." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "Unrelated failures: _is_hf_initialized regression vs mmap/OOM on Strix Halo." + "reason": "One is an HF init flag regression, the other is an AppleSilicon mmap/OOM workaround." }, { - "left": "issue:43611", - "right": "issue:44534", + "left": "issue:43296", + "right": "issue:43531", "accept": false, - "reason": "Different code paths: base_model_prefix loading vs non-persistent buffer initialization." + "reason": "Both are Qwen-related, but the bugs are unrelated training/loading failures." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:43425", + "right": "issue:44403", "accept": false, - "reason": "Same model family, but distinct bugs: iterable-dataset multi-image training vs rope index scaling." + "reason": "Torch version incompatibility is not the same as noisy loading output." }, { - "left": "issue:42796", - "right": "issue:43746", + "left": "issue:43611", + "right": "issue:44534", "accept": false, - "reason": "Different subsystems: tokenizer BPE markers vs PEFT/local-checkpoint loading." + "reason": "Both are v5 regressions, but they affect different loading internals and are not the same bug." }, { - "left": "issue:43425", - "right": "issue:44403", + "left": "issue:40444", + "right": "issue:45325", "accept": false, - "reason": "Torch compatibility issue vs noisy loading output; not the same bug." + "reason": "Different Qwen2.5-VL issues: one is dataset/multi-image finetuning, the other is rope-index scaling." }, { - "left": "issue:43296", - "right": "issue:43531", + "left": "issue:42796", + "right": "issue:43746", "accept": false, - "reason": "Different model/load failures: PaddleOCR-VL loading in vLLM vs Qwen3-MoE sliding window." + "reason": "Tokenizer BPE-marker handling is unrelated to PEFT checkpoint loading." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different behavior (Granite IDs vs CodeLlama round-trip spacing)." + "reason": "Both tokenizer bugs, but different models and different round-trip behavior." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Distinct runtime failures: Whisper execution vs missing Float8 storage object." + "reason": "Whisper runtime failure is unrelated to a Float8 storage lookup error." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Different issues: extra load-time noise vs tokenizer loading failure for a specific model." - }, - { - "left": "issue:43632", - "right": "issue:44534", - "accept": false, - "reason": "Different regressions with no shared failing code path." + "reason": "Generic loading noise is not the same as a specific tokenizer load failure." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Different training scenarios and failure modes; not the same bug." + "reason": "Both are VL training-related, but they are different underlying failures." }, { - "left": "issue:40990", - "right": "issue:44811", + "left": "issue:43632", + "right": "issue:44534", "accept": false, - "reason": "Unrelated issues: perplexity anomaly vs Whisper batch_decode behavior." + "reason": "Different v5 regressions affecting different code paths." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Gradient checkpointing eval-mode error is different from missing supports_gradient_checkpointing flag." + "reason": "Eval-mode checkpointing restriction is not the same as a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity regression and batch_decode special-token handling are unrelated." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key handling vs Gemma-4 FSDP2 attention KeyError are unrelated." + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 attention state are unrelated." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "Both are v5-related type errors, but they are clearly different bugs in different loading paths." + "reason": "Unexpected init argument on load is unrelated to missing Float8 storage support." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Both involve special tokens, but one is runtime post-processor syncing and the other is add_special_tokens behavior for a specific tokenizer." + "reason": "Runtime post-processor updates and a specific add_special_tokens regression are different bugs." }, { - "left": "issue:41553", - "right": "issue:43582", + "left": "issue:43482", + "right": "issue:45538", "accept": false, - "reason": "Bad tokenizer error message vs AppleSilicon allocator TypeError; unrelated." + "reason": "Qwen2.5-GGUF loading and CLIPTokenizer max_length are unrelated." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Different generation/training bugs with no shared code path." - }, - { - "left": "issue:43482", - "right": "issue:45538", - "accept": false, - "reason": "GGUF loading regression vs CLIPTokenizer max-length bug are unrelated." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multi-image finetuning." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "Different model/config failures; no evidence of the same underlying bug." + "reason": "Both are load failures, but different models and different root causes." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Different issues: dequantized model offload loading vs Qwen3 MoE training memory use." + "reason": "Device-map offload failure and MoE training memory inefficiency are different problems." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Both are backend/package-related, but they are not the same installation bug." + "reason": "Both involve TensorFlow packaging, but one is missing backend inclusion and the other is an undefined backend error." }, { - "left": "issue:44829", - "right": "issue:45479", + "left": "issue:41553", + "right": "issue:43582", "accept": false, - "reason": "Both mention degenerate training, but one is FA3 sequence-classification behavior and the other is a loss formulation bug." + "reason": "Tokenizer loading error message and AppleSilicon warmup TypeError are unrelated." }, { "left": "issue:43818", "right": "issue:46402", "accept": false, - "reason": "Different multimodal model loading defects with different symptoms and code paths." - }, - { - "left": "issue:43818", - "right": "issue:45357", - "accept": false, - "reason": "Both touch multimodal weights, but one is missing temporal attention and the other is wrong visual-encoder save keys." + "reason": "Different multimodal model loading bugs with different affected components." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "Distinct internal regressions: ModelOutput key assignment vs buffer initialization." + "reason": "ModelOutput key handling is unrelated to non-persistent buffer initialization." }, { "left": "issue:44462", "right": "issue:45812", "accept": false, - "reason": "Related to tokenizer loading, but one is ignored tokenizer.json and the other is wrong Granite token IDs." + "reason": "Both are AutoTokenizer regressions, but not the same concrete tokenizer bug." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Different problems: sliding-window behavior vs unexpected _is_hf_initialized argument." + "reason": "Qwen3-MoE sliding-window issue and init_empty_weights argument error are unrelated." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Both are vision-language model issues, but they touch different loading/save paths and different root causes." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Different training pathologies: flash-attention-3 degeneracy vs single-label zero-loss behavior." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Different batching/padding bugs in different models and code paths." + "reason": "Both are batching/padding-related, but the models and failure mechanisms differ." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Different Qwen/model-loading failures; one is cudaErrorAssert during device mapping, the other is a Mistral tokenizer patch crash." + "reason": "Auto device mapping CUDA assert and Mistral regex patch crash are unrelated." } ] }, @@ -3831,8 +3829,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -3888,18 +3886,18 @@ "soft_pairs": [ "issue:43644|issue:44079", "issue:40444|issue:45381", - "issue:44265|issue:45072", "issue:43611|issue:43646", + "issue:44265|issue:45072", "issue:43749|issue:44079", "issue:45325|issue:45381", "issue:42175|issue:43296", - "issue:42796|issue:45381", "issue:43994|issue:44898", "issue:43792|issue:43957", - "issue:45325|issue:45834", + "issue:42796|issue:45381", "issue:43454|issue:44466", - "issue:43632|issue:45072", + "issue:45325|issue:45834", "issue:42915|issue:43482", + "issue:43632|issue:45072", "issue:43701|issue:45325", "issue:44291|issue:44877", "issue:44479|issue:45823", @@ -3908,99 +3906,93 @@ "issue:43606|issue:44265", "issue:43262|issue:43746", "issue:40990|issue:45584", - "issue:42796|issue:46209", - "issue:43881|issue:44488", "issue:34634|issue:35707", + "issue:42796|issue:46209", "issue:35707|issue:41720", - "issue:43454|issue:43550", + "issue:43881|issue:44488", "issue:44186|issue:44265", "issue:39692|issue:40444", + "issue:43454|issue:43550", "issue:43606|issue:43819", "issue:41720|issue:45823", "issue:46305|issue:46315" ], - "prepared_review_unit_hash": "9c16d55ced78fe817e99c3a4ec49852d5fe380c1681383b2ce01588fa722772f", + "prepared_review_unit_hash": "3ca0f4e2861c7c5455f72ef1c394bc2549fb6361d568067035843c359e4fb2de", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37187, - "estimated_input_tokens": 9297, - "estimated_eval_tokens": 18850 + "serialized_chars": 37195, + "estimated_input_tokens": 9299, + "estimated_eval_tokens": 18854 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a loose theme cluster around many unrelated Transformers regressions and model-specific bugs, not a tight duplicate set. I found no convincing duplicate pairs, so all soft edges are rejected. If forced to pick one issue as the cluster representative, 43611 is the broadest loading-regression report, but it does not actually cover the rest of the cluster.", - "confidence": 0.19, - "canonical_issue_reason": "issue:43611 is the broadest and most general report in the set (Transformers 5.0.0 breaking model loading via base_model_prefix), so it is the best single representative if one must be chosen.", + "summary": "Mostly a heterogeneous cluster of unrelated bug reports. A few issues are superficially close within the Transformers v5 loading regressions and Qwen2.5-VL position-id bugs, but none are close enough to treat as the same underlying fix target. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43611 is the broadest representative of the Transformers 5.0 loading-regression subgroup; several nearby reports mention model-loading breakage, but this one is the clearest umbrella issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43611 best fits as a representative issue because it describes a generic v5 loading regression rather than a narrow model-specific symptom, but the surrounding items are too diverse to treat it as a true umbrella duplicate.", + "best_issue_reason": "issue:43611 is the best single issue to anchor the cluster because it describes a central v5 model-loading failure rather than a model-specific edge case.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Different failure modes: buffer serialization/junk vs ModelOutput key assignment." + "reason": "Non-persistent buffer serialization junk is a different bug from ModelOutput key assignment." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen2.5-VL regressions, but one is multi-image finetuning and the other is video position ids; not the same concrete bug." + "reason": "Qwen2.5-VL multi-image finetuning failure is unrelated to video position-id regression." }, { - "left": "issue:44265", - "right": "issue:45072", + "left": "issue:43611", + "right": "issue:43646", "accept": false, - "reason": "Unrelated code paths: torch.export with torch_compilable_check vs bfloat16 dtype mismatches." + "reason": "Both are v5 loading regressions, but one is about base_model_prefix and the other custom initialization; not the same code-path bug." }, { - "left": "issue:43611", - "right": "issue:43646", + "left": "issue:44265", + "right": "issue:45072", "accept": false, - "reason": "Both are v5 loading/init regressions, but the concrete breakage differs: base_model_prefix loading vs custom model initialization." + "reason": "torch.export/torch_compilable_check failure is unrelated to bfloat16 dtype mismatch." }, { "left": "issue:43749", "right": "issue:44079", "accept": false, - "reason": "FSDP CPU-efficient loading vs ModelOutput key handling are unrelated." + "reason": "FSDP CPU RAM efficient loading is not the same as ModelOutput key handling." }, { "left": "issue:45325", "right": "issue:45381", "accept": false, - "reason": "Related area, but still-image temporal scaling and video vision_position_ids are distinct symptoms and likely distinct fixes." + "reason": "Both touch Qwen2.5-VL position ids, but one is still-image temporal scaling and the other is video vision_position_ids; too different to merge." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "TensorFlow packaging/backend issue vs PaddleOCR-VL load failure in vLLM." - }, - { - "left": "issue:42796", - "right": "issue:45381", - "accept": false, - "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL position-id regressions." + "reason": "TensorFlow backend packaging is unrelated to PaddleOCR-VL model loading." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Different models and bugs: SigLIP2 bad outputs vs Perceiver resolution/interpolation failure." + "reason": "SigLIP2 nonsensical output and Perceiver resolution failure are different model-specific bugs." }, { "left": "issue:43792", @@ -4009,136 +4001,142 @@ "reason": "Whisper runtime failure is not the same as meta-device loading breakage." }, { - "left": "issue:45325", - "right": "issue:45834", + "left": "issue:42796", + "right": "issue:45381", "accept": false, - "reason": "Both are vision/model-input issues, but one is Qwen2.5-VL position ids and the other is Kosmos2.5 OCR indexing." + "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL video position ids." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Both mention lm_head/tied weights, but one is a model-specific tying bug and the other is device-dependent serialization inconsistency." + "reason": "Tied lm_head weights not being set is different from inconsistent serialization behavior." }, { - "left": "issue:43632", - "right": "issue:45072", + "left": "issue:45325", + "right": "issue:45834", "accept": false, - "reason": "_is_hf_initialized breakage is unrelated to dtype mismatch during inference." + "reason": "Qwen2.5-VL rope-index scaling and Kosmos2.5 OCR index errors are different bugs." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Different models and failures: FP8 config handling vs GGUF loading in v5." + "reason": "Qwen3Moe FP8 config failure is unrelated to Qwen2.5-GGUF loading in v5." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "_is_hf_initialized flag breakage is unrelated to torch.export dtype mismatches." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to Qwen2.5-VL position ids." + "reason": "Checkpoint resume key mismatch has no relation to Qwen2.5-VL rope indexing." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "Unexpected init_empty_weights argument vs strict config loading are different load-time problems." + "reason": "init_empty_weights argument mismatch and strict config loading are separate issues." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Different models and failure surfaces: Qwen VL video regression vs Gemma4 device-map mismatch." + "reason": "Qwen VL video-input regression and Gemma4 PLE device mismatch are different failures." }, { "left": "issue:43296", "right": "issue:45800", "accept": false, - "reason": "PaddleOCR-VL load failure and torch/transformers compatibility are not the same bug." + "reason": "PaddleOCR-VL loading failure is unrelated to torch/transformers version compatibility." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Quantization/offload behavior and tokenization version changes are unrelated." + "reason": "Quantized offloading malfunction and tokenization slowdown are different subsystems and bugs." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU offload device mismatch vs torch.export/torch_compilable_check are different issues." + "reason": "CPU offload device mismatch is unrelated to torch.export on compilable checks." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sample-rate defaulting is unrelated to PEFT checkpoint loading." + "reason": "Audio chat-template sampling-rate default and PEFT checkpoint loading are unrelated." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Perplexity anomaly and Whisper empty-transcription failure are unrelated." - }, - { - "left": "issue:42796", - "right": "issue:46209", - "accept": false, - "reason": "Tokenizer marker handling vs FP8 inference on B200 are different problems." + "reason": "High perplexity on gpt-oss-20b is not the same as Whisper empty-transcription generation failure." }, { - "left": "issue:43881", - "right": "issue:44488", + "left": "issue:34634", + "right": "issue:35707", "accept": false, - "reason": "Different models and symptoms: glm-4v loading failure vs sleng-bert loading failure." + "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are different features and failures." }, { - "left": "issue:34634", - "right": "issue:35707", + "left": "issue:42796", + "right": "issue:46209", "accept": false, - "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are unrelated." + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are unrelated." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation bug vs Qwen3 auto device-map cuda assert are unrelated." + "reason": "Progressive generation inputs_embeds/past_key_values is unrelated to Qwen3 auto-device-map cuda asserts." }, { - "left": "issue:43454", - "right": "issue:43550", + "left": "issue:43881", + "right": "issue:44488", "accept": false, - "reason": "lm_head tying bug and Bamba SDPA/torch.compile failure are different code paths." + "reason": "glm-4v-9b loading failure and sleng-bert loading failure are different model-specific issues." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer padding/truncation crash and torch.export failure are unrelated." + "reason": "LayoutLMv2Tokenizer crashes are unrelated to torch.export/torch_compilable_check." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "SigLIP2 docs/example problems and Qwen2.5-VL multi-image finetuning failure are different bugs." + "reason": "SigLIP2 doc-example errors and Qwen2.5-VL iterable-dataset multi-image finetuning failure are different problems." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "lm_head weight tying bug is unrelated to Bamba torch.compile SDPA failure." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "CPU offload device mismatch is unrelated to DAC.from_latents/STE mismatch." + "reason": "Bark CPU-offload device mismatch is not the same as DAC.from_latents / missing STE mismatch." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Different model families and failure modes despite both involving device maps." + "reason": "Qwen3 auto device mapping cudaErrorAssert and Gemma4 PLE forward mismatch are different device issues." }, { "left": "issue:46305", "right": "issue:46315", "accept": false, - "reason": "Streaming mask off-by-one and tokenizer performance regression are unrelated." + "reason": "Sliding-window off-by-one is unrelated to convert_tokens_to_ids performance regression." } ] }, @@ -4157,8 +4155,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -4211,8 +4209,8 @@ "issue:46402" ], "soft_pairs": [ - "issue:42796|issue:44977", "issue:43377|issue:44610", + "issue:42796|issue:44977", "issue:44568|issue:45920", "issue:43454|issue:45072", "issue:43257|issue:45923", @@ -4222,241 +4220,241 @@ "issue:40990|issue:42947", "issue:43454|issue:43720", "issue:43572|issue:45759", - "issue:41720|issue:43322", "issue:43749|issue:44534", - "issue:43526|issue:43883", + "issue:41720|issue:43322", "issue:42617|issue:42915", "issue:44466|issue:45357", - "issue:43881|issue:44991", + "issue:43526|issue:43883", "issue:41720|issue:44977", "issue:26413|issue:28530", + "issue:43881|issue:44991", "issue:44387|issue:45704", "issue:36331|issue:42617", "issue:40990|issue:45701", - "issue:43450|issue:45701", "issue:36331|issue:42915", "issue:44877|issue:44991", + "issue:43450|issue:45701", "issue:44186|issue:45072", "issue:43065|issue:46402", "issue:43873|issue:44756", + "issue:43550|issue:43720", "issue:43577|issue:43708", - "issue:43404|issue:45357", - "issue:43550|issue:43720" + "issue:43404|issue:45357" ], - "prepared_review_unit_hash": "523e00551cf52a4d1db5e9869012fbd27aea261665c55ff8d875b956322e27a2", + "prepared_review_unit_hash": "0115a2b4f30680f330e1e4ec56267e40bb24cbd2c0759dcf40cecc5fe013e4b6", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36502, - "estimated_input_tokens": 9126, - "estimated_eval_tokens": 18508 + "serialized_chars": 36499, + "estimated_input_tokens": 9125, + "estimated_eval_tokens": 18506 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes tokenizer regressions, loading/offload problems, multimodal weight-tying bugs, and training-loop edge cases. None of the soft pairs appear to be the same concrete bug/change, so I would not merge any of them as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "issue:45920 is the broadest concrete issue in the set: it is an open, cross-model tokenizer regression with inbound references, so it is the best umbrella if a single issue must be chosen.", + "summary": "The cluster is loosely centered on Transformers v5-era regressions (tokenization, loading, offloading, trainer behavior), but the soft links are mostly subsystem-level similarity rather than true duplicates. I would not merge any of the listed pairs as the same underlying bug.", + "confidence": 0.58, + "canonical_issue_reason": "issue:45920 is the broadest and most representative tokenizer-regression report in the set; it spans multiple model families and matches the cluster's strongest recurring theme. However, the cluster overall is still quite heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the strongest representative issue because it describes a real shared regression pattern rather than a model-specific symptom, and it has the clearest chance of aggregating related tokenizer reports.", + "best_issue_reason": "issue:45920 is the best single representative issue because it is multi-model, v5-specific, and broad enough to anchor the tokenizer-regression subset better than the narrower tickets.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42796", - "right": "issue:44977", + "left": "issue:43377", + "right": "issue:44610", "accept": false, - "reason": "Both are tokenizer/model-behavior bugs, but one is a BPE-marker handling issue and the other is a flash-attention generation issue; different code paths." + "reason": "Different bugs: batched-vs-single encoder output mismatch vs processor/model resolution mismatch." }, { - "left": "issue:43377", - "right": "issue:44610", + "left": "issue:42796", + "right": "issue:44977", "accept": false, - "reason": "Both involve vision inputs, but one is batched-vs-single padding-mask correctness and the other is a processor/model resolution mismatch." + "reason": "Both are Qwen/Tokenizer-adjacent only at a very broad level; one is a tokenizer BPE-marker issue, the other is flash-attention generation failure." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is missing BOS/EOS insertion for a specific tokenizer while the other is wrong token IDs across several models." + "reason": "Both involve tokenization regressions, but one is missing BOS/EOS insertion for a specific tokenizer while the other is wrong token IDs across several models." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "Both mention dtype/weight handling, but one is lm_head tying in a multimodal model and the other is a bfloat16 dtype mismatch across different models." + "reason": "Unrelated: tied lm_head weights vs bfloat16 dtype mismatch in different model wrappers." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Both are model-loading/config issues, but one is Qwen3 MoE weight conversion with accelerate+deepspeed and the other is a missing gradient-checkpointing flag." + "reason": "Different concerns: MoE weight conversion/loading vs a missing gradient-checkpointing support flag." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Both are offload/quantization related, but one is a specific dequantized-model load failure with device mapping and the other is a broader quantized offloading OOM complaint." + "reason": "Both touch offloading/quantization, but the concrete failure modes and code paths differ too much to treat as duplicates." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Both are about input handling for vision models, but one is MIMI padding-mask batching and the other is Perceiver image resolution/interpolation behavior." + "reason": "Different model families and symptoms: missing padding-mask support vs non-default-resolution image classification failure." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "One is a Llava Next loading segfault and the other is sinq quantization save/load breakage; not the same failure mode." + "reason": "Segfault while loading Llava Next is not the same as quantized-model save/load corruption." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Both concern model quality/perplexity-like symptoms, but one is a specific perplexity regression and the other is ineffective gradient checkpointing with PEFT LoRA." + "reason": "Perplexity anomaly and gradient-checkpointing ineffectiveness are unrelated bugs." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "One is weight tying in a multimodal generation model; the other is packed-weight unpacking during accelerate loading for BitNet. Different bugs." + "reason": "Tied-weight serialization is unrelated to packed-weight unpacking during accelerate loading." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "One is a missing pad_token_idx config regression; the other is improper unwrapping of text_config in composite Qwen multimodal configs." - }, - { - "left": "issue:41720", - "right": "issue:43322", - "accept": false, - "reason": "Different runtime failures: CUDA assert on Qwen3 auto device mapping versus a segmentation fault while loading Llava Next." + "reason": "Config missing pad_token_idx is different from failing to unwrap text_config for composite multimodal configs." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "One is broken FSDP CPU RAM efficient loading; the other is non-persistent buffers being filled with junk. Similar upgrade scope, but not the same bug." + "reason": "FSDP CPU RAM-efficient loading and non-persistent buffer initialization are separate regressions." }, { - "left": "issue:43526", - "right": "issue:43883", + "left": "issue:41720", + "right": "issue:43322", "accept": false, - "reason": "One is a BeitImageProcessorFast label-reduction bug; the other is a missing all_tied_weights_keys attribute in a causal LM." + "reason": "Different loading failures on different models; no shared concrete bug path." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "One is a failure to run a 3d_parallel script; the other is a Qwen3Moe FineGrainedFP8Config loading failure. No shared code-path evidence." + "reason": "3d_parallel execution failure is not the same as Qwen3MoE FP8 loading failure." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Both are v5 save/load regressions around multimodal weights, but they affect different serialized keys and different model components." + "reason": "Both are serialization-related, but one is lm_head tying/serialization and the other is visual-encoder key saving in a specific multimodal model." }, { - "left": "issue:43881", - "right": "issue:44991", + "left": "issue:43526", + "right": "issue:43883", "accept": false, - "reason": "Both are tokenizer-loading problems, but they are model-specific and describe different breakages: a load failure for glm-4v-9b versus wrong tokenizer loading for est-roberta." + "reason": "Image processor label reduction bug is unrelated to missing all_tied_weights_keys on a model class." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Both touch Qwen-related inference, but one is a CUDA device-mapping crash and the other is flash-attention generation corruption." + "reason": "Both mention Qwen, but one is device-mapping assert and the other is flash-attention generation behavior." }, { "left": "issue:26413", "right": "issue:28530", "accept": false, - "reason": "One is resume_from_checkpoint epoch-iterator failure; the other is early-stopping/metric_for_best_model handling. Different trainer logic." + "reason": "Trainer checkpoint-resume sample iterator failure is not the same as early-stopping metric resolution." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Model loading failure and tokenizer loading failure affect different code paths and likely different root causes." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "One is increased reserved memory under int4 quantization; the other is a T5/apex fused RMSNorm memory leak. Different sources of memory issues." + "reason": "Both involve memory, but one is int4 quantization reservation/OOM and the other is an apex RMSNorm leak." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "A custom compute_loss signature change is unrelated to a 3d_parallel execution failure." + "reason": "Unexpected compute_loss keyword argument is unrelated to the 3d_parallel runtime failure." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "One is a perplexity regression on a specific model/dataset; the other is a broad tokenizer-behavior change. Not the same bug." - }, - { - "left": "issue:43450", - "right": "issue:45701", - "accept": false, - "reason": "Video processor batch-shape handling is unrelated to tokenizer behavior changes." + "reason": "A model-specific perplexity regression is not the same as a general tokenization-change report." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Trainer compute_loss signature breakage and Qwen3Moe FP8 loading are separate issues." + "reason": "Trainer API signature breakage is unrelated to Qwen3MoE FP8 loading." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "One is strict config blocking a model config; the other is tokenizer loading failure. Different layers of the stack." + "reason": "Strict config loading and tokenizer loading are separate problems." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape bug is unrelated to tokenization changes." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "NER/padding/truncation crashes in a tokenizer are unrelated to dtype mismatches in bfloat16 inference." + "reason": "Tokenizer crash on NER/padding is unrelated to dtype mismatch in inference." }, { "left": "issue:43065", "right": "issue:46402", "accept": false, - "reason": "Sam3PixelDecoder's dummy Conv2d and missing Sam3Model text-encoder weights are distinct model issues." + "reason": "A dummy Conv2d in the pixel decoder is not the same issue as missing text-encoder weight loading." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Both are memory/offload-related, but one is quantization offloading and the other is mmap OOM avoidance on Strix Halo; not the same fix." + "reason": "Offloading/quantization behavior and Strix Halo mmap OOM are different root causes." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile/SDPA failure is unrelated to accelerate weight-unpacking failure." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "One is dtype not propagating to Blip2 components; the other is incorrect max_steps recomputation when resuming checkpoint with batch-size changes." + "reason": "Model dtype propagation and resume_from_checkpoint max_steps calculation are unrelated bugs." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "One is lm_head weight tying in Mistral3ForConditionalGeneration; the other is incorrect visual-encoder keys on save_pretrained for Qwen3.5." - }, - { - "left": "issue:43550", - "right": "issue:43720", - "accept": false, - "reason": "torch.compile with SDPA on Bamba is a different failure from BitNet packed-weight unpacking during accelerate loading." + "reason": "Tied lm_head weights in Mistral3 are not the same as incorrect visual encoder keys in Qwen3.5 save_pretrained." } ] }, @@ -4475,8 +4473,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -4535,46 +4533,46 @@ "issue:44423|issue:45406", "issue:44186|issue:44610", "issue:44610|issue:44898", - "issue:43295|issue:45701", "issue:43550|issue:43606", + "issue:43295|issue:45701", "issue:42915|issue:43317", "issue:45632|issue:45684", "issue:36010|issue:42617", "issue:41720|issue:45161", - "issue:44671|issue:44743", "issue:43531|issue:43572", + "issue:44671|issue:44743", "issue:42915|issue:46291", "issue:28530|issue:33666", - "issue:43526|issue:43881", "issue:44186|issue:44361", - "issue:43540|issue:43782", "issue:36331|issue:41720", + "issue:43526|issue:43881", + "issue:43540|issue:43782", "issue:42915|issue:43322", "issue:43316|issue:44913", "issue:28530|issue:30064", "issue:29159|issue:36032", "issue:33357|issue:35707", - "issue:43388|issue:45059", "issue:30064|issue:34634", + "issue:43388|issue:45059", "issue:43550|issue:43653", "issue:43606|issue:43720", "issue:43873|issue:45704", "issue:44387|issue:45701" ], - "prepared_review_unit_hash": "a9e6aa47643c468bbfb384dda376b27c38ce9447980fb1ccd0cad1da037e1260", + "prepared_review_unit_hash": "a548408dcafc7e21bc57994bee5c776716d3ba3b711b05eb3a4ae098cdb5822d", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36675, + "serialized_chars": 36676, "estimated_input_tokens": 9169, "estimated_eval_tokens": 18594 }, @@ -4583,150 +4581,150 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster does not look like a duplicate set; it is mostly a collection of unrelated Transformers issues with superficial similarity from broad subsystem keywords (tokenizers, quantization, offloading, multimodal).", - "confidence": 0.19, - "canonical_issue_reason": "Issue #43873 is the best umbrella candidate because it is broad, actively discussed, and sits in a recurring failure area (offloading + quantization), but it still does not appear to be the same underlying bug as the others.", + "summary": "This cluster is highly heterogeneous: most pairs share only loose topical similarity (tokenizers, model loading, offloading, multimodal inference, configs) but do not look like the same underlying bug or change. I would not merge any soft pairs; the best triage anchor is the broad offloading/quantization issue.", + "confidence": 0.34, + "canonical_issue_reason": "issue:43873 is the broadest and most central umbrella in the cluster, covering the recurring offloading/quantization/device-map failures better than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "#43873 is the strongest issue to anchor triage due to its generality and higher discussion/inbound activity, though the cluster does not form a true duplicate group.", + "best_issue_reason": "Among the issues here, 43873 is the most reusable representative for duplicate triage because it is broad, active, and closest to the repeated loading/offload failure pattern.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Different models and failures: Gemma3n variable-name collision vs Gemma4 PLE device mismatch." + "reason": "Both involve Gemma models, but one is a Gemma3n forward variable-collision bug and the other is a Gemma4 device-mapping mismatch; different code paths and failures." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Unrelated problems: Qwen3 auto device mapping CUDA assert vs inability to run 3d_parallel.py." + "reason": "Qwen3 auto-device-map CUDA assert and a 3d_parallel.py execution failure are unrelated problems." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Both involve serving/multimodal paths, but the concrete crashes differ (_tokenizer missing vs str.to AttributeError)." + "reason": "Continuous-batching multimodal crash with a string `.to()` error is not the same as Gemma4Processor missing `_tokenizer`." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Tokenizer crash on NER/padding vs processor output resolution mismatch; not the same bug." + "reason": "LayoutLMv2 tokenizer crashes on NER/padding, while OmDet-Turbo has an image-size mismatch; different processors and root causes." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both are image-size related, but they affect different models and different code paths." + "reason": "OmDet-Turbo input-size mismatch and Perceiver interpolation failure are unrelated model/processor bugs." }, { - "left": "issue:43295", - "right": "issue:45701", + "left": "issue:43550", + "right": "issue:43606", "accept": false, - "reason": "A regression in processor/tokenizer access is not the same as a vague tokenization behavior change." + "reason": "Bamba torch.compile/SDPA failure and Bark-small CPU offload device mismatch are different concrete issues." }, { - "left": "issue:43550", - "right": "issue:43606", + "left": "issue:43295", + "right": "issue:45701", "accept": false, - "reason": "Different concrete failures: SDPA/torch.compile vs CPU offload device mismatch." + "reason": "Both mention version-related tokenizer behavior, but one is a regression in processor/tokenizer access and the other is a broad tokenization-change report; not the same bug." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Both are loading/quantization-adjacent, but one is FineGrainedFP8 model failure and the other is dequantized gpu+cpu offload loading." + "reason": "Qwen3MoE FineGrainedFP8Config failure is distinct from dequantized-model loading with gpu+cpu offload." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "Cache-path collision and read-only permission propagation are unrelated." + "reason": "Cache-path collisions for trust_remote_code and permission propagation in save_pretrained are unrelated filesystem behaviors." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to a 3d_parallel.py runtime issue." + "reason": "GenerationMixin import failure and a 3d_parallel.py runtime issue are not the same underlying bug." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Different distributed-training problems: device-map CUDA assert vs tensor-parallel-only failure." + "reason": "Both are device/distributed-related, but one is a Qwen3 auto-device-map CUDA assert and the other is GPT-OSS MoE TP behavior; different failures." }, { - "left": "issue:44671", - "right": "issue:44743", + "left": "issue:43531", + "right": "issue:43572", "accept": false, - "reason": "CamemBERT masked-LM prediction regression is unrelated to Qwen3 recurrent-state caching." + "reason": "Qwen3-MoE sliding-window behavior and StableLmConfig pad_token_idx regression are unrelated." }, { - "left": "issue:43531", - "right": "issue:43572", + "left": "issue:44671", + "right": "issue:44743", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to missing pad_token_idx in StableLmConfig." + "reason": "CamemBERT masked-LM prediction errors and Qwen3.5 recurrent-state reset are different model-specific bugs." }, { "left": "issue:42915", "right": "issue:46291", "accept": false, - "reason": "FP8 model loading and pipelines import failing on kernels version requirements are unrelated." + "reason": "A Qwen3MoE FP8 config issue and a kernels import/version requirement problem are unrelated." }, { "left": "issue:28530", "right": "issue:33666", "accept": false, - "reason": "Early stopping metric selection and Qwen2-VL multi-GPU training are unrelated." + "reason": "Early-stopping metric selection and Qwen2-VL multi-GPU training are not the same issue." }, { - "left": "issue:43526", - "right": "issue:43881", + "left": "issue:44186", + "right": "issue:44361", "accept": false, - "reason": "BeitImageProcessorFast label reduction and glm-4v loading failure are different bugs." + "reason": "Both are tokenizer failures, but the affected tokenizers, triggers, and failure modes differ." }, { - "left": "issue:44186", - "right": "issue:44361", + "left": "issue:36331", + "right": "issue:41720", "accept": false, - "reason": "Both mention tokenizers, but they are separate tokenizer/model failures with different symptoms." + "reason": "Trainer `compute_loss` signature change is unrelated to Qwen3 device-mapping CUDA assertions." }, { - "left": "issue:43540", - "right": "issue:43782", + "left": "issue:43526", + "right": "issue:43881", "accept": false, - "reason": "Different Qwen3 family issues with different modalities and failure modes; not one underlying bug." + "reason": "BeitImageProcessorFast label reduction and glm-4v-9b load failure are different bugs." }, { - "left": "issue:36331", - "right": "issue:41720", + "left": "issue:43540", + "right": "issue:43782", "accept": false, - "reason": "CustomTrainer compute_loss signature mismatch is unrelated to Qwen3 device mapping." + "reason": "Qwen3OmniMoe video-input ValueError and Qwen3VL weight_only load error affect different modalities and paths." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "FP8 loading failure is unrelated to Llava Next segmentation faults." + "reason": "Qwen3MoE FP8 config failure and Llava Next segmentation fault are unrelated." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "API discrepancy in Gemma3TextConfig is unrelated to GPTNeoX rotary_pct reload behavior." + "reason": "Gemma3TextConfig API mismatch and GPTNeoXConfig rotary_pct reload behavior are different config regressions." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Early stopping metric lookup and void segmentation-map processing are unrelated." + "reason": "Early stopping metric lookup and void segmentation-map processing are unrelated issues." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Both are tokenizer-related, but slow/fast inconsistency and T5 add_special_tokens conflict are different code paths." + "reason": "Slow/fast tokenizer inconsistency and T5Tokenizer `add_special_tokens` attribute conflict are different tokenizer bugs." }, { "left": "issue:33357", @@ -4735,40 +4733,40 @@ "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are unrelated." }, { - "left": "issue:43388", - "right": "issue:45059", + "left": "issue:30064", + "right": "issue:34634", "accept": false, - "reason": "Metric gathering label truncation and SAM3 text/bbox behavior are different failures." + "reason": "Void segmentation-map processing and Bark voice_preset handling are not the same bug." }, { - "left": "issue:30064", - "right": "issue:34634", + "left": "issue:43388", + "right": "issue:45059", "accept": false, - "reason": "Image processor void-segmentation failure vs BarkProcessor voice_preset issue are unrelated." + "reason": "Metric gathering label truncation and SAM3 text/bounding-box odd behavior are different code paths." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "SDPA/compile failure is not the same as BigBirdTokenizer special-token registration." + "reason": "Bamba SDPA/compile failure and BigBirdTokenizer mask-token registration are unrelated." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "Both touch offloading/loading, but the concrete bugs differ: device mismatch vs packed-weight unpacking." + "reason": "Both touch loading/offload, but one is a CPU-offload device mismatch and the other is packed-weight unpacking during accelerate loading; not the same concrete fix." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantized offloading behavior and T5 apex RMSNorm memory leak are unrelated." + "reason": "Quantization/offloading problems and a T5 apex RMSNorm memory leak are unrelated." }, { "left": "issue:44387", "right": "issue:45701", "accept": false, - "reason": "Int4 quantization OOM and tokenization changes across versions are different issues." + "reason": "Int4 quantization OOM in Transformers 5.x and tokenization changes across versions are different regression classes." } ] }, @@ -4787,8 +4785,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -4853,226 +4851,240 @@ "issue:34567|issue:35707", "issue:44451|issue:44991", "issue:43531|issue:44589", - "issue:44077|issue:44625", "issue:45759|issue:45859", + "issue:44077|issue:44625", "issue:44625|issue:45479", "issue:44898|issue:45072", "issue:44514|issue:44545", "issue:45588|issue:45759", - "issue:44368|issue:45357", "issue:44062|issue:44451", "issue:44987|issue:45042", + "issue:44368|issue:45357", "issue:39692|issue:44610", - "issue:43421|issue:43927", "issue:44918|issue:44936", + "issue:43421|issue:43927", "issue:43381|issue:43844", "issue:43746|issue:46050", "issue:43782|issue:45381", "issue:42886|issue:44843", - "issue:45704|issue:45923" + "issue:45704|issue:45923", + "issue:43421|issue:43653", + "issue:45704|issue:45859" ], - "prepared_review_unit_hash": "b1311fa8675bbf886384ab7d9ee47df2cbff8c12091e46589eb0bc55464de79b", + "prepared_review_unit_hash": "9a8ae283047095734ed7efc91bd37660a089df23fca4ea3f29312de5cd8f2528", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36432, - "estimated_input_tokens": 9108, - "estimated_eval_tokens": 18472 + "soft_pair_count": 30, + "serialized_chars": 37039, + "estimated_input_tokens": 9260, + "estimated_eval_tokens": 18776 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly unrelated issues. I found one clear duplicate issue pair (`44514`/`44545`) and one likely same offline-tokenizer regression pair (`42886`/`44843`); the rest are different bugs touching different code paths or models.", - "confidence": 0.78, - "canonical_issue_reason": "issue:44514 is the clearest representative bug report among the true duplicate pair; `issue:44545` is the same batched `apply_chat_template(..., padding=False)` crash.", + "summary": "These candidates are mostly unrelated issue reports. The only clear duplicate pair is the two Qwen2_5_VLProcessor batched padding=False crash reports; everything else appears to be different bugs or different code paths within the same subsystem.", + "confidence": 0.97, + "canonical_issue_reason": "issue:44545 is the clearest canonical issue: it describes the exact Qwen2_5_VLProcessor.apply_chat_template batched padding=False crash, and it matches 44514 verbatim with slightly more reference activity.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44514 is the best issue anchor because it states the bug directly and matches the duplicate partner exactly.", + "best_issue_reason": "issue:44545 is the best issue representative because it is the exact duplicate target for the only accepted soft pair and has the strongest engagement signal among the two identical reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Different models and failures: Qwen3 auto device mapping CUDA assert vs GraniteSpeech PEFT local checkpoint loading." + "reason": "Different bugs: Qwen3 auto device-mapping CUDA assert vs GraniteSpeech local checkpoint/PEFT loading." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Different processors and runtime contexts; one is a missing `_tokenizer` attribute in `Gemma4Processor`, the other is a Qwen3.5 streaming inference failure." + "reason": "Different model and failure mode: Gemma4Processor missing _tokenizer vs Qwen3.5 streaming chat/completions failure." }, { "left": "issue:45357", "right": "issue:46402", "accept": false, - "reason": "Both are save/load regressions, but they affect different models and different weights paths (`visual encoder` vs `text encoder`)." + "reason": "Both are model-loading regressions, but they target different models and different missing-weight paths." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A typo fix in cache strategies is unrelated to docs still listing removed pipeline tasks." + "reason": "A typo in cache strategy docs is unrelated to outdated pipeline-task documentation." }, { "left": "issue:43531", "right": "issue:45759", "accept": false, - "reason": "Both mention Qwen, but one is a sliding-window bug in Qwen3-MoE and the other is `from_config` unwrapping for composite multimodal configs." + "reason": "Different config bugs: sliding_window handling in Qwen3-MoE vs text_config unwrapping for composite Qwen multimodal configs." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Unrelated subsystems: Bark CPU offload/device mismatch vs BigBirdTokenizer special-token registration/decode output." + "reason": "Unrelated: CPU offload device mismatch vs tokenizer mask-token special registration." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "Import warning for local paths ending in slash is unrelated to a multimodal token-counting video-branch bug." + "reason": "Different code paths: local-path trailing-slash import warning vs video token counting using undefined names." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "BarkProcessor voice preset bug is unrelated to Qwen3 auto device mapping CUDA assertions." + "reason": "BarkProcessor voice_preset bug is unrelated to Qwen3 auto device mapping on A800." }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "TrainerState token counting and progressive generation with `inputs_embeds`/`past_key_values` are different issues." + "reason": "TrainerState token counters not updating is a different issue from progressive generation with inputs_embeds/past_key_values." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer load failures, but they involve different models and no shared concrete code path is evident." + "reason": "Both are load failures, but they involve different models and likely different tokenizer/model compatibility problems." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Qwen3 sliding-window behavior is unrelated to the Float8 storage-object loading error." + "reason": "Different failures: Qwen3-MoE sliding_window handling vs Float8 storage lookup error." }, { - "left": "issue:44077", - "right": "issue:44625", + "left": "issue:45759", + "right": "issue:45859", "accept": false, - "reason": "Optional `post_init` handling in PatchTSTMixer is unrelated to Qwen3.5 `num_labels` propagation." + "reason": "Different bugs: multimodal config unwrapping vs rotary embedding CPU-offload compatibility." }, { - "left": "issue:45759", - "right": "issue:45859", + "left": "issue:44077", + "right": "issue:44625", "accept": false, - "reason": "Different Qwen3.5 MoE problems: config unwrapping vs CPU-offload rotary embedding compatibility." + "reason": "PatchTSMixer post_init policy and Qwen3.5 num_labels propagation are unrelated." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "`num_labels` propagation and zero-loss classification are related to classification config, but they are not the same concrete bug." + "reason": "Both touch sequence-classification config, but one is num_labels propagation and the other is a degenerate-loss bug for num_labels=1." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Perceiver image-size mismatch and bfloat16 dtype mismatches are different failures in different models." + "reason": "Different model families and symptoms: Perceiver positional interpolation vs bf16 dtype mismatches." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Same exact crash: `Qwen2_5_VLProcessor.apply_chat_template` fails on batched input when `padding=False`." + "reason": "Same exact bug: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding=False." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention `s_aux=None` handling is unrelated to Qwen3.5 config unwrapping." - }, - { - "left": "issue:44368", - "right": "issue:45357", - "accept": false, - "reason": "A warning about `tie_word_embeddings=False` during LoRA fine-tuning is not the same as incorrect visual encoder key saving." + "reason": "Different code paths: flash_attention sink handling vs Qwen composite config loading." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "`AddedToken(..., special=...)` argument conflict is unrelated to a model loading failure for ScandiBERT." + "reason": "AddedToken constructor argument conflict is unrelated to ScandiBERT loading failure." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Tokenizer/model load regression and PIL backend torchvision dependency are unrelated." + "reason": "Different regressions: loading physical-intelligence/fast vs PIL image processors incorrectly requiring torchvision." }, { - "left": "issue:39692", - "right": "issue:44610", + "left": "issue:44368", + "right": "issue:45357", "accept": false, - "reason": "Both mention processor/model mismatches, but one is a docs example with quantization errors and the other is an OmDet-Turbo input-size mismatch." + "reason": "A warning about tie_word_embeddings is not the same as incorrect visual-encoder key saving." }, { - "left": "issue:43421", - "right": "issue:43927", + "left": "issue:39692", + "right": "issue:44610", "accept": false, - "reason": "Runtime special-token post-processor updates and DiaConfig token ID persistence are different tokenization/config bugs." + "reason": "Docs-example errors for SigLIP2 are unrelated to OmDet-Turbo's processor/model image-size mismatch." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "TRL SFT unpacking of Qwen3.5 embeddings and post-train `evaluate()` failure are different trainer issues." + "reason": "Different trainer issues: SFT embedding unpacking vs evaluate() failing after train()." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Runtime special-token/post-processor updates are unrelated to DiaConfig save/load token-ID loss." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Gradient checkpointing in eval mode and ZeRO-3 gradient growth are distinct training behaviors." + "reason": "Eval-mode gradient checkpointing restriction is unrelated to gradient blow-up under ZeRO-3." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "Local checkpoint loading with PEFT adapters and quantized `sinq` save/load breakage are not the same bug." + "reason": "Different persistence bugs: PEFT local checkpoint loading vs quantized save_pretrained/from_pretrained breakage." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "`weight_only=True` load failure and wrong video `vision_position_ids` are different Qwen-VL issues." + "reason": "Different Qwen VL bugs: weight_only load error vs wrong video vision_position_ids." }, { "left": "issue:42886", "right": "issue:44843", - "accept": true, - "reason": "Likely the same offline tokenizer-loading regression: offline cache loading breaks because `AutoTokenizer.from_pretrained` calls `model_info()` unconditionally in `_patch_mistral_regex`." + "accept": false, + "reason": "Both concern offline mode, but one is cache loading generally and the other is an unconditional model_info call in a specific regex patch." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "T5 RMSNorm memory leak and missing gradient-checkpointing support in a trust_remote_code model are unrelated." + "reason": "Apex RMSNorm memory leak and missing gradient-checkpointing support are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Special-token post-processor refresh and BigBirdTokenizer mask-token registration are different tokenizer bugs." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "Different issues: T5 apex RMSNorm leak vs Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility." } ] }, "evaluator_result": { "accept": true, - "feedback": "The clustering is grounded in the packet and the soft-edge verdicts are conservative. The two accepted pairs are plausibly duplicate/near-duplicate bug reports, and the rejected pairs are mostly clearly distinct by model, subsystem, or failure mode. No clear overstatement of evidence." + "feedback": "Grounded and conservative overall. The only accepted soft pair is the clear duplicate between issue:44514 and issue:44545, and choosing issue:44545 as canonical/best is supported by the identical title plus slightly higher reference activity. The other pairwise rejections are phrased cautiously enough from the packet metadata alone." }, "evaluator_used": true, "retried": false, @@ -5088,13 +5100,14 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ "issue:30064", "issue:36010", + "issue:36331", "issue:39692", "issue:41720", "issue:43295", @@ -5135,7 +5148,6 @@ "issue:45685", "issue:45704", "issue:45834", - "issue:45859", "issue:45907", "issue:45923", "issue:46032", @@ -5143,11 +5155,9 @@ "issue:46463" ], "soft_pairs": [ - "issue:46032|issue:46463", - "issue:43421|issue:43653", - "issue:45704|issue:45859", - "issue:41720|issue:45704", "issue:43531|issue:43645", + "issue:41720|issue:45704", + "issue:46032|issue:46463", "issue:45663|issue:45923", "issue:43421|issue:44186", "issue:43994|issue:44671", @@ -5156,14 +5166,14 @@ "issue:45081|issue:45356", "issue:43596|issue:44220", "issue:45356|issue:45907", - "issue:43540|issue:46294", "issue:30064|issue:36010", + "issue:43540|issue:46294", "issue:43720|issue:43819", "issue:43653|issue:43720", "issue:44220|issue:44661", - "issue:43618|issue:43697", "issue:43295|issue:45834", "issue:43550|issue:45072", + "issue:43618|issue:43697", "issue:44062|issue:44488", "issue:43550|issue:43819", "issue:43540|issue:44008", @@ -5172,230 +5182,225 @@ "issue:43531|issue:43643", "issue:43322|issue:45685", "issue:45183|issue:45479", - "issue:43335|issue:43756", "issue:43643|issue:44877", - "issue:44843|issue:45081" + "issue:43335|issue:43756", + "issue:44843|issue:45081", + "issue:30064|issue:36331" ], - "prepared_review_unit_hash": "c99b3bb4a6c5781028640ed7512d56cee28880ea68db9d05b3a3dab49be13c80", + "prepared_review_unit_hash": "419301f45f1954ac0c6e3d8befcb27aaab487f9de2bc271d04b0753cd8c3af0d", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37789, - "estimated_input_tokens": 9448, - "estimated_eval_tokens": 19152 + "soft_pair_count": 31, + "serialized_chars": 37522, + "estimated_input_tokens": 9381, + "estimated_eval_tokens": 19018 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a duplicate set; the items cover many unrelated bugs across tokenizers, multimodal processors, config loading, model-specific inference, and training/offload paths.", - "confidence": 0.97, - "canonical_issue_reason": "No strong duplicate center emerges. The issues are mostly unrelated, so there is no single canonical issue for the cluster.", + "summary": "This cluster is heterogeneous: tokenizer/loading regressions, multimodal processor bugs, model-specific inference issues, and unrelated trainer/config problems. No soft pair looks like a true duplicate or mergeable change, so I would not collapse any of them.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44843 is the closest cluster anchor because it\u2019s a framework-level tokenizer-loading regression with broad impact; the rest split into separate model- or subsystem-specific bugs, so there is no true canonical duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46032 is the most self-contained and concrete report: it names a specific model component, a precise failing condition, and a clear incorrect-behavior symptom.", + "best_issue_reason": "44843 is the most reusable representative for this mixed set, but only as a loose anchor for the tokenizer/loading subtheme rather than a real duplicate canonical.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:46032", - "right": "issue:46463", - "accept": false, - "reason": "Different subsystems and failures: Mamba2 caching correctness vs T5 decoder input validation messaging." - }, - { - "left": "issue:43421", - "right": "issue:43653", - "accept": false, - "reason": "Both involve tokenizers, but one is runtime special-token/post-processor syncing and the other is BigBird mask-token registration/decoding." - }, - { - "left": "issue:45704", - "right": "issue:45859", + "left": "issue:43531", + "right": "issue:43645", "accept": false, - "reason": "Unrelated issues: a T5 Apex RMSNorm memory leak versus Qwen3 MoE CPU-offload incompatibility." + "reason": "Different bugs: Qwen3-MoE sliding-window handling vs Jupyter/custom-model initialization regression." }, { "left": "issue:41720", "right": "issue:45704", "accept": false, - "reason": "Different models and bugs: Qwen3 auto device-mapping CUDA assert versus T5 Apex RMSNorm memory leak." + "reason": "Unrelated failures: Qwen3 auto-device-map CUDA assert vs T5 RMSNorm memory leak." }, { - "left": "issue:43531", - "right": "issue:43645", + "left": "issue:46032", + "right": "issue:46463", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior is unrelated to Jupyter custom-model initialization regressions." + "reason": "Different code paths and symptoms: Mamba2 cache correctness vs T5 decoder-input validation error text." }, { "left": "issue:45663", "right": "issue:45923", "accept": false, - "reason": "Gemma-4 FSDP2 shared_kv_state logic and Nemotron trust_remote_code gradient-checkpointing flag are different code paths." + "reason": "Gemma4 FSDP2 shared_kv_states bug is unrelated to a missing supports_gradient_checkpointing flag." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer-related, but they target different tokenizer APIs and different crash modes." + "reason": "Both tokenizer-related, but one is post-processor update behavior and the other is a LayoutLMv2 NER/padding crash." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Different model families and symptoms: SigLIP2 pipeline/output mismatch versus CamemBERT masked-LM prediction regression." + "reason": "Different model-specific regressions: SigLIP2 inference nonsense vs CamemBERT MLM predictions." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "PEFT local checkpoint loading for GraniteSpeech is not the same bug as quantization/offloading behavior." + "reason": "Local checkpoint/PEFT loading failure is not the same as an offloading-with-quantization issue." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "Documentation/example errors are unrelated to quantization offloading bugs." + "reason": "A docs/example mismatch and quantization failure is distinct from the runtime offloading bug in 43873." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Both mention fix_mistral_regex, but one is an AttributeError crash in patching and the other is a tokenizer regression with codec handling/warning text." + "reason": "Same helper area, but different root causes: backend_tokenizer AttributeError vs codec/ warning regression in Kimi tokenizer." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "BertModel DeepSpeed init failure is unrelated to _torch_extract_fbank_features audio preprocessing." + "reason": "Deepspeed/BertModel zero3 init IndexError is unrelated to fbank feature extraction." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Tokenizer regression around Mistral regex/codec handling is unrelated to list_repo_files revision forwarding." + "reason": "Tokenizer regression vs repo-file revision forwarding in CI are unrelated." }, { - "left": "issue:43540", - "right": "issue:46294", + "left": "issue:30064", + "right": "issue:36010", "accept": false, - "reason": "Both are video-related, but one is Qwen3OmniMoe video-input processing and the other is Qwen3.5 chat-template token/feature mismatch." + "reason": "Image processing for void segmentation maps has nothing to do with GenerationMixin import failures." }, { - "left": "issue:30064", - "right": "issue:36010", + "left": "issue:43540", + "right": "issue:46294", "accept": false, - "reason": "Image segmentation-map processing and a GenerationMixin import error are completely different failures." + "reason": "Both involve video, but one is a Qwen3OmniMoe processing ValueError and the other is a Qwen3.5 chat-template token/feature mismatch." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "BitNet packed-weight loading and DAC latent reconstruction mismatch do not share the same code path." + "reason": "BitNet packed-weight loading and DAC latent reconstruction are different model-specific bugs." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Tokenizer special-token registration and BitNet accelerate-loading unpacking are unrelated." + "reason": "BigBirdTokenizer special-token registration and BitNet accelerate loading are unrelated." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio feature extraction and tokenizer-mapping add-new-model logic are different components and bugs." - }, - { - "left": "issue:43618", - "right": "issue:43697", - "accept": false, - "reason": "CLIPOutput attentions assignment and RTDetrV2 output changes are distinct model-behavior issues." + "reason": "A torchaudio feature-extraction issue is not the same as add-new-model-like failing in TOKENIZER_MAPPING_NAMES." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Processor/tokenizer regression with custom models is not the same as Kosmos2.5 long-OCR index errors." + "reason": "Processor.tokenizer regression in custom models is different from Kosmos2.5 OCR long-input indexing." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "torch.compile SDPA failure in Bamba-9B-v2 is unrelated to dtype mismatch issues in SwitchTransformers/TimmWrapperModel." + "reason": "torch.compile/SDPA failure in Bamba-9B-v2 is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Missing CLIPOutput attentions is not the same bug as RTDetrV2 producing different outputs." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "AddedToken keyword conflicts are not the same bug as failing to load cjvt/sleng-bert." + "reason": "AddedToken constructor argument collision is unrelated to sleng-bert loading failure." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Different models and execution paths: SDPA compile failure versus DAC latent/STE mismatch." + "reason": "These are separate bugs in different models and execution paths: SDPA compile vs DAC latent conversion." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Video-input processing failure in Qwen3OmniMoe is unrelated to Gemma3n variable-name collision in forward()." + "reason": "Video-input failure in Qwen3OmniMoe is unrelated to Gemma 3n's audio_mel_mask variable collision." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes and Perceiver image-classification interpolation failures are unrelated." + "reason": "Tokenizer crash in LayoutLMv2 is unrelated to Perceiver image-classification resolution handling." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "BigBird special-token decode behavior is unrelated to DAC latent reconstruction correctness." + "reason": "BigBirdTokenizer special-token registration and DAC latent mismatch do not share the same underlying bug." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to AutoConfig trust_remote_code field loss." + "reason": "Qwen3-MoE sliding_window behavior and AutoConfig trust_remote_code field loss are unrelated." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Llava Next loading segfaults and MPS histogram dtype support are different layers and failure modes." + "reason": "Llava Next load segfault and MPS histogram_mps type support are different failures." }, { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Input validation error messaging and sequence-classification zero-loss behavior are different bugs." + "reason": "Input-validation XOR error messaging and single-label-classification zero loss are different problems." }, { - "left": "issue:43335", - "right": "issue:43756", + "left": "issue:43643", + "right": "issue:44877", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are unrelated model-configuration issues." + "reason": "Both are config-loading related, but one is remote-code field loss and the other is strict config blocking granite_speech load." }, { - "left": "issue:43643", - "right": "issue:44877", + "left": "issue:43335", + "right": "issue:43756", "accept": false, - "reason": "Trust-remote-code missing fields and strict-config loading failure are separate configuration-handling bugs." + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are different model-architecture bugs." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Offline model_info calls in _patch_mistral_regex and backend_tokenizer AttributeError in the same helper are different concrete failures." + "reason": "Same tokenizer patch area, but offline model_info calls and backend_tokenizer AttributeError are distinct regressions." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Void segmentation-map processing and CustomTrainer.compute_loss kwargs are unrelated." } ] }, @@ -5414,15 +5419,15 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ - "issue:30064", + "issue:28218", + "issue:29159", "issue:34634", "issue:36010", - "issue:36331", "issue:39692", "issue:41720", "issue:43388", @@ -5469,224 +5474,217 @@ "issue:46512" ], "soft_pairs": [ - "issue:30064|issue:36331", + "issue:41720|issue:44991", "issue:43716|issue:44008", "issue:43388|issue:46032", - "issue:41720|issue:44991", "issue:44821|issue:45084", "issue:43950|issue:45987", "issue:44448|issue:44671", - "issue:43582|issue:46512", "issue:39692|issue:45834", + "issue:43582|issue:46512", "issue:45584|issue:45834", "issue:44186|issue:44442", - "issue:43421|issue:44811", "issue:43644|issue:43749", + "issue:43421|issue:44811", "issue:43540|issue:45381", "issue:43782|issue:44977", "issue:43577|issue:43818", "issue:43606|issue:45072", "issue:44479|issue:45834", - "issue:46294|issue:46463", - "issue:43825|issue:43994", "issue:44485|issue:44509", "issue:34634|issue:36010", "issue:44805|issue:44936", + "issue:43825|issue:43994", + "issue:46294|issue:46463", "issue:44991|issue:46291", "issue:44561|issue:45020", - "issue:43577|issue:43582", - "issue:45987|issue:46192" + "issue:45987|issue:46192", + "issue:28218|issue:29159" ], - "prepared_review_unit_hash": "138dd2062ec2d51b7753dc72aa1c4db67dbccb7f2657646b5320b93f4d0538ee", + "prepared_review_unit_hash": "22ad27aaed99290bec0c6b3db069f15066558280973e2d6e3225c3ee334c86c5", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35966, - "estimated_input_tokens": 8992, - "estimated_eval_tokens": 18240 + "soft_pair_count": 26, + "serialized_chars": 35644, + "estimated_input_tokens": 8911, + "estimated_eval_tokens": 18078 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "None of the proposed soft pairs look like true duplicates; they mostly span different models, subsystems, or failure modes. No PRs are present, and there is no clear single canonical issue for the set.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue: the candidate issues are largely unrelated, and the soft pairs do not describe the same underlying defect.", - "canonical_pr_reason": "No pull requests were provided.", - "best_issue_reason": "If a representative issue must be picked, issue:45020 is the broadest umbrella-style report, but it still is not a true duplicate target for the others.", - "best_pr_reason": "No pull requests were provided.", + "summary": "This cluster is mostly heterogeneous: tokenizer, vision, audio, cache, dtype, and trainer issues mixed together. I do not see a true duplicate set, though 44561 and 45020 are the closest thematically around remote_code regressions.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the only near-related theme (remote_code breakages in recent transformers versions), so it is the best single representative even though the cluster is not a clean duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 has the widest scope and most general framing among the items here; it can absorb narrower remote_code breakage reports better than the more specific 44561. The rest of the cluster is unrelated.", + "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:30064", - "right": "issue:36331", + "left": "issue:41720", + "right": "issue:44991", "accept": false, - "reason": "Different bugs in different areas: image processor void segmentation maps vs CustomTrainer loss signature handling." + "reason": "Both mention model loading/runtime failures, but one is a Qwen device-map CUDA assert and the other is a tokenizer load failure for a specific RoBERTa model; different bugs and code paths." }, { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Both are multimodal/model-loading problems, but one is a dtype mismatch in Mistral-3 preprocessing and the other is a Gemma 3n variable-name collision." + "reason": "Different modalities and failures: Mistral-3 image preprocessor dtype mismatch vs Gemma3n variable-name collision causing an AttributeError." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "Trainer metric gathering dropping tuple labels is unrelated to Mamba2Mixer use_cache correctness." - }, - { - "left": "issue:41720", - "right": "issue:44991", - "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and tokenizer loading for EMBEDDIA/est-roberta are unrelated failures." + "reason": "One is a metrics/gathering bug dropping tuple labels; the other is a Mamba2 cache correctness bug for seq_len > 1. Not the same underlying issue." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor URL loading and a Jinja/template compilation error are different code paths." + "reason": "AutoImageProcessor URL loading and a template-node compilation TypeError are unrelated failures." }, { "left": "issue:43950", "right": "issue:45987", "accept": false, - "reason": "Both concern loading/runtime state, but one is silent corruption of non-persistent buffers and the other is a StaticCache return-type bug." + "reason": "from_pretrained corrupting non-persistent buffers is a model-loading regression; StaticCache.get_seq_length returning a tensor is a cache API contract bug." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Model output regression in Pegasus vs masked-LM prediction regression in CamemBERT are distinct model-specific bugs." + "reason": "Both are v4/v5 behavioral regressions, but they affect different models and different inference paths; not one shared bug." }, { - "left": "issue:43582", - "right": "issue:46512", + "left": "issue:39692", + "right": "issue:45834", "accept": false, - "reason": "AppleSilicon warmup TypeError and stale config.dtype after from_config are different initialization issues." + "reason": "A documentation example error for SigLIP2 is unrelated to a Kosmos2.5 OCR index error." }, { - "left": "issue:39692", - "right": "issue:45834", + "left": "issue:43582", + "right": "issue:46512", "accept": false, - "reason": "Documentation example errors are unrelated to Kosmos2.5 OCR index errors." + "reason": "Apple Silicon warmup TypeError and stale config dtype after from_config are different dtype-related problems with different code paths." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty-transcription generation failure after token alignment is unrelated to Kosmos2.5 long-OCR indexing." + "reason": "Whisper empty-transcription generation after align_special_tokens is unrelated to Kosmos2.5 long OCR index errors." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes on NER padding/truncation vs AutoTokenizer failing to load a different tokenizer are not the same bug." + "reason": "LayoutLMv2 tokenizer crashes on NER/padding are unrelated to AutoTokenizer failing to load a FastSpeech2Conformer tokenizer." }, { - "left": "issue:43421", - "right": "issue:44811", + "left": "issue:43644", + "right": "issue:43749", "accept": false, - "reason": "Runtime special-token post-processor update behavior is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Non-persistent buffer corruption in from_pretrained is not the same as FSDP CPU RAM efficient loading being broken." }, { - "left": "issue:43644", - "right": "issue:43749", + "left": "issue:43421", + "right": "issue:44811", "accept": false, - "reason": "Non-persistent buffer corruption in from_pretrained and FSDP_CPU_RAM_EFFICIENT_LOADING are different regression classes." + "reason": "Both concern tokenizer/processor behavior, but one is about post-processor updates after special token changes and the other is Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Both involve video inputs in Qwen variants, but the symptoms and affected model paths differ enough that they are not the same underlying bug." + "reason": "Both are video-input bugs in Qwen-family models, but they describe different model variants and different failure mechanisms." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "Qwen3VL weight_only loading error and Qwen3.5 flash-attention generation failure are different concrete problems." + "reason": "Qwen3VL weight_only loading failure and Qwen3.5 flash-attention generation failure are distinct issues." }, { "left": "issue:43577", "right": "issue:43818", "accept": false, - "reason": "Blip2 dtype loading mismatch and Video-LLaVA missing temporal attention/shared weights are unrelated model architecture issues." + "reason": "Blip2 dtype propagation and Video-LLaVA temporal-attention/weight-sharing defects are unrelated." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Both mention device/dtype mismatches, but they affect different models and different execution paths, so they are not the same bug." + "reason": "Bark CPU offload device mismatch and bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are different failures." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Qwen video-input regression is unrelated to Kosmos2.5 long OCR input indexing." - }, - { - "left": "issue:46294", - "right": "issue:46463", - "accept": false, - "reason": "Qwen3.5 video chat-template token mismatch and T5 decoder-input validation error are different validation issues." - }, - { - "left": "issue:43825", - "right": "issue:43994", - "accept": false, - "reason": "One is a pipeline error message/regression about translation tasks, the other is a SigLIP2 model/output issue involving pipeline use but not the same defect." + "reason": "A Qwen video-input regression is unrelated to a Kosmos2.5 OCR index error." }, { "left": "issue:44485", "right": "issue:44509", "accept": false, - "reason": "GLM-5 RoPE implementation discussion is unrelated to stale v5 docs mentioning removed pipeline tasks." + "reason": "GLM-5 RoPE implementation discussion and docs mentioning removed pipeline tasks do not describe the same bug." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated issues." + "reason": "BarkProcessor voice_preset failure and GenerationMixin import breakage are unrelated." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "A tensor mask shape mismatch and trainer.evaluate failing after train are different failure points." + "reason": "Mask shape IndexError in tensor indexing is unrelated to trainer.evaluate failing after train." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message wording and SigLIP2 nonsensical outputs with AutoModel/pipeline are different issues." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen3.5 video chat-template mismatch and a T5 decoder-input validation error are unrelated." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer loading for a specific model and importing transformers.pipelines with kernels version requirements are unrelated import/load issues." + "reason": "Tokenizer loading for a specific model and a kernels dependency/version import failure are unrelated." }, { "left": "issue:44561", "right": "issue:45020", "accept": false, - "reason": "Both touch remote_code compatibility, but 44561 is a specific is_torch_fx_available removal regression while 45020 is a broader recent-versions report, so they are not clearly the same bug." + "reason": "These are the closest pair thematically, but 45020 is a broad umbrella report while 44561 is a specific root-cause report; the shared theme is not enough to confirm one concrete duplicate bug." }, { - "left": "issue:43577", - "right": "issue:43582", + "left": "issue:45987", + "right": "issue:46192", "accept": false, - "reason": "Blip2 dtype mismatch and AppleSilicon caching_allocator_warmup TypeError are distinct runtime problems." + "reason": "A cache length API return-type bug is unrelated to return_tensors being ignored when text_kwargs is passed." }, { - "left": "issue:45987", - "right": "issue:46192", + "left": "issue:28218", + "right": "issue:29159", "accept": false, - "reason": "StaticCache return type corruption and text_kwargs silently overriding return_tensors are unrelated API bugs." + "reason": "Tokenizer extra-space-after-added-token and slow-vs-fast tokenizer inconsistency are related to tokenization behavior, but not the same concrete bug." } ] }, @@ -5705,13 +5703,11 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ - "issue:28218", - "issue:29159", "issue:34634", "issue:36331", "issue:42175", @@ -5720,6 +5716,7 @@ "issue:43452", "issue:43526", "issue:43550", + "issue:43577", "issue:43582", "issue:43638", "issue:43653", @@ -5761,8 +5758,8 @@ ], "soft_pairs": [ "issue:44610|issue:45072", - "issue:28218|issue:29159", "issue:43720|issue:45072", + "issue:43577|issue:43582", "issue:44671|issue:45701", "issue:43825|issue:43827", "issue:43782|issue:45325", @@ -5772,205 +5769,205 @@ "issue:43262|issue:43819", "issue:43906|issue:46209", "issue:45003|issue:46285", + "issue:34634|issue:36331", "issue:43653|issue:44912", "issue:43526|issue:43697", - "issue:34634|issue:36331", "issue:45081|issue:45406", "issue:44060|issue:44368", - "issue:45372|issue:46326", "issue:43742|issue:45020", + "issue:45372|issue:46326", "issue:45084|issue:45538", + "issue:44589|issue:45042", "issue:43526|issue:44554", "issue:43329|issue:45600", - "issue:44589|issue:45042", "issue:43452|issue:44991", - "issue:43550|issue:44912", - "issue:43638|issue:44936" + "issue:43638|issue:44936", + "issue:43550|issue:44912" ], - "prepared_review_unit_hash": "ee53b06e5cb52fb841245ba20e8efff6f0b2a3d6e2edab5ee00fbdc46f0a4859", + "prepared_review_unit_hash": "cbb1a42efaaa0bdf8de71b73d9da3aee52abce27a9bdd8065bab636d8d9303cb", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35802, - "estimated_input_tokens": 8951, - "estimated_eval_tokens": 18158 + "serialized_chars": 35295, + "estimated_input_tokens": 8824, + "estimated_eval_tokens": 17904 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: most items are unrelated transformer bugs spanning tokenizers, model loading, quantization, training, audio/vision processors, and backend-specific failures. None of the soft edges look like the same concrete underlying bug or a single PR-worthy fix.", - "confidence": 0.9, - "canonical_issue_reason": "No single issue is a safe canonical for the whole cluster; the items do not form one duplicate bug family.", + "summary": "This cluster is a grab-bag of largely unrelated Transformers bugs; none of the soft pairs look like true duplicates. The closest umbrella issue is the broad remote_code/model-loading regression report, but it does not absorb the rest of the set.", + "confidence": 0.22, + "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the set and the best anchor for related loading/regression triage.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue stands out because the cluster is not a coherent duplicate set.", + "best_issue_reason": "issue:45020 is the most generally reusable triage target because it describes the widest failure class; the other issues are mostly narrow, model-specific bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "Different failures: image processor input-size mismatch vs bfloat16 dtype mismatch." + "reason": "Different failures: processor image-size mismatch vs dtype mismatch in inference." }, { - "left": "issue:28218", - "right": "issue:29159", + "left": "issue:43720", + "right": "issue:45072", "accept": false, - "reason": "Both are tokenizer-related, but one is added-token spacing and the other is slow/fast inconsistency; not the same bug." + "reason": "BitNet packed-weight loading bug is unrelated to the dtype mismatch report." }, { - "left": "issue:43720", - "right": "issue:45072", + "left": "issue:43577", + "right": "issue:43582", "accept": false, - "reason": "Unrelated: packed-weight loading failure vs dtype mismatch in inference." + "reason": "Blip2 dtype initialization and Apple Silicon allocator warmup are separate code paths." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "One is a CamemBERT masked-LM regression; the other is broad tokenization changes across versions." + "reason": "One is a CamemBERT masked-LM regression; the other is a broad tokenization-change report." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "Runtime error-message bug and documentation staleness are different fixes, not one underlying defect." + "reason": "Related theme, but one is an error-message bug and the other is docs referencing removed APIs." }, { "left": "issue:43782", "right": "issue:45325", "accept": false, - "reason": "Different Qwen VL models and different code paths: weight_only loading vs rope/position-id scaling." + "reason": "Different Qwen VL bugs: weight-only loading vs rope-index scaling." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Both involve loading, but one is FSDP CPU RAM efficient loading and the other is init_empty_weights argument handling." + "reason": "Both are loading-related, but they hit different mechanisms and should not be merged." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Package/backend installation issue vs tokenizer regex crash; no shared concrete bug." + "reason": "TensorFlow backend packaging and Mistral tokenizer regex crashing are unrelated." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "Apple Silicon allocator TypeError and CUDA AMP crash are unrelated backend-specific failures." + "reason": "Allocator warmup TypeError is not the same as the CUDA AMP crash in LwDetrImageLoss." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio chat-template sampling-rate default vs DAC latent reconstruction mismatch; different subsystems." + "reason": "Audio chat-template sampling-rate default and DAC latent reconstruction are different bugs." }, { "left": "issue:43906", "right": "issue:46209", "accept": false, - "reason": "An isolated reproduction of an older issue is not the same as an FP8 B200 inference bug." + "reason": "Isolated reproduction of an older issue is unrelated to FP8 inference on B200." }, { "left": "issue:45003", "right": "issue:46285", "accept": false, - "reason": "Unsafe sys.modules access and CLIP text-model structure inconsistency are different model-loading problems." + "reason": "Unsafe sys.modules access and CLIP text-model structure inconsistency are different defects." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Bark voice preset and CustomTrainer loss signature are unrelated." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "Tokenizer special-token registration bug vs MXFP4 quantization fallback; unrelated." + "reason": "Tokenizer special-token registration and MXFP4 quantization fallback are unrelated." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "BEiT label reduction bug vs RTDetrV2 output drift; different model families and symptoms." - }, - { - "left": "issue:34634", - "right": "issue:36331", - "accept": false, - "reason": "BarkProcessor voice_preset failure and Trainer compute_loss signature mismatch are unrelated." + "reason": "BEiT label reduction bug and RTDetrV2 output drift are distinct model issues." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Tokenizer regex patch crash vs Gemma4Processor missing _tokenizer; distinct code paths." + "reason": "Both involve tokenizer/processor loading, but the crash causes and affected components differ." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Different warning conditions and different Qwen model variants; not the same bug." + "reason": "These are similar warning reports, but they describe different tied-weights problems on different Qwen variants." }, { - "left": "issue:45372", - "right": "issue:46326", + "left": "issue:43742", + "right": "issue:45020", "accept": false, - "reason": "Mistral_common import error in Gemma 4 loading is separate from continue_final_message behavior." + "reason": "MobileLLM key error is too specific to be the same as the broad remote_code regression." }, { - "left": "issue:43742", - "right": "issue:45020", + "left": "issue:45372", + "right": "issue:46326", "accept": false, - "reason": "Model-specific key error loading MobileLLM-125M vs broad remote_code breakage." + "reason": "Both mention mistral_common, but one is an ImportError during Gemma 4 loading and the other is a continue_final_message bug." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template compilation TypeError and CLIPTokenizer max_length issue are unrelated." + "reason": "Template-node compilation failure and CLIPTokenizer max-length behavior are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage import failure and PIL/torchvision dependency handling are unrelated." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "Image processor label reduction bug is unrelated to the MPS attention correctness issue." + "reason": "BEiT label reduction and an MPS attention correctness bug do not match." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Multimodal token counting bug and stale auto_mappings config reference are different defects." - }, - { - "left": "issue:44589", - "right": "issue:45042", - "accept": false, - "reason": "Missing Float8 storage object and unwanted torchvision dependency are unrelated." + "reason": "Undefined video-token helper variables and removed auto-mapping config references are separate issues." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "gguf_file loader breakage and tokenizer load failure are both loading-related but not the same concrete bug." + "reason": "Both involve model/tokenizer loading, but the failing paths and root causes differ." }, { - "left": "issue:43550", - "right": "issue:44912", + "left": "issue:43638", + "right": "issue:44936", "accept": false, - "reason": "torch.compile+SDPA failure is unrelated to MXFP4 quantization fallback." + "reason": "DeepSpeed zero3 index error and trainer.evaluate() after train are different trainer bugs." }, { - "left": "issue:43638", - "right": "issue:44936", + "left": "issue:43550", + "right": "issue:44912", "accept": false, - "reason": "DeepSpeed ZeRO-3 loading IndexError and train/evaluate lifecycle failure are different trainer bugs." + "reason": "SDPA torch.compile failure and MXFP4 quantization fallback are not the same bug." } ] }, @@ -5989,8 +5986,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -6046,16 +6043,16 @@ "soft_pairs": [ "issue:43746|issue:44811", "issue:43819|issue:43927", - "issue:43650|issue:43906", - "issue:43257|issue:43329", - "issue:44387|issue:44617", "issue:36010|issue:43645", + "issue:43257|issue:43329", "issue:42175|issue:46291", + "issue:44387|issue:44617", + "issue:43650|issue:43906", "issue:38175|issue:44610", "issue:43720|issue:44610", "issue:36010|issue:36331", - "issue:43854|issue:44977", "issue:43531|issue:45042", + "issue:43854|issue:44977", "issue:39692|issue:44912", "issue:46421|issue:46424", "issue:45459|issue:45803", @@ -6063,219 +6060,219 @@ "issue:43262|issue:44811", "issue:43257|issue:43696", "issue:43643|issue:45759", - "issue:43454|issue:43577", "issue:43329|issue:44479", "issue:43329|issue:43450", "issue:43749|issue:44756", "issue:43257|issue:43931", + "issue:43454|issue:43577", "issue:34567|issue:40990", "issue:43606|issue:45859", "issue:42915|issue:43742", "issue:44291|issue:45042", "issue:29159|issue:34567" ], - "prepared_review_unit_hash": "1628bbd263bdd42b37964f23febf67276ae804dfa69456652728dc1dc6e35369", + "prepared_review_unit_hash": "bf1abea57acbdb3e9bec7d8945c75af7db029eddfe0617d0fd7443996552c481", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36537, - "estimated_input_tokens": 9135, - "estimated_eval_tokens": 18526 + "serialized_chars": 36532, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are a heterogeneous set of unrelated issues across tokenization, training, model loading, multimodal/video, quantization, and caching. There is no single underlying bug cluster, so no canonical duplicate stands out and every soft edge should be rejected.", - "confidence": 0.96, - "canonical_issue_reason": "No canonical issue: these reports cover distinct failures in different subsystems and models rather than one shared bug.", + "summary": "This looks like a loose embedding cluster of unrelated Transformers issues. The soft pairs mostly share only broad themes (tokenization, loading, multimodal/video, OOM, CPU offload) but not the same concrete bug or code path, so none should be merged as duplicates. No PRs are present in the cluster.", + "confidence": 0.95, + "canonical_issue_reason": "If a single anchor is needed, issue 43257 is the best representative because it has the strongest discussion/reference activity and is a concrete, actionable model-loading regression. That said, the cluster itself is not a true duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a good cluster representative; the list is too diverse to treat any one item as the duplicate anchor.", + "best_issue_reason": "Issue 43257 is the most suitable global issue anchor: it is concrete, well-discussed, and closer to the dominant loading/regression theme than the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "Different components and failures: PEFT/local checkpoint loading vs Whisper processor decoding options." + "reason": "Different subsystems and symptoms: local checkpoint/PEFT loading vs Whisper batch_decode token filtering." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "Unrelated model logic: DAC latent reconstruction mismatch vs DiaConfig losing custom token IDs." + "reason": "Unrelated bugs: DAC latent reconstruction mismatch vs DiaConfig losing custom token IDs after save/load." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:36010", + "right": "issue:43645", "accept": false, - "reason": "A placeholder/data issue is unrelated to an isolated reproduction of another upstream bug." + "reason": "Both touch model import/init behavior, but one is a GenerationMixin import error and the other is a Jupyter custom-model initialization regression." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Different Qwen3-MoE loading problem vs multimodal video token-count helper bug." + "reason": "Different concrete failures: Qwen3 MoE weight conversion under accelerate/deepspeed vs undefined variables in multimodal video token counting." }, { - "left": "issue:44387", - "right": "issue:44617", + "left": "issue:42175", + "right": "issue:46291", "accept": false, - "reason": "Both mention OOM, but one is int4 quantization memory regression and the other is a model-specific video inference OOM." + "reason": "Different dependency/import problems: pip extras missing tensorflow backend vs pipelines import failing due to kernels package requirements." }, { - "left": "issue:36010", - "right": "issue:43645", + "left": "issue:44387", + "right": "issue:44617", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to notebook custom-model initialization behavior." + "reason": "Both mention OOM, but the underlying code paths are different: int4 quantization memory growth vs Sam3Video runtime OOM." }, { - "left": "issue:42175", - "right": "issue:46291", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "Package extra/backend composition issue is unrelated to a kernels dependency/version requirement on pipelines import." + "reason": "No clear bug overlap; one is a placeholder/data issue, the other is an isolated reproduction of a separate issue." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Different multimodal model bugs: zero probabilities vs processor producing the wrong input size." + "reason": "Different model-specific bugs: unexpected zero probabilities in SigLIP2 vs processor/model input-size mismatch in OmDet-Turbo." }, { "left": "issue:43720", "right": "issue:44610", "accept": false, - "reason": "Packed-weight loading in BitNet is unrelated to OmDet-Turbo processor image-size mismatch." + "reason": "Packed-weight unpacking during accelerate loading is unrelated to OmDet-Turbo processor shape mismatch." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "Importing GenerationMixin and an unexpected trainer keyword argument are separate API regressions." + "reason": "Different APIs and failures: GenerationMixin import resolution vs Trainer.compute_loss signature change." }, { - "left": "issue:43854", - "right": "issue:44977", + "left": "issue:43531", + "right": "issue:45042", "accept": false, - "reason": "Different models and symptoms: GLM-4.7 load failure in tests vs Qwen3.5 flash-attention generation issues." + "reason": "Different areas entirely: Qwen3-MoE sliding_window handling vs PIL image processor torchvision dependency regression." }, { - "left": "issue:43531", - "right": "issue:45042", + "left": "issue:43854", + "right": "issue:44977", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to PIL backend unexpectedly requiring torchvision." + "reason": "Both are model-loading/generation issues, but they affect different models and failure modes; not the same underlying bug." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "Doc-example/model-processor mismatch and MXFP4 quantization fallback are different bugs." + "reason": "Unrelated: documentation example errors for SigLIP2 vs MXFP4 quantization fallback on gpt-oss-20b." }, { "left": "issue:46421", "right": "issue:46424", "accept": false, - "reason": "Both are cache-related, but one is lazy-init recompilation and the other is max_cache_length handling in quantized cache." + "reason": "Both involve cache behavior, but one is lazy-init recompilation with chunked prefill while the other is max_cache_length being ignored except for quantized cache." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Both involve error handling, but the tokenizer protobuf path and Fuyu tensor conversion are different code paths." + "reason": "Both discuss bad exception handling, but they are different code paths and different failure surfaces." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Trailing-slash model-path warning is unrelated to Qwen3 MOE weight conversion during accelerate/deepspeed loading." + "reason": "No overlap: trailing-slash local-path warning vs Qwen3 MoE loading under accelerate/deepspeed." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio chat-template sampling rate defaults are unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Different audio processor issues: chat-template sample-rate default vs batch_decode ignoring skip_special_tokens." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MOE weight conversion and GPT-oss OOM are different bugs with different failure modes." + "reason": "Qwen3 MoE weight conversion failure is unrelated to GPT-oss-20b CUDA OOM." }, { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Both concern config/model construction, but they are different APIs and distinct missing-field vs text_config unwrap problems." - }, - { - "left": "issue:43454", - "right": "issue:43577", - "accept": false, - "reason": "AyaVision weight tying and Blip2 dtype propagation are unrelated model-loading issues." + "reason": "Both are config-loading issues, but one is missing fields with trust_remote_code and the other is unwrapping text_config for composite Qwen multimodal configs." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "The undefined video token helper bug is not clearly the same concrete regression as the broader v5.3.0 video-input failure." + "reason": "Related area (video multimodal), but not the same bug: undefined video-token helper variables vs a broader video input regression across models." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Undefined helper variables in video token counting and batched video shape mismatch are different problems." + "reason": "Different video bugs: token counting internals vs batched video processor output shape." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "FSDP RAM-efficient loading and Strix Halo mmap OOM are different memory/loading issues." + "reason": "Both concern loading/memory, but one is FSDP CPU RAM-efficient loading breakage and the other is mmap-related OOM on Strix Halo." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Qwen3 MOE loading under accelerate/deepspeed is unrelated to Qwen3-VL weight-shape mismatch." + "reason": "Different Qwen-family problems and different failure modes: MoE accelerate/deepspeed conversion vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "Different model internals: AyaVision lm_head tying bug vs Blip2 dtype propagation issue." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "Trainer token-count tracking and GPT-oss perplexity are unrelated." + "reason": "Trainer token counting bug is unrelated to GPT-oss-20b perplexity regression." }, { "left": "issue:43606", "right": "issue:45859", "accept": false, - "reason": "CPU offload incompatibility in Bark and Qwen3_5Moe rotary embedding are different model-specific bugs." + "reason": "Both mention CPU offload, but they affect different models and code paths; not a mergeable duplicate." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe is unrelated to a MobileLLM key error." + "reason": "Different loading failures: Qwen3Moe FineGrainedFP8Config vs MobileLLM key error." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "init_empty_weights argument handling is unrelated to PIL backend's torchvision dependency issue." + "reason": "Unrelated: init_empty_weights compatibility issue vs image-processor torchvision dependency regression." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to TrainerState token accounting not updating." + "reason": "Tokenizer fast/slow inconsistency is unrelated to TrainerState num_input_tokens_seen not updating." } ] }, @@ -6294,8 +6291,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -6326,6 +6323,7 @@ "issue:44062", "issue:44291", "issue:44479", + "issue:44488", "issue:44560", "issue:44568", "issue:44821", @@ -6348,244 +6346,251 @@ "issue:46512" ], "soft_pairs": [ - "issue:44568|issue:45356", "issue:43525|issue:46291", "issue:42915|issue:44821", + "issue:44568|issue:45356", "issue:43404|issue:45127", - "issue:43927|issue:44912", "issue:44987|issue:45538", + "issue:43927|issue:44912", "issue:43697|issue:45357", "issue:44849|issue:45701", "issue:42915|issue:45538", - "issue:43577|issue:46512", "issue:43425|issue:46291", + "issue:43577|issue:46512", "issue:43452|issue:43582", "issue:42175|issue:43824", - "issue:33357|issue:34567", "issue:43317|issue:45538", - "issue:41720|issue:43540", + "issue:33357|issue:34567", "issue:28282|issue:34689", "issue:45137|issue:45405", + "issue:41720|issue:43540", "issue:44062|issue:44991", "issue:44479|issue:45381", "issue:43540|issue:44479", - "issue:43540|issue:44560", "issue:43742|issue:44291", + "issue:43540|issue:44560", "issue:44568|issue:45488", "issue:28282|issue:29127", "issue:44987|issue:45907", "issue:24643|issue:29159", "issue:42915|issue:43749", - "issue:38175|issue:44898" + "issue:38175|issue:44898", + "issue:42175|issue:44488" ], - "prepared_review_unit_hash": "7363016808a8ba7f5085d7949be686bc5323b4425a63cf95382cafe4cccd78aa", + "prepared_review_unit_hash": "655cce5335cd01dae046c354f612cb14a0c3ef8f74a14f571c8361f69967e93d", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35935, - "estimated_input_tokens": 8984, - "estimated_eval_tokens": 18224 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36787, + "estimated_input_tokens": 9197, + "estimated_eval_tokens": 18650 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: tokenizer, loading, dtype, video-input, and backend bugs are mixed together. Only one soft pair looks like a true duplicate family, centered on the Qwen2.5-VL video-input regression.", - "confidence": 0.69, - "canonical_issue_reason": "Issue 44479 is the best canonical anchor because it is the broader umbrella report for the Qwen video-input regression family and plausibly subsumes 45381; the other items in the cluster do not form a coherent duplicate set.", + "summary": "This cluster is mostly noisy and spans many unrelated bugs (tokenizers, model loading, dtype propagation, quantization, training, and video processing). The only clearly overlapping pair is the Qwen VL video-input regression reports.", + "confidence": 0.89, + "canonical_issue_reason": "issue:44479 is the broadest and most representative bug report: it describes a shared video-input regression across multiple Qwen VL model variants, which matches the only likely duplicate in the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "44479 is the strongest representative issue in this mixed cluster: it is specific enough to be actionable, broad enough to cover the only clear duplicate-like pair, and earlier than the more specific follow-up report.", + "best_issue_reason": "issue:44479 is the best standalone issue because it is the most general and actionable description of the common underlying regression.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44568", - "right": "issue:45356", - "accept": false, - "reason": "Different tokenizer regressions: one is BOS/EOS insertion behavior, the other is codec/ warning handling for Kimi-K2.5." - }, { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "Unrelated failures: missing pad_token_id on Llama4Config vs kernels import requiring version/revision." + "reason": "Unrelated failures: Llama4 config attribute vs kernels version/revision requirement." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "Different subsystems and symptoms: FP8 config load failure vs loading AutoImageProcessor from a URL." + "reason": "Different subsystems and symptoms: FP8 config load error vs image processor URL loading." }, { - "left": "issue:43404", - "right": "issue:45127", + "left": "issue:44568", + "right": "issue:45356", "accept": false, - "reason": "Both mention weights/tied embeddings, but one is a missing tie in a multimodal model and the other is LoRA merge collapse after vocab extension; not the same bug." + "reason": "Both are tokenizer regressions, but they affect different models and different mechanics." }, { - "left": "issue:43927", - "right": "issue:44912", + "left": "issue:43404", + "right": "issue:45127", "accept": false, - "reason": "Config/token ID persistence bug vs MXFP4 quantization fallback bug; no shared code-path evidence." + "reason": "Both involve tied weights, but one is a Mistral3 model bug and the other is LoRA merge with extended vocabulary." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Model load regression vs CLIP tokenizer max_length constant; unrelated issues." + "reason": "Model-loading regression vs CLIPTokenizer max_length bug; no shared underlying code path." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Config save/load token IDs issue vs quantization fallback behavior." }, { "left": "issue:43697", "right": "issue:45357", "accept": false, - "reason": "Different regressions: RTDetr output drift in v5 vs incorrect visual encoder keys on save_pretrained." + "reason": "Different bugs: RTDetrV2 output mismatch vs Qwen3.5 save_pretrained key names." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "Qwen3.5 hidden-states bug is not the same as a broad version-dependent tokenization change." + "reason": "Qwen3.5 hidden-states bug is not the same as a broad tokenization-version change report." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "FP8 model loading problem vs tokenizer max_length issue; no common bug." + "reason": "FineGrainedFP8Config failure is unrelated to CLIPTokenizer model_max_length." }, { - "left": "issue:43577", - "right": "issue:46512", + "left": "issue:43425", + "right": "issue:46291", "accept": false, - "reason": "Both involve dtype, but one is weights staying float32 after load while the other is stale config metadata after from_config; distinct fixes." + "reason": "Torch compatibility issue vs kernels package version/revision requirement." }, { - "left": "issue:43425", - "right": "issue:46291", + "left": "issue:43577", + "right": "issue:46512", "accept": false, - "reason": "Torch compatibility problem vs kernels version/revision requirement; not the same defect." + "reason": "Related dtype theme, but one is Blip2 loading and the other is generic from_config config staleness." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "Tokenizer/model gguf loading bug vs Apple Silicon allocator warmup TypeError; unrelated." + "reason": "Tokenizer/model loading issue vs Apple Silicon allocator TypeError." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Package/backend installation issue vs missing model import; different failure modes." - }, - { - "left": "issue:33357", - "right": "issue:34567", - "accept": false, - "reason": "MacOS bus error on CLIP model is unrelated to TrainerState token counter not updating." + "reason": "Packaging/backend inclusion bug vs missing class import error." }, { "left": "issue:43317", "right": "issue:45538", "accept": false, - "reason": "GPU/CPU offload loading failure vs tokenizer max_length constant; no overlap in bug." + "reason": "Dequantized device_map offload bug is unrelated to CLIPTokenizer max_length." }, { - "left": "issue:41720", - "right": "issue:43540", + "left": "issue:33357", + "right": "issue:34567", "accept": false, - "reason": "Different Qwen3-family issues: auto device mapping CUDA assert vs video-input ValueError." + "reason": "Mac bus error in model loading vs TrainerState counter not updating." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Missing PyTorch import error vs a specific model-loading regression for Llama 3.2 Vision; not duplicates." + "reason": "Generic missing-PyTorch import error vs a specific Llama 3.2 Vision model loading regression." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error is unrelated to the PEFT version bump issue." + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT version bump." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both mention Qwen and runtime failures, but one is auto device mapping CUDA assert and the other is video input processing." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "AddedToken constructor argument collision vs tokenizer load failure for a specific model; different root causes." + "reason": "AddedToken keyword collision vs tokenizer loading for a specific model." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "Both describe the same Qwen2.5-VL video-input regression in v5.3.0; 45381 is a specific manifestation of the broader bug reported in 44479." + "reason": "Both describe the same Qwen VL video-input regression in v5.3.0; 45381 is a specific symptom of the broader regression in 44479." }, { "left": "issue:43540", "right": "issue:44479", "accept": false, - "reason": "Both concern video inputs, but they hit different Qwen model families and different failure paths." + "reason": "Different Qwen model families and different failures; too broad to be the same bug." }, { - "left": "issue:43540", - "right": "issue:44560", + "left": "issue:43742", + "right": "issue:44291", "accept": false, - "reason": "Two different Qwen3OmniMoe video-input failures with different symptoms; not enough evidence they share one root cause." + "reason": "Loading MobileLLM-125M key error is unrelated to init_empty_weights argument handling." }, { - "left": "issue:43742", - "right": "issue:44291", + "left": "issue:43540", + "right": "issue:44560", "accept": false, - "reason": "Model key-loading error vs init_empty_weights unexpected argument; separate loading bugs." + "reason": "Both are video-related Qwen issues, but the affected models and failure modes differ." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Tokenizer special-token insertion bug vs a LlamaTokenizer pre-tokenizer override regression; related theme but different bug." + "reason": "Tokenizer special-token bug vs LlamaTokenizer pre-tokenizer override regression." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "Environment import error vs LayoutLMv3 error-message clarity issue; not duplicates." + "reason": "Missing PyTorch import error vs LayoutLMv3 error-message clarity issue." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "Model loading regression vs list_repo_files revision forwarding; different code paths." + "reason": "Different API regressions: model loading vs list_repo_files not forwarding revision." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed training weight-shape error is unrelated to tokenizer slow/fast inconsistency." + "reason": "DeepSpeed weight shape error vs tokenizer fast/slow inconsistency." }, { "left": "issue:42915", "right": "issue:43749", "accept": false, - "reason": "Different training/loading bugs: Qwen3Moe FP8 config failure vs FSDP CPU RAM efficient loading regression." + "reason": "FP8 config failure is unrelated to FSDP CPU RAM efficient loading." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Different vision-model bugs: zero probabilities in SigLIP2 vs Perceiver non-default resolution failure." + "reason": "Siglip2 zero probabilities and Perceiver non-default resolution failure are different model bugs." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Backend packaging issue vs model loading failure for a specific checkpoint." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is well grounded in the packet: 44479 and 45381 are the only convincingly related pair, and the other soft pairs are treated conservatively with clear reasons for non-identity. The umbrella/canonical framing for 44479 is supported by its broader Qwen video-input regression title and earlier date relative to 45381." + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported overlap is issue:44479 with issue:45381, and the analyst correctly treats the rest as unrelated or only superficially similar. No soft-edge verdict appears overstated." }, "evaluator_used": true, "retried": false, @@ -6601,8 +6606,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -6610,7 +6615,6 @@ "issue:28282", "issue:34567", "issue:36331", - "issue:42175", "issue:42796", "issue:43257", "issue:43404", @@ -6619,8 +6623,7 @@ "issue:43526", "issue:43531", "issue:43577", - "issue:43643", - "issue:43697", + "issue:43720", "issue:43742", "issue:43756", "issue:43906", @@ -6630,18 +6633,19 @@ "issue:44488", "issue:44568", "issue:44589", + "issue:44625", "issue:44737", "issue:44792", "issue:44821", + "issue:44877", + "issue:44898", "issue:44913", "issue:44964", "issue:45042", "issue:45072", "issue:45081", - "issue:45216", "issue:45325", "issue:45357", - "issue:45375", "issue:45405", "issue:45479", "issue:45588", @@ -6657,236 +6661,223 @@ ], "soft_pairs": [ "issue:44737|issue:46463", - "issue:42175|issue:44488", - "issue:43577|issue:45072", "issue:43493|issue:43994", - "issue:43257|issue:43756", "issue:44291|issue:44821", - "issue:43906|issue:46468", + "issue:43257|issue:43756", "issue:34567|issue:36331", + "issue:43577|issue:45072", + "issue:43906|issue:46468", "issue:27957|issue:28282", "issue:44062|issue:45081", "issue:44568|issue:46491", - "issue:44737|issue:45588", "issue:44792|issue:44964", "issue:43404|issue:45987", - "issue:43756|issue:46399", "issue:43526|issue:45479", + "issue:44737|issue:45588", + "issue:45042|issue:45081", "issue:42796|issue:43906", + "issue:43742|issue:44589", + "issue:43756|issue:46399", "issue:43756|issue:45325", "issue:45823|issue:46463", - "issue:45042|issue:45081", - "issue:43742|issue:44589", "issue:43452|issue:44488", "issue:43531|issue:44821", "issue:45405|issue:45600", "issue:45357|issue:45920", "issue:44913|issue:46121", - "issue:43643|issue:45375", - "issue:43697|issue:45216" + "issue:44625|issue:44877", + "issue:43720|issue:44898" ], - "prepared_review_unit_hash": "8a1d520514291394ae51a608e09e33f88cf154ac601c11917fb8b7fe85f7f7e7", + "prepared_review_unit_hash": "266d99a667e865cbda1096fcafd69142cacca2b37853d0ae8bcacf4997a925aa", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36556, - "estimated_input_tokens": 9139, - "estimated_eval_tokens": 18534 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35635, + "estimated_input_tokens": 8909, + "estimated_eval_tokens": 18074 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issue set with one plausible duplicate pair in SigLIP2; the rest of the soft edges are too broad or model-specific to merge safely.", - "confidence": 0.92, - "canonical_issue_reason": "No single global canonical issue: the cluster spans many unrelated bugs. If a representative were needed, issue 43493 is the clearest root report for the only overlapping SigLIP2 pair.", + "summary": "Mostly unrelated bug reports; the only soft duplicate candidate I\u2019d accept is the SigLIP2 pair, where both issues point to the same model implementation producing wrong outputs. All other pairs look like different code paths, models, or failure modes.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43493 is the best canonical issue because it describes the underlying SigLIP2 implementation discrepancy against the original JAX model, while the paired report is a downstream symptom of the same problem.", "canonical_pr_reason": null, - "best_issue_reason": "No clear global best issue because the cluster is not one problem; issue 43493 is only the best representative for the accepted SigLIP2 overlap, not for the whole set.", + "best_issue_reason": "issue:43493 is the strongest representative of the cluster\u2019s only true duplicate thread: it is the most root-cause-focused and specific report for the SigLIP2 bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44737", "right": "issue:46463", "accept": false, - "reason": "Different model/code-path bugs: XLNet CPU device placement vs T5 decoder-input validation messaging." - }, - { - "left": "issue:42175", - "right": "issue:44488", - "accept": false, - "reason": "Packaging/backend availability issue versus a specific model load failure; not the same bug." - }, - { - "left": "issue:43577", - "right": "issue:45072", - "accept": false, - "reason": "Both mention dtype, but they affect different models and inference paths." + "reason": "Different bugs: XLNet CPU device placement in rotary encoding vs T5 decoder input validation error message." }, { "left": "issue:43493", "right": "issue:43994", "accept": true, - "reason": "Both report SigLIP2 producing incorrect outputs; one is a reference-vs-HF discrepancy, the other a concrete bad-output symptom on the same model family." + "reason": "Both report SigLIP2 producing incorrect results; 43493 frames the HF-vs-JAX implementation mismatch, and 43994 is the user-facing symptom in AutoModel/pipeline." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Unrelated failures: init_empty_weights/_is_hf_initialized loading error vs AutoImageProcessor URL loading." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "Different model families and failures: MoE weight conversion vs RoPE layer dropping." + "reason": "Different models and failures: Qwen3 MOE loading/conversion vs Smollm3 RoPE-layer handling." }, { - "left": "issue:44291", - "right": "issue:44821", + "left": "issue:34567", + "right": "issue:36331", "accept": false, - "reason": "Unexpected init_empty_weights argument vs loading an image processor from URL are unrelated." + "reason": "Both trainer-related, but distinct issues: TrainerState token counting vs custom compute_loss signature mismatch." }, { - "left": "issue:43906", - "right": "issue:46468", + "left": "issue:43577", + "right": "issue:45072", "accept": false, - "reason": "An isolated reproduction of one issue and an example-breakage report are not clearly the same underlying bug." + "reason": "Separate dtype bugs in different model families; no shared concrete code path." }, { - "left": "issue:34567", - "right": "issue:36331", + "left": "issue:43906", + "right": "issue:46468", "accept": false, - "reason": "Trainer state update bug versus a custom loss API signature break." + "reason": "Different reports for different models/examples; too little evidence of the same underlying bug." }, { "left": "issue:27957", "right": "issue:28282", "accept": false, - "reason": "FlashAttention/model support issue versus a missing PyTorch dependency error." + "reason": "XLMRoberta flash-attention issue is unrelated to AutoModel PyTorch import detection." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Different tokenizer failures in different code paths; only broad tokenizer similarity." + "reason": "Both tokenizer-related, but one is AddedToken construction and the other is Mistral regex patching; distinct root causes." }, { "left": "issue:44568", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they concern different models and distinct round-trip/add-special-tokens behavior." - }, - { - "left": "issue:44737", - "right": "issue:45588", - "accept": false, - "reason": "XLNet positional encoding device bug is unrelated to flash-attention sink handling." + "reason": "Different tokenizer regressions: special-token insertion vs CodeLlama encode/decode space preservation." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Different multimodal model failures with different symptoms and likely different fixes." + "reason": "Different multimodal model failures; no indication they share the same code path." }, { "left": "issue:43404", "right": "issue:45987", "accept": false, - "reason": "Weight tying in Mistral3 is unrelated to StaticCache seq-length return type." + "reason": "Mistral3 lm_head tying bug is unrelated to StaticCache sequence-length return type." }, { - "left": "issue:43756", - "right": "issue:46399", + "left": "issue:43526", + "right": "issue:45479", "accept": false, - "reason": "Different Qwen/Smollm bugs: RoPE layer count vs kernel registration/decorator error." + "reason": "Different classification/processor issues with different affected components and symptoms." }, { - "left": "issue:43526", - "right": "issue:45479", + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU tensor creation is unrelated to flash_attention sink-less model handling." + }, + { + "left": "issue:45042", + "right": "issue:45081", "accept": false, - "reason": "Image processor label reduction bug is not the same as classification loss degeneracy." + "reason": "Image processor torchvision dependency regression is unrelated to Mistral tokenizer regex patching." }, { "left": "issue:42796", "right": "issue:43906", "accept": false, - "reason": "Tokenizer BPE-marker handling and an unrelated isolated reproduction do not match." + "reason": "Ministral tokenizer BPE-marker handling is unrelated to an isolated reproduction of another issue." }, { - "left": "issue:43756", - "right": "issue:45325", + "left": "issue:43742", + "right": "issue:44589", "accept": false, - "reason": "Different RoPE-related model bugs with different models and mechanics." + "reason": "Model loading key error and Float8 storage lookup failure are different loading paths and different failures." }, { - "left": "issue:45823", - "right": "issue:46463", + "left": "issue:43756", + "right": "issue:46399", "accept": false, - "reason": "Device-map forward mismatch versus decoder-input validation are unrelated." + "reason": "Smollm3 RoPE-layer discrepancy and Qwen3.5 kernel registration crash are separate model-specific bugs." }, { - "left": "issue:45042", - "right": "issue:45081", + "left": "issue:43756", + "right": "issue:45325", "accept": false, - "reason": "Image-processor torchvision dependency regression versus Mistral regex patch crash." + "reason": "Both mention RoPE, but they affect different models and different logic (layer selection vs temporal position scaling)." }, { - "left": "issue:43742", - "right": "issue:44589", + "left": "issue:45823", + "right": "issue:46463", "accept": false, - "reason": "Model-loading key error and missing Float8 storage class are unrelated." + "reason": "Gemma4 PLE device mismatch is unrelated to T5 decoder input validation." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "GGUF loading breakage and a specific model load failure do not look like the same bug." - }, - { - "left": "issue:43531", - "right": "issue:44821", - "accept": false, - "reason": "Qwen3-MoE sliding-window bug versus AutoImageProcessor URL loading are different areas." + "reason": "gguf_file loading regression and a specific model load failure are different issues without a clear shared root cause." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "PEFT version bump issue and stale auto_mappings references are unrelated CI/config problems." + "reason": "PEFT version bump and stale auto_mappings references are unrelated maintenance/CI problems." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Qwen3.5 save_pretrained visual-key regression is unrelated to tokenizer token-ID regression across models." + "reason": "Qwen3.5 save_pretrained key regression is unrelated to tokenizer producing wrong token IDs." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "GPTNeoX config reload defaulting and rope-parameter validation TypeError are different bugs." + "reason": "GPTNeoX rotary_pct reload bug and rope-parameter JSON list handling are distinct config/validation issues." }, { - "left": "issue:43643", - "right": "issue:45375", + "left": "issue:44625", + "right": "issue:44877", "accept": false, - "reason": "Generic trust_remote_code field loss and a specific config field dropped by @strict are not the same code-path." + "reason": "num_labels propagation bug and strict granite_speech config loading are different configuration problems." }, { - "left": "issue:43697", - "right": "issue:45216", + "left": "issue:43720", + "right": "issue:44898", "accept": false, - "reason": "RTDetrV2 output differences and Qwen3.5 save_pretrained regression are unrelated." + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to Perceiver interpolation failures." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The heterogeneous-cluster conclusion fits the issue titles, and the lone accepted SigLIP2 pair is a plausible duplicate/overlap without overstating certainty. The rejected soft edges are generally explained with model- or code-path-specific differences, which is appropriately cautious." + "feedback": "The summary is grounded in the packet and stays mostly conservative. The only accepted soft edge, issue:43493 vs issue:43994, is reasonable from the titles alone because both concern SigLIP2 and the second appears to be a symptom/report of wrong outputs from the same model family. The other rejections are also cautious and consistent with the packet. One minor note: phrases like \"only true duplicate thread\" and \"same underlying bug\" are a bit stronger than the evidence strictly proves, but not enough to overturn the overall assessment." }, "evaluator_used": true, "retried": false, @@ -6902,8 +6893,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -6917,12 +6908,14 @@ "issue:42898", "issue:43295", "issue:43452", + "issue:43531", "issue:43577", + "issue:43643", "issue:43646", "issue:43653", "issue:43688", "issue:43697", - "issue:43720", + "issue:43742", "issue:43749", "issue:43819", "issue:44361", @@ -6933,19 +6926,18 @@ "issue:44493", "issue:44514", "issue:44568", - "issue:44625", "issue:44704", "issue:44743", "issue:44811", "issue:44843", - "issue:44877", - "issue:44898", "issue:44936", "issue:44991", "issue:45005", "issue:45200", + "issue:45216", "issue:45356", "issue:45372", + "issue:45375", "issue:45381", "issue:45542", "issue:45823", @@ -6956,238 +6948,248 @@ "issue:46491" ], "soft_pairs": [ - "issue:43720|issue:44898", - "issue:44625|issue:44877", + "issue:43697|issue:45216", + "issue:43643|issue:45375", "issue:45200|issue:46463", "issue:24643|issue:29127", "issue:41762|issue:44936", "issue:44493|issue:46050", "issue:44466|issue:45005", - "issue:42898|issue:46491", "issue:28282|issue:32090", + "issue:42898|issue:46491", "issue:43819|issue:44811", "issue:43452|issue:44843", "issue:44514|issue:45381", "issue:28282|issue:33357", - "issue:43688|issue:43749", "issue:44936|issue:45823", - "issue:45372|issue:46170", - "issue:38175|issue:43295", + "issue:43688|issue:43749", "issue:44361|issue:44442", + "issue:38175|issue:43295", + "issue:45372|issue:46170", "issue:44484|issue:44704", "issue:43653|issue:44361", "issue:45542|issue:46178", - "issue:43452|issue:43577", "issue:43295|issue:45356", + "issue:43452|issue:43577", "issue:43577|issue:43697", "issue:43653|issue:44568", "issue:44464|issue:44743", + "issue:43531|issue:43742", "issue:43646|issue:43749", "issue:44843|issue:44991" ], - "prepared_review_unit_hash": "fd1ec48830ce32fd6c30494e410cf91144e0b18c424398cbe3f814578aa7568b", + "prepared_review_unit_hash": "5a4745431ab199b1c6b0901814381df2eeab557d9f60010565522aa28eae5822", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35873, - "estimated_input_tokens": 8969, - "estimated_eval_tokens": 18194 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36729, + "estimated_input_tokens": 9183, + "estimated_eval_tokens": 18622 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers bug reports and regressions across tokenizers, configs, loaders, trainers, and model-specific code paths. I don\u2019t see a true duplicate group here, so no soft pair is safe to merge.", - "confidence": 0.83, - "canonical_issue_reason": "issue:43646 is the broadest, most general regression in the set (custom model initialization in v5), so it is the closest thing to a representative issue, though it does not actually cover the others.", + "summary": "This cluster is mostly heterogeneous: many separate v5 regressions, tokenizer/processor edge cases, and training/loading bugs. The only plausible duplicate-style pair is the tied-weights serialization reports; the rest look distinct.", + "confidence": 0.72, + "canonical_issue_reason": "issue:43646 is the broadest umbrella regression report in the set and the best single triage anchor, but it should not be treated as proof that the whole cluster is one duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the most representative by scope, but the cluster is not cohesive enough to treat it as a real canonical duplicate.", + "best_issue_reason": "issue:43646 is the most general, high-signal issue here; it best represents the cluster\u2019s broad v5 regression theme without being overly specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43720", - "right": "issue:44898", + "left": "issue:43697", + "right": "issue:45216", "accept": false, - "reason": "Different models and failure modes: BitNet accelerate loading vs Perceiver image classification." + "reason": "Different model families and different failure modes: output drift in RTDetrV2 vs checkpoint save corruption in Qwen3.5." }, { - "left": "issue:44625", - "right": "issue:44877", + "left": "issue:43643", + "right": "issue:45375", "accept": false, - "reason": "Both are config-related, but one is num_labels propagation and the other is strict config loading for granite_speech." + "reason": "Both involve config/strict handling, but one is missing fields from trust_remote_code and the other drops a specific vision config field; not the same bug." }, { "left": "issue:45200", "right": "issue:46463", "accept": false, - "reason": "Gemma 4 multimodal token-type defaults vs T5 decoder input validation; unrelated code paths." + "reason": "Gemma 4 multimodal token defaults and T5 decoder-input validation are unrelated code paths and symptoms." }, { "left": "issue:24643", "right": "issue:29127", "accept": false, - "reason": "DeepSpeed training weight shape error vs LayoutLMv3 box-info error-message clarity." + "reason": "DeepSpeed training crash and a LayoutLMv3 error-message clarity issue are not the same underlying problem." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "Gemma3 ZeRO-3 loading crash and trainer evaluate-after-train are different bugs." + "reason": "ZeRO-3 load-time index error for Gemma3 is unrelated to trainer.evaluate() failing after train()." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Unexpected position-id keys during loading is not the same as quantized model save/load breakage." + "reason": "Unexpected position-id keys and sinq save/load corruption are different serialization/model-state bugs." }, { "left": "issue:44466", "right": "issue:45005", - "accept": false, - "reason": "Both mention tied weights, but one is serialization behavior and the other is translation-model tied-weight issues; not the same concrete bug." + "accept": true, + "reason": "Both describe the same tied-weights serialization regression in v5, with 44466 as a concrete device-dependent instance and 45005 as the broader report." }, { - "left": "issue:42898", - "right": "issue:46491", + "left": "issue:28282", + "right": "issue:32090", "accept": false, - "reason": "Both are tokenizer regressions, but clean_up_tokenization_spaces and CodeLlama leading-space round-trip are different behaviors." + "reason": "Importing AutoModel without PyTorch and a NoneType GPU broadcast error in Trainer are unrelated failures." }, { - "left": "issue:28282", - "right": "issue:32090", + "left": "issue:42898", + "right": "issue:46491", "accept": false, - "reason": "Missing torch import dependency vs NoneType broadcast failure in Trainer/DeepSpeed." + "reason": "clean_up_tokenization_spaces behavior change and CodeLlama leading-space round-trip loss are different tokenizer regressions." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC from_latents mismatch and Whisper batch_decode skip_special_tokens are unrelated model methods." + "reason": "DAC latent decoding mismatch and Whisper batch_decode skipping special tokens are separate model-specific bugs." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "gguf_file loading breakage is distinct from offline-mode model_info() calls in _patch_mistral_regex." + "reason": "Both are tokenizer loading regressions, but one is gguf_file support and the other is an offline hub call in _patch_mistral_regex." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Same family, but batched chat template padding crash and wrong video vision_position_ids are different bugs." + "reason": "Processor batched-template crashing and video vision_position_ids being wrong are distinct Qwen2.5-VL issues." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "Torch not found ImportError is unrelated to MacOS bus error on CLIP." + "reason": "Missing PyTorch import support and a MacOS bus error on CLIP are not the same failure." }, { - "left": "issue:43688", - "right": "issue:43749", + "left": "issue:44936", + "right": "issue:45823", "accept": false, - "reason": "Auxiliary-loss normalization and FSDP CPU RAM efficient loading are unrelated." + "reason": "Trainer lifecycle failure and Gemma4 device-map mismatch are different execution paths and symptoms." }, { - "left": "issue:44936", - "right": "issue:45823", + "left": "issue:43688", + "right": "issue:43749", "accept": false, - "reason": "Trainer lifecycle failure and Gemma4 device mismatch are different code paths." + "reason": "Auxiliary-loss normalization and FSDP CPU RAM-efficient loading are unrelated bugs." }, { - "left": "issue:45372", - "right": "issue:46170", + "left": "issue:44361", + "right": "issue:44442", "accept": false, - "reason": "Gemma 4 processor import regression from mistral_common differs from Windows/ROCm torch.distributed dependency issues." + "reason": "MLukeTokenizer AttributeError and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are separate tokenizer issues." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Zero probabilities in SigLIP2 is unrelated to processor.tokenizer regression in custom model code." + "reason": "Unexpected zero probabilities in SigLIP2 and a processor.tokenizer regression are different model/processor problems." }, { - "left": "issue:44361", - "right": "issue:44442", + "left": "issue:45372", + "right": "issue:46170", "accept": false, - "reason": "Tokenizer AttributeError on tasks and AutoTokenizer loading FastSpeech2ConformerTokenizer are not the same bug." + "reason": "Gemma 4 processor import breakage from mistral_common and Windows ROCm torch.distributed issues are unrelated dependency failures." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "save_pretrained shard-size question is unrelated to AutoProcessor kwargs forwarding." + "reason": "A save_pretrained shard-size question and missing kwargs passthrough in AutoProcessor.from_pretrained are not duplicates." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "BigBird special-token registration issue and MLukeTokenizer AttributeError are different tokenizer failures." + "reason": "BigBird special-token registration and MLukeTokenizer task errors are different tokenizer bugs." }, { "left": "issue:45542", "right": "issue:46178", "accept": false, - "reason": "TensorBoard-without-TensorFlow backend error and tensorflow_text backend mapping error are separate dependency-handling bugs." + "reason": "Both concern backend detection, but they target different missing backend cases and are not the same underlying defect." }, { - "left": "issue:43452", - "right": "issue:43577", + "left": "issue:43295", + "right": "issue:45356", "accept": false, - "reason": "gguf_file loading regression is unrelated to Blip2 dtype not propagating to submodules." + "reason": "Processor/tokenizer regression in v4.57.5 and Kimi-K2.5 codec/warning issues are distinct." }, { - "left": "issue:43295", - "right": "issue:45356", + "left": "issue:43452", + "right": "issue:43577", "accept": false, - "reason": "Both are regressions, but one is processor.tokenizer access and the other is Kimi-K2.5 codec/regex handling." + "reason": "gguf_file loading breakage and Blip2 dtype remaining float32 are unrelated." }, { "left": "issue:43577", "right": "issue:43697", "accept": false, - "reason": "Blip2 dtype propagation and RTDetrV2 output changes are different model-specific regressions." + "reason": "A dtype propagation bug in Blip2 and RTDetrV2 output drift are different problems." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Different tokenizer bugs: BigBird mask token special registration vs mdeberta add_special_tokens behavior." + "reason": "Special-token registration for BigBird and missing BOS/EOS insertion for mdeberta are separate tokenizer behaviors." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is not the same as recurrent-state reset in modular_qwen3_5." + "reason": "Chunked generation with compiled forward and Qwen3_5 recurrent-state reset are different generation/cache bugs." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and MobileLLM-125M load-time KeyError are not the same issue." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Generic custom model initialization regression and FSDP CPU RAM loading are unrelated." + "reason": "Custom model initialization breakage and FSDP CPU RAM-efficient loading are different failures." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Offline Mistral regex patch bug is unrelated to tokenizer loading failure for EMBEDDIA/est-roberta." + "reason": "Offline-mode model_info() call in tokenizer patching and est-roberta tokenizer loading regression are different root causes." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The summary correctly treats the cluster as mostly heterogeneous and only accepts the tied-weights serialization pair, which is the one clearly plausible duplicate-style match from the titles. The rejections are cautious rather than overstated." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -7199,8 +7201,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -7215,8 +7217,9 @@ "issue:43295", "issue:43452", "issue:43504", - "issue:43531", + "issue:43525", "issue:43550", + "issue:43572", "issue:43577", "issue:43606", "issue:43643", @@ -7226,7 +7229,6 @@ "issue:43742", "issue:43756", "issue:43927", - "issue:43994", "issue:44361", "issue:44451", "issue:44464", @@ -7238,6 +7240,7 @@ "issue:44610", "issue:44625", "issue:44821", + "issue:44849", "issue:44977", "issue:45290", "issue:45440", @@ -7247,23 +7250,21 @@ "issue:45823", "issue:46002", "issue:46050", - "issue:46192", "issue:46209", "issue:46365", "issue:46402" ], "soft_pairs": [ - "issue:43531|issue:43742", "issue:29159|issue:33357", - "issue:43756|issue:46402", - "issue:45491|issue:46002", "issue:43688|issue:45704", + "issue:43756|issue:46402", "issue:43688|issue:45440", "issue:43653|issue:43927", "issue:44514|issue:45290", "issue:43550|issue:43927", "issue:43688|issue:45701", "issue:44479|issue:44625", + "issue:45491|issue:46002", "issue:43550|issue:44361", "issue:44464|issue:44977", "issue:43643|issue:44561", @@ -7271,241 +7272,242 @@ "issue:33666|issue:36010", "issue:43742|issue:44821", "issue:44466|issue:46050", - "issue:43504|issue:43720", "issue:43504|issue:43550", "issue:32090|issue:33666", + "issue:43504|issue:43720", "issue:43577|issue:44554", "issue:44977|issue:46209", "issue:43504|issue:44610", "issue:44554|issue:46365", "issue:43262|issue:43295", "issue:43720|issue:43927", - "issue:45823|issue:46050", "issue:43606|issue:43927", "issue:43452|issue:44451", + "issue:45823|issue:46050", "issue:33666|issue:36331", - "issue:43994|issue:46192" + "issue:43525|issue:43572", + "issue:44849|issue:46209" ], - "prepared_review_unit_hash": "e5f1a2995ba24a60ca68af5a2f340b9943d42790cc66910ed35cb300271b45af", + "prepared_review_unit_hash": "7aba7b1092b65d9cc20a92a2363bc35d4d1dad48c5c0bae8689a8b1d16512707", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 36774, - "estimated_input_tokens": 9194, - "estimated_eval_tokens": 18644 + "serialized_chars": 36763, + "estimated_input_tokens": 9191, + "estimated_eval_tokens": 18638 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: it mixes unrelated tokenizer, model-loading, generation, device-map, precision, and processor regressions. The soft links look like broad textual similarity rather than the same underlying bug, so none should be merged as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is heterogeneous: most items are distinct model-, tokenizer-, config-, or runtime-specific regressions that only share broad keywords like loading, generation, or chat templating. None of the soft-similarity pairs look like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue stands out because the set spans many unrelated code paths (tokenizers, config loading, generation, vision/video processors, trainer/runtime errors). The apparent overlaps are only surface-level.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "If forced to pick a representative anchor, issue:43643 is the most umbrella-like because it is a framework-level config/loading regression and closest to the recurring trust_remote_code/load-path theme, but it still is not a true duplicate hub.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43531", - "right": "issue:43742", - "accept": false, - "reason": "Different bugs: Qwen3-MoE sliding_window behavior vs a MobileLLM load/key error." - }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to a MacOS CLIP bus error." - }, - { - "left": "issue:43756", - "right": "issue:46402", - "accept": false, - "reason": "Different failure modes: RoPE-layer dropping in Smollm3 vs missing text-encoder weight loading in Sam3Model." + "reason": "Tokenizer inconsistency vs MacOS bus error on a CLIP model; different failure modes and code paths." }, { - "left": "issue:45491", - "right": "issue:46002", + "left": "issue:43688", + "right": "issue:45704", "accept": false, - "reason": "Both mention NaNs, but one is sliding-window attention on mixed-length batches and the other is classification-head init/checkout behavior." + "reason": "Both mention model math bugs, but one is auxiliary-loss normalization and the other is a T5 RMSNorm memory leak." }, { - "left": "issue:43688", - "right": "issue:45704", + "left": "issue:43756", + "right": "issue:46402", "accept": false, - "reason": "Aux-loss normalization is unrelated to a T5/apex RMSNorm memory leak." + "reason": "Smollm3 RoPE-layer truncation and Sam3 text-encoder weights not loading are unrelated model bugs." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Different model parity issues: auxiliary-loss normalization vs DeepseekV3 implementation divergence." + "reason": "Aux-loss normalization in MoE models is not the same as DeepSeekV3MoE diverging from the remote implementation." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "Both involve token handling, but one is a BigBird special-token registration bug and the other is config persistence of custom token IDs." + "reason": "Tokenizer special-token registration vs config losing custom token IDs after save/load are different bugs." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Same API surface, but different triggers and likely different code paths: batched padding=False crash vs tool-call assistant message with no content." + "reason": "Both are apply_chat_template crashes, but the edge cases differ: batched padding=False vs assistant tool-call messages with no content." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "Unrelated: Bamba torch.compile/SDPA bug vs DiaConfig save/load token-ID regression." + "reason": "torch.compile+SDPA failure for Bamba is unrelated to DiaConfig token-ID persistence." }, { "left": "issue:43688", "right": "issue:45701", "accept": false, - "reason": "Aux-loss normalization is not the same as a broad tokenization-version regression." + "reason": "Aux-loss normalization and tokenization-version changes are different subsystems and failure classes." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Different regressions: video-input preprocessing vs num_labels propagation in Qwen3.5." + "reason": "Qwen3.5 video-input regression and num_labels propagation are separate model/config issues." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both involve NaNs, but one is Gemma3 sliding-window attention and the other is classification-head init/checkout behavior." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "Different subsystems: compiled SDPA failure vs MLukeTokenizer AttributeError." + "reason": "torch.compile/SDPA failure is unrelated to MLukeTokenizer AttributeError during tasks." }, { "left": "issue:44464", "right": "issue:44977", "accept": false, - "reason": "Both affect generation, but one is compiled-forward chunked generation and the other is flash-attention behavior on Qwen3.5." + "reason": "Chunked generation with compiled forward and Qwen3.5 flash-attention generation bugs are different concrete code paths." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both mention trust_remote_code, but one returns incomplete config fields and the other is a v5 removal breaking remote-code models." + "reason": "Both affect trust_remote_code flows, but one is missing fields from AutoConfig and the other is removal of is_torch_fx_available." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Different bug classes: auto device mapping cuda assert vs compiled-forward chunked generation inconsistency." + "reason": "Qwen3 auto device-map cuda assert and compiled-forward chunked generation are unrelated runtime issues." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training issue is unrelated to an import failure for GenerationMixin." + "reason": "Multi-GPU training for Qwen2-VL and GenerationMixin import failure are not the same bug." }, { "left": "issue:43742", "right": "issue:44821", "accept": false, - "reason": "Different loaders and surfaces: MobileLLM checkpoint load error vs AutoImageProcessor URL loading." + "reason": "MobileLLM key error while loading and AutoImageProcessor-from-URL failure are different loading paths." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "Both concern save/load, but one is tied-weight serialization and the other is quantized sinq checkpoint restoration." - }, - { - "left": "issue:43504", - "right": "issue:43720", - "accept": false, - "reason": "Different model-loading failures: legacy field handling in Beit vs packed-weight unpacking in BitNet." + "reason": "lm_head.weight serialization inconsistency and sinq quantized save/load breakage are distinct persistence bugs." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "No overlap beyond being model bugs; Beit load failure and Bamba compile/SDPA failure are unrelated." + "reason": "Beit legacy-field loading failure and Bamba torch.compile/SDPA failure are unrelated." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Trainer NoneType broadcast error is not the same as a Qwen2-VL multi-GPU training request/issue." + "reason": "Trainer _gpu_broadcast_one NoneType error is not the same as Qwen2-VL multi-GPU training behavior." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Both are loading-time problems, but one is a legacy config field and the other is packed-weight unpacking for BitNet." }, { "left": "issue:43577", "right": "issue:44554", "accept": false, - "reason": "Different concerns: dtype propagation in Blip2 vs an MPS attention correctness issue." + "reason": "BLIP2 dtype propagation and an MPS attention correctness bug are different model/runtime issues." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Different precision/runtime problems: Qwen3.5 flash-attention generation vs FP8 inference on B200." + "reason": "Qwen3.5 flash-attention generation and FP8 inference on B200 are not the same failure." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Beit legacy-field load failure is unrelated to OmDet processor/model input-size mismatch." + "reason": "Beit pretrained loading via legacy field and OmDet-Turbo processor/model input-size mismatch are separate problems." }, { "left": "issue:44554", "right": "issue:46365", "accept": false, - "reason": "MPS attention correctness issue is unrelated to importing Trainer conflicting with datasets." + "reason": "MPS attention correctness and importing Trainer being blocked by datasets are unrelated." }, { "left": "issue:43262", "right": "issue:43295", "accept": false, - "reason": "Both are processor regressions, but one is default sampling-rate selection and the other is processor.tokenizer/image-passing API breakage." + "reason": "Audio apply_chat_template sample-rate defaulting and processor.tokenizer/image-handling regression are different processor bugs." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "Packed-weight loading in BitNet is unrelated to DiaConfig custom token-ID persistence." - }, - { - "left": "issue:45823", - "right": "issue:46050", - "accept": false, - "reason": "Different device/serialization bugs: Gemma4 PLE auto-device mismatch vs sinq checkpoint save/load regression." + "reason": "BitNet packed-weight loading and DiaConfig token-ID persistence do not share the same code path." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch is unrelated to token-ID persistence." + "reason": "Bark CPU offload device mismatch and DiaConfig save/load token-ID loss are unrelated." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "Both are loading-related, but gguf_file regressions and ScandiBERT loading failure are different root causes." + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and inability to load a specific pretrained model are different loading regressions." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Gemma4 PLE device mismatch and sinq quantized save/load breakage are distinct issues." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Multi-GPU training issue is unrelated to a CustomTrainer compute_loss signature mismatch." + "reason": "Qwen2-VL multi-GPU training and CustomTrainer.compute_loss signature mismatch are unrelated training bugs." }, { - "left": "issue:43994", - "right": "issue:46192", + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config and missing pad_token_idx in StableLmConfig are similar in theme but model-specific and not the same bug." + }, + { + "left": "issue:44849", + "right": "issue:46209", "accept": false, - "reason": "Siglip2 output quality regression is unrelated to return_tensors being ignored when text_kwargs is passed." + "reason": "Qwen3.5 output_hidden_states bug and FP8 inference failure on B200 are separate model/runtime regressions." } ] }, @@ -7524,8 +7526,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -7540,7 +7542,6 @@ "issue:43334", "issue:43493", "issue:43504", - "issue:43525", "issue:43540", "issue:43572", "issue:43575", @@ -7553,6 +7554,8 @@ "issue:43825", "issue:43844", "issue:43927", + "issue:43994", + "issue:44291", "issue:44361", "issue:44464", "issue:44466", @@ -7573,265 +7576,257 @@ "issue:45584", "issue:45759", "issue:45812", - "issue:45823", - "issue:46209", + "issue:46192", "issue:46424", "issue:46489" ], "soft_pairs": [ - "issue:43525|issue:43572", - "issue:44849|issue:46209", + "issue:43994|issue:46192", + "issue:43720|issue:44361", "issue:43278|issue:43688", "issue:43493|issue:43825", - "issue:43720|issue:44361", "issue:44964|issue:45405", "issue:39692|issue:43493", "issue:28218|issue:33357", "issue:44625|issue:45584", "issue:44849|issue:45357", - "issue:43493|issue:43844", - "issue:44568|issue:46489", "issue:39692|issue:43295", - "issue:45200|issue:45229", + "issue:44568|issue:46489", + "issue:43493|issue:43844", "issue:43278|issue:43844", "issue:44488|issue:44843", "issue:43540|issue:44977", "issue:43646|issue:44466", "issue:44568|issue:45812", - "issue:44464|issue:46424", - "issue:33357|issue:33666", "issue:43824|issue:44589", + "issue:33357|issue:33666", + "issue:45200|issue:45229", "issue:44589|issue:45520", + "issue:44464|issue:46424", "issue:43257|issue:43299", "issue:43334|issue:43572", "issue:43927|issue:44361", - "issue:43504|issue:43606", "issue:43334|issue:45759", - "issue:45229|issue:45823", + "issue:43504|issue:43606", "issue:43278|issue:43701", "issue:44792|issue:45405", - "issue:43575|issue:45405" + "issue:43575|issue:45405", + "issue:43572|issue:44291" ], - "prepared_review_unit_hash": "cf4534343154e46c6667551adb947cbfd29c736bc35ae5a9aaea8a20e09b7a9d", + "prepared_review_unit_hash": "6e0ff0e526042443d30f4bdde563727047863a0c0f0958d0abf3aed4bfa8174e", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37290, - "estimated_input_tokens": 9323, - "estimated_eval_tokens": 18902 + "soft_pair_count": 31, + "serialized_chars": 37103, + "estimated_input_tokens": 9276, + "estimated_eval_tokens": 18808 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The soft pairs cluster around tokenizer/config/model-loading regressions, but they point to different concrete failures and code paths, so none are safe duplicate merges. The clearest representative issue is the Granite tokenizer regression because it describes a broad v4\u2192v5 tokenization break in core AutoTokenizer behavior.", - "confidence": 0.67, - "canonical_issue_reason": "issue:45812 is the strongest representative: it describes a broad, silent AutoTokenizer regression affecting token IDs across an entire model family, which is more central than the niche model/config-specific reports.", + "summary": "These issues are thematically unrelated overall; the soft-similarity edges mostly connect different model families or different failure modes, so none should be merged as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45812 is the best overall issue to anchor this cluster because it captures the most general and high-impact tokenizer regression with a clear code-path description.", + "best_issue_reason": "issue:28218 is the oldest, broadest tokenizer-regression report in the set, but even it is not a true duplicate representative for the rest of the items.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43525", - "right": "issue:43572", + "left": "issue:43994", + "right": "issue:46192", "accept": false, - "reason": "Both are missing-pad-token config bugs, but one is Llama4 `pad_token_id` and the other is StableLm `pad_token_idx`; different models and different fields." + "reason": "Different bugs: SigLIP2 inference/pipeline mismatch vs return_tensors being ignored when text_kwargs is passed." }, { - "left": "issue:44849", - "right": "issue:46209", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "Qwen3.5 hidden-states behavior and FP8 inference on B200 are unrelated failures in different subsystems." + "reason": "BitNet packed-weight loading failure is unrelated to MLukeTokenizer\u2019s task-time AttributeError." }, { "left": "issue:43278", "right": "issue:43688", "accept": false, - "reason": "Embedding dtype drift and auxiliary-loss normalization are distinct training/eval bugs with different fixes." + "reason": "BF16\u2192FP32 eval dtype drift is a different issue from auxiliary-loss normalization in MoE models." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 implementation parity and a pipeline error-message regression are unrelated issues." - }, - { - "left": "issue:43720", - "right": "issue:44361", - "accept": false, - "reason": "Packed-weight unpacking in accelerate loading and an MLukeTokenizer AttributeError do not share the same underlying bug." + "reason": "SigLIP2 HF-vs-JAX discrepancy is unrelated to the pipeline translation error-message regression." }, { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "Model loading for Phi-4 multimodal and a PEFT version requirement bump are unrelated." + "reason": "Model loading failure and an unsupported PEFT version bump are not the same underlying bug." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "Both mention SigLIP2, but one is documentation/example mistakes and the other is implementation discrepancy; not the same bug." + "reason": "Docs/example errors for SigLIP2 are separate from the SigLIP2 implementation discrepancy report." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer spacing after added tokens and a MacOS bus error on a CLIP model are unrelated." + "reason": "Tokenizer added-space bug does not match a MacOS bus error in a CLIP model." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "Qwen3.5 `num_labels` propagation and Whisper empty-transcription generation failure are different code paths." + "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription behavior are different code paths." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Qwen3.5 hidden-states output and incorrect save_pretrained visual encoder keys are separate regressions." + "reason": "Hidden-states output bug is unrelated to incorrect visual-encoder key saving." }, { - "left": "issue:43493", - "right": "issue:43844", + "left": "issue:39692", + "right": "issue:43295", "accept": false, - "reason": "SigLIP2 parity mismatch and DeepSpeed gradient growth are unrelated training/model bugs." + "reason": "SigLIP2 docs example problems are not the same as the processor.tokenizer regression in custom model code." }, { "left": "issue:44568", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer regressions, but one is mdeberta special-token insertion and the other is DeepSeek-Coder v1 token output; different model families and code paths." + "reason": "These are distinct tokenizer regressions for different model families and different failure modes." }, { - "left": "issue:39692", - "right": "issue:43295", - "accept": false, - "reason": "SigLIP2 example issues and the processor.tokenizer regression are not the same underlying failure." - }, - { - "left": "issue:45200", - "right": "issue:45229", + "left": "issue:43493", + "right": "issue:43844", "accept": false, - "reason": "Missing mm_token_type_ids defaults and Gemma4 multi-GPU OOM are different problems." + "reason": "SigLIP2 discrepancy and gradient blow-up under ZeRO-3 do not share a concrete root cause." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "A dtype mismatch in embeddings and exploding gradients under ZeRO-3 are not the same bug." + "reason": "Eval dtype change is unrelated to abnormal gradient growth with HfDeepSpeedConfig." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "Loading a specific model and an offline-mode `model_info()` call regression are unrelated." + "reason": "Model load failure for cjvt/sleng-bert is unrelated to the offline model_info call in mistral regex patching." }, { "left": "issue:43540", "right": "issue:44977", "accept": false, - "reason": "Qwen3OmniMoe video-processing failure and Qwen3.5 flash-attention generation are separate issues." + "reason": "Qwen3OmniMoe video processing error and Qwen3.5 flash-attention generation bug are different problems." }, { "left": "issue:43646", "right": "issue:44466", "accept": false, - "reason": "Custom model initialization breakage and tied-weight serialization inconsistency are different regressions." + "reason": "Custom model initialization breakage is separate from device-dependent lm_head serialization inconsistency." }, { "left": "issue:44568", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer bugs, but `mdeberta-v3-base` special-token handling and Granite token-ID regression affect different tokenizer implementations and failure modes." + "reason": "MDeBERTa special-token handling and Granite tokenizer regression are different tokenizer paths." }, { - "left": "issue:44464", - "right": "issue:46424", + "left": "issue:43824", + "right": "issue:44589", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency and quantized-cache `max_cache_length` behavior are unrelated." + "reason": "Importing a missing Qwen2.5-VL class is unrelated to a Float8 storage-object lookup failure." }, { "left": "issue:33357", "right": "issue:33666", "accept": false, - "reason": "A MacOS bus error on CLIP and Qwen2-VL multi-GPU training are unrelated." + "reason": "Mac bus error in CLIP inference is not the same as Qwen2-VL multi-GPU training support." }, { - "left": "issue:43824", - "right": "issue:44589", + "left": "issue:45200", + "right": "issue:45229", "accept": false, - "reason": "Missing Qwen2.5 VL import and a Float8 storage lookup error are different failures." + "reason": "mm_token_type_ids defaulting for text-only fine-tuning is unrelated to Gemma4 multi-GPU OOM." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage import failure and `flash_attn` KeyError on Python 3.13 are unrelated environment issues." + "reason": "Float8 storage resolution and flash_attn import KeyError are separate backend issues." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency is different from cache-length handling in quantized cache." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Both involve Qwen3 MoE loading, but one is accelerate/deepspeed weight conversion and the other is Qwen3VL dev loading; not the same concrete bug." + "reason": "Accelerate/DeepSpeed MOE conversion failure is not the same as the Qwen3VL MOE loading regression." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Qwen3-VL `pad_token_id` missing and StableLm `pad_token_idx` missing are separate config regressions." + "reason": "Qwen3-VL missing pad_token_id and StableLmConfig missing pad_token_idx are unrelated config bugs." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "DiaConfig token-ID persistence and MLukeTokenizer task handling are different issues." - }, - { - "left": "issue:43504", - "right": "issue:43606", - "accept": false, - "reason": "BEiT legacy-field loading and bark-small CPU offload device mismatch are unrelated." + "reason": "DiaConfig save/load token-ID loss is unrelated to MLukeTokenizer\u2019s AttributeError on tasks." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Both concern Qwen multimodal configs, but one is a missing `pad_token_id` attribute and the other is failing to unwrap `text_config`; different root causes." + "reason": "These both involve multimodal Qwen configs, but one is a missing pad_token_id bug and the other is failure to unwrap text_config." }, { - "left": "issue:45229", - "right": "issue:45823", + "left": "issue:43504", + "right": "issue:43606", "accept": false, - "reason": "Gemma4 multi-GPU OOM and PLE device mismatch during forward are different runtime failures." + "reason": "Legacy-field load failure for Beit is unrelated to the CPU-offload device mismatch in bark-small." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "Embedding dtype changes and resume-from-checkpoint key mismatch are unrelated." + "reason": "Eval dtype drift and resume_from_checkpoint key mismatch are different state-management bugs." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "A janus generation test failure and a PEFT minimum-version bump are not the same bug." + "reason": "A failing Janus test and a bumped minimum PEFT version are not the same change." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "Qwen2-57B tensor-parallel OOM and a dependency version bump are unrelated." + "reason": "Qwen2 TP OOM and PEFT version requirements are unrelated issues." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to init_empty_weights receiving an unexpected argument." } ] }, @@ -7850,8 +7845,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -7866,10 +7861,10 @@ "issue:38175", "issue:42490", "issue:43065", + "issue:43232", "issue:43493", "issue:43504", "issue:43540", - "issue:43572", "issue:43606", "issue:43653", "issue:43696", @@ -7880,17 +7875,18 @@ "issue:43994", "issue:44112", "issue:44186", - "issue:44291", "issue:44423", "issue:44464", "issue:44493", "issue:44554", "issue:44734", + "issue:44849", "issue:44964", "issue:45072", "issue:45084", "issue:45198", "issue:45200", + "issue:45229", "issue:45406", "issue:45446", "issue:45520", @@ -7904,15 +7900,15 @@ "issue:46369" ], "soft_pairs": [ - "issue:43572|issue:44291", "issue:44493|issue:46285", + "issue:45229|issue:45823", "issue:43606|issue:44112", "issue:43504|issue:43653", "issue:44464|issue:45584", "issue:43994|issue:44493", "issue:43873|issue:46050", - "issue:45588|issue:46050", "issue:29942|issue:30333", + "issue:45588|issue:46050", "issue:44423|issue:44734", "issue:43844|issue:44964", "issue:43065|issue:46369", @@ -7924,190 +7920,197 @@ "issue:44554|issue:45446", "issue:15354|issue:38175", "issue:43540|issue:45200", - "issue:45200|issue:45823", - "issue:45072|issue:45198", "issue:43493|issue:43701", + "issue:45072|issue:45198", "issue:37428|issue:43957", - "issue:44186|issue:46192" + "issue:45200|issue:45823", + "issue:44186|issue:46192", + "issue:43232|issue:44849" ], - "prepared_review_unit_hash": "368b4bfcdcfea8e8470ea46e03e05e02d4d0989caf65e6ace2880c1fe64cda6e", + "prepared_review_unit_hash": "7ba4b9969eb8b59d9196424737acb0182b8ecb75616f0e7b0215426fdb216c51", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 25, - "serialized_chars": 34800, - "estimated_input_tokens": 8700, - "estimated_eval_tokens": 17656 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35586, + "estimated_input_tokens": 8897, + "estimated_eval_tokens": 18050 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: most items are unrelated Transformers bugs spanning tokenizers, model loading, serving, quantization, and attention backends. The soft pairs mostly share broad subsystem vocabulary but not the same concrete failure mode, so none should be merged as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "Issue 43873 is the broadest, actively discussed bug report in the set and has the strongest chance of serving as a representative anchor, though this cluster is not truly duplicate-heavy.", + "summary": "The cluster is heterogeneous: tokenizer quirks, serving crashes, model-specific regressions, quantization/offload issues, and CI failures are all mixed together. None of the soft pairs look like the same underlying bug/change, so I rejected them all.", + "confidence": 0.91, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated subsystems and model families, with no strong duplicate center to anchor the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the best representative issue because it is open, well-described, and broadly scoped; however, it is still only a loose thematic fit for the rest of the cluster.", + "best_issue_reason": "issue:43873 is the most broadly scoped and still-open bug report in the set, but it is not a duplicate representative for the rest of the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43572", - "right": "issue:44291", + "left": "issue:44493", + "right": "issue:46285", "accept": false, - "reason": "Both are model-loading/config compatibility problems, but they fail in different places: missing pad_token_idx vs an unexpected _is_hf_initialized argument." + "reason": "Both mention key/position-related behavior, but one is a broad model-key regression and the other is a CLIP class structure inconsistency; not the same bug." }, { - "left": "issue:44493", - "right": "issue:46285", + "left": "issue:45229", + "right": "issue:45823", "accept": false, - "reason": "These concern different tokenizer/model-structure regressions; one is a key/position-id warning, the other is a CLIP text-model attribute layout change." + "reason": "Same Gemma4 family, but one is CUDA OOM in multi-GPU inference and the other is a PLE device mismatch in forward; different failure modes and code paths." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "Both mention device-related CI failures, but they affect different models and code paths, so they are not the same underlying bug." + "reason": "Device/offload test wording overlaps, but one is a bark-small CPU offload mismatch and the other is a GraniteSpeech CI test issue; unrelated concrete problems." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "A legacy preset-load issue for BEiT segmentation is unrelated to BigBirdTokenizer special-token registration and decode output." + "reason": "Different components and symptoms: BEiT legacy-field loading vs BigBirdTokenizer special-token registration/decode output." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Both are generation bugs, but one is about compiled forward/chunked generation and the other is Whisper empty-transcription handling after token alignment." + "reason": "One is chunked generation with compiled forward, the other is Whisper empty-transcription handling after align_special_tokens; not the same code path." }, { "left": "issue:43994", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 output corruption and an unexpected key/position-id issue are different model behaviors with no shared concrete failure path." + "reason": "Both are model-output correctness complaints, but one is a SigLIP2 inference regression and the other is a generic unexpected key/position-id issue." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Quantization/offloading and quantization save/load breakage are related themes but distinct bugs affecting different lifecycle stages." + "reason": "Both involve quantization/offloading, but one is offload behavior with quantization and the other is save/load corruption for sinq quantization; different bugs." }, { - "left": "issue:45588", - "right": "issue:46050", + "left": "issue:29942", + "right": "issue:30333", "accept": false, - "reason": "One is a flash-attention integration crash on sink-less models; the other is a sinq serialization/deserialization issue." + "reason": "Flash Attention 2 test failures and MLflow reporting on job failure are unrelated." }, { - "left": "issue:29942", - "right": "issue:30333", + "left": "issue:45588", + "right": "issue:46050", "accept": false, - "reason": "Flash Attention test failures and MLflow reporting on failed jobs are unrelated." + "reason": "Flash-attention sink handling and sinq save/load breakage are separate subsystems with different symptoms." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Both are serve bugs, but one crashes on multimodal continuous batching while the other fails in KV-cache continuation indexing." + "reason": "Both are serve crashes, but one is multimodal continuous-batching string handling and the other is KV-cache continuation indexing; distinct defects." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "Gradient explosion under DeepSpeed ZeRO-3 is unrelated to a latest-transformers model-loading failure for Phi-4 multimodal." + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading failure are unrelated." }, { "left": "issue:43065", "right": "issue:46369", "accept": false, - "reason": "Both involve SAM3, but one is about a dummy Conv2d placeholder in a decoder while the other is a memory overflow when running the litetext model." + "reason": "Same model family only in a very broad sense; dummy Conv2d setup and memory overflow are not the same underlying issue." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "A local-path import warning is not the same bug as GPT-oss-20b CUDA OOM." + "reason": "Import warning on trailing slash paths is unrelated to GPT-oss CUDA OOM." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Tokenizer slow/fast inconsistency and a Trainer _gpu_broadcast_one NoneType error are unrelated." + "reason": "Tokenizer slow/fast inconsistency and Trainer NoneType broadcast errors are different bugs." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4 serve processor missing _tokenizer is unrelated to rope-parameter list typing in configuration validation." + "reason": "Gemma4Processor missing _tokenizer and rope-validation list typing are unrelated." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "An added-token spacing bug is unrelated to TrainerState num_input_tokens_seen not updating." + "reason": "Tokenizer spacing after added tokens is not the same as TrainerState.num_input_tokens_seen not updating." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "A template-node compilation error is unrelated to a flash_attn import KeyError on Python 3.13." + "reason": "Template compilation errors and flash_attn import KeyError are separate failures." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "An MPS attention correctness issue is not the same as a PyTorch version check bug in flex_attention imports." + "reason": "MPS attention correctness and a PyTorch version check for AuxRequest are unrelated." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript export with GeneratorExp is unrelated to SigLIP2 zero-probability outputs." + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are completely different issues." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Qwen3OmniMoe video input validation and Gemma4 mm_token_type_ids defaulting are different multimodal bugs." + "reason": "Qwen3OmniMoe video processing ValueError and Gemma4 mm_token_type_ids defaults are unrelated model/input bugs." }, { - "left": "issue:45200", - "right": "issue:45823", + "left": "issue:43493", + "right": "issue:43701", "accept": false, - "reason": "Both are Gemma4 issues, but one is a missing default for token type ids and the other is a device-map forward mismatch." + "reason": "SigLIP2 HF-vs-JAX discrepancy and resume_from_checkpoint key mismatch are different problems." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "Bfloat16 dtype mismatches in inference and Wav2Vec2 save/tokenization failures do not share the same root cause." + "reason": "dtype mismatches in inference and Wav2Vec2 save/tokenization failures do not share the same code path." }, { - "left": "issue:43493", - "right": "issue:43701", + "left": "issue:37428", + "right": "issue:43957", "accept": false, - "reason": "A SigLIP2 implementation discrepancy and a resume_from_checkpoint key mismatch are unrelated." + "reason": "flash attention import regression and meta-device loading breakage are unrelated." }, { - "left": "issue:37428", - "right": "issue:43957", + "left": "issue:45200", + "right": "issue:45823", "accept": false, - "reason": "A missing flash-attention utility import and meta-device model-loading failures are different compatibility issues." + "reason": "Both touch Gemma4, but one is missing mm_token_type_ids defaults and the other is a PLE device mismatch during forward." }, { "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "LayoutLMv2 NER padding/truncation crashes and return_tensors being ignored in text_kwargs are different tokenizer API bugs." + "reason": "LayoutLMv2 tokenizer NER/padding crash and return_tensors being ignored are different tokenizer issues." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and Qwen3.5 output_hidden_states behavior are not the same bug." } ] }, @@ -8126,8 +8129,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -8138,7 +8141,7 @@ "issue:39692", "issue:42490", "issue:42915", - "issue:43232", + "issue:43278", "issue:43295", "issue:43334", "issue:43450", @@ -8151,6 +8154,7 @@ "issue:43653", "issue:43824", "issue:43844", + "issue:43856", "issue:43873", "issue:43950", "issue:43957", @@ -8161,7 +8165,6 @@ "issue:44554", "issue:44610", "issue:44792", - "issue:44849", "issue:44898", "issue:44987", "issue:45127", @@ -8182,240 +8185,237 @@ ], "soft_pairs": [ "issue:44521|issue:45290", - "issue:43232|issue:44849", - "issue:45823|issue:46192", "issue:43295|issue:43450", + "issue:45823|issue:46192", "issue:43295|issue:44610", "issue:45381|issue:46209", "issue:43295|issue:43873", "issue:44554|issue:45356", - "issue:39692|issue:46192", "issue:43572|issue:44987", + "issue:39692|issue:46192", "issue:45520|issue:46178", "issue:43550|issue:43844", - "issue:44466|issue:45127", "issue:43572|issue:43957", "issue:37428|issue:43824", + "issue:44466|issue:45127", "issue:44898|issue:45412", - "issue:43950|issue:46285", "issue:43450|issue:43575", - "issue:45200|issue:45325", + "issue:43950|issue:46285", "issue:42490|issue:43493", - "issue:43452|issue:46512", - "issue:43334|issue:43525", "issue:44792|issue:45563", + "issue:43334|issue:43525", "issue:42915|issue:43824", "issue:38175|issue:43493", + "issue:43452|issue:46512", "issue:43824|issue:43975", + "issue:43653|issue:44112", "issue:28218|issue:35141", - "issue:43653|issue:44112" + "issue:45200|issue:45325", + "issue:43278|issue:43856" ], - "prepared_review_unit_hash": "12a292c44111c34cf235a5c54bdf3ab7961d8c7dd64b4ac2fc9dde0ee665e003", + "prepared_review_unit_hash": "377143ceb956ffb4d214101ea46bdf9f32128e999d2193f28ae3663d4bc35b2f", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36454, - "estimated_input_tokens": 9114, - "estimated_eval_tokens": 18484 + "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": "Mostly heterogeneous transformer issues with a single strong duplicate-like family around missing `pad_token_id` on config objects. The SigLIP2 pair is not a safe duplicate from the packet alone, and there are no PRs here.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43334 is the clearest anchor: it reports a concrete load-time `AttributeError` from a missing `pad_token_id` on a config object, and issue:43525 appears to be the same config-attribute regression on a different model.", + "summary": "This cluster is a broad semantic false-positive: the issues span unrelated bugs in tokenizers, multimodal processors, model loading, quantization, config attributes, and backend imports. None of the proposed soft pairs looks like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43295 is the broadest regression report in the set and touches the processor/tokenizer API surface, but the cluster is too heterogeneous for it to be a true canonical duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43334 is the best standalone issue because it has the most explicit repro and the most concrete failure mode, making it the strongest candidate for deduping nearby config-attribute regressions.", + "best_issue_reason": "No single issue is a strong global canonical for this cluster; 43295 is only the closest broad representative, not an actual duplicate umbrella.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "Both touch chat templating, but one is multimodal assistant-mask handling and the other is tool-call/no-content crash; different code paths and symptoms." + "reason": "Both involve chat templates, but one is about assistant masks for multimodal inputs and the other is a crash with tool-call messages; different code paths." }, { - "left": "issue:43232", - "right": "issue:44849", + "left": "issue:43295", + "right": "issue:43450", "accept": false, - "reason": "Generation kwargs sync behavior and hidden-state output regression are unrelated bugs." + "reason": "Processor/tokenizer regression vs batched video shape bug; different failure modes and subsystems." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Device-map forward mismatch and `return_tensors`/`text_kwargs` handling are different problems." - }, - { - "left": "issue:43295", - "right": "issue:43450", - "accept": false, - "reason": "Processor/tokenizer regression versus batched video shape bug; different inputs and failures." + "reason": "Device-mapping mismatch in Gemma4 vs return_tensors being ignored when text_kwargs is passed; unrelated behavior." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "A processor/tokenizer API regression is not the same as an OmDet-Turbo resolution mismatch." + "reason": "Custom model processor/tokenizer access issue vs OmDet processor image-size mismatch; not the same bug." }, { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Qwen2.5-VL video position IDs and FP8 inference on B200 are unrelated." + "reason": "Wrong Qwen2.5-VL video position ids vs FP8 inference failure on B200; completely different problems." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor/tokenizer regression and quantization offloading behavior are not the same bug." + "reason": "Processor/tokenizer regression vs quantization offloading issue; no concrete overlap." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "MPS attention correctness and Kimi tokenizer codec regression are different failure modes." + "reason": "MPS attention correctness bug vs tokenizer codec regression; unrelated." }, { - "left": "issue:39692", - "right": "issue:46192", + "left": "issue:43572", + "right": "issue:44987", "accept": false, - "reason": "SigLIP2 doc-example errors and ignored `return_tensors` are unrelated." + "reason": "Missing pad_token_idx in StableLmConfig vs a model-loading failure for physical-intelligence/fast; different root causes." }, { - "left": "issue:43572", - "right": "issue:44987", + "left": "issue:39692", + "right": "issue:46192", "accept": false, - "reason": "Missing padding config field and `physical-intelligence/fast` load failure are separate issues." + "reason": "SigLIP2 docs example errors vs ignored return_tensors; one is documentation/example drift, the other API handling." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "Different backend/import errors in different code paths." + "reason": "flash_attn import key mapping bug vs unsupported tensorflow_text backend; different import-system issues." }, { "left": "issue:43550", "right": "issue:43844", "accept": false, - "reason": "torch.compile SDPA failure and ZeRO-3 gradient inflation are distinct training/runtime bugs." - }, - { - "left": "issue:44466", - "right": "issue:45127", - "accept": false, - "reason": "Tied-weight serialization inconsistency and LoRA merge collapse are related in theme but not the same concrete bug." + "reason": "Bamba SDPA compile failure vs gradient blow-up under DeepSpeed ZeRO-3; not the same code-path." }, { "left": "issue:43572", "right": "issue:43957", "accept": false, - "reason": "Missing padding config field and meta-device loading breakage are separate regressions." + "reason": "StableLm config attribute regression vs meta-device loading bug; only broadly about loading, not duplicate." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Different import errors in different modules/models." + "reason": "Different missing-import errors for different symbols/models; same symptom class, not same bug." }, { - "left": "issue:44898", - "right": "issue:45412", + "left": "issue:44466", + "right": "issue:45127", "accept": false, - "reason": "Perceiver positional-encoding failure and RT-DETR memory release are unrelated." + "reason": "Serialization inconsistency of lm_head.weight vs LoRA merge collapse with extended vocab; related to tied weights but distinct failures." }, { - "left": "issue:43950", - "right": "issue:46285", + "left": "issue:44898", + "right": "issue:45412", "accept": false, - "reason": "Non-persistent buffer corruption during load and CLIP text-model structure inconsistency are different regressions." + "reason": "Perceiver interpolation bug vs RT-DETR memory leak; unrelated." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Batched video shape bug and tensor-parallel OOM are not the same underlying issue." + "reason": "Batched video shape bug vs tensor-parallel OOM loading a large model; different layers and symptoms." }, { - "left": "issue:45200", - "right": "issue:45325", + "left": "issue:43950", + "right": "issue:46285", "accept": false, - "reason": "Gemma4 text-only token-type defaults and Qwen2.5-VL rope/position scaling are unrelated multimodal bugs." + "reason": "Non-persistent buffer corruption in from_pretrained vs CLIPTextModel structure inconsistency; not the same issue." }, { "left": "issue:42490", "right": "issue:43493", "accept": false, - "reason": "Trailing-slash import warning and SigLIP2 implementation mismatch are unrelated." + "reason": "Trailing-slash import warning vs SigLIP2 implementation discrepancy; unrelated." }, { - "left": "issue:43452", - "right": "issue:46512", + "left": "issue:44792", + "right": "issue:45563", "accept": false, - "reason": "gguf_file loading breakage and stale `config.dtype` after `from_config()` are separate loading issues." + "reason": "Janus image-generation test failure vs stale generate() warning; one is a functional failure, the other a warning bug." }, { "left": "issue:43334", "right": "issue:43525", - "accept": true, - "reason": "Both report the same missing-`pad_token_id` config regression on model load, so they plausibly share one fix." - }, - { - "left": "issue:44792", - "right": "issue:45563", "accept": false, - "reason": "Janus image-generation test failure and stale paged-generate warning are different generation issues." + "reason": "Both mention missing pad_token_id, but they are model-specific config bugs for different architectures and likely different fixes." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "FP8 config failure for Qwen3MoE and a Qwen2.5-VL import error are unrelated." + "reason": "FP8 config failure for Qwen3Moe vs Qwen2.5-VL import error; no overlap beyond being model issues." }, { "left": "issue:38175", "right": "issue:43493", "accept": false, - "reason": "They share only a broad SigLIP2/correctness theme; one is a zero-probability symptom and the other is a broader HF-vs-JAX discrepancy, which is not enough to treat them as the same bug." + "reason": "Both are SigLIP2-related, but one reports zero probabilities and the other a broader HF-vs-JAX discrepancy; not clearly the same concrete bug." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file loading breakage vs stale model.config.dtype after from_config(dtype=...); separate loading APIs." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Qwen2_5_VL import failure and DeepSeek tokenizer detokenization are unrelated model-specific bugs." + "reason": "ImportError for a missing Qwen class vs DeepSeek tokenizer detokenization regression; unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Tokenizer special-token registration bug vs CI failure in GraniteSpeech; different objects and causes." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Tokenizer spacing after added tokens and resized-embedding reinitialization are different issues." + "reason": "Tokenizer spacing after added token vs embedding reinitialization after resize; different tokenizer/model behaviors." }, { - "left": "issue:43653", - "right": "issue:44112", + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaulting issue vs Qwen2.5-VL rope/position-id scaling bug; unrelated multimodal details." + }, + { + "left": "issue:43278", + "right": "issue:43856", "accept": false, - "reason": "BigBirdTokenizer special-token registration and a GraniteSpeech CI test flake are unrelated." + "reason": "dtype mismatch between train/eval vs inefficient memory usage in Qwen3 MoE training; not the same bug." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "Reject: the summary is mostly grounded, but one soft-edge acceptance is too aggressive. issue:38175 vs issue:43493 share a SigLIP2/correctness theme, yet the packet only gives a zero-probability symptom on one side and a broader HF-vs-JAX discrepancy on the other; that is not enough to conservatively call them the same bug. The other soft-edge rejections look fine, and there are no PRs in the packet." - }, - "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -8427,8 +8427,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -8438,7 +8438,6 @@ "issue:38175", "issue:41720", "issue:43262", - "issue:43278", "issue:43335", "issue:43388", "issue:43425", @@ -8450,7 +8449,6 @@ "issue:43653", "issue:43819", "issue:43827", - "issue:43856", "issue:43901", "issue:44112", "issue:44206", @@ -8463,8 +8461,10 @@ "issue:44743", "issue:44857", "issue:45083", + "issue:45127", "issue:45198", "issue:45200", + "issue:45245", "issue:45290", "issue:45412", "issue:45520", @@ -8483,22 +8483,21 @@ ], "soft_pairs": [ "issue:43526|issue:46192", - "issue:43278|issue:43856", "issue:43388|issue:43526", - "issue:41720|issue:46209", "issue:44442|issue:45198", + "issue:41720|issue:46209", "issue:43262|issue:43450", "issue:45520|issue:46170", "issue:38175|issue:45412", "issue:43819|issue:44112", - "issue:44206|issue:45584", "issue:44561|issue:46234", + "issue:44206|issue:45584", "issue:45759|issue:46285", - "issue:45290|issue:46326", "issue:43827|issue:43901", "issue:44610|issue:44857", - "issue:44514|issue:44521", + "issue:45290|issue:46326", "issue:43819|issue:45083", + "issue:44514|issue:44521", "issue:44857|issue:45593", "issue:43335|issue:43550", "issue:43611|issue:46402", @@ -8508,205 +8507,206 @@ "issue:29942|issue:32090", "issue:44448|issue:44743", "issue:43335|issue:43653", - "issue:45200|issue:45588", - "issue:43550|issue:44112" + "issue:43550|issue:44112", + "issue:45127|issue:45245", + "issue:45200|issue:45588" ], - "prepared_review_unit_hash": "b12d83b13dcd2c69aadc4efaf66e36119045a3778facfa494c3a6f5021397dc9", + "prepared_review_unit_hash": "a757eb708f730b8aa8a178720c929631c788133231a760017d7ac3671c9f7f08", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36313, - "estimated_input_tokens": 9079, - "estimated_eval_tokens": 18414 + "serialized_chars": 36373, + "estimated_input_tokens": 9094, + "estimated_eval_tokens": 18444 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is noisy and mostly non-duplicate: it mixes unrelated issues across multimodal processors/chat templates, model loading, hardware/version regressions, and CI/docs noise. No soft pair looks like the same underlying bug or change.", - "confidence": 0.9, - "canonical_issue_reason": "issue:44521 is the most active and central representative of the cluster, with broader multimodal chat-template scope and inbound references, making it the best umbrella-style issue here.", + "summary": "This cluster is mostly a loose lexical grouping of unrelated issue reports. I would not merge any soft edge as a duplicate; the only reasonable representative is the active chat-template multimodal issue (44521), but even that is not a true hub for the others.", + "confidence": 0.69, + "canonical_issue_reason": "issue:44521 has the most discussion/inbound references and is the most representative of a recurring API failure theme, but the overall set is too heterogeneous to justify a strong canonical duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the strongest representative issue, but only as a loose cluster anchor rather than a true duplicate target.", + "best_issue_reason": "issue:44521 is the best single representative by activity and breadth of impact; it is the clearest candidate to anchor triage, though not a duplicate of the rest.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43526", "right": "issue:46192", "accept": false, - "reason": "Related processor behavior, but one is reduce_labels truncation and the other is return_tensors being ignored; different bugs and code paths." - }, - { - "left": "issue:43278", - "right": "issue:43856", - "accept": false, - "reason": "One is a dtype drift during evaluate, the other is MoE training memory usage; not the same underlying issue." + "reason": "Both touch label/text processing, but one is a BeitImageProcessorFast label truncation bug and the other is a processor kwargs handling bug; different code paths and symptoms." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "Both involve labels, but one is gather_for_metrics dropping tuple elements and the other is BeitImageProcessorFast reduce_labels; distinct failures." + "reason": "Both involve labels in vision workflows, but one is gather_for_metrics dropping tuple labels and the other is reduce_labels returning only one label; not the same underlying defect." }, { - "left": "issue:41720", - "right": "issue:46209", + "left": "issue:44442", + "right": "issue:45198", "accept": false, - "reason": "Both are Qwen3 hardware/runtime failures, but one is a CUDA assert during device mapping and the other is FP8 inference on B200; not the same bug." + "reason": "Both are tokenizer-related failures, but one is AutoTokenizer loading FastSpeech2ConformerTokenizer and the other is Wav2Vec2 save/tokenization behavior; different models and failure modes." }, { - "left": "issue:44442", - "right": "issue:45198", + "left": "issue:41720", + "right": "issue:46209", "accept": false, - "reason": "Different speech-tokenizer/model loading failures with different APIs and symptoms." + "reason": "Both are hardware/runtime compatibility problems, but one is a Qwen3 auto device-map cudaErrorAssert on A800 and the other is an FP8 inference issue on B200; not the same bug." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate default vs batched video shape output; same processor area, but different concrete bugs." + "reason": "Audio chat-template sampling-rate handling and batched video shape handling are unrelated bugs in different processors." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Separate import/runtime compatibility problems: flash_attn on Python 3.13 versus Windows ROCm torch.distributed dependency issues." + "reason": "One is a flash_attn import key error on Python 3.13; the other is a Windows/ROCm torch.distributed dependency issue. Different platform failures." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "Zero-probability inference output vs model memory not releasing; unrelated issues." + "reason": "Unexpected zero probabilities in a vision model and RT-DETR memory not being released are unrelated issues." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent mismatch and GraniteSpeech CI device-override failure are different problems." - }, - { - "left": "issue:44206", - "right": "issue:45584", - "accept": false, - "reason": "Both are audio-related, but one is an unsupported feature-extractor arg and the other is empty-transcription generation after align_special_tokens." + "reason": "DAC latent/forward mismatch and GraniteSpeech stale device override CI failure do not share the same code path or symptom." }, { "left": "issue:44561", "right": "issue:46234", "accept": false, - "reason": "Both mention missing utilities, but they are different symbols and different regression surfaces." + "reason": "Both reference import_utils helpers, but they concern different missing/removed availability checks and different downstream breakages." }, { - "left": "issue:45759", - "right": "issue:46285", + "left": "issue:44206", + "right": "issue:45584", "accept": false, - "reason": "Both concern model/config structure, but one is text_config unwrapping for composite Qwen configs and the other is CLIP text model structure inconsistency." + "reason": "A feature-extractor center-arg regression and a Whisper empty-transcription generation failure are distinct bugs." }, { - "left": "issue:45290", - "right": "issue:46326", + "left": "issue:45759", + "right": "issue:46285", "accept": false, - "reason": "Both are chat-template/tokenizer crashes, but they hit different inputs and different code paths." + "reason": "Both concern model/config structure changes, but one is from_config unwrapping composite Qwen multimodal configs and the other is CLIP text model attribute inconsistency." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs updates around old pipeline APIs, but they concern different deprecated arguments and pages." + "reason": "Both are documentation-related, but they cover different docs and different deprecated API references; not the same change." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "Object-detection adjacent, but one is wrong processor output size and the other is an AMP CUDA crash in a loss function." + "reason": "One is a processor producing the wrong image size; the other is an AMP/CUDA crash in LwDetrImageLoss." }, { - "left": "issue:44514", - "right": "issue:44521", + "left": "issue:45290", + "right": "issue:46326", "accept": false, - "reason": "Both are multimodal apply_chat_template bugs, but one is a batched padding=False crash and the other is zeroed assistant masks; not the same bug." + "reason": "Both are chat-template/tokenizer issues, but one crashes on assistant tool-call messages and the other concerns continue_final_message for mistral_common tokenizers." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "Different helper/functionality: DAC latent mismatch versus qwen3_omni_moe feature-length calculation." + "reason": "DAC latent mismatch and a qwen3_omni_moe feature-length helper behaving unexpectedly are different models and different logic." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both involve apply_chat_template in multimodal settings, but one is a batching/padding crash and the other is all-zero assistant masks; related area, not the same bug." }, { "left": "issue:44857", "right": "issue:45593", "accept": false, - "reason": "One is a float16 AMP crash, the other is missing auxiliary losses when denoising is off; unrelated." + "reason": "An AMP loss crash and missing auxiliary losses when denoising is off are separate detection/training issues." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and Bamba torch.compile/SDPA failure are unrelated model bugs." + "reason": "SwitchTransformers sparse-layer construction and Bamba torch.compile SDPA failure are unrelated model/runtime bugs." }, { "left": "issue:43611", "right": "issue:46402", "accept": false, - "reason": "Both involve model loading, but one is base_model_prefix handling and the other is Sam3 text encoder weights not loading." + "reason": "Both are loading failures, but one is a broad base_model_prefix regression and the other is Sam3Model text encoder weights not loading; different root causes." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "CPU offload device mismatch and feature-extractor output-length behavior are unrelated." + "reason": "CPU offload device mismatch in bark-small and qwen3_omni_moe feature-length behavior are unrelated." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "The former is a specific flash_attention_utils import error, while the latter is a broad Torch 2.10 incompatibility report." + "reason": "Missing _flash_supports_window_size import and Torch 2.10 incompatibility are not the same concrete failure." }, { "left": "issue:45759", "right": "issue:46512", "accept": false, - "reason": "Both are from_config semantics, but one is text_config unwrapping and the other is stale config.dtype after initialization." + "reason": "from_config unwrapping composite multimodal configs and stale model.config.dtype after from_config are separate configuration issues." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and a Trainer _gpu_broadcast_one NoneType TypeError are different issues." + "reason": "Flash Attention 2 test failures and Trainer _gpu_broadcast_one NoneType errors are unrelated training/runtime problems." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Pegasus output differences across versions and Qwen3.5 recurrent-state reset are unrelated model behaviors." + "reason": "Pegasus v4/v5 output differences and Qwen recurrent state reset with cache are distinct model-behavior bugs." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "SwitchTransformers sparse-layer bug and BigBirdTokenizer special-token decode bug do not share the same root cause." + "reason": "SwitchTransformers sparse-layer creation and BigBirdTokenizer special-token registration are unrelated." }, { - "left": "issue:45200", - "right": "issue:45588", + "left": "issue:43550", + "right": "issue:44112", "accept": false, - "reason": "mm_token_type_ids defaults and flash_attention s_aux=None crashes are unrelated." + "reason": "torch.compile+SDPA failure in Bamba and a GraniteSpeech CI device override test failure do not share the same underlying defect." }, { - "left": "issue:43550", - "right": "issue:44112", + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge/collapsed model behavior and the 2^24 category limit runtime error are unrelated." + }, + { + "left": "issue:45200", + "right": "issue:45588", "accept": false, - "reason": "Bamba torch.compile/SDPA failure and GraniteSpeech CI device-override failure are separate code paths." + "reason": "Gemma 4 token-type defaults and flash_attention.py s_aux=None handling are different multimodal/integration bugs." } ] }, @@ -8725,8 +8725,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -8762,10 +8762,10 @@ "issue:44779", "issue:44855", "issue:44857", + "issue:44977", "issue:45072", "issue:45083", - "issue:45127", - "issue:45245", + "issue:45229", "issue:45290", "issue:45310", "issue:45381", @@ -8780,28 +8780,27 @@ "issue:46234" ], "soft_pairs": [ - "issue:45127|issue:45245", "issue:44855|issue:45083", "issue:37428|issue:44351", "issue:43335|issue:43504", - "issue:43819|issue:44493", "issue:43278|issue:43825", - "issue:44206|issue:45823", + "issue:43819|issue:44493", "issue:39692|issue:43526", "issue:29127|issue:32090", + "issue:44206|issue:45823", "issue:43335|issue:45083", "issue:45588|issue:45859", "issue:26413|issue:31501", "issue:43720|issue:44112", "issue:43335|issue:44112", - "issue:45459|issue:46234", - "issue:45310|issue:45478", "issue:43335|issue:45593", + "issue:45310|issue:45478", "issue:39692|issue:44493", + "issue:45459|issue:46234", "issue:45290|issue:45381", + "issue:34689|issue:35141", "issue:44448|issue:44779", "issue:43606|issue:44857", - "issue:34689|issue:35141", "issue:42490|issue:43854", "issue:44360|issue:44561", "issue:44857|issue:45072", @@ -8809,222 +8808,230 @@ "issue:28218|issue:32090", "issue:45920|issue:46192", "issue:43701|issue:45083", - "issue:40444|issue:43526" + "issue:40444|issue:43526", + "issue:44448|issue:44977", + "issue:44977|issue:45229" ], - "prepared_review_unit_hash": "8b9173daa05fb266d71e91f111453504f36ef199380280f86418653781286db8", + "prepared_review_unit_hash": "a52c5cf8d24243b5326ba3505407391a67c8f0d6c2732eb742ec467ba0ebdd3b", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36937, - "estimated_input_tokens": 9235, - "estimated_eval_tokens": 18726 + "soft_pair_count": 31, + "serialized_chars": 37152, + "estimated_input_tokens": 9288, + "estimated_eval_tokens": 18832 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of unrelated bug reports across different models, tokenizer behavior, loading paths, and training/runtime failures. The only clear duplicate pair is the two identical Qwen3.5 MoE `from_pretrained` regression reports.", - "confidence": 0.95, - "canonical_issue_reason": "Issue 45310 is the earliest report of the exact same Qwen3.5 MoE `from_pretrained` regression, with the duplicate 45478 repeating the same title and symptom.", + "summary": "This cluster is mostly heterogeneous and does not form a single duplicate set. The only clear duplicate pair is the Qwen3.5 MoE `from_pretrained` regression; the rest are related only at a broad subsystem level or are plainly unrelated.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45310 is the clearest canonical anchor because it directly names the shared Qwen3.5 MoE `from_pretrained` regression and is the earlier, slightly richer report.", "canonical_pr_reason": null, - "best_issue_reason": "45310 is the best representative issue because it is the original, clearest instance of the duplicated Qwen3.5 MoE loading bug.", + "best_issue_reason": "issue:45310 is the best issue to serve as the cluster representative because it is the only unambiguous duplicate target in the set and describes a concrete, reusable bug report.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:45127", - "right": "issue:45245", - "accept": false, - "reason": "Unrelated bugs: LoRA/extended-vocab collapse vs a category-count limit error." - }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "Different code paths and failures: Python import parsing error vs a feature-extractor helper behavior bug." + "reason": "Different problems: Python 3.13 import/parsing failure vs a helper length-calculation behavior change." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different missing symbols and likely different breakages." + "reason": "Both are import errors, but they are for different missing symbols and unrelated code paths." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "Different models and problems: SwitchTransformers config bug vs BEiT pretrained preset loading issue." - }, - { - "left": "issue:43819", - "right": "issue:44493", - "accept": false, - "reason": "Different failures: DAC latent decoding mismatch vs position-id key warnings across models." + "reason": "Config construction bug for SwitchTransformers vs BEiT pretrained-loading legacy-field issue." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Training/eval dtype drift is unrelated to a pipeline error message regression." + "reason": "Unrelated bugs: dtype drift during evaluation vs a pipeline error-message regression." }, { - "left": "issue:44206", - "right": "issue:45823", + "left": "issue:43819", + "right": "issue:44493", "accept": false, - "reason": "Different model families and failures: unsupported feature-extractor arg vs device mismatch in Gemma4 PLE." + "reason": "Different model-path bugs: DAC latent reconstruction mismatch vs position-id key warnings across many models." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "Documentation/example issues for SigLIP2 are not the same bug as BEiT Fast label reduction." + "reason": "Docs/example cleanup and quantization issues are not the same as a BEiT fast processor label bug." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "Clarity of a LayoutLMv3 error message is unrelated to a Trainer broadcast TypeError." + "reason": "Error-message clarity for LayoutLMv3 boxes is unrelated to Trainer broadcasting `NoneType`." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Feature-extractor argument validation and Gemma4 device placement are different failures." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug and Qwen3 Omni feature-length helper bug are different." + "reason": "SwitchTransformers sparse-layer construction is unrelated to qwen3_omni_moe output-length behavior." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Both involve Qwen3.5 MoE-adjacent code, but one is flash-attention handling and the other is CPU offload rotary embedding; not the same bug." + "reason": "Different Qwen/MoE issues: flash-attention auxiliary-state handling vs CPU-offload rotary embedding compatibility." }, { "left": "issue:26413", "right": "issue:31501", "accept": false, - "reason": "Checkpoint resume failure and FineWeb training startup failure are different symptoms and likely different root causes." + "reason": "Checkpoint resume epoch-iterator failure is not the same as a training startup problem." }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "BitNet packed-weight loading and GraniteSpeech CI device override are unrelated." + "reason": "Accelerate packed-weight loading and a GraniteSpeech CI test failure are different code paths." }, { "left": "issue:43335", "right": "issue:44112", "accept": false, - "reason": "SwitchTransformers configuration and GraniteSpeech CI failure are unrelated." + "reason": "Different subsystems and different failure modes; only broad CI/device language overlaps." }, { - "left": "issue:45459", - "right": "issue:46234", + "left": "issue:43335", + "right": "issue:45593", "accept": false, - "reason": "Tokenizer protobuf error handling and a missing import-utils symbol are different API regressions." + "reason": "Sparse-layer config bug does not match the D-FINE auxiliary-loss omission." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "Same exact Qwen3.5 MoE `from_pretrained` error report; same title, same model, same failure mode." + "reason": "Same underlying bug: identical Qwen3.5 MoE `from_pretrained` error on transformers>=5.4.0." }, { - "left": "issue:43335", - "right": "issue:45593", + "left": "issue:39692", + "right": "issue:44493", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and D-FINE auxiliary-loss handling are unrelated." + "reason": "Documentation/example issues are unrelated to the position-id key regression." }, { - "left": "issue:39692", - "right": "issue:44493", + "left": "issue:45459", + "right": "issue:46234", "accept": false, - "reason": "SigLIP2 docs/example issues are not the same as position-id key warnings across models." + "reason": "Tokenizer protobuf error masking and a missing import-utils symbol are unrelated." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat template tool-call crash and Qwen2.5-VL video position-id bug are different code paths." + "reason": "Chat-template tool-call crash is a different bug from Qwen2.5-VL video position-id handling." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Model-loading breakage for Llama vision and embedding reinitialization after resize are different issues." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Pegasus v4/v5 output difference and DeepSeek tokenizer regression are separate tokenizer/model behavior issues." + "reason": "Both affect outputs, but one is Pegasus generation drift and the other is DeepSeek tokenizer regression; not the same code path." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "CPU-offload device mismatch for Bark is unrelated to an AMP/CUDA crash in LwDetrImageLoss." - }, - { - "left": "issue:34689", - "right": "issue:35141", - "accept": false, - "reason": "Llama 3.2 Vision loading breakage and embedding reinitialization after resize are different bugs." + "reason": "CPU offload device mismatch is unrelated to float16 AMP loss crashes on CUDA." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "Local-path import warning and GLM-4.7-Flash unit-test loading failure are unrelated." + "reason": "Local-path import warning and GLM-4.7-Flash loading failure are different model-loading problems." }, { "left": "issue:44360", "right": "issue:44561", "accept": false, - "reason": "DSA indexer activation choice and removal of `is_torch_fx_available` affect different subsystems." + "reason": "DSA indexer discussion is unrelated to the removal of `is_torch_fx_available` breaking trust_remote_code models." }, { "left": "issue:44857", "right": "issue:45072", "accept": false, - "reason": "LwDetr AMP/CUDA crash and bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel are different failures." + "reason": "Different bugs in different models: LwDETR AMP crash vs dtype mismatch in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "Zero probabilities in SigLIP2 and unexpected position-id keys are unrelated model bugs." + "reason": "Unexpected zero probabilities in SigLIP2 is not the same as the position-id key regression." }, { "left": "issue:28218", "right": "issue:32090", "accept": false, - "reason": "Tokenizer spacing after added tokens is unrelated to a Trainer broadcast NoneType error." + "reason": "Tokenizer added-token spacing bug is unrelated to Trainer broadcast failures." }, { "left": "issue:45920", "right": "issue:46192", "accept": false, - "reason": "AutoTokenizer wrong token IDs and `return_tensors` being ignored are different tokenizer API issues." + "reason": "Broad AutoTokenizer ID regression across models is not the same as silently ignoring `return_tensors` when `text_kwargs` is passed." }, { "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "Resume-from-checkpoint key mismatch and Qwen3 Omni feature-length behavior are unrelated." + "reason": "Resume-from-checkpoint key mismatch and qwen3_omni_moe length helper behavior are unrelated." }, { "left": "issue:40444", "right": "issue:43526", "accept": false, - "reason": "Qwen2.5-VL iterable dataset multi-image finetuning failure is unrelated to BEiT Fast label reduction." + "reason": "IterableDataset multi-image finetuning failure is different from BEiT `reduce_labels` returning only one label." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Pegasus v4/v5 output differences are not the same underlying bug as Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU OOM are different failure classes." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is grounded in the packet: 45310 and 45478 have identical titles and matching Qwen3.5 MoE from_pretrained symptoms, and 45310 is earlier. The other pairwise rejections stay conservative and do not overclaim duplicate evidence." + "feedback": "The evaluation is grounded in the packet: the only accepted soft pair is the Qwen3.5 MoE `from_pretrained` regression between issue:45310 and issue:45478, and the rest are conservatively rejected based on clearly different titles/code paths. The canonical/representative issue choice is supported by the earlier date and slightly fuller report for issue:45310." }, "evaluator_used": true, "retried": false, @@ -9040,8 +9047,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -9094,239 +9101,231 @@ "issue:46399" ], "soft_pairs": [ - "issue:44977|issue:45229", - "issue:44448|issue:44977", "issue:32090|issue:33357", "issue:29942|issue:35141", - "issue:43746|issue:46209", "issue:43550|issue:45083", + "issue:43746|issue:46209", "issue:44855|issue:46399", "issue:42490|issue:43575", "issue:43927|issue:45698", "issue:10000|issue:15354", "issue:27957|issue:34689", - "issue:44561|issue:44779", "issue:37428|issue:44987", + "issue:44561|issue:44779", "issue:43278|issue:43329", - "issue:43873|issue:45229", "issue:44857|issue:45588", + "issue:43873|issue:45229", + "issue:43278|issue:43381", "issue:41720|issue:45229", "issue:44977|issue:45563", "issue:44568|issue:45491", - "issue:43278|issue:43381", - "issue:45588|issue:45823", - "issue:45372|issue:45588", "issue:43976|issue:45440", - "issue:43412|issue:45588", - "issue:45372|issue:45406", "issue:44448|issue:44466", + "issue:45588|issue:45823", + "issue:45372|issue:45588", "issue:45198|issue:45803", "issue:25251|issue:32090", + "issue:43412|issue:45588", "issue:32090|issue:33290", + "issue:45372|issue:45406", "issue:45245|issue:45563" ], - "prepared_review_unit_hash": "e37e724b422883d81b3145d1a9d82709e6b9011cce23c892a443239559237486", + "prepared_review_unit_hash": "ff3c8db3f2186ff49d70e67a7252d45b1b4e395b54944542fa5572d4fbf14264", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36336, - "estimated_input_tokens": 9084, - "estimated_eval_tokens": 18424 + "soft_pair_count": 28, + "serialized_chars": 35735, + "estimated_input_tokens": 8934, + "estimated_eval_tokens": 18124 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The proposed soft pairs are mostly superficial lexical matches rather than the same underlying bug. They span unrelated models, code paths, and failure modes, so none should be merged as duplicates. No single canonical issue emerges from this set.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "These items are mostly unrelated bug reports spanning different models, features, and failure modes. I found no soft-edge pair that looks like the same underlying bug or change, so all candidate links should be rejected.", + "confidence": 0.87, + "canonical_issue_reason": "issue:43873 is the broadest, highest-signal report in the set (quantization/offloading behavior) with the strongest discussion and inbound reference activity, so it is the best anchor for the cluster despite not matching the others as duplicates.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:43873 is the most reusable representative issue here: it has the broadest scope, substantial follow-up interest, and a clear user-facing regression to anchor triage.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44977", - "right": "issue:45229", - "accept": false, - "reason": "Both mention generation/runtime failure, but one is a flash-attention correctness crash for Qwen3.5 and the other is a multi-GPU OOM for Gemma4; different model and root cause." - }, - { - "left": "issue:44448", - "right": "issue:44977", - "accept": false, - "reason": "v4 vs v5 output regression for Pegasus is a broad behavior mismatch, while 44977 is a flash-attention generation crash on Qwen3.5; not the same bug." - }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Trainer broadcast NoneType error and MacOS bus error in CLIP are unrelated failure modes with different stacks and symptoms." + "reason": "Different failures: Trainer GPU broadcast NoneType vs MacOS bus error in a community CLIP model." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures and token embedding reinit after resize are separate issues affecting different parts of the model lifecycle." + "reason": "Flash Attention 2 test failures are unrelated to the token-embedding reinitialization bug in post_init." }, { - "left": "issue:43746", - "right": "issue:46209", + "left": "issue:43550", + "right": "issue:45083", "accept": false, - "reason": "Local checkpoint loading with PEFT adapters is unrelated to FP8 inference failures on B200." + "reason": "Different code paths: torch.compile/SDPA on Bamba vs feature-extract length helper in qwen3_omni_moe." }, { - "left": "issue:43550", - "right": "issue:45083", + "left": "issue:43746", + "right": "issue:46209", "accept": false, - "reason": "torch.compile/SDPA bug in Bamba-9B-v2 is not the same as the qwen3_omni_moe helper length calculation issue." + "reason": "PEFT local checkpoint loading and FP8 inference on B200 are different issues with different root causes." }, { "left": "issue:44855", "right": "issue:46399", "accept": false, - "reason": "Python 3.13 import/IndentationError in DebertaV2Model is a different code-path from Qwen3.5 rotary-pos registration missing a decorator." + "reason": "Both mention decorators/JIT, but they describe different models and distinct parsing/registration failures." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "Local-path import warning and tp-related OOM on Qwen2-57B are unrelated." + "reason": "Local-path trailing-slash warning is unrelated to the tensor-parallel OOM on Qwen2-57B-A14B-Instruct." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "One is custom token IDs being lost on save/load, the other is the wrong custom module being loaded after save_pretrained; both involve persistence but not the same defect." + "reason": "Both involve save/load, but one loses custom token IDs while the other loads the wrong custom module." }, { "left": "issue:10000", "right": "issue:15354", "accept": false, - "reason": "DistilBertModel issue and torch.jit GeneratorExp export failure are unrelated model/runtime problems." + "reason": "Different models and different problems: DistilBert raises an issue vs GeneratorExp unsupported by torch.jit.script." }, { "left": "issue:27957", "right": "issue:34689", "accept": false, - "reason": "Flash Attention 2 support on XLMRoberta and Llama 3.2 90B Vision model loading breakage are different issues." + "reason": "Flash Attention 2 on XLMRoberta is unrelated to the Llama 3.2 90B Vision model-loading regression." }, { - "left": "issue:44561", - "right": "issue:44779", + "left": "issue:37428", + "right": "issue:44987", "accept": false, - "reason": "Removal of is_torch_fx_available breaking trust_remote_code is not the same as Deepseek tokenizer producing incorrect results." + "reason": "ImportError for a flash-attention helper is a different failure from loading physical-intelligence/fast on v5.1.0." }, { - "left": "issue:37428", - "right": "issue:44987", + "left": "issue:44561", + "right": "issue:44779", "accept": false, - "reason": "Missing _flash_supports_window_size import and transformers>=5.1.0 load failure for physical-intelligence/fast are both import-related but not the same concrete bug." + "reason": "Removal of a utility function breaking trust_remote_code is not the same bug as DeepSeek tokenizer output drift." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "BF16-to-FP32 dtype change during evaluate is unrelated to undefined variables in the video branch of _get_num_multimodal_tokens." + "reason": "Embedding dtype drift in eval is unrelated to the undefined video-token helper names in multimodal token counting." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetr float16 AMP crash and flash_attention s_aux=None crash are different subsystems and different bugs." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Quantization/offloading behavior and Gemma4 multi-GPU CUDA OOM are different resource-management problems." + "reason": "Both are memory/offload themed, but one is quantization offloading behavior and the other is a Gemma4 multi-GPU OOM." }, { - "left": "issue:44857", - "right": "issue:45588", + "left": "issue:43278", + "right": "issue:43381", "accept": false, - "reason": "AMP/CUDA crash in LwDetrImageLoss and flash_attention.py AttributeError on s_aux=None are different code paths and failures." + "reason": "A dtype regression and an eval-mode gradient-checkpointing restriction are not the same underlying issue." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert is not the same underlying issue as Gemma4 multi-GPU OOM." + "reason": "Both involve distributed inference failures, but the models, errors, and code paths are distinct." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Flash-attention generation crash and stale num_return_sequences warning in paged generate are unrelated." + "reason": "Flash-attention generation corruption is unrelated to a stale warning in paged generate()." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Tokenizer special-token regression and Gemma3 NaN embeddings from sliding-window attention are different bugs." - }, - { - "left": "issue:43278", - "right": "issue:43381", - "accept": false, - "reason": "Embedding dtype drift in evaluate is unrelated to gradient checkpointing being disallowed in eval mode." + "reason": "Tokenizer special-token handling and Gemma3 NaN embeddings are different bugs with different fixes." }, { - "left": "issue:45588", - "right": "issue:45823", + "left": "issue:43976", + "right": "issue:45440", "accept": false, - "reason": "flash_attention.py crashing on s_aux=None and Gemma4 PLE device mismatch with device_map=\"auto\" are different failures." + "reason": "Python-version compatibility breakage is unrelated to the DeepSeekV3MoE parity divergence." }, { - "left": "issue:45372", - "right": "issue:45588", + "left": "issue:44448", + "right": "issue:44466", "accept": false, - "reason": "Gemma4 processor loading import error for ReasoningEffort and flash_attention AttributeError are separate dependencies/code paths." + "reason": "Both are v5 regressions, but one is output drift and the other is lm_head serialization/tied-weights handling." }, { - "left": "issue:43412", - "right": "issue:45588", + "left": "issue:45588", + "right": "issue:45823", "accept": false, - "reason": "executorch export guard/erfinv support and sink-less flash_attention crash are unrelated." + "reason": "Gemma-related, but one is a flash_attention null-aux crash and the other is a device-map mismatch in forward." }, { "left": "issue:45372", - "right": "issue:45406", - "accept": false, - "reason": "Both concern Gemma4 loading, but one is a missing mistral_common import and the other is a missing _tokenizer attribute in serve; not the same bug." - }, - { - "left": "issue:44448", - "right": "issue:44466", + "right": "issue:45588", "accept": false, - "reason": "Pegasus v4/v5 output divergence and lm_head.weight serialization inconsistency are separate v5 regressions." + "reason": "Gemma 4 processor import dependency failure is not the same as the flash_attention s_aux=None crash." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Wav2Vec2 save_pretrained/tokenization failure and Fuyu bare-except exception swallowing are unrelated." + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare-except error handling." }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "top_k changing pipeline output shape is unrelated to Trainer _gpu_broadcast_one NoneType errors." + "reason": "Pipeline top_k nesting behavior and Trainer GPU broadcast errors are unrelated." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export/training-guard failure is a different code path from the flash_attention crash." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "Trainer broadcast NoneType and adafactor+deepspeed OOM are different training failures." + "reason": "Trainer broadcast NoneType and deepspeed Adafactor OOM are different problems." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma 4 processor import failure and Gemma4Processor missing _tokenizer in serve are separate breakages." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "2^24 category limit error and stale paging warning are unrelated generate-time issues." + "reason": "Category-count runtime limit and stale paged-generate warning are unrelated." } ] }, @@ -9345,8 +9344,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -9402,23 +9401,23 @@ "soft_pairs": [ "issue:43653|issue:44743", "issue:42490|issue:43873", - "issue:43065|issue:45412", "issue:33357|issue:39692", - "issue:44206|issue:44479", - "issue:43065|issue:43994", + "issue:43065|issue:45412", + "issue:29127|issue:34567", "issue:44361|issue:45198", + "issue:44206|issue:44479", "issue:44186|issue:45198", - "issue:29127|issue:34567", + "issue:43065|issue:43994", "issue:43611|issue:46121", "issue:44871|issue:44977", "issue:42915|issue:42994", - "issue:45459|issue:45712", "issue:43278|issue:43575", + "issue:45459|issue:45712", "issue:45412|issue:46369", "issue:29942|issue:30064", - "issue:45372|issue:45459", - "issue:43994|issue:45412", "issue:43381|issue:43701", + "issue:43994|issue:45412", + "issue:45372|issue:45459", "issue:43066|issue:44462", "issue:44265|issue:45198", "issue:43756|issue:45440", @@ -9436,226 +9435,226 @@ "issue:44442|issue:45072", "issue:42490|issue:43381" ], - "prepared_review_unit_hash": "a748152172d8c7a50c5df8f188883319352ac4d9ab5618f5caa72607e6edcc56", + "prepared_review_unit_hash": "cd089768a08ee7335e57c35308530eed076886ae2c25194da8a7f99018610067", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 35, - "serialized_chars": 38114, - "estimated_input_tokens": 9529, - "estimated_eval_tokens": 19314 + "serialized_chars": 38119, + "estimated_input_tokens": 9530, + "estimated_eval_tokens": 19316 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly unrelated and do not form a real duplicate cluster; the soft pairs share only superficial subsystem similarity (tokenizers, quantization, vision, flash-attn, etc.).", - "confidence": 0.67, - "canonical_issue_reason": "No true canonical duplicate exists here; if one issue must anchor the set, issue 43873 is the broadest active bug report with the most follow-up activity.", + "summary": "This cluster is a loose collection of unrelated Transformers bug reports spanning tokenizers, model loading, quantization, multimodal processors, and training/runtime regressions. None of the suggested soft edges look like true duplicates of the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the best standalone representative because it is open, broadly framed, and has the strongest discussion/reference footprint.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "Different bugs: BigBirdTokenizer special-token decoding vs Qwen3.5 recurrent-state reset." + "reason": "Different problems: BigBird tokenizer special-token registration vs Qwen3.5 recurrent-state reset in cached generation." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Unrelated: local-path import warning vs quantization/offloading behavior." - }, - { - "left": "issue:43065", - "right": "issue:45412", - "accept": false, - "reason": "Different models and failures: Sam3PixelDecoder dummy Conv2d vs RT-DETR memory release." + "reason": "Local path trailing-slash import warning is unrelated to quantization offloading behavior." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS CLIP bus error is unrelated to a SigLIP2 documentation example issue." + "reason": "MacOS bus error in a community CLIP model is not the same as SigLIP2 doc example errors." }, { - "left": "issue:44206", - "right": "issue:44479", + "left": "issue:43065", + "right": "issue:45412", "accept": false, - "reason": "Different code paths: LasrFeatureExtractor center-arg crash vs Qwen VL video regression." + "reason": "Dummy Conv2d in Sam3PixelDecoder and RT-DETR memory leak are different model-specific bugs." }, { - "left": "issue:43065", - "right": "issue:43994", + "left": "issue:29127", + "right": "issue:34567", "accept": false, - "reason": "Sam3PixelDecoder dummy layer is not the same as SigLIP2 AutoModel/pipeline outputs." + "reason": "LayoutLMv3 error messaging and TrainerState token counting are unrelated subsystems." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "MLukeTokenizer task AttributeError differs from Wav2Vec2 save/tokenization failure." + "reason": "MLukeTokenizer task failure is unrelated to Wav2Vec2 save/tokenization failures." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Unsupported center arg in LasrFeatureExtractor is not the same as the video-input regression." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "LayoutLMv2 tokenizer crash is unrelated to Wav2Vec2 save_pretrained/tokenization." + "reason": "Different components and failures: LayoutLMv2 tokenizer padding/truncation vs Wav2Vec2 save_pretrained/tokenization." }, { - "left": "issue:29127", - "right": "issue:34567", + "left": "issue:43065", + "right": "issue:43994", "accept": false, - "reason": "LayoutLMv3 error messaging and TrainerState token counting are unrelated bugs." + "reason": "Sam3PixelDecoder dummy layer is unrelated to SigLIP2 wrong outputs through AutoModel/pipeline." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "Base_model_prefix loading regression is not the same as rope-parameter JSON list handling." + "reason": "base_model_prefix loading regression and RoPE-parameter JSON list handling are different code paths." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma eos_token_id mismatch is unrelated to Qwen3.5 flash-attention generation failure." + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation issues." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "FineGrainedFP8Config load failure and quantized saving failure are separate problems." + "reason": "Both involve quantization, but one is FineGrainedFP8Config failure and the other is saving quantized models; not the same bug." }, { - "left": "issue:45459", - "right": "issue:45712", + "left": "issue:43278", + "right": "issue:43575", "accept": false, - "reason": "Tokenizer protobuf error masking is unrelated to leftover dummy classes in repo checks." + "reason": "Embedding dtype drift in eval and TP-related OOM on Qwen2-57B are different failures." }, { - "left": "issue:43278", - "right": "issue:43575", + "left": "issue:45459", + "right": "issue:45712", "accept": false, - "reason": "Embedding dtype drift in eval is not the same as Qwen2-57B tensor-parallel OOM." + "reason": "Tokenizer error masking with missing protobuf is unrelated to leftover dummy classes and repo checks." }, { "left": "issue:45412", "right": "issue:46369", "accept": false, - "reason": "RT-DETR GC/memory leak is different from SAM3 litetext memory overflow." + "reason": "RT-DETR GC/memory leak and sam3-litetext memory overflow are different model-specific memory issues." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to void segmentation map processing." + "reason": "Flash Attention 2 test failures are not the same as void segmentation map processing failures." }, { - "left": "issue:45372", - "right": "issue:45459", + "left": "issue:43381", + "right": "issue:43701", "accept": false, - "reason": "Gemma processor import breakage is unrelated to protobuf decode-error masking." + "reason": "Eval-mode gradient checkpointing limitation and resume_from_checkpoint key mismatch are unrelated." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 nonsensical outputs and RT-DETR memory leak are different bugs." + "reason": "SigLIP2 wrong outputs and RT-DETR memory leak are different bugs in different models." }, { - "left": "issue:43381", - "right": "issue:43701", + "left": "issue:45372", + "right": "issue:45459", "accept": false, - "reason": "Eval-mode gradient checkpointing and resume checkpoint key mismatch are separate issues." + "reason": "Gemma 4 processor import failure from mistral_common is unrelated to protobuf-missing tokenizer error masking." }, { "left": "issue:43066", "right": "issue:44462", "accept": false, - "reason": "Wrong tokenizer decoder type is not the same as AutoTokenizer ignoring tokenizer.json." + "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring repository tokenizer.json." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "torch.export failures around torch_compilable_check are unrelated to Wav2Vec2 save/tokenization." + "reason": "torch.export failures with torch_compilable_check are unrelated to Wav2Vec2 tokenization/save issues." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Smollm3 RoPE-layer mismatch and DeepSeekV3MoE divergence are distinct model-implementation bugs." + "reason": "Smollm3 RoPE-layer discrepancy and DeepseekV3MoE divergence are separate architecture implementation bugs." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBird mask-token registration does not match FastSpeech2ConformerTokenizer loading." + "reason": "BigBird mask-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Runtime special-token post-processor updates are unrelated to Gemma3 NaN embeddings." + "reason": "Runtime special-token post-processing and Gemma3 mixed-length NaN embeddings are unrelated." }, { "left": "issue:27957", "right": "issue:29942", "accept": false, - "reason": "XLMRoberta FA2 support request is not the same as failing FA2 tests." + "reason": "XLMRoberta FA2 support and FA2 test failures are not the same concrete bug." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "LayoutLMv3 error clarity and MacOS CLIP bus error are unrelated." + "reason": "LayoutLMv3 error clarity and CLIP MacOS bus error are unrelated." }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "DeepSeekV3MoE divergence is unrelated to D-FINE auxiliary-loss behavior." + "reason": "DeepseekV3MoE implementation drift is unrelated to D-FINE auxiliary-loss behavior." }, { "left": "issue:43994", "right": "issue:46369", "accept": false, - "reason": "SigLIP2 inference regression is not the same as SAM3 litetext memory overflow." + "reason": "SigLIP2 nonsensical outputs and sam3-litetext memory overflow are different issues." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Kosmos2.5 OCR length indexing." + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Kosmos2.5 OCR long-input index error." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "LayoutLMv3 error handling and output-embedding reinitialization are different bugs." + "reason": "LayoutLMv3 error messaging and embedding reinitialization after resize are different bugs." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "Qwen2_5_VL import failure is not the same as a generic dependency issue." + "reason": "Specific Qwen2.5-VL import failure is too different from a generic dependency issue." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "BeitImageProcessorFast label reduction and LayoutLMv2 tokenizer crashes are unrelated." + "reason": "Different tokenizer/image-processor problems with no shared code path." }, { "left": "issue:43575", @@ -9667,13 +9666,13 @@ "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "Tokenizer loading failure and bfloat16 dtype mismatches are different code paths." + "reason": "AutoTokenizer loading failure and dtype mismatch in inference are unrelated." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Local-path slash import warning is unrelated to eval-mode gradient checkpointing." + "reason": "Trailing-slash local model import warning is unrelated to eval-mode gradient checkpointing." } ] }, @@ -9692,8 +9691,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -9747,231 +9746,231 @@ "issue:46399" ], "soft_pairs": [ - "issue:42915|issue:43931", "issue:43575|issue:44387", + "issue:42915|issue:43931", "issue:43761|issue:45859", "issue:46121|issue:46399", - "issue:43526|issue:44857", "issue:43334|issue:43531", + "issue:43526|issue:44857", "issue:44977|issue:45245", "issue:43262|issue:44206", "issue:44464|issue:45491", "issue:27957|issue:29127", "issue:27957|issue:33290", "issue:45588|issue:45600", - "issue:44743|issue:44811", "issue:43646|issue:44077", "issue:33290|issue:34689", + "issue:44743|issue:44811", "issue:33290|issue:35141", "issue:33290|issue:34567", "issue:43381|issue:43873", "issue:44442|issue:44898", "issue:44062|issue:45907", "issue:45003|issue:45440", - "issue:43723|issue:45479", "issue:42490|issue:44743", + "issue:43723|issue:45479", "issue:43316|issue:43525", - "issue:43761|issue:44079", "issue:42175|issue:46234", "issue:34689|issue:43611", - "issue:43720|issue:44442" + "issue:43720|issue:44442", + "issue:43761|issue:44079" ], - "prepared_review_unit_hash": "11913479c049d9f2820e608cedeb3d59cb8671fdbdcd8bfb27a10c486ff31585", + "prepared_review_unit_hash": "514fecb6b45c35431e57dfeb69e531d01671a3ab7033e546a3692322f6dd95cd", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36119, - "estimated_input_tokens": 9030, - "estimated_eval_tokens": 18316 + "serialized_chars": 36113, + "estimated_input_tokens": 9029, + "estimated_eval_tokens": 18314 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of unrelated Transformers issues. None of the soft pairs look like the same underlying bug or change, so all soft edges are rejected. The best anchor is the active, widely referenced quantization/offloading issue.", - "confidence": 0.78, - "canonical_issue_reason": "Issue 43873 is the most active and externally referenced report in the set, so it is the strongest cluster anchor despite the broad topic mix.", + "summary": "This is a heterogeneous similarity cluster: the soft pairs share only broad keywords, not the same underlying bug or fix path. No pair looks safe to merge as a duplicate; issue 43873 is the broadest open anchor if one must pick a representative issue.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43873 is the broadest, still-open report with inbound references, so it is the best cluster anchor; the others are narrower and cover unrelated failure modes.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 has the highest discussion/reference signal and a clear user-facing regression-style failure mode, making it the most suitable representative issue.", + "best_issue_reason": "Issue 43873 is the best global issue candidate because it is open, broadly framed, and already attracting references; it is the most plausible triage hub even though the set is not a true duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42915", - "right": "issue:43931", + "left": "issue:43575", + "right": "issue:44387", "accept": false, - "reason": "Different models and failure modes: FP8 config failure vs Qwen3-VL weight-shape mismatch." + "reason": "Both mention OOM, but one is tensor-parallel model loading and the other is int4 quantization memory growth; different code paths." }, { - "left": "issue:43575", - "right": "issue:44387", + "left": "issue:42915", + "right": "issue:43931", "accept": false, - "reason": "Both mention OOM, but one is tensor-parallel model loading and the other is int4 quantization memory growth; not the same bug." + "reason": "Different models and failures: FineGrainedFP8Config crash vs weight-shape mismatch during Qwen3-VL loading." }, { "left": "issue:43761", "right": "issue:45859", "accept": false, - "reason": "Unrelated code paths: CLIPVision hidden states regression vs Qwen3.5 MoE rotary embedding CPU offload incompatibility." + "reason": "A CLIPVision hidden-states regression and a Qwen3 MoE CPU-offload incompatibility are unrelated bugs." }, { "left": "issue:46121", "right": "issue:46399", "accept": false, - "reason": "Both are rope-related, but one is a TypeError in validation dict conversion and the other is a kernel registration ValueError." + "reason": "Both are rope-related, but one is JSON/list validation in config conversion and the other is a missing decorator for rotary-pos-emb registration." }, { - "left": "issue:43526", - "right": "issue:44857", + "left": "issue:43334", + "right": "issue:43531", "accept": false, - "reason": "Different subsystems: Beit label reduction bug vs LW-DETR AMP/CUDA loss crash." + "reason": "Pad-token-id API mismatch and sliding-window recurrent-state behavior are different model issues." }, { - "left": "issue:43334", - "right": "issue:43531", + "left": "issue:43526", + "right": "issue:44857", "accept": false, - "reason": "Different Qwen model families and different issues: missing pad_token_id vs sliding-window behavior." + "reason": "BEiT label reduction and LwDetr float16 AMP crash affect different components and code paths." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Flash-attention generation bug is unrelated to the 2^24 category-count runtime error." + "reason": "Flash-attention generation failure and category-cardinality limit are unrelated." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Both audio-related, but one is chat-template sample-rate defaulting and the other is an unsupported center argument crash." + "reason": "Audio chat-template sampling-rate defaulting and unsupported center-arg crash are different processor bugs." }, { "left": "issue:44464", "right": "issue:45491", "accept": false, - "reason": "Different generation failures: compiled-forward inconsistency vs NaNs from mixed-length sliding-window padding." + "reason": "Compiled-forward inconsistency and Gemma3 sliding-window NaNs are distinct symptoms with different causes." }, { "left": "issue:27957", "right": "issue:29127", "accept": false, - "reason": "XLM-R flash-attention support and LayoutLMv3 error clarity are unrelated." + "reason": "XLM-Roberta flash-attention support and LayoutLMv3 error messaging are unrelated." }, { "left": "issue:27957", "right": "issue:33290", "accept": false, - "reason": "Flash-attention model support vs DeepSpeed Adafactor OOM are not the same bug." + "reason": "Flash-attention model support and Adafactor/deepspeed OOM are not the same bug." }, { "left": "issue:45588", "right": "issue:45600", "accept": false, - "reason": "Flash-attention sink handling and auto_mappings cleanup are different issues with different code paths." - }, - { - "left": "issue:44743", - "right": "issue:44811", - "accept": false, - "reason": "Qwen recurrent-state reset bug is unrelated to Whisper processor skip_special_tokens behavior." + "reason": "A flash_attention sink-less-model crash and a removed auto-mapping reference in CI are different issues." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "A general custom initialization regression is not the same as patchtsmixer's post_init restriction." + "reason": "Both touch initialization semantics, but one is a broad v5 custom-model init regression and the other is a specific patchtsmixer post_init policy change." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "Different regressions: Adafactor+DeepSpeed OOM vs Llama 3.2 Vision loading breakage." + "reason": "OOM with Adafactor/deepspeed and Llama 3.2 Vision load failure are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Recurrent-state reset in cache handling and Whisper batch_decode skip_special_tokens behavior are separate paths." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "OOM in optimizer setup is unrelated to embedding reinitialization during post_init." + "reason": "Training OOM and embedding reinitialization in post_init are different bugs." }, { "left": "issue:33290", "right": "issue:34567", "accept": false, - "reason": "Memory exhaustion and token-count tracking are unrelated." + "reason": "OOM under deepspeed is unrelated to TrainerState token counter not updating." }, { "left": "issue:43381", "right": "issue:43873", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction is a different issue from quantization/offloading behavior." + "reason": "Gradient checkpointing in eval mode and quantization/offloading behavior are different subsystems." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "Tokenizer loading failure and Perceiver non-default resolution failure are unrelated." + "reason": "Tokenizer loading failure and Perceiver non-default-resolution failure are not the same issue." }, { "left": "issue:44062", "right": "issue:45907", "accept": false, - "reason": "AddedToken keyword collision is unrelated to list_repo_files revision forwarding." + "reason": "AddedToken special-argument duplication and list_repo_files revision forwarding are unrelated API bugs." }, { "left": "issue:45003", "right": "issue:45440", "accept": false, - "reason": "Unsafe sys.modules access and DeepSeekV3 implementation divergence are different bugs." - }, - { - "left": "issue:43723", - "right": "issue:45479", - "accept": false, - "reason": "Tokenizer loading in v5 and degenerate sequence-classification loss are unrelated." + "reason": "Unsafe sys.modules access and DeepseekV3MoE divergence from remote implementation are different code-path problems." }, { "left": "issue:42490", "right": "issue:44743", "accept": false, - "reason": "Local-path import warning and Qwen recurrent-state reset are different code paths." + "reason": "Trailing-slash import warning and recurrent-state reset do not describe the same bug." }, { - "left": "issue:43316", - "right": "issue:43525", + "left": "issue:43723", + "right": "issue:45479", "accept": false, - "reason": "Both are config-API mismatches, but they affect different models and different missing fields." + "reason": "AutoTokenizer loading regression and degenerate zero-loss classification are unrelated." }, { - "left": "issue:43761", - "right": "issue:44079", + "left": "issue:43316", + "right": "issue:43525", "accept": false, - "reason": "CLIPVision hidden-state regression is unrelated to ModelOutput key assignment." + "reason": "Gemma3TextConfig API discrepancy and Llama4Config missing pad_token_id are similar in theme only, not the same concrete bug." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Package/backend installation issue and missing import-utils helper are not the same underlying problem." + "reason": "Backend extras installation and a missing import_utils helper are unrelated packaging/API issues." }, { "left": "issue:34689", "right": "issue:43611", "accept": false, - "reason": "Different version-specific loading regressions with different root causes." + "reason": "Llama 3.2 Vision loading breakage and base_model_prefix loading regression are different model-loading failures." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "BitNet packed-weight loading and tokenizer loading failure are unrelated." + "reason": "Accelerate packed-weights handling and tokenizer loading are separate bugs." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment bug affect different code paths." } ] }, @@ -9990,8 +9989,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -10046,25 +10045,25 @@ ], "soft_pairs": [ "issue:43232|issue:43388", - "issue:43931|issue:45216", "issue:44077|issue:44479", "issue:43761|issue:44188", "issue:43493|issue:43827", - "issue:42907|issue:45698", + "issue:43931|issue:45216", "issue:44987|issue:46178", + "issue:42907|issue:45698", "issue:44442|issue:45803", - "issue:42796|issue:45290", "issue:43873|issue:44871", + "issue:42796|issue:45290", "issue:43749|issue:43931", "issue:43316|issue:45759", - "issue:44464|issue:46421", "issue:43701|issue:46506", + "issue:44464|issue:46421", "issue:44190|issue:44496", "issue:43643|issue:44913", "issue:44855|issue:45137", "issue:46234|issue:46291", - "issue:43295|issue:43825", "issue:25251|issue:28282", + "issue:43295|issue:43825", "issue:43550|issue:43701", "issue:42617|issue:43575", "issue:43606|issue:43701", @@ -10072,190 +10071,190 @@ "issue:27957|issue:32090", "issue:16998|issue:22355" ], - "prepared_review_unit_hash": "4790d2dcdb722b6cb39b2fbef0cc4a304d80aaaa441f89472287357926a08624", + "prepared_review_unit_hash": "cc98b05da001c39d42f9d2aea6d103984c59f1ca775134a8776c867877bd1a93", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 26, - "serialized_chars": 35523, - "estimated_input_tokens": 8881, - "estimated_eval_tokens": 18018 + "serialized_chars": 35526, + "estimated_input_tokens": 8882, + "estimated_eval_tokens": 18020 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No duplicate cluster emerges here: the items are a heterogeneous set of unrelated Transformers issues across loading, generation, configs, quantization, and docs. None of the soft edges look like the same underlying bug/change, so no merges are warranted.", + "summary": "The cluster is heterogeneous: the items cover unrelated bugs across generation, loading, tokenization, config, compilation, and docs. I do not see any true duplicate pair among the soft edges, so all candidate merges should be rejected.", "confidence": 0.98, - "canonical_issue_reason": "No single canonical issue is appropriate; the cluster is not a duplicate set and the soft-linked issues describe different code paths and failure modes.", - "canonical_pr_reason": "No pull requests are present in the cluster.", - "best_issue_reason": "No issue is a clear canonical representative because the cluster spans unrelated bugs rather than one underlying defect.", - "best_pr_reason": "No pull requests are present in the cluster.", + "canonical_issue_reason": "No canonical issue: this is not a duplicate cluster, so there is no single issue that cleanly represents the whole set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43761 is the strongest standalone report here: it is a specific, reproducible regression with a clear symptom (`hidden_states=None` despite `output_hidden_states=True`). It is still not a duplicate hub for the rest of the items.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Different subsystems and failures: generation kwargs after sync_gpus vs metric gathering dropping tuple labels." - }, - { - "left": "issue:43931", - "right": "issue:45216", - "accept": false, - "reason": "Both involve Qwen models, but one is a loading shape-mismatch and the other is a save_pretrained regression." + "reason": "Different areas: generation kwargs update vs metric gathering/last-batch label truncation." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "Unrelated: patchtsmixer post_init API change vs a v5 video-input regression for Qwen VL models." + "reason": "Unrelated bugs: patchtsmixer post_init validation vs Qwen video-input regression." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "Different code paths: CLIPVision hidden_states bug vs torch.compile attention-kernel branching divergence." + "reason": "Different code paths: CLIPVision hidden states vs attention-kernel branching under torch.compile." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation mismatch is unrelated to pipeline removal/docs referencing pipeline()." + "reason": "Model implementation parity vs docs still referencing removed pipelines; not the same bug." }, { - "left": "issue:42907", - "right": "issue:45698", + "left": "issue:43931", + "right": "issue:45216", "accept": false, - "reason": "Both touch save/load behavior, but one is dequantized model saving and the other is custom module resolution after save_pretrained." + "reason": "Both are Qwen VL-related, but one is shape mismatch loading and the other is save_pretrained checkpoint corruption." }, { "left": "issue:44987", "right": "issue:46178", "accept": false, - "reason": "Separate failures: loading a specific model repo vs backend lookup error for tensorflow_text." + "reason": "Model loading failure for a specific model vs backend-mapping error for tensorflow_text; unrelated causes." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Dequantized Ministral save failure vs wrong custom module loaded after save_pretrained; different defects." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer loading failure and a bare-except bug in Fuyu tensor conversion are different bugs." + "reason": "Tokenizer loading failure vs Fuyu tensor conversion swallowing exceptions; not the same underlying issue." }, { - "left": "issue:42796", - "right": "issue:45290", + "left": "issue:43873", + "right": "issue:44871", "accept": false, - "reason": "Tokenizer BPE-marker handling is unrelated to chat template crashes on tool-call assistant messages." + "reason": "Quantization/offloading behavior vs Gemma eos_token_id config mismatch; no shared bug." }, { - "left": "issue:43873", - "right": "issue:44871", + "left": "issue:42796", + "right": "issue:45290", "accept": false, - "reason": "Quantization/offload behavior is unrelated to Gemma-3 eos_token_id config inconsistency." + "reason": "Tokenizer BPE marker handling vs chat-template tool-call crash; separate features and failures." }, { "left": "issue:43749", "right": "issue:43931", "accept": false, - "reason": "Different model-loading problems: FSDP CPU RAM loading vs Qwen3-VL weight-shape mismatch." + "reason": "FSDP CPU-RAM-efficient loading vs Qwen3-VL shape mismatch; different loading problems." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Related multimodal config area, but different models and bugs: Gemma3TextConfig API mismatch vs Qwen text_config unwrapping." + "reason": "API discrepancy in Gemma3TextConfig vs missing text_config unwrap in multimodal AutoModelForCausalLM.from_config." }, { - "left": "issue:44464", - "right": "issue:46421", + "left": "issue:43701", + "right": "issue:46506", "accept": false, - "reason": "Shared chunked-generation vocabulary, but one is output inconsistency and the other is spurious recompilation from cache lazy init." + "reason": "Checkpoint-key mismatch vs CPU/CUDA output divergence; unrelated symptoms." }, { - "left": "issue:43701", - "right": "issue:46506", + "left": "issue:44464", + "right": "issue:46421", "accept": false, - "reason": "Resume checkpoint key mismatch is unrelated to CPU/CUDA output differences in PPDocLayoutV3." + "reason": "Both involve chunking/compile, but one is inconsistent generation outputs and the other is recompilation from cache init; not same bug." }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Local dataset loading failure is unrelated to unrecognized model/config.json errors." + "reason": "Local dataset loading in a training example vs unrecognized model config loading; unrelated." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "AutoConfig missing fields with trust_remote_code is not the same as GPTNeoXConfig rotary_pct not persisting on reload." + "reason": "trust_remote_code config-field loss vs rotary_pct reset on reload; different serialization/config issues." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Python 3.13 torch.jit parsing/IndentationError is unrelated to DeepSpeed ZeRO3 deque underflow." + "reason": "Python 3.13 JIT parse/IndentationError vs DeepSpeed ZeRO3 deque underflow; unrelated runtime failures." }, { "left": "issue:46234", "right": "issue:46291", "accept": false, - "reason": "Missing import_utils symbol and kernels v0.15 import failure are different dependency/import issues." + "reason": "Missing import_utils helper vs kernels version/revision import failure in pipelines; different packaging problems." }, { - "left": "issue:43295", - "right": "issue:43825", + "left": "issue:25251", + "right": "issue:28282", "accept": false, - "reason": "Different regressions: processor.tokenizer/image-tokenizer compatibility vs an error message wrongly mentioning translation pipelines." + "reason": "Pipeline output nesting bug vs missing PyTorch dependency ImportError; not the same issue." }, { - "left": "issue:25251", - "right": "issue:28282", + "left": "issue:43295", + "right": "issue:43825", "accept": false, - "reason": "Pipeline top_k output shape change is unrelated to PyTorch-not-found ImportError." + "reason": "Custom model/processor regression vs incorrect v5 translation error message; different problems." }, { "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "torch.compile+SDPA failure is unrelated to resume_from_checkpoint key mismatch." + "reason": "torch.compile + SDPA failure vs resume_from_checkpoint key mismatch; unrelated code paths." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py execution failure and tensor-parallel OOM are not the same concrete bug." + "reason": "3d_parallel script failure vs tensor-parallel OOM loading Qwen2-57B; different resource/launch issues." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch is unrelated to checkpoint resume key mismatch." + "reason": "CPU offload device mismatch vs checkpoint key mismatch; no shared underlying defect." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "Tensor-parallel OOM for Qwen2-57B is unrelated to MobileLLM key errors on load." + "reason": "TP OOM loading a large model vs key error loading MobileLLM; unrelated model-loading failures." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 model support and Trainer _gpu_broadcast_one NoneType errors are unrelated." + "reason": "FlashAttention2 support question vs Trainer broadcast NoneType error; different subsystems." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "DeBERTa model_max_length behavior is unrelated to a missing transformers.onnx module." + "reason": "model_max_length question vs missing transformers.onnx module; unrelated." } ] }, @@ -10274,8 +10273,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -10330,23 +10329,23 @@ ], "soft_pairs": [ "issue:43316|issue:43716", + "issue:43295|issue:44871", "issue:45632|issue:46424", "issue:44610|issue:46369", - "issue:43295|issue:44871", "issue:42175|issue:43316", - "issue:43650|issue:46202", "issue:45803|issue:46365", "issue:44871|issue:45584", - "issue:43065|issue:44617", "issue:44077|issue:44534", "issue:43540|issue:44514", + "issue:43065|issue:44617", + "issue:43650|issue:46202", "issue:44162|issue:44792", "issue:45803|issue:46170", "issue:44871|issue:46050", "issue:41950|issue:43475", "issue:43646|issue:44913", - "issue:43404|issue:43441", "issue:43452|issue:44496", + "issue:43404|issue:43441", "issue:43761|issue:44479", "issue:45127|issue:45200", "issue:43388|issue:43688", @@ -10357,196 +10356,196 @@ "issue:44734|issue:45406", "issue:44617|issue:45584" ], - "prepared_review_unit_hash": "d66fe126a21896b38a8a45af8ced835501104cd35547d94ec9b811799a54aca3", + "prepared_review_unit_hash": "7bede5940b667dc993b307f45ed08e524764215b80afa9bbc2f21fdc5c023e83", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35842, - "estimated_input_tokens": 8961, - "estimated_eval_tokens": 18178 + "serialized_chars": 35847, + "estimated_input_tokens": 8962, + "estimated_eval_tokens": 18180 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a mixed bag of unrelated Transformers issues; none of the soft-edge pairs look like the same underlying bug or change, so all candidate merges are rejected. For a representative issue, I\u2019d pick 45698 because it\u2019s the most broadly framed and most actively discussed save/load regression in the set.", - "confidence": 0.93, - "canonical_issue_reason": "issue:45698 is the strongest representative by activity and breadth: it has the most discussion, inbound references, and a generic from_pretrained/save_pretrained loading failure that is easy to recognize as a real user-facing regression.", + "summary": "Mostly a mixed bag of unrelated Transformers regressions and model-specific bugs; the soft pairs share broad subsystems at best, but not the same concrete failure mode. I would not merge any of the suggested pairs.", + "confidence": 0.43, + "canonical_issue_reason": "Issue 45698 is the strongest representative because it is open, active, and describes a broad save/load regression with custom modules that is likely to attract duplicate reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45698 is the best single issue to anchor the cluster because it is open, high-activity, and describes a clear, broadly impactful model-loading bug.", + "best_issue_reason": "Issue 45698 is the best overall triage target for the cluster because it is the most active and generic failure mode among the issues listed.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Different models and failures: Gemma3TextConfig API mismatch vs Mistral-3 image preprocessor dtype mismatch." + "reason": "Different bugs: config API discrepancy vs image-preprocessor dtype mismatch." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Processor.tokenizer regression and Gemma eos_token_id mismatch are unrelated." }, { "left": "issue:45632", "right": "issue:46424", "accept": false, - "reason": "Both involve caching, but one is a remote-code path collision and the other is a max_cache_length behavior bug." + "reason": "Cache-path collision and max_cache_length behavior are different cache issues." }, { "left": "issue:44610", "right": "issue:46369", "accept": false, - "reason": "Unrelated SAM3/OmDet model issues: wrong input resolution vs memory overflow." - }, - { - "left": "issue:43295", - "right": "issue:44871", - "accept": false, - "reason": "Different regressions: processor.tokenizer access/image-tokenizer handling vs Gemma eos_token_id config inconsistency." + "reason": "Different models and failure modes: input size mismatch vs OOM." }, { "left": "issue:42175", "right": "issue:43316", "accept": false, - "reason": "Backend packaging problem vs model config API discrepancy; no shared underlying bug." - }, - { - "left": "issue:43650", - "right": "issue:46202", - "accept": false, - "reason": "Generic noise title vs a Gemma pipeline roundtrip bug; not the same issue." + "reason": "Backend installation dependency issue vs config API discrepancy." }, { "left": "issue:45803", "right": "issue:46365", "accept": false, - "reason": "Different code paths: Fuyu tensor conversion error handling vs Trainer import dependency conflict." + "reason": "Bare-except exception handling bug vs Trainer import conflict are unrelated." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Both are generation-related, but one is eos_token_id config and the other is empty transcription handling in Whisper." - }, - { - "left": "issue:43065", - "right": "issue:44617", - "accept": false, - "reason": "Same product family, but dummy Conv2d in the decoder and CUDA OOM are not the same bug." + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure are different code paths." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "Optional post_init validation vs non-persistent buffers initialization; different regressions." + "reason": "Both are v5-related, but one is post_init policy and the other is non-persistent buffer initialization." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Both are multimodal/video-related, but they fail in different processors and different call paths." + "reason": "Both involve Qwen processors, but video-input validation and batched chat-template padding are separate bugs." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "SAM3 PixelDecoder dummy layer and SAM3 video OOM are not the same problem." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "The titles indicate unrelated issues with no shared failure mechanism." }, { "left": "issue:44162", "right": "issue:44792", "accept": false, - "reason": "Unrelated model breakage vs a janus test failure; no shared concrete defect." + "reason": "ESM2 breakage and janus image-generation test failure are different models/issues." }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Bare except swallowing errors is unrelated to the Windows/ROCm distributed import issue." + "reason": "Different root causes: tensor conversion exception handling vs Windows/ROCm import dependency failure." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "Gemma token-id inconsistency and sinq save/load corruption are different bugs." + "reason": "Gemma config mismatch and sinq save/load quantization failure are separate save/load bugs." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Both mention video/SAM-like models, but one is processor lookup and the other is a missing output attribute." + "reason": "Both are video-related, but one misroutes processor selection and the other is a missing vision-encoder attribute." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "Broad v5 custom-init regression vs a specific GPTNeoX config reload issue; not the same code-path." + "reason": "General custom model initialization regression vs GPTNeoX config reload defaulting are different issues." }, { - "left": "issue:43404", - "right": "issue:43441", + "left": "issue:43452", + "right": "issue:44496", "accept": false, - "reason": "Same model family, but weight tying and FlashAttention failures are distinct problems." + "reason": "gguf loading failure and unrecognized Olmo model config are separate model-loading problems." }, { - "left": "issue:43452", - "right": "issue:44496", + "left": "issue:43404", + "right": "issue:43441", "accept": false, - "reason": "Different loading failures: gguf_file handling vs missing model_type in a config.json." + "reason": "Same model family, but weight tying and FlashAttention failures are distinct code paths." }, { "left": "issue:43761", "right": "issue:44479", "accept": false, - "reason": "Different regressions across different models and modalities; not one underlying bug." + "reason": "CLIP vision hidden-state regression and Qwen video-input regression are unrelated." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "LoRA/extended-vocab collapse is unrelated to Gemma 4 mm_token_type_ids defaulting." + "reason": "LoRA/extended-vocab collapse and Gemma4 mm_token_type_ids defaulting are different bugs." }, { "left": "issue:43388", "right": "issue:43688", "accept": false, - "reason": "Metric gathering label truncation and auxiliary loss normalization are separate training bugs." + "reason": "Metric gathering label truncation and auxiliary-loss normalization are unrelated training issues." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both involve cache/state handling, but one is a global GenerationMixin cache regression and the other is a Qwen3.5 recurrent-state reset bug." + "reason": "Generic GenerationMixin cache issue and Qwen recurrent-state reset are not the same concrete bug." }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "StableLm config field regression and tensorflow_text backend registration are unrelated." + "reason": "StableLm config field regression and tensorflow_text backend mapping error are unrelated." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "Legacy field loading for BEiT presets is a different issue from custom-module misloading after save/load." + "reason": "Legacy preset-field loading failure and wrong custom-module loading after save_pretrained are distinct serialization bugs." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "TensorBoard/TensorFlow backend confusion is unrelated to a missing import_utils symbol." + "reason": "TensorFlow backend detection and missing import_utils symbol are different import/backend issues." }, { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "Both are serve crashes, but one is KV-cache tensor indexing and the other is a missing Gemma4Processor attribute." + "reason": "Serve KV-cache indexing crash and Gemma4Processor missing _tokenizer are separate serve regressions." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "SAM3 video OOM and Whisper empty-transcription failure are unrelated bugs." + "reason": "SAM3 video OOM and Whisper empty-transcription failure are unrelated." } ] }, @@ -10565,8 +10564,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -10576,6 +10575,7 @@ "issue:41720", "issue:42490", "issue:43065", + "issue:43335", "issue:43525", "issue:43531", "issue:43540", @@ -10598,7 +10598,6 @@ "issue:44610", "issue:44617", "issue:44734", - "issue:44779", "issue:44792", "issue:44855", "issue:44877", @@ -10616,18 +10615,17 @@ "issue:45588", "issue:45600", "issue:46369", - "issue:46421", - "issue:46489" + "issue:46421" ], "soft_pairs": [ "issue:43673|issue:46421", "issue:43525|issue:43531", "issue:44792|issue:44977", "issue:45381|issue:46369", - "issue:43931|issue:44060", "issue:44792|issue:45600", - "issue:44190|issue:44964", "issue:44534|issue:44913", + "issue:44190|issue:44964", + "issue:43931|issue:44060", "issue:43761|issue:44242", "issue:45563|issue:45584", "issue:35141|issue:42490", @@ -10635,222 +10633,216 @@ "issue:44479|issue:44734", "issue:43827|issue:43994", "issue:43065|issue:44610", - "issue:41720|issue:45245", "issue:44933|issue:45362", + "issue:41720|issue:45245", "issue:44617|issue:46369", "issue:44464|issue:45563", "issue:43644|issue:44913", - "issue:44521|issue:45381", "issue:43550|issue:44190", "issue:43976|issue:44855", + "issue:44521|issue:45381", "issue:44077|issue:44877", "issue:25251|issue:27957", "issue:45127|issue:45357", "issue:43540|issue:44734", - "issue:44779|issue:46489" + "issue:43335|issue:44190" ], - "prepared_review_unit_hash": "10844b3a11780ac1120281309297923d26889b88fdda3995c46c7908f1f3718b", + "prepared_review_unit_hash": "42bad97681888888bc9f9bcdac8d5a32b29c659f5b4ebba3706a72799278df13", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35898, - "estimated_input_tokens": 8975, - "estimated_eval_tokens": 18206 + "serialized_chars": 35345, + "estimated_input_tokens": 8837, + "estimated_eval_tokens": 17930 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous, but one clear duplicate pair stands out: the v5 non-persistent-buffer regression. Most other pairs are related only at a broad subsystem or model-family level and should stay separate.", - "confidence": 0.78, - "canonical_issue_reason": "issue:43644 is the best canonical issue because it is the earliest and clearest report of the shared \"non-persistent buffers filled with junk\" v5 regression; issue:44534 is effectively the same bug with later wording.", + "summary": "This cluster is heterogeneous: the soft pairs share only broad surface themes (cache/chunked prefill, multimodal/video, config/loading, Python/version issues) but not the same concrete bug or fix. I did not find any true duplicate issue pairs.", + "confidence": 0.93, + "canonical_issue_reason": "No clear canonical issue: the set spans many unrelated bugs/regressions, and the apparent similarities are superficial rather than the same underlying problem.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the strongest issue to keep as the representative anchor for the only clear duplicate pair in this set.", + "best_issue_reason": "issue:44521 is the strongest single representative only because it is open and has the most inbound references/discussion, but it is still just a weak stand-in for this heterogeneous set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43673", "right": "issue:46421", "accept": false, - "reason": "Both mention chunked prefill/cache behavior, but one is a missing cache regression and the other is a recompilation/perf issue; not clearly the same bug." + "reason": "Both mention chunked prefill/cache, but one is about a missing GenerationMixin cache in v5.0.0 and the other about lazy cache init causing recompilation; different failure modes." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Different models and different failures: missing pad_token_id vs sliding_window handling." + "reason": "Different model/config bugs: missing pad_token_id on Llama4Config vs sliding_window handling in Qwen3-MoE." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Different model families and symptoms; janus test failure is not the same as Qwen3.5 flash-attention generation failure." + "reason": "A failed janus test and a Qwen3.5 flash-attention generation issue are not the same code-path problem." }, { "left": "issue:45381", "right": "issue:46369", "accept": false, - "reason": "Unrelated issues: Qwen2.5-VL video position IDs vs Sam3-lite text memory overflow." + "reason": "Both involve multimodal/video-ish models, but one is a vision position-id regression and the other is a memory overflow on sam3-litetext." }, { - "left": "issue:43931", - "right": "issue:44060", + "left": "issue:44792", + "right": "issue:45600", "accept": false, - "reason": "Different bugs in Qwen3-VL loading vs tied-weights warning behavior." + "reason": "Test failure for janus vs broken auto_mappings references for removed Sam3LiteText configs; unrelated bugs." }, { - "left": "issue:44792", - "right": "issue:45600", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "Different components: Janus image-generation test vs Sam3LiteText auto-mapping/CI breakage." + "reason": "Non-persistent buffer junk initialization and rotary_pct reloading are different config/state bugs." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Different workflows and different models: local dataset loading vs Phi-4 multimodal model loading." + "reason": "A local dataset loading failure and a model weight-loading mismatch are different issues." }, { - "left": "issue:44534", - "right": "issue:44913", + "left": "issue:43931", + "right": "issue:44060", "accept": false, - "reason": "One is a v5 buffer initialization regression; the other is a GPTNeoX config reload bug." + "reason": "Different Qwen model families and different symptoms: shape mismatch on load vs tied-weights warning logic." }, { "left": "issue:43761", "right": "issue:44242", "accept": false, - "reason": "Different failures in different model paths: CLIP hidden_states regression vs MoE load-balancing loss omission." + "reason": "CLIP hidden_states regression and load-balancing loss omission are separate model behaviors." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "Paged generate warning handling and Whisper empty-transcription generation are unrelated." + "reason": "Stale num_return_sequences warning in paged generate is unrelated to Whisper failing on empty transcription." }, { "left": "issue:35141", "right": "issue:42490", "accept": false, - "reason": "Different areas entirely: embedding resize/post_init reinit vs local-path trailing-slash import warning." + "reason": "Embedding resize/post_init reinit bug is unrelated to the trailing-slash local model import warning." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 JIT parsing error is unrelated to the flash-attention s_aux=None crash." + "reason": "Python 3.13 decorator parsing failure and flash_attention s_aux=None crash are different problems." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Both are runtime regressions, but one is multimodal video input handling and the other is serve/KV-cache continuation indexing." + "reason": "Qwen VL video regression and serve KV-cache continuation indexing are distinct code paths." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs referencing pipeline() is not the same as SigLIP pipeline output corruption." + "reason": "Docs still referencing pipeline() after removals is not the same as SigLIP pipeline producing nonsense outputs." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "Different model paths: Sam3PixelDecoder dummy Conv2d vs OmDet-Turbo processor/model input-size mismatch." + "reason": "Dummy Conv2d in Sam3PixelDecoder and OmDet-Turbo 640px vs 224px mismatch are unrelated vision bugs." }, { - "left": "issue:41720", - "right": "issue:45245", + "left": "issue:44933", + "right": "issue:45362", "accept": false, - "reason": "Qwen3 device-mapping CUDA assert and category-count limit are unrelated." + "reason": "Nonexistent import from image_utils and Qwen3.5 chat crashes are different failures." }, { - "left": "issue:44933", - "right": "issue:45362", + "left": "issue:41720", + "right": "issue:45245", "accept": false, - "reason": "Missing import from image_utils is unrelated to Qwen3.5 chat crashes." + "reason": "Qwen3 auto device mapping CUDA assert and category-count overflow have no shared underlying fix." }, { "left": "issue:44617", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but one is Sam3Video OOM and the other is Sam3LiteText overflow; not the same bug." + "reason": "Both mention memory overflow, but they affect different models and likely different causes." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Compiled-forward chunked-generation inconsistency is a different issue from the stale num_return_sequences warning." + "reason": "Compiled chunked generation output inconsistency is unrelated to a stale generate warning." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffer junk and GPTNeoX rotary_pct reload loss are separate config/runtime bugs." - }, - { - "left": "issue:44521", - "right": "issue:45381", - "accept": false, - "reason": "Different multimodal problems: assistant_mask generation vs video vision_position_ids." + "reason": "Non-persistent buffer initialization junk and GPTNeoX rotary_pct reload default are different state/config bugs." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "torch.compile+SDPA failure is unrelated to local dataset loading." + "reason": "torch.compile/SDPA failure in Bamba is unrelated to local dataset loading in the training script." }, { "left": "issue:43976", "right": "issue:44855", "accept": false, - "reason": "Python version support regression is unrelated to the DebertaV2 @torch.jit.script parsing failure." + "reason": "Python version support mismatch and a Python 3.13 decorator parsing error are not the same issue." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal, but all-zero assistant_masks and wrong vision_position_ids are different bugs in different processing steps." }, { "left": "issue:44077", "right": "issue:44877", "accept": false, - "reason": "Optional post_init restriction and strict granite_speech config loading are different config-policy issues." + "reason": "patchtsmixer post_init policy and granite_speech strict config loading are unrelated configuration issues." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "Pipeline top_k nesting behavior and XLMRoberta Flash Attention 2 are unrelated." + "reason": "Pipeline top_k nesting bug and XLMRoberta Flash Attention 2 support are entirely different topics." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "Different bugs: LoRA/extended-vocab model collapse vs incorrect visual encoder keys in save_pretrained." + "reason": "LoRA merge with extended vocab collapse and incorrect visual encoder keys in save_pretrained are different regressions." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Different failures: Qwen3OmniMoe video-processing ValueError vs serve KV-cache continuation indexing crash." + "reason": "Qwen3OmniMoe video-input ValueError and serve KV cache continuation indexing are unrelated." }, { - "left": "issue:44779", - "right": "issue:46489", + "left": "issue:43335", + "right": "issue:44190", "accept": false, - "reason": "Related tokenizer family, but not the same concrete bug: generic DeepSeek v5 regression vs a specific DeepSeek-Coder v1 gap in an earlier fix." - }, - { - "left": "issue:43644", - "right": "issue:44534", - "accept": true, - "reason": "Same underlying v5 regression: non-persistent buffers are initialized with junk; 44534 is a later duplicate of 43644." + "reason": "Sparse layer creation in SwitchTransformersConfig and local dataset loading are not the same bug." } ] }, @@ -10869,8 +10861,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -10882,7 +10874,6 @@ "issue:42796", "issue:43116", "issue:43295", - "issue:43335", "issue:43404", "issue:43643", "issue:43650", @@ -10902,6 +10893,7 @@ "issue:44683", "issue:44734", "issue:44737", + "issue:44779", "issue:44938", "issue:44964", "issue:45003", @@ -10920,32 +10912,33 @@ "issue:46234", "issue:46421", "issue:46424", + "issue:46489", "issue:46506" ], "soft_pairs": [ - "issue:43335|issue:44190", + "issue:44779|issue:46489", "issue:44462|issue:45907", "issue:44112|issue:45600", - "issue:42796|issue:44521", "issue:44737|issue:45803", + "issue:42796|issue:44521", "issue:43116|issue:44190", - "issue:45632|issue:46234", "issue:44938|issue:45542", "issue:43295|issue:45276", "issue:45276|issue:45584", - "issue:43404|issue:45216", + "issue:45632|issue:46234", "issue:36246|issue:41628", "issue:45245|issue:45584", "issue:46082|issue:46121", "issue:36246|issue:38617", - "issue:45632|issue:46421", + "issue:43404|issue:45216", "issue:42175|issue:44279", "issue:44246|issue:45003", - "issue:44683|issue:45446", + "issue:45632|issue:46421", "issue:44186|issue:44734", + "issue:44683|issue:45446", "issue:41950|issue:44016", - "issue:44295|issue:46506", "issue:43643|issue:44496", + "issue:44295|issue:46506", "issue:43854|issue:44964", "issue:44964|issue:45600", "issue:43673|issue:46424", @@ -10953,208 +10946,208 @@ "issue:43650|issue:44220", "issue:44016|issue:44220" ], - "prepared_review_unit_hash": "c051c9d7ec195a5ea440a844b78f6abd89bd7d87bd152118deb00030a6418903", + "prepared_review_unit_hash": "bbc0de35a1b8d2543d86afa9a89430d026b76fc1a6548d54056f0a93642cd5dd", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35879, - "estimated_input_tokens": 8970, - "estimated_eval_tokens": 18196 + "serialized_chars": 36439, + "estimated_input_tokens": 9110, + "estimated_eval_tokens": 18476 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: the soft-similarity pairs usually share a broad area like imports, caching, or multimodal processing, but they do not look like the same concrete bug. I would not merge any of the proposed soft edges.", - "confidence": 0.84, - "canonical_issue_reason": "If a representative issue must be chosen, issue 44964 is the broadest recent multimodal model-loading regression in the set; however, the cluster does not form a true duplicate group.", + "summary": "This cluster is mostly a grab bag of unrelated transformers issues; none of the soft pairs look like true duplicates. If forced to anchor the cluster, issue #44521 is the strongest representative because it has a concrete bug report, active discussion, inbound references, and a clear multimodal API path.", + "confidence": 0.78, + "canonical_issue_reason": "Issue #44521 is the best anchor: it describes a concrete, reproducible bug in `apply_chat_template` for multimodal inputs, has multiple references/discussion, and is specific enough to guide triage.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44964 is the most central user-facing failure among these issues, but it is still not a duplicate of the others.", + "best_issue_reason": "Issue #44521 is the most suitable global representative because it is substantive, well-scoped, and directly actionable; the rest are mostly isolated model/import/config regressions with no clear duplicate relationship.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43335", - "right": "issue:44190", + "left": "issue:44779", + "right": "issue:46489", "accept": false, - "reason": "Different bugs: SwitchTransformers config layer construction vs an image-classification example failing to read a local dataset." + "reason": "Both are tokenizer regressions, but one is DeepSeek generally and the other is DeepSeek-Coder v1 with a specific v5+ gap; not the same concrete bug." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "One is AutoTokenizer ignoring tokenizer.json; the other is list_repo_files not forwarding revision. Different code paths and symptoms." + "reason": "`AutoTokenizer` loading `tokenizer.json` and `list_repo_files` revision forwarding are unrelated code paths." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "Unrelated CI failures in different areas: GraniteSpeech test device override vs auto_mappings referencing removed configs." - }, - { - "left": "issue:42796", - "right": "issue:44521", - "accept": false, - "reason": "Both touch tokenization, but one is BPE marker handling and the other is assistant mask generation for multimodal chat templates." + "reason": "Both mention CI, but one is a GraniteSpeech test failure and the other is stale auto-mapping references; different bugs." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "XLNet device placement bug vs FuyuBatchFeature exception handling bug; no shared concrete failure." + "reason": "CPU device placement in XLNet and a bare `except` in Fuyu tensor conversion are unrelated issues." }, { - "left": "issue:43116", - "right": "issue:44190", + "left": "issue:42796", + "right": "issue:44521", "accept": false, - "reason": "Both are example-script problems, but one is multi-label classification output logic and the other is image-classification local dataset loading." + "reason": "Tokenizer BPE marker handling and multimodal assistant mask generation are different failure modes and code paths." }, { - "left": "issue:45632", - "right": "issue:46234", + "left": "issue:43116", + "right": "issue:44190", "accept": false, - "reason": "Trust-remote-code cache path collision is unrelated to a missing import-utils availability flag." + "reason": "Example script classification output and local dataset loading in a different script are not the same bug." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Python 3.14 import failure is not the same as the TensorFlow/tensorboard backend dependency issue." + "reason": "Python 3.14 import failure and missing TensorFlow backend wiring are separate environment/dependency issues." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Different multimodal/model bugs: processor.tokenizer regression vs gemma4 embedding resizing behavior." + "reason": "Processor/tokenizer regression and Gemma embedding resizing are different model behaviors with no shared concrete fix." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Gemma4 embedding resize issue is unrelated to Whisper empty-transcription generation after align_special_tokens." + "reason": "Gemma embedding resizing and Whisper empty transcription generation are unrelated." }, { - "left": "issue:43404", - "right": "issue:45216", + "left": "issue:45632", + "right": "issue:46234", "accept": false, - "reason": "Mistral3 lm_head tying bug is separate from Qwen3.5 save_pretrained checkpoint corruption." + "reason": "A `trust_remote_code` cache-path collision and a missing availability helper are different utilities." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different modules." + "reason": "Both are import errors, but they involve different symbols and different modules." }, { "left": "issue:45245", "right": "issue:45584", "accept": false, - "reason": "Category cardinality limit error is unrelated to Whisper generation on empty transcription." + "reason": "Category-count overflow and Whisper empty-transcription failure are unrelated." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "LlamaConfig head_dim validation and rope-parameter JSON list handling are distinct config-validation bugs." + "reason": "Llama head-dim validation and rope-validation list handling are distinct config-validation bugs." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Different missing imports: Qwen2_5_VLImageProcessor vs layer_type_validation." + "reason": "Different missing imports in different modules; not the same underlying defect." }, { - "left": "issue:45632", - "right": "issue:46421", + "left": "issue:43404", + "right": "issue:45216", "accept": false, - "reason": "Cache-path collision for local models is unrelated to chunked-prefill cache lazy initialization recompilation." + "reason": "One is a weight-tying bug in Mistral3 generation, the other is a save_pretrained regression in Qwen3.5." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Specific packaging bug around torch extras vs a generic dependency issue; no evidence of the same defect." + "reason": "A missing TensorFlow extra and a generic dependency issue are too broad and not the same bug." }, { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Import slowness and unsafe sys.modules access are different problems with different failure modes." + "reason": "Import latency and unsafe `sys.modules` access may be related at a high level, but they are not the same concrete problem." }, { - "left": "issue:44683", - "right": "issue:45446", + "left": "issue:45632", + "right": "issue:46421", "accept": false, - "reason": "Both involve flex_attention, but one is compiled flex_attention on torch>=2.9 and the other is a wrong version check for AuxRequest import." + "reason": "Both mention cache, but one is a remote-code path collision and the other is chunked-prefill recompilation from lazy cache init." }, { "left": "issue:44186", "right": "issue:44734", "accept": false, - "reason": "LayoutLMv2Tokenizer NER/padding crash is unrelated to KV-cache continuation tensor indexing in transformers serve." + "reason": "Tokenizer crashes for NER padding/truncation and KV-cache continuation indexing in serve are different subsystems." }, { - "left": "issue:41950", - "right": "issue:44016", + "left": "issue:44683", + "right": "issue:45446", "accept": false, - "reason": "Video-classification pipeline image-processor lookup vs a notebook syntax error are not the same bug." + "reason": "Both touch flex_attention, but one is a torch>=2.9 runtime failure and the other is a bad version check for an import gate." }, { - "left": "issue:44295", - "right": "issue:46506", + "left": "issue:41950", + "right": "issue:44016", "accept": false, - "reason": "Position_ids buffer reading error and CPU/CUDA output divergence in PPDocLayoutV3 are different issues." + "reason": "Video-classification pipeline behavior and a notebook syntax error are unrelated." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "trust_remote_code missing fields and missing model_type in config.json are separate config-loading failures." + "reason": "Missing fields with `trust_remote_code=True` and an unrecognized model/config missing `model_type` are different loader failures." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "Reading registered `position_ids` and CPU/CUDA output divergence in PPDocLayoutV3 are unrelated." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both are model-loading failures, but for different models and likely different incompatibilities." + "reason": "Different model-loading failures for different models; no evidence of a shared code-path bug." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "Phi-4 multimodal loading regression is unrelated to stale auto_mappings entries breaking CI." + "reason": "Phi-4 multimodal loading and stale Sam3LiteText auto-mapping references are unrelated." }, { "left": "issue:43673", "right": "issue:46424", "accept": false, - "reason": "Both mention cache, but one is missing cache during chunked_prefill and the other is max_cache_length semantics for quantized cache." + "reason": "Chunked-prefill cache initialization and `max_cache_length` honoring are both cache-related but not the same defect." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "CI test failure in GraniteSpeech does not match a user-facing multimodal model load regression." + "reason": "CI test flakiness in GraniteSpeech and loading Phi-4 multimodal are unrelated." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "Placeholder data issue vs _torch_extract_fbank_features failure; no shared code path." + "reason": "The placeholder/spam report and `_torch_extract_fbank_features()` are unrelated." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax error and audio feature extraction bug are unrelated." + "reason": "Notebook syntax error and feature-extraction bug are different issues." } ] }, @@ -11173,8 +11166,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -11209,11 +11202,10 @@ "issue:44987", "issue:45042", "issue:45083", - "issue:45103", "issue:45216", "issue:45335", - "issue:45372", "issue:45375", + "issue:45405", "issue:45584", "issue:45588", "issue:45593", @@ -11230,230 +11222,248 @@ "soft_pairs": [ "issue:27957|issue:35141", "issue:43479|issue:45375", - "issue:45083|issue:46506", "issue:43531|issue:45600", + "issue:45083|issue:46506", "issue:43295|issue:45335", "issue:43976|issue:44987", - "issue:46123|issue:46369", - "issue:43526|issue:43901", "issue:32090|issue:35141", "issue:28282|issue:35141", - "issue:42907|issue:43441", + "issue:43526|issue:43901", "issue:44496|issue:45042", - "issue:43441|issue:45588", + "issue:46123|issue:46369", + "issue:42907|issue:43441", "issue:44016|issue:44062", - "issue:43299|issue:45216", + "issue:43441|issue:45588", "issue:43673|issue:44336", "issue:44792|issue:46424", "issue:43381|issue:45593", "issue:43824|issue:44933", "issue:43976|issue:44279", "issue:43901|issue:44496", + "issue:43299|issue:45216", "issue:42796|issue:43650", - "issue:45335|issue:45584", - "issue:44279|issue:46515", "issue:44496|issue:45759", + "issue:44279|issue:46515", + "issue:45335|issue:45584", "issue:44075|issue:44368", "issue:43577|issue:45216", "issue:46082|issue:46399", - "issue:45103|issue:45372" + "issue:44279|issue:45405" ], - "prepared_review_unit_hash": "318d4d3e9d7cb40e221eae46d0d84e8cc99114de1d2dfa8f242e42b2fec8e14f", + "prepared_review_unit_hash": "c985b7a90d8c903ced6e04837aba8f9bdab7306e951f197a4afde8f2cb43ca46", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36524, - "estimated_input_tokens": 9131, - "estimated_eval_tokens": 18518 + "serialized_chars": 35834, + "estimated_input_tokens": 8959, + "estimated_eval_tokens": 18174 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab-bag of unrelated Transformers issues across different models, configs, and subsystems. None of the soft pairs appears to be the same underlying bug/change, so all candidate edges are rejected.", - "confidence": 0.92, - "canonical_issue_reason": "No single canonical issue stands out: the items span many distinct regressions and model-specific bugs, and none of the soft pairs is a true duplicate.", + "summary": "This cluster is highly heterogeneous: the paired issues share broad vocabulary in a few cases (FlashAttention, model/config loading, tokenizer behavior), but they do not look like the same underlying bug/change. No soft edge is strong enough to merge.", + "confidence": 0.91, + "canonical_issue_reason": "No single issue cleanly represents the cluster because it contains many unrelated bugs. If forced to pick one, issue 45588 is the most concrete runtime code-path failure and one of the few directly tied to an implementation crash rather than a broad symptom.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45584 is the strongest standalone representative because it is a concrete, well-scoped open bug with multiple inbound references and an actionable failure mode.", + "best_issue_reason": "Issue 45588 is the best standalone representative only in the weak sense that it is specific, reproducible, and points at a named failing integration path. It is still not a good global canonical for this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "Different bugs: Flash Attention 2 support for XLMRoberta vs output embedding reinit after resize when tie_word_embedding is False." + "reason": "Different subsystems and failure modes: FlashAttention support for XLMRoberta vs embedding reinitialization after resize_token_embeddings." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Both are config-related, but they concern different models and different failure modes: default None handling vs a missing vision field dropped by strict validation." + "reason": "Both are config-related, but they concern different models and different bugs (default sub-config initialization vs a missing field dropped by strict config handling)." }, { - "left": "issue:45083", - "right": "issue:46506", + "left": "issue:43531", + "right": "issue:45600", "accept": false, - "reason": "Unrelated: a qwen3_omni_moe helper-length behavior issue vs CPU/CUDA result divergence in PPDocLayoutV3." + "reason": "Qwen3-MoE sliding window behavior is unrelated to stale auto_mappings references for removed Sam3LiteText configs." }, { - "left": "issue:43531", - "right": "issue:45600", + "left": "issue:45083", + "right": "issue:46506", "accept": false, - "reason": "Different root causes and areas: Qwen3-MoE sliding_window behavior vs stale auto_mappings references breaking CI." + "reason": "Different models and code paths: a feature-extraction length helper in qwen3_omni_moe vs CPU/CUDA numeric divergence in PPDocLayoutV3." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Distinct regressions: processor.tokenizer/image passing broke in v4.57.5, while t5gemma resize_token_embeddings fails to update decoder.embed_tokens." + "reason": "One is a regression in processor/tokenizer image handling; the other is decoder embedding resize behavior in t5gemma." }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Not the same bug: one is a Python-version compatibility problem, the other is a specific model loading failure for physical-intelligence/fast." - }, - { - "left": "issue:46123", - "right": "issue:46369", - "accept": false, - "reason": "Different symptoms and code paths: final partial batch handling in MaskGenerationPipeline vs memory overflow in sam3-litetext." - }, - { - "left": "issue:43526", - "right": "issue:43901", - "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to a docs/behavior mismatch in TextClassificationPipeline." + "reason": "Version/platform compatibility issue vs a model-loading failure for a specific repo; not the same bug." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer _gpu_broadcast_one rejecting NoneType is unrelated to token embedding reinitialization after resize." + "reason": "Trainer broadcast TypeError is unrelated to embedding reinitialization after resizing token embeddings." }, { "left": "issue:28282", "right": "issue:35141", "accept": false, - "reason": "Missing PyTorch import support is a dependency issue, not the same as the embedding reinit bug." + "reason": "Missing PyTorch import error is unrelated to token-embedding reinitialization." }, { - "left": "issue:42907", - "right": "issue:43441", + "left": "issue:43526", + "right": "issue:43901", "accept": false, - "reason": "Both mention Ministral/flash attention, but one is about saving dequantized models and the other is a runtime flash-attention failure; different code paths." + "reason": "ImageProcessor label reduction bug and pipeline docs mismatch are unrelated." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "An unrecognized model/config problem is unrelated to PIL image processors incorrectly requiring torchvision." + "reason": "Model config recognition failure is unrelated to PIL backend torchvision dependency handling." }, { - "left": "issue:43441", - "right": "issue:45588", + "left": "issue:46123", + "right": "issue:46369", "accept": false, - "reason": "Same broad flash-attention area, but different concrete failures and models: Ministral-3 support vs s_aux=None crash in sink-less models." + "reason": "Partial-batch handling in MaskGenerationPipeline is unrelated to memory overflow in sam3-litetext." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Both involve Ministral-family models, but one is about saving dequantized weights and the other about FlashAttention loading/runtime failure." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Notebook syntax error and AddedToken keyword-argument conflict are clearly different issues." + "reason": "Notebook syntax error is unrelated to AddedToken keyword argument duplication." }, { - "left": "issue:43299", - "right": "issue:45216", + "left": "issue:43441", + "right": "issue:45588", "accept": false, - "reason": "Different Qwen regressions: one blocks model loading, the other corrupts save_pretrained output." + "reason": "Both mention FlashAttention, but the reported failures are different and one is model-specific while the other is a generic sink-less-model AttributeError; insufficient evidence they are the same bug." }, { "left": "issue:43673", "right": "issue:44336", "accept": false, - "reason": "Generation cache missing in v5.0.0 during chunked_prefill has nothing to do with ANSI codes in loading_report." + "reason": "Generation cache missing during chunked prefill is unrelated to ANSI code emission in loading_report." }, { "left": "issue:44792", "right": "issue:46424", "accept": false, - "reason": "A janus test failure is not the same as max_cache_length only being honored for Quantized Cache." + "reason": "Janus image-generation test failure is unrelated to cache config max length semantics." }, { "left": "issue:43381", "right": "issue:45593", "accept": false, - "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated." + "reason": "Gradient checkpointing in eval mode and missing auxiliary losses in D-FINE are unrelated behaviors." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Missing exported Qwen2_5_VL symbol vs a nonexistent image_utils import are different import failures." + "reason": "Both are import-related, but they concern different missing symbols/modules and different underlying causes." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "These are both dependency-themed but not the same defect: Python compatibility vs a generic transformers dependency complaint." + "reason": "Python version support issue is unrelated to a generic dependency issue." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Docs mentioning return_all_scores is unrelated to model config loading failing on an unrecognized repository." + "reason": "Documentation mismatch for TextClassificationPipeline is unrelated to model config recognition failure." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both involve Qwen multimodal models, but one is a load-time breakage and the other is incorrect save_pretrained checkpoint output." }, { "left": "issue:42796", "right": "issue:43650", "accept": false, - "reason": "Tokenizer BPE-marker handling and a vague data request do not describe the same bug." + "reason": "Tokenizer BPE marker handling is unrelated to the vague 'ADD THE DATA' issue." }, { - "left": "issue:45335", - "right": "issue:45584", + "left": "issue:44496", + "right": "issue:45759", "accept": false, - "reason": "Different model behaviors: token embedding resize not propagating vs Whisper empty transcription failure after align_special_tokens." + "reason": "Unrecognized model config and failing to unwrap text_config are different multimodal-config bugs." }, { "left": "issue:44279", "right": "issue:46515", "accept": false, - "reason": "A generic dependency issue is not the same as a torch 2.4.1 incompatibility report." + "reason": "General dependency issue is not the same as torch 2.4.1 incompatibility." }, { - "left": "issue:44496", - "right": "issue:45759", + "left": "issue:45335", + "right": "issue:45584", "accept": false, - "reason": "One is about an unrecognized model config, the other about AutoModelForCausalLM.from_config not unwrapping text_config in composite multimodal configs." + "reason": "Embedding resize not propagating to decoder.embed_tokens is unrelated to Whisper empty-transcription generation after token alignment." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "SGD args being ignored is unrelated to a tie_word_embeddings warning in Qwen3.5 LoRA fine-tuning." + "reason": "Optimizer argument handling is unrelated to a LoRA warning about tie_word_embeddings." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "BLIP2 dtype propagation and Qwen3.5 checkpoint saving are unrelated issues." }, { "left": "issue:46082", "right": "issue:46399", "accept": false, - "reason": "LlamaConfig head_dim validation and a Qwen3.5 rotary-pos-emb registration error are different concrete problems." + "reason": "LlamaConfig head_dim validation is unrelated to a Qwen3.5 kernel registration/decorator crash." }, { - "left": "issue:45103", - "right": "issue:45372", + "left": "issue:46082", + "right": "issue:46369", + "accept": false, + "reason": "LlamaConfig validation is unrelated to sam3-litetext memory overflow." + }, + { + "left": "issue:43577", + "right": "issue:45584", "accept": false, - "reason": "auto_docstring crashing with future annotations and a mistral_common import breakage are unrelated failures." + "reason": "Dtype loading inconsistency for BLIP2 is unrelated to Whisper generation on empty transcription." + }, + { + "left": "issue:44075", + "right": "issue:46515", + "accept": false, + "reason": "SGD argument usage and torch version incompatibility are unrelated." } ] }, @@ -11472,8 +11482,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -11495,7 +11505,6 @@ "issue:43976", "issue:44016", "issue:44242", - "issue:44279", "issue:44291", "issue:44297", "issue:44351", @@ -11507,17 +11516,17 @@ "issue:44779", "issue:45020", "issue:45081", + "issue:45103", "issue:45200", "issue:45276", "issue:45278", "issue:45325", "issue:45357", + "issue:45372", "issue:45405", "issue:45542", "issue:45593", - "issue:45600", "issue:45656", - "issue:45712", "issue:45812", "issue:45823", "issue:45834", @@ -11527,217 +11536,210 @@ "issue:46512" ], "soft_pairs": [ - "issue:44279|issue:45405", - "issue:38175|issue:44617", - "issue:43650|issue:46209", + "issue:45103|issue:45372", "issue:41950|issue:45020", - "issue:43299|issue:45357", "issue:44351|issue:45278", + "issue:38175|issue:44617", + "issue:43650|issue:46209", "issue:44242|issue:45200", - "issue:44521|issue:45325", + "issue:43299|issue:45357", "issue:43295|issue:44779", "issue:39692|issue:43901", + "issue:44521|issue:45325", "issue:41950|issue:45823", "issue:44683|issue:45542", - "issue:44779|issue:45812", "issue:43976|issue:45405", + "issue:44779|issue:45812", "issue:46082|issue:46512", "issue:43299|issue:43854", "issue:43479|issue:43525", "issue:44297|issue:44462", "issue:44466|issue:45276", "issue:44016|issue:44291", - "issue:43441|issue:45834", "issue:45081|issue:45907", + "issue:43441|issue:45834", "issue:43653|issue:45593", "issue:42994|issue:43317", - "issue:45600|issue:45712", "issue:33290|issue:45656" ], - "prepared_review_unit_hash": "f0436fc5996b8032c3a98d91f57a2294ad621e0bc15b75f04bdad42d0dfea914", + "prepared_review_unit_hash": "7f409c6712cd5a00e4f16551e622639afce881f370a84caba09ab0102004771d", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 26, - "serialized_chars": 35453, - "estimated_input_tokens": 8864, - "estimated_eval_tokens": 17984 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 25, + "serialized_chars": 34690, + "estimated_input_tokens": 8673, + "estimated_eval_tokens": 17602 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is not a true duplicate cluster: the items are mostly unrelated Transformers bug reports, with only superficial overlap around regressions, tokenizers, multimodal models, and deepspeed. None of the soft pairs look like the same underlying bug or change.", - "confidence": 0.92, - "canonical_issue_reason": "issue:44521 is the strongest representative by engagement and breadth of impact: it has the most discussion/inbound refs in this set and is a concrete, user-facing regression in a core multimodal path.", + "summary": "This cluster is heterogeneous: most items are distinct regressions or docs/issues touching broad Transformer subsystems, not duplicates of one another. None of the soft pairs look like the same concrete bug/change, so all are rejected.", + "confidence": 0.91, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the best overall issue candidate because it is active, clearly described, and centrally about a recurring Transformers regression surface (multimodal chat-template behavior).", + "best_issue_reason": "issue:44779 is the closest broad anchor because it describes a clear v4\u2192v5 tokenizer regression, which is the most generally representative problem theme in this set; however, it is still not a true canonical duplicate for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44279", - "right": "issue:45405", - "accept": false, - "reason": "Different problems: a dependency/import issue versus a PEFT minimum-version bump. No shared code-path or fix target." - }, - { - "left": "issue:38175", - "right": "issue:44617", + "left": "issue:45103", + "right": "issue:45372", "accept": false, - "reason": "Unrelated symptoms and subsystems: SigLIP2 probability output vs Sam3Video CUDA OOM." + "reason": "Different failures: an auto-docstring annotation crash vs a Gemma 4 processor import/loading regression." }, { - "left": "issue:43650", - "right": "issue:46209", + "left": "issue:41950", + "right": "issue:45020", "accept": false, - "reason": "The titles are unrelated and point to different failures: vague data request vs FP8 inference on B200." + "reason": "One is a video-classification processor lookup bug; the other is a broad remote_code compatibility regression." }, { - "left": "issue:41950", - "right": "issue:45020", + "left": "issue:44351", + "right": "issue:45278", "accept": false, - "reason": "One is a video-classification processor lookup bug; the other is a broad remote_code regression. Different concrete code paths." + "reason": "Both mention imports, but one is missing HybridCache specifically while the other is a wide set of import errors after upgrade." }, { - "left": "issue:43299", - "right": "issue:45357", + "left": "issue:38175", + "right": "issue:44617", "accept": false, - "reason": "Both are Qwen-related, but one is a model loading regression and the other is incorrect save_pretrained visual-encoder keys. Not the same bug." + "reason": "Unrelated symptoms and code paths: zero probabilities in SigLIP2 vs CUDA OOM in Sam3Video." }, { - "left": "issue:44351", - "right": "issue:45278", + "left": "issue:43650", + "right": "issue:46209", "accept": false, - "reason": "Shared theme of imports after upgrade, but one is a missing HybridCache export and the other is a broad import-error report. Too generic to merge." + "reason": "Different problems entirely: vague data handling vs FP8 inference failure on B200." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Different model behaviors: load-balancing loss gating vs Gemma4 multimodal token-type defaults." + "reason": "Both involve training losses/config, but one is router load-balancing loss and the other is mm_token_type_ids defaults." }, { - "left": "issue:44521", - "right": "issue:45325", + "left": "issue:43299", + "right": "issue:45357", "accept": false, - "reason": "Both are multimodal issues, but one is assistant mask generation and the other is temporal position-id scaling in Qwen2.5-VL. Distinct code paths." + "reason": "Same model family, but one is a load regression and the other is a save_pretrained key-serialization bug." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Different regressions: processor.tokenizer/image handling vs Deepseek tokenizer tokenization errors." + "reason": "Both are tokenizer/processor regressions, but one is processor.tokenizer / image-input API breakage and the other is wrong Deepseek tokenization." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Both touch docs/examples, but they concern different examples and different API mismatches." + "reason": "Both are documentation-related, but they refer to different docs and different bugs." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal/template issues, but assistant mask generation and temporal RoPE scaling are different code paths." }, { "left": "issue:41950", "right": "issue:45823", "accept": false, - "reason": "Different subsystems: processor lookup in a pipeline vs Gemma4 PLE device placement during forward." + "reason": "Different subsystems: video pipeline processor lookup vs Gemma4 PLE device placement." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "One is compiled flex_attention on new torch, the other is a TensorFlow backend/package-install issue. No overlap." + "reason": "No overlap beyond general framework/runtime failure: flex_attention compilation vs tensorboard/TensorFlow backend handling." }, { - "left": "issue:44779", - "right": "issue:45812", + "left": "issue:43976", + "right": "issue:45405", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families with different failure modes and likely different fixes." + "reason": "Separate dependency/version issues affecting unrelated packages (Python support vs PEFT minimum version)." }, { - "left": "issue:43976", - "right": "issue:45405", + "left": "issue:44779", + "right": "issue:45812", "accept": false, - "reason": "Unrelated: Python compatibility regression vs PEFT version bump." + "reason": "Both are tokenizer regressions, but they affect different model families and do not clearly point to the same underlying bug." }, { "left": "issue:46082", "right": "issue:46512", "accept": false, - "reason": "Different config semantics: head_dim validation vs stale config.dtype after from_config(dtype=...)." + "reason": "Different configuration/state issues: rejecting explicit head_dim vs stale config.dtype after from_config(dtype=...)." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Both mention model loading, but the affected models and failure mechanisms differ completely." + "reason": "Different models and bugs: Qwen3VL MOE load regression vs GLM-4.7-Flash unit-test loading issue." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Different config classes and fields: default vision/audio configs vs missing pad_token_id on Llama4Config." + "reason": "Both are config-related, but one is default subconfig initialization and the other is a missing pad_token_id attribute." }, { "left": "issue:44297", "right": "issue:44462", "accept": false, - "reason": "Related to tokenizer serialization/loading, but one is tokenizer_class written to config and the other is AutoTokenizer ignoring tokenizer.json. Not the same concrete bug." + "reason": "Both are tokenizer-related, but one is save_pretrained metadata mismatch and the other is AutoTokenizer ignoring tokenizer.json." }, { "left": "issue:44466", "right": "issue:45276", "accept": false, - "reason": "Different model-state issues: device-dependent lm_head serialization vs resize_token_embeddings not updating related embeddings." + "reason": "Different serialization/resize bugs: tied lm_head weight saving vs resize_token_embeddings not propagating to Gemma4 embeddings." }, { "left": "issue:44016", "right": "issue:44291", "accept": false, - "reason": "Notebook syntax error and init_empty_weights TypeError are unrelated." + "reason": "A notebook syntax error is unrelated to the init_empty_weights TypeError regression." }, { - "left": "issue:43441", - "right": "issue:45834", + "left": "issue:45081", + "right": "issue:45907", "accept": false, - "reason": "Different model-specific failures: FlashAttention issue for Ministral-3 vs OCR index error in Kosmos2.5." + "reason": "Different targets and failure modes: Mistral tokenizer regex patch crash vs list_repo_files not forwarding revision." }, { - "left": "issue:45081", - "right": "issue:45907", + "left": "issue:43441", + "right": "issue:45834", "accept": false, - "reason": "One is a tokenizer regex patch crash; the other is list_repo_files not forwarding revision. Different APIs." + "reason": "Unrelated model-specific inference bugs: FlashAttention failure vs long OCR index error." }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "Tokenizer special-token registration and D-FINE auxiliary losses are unrelated." + "reason": "Different training/evaluation bugs: BigBirdTokenizer special-token registration vs D-FINE auxiliary-loss handling." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Both concern quantization, but one is saving a quantized model and the other is loading/dequantizing with device_map offload. Different failure points." - }, - { - "left": "issue:45600", - "right": "issue:45712", - "accept": false, - "reason": "Both are CI/cleanup-related, but one is stale auto_mappings references and the other is leftover dummy classes. Different fixes." + "reason": "Both involve quantization/offload, but one is saving a quantized model and the other is loading a dequantized model with device_map=auto." }, { "left": "issue:33290", "right": "issue:45656", "accept": false, - "reason": "Both mention deepspeed/optimizer behavior, but one is OOM with Adafactor and the other is a double optimizer step. Not the same bug." + "reason": "Both mention DeepSpeed and optimizers, but one is Adafactor OOM and the other is an optimizer.step() being invoked twice." } ] }, @@ -11756,8 +11758,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -11767,7 +11769,6 @@ "issue:43299", "issue:43441", "issue:43525", - "issue:43531", "issue:43643", "issue:43673", "issue:43723", @@ -11804,12 +11805,15 @@ "issue:45561", "issue:45563", "issue:45593", + "issue:45600", + "issue:45712", "issue:45800", "issue:45823", "issue:46291", "issue:46399" ], "soft_pairs": [ + "issue:45600|issue:45712", "issue:44623|issue:44749", "issue:44261|issue:44263", "issue:44623|issue:45397", @@ -11829,236 +11833,232 @@ "issue:43525|issue:43643", "issue:44297|issue:45397", "issue:44589|issue:45362", + "issue:44188|issue:44336", "issue:43784|issue:45542", "issue:41628|issue:43976", - "issue:44188|issue:44336", "issue:44336|issue:44373", "issue:45561|issue:45563", "issue:45103|issue:46399", "issue:44246|issue:44496", - "issue:44242|issue:45593", "issue:44295|issue:45397", - "issue:42491|issue:43299", - "issue:43531|issue:45362" + "issue:44242|issue:45593", + "issue:42491|issue:43299" ], - "prepared_review_unit_hash": "978e9fa2f0313153f32289f2ca7d1da367994f633ce7180b954f49519cb08f40", + "prepared_review_unit_hash": "de0ba74d655b4538c9c49494c893e6be0b2b933cc9e076e764e535a8e7a3a8d0", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 35941, - "estimated_input_tokens": 8986, - "estimated_eval_tokens": 18228 + "serialized_chars": 36597, + "estimated_input_tokens": 9150, + "estimated_eval_tokens": 18556 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous Transformers issue reports, not one tight duplicate set. The only soft edge that looks like the same underlying bug is the Qwen3 MoE loading regression pair; the rest are different code paths, models, or unrelated docs/tooling problems.", - "confidence": 0.9, - "canonical_issue_reason": "issue:43299 is the broader, cleaner statement of the Qwen3 MoE loading regression in v5; issue:42491 reads like a downstream symptom of the same breakage on a specific LoRA setup.", + "summary": "This cluster is a loose soft-similarity bucket of mostly unrelated Transformers issues: packaging/backends, model loading, tokenizer regressions, docs, and kernel/runtime bugs. None of the soft pairs look like true duplicates, so all edges should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43299 is the clearest concrete regression report and the closest to the only near-duplicate-looking subgroup (Qwen3 MoE loading), though the cluster overall is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43299 is the best representative issue because it describes the core compatibility break more generally and is likely to attract the right fix and discussion.", + "best_issue_reason": "issue:43299 is the most representative standalone issue here: specific version, specific model family, and an actionable load failure.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Different file areas and failure modes: stale auto_mappings references vs leftover dummy classes leaking into dir(transformers)." + }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "Different bugs: missing files on processor.save_pretrained vs a slowdown/regression when filtering data after upgrade." + "reason": "processor.save_pretrained missing files is unrelated to the Chinese report about post-upgrade filtering slowdown." }, { "left": "issue:44261", "right": "issue:44263", "accept": false, - "reason": "Both are GLM-related, but they affect different internals and different failure modes; not the same underlying bug." + "reason": "Different models and code paths: MLA q_a_layernorm precision issue vs GlmMoeDsaIndexer torch.split behavior." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "One is serialization output from processor.save_pretrained, the other is a from_pretrained loading issue for Gemma4." + "reason": "Save-pretrained missing files and gemma-4 zero3 from_pretrained are distinct load/save bugs on different models." }, { "left": "issue:44361", "right": "issue:45593", "accept": false, - "reason": "Unrelated model/task failures: MLukeTokenizer AttributeError vs D-FINE auxiliary-loss behavior." + "reason": "Tokenizer AttributeError on tasks is unrelated to D-FINE missing auxiliary losses when denoising is off." }, { "left": "issue:45542", "right": "issue:45800", "accept": false, - "reason": "Different issues: a missing TensorFlow backend dependency vs a version incompatibility between torch and transformers." + "reason": "Packaging/backend installation mismatch is not the same as a torch/transformers version incompatibility." }, { "left": "issue:44016", "right": "issue:45823", "accept": false, - "reason": "Notebook syntax error and Gemma4 device-mismatch forward bug are unrelated." + "reason": "A notebook syntax error is a different problem from a Gemma4 forward device-map mismatch." }, { "left": "issue:42175", "right": "issue:43976", "accept": false, - "reason": "Dependency/back-end installation issue is not the same as a Python version support incompatibility." + "reason": "Missing TensorFlow in the torch extra and Python-version support regression are different environment issues." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Different model families and different symptoms; both are output correctness bugs, but not the same code-path problem." + "reason": "CamemBERT masked-LM predictions and Deepseek tokenizer outputs are different components and bugs." }, { "left": "issue:43441", "right": "issue:44683", "accept": false, - "reason": "Both involve attention backends, but one is Ministral-3 FlashAttention in v5 RC and the other is compiled flex_attention on torch>=2.9; too different to merge." + "reason": "Both involve attention kernels, but the triggers and code paths differ: FlashAttention on Ministral-3 vs compiled flex_attention on torch>=2.9." }, { "left": "issue:43901", "right": "issue:43994", "accept": false, - "reason": "Docs mismatch for return_all_scores is unrelated to SigLIP2 incorrect outputs." + "reason": "Docs drift about return_all_scores is not the same as incorrect SigLIP2 model/pipeline outputs." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer loading regression and ESM2 model breakage are separate issues." + "reason": "AutoTokenizer loading failure and ESM2 breakage are different models and failure modes." }, { "left": "issue:46291", "right": "issue:46399", "accept": false, - "reason": "Both are kernels-related, but one fails importing pipelines due to kernels metadata requirements while the other crashes on a missing decorator in Qwen3.5 use_kernels flow." + "reason": "Kernel package metadata import failure is unrelated to the apply_rotary_pos_emb decorator registration crash." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Generation cache/chunked_prefill regression is not the same as CLIPVisionModel hidden_states handling." + "reason": "Generation cache missing during chunked_prefill is a different bug from CLIPVisionModel returning hidden_states=None." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Llama4Config pad_token_id missing and glm-4v-9b loading failure are different model-loading bugs." + "reason": "Missing pad_token_id on Llama4Config and glm-4v-9b loading failure are not the same underlying issue." }, { "left": "issue:43824", "right": "issue:43976", "accept": false, - "reason": "Missing Qwen2_5_VL export and Python support mismatch are unrelated." + "reason": "Missing Qwen2.5-VL symbol import and Python version incompatibility are unrelated." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "A documentation issue and an import performance issue do not share the same underlying bug." + "reason": "A documentation mismatch and intermittent import slowness are not duplicates." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Config attribute missing vs missing fields returned by AutoConfig with trust_remote_code=True are distinct problems." + "reason": "Both mention missing config fields, but one is a specific Llama4Config attribute and the other is a trust_remote_code serialization gap." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer save metadata mismatch is unrelated to Gemma4 zero3 from_pretrained loading." + "reason": "tokenizer_class mismatch in save_pretrained is unrelated to gemma-4 zero3 from_pretrained failures." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage lookup error and Qwen3.5 chat crash are different failures." + "reason": "Float8 storage lookup errors and Qwen3.5 chat crashes are different bugs with no shared code path." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence on torch.compile and ANSI code leakage from loading_report are unrelated." }, { "left": "issue:43784", "right": "issue:45542", "accept": false, - "reason": "sentence-transformers import NameError and TensorFlow backend dependency installation are unrelated." + "reason": "sentence-transformers import NameError and undefined tf backend from missing TensorFlow are different import-time failures." }, { "left": "issue:41628", "right": "issue:43976", "accept": false, - "reason": "Cannot import AutoImageProcessor and Python version support incompatibility are separate issues." - }, - { - "left": "issue:44188", - "right": "issue:44336", - "accept": false, - "reason": "Attention-kernel divergence under torch.compile is unrelated to ANSI codes emitted by loading_report." + "reason": "AutoImageProcessor import error is unrelated to the Python version support issue." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "Loading-report terminal coloring and wrong position_ids docstring are unrelated." + "reason": "ANSI output leakage and a wrong docstring for position_ids are different severity and different subsystems." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "pytest-xdist race in testing utils and a stale generate() warning are different bugs." + "reason": "A pytest-xdist race in testing utils is unrelated to a stale warning in paged generate()." }, { "left": "issue:45103", "right": "issue:46399", "accept": false, - "reason": "Auto-docstring crash with future annotations and Qwen3.5 kernel registration failure are unrelated." + "reason": "auto_docstring crashes under future annotations and kernel decorator registration errors are different code paths." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Slow import and unrecognized model/config loading failure are different issues." + "reason": "Intermittent import slowness is not the same as an unrecognized model/config.json error." }, { - "left": "issue:44242", - "right": "issue:45593", + "left": "issue:44295", + "right": "issue:45397", "accept": false, - "reason": "Load-balancing loss bug and D-FINE auxiliary-loss bug do not look like the same underlying change." + "reason": "position_ids buffer-reading error and gemma-4 zero3 from_pretrained are distinct bugs." }, { - "left": "issue:44295", - "right": "issue:45397", + "left": "issue:44242", + "right": "issue:45593", "accept": false, - "reason": "position_ids buffer access error and Gemma4 zero3 loading failure are unrelated." + "reason": "Both concern missing losses, but one is load-balancing loss with router logits and the other is D-FINE auxiliary losses with denoising; different models and logic." }, { "left": "issue:42491", "right": "issue:43299", - "accept": true, - "reason": "Both describe the same Qwen3 MoE loading regression after the hf4.x -> hf5.x/v5 transition; 42491 is a downstream LoRA-specific manifestation of the broader breakage in 43299." - }, - { - "left": "issue:43531", - "right": "issue:45362", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and Qwen3.5 chat crash are not the same concrete bug." + "reason": "Both are Qwen3 MoE-related, but one is LoRA compatibility across hf4.x/hf5.x and the other is base Qwen3VL MoE loading on v5; not the same concrete bug." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The non-duplicate verdicts are appropriately cautious, and the single accepted soft edge (42491 vs 43299) is reasonably supported by the titles as a shared Qwen3 MoE loading regression across versions." - }, - "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 }, @@ -12070,8 +12070,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -12084,6 +12084,7 @@ "issue:43352", "issue:43475", "issue:43526", + "issue:43531", "issue:43550", "issue:43596", "issue:43673", @@ -12110,14 +12111,13 @@ "issue:45200", "issue:45250", "issue:45292", + "issue:45362", "issue:45375", - "issue:45412", "issue:45561", "issue:45584", "issue:45701", "issue:45759", "issue:45859", - "issue:46123", "issue:46132", "issue:46369", "issue:46399", @@ -12125,11 +12125,12 @@ "issue:46506" ], "soft_pairs": [ + "issue:43531|issue:45362", "issue:43526|issue:45003", "issue:41950|issue:44220", "issue:44246|issue:44871", - "issue:42886|issue:46132", "issue:44336|issue:44485", + "issue:42886|issue:46132", "issue:43550|issue:46506", "issue:36246|issue:42757", "issue:36246|issue:44351", @@ -12137,240 +12138,239 @@ "issue:44336|issue:46424", "issue:45003|issue:45561", "issue:44779|issue:45292", - "issue:43475|issue:46369", "issue:45859|issue:46399", "issue:36246|issue:43596", - "issue:44242|issue:44857", + "issue:43475|issue:46369", "issue:44295|issue:44749", + "issue:44242|issue:44857", "issue:43673|issue:44561", "issue:43901|issue:45003", - "issue:44162|issue:45200", "issue:43352|issue:45250", + "issue:44162|issue:45200", "issue:43257|issue:45137", "issue:44496|issue:45003", "issue:45292|issue:45584", - "issue:45292|issue:45701", "issue:43262|issue:44246", + "issue:45292|issue:45701", "issue:45375|issue:45759", "issue:44877|issue:45375", "issue:44038|issue:46506", - "issue:45412|issue:46123", "issue:44561|issue:46399" ], - "prepared_review_unit_hash": "c53b095b30f4241ee2e6f946a6276353fec614620070e1067ff056b1cb488aa3", + "prepared_review_unit_hash": "cc0b5c61d2798b2bde2ddb10037952f5b2be177d4b03d13afd0143f9db4c81b4", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36887, - "estimated_input_tokens": 9222, - "estimated_eval_tokens": 18700 + "serialized_chars": 36813, + "estimated_input_tokens": 9204, + "estimated_eval_tokens": 18664 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is heterogeneous: most items are distinct regressions across different subsystems (imports, tokenization, config loading, generation, DeepSpeed, vision/audio pipelines). None of the soft-similarity pairs look like the same underlying bug or change closely enough to merge.", - "confidence": 0.9, - "canonical_issue_reason": "Issue 45701 is the broadest user-facing regression in the set and best represents the cluster's repeated theme of version-dependent behavior changes, but the overall set is still quite mixed.", + "summary": "This cluster is highly heterogeneous: the soft links are mostly title-level similarity across unrelated bugs (Qwen, DeepSpeed, cache handling, docs, audio, vision, import errors). None of the proposed soft edges look like true duplicates, so all should be rejected. If forced to name a representative issue, 45003 is the broadest hub-like report, but it is not a real canonical duplicate anchor.", + "confidence": 0.92, + "canonical_issue_reason": "issue:45003 is the broadest, most central infrastructure-style report in the set and has the most superficial overlap with other items, but the cluster is too diverse for a true duplicate canonical.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45701 is the best single representative because it is broad, current, and aligned with several other v5-era behavior-regression reports; still, it is only a loose fit for the heterogeneous cluster.", + "best_issue_reason": "issue:45003 is the best overall representative only in the sense of being the broadest hub, not because it duplicates the others; it should not be treated as a merged target for this set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Both mention Qwen, but one is a sliding_window bug in Qwen3-MoE and the other is a chat crash in Qwen3.5-35B; different failures and code paths." + }, { "left": "issue:43526", "right": "issue:45003", "accept": false, - "reason": "One is a BEiT image-processor label reduction bug; the other is a sys.modules access issue in modeling utils. Different components and failure modes." + "reason": "BeitImageProcessorFast label reduction vs generic sys.modules access; unrelated components and symptoms." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification pipeline processor lookup and audio feature extraction are unrelated subsystems with no shared code path." + "reason": "Video-classification processor lookup vs audio fbank extraction; different pipeline areas." }, { "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Import performance regression is unrelated to Gemma-3 eos_token_id config inconsistency." + "reason": "Import slowness report vs Gemma eos_token_id config mismatch; unrelated bugs." }, { - "left": "issue:42886", - "right": "issue:46132", + "left": "issue:44336", + "right": "issue:44485", "accept": false, - "reason": "Offline cache loading and AttentionInterface registration behavior are different bugs in different layers." + "reason": "ANSI output in loading_report vs GLM-5 RoPE implementation; not the same issue." }, { - "left": "issue:44336", - "right": "issue:44485", + "left": "issue:42886", + "right": "issue:46132", "accept": false, - "reason": "ANSI escape output in loading_report is unrelated to GLM-5 RoPE implementation." + "reason": "Offline cache-loading failure vs AttentionInterface.register behavior change; different APIs and code paths." }, { "left": "issue:43550", "right": "issue:46506", "accept": false, - "reason": "torch.compile+SDPA failure in Bamba and CPU/CUDA result mismatch in PPDocLayoutV3 are different model/runtime issues." + "reason": "Bamba torch.compile/SDPA failure vs PPDocLayoutV3 CPU/CUDA divergence; different models and backend problems." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but for different missing symbols and different root causes." + "reason": "Both are import errors, but for different missing symbols in different packages; not the same underlying bug." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Missing Qwen2_5_VLImageProcessor and missing HybridCache are separate export problems." + "reason": "Missing Qwen2.5-VL image processor vs missing HybridCache export; different symbols and fixes." }, { "left": "issue:44734", "right": "issue:46424", "accept": false, - "reason": "KV cache continuation crash in serve is not the same as max_cache_length handling in cache config." + "reason": "KV-cache continuation tensor indexing crash vs max_cache_length behavior for Quantized Cache; different cache logic." }, { "left": "issue:44336", "right": "issue:46424", "accept": false, - "reason": "Terminal ANSI formatting and cache-length configuration are unrelated." + "reason": "Terminal-color formatting bug vs cache-length semantics; unrelated." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "sys.modules access bug in modeling_utils and xdist race on captured_info.txt are unrelated." + "reason": "sys.modules access bug vs pytest-xdist captured_info.txt race; different subsystems." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "Tokenizer output regression and resize_token_embeddings/output_embeddings mismatch are different behaviors in different code paths." - }, - { - "left": "issue:43475", - "right": "issue:46369", - "accept": false, - "reason": "SAM 3 missing attribute and SAM3-litetext memory overflow are not the same bug." + "reason": "DeepSeek tokenizer correctness regression vs resize_token_embeddings not updating output embeddings; distinct behaviors." }, { "left": "issue:45859", "right": "issue:46399", "accept": false, - "reason": "CPU offload incompatibility for a rotary embedding and a missing decorator registration crash are separate Qwen3.5 issues." + "reason": "CPU offload incompatibility vs missing decorator for kernel registration; both Qwen3.5-related but not the same defect." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "Missing Qwen import and DeepSpeed/BertModel index error have no shared root cause." + "reason": "Qwen2.5-VL import failure vs DeepSpeed ZeRO3 BertModel index error; unrelated." }, { - "left": "issue:44242", - "right": "issue:44857", + "left": "issue:43475", + "right": "issue:46369", "accept": false, - "reason": "Router logits load-balancing loss logic and LwDetrImageLoss AMP crash are unrelated." + "reason": "SAM3 video encoder attribute error vs SAM3-litetext memory overflow; different failure modes." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "Position_ids buffer access error and Chinese report about slower filtering after upgrade are not the same issue." + "reason": "Position_ids buffer access error vs a transformers performance regression after upgrade; unrelated." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "MoE load-balancing loss omission vs float16 AMP crash in LwDetrImageLoss; different bugs." }, { "left": "issue:43673", "right": "issue:44561", "accept": false, - "reason": "Generation cache missing in chunked prefill and removal of is_torch_fx_available affecting trust_remote_code are different regressions." + "reason": "Chunked-prefill cache regression vs v5 removal of is_torch_fx_available breaking trust_remote_code; different breakages." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "Documentation mismatch for return_all_scores and sys.modules unsafe access are unrelated." + "reason": "Docs mismatch for return_all_scores vs sys.modules access bug; unrelated." }, { - "left": "issue:44162", - "right": "issue:45200", + "left": "issue:43352", + "right": "issue:45250", "accept": false, - "reason": "ESM2 breakage and Gemma 4 token-type defaults are unrelated model-specific problems." + "reason": "Specific Nemotron Flash Attention 2.0 unsupported error vs a generic Flash Attention 2.0 topic; not the same concrete issue." }, { - "left": "issue:43352", - "right": "issue:45250", + "left": "issue:44162", + "right": "issue:45200", "accept": false, - "reason": "A model-specific Flash Attention 2 unsupported error is not the same as a generic Flash Attention 2 issue." + "reason": "ESM2 breakage vs Gemma 4 mm_token_type_ids defaulting; unrelated models and code paths." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MOE weight conversion with accelerate/deepspeed and ZeRO3 empty-deque failure are different DeepSpeed loading bugs." + "reason": "Qwen3 MOE weight conversion issue vs ZeRO3 deque underflow; different DeepSpeed failures." }, { "left": "issue:44496", "right": "issue:45003", "accept": false, - "reason": "Unrecognized model config loading and unsafe sys.modules access are distinct loading failures." + "reason": "Unrecognized model config vs unsafe sys.modules access; unrelated." }, { "left": "issue:45292", "right": "issue:45584", "accept": false, - "reason": "Embedding resize mismatch and Whisper empty-transcription failure are unrelated." + "reason": "Embedding resize/output-head sync bug vs Whisper empty-transcription generation failure; different subsystems." }, { - "left": "issue:45292", - "right": "issue:45701", + "left": "issue:43262", + "right": "issue:44246", "accept": false, - "reason": "Resize-token-embeddings/output-embeddings mismatch is not the same as a broad tokenization-version-change report." + "reason": "Audio apply_chat_template sampling-rate bug vs import slowness; unrelated." }, { - "left": "issue:43262", - "right": "issue:44246", + "left": "issue:45292", + "right": "issue:45701", "accept": false, - "reason": "Audio chat-template sampling-rate default and import performance are unrelated." + "reason": "resize_token_embeddings/output embedding mismatch vs tokenizer behavior changes across versions; distinct issues." }, { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Missing strict config field for Qwen3_5MoeVisionConfig and AutoModelForCausalLM text_config unwrapping are different composite-config bugs." + "reason": "Missing Qwen3.5MoeVisionConfig field vs composite config not unwrapping text_config; similar family, but different config defects." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Granite speech strict-config loading and Qwen vision config field dropping are separate strictness/config-schema issues." + "reason": "Strict granite_speech config loading vs Qwen vision config field loss; different models and symptoms." }, { "left": "issue:44038", "right": "issue:46506", "accept": false, - "reason": "Qwen3-VL-Moe v5 bug and PPDocLayoutV3 CPU/CUDA divergence are different model-specific failures." - }, - { - "left": "issue:45412", - "right": "issue:46123", - "accept": false, - "reason": "RT-DETR memory leak and MaskGenerationPipeline final-batch dropping are unrelated." + "reason": "Qwen3-VL-Moe v5 bug vs PPDocLayoutV3 CPU/CUDA divergence; unrelated." }, { "left": "issue:44561", "right": "issue:46399", "accept": false, - "reason": "Removal of is_torch_fx_available breaking remote-code models and Qwen3.5 rotary registration crash are different breakages." + "reason": "Removal of is_torch_fx_available breaking trust_remote_code vs Qwen3.5 apply_rotary_pos_emb registration problem; different regressions." } ] }, @@ -12389,8 +12389,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -12427,6 +12427,7 @@ "issue:45229", "issue:45335", "issue:45397", + "issue:45412", "issue:45459", "issue:45563", "issue:45584", @@ -12443,26 +12444,27 @@ "issue:46431" ], "soft_pairs": [ - "issue:45685|issue:46431", - "issue:44987|issue:45459", + "issue:45412|issue:46123", "issue:44466|issue:45335", + "issue:44987|issue:45459", "issue:45137|issue:45600", - "issue:44995|issue:45200", + "issue:45685|issue:46431", "issue:43475|issue:43540", + "issue:44995|issue:45200", "issue:44060|issue:44077", + "issue:45397|issue:45656", "issue:44479|issue:46170", "issue:43653|issue:44162", - "issue:45397|issue:45656", - "issue:42886|issue:45229", "issue:44368|issue:44977", - "issue:43906|issue:46202", + "issue:42886|issue:45229", "issue:44483|issue:46202", + "issue:43906|issue:46202", "issue:44995|issue:45588", - "issue:44368|issue:45563", "issue:43065|issue:43408", + "issue:44368|issue:45563", "issue:44188|issue:46202", - "issue:43931|issue:44368", "issue:44877|issue:46170", + "issue:43931|issue:44368", "issue:45987|issue:46082", "issue:43065|issue:46123", "issue:44188|issue:46421", @@ -12472,223 +12474,223 @@ "issue:43646|issue:44162", "issue:44749|issue:45656", "issue:44623|issue:45656", - "issue:43232|issue:43673", - "issue:43701|issue:44749" + "issue:43701|issue:44749", + "issue:43232|issue:43673" ], - "prepared_review_unit_hash": "550912fd135284ed9254da971c89f5a4148504038e6f9d1a2062f5e40ad9212b", + "prepared_review_unit_hash": "51928517ff6df04990dc7001f8acd62f7a0ed8a50fb0b7b1bdac3d7f41e4c05b", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36403, - "estimated_input_tokens": 9101, - "estimated_eval_tokens": 18458 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37276, + "estimated_input_tokens": 9319, + "estimated_eval_tokens": 18894 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues do not form a true duplicate cluster; the soft pairs are mostly broad semantic overlaps across unrelated bugs. I would anchor the set on the broadest regression report, with no PR canonical candidate present.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44479 is the broadest, most actionable regression report in this set (multi-model v5.3 video-input failure) and is the best anchor if one issue must represent the cluster.", + "summary": "These items are mostly unrelated standalone issues that happen to share broad themes like generation, caching, tokenizers, or Gemma/Qwen/SAM subsystems. The soft pairs are not the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44466 is the clearest standalone regression report here: it has a precise version boundary, a focused serialization/tied-weights symptom, and good reproducibility, so it is the best anchor among otherwise unrelated issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 has the clearest user impact, strongest reproduction value, and broadest scope among the listed issues.", + "best_issue_reason": "Issue 44466 is the strongest single triage candidate because it is specific, regression-oriented, and well-scoped; the rest of the set does not form a coherent duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45685", - "right": "issue:46431", + "left": "issue:45412", + "right": "issue:46123", "accept": false, - "reason": "Different bugs: MPS histogram/int support in MoE vs missing chat template in a processor." + "reason": "Different bugs: RT-DETR memory retention vs MaskGenerationPipeline final partial-batch handling." }, { - "left": "issue:44987", - "right": "issue:45459", + "left": "issue:44466", + "right": "issue:45335", "accept": false, - "reason": "Both touch tokenization/loading, but one is a model load regression and the other is error masking when protobuf is absent." + "reason": "Both involve weights/embedding behavior, but one is serialization inconsistency and the other is resize_token_embeddings not propagating to decoder.embed_tokens." }, { - "left": "issue:44466", - "right": "issue:45335", + "left": "issue:44987", + "right": "issue:45459", "accept": false, - "reason": "Related to tied weights, but one is serialization inconsistency and the other is resize_token_embeddings not propagating to decoder embeddings." + "reason": "Different tokenizer-loading failures: one breaks loading a specific model, the other hides real tokenizer errors when protobuf is missing." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "Unrelated subsystems: DeepSpeed ZeRO3 deque error vs stale auto_mappings references to removed configs." + "reason": "Unrelated: DeepSpeed ZeRO3 deque underflow vs stale Sam3LiteText references in auto_mappings.py." }, { - "left": "issue:44995", - "right": "issue:45200", + "left": "issue:45685", + "right": "issue:46431", "accept": false, - "reason": "Different model behaviors: stale cache on second forward pass vs missing mm_token_type_ids default in Gemma 4 fine-tuning." + "reason": "Different areas and failures: MPS histogram on Int vs chat-template absence in a processor." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "Both are SAM/Qwen video issues, but the concrete failures and code paths are different." + "reason": "Both are video-related, but one is a missing attribute in SAM3 output and the other is a Qwen3OmniMoe input-processing ValueError." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "Different root causes: stale indexer cache on a second forward pass vs Gemma 4 mm_token_type_ids defaulting." }, { "left": "issue:44060", "right": "issue:44077", "accept": false, - "reason": "Different model families and failures: tied-weights warning in Qwen3-Next vs patchtsmixer post_init policy." + "reason": "Only broad theme overlap; one is a tied-weights warning bug, the other is a deprecation/allowance issue for optional post_init." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Different bugs and code paths: gemma-4 zero3 from_pretrained vs optimizer step being called twice under deepspeed." }, { "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Video-input regression vs Windows/ROCm distributed import failure; no shared code path." + "reason": "Different failure modes: video input regression in v5.3 vs Windows/ROCm import or torch.distributed dependency issues." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "Tokenizer special-token registration bug vs a broader ESM2 breakage; not the same defect." + "reason": "Different tokenizer/model problems: BigBirdTokenizer special-token registration vs ESM2 being broken broadly." }, { - "left": "issue:45397", - "right": "issue:45656", + "left": "issue:44368", + "right": "issue:44977", "accept": false, - "reason": "Different failures: Gemma-4 ZeRO3 loading vs optimizer step being called twice under DeepSpeed." + "reason": "One is a tie_word_embeddings warning during LoRA fine-tuning, the other is a flash-attention generation regression." }, { "left": "issue:42886", "right": "issue:45229", "accept": false, - "reason": "Offline cache tokenization failure vs Gemma4 multi-GPU CUDA OOM; unrelated." - }, - { - "left": "issue:44368", - "right": "issue:44977", - "accept": false, - "reason": "One is a tie_word_embeddings warning during LoRA fine-tuning, the other is a flash-attention generation regression." + "reason": "Unrelated: offline cache tokenizer load failure vs Gemma4 multi-GPU OOM." }, { - "left": "issue:43906", + "left": "issue:44483", "right": "issue:46202", "accept": false, - "reason": "The reproduction issue is too vague to equate; Gemma roundtrip failure on '\u2581' is a specific tokenizer/pipeline bug." + "reason": "Different APIs and symptoms: chat/completions request validation vs Gemma pipeline roundtrip on the \u2581 token." }, { - "left": "issue:44483", + "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "Chat/completions API acceptance failure is unrelated to Gemma tokenizer roundtrip behavior." + "reason": "The latter is a Gemma pipeline roundtrip bug; the former is only an isolated reproduction of a different issue." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "Different code paths: stale indexer cache in GlmMoeDsa vs flash-attention aux-state handling." + "reason": "Different bugs: stale cache on repeated forward pass vs flash_attention.py crashing on s_aux=None." }, { - "left": "issue:44368", - "right": "issue:45563", + "left": "issue:43065", + "right": "issue:43408", "accept": false, - "reason": "Warning about tied embeddings during fine-tuning is unrelated to paged generate's stale num_return_sequences warning." + "reason": "SAM3 PixelDecoder dummy Conv2d vs sam3_video/sam3_tracker config mismatch are not the same defect." }, { - "left": "issue:43065", - "right": "issue:43408", + "left": "issue:44368", + "right": "issue:45563", "accept": false, - "reason": "Same SAM3 family, but dummy Conv2d in a decoder and a model-type warning are not the same bug." + "reason": "One is a tie-weights warning during fine-tuning; the other is a stale paged-generate warning about num_return_sequences." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "Torch.compile attention-kernel divergence is unrelated to Gemma tokenizer roundtrip failure." + "reason": "Different concerns: torch.compile branching divergence in attention kernels vs Gemma pipeline token roundtrip failure." }, { - "left": "issue:43931", - "right": "issue:44368", + "left": "issue:44877", + "right": "issue:46170", "accept": false, - "reason": "Shape-mismatch model loading for Qwen3-VL is unrelated to Qwen3.5 tie-word-embeddings warnings." + "reason": "Strict config loading vs Windows ROCm import/type errors are unrelated." }, { - "left": "issue:44877", - "right": "issue:46170", + "left": "issue:43931", + "right": "issue:44368", "accept": false, - "reason": "Strict config loading for granite_speech vs Windows/ROCm distributed dependency issue; different defects." + "reason": "Model weight-shape mismatch on load vs tied-weights warning during LoRA fine-tuning are different problems." }, { "left": "issue:45987", "right": "issue:46082", "accept": false, - "reason": "Cache return-type contract bug vs LlamaConfig head_dim validation; different layers of the stack." + "reason": "StaticCache return-type contract bug vs LlamaConfig head_dim validation are unrelated." }, { "left": "issue:43065", "right": "issue:46123", "accept": false, - "reason": "Both are pipeline/model issues, but dummy Conv2d and is_last handling in MaskGenerationPipeline are unrelated." + "reason": "Different subsystems and failures: SAM3 decoder placeholder layer vs final partial-batch result dropping." }, { "left": "issue:44188", "right": "issue:46421", "accept": false, - "reason": "Both mention cache/compile behavior, but one is kernel divergence under torch.compile and the other is lazy cache init causing recompilation." + "reason": "Both mention cache/compile behavior, but one is attention-kernel divergence and the other is spurious recompilation from lazy cache init." }, { "left": "issue:45584", "right": "issue:46170", "accept": false, - "reason": "Whisper empty-transcription bug is unrelated to a Windows ROCm dependency/import failure." + "reason": "Whisper empty transcription after align_special_tokens is unrelated to Windows ROCm import failures." }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Both involve save_pretrained, but tokenizer metadata mismatch and missing processor files are different problems." + "reason": "Both are save_pretrained-related, but one is tokenizer_config metadata mismatch and the other is missing processor files." }, { "left": "issue:44361", "right": "issue:44749", "accept": false, - "reason": "Tokenizer AttributeError on tasks vs a performance regression in data filtering after upgrade; not the same issue." + "reason": "MLukeTokenizer AttributeError and a Chinese report of performance slowdown after upgrading are not the same bug." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "General v5 custom-model init breakage is not the same concrete bug as the ESM2 report." + "reason": "Generic Transformers 5.0 custom init breakage does not match the separate ESM2 breakage report." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "Performance slowdown in data filtering and double optimizer stepping are unrelated." + "reason": "Performance regression after upgrading vs duplicate optimizer-step behavior under deepspeed are unrelated." }, { - "left": "issue:44623", - "right": "issue:45656", + "left": "issue:43701", + "right": "issue:44749", "accept": false, - "reason": "processor.save_pretrained missing files is unrelated to DeepSpeed optimizer-step duplication." + "reason": "Checkpoint key mismatch during resume is unrelated to a filtering performance regression." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Both are generation/cache-related, but one is sync_gpus kwargs handling and the other is chunked_prefill cache missing in v5.0.0." - }, - { - "left": "issue:43701", - "right": "issue:44749", - "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to the filtering slowdown regression." + "reason": "Both concern generation/caching, but the concrete failures are different: sync_gpus update timing vs cache missing during chunked_prefill." } ] }, @@ -12707,8 +12709,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -12716,6 +12718,7 @@ "issue:30990", "issue:36010", "issue:41628", + "issue:42175", "issue:42371", "issue:42673", "issue:42831", @@ -12753,260 +12756,265 @@ "issue:45397", "issue:45406", "issue:45478", + "issue:45712", "issue:45753", - "issue:46077", - "issue:46123", "issue:46132", "issue:46202", "issue:46326", "issue:46399" ], "soft_pairs": [ - "issue:44869|issue:45356", "issue:44871|issue:46132", + "issue:44869|issue:45356", "issue:42994|issue:43482", "issue:45230|issue:45397", "issue:43475|issue:43643", "issue:44038|issue:44623", - "issue:28282|issue:30990", "issue:42371|issue:45092", + "issue:28282|issue:30990", "issue:43475|issue:45381", "issue:41628|issue:44933", "issue:45375|issue:45478", - "issue:45406|issue:46202", "issue:42994|issue:43525", - "issue:43784|issue:46399", "issue:45310|issue:45375", + "issue:45406|issue:46202", "issue:43723|issue:46326", + "issue:43784|issue:46399", "issue:42673|issue:43704", - "issue:43873|issue:44945", "issue:44964|issue:45137", + "issue:43873|issue:44945", "issue:43232|issue:44188", - "issue:45081|issue:45103", "issue:41628|issue:44351", + "issue:45081|issue:45103", "issue:44945|issue:45753", "issue:42994|issue:43824", "issue:44222|issue:45310", "issue:44038|issue:45310", - "issue:43232|issue:44792", "issue:36010|issue:41628", + "issue:43232|issue:44792", "issue:42831|issue:44222", - "issue:46077|issue:46123" + "issue:42175|issue:45712" ], - "prepared_review_unit_hash": "339425e2d834d7cd2e79ce40bfbc9982b34fa7886f6b5ca57478ab22f2a39de0", + "prepared_review_unit_hash": "886f0687ca4233881602d421cbc6e81795e49d716da5a1b9388a5a32cda94572", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36463, - "estimated_input_tokens": 9116, - "estimated_eval_tokens": 18488 + "serialized_chars": 36485, + "estimated_input_tokens": 9122, + "estimated_eval_tokens": 18500 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many items are unrelated model-loading/import/tokenizer regressions that only share broad Transformers similarity. The only clear duplicate pair in the provided soft edges is the repeated Qwen3ForCausalLM VRAM leak report.", - "confidence": 0.82, - "canonical_issue_reason": "No single canonical issue cleanly represents the whole set because the items span many distinct bugs; if an anchor is needed, the Qwen3.5 MoE from_pretrained regression is the strongest standalone candidate.", + "summary": "This cluster is mostly heterogeneous transformer-related bugs. I only found two clear duplicate pairs: the Qwen3ForCausalLM VRAM leak reports and the Qwen3.5 MoE from_pretrained error reports. Everything else is too broad or only superficially similar.", + "confidence": 0.92, + "canonical_issue_reason": "issue:42673 is the clearest canonical issue: it has the fuller report for the exact duplicate pair with issue:43704, older age, and more discussion, making it the best representative of the duplicate subcluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the best representative issue: it is specific, version-scoped, and has a clear concrete failure mode in a high-signal model-loading regression.", + "best_issue_reason": "issue:42673 is the strongest issue representative overall because it is an exact duplicate target with substantial discussion and a specific reproducible symptom; the other apparent duplicate pair (issue:45310 / issue:45478) is separate and later.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44869", - "right": "issue:45356", + "left": "issue:44871", + "right": "issue:46132", "accept": false, - "reason": "Both are tokenizer/encoding-adjacent, but they describe different models and different failure paths." + "reason": "Different bugs: Gemma eos-token configuration mismatch vs AttentionInterface behavior change." }, { - "left": "issue:44871", - "right": "issue:46132", + "left": "issue:44869", + "right": "issue:45356", "accept": false, - "reason": "Gemma EOS config inconsistency and AttentionInterface behavior are unrelated bugs." + "reason": "Both tokenizer/decoding-related, but they describe different failure modes and code paths." }, { "left": "issue:42994", "right": "issue:43482", "accept": false, - "reason": "Quantized model saving and Qwen2.5-GGUF loading are separate code paths." + "reason": "Different model issues: quantized save_pretrained failure vs Qwen2.5-GGUF loading failure." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "A generic bug report does not match the specific Gemma-4 zero3 loading issue." + "reason": "Too vague on the left and a specific ZeRO3/gemma-4 from_pretrained issue on the right; no clear same bug." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "SAM 3 video attribute error and trust_remote_code missing fields are not the same bug." + "reason": "One is a SAM 3 video attribute error, the other is missing fields from trust_remote_code config loading." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL-Moe loading failure and processor.save_pretrained missing files are different problems." + "reason": "Different subsystems: Qwen3-VL-Moe bug vs processor.save_pretrained missing files." }, { - "left": "issue:28282", - "right": "issue:30990", + "left": "issue:42371", + "right": "issue:45092", "accept": false, - "reason": "PyTorch-missing ImportError and Sentence Transformers loading stall are distinct issues." + "reason": "TF32 API guidance is unrelated to InternVL2 meta-initialization incompatibility." }, { - "left": "issue:42371", - "right": "issue:45092", + "left": "issue:28282", + "right": "issue:30990", "accept": false, - "reason": "TF32 API guidance and InternVL2 meta-init incompatibility are unrelated." + "reason": "Both mention loading, but one is a PyTorch import error and the other a sentence-transformers stall." }, { "left": "issue:43475", "right": "issue:45381", "accept": false, - "reason": "Different model families and different symptoms; not the same underlying bug." + "reason": "Different multimodal models and distinct video/position-id failures." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both mention imports, but they refer to different missing symbols and likely different API removals." + "reason": "Both are import-related, but they target different missing symbols and likely different export issues." }, { "left": "issue:45375", "right": "issue:45478", "accept": false, - "reason": "Same model family, but one is a missing config field and the other is a broader from_pretrained error without enough evidence they are the same bug." - }, - { - "left": "issue:45406", - "right": "issue:46202", - "accept": false, - "reason": "Gemma4 processor attribute failure and Gemma pipeline roundtrip issue are different failure modes." + "reason": "Same model family, but the reported missing field vs from_pretrained error are not clearly the same underlying bug." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized saving failure and missing pad_token_id in Llama4Config are unrelated." + "reason": "Unrelated: quantized model saving failure vs Llama4Config missing pad_token_id." }, { - "left": "issue:43784", - "right": "issue:46399", + "left": "issue:45310", + "right": "issue:45375", "accept": false, - "reason": "Sentence-transformers import NameError and Qwen3.5 kernel registration error are unrelated." + "reason": "Related Qwen3.5 MoE area, but one is a generic from_pretrained error and the other is a missing config field report; not enough evidence to merge." }, { - "left": "issue:45310", - "right": "issue:45375", + "left": "issue:45406", + "right": "issue:46202", "accept": false, - "reason": "Related model family, but the titles point to different specific loading/config bugs." + "reason": "Gemma4 processor attribute error and Gemma pipeline roundtrip failure are different code paths." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "Tokenizer loading regression and continue_final_message behavior are not the same code-path problem." + "reason": "Tokenizer loading issue in v5 vs continue_final_message behavior for mistral_common tokenizers are different bugs." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "Import-time NameError in sentence-transformers vs Qwen3.5 rotary registration crash; different causes." }, { "left": "issue:42673", "right": "issue:43704", "accept": true, - "reason": "Exact same title and same reported Qwen3ForCausalLM VRAM leak bug." + "reason": "Exact duplicate reports of the same Qwen3ForCausalLM VRAM leak with the same title and symptom." }, { - "left": "issue:43873", - "right": "issue:44945", + "left": "issue:44964", + "right": "issue:45137", "accept": false, - "reason": "Quantization offloading and pipeline-parallel output correctness are separate issues." + "reason": "Different failures: model loading vs DeepSpeed ZeRO3 deque underflow." }, { - "left": "issue:44964", - "right": "issue:45137", + "left": "issue:43873", + "right": "issue:44945", "accept": false, - "reason": "Model loading failure and DeepSpeed ZeRO3 deque error are unrelated." + "reason": "Quantization/offloading behavior vs incorrect output under pipeline parallelism are distinct problems." }, { "left": "issue:43232", "right": "issue:44188", "accept": false, - "reason": "Generation kwargs handling and torch.compile attention-kernel divergence are different bugs." + "reason": "Generation kwargs update issue vs attention-kernel divergence under torch.compile; unrelated code paths." }, { - "left": "issue:45081", - "right": "issue:45103", + "left": "issue:41628", + "right": "issue:44351", "accept": false, - "reason": "Mistral tokenizer regex crash and auto_docstring annotation crash are unrelated." + "reason": "Both are import errors, but they concern different missing top-level symbols." }, { - "left": "issue:41628", - "right": "issue:44351", + "left": "issue:45081", + "right": "issue:45103", "accept": false, - "reason": "Both are import errors, but for different missing classes and different fixes." + "reason": "Tokenizer regex crash and auto_docstring annotation crash are unrelated." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallel generation correctness and Qwen3_5 image looping are distinct problems." + "reason": "Pipeline-parallelism output bug vs Qwen3_5 image-specific infinite loop; not the same issue." }, { "left": "issue:42994", "right": "issue:43824", "accept": false, - "reason": "Quantized saving failure and missing Qwen2.5-VL import are unrelated." + "reason": "Quantized saving failure vs missing Qwen2_5_VLForConditionalGeneration import are different issues." }, { "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Both involve Qwen/FP8/loading themes, but the concrete bugs are different." + "reason": "FP8 save_pretrained on MoE vs Qwen3.5 MoE from_pretrained error; related area but not the same concrete bug." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Different model variants and different failure modes; not mergeable as one fix." + "reason": "Different Qwen/MoE reports with different triggers and symptoms." }, { - "left": "issue:43232", - "right": "issue:44792", + "left": "issue:36010", + "right": "issue:41628", "accept": false, - "reason": "Generation kwargs handling and Janus image test failure are unrelated." + "reason": "GenerationMixin import error and AutoImageProcessor import error are distinct missing-symbol regressions." }, { - "left": "issue:36010", - "right": "issue:41628", + "left": "issue:43232", + "right": "issue:44792", "accept": false, - "reason": "Different import errors for different APIs." + "reason": "Generation helper bug vs janus image generation test failure; no clear common root cause." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FP8 accuracy issue and FP8 save_pretrained failure are different bugs." + "reason": "FP8 accuracy issue and FP8 save_pretrained issue are different failure modes." }, { - "left": "issue:46077", - "right": "issue:46123", + "left": "issue:42175", + "right": "issue:45712", "accept": false, - "reason": "Gemma4 bidirectional masking and MaskGenerationPipeline batch handling are unrelated." + "reason": "Backend installation packaging issue vs dummy class cleanup; not the same bug." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same title and same Qwen3.5 MoE from_pretrained error; clear duplicate reports." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The only accepted pair (42673/43704) is an exact title duplicate, and the other pairwise rejections are reasonably cautious given the limited evidence. One note: 45310 and 45478 share the same title in the packet, but that pair was not among the provided soft edges, so it does not invalidate this evaluation." + "feedback": "The summary is grounded in the packet and stays conservative overall. The two accepted duplicate pairs are well supported by matching titles/symptoms, and the other soft-edge rejections are cautious rather than overstated." }, "evaluator_used": true, "retried": false, @@ -13022,21 +13030,20 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ - "issue:27957", "issue:30333", "issue:30990", "issue:33290", - "issue:34689", "issue:35141", "issue:42175", "issue:42222", "issue:42898", "issue:43278", + "issue:43295", "issue:43377", "issue:43479", "issue:43504", @@ -13053,7 +13060,6 @@ "issue:44242", "issue:44295", "issue:44297", - "issue:44448", "issue:44496", "issue:44568", "issue:44623", @@ -13074,16 +13080,18 @@ "issue:45656", "issue:45712", "issue:45812", + "issue:46077", + "issue:46123", "issue:46231" ], "soft_pairs": [ - "issue:42175|issue:45712", "issue:44496|issue:45561", - "issue:43504|issue:45561", "issue:42175|issue:44933", + "issue:43504|issue:45561", "issue:45561|issue:45600", "issue:43976|issue:44933", "issue:43278|issue:45137", + "issue:46077|issue:46123", "issue:43377|issue:44242", "issue:30990|issue:35141", "issue:42898|issue:44568", @@ -13094,243 +13102,228 @@ "issue:43479|issue:44079", "issue:42898|issue:45812", "issue:43723|issue:44568", - "issue:44016|issue:46231", "issue:43874|issue:45092", + "issue:44016|issue:46231", "issue:43784|issue:43824", "issue:42898|issue:44779", "issue:44734|issue:45103", "issue:42898|issue:45356", "issue:44295|issue:45656", - "issue:44060|issue:45125", "issue:44075|issue:46231", - "issue:42898|issue:44448", + "issue:42898|issue:43295", + "issue:44060|issue:45125", "issue:45405|issue:45712", "issue:44222|issue:44297", - "issue:44933|issue:45712", - "issue:27957|issue:30333", - "issue:30333|issue:34689", - "issue:44297|issue:45656" + "issue:44933|issue:45712" ], - "prepared_review_unit_hash": "dfbd592005302fd1067cbb5db3a05c9df013909d8973dfa56200655198b7a3ed", + "prepared_review_unit_hash": "038620d7d0dbccf517b358e866e579dec73ecd03ca001705e6d8bfd2edfdf75d", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37748, - "estimated_input_tokens": 9437, - "estimated_eval_tokens": 19130 + "soft_pair_count": 30, + "serialized_chars": 36943, + "estimated_input_tokens": 9236, + "estimated_eval_tokens": 18728 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a heterogeneous grab bag of unrelated issues across tokenization, model loading, training, packaging, serving, and docs. None of the suggested soft edges look like true duplicates or the same underlying fix.", - "confidence": 0.98, - "canonical_issue_reason": "No single canonical issue: the items span unrelated bugs in different subsystems, so there is no clear duplicate target to anchor the cluster.", + "summary": "This is a loose cluster of mostly unrelated Transformers bug reports, concentrated around tokenizer/model-regression issues but spanning many separate code paths. None of the soft pairs look like the same underlying bug, so all proposed merges are rejected.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45812 is the clearest high-signal, user-facing regression here: a silent AutoTokenizer token-id break across Granite models with a well-scoped tokenizer path and broad impact.", "canonical_pr_reason": null, - "best_issue_reason": "No clear best issue either; none of the issues is representative enough of the rest to serve as a global duplicate canonical.", + "best_issue_reason": "issue:45812 is the strongest standalone anchor in the set because it describes a concrete, broadly impactful regression with a specific reproducer and failure mode.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42175", - "right": "issue:45712", - "accept": false, - "reason": "Different problems: backend/extras installation vs leftover dummy PT classes leaking into namespace and failing checks." - }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Unrelated: model config loading error vs pytest-xdist file race in testing utils." + "reason": "Different problems: a model-loading/config error versus a pytest-xdist captured-info race." }, { - "left": "issue:43504", - "right": "issue:45561", + "left": "issue:42175", + "right": "issue:44933", "accept": false, - "reason": "Beit pretrained-load legacy field bug is unrelated to the captured_info.txt race." + "reason": "Unrelated: missing TensorFlow in an extra install versus a nonexistent import from image_utils." }, { - "left": "issue:42175", - "right": "issue:44933", + "left": "issue:43504", + "right": "issue:45561", "accept": false, - "reason": "Packaging/backend dependency issue vs a missing import in image_utils; different code paths." + "reason": "Legacy-field model preset loading failure does not match the xdist file-race bug." }, { "left": "issue:45561", "right": "issue:45600", "accept": false, - "reason": "Test-worker file race vs stale auto_mappings references breaking CI; not the same bug." + "reason": "Test harness race on captured_info.txt versus stale auto_mappings references to removed configs." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version compatibility regression vs missing import; separate failures." + "reason": "Python-version compatibility issue versus an import path bug in image_utils." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Embedding dtype mismatch in eval vs DeepSpeed ZeRO3 deque error; unrelated." + "reason": "Embedding dtype drift in eval is a different failure mode from a DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Gemma4 attention-mask construction bug versus a batch-iteration result-dropping bug." }, { "left": "issue:43377", "right": "issue:44242", "accept": false, - "reason": "MIMI padding-mask batching bug vs MoE router-loss condition; different model behaviors." + "reason": "Different model paths: MIMI padding-mask mismatch versus MoE load-balancing-loss gating." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Loading hang in Sentence Transformers vs embedding reinit after resize; no shared code-path." + "reason": "Sentence-transformers loading hang is unrelated to token-embedding reinitialization in post_init." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Tokenization cleanup behavior change vs special-tokens insertion regression; distinct tokenizer logic." + "reason": "Tokenizer cleanup-space behavior change is not the same as missing BOS/EOS insertion on add_special_tokens." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLflow reporting after failure vs OOM in Adafactor/DeepSpeed; unrelated training issues." + "reason": "MLflow failure reporting and Adafactor OOM in DeepSpeed are distinct training issues." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "save_pretrained missing files vs Qwen3.5-MoE from_pretrained error; different subsystems." + "reason": "Processor save_pretrained file omissions do not match the Qwen3.5-Moe from_pretrained error." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Multimodal config default initialization vs Whisper timestamp decode crash; unrelated." + "reason": "Default multimodal config initialization is unrelated to Whisper timestamp decoding on trailing replacement chars." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "VitPose model breakage vs Gemma 4 multimodal token defaults; different models and failures." + "reason": "Broken vitpose presets and Gemma 4 mm_token_type_ids defaults are different model families and bugs." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Config initialization bug vs ModelOutput key assignment bug; not the same underlying issue." + "reason": "Config-default initialization bug versus ModelOutput key assignment semantics." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but one is cleanup-space behavior and the other is Granite token IDs; distinct root causes." + "reason": "Both are tokenizer regressions, but one is cleanup-space behavior and the other is wrong token IDs for Granite models." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Tokenizer loading in v5 vs special-token behavior for mdeberta; separate failure modes." + "reason": "AutoTokenizer loading regression in v5 is not the same bug as missing special-token insertion for mdeberta-v3." }, { - "left": "issue:44016", - "right": "issue:46231", + "left": "issue:43874", + "right": "issue:45092", "accept": false, - "reason": "Notebook syntax error vs documentation parameter-name mismatch; not duplicates." + "reason": "Missing image-patch helper is unrelated to old InternVL2 checkpoints breaking under meta initialization." }, { - "left": "issue:43874", - "right": "issue:45092", + "left": "issue:44016", + "right": "issue:46231", "accept": false, - "reason": "Missing image processor method vs remote-code meta-init incompatibility; different code paths." + "reason": "Notebook syntax error and documentation parameter-name mismatch are not duplicate code bugs." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Import failure in sentence-transformers vs missing Qwen2.5-VL export; unrelated import issues." + "reason": "Import NameError in sentence-transformers is unrelated to Qwen2.5-VL import availability." }, { "left": "issue:42898", "right": "issue:44779", "accept": false, - "reason": "Different tokenizer regressions: cleanup_spaces vs Deepseek tokenization correctness." + "reason": "Cleanup-space regression and Deepseek tokenizer incorrect results are separate tokenizer failures." }, { "left": "issue:44734", "right": "issue:45103", "accept": false, - "reason": "Serving KV-cache indexing crash vs auto_docstring annotation handling crash; unrelated." + "reason": "Serve KV-cache tensor indexing crash versus auto_docstring future-annotations crash are different subsystems." }, { "left": "issue:42898", "right": "issue:45356", "accept": false, - "reason": "Both tokenizer regressions, but Kimi codec handling/warning issue is not the same as clean_up_tokenization_spaces behavior." + "reason": "Cleanup-space behavior change does not match the Kimi-K2.5 codec-handling regression." }, { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "Reading registered position_ids buffer vs optimizer step called twice with DeepSpeed; different bugs." - }, - { - "left": "issue:44060", - "right": "issue:45125", - "accept": false, - "reason": "Incorrect tied-weights warning in Qwen3-Next vs missing _tp_plan in Qwen3_5MoE; separate model issues." + "reason": "position_ids buffer access error is a different issue from DeepSpeed causing optimizer.step to run twice." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "Optimizer SGD args ignored vs documentation typo; not the same underlying change." + "reason": "Optimizer argument handling bug and a docs typo in set_training are not the same code-path problem." }, { "left": "issue:42898", - "right": "issue:44448", + "right": "issue:43295", + "accept": false, + "reason": "Tokenizer cleanup-space behavior is unrelated to processor.tokenizer / image-passing regression in custom model code." + }, + { + "left": "issue:44060", + "right": "issue:45125", "accept": false, - "reason": "v5 tokenizer cleanup regression vs Pegasus v4/v5 output difference; different tokenizer/model behavior." + "reason": "Both mention Qwen3 variants, but the tied-weights warning and missing tensor-parallel plan are different defects." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "Dependency version bump issue vs leftover dummy class cleanup; unrelated maintenance problems." + "reason": "PEFT version pinning issue and leftover dummy-class cleanup are unrelated maintenance problems." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "FP8 MoE save_pretrained bug vs tokenizer_config.json class mismatch; different serialization paths." - }, - { - "left": "issue:27957", - "right": "issue:30333", - "accept": false, - "reason": "Flash Attention 2 support for XLM-Roberta vs MLflow reporting on failed training jobs; unrelated." - }, - { - "left": "issue:30333", - "right": "issue:34689", - "accept": false, - "reason": "MLflow failure-reporting bug vs Llama 3.2 Vision model-loading regression; different areas." + "reason": "FP8 save_pretrained behavior and tokenizer_class metadata mismatch affect different save paths." }, { - "left": "issue:44297", - "right": "issue:45656", + "left": "issue:44933", + "right": "issue:45712", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch vs DeepSpeed optimizer double-step; no shared bug." + "reason": "Missing import from image_utils is unrelated to stale dummy classes leaking into dir(transformers)." } ] }, @@ -13349,14 +13342,16 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ "issue:22355", + "issue:27957", "issue:28282", "issue:30333", + "issue:34689", "issue:35141", "issue:36331", "issue:41628", @@ -13376,10 +13371,8 @@ "issue:44279", "issue:44295", "issue:44297", - "issue:44360", "issue:44373", "issue:44448", - "issue:44483", "issue:44484", "issue:44485", "issue:44623", @@ -13392,7 +13385,7 @@ "issue:45245", "issue:45310", "issue:45381", - "issue:45406", + "issue:45656", "issue:45834", "issue:46082", "issue:46132", @@ -13404,11 +13397,14 @@ "issue:46431" ], "soft_pairs": [ + "issue:27957|issue:30333", + "issue:30333|issue:34689", + "issue:44297|issue:45656", "issue:44448|issue:46132", - "issue:42796|issue:43723", "issue:44222|issue:44623", - "issue:43479|issue:43644", "issue:30333|issue:36331", + "issue:43479|issue:43644", + "issue:42796|issue:43723", "issue:42831|issue:44623", "issue:30333|issue:35141", "issue:22355|issue:28282", @@ -13421,9 +13417,9 @@ "issue:43723|issue:44977", "issue:45381|issue:46305", "issue:45071|issue:46082", - "issue:43232|issue:46424", - "issue:44206|issue:45103", "issue:44263|issue:45092", + "issue:44206|issue:45103", + "issue:43232|issue:46424", "issue:44222|issue:44484", "issue:44734|issue:46421", "issue:44373|issue:44485", @@ -13431,218 +13427,222 @@ "issue:44295|issue:44623", "issue:44190|issue:46082", "issue:43723|issue:44871", - "issue:44279|issue:46328", - "issue:44483|issue:45406", - "issue:44263|issue:44360" + "issue:44279|issue:46328" ], - "prepared_review_unit_hash": "90cbf2ebe5613a0ce0c7e82e86c796e0b757887f0aee7c35521901e20c24fd3e", + "prepared_review_unit_hash": "16055b342215787629e40ebc54d606e00a53d57d0f8d3d2f31c79b9800d4d0b3", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36478, - "estimated_input_tokens": 9120, - "estimated_eval_tokens": 18496 + "soft_pair_count": 31, + "serialized_chars": 36724, + "estimated_input_tokens": 9181, + "estimated_eval_tokens": 18618 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The proposed soft links are all false positives: the items span many unrelated bugs and subsystems (imports, training, generation, tokenization, cache, multimodal, FP8, docs). There is no single duplicate issue or PR candidate in this set.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "This cluster is heterogeneous: the soft edges are mostly weak topical overlaps (imports, tokenizers, config, generation, cache, FP8, etc.) rather than the same underlying bug. No PRs are present, and there is no clear duplicate canonical issue.", + "confidence": 0.95, + "canonical_issue_reason": "No single issue is a defensible canonical: the items cover many unrelated failure modes, so the cluster should likely be split instead of deduplicated.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No issue is a strong global representative; the set is too broad to treat as one bug report.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44448", - "right": "issue:46132", + "left": "issue:27957", + "right": "issue:30333", "accept": false, - "reason": "Different areas: model output differences vs AttentionInterface registration behavior." + "reason": "Different problems: Flash Attention 2 model support vs MLflow reporting on failed training jobs." }, { - "left": "issue:42796", - "right": "issue:43723", + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Unrelated: MLflow job-status reporting vs Llama 3.2 Vision model-loading regression." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to optimizer stepping twice under deepspeed." + }, + { + "left": "issue:44448", + "right": "issue:46132", "accept": false, - "reason": "Tokenizer BPE-marker handling and AutoTokenizer import/loading are unrelated failures." + "reason": "Different code paths: Pegasus output changes vs AttentionInterface.register altering callback behavior." }, { "left": "issue:44222", "right": "issue:44623", "accept": false, - "reason": "FP8 MoE save_pretrained and processor save_pretrained missing-files are different code paths." + "reason": "Both mention save_pretrained, but one is FP8 MoE model serialization and the other is missing processor files." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "Training-job MLflow reporting is unrelated to CustomTrainer.compute_loss signature changes." }, { "left": "issue:43479", "right": "issue:43644", "accept": false, - "reason": "Multimodal config defaults vs non-persistent buffer initialization junk are unrelated." + "reason": "Config default initialization bug vs non-persistent buffer initialization bug." }, { - "left": "issue:30333", - "right": "issue:36331", + "left": "issue:42796", + "right": "issue:43723", "accept": false, - "reason": "MLflow reporting on failed jobs and Trainer compute_loss kwargs are unrelated bugs." + "reason": "Tokenizer BPE-marker handling and AutoTokenizer loading regression are different bugs." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FP8 accuracy issue and processor serialization missing files do not match." + "reason": "FP8 accuracy regression is unrelated to processor.save_pretrained missing files." }, { "left": "issue:30333", "right": "issue:35141", "accept": false, - "reason": "Failed-job MLflow tracking and embedding reinitialization are different problems." + "reason": "MLflow failure reporting and embedding resize/post_init reinitialization are unrelated." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "ONNX import failure is unrelated to missing PyTorch dependency errors." + "reason": "Both are import errors, but for different missing dependencies and modules." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "Category cardinality limit and AttentionInterface behavior have no common bug path." + "reason": "Category-cardinality runtime error is unrelated to AttentionInterface.register behavior changes." }, { "left": "issue:46209", "right": "issue:46431", "accept": false, - "reason": "FP8 inference on B200 and missing chat template on a processor are unrelated." + "reason": "FP8 inference on B200 is unrelated to processor chat-template absence." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Separate import errors from different missing symbols in different packages." + "reason": "Different import failures in different packages/modules; not the same bug." }, { "left": "issue:43475", "right": "issue:46305", "accept": false, - "reason": "SAM3 output attribute error and MoonshineStreaming off-by-one mask bug are unrelated." + "reason": "SAM 3 Vision missing attribute and MoonshineStreaming mask off-by-one are unrelated." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer config save mismatch and Qwen3.5 MoE from_pretrained error are distinct issues." + "reason": "Tokenizer-class metadata mismatch is not the same as Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:43122", "right": "issue:46209", "accept": false, - "reason": "Tokenization regression and FP8 B200 inference failure are different subsystems." + "reason": "Tokenization differences across versions are unrelated to FP8 inference failures on B200." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "AutoTokenizer loading issue and flash-attention generation bug are unrelated." + "reason": "Tokenizer loading regression and Qwen3.5 flash-attention generation failure are different code paths." }, { "left": "issue:45381", "right": "issue:46305", "accept": false, - "reason": "Qwen2.5-VL vision_position_ids and MoonshineStreaming windowing are separate bugs." + "reason": "Both mention positional handling, but they are model/modality-specific and not the same defect." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "PretrainedConfig type checking and LlamaConfig head_dim validation are not the same defect." + "reason": "Config type-checking regression is unrelated to LlamaConfig head_dim validation." }, { - "left": "issue:43232", - "right": "issue:46424", + "left": "issue:44263", + "right": "issue:45092", "accept": false, - "reason": "Generation kwargs update after sync_gpus and cache max length handling are different code paths." + "reason": "torch.split return values bug is unrelated to InternVL2 meta-initialization incompatibility." }, { "left": "issue:44206", "right": "issue:45103", "accept": false, - "reason": "Feature-extractor center-arg crash and auto_docstring annotation crash are unrelated." + "reason": "Feature extractor argument handling and auto-docstring annotation crashes are unrelated." }, { - "left": "issue:44263", - "right": "issue:45092", + "left": "issue:43232", + "right": "issue:46424", "accept": false, - "reason": "GlmMoeDsaIndexer torch.split behavior and InternVL2 meta-init incompatibility do not match." + "reason": "Generation kwargs update after sync_gpus is not the same bug as cache max_cache_length handling." }, { "left": "issue:44222", "right": "issue:44484", "accept": false, - "reason": "FP8 MoE save_pretrained and max_shard_size default discussion are unrelated." + "reason": "FP8 MoE save_pretrained issue is unrelated to the default 50GB shard-size question." }, { "left": "issue:44734", "right": "issue:46421", "accept": false, - "reason": "Serve KV-cache continuation crash and lazy cache recompilation are different cache bugs." + "reason": "KV-cache continuation indexing crash and chunked-prefill cache recompilation are different cache bugs." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "Docstring correction for position_ids is unrelated to GLM-5 RoPE implementation." + "reason": "Docstring typo/accuracy issue is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:43232", "right": "issue:45834", "accept": false, - "reason": "Generation bookkeeping bug and Kosmos2.5 OCR long-input index error are unrelated." + "reason": "Generation sync_gpus bug is unrelated to Kosmos2.5 long-OCR index errors." }, { "left": "issue:44295", "right": "issue:44623", "accept": false, - "reason": "position_ids buffer access and processor.save_pretrained file output are different issues." + "reason": "Position_ids buffer read error is unrelated to missing processor save_pretrained files." }, { "left": "issue:44190", "right": "issue:46082", "accept": false, - "reason": "Dataset-loading script bug and LlamaConfig head_dim validation are unrelated." + "reason": "Local dataset loading in an example script is unrelated to LlamaConfig head_dim validation." }, { "left": "issue:43723", "right": "issue:44871", "accept": false, - "reason": "AutoTokenizer loading regression and Gemma-3 eos_token_id inconsistency are different problems." + "reason": "Tokenizer loading regression and eos_token_id config inconsistency are different failures." }, { "left": "issue:44279", "right": "issue:46328", "accept": false, - "reason": "Generic dependency issue and SequenceFeatureExtractor.pad performance bug are unrelated." - }, - { - "left": "issue:44483", - "right": "issue:45406", - "accept": false, - "reason": "Chat/completions request rejection and Gemma4Processor missing _tokenizer are different failures." - }, - { - "left": "issue:44263", - "right": "issue:44360", - "accept": false, - "reason": "GlmMoeDsaIndexer torch.split issue and DSA indexer ReLU discussion are not the same bug." + "reason": "A vague dependency issue is not the same as a SequenceFeatureExtractor.pad performance regression." } ] }, @@ -13661,8 +13661,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -13692,20 +13692,20 @@ "issue:43784", "issue:43824", "issue:44038", - "issue:44117", - "issue:44188", "issue:44261", "issue:44263", + "issue:44360", "issue:44368", "issue:44373", + "issue:44483", "issue:44568", - "issue:44704", "issue:44829", "issue:44913", "issue:45092", "issue:45278", "issue:45310", "issue:45357", + "issue:45406", "issue:45593", "issue:45753", "issue:46002", @@ -13716,10 +13716,12 @@ "issue:46402" ], "soft_pairs": [ + "issue:44483|issue:45406", + "issue:44263|issue:44360", "issue:42907|issue:46050", - "issue:30990|issue:38175", "issue:45092|issue:45357", "issue:44261|issue:45092", + "issue:30990|issue:38175", "issue:43644|issue:46305", "issue:42831|issue:45310", "issue:44829|issue:46002", @@ -13729,232 +13731,230 @@ "issue:43479|issue:44913", "issue:43408|issue:46402", "issue:22355|issue:27957", - "issue:42796|issue:43122", "issue:42757|issue:45278", "issue:39401|issue:44568", - "issue:42886|issue:43066", + "issue:42796|issue:43122", "issue:42831|issue:45753", + "issue:42886|issue:43066", "issue:22355|issue:28218", "issue:44368|issue:46305", "issue:27957|issue:30990", "issue:42886|issue:42994", "issue:43232|issue:45593", "issue:43408|issue:46369", - "issue:43550|issue:46132", "issue:43646|issue:46002", + "issue:43550|issue:46132", "issue:43576|issue:43784", - "issue:42371|issue:44263", - "issue:44188|issue:44373", - "issue:44117|issue:44704" + "issue:42371|issue:44263" ], - "prepared_review_unit_hash": "88317c3c7c92a730b76895325da01ec4b5f7a1890b6adfce93c2be1727b307d6", + "prepared_review_unit_hash": "3b02d830f508d6e19d31df2c7ae0dc1e8cb83edeb578de6745923a41d930b682", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36288, - "estimated_input_tokens": 9072, - "estimated_eval_tokens": 18400 + "serialized_chars": 36250, + "estimated_input_tokens": 9063, + "estimated_eval_tokens": 18382 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a loose collection of mostly unrelated Transformers regression reports. The only common thread is that several involve v5-era import/tokenizer/config/save-load regressions, but the soft pairs are not close enough to treat as true duplicates.", - "confidence": 0.36, - "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like issue in the set: it captures multiple import failures after a version upgrade and best matches the recurring migration/regression theme.", + "summary": "This cluster is heterogeneous: most items are unrelated model, tokenizer, import, or config bugs, and there are no pull requests. I treated the broad import-regression report as the best issue anchor, but rejected all soft edges as non-duplicates.", + "confidence": 0.68, + "canonical_issue_reason": "Issue #45278 is the broadest umbrella item here: it captures multiple import failures after a version upgrade, making it the most representative anchor for this mixed set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single issue to anchor this cluster because it is broad, user-facing, and plausibly subsumes narrower import-related reports better than the model-specific bugs.", + "best_issue_reason": "Best general issue candidate for dedup triage because it summarizes a class of upgrade-related import breakages; the other issues are narrower, model- or tokenizer-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42907", - "right": "issue:46050", + "left": "issue:44483", + "right": "issue:45406", "accept": false, - "reason": "Both concern quantized model save/load, but they appear to target different quantization paths and symptoms; not enough evidence of the same underlying bug." + "reason": "Different failures: chat/completions request validation vs Gemma4Processor missing _tokenizer." }, { - "left": "issue:30990", - "right": "issue:38175", + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Same subsystem (DSA indexer) but different bugs: torch.split behavior vs missing ReLU." + }, + { + "left": "issue:42907", + "right": "issue:46050", "accept": false, - "reason": "Different models and different failures: loading hangs versus zero-probability outputs." + "reason": "Different quantization save/load paths: dequantized Ministrals/Devstrals vs sinq models." }, { "left": "issue:45092", "right": "issue:45357", "accept": false, - "reason": "Both are multimodal/vision-related regressions, but one is a load-time meta-init incompatibility and the other is a save_pretrained key serialization bug." + "reason": "Different models and code paths: InternVL2 meta-init incompatibility vs Qwen3.5 visual encoder save keys." }, { "left": "issue:44261", "right": "issue:45092", "accept": false, - "reason": "Completely different code paths and symptoms; this is not the same bug." + "reason": "Unrelated bugs: MLA RMS norm epsilon precision issue vs InternVL2 meta initialization failure." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Different symptoms and models: stuck loading sentence-transformers vs zero probabilities in SigLIP2." }, { "left": "issue:43644", "right": "issue:46305", "accept": false, - "reason": "Non-persistent buffer corruption is unrelated to the MoonshineStreaming off-by-one mask issue." + "reason": "Different code paths: non-persistent buffer initialization junk vs sliding-window mask off-by-one." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "FineGrainedFP8 accuracy issues and Qwen3.5 Moe from_pretrained errors are different failures." + "reason": "Different failures: FineGrainedFP8 accuracy regression vs Qwen3.5 MoE from_pretrained error." }, { "left": "issue:44829", "right": "issue:46002", "accept": false, - "reason": "Training divergence with flash_attention_3 is not the same as NaNs during checkout/init." + "reason": "Different issues: Flash Attention 3 training degeneration vs NaNs in classification heads after init." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "A tensor-splitting bug and a docstring issue are unrelated." + "reason": "One is a functional DSA-indexer bug; the other is only a docstring fix." }, { "left": "issue:41628", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and likely different root causes." + "reason": "Both are import errors, but for different missing symbols and likely different registration fixes." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "These are different model families and different regression types." + "reason": "Different bugs: FineGrainedFP8 accuracy issue vs Qwen3-VL-Moe failure in v5." }, { "left": "issue:43479", "right": "issue:44913", "accept": false, - "reason": "Both involve config defaults on reload/init, but the affected models and behavior are different." + "reason": "Different config bugs in different models: default multimodal configs vs GPTNeoX rotary_pct reload." }, { "left": "issue:43408", "right": "issue:46402", "accept": false, - "reason": "These are different SAM3 issues: a warning about model type mismatch versus missing text encoder weights." - }, - { - "left": "issue:22355", - "right": "issue:27957", - "accept": false, - "reason": "No shared underlying bug; one is a missing module import, the other is a Flash Attention 2 model issue." + "reason": "Same model family, but different problems: wrong model-type warning vs text encoder weights not loading." }, { - "left": "issue:42796", - "right": "issue:43122", + "left": "issue:22355", + "right": "issue:27957", "accept": false, - "reason": "Tokenizer behavior regressions, but one is a model-specific BPE marker problem and the other is a broader version-to-version tokenization change." + "reason": "Unrelated: missing transformers.onnx module vs XLMRoberta Flash Attention 2 behavior." }, { "left": "issue:42757", "right": "issue:45278", "accept": false, - "reason": "Related only at a very broad level; one is a specific missing dependency import, the other is a generic report of many import errors after upgrade." + "reason": "Both are upgrade-related import issues, but not the same concrete missing symbol or failure path." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different tokenization semantics." + "reason": "Different tokenizer bugs: wrong offset mapping vs BOS/EOS not added." }, { - "left": "issue:42886", - "right": "issue:43066", + "left": "issue:42796", + "right": "issue:43122", "accept": false, - "reason": "Cache/offline loading failure is unrelated to wrong decoder type selection." + "reason": "Both concern tokenization drift, but they describe different model-specific bugs and outputs." }, { "left": "issue:42831", "right": "issue:45753", "accept": false, - "reason": "An FP8 accuracy issue and an infinite-loop multimodal bug are unrelated." + "reason": "Unrelated: FP8 accuracy regression vs Qwen3_5 image-triggered infinite loop." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Different regressions: offline cache loading failure vs wrong tokenizer decoder type in v5." }, { "left": "issue:22355", "right": "issue:28218", "accept": false, - "reason": "Missing module import versus tokenizer spacing behavior; not the same bug." + "reason": "Completely different topics: missing onnx import vs tokenizer spacing after added token." }, { "left": "issue:44368", "right": "issue:46305", "accept": false, - "reason": "A configuration warning is unrelated to a streaming attention mask off-by-one error." + "reason": "Different issues: tie_word_embeddings warning vs sliding-window mask off-by-one." }, { "left": "issue:27957", "right": "issue:30990", "accept": false, - "reason": "Different models and different problems; not duplicates." + "reason": "Different problems in different contexts: Flash Attention 2 support vs loading hang." }, { "left": "issue:42886", "right": "issue:42994", "accept": false, - "reason": "Offline cache loading and quantized model save failure are different issues." + "reason": "Unrelated failures: offline tokenizer cache vs quantized model save failure." }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Generation-state handling and D-FINE auxiliary losses are unrelated." + "reason": "Different subsystems entirely: generation kwargs sync logic vs D-FINE denoising losses." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "Model-type warning versus memory overflow are different symptoms and code paths." + "reason": "Same family (sam3) but different bugs: tracker warning vs memory overflow in litetext." }, { - "left": "issue:43550", - "right": "issue:46132", + "left": "issue:43646", + "right": "issue:46002", "accept": false, - "reason": "torch.compile SDPA failure is not the same as AttentionInterface.register changing registered-function behavior." + "reason": "Different regressions: custom model init breakage vs NaNs in classification heads." }, { - "left": "issue:43646", - "right": "issue:46002", + "left": "issue:43550", + "right": "issue:46132", "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are not the same concrete bug." + "reason": "Different code paths: Bamba torch.compile/SDPA failure vs AttentionInterface.register behavior change." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "Broken CLI command and sentence-transformers import NameError are unrelated." + "reason": "Unrelated: transformers env CLI issue vs sentence-transformers import NameError." }, { "left": "issue:42371", "right": "issue:44263", "accept": false, - "reason": "TF32 API guidance/warning is unrelated to the torch.split indexing bug." - }, - { - "left": "issue:44188", - "right": "issue:44373", - "accept": false, - "reason": "Attention-kernel divergence under torch.compile is unrelated to a docstring issue." - }, - { - "left": "issue:44117", - "right": "issue:44704", - "accept": false, - "reason": "A tokenizer mapping assumption bug and missing kwargs forwarding in AutoProcessor are different failures." + "reason": "No meaningful overlap: TF32 settings warning vs DSA-indexer torch.split bug." } ] }, @@ -13973,8 +13973,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -13985,7 +13985,6 @@ "issue:36296", "issue:39401", "issue:41628", - "issue:42175", "issue:42757", "issue:42907", "issue:43116", @@ -14001,6 +14000,8 @@ "issue:43976", "issue:44038", "issue:44060", + "issue:44117", + "issue:44188", "issue:44222", "issue:44261", "issue:44295", @@ -14015,7 +14016,6 @@ "issue:44625", "issue:44704", "issue:44829", - "issue:44938", "issue:44945", "issue:44964", "issue:45200", @@ -14027,28 +14027,30 @@ "issue:46326" ], "soft_pairs": [ + "issue:44188|issue:44373", + "issue:44117|issue:44704", "issue:44295|issue:44297", "issue:30990|issue:33357", - "issue:43756|issue:44060", - "issue:42907|issue:44945", "issue:44261|issue:44373", + "issue:42907|issue:44945", + "issue:43756|issue:44060", "issue:41628|issue:45278", "issue:42907|issue:43937", "issue:43976|issue:44945", "issue:39401|issue:44488", "issue:43122|issue:44521", - "issue:42907|issue:44964", - "issue:44704|issue:45290", "issue:43784|issue:45600", + "issue:44704|issue:45290", + "issue:42907|issue:44964", "issue:43122|issue:46326", "issue:39401|issue:44521", "issue:44360|issue:44485", - "issue:43116|issue:45200", "issue:44038|issue:44297", + "issue:43116|issue:45200", "issue:44625|issue:44829", + "issue:44038|issue:44295", "issue:43408|issue:43475", "issue:42907|issue:43452", - "issue:44038|issue:44295", "issue:43122|issue:44514", "issue:44038|issue:44222", "issue:43824|issue:44351", @@ -14056,223 +14058,228 @@ "issue:22355|issue:29159", "issue:42757|issue:44351", "issue:36296|issue:43116", - "issue:43278|issue:44945", - "issue:42175|issue:44938" + "issue:43278|issue:44945" ], - "prepared_review_unit_hash": "0c2f1e2cca87744bc9dc100cae71472efb2990959fdd4c51f23d4816c050c0f3", + "prepared_review_unit_hash": "59344a7c77e1baf2993465226de7cb00ed408eb67c068d51d05f913125ef1547", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36027, - "estimated_input_tokens": 9007, - "estimated_eval_tokens": 18270 + "soft_pair_count": 32, + "serialized_chars": 36393, + "estimated_input_tokens": 9099, + "estimated_eval_tokens": 18454 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft-similar pairs share only surface keywords (imports, tokenizer, save/load, multimodal, Qwen/SAM), but they describe different bugs or regressions. I would not merge any of the suggested pairs as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue stands out because the cluster does not appear to represent one underlying bug; it is a set of unrelated regression reports that only look similar lexically.", + "summary": "The candidate pairs are mostly superficial similarities across unrelated bugs. This set spans import errors, tokenizer behavior, multimodal chat templates, training, saving, and model-specific regressions, so no pair looks like the same underlying issue/change.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No strong best issue candidate; the most-connected items are still about different code paths, so any one issue would be a misleading representative.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Both mention position-related behavior, but one is a torch.compile attention-kernel divergence bug and the other is a docstring issue; not the same bug." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Both concern from_pretrained plumbing, but one is TOKENIZER_MAPPING_NAMES None handling and the other is cached_file kwargs passthrough; different code paths." + }, { "left": "issue:44295", "right": "issue:44297", "accept": false, - "reason": "One is a position_ids buffer access error; the other is tokenizer.save_pretrained writing the wrong tokenizer_class. Different code paths and failures." + "reason": "Both are about position_ids/tokenizer save behavior, but one is reading a registered buffer and the other is tokenizer_class metadata mismatch; distinct failures." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence Transformers hanging on load and a MacOS bus error with a CLIP model are different symptoms and likely different root causes." + "reason": "Different symptoms and subsystems: Sentence Transformers loading hang vs MacOS bus error with a CLIP model." }, { - "left": "issue:43756", - "right": "issue:44060", + "left": "issue:44261", + "right": "issue:44373", "accept": false, - "reason": "Smollm3 RoPE-layer omission and Qwen3-Next tied-weights warnings concern different model implementations and distinct bugs." + "reason": "LayerNorm epsilon precision mismatch is unrelated to a wrong docstring for position_ids." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Saving dequantized models and incorrect output under pipeline parallelism are unrelated failures in different execution paths." + "reason": "Dequantized save_pretrained failure is unrelated to incorrect LLM output under pipeline parallelism." }, { - "left": "issue:44261", - "right": "issue:44373", + "left": "issue:43756", + "right": "issue:44060", "accept": false, - "reason": "A layernorm epsilon precision issue is not the same as a wrong docstring for position_ids." + "reason": "Both involve model internals, but Smollm3 RoPE layer dropping and Qwen3-Next tied-weights warnings are different model-specific bugs." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "A missing AutoImageProcessor import and broad import errors after upgrade are related only at a high level; the concrete failing symbols and causes differ." + "reason": "Both are import-related, but AutoImageProcessor import failure and many import errors after a major version upgrade are too broad and not the same issue." }, { "left": "issue:42907", "right": "issue:43937", "accept": false, - "reason": "Dequantized-save failures and invalid GenerationConfig errors are separate model-saving vs generation-validation problems." + "reason": "Saving dequantized models and invalid GenerationConfig errors are unrelated." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python-version compatibility failure and wrong LLM output under pipeline parallelism are not the same bug." + "reason": "Python version compatibility breakage is unrelated to incorrect outputs under pipeline parallelism." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Qwen3 tokenizer offset mapping is unrelated to loading cjvt/sleng-bert." + "reason": "Tokenizer offset_mapping bug for Qwen3 is unrelated to a model loading failure for cjvt/sleng-bert." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Tokenizer-version behavior changes and zero assistant_masks for multimodal inputs are different tokenization/template bugs." + "reason": "Different tokenizer/template bugs: version-dependent tokenization changes vs multimodal assistant_masks all-zero output." }, { - "left": "issue:42907", - "right": "issue:44964", + "left": "issue:43784", + "right": "issue:45600", "accept": false, - "reason": "A dequantized-save regression is unrelated to loading microsoft/Phi-4-multimodal-instruct." + "reason": "Importing sentence-transformers with latest transformers and Sam3LiteText auto_mappings CI breakage are different import/config issues." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "Passing kwargs to cached_file in AutoProcessor and crashing on tool-call assistant messages are different apply-chat-template/loader issues." + "reason": "AutoProcessor kwargs forwarding and chat_template crashes on tool-call assistant messages are different bugs." }, { - "left": "issue:43784", - "right": "issue:45600", + "left": "issue:42907", + "right": "issue:44964", "accept": false, - "reason": "A NameError during sentence-transformers import and stale auto_mappings references are distinct import/configuration problems." + "reason": "Model saving failure is unrelated to loading a specific multimodal model." }, { "left": "issue:43122", "right": "issue:46326", "accept": false, - "reason": "Tokenizer behavior drift across versions and continue_final_message bugs for mistral_common tokenizers are not the same defect." + "reason": "Different tokenizer behavior regressions: version-to-version tokenization drift vs continue_final_message with mistral_common." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Wrong offset_mapping and all-zero assistant_masks are separate tokenizer/template behaviors." + "reason": "Qwen3 offset mapping and multimodal assistant mask generation are separate tokenizer/template problems." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "A missing ReLU in DSA indexing and GLM-5 RoPE implementation are unrelated model-internal issues." + "reason": "A missing ReLU in the DSA indexer is unrelated to GLM-5 RoPE implementation concerns." }, { - "left": "issue:43116", - "right": "issue:45200", + "left": "issue:44038", + "right": "issue:44297", "accept": false, - "reason": "Empty results in the classification example and missing default mm_token_type_ids are different task-specific bugs." + "reason": "Qwen3-VL-Moe bug and tokenizer.save_pretrained metadata mismatch affect different components." }, { - "left": "issue:44038", - "right": "issue:44297", + "left": "issue:43116", + "right": "issue:45200", "accept": false, - "reason": "A Qwen3-VL-Moe load bug and tokenizer_class mismatch on save_pretrained affect different subsystems." + "reason": "Empty results in an example classification script and Gemma 4 mm_token_type_ids defaults are different issues." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "num_labels propagation in Qwen3.5 config and degenerate training with flash_attention_3 are unrelated." + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 training degeneration are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe and position_ids buffer reading are unrelated." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "A model-type warning for sam3_tracker and a missing fpn_position_embeddings attribute are different SAM3 issues." + "reason": "Model-type warning for sam3_video vs sam3_tracker is not the same as missing fpn_position_embeddings on Sam3VisionEncoderOutput." }, { "left": "issue:42907", "right": "issue:43452", "accept": false, - "reason": "Saving dequantized models and gguf_file breaking from_pretrained loading are separate save/load paths." - }, - { - "left": "issue:44038", - "right": "issue:44295", - "accept": false, - "reason": "Qwen3-VL-Moe loading and position_ids buffer access are unrelated bugs." + "reason": "Saving dequantized models and gguf_file failures in from_pretrained are different code paths." }, { "left": "issue:43122", "right": "issue:44514", "accept": false, - "reason": "Tokenizer output changes across versions and apply_chat_template crashing on batched input with padding=False are different failures." + "reason": "Version-dependent tokenization differences are unrelated to batched apply_chat_template padding=False crashes." }, { "left": "issue:44038", "right": "issue:44222", "accept": false, - "reason": "Qwen3-VL-Moe load failure and FP8 save_pretrained on moe models do not look like the same concrete bug." + "reason": "Qwen3-VL-Moe loading issues are distinct from FP8 save_pretrained for moe." }, { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "Two import errors for different symbols (Qwen2_5_VLForConditionalGeneration vs HybridCache) are not the same underlying issue." + "reason": "Both are import failures, but Qwen2_5_VLForConditionalGeneration missing and HybridCache missing are different symbols and regressions." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "Both involve Qwen3.5 regressions, but one is save_pretrained checkpoint corruption and the other is from_pretrained load failure; not the same code-path bug." + "reason": "Both involve Qwen3.5 save_pretrained behavior, but one is a saved-checkpoint regression and the other is a from_pretrained error; not clearly the same bug." }, { "left": "issue:22355", "right": "issue:29159", "accept": false, - "reason": "Missing transformers.onnx and tokenizer inconsistency are unrelated issues." + "reason": "A missing transformers.onnx module and tokenizer slow/fast inconsistency are unrelated." }, { "left": "issue:42757", "right": "issue:44351", "accept": false, - "reason": "An import removed from huggingface_hub and a missing transformers.HybridCache symbol are different package-level import problems." + "reason": "Both are import errors, but is_offline_mode from huggingface_hub and HybridCache from transformers are different dependency issues." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training bugs and example-script empty classification results are not the same defect." + "reason": "Tensor parallel training bug and empty multi-label classification example output are unrelated." }, { "left": "issue:43278", "right": "issue:44945", "accept": false, - "reason": "BF16-to-FP32 dtype drift during evaluate and incorrect output with pipeline parallelism are different problems." - }, - { - "left": "issue:42175", - "right": "issue:44938", - "accept": false, - "reason": "Missing TensorFlow in a torch install extra and Python 3.14 load failure are unrelated environment/compatibility issues." + "reason": "Embedding dtype change during evaluate is unrelated to incorrect outputs under pipeline parallelism." } ] }, @@ -14291,8 +14298,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -14301,6 +14308,7 @@ "issue:33453", "issue:38617", "issue:39401", + "issue:42175", "issue:42222", "issue:42371", "issue:42673", @@ -14315,7 +14323,6 @@ "issue:43825", "issue:43827", "issue:43874", - "issue:43906", "issue:43937", "issue:43976", "issue:43992", @@ -14325,12 +14332,12 @@ "issue:44188", "issue:44355", "issue:44380", - "issue:44466", "issue:44485", "issue:44625", "issue:44704", "issue:44779", "issue:44936", + "issue:44938", "issue:45072", "issue:45083", "issue:45278", @@ -14346,22 +14353,23 @@ "issue:46328" ], "soft_pairs": [ + "issue:42175|issue:44938", "issue:43493|issue:45753", "issue:22355|issue:29127", "issue:43992|issue:44704", "issue:43066|issue:44779", - "issue:42673|issue:45072", "issue:43937|issue:44485", "issue:38617|issue:45278", "issue:44188|issue:46328", + "issue:42673|issue:45072", "issue:43611|issue:46002", "issue:44355|issue:46231", "issue:42371|issue:43874", "issue:43116|issue:43493", "issue:44038|issue:45478", "issue:44625|issue:46002", - "issue:42222|issue:43441", "issue:43976|issue:45446", + "issue:42222|issue:43441", "issue:44075|issue:44936", "issue:45357|issue:45478", "issue:43116|issue:43825", @@ -14373,225 +14381,210 @@ "issue:44380|issue:46077", "issue:44117|issue:45588", "issue:43827|issue:45446", - "issue:39401|issue:44779", - "issue:42673|issue:43906", - "issue:43673|issue:46328", - "issue:44466|issue:46002" + "issue:39401|issue:44779" ], - "prepared_review_unit_hash": "193848b0870522285908cb5af893fb063c5181e5a0ae30c1b7374a77038b309e", + "prepared_review_unit_hash": "fbcf486132484307c976f8a244ac7014896b73d96f5462ca4e614e50fb1ced09", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36757, - "estimated_input_tokens": 9190, - "estimated_eval_tokens": 18636 + "soft_pair_count": 29, + "serialized_chars": 36051, + "estimated_input_tokens": 9013, + "estimated_eval_tokens": 18282 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a broad false-positive mix of unrelated Transformers issues: tokenizer regressions, import/runtime breakages, model-specific bugs, training problems, and docs/perf nits. None of the soft pairs look like the same underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": "issue:45278 is the least-bad umbrella because it is the broadest v5 upgrade regression report and overlaps with several import/load failures, but the cluster is too heterogeneous for a true canonical duplicate.", + "summary": "These are mostly loosely similar regression reports, but none of the soft pairs look like the same underlying bug/change closely enough to merge. I rejected all candidate duplicate links.", + "confidence": 0.86, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most central/representative topic-wise, but only as a broad migration-breakage anchor; it does not actually subsume the rest of the mixed cluster cleanly.", + "best_issue_reason": "No single issue is a clear canonical for this set: the items span many unrelated bugs, regressions, and doc/API problems rather than one duplicate cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Both are import/loading failures, but one is a missing backend dependency and the other is a Python 3.14 startup failure; different causes and fixes." + }, { "left": "issue:43493", "right": "issue:45753", "accept": false, - "reason": "Different models and failures: SigLIP2 HF-vs-JAX discrepancy vs Qwen3_5 infinite loop on one image." + "reason": "Both mention Qwen3.5, but one is a model discrepancy vs JAX and the other is an infinite-loop image bug; not the same defect." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "ONNX import failure vs LayoutLMV3 error-message clarity; unrelated code paths." + "reason": "Unrelated: missing `transformers.onnx` module import vs a layoutlmv3 validation error message issue." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "Missing UMT5 encoder weight vs AutoProcessor kwargs forwarding; different APIs and bugs." + "reason": "Both involve config propagation/loading paths, but one is missing `embed_tokens.weight` in UMT5Encoder and the other is AutoProcessor kwargs not reaching `cached_file`." }, { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Both tokenizer-related, but one is decoder-type selection in v5 and the other is Deepseek output correctness; not the same bug." - }, - { - "left": "issue:42673", - "right": "issue:45072", - "accept": false, - "reason": "VRAM leak in Qwen3ForCausalLM threads vs dtype mismatch in CI inference; unrelated issues." + "reason": "Both are tokenizer regressions in v5, but they affect different code paths and symptoms: decoder type vs incorrect Deepseek tokenization." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "Same model family only at a high level; invalid GenerationConfig vs RoPE implementation discussion are different problems." + "reason": "Both concern GLM-5, but one is invalid GenerationConfig handling and the other is RoPE implementation discussion; not the same bug." }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "Specific missing symbol import vs broad post-upgrade import breakages; not the same concrete failure." + "reason": "Both are import errors after version changes, but the former is a specific missing symbol and the latter is a broad upgrade breakage report." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "torch.compile attention-kernel divergence vs SequenceFeatureExtractor.pad performance; unrelated subsystems." + "reason": "Different areas entirely: attention-kernel divergence under `torch.compile` vs a performance issue in `SequenceFeatureExtractor.pad`." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "Both are runtime issues, but one is VRAM leakage in `Qwen3ForCausalLM` and the other is dtype mismatches in CI inference for different models." }, { "left": "issue:43611", "right": "issue:46002", "accept": false, - "reason": "base_model_prefix loading regression vs NaNs in classification heads; different symptoms and code paths." + "reason": "Base-model-prefix loading regression vs NaNs in classification heads; different failure modes and likely different fixes." }, { "left": "issue:44355", "right": "issue:46231", "accept": false, - "reason": "Compiled Python file errors vs documentation parameter-name mismatch; unrelated." + "reason": "Compiled Python load errors are unrelated to a docs/API parameter-name mismatch in `set_training()`." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 API-warning behavior vs missing image-patch count method in GLM46V; unrelated." + "reason": "TF32 API settings warning is unrelated to a missing image-processor method breaking multimodal token counting." }, { "left": "issue:43116", "right": "issue:43493", "accept": false, - "reason": "Multi-label classification example failure vs SigLIP2 implementation discrepancy; no shared bug." + "reason": "Example-script multi-label output bug is unrelated to the SigLIP2 implementation discrepancy." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3 MoE, but one is Qwen3-VL-Moe and the other is Qwen3.5 from_pretrained; not enough to treat as one bug." + "reason": "Both mention Qwen3.5/Moe and Transformers v5, but one is a general bug report and the other is a specific from_pretrained error; not enough evidence of the same root cause." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "num_labels propagation issue vs NaNs after checkout/init; different failures." + "reason": "`num_labels` propagation in Qwen3.5 is unrelated to NaNs in classification heads." }, { - "left": "issue:42222", - "right": "issue:43441", + "left": "issue:43976", + "right": "issue:45446", "accept": false, - "reason": "VitPose model breakage vs Ministral FlashAttention bug; unrelated models and code paths." + "reason": "Python version support breakage and a PyTorch version check bug are different compatibility issues." }, { - "left": "issue:43976", - "right": "issue:45446", + "left": "issue:42222", + "right": "issue:43441", "accept": false, - "reason": "Python version compatibility problem vs incorrect PyTorch version guard in flex_attention; different version checks." + "reason": "Different model families and symptoms: vitpose model breakage vs Ministral FlashAttention failure." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "SGD args ignored vs trainer.evaluate failing after train; unrelated trainer/optimizer bugs." + "reason": "Optimizer SGD argument handling is not the same as `trainer.evaluate()` failing after training; different code paths." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Save_pretrained visual-key regression vs from_pretrained error in Qwen3.5 Moe; related model family but different concrete failures." + "reason": "Both are Qwen3.5 Moe regressions, but one is save_pretrained key serialization and the other is from_pretrained loading; distinct bugs." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "Classification example returning empty results vs pipeline translation-task error message; unrelated." + "reason": "Example script behavior is unrelated to pipeline() task-support error messaging." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Qwen3_omni_moe feature-length helper vs flex_attention PyTorch version check; different subsystems." + "reason": "Different subsystems: qwen3_omni_moe feature-length helper vs flex_attention import/version-check logic." }, { "left": "issue:42673", "right": "issue:42796", "accept": false, - "reason": "Qwen3 VRAM leak vs Ministral tokenizer BPE markers; unrelated." + "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to Ministral tokenizer BPE marker handling." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Training optimizer-state corruption vs missing GenerationMixin cache in chunked prefill; different paths." + "reason": "FSDP2/PEFT optimizer-state mismatch is unrelated to missing GenerationMixin cache during chunked prefill." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "Invalid GenerationConfig vs stale num_return_sequences warning; not the same defect." + "reason": "GenerationConfig validation and stale `num_return_sequences` warning are different generation issues." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Tokenizer-loading regression vs Qwen3 tokenizer offset mapping bug; both tokenizer issues but different concrete behaviors." + "reason": "Tokenizer loading regression is not the same as Qwen3 wrong offset mapping; both are tokenizer-related but distinct." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "GPT2 attention scaling ignored in backend dispatch vs Gemma4 causal-mask construction; different models and mechanisms." + "reason": "Both involve attention behavior, but GPT2 scaling under SDPA/FlashAttention is not the same as Gemma4 using causal masks for bidirectional attention." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES None-handling vs flash_attention s_aux=None crash; unrelated." + "reason": "Tokenizer mapping lookup failure is unrelated to a flash-attention crash on `s_aux=None`." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing removed pipeline() vs incorrect PyTorch version gate; unrelated." + "reason": "Docs still referencing removed pipeline() APIs is unrelated to a PyTorch version check bug." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Qwen3 offset_mapping bug vs Deepseek tokenizer regression; different model-specific tokenizer issues." - }, - { - "left": "issue:42673", - "right": "issue:43906", - "accept": false, - "reason": "VRAM leak in dataloader threads vs isolated reproduction of another issue; no evidence of same bug." - }, - { - "left": "issue:43673", - "right": "issue:46328", - "accept": false, - "reason": "Generation cache missing in chunked_prefill vs SequenceFeatureExtractor.pad performance; unrelated." - }, - { - "left": "issue:44466", - "right": "issue:46002", - "accept": false, - "reason": "Device-dependent lm_head.weight serialization vs NaNs in classification heads; different bug class." + "reason": "Both are tokenizer bugs, but one is offset mapping and the other is incorrect tokenization results in Deepseek; not the same issue." } ] }, @@ -14610,8 +14603,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -14620,6 +14613,7 @@ "issue:41628", "issue:42222", "issue:42371", + "issue:42490", "issue:42617", "issue:42673", "issue:42831", @@ -14627,16 +14621,17 @@ "issue:43122", "issue:43408", "issue:43450", - "issue:43582", + "issue:43673", "issue:43742", + "issue:43906", "issue:43937", "issue:43986", - "issue:44075", "issue:44117", "issue:44220", "issue:44261", "issue:44291", "issue:44410", + "issue:44466", "issue:44484", "issue:44496", "issue:44829", @@ -14653,9 +14648,8 @@ "issue:45405", "issue:45446", "issue:45478", - "issue:45542", "issue:45561", - "issue:45656", + "issue:46002", "issue:46178", "issue:46209", "issue:46231", @@ -14664,238 +14658,239 @@ "issue:46512" ], "soft_pairs": [ + "issue:42673|issue:43906", + "issue:43673|issue:46328", "issue:44291|issue:45071", + "issue:44466|issue:46002", "issue:43450|issue:43986", "issue:43937|issue:44977", "issue:45405|issue:45446", "issue:44410|issue:45310", + "issue:45341|issue:46328", "issue:45341|issue:45561", "issue:43122|issue:46298", - "issue:45341|issue:46328", "issue:38617|issue:41628", "issue:44117|issue:45081", - "issue:45071|issue:46512", "issue:42222|issue:46178", + "issue:45071|issue:46512", "issue:44829|issue:45216", "issue:45362|issue:46328", "issue:43408|issue:43742", "issue:42617|issue:45341", - "issue:42673|issue:43582", + "issue:42490|issue:42673", "issue:42371|issue:44261", "issue:44877|issue:46231", "issue:44484|issue:45310", "issue:39401|issue:46209", - "issue:38617|issue:44220", "issue:43937|issue:44496", + "issue:38617|issue:44220", "issue:44933|issue:45278", - "issue:43066|issue:43122", "issue:42371|issue:42831", - "issue:44410|issue:45478", - "issue:42222|issue:45542", - "issue:44075|issue:45656" + "issue:43066|issue:43122", + "issue:44410|issue:45478" ], - "prepared_review_unit_hash": "7a4e1d73d92ad56e771404c29a6d52a08951a96dbaa9450a63d23d44f3ec2e5d", + "prepared_review_unit_hash": "25ba51af11ca9d0438db341ccc0ffae3e257e65d585c2c9b6feacb22bca606c2", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35471, - "estimated_input_tokens": 8868, - "estimated_eval_tokens": 17992 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36439, + "estimated_input_tokens": 9110, + "estimated_eval_tokens": 18476 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is dominated by Transformers v5-era regressions, especially import/config/tokenizer loading issues, but the soft-pair candidates are mostly only subsystem-similar rather than the same concrete bug. I rejected all soft edges.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45278 is the broadest representative of the cluster\u2019s recurring upgrade/import breakages and best captures the shared theme across several narrow regressions.", + "summary": "This cluster is mostly heterogeneous: tokenizer, import/API, model-loading, training, and testing-utils reports that are not the same underlying bug. No soft-edge pair looks mergeable with high confidence; the clearest duplicate-like subcluster is the repeated Qwen3.5 MoE `from_pretrained` regression (45310/45478), though it is not one of the provided soft pairs.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45310 is the best canonical anchor for the only clear duplicate subcluster: it is the earlier, concrete report of the Qwen3.5 MoE `from_pretrained` error, and 45478 appears to be the later twin report with the same title.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it is the most general report and most likely to subsume related import failures after the 4.57.x -> 5.x transition.", + "best_issue_reason": "Issue 45278 is the broadest triage anchor for the cluster as a whole because it captures the general import-regression theme after upgrading, even though it does not truly subsume the unrelated model- and test-specific bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "Both mention Qwen3 and threading/repro context, but one is a VRAM leak in dataloader threads and the other is an isolated reproduction of a different issue; not the same bug." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Chunked-prefill cache handling and feature-extractor padding performance are unrelated code paths and symptoms." + }, { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "Both are v5 loading/config regressions, but one is an init_empty_weights argument error and the other is a PretrainedConfig type-checking issue; different code paths." + "reason": "Both are v5-era regressions, but one is an init-empty-weights argument error and the other is a PreTrainedConfig type-checking break; different failures." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "One is tied-weights serialization/device-dependent state, the other is NaNs during checkout+init; not the same underlying defect." }, { "left": "issue:43450", "right": "issue:43986", "accept": false, - "reason": "Both involve video processing, but one is an output-shape bug and the other is a torchvision-missing crash in AutoProcessor; not the same defect." + "reason": "Both involve video models, but one is batched video processor shape output and the other is AutoProcessor crashing without torchvision; distinct issues." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "Different models and symptoms: GLM-5 invalid GenerationConfig vs Qwen3.5 flash-attention generation failure." + "reason": "GLM-5 invalid GenerationConfig and Qwen3.5 flash-attention generation failures are different model-specific bugs." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "Unrelated concerns: PEFT version pinning versus a PyTorch version check for flex_attention import." + "reason": "A PEFT minimum-version bump and a PyTorch version guard for AuxRequest are separate dependency/version checks." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Both are Qwen-family loading problems, but one is missing projections in qwen3next and the other is a Qwen3.5 MoE from_pretrained failure; not the same bug." + "reason": "Qwen3Next missing projections is a model-architecture loading issue, while Qwen3.5 MoE `from_pretrained` is a different model family regression." }, { "left": "issue:45341", - "right": "issue:45561", + "right": "issue:46328", "accept": false, - "reason": "Both touch testing_utils, but one is a generic small bug and the other is a pytest-xdist race on captured_info.txt; different issues." + "reason": "Testing utility bug and feature-extractor padding performance bug are unrelated." }, { - "left": "issue:43122", - "right": "issue:46298", + "left": "issue:45341", + "right": "issue:45561", "accept": false, - "reason": "Different topics entirely: tokenizer behavior regression versus a noisy logging message." + "reason": "Both touch testing utils, but one is a generic bug report and the other is a specific xdist race on `captured_info.txt`; different problems." }, { - "left": "issue:45341", - "right": "issue:46328", + "left": "issue:43122", + "right": "issue:46298", "accept": false, - "reason": "Testing_utils bug versus SequenceFeatureExtractor.pad inefficiency; no shared underlying defect." + "reason": "Tokenizer behavior regression and noisy logging are unrelated." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols from different modules and do not look like the same code-path failure." + "reason": "Both are import errors, but they concern different missing symbols from different APIs and do not indicate the same bug." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer-related, but one is TOKENIZER_MAPPING_NAMES returning None and the other is a Mistral regex patch crash; distinct causes." + "reason": "Tokenizer mapping fallback behavior and Mistral regex patching/backend_tokenizer crash are different tokenizer-loading issues." }, { - "left": "issue:45071", - "right": "issue:46512", + "left": "issue:42222", + "right": "issue:46178", "accept": false, - "reason": "One is a type-checking break in PretrainedConfig, the other is stale dtype metadata after from_config; not the same bug." + "reason": "VitPose breakage and a tensorflow_text backend-mapping error are separate subsystem failures." }, { - "left": "issue:42222", - "right": "issue:46178", + "left": "issue:45071", + "right": "issue:46512", "accept": false, - "reason": "Vitpose model breakage and a tensorflow_text backend mapping error are unrelated." + "reason": "One is a PreTrainedConfig type-checking regression, the other is stale `config.dtype` after `from_config(dtype=...)`; related theme, but not the same bug." }, { "left": "issue:44829", "right": "issue:45216", "accept": false, - "reason": "Flash-attention training degeneration and Qwen3.5 save_pretrained checkpoint corruption are different failures." + "reason": "Flash-attention training degeneration and incorrect `save_pretrained` checkpoint serialization are different Qwen3.5-related problems." }, { "left": "issue:45362", "right": "issue:46328", "accept": false, - "reason": "A Qwen3.5 chat crash is unrelated to a SequenceFeatureExtractor.pad performance issue." + "reason": "Chat/model crash for Qwen3.5-35B and a feature-extractor padding performance issue are unrelated." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "Different model families and symptoms: sam3_video/sam3_tracker warning versus MobileLLM loading KeyError." - }, - { - "left": "issue:42617", - "right": "issue:45341", - "accept": false, - "reason": "3d_parallel.py execution failure is unrelated to the testing_utils bug." - }, - { - "left": "issue:42673", - "right": "issue:43582", - "accept": false, - "reason": "A VRAM leak in Qwen3ForCausalLM and an Apple Silicon TypeError in caching_allocator_warmup are different problems." + "reason": "SAM3 model-type warning and MobileLLM key error are different loading/configuration bugs." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API guidance and MLA q_a_layernorm precision/config error are unrelated." + "reason": "TF32 API guidance and MLA `rms_norm_eps` precision error are unrelated." }, { "left": "issue:44877", "right": "issue:46231", "accept": false, - "reason": "Strict config loading for granite_speech versus a documentation parameter-name mismatch; not the same bug." + "reason": "Strict config loading for granite_speech and a documentation parameter-name mismatch are not the same issue." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "A missing import from `image_utils` may be part of broader upgrade fallout, but this is not the same concrete bug as the general 'many import errors' report." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "Max shard size behavior and Qwen3.5 MoE from_pretrained failure are unrelated." + "reason": "`max_shard_size` question and Qwen3.5 MoE `from_pretrained` failure are unrelated." }, { "left": "issue:39401", "right": "issue:46209", "accept": false, - "reason": "Tokenizer offset mapping bug versus FP8 inference failure on B200; different subsystems and symptoms." - }, - { - "left": "issue:38617", - "right": "issue:44220", - "accept": false, - "reason": "ImportError for layer_type_validation and a feature-extraction function issue are not the same underlying problem." + "reason": "Qwen3 tokenizer offset mapping and FP8 inference failure on B200 are unrelated." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "GenerationConfig invalid for GLM-5 is unrelated to an unrecognized-model config.json error for Olmo-Hybrid." - }, - { - "left": "issue:44933", - "right": "issue:45278", - "accept": false, - "reason": "Both mention import problems, but one is a specific image_utils import and the other is a broad report of many import errors after upgrade; too general to treat as the same bug." + "reason": "Invalid GenerationConfig for GLM-5 and unrecognized model/model_type config loading are different classes of loading errors." }, { - "left": "issue:43066", - "right": "issue:43122", + "left": "issue:38617", + "right": "issue:44220", "accept": false, - "reason": "Tokenizer decoder type regression and differing tokenization outputs are related in theme but not the same concrete failure." + "reason": "ImportError for `layer_type_validation` and a feature-extractor helper bug are unrelated." }, { - "left": "issue:42371", - "right": "issue:42831", + "left": "issue:44933", + "right": "issue:45278", "accept": false, - "reason": "TF32 behavior settings and FineGrainedFP8 accuracy issues are unrelated." + "reason": "General import errors after upgrade are broader than, and not identical to, the specific missing import from `image_utils`." }, { - "left": "issue:44410", - "right": "issue:45478", + "left": "issue:42371", + "right": "issue:42831", "accept": false, - "reason": "Although both are Qwen3.5/Qwen3-related, they describe different loading failures in different model variants." + "reason": "TF32 settings API migration and FineGrainedFP8 accuracy regression are unrelated." }, { - "left": "issue:42222", - "right": "issue:45542", + "left": "issue:43066", + "right": "issue:43122", "accept": false, - "reason": "Vitpose model breakage and missing TensorFlow backend handling are unrelated." + "reason": "Wrong tokenizer decoder type and tokenization differences across versions are distinct tokenizer regressions." }, { - "left": "issue:44075", - "right": "issue:45656", + "left": "issue:44410", + "right": "issue:45478", "accept": false, - "reason": "Optimizer SGD args ignored and optimizer step called twice with deepspeed are different optimizer bugs." + "reason": "Qwen3Next projection mismatch and Qwen3.5 MoE `from_pretrained` error are different model-loading failures." } ] }, @@ -14914,8 +14909,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -14938,6 +14933,7 @@ "issue:43931", "issue:43986", "issue:44038", + "issue:44075", "issue:44117", "issue:44261", "issue:44263", @@ -14955,25 +14951,26 @@ "issue:45071", "issue:45341", "issue:45372", - "issue:45375", "issue:45440", "issue:45464", "issue:45478", "issue:45542", "issue:45588", + "issue:45656", "issue:45753", - "issue:46050", "issue:46123", "issue:46172", "issue:46231", "issue:46369" ], "soft_pairs": [ - "issue:44483|issue:45464", + "issue:42222|issue:45542", + "issue:44075|issue:45656", "issue:42175|issue:42222", - "issue:43012|issue:43577", + "issue:44483|issue:45464", "issue:43097|issue:44493", "issue:43673|issue:45341", + "issue:43012|issue:43577", "issue:43874|issue:44261", "issue:44263|issue:44485", "issue:41628|issue:45071", @@ -14987,249 +14984,253 @@ "issue:43931|issue:44038", "issue:43097|issue:43452", "issue:45003|issue:45341", - "issue:45753|issue:46369", "issue:43986|issue:45341", + "issue:45753|issue:46369", "issue:44933|issue:46231", "issue:43874|issue:44263", "issue:43673|issue:46123", "issue:43986|issue:44484", "issue:44117|issue:44987", "issue:44261|issue:44485", + "issue:43531|issue:45341", "issue:45440|issue:46172", "issue:43986|issue:45372", - "issue:43531|issue:45341", - "issue:43097|issue:45588", "issue:43064|issue:44315", "issue:38617|issue:42757", - "issue:43097|issue:46050", - "issue:44410|issue:45375" + "issue:43097|issue:45588" ], - "prepared_review_unit_hash": "6b0f39666913e904e4732193566a6427b574314197efaf8503c5d855af199766", + "prepared_review_unit_hash": "ccdb07e0fbc89448eaa3fb61eed4c3559a0155469f01c4e9376fc63ff8f503e6", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37684, - "estimated_input_tokens": 9421, - "estimated_eval_tokens": 19098 + "serialized_chars": 37602, + "estimated_input_tokens": 9401, + "estimated_eval_tokens": 19058 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is a loose collection of mostly unrelated Transformers issues, and the soft pairs are best treated as false positive similarities rather than duplicates.", - "confidence": 0.93, - "canonical_issue_reason": "Issue 43097 is the best weak canonical because it is one of the most broadly scoped, discussion-heavy v5 regressions in the set and has inbound references; however, the overall group is too heterogeneous for a strong canonical issue.", + "summary": "This is a loose similarity cluster rather than a true duplicate set. The issues span unrelated import errors, model-specific loading bugs, optimizer/runtime regressions, and docs/test nits, so I rejected every soft edge as not the same underlying bug or change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43097 is the clearest representative of the set: it has a specific regression claim, clear version context, and broader downstream impact than the more model-specific or incidental reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43097 is the closest thing to a representative issue in this cluster because it concerns a high-level, user-visible v5 breaking change rather than a narrow model-specific edge case.", + "best_issue_reason": "issue:43097 is the best single issue to anchor the cluster because it is well-scoped, user-facing, and described as a concrete v5 regression with some follow-up traffic.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44483", - "right": "issue:45464", + "left": "issue:42222", + "right": "issue:45542", "accept": false, - "reason": "Both are chat/completions failures, but one is a v5 API acceptance problem and the other is a streaming inference failure on Qwen3.5; different concrete bugs." + "reason": "Both mention TensorFlow/backend behavior, but one is about broken vitpose models and the other about installing torch-only extras; not the same code-path bug." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Different optimizer problems: SGD args ignored vs deepspeed causing a double optimizer step." }, { "left": "issue:42175", "right": "issue:42222", "accept": false, - "reason": "One is missing TensorFlow from the torch extra; the other is a broken VitPose model path. Different subsystems and symptoms." + "reason": "Package/backend installation issue versus model breakage; too broad and not the same failure." }, { - "left": "issue:43012", - "right": "issue:43577", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "Both involve dtype/precision, but one is a PyTorch warning during bfloat16 compilation and the other is BLIP2 loading leaving submodules in float32." + "reason": "Both hit chat/completions, but one is a general API rejection and the other is a specific streaming inference failure on Qwen3.5-0.8B." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "One is an explicit API removal warning around tie_embeddings_and_encoder_decoder; the other is unexpected position-id keys during model loading. Not the same bug." + "reason": "One is a removed config method in v5.0.0; the other is unexpected position-id keys across many models. Different regressions." }, { "left": "issue:43673", "right": "issue:45341", "accept": false, - "reason": "Chunked prefill cache handling and a testing_utils bug are unrelated code paths." + "reason": "Chunked-prefill cache handling and a testing_utils bug are unrelated." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Both involve dtype behavior, but one is a warning during bfloat16 compilation and the other is model components staying float32 when loading BLIP2." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "A missing image-patch method in GLM46V image processing is unrelated to an MLA layer norm epsilon/config mismatch." + "reason": "Different GLM-family bugs: missing image-patch method vs missing rms_norm_eps in layernorm config." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "Both mention GLM, but one is a torch.split/indexing bug and the other is a RoPE implementation issue; different fixes." + "reason": "A torch.split return issue in a DSA indexer is not the same as a RoPE implementation bug." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "An AutoImageProcessor import failure is unrelated to PretrainedConfig type checking breakage." + "reason": "Importing AutoImageProcessor and PreTrainedConfig type checking are unrelated failures." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "Type-checking regression and undefined tf backend error are different failures with different triggers." + "reason": "Type-checking breakage in config classes is unrelated to a TensorFlow backend installation problem." }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "Both mention Flash Attention 2, but they affect different model classes and likely different integration points." + "reason": "Both mention Flash Attention 2, but one is XLM-Roberta support and the other is NemotronHForCausalLM support; different model paths." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "max_shard_size default behavior and Qwen3.5-Moe from_pretrained errors are unrelated." + "reason": "Save-pretrained shard size behavior is unrelated to Qwen3.5 MoE from_pretrained loading." }, { "left": "issue:44263", "right": "issue:46172", "accept": false, - "reason": "Different model families and different math/indexing bugs." + "reason": "Different DeepSeek/GLM internals and different bugs; no shared code-path evidence." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "A missing ReLU in the DSA indexer is unrelated to a Deepseek transpose/dtype issue." + "reason": "A missing ReLU in the DSA indexer is unrelated to the deepseek-v4 dtype/transpose issue." }, { "left": "issue:43931", "right": "issue:44410", "accept": false, - "reason": "Both are Qwen-family issues, but one is a weight-shape mismatch and the other is missing projections in a specific layer." + "reason": "Both are Qwen-family loading issues, but one is a weight-shape mismatch and the other is missing projections in layer 0." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "One is a FlashAttention compatibility failure for Ministral-3; the other is a flash_attention_3 training degeneration bug. Not mergeable into one fix." + "reason": "Different FlashAttention problems: one is model support in v5 RC, the other is degenerate training with flash_attention_3." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "A Qwen3-VL model loading mismatch and a Qwen3-VL-Moe bug are different concrete problems." + "reason": "Different Qwen-VL/Qwen3-VL-MoE loading issues with no concrete overlap in failure mode." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "Tie-embeddings API removal and gguf_file tokenizer/model loading failures are unrelated." + "reason": "Config method removal and gguf_file handling are unrelated." }, { "left": "issue:45003", "right": "issue:45341", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to a testing_utils issue." + "reason": "Unsafe sys.modules access in modeling_utils is not the same as a testing_utils bug." }, { - "left": "issue:45753", - "right": "issue:46369", + "left": "issue:43986", + "right": "issue:45341", "accept": false, - "reason": "An infinite loop on a specific image and a memory overflow in sam3-litetext are different failure modes and models." + "reason": "Missing torchvision during AutoProcessor loading is unrelated to the testing_utils issue." }, { - "left": "issue:43986", - "right": "issue:45341", + "left": "issue:45753", + "right": "issue:46369", "accept": false, - "reason": "Video AutoProcessor loading without torchvision and a testing_utils bug do not describe the same bug." + "reason": "An infinite loop on a specific image is not the same as a memory overflow in sam3-litetext." }, { "left": "issue:44933", "right": "issue:46231", "accept": false, - "reason": "A nonexistent image_utils import and a documentation parameter mismatch are not the same issue." + "reason": "A bad import from image_utils and a docs/parameter-name mismatch are unrelated." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different GLM components with different runtime failures." + "reason": "Different GLM bugs with different failing methods and symptoms." }, { "left": "issue:43673", "right": "issue:46123", "accept": false, - "reason": "Generation cache missing in chunked prefill is unrelated to MaskGenerationPipeline final-batch handling." + "reason": "Generation cache missing in chunked_prefill and MaskGenerationPipeline dropping partial-batch results are different pipeline bugs." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "A torchvision dependency crash and a shard-size question are unrelated." + "reason": "AutoProcessor without torchvision and max shard size in save_pretrained have no common underlying bug." }, { "left": "issue:44117", "right": "issue:44987", "accept": false, - "reason": "Tokenizer mapping returning None is broader import/loading metadata behavior, while physical-intelligence/fast failure is a specific model-loading regression; too weak to merge." + "reason": "Tokenizer mapping returning None may be related in theme, but the reported failure modes are not clearly the same bug." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "Different GLM bugs: config epsilon precision mismatch versus RoPE implementation." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." }, { "left": "issue:45440", "right": "issue:46172", "accept": false, - "reason": "Both touch DeepSeek models, but they are distinct implementation mismatches with different code paths." + "reason": "DeepSeekV3MoE divergence from remote impl and a deepseek-v4 comb dtype/transpose issue are different problems." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Missing torchvision for video processors and missing ReasoningEffort import for Gemma 4 are separate dependency/import issues." - }, - { - "left": "issue:43531", - "right": "issue:45341", - "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." + "reason": "Missing torchvision for video AutoProcessor loading is unrelated to the ReasoningEffort import error in Gemma 4 processor loading." }, { - "left": "issue:43097", - "right": "issue:45588", + "left": "issue:43064", + "right": "issue:44315", "accept": false, - "reason": "Tie-embeddings removal and a flash_attention.py s_aux=None crash are different problems." + "reason": "Wrong optimizer states on nonzero ranks with FSDP2/PEFT is unrelated to Liger Kernel not being applied in model_init." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are ImportErrors, but they involve different missing symbols from different modules and are not the same underlying bug." + "reason": "Both are import errors from upstream packages, but they affect different missing symbols and different dependencies." }, { "left": "issue:43097", - "right": "issue:46050", - "accept": false, - "reason": "A breaking API removal and a quantization save/load regression are unrelated." - }, - { - "left": "issue:44410", - "right": "issue:45375", + "right": "issue:45588", "accept": false, - "reason": "Missing projections in a Qwen3Next layer and a dropped config field in Qwen3_5MoeVisionConfig are different issues, even if both are model-structure related." + "reason": "Removed tie_embeddings method and flash_attention.py crashing on s_aux=None are unrelated failures." } ] }, @@ -15248,14 +15249,12 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ - "issue:36683", "issue:38617", - "issue:42175", "issue:42890", "issue:42907", "issue:43012", @@ -15266,7 +15265,6 @@ "issue:43352", "issue:43502", "issue:43550", - "issue:43582", "issue:43824", "issue:43825", "issue:43873", @@ -15290,11 +15288,13 @@ "issue:44936", "issue:44938", "issue:45103", + "issue:45375", "issue:45478", "issue:45479", "issue:45584", "issue:45712", "issue:45753", + "issue:46050", "issue:46132", "issue:46172", "issue:46178", @@ -15303,11 +15303,13 @@ "issue:46421" ], "soft_pairs": [ + "issue:43097|issue:46050", "issue:43502|issue:43992", + "issue:44410|issue:45375", "issue:44038|issue:44410", - "issue:42890|issue:46298", "issue:38617|issue:44351", "issue:44062|issue:44117", + "issue:42890|issue:46298", "issue:43097|issue:43873", "issue:44936|issue:46365", "issue:44704|issue:45103", @@ -15318,12 +15320,12 @@ "issue:43992|issue:44484", "issue:43208|issue:43550", "issue:43066|issue:43906", - "issue:43012|issue:43550", "issue:42907|issue:44655", "issue:43012|issue:44368", + "issue:43012|issue:43550", "issue:45712|issue:46298", - "issue:43825|issue:44655", "issue:42907|issue:45753", + "issue:43825|issue:44655", "issue:43097|issue:46298", "issue:43352|issue:44380", "issue:43352|issue:43976", @@ -15333,236 +15335,234 @@ "issue:44077|issue:46365", "issue:44380|issue:45479", "issue:43064|issue:44380", - "issue:44222|issue:45478", - "issue:36683|issue:43582", - "issue:42175|issue:46365" + "issue:44222|issue:45478" ], - "prepared_review_unit_hash": "fb45eb4ef6942f0ec75f741f8601795b8001a72e9886bbe3e9b056b4d6d88047", + "prepared_review_unit_hash": "49a1ec11dd89819d7c46d27d9edc917c4679d70c492407559534950f4a8ceb82", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 33, - "serialized_chars": 37401, - "estimated_input_tokens": 9351, - "estimated_eval_tokens": 18958 + "serialized_chars": 36859, + "estimated_input_tokens": 9215, + "estimated_eval_tokens": 18686 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are not a duplicate cluster; they span many unrelated regressions, model-specific bugs, and API/import failures. No soft-edge pair looks like the same underlying issue, so all candidate links should be rejected.", - "confidence": 0.93, - "canonical_issue_reason": "If a representative issue is needed, issue 43873 is the broadest and most discussion-rich bug report here, but it is not a true duplicate target for the rest of the set.", + "summary": "The set is heterogeneous: most pairs are only loosely similar by subsystem or model family, not the same underlying defect. I would not merge any soft edge as a duplicate based on the available evidence.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45478 is the most concrete, version-scoped runtime failure in the set; the rest are mostly distinct bugs or ancillary warnings, so it is the best single representative if one issue must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the strongest general representative because it has the most activity and describes a concrete runtime bug with broader user impact, though the collection is too heterogeneous for a real canonical duplicate.", + "best_issue_reason": "Issue 45478 is the most actionable and specific report here: it names a clear model family, a version boundary, and a failing from_pretrained path.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Different bugs: a removed config flag warning versus a quantized model save/load regression." + }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "Different failures: local_files_only network leakage vs missing embed_tokens weight in UMT5Encoder." + "reason": "Different failure modes: stray API requests with local_files_only versus a missing encoder weight on load." }, { - "left": "issue:44038", - "right": "issue:44410", + "left": "issue:44410", + "right": "issue:45375", "accept": false, - "reason": "Different model families and symptoms: Qwen3-VL-Moe version bug vs missing qkv/gate projections in qwen3next." + "reason": "Same broad Qwen family, but one is missing attention projections and the other is a config field being dropped; not the same bug." }, { - "left": "issue:42890", - "right": "issue:46298", + "left": "issue:44038", + "right": "issue:44410", "accept": false, - "reason": "Set_seed test flakiness is unrelated to a noisy logging message." + "reason": "Both are Qwen-related, but one is a generic v5/Qwen3-VL-Moe issue and the other is a specific missing-projection defect." }, { "left": "issue:38617", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they concern different symbols and different underlying API breakages." + "reason": "Separate import errors for different symbols; no shared code-path." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Different tokenizer bugs: AddedToken keyword collision vs TOKENIZER_MAPPING_NAMES returning None." + "reason": "Tokenizer construction conflict versus tokenizer mapping returning None; different root causes." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "Test flakiness from missing seeding is unrelated to a noisy logging message." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "Missing tie_embeddings_and_encoder_decoder API vs offloading/quantization behavior are unrelated." + "reason": "A v5 removal warning is not the same as a quantization/offloading behavior issue." }, { "left": "issue:44936", "right": "issue:46365", "accept": false, - "reason": "Trainer.evaluate() failing after train() is not the same as importing Trainer being blocked by datasets." + "reason": "Trainer evaluation failure after training is unrelated to importing Trainer failing because of datasets." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "AutoProcessor kwargs forwarding bug is separate from auto_docstring crashes under future annotations." + "reason": "Different subsystems: cached_file kwargs forwarding versus auto_docstring annotation handling." }, { "left": "issue:43986", "right": "issue:44117", "accept": false, - "reason": "Video-model AutoProcessor crash without torchvision is unrelated to tokenizer mapping None handling." + "reason": "Video-model loading without torchvision is unrelated to tokenizer mapping assumptions." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "GPT2 attention scaling under SDPA/FlashAttention is unrelated to cache lazy initialization recompilation." + "reason": "Backend attention scaling bug versus cache lazy-init recompilation issue; different code paths." }, { "left": "issue:43824", "right": "issue:43986", "accept": false, - "reason": "Missing Qwen2.5-VL export and video-model AutoProcessor crash are different import-path issues." + "reason": "Both are import/load problems, but they affect different objects and dependencies." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Python 3.14 load failure is unrelated to tensorflow_text being absent from BACKENDS_MAPPING." + "reason": "Python version load failure and tensorflow_text backend validation error are unrelated." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "Missing embed token weight in a model loader is not the same as save_pretrained shard-size behavior." + "reason": "Missing encoder weights versus max_shard_size semantics; not the same defect." }, { "left": "issue:43208", "right": "issue:43550", "accept": false, - "reason": "xLSTM training bugs and Bamba torch.compile SDPA failure are different model/runtime problems." + "reason": "xLSTM training blockers and Bamba torch.compile/SDPA failure are different model-specific bugs." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as the reproduced issue 38071." - }, - { - "left": "issue:43012", - "right": "issue:43550", - "accept": false, - "reason": "bfloat16 compile warning and Bamba SDPA compile failure are not the same code-path bug." + "reason": "Wrong tokenizer decoder type is not the same as an isolated reproduction report." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Dequantized model save failures and pipeline save_pretrained limitations are distinct save-path problems." + "reason": "Save/load of dequantized models versus saving Pipeline objects; different object types and failure modes." }, { "left": "issue:43012", "right": "issue:44368", "accept": false, - "reason": "A PyTorch precision warning is unrelated to tie_word_embeddings warning behavior in LoRA fine-tuning." + "reason": "Both are warning-related, but they concern different config fields and different behaviors." }, { - "left": "issue:45712", - "right": "issue:46298", + "left": "issue:43012", + "right": "issue:43550", "accept": false, - "reason": "Dummy class cleanup/check_repo leakage is unrelated to a generic noisy log message." + "reason": "A precision warning is unrelated to a compile-time SDPA failure." }, { - "left": "issue:43825", - "right": "issue:44655", + "left": "issue:45712", + "right": "issue:46298", "accept": false, - "reason": "pipeline translation support messaging is separate from inability to save Pipeline objects." + "reason": "Repository dummy-class cleanup is unrelated to logging noise." }, { "left": "issue:42907", "right": "issue:45753", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals is unrelated to Qwen3_5 infinite generation loops." + "reason": "Model save/load failure and a specific-image infinite loop are not the same bug." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline task-support messaging and pipeline serialization are distinct issues." }, { "left": "issue:43097", "right": "issue:46298", "accept": false, - "reason": "Removed tie_embeddings_and_encoder_decoder API is not a logging-noise issue." + "reason": "Removed flag behavior and noisy logging do not share a concrete code-path problem." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Flash Attention support for NemotronH is unrelated to GPT2 attention scaling under backend selection." + "reason": "Different attention-backend problems on different models; not mergeable as one fix." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "Flash Attention support for a model is unrelated to the Python version compatibility breakage." + "reason": "Flash-Attention support on one model is unrelated to Python-version compatibility." }, { "left": "issue:44162", "right": "issue:46172", "accept": false, - "reason": "ESM2 regression and deepseek-v4 comb.to(dtype) issue are different model-specific bugs." + "reason": "ESM2 breakage and a DeepSeek tensor-transpose issue are separate model bugs." }, { "left": "issue:45584", "right": "issue:46365", "accept": false, - "reason": "Whisper empty transcription failure is unrelated to Trainer import conflicts with datasets." + "reason": "Whisper generation edge case is unrelated to importing Trainer." }, { "left": "issue:44380", "right": "issue:46132", "accept": false, - "reason": "GPT2 backend scaling bug is different from AttentionInterface.register changing registered function behavior." + "reason": "Attention scaling config ignored versus register() changing function behavior; different APIs." }, { "left": "issue:44077", "right": "issue:46365", "accept": false, - "reason": "patchtsmixer post_init policy is unrelated to Trainer import failure caused by datasets." + "reason": "PatchTSMixer post_init validation and Trainer import failure have no shared root cause." }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Attention scaling under SDPA/FlashAttention is different from degenerate zero-loss sequence-classification behavior." + "reason": "Attention backend scaling bug and degenerate classification loss are unrelated." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "FSDP2 optimizer-state corruption and GPT2 attention scaling are separate training/runtime issues." + "reason": "FSDP2/PEFT optimizer-state corruption is not the same as attention scaling behavior under SDPA/FlashAttention." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "FP8 save_pretrained moe and Qwen3.5 Moe from_pretrained error are related to different save/load paths." - }, - { - "left": "issue:36683", - "right": "issue:43582", - "accept": false, - "reason": "Gemma3Config missing vocab_size and Apple Silicon caching_allocator_warmup TypeError are unrelated." - }, - { - "left": "issue:42175", - "right": "issue:46365", - "accept": false, - "reason": "TensorFlow missing from a torch install is unrelated to Trainer import being affected by datasets." + "reason": "Both mention MoE/quantization context, but one is FP8 save_pretrained while the other is a Qwen3.5 from_pretrained error; not the same concrete bug." } ] }, @@ -15581,14 +15581,15 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ "issue:36010", "issue:36683", "issue:38617", + "issue:42175", "issue:42490", "issue:42757", "issue:42898", @@ -15600,15 +15601,14 @@ "issue:43352", "issue:43475", "issue:43502", - "issue:43525", "issue:43526", + "issue:43582", "issue:43611", "issue:43618", "issue:43756", "issue:43761", "issue:43824", "issue:43867", - "issue:43901", "issue:43937", "issue:43976", "issue:44038", @@ -15628,6 +15628,7 @@ "issue:45593", "issue:45600", "issue:45698", + "issue:46050", "issue:46077", "issue:46082", "issue:46172", @@ -15635,259 +15636,252 @@ "issue:46365" ], "soft_pairs": [ + "issue:36683|issue:43582", + "issue:42175|issue:46365", "issue:42490|issue:43012", - "issue:44483|issue:44514", - "issue:43012|issue:44683", "issue:38617|issue:43502", "issue:43208|issue:43526", + "issue:43012|issue:44683", + "issue:44483|issue:44514", "issue:42757|issue:43824", "issue:45600|issue:46298", "issue:45698|issue:46172", + "issue:44038|issue:44230", "issue:43618|issue:44380", "issue:43618|issue:43761", - "issue:44038|issue:44230", "issue:44483|issue:45290", "issue:43299|issue:43867", "issue:42994|issue:44655", - "issue:36683|issue:43475", "issue:42898|issue:43066", "issue:44625|issue:46365", + "issue:36683|issue:43475", "issue:43756|issue:44485", "issue:45276|issue:46077", - "issue:43352|issue:43937", "issue:36010|issue:42757", + "issue:43352|issue:43937", "issue:43611|issue:44655", - "issue:43901|issue:44655", + "issue:44655|issue:46050", "issue:36683|issue:44077", "issue:43867|issue:45216", "issue:43976|issue:44938", "issue:44625|issue:46077", "issue:43352|issue:45593", - "issue:46077|issue:46082", - "issue:36683|issue:43525", - "issue:43618|issue:45216", - "issue:36683|issue:44625" + "issue:46077|issue:46082" ], - "prepared_review_unit_hash": "6ea1d2e402091d3e1071351a31720109868e0d43e3cabf899d28a0eed199a645", + "prepared_review_unit_hash": "736308b63d9a42ca755562993f9e91399fb1bfd732380bc16df8071df455eff8", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 32, - "serialized_chars": 36268, - "estimated_input_tokens": 9067, - "estimated_eval_tokens": 18390 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36531, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: it mixes unrelated import errors, model-specific attribute/config bugs, serialization issues, attention/backend regressions, and a few docs/logging items. None of the soft pairs look like the same underlying bug/change, so they should stay separate.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is mostly a heterogeneous mix of unrelated issues; the soft-similarity links do not look like true duplicates, so I rejected all of them. The most representative issue is the high-activity save/load regression report.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45698 is the most active and broadly framed bug report in the set, with detailed repros and inbound references around a save_pretrained/from_pretrained regression.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45698 is the best global representative because it has the strongest discussion signal and describes a concrete, reusable serialization/loading failure mode.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42490", - "right": "issue:43012", + "left": "issue:36683", + "right": "issue:43582", "accept": false, - "reason": "Different problems: local-path slash import warning vs bfloat16 compile warning." + "reason": "Different failures: Gemma3 config missing vocab_size vs Apple Silicon TypeError in caching_allocator_warmup." }, { - "left": "issue:44483", - "right": "issue:44514", + "left": "issue:42175", + "right": "issue:46365", "accept": false, - "reason": "API request handling failure and batched chat-template crash are distinct code paths." + "reason": "Both mention dependency/import problems, but one is torch backend packaging and the other is Trainer import conflict with datasets." }, { - "left": "issue:43012", - "right": "issue:44683", + "left": "issue:42490", + "right": "issue:43012", "accept": false, - "reason": "Both involve compile-time warnings/failures, but one is a warning and the other is a flex_attention compile break." + "reason": "Unrelated warnings: local path slash import warning vs bfloat16 compile warning." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "Different import/API behaviors: missing symbol import vs network calls despite local_files_only." + "reason": "Different code paths: missing configuration_utils symbol vs local_files_only still making API requests." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "Unrelated model-training bugs vs an image processor label-reduction bug." + "reason": "Distinct model bugs in different components (xLSTM training issues vs BeitImageProcessorFast label reduction)." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "One is a PyTorch warning, the other is a flex_attention compile failure on newer torch." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Different surfaces: /v1/chat/completions request rejection vs Qwen2_5_VLProcessor batched chat template crash." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but for different symbols/modules and different root causes." + "reason": "Both are import errors, but they are for different missing symbols from different modules." }, { "left": "issue:45600", "right": "issue:46298", "accept": false, - "reason": "CI break from removed config references is unrelated to a noisy logging message." + "reason": "CI break from removed Sam3LiteText config refs is unrelated to a noisy logging message." }, { "left": "issue:45698", "right": "issue:46172", "accept": false, - "reason": "Save/load custom module mismatch is unrelated to the deepseek-v4 dtype/transpose bug." + "reason": "Serialization/loading regression vs a DeepSeek-V4 dtype/transpose bug; not the same issue." }, { - "left": "issue:43618", - "right": "issue:44380", + "left": "issue:44038", + "right": "issue:44230", "accept": false, - "reason": "Missing CLIP attentions and GPT2 attention-scaling being ignored are different regressions." + "reason": "Same model family, but one is a general Qwen3-VL-Moe bug and the other is fp8 inference support." }, { "left": "issue:43618", - "right": "issue:43761", + "right": "issue:44380", "accept": false, - "reason": "Different outputs are missing: attentions vs hidden_states." + "reason": "CLIP attentions missing is unrelated to GPT2 attention scaling being ignored under SDPA/FlashAttention." }, { - "left": "issue:44038", - "right": "issue:44230", + "left": "issue:43618", + "right": "issue:43761", "accept": false, - "reason": "Same model family, but one is a concrete loading bug and the other is fp8 support work." + "reason": "Related CLIP regressions, but different outputs and likely separate forward-path bugs." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "Chat-completions request rejection and assistant tool-call tokenization crash are unrelated." + "reason": "Chat-completions endpoint validation issue vs assistant tool-call template crash; different failure modes." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Qwen3VL Moe loading failure and sorted state_dict loading error are distinct bugs." + "reason": "Both are load/save problems, but one is Qwen3.5 checkpoint regression and the other is a sorted-state_dict loading error." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Quantized model saving failure and Pipeline save_pretrained failure hit different save paths." - }, - { - "left": "issue:36683", - "right": "issue:43475", - "accept": false, - "reason": "Missing vocab_size on Gemma3Config is unrelated to missing fpn_position_embeddings on SAM3 output." + "reason": "Quantized model saving failure is not the same as Pipeline.save_pretrained failing." }, { "left": "issue:42898", "right": "issue:43066", "accept": false, - "reason": "Tokenization-space cleanup behavior and tokenizer decoder type are different v5 changes." + "reason": "Tokenizer behavior change and wrong decoder type are both v5 tokenizer issues, but not the same bug." }, { "left": "issue:44625", "right": "issue:46365", "accept": false, - "reason": "Qwen3.5 num_labels propagation and Trainer import being blocked by datasets are unrelated." + "reason": "Qwen3.5 num_labels propagation bug is unrelated to the Trainer import conflict." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Different model-specific attribute errors in Gemma3 config and SAM 3 Video output classes." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Both mention RoPE, but they concern different models and different underlying issues." + "reason": "Both mention RoPE, but one is a Smollm3 architecture mismatch and the other is a GLM-5 implementation issue." }, { "left": "issue:45276", "right": "issue:46077", "accept": false, - "reason": "Embedding resize propagation and bidirectional attention masking are unrelated Gemma4 bugs." + "reason": "Same family (Gemma4), but resize_token_embeddings and bidirectional attention are separate code paths." }, { - "left": "issue:43352", - "right": "issue:43937", + "left": "issue:36010", + "right": "issue:42757", "accept": false, - "reason": "FlashAttention support for Nemotron and invalid GenerationConfig for GLM-5 are unrelated." + "reason": "Import errors in different modules: GenerationMixin from transformers.generation vs is_offline_mode from huggingface_hub." }, { - "left": "issue:36010", - "right": "issue:42757", + "left": "issue:43352", + "right": "issue:43937", "accept": false, - "reason": "Both are import errors, but for different packages and symbols." + "reason": "Nemotron Flash Attention support and GLM-5 GenerationConfig validation are unrelated." }, { "left": "issue:43611", "right": "issue:44655", "accept": false, - "reason": "base_model_prefix loading regression is unrelated to Pipeline save_pretrained failure." + "reason": "Model loading with base_model_prefix is distinct from saving Pipeline objects." }, { - "left": "issue:43901", - "right": "issue:44655", + "left": "issue:44655", + "right": "issue:46050", "accept": false, - "reason": "Docs mismatch for return_all_scores is not the same as broken Pipeline serialization." + "reason": "Both involve save_pretrained/from_pretrained, but one is Pipeline persistence and the other is quantized sinq round-trip." }, { "left": "issue:36683", "right": "issue:44077", "accept": false, - "reason": "Missing Gemma3Config attribute and patchtsmixer post_init validation are unrelated." + "reason": "No meaningful overlap: Gemma3 config attribute error vs patchtsmixer post_init validation." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Both concern loading/saving, but one is sorted state_dict loading and the other is incorrect saved checkpoints." + "reason": "Both are checkpoint/loading regressions, but they affect different mechanisms and models." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Both are Python-version compatibility issues, but for different supported-version ranges and failures." + "reason": "Separate Python-version compatibility problems for different releases." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "Qwen3.5 label propagation and Gemma4 attention-mask behavior are unrelated." + "reason": "Different Gemma4/Qwen3.5 bugs with no shared failing code path." }, { "left": "issue:43352", "right": "issue:45593", "accept": false, - "reason": "FlashAttention support and D-FINE auxiliary-loss logic are different subsystems." + "reason": "Nemotron Flash Attention support is unrelated to D-FINE auxiliary losses." }, { "left": "issue:46077", "right": "issue:46082", "accept": false, - "reason": "Different config-level behaviors: attention mask construction vs head_dim validation." - }, - { - "left": "issue:36683", - "right": "issue:43525", - "accept": false, - "reason": "Different missing attributes on different configs and model families." - }, - { - "left": "issue:43618", - "right": "issue:45216", - "accept": false, - "reason": "CLIP attentions regression and Qwen3.5 checkpoint corruption are unrelated." - }, - { - "left": "issue:36683", - "right": "issue:44625", - "accept": false, - "reason": "Missing Gemma3Config vocab_size and Qwen3.5 num_labels propagation are unrelated config bugs." + "reason": "Different model/config issues: Gemma4 bidirectional attention vs LlamaConfig head_dim validation." } ] }, @@ -15906,8 +15900,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -15921,8 +15915,8 @@ "issue:43064", "issue:43065", "issue:43502", + "issue:43525", "issue:43575", - "issue:43576", "issue:43618", "issue:43723", "issue:43867", @@ -15937,15 +15931,16 @@ "issue:44355", "issue:44360", "issue:44387", - "issue:44485", "issue:44492", "issue:44556", + "issue:44625", "issue:44855", "issue:44857", "issue:44938", "issue:44964", "issue:45071", "issue:45125", + "issue:45216", "issue:45278", "issue:45310", "issue:45357", @@ -15955,12 +15950,14 @@ "issue:45561", "issue:45600", "issue:45850", - "issue:46132", "issue:46202", "issue:46285", "issue:46365" ], "soft_pairs": [ + "issue:36683|issue:43525", + "issue:43618|issue:45216", + "issue:36683|issue:44625", "issue:43992|issue:45310", "issue:43064|issue:44492", "issue:43010|issue:45278", @@ -15975,8 +15972,8 @@ "issue:43010|issue:45850", "issue:43992|issue:45478", "issue:43867|issue:45357", - "issue:36296|issue:43575", "issue:43010|issue:43976", + "issue:36296|issue:43575", "issue:43618|issue:44077", "issue:28282|issue:33453", "issue:43618|issue:46285", @@ -15989,241 +15986,235 @@ "issue:43723|issue:44556", "issue:44387|issue:44556", "issue:44222|issue:45125", - "issue:43010|issue:43065", - "issue:43576|issue:46132", - "issue:43502|issue:44485", - "issue:45125|issue:45478" + "issue:43010|issue:43065" ], - "prepared_review_unit_hash": "6a7153fd36c3f7fdbe1b8df79c3336dc23879f50b4e883a929fd558203eef9e7", + "prepared_review_unit_hash": "077aa8ca16956a1997f861fb4a558384e1e59d43b904c3a602837026427254b0", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 36741, - "estimated_input_tokens": 9186, - "estimated_eval_tokens": 18628 + "serialized_chars": 36821, + "estimated_input_tokens": 9206, + "estimated_eval_tokens": 18668 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are only thematically related v5 regression reports, not true duplicates. The only clear duplicate-style match is the import-time slowdown pair around removing `import *` usage.", - "confidence": 0.66, - "canonical_issue_reason": "issue:45278 is the broadest concrete regression report here: it captures multiple import/load failures after upgrading from 4.57.0 to 5.5.0 and works as the best umbrella issue for the cluster.", + "summary": "This set is mostly a loose thematic cluster of unrelated Transformer regressions and CI issues. A few items share broad language around import/loading/version-upgrade problems, but none of the listed soft pairs look like true duplicates of the same concrete bug path.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 45278 is the broadest representative of the cluster\u2019s recurring theme: import/load regressions after upgrading Transformers. It is still not a true duplicate umbrella, but it is the most central summary issue among the set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most useful representative because it is specific enough to describe the upgrade regression while still broad enough to cover closely related import/load breakages.", + "best_issue_reason": "Issue 45278 best fits as the global representative because it captures the widest version-upgrade regression surface and is the closest to the common thread across the issues, even though the cluster is not actually duplicate-heavy.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Different model configs and different missing attributes; both are config errors, but not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "One is about CLIPOutput attentions assignment, the other about Qwen3.5 save_pretrained checkpoint correctness; unrelated code paths." + }, + { + "left": "issue:36683", + "right": "issue:44625", + "accept": false, + "reason": "Both involve config propagation, but they affect different models and different fields." + }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "Different models and different failures: missing `embed_tokens.weight` in UMT5 vs Qwen3.5 MoE `from_pretrained` error." + "reason": "Different model-loading failures for different architectures; no shared concrete bug." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "Optimizer-state corruption under FSDP2/PEFT is unrelated to a cache-strategy typo." + "reason": "Trainer/FSDP/PEFT optimizer-state bug versus a typo in cache strategy text; unrelated." }, { "left": "issue:43010", "right": "issue:45278", "accept": false, - "reason": "One is a cache/layer `no_grad` correctness change; the other is a broad import regression report." + "reason": "Decorating cache update methods with no_grad is a specific runtime behavior bug; 45278 is a broad import-error regression report." }, { "left": "issue:44075", "right": "issue:46365", "accept": false, - "reason": "SGD argument handling is unrelated to importing `Trainer` being blocked by `datasets`." + "reason": "SGD argument handling is unrelated to an import failure that disappears after uninstalling datasets." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Both affect importing, but one is a compiled-file error and the other is a TorchScript parsing issue in a specific model." + "reason": "Both are import-related, but one is about compiled Python files and the other about a Python 3.13 torch.jit parsing issue." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "State-dict loading order bug and missing config field serialization are different code paths." + "reason": "State-dict loading error versus a missing Qwen3.5 VisionConfig field; different failure mechanisms." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "Checkpoint loading failure and Gemma pipeline roundtrip tokenization failure are not the same bug." + "reason": "Checkpoint loading sort-order bug versus Gemma pipeline roundtrip tokenization; not the same issue." }, { "left": "issue:44938", "right": "issue:45071", "accept": false, - "reason": "Python 3.14 import failure and `PretrainedConfig` type-checking breakage are unrelated regressions." + "reason": "Python 3.14 import failure and PretrainedConfig type-checking regression are distinct problems." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "Gemma3 config attribute missing is unrelated to FSDP2 optimizer-state placement." + "reason": "Different subsystems: config attribute loading versus training/optimizer state under FSDP2." }, { "left": "issue:41669", "right": "issue:44246", - "accept": true, - "reason": "Both report the same import-time slowdown regression; the `import *` cleanup is a plausible direct fix for the slower `transformers` import." + "accept": false, + "reason": "Both mention import performance, but one is about removing import * in models and the other is a general slow-import report." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "A missing `set_seed()` in SamHQ tests is unrelated to xdist races on `captured_info.txt`." + "reason": "A flaky test from missing set_seed is not the same as an xdist race on captured_info.txt." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "A specific cache/layer update bug is not the same as the vague 'too many models break' report." + "reason": "One is a specific no_grad decorator request; the other is a vague broad complaint about v5 breaking many models." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Different models and different loading failures; not enough evidence they share one concrete code path." + "reason": "Both concern model loading, but they are different models and different missing keys/errors." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Loading a sorted state_dict and saving incorrect visual encoder keys are separate serialization bugs." + "reason": "State-dict ordering/load failure is not the same as incorrect visual encoder keys being saved." }, { - "left": "issue:36296", - "right": "issue:43575", + "left": "issue:43010", + "right": "issue:43976", "accept": false, - "reason": "Tensor-parallel training bug and TP-related OOM on a specific Qwen model are too different to merge safely." + "reason": "No_grad decoration on cache update methods is unrelated to Python-version compatibility failures." }, { - "left": "issue:43010", - "right": "issue:43976", + "left": "issue:36296", + "right": "issue:43575", "accept": false, - "reason": "Cache-update semantics have nothing to do with the Python version support regression." + "reason": "Both mention tensor parallelism, but one is a training bug and the other is an OOM during model load." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIP output attentions and patchtsmixer `post_init` eligibility are unrelated model issues." + "reason": "CLIPOutput attentions regression and patchtsmixer post_init policy are unrelated." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "PyTorch-missing import error and tokenizer-loading regression are different failures." + "reason": "AutoModel PyTorch dependency error and tokenizer-loading regression are different issues." }, { "left": "issue:43618", "right": "issue:46285", "accept": false, - "reason": "Both involve CLIP, but one is about attentions output and the other is a model-structure inconsistency." + "reason": "CLIP attentions assignment and CLIP text-model structure inconsistency are related only at a broad CLIP level, not the same concrete bug." }, { "left": "issue:43618", "right": "issue:44857", "accept": false, - "reason": "CLIP attentions assignment and LwDetr AMP crash are unrelated." + "reason": "No connection between missing CLIP attentions and a float16 AMP crash in LwDetrImageLoss." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "A missing ReLU in the DSA indexer is unrelated to DeepSeekV3MoE implementation divergence." + "reason": "A missing ReLU in the DSA indexer is unrelated to a DeepSeekV3MoE implementation divergence." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "Local-files-only network requests and `AddedToken` keyword collisions are separate tokenizer/runtime bugs." + "reason": "Local-files-only network leakage and AddedToken keyword conflicts are different tokenizer/runtime bugs." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "SamHQ test seeding issues are unrelated to loading Phi-4 multimodal models." + "reason": "A flaky integration test and a Phi-4 multimodal loading failure are unrelated." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "Test seeding instability is unrelated to a GraniteSpeech CI device-override stale test." + "reason": "Missing set_seed in a SamHQ test is not the same as a stale device override failure in GraniteSpeech CI." }, { "left": "issue:42890", "right": "issue:45600", "accept": false, - "reason": "Missing `set_seed()` in a test is not the same as broken `auto_mappings` references." + "reason": "Test seeding instability and auto_mappings referencing removed configs are different CI problems." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Tokenizer loading failures and checkpoint reload incompatibility are both loading-related but not the same bug." + "reason": "Tokenizer loading in v5 and checkpoint reload compatibility after upgrading versions are both loading regressions, but not the same bug." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Int4 CUDA memory regression is unrelated to checkpoint reload failures." + "reason": "Int4 CUDA memory regression and checkpoint reload failure are unrelated." }, { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "FP8 save-pretrained behavior for MoE and missing `_tp_plan` for TP are distinct issues." + "reason": "FP8 save_pretrained behavior and missing tensor-parallel plan metadata are different Qwen3.5 issues." }, { "left": "issue:43010", "right": "issue:43065", "accept": false, - "reason": "Cache update semantics do not match the dummy `Conv2d` issue in Sam3PixelDecoder." - }, - { - "left": "issue:43576", - "right": "issue:46132", - "accept": false, - "reason": "`transformers env` CLI breakage and `AttentionInterface.register` behavior change are unrelated." - }, - { - "left": "issue:43502", - "right": "issue:44485", - "accept": false, - "reason": "Offline-loading network requests and GLM-5 RoPE implementation are unrelated." - }, - { - "left": "issue:45125", - "right": "issue:45478", - "accept": false, - "reason": "Both mention Qwen3.5 MoE, but one is specifically tensor-parallel planning while the other is a broader `from_pretrained` error with no concrete shared code path." + "reason": "Cache update no_grad semantics and a dummy Conv2d in Sam3PixelDecoder are unrelated code changes." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "Grounded overall: issue:45278 is a defensible umbrella for the cluster, and the summary does not overclaim beyond the packet. The soft-edge verdicts are appropriately conservative, with the lone accepted pair (issue:41669/issue:44246) reasonably supported by the shared import-time slowdown theme. No mergeability problem appears in the issue pairs." - }, - "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 }, @@ -16235,8 +16226,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -16272,7 +16263,6 @@ "issue:44485", "issue:44492", "issue:44556", - "issue:44623", "issue:44655", "issue:44749", "issue:44908", @@ -16280,16 +16270,20 @@ "issue:44961", "issue:45071", "issue:45125", - "issue:45230", "issue:45362", "issue:45375", "issue:45468", + "issue:45478", "issue:45850", + "issue:46132", "issue:46143", "issue:46144", "issue:46209" ], "soft_pairs": [ + "issue:43576|issue:46132", + "issue:43502|issue:44485", + "issue:45125|issue:45478", "issue:44355|issue:44749", "issue:42548|issue:44351", "issue:30333|issue:33453", @@ -16305,8 +16299,8 @@ "issue:43827|issue:44961", "issue:44355|issue:44655", "issue:33453|issue:34689", - "issue:44315|issue:44360", "issue:27957|issue:33453", + "issue:44315|issue:44360", "issue:42757|issue:43576", "issue:43576|issue:45071", "issue:43576|issue:43824", @@ -16319,244 +16313,255 @@ "issue:46143|issue:46144", "issue:36296|issue:43901", "issue:41628|issue:42548", - "issue:44315|issue:46209", "issue:36296|issue:43526", + "issue:44315|issue:46209", "issue:43673|issue:44492", - "issue:45362|issue:45850", - "issue:44623|issue:45230" + "issue:45362|issue:45850" ], - "prepared_review_unit_hash": "1015071dbee31d57587bc26381d87a835f000d6ea87e18e33710fbf4830ba2ae", + "prepared_review_unit_hash": "7d319c0a0acc624ca02e54c3f93f1459abb2b4f62d1d113b5edc16d9275d4682", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37046, - "estimated_input_tokens": 9262, - "estimated_eval_tokens": 18780 + "soft_pair_count": 36, + "serialized_chars": 37714, + "estimated_input_tokens": 9429, + "estimated_eval_tokens": 19114 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a duplicate set; it mixes unrelated import regressions, model-loading bugs, training/runtime issues, docs problems, and feature requests. None of the soft pairs look like the same underlying bug/change with a fix that could plausibly be merged as one PR.", - "confidence": 0.92, - "canonical_issue_reason": "No single issue is a safe canonical because the items span many distinct code paths and problem types rather than one underlying defect.", + "summary": "This cluster is not a true duplicate set: most items are unrelated regressions, import errors, docs issues, or model-specific bugs that only share Transformers/v5 similarity. No soft edge looks like the same underlying bug or change.", + "confidence": 0.77, + "canonical_issue_reason": "No single issue cleanly represents the whole cluster, but issue 43576 is the best anchor because it is a concrete, user-facing v5 regression with some discussion/inbound attention and broader reach than the model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "No clear best issue: the closest matches are still only broad thematic overlaps (for example, generic v5 regressions or model-loading failures), not the same concrete bug.", + "best_issue_reason": "Issue 43576 is the most generally representative and actionable item in the set; the others are mostly narrower, separate failures or docs problems.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Both touch v5-era behavior, but one is a broken env command and the other is AttentionInterface.register semantics; different code paths and symptoms." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "local_files_only network leakage is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5 MoE, but one is tensor-parallel plan support and the other is a from_pretrained loading failure; too different to merge as one bug." + }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled-file import/runtime errors vs. a preprocessing slowdown regression; different symptoms and code paths." + "reason": "Compiled Python file errors and post-upgrade slowdown are different problems." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different missing symbols with no evidence of the same root cause." + "reason": "Different missing imports for different symbols; likely separate export/version issues." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLFlow reporting on failed training jobs is unrelated to tokenizer loading regression." + "reason": "MLflow reporting after failed training and tokenizer loading regression are unrelated." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Different Qwen3.5 components: tensor-parallel plan vs. missing config field in vision config." + "reason": "Qwen3.5 MoE tensor-parallel support and missing vision config field are different defects." }, { "left": "issue:43010", "right": "issue:43232", "accept": false, - "reason": "Both touch generation/cache behavior, but one is about no_grad decoration and the other about sync_gpus kwargs updates." + "reason": "Both involve generation/cache internals, but one is about no_grad decoration and the other about sync_gpus/update_model_kwargs behavior; not the same bug." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "local_files_only network leakage and FP8 save_pretrained issues are unrelated." + "reason": "local_files_only network calls and FP8 save_pretrained on MoE are separate issues." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Loading hangs vs. documentation confusion; not the same bug." + "reason": "Loading hang for Sentence Transformers is unrelated to documentation confusion in model.forward." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Both are loading regressions after v5 upgrades, but they affect different model/load paths and are not clearly the same defect." + "reason": "Both are version-regression complaints, but they hit different model families and failure modes." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "Network-request bug vs. unrelated placeholder issue." + "reason": "No relation between offline-mode network requests and the unrelated 'racoon' issue." }, { "left": "issue:42831", "right": "issue:43874", "accept": false, - "reason": "FP8 accuracy regression and a missing image-processor method causing AttributeError are different failures." + "reason": "FineGrainedFP8 accuracy and a missing image-processor method are different problems." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLFlow job-status reporting and tensor-parallel training are unrelated." + "reason": "MLflow reporting and tensor-parallel training bugs are unrelated." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled Python file handling vs. PretrainedConfig type-checking regression are different issues." + "reason": "Compiled-file loading errors and PretrainedConfig type-checking regressions are different issues." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs still mentioning pipeline() is unrelated to the racoon placeholder issue." + "reason": "Docs still referencing pipeline() and the unrelated 'racoon' issue do not match." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "Compiled-file errors and pipeline save_pretrained are different code paths and symptoms." + "reason": "Both are loading/saving related at a high level, but compiled-file errors and save_pretrained failure are different code paths." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer-loading regression vs. model-loading failure for a specific Llama vision model; not the same bug." + "reason": "Both are loading regressions, but tokenizer loading and Llama 3.2 Vision model loading are distinct failures." }, { - "left": "issue:44315", - "right": "issue:44360", + "left": "issue:27957", + "right": "issue:33453", "accept": false, - "reason": "Liger Kernel application bug and a DSA indexer ReLU discussion are unrelated." + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to tokenizer loading regression." }, { - "left": "issue:27957", - "right": "issue:33453", + "left": "issue:44315", + "right": "issue:44360", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support issue is unrelated to tokenizer loading regression." + "reason": "Liger Kernel application and a DSA indexer ReLU discussion are unrelated." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "Missing huggingface_hub symbol import and broken `transformers env` CLI are different failures." + "reason": "Missing huggingface_hub import and broken env command are different failures." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "CLI breakage in v5 and PretrainedConfig type checking are unrelated." + "reason": "A broken CLI env command and a PretrainedConfig typing regression are not the same bug." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "Broken env command and missing Qwen2.5-VL export are different API surfaces." + "reason": "CLI env breakage is unrelated to Qwen2.5-VL import errors." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Docs confusion about forward args vs. missing layer_type_validation import; not the same problem." + "reason": "Documentation confusion and missing layer_type_validation import are separate issues." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "Placeholder racoon issue vs. Gemma4 audio positional encoding bug are unrelated." + "reason": "Unrelated placeholder issue versus Gemma-4 audio positional encoding bug." }, { "left": "issue:43576", "right": "issue:44938", "accept": false, - "reason": "Broken env command and Python 3.14 load failure are different compatibility issues." + "reason": "Broken env command and Python 3.14 load failure are different problems." }, { "left": "issue:44485", "right": "issue:44961", "accept": false, - "reason": "GLM-5 RoPE implementation discussion is unrelated to the racoon placeholder issue." + "reason": "GLM-5 RoPE implementation and 'racoon' are unrelated." }, { "left": "issue:44908", "right": "issue:44961", "accept": false, - "reason": "Scheduler kwargs bug and unrelated placeholder issue are not duplicates." + "reason": "Scheduler kwargs bug and 'racoon' are unrelated." }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "FP8 save_pretrained bug and FP8 inference support request are related by theme only, not the same concrete bug." + "reason": "Both are FP8/MoE-adjacent, but one is save_pretrained behavior and the other is inference support; not the same concrete fix." }, { "left": "issue:46143", "right": "issue:46144", "accept": false, - "reason": "Both are RoFormer-related, but one is kwargs forwarding and the other is attention-interface usage; different code-path issues." + "reason": "Both concern RoFormer, but one is kwargs forwarding and the other is attention-interface integration; separate code-path problems." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor-parallel training bug and text-classification docs mismatch are unrelated." + "reason": "Tensor-parallel training bug and text-classification docs mention are unrelated." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "Different missing top-level exports; no evidence of one shared root cause." + "reason": "Different import symbols failing from transformers; likely separate export regressions." }, { - "left": "issue:44315", - "right": "issue:46209", + "left": "issue:36296", + "right": "issue:43526", "accept": false, - "reason": "Liger Kernel application bug and FP8 B200 inference failure are unrelated." + "reason": "Tensor-parallel training and BeitImageProcessorFast reduce_labels behavior are unrelated." }, { - "left": "issue:36296", - "right": "issue:43526", + "left": "issue:44315", + "right": "issue:46209", "accept": false, - "reason": "Tensor-parallel training bug and BeitImageProcessorFast label reduction bug are unrelated." + "reason": "Liger Kernel/model_init behavior and FP8 inference on B200 are different bugs." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Generation cache missing in chunked_prefill and a cache-strategy typo are not the same defect." + "reason": "Generation cache missing in v5 and a cache-strategy typo are not the same issue." }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Specific Qwen3.5-35B chat crash vs. a broad generic complaint that v5 breaks many models; not the same bug." - }, - { - "left": "issue:44623", - "right": "issue:45230", - "accept": false, - "reason": "Missing files in processor.save_pretrained and a generic bug report are not duplicate issues." + "reason": "Both are broad v5 complaints, but one is a specific Qwen3.5 chat crash and the other is a generic 'too many models' report." } ] }, @@ -16575,8 +16580,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -16606,6 +16611,7 @@ "issue:44393", "issue:44492", "issue:44556", + "issue:44623", "issue:44655", "issue:44704", "issue:44829", @@ -16629,14 +16635,15 @@ "issue:46328" ], "soft_pairs": [ + "issue:44623|issue:45230", "issue:44273|issue:45542", "issue:44038|issue:45125", "issue:44393|issue:45071", "issue:44273|issue:45003", "issue:44393|issue:45753", "issue:43576|issue:44556", - "issue:43525|issue:45850", "issue:41669|issue:43673", + "issue:43525|issue:45850", "issue:44230|issue:46095", "issue:44556|issue:45092", "issue:43572|issue:45850", @@ -16657,294 +16664,300 @@ "issue:41669|issue:44351", "issue:44492|issue:46328", "issue:41669|issue:45341", - "issue:44393|issue:45478", "issue:43976|issue:45850", + "issue:44393|issue:45478", "issue:43976|issue:46218", "issue:44077|issue:44861", "issue:43827|issue:44908", "issue:31795|issue:33453", "issue:44315|issue:44829", "issue:43576|issue:44273", - "issue:44655|issue:44861", "issue:44360|issue:45468", + "issue:44655|issue:44861", "issue:45230|issue:45310", "issue:44273|issue:44908", "issue:43576|issue:46218" ], - "prepared_review_unit_hash": "2befc357ecd76a3a639e9dbdaa9514927e3e41e2a1fbb7cead6c1999fad7804d", + "prepared_review_unit_hash": "301aeb561b30cc1e517103e43297bc9079034010501c1fa4016145c25201d468", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 41, - "serialized_chars": 38848, - "estimated_input_tokens": 9712, - "estimated_eval_tokens": 19680 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 39694, + "estimated_input_tokens": 9924, + "estimated_eval_tokens": 20104 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pair looks like the same concrete underlying bug/change; most are only loosely related by model family, v5 migration, or import/loading themes. I\u2019d anchor the cluster on the broad lazy-loading/import behavior issue rather than a model-specific regression.", - "confidence": 0.84, - "canonical_issue_reason": "issue:44273 is the broadest open issue in the set about loading/lazy-loading behavior, which best fits the cluster\u2019s recurring import/load regressions.", + "summary": "The cluster is mostly a grab-bag of unrelated Transformers regression reports, with only a few weakly similar v5/import/lazy-loading complaints. I would not merge any of the soft edges; none are clearly the same underlying bug or change.", + "confidence": 0.87, + "canonical_issue_reason": "No single issue cleanly represents this set: the items span docs, model-specific bugs, import/loading regressions, and unrelated API breakages. If forced, issue:45850 is the broadest umbrella for the v5-breakage reports, but it is too vague to be a true canonical duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44273 is the strongest representative to keep: broad scope, still open, and already has active discussion around the same general loading path.", + "best_issue_reason": "issue:45850 is the closest umbrella representative because it captures the broad v5 regression theme seen across several reports, but it is still not specific enough to serve as a reliable duplicate anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "45230 is only a generic bug report; there is no evidence it refers to the same save_pretrained/file-missing problem." + }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Both mention import/loading pain, but one is lazy loading and the other is a tf-backend error from tensorboard-only installs; different bugs." + "reason": "Both mention runtime behavior, but one is lazy-loading/import behavior and the other is a TensorFlow backend/package issue; different bugs." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "Same model family, but different failures: a Qwen3-VL-Moe bug versus missing _tp_plan for tensor parallelism." + "reason": "Both involve Qwen3/MoE, but one is a model-support regression and the other is missing tensor-parallel metadata; not the same defect." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "One is a Qwen3-VL bbox/output bug; the other is a global PretrainedConfig type-checking regression." + "reason": "These are unrelated: a Qwen3-VL bbox-generation issue versus a PretrainedConfig type-checking regression." }, { "left": "issue:44273", "right": "issue:45003", "accept": false, - "reason": "Lazy loading/import slowness is not the same as unsafe sys.modules access in modeling_utils." + "reason": "Lazy loading/import behavior and unsafe sys.modules access are different code paths and different failure modes." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Both are Qwen3-VL-related, but one is 2D bbox output corruption and the other is an infinite loop on a specific image." + "reason": "Both are Qwen-related, but one is 2D bbox output error and the other is an infinite-loop case on a specific image; not the same bug." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "Both are v5 upgrade complaints, but one is the transformers env command and the other is checkpoint reload compatibility." + "reason": "Both are v5 upgrade regressions, but one is the env command and the other is checkpoint reload compatibility; different concrete problems." }, { - "left": "issue:43525", - "right": "issue:45850", + "left": "issue:41669", + "right": "issue:43673", "accept": false, - "reason": "Specific Llama4 pad_token_id attribute error is not the same bug as a generic 'many models break' report." + "reason": "Import-time slowdown from import * removal and missing GenerationMixin cache during chunked prefill are unrelated." }, { - "left": "issue:41669", - "right": "issue:43673", + "left": "issue:43525", + "right": "issue:45850", "accept": false, - "reason": "Import-star performance cost and missing GenerationMixin cache during chunked prefill are unrelated code-paths." + "reason": "45850 is a vague umbrella complaint; it does not match the specific Llama4Config pad_token_id AttributeError." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 Qwen3-VL/Moe inference support and a DeepSeek weight-conversion question are different requests." + "reason": "FP8 Qwen3-VL/MoE support and a DeepSeek weight-conversion question are different topics entirely." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both concern v5/v4 checkpoint compatibility, but they target different model-loading failures and different mechanisms." + "reason": "Both concern upgrade/reload compatibility, but one is general checkpoint reload after v4.57 and the other is remote-code/meta-init incompatibility for InternVL2; not clearly the same bug." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "A missing StableLmConfig pad_token_idx attribute is not the same as a broad v5 breakage complaint." + "reason": "The StableLm pad_token_idx regression is too specific to be the same as the broad, vague v5-breakage complaint." }, { "left": "issue:44492", "right": "issue:45850", "accept": false, - "reason": "A typo in cache-strategy docs/config is too narrow to match a generic 'too many models break' issue." + "reason": "A typo in cache strategies is not the same underlying issue as a generic report that v5 breaks many models." }, { "left": "issue:45278", "right": "issue:45850", "accept": false, - "reason": "General import errors after upgrade are not the same concrete bug as a broad complaint about model breakage." + "reason": "Both are broad v5 regression reports, but one is specifically about import errors after upgrade and the other is a vague general complaint; not the same concrete bug." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API guidance and a cache-strategies typo are unrelated documentation issues." + "reason": "TF32 API guidance and a cache-strategy typo are unrelated documentation/API issues." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import-star cleanup/performance is not the same as failing to import a specific Qwen3_5 model class." + "reason": "Import slowness from model imports and failure to import Qwen3_5ForCausalLM are distinct issues." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "A sam3_tracker/sam3_video config warning is unrelated to Liger Kernel not being applied with model_init." + "reason": "A wrong model-type warning and Liger Kernel not being applied with model_init are separate behaviors." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Missing set_seed in SAM HQ tests and inverse_sqrt scheduler kwargs are unrelated test/scheduler issues." + "reason": "Missing set_seed in tests and inverse_sqrt scheduler kwargs handling are unrelated test/stability and scheduler issues." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention 2 unsupported for NemotronH is unrelated to Qwen3-VL output errors." + "reason": "Flash Attention 2.0 support for Nemotron and Qwen3-VL bbox hallucination are different model-specific problems." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Both concern import behavior, but one is intermittent import slowness while the other is lazy-loading correctness; not the same bug." + "reason": "Both touch import/lazy-loading performance, but one is a general intermittent import slowness report and the other is a specific lazy-loading bug; insufficient evidence they are the same defect." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "A broken env command and scheduler kwargs handling are unrelated." + "reason": "The env command regression and scheduler kwargs handling are unrelated." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python 3.9/v5 compatibility is unrelated to Qwen3-VL bounding-box hallucination/output errors." + "reason": "Python version compatibility and Qwen3-VL bbox output are unrelated." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "Position_ids buffer access and a generic bug report do not identify the same failure." + "reason": "A position_ids buffer access error is not tied to the generic bug report." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer class metadata mismatch is not the same as an unspecified bug report." + "reason": "Tokenizer save_pretrained metadata mismatch is a different issue from the generic bug report." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler kwargs and PretrainedConfig type checking are separate regressions." + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are unrelated." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs affect different subsystems and code paths." + "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs are different APIs and different bugs." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-star usage/performance is not the same as a missing top-level HybridCache symbol." + "reason": "Model import performance and missing HybridCache import are unrelated import-path issues." }, { "left": "issue:44492", "right": "issue:46328", "accept": false, - "reason": "A cache-strategy typo and a SequenceFeatureExtractor.pad performance issue are unrelated." + "reason": "Cache-strategy typo and feature-extractor pad performance are not the same code-path problem." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Model import cleanup and a testing_utils bug are unrelated." + "reason": "Import performance regression and a testing_utils bug do not match." }, { - "left": "issue:44393", - "right": "issue:45478", + "left": "issue:43976", + "right": "issue:45850", "accept": false, - "reason": "Different Qwen3 model variants and different failures: bbox output error versus from_pretrained error." + "reason": "Python support issues are not the same as a broad claim that v5 breaks too many models." }, { - "left": "issue:43976", - "right": "issue:45850", + "left": "issue:44393", + "right": "issue:45478", "accept": false, - "reason": "Python version compatibility and a vague 'too many models break' complaint are not the same bug." + "reason": "They are both Qwen-related, but one is bbox output hallucination and the other is a Qwen3.5 MoE from_pretrained error; not the same underlying bug." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "Python 3.9 incompatibility is unrelated to the Qwen3_5 import failure." + "reason": "Python compatibility and missing Qwen3_5ForCausalLM import are unrelated." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init policy and _get_tied_weight_keys crashing on list input are different model-initialization issues." + "reason": "Optional post_init enforcement for patchtsmixer and _get_tied_weight_keys crashing on list are distinct implementation issues." }, { "left": "issue:43827", "right": "issue:44908", "accept": false, - "reason": "Docs still referencing pipeline() after removals is unrelated to scheduler kwargs propagation." + "reason": "Pipeline docs referencing removed pipeline() and scheduler kwargs handling are unrelated." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion for model.forward inputs is unrelated to tokenizer loading regression." + "reason": "Documentation confusion around forward arguments and tokenizer loading regression are different bugs." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "Liger Kernel application with model_init and flash_attention_3 training degeneration are different code paths and symptoms." + "reason": "Liger Kernel application with model_init and degenerate training under flash_attention_3 are different failure modes." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "The env command bug and lazy-loading failure are separate v5 regressions." + "reason": "The env command regression and lazy-loading issue are not the same concrete problem." }, { - "left": "issue:44655", - "right": "issue:44861", + "left": "issue:44360", + "right": "issue:45468", "accept": false, - "reason": "Pipeline save_pretrained and tied-weights-key handling are unrelated save/serialization bugs." + "reason": "A ReLU discussion and Gemma-4 audio positional encoding are unrelated." }, { - "left": "issue:44360", - "right": "issue:45468", + "left": "issue:44655", + "right": "issue:44861", "accept": false, - "reason": "A DSA indexer discussion and Gemma-4 audio positional encoding are unrelated." + "reason": "Pipeline save_pretrained support and tied-weight key handling are separate save/serialization issues." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "A generic bug report and a specific Qwen3.5 Moe from_pretrained error do not match the same underlying issue." + "reason": "A generic bug report does not establish the same Qwen3.5 MoE from_pretrained error." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy loading not working and inverse_sqrt scheduler kwargs are unrelated." + "reason": "Lazy loading/import behavior and scheduler kwargs handling are unrelated." }, { "left": "issue:43576", "right": "issue:46218", "accept": false, - "reason": "Broken env command and Qwen3_5 import failure are different v5 problems, not one bug." + "reason": "Broken env command and missing Qwen3_5ForCausalLM import are different issues." } ] }, @@ -16963,8 +16976,8 @@ "review_unit_count": 54, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-463", "nodes": [ @@ -17025,12 +17038,12 @@ "issue:31515|issue:31795", "issue:43704|issue:44485", "issue:44485|issue:45468", - "issue:45850|issue:46218", "issue:42371|issue:43704", + "issue:45850|issue:46218", "issue:43502|issue:43519", "issue:43519|issue:44623", - "issue:43519|issue:43704", "issue:43298|issue:46095", + "issue:43519|issue:43704", "issue:44263|issue:45468", "issue:43606|issue:45468", "issue:44261|issue:45468", @@ -17065,316 +17078,316 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "15014a201888734b3e13035f7b892768ea66d186eb9ab8eeec45032266c4ae85", + "prepared_review_unit_hash": "c01201a6112e71ed5363463a43e39a92e7ad5ca503ee423b41b207f042350fcc", "component_budget": { "node_count": 463, "item_count": 463, "soft_pair_count": 1704, - "serialized_chars": 883318, - "estimated_input_tokens": 220830, - "estimated_eval_tokens": 441916 + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 47, - "serialized_chars": 40312, - "estimated_input_tokens": 10078, - "estimated_eval_tokens": 20412 + "serialized_chars": 40307, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous; none of the soft pairs look like the same underlying bug or change. I would anchor the cluster on the concrete loading bug in issue 43502.", - "confidence": 0.71, - "canonical_issue_reason": "Issue 43502 is the most concrete, user-facing, and actionable bug report in the set: a specific loading path violates `local_files_only=True` and causes unwanted API calls.", + "summary": "The cluster is mostly a loose grab-bag of unrelated issues, with one dominant theme around Transformers v5/model-loading regressions and tokenizer/config breakages. No soft-edge pair looks like the same concrete bug, so none are merged.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45850 is the broadest umbrella report in the cluster and best matches the dominant v5 regression / model-breakage theme.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43502 is the strongest standalone issue candidate because it describes a specific reproducible defect with clear scope and likely code-path ownership.", + "best_issue_reason": "issue:45850 is the best single representative because it summarizes the widespread version-5 breakage pattern seen across several other issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Different bug domains: Qwen3VL timestamp calculation vs GLM-5 RoPE implementation." + "reason": "Both are model-specific correctness bugs, but one is Qwen3VL timestamp calculation and the other is GLM-5 RoPE implementation; different code paths and symptoms." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor timestamp math and scheduler kwargs handling are unrelated code paths." + "reason": "Qwen3VL timestamp math and inverse_sqrt scheduler kwargs are unrelated subsystems." }, { "left": "issue:46018", "right": "issue:46172", "accept": false, - "reason": "Shared DeepSeek-v4 context is too broad; one is expert gating, the other is a tensor transpose bug." + "reason": "DeepSeek-V4 shared-expert gating and a comb.transpose dtype issue are different DeepSeek problems with different fixes." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-star cleanup/performance issue is unrelated to a Qwen3ForCausalLM VRAM leak." + "reason": "Import-star cleanup/performance issue is unrelated to a Qwen3ForCausalLM VRAM leak in dataloader threads." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Documentation confusion about forward args is not the same as a tokenizer-loading regression." + "reason": "Documentation confusion for forward args is not the same as a model-loading regression for Llama 3.2 Vision." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Slow checkpoint loading and forward API docs are different issues with no shared concrete failure." + "reason": "Slow checkpoint loading and forward-API documentation confusion are different issue types and different code paths." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak in multithreaded loading is unrelated to RoPE implementation correctness." + "reason": "VRAM leak in Qwen3 threading and GLM-5 RoPE implementation are unrelated bugs." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "Both mention model internals, but RoPE implementation and audio positional encoding are distinct bugs." + "reason": "GLM-5 RoPE math and Gemma-4 audio positional encoding are different model implementations." }, { - "left": "issue:45850", - "right": "issue:46218", + "left": "issue:42371", + "right": "issue:43704", "accept": false, - "reason": "A broad \u201cmany models break\u201d complaint is not the same defect as an import failure for one model class." + "reason": "TF32 settings guidance and a Qwen3 memory leak do not share the same bug." }, { - "left": "issue:42371", - "right": "issue:43704", + "left": "issue:45850", + "right": "issue:46218", "accept": false, - "reason": "TF32 configuration guidance and a VRAM leak are unrelated issues." + "reason": "A broad v5 breakage meta-issue is not the same as a Qwen3.5 import failure." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Loading requests despite `local_files_only=True` is unrelated to Qwen3VL timestamp calculation." + "reason": "local_files_only network requests and Qwen3VL timestamp calculation are unrelated." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Processor timestamp math and missing processor save files are different failures." + "reason": "Timestamp calculation in a processor and missing save_pretrained files are different defects." }, { - "left": "issue:43519", - "right": "issue:43704", + "left": "issue:43298", + "right": "issue:46095", "accept": false, - "reason": "These affect different models and different failure modes: timestamps vs memory leak." + "reason": "Audio continuous batching support and DeepSeek-V4 weight conversion are different feature requests/bugs." }, { - "left": "issue:43298", - "right": "issue:46095", + "left": "issue:43519", + "right": "issue:43704", "accept": false, - "reason": "Audio continuous batching support and DeepSeek-v4 weight conversion are not the same problem." + "reason": "Qwen3VL timestamp math and Qwen3 threading VRAM leak are distinct model/runtime issues." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "Torch split return handling and Gemma-4 audio positional encoding are unrelated defects." + "reason": "torch.split return handling in one model is unrelated to Gemma-4 audio positional encoding." }, { "left": "issue:43606", "right": "issue:45468", "accept": false, - "reason": "CPU offload device mismatch and Gemma-4 audio positional encoding do not describe the same bug." + "reason": "CPU offload device mismatch for bark-small is a different failure mode than Gemma-4 audio positional encoding." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "Missing RMS norm epsilon in MLA is not the same as Gemma-4 audio positional encoding behavior." + "reason": "Missing rms_norm_eps precision issue in MLA q_a_layernorm is not the same as Gemma-4 audio positional encoding." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "Importing HybridCache and importing Qwen3_5ForCausalLM are separate import failures." + "reason": "Missing HybridCache import and missing Qwen3_5ForCausalLM import are different missing-symbol regressions." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Flash Attention support and a missing model import are different issues." + "reason": "Flash Attention 2 support for NemotronH and Qwen3.5 import failure are unrelated." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Both concern tokenizer/v5 behavior, but one is broad behavioral difference and the other is a specific decoder-type mismatch." + "reason": "Unexpected tokenizer behavior differences and wrong tokenizer decoder type are both tokenizer-adjacent, but not the same bug." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Scheduler kwargs handling is unrelated to the nonsensical unrelated issue title." + "reason": "Scheduler kwargs propagation and the unrelated 'Unemployment' issue are not duplicates." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "A cache-strategy typo is not the same as an unrelated low-quality issue." + "reason": "Cache-strategy typo and a nonsensical unrelated issue title do not match." }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attentions regression is unrelated to the unrelated placeholder issue." + "reason": "CLIPOutput attentions regression and the unrelated 'Unemployment' issue are not the same bug." }, { "left": "issue:44998", "right": "issue:45850", "accept": false, - "reason": "An unrelated/invalid issue title does not match a broad Transformers 5.0 breakage report." + "reason": "A spam/off-topic issue is not the same as a broad v5 breakage report." }, { "left": "issue:44938", "right": "issue:44998", "accept": false, - "reason": "Python 3.14 loading failure and the unrelated placeholder issue are not duplicates." + "reason": "Python 3.14 load failure and the unrelated 'Unemployment' issue are unrelated." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weight key handling and chat-template tool-call crashes are separate code paths." + "reason": "_get_tied_weight_keys crash and apply_chat_template tool-call crash affect different code paths." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Missing `pad_token_id` in config and tied-weight key crashes are unrelated." + "reason": "Llama4 pad_token_id AttributeError and tied-weight-key handling are different configuration/model-loading bugs." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Slow checkpoint loading and an ImportError for `layer_type_validation` are different defects." + "reason": "Slow from_pretrained checkpoint loading and missing layer_type_validation import are unrelated." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "StableLm config pad token naming and tied-weight list handling are not the same bug." + "reason": "StableLmConfig missing pad_token_idx and tied-weight-key crash are unrelated config issues." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 inference support and missing processor save files are unrelated." + "reason": "FP8 Qwen3-VL/MoE support and processor.save_pretrained missing files are different topics." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow reporting on failed training jobs is unrelated to slow checkpoint loading." + "reason": "MLFlow reporting on failed training jobs and slow checkpoint loading are unrelated." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Checkpoint load performance and tensor-parallel training bugs are different issues." + "reason": "Slow checkpoint loading and tensor-parallel training bug are different defects." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "FP8 support for Qwen3-VL/Qwen3.5 MoE and a from_pretrained error are not the same defect." + "reason": "FP8 support request and Qwen3.5 Moe from_pretrained error are different issues." }, { "left": "issue:31515", "right": "issue:33453", "accept": false, - "reason": "Slow checkpoint loading and tokenizer loading regression are different loading problems." + "reason": "Slow checkpoint loading and tokenizer loading regression are different problems." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Audio continuous batching support and FP8 inference support do not describe the same bug." + "reason": "Continuous batching for audio models and FP8 inference support are unrelated feature gaps." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba init_weights behavior is unrelated to RoFormer attention interface usage." + "reason": "Mamba init_weights behavior and RoFormer attention interface usage are unrelated model implementations." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former ignore_value handling is unrelated to DeepSeek-V4 shared expert gating." + "reason": "Mask2Former ignore_value misuse and DeepSeek-V4 shared expert gating are different model bugs." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "Incorrect `norm_eps` typing and tokenizer regex warning behavior are different issues." + "reason": "ModernBERT config type hint and tokenizer regex warning are not the same bug." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Mask2Former loss masking and missing kwargs in loss calculation are unrelated." + "reason": "Mask2Former ignore_value bug and missing kwargs in loss calculation are unrelated." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss kwargs propagation and RoFormer attention interface usage are different code paths." + "reason": "Loss-function kwargs propagation and RoFormer attention interface usage are different code paths." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former ignore_value handling does not match tokenizer regex warning regression." + "reason": "Mask2Former ignore_value and tokenizer regex warnings are unrelated." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Mask2Former segmentation loss behavior is unrelated to RoFormer attention implementation." + "reason": "Mask2Former ignore_value bug and RoFormer attention interface issue are different models and fixes." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Mask2Former ignore_value and a quantization-operation typo are not the same defect." + "reason": "Mask2Former ignore_value and a quantization_operation typo are unrelated." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwargs propagation and config type-hint correctness are unrelated." + "reason": "Loss kwargs propagation and ModernBERT norm_eps type hint are unrelated." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Config type hints and a typo in quantization terminology are separate issues." + "reason": "ModernBERT type-hint mismatch and quantization typo are unrelated." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "A question about Llama4TextL2Norm semantics is unrelated to ModernBERT config typing." + "reason": "Llama4TextL2Norm/RMS norm question and a ModernBERT type hint issue are unrelated." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value behavior and ModernBERT norm_eps typing are unrelated." + "reason": "Mask2Former ignore_value and ModernBERT norm_eps typing are different bugs." } ] }, @@ -17393,8 +17406,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10519-3", "nodes": [ @@ -17407,7 +17420,7 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "3bbc2388f919baba56799ffe856817f2848ce4de7e9b684e15bce16dc6e9da6f", + "prepared_review_unit_hash": "3ec353ecfd76e7191747003ad58b33c0e441f043e14593c3e79c78521b1323df", "component_budget": { "node_count": 3, "item_count": 3, @@ -17429,30 +17442,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues are not duplicates: they describe unrelated feature requests for truncation behavior, learning-rate scheduling, and Wav2Vec2 word offsets. No soft pair looks like the same underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "These three issues are not duplicates: they cover unrelated topics across tokenizer truncation, learning-rate scheduling, and Wav2Vec2 word timestamps. No single issue serves as a meaningful cluster canonical.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the issues describe different feature requests in different areas and do not share the same underlying problem.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #10519 is the most concrete and narrowly scoped request, so it is the best representative issue if one must be chosen.", + "best_issue_reason": "None of the issues is a good cluster representative; if forced, 10519 is the most self-contained, but it is still unrelated to the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different domains and requests: text truncation API behavior vs Wav2Vec2 time offsets." + "reason": "Different subjects: tokenizer truncation behavior vs. Wav2Vec2 word boundary timestamps." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "Unrelated features: tokenizer truncation option vs ReduceLROnPlateau-like scheduler functionality." + "reason": "Unrelated requests: text truncation option vs. ReduceLROnPlateau-like scheduler functionality." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "No shared underlying problem; one is about optimizer scheduling, the other about extracting word time boundaries from speech models." + "reason": "No common underlying bug or change; one is about LR scheduling, the other about audio word offsets." } ] }, @@ -17471,8 +17484,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-13244-3", "nodes": [ @@ -17484,12 +17497,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "6b54522640216f389fef5d167d6be499dd8871cf8cd18a74f7f75e931b87b780", + "prepared_review_unit_hash": "7ef7defb32fd41c2cc5f9c3ecd5f2c85ab87db28085daf81b5e84c01ea1da200", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2478, + "serialized_chars": 2477, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -17497,7 +17510,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2478, + "serialized_chars": 2477, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -17506,24 +17519,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These issues are only loosely related by ML/model-loading themes, but they describe different bugs and should not be deduplicated.", - "confidence": 0.87, - "canonical_issue_reason": null, + "summary": "These three issues are not duplicates: one is about Tapas tokenization differences, one is about a misleading private-model auth error, and one is about failing to load bert-base-uncased. The only commonality is vague model/tokenization/HF-loading similarity, not the same underlying bug.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue clearly represents a shared bug here; the cluster appears to be a loose similarity grouping rather than true duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "None is a true duplicate; if a representative is needed, 16618 is the broadest model-loading report, but it does not subsume the others.", + "best_issue_reason": "Issue #16618 is the most general of the three and the only one directly about model loading, but it still does not cover the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tapas tokenization vs. model loading are different subsystems and different failure modes; no evidence they are the same bug." + "reason": "Tapas tokenization mismatch is a different problem from a model-loading failure for bert-base-uncased; they affect different code paths and symptoms." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "Both mention loading models, but one is specifically about private-model auth/error messaging and the other is a generic inability to load 'bert-base-uncased'; not enough to treat as the same underlying issue." + "reason": "A bad private-model login error and a failure to load a public model are distinct issues with different root causes and user-facing behavior." } ] }, @@ -17542,8 +17555,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-26350-2", "nodes": [ @@ -17553,12 +17566,12 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "1e65ed008dd3da756bfbfc444e078869e85508e27cbd35febff1726b007f9ed2", + "prepared_review_unit_hash": "fcba337161f68748bd79fd7701cf80c90e5ae9e5d538ceb6fe8b66b7a749b5b4", "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 }, @@ -17566,7 +17579,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1590, + "serialized_chars": 1589, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -17575,18 +17588,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two similarly phrased but substantively different community-contribution issues: one for Flash Attention 2 support across more architectures, the other for GGUF support across more architectures. The shared wording reflects the same kind of request pattern, not the same underlying change.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the requests target different features/code paths (Flash Attention 2 vs GGUF support) and are not duplicates.", + "summary": "These are two distinct feature requests: one for Flash Attention 2 architecture support and one for GGUF architecture support. They share only a broad \u201ccommunity contribution\u201d framing, not the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No single best representative for a duplicate cluster; both issues are separate feature requests with only superficial title similarity.", + "best_issue_reason": "issue #33260 is the cleaner standalone request: it is more narrowly scoped and has more inbound references, so it is the better representative issue in this cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Different underlying enhancements and likely different implementation areas; the similarity is only that both ask for broader architecture support." + "reason": "Different features and code paths (Flash Attention 2 vs GGUF support); not the same bug or change." } ] }, @@ -17605,8 +17618,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-27088-8", "nodes": [ @@ -17645,184 +17658,184 @@ "issue:44370|issue:46226", "issue:46222|issue:46226" ], - "prepared_review_unit_hash": "d3834b689ccaf7c61483e9b0f6770a5272fdd9ba92fe496d18ddc06756e10c4c", + "prepared_review_unit_hash": "6466810d1b85a4e72b8732a1bb3b40f6adffb2cb18904d37de8ff98180b60390", "component_budget": { "node_count": 8, "item_count": 8, "soft_pair_count": 24, - "serialized_chars": 13525, - "estimated_input_tokens": 3382, - "estimated_eval_tokens": 7020 + "serialized_chars": 13531, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 }, "budget": { "node_count": 8, "item_count": 8, "soft_pair_count": 24, - "serialized_chars": 13525, - "estimated_input_tokens": 3382, - "estimated_eval_tokens": 7020 + "serialized_chars": 13531, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is not a single duplicate set: it appears to contain at least two separate translation-request groups, plus several language-specific issues that are clearly distinct. The strongest duplicate evidence is within the identical placeholder-title issues, and the exact Thai-title pair.", - "confidence": 0.61, - "canonical_issue_reason": null, + "summary": "Mostly a set of separate i18n translation requests by language, not one true duplicate cluster. The only clear duplicate subgroups are the three identical placeholder-template issues and the two identical Thai-titled issues.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43834 is the clearest representative of the duplicate placeholder-template subgroup; issue:44365 and issue:44370 match it closely, while the language-specific requests are distinct and should not be collapsed into the same issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46222 is the best representative overall: it is open, has inbound references, and is an exact duplicate of issue:46226.", + "best_issue_reason": "issue:43834 is the best overall issue to keep as the canonical representative because it is the most generic/central of the exact-duplicate subgroup and has the broadest overlap with the other template-like entries.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different translation requests; same broad i18n theme is not enough to treat Turkish and a generic placeholder issue as duplicates." + "reason": "Different language requests; same translation theme but not the same underlying issue." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different translation requests; the titles point to different language-specific issues." + "reason": "Different language requests; not a duplicate bug/change." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Turkish translation request vs a separate placeholder-title request; not the same underlying issue." + "reason": "Turkish translation request vs generic template issue; related topic, but not the same issue." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Different language translation requests; not a duplicate." + "reason": "Different language request (Turkish vs Romanian); not a duplicate." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish vs Spanish translation requests are distinct issues." + "reason": "Different language request (Turkish vs Spanish); not the same underlying issue." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish translation request and a separate placeholder-title request; too broad to merge as duplicates." + "reason": "Spanish translation request is distinct from the generic template issue." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Different translation requests; no evidence they are the same concrete issue." + "reason": "Different language request; only broadly similar by template." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Spanish vs Romanian translation requests are distinct." + "reason": "Different language request (Spanish vs Romanian); not duplicates." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Different language/request; not the same issue." + "reason": "Different language request; same theme, different issue." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian translation request vs a separate placeholder-title issue; not enough to merge." + "reason": "Romanian translation request vs generic template issue; related but not the same issue." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Different translation requests; same subsystem, but not the same issue." + "reason": "Romanian translation request vs generic template issue; not a duplicate." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Romanian translation request is distinct from the placeholder-title request." + "reason": "Romanian request is distinct from the generic placeholder issue." }, { "left": "issue:38435", "right": "issue:46222", "accept": false, - "reason": "Different language/request; similar template text does not make them duplicates." + "reason": "Different language/request content; not the same underlying issue." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "Different language/request; not the same underlying issue." + "reason": "Different language/request content; not duplicates." }, { "left": "issue:43834", "right": "issue:44370", "accept": true, - "reason": "Same placeholder title and near-identical body; this looks like the same translation-request issue reopened or duplicated." + "reason": "Same placeholder title/body pattern; looks like the same template issue duplicated." }, { "left": "issue:43834", "right": "issue:44365", "accept": true, - "reason": "Same placeholder-title translation request with near-identical content." + "reason": "Same generic template issue with identical title/body structure; likely duplicate." }, { "left": "issue:43834", "right": "issue:46222", "accept": false, - "reason": "Body similarity alone is not enough; these appear to be different language-specific translation requests." + "reason": "Placeholder-template issue vs unrelated Thai-titled issue; not the same content." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Different language-specific request; not the same concrete issue." + "reason": "Placeholder-template issue vs unrelated Thai-titled issue; not a duplicate." }, { "left": "issue:44365", "right": "issue:44370", "accept": true, - "reason": "Identical placeholder-title issues with almost the same body; likely duplicates." + "reason": "Identical placeholder-template issues; same underlying issue." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Different translation-request topics; not clearly the same issue." + "reason": "Generic placeholder issue vs Thai-titled issue; similar only at a high level." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Different language/request; reject as duplicate." + "reason": "Generic placeholder issue vs Thai-titled issue; not the same issue." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "The requests appear to be for different translation targets; not a safe duplicate match." + "reason": "Generic placeholder issue vs Thai-titled issue; not a duplicate." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Different translation-request topics; not the same underlying issue." + "reason": "Generic placeholder issue vs Thai-titled issue; not the same underlying issue." }, { "left": "issue:46222", "right": "issue:46226", "accept": true, - "reason": "Exact same title/body pattern; this is the clearest duplicate pair in the cluster." + "reason": "Same exact Thai title/body pattern; these appear to be duplicate issues." } ] }, "evaluator_result": { "accept": true, - "feedback": "The cluster summary is grounded in the packet: the Thai-titled pair 46222/46226 is an exact match by title/body similarity and inbound refs support 46222 as the better representative, while the placeholder-title issues form a separate near-duplicate group. The rejected Turkish/Spanish/Romanian pairings are conservatively phrased and do not overclaim beyond the available evidence." + "feedback": "The summary stays grounded in the packet: it separates the language-specific requests, and only groups the near-identical placeholder-template pairings and the Thai pair. The soft-edge accepts are conservative enough given the very high textual similarity and matching titles." }, "evaluator_used": true, "retried": false, @@ -17838,8 +17851,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-36817-16", "nodes": [ @@ -17898,250 +17911,250 @@ "issue:43837|issue:45177", "issue:45322|issue:46133" ], - "prepared_review_unit_hash": "92adec6eebb4d34a31dbaf30fcff255c9c6f6c4c59b51fa947135d6f86345985", + "prepared_review_unit_hash": "08ee66004d9577990c1ac2896cab2a23192adc6315199d2f9a6f51eafff5836a", "component_budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21655, - "estimated_input_tokens": 5414, - "estimated_eval_tokens": 11084 + "serialized_chars": 21663, + "estimated_input_tokens": 5416, + "estimated_eval_tokens": 11088 }, "budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21655, - "estimated_input_tokens": 5414, - "estimated_eval_tokens": 11084 + "serialized_chars": 21663, + "estimated_input_tokens": 5416, + "estimated_eval_tokens": 11088 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are all model-addition/support requests for different concrete models or modalities, so the high textual similarity is mostly template-like rather than duplicate evidence. I found no soft pair that looks like the same underlying bug/change.", - "confidence": 0.94, - "canonical_issue_reason": "Issue 43671 is the clearest representative support-request thread in this set, with substantial discussion and a concrete model target; the other issues are different model requests rather than true duplicates.", + "summary": "This cluster is mostly a loose pile of unrelated model-request issues that only resemble each other at a very broad \u201cadd model support\u201d level. There are no convincing duplicates here, so all soft edges should be rejected.", + "confidence": 0.88, + "canonical_issue_reason": "No true duplicate cluster exists; if one issue must represent the set, 46133 is the most active and detailed generic model-request thread among otherwise heterogeneous requests.", "canonical_pr_reason": null, - "best_issue_reason": "43671 is the most suitable anchor only in a loose clustering sense, because it is an active standalone model-support request; however, this cluster does not contain a true duplicate set.", + "best_issue_reason": "issue:46133 is the best triage representative because it has the strongest discussion/inbound activity, but it still does not unify the other requests as duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "Different models (EuroBert vs DEIMv2); same request pattern, not the same change." + "reason": "Different model requests (EuroBert vs DEIMv2); same broad theme, not the same change." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert config addition vs Qwen3-TTS support; unrelated targets." + "reason": "EuroBert config support vs Qwen3-TTS support are different models and code paths." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert and RF-DETR are different model additions." + "reason": "EuroBert config vs RF-DETR support are unrelated model additions." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR model vs DEIMv2 model; no shared concrete bug/change." + "reason": "RF-DETR and DEIMv2 are distinct models with different integration work." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "Different model families and tasks (detection vs audio model support)." + "reason": "Object-detection model request vs audio model request; not the same underlying bug/change." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR support request and weight-sparse transformer request are unrelated." + "reason": "RF-DETR support and weight-sparse transformer support are unrelated." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR vs Qwen3-ASR; different model targets and modalities." + "reason": "RF-DETR vs Qwen3-ASR are different model families and implementations." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR vs Qwen3-TTS; only share 'add support' wording." + "reason": "RF-DETR and Qwen3-TTS are unrelated model support requests." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT addition and Qwen3-ASR support are different model requests." + "reason": "NeoBERT and Qwen3-ASR are different model additions." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and OpenAI weight-sparse transformer are unrelated models." + "reason": "NeoBERT support and sparse-transformer support are distinct changes." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT vs LFM2.5 Audio 1.5B; different architecture and modality." + "reason": "BERT-style model request vs audio model request; not duplicates." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "RF-DETR and NeoBERT are distinct model additions." + "reason": "RF-DETR and NeoBERT are separate model requests." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT request is not the same change as ModernVBERT models." + "reason": "NeoBERT and ModernVBERT are different models, even if both are BERT-like." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture support and Qwen3-ASR support are separate model requests." + "reason": "MiniCPM3 architecture vs Qwen3-ASR support are unrelated." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 vs OpenAI weight-sparse transformer; different concrete models." + "reason": "MiniCPM3 architecture work is not the same as sparse-transformer support." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different model support requests." + "reason": "MiniCPM3 and ModernVBERT are distinct model requests." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 vs Voxtral TTS; different model and modality." + "reason": "MiniCPM3 architecture vs Voxtral TTS support are different code paths." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 addition and Qwen3-TTS support are unrelated." + "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT models and Qwen3-ASR are different model targets." + "reason": "ModernVBERT models and Qwen3-ASR support are different additions." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT request vs Qwen3-TTS request; no shared code-path change." + "reason": "ModernVBERT and Qwen3-TTS are separate model families." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models and BERT-like models with RoPE are related only at a broad taxonomy level, not as the same issue." + "reason": "ModernVBERT support and a generic BERT-with-RoPE request are not the same concrete change." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like RoPE support and Voxtral TTS support are unrelated." + "reason": "BERT-like models with RoPE vs Voxtral TTS support are unrelated." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are different modalities and separate integrations." + "reason": "Qwen3-TTS and Qwen3-ASR are related by family name but are distinct audio modalities and separate support work." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and Voxtral-4B-TTS are different model support requests, not duplicates." + "reason": "Qwen3-TTS and Voxtral TTS are different model integrations." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS support and DeepSeek-OCR2 support are unrelated." + "reason": "TTS support vs OCR model support are unrelated changes." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS and weight-sparse transformer are different model requests." + "reason": "Qwen3-TTS support is not the same as sparse-transformer support." }, { "left": "issue:43671", "right": "issue:43909", "accept": false, - "reason": "Qwen3-TTS support and LFM2.5 Audio support are separate audio-model requests." + "reason": "TTS and audio-model request are different model-specific integrations." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR vs LFM2.5 Audio; both audio-related, but not the same model or fix." + "reason": "Qwen3-ASR and LFM2.5 Audio are separate audio models and requests." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR support and EUPE are different model requests." + "reason": "ASR support and EUPE vision/perception encoder support are unrelated." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "ASR support request and weight-sparse transformer request are unrelated." + "reason": "Qwen3-ASR and sparse-transformer support are different code-path changes." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are unrelated model requests." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer support and EUPE support are not the same underlying change." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral TTS are distinct modalities/models." + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different speech-model integrations." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS are different model additions." + "reason": "NeoBERT and Voxtral TTS are unrelated model additions." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR support and DeepSeek-OCR2 support are unrelated." + "reason": "ASR support and DeepSeek-OCR2 support are different model requests." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are different vision-language model requests." - }, - { - "left": "issue:43909", - "right": "issue:44205", - "accept": false, - "reason": "LFM2.5 Audio 1.5B and SAM3-LiteText are different models and tasks." - }, - { - "left": "issue:44121", - "right": "issue:45322", - "accept": false, - "reason": "Weight-sparse transformer and EUPE are unrelated model requests." + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." } ] }, @@ -18160,8 +18173,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-38740-2", "nodes": [ @@ -18171,40 +18184,40 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "987160f7950425cc65b68baf03a2e6abf5619981ce8b0f885deb4a760bd935c1", + "prepared_review_unit_hash": "36daff859e69e3fbee1771f7029c1e67aee48229f6387d9dc936e5f16d55f3ce", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1505, - "estimated_input_tokens": 377, - "estimated_eval_tokens": 1010 + "serialized_chars": 1504, + "estimated_input_tokens": 376, + "estimated_eval_tokens": 1008 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1505, - "estimated_input_tokens": 377, - "estimated_eval_tokens": 1010 + "serialized_chars": 1504, + "estimated_input_tokens": 376, + "estimated_eval_tokens": 1008 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are unrelated: one is a docs request about adding Pruna as an optimization framework, the other is a Vision RoPE refactor. There is no duplicate or mergeable relationship here.", + "summary": "The two issues are unrelated: one is a docs request for adding Pruna as an optimization framework, and the other is a Vision RoPE refactor. No duplicate/merge relationship is supported.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "Issue 46443 is the better canonical representative because it is the substantive technical item; issue 38740 is a separate documentation request about a different topic.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46443 is the more substantive technical issue, but it is not a duplicate of issue:38740 and should not be treated as a canonical cluster target.", + "best_issue_reason": "Issue 46443 is the strongest issue in the set for canonical tracking since it describes an actual code refactor rather than a documentation addition.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Different topics and code areas: docs/framework listing vs vision positional-encoding refactor; no evidence they describe the same underlying bug or change." + "reason": "Different topics and scopes: documentation about Pruna vs a Vision RoPE refactor. They do not describe the same underlying bug or change." } ] }, @@ -18223,8 +18236,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41084-2", "nodes": [ @@ -18234,12 +18247,12 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "4d2175abb49e1fe22db6006c1bb386443e8d2761a04fed01edfaf96bcb5e1e51", + "prepared_review_unit_hash": "b109d2da563161a29894e068cf438e6bf170931298b1c7d903b240f244adecc8", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1480, + "serialized_chars": 1478, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -18247,7 +18260,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1480, + "serialized_chars": 1478, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -18256,18 +18269,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one issue is about block decoding, the other is about DeepSeek v3.2 support. The titles indicate different feature areas and no evidence shows the same underlying bug or change.", + "summary": "These two issues are not duplicates: one requests Set Block Decoding support, while the other requests DeepSeek v3.2 support. They refer to different model/features with no shared targets or code-path evidence.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "Issue #43037 is the more concrete, actionable request; #41084 is a broader/ambiguous decoding topic.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "#43037 is the best representative issue in this set because it is specific and clearly scoped to a single model-support request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different requests in different areas: decoding behavior vs adding support for a specific model/version. No shared targets, files, or concrete code-path overlap." + "reason": "Different underlying requests: block decoding vs DeepSeek v3.2 support. No shared files, targets, or evidence of the same bug/change." } ] }, @@ -18285,9 +18298,9 @@ "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", - "reason": null, - "source": "cache", - "cache_hit": true, + "reason": null, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43598-2", "nodes": [ @@ -18297,7 +18310,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "7b5c373286dc37544c80165a751eb48a42ff66b09fd2e5f4bdde9599edb8ee31", + "prepared_review_unit_hash": "3f7d0ebd870fcf6b5ffa56130a244e9f9b95dbe16d8f89da5e16e2295e806067", "component_budget": { "node_count": 2, "item_count": 2, @@ -18319,18 +18332,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two different closed feature requests in Trainer, not duplicates: one is about making `remove_unused_column` more customizable, the other about FLOPs estimation for custom trainers.", - "confidence": 0.94, - "canonical_issue_reason": "No clear canonical duplicate exists here; the issues target different behaviors and different code paths.", + "summary": "These two issues both mention custom trainers, but they target different Trainer behaviors: one is about `remove_unused_column` customizability, the other about FLOPs estimation. They do not look like the same underlying bug.", + "confidence": 0.91, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue #43598 is the better representative of the cluster only because it is a broader Trainer customizability request, but it is still not a canonical match for #43600.", + "best_issue_reason": "issue 43598 is the more concrete and actionable report, but it addresses a different code path than FLOPs estimation, so there is no true canonical duplicate here.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "They discuss different problems: column pruning customizability versus FLOPs estimation for custom trainers. No shared concrete fix path is evident." + "reason": "Shared mention of custom trainers is too broad; one issue concerns `remove_unused_column` behavior while the other concerns FLOPs counting, so they are not the same bug." } ] }, @@ -18349,8 +18362,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43684-2", "nodes": [ @@ -18360,7 +18373,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "4008d29377c7f9a3871b0346d6aa61df4d6e4579629bd64b7b2f47450d5c0bcc", + "prepared_review_unit_hash": "c6f52a94191bbea14f6d73945f562802be2f94f512a4ffe92144a609cc91eb84", "component_budget": { "node_count": 2, "item_count": 2, @@ -18382,18 +18395,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate feature requests, not duplicates: one asks for Qwen3-Omni auto-class registration, the other asks for AutoModelForSequenceClassification support for Qwen3.5. They target different model families and different auto classes.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "These are related at a high level (both ask for Qwen family AutoModel support), but they target different model/config combinations and different auto classes, so they should not be merged as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43684 is the broader, more foundational request: register Qwen3-Omni with AutoModel/AutoModelForConditionalGeneration, which is the clearer canonical tracking item for auto-class registration work.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the broader of the two requests, but it still does not subsume issue 44405; they are different additions in different code paths.", + "best_issue_reason": "Issue 43684 is the best representative of the cluster because it is the more general auto-registration gap, while 44405 is a separate, narrower request for sequence-classification support on Qwen3.5.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying changes: Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration vs. sequence-classification support for Qwen3.5. They do not look mergeable into one fix." + "reason": "Different underlying changes: one is Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration, the other is Qwen3.5 AutoModelForSequenceClassification support. They involve different model families/configs and different auto heads, so they are not the same bug/change." } ] }, @@ -18412,8 +18425,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43835-4", "nodes": [ @@ -18430,12 +18443,12 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "0a486399d8626513c5add968250425b818f9f6e437cd1f7781e160a50920ca20", + "prepared_review_unit_hash": "c0a00ce36dfba26aaf9ef10c63c055ebfb5ff6397b71c29516c00218ca479d05", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4265, + "serialized_chars": 4266, "estimated_input_tokens": 1067, "estimated_eval_tokens": 2390 }, @@ -18443,7 +18456,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4265, + "serialized_chars": 4266, "estimated_input_tokens": 1067, "estimated_eval_tokens": 2390 }, @@ -18452,55 +18465,58 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster is dominated by low-information issue titles, with three near-identical \"huggingface \" issues and one unrelated-looking \"Fett\" issue. There is not enough concrete problem detail to confidently merge any pair.", - "confidence": 0.28, - "canonical_issue_reason": "issue:43845 is the earliest of the repeated \"huggingface \" issues and is the most reasonable representative of that likely duplicate set, though the evidence is very weak.", + "summary": "The three \"huggingface\" issues are likely duplicates of one another based on identical titles and high similarity, but the evidence is limited to metadata. \"Fett\" looks separate. I\u2019d anchor the cluster on the earliest huggingface issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43845 is the earliest of the repeated \"huggingface\" issues and is the best anchor for the likely duplicate set, given the identical titles and high similarity.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43845 is the best available issue to anchor the cluster because it matches the repeated title pattern and was opened first; however, the lack of body detail makes this only a tentative choice.", + "best_issue_reason": "issue:43845 is the cleanest representative of the likely duplicate cluster because it is earliest and shares the exact repeated title with the other huggingface issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles and no body details or explicit links tying them to the same underlying issue." + "reason": "Different title and no concrete content overlap; soft similarity alone is not enough to treat them as duplicates." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no concrete evidence they describe the same bug." + "reason": "Different title and no concrete evidence of the same underlying report." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles and no concrete evidence they describe the same bug." + "reason": "Different title and only weak similarity evidence; not enough to accept as duplicates." }, { "left": "issue:43845", "right": "issue:43846", - "accept": false, - "reason": "The titles match, but there is no issue text or linkage proving they are the same underlying report; keep strict and reject." + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, though the packet does not expose the underlying issue text." }, { "left": "issue:43845", "right": "issue:43859", - "accept": false, - "reason": "The titles match, but there is no issue text or linkage proving they are the same underlying report; keep strict and reject." + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, with no evidence they are different reports." }, { "left": "issue:43846", "right": "issue:43859", - "accept": false, - "reason": "The titles match, but there is no issue text or linkage proving they are the same underlying report; keep strict and reject." + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, but the available evidence is limited to metadata." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The cluster anchor is reasonable, but several soft-edge justifications overstate the evidence. The packet only gives identical titles, body lengths, and soft-similarity scores; it does not provide concrete issue content showing the reports are the same. Phrases like \"clearly overlapping issue report\" are too strong here. Please keep the duplicate verdicts more conservative, e.g. \"likely duplicate based on identical title and high similarity.\"" + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -18512,8 +18528,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-44371-5", "nodes": [ @@ -18534,7 +18550,7 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "8cff16aca8cf9015d491d5e4177a3409ff14055d6b81c5d3a61d2be8132fe514", + "prepared_review_unit_hash": "6e65ec30bc4253951e92f3965429250922f6ba2c2e9eb38cd178d4ddc7acffc0", "component_budget": { "node_count": 5, "item_count": 5, @@ -18556,66 +18572,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster looks like unrelated low-information/spam issues rather than a true duplicate set. The only strong similarity is generic or identical spam-like titles, with no substantive bodies, links, or code context to prove a shared underlying problem.", - "confidence": 0.91, - "canonical_issue_reason": "Issue 44572 is the most representative anchor only because it shares the exact '' title with 44573 and fits the dominant spam-like pattern in the cluster; however, the issues do not provide enough real content to justify a genuine canonical bug report.", + "summary": "This cluster is dominated by low-information spam/placeholder issues. The only exact duplicate-looking items are the two \"\" issues, but there is still no evidence they describe the same underlying bug or change beyond identical placeholder titles, so I would not merge any pair on that basis alone.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44371 is the earliest exact \"\" placeholder entry and is as representative as any other low-information spam issue in this cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44572 is the best available issue anchor for triage, but only as a placeholder: it is one of the exact '' duplicates and there is no meaningful technical content to prefer any other issue.", + "best_issue_reason": "44371 is the cleanest canonical representative because it is an exact placeholder title and the earliest of the identical spam-like issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are generic spam-title issues with no body/context; identical wording alone is not enough to prove the same underlying report." + "reason": "Both are titled \"\", but that is not enough to establish the same underlying bug or change; these look like separate spam placeholders." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are generic spam-title issues with no substantive evidence of the same bug or request." + "reason": "Identical placeholder titles alone are insufficient to confirm duplication of the same concrete problem." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is '' and the other is a generic documentation suggestion; no concrete shared issue is visible." + "reason": "One is generic spam and the other is a vague documentation suggestion; no concrete shared issue is visible." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "'Check out \"Google LLC\"' is unrelated spam-like content; no shared underlying problem with ''." + "reason": "The titles are unrelated; there is no indication they describe the same bug or requested change." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Documentation-enhancement wording does not match the spam issue in any concrete way." + "reason": "A vague docs suggestion and a spam placeholder do not appear to be the same underlying item." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "A documentation suggestion and a spam placeholder are not the same underlying issue." + "reason": "A documentation enhancement request is not the same as an unrelated spam placeholder." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "They have the same spam-like title, but there is no actual issue description or target to confirm a true duplicate report." + "reason": "Although both are titled \"\", there is no evidence they refer to the same underlying report rather than two separate spam entries." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "The titles indicate different spam content and no shared technical defect." + "reason": "The titles do not match and there is no shared concrete problem or change target." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Different spam-like topics with no evidence of the same underlying bug or change." + "reason": "No meaningful overlap beyond being closed issues; they do not look like duplicates." } ] }, @@ -18634,8 +18650,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45145-2", "nodes": [ @@ -18645,46 +18661,46 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "49aac96638fd4cc54a3e7965deb168173b9905e7a1f28b65025580cc3be38450", + "prepared_review_unit_hash": "ccc52389b5c8a5d1d1f9864401cbd918f5aa7bc7727bea700c4a80a3886c7a12", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1644, - "estimated_input_tokens": 411, - "estimated_eval_tokens": 1078 + "serialized_chars": 1646, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1644, - "estimated_input_tokens": 411, - "estimated_eval_tokens": 1078 + "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": "Two closed issues with near-identical titles and highly similar bodies, clearly describing the same energy-optimization writeup/topic. They should be treated as duplicates, with the earlier and slightly more referenced issue as canonical.", - "confidence": 0.99, - "canonical_issue_reason": "Issue #45145 was created earlier and already has an inbound reference, so it is the better canonical record for the same duplicated topic.", + "summary": "The two issues have the same title and nearly identical metadata, so they are very likely duplicate reports of the same energy-optimization proposal. Issue #45145 is the better canonical issue because it was created first and has an inbound reference.", + "confidence": 0.98, + "canonical_issue_reason": "#45145 was opened first and appears to be the original report; #45151 is effectively the same issue with no distinguishing links or files.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #45145 is the best issue representative because it is the earliest instance and has marginally more reference activity.", + "best_issue_reason": "#45145 is the strongest representative because it is earlier and slightly better connected (it has an inbound reference).", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", "accept": true, - "reason": "Same title, same subject matter, and highly similar body content; these look like duplicate reports of the same underlying issue/topic." + "reason": "Same title, same topic, and no evidence of distinct scope; these look like duplicate copies of the same issue." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is grounded in the packet: the titles are identical, the bodies are similar in length, and the soft-similarity signal supports treating them as duplicates. The canonical choice is also conservative (earlier issue with one inbound reference). One small caution: avoid overstating body-content certainty since the packet does not expose the full bodies here." + "feedback": "The conclusion is grounded in the packet: the issues have identical titles, close creation times, no distinct links/files, and one has only a slight inbound-reference advantage. The duplicate/ canonical-candidate call is conservative enough given the limited evidence, and the soft-edge accept does not overstate beyond the available metadata." }, "evaluator_used": true, "retried": false, @@ -18754,8 +18770,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45672-2", "nodes": [ @@ -18765,7 +18781,7 @@ "soft_pairs": [ "issue:45672|issue:45674" ], - "prepared_review_unit_hash": "01eabb7a2406b21eb5a70fe19ed9beda314b82ec002ba6a512cd3e9f98bedf69", + "prepared_review_unit_hash": "7a3b6b2ef07228a7b3922690a274594ce9fbae157b9f41cb05b1f97e9c3a2bae", "component_budget": { "node_count": 2, "item_count": 2, @@ -18787,18 +18803,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: both involve 4-bit inference, but one is a Gemma4/audio-module uint8 dtype crash and the other is a BitsAndBytesConfig regression where llm_int8_skip_modules removes the default lm_head exclusion.", - "confidence": 0.9, - "canonical_issue_reason": "Issue 45674 is the better canonical representative because it describes a specific, reproducible configuration bug in BitsAndBytesConfig with a clear failure mode and broader applicability than the Gemma4-specific dtype crash.", + "summary": "These are two separate 4-bit inference bugs in the same general area, but they do not look like duplicates: one is a Gemma4 audio-module uint8/torch.finfo TypeError, the other is a BitsAndBytesConfig skip-modules regression that drops the default lm_head exclusion and triggers an AssertionError.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue is evident here because the issues describe different failure modes and different root causes.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45674 is the strongest issue to anchor the cluster if one must be chosen: it is more general, has a precise causal mechanism, and is less model-specific than 45672.", + "best_issue_reason": "issue 45674 is the clearest representative of the cluster\u2019s broader 4-bit inference/BitsAndBytes regression theme, with a more specific and actionable root-cause description.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Both are 4-bit inference failures, but the underlying bugs differ: 45672 is a torch.finfo(uint8) TypeError in Gemma4/audio modules, while 45674 is an AssertionError caused by llm_int8_skip_modules clearing the default lm_head exclusion in BitsAndBytesConfig." + "reason": "The reports concern different code paths and symptoms: Gemma4 audio modules failing in torch.finfo on uint8 weights versus BitsAndBytesConfig clearing the default lm_head exclusion and asserting in 4-bit inference. Similar subsystem, but not the same bug." } ] }, @@ -18817,8 +18833,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45762-2", "nodes": [ @@ -18828,7 +18844,7 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "79b17fe319b3d341192f7fca9320edb56a5fda66a8838aa41bc3130a0ff8f46f", + "prepared_review_unit_hash": "cc53002243c7fafdf54318f45149fca1c891d85abf4cfade38755b4481ed1546", "component_budget": { "node_count": 2, "item_count": 2, @@ -18850,18 +18866,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues are unrelated one-line reports containing different personal-name/date strings; there is no evidence they describe the same bug or request.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue: the issues do not share a concrete problem statement or context, so they should not be treated as duplicates.", + "summary": "Both items are bare-name/date issues with no body or technical context, so they do not appear to describe the same bug or change and are not good canonical triage candidates.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No best representative for a duplicate cluster: neither issue appears to be a duplicate of the other, so there is no meaningful canonical issue to select.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Reject: the titles are different personal-name/date strings with no shared bug, feature, or code context, so they do not look like the same underlying issue." + "reason": "Titles are different personal-name entries with different dates and no shared bug report content; likely unrelated/spam rather than the same underlying issue." } ] }, @@ -18880,8 +18896,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45764-2", "nodes": [ @@ -18891,40 +18907,40 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "89d469ab2f72b5f340245d533b1fbbc40eb7b2b2ca00f5a75459fa92a853488c", + "prepared_review_unit_hash": "9c051256cc1dca01ffc7e8a478879cf42bd3fc9cf8e3c9582ae8054c381e477a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1481, - "estimated_input_tokens": 371, - "estimated_eval_tokens": 998 + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1481, - "estimated_input_tokens": 371, - "estimated_eval_tokens": 998 + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues do not appear to describe the same bug or request; they are separate, content-free entries with different personal-name titles. The soft similarity is not enough to treat them as duplicates.", + "summary": "The two issues do not look like duplicates; they appear to be unrelated name/date entries rather than the same bug or request. No PRs are present.", "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good canonical representative because neither contains any real problem description, repro, or actionable content.", + "best_issue_reason": "No single issue is a plausible canonical representative here; the items do not share a clear underlying problem or change.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "Only superficial similarity in title format; the names and dates differ and there is no shared bug/report content to indicate the same underlying issue." + "reason": "The titles are unrelated personal-name/date entries with no shared subject, code path, or explicit linkage; this does not look like the same underlying issue." } ] }, @@ -18943,8 +18959,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45810-2", "nodes": [ @@ -18954,7 +18970,7 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "44a66418904ef2aa6154220b04e28c09f8feadc2b0544e69d6856a3a5bb0baf6", + "prepared_review_unit_hash": "dd379e76b76a6de5d7299e6a03345af513ec3f51f706e2aa32b25de36259d837", "component_budget": { "node_count": 2, "item_count": 2, @@ -18976,18 +18992,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one requests a new token-classification model class, while the other requests FP8 training support. They address different code paths and different user needs.", - "confidence": 0.97, + "summary": "These are two unrelated feature requests: one asks for a new Qwen3_5 token-classification/value-model class, the other asks for FP8 training support. They should not be clustered together.", + "confidence": 0.99, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue #45810 is the more concrete, narrowly scoped feature request and is a better representative issue for this set, though it is still unrelated to #46295.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different requests in different areas: model class addition vs FP8 training support. They do not describe the same underlying bug or change." + "reason": "Different underlying requests and code changes: model/class support for token classification vs numerical precision/training support. No shared target or concrete overlapping bug/change." } ] }, @@ -19006,8 +19022,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45869-5", "nodes": [ @@ -19027,91 +19043,91 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "d31f0c1f20bdc44990fca5675a43ca11b534e8871111b2c8c21b902180ef21cf", + "prepared_review_unit_hash": "04242ae3784a557f29efbe57071946174e149fc4fc7387eb789313cb17d81950", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5723, - "estimated_input_tokens": 1431, - "estimated_eval_tokens": 3118 + "serialized_chars": 5725, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5723, - "estimated_input_tokens": 1431, - "estimated_eval_tokens": 3118 + "serialized_chars": 5725, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster splits into two likely duplicate issue pairs (45870/45871 and 45872/45873) plus a separate dataset promo issue (45869). The duplicate calls should be conservative: same title and identical body length make them likely duplicates, but the packet does not prove exact match.", - "confidence": 0.89, - "canonical_issue_reason": "No single cluster-wide canonical issue: the evidence supports two separate likely-duplicate pairs, while 45869 is a different topic. Treat the duplicates pairwise rather than collapsing all five into one set.", + "summary": "There are two separate exact-duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is about a dataset bundle and does not look like the same underlying request as the DataForge example/service issues.", + "confidence": 0.98, + "canonical_issue_reason": "45870 is a strong canonical representative for the 'Example: Financial time series prediction with DataForge' pair because it is an exact duplicate of 45871 and is the earlier issue; 45872/45873 form a separate duplicate pair, so there is no single canonical issue for all five items.", "canonical_pr_reason": null, - "best_issue_reason": "If forced to choose one representative, 45870 is a reasonable pick because it is the earlier issue in one of the likely-duplicate pairs; however, there is no true cluster-wide best because the cluster contains multiple topics.", + "best_issue_reason": "45870 is the best overall issue to keep as the representative of its duplicate pair: it is earliest, fully duplicated by 45871, and its title is stable and specific. The other items are either a separate duplicate pair (45872/45873) or a different topic (45869).", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Dataset bundle promo versus a DataForge example issue; same general finance theme, but not the same underlying request." + "reason": "Different topics: a financial data bundle promo versus a DataForge example issue; similar keywords but not the same underlying request." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Related finance theme only; the dataset promo is a different issue from the DataForge example post." + "reason": "Different topics: dataset bundle promotion versus DataForge example; not the same bug/change." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Same title and identical body length, so these look like likely duplicates, though the body content is not visible here." + "reason": "Exact duplicate title and matching metadata/body length indicate the same underlying issue." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Different issue topic: example prediction post versus scraping-service request." + "reason": "Both mention financial time series/DataForge, but one is an example issue and the other is a scraping-service request; not the same underlying item." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Same broad DataForge framing, but the concrete request differs from the example-prediction issue." + "reason": "Same as 45870/45872: related theme, but different request and not the same issue." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Different concrete topic: example prediction post versus scraping-service request." + "reason": "Different concrete request: example prediction issue versus scraping-service issue." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Same broad theme only; not the same underlying issue as the example-prediction pair." + "reason": "Different concrete request: example prediction issue versus scraping-service issue." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Same title and identical body length, so these look like likely duplicates, though the body content is not visible here." + "reason": "Exact duplicate title and matching metadata/body length indicate the same underlying issue." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster-level split is plausible, but the analyst overstates the evidence. The packet only shows titles, body lengths, and soft-similarity scores; it does not establish \"exact duplicate\" or \"matching issue content.\" The two accepted soft edges (45870/45871 and 45872/45873) should be stated more conservatively as likely duplicates based on identical titles/body lengths, not exact matches. Because the verdicts are not conservative enough, reject." + "accept": true, + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 have identical titles and matching body lengths, while 45869 is clearly a different topic. The soft-edge verdicts are conservative, with no obvious overstatement beyond the evidence available." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -19124,8 +19140,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-46439-2", "nodes": [ @@ -19135,7 +19151,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "a82533d9dd6b65b73782c4bb8512f57cd7f666424af3664cc394164891420a8d", + "prepared_review_unit_hash": "366a0731ebf344bedb4ce1324517c0534f5993c8cbb193dd4b2cd1e1e6e61275", "component_budget": { "node_count": 2, "item_count": 2, @@ -19157,27 +19173,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "#46439 is the better canonical issue because it describes the broader static-cache/linear-attention failure mode, while #46441 is a narrower Qwen3.5 symptom report. The two look possibly related, but the evidence is too weak to treat them as the same bug.", - "confidence": 0.8, - "canonical_issue_reason": "issue #46439 is the broader root-cause report: static cache `early_initialization` corrupting linear-attention layers with wrong shapes and runtime errors.", + "summary": "Both issues concern static-cache failures around linear-attention models, but they describe different failure modes: one is shape corruption during `early_initialization`, the other is a `KeyError` for `linear_attention` on Qwen3.5. They are related in subsystem, but not the same underlying bug.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it describes a more general and concrete static-cache code-path failure with a clear runtime consequence and root area (`early_initialization` / shape corruption).", "canonical_pr_reason": null, - "best_issue_reason": "#46439 is the best representative because it captures the underlying bug generically, not just one model-specific manifestation.", + "best_issue_reason": "Issue 46439 is the strongest representative of the cluster because it is more broadly applicable and closer to an actionable engine-level bug than the Qwen3.5-specific `KeyError` report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "This pair is only supported by `soft_similarity`; there are no shared filenames or explicit targets, and `deterministic_accept` is false. #46441 may be a related manifestation, but the evidence is not strong enough to confirm the same underlying bug." + "reason": "Related static-cache/linear-attention breakage, but the symptoms differ materially: wrong tensor shapes in early initialization vs a missing `linear_attention` key on Qwen3.5. Not the same concrete bug." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The canonical issue choice is plausible, but the soft-edge acceptance is too strong for the packet. The only explicit pair evidence is `soft_similarity` with no shared filenames/targets and `deterministic_accept: false`, so saying the issues \u201cpoint to the same\u201d bug overstates what\u2019s grounded here. A more conservative verdict would be to note possible relatedness without accepting the edge." - }, - "evaluator_used": true, - "retried": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -19231,154 +19244,154 @@ "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "6530791bb1cc72949c01baee523d6c52194a7b89a5959792d13f0694a395ac7a", + "prepared_review_unit_hash": "fa1b7312c5d72dec8367e717985896f78bffc8022eaee6a6b9b2500cfd97cfad", "component_budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 19, - "serialized_chars": 38084, - "estimated_input_tokens": 9521, - "estimated_eval_tokens": 19298 + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 }, "budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 19, - "serialized_chars": 38084, - "estimated_input_tokens": 9521, - "estimated_eval_tokens": 19298 + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly unrelated PRs spanning new model additions, refactors, tests, and docs. The only true duplicate-like pair is the two `fixed_cross_entropy` PRs, which target the same issue and the same code path.", - "confidence": 0.95, + "summary": "One true duplicate pair stands out: the two `loss_utils.py` PRs both extend `fixed_cross_entropy` with the same new arguments on the same code path. The rest of the cluster is a grab bag of unrelated model additions, test fixes, docs, and refactors.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43254 is the cleanest canonical representative for the loss-helper fix: it directly adds supported kwargs to `fixed_cross_entropy` in a single focused change tied to issue #43240. #43251 is essentially the same fix with a slightly different signature style.", + "canonical_pr_reason": "PR 43251 is the clearest canonical: it is issue-linked to 43240, edits the target loss helper directly, and presents the fuller/primary implementation of the cross-entropy kwargs fix.", "best_issue_reason": null, - "best_pr_reason": "PR #43254 is the best global PR candidate because it is the most direct, self-contained implementation of the underlying bug fix and is easier to merge or supersede than the alternate variant.", + "best_pr_reason": "PR 43251 is the best overall representative because it is the most directly scoped fix in the cluster and the only one with an explicit issue target; it also subsumes the near-identical 43254 change.", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "Unrelated: VideoPrism model addition vs an Executorch dynamic-shapes test." + "reason": "Unrelated PRs: VideoPrism model/docs/weights conversion vs ExecuTorch dynamic-shape export tests. Different subsystems and code paths." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Tokenizer method addition vs adding a new MobileLLM model; different changes." + "reason": "Different artifacts: Mistral tokenizer method addition vs adding the MobileLLM model. No shared bug or mergeable code-path overlap." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Mistral tokenizer method fix and Executorch export test are unrelated code paths." + "reason": "Tokenizer API change vs ExecuTorch test coverage; these do not fix the same underlying issue." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness vs a new model implementation; not the same bug or change." + "reason": "Processor auto-loading robustness changes vs adding a new model implementation. Same broad area of transformers, but not the same bug or change." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss helper kwargs fix vs MobileLLM model addition; no shared underlying issue." + "reason": "Loss helper kwargs plumbing and a new model addition are unrelated changes." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Cross-entropy helper fix is unrelated to Falcon/Mistral4 test and refactor work." + "reason": "Cross-entropy helper fix vs Mistral4 test/model refactor. No shared concrete code-path problem." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Executorch dynamic-shape test vs Mistral4 test fixes; different functionality." + "reason": "ExecuTorch export test vs Mistral4 test/model changes; both are tests/infra-heavy, but they address different issues." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Executorch export testing and MobileLLM model addition are unrelated." + "reason": "Dynamic-shape export test coverage and MobileLLM model addition are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New MobileLLM model vs config-resolution ordering fix; different problems." + "reason": "MobileLLM model addition vs fixing config resolution order in `modeling_utils.py`; different code paths and goals." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "New model implementation vs Falcon output-interface refactor; not the same change." + "reason": "New model implementation vs Falcon output-interface refactor; no common bug or combined PR candidate." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor vs a model-creation skill/meta PR; no shared code-path bug." + "reason": "Falcon refactor vs a repo skill/documentation PR. Not the same underlying change." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon output refactor and Mistral4 test fixes are separate model-specific changes." + "reason": "Falcon output-collection refactor and Mistral4 fixes are distinct model-specific changes." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test fixes vs CLIP-like `layrnorm` rename; unrelated subsystems." + "reason": "Mistral4 test/model work vs a CLIP-family typo rename/mapping fix; different model families and bugs." }, { "left": "pull_request:44827", "right": "pull_request:46267", "accept": false, - "reason": "Mistral4 test fixes vs chat parsing/serve update; different areas." + "reason": "Mistral4-specific fixes vs chat parsing/serve updates; unrelated functionality." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 test fixes and TDT loss kernel integration are unrelated." + "reason": "Mistral4 changes vs adding a TDT loss kernel integration; distinct code paths and concerns." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test fixes vs config resolution order fix; different code paths." + "reason": "Mistral4 model/test updates and config-resolution ordering fix are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Shared auto/modeling filenames are superficial; the PRs address unrelated features." + "reason": "Mistral4 fixes vs model-creation skill/docs; not the same bug/change." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "MobileLLM model addition vs model-creation skill/meta changes; not duplicates." + "reason": "Adding MobileLLM and adding a model-creation skill are different kinds of changes with no shared code-path issue." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Same underlying bug: `fixed_cross_entropy` needed to forward supported kwargs to `nn.functional.cross_entropy`, in the same file and with the same issue target #43240." + "reason": "Same file, same helper, same external behavior change: both extend `fixed_cross_entropy` to pass supported kwargs to `nn.functional.cross_entropy`. One is a near-duplicate variant of the same fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative: it correctly identifies the single shared-issue duplicate-like pair (`43251`/`43254`) and treats the remaining PR pairs as unrelated on the evidence shown. No soft-edge verdict looks overstated, and the accepted pair is plausibly mergeable as the same bugfix." + "feedback": "Grounded and conservative overall. The only accepted pair, 43251/43254, is strongly supported by the packet: same file, same helper, same issue target, and near-identical kwargs plumbing. The other verdicts stay safely on the conservative side by rejecting pairs that are clearly different subsystems or bugfixes." }, "evaluator_used": true, "retried": false, @@ -19407,46 +19420,46 @@ "pull_request:40546|pull_request:42781", "pull_request:42781|pull_request:46331" ], - "prepared_review_unit_hash": "695ba08c3102bf8e3de92d0b2ead5719b0450ea878a412824653f5ee9a6f24c6", + "prepared_review_unit_hash": "c90d1d6a85fbfdba61b56a41ffab7bb9eeed402e89f9e3cdcd459b91bb018cf1", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10867, - "estimated_input_tokens": 2717, - "estimated_eval_tokens": 5690 + "serialized_chars": 10855, + "estimated_input_tokens": 2714, + "estimated_eval_tokens": 5684 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10867, - "estimated_input_tokens": 2717, - "estimated_eval_tokens": 5690 + "serialized_chars": 10855, + "estimated_input_tokens": 2714, + "estimated_eval_tokens": 5684 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related only at a broad model-family or subsystem level; they do not appear to fix the same concrete bug or implement the same exact change. The VibeVoice PRs are separate feature additions, and the Parakeet PR is a different model entirely.", - "confidence": 0.86, + "summary": "This cluster is VibeVoice-centric, but the soft edges are not true duplicates: 40546 is the base VibeVoice implementation, 42781 is a later realtime/tokenizer extension, and 46331 is a separate Parakeet-RNNT PR that only overlaps on generic auto-registration files.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42781 is the best representative because it is the broadest VibeVoice-related change set here and adds the realtime variant plus the acoustic tokenizer alongside the core model integration.", + "canonical_pr_reason": "PR 40546 is the core VibeVoice implementation with the broadest base model/docs/auto wiring, so it is the cleanest cluster anchor.", "best_issue_reason": null, - "best_pr_reason": "PR 42781 is the strongest cluster representative by scope and completeness, but it still should not be merged with the other PRs as a duplicate.", + "best_pr_reason": "PR 40546 is the best representative of the cluster because it is the primary VibeVoice implementation; 42781 is a follow-on realtime feature set, and 46331 is unrelated.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Both are VibeVoice-related, but one implements the base model and the other adds the realtime variant/acoustic tokenizer; they are distinct feature changes and not the same concrete PR." + "reason": "Related VibeVoice work, but not the same underlying change: 40546 adds the base model, while 42781 adds realtime/acoustic-tokenizer support and other extra plumbing, so they would not plausibly merge as one duplicate PR." }, { "left": "pull_request:42781", "right": "pull_request:46331", "accept": false, - "reason": "These target different model families (VibeVoice vs Parakeet). Shared auto-mapping files are incidental, not evidence of the same underlying change." + "reason": "Different model families and code paths: 42781 is VibeVoice realtime, while 46331 is Parakeet-RNNT; shared auto-file edits are generic and not enough to treat them as the same change." } ] }, @@ -19476,12 +19489,12 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "d7a7ace2a88e9af4f74a49b7e70263c07a60b057ca1ee226a665405ef720d0d6", + "prepared_review_unit_hash": "99d3314828221bd1f910a894967781fdeb84a6c2f837b397684f0fb3613856a6", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4674, + "serialized_chars": 4675, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -19489,7 +19502,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4674, + "serialized_chars": 4675, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -19498,18 +19511,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests 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 clustered as duplicates.", - "confidence": 0.98, + "summary": "The two PRs are unrelated: one adds TimesFM covariate forecasting support, the other adds DINOv3 image classification support. No duplicate triage match here.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: these are unrelated changes affecting different model families and features.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "Neither PR is a duplicate of the other; if forced to pick a representative, the DINOv3 PR is the narrower, self-contained feature addition.", + "best_pr_reason": "pull_request:41224 is the more substantial standalone PR in this set, with a concrete model-support change and more review activity, but it is not a duplicate of pull_request:40755.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different domains and code paths: TimesFM forecasting with covariates vs DINOv3 image classification. Shared similarity is only broad ML/model docs, not the same underlying bug or change." + "reason": "Different models, different code paths, and different feature goals: TimesFM forecasting with covariates vs DINOv3 image classification support. They do not look mergeable into one PR." } ] }, @@ -19604,12 +19617,12 @@ "pull_request:41797|pull_request:45186", "pull_request:45186|pull_request:45630" ], - "prepared_review_unit_hash": "db2f581d54bcc1161c86cabf99725499942b2d82fa17a9240c9324881cb18c93", + "prepared_review_unit_hash": "781cfc97896dc60a6f2df0187645d455c39f6b37b79a9fff0c900dd5fbfe0a38", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10436, + "serialized_chars": 10434, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -19617,7 +19630,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10436, + "serialized_chars": 10434, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -19626,24 +19639,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three separate model-addition PRs (DeepSeek OCR, Isaac, Kimi2-6) with similar Transformers boilerplate, but they target different models and different codebases, so they are not duplicates.", + "summary": "These are three unrelated model-addition pull requests that only overlap on the standard Transformers scaffolding for new models. They are not duplicates and should not be merged together.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45630 is the strongest representative of the cluster because it is the most recent and most extensively discussed, but it is still a distinct model integration rather than a duplicate of the others.", + "canonical_pr_reason": "No canonical PR in this set: each pull request adds a different model family (DeepSeekOCR, Isaac, Kimi2-6) with distinct configs, processors, and tests.", "best_issue_reason": null, - "best_pr_reason": "PR #45630 is the best single PR to keep as the representative item since it has the highest review activity and the broadest implementation set, while the others are separate model additions.", + "best_pr_reason": "PR 45630 is the most active and broadest of the three, but it is not a better duplicate representative because it targets a different model than the others.", "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "Different models and different implementations: DeepSeek OCR vs Isaac. Shared auto/docs scaffolding is expected for model integrations and does not indicate the same underlying change." + "reason": "Different model implementations and docs: DeepSeekOCR vs Isaac. Shared framework files are just normal new-model registration, not the same underlying change." }, { "left": "pull_request:45186", "right": "pull_request:45630", "accept": false, - "reason": "Different model additions: Isaac vs Kimi2-6. They share Hugging Face registration boilerplate, but they are not fixing the same code-path problem and should not be merged as duplicates." + "reason": "Different model additions with different code paths and assets: Isaac vs Kimi2-6. Similar filenames reflect boilerplate for model onboarding, not a duplicate fix." } ] }, @@ -19673,7 +19686,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "6be76ce29fb2d5d935d6f7d81edfa74b1909daddf5b659a8e38ab184b998bb99", + "prepared_review_unit_hash": "7eca033038917baa9d7f82bf4618da6a5c187e5b764116f4c6f8b1d8fb4b9aa0", "component_budget": { "node_count": 2, "item_count": 2, @@ -19695,18 +19708,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs both update GitHub Actions versions, but they target different action families and different change sets: one is a broad Node 24 compatibility sweep, the other is a docker/build-push-action upgrade. They should not be treated as duplicates.", + "summary": "Both items are GitHub Actions dependency bumps, but they target different action families and different scopes. PR 42975 is a broad workflow-tooling update for Node 24 compatibility, while PR 42976 only bumps docker/build-push-action in several Docker image workflows. They are related, but not the same underlying change.", "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #42975 is the broader representative: it updates multiple workflow actions for Node 24 compatibility across many files, whereas #42976 is a narrower docker/build-push-action version bump.", + "canonical_pr_reason": "PR 42975 is the broader, more representative workflow-action upgrade and has the clearest stated purpose (Node 24 compatibility); PR 42976 is a narrower, separate action-version bump.", "best_issue_reason": null, - "best_pr_reason": "PR #42975 is the best cluster representative because its title and diff most clearly describe the general GitHub Actions upgrade theme and wider impact.", + "best_pr_reason": "PR 42975 is the best representative artifact for this cluster because it covers the wider GitHub Actions compatibility migration rather than a single action upgrade.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "They share workflow files, but the concrete changes are different: #42975 upgrades checkout/cache/upload-artifact for Node 24 compatibility, while #42976 upgrades docker/build-push-action to v6. Different action families, different fix, not one mergeable PR." + "reason": "Reject: both are workflow maintenance, but they update different actions (general GitHub Actions/runtime compatibility vs docker/build-push-action) and do not appear to fix the same concrete problem or belong in one merged PR." } ] }, @@ -19743,70 +19756,70 @@ "pull_request:43448|pull_request:45490", "pull_request:43451|pull_request:45490" ], - "prepared_review_unit_hash": "8399333aa779fbd9293d204931aab73d795f9027438a38cd09bf673bb861a9ac", + "prepared_review_unit_hash": "9dc2d25618202401aeec09219d72cff8990a5e1c675f2a2437d7405a222229d0", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 14253, - "estimated_input_tokens": 3564, - "estimated_eval_tokens": 7384 + "serialized_chars": 14247, + "estimated_input_tokens": 3562, + "estimated_eval_tokens": 7380 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 14253, - "estimated_input_tokens": 3564, - "estimated_eval_tokens": 7384 + "serialized_chars": 14247, + "estimated_input_tokens": 3562, + "estimated_eval_tokens": 7380 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four pull requests are unrelated model-addition PRs that happen to touch shared infrastructure (docs and auto-mapping files). They are not duplicates and should be triaged separately.", - "confidence": 0.98, + "summary": "These are four separate model-addition PRs (ViT NEPA, Molmo, Molmo2, CTSM), not duplicates. The shared docs/auto-registry edits are expected for new model integrations and do not indicate the same underlying change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43451 (Add Molmo2) is the strongest representative of the cluster only in the sense that it is the most substantial and well-scoped PR with the highest review activity; however, it is still a distinct model addition, not a duplicate of the others.", + "canonical_pr_reason": "No canonical PR: each PR adds a different model family with distinct files, configs, and modeling code, so they are not the same change.", "best_issue_reason": null, - "best_pr_reason": "43451 is the most complete and central-looking PR among these, with the broadest model-specific implementation surface and the most discussion, so it is the best single representative if one must be chosen.", + "best_pr_reason": "No single best PR for deduplication; all are legitimate but unrelated integrations, so none should represent the cluster.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "ViT NEPA and Molmo are different models with different codepaths and files; overlap is limited to shared framework plumbing, not the same bug/change." + "reason": "Different model families (ViT NEPA vs Molmo) with distinct modeling and test code; shared auto/docs updates are only boilerplate for registration." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "ViT NEPA vs Molmo2 are separate model additions; shared auto/docs edits do not indicate the same underlying implementation work." + "reason": "ViT NEPA and Molmo2 are unrelated model additions; they do not fix the same code-path problem." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "ViT NEPA and CTSM are unrelated domains and architectures, so this is not the same change." + "reason": "CTSM time-series integration is a different subsystem and architecture from ViT NEPA, so these are not duplicates." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo and Molmo2 are distinct model families and the diffs add different processors, configs, and model code; they are not one mergeable fix." + "reason": "Molmo and Molmo2 are related by name but are separate model integrations with different files and functionality, not one mergeable duplicate PR." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are unrelated model additions with different implementation targets." + "reason": "Molmo and CTSM are unrelated model additions; overlapping registry/docs edits are not enough to treat them as the same change." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM affect different model stacks and codepaths; shared auto-mapping edits are incidental." + "reason": "Molmo2 and CTSM address different models and code paths, so they should not be merged as duplicates." } ] }, @@ -19843,7 +19856,7 @@ "pull_request:43665|pull_request:46380", "pull_request:44815|pull_request:46380" ], - "prepared_review_unit_hash": "103f069f88922f109becc9b01a85bbcaf6753b9cc430bb0bbd7b0dee9c041ea3", + "prepared_review_unit_hash": "6a10945238272666b9a87d5e57272e7010c85c7fdf77577fdb776d01844bcbfa", "component_budget": { "node_count": 4, "item_count": 4, @@ -19865,48 +19878,48 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are mostly unrelated despite generic \"fix\" titles: one Whisper tokenizer test/code change, one test-only CLIP/ViT patch, one FP8 dequantization fix, and one generation/watermarking validation fix. No pair looks like the same concrete bug/change, so all soft edges should be rejected.", - "confidence": 0.28, + "summary": "These PRs are not duplicates: they touch different subsystems and fix different problems. 46380 and 44815 are the strongest real bug-fix candidates, while 43665 looks like test-only noise.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46380 is the strongest standalone bug fix in the set because it changes production generation/validation code with a clear concrete failure mode; the others are either test-only or address different subsystems.", + "canonical_pr_reason": "No issue links are present, and the PRs are unrelated enough that there is no true canonical duplicate. If one PR must represent the cluster, 46380 is the clearest concrete bug fix.", "best_issue_reason": null, - "best_pr_reason": "PR 46380 is the best representative of the cluster as a substantive fix with a clear code-path issue, but it is not a duplicate of the other PRs.", + "best_pr_reason": "46380 is the best representative overall: it fixes a specific runtime validation bug in generation config handling, with a coherent code-path and test coverage. 44815 is also substantive, but it targets a different FP8/dequantization path.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Different changes: Whisper tokenizer decode logic vs CLIP/ViT test edits; no shared bug or code path." + "reason": "Whisper tokenizer decode/timestamp logic vs. CLIP/ViT test changes; different models, different bug classes, and 43665 appears to be test-only." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper tokenization/tests are unrelated to FP8 dequantization and model-loading changes." + "reason": "Whisper decoding behavior is unrelated to FP8 dequantization and model-loading changes." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "Whisper decode handling and generation watermarking validation are separate subsystems and bugs." + "reason": "Whisper tokenizer code vs. generation watermarking/config validation; distinct code paths and fixes." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Test-only CLIP/ViT patch does not match the FP8 dequantization/runtime fix." + "reason": "Test assertions in CLIP/ViT are unrelated to FP8 kernel/model-loading behavior." }, { "left": "pull_request:43665", "right": "pull_request:46380", "accept": false, - "reason": "CLIP/ViT test edits are not the same underlying issue as generation watermarking/config validation." + "reason": "Test-only edits for CLIP/ViT do not match generation config validation in 46380." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "Both are fixes, but they target different concrete problems in FP8 quantization vs generation watermarking/loading validation." + "reason": "FP8 dequant/padding and model-loading changes are unrelated to generation config watermarking validation." } ] }, @@ -19936,7 +19949,7 @@ "soft_pairs": [ "pull_request:43488|pull_request:43532" ], - "prepared_review_unit_hash": "6b2c6326e79bf2b75fa2fd8c34a2f761e29e1485d130a2d612cff39f9c9b5be4", + "prepared_review_unit_hash": "b3d449f812f0369e4dad745f57d716e1603376d008038cf26a6c3cd31682eedd", "component_budget": { "node_count": 2, "item_count": 2, @@ -19958,18 +19971,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two pull requests are not duplicates: one is a small repo-bot/version/dependency test tweak, while the other rewires CI/workflow behavior and touches unrelated utility/test files. Their only overlap is a shared Vit test file, which is not enough to indicate the same underlying fix.", - "confidence": 0.97, + "summary": "The two pull requests are not duplicates: one is a small repo-bot/test formatting probe, while the other changes CI workflows and related utility code. The only overlap is a shared test file, which is not enough to merge them as the same change.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs target different code paths and purposes, with no evidence they fix the same concrete bug or implement the same change.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "If a representative is needed, PR 43532 is the better cluster anchor because it is the more substantial end-to-end workflow/utility change; PR 43488 is a narrow bot-check/test setup tweak.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "Reject: they address different changesets and code paths; the shared Vit test file is incidental, not evidence of the same fix." + "reason": "Reject: these are unrelated PRs. 43488 only tweaks setup/versioning and a test hook, while 43532 rewrites CI workflow logic and touches multiple supporting files. Shared test filename alone does not indicate the same underlying change." } ] }, @@ -19999,12 +20012,12 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "560853986ec4bab3b1b6d84bb738c09ea36ee0599d2bc0dfc5f09dfbaac011b9", + "prepared_review_unit_hash": "a80accd323244a28022ca51524a27be0753c81477b0a4362023f3804552339ab", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6083, + "serialized_chars": 6082, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -20012,7 +20025,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6083, + "serialized_chars": 6082, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -20021,18 +20034,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two related but distinct feature PRs: one adds a Promptable Concept Segmentation pipeline and the other adds a Promptable Visual Segmentation pipeline. They share scaffolding files, but they target different pipelines, model integrations, and tests, so they should not be deduplicated.", - "confidence": 0.96, + "summary": "Two open PRs add similar new segmentation pipelines and share doc/registry/test plumbing, but they target different features and code paths: Promptable Concept Segmentation vs Promptable Visual Segmentation. They should not be merged as duplicates.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: the two pull requests implement different segmentation features and should remain separate.", + "canonical_pr_reason": "PR #43613 is the broader and more mature change set: it covers Promptable Visual Segmentation and integrates more model families (SAM, SAM2, EdgeTAM, SAM3 tracker), so it is the better representative of this cluster if one must be chosen.", "best_issue_reason": null, - "best_pr_reason": "No global best PR for duplicate consolidation; neither PR is a duplicate of the other, and each addresses a different pipeline addition.", + "best_pr_reason": "PR #43613 is the better anchor for the cluster because it is the more general segmentation pipeline addition and has the larger supporting surface area. That said, it is still a distinct feature from #43612, not the same underlying change.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different feature additions with different pipeline names, model files, and tests; they share project scaffolding but not the same concrete code-path change." + "reason": "These are different pipeline additions with different intended behaviors and model integrations. Shared docs/registration/test scaffolding is not enough to treat them as the same concrete change." } ] }, @@ -20062,40 +20075,40 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "ba14b2913c7e15f983fda48df2a228d3728d630395bf3ef33ccaabd0606353f7", + "prepared_review_unit_hash": "9de4cac0da3dc8dc546304bd9478fb01821dce859f6c35eadba3bf083d6fc4b4", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4292, - "estimated_input_tokens": 1073, - "estimated_eval_tokens": 2402 + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4292, - "estimated_input_tokens": 1073, - "estimated_eval_tokens": 2402 + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch `src/transformers/trainer.py`, but they address different trainer extension points: one changes gradient-accumulation loss scaling, the other factors out signature-column inspection for column filtering. They are not duplicates.", - "confidence": 0.97, + "summary": "Two Trainer.py extensibility PRs that look superficially similar but address different code paths: gradient-accumulation loss scaling vs. signature-column filtering.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43651 is the more concrete bugfix-oriented change: it adjusts training-step loss scaling behavior and adds `_loss_is_scaled_for_ga` to let custom trainers control that path.", + "canonical_pr_reason": "PR 43651 is the stronger representative because it makes a concrete runtime change to training_step and adds the new _loss_is_scaled_for_ga hook for custom trainers.", "best_issue_reason": null, - "best_pr_reason": "PR 43651 is the strongest standalone representative because it targets a specific runtime behavior change in `training_step`, whereas PR 43663 is a separate refactor/extension for signature-column handling.", + "best_pr_reason": "PR 43651 is the best overall PR in this cluster as it introduces a clearer, more self-contained Trainer behavior change with the broader extensibility hook.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Same file and similar trainer-extension theme, but different code paths and purposes: gradient accumulation loss scaling vs. signature column filtering. They would not plausibly be merged as one PR." + "reason": "Reject: they touch different Trainer internals and solve different problems. 43651 changes loss normalization in training_step; 43663 factors out signature-column inspection for column filtering. Shared filename and similar extensibility framing are not enough to treat them as the same change." } ] }, @@ -20125,7 +20138,7 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "b4305d4991b67fa328e79ca996a9f3074592ee7cba36752dd479a9a1bd91ec68", + "prepared_review_unit_hash": "10181352bff315da093a668b6561586643ebfa37711b1f85ca91ec2ac760d2f8", "component_budget": { "node_count": 2, "item_count": 2, @@ -20147,18 +20160,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same file and same symptom label, but they appear to fix different TypeAdapter annotation/runtime issues rather than the same concrete bug.", - "confidence": 0.85, + "summary": "Two open PRs touch the same CLI file and both mention NameError-ish typing issues, but they fix different concrete problems: one is about annotation/runtime typing in `serve.py`, the other is specifically about `TypeAdapter` when `pydantic` is absent.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:43842 is the more specific and directly targeted fix: it explicitly guards the TypeAdapter import under TYPE_CHECKING and quotes the annotation for the pydantic-absent path.", + "canonical_pr_reason": "No single canonical PR: 43656 targets a broader annotation/runtime typing cleanup, while 43842 specifically fixes the `TypeAdapter` import/type-guard issue when `pydantic` is not installed.", "best_issue_reason": null, - "best_pr_reason": "pull_request:43842 is the clearest representative of the cluster because its diff and title both match the concrete NameError scenario, whereas pull_request:43656 looks like a broader annotation/runtime typing cleanup in the same module.", + "best_pr_reason": "PR 43842 is the more focused and explicit bug fix: it names the failing `TypeAdapter` path, adds the guarded import, and has a narrower, clearer issue target.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Do not merge: both are in cli/serve.py and mention TypeAdapter NameError, but the diffs suggest different fixes\u201443656 is a broader future-annotations/type-hint cleanup, while 43842 specifically handles pydantic-not-installed imports. They are related, not the same underlying change." + "reason": "Reject: although both modify `src/transformers/cli/serve.py` and are related to typing/NameError problems, they address different runtime failures and are not the same underlying bug or mergeable change." } ] }, @@ -20350,12 +20363,12 @@ "pull_request:44044|pull_request:44129", "pull_request:44044|pull_request:44154" ], - "prepared_review_unit_hash": "340ba267cd45dd836af1c2e12e7849e60e222f2b0bd464fb6a762f8ea826c8b9", + "prepared_review_unit_hash": "158a1d3a071672acfe5c5dc8239d8f01ccd24567f936b7e429c0766ff8c218ef", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84895, + "serialized_chars": 84894, "estimated_input_tokens": 21224, "estimated_eval_tokens": 42704 }, @@ -20363,7 +20376,7 @@ "node_count": 27, "item_count": 27, "soft_pair_count": 72, - "serialized_chars": 75091, + "serialized_chars": 75090, "estimated_input_tokens": 18773, "estimated_eval_tokens": 37802 }, @@ -20372,339 +20385,351 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a wave of model-specific PRs migrating transformer outputs to the new capture/decorator tracing pattern, usually under umbrella issue #43979. The only clear duplicates are within the GPT-J/CodeGen family and the GPT-Neo family; the rest are same-theme but distinct model-specific refactors.", - "confidence": 0.94, - "canonical_issue_reason": null, - "canonical_pr_reason": "PR #44066 is the clearest representative of the shared change: it directly targets #43979 and applies the standardized output-tracing refactor on the central GPT-J/CodeGen path.", - "best_issue_reason": null, - "best_pr_reason": "#44066 is the best single PR to stand in for this cluster because it most directly captures the underlying standardized output-tracing change with explicit issue linkage and concrete code-path edits.", + "summary": "This cluster is mostly standardized output-tracing refactors spread across different model families. The clearest duplicate-like grouping is the GPT-J/CodeGen line around 44066/44085/44722; most other PRs are separate model-specific slices under umbrella issue #43979 rather than the same concrete change.", + "confidence": 0.79, + "canonical_issue_reason": "Issue #43979 is the shared umbrella tracker referenced by most PRs and best represents the underlying standardized output-tracing effort.", + "canonical_pr_reason": "PR #44722 is the strongest representative of the GPT-J/CodeGen subcluster: it is explicit, issue-linked, and spans the shared GPT-J code path plus the copied CodeGen block.", + "best_issue_reason": "As the common tracking issue, #43979 is the best issue-level anchor for the cluster; the individual PRs are implementation slices rather than separate problem statements.", + "best_pr_reason": "PR #44722 is the best single PR exemplar in this set because it is the most complete GPT-J/CodeGen-family refactor and most clearly aligns with the umbrella issue.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families (FNet/CVT vs GPT-J); same refactor theme but not the same concrete code-path change." + "reason": "Shared umbrella issue is not enough; FNet/CVT and GPT-J are different model code paths." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Both are output-tracing refactors, but they touch unrelated model code paths." + "reason": "FNet/CVT refactor is unrelated to DeBERTa; same issue target does not make them duplicates." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Same GPT-J output-tracing refactor on the same code path; 44085 looks like a narrower variant of the same change." + "reason": "Both touch GPT-J output-tracing in the same file, so they overlap on the same concrete code path and are plausibly mergeable." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "ResNet and EfficientNet are separate model implementations; this is the same pattern, not the same fix." + "reason": "RegNet and EfficientNet are separate model implementations, not the same bug/change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet and GPT-J are unrelated code paths despite the shared tracing refactor theme." + "reason": "EfficientNet output tracing and GPT-J/CodeGen output tracing are unrelated model-specific refactors." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "Different model families and different implementation files; not the same underlying bug/change." + "reason": "GPT-J/CodeGen and EfficientNet touch different code paths and are not mergeable duplicates." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTa-v2 are unrelated model code paths." + "reason": "MobileNetV2 and DeBERTa are unrelated implementations despite the shared umbrella issue." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen output tracing is unrelated to MGP-STR output tracing." + "reason": "GPT-J/CodeGen and MGP-STR are different models with different output paths." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen and MPT are separate model refactors, not one mergeable fix." + "reason": "GPT-J/CodeGen and MPT are separate model families; same refactor theme is too broad to merge." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo output-tracing refactor for the same model path; these look like duplicate variants of one change." + "reason": "Both are GPT-Neo output-tracing refactors on the same code path and are close enough to be duplicate change proposals." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen and GPT-Neo are different model code paths." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44066", - "accept": false, - "reason": "Different models and different code paths; shared tracing framework alone is not enough." + "reason": "GPT-J/CodeGen and GPT-Neo are different concrete implementations." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet output tracing and GPT-J output tracing are separate refactors." + "reason": "ResNet and GPT-J/CodeGen are unrelated model code paths." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "Different models; same broad tracing migration, but not the same concrete change." + "reason": "ResNet and GPT-J are not the same underlying change." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J are unrelated code paths." + "reason": "MGP-STR and GPT-J/CodeGen are separate models with no shared code path." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet are unrelated model implementations." + "reason": "GPT-Neo and ResNet are unrelated refactors." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet and MPT touch different model stacks, so these are not duplicates." + "reason": "ResNet and MPT are different model implementations." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are unrelated model changes." + "reason": "GPT-Neo and Flaubert/XLM are unrelated concrete code paths." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPT-J are different concrete refactors, even if they share the same output-tracing architecture." + "reason": "GPT-Neo and GPT-J/CodeGen are distinct model families." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are separate model-specific changes." + "reason": "GPT-Neo and Flaubert/XLM are unrelated refactors." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT and GPT-J are unrelated code paths." + "reason": "MPT and GPT-J/CodeGen are unrelated implementations." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo and GPT-J are different model-specific refactors." + "reason": "GPT-Neo and GPT-J are different model code paths." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and MGP-STR are unrelated model implementations." + "reason": "GPT-Neo and MGP-STR are unrelated model-specific changes." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo and MGP-STR are different code paths." + "reason": "GPT-Neo and MGP-STR do not share the same concrete bug/change." }, { - "left": "pull_request:44071", - "right": "pull_request:44086", + "left": "pull_request:44098", + "right": "pull_request:44101", "accept": false, - "reason": "MPT and MGP-STR are unrelated model changes." + "reason": "ViLT and XLM/Flaubert are different models; same refactor style is too broad." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-2 output tracing and MGP-STR output tracing are separate fixes." + "reason": "GPT-2 and MGP-STR are unrelated model implementations." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J and MGP-STR are unrelated model code paths." + "reason": "GPT-J and MGP-STR are different concrete code paths." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert/XLM are separate model refactors." + "reason": "ViLT and XLM/Flaubert are separate model families." }, { - "left": "pull_request:44059", - "right": "pull_request:44722", + "left": "pull_request:44017", + "right": "pull_request:44161", "accept": false, - "reason": "GPT-2 and GPT-J are different concrete implementations." + "reason": "SegFormer and LongT5 are unrelated model-specific refactors." }, { - "left": "pull_request:44018", - "right": "pull_request:44071", + "left": "pull_request:44017", + "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and MPT are unrelated model paths." + "reason": "SegFormer and GPT-J/CodeGen are unrelated concrete changes." }, { - "left": "pull_request:44074", - "right": "pull_request:44101", + "left": "pull_request:44017", + "right": "pull_request:44073", "accept": false, - "reason": "TextNet and Flaubert/XLM are different model code paths." + "reason": "SegFormer and VisualBert are different model code paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44017", + "left": "pull_request:44068", + "right": "pull_request:44071", "accept": false, - "reason": "RegNet and SegFormer are separate vision model implementations." + "reason": "GPT-Neo and MPT are unrelated models." }, { "left": "pull_request:43997", - "right": "pull_request:44056", + "right": "pull_request:44073", "accept": false, - "reason": "RegNet and MPNet are unrelated model changes." + "reason": "RegNet and VisualBert are unrelated model implementations." }, { "left": "pull_request:44071", - "right": "pull_request:44076", + "right": "pull_request:44086", "accept": false, - "reason": "MPT and ImageGPT are different code paths." + "reason": "MPT and MGP-STR are separate model families." }, { - "left": "pull_request:44074", - "right": "pull_request:44161", + "left": "pull_request:44073", + "right": "pull_request:44101", + "accept": false, + "reason": "VisualBert and XLM/Flaubert are unrelated code paths." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44071", + "accept": false, + "reason": "SegFormer and MPT are different models." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", "accept": false, - "reason": "TextNet and LongT5 are unrelated refactors." + "reason": "GPT-2 and MPT are unrelated refactors." }, { "left": "pull_request:44056", - "right": "pull_request:44161", + "right": "pull_request:44101", "accept": false, - "reason": "MPNet and LongT5 are separate model-specific changes." + "reason": "MPNet and XLM/Flaubert are separate model code paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44018", + "left": "pull_request:44018", + "right": "pull_request:44059", "accept": false, - "reason": "RegNet and GPT-Neo are unrelated model implementations." + "reason": "GPT-Neo and GPT-2 are different model families." }, { - "left": "pull_request:44071", - "right": "pull_request:44085", + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-2 and GPT-Neo are not the same concrete change." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", "accept": false, - "reason": "MPT and GPT-J are different concrete code paths." + "reason": "XLM/Flaubert and LongT5 are unrelated model implementations." }, { "left": "pull_request:43997", - "right": "pull_request:44074", + "right": "pull_request:44017", "accept": false, - "reason": "RegNet and TextNet are separate model refactors." + "reason": "RegNet and SegFormer are different model paths." }, { "left": "pull_request:44010", "right": "pull_request:44071", "accept": false, - "reason": "SqueezeBERT and MPT are unrelated model code paths." + "reason": "SqueezeBert and MPT are unrelated refactors." }, { "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "SqueezeBERT and VisualBERT are different model-specific refactors." + "reason": "SqueezeBert and VisualBert are separate model code paths." }, { "left": "pull_request:44018", "right": "pull_request:44129", "accept": false, - "reason": "GPT-Neo and SpeechT5 are unrelated code paths." + "reason": "GPT-Neo and SpeechT5 are unrelated model changes." }, { "left": "pull_request:43996", "right": "pull_request:44161", "accept": false, - "reason": "FNet/CVT and LongT5 are different model implementations." + "reason": "FNet/CVT and LongT5 are unrelated implementations." }, { "left": "pull_request:43996", "right": "pull_request:44017", "accept": false, - "reason": "FNet/CVT and SegFormer are separate model changes." + "reason": "FNet/CVT and SegFormer are different model families." }, { "left": "pull_request:44010", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBERT and DeBERTa-v2 are unrelated model code paths." + "reason": "SqueezeBert and DeBERTa are unrelated concrete changes." }, { "left": "pull_request:44044", "right": "pull_request:44074", "accept": false, - "reason": "DeBERTa-v2 and TextNet are different model-specific refactors." + "reason": "DeBERTa and TextNet are separate model-specific refactors." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor on overlapping files and the same issue target #43979." + "reason": "Same GPT-J/CodeGen output-tracing refactor family; 44722 is a later, compatible version of the same change." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Both PRs patch GPT-J output tracing in the same file and appear to be alternate variants of the same change." + "reason": "Both edits are GPT-J output-tracing refactors on the same file and same underlying code path." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J and SpeechT5 are unrelated model code paths." + "reason": "GPT-J and SpeechT5 are unrelated model implementations." }, { "left": "pull_request:43996", "right": "pull_request:44007", "accept": false, - "reason": "FNet/CVT and ResNet are separate model implementations." + "reason": "FNet/CVT and RegNet are unrelated model code paths." }, { "left": "pull_request:43996", "right": "pull_request:44129", "accept": false, - "reason": "FNet/CVT and SpeechT5 are unrelated model refactors." + "reason": "FNet/CVT and SpeechT5 are not the same underlying bug/change." }, { "left": "pull_request:44044", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa-v2 and SpeechT5 are different concrete code paths." + "reason": "DeBERTa and SpeechT5 are different models." }, { "left": "pull_request:44044", "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa-v2 and VITS are unrelated model changes." + "reason": "DeBERTa and VITS are unrelated concrete changes." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The cluster summary is conservative and grounded in the packet: most PRs are model-specific output-tracing refactors tied to issue #43979, while the accepted soft pairs stay within the same model family/code path and look mergeable into a single bugfix PR. No verdict appears overstated." + "accept": false, + "feedback": "Reject: the soft-edge call on 44066 vs 44085 is too strong. Both PRs touch the same GPT-J file (`src/transformers/models/gptj/modeling_gptj.py`) and share the umbrella issue target, so describing them as \u201cdifferent model families and different code paths\u201d overstates the separation. At minimum, this pair looks plausibly mergeable or duplicate-like enough that a conservative verdict would not rule it out so firmly. The summary is mostly grounded, but the exclusivity claim (\u201conly true duplicate-like groups\u201d) should also be softened." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -20739,12 +20764,12 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "d83d177be6d033edac1a8c550da8098784a05446e6f76ee1b2bd8a55ceacc2ba", + "prepared_review_unit_hash": "c37762392b7782bbcf0d8640b918d5640f8eff8cfff7939a38c6d0ed780fb6fe", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84895, + "serialized_chars": 84894, "estimated_input_tokens": 21224, "estimated_eval_tokens": 42704 }, @@ -20761,48 +20786,48 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are model-specific refactors to standardized output tracing/capture_outputs across different architectures. They share a common refactoring pattern, but not the same concrete code-path bug or mergeable change set, so the cluster should not be merged by soft similarity alone.", - "confidence": 0.84, + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they touch different architectures and code paths. They are related thematically, not duplicates of the same underlying change.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "SpeechT5 (44129) is the strongest representative because it has the clearest explicit issue linkage and a substantial output-tracing refactor, making it the most anchored to a specific tracked change among the set.", + "canonical_pr_reason": "pull_request:44129 is the clearest representative: it has an explicit issue target, a focused title, and a concrete output-tracing refactor in one model implementation.", "best_issue_reason": null, - "best_pr_reason": "44129 is the best overall PR candidate for the cluster because it is the most explicitly scoped and traceable; the others are similar migrations but are narrower, less anchored, or even more obviously just parallel model-specific cleanup PRs.", + "best_pr_reason": "pull_request:44129 is the best single representative for the cluster because it is narrowly scoped and explicitly tied to the standardized output capture work, but it is not a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "Both are output-capture refactors, but for different model implementations (FocalNet vs ImageGPT). They are parallel migrations, not the same concrete bug or a PR that could plausibly be merged as one change." + "reason": "Both are output-capture refactors, but they target different models (FocalNet vs ImageGPT) and different code paths, so they are not the same change." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "VisualBert and TextNet are different code paths and architectures. The refactor pattern is similar, but this is not the same underlying change." + "reason": "VisualBert and TextNet are separate model implementations; this is the same refactor pattern, not a shared underlying bug or mergeable duplicate PR." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Both migrate output tracing, but MPNet and VisualBert are separate model-specific implementations. Too broad to treat as duplicate PRs." + "reason": "MPNet and VisualBert are unrelated model-specific output-tracing updates; they do not fix the same concrete defect." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "Same refactor theme, but MPNet and TextNet are unrelated model code paths. Not the same concrete change." + "reason": "These are analogous refactors in different model files, not duplicate fixes to one code-path problem." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are unrelated models. Similar decorator/capture_outputs migration does not make them duplicates." + "reason": "RegNet and SqueezeBert each change their own model forward/output handling, so they are not the same underlying change." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 are different model implementations; even with a similar tracing refactor, they are separate PRs and should not be merged as duplicates." + "reason": "SqueezeBert and SpeechT5 both involve output handling, but they modify different model internals and are not the same concrete fix." } ] }, @@ -20885,274 +20910,292 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "59ebef6d695243ffa830690f686f4f295b4a59c84dceec699ea8b230cfa0bb8f", + "prepared_review_unit_hash": "06b9e3e1609830c6189c1d7130612d00c47831fd663a6368f549830fd4371a14", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41325, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41333, + "estimated_input_tokens": 10334, + "estimated_eval_tokens": 20924 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41325, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41333, + "estimated_input_tokens": 10334, + "estimated_eval_tokens": 20924 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs share a broad \u201coutput tracing / can_return_tuple\u201d refactor theme, but they target different model implementations and code paths. I would not deduplicate any of them; all soft edges should be rejected.", - "confidence": 0.98, + "summary": "This cluster is a family of similar-looking but distinct model-specific output-tracing refactors. They share naming/style and utility usage, but each PR targets a different code path and should not be merged as duplicates.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: each change is model-specific (different files, forward signatures, and tracing behavior), so they are not interchangeable duplicates.", + "canonical_pr_reason": "PR 44025 is the clearest representative of the cluster: it applies the refactor to a primary model plus its paired prompt variant, making it a good exemplar of the shared pattern.", "best_issue_reason": null, - "best_pr_reason": "No global best PR for deduplication; the cluster is only thematically similar, not a single underlying fix/change.", + "best_pr_reason": "PR 44025 is the best standalone representative of the theme because it is explicit, self-contained, and spans a closely related model pair without being overly broad.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different models and files (timm_backbone vs upernet); same refactor pattern, not the same change." + "reason": "Same refactor pattern, but Timm Backbone and UperNet are different model code paths." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Distinct model implementations and code paths; only the tracing refactor theme overlaps." + "reason": "Different model families; same tracing cleanup theme is not enough for a duplicate PR." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Unrelated model families (timm_backbone vs univnet); not a duplicate bug/change." + "reason": "Timm Backbone vs UnivNet are unrelated implementations." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; too broad a similarity to count as the same fix." + "reason": "Different architectures and files; not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Different model code paths (timm_backbone vs rwkv); shared refactor wording is insufficient." + "reason": "Timm Backbone and RWKV are separate model paths." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "MobileNetV1 and UnivNet are unrelated model code paths; not the same underlying change." + "reason": "MobileNetV1 and UnivNet are different code paths." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different models and implementations; only broadly similar tracing cleanup." + "reason": "MobileNetV1 vs UperNet is only a broad thematic match." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and files; not one mergeable duplicate PR." + "reason": "MobileNetV1 and DepthAnything target different model families." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Mobilenet_v1 vs vision_encoder_decoder are separate code paths and changes." + "reason": "MobileNetV1 vs VisionEncoderDecoder are distinct implementations." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Different models; same refactor theme does not make them duplicates." + "reason": "MobileNetV1 vs SpeechEncoderDecoder are separate code paths." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Vision-text dual encoder and UnivNet are unrelated; no shared concrete bug." + "reason": "Vision-Text Dual Encoder and UnivNet are unrelated model changes." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different model families/code paths; not the same fix." + "reason": "Different model families; not the same fix." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Vision-text dual encoder vs depth_anything are separate changes, not one duplicate PR." + "reason": "Vision-Text Dual Encoder and DepthAnything are separate code paths." }, { "left": "pull_request:44000", "right": "pull_request:44026", "accept": false, - "reason": "Distinct models and forward logic; only the refactor motif overlaps." + "reason": "Different architectures and files; same output-tracing refactor is too broad." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "UnivNet and UperNet are unrelated model implementations; no same underlying bug/change." + "reason": "UnivNet and UperNet are unrelated models." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Different architectures and code paths; not mergeable as one PR." + "reason": "UnivNet vs DepthAnything are distinct implementations." }, { "left": "pull_request:44001", "right": "pull_request:44026", "accept": false, - "reason": "Unrelated model code paths; shared output-tracing refactor is too broad." + "reason": "Different code paths; not mergeable as one PR." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "UperNet and RWKV are different model families; not the same concrete fix." + "reason": "UperNet and RWKV are different model families." }, { "left": "pull_request:44002", "right": "pull_request:44025", "accept": false, - "reason": "Different files and behaviors; only broad thematic similarity." + "reason": "UperNet and DepthAnything are separate changes." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "UnivNet vs Mamba/FalconMamba are separate code paths, not duplicate changes." + "reason": "UnivNet vs Mamba/Falcon-Mamba are unrelated implementations." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different model families and implementations; not the same bug or feature change." + "reason": "UperNet vs Mamba/Falcon-Mamba are different code paths." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Vision-text dual encoder and Mamba/FalconMamba are distinct changes; similarity is only superficial." + "reason": "Vision-Text Dual Encoder and Mamba/Falcon-Mamba are unrelated." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "Different model architectures (Mamba/FalconMamba vs SuperPoint); not one underlying fix." + "reason": "Mamba/Falcon-Mamba and SuperPoint are separate model families." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Different code paths and model types; same refactor idea is not enough for duplication." + "reason": "Mamba/Falcon-Mamba vs CodeGen is not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Timm_backbone and CodeGen are unrelated implementations; not a duplicate PR." + "reason": "Timm Backbone and CodeGen are unrelated code paths." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "UnivNet vs CodeGen share only the refactor theme." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "UperNet vs CodeGen are different implementations." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Vision-Text Dual Encoder vs CodeGen are distinct model paths." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "DepthAnything and VisionEncoderDecoder are different model families and changes." + "reason": "DepthAnything and VisionEncoderDecoder are different code paths." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "DepthAnything vs SpeechEncoderDecoder are separate code paths; not the same change." + "reason": "DepthAnything and SpeechEncoderDecoder are separate implementations." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Different model code paths and behavior; only shared refactor style." + "reason": "DepthAnything vs SuperPoint is only a broad pattern match." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "DepthAnything and RWKV are unrelated models; not a duplicate fix." + "reason": "DepthAnything and RWKV are unrelated model families." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Different model implementations; same output-tracing theme does not imply duplication." + "reason": "DepthAnything and DPR are different code paths." }, { "left": "pull_request:44026", - "right": "pull_request:44027", + "right": "pull_request:44029", "accept": false, - "reason": "VisionEncoderDecoder and SpeechEncoderDecoder are separate model code paths, not one bug/change." + "reason": "VisionEncoderDecoder vs RWKV are unrelated implementations." }, { "left": "pull_request:44026", - "right": "pull_request:44028", + "right": "pull_request:44027", "accept": false, - "reason": "Different models and files; too broad to collapse into one PR." + "reason": "VisionEncoderDecoder and SpeechEncoderDecoder are separate wrappers, not the same fix." }, { "left": "pull_request:44026", - "right": "pull_request:44029", + "right": "pull_request:44028", "accept": false, - "reason": "Distinct architectures and forward behavior; not the same fix." + "reason": "VisionEncoderDecoder vs SuperPoint are different model paths." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "VisionEncoderDecoder vs DPR are different code paths; not duplicates." + "reason": "VisionEncoderDecoder and DPR are unrelated code paths." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "SpeechEncoderDecoder and SuperPoint are unrelated implementations." + "reason": "SpeechEncoderDecoder vs SuperPoint are separate implementations." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Different model families; shared refactor pattern is insufficient." + "reason": "SpeechEncoderDecoder and RWKV are unrelated model families." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "SpeechEncoderDecoder vs DPR are separate code paths and changes." + "reason": "SpeechEncoderDecoder and DPR are distinct code paths." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "SuperPoint and RWKV are unrelated models; not the same underlying issue." + "reason": "SuperPoint and RWKV are different model families." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "Different model implementations and forward logic; not a duplicate." + "reason": "SuperPoint and DPR are separate implementations." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "RWKV and DPR are distinct model families; only the refactor theme overlaps." + "reason": "RWKV and DPR are unrelated model code paths." } ] }, @@ -21182,12 +21225,12 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "24af8c5e8bcbc8eddee09eafd2d05d2e87dae9e1a7a9f414f2d4fc95cfc818f9", + "prepared_review_unit_hash": "6c9621482385977d79ffe69e483c88eb31dd67dc617f93ebddbf9bbf06710bbf", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6552, + "serialized_chars": 6551, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -21195,7 +21238,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6552, + "serialized_chars": 6551, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -21204,18 +21247,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests are unrelated model additions: one implements PenguinVL, the other adds Voxtral TTS support. Their overlap is limited to standard Transformers plumbing and docs, not the same underlying change.", - "confidence": 0.96, + "summary": "These are two unrelated model-addition PRs: one implements PenguinVL, the other adds Voxtral TTS support. They share only common Transformers integration scaffolding, not the same underlying change or code path.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR 45401 is the better canonical representative only because it is explicitly tied to issue 45027 and has a tighter, well-scoped model addition; PR 44662 is a separate unrelated implementation for a different model.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "PR 45401 is the better overall PR representative for the cluster because it has an explicit tracking issue and a narrower, more clearly defined feature scope. PR 44662 is a distinct PenguinVL implementation and should not be merged with it.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different models and use cases: PenguinVL is a vision-language implementation, while Voxtral TTS is a text-to-speech integration. Shared auto-registration/docs changes are generic scaffolding, not the same concrete code path or mergeable duplicate change." + "reason": "Different models and different feature work: PenguinVL implementation vs Voxtral TTS support. Shared auto/model registration files are generic integration overlap, not evidence of the same underlying bug or change." } ] }, @@ -21245,12 +21288,12 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "e25d45ebb9dc16cd8912707d244d2db52688c4922421e80c7eff34126a0f694f", + "prepared_review_unit_hash": "065146b3ef00eeaf2b853102a8ba0d028b1f00f78d09a30b95e284dcd3224716", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4406, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -21258,7 +21301,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4406, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -21267,18 +21310,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related by the broad theme of module/kernel fusion, but they appear to be different features: one is a general Module Fusion API, the other is a specific n-to-1 kernel fusion mechanism driven by KernelConfig. They should not be treated as duplicates.", - "confidence": 0.91, + "summary": "Two open PRs are about fusion, but they are not duplicates: PR 44979 introduces a generic Module Fusion API, while PR 45363 implements KernelConfig-based n-to-1 kernel fusion integration and an example. Same broad area, different layers/code paths.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45363 is the more concrete and integrated change: it wires KernelConfig through modeling/integration paths and adds an end-to-end example, making it the better representative of the fusion work.", + "canonical_pr_reason": "PR 45363 is the more concrete, end-user-facing change: it wires fusion through KernelConfig and modeling hooks rather than just adding the abstraction layer.", "best_issue_reason": null, - "best_pr_reason": "PR #45363 is the best overall PR in the cluster because it has the clearest user-facing behavior change and the most complete implementation surface; PR #44979 looks like a separate foundational API proposal.", + "best_pr_reason": "PR 45363 is the best representative of the cluster because it captures the actual runtime integration path and practical usage example for fusion.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Do not merge: they address different fusion layers. #44979 introduces a general Module Fusion API, while #45363 implements n-to-1 kernel fusion via KernelConfig and related integration hooks. They share a theme but not the same concrete change or code path." + "reason": "They overlap on fusion concepts but target different implementations: 44979 adds a generic Module Fusion API, while 45363 adds KernelConfig-driven n-to-1 kernel fusion integration. They are related, not the same underlying change." } ] }, @@ -21308,7 +21351,7 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "505432a8fdebded8ce4f1ff2c6124293529fca0778fee6b3f6fd13d198cc355b", + "prepared_review_unit_hash": "5256a53f486842f1b25657abfee6c5fd24498960bf8fcdcece93743fa09ccb1d", "component_budget": { "node_count": 2, "item_count": 2, @@ -21330,18 +21373,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests fix the same AttributeError in `_patch_mistral_regex` by switching from `tokenizer.backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; PR #45317 is the more complete implementation because it also adds a test.", - "confidence": 0.98, + "summary": "Both pull requests fix the same `_patch_mistral_regex` AttributeError in `tokenization_utils_tokenizers.py` for the same tracked issue; the second PR is a slightly more complete version with tests.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45317 is the best canonical PR because it contains the same code fix as #45086 plus a test file, making it the more complete and reviewable patch.", + "canonical_pr_reason": "PR #45317 is the stronger canonical PR: it fixes the same bug as #45086 and also adds test coverage, making it the more complete representative of the change.", "best_issue_reason": null, - "best_pr_reason": "PR #45317 is the strongest representative of the cluster: same bug fix, same target issue, and additional test coverage.", + "best_pr_reason": "PR #45317 is the best overall PR because it addresses the same concrete code-path bug while including validation tests; #45086 is effectively an earlier, narrower variant of the same fix.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying bug and same code-path fix in the same file; #45317 is essentially the tested version of #45086 and could plausibly be merged as one PR." + "reason": "Same underlying `_patch_mistral_regex` AttributeError fix in the same function/file, same explicit issue target, and the second PR only adds tests rather than changing the problem being fixed." } ] }, @@ -21371,7 +21414,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "7e3506ae12fd2eb73e7ad58e1f868ed9d8d0bdf38af87dbcdd5913b1fc512f7c", + "prepared_review_unit_hash": "aa9aa39d1dc95db112c3bd0d9af1c5ff9276c01b643e7ec369f5c32c472f3d44", "component_budget": { "node_count": 2, "item_count": 2, @@ -21393,18 +21436,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related NemotronH refactor PRs, but they address different changes: one is a narrow checkpoint conversion/mapping update, the other is a broader architecture split into NemotronHDense/NemotronHSparse and related docs/tests. They should not be deduplicated or merged as the same underlying change.", - "confidence": 0.93, + "summary": "Both PRs are about NemotronH, but they address different changes: #45115 is a narrow conversion/mapping and inheritance refactor, while #45569 is a broader model split into dense/sparse variants with docs and auto-mapping updates. They are related, but not the same underlying fix/change set.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "#45569 is the broader, more central NemotronH rework and better represents the cluster topic; #45115 is a narrower supporting refactor focused on conversion mappings and inheritance cleanup.", "best_issue_reason": null, - "best_pr_reason": "pull_request:45569 is the better representative PR because it is the broader, more central NemotronH refactor with model, auto-mapping, docs, and tests; pull_request:45115 is a narrower conversion-mapping adjustment.", + "best_pr_reason": "#45569 is the best single PR to keep because it covers the main architectural change and the new dense/sparse model support, whereas #45115 only adjusts conversion naming and model wiring.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "They overlap on NemotronH-related files, but the concrete changes differ: 45115 is about conversion key renaming for checkpoints, while 45569 refactors the model into dense/sparse variants and updates docs/tests. Same subsystem, not the same fix." + "reason": "Related NemotronH work, but not the same concrete change: #45115 refactors conversion mapping/inheritance compatibility, while #45569 introduces dedicated dense/sparse model packages, docs, and auto-mapping updates." } ] }, @@ -21434,7 +21477,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "edb8d48208ed200f9a92c4ced2643a1bb43b1a842cafeb0d9a3d8da2730a6d9d", + "prepared_review_unit_hash": "e41fe80370b097da6b802665eecf683f575f96bf9c4a434717910a8e77a2b508", "component_budget": { "node_count": 2, "item_count": 2, @@ -21456,18 +21499,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two related but distinct PRs in the heterogeneous-model feature area: one for per-layer config support and one for per-layer modeling/cache handling. They share files, but they are not the same concrete change.", - "confidence": 0.89, + "summary": "Two related but distinct pull requests: one adds heterogeneous config support, the other adds broader heterogeneous modeling/cache support. They share implementation files, but they are not the same concrete change and are better treated as separate work items.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45332 is the broader umbrella change: it includes heterogeneous modeling/cache logic plus supporting config and test updates, so it is the best representative of the cluster.", + "canonical_pr_reason": "PR #45332 is the broader end-to-end feature PR and better represents the cluster; PR #45333 is a narrower config-only piece that looks like a prerequisite or companion change rather than the same PR.", "best_issue_reason": null, - "best_pr_reason": "PR #45332 is the best overall representative because it covers the main end-to-end heterogeneous model support path rather than only the config layer.", + "best_pr_reason": "PR #45332 is the best representative PR because it covers the larger heterogeneous model support effort across config, modeling, cache, and tests.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related feature work, but not the same underlying bug/change: #45333 adds per-layer config support, while #45332 adds heterogeneous modeling/cache behavior. They are adjacent components, not one fix." + "reason": "Related feature work, but not the same underlying bug/change: one is config support and the other is modeling/cache support. Shared files alone are not enough to treat them as duplicates." } ] }, @@ -21497,7 +21540,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "4098efa219391bc4168c41c47419f95d5be59feac60f1709ae7163bdd42078a8", + "prepared_review_unit_hash": "56b29307a2e2f3161c4a4347551bc2c607d622c537ebdefa059a03844f13d7f3", "component_budget": { "node_count": 2, "item_count": 2, @@ -21519,18 +21562,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both typing-related PRs, but they address different changes: one is a broad repo-wide type-checking cleanup, the other adds/adjusts specific internal typing protocols for modeling_utils/PEFT integration. They should not be deduplicated.", - "confidence": 0.87, + "summary": "Both PRs are typing-related, but they address different scopes: one is a broad type-checking sweep across many transformers modules, while the other adds shared typing protocols/aliases and a smaller set of typing refinements. They are related, but not the same concrete change.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the two pull requests are distinct changes with different scope and implementation targets.", + "canonical_pr_reason": "PR 45415 is the broader, more representative typing cleanup: it adds type checking coverage/config and touches many modules in one pass. PR 45425 is a narrower follow-up focused on shared typing helpers and protocol definitions.", "best_issue_reason": null, - "best_pr_reason": "PR #45415 is the better representative of the cluster because it is the broader, more general typing cleanup across the codebase.", + "best_pr_reason": "PR 45415 is the best representative artifact for the cluster because it captures the main end-to-end type-checking initiative and has the widest surface area.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Both are typing-focused, but #45415 is a broad type-checking pass touching many files and pyproject config, while #45425 adds specific typing protocols and annotations for modeling_utils/PEFT-related code. They are not the same underlying change and would not plausibly be merged into one PR." + "reason": "Both are typing improvements, but they do not fix the same concrete bug or change. 45415 is a broad type-checking sweep across many files, while 45425 adds typing aliases/protocols and related fixes in a smaller set of files. They could be separate PRs and are not duplicates." } ] }, @@ -21560,12 +21603,12 @@ "soft_pairs": [ "pull_request:45513|pull_request:46084" ], - "prepared_review_unit_hash": "e72213c41f7ead1564a6a49b72c5b4f1d93501f5b492cc5ef7071a502ec3a4ed", + "prepared_review_unit_hash": "2b0e9c24acf751d11367a2187195f5e7812164fe7c643c66d126d311ad74d979", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4857, + "serialized_chars": 4858, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -21573,7 +21616,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4857, + "serialized_chars": 4858, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -21582,18 +21625,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are both cache-related, but they fix different model families and different code paths: Qwen/Olmo linear-attention cached multi-token forward vs. Mamba2 use_cache for seq_len > 1. Not a duplicate cluster.", - "confidence": 0.95, + "summary": "Two open PRs about cache-aware multi-token forward paths, but they fix different model families and different code paths: Qwen/Olmo linear attention vs Mamba2 cache state handling. They are superficially similar, not duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45513 is the better canonical PR for this cluster because it is the more central, multi-model fix for cached multi-token forward behavior in the GDN linear-attention path across Qwen3.5/Olmo/Qwen3-next variants.", + "canonical_pr_reason": "PR #45513 is the more central representative of the cluster: it has broader scope across several related Qwen/Olmo model files, more discussion/inbound references, and the main cache-forward fix theme the cluster appears to revolve around.", "best_issue_reason": null, - "best_pr_reason": "PR 45513 is the best representative of the cluster\u2019s shared theme (cached forward logic), but only as a loose thematic anchor; it is not actually the same bug as PR 46084.", + "best_pr_reason": "PR #45513 is the best standalone representative because it is the broader, higher-activity cache-forward fix. PR #46084 is a separate Mamba2 bug fix and should not be merged into this cluster despite the similar cache/seq_len wording.", "soft_edge_verdicts": [ { "left": "pull_request:45513", "right": "pull_request:46084", "accept": false, - "reason": "Different underlying bug and code path: 45513 fixes multi-token cached forward in GDN linear attention for Qwen/Olmo models, while 46084 fixes Mamba2 cache behavior when seq_len > 1. They are not mergeable as one PR." + "reason": "Both mention cached multi-token forward paths, but they target different architectures and concrete implementations: Qwen/Olmo GDN linear attention vs Mamba2 CUDA kernel cache states. The diffs do not indicate the same underlying bug or a mergeable single PR." } ] }, @@ -21623,7 +21666,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "e1b9b3bd679fb85c046917d48d11a1bb1729db3f3f3fa1053907d6d9cc5797ec", + "prepared_review_unit_hash": "d05ddc9be153b248a3e8aec9d1d65c6ff423a13610e2a8e4bfbeec0eafe1a8b1", "component_budget": { "node_count": 2, "item_count": 2, @@ -21645,18 +21688,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are related but not duplicates: 45548 fixes EP loading/serialization for DeepSpeed ZeRO-3 via accelerate launch, while 45662 fixes a separate EP/FSDP2 issue where experts get overwritten by rank-0 broadcast.", + "summary": "Both PRs touch expert-parallel loading, but they address different bugs and different code paths: 45548 is about distributed_config serialization and accelerate/ZeRO-3 loading, while 45662 is about expert parameters being overwritten during FSDP2 rank-0 broadcast.", "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: the two patches address different concrete failure modes in different distributed loading paths.", + "canonical_pr_reason": "No single canonical PR: the two patches solve related EP loading problems, but on different mechanisms and paths, so they should not be merged as one duplicate fix.", "best_issue_reason": null, - "best_pr_reason": "45548 is the more general EP-loading/configuration fix, but it does not cover the distinct FSDP2 silent-overwrite bug in 45662.", + "best_pr_reason": "45662 is the stronger representative of the cluster because it targets a concrete, reproducible parameter-overwrite bug with a broader fix across loading/integration code, whereas 45548 is narrower and focused on serialization/launch behavior.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Different underlying bugs and code paths: ZeRO-3/accelerate loading vs FSDP2 rank-0 broadcast overwriting experts. Similar EP terminology is not enough to merge them as one change." + "reason": "Different underlying bugs. 45548 removes distributed_config from serialization and adds has_ep for accelerate/ZeRO-3 loading; 45662 wraps sharded params to prevent expert weights being overwritten during FSDP2 rank-0 broadcast. Shared EP context is not enough to treat them as duplicates." } ] }, @@ -21686,12 +21729,12 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "9a996ec2fd0d1adeeaf1124c22a95a5a34c13fc75a81328f3d07d9fb6aff891c", + "prepared_review_unit_hash": "651095e3ac14c7915455b92951552488d47b90970383e1f68f78f44de3fa26e4", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4590, + "serialized_chars": 4589, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -21699,7 +21742,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4590, + "serialized_chars": 4589, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -21708,18 +21751,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs target the same xdist-related captured_info logging collision in testing_utils; PR 45645 is the broader, more complete fix because it also updates CI log collection to read multiple captured_info files.", - "confidence": 0.95, + "summary": "Both PRs address the same xdist-related captured_info logging collision in testing utilities; PR 45645 is the broader, more complete fix and includes CI log handling, while PR 45639 is a narrower subset.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the strongest canonical PR because it covers the core xdist-safe output change plus the downstream CI/workflow handling needed to preserve and display the split debug logs.", + "canonical_pr_reason": "pull_request:45645 is the best canonical PR because it covers the core testing_utils xdist-safe file change plus the CI workflow update to aggregate multiple captured_info files, making it the most complete representation of the fix.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 is the best representative of the cluster: it addresses the same underlying bug as 45639 and adds the operational CI changes needed for the fix to work end-to-end.", + "best_pr_reason": "pull_request:45645 is the strongest representative of the cluster: it fixes the concrete xdist collision in the code path and also preserves CI debug log collection, whereas 45639 only addresses the utility-side portion.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Same underlying captured_info xdist collision bug in the same code path; 45645 looks like a superset of 45639 and they could plausibly be merged into one PR." + "reason": "Both PRs fix the same underlying captured_info/xdist collision in testing_utils, and 45645 extends the same fix to CI log collection; they could plausibly be merged into one PR." } ] }, @@ -21752,7 +21795,7 @@ "pull_request:45670|pull_request:46401", "pull_request:46320|pull_request:46401" ], - "prepared_review_unit_hash": "91c1db89eb77573af43f47d0fd35137a6713dee66da5c3450f49cee2db75ad63", + "prepared_review_unit_hash": "e5eea6e37be9354b6a818b6aa4e174e902d5adf5d4c95b9538580091a0e8e4b0", "component_budget": { "node_count": 3, "item_count": 3, @@ -21774,30 +21817,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related only at a high level: two PRs add attn_implementation support for different model families, while one PR is a separate AutoModel mapping tweak. They are not duplicates of the same concrete code change.", - "confidence": 0.91, + "summary": "These PRs are related only at a high level: two add attn_implementation dispatch support for different model families, while one is a separate AutoModel mapping fix. They do not look like duplicates or mergeable into one change.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:46401 is the best representative PR because it is the broader attn_implementation dispatch change across multiple BLIP-family models, with corresponding model and test updates.", + "canonical_pr_reason": "No true duplicate cluster here. If a representative PR is needed, PR 46401 is the broadest and most developed attn-dispatch change across multiple BLIP-family models.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46401 is the most substantial and central feature PR in the cluster; it covers several related BLIP-family code paths and is better suited as the anchor than the narrower RecurrentGemma PR or the unrelated AutoModel mapping tweak.", + "best_pr_reason": "PR 46401 is the strongest representative of the cluster because it has the broadest scope, touches multiple model implementations/tests, and has the most discussion/review activity.", "soft_edge_verdicts": [ { "left": "pull_request:45670", "right": "pull_request:46320", "accept": false, - "reason": "Different changes: 45670 adds a GLM-ASR AutoModel mapping entry, while 46320 implements attn_implementation dispatch for RecurrentGemma. No shared bug or mergeable common fix." + "reason": "Different changes: AutoModel registry wiring for GLMASR versus RecurrentGemma attention-dispatch support. Not the same code-path bug or fix." }, { "left": "pull_request:45670", "right": "pull_request:46401", "accept": false, - "reason": "Different scopes and code paths: 45670 is an AutoModel registry nit, while 46401 adds attn_implementation support for BLIP-2/InstructBLIP variants. Not the same underlying change." + "reason": "Unrelated fixes: AutoModel mapping for GLMASR versus BLIP-family attn_implementation dispatch support." }, { "left": "pull_request:46320", "right": "pull_request:46401", "accept": false, - "reason": "Both are about attn_implementation support, but for different model families and separate modeling files/tests. They are related feature work, not the same concrete fix that should be merged into one PR." + "reason": "Both concern attention dispatch, but for different model families and different implementation paths; they are not the same concrete fix and would not plausibly be merged as one PR." } ] }, @@ -21827,40 +21870,40 @@ "soft_pairs": [ "pull_request:45720|pull_request:46259" ], - "prepared_review_unit_hash": "947b8734b29328579057d0be70042289bb5f32c99765d6b2f582a035beb67bff", + "prepared_review_unit_hash": "c08547e1f95f9ad4450748d63c460d23ef9b8f46051f1246c73b9bc368724b0e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7189, - "estimated_input_tokens": 1798, - "estimated_eval_tokens": 3852 + "serialized_chars": 7188, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7189, - "estimated_input_tokens": 1798, - "estimated_eval_tokens": 3852 + "serialized_chars": 7188, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated open PRs in the same codebase: one implements TRF003 across several model forward methods, while the other adds a new linter rule TRF018 for image/video processor flags. They share some surrounding repo areas but not the same bug or change.", - "confidence": 0.95, + "summary": "Two unrelated linter-rule PRs were grouped together because they both modify Transformers modeling code, but they target different concrete problems: TRF003 for return_dict/can_return_tuple in model forwards vs. TRF018 for dead do_* flags in image/video processors. They should not be deduplicated.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR for a duplicate cluster: PR #45720 and PR #46259 address different rules, different code paths, and different subsystems.", + "canonical_pr_reason": "No canonical PR: 45720 and 46259 address different rules, different code paths, and different model families, so they are not duplicates.", "best_issue_reason": null, - "best_pr_reason": "PR #46259 is the cleaner representative because it introduces a self-contained, documented rule change with a clear scope; PR #45720 is a broader partial implementation across many model files.", + "best_pr_reason": "PR 45720 is the better representative of the cluster\u2019s linter-rule work because it has broader impact across many model forwards and introduces the reusable can_return_tuple decorator; 46259 is a separate processor-specific rule.", "soft_edge_verdicts": [ { "left": "pull_request:45720", "right": "pull_request:46259", "accept": false, - "reason": "Different underlying changes: TRF003 modifies model forward signatures/tuple-return behavior, while TRF018 adds a linter rule for dead do_* flags in image/video processors. They should not be merged as duplicates." + "reason": "Both are linter-related, but they fix different issues: forward return handling in text/audio models vs. unused do_* preprocessing flags in image/video processors. Not the same change and not mergeable as one PR." } ] }, @@ -21956,7 +21999,7 @@ "soft_pairs": [ "pull_request:45932|pull_request:46100" ], - "prepared_review_unit_hash": "fbcd9bcbba11f9d1132da41a6af321dd6cbbaa297db327c42b3e3eebb0e41a7f", + "prepared_review_unit_hash": "2e1556eeaaa84146eefb95b6c51edcaccba5504a488b1a80ce7733394bc733c8", "component_budget": { "node_count": 2, "item_count": 2, @@ -21978,18 +22021,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are not duplicates: #45932 is a broad workflow/model patch, while #46100 is a small test-file migration with intentional failing assertions. They do not look like the same underlying change.", - "confidence": 0.96, + "summary": "The two PRs are not duplicates: one is a broad testing/infra and model-change PR, while the other is a small ViT test-edit PR with intentional failing assertions. The similarity is superficial.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR for deduping here; the PRs target different code paths and purposes, with #45932 being the more substantive multi-file change and #46100 being a narrow test-only edit.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "#45932 is the better representative artifact because it contains the larger, real implementation/workflow change; #46100 appears to be a local test-migration or placeholder change rather than a production fix.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "Different intents and code paths: #45932 modifies CI/workflow and multiple model/config files, while #46100 only edits a ViT test file with intentional failing assertions. They are not the same underlying bug or change." + "reason": "Reject: they do not address the same code-path or bug. PR 45932 touches CI/workflow plumbing and several model files, while PR 46100 only edits a ViT test file and adds failing assertions for migration. Shared model/test surface is too broad and not a plausible single merged fix." } ] }, @@ -22022,7 +22065,7 @@ "pull_request:46073|pull_request:46196", "pull_request:46154|pull_request:46196" ], - "prepared_review_unit_hash": "8cc77eaf8f2c4aa7a3599a30514711d515939197376a125cfc34db36c42475a7", + "prepared_review_unit_hash": "9df4f97523bdf2b2ef90866a130bfb7224a30675f218ea0df4dd801627bc3e32", "component_budget": { "node_count": 3, "item_count": 3, @@ -22044,30 +22087,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three PRs are about adding SDPA dispatch support, but they target different model families and code paths, so they are related theme-wise rather than duplicates.", - "confidence": 0.82, + "summary": "All three PRs are separate model-specific SDPA dispatch enablement changes. They share a common theme, but they do not look like the same concrete bug/fix and would not plausibly be merged as one PR.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46196 is the broadest and most complete representative of the SDPA-dispatch work, covering multiple affected model implementations and tests, so it is the best canonical PR among these candidates.", + "canonical_pr_reason": "PR 46196 is the broadest representative of the shared SDPA-dispatch theme, spanning multiple model families and their tests, so it is the best cluster anchor among the three.", "best_issue_reason": null, - "best_pr_reason": "PR 46196 is the best overall representative because it spans the widest set of model changes and test coverage, making it the strongest canonical example for this SDPA support effort.", + "best_pr_reason": "PR 46196 is the most comprehensive and central example of the shared feature work, covering several affected model implementations rather than a single model.", "soft_edge_verdicts": [ { "left": "pull_request:46073", "right": "pull_request:46154", "accept": false, - "reason": "Both add SDPA dispatch, but CTRL and T5 are different model implementations with different attention code paths; they are not the same underlying change." + "reason": "Both add SDPA dispatch support, but for different model implementations (CTRL vs T5) with different code paths and tests; same theme, not the same change." }, { "left": "pull_request:46073", "right": "pull_request:46196", "accept": false, - "reason": "CTRL support and wav2vec2/conformer + SeamlessM4T support are separate model-specific implementations; they could not plausibly be merged into one PR as a duplicate." + "reason": "These target different model families and modules; 46196 is a broader multi-model change, not the same concrete fix as the CTRL-only PR." }, { "left": "pull_request:46154", "right": "pull_request:46196", "accept": false, - "reason": "T5 SDPA dispatch is a separate model-specific fix from the wav2vec2/seamless M4T work, touching different code paths and test suites." + "reason": "T5-specific SDPA dispatch work is distinct from the broader wav2vec2/SeamlessM4T patch; they share a feature goal but not a mergeable single code-path change." } ] }, @@ -22097,7 +22140,7 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "1a3099bd0e1cc2c367192fa6fcf9636f28bc24ba71205f08c09302fcc1e0ac62", + "prepared_review_unit_hash": "cec03688df403c6cd08df80f1572f22f0d74c3a16374c0bad5fd2780bbe55309", "component_budget": { "node_count": 2, "item_count": 2, @@ -22119,18 +22162,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs overlap in distributed files, but they are not the same change: #46126 is a focused EP + Trainer integration, while #46269 is a broad distributed branch base touching many subsystems. The soft link should be rejected.", - "confidence": 0.94, + "summary": "These are related distributed-training PRs, but they are not duplicates: one is a focused EP + Trainer integration, while the other is a much broader distributed branch base touching many subsystems. They share some files because the larger PR likely builds the foundation the smaller one uses, not because they fix the same concrete change.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "#46269 is the broader umbrella distributed branch base and is the better representative PR for the cluster, but it does not duplicate #46126\u2019s narrower EP/Trainer integration work.", + "canonical_pr_reason": "PR #46269 is the better canonical representative because it is the broader foundational distributed branch, spans many more affected components, and appears to subsume the shared infrastructure that #46126 builds on.", "best_issue_reason": null, - "best_pr_reason": "#46269 is the strongest single PR to keep as the representative of this cluster because it subsumes a wide distributed refactor/integration set; #46126 is a narrower follow-up feature PR, not the same concrete change.", + "best_pr_reason": "PR #46269 is the best single representative for the cluster since it is the larger umbrella change and the shared-file overlap is better explained by dependency on the same distributed infrastructure than by duplicate scope.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Shared filenames reflect a common distributed codebase, but the PRs target different work: EP+Trainer integration versus a large distributed branch base. They are not the same concrete fix/change and would not plausibly be merged into one PR as duplicates." + "reason": "Reject: #46126 is a targeted EP + Trainer integration PR, while #46269 is a broad distributed branch base spanning FSDP, tensor parallel, integrations, and model changes. They overlap in infrastructure files but do not appear to be the same concrete change or a mergeable single PR." } ] }, @@ -22178,29 +22221,29 @@ "pull_request:46502|pull_request:46504", "pull_request:46503|pull_request:46504" ], - "prepared_review_unit_hash": "831fe701739fb2510d935a669e75fcd89dfe2bc06f10cac236a05dab72c0e89f", + "prepared_review_unit_hash": "4f2b0604cdb15736b4404874325136cce6a36f1ed6a29a12f33a1613f4f474a6", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 14219, - "estimated_input_tokens": 3555, - "estimated_eval_tokens": 7366 + "serialized_chars": 14226, + "estimated_input_tokens": 3557, + "estimated_eval_tokens": 7370 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 15, - "serialized_chars": 14219, - "estimated_input_tokens": 3555, - "estimated_eval_tokens": 7366 + "serialized_chars": 14226, + "estimated_input_tokens": 3557, + "estimated_eval_tokens": 7370 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are six small but unrelated workflow-maintenance pull requests. They all look superficially similar because they edit CI YAML files, but each changes a different workflow and a different behavior, so none should be merged as duplicates.", + "summary": "These are small, unrelated CI/workflow maintenance PRs. They share a superficial \u201cchore: update workflow\u201d pattern, but they do not fix the same concrete bug or change, so there is no duplicate cluster here.", "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, @@ -22209,93 +22252,93 @@ "soft_edge_verdicts": [ { "left": "pull_request:46138", - "right": "pull_request:46432", + "right": "pull_request:46504", "accept": false, - "reason": "Different workflows and different changes: self-comment CI stdin handling vs SSH runner hardening/secrets cleanup." + "reason": "Both harden shell input handling in workflows, but they affect different jobs/files and different failure modes; not the same underlying change." }, { "left": "pull_request:46138", - "right": "pull_request:46501", + "right": "pull_request:46432", "accept": false, - "reason": "Unrelated workflow edits: self-comment CI command invocation vs anti-slop permission scoping." + "reason": "One updates stdin handling for a Python helper; the other changes SSH runner setup and token handling. Different code paths and unrelated fixes." }, { "left": "pull_request:46138", - "right": "pull_request:46502", + "right": "pull_request:46503", "accept": false, - "reason": "Different workflow jobs and behaviors: self-comment CI vs reviewer assignment checkout ref." + "reason": "Benchmark trigger logic is unrelated to the self-comment CI message-passing change." }, { "left": "pull_request:46138", - "right": "pull_request:46503", + "right": "pull_request:46501", "accept": false, - "reason": "Different workflow logic: self-comment CI input handling vs benchmark trigger conditions." + "reason": "Permissions restructuring for anti-slop has no overlap with PR comment piping." }, { "left": "pull_request:46138", - "right": "pull_request:46504", + "right": "pull_request:46502", "accept": false, - "reason": "Different code paths: self-comment CI stdin piping vs build-ci-docker-images commit-message parsing." + "reason": "Reviewer checkout ref selection is a different workflow concern than handling PR comment input." }, { "left": "pull_request:46432", - "right": "pull_request:46501", + "right": "pull_request:46503", "accept": false, - "reason": "Both touch workflows, but one hardens SSH runner setup and the other changes anti-slop permissions; not the same fix." + "reason": "SSH runner security/setup changes and benchmark dispatch support are separate workflow changes." }, { "left": "pull_request:46432", - "right": "pull_request:46502", + "right": "pull_request:46501", "accept": false, - "reason": "Distinct workflow concerns: SSH runner hardening vs reviewer assignment checkout behavior." + "reason": "Different workflow files and unrelated permission changes; not the same bug." }, { "left": "pull_request:46432", - "right": "pull_request:46503", + "right": "pull_request:46502", "accept": false, - "reason": "Different workflows and intents: SSH runner secret handling vs benchmark dispatch conditions." + "reason": "Checkout base ref behavior and SSH runner hardening are unrelated." }, { "left": "pull_request:46432", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and purposes: SSH runner setup vs Docker image build trigger parsing." + "reason": "Both touch workflow scripts, but one is SSH runner security/setup and the other is commit-message parsing in a Docker-image workflow." }, { "left": "pull_request:46501", - "right": "pull_request:46502", + "right": "pull_request:46503", "accept": false, - "reason": "Separate workflow files with different fixes: anti-slop permissions vs assign-reviewers checkout ref." + "reason": "Anti-slop permissions and benchmark workflow dispatch are unrelated changes." }, { "left": "pull_request:46501", - "right": "pull_request:46503", + "right": "pull_request:46502", "accept": false, - "reason": "Different workflow behavior: permission scoping for anti-slop vs benchmark_v2 event gating." + "reason": "They modify different workflow behaviors with no shared bug or code path." }, { "left": "pull_request:46501", "right": "pull_request:46504", "accept": false, - "reason": "Not the same change: anti-slop permissions vs build-ci-docker-images commit-message handling." + "reason": "No shared underlying issue; one is permissions scoping, the other is shell-safe commit message inspection." }, { "left": "pull_request:46502", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow targets and code paths: reviewer assignment checkout vs benchmark dispatch conditions." + "reason": "Reviewer assignment checkout ref and benchmark dispatch are unrelated workflow updates." }, { "left": "pull_request:46502", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and fixes: reviewer assignment checkout ref vs Docker build trigger parsing." + "reason": "Different CI jobs and different behavioral fixes; not duplicates." }, { "left": "pull_request:46503", "right": "pull_request:46504", "accept": false, - "reason": "Both are workflow tweaks, but one changes benchmark triggers and the other changes build-image commit-message detection; not duplicates." + "reason": "Benchmark dispatch support and commit-message parsing in build-image automation are separate changes." } ] }, @@ -22325,12 +22368,12 @@ "soft_pairs": [ "pull_request:46152|pull_request:46207" ], - "prepared_review_unit_hash": "c92d4c1779fd49d8e68ffdd9b79deca45c6469cdfeaf2f6c84bdf508885d9644", + "prepared_review_unit_hash": "7540af6984132a967ef84606ef296b44a8618f79389a899c45ae632e80a2c917", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4276, + "serialized_chars": 4275, "estimated_input_tokens": 1069, "estimated_eval_tokens": 2394 }, @@ -22338,7 +22381,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4276, + "serialized_chars": 4275, "estimated_input_tokens": 1069, "estimated_eval_tokens": 2394 }, @@ -22347,18 +22390,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about replacing file-open source inspection patterns, but they touch different code paths and helpers: one is in import_utils for resolving package resources, the other is in modeling_utils for attention-implementation support detection. They are related in theme, not duplicate enough to merge together.", - "confidence": 0.91, + "summary": "Two unrelated source-reading refactors in different code paths; they share the same general theme but not the same underlying bug/change.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46207 is the cleaner representative because it targets a specific, well-scoped bug in one model utility path and has a focused title/diff.", + "canonical_pr_reason": "PR 46207 is the better canonical PR because it is more concrete and narrowly scoped: it updates the _can_set_attn_implementation source-detection path and includes a test change, making it easier to triage than the broader import-utils refactor.", "best_issue_reason": null, - "best_pr_reason": "PR 46207 is the best fit for the cluster because it is a concrete behavioral change in a single code path, whereas PR 46152 is broader infrastructure work in a different module.", + "best_pr_reason": "PR 46207 is the strongest overall representative of the cluster because it has a clear, specific behavior change in one code path and a matching test update; PR 46152 is a separate implementation cleanup in another module.", "soft_edge_verdicts": [ { "left": "pull_request:46152", "right": "pull_request:46207", "accept": false, - "reason": "Same general theme of avoiding open() for source inspection, but they modify different files, different helpers, and different runtime paths; they do not look like the same underlying bug or a single mergeable PR." + "reason": "Both avoid reading source via open(), but they change different modules and different logic paths. They are similar refactors, not the same concrete bug or mergeable change." } ] }, @@ -22390,7 +22433,7 @@ "pull_request:46436|pull_request:46474", "pull_request:46173|pull_request:46474" ], - "prepared_review_unit_hash": "16f86cb2347779e32556908dc2d533de640e90abc199e2e88ce50a83a484cdd8", + "prepared_review_unit_hash": "f21f88628ca440bf95d7ecd8614236e429fa557368f6b97d17d2b1104a059a70", "component_budget": { "node_count": 3, "item_count": 3, @@ -22412,30 +22455,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two PRs are near-duplicates of the same StaticLayer type-hint fix; one PR is unrelated and should stay separate.", - "confidence": 0.98, + "summary": "Two PRs (46173 and 46474) are effectively the same StaticLayer.get_seq_length typing fix, while 46436 is a separate behavioral change in static-cache setup. I treat 46173/46474 as duplicates and keep 46173 as the canonical representative; 46436 is the strongest standalone PR in the cluster.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:46173 is the best representative of the duplicate StaticLayer get_seq_length type-hint fix: it matches the same file, same issue target (#45987), and the diff is effectively identical to pull_request:46474.", + "canonical_pr_reason": "PR 46173 is the earlier, slightly more active version of the identical StaticLayer.get_seq_length return-type fix, and it directly targets the linked issue #45987.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46173 is the strongest canonical PR for the cluster because it is the same change as the other duplicate, has the explicit issue link to #45987, and carries more discussion/context than the later twin.", + "best_pr_reason": "PR 46436 is the only substantive behavior change in the cluster: it makes generation respect the configured static cache max length instead of only the computed maximum.", "soft_edge_verdicts": [ { "left": "pull_request:46436", "right": "pull_request:46474", "accept": false, - "reason": "Different bugs: 46436 changes generation cache preparation to honor max_cache_len config, while 46474 only widens the StaticLayer.get_seq_length type annotation. Same subsystem, but not the same underlying change." + "reason": "Different fixes and code paths: 46436 changes generation cache initialization to honor max_cache_len, while 46474 only relaxes a type hint in cache_utils. Same subsystem, but not the same underlying bug/change." }, { "left": "pull_request:46173", "right": "pull_request:46474", "accept": true, - "reason": "Same concrete fix: both patch src/transformers/cache_utils.py to change StaticLayer.get_seq_length from int to int | torch.Tensor, with effectively identical diff and the same issue target #45987." + "reason": "These are the same concrete change: both update StaticLayer.get_seq_length's type annotation and implementation in cache_utils, with the same linked issue #45987." } ] }, "evaluator_result": { "accept": true, - "feedback": "The assessment is grounded in the packet. The 46173/46474 pair is correctly treated as the same StaticLayer get_seq_length type-hint fix: same file, same issue target, and effectively identical diff. The 46436/46474 pair is correctly kept separate because it changes generation cache preparation behavior in different files and targets a different issue. The canonical PR choice is conservative and supported by the available metadata." + "feedback": "The summary is grounded in the packet. 46173 and 46474 share the same file, issue target, and identical diff preview, so treating them as duplicates is conservative. 46436 is a separate change in generation cache handling, with different files and a different issue target, so keeping it separate is also well supported." }, "evaluator_used": true, "retried": false, @@ -22525,40 +22568,40 @@ "soft_pairs": [ "pull_request:46376|pull_request:46417" ], - "prepared_review_unit_hash": "bf87f9af90f994978f7b99e59a1fe902fe2ba75fe85440e929164698da2f1ed1", + "prepared_review_unit_hash": "9cf1417a8ad023e3fbc849f514b594650edca756cbdef0664d1f19d2beb4f090", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5959, - "estimated_input_tokens": 1490, - "estimated_eval_tokens": 3236 + "serialized_chars": 5938, + "estimated_input_tokens": 1485, + "estimated_eval_tokens": 3226 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5959, - "estimated_input_tokens": 1490, - "estimated_eval_tokens": 3236 + "serialized_chars": 5938, + "estimated_input_tokens": 1485, + "estimated_eval_tokens": 3226 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are unrelated: one adds multimodal support to continuous batching internals, while the other introduces a new HunYuan VL model package, docs, and tests. They do not look like the same underlying change or bug fix, so the soft link should be rejected.", - "confidence": 0.98, + "summary": "The two PRs are unrelated: one adds multimodal support to continuous batching infrastructure, while the other introduces the HunYuan VL model, docs, processors, and tests. They share only a broad ML/transformers context, not the same concrete change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR 46417 is the stronger representative of the cluster because it is a complete, self-contained model addition with code, docs, and tests; PR 46376 is a draft infrastructure feature that is not comparable as a duplicate candidate.", "best_issue_reason": null, - "best_pr_reason": "PR #46417 is the stronger standalone change: it is a focused model addition with accompanying docs, auto-mapping updates, and tests, whereas PR #46376 is a draft, broad infrastructure change still in progress.", + "best_pr_reason": "PR 46417 is the best standalone PR in this set: it has a clear deliverable, broader implementation coverage, and appears more complete than the draft multimodal batching PR.", "soft_edge_verdicts": [ { "left": "pull_request:46376", "right": "pull_request:46417", "accept": false, - "reason": "Different scopes and code paths: continuous batching multimodal support vs. adding a new HunYuan VL model stack. They are not plausibly mergeable into one duplicate PR." + "reason": "Different changes and code paths: continuous batching multimodal support vs. adding the HunYuan VL model. They are not the same underlying bug or change, and would not plausibly be merged into one PR." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index c4c8880d4242752f28ae2aa859c3d813ec722880..d0b4e763ab14bfb3d1871bf64a89f993e8b01877 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260610t060026z", + "analysis_id": "hybrid-model-20260610t120053z", "archived_artifacts": { - "hybrid": "snapshots/20260610T060026Z/analysis-runs/hybrid-model-20260610t060026z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260610T060026Z/analysis-runs/hybrid-model-20260610t060026z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-06-10T06:04:35Z", + "published_at": "2026-06-10T12:12:03Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260610T060026Z", + "snapshot_id": "20260610T120053Z", "variant": "hybrid" } diff --git a/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..30c3eed06d5a76b1796f58cf9a81bbe837204f3b --- /dev/null +++ b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json @@ -0,0 +1,2138 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260610T120053Z", + "generated_at": "2026-06-10T12:12:01Z", + "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-41115-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41115.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41115, + "canonical_pr_number": 45613, + "issue_numbers": [ + 41115 + ], + "pr_numbers": [ + 41116, + 45613 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41116, + "right_pr_number": 45613, + "code_similarity": 0.797, + "size_similarity": 0.885, + "file_overlap": 0.75, + "area_overlap": 0.831, + "patch_similarity": 0.693, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "docs/source/en/model_doc/minicpm3.md", + "src/transformers/models/auto/modeling_auto.py", + "src/transformers/models/minicpm3/__init__.py", + "src/transformers/models/minicpm3/configuration_minicpm3.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/minicpm3/modular_minicpm3.py", + "tests/models/minicpm3/__init__.py", + "tests/models/minicpm3/test_modeling_minicpm3.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 713, + 720 + ] + ], + "right_ranges": [ + [ + 730, + 737 + ] + ] + }, + { + "filename": "docs/source/en/model_doc/minicpm3.md", + "left_ranges": [ + [ + 1, + 67 + ] + ], + "right_ranges": [ + [ + 1, + 45 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 285, + 291 + ], + [ + 700, + 706 + ], + [ + 1298, + 1304 + ] + ], + "right_ranges": [ + [ + 720, + 726 + ], + [ + 1331, + 1337 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/__init__.py", + "left_ranges": [ + [ + 1, + 29 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/configuration_minicpm3.py", + "left_ranges": [ + [ + 1, + 141 + ] + ], + "right_ranges": [ + [ + 1, + 126 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 544 + ] + ], + "right_ranges": [ + [ + 1, + 522 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modular_minicpm3.py", + "left_ranges": [ + [ + 1, + 444 + ] + ], + "right_ranges": [ + [ + 1, + 342 + ] + ] + }, + { + "filename": "tests/models/minicpm3/test_modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 109 + ] + ], + "right_ranges": [ + [ + 1, + 136 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45647-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45647.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 45647, + "canonical_pr_number": 45738, + "issue_numbers": [ + 45647 + ], + "pr_numbers": [ + 45737, + 45738 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 45737, + "right_pr_number": 45738, + "code_similarity": 0.526, + "size_similarity": 0.11, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.024, + "shared_filenames": [ + "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py", + "left_ranges": [ + [ + 491, + 497 + ] + ], + "right_ranges": [ + [ + 491, + 497 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45987-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45987.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 45987, + "canonical_pr_number": 46173, + "issue_numbers": [ + 45987 + ], + "pr_numbers": [ + 46173, + 46474 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 46173, + "right_pr_number": 46474, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/cache_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cache_utils.py", + "left_ranges": [ + [ + 69, + 75 + ], + [ + 375, + 381 + ] + ], + "right_ranges": [ + [ + 69, + 75 + ], + [ + 375, + 381 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46291-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46291.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 46291, + "canonical_pr_number": 46293, + "issue_numbers": [ + 46291 + ], + "pr_numbers": [ + 46293, + 46478 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 46293, + "right_pr_number": 46478, + "code_similarity": 0.483, + "size_similarity": 0.325, + "file_overlap": 0.333, + "area_overlap": 0.862, + "patch_similarity": 0.111, + "shared_filenames": [ + "src/transformers/integrations/hub_kernels.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 89, + 180 + ], + [ + 233, + 247 + ] + ], + "right_ranges": [ + [ + 89, + 181 + ], + [ + 234, + 248 + ], + [ + 263, + 269 + ], + [ + 360, + 366 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45561-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45561, + "canonical_pr_number": 45645, + "issue_numbers": [ + 45561 + ], + "pr_numbers": [ + 45639, + 45645 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45639, + "right_pr_number": 45645, + "code_similarity": 0.494, + "size_similarity": 0.619, + "file_overlap": 0.5, + "area_overlap": 0.434, + "patch_similarity": 0.451, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3595, + 3602 + ], + [ + 3819, + 3853 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 86 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45081-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45081, + "canonical_pr_number": 45317, + "issue_numbers": [ + 45081 + ], + "pr_numbers": [ + 45086, + 45317 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45086, + "right_pr_number": 45317, + "code_similarity": 0.589, + "size_similarity": 0.34, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.136, + "shared_filenames": [ + "src/transformers/tokenization_utils_tokenizers.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/tokenization_utils_tokenizers.py", + "left_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ], + "right_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-43698-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43698.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43698, + "canonical_pr_number": 43779, + "issue_numbers": [ + 43698 + ], + "pr_numbers": [ + 43779, + 43816 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 43779, + "right_pr_number": 43816, + "code_similarity": 0.418, + "size_similarity": 0.538, + "file_overlap": 1.0, + "area_overlap": 0.02, + "patch_similarity": 0.02, + "shared_filenames": [ + "src/transformers/integrations/integration_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/integration_utils.py", + "left_ranges": [ + [ + 2227, + 2241 + ], + [ + 2303, + 2309 + ] + ], + "right_ranges": [ + [ + 2278, + 2291 + ], + [ + 2309, + 2322 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [], + "duplicate_prs": [ + { + "cluster_id": "cluster-41115-3", + "canonical_pr_number": 45613, + "duplicate_pr_numbers": [ + 41116 + ], + "target_issue_number": 41115, + "reason": "PRs in cluster-41115-3 are treated as duplicates because they converge on issue #41115 with closing_reference evidence." + }, + { + "cluster_id": "cluster-41211-3", + "canonical_pr_number": 44339, + "duplicate_pr_numbers": [ + 41356 + ], + "target_issue_number": 41211, + "reason": "PRs in cluster-41211-3 are treated as duplicates because they converge on issue #41211 with closing_reference evidence." + }, + { + "cluster_id": "cluster-43240-3", + "canonical_pr_number": 43251, + "duplicate_pr_numbers": [ + 43254 + ], + "target_issue_number": 43240, + "reason": "PRs in cluster-43240-3 are treated as duplicates because they converge on issue #43240 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43698-3", + "canonical_pr_number": 43779, + "duplicate_pr_numbers": [ + 43816 + ], + "target_issue_number": 43698, + "reason": "PRs in cluster-43698-3 are treated as duplicates because they converge on issue #43698 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43979-11", + "canonical_pr_number": 44007, + "duplicate_pr_numbers": [ + 43996, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "target_issue_number": 43979, + "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45561-3", + "canonical_pr_number": 45645, + "duplicate_pr_numbers": [ + 45639 + ], + "target_issue_number": 45561, + "reason": "PRs in cluster-45561-3 are treated as duplicates because they converge on issue #45561 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45647-3", + "canonical_pr_number": 45738, + "duplicate_pr_numbers": [ + 45737 + ], + "target_issue_number": 45647, + "reason": "PRs in cluster-45647-3 are treated as duplicates because they converge on issue #45647 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-45987-3", + "canonical_pr_number": 46173, + "duplicate_pr_numbers": [ + 46474 + ], + "target_issue_number": 45987, + "reason": "PRs in cluster-45987-3 are treated as duplicates because they converge on issue #45987 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46291-3", + "canonical_pr_number": 46293, + "duplicate_pr_numbers": [ + 46478 + ], + "target_issue_number": 46291, + "reason": "PRs in cluster-46291-3 are treated as duplicates because they converge on issue #46291 with closing_reference evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-41211-3", + "issue_number": 41211, + "reason": "Issue #41211 is the strongest global issue candidate because it is open, belongs to a cluster with 3 artifacts, and carries 6 discussion comments plus 6 inbound references.", + "score": 148.0 + }, + "best_pr": { + "cluster_id": "cluster-43979-11", + "pr_number": 44007, + "reason": "PR #44007 is the strongest global PR candidate because it is open, belongs to a cluster with 11 artifacts, links to 1 issues, and carries 2 review/discussion events.", + "score": 238.886 + } +} diff --git a/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..57680c707ddb183bf3d810d3584519a7878f92b7 --- /dev/null +++ b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,22616 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260610T120053Z", + "generated_at": "2026-06-10T12:12:01Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43425", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43866", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44805", + "issue:44918", + "issue:45081", + "issue:45161", + "issue:45237", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45800", + "issue:45823", + "issue:45859", + "issue:45874", + "issue:45901", + "issue:45902", + "issue:46015", + "issue:46165", + "issue:46294", + "issue:46304", + "issue:46402", + "issue:46468", + "issue:46515" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:46468", + "issue:45237|issue:45507", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:44509|issue:44512", + "issue:35545|issue:36010", + "issue:44918|issue:46165", + "issue:44805|issue:45823", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:46294", + "issue:44805|issue:45161", + "issue:46165|issue:46468", + "issue:44560|issue:46165", + "issue:43596|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43039|issue:45161", + "issue:43975|issue:44457", + "issue:44292|issue:44457", + "issue:34928|issue:44805", + "issue:24643|issue:30277", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45874|issue:46294", + "issue:44292|issue:45237", + "issue:39290|issue:45081", + "issue:45800|issue:46515", + "issue:45823|issue:45874", + "issue:43425|issue:44292", + "issue:34928|issue:41720", + "issue:43828|issue:45161", + "issue:43828|issue:44292", + "issue:43541|issue:43716", + "issue:43716|issue:46468", + "issue:43716|issue:45685", + "issue:44918|issue:45161", + "issue:44292|issue:44918", + "issue:45507|issue:45685", + "issue:43716|issue:44292", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43828|issue:45685", + "issue:44918|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:34928|issue:36010", + "issue:43039|issue:44805", + "issue:45237|issue:46015", + "issue:43866|issue:46402", + "issue:45800|issue:46304" + ], + "prepared_review_unit_hash": "a370e3b41d1e9d19bffbc775c368791147f3d06736cc5b802b108fbe72314601", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 46528, + "estimated_input_tokens": 11632, + "estimated_eval_tokens": 23520 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous. The packet supports two likely duplicate issue pairs: the ZeRO-3/BertModel index-out-of-bounds reports and the torch 2.4.1 incompatibility reports. The docs pair is related but too broad to call a true duplicate from this packet alone.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43596 is the clearest representative of the shared ZeRO-3/BertModel index-out-of-bounds failure; issue:43638 looks like the same underlying bug with a slightly different model/setup description.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43596 is the best global issue canonical because it states the concrete failure mode most cleanly and matches the strongest duplicate pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failures: a missing config attribute vs a streaming chat/completions bug." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Both are runtime/backend issues, but they involve different models and different failure modes." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "One is an image dtype mismatch; the other is an AMD GPU support problem for GPT-OSS." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Autocast dtype mismatch for Phi-tiny-MoE is unrelated to the GPT-OSS AMD runtime issue." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Different models and different breakage: AMD GPU support vs a broken official Nemotron example." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "These describe different model/runtime problems and do not point to the same code path." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ONNX export errors." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "Different models and different export/runtime failures; not the same bug." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "ONNX export error is unrelated to Qwen3 auto device mapping CUDA asserts." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Table QA crash and outdated RAG examples are different issues." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are stale v5 docs issues, but they reference different removed artifacts and are too broad to treat as the same duplicate." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ONNX export failure and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "TRL SFT embedding unpacking and resume-training/trust-remote-code behavior are different problems." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Different symptoms and subsystems: a mask-shape IndexError vs device-map forward mismatch." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both mention indexing errors, but the reported failures and triggers are different enough to keep separate." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 Bert indexing error is unrelated to model resume with trust_remote_code." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Same broad area, but not the same concrete bug: BertModel ZeRO-3 init error vs a mask/tensor shape mismatch." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal/input-shape issues with no shared failure path." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Mask-shape IndexError is not the same as GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Resume-training issue and Nemotron example breakage are unrelated." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Video StopIteration in chat templating is unrelated to resume-training behavior." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "Same ZeRO-3 theme does not make these the same bug; the underlying failures differ." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "BarkProcessor voice_preset and ModernBERT ONNX export are different model-specific issues." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Both use DeepSpeed, but the collective mismatch and Bert zero-sized index error are different code paths." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Both report the same ZeRO-3/BertModel index-out-of-bounds failure during model init/load." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross_entropy routing and GPT-OSS tensor-parallel failure are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Detokenization bug and LoRA merge/save mismatch are different model behaviors." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen NVFP4 runtime error and LoRA save/reload inconsistency are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "FSDP checkpointing mismatch and mask/index errors are not the same bug." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "DeepSpeed training with 2-D weight error and collective mismatch are distinct failure modes." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": false, + "reason": "Same module, but CPU offload incompatibility and uninitialized-memory-after-to_empty are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "Bert ZeRO-3 indexing error is unrelated to GPT-OSS TP failure." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "Activation-checkpointing/FSDP mismatch is unrelated to the CustomTrainer compute_loss signature change." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 Bert loading error and GPT-OSS tensor-parallel failure are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific issues with no shared bug." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Reverse embedding lookup cost and video token mismatch are different multimodal issues." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Qwen NVFP4 runtime error and GPT-OSS AMD GPU failure are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3 config/vLLM attribute error is unrelated to the Mistral regex patch crash." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": true, + "reason": "Both report the same torch 2.4.1 incompatibility regression; the titles differ only in wording/detail." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Device-map forward mismatch and reverse embedding lookup slowdown are different Gemma4 problems." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Torch version incompatibility and Qwen NVFP4 runtime failure are not the same issue." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP activation checkpointing and Qwen3 auto device mapping CUDA assert are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch and MPS histogram integer error are different backend-specific bugs." + }, + { + "left": "issue:44918", + "right": "issue:45685", + "accept": false, + "reason": "TRL SFT embedding issue and MPS histogram failure do not share a code path." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Liger Kernel cross_entropy usage is unrelated to a Phi MoE autocast mismatch." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "DeepSpeed collective mismatch and compute_loss signature incompatibility are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "Activation checkpointing/FSDP mismatch and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger Kernel cross_entropy routing and an index/mask shape error are unrelated." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "GPT-OSS AMD runtime issue and outdated RAG examples are unrelated." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "Corrupted checkpoint and missing Sam3Model text encoder weights are different load-time problems." + }, + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Torch 2.4.1 compatibility and USE_DISTRIBUTED=0 support regression are different compatibility issues." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Mostly grounded, but not conservative enough overall. The ZeRO-3/BertModel pair (43596/43638) and torch 2.4.1 pair (45800/46515) look supportable from the titles alone, but the docs pair (44509/44512) is too broad to call a true duplicate from this packet: both are stale-v5 documentation issues, yet they reference different removed artifacts. The summary also overstates the result by saying there are 'true duplicate pairs' rather than 'likely related' pairs." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 2, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36010", + "issue:39290", + "issue:41093", + "issue:41720", + "issue:41762", + "issue:43039", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43792", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44387", + "issue:44512", + "issue:44560", + "issue:44589", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45161", + "issue:45237", + "issue:45362", + "issue:45464", + "issue:45685", + "issue:45800", + "issue:45812", + "issue:45823", + "issue:45901", + "issue:45920", + "issue:46015", + "issue:46294", + "issue:46304" + ], + "soft_pairs": [ + "issue:43296|issue:46304", + "issue:41762|issue:43596", + "issue:39290|issue:41720", + "issue:44589|issue:46304", + "issue:43901|issue:44512", + "issue:45823|issue:46294", + "issue:45237|issue:45901", + "issue:44841|issue:45823", + "issue:43039|issue:44560", + "issue:43975|issue:46015", + "issue:44863|issue:46015", + "issue:45020|issue:46304", + "issue:45812|issue:45920", + "issue:43866|issue:44863", + "issue:44292|issue:46015", + "issue:43872|issue:46304", + "issue:39290|issue:43531", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43854|issue:43866", + "issue:43425|issue:45070", + "issue:43596|issue:45800", + "issue:43975|issue:45901", + "issue:43792|issue:45070", + "issue:45362|issue:45464", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:45464|issue:45685", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:43404|issue:43866", + "issue:44841|issue:45084", + "issue:36010|issue:39290", + "issue:41093|issue:41762", + "issue:45800|issue:45823", + "issue:43296|issue:43366", + "issue:43572|issue:45070", + "issue:44805|issue:44918", + "issue:44387|issue:45005", + "issue:43541|issue:45161", + "issue:43828|issue:45464" + ], + "prepared_review_unit_hash": "019191a6dede7d5eec4b1c15214c43516e47c0a245ca83adc75399689cf3eade", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 41, + "serialized_chars": 39731, + "estimated_input_tokens": 9933, + "estimated_eval_tokens": 20122 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous. Only two soft edges look like true duplicates: the ZeRO-3 load-time IndexError pair (41762/43596) and the broad AutoTokenizer wrong-token-ID regression pair (45812/45920). The mask-shape IndexError pairs are rejected because the titles point to different failure shapes and code paths, not the same underlying bug.", + "confidence": 0.88, + "canonical_issue_reason": "45920 is the best canonical issue because it captures the broad multi-model AutoTokenizer v5 wrong-token-ID regression and subsumes the narrower Granite-specific report in 45812.", + "canonical_pr_reason": null, + "best_issue_reason": "45920 is the strongest overall representative issue: it is the broadest tokenizer regression report and matches the accepted duplicate pair against 45812.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Different bug classes: vLLM model load/support failure vs a PyTorch distributed build/runtime compatibility break." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Both describe the same ZeRO-3 load-time IndexError on zero-sized tensors; only the model/context differs." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Both are model-loading failures, but the titles indicate different underlying bugs and error paths." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Float8 storage lookup failure is unrelated to the distributed build/runtime incompatibility." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs-related, but they describe different documentation regressions." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Different model families and different symptoms: device mismatch vs video/chat-template feature mismatch." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Completely different subsystems and error surfaces; not the same bug." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Processor failure and device-map mismatch are different code paths and different problems." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger-kernel cross-entropy behavior is unrelated to the video StopIteration error." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Tokenizer detokenization regression and outdated RAG examples are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "Checkpoint-loading failure and outdated documentation are unrelated." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Remote-code breakage is a different compatibility problem than the USE_DISTRIBUTED=0 issue." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same AutoTokenizer v5 wrong-token-ID regression; 45920 is the broader multi-model report." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Corrupted checkpoint vs implementation load failure are not the same underlying bug." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Model runtime failure and RAG docs are unrelated." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes constructor incompatibility is unrelated to the PyTorch distributed build/runtime issue." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but one is a missing config attribute and the other is a different Qwen3-MoE sliding_window bug." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both are IndexErrors, but one is a mask-shape mismatch and the other is a zero-sized tensor access under ZeRO-3." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper runtime failure and table QA crash are unrelated." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Different model-loading failures with different causes." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch version incompatibility and a pydantic field regression are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "ZeRO-3 loading IndexError and torch/transformers version compatibility are different issues." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization regression and table QA crash are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Different subsystems and different errors." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Both are Qwen3.5-related, but they describe different failures and different interfaces." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Different model/runtime problem versus table QA crash." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Different models and different failure modes." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Qwen3.5 streaming failure is unrelated to the MPS unsupported-op error." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Different model/math-path error versus MPS unsupported-op error." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Both involve MoE/precision/tracing themes, but the concrete failures and code paths differ." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Weight-tying bug and corrupted checkpoint are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Similar-looking IndexErrors, but the titles point to different operations and different failure points." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "GenerationMixin import error is unrelated to the Gemma3 sliding_window_pattern failure." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are IndexErrors, but one is a mask-shape mismatch and the other is a ZeRO-3 zero-sized tensor access." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch/transformers version compatibility is unrelated to Gemma4 device placement." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL load failure and gpt-oss GGUF support are different model-support issues." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Missing pad_token_idx in a config and a pydantic-field regression are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Different stack traces and different failing operations; not the same bug." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "int4 OOM regression and tied-weights translation-model issue are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Different MoE-related failures with different symptoms and code paths." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Autocast dtype mismatch and Qwen3.5 streaming API failure are unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Most of the pairwise calls are conservative and grounded in the titles, and the tokenizer anchor on 45920 looks reasonable. However, the summary overstates the evidence by saying there are genuine duplicate mask-shape IndexError pairs, even though the corresponding verdicts are both rejects. Please align the summary with the actual soft-edge outcomes and avoid calling those cases duplicates unless you can justify it from the packet." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 3, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36331", + "issue:39186", + "issue:39692", + "issue:40990", + "issue:41093", + "issue:41762", + "issue:42915", + "issue:43054", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43644", + "issue:43645", + "issue:43716", + "issue:43873", + "issue:43950", + "issue:43975", + "issue:43994", + "issue:44291", + "issue:44292", + "issue:44360", + "issue:44479", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44918", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45183", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45632", + "issue:45698", + "issue:45759", + "issue:45800", + "issue:45902", + "issue:46294", + "issue:46463", + "issue:46489", + "issue:46491" + ], + "soft_pairs": [ + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:43531|issue:45070", + "issue:43716|issue:46294", + "issue:45020|issue:45800", + "issue:43541|issue:44589", + "issue:44479|issue:46294", + "issue:41093|issue:43596", + "issue:43054|issue:43994", + "issue:45759|issue:45902", + "issue:39186|issue:44918", + "issue:43716|issue:44918", + "issue:40990|issue:44960", + "issue:43645|issue:45399", + "issue:45632|issue:45698", + "issue:43404|issue:45005", + "issue:42915|issue:45399", + "issue:44530|issue:45005", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:44863|issue:45005", + "issue:45183|issue:46463", + "issue:46489|issue:46491", + "issue:44918|issue:45507", + "issue:44492|issue:44512", + "issue:43975|issue:44292", + "issue:44292|issue:45507", + "issue:43950|issue:44534", + "issue:39692|issue:43054", + "issue:43644|issue:43950", + "issue:36331|issue:41762" + ], + "prepared_review_unit_hash": "80bed78dbc4b86e7f9d2a05dc6818c90ecf30c49e8f735370d050e6af059dc47", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36940, + "estimated_input_tokens": 9235, + "estimated_eval_tokens": 18726 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a noisy mix of distinct Transformers issues that only share broad wording (v5 regressions, model-loading bugs, tokenizer bugs, etc.). I found one clear duplicate family around non-persistent buffer corruption and one likely duplicate around input validation error messaging; the rest are too different to merge.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 43950 is the clearest representative of the strongest duplicate family here: it states the non-persistent-buffer corruption regression explicitly and matches the near-identical v5 reports in the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "43950 is the best issue anchor because it names the root bug most precisely and is the most actionable summary of the duplicate buffer-corruption reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Different bugs: unexpected init argument handling vs pydantic field compatibility." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "Unrelated symptoms and code paths: GGUF architecture support vs streaming chat/completions failure." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to pydantic config regression." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal issues: image preprocessor dtype mismatch vs video feature/token count mismatch." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote-code breakage and torch/transformers incompatibility are separate root causes." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Grouped_mm tracing failure is not the same as missing Float8 storage type support." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-related, but one is a v5.3 regression across models and the other is a specific chat-template mismatch." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both are shape/index errors, but they involve different model setups and failure modes." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Different SigLIP2 problems: embedding quality complaint vs incorrect pipeline/AutoModel output." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Composite-config unwrapping bug and uninitialized rotary embedding memory are distinct issues." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP weight-dimension error is unrelated to TRL SFT embedding unpacking." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 preprocessing dtype issue and Qwen3.5 SFT trainer embedding issue are different." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "High perplexity on GPT-OSS and GLM5 are unrelated model-specific reports." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Jupyter custom-model initialization regression is not the same as flash-attn fallback gating." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both involve custom-code loading, but one is a local leaf-dir cache collision and the other is wrong module selection after save/load." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but they are model-family-specific reports without clear evidence of the same bug." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Qwen3MoE FP8 failure is unrelated to flash-attn2 fallback blocking." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash is not the same as tied-weight regressions." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization offloading behavior is unrelated to tied-weight issues." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA indexer discussion and documentation wording are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "Checkpoint-loading failure for NemotronH is not a tied-weight regression." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": true, + "reason": "Same validation code path: 46463 looks like a concrete T5-family instance of the broader input_ids/inputs_embeds XOR error-message bug." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions affecting different model families and round-trip behaviors." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "TRL embedding unpacking and GraniteMoEHybrid invalid-method calls are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A typo in docs is not the same as a removed command mention." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "DeepSeek-Coder detokenization and Qwen-3-8B-NVFP4 runtime failure are different problems." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "NVFP4 runtime failure and GraniteMoEHybrid method selection are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Near-identical regression: v5 corrupts/fills non-persistent buffers with junk; these are the same underlying bug." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "SigLIP2 doc-example errors and text-embedding quality complaints are not the same bug." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Same non-persistent-buffer corruption regression in Transformers v5, just described from different angles." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Unexpected trainer kwarg and DeepSpeed zero-3 index error are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it identifies the non-persistent-buffer corruption family and the input-validation error-message family without overextending beyond the evidence. The accepted soft-edge pairings also look defensible as same-bug families, and the other rejections are suitably cautious." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 4, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:24643", + "issue:28530", + "issue:38175", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43493", + "issue:43643", + "issue:43782", + "issue:43824", + "issue:43828", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44368", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44589", + "issue:44661", + "issue:44910", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45356", + "issue:45399", + "issue:45488", + "issue:45520", + "issue:45663", + "issue:45685", + "issue:45701", + "issue:45715", + "issue:45823", + "issue:45859", + "issue:45902", + "issue:45920", + "issue:46315", + "issue:46489" + ], + "soft_pairs": [ + "issue:43828|issue:44928", + "issue:43872|issue:44589", + "issue:45920|issue:46489", + "issue:43950|issue:45902", + "issue:43866|issue:44403", + "issue:43054|issue:45701", + "issue:38175|issue:43054", + "issue:41553|issue:43883", + "issue:44910|issue:45859", + "issue:42947|issue:44387", + "issue:45663|issue:45902", + "issue:43054|issue:43493", + "issue:41553|issue:43881", + "issue:44403|issue:44451", + "issue:24643|issue:28530", + "issue:43975|issue:45520", + "issue:44403|issue:44488", + "issue:42947|issue:43856", + "issue:43643|issue:45020", + "issue:41762|issue:42915", + "issue:45685|issue:45823", + "issue:43782|issue:44960", + "issue:43824|issue:43872", + "issue:44991|issue:45399", + "issue:43881|issue:43883", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:45081|issue:45399", + "issue:42491|issue:44368", + "issue:45715|issue:46315", + "issue:45488|issue:46489", + "issue:45020|issue:45356" + ], + "prepared_review_unit_hash": "e5cdb599d7e2ea15e317bb8ff8309e35275c065f421906120607b3edb620de1c", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 38002, + "estimated_input_tokens": 9501, + "estimated_eval_tokens": 19258 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct bugs. One soft duplicate looks plausible in the non-persistent-buffer/materialization path: a broad `from_pretrained()` buffer corruption regression and a model-specific `inv_freq` uninitialized-memory failure. The strongest issue candidate overall is the broad tokenizer regression affecting multiple model families.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45920 is the best umbrella issue: it describes a current v5 tokenizer regression with wrong token IDs across multiple model families, which is broader and more representative than the more model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45920 is the most actionable global representative because it names the shared failure mode, cites several affected models, and matches the recurring tokenizer-breakage theme in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Both involve model/runtime failures, but the concrete bugs differ: one is an autocast dtype mismatch, the other is a flash-attention packed-sequence/position_ids crash." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Both are quantization/storage compatibility errors, but they point to different failure modes and different tensor/storage classes." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but 45920 is a broad multi-model wrong-ID bug while 46489 is a specific DeepSeek-Coder v1 issue tied to a particular PR gap." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": true, + "reason": "These look like the same underlying non-persistent-buffer/materialization regression: 43950 is the generic `from_pretrained()` buffer corruption report, and 45902 is a concrete Qwen3_5Moe manifestation reading uninitialized buffer memory after meta materialization." + }, + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "One reports a corrupted checkpoint, the other just noisy loading output; they do not describe the same bug." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "SigLIP2 embedding quality and a general tokenization-version change are not the same failure mode." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is degraded text embeddings; too different to merge." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Different model-loading problems for different models and code paths." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Both touch Qwen3.5 MoE, but one is a flash-attention memory access crash and the other is CPU-offload incompatibility in rotary embeddings." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Gradient checkpointing/LoRA and int4 quantization OOM are separate training-memory regressions." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "FSDP2 shared_kv_state rebuilding and non-persistent buffer materialization are unrelated code paths." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Both concern SigLIP2, but one is relative embedding quality and the other is an implementation discrepancy versus JAX; not clearly the same bug." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Bad error message for Voxtral AutoTokenizer and glm-4v-9b loading failure are distinct issues." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic loading noise versus a specific model load failure." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "DeepSpeed training weight-shape error and early-stopping metric detection are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer detokenization and a flash-attn import error are unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Noise on load versus a specific tokenizer load failure for a different model." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "These are both training/memory-related but cover different regressions and model families." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Missing fields in `AutoConfig` with `trust_remote_code=True` is not the same as a broad remote-code compatibility regression." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "DeepSpeed ZeRO-3 loading crash and FineGrainedFP8Config failure are separate model-initialization bugs." + }, + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "MPS histogram integer support and Gemma4 device placement mismatch are different backend issues." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight-only loading and GLM5 are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "ImportError for a model class and bitsandbytes storage incompatibility are unrelated." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading for a specific model and flash-attn fallback gating are unrelated." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different models and different failure modes during loading." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Remote-code breakage and Gemma4 device mismatch are separate regressions." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Model-registration logic for tokenizer mappings is not the same as a remote-code loading regression." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Dtype mismatch in autocast and device mismatch in Gemma4 are different runtime errors." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer regex patch crash and flash-attn fallback checks are unrelated." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA/qwen3_moe hf4.x vs hf5.x compatibility and a warning about `tie_word_embeddings` are not the same issue." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "Both are tokenizer performance regressions, but one is `convert_ids_to_tokens` and the other is `convert_tokens_to_ids`; different code paths." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Both involve DeepSeek tokenizer breakage, but 45488 is a LlamaTokenizer pre-tokenizer override bug while 46489 is a DeepSeek-Coder v1 regression tied to a specific fix gap." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Generic remote-code breakage and Kimi-K2.5 tokenizer codec handling are different problems." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The chosen umbrella issue (45920) is well supported by its title as a broad v5 tokenizer regression affecting multiple model families, and the single accepted soft pair (43950/45902) is plausibly the same non-persistent-buffer/materialization path. The rejected pairs are also framed cautiously without overstating equivalence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 5, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43742", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43872", + "issue:43940", + "issue:43950", + "issue:44155", + "issue:44220", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44805", + "issue:44863", + "issue:44928", + "issue:44991", + "issue:45305", + "issue:45313", + "issue:45325", + "issue:45399", + "issue:45584", + "issue:45588", + "issue:45663", + "issue:45823", + "issue:45834", + "issue:45902", + "issue:45923", + "issue:46402", + "issue:46515" + ], + "soft_pairs": [ + "issue:43742|issue:45399", + "issue:43827|issue:44509", + "issue:43404|issue:43950", + "issue:43872|issue:44291", + "issue:42915|issue:43872", + "issue:43792|issue:44220", + "issue:42947|issue:45923", + "issue:36331|issue:41093", + "issue:43257|issue:45305", + "issue:41553|issue:44991", + "issue:43940|issue:45313", + "issue:43854|issue:44863", + "issue:44928|issue:45588", + "issue:41093|issue:44805", + "issue:44479|issue:44560", + "issue:41553|issue:44488", + "issue:44863|issue:46402", + "issue:24643|issue:30064", + "issue:43596|issue:46515", + "issue:44530|issue:44863", + "issue:43366|issue:43828", + "issue:45325|issue:45902", + "issue:43638|issue:46515", + "issue:43366|issue:43531", + "issue:45663|issue:45823", + "issue:44155|issue:45584", + "issue:43716|issue:44560", + "issue:43950|issue:45834", + "issue:43366|issue:44589" + ], + "prepared_review_unit_hash": "f9a54e435c702397705f9c31d69855ff3b4cf86b6a3ad617d2f7f7a0b1e4dbc4", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36669, + "estimated_input_tokens": 9168, + "estimated_eval_tokens": 18592 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous issue cluster with two clear duplicate subpairs: docs references to removed v5 pipeline tasks, and the `_is_hf_initialized` constructor regression. The rest look like related-at-a-distance model-loading or training failures, but not the same underlying bug.", + "confidence": 0.64, + "canonical_issue_reason": "Issue 43872 is the strongest concrete bug report here: it captures the `_is_hf_initialized` loading regression clearly, and 44291 is essentially a narrower reproducer of the same constructor failure.", + "canonical_pr_reason": null, + "best_issue_reason": "43872 is the cleanest representative issue in the set: specific failure, clear symptom, and the best duplicate anchor among the items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "Different problems: tokenizer/key loading failure vs flash-attn fallback gating; same broad loading area but not the same bug." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both report the same docs regression: v5 docs still mention removed pipeline tasks / pipeline() references." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Different code paths: tied lm_head weight bug vs silent corruption of non-persistent buffers during from_pretrained()." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same underlying `_is_hf_initialized` constructor/loading regression, just with different reproductions and wording." + }, + { + "left": "issue:42915", + "right": "issue:43872", + "accept": false, + "reason": "Both mention loading failures, but one is a Qwen3-MoE + FP8 config issue and the other is a constructor kwarg regression; not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Related audio area, but a general model run failure is not the same as a specific fbank extraction issue." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "Different training behaviors: LoRA gradient checkpointing ineffectiveness vs missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unexpected trainer kwarg vs tensor-mask shape mismatch; unrelated failures." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Different DeepSpeed issues: weight conversion/loading vs gradient averaging under GAS." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer/model load failures on different models; too generic to be the same bug." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are ZeRO-3 weight-loading failures, but the affected model paths and symptoms differ enough that they do not clearly collapse to one bug." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures: unit-test load issue for GLM-4.7-Flash vs NemotronH checkpoint load failure." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Unrelated failures: RLHF gradient explosion from position_ids/SDPA behavior vs flash_attention crash on s_aux=None." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": false, + "reason": "Both are generic mask/tensor index errors, but there is no evidence they share the same root cause or code path." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Different regressions in the video stack: one is a broad v5 input regression across multiple models, the other is a StopIteration in Qwen3-VL embedding video handling." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "One is a bad error message for Voxtral tokenizer loading; the other is a tokenizer load failure for a different model." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Both are model-load related, but NemotronH checkpoint loading and Sam3 text-encoder weights not loading are distinct failures." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Completely unrelated: DeepSpeed training RuntimeError vs image processor void segmentation-map handling." + }, + { + "left": "issue:43596", + "right": "issue:46515", + "accept": false, + "reason": "Different failures: ZeRO-3/Bert index error vs torch version incompatibility." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "PagedAttention invalid group type and NemotronH checkpoint loading are different model/runtime issues." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF support request vs autocast dtype mismatch; unrelated." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Both touch rotary/position handling in Qwen models, but the symptoms and failure points differ and do not look like the same bug." + }, + { + "left": "issue:43638", + "right": "issue:46515", + "accept": false, + "reason": "Non-pretrained Bert ZeRO-3 index error is unrelated to the torch 2.4.1 incompatibility report." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "GGUF architecture support and Qwen3-MoE sliding_window behavior are unrelated." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states bug vs PLE device mismatch; different forward-path defects." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "AudioFlamingo3 track leakage vs Whisper empty transcription after align_special_tokens; not the same underlying issue." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different video/model issues: dtype mismatch in image preprocessing vs StopIteration in Qwen3-VL video embedding." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Silent buffer corruption during from_pretrained() is unrelated to OCR long-input index errors." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "GGUF gpt-oss support request and Float8 storage lookup failure are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded and conservative. The two accepted soft pairs are well-supported by the packet: 43827/44509 is the same v5 docs regression, and 43872/44291 is the same `_is_hf_initialized` loading/construction bug. The other rejections stay appropriately cautious and do not overclaim duplication from broad thematic similarity alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 6, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43317", + "issue:43366", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43720", + "issue:43824", + "issue:43957", + "issue:43994", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44560", + "issue:44661", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45030", + "issue:45276", + "issue:45313", + "issue:45357", + "issue:45507", + "issue:45685", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:46459", + "issue:46468", + "issue:46489", + "issue:46512", + "issue:46515" + ], + "soft_pairs": [ + "issue:44683|issue:46515", + "issue:43824|issue:43957", + "issue:44560|issue:45685", + "issue:42915|issue:43957", + "issue:43645|issue:43646", + "issue:43550|issue:44910", + "issue:44560|issue:44918", + "issue:42617|issue:43366", + "issue:45812|issue:46489", + "issue:45823|issue:46515", + "issue:36010|issue:42915", + "issue:44661|issue:45823", + "issue:44560|issue:44805", + "issue:38175|issue:43994", + "issue:43632|issue:45834", + "issue:44877|issue:45030", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:44661|issue:44805", + "issue:43317|issue:44756", + "issue:38175|issue:39692", + "issue:43644|issue:44534", + "issue:42491|issue:43720", + "issue:41720|issue:42915", + "issue:46459|issue:46512", + "issue:42915|issue:44756", + "issue:44451|issue:44488", + "issue:42491|issue:45276", + "issue:44960|issue:44977", + "issue:45313|issue:45357", + "issue:45507|issue:46468", + "issue:44488|issue:44991", + "issue:43644|issue:43645", + "issue:39692|issue:43994" + ], + "prepared_review_unit_hash": "616cbca3151856ec32c02a2c9fc6a12e4acb39b7f577c08ccb04e6b394fdc1e4", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37648, + "estimated_input_tokens": 9412, + "estimated_eval_tokens": 19080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous issue cluster with a few tight duplicate pairs. The strongest matches are the custom-model init reports (43645/43646), the non-persistent-buffer regression (43644/44534), the SigLIP2 bad-output reports (38175/43994), and the stricter-config-loading failures (44877/45030).", + "confidence": 0.83, + "canonical_issue_reason": "issue:43646 is the best canonical issue because it is the broadest, clearest statement of the custom-model initialization regression, and it cleanly subsumes the narrower Jupyter-specific report in 43645.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best issue anchor for this cluster because it captures a concrete, general bug rather than a niche manifestation, while still representing one of the few clearly duplicate threads present.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44683", + "right": "issue:46515", + "accept": false, + "reason": "Different failures: compiled flex_attention on torch 2.9 vs a torch 2.4.1 incompatibility report." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "A missing import for a specific class and a meta-device loading regression are different code paths." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Different model families and different failure modes; too little evidence of one underlying bug." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8 loading and generic device='meta' loading breakage are not the same issue." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same underlying custom-model initialization regression; 43645 is the Jupyter-specific manifestation of 43646." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and Qwen3.5 flash-attention packed-sequence crash are different bugs." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Qwen3.5 video StopIteration and TRL embedding unpacking are different symptoms and likely different code paths." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3D parallel execution failure is unrelated to GGUF gpt-oss architecture support." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but on different model families with different described failures." + }, + { + "left": "issue:45823", + "right": "issue:46515", + "accept": false, + "reason": "Gemma4 device mismatch and a torch 2.4.1 incompatibility are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "GenerationMixin import failure and Qwen3Moe FP8 loading are different problems." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Tokenizer-mapping edge case and Gemma4 PLE device mismatch are not the same underlying bug." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different symptoms: Qwen3 video StopIteration versus a tensor/mask shape mismatch index error." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Same SigLIP2 bad-output regression; both describe near-zero/nonsensical predictions for the same model family." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "HF v5 initialization flag breakage and Kosmos2.5 long-OCR indexing are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": true, + "reason": "Both report loading failures caused by stricter config validation in v5." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "DeepSpeed zero3 indexing failure and tokenizer-mapping logic are different initialization bugs." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Zero3 empty-tensor indexing and add-new-model-like tokenizer mapping are unrelated." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer mapping edge case does not match the separate shape-mismatch indexing error." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Dequantized-model offload loading and Strix Halo mmap OOM are different load-path problems." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Related SigLIP2 area, but docs-example errors and bad model outputs are not the same concrete bug." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same non-persistent-buffer junk regression in Transformers v5." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility and BitNet packed-weight loading are separate issues." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Auto device-mapping CUDA assert and FineGrainedFP8Config failure do not share the same underlying cause." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Closely related dtype-handling themes, but one is about ignored load dtype and the other about stale config.dtype after from_config." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "FP8 loading and Strix Halo mmap/OOM are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are model-load failures, but the repo-specific evidence is too thin to treat them as one bug." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "LoRA training compatibility and Gemma4 embedding-resize behavior are unrelated." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "GLM5 and Qwen3.5 flash-attention failures are different models and failure modes." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "ZeRO-3 weight loading and save_pretrained key serialization are different directions of failure." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "GraniteMoEHybrid invalid-method calls and the Nemotron example breakage are not the same bug." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/load regressions, but they target different repos and lack evidence of a shared root cause." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "These are related v5 regressions, but non-persistent buffer corruption and notebook custom-model init breakage are not the same concrete bug." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "SigLIP2 docs-example problems and wrong model outputs are related at a high level, but not the same defect." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the high-level summary overstates the evidence by calling the Qwen3.5 packed-sequence pair a true duplicate, but that pair was explicitly rejected in the soft verdicts. The rest is mostly conservative, and the 43645/43646 custom-model init linkage is grounded, but the summary is not fully supported by the packet." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 7, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:44075", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44488", + "issue:44756", + "issue:44863", + "issue:44910", + "issue:44912", + "issue:44960", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45588", + "issue:45663", + "issue:45685", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:45859", + "issue:46050", + "issue:46468" + ], + "soft_pairs": [ + "issue:43856|issue:43873", + "issue:43708|issue:44075", + "issue:43854|issue:43931", + "issue:43856|issue:44387", + "issue:43257|issue:43854", + "issue:42491|issue:45292", + "issue:44960|issue:46050", + "issue:45488|issue:45812", + "issue:42491|issue:43931", + "issue:45362|issue:45406", + "issue:45276|issue:45335", + "issue:44164|issue:44291", + "issue:44910|issue:45588", + "issue:43404|issue:43454", + "issue:43425|issue:43957", + "issue:41720|issue:44155", + "issue:44402|issue:44488", + "issue:39692|issue:44960", + "issue:42947|issue:45704", + "issue:43873|issue:44387", + "issue:36010|issue:39404", + "issue:45663|issue:45759", + "issue:44291|issue:44756", + "issue:45663|issue:45859", + "issue:44387|issue:44756", + "issue:44292|issue:44912", + "issue:44155|issue:45823", + "issue:43957|issue:44756", + "issue:42947|issue:43381", + "issue:43335|issue:45663", + "issue:45276|issue:45292", + "issue:44863|issue:46468", + "issue:45685|issue:46468", + "issue:44164|issue:44756" + ], + "prepared_review_unit_hash": "8f552334bfc14c7c490ab16da6d3760170a74f26ed847634510591eb2f039ca6", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37366, + "estimated_input_tokens": 9342, + "estimated_eval_tokens": 18940 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are unrelated bugs across model loading, quantization/offloading, tokenizer regressions, checkpointing, and model-specific architecture issues. I found no strong duplicate chain; all soft edges are too weak to merge strictly.", + "confidence": 0.94, + "canonical_issue_reason": "No canonical issue: the cluster does not form a single duplicate family; the titles span many unrelated subsystems and model families.", + "canonical_pr_reason": null, + "best_issue_reason": "No clear best representative issue for deduplication; none of the issues is a convincing target for the rest of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/quantization, but one is Qwen3 MoE training memory use and the other is quantized offloading behavior; different bugs." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer resume/max_steps calculation and SGD arguments being ignored are unrelated code paths." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Both are load failures, but for different models and different failure modes." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both involve quantization-related memory pressure, but the reported problems and contexts differ materially." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Qwen3 MoE weight conversion with accelerate/deepspeed is unrelated to GLM-4.7-Flash loading failures." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "LoRA model compatibility on hf5.x is unrelated to resize_token_embeddings not updating output embeddings." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "GLM5 model support and sinq save/load corruption are separate issues." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different underlying mechanisms." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "A LoRA compatibility regression and a Qwen3-VL shape-mismatch load error are not the same bug." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Qwen3.5 chat crashing and Gemma4Processor missing _tokenizer are unrelated model/serving bugs." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Both are resize_token_embeddings issues, but they affect different model families and different embedding components." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Extra_state save/load handling and init_empty_weights _is_hf_initialized are distinct model-loading internals." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both concern flash attention, but one is a Qwen3.5 position_ids crash and the other is an s_aux=None AttributeError." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both are weight-tying bugs, but they occur in different model classes and likely different code paths." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 incompatibility and meta-device loading regressions are unrelated failures." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA asserts and AudioFlamingo3 track leakage are unrelated." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Tokenizer vocab mismatch and inability to load a different model are not the same underlying bug." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "SigLIP2 documentation errors and GLM5 model support are unrelated." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to apex RMSNorm memory leak behavior." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both touch quantization/memory, but one is offloading behavior and the other is reserved-memory/OOM growth." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "An import-path regression for GenerationMixin and Whisper return_language pipeline behavior are different issues." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Gemma-4 FSDP2 shared_kv_states and Qwen multimodal config unwrapping are separate model-architecture problems." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "init_empty_weights argument handling and mmap OOM on Strix Halo are unrelated." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Gemma-4 FSDP2 KeyError and Qwen3_5MoE CPU offload incompatibility are different code paths." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention OOM, but the causes are different: int4 reserved-memory growth vs mmap behavior on Strix Halo." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are quantization-load regressions, but with different formats, models, and failure modes." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo3 batched inference leakage and Gemma4 device-mismatch during forward are unrelated." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are separate problems." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both mention gradient checkpointing, but one is ineffective under PEFT LoRA and the other is disallowed in eval mode." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug and Gemma-4 FSDP2 attention state bug are unrelated." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": false, + "reason": "Closely related surface area, but Gemma4 versus t5gemma suggests separate resize paths rather than one concrete duplicate bug." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint-loading failures and a broken official example are not clearly the same bug." + }, + { + "left": "issue:45685", + "right": "issue:46468", + "accept": false, + "reason": "An MPS histogram op issue and a Nemotron example failure are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "save/from_pretrained extra_state handling and 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": 1, + "component_count": 20, + "review_unit_index": 8, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:39404", + "issue:42796", + "issue:43381", + "issue:43425", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43716", + "issue:43828", + "issue:43854", + "issue:43901", + "issue:43906", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44877", + "issue:44898", + "issue:44912", + "issue:44918", + "issue:44928", + "issue:45084", + "issue:45216", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45701", + "issue:45812", + "issue:45823", + "issue:46050", + "issue:46315", + "issue:46326", + "issue:46491" + ], + "soft_pairs": [ + "issue:42796|issue:46326", + "issue:44479|issue:44877", + "issue:43854|issue:45216", + "issue:43381|issue:44928", + "issue:45684|issue:45698", + "issue:45588|issue:45663", + "issue:43646|issue:43708", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:43828|issue:43957", + "issue:44402|issue:44451", + "issue:43425|issue:43828", + "issue:46315|issue:46491", + "issue:45812|issue:46315", + "issue:44589|issue:44756", + "issue:44683|issue:45823", + "issue:39404|issue:45701", + "issue:43611|issue:43645", + "issue:43425|issue:45823", + "issue:43425|issue:43716", + "issue:44530|issue:45084", + "issue:43906|issue:44530", + "issue:44479|issue:44805", + "issue:44265|issue:44898", + "issue:44589|issue:45084", + "issue:39404|issue:44912", + "issue:43901|issue:44509", + "issue:43425|issue:44918", + "issue:43906|issue:46326", + "issue:46050|issue:46326", + "issue:43994|issue:44079" + ], + "prepared_review_unit_hash": "0898fc6b154ced86ec1140da651ea4f714278b23d9a715725ebc01ab2b5ce87b", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36519, + "estimated_input_tokens": 9130, + "estimated_eval_tokens": 18516 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a collection of unrelated issue threads that only share broad Transformers subsystems. I only see one strong duplicate-like pair: the `resize_token_embeddings` regression affecting output embeddings, with one issue phrased generally and the other as a T5Gemma-specific manifestation.", + "confidence": 0.77, + "canonical_issue_reason": "Issue 45292 is the broadest formulation of the `resize_token_embeddings` bug; 45335 looks like the same underlying propagation problem reported on a specific model path (`decoder.embed_tokens`).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the best representative because it describes the core failure mode in a general way and can subsume the model-specific variant.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Both are tokenizer-related, but one is about BPE marker handling and the other about `continue_final_message`; different code paths and symptoms." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Different models and failures: video input regression vs strict config loading for `granite_speech`." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different bug reports on different models: GLM-4.7-Flash unit-test loading vs Qwen3.5 save_pretrained regression." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Unrelated training-time problems: gradient checkpointing in eval mode vs Qwen3.5 RLHF NaN explosion from position_ids/SDPA behavior." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Both involve custom-model save/load, but one is permissions propagation and the other is wrong module resolution after save; not the same bug." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Different failures in different areas: flash-attention crash on `s_aux=None` vs Gemma-4 FSDP2 shared_kv_states KeyError." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization in notebooks is unrelated to `resume_from_checkpoint` max_steps calculation." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same underlying `resize_token_embeddings` propagation bug; the second issue is a model-specific instance of the same failure to update output embeddings." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Both mention custom models, but one is notebook initialization and the other is save/from_pretrained handling of `extra_state`." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Different dtype/device issues on different models and code paths: autocast mismatch vs meta-device loading." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both are model/tokenizer loading complaints, but the reported failures and models are different." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch version incompatibility is not the same as a Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Tokenizer regression in performance vs a round-trip spacing bug; different symptoms and likely different fixes." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Granite token IDs regression and slow `convert_tokens_to_ids` are separate tokenizer issues." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Storage-object lookup failure is unrelated to Strix Halo mmap OOM avoidance." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and Gemma4 device-mismatch under `device_map='auto'` are distinct bugs." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline `return_language` regression is unrelated to general tokenization changes across versions." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers 5 migration issues, but one is `base_model_prefix` loading and the other is notebook initialization of custom models." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "Torch 2.10 incompatibility and Gemma4 forward device mismatch are different concrete problems." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch version compatibility and Mistral-3 image preprocessor/model dtype mismatch are not the same bug." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and compile failure for non-template nodes are different failures." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "An isolated reproduction of a prior issue is not the same as the Qwen3.5 PagedAttentionCache crash." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression and a mask-shape IndexError are different failure modes." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "`torch.export` failing with `torch_compilable_check` is unrelated to Perceiver image classification failing at non-default resolution." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup and non-template-node compile failure are separate issues." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline language return regression is unrelated to MXFP4 quantization fallback on git-oss-20b." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs mismatches, but they concern different pipeline APIs and different removed/deprecated behaviors." + }, + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch compatibility and Qwen3.5 input-embedding unpacking in TRL SFT are different bugs." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "Isolated reproduction of another issue is not the same as the mistral_common `continue_final_message` tokenizer bug." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantized save/load corruption and tokenizer message-finalization behavior are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Siglip2 wrong outputs via AutoModel/pipeline and `ModelOutput` key assignment are different code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the soft-edge decisions are generally conservative. The only accepted pair, `issue:45292` / `issue:45335`, is well-supported by the titles as the same `resize_token_embeddings` propagation problem affecting output embeddings / `decoder.embed_tokens`. I don\u2019t see any soft-edge verdict that overstates duplicate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 9, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:32090", + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43299", + "issue:43404", + "issue:43421", + "issue:43452", + "issue:43541", + "issue:43582", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43646", + "issue:43716", + "issue:43856", + "issue:43931", + "issue:44164", + "issue:44451", + "issue:44462", + "issue:44479", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44756", + "issue:44811", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45584", + "issue:45759", + "issue:45859", + "issue:46178", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46489" + ], + "soft_pairs": [ + "issue:36032|issue:43452", + "issue:43299|issue:43931", + "issue:44671|issue:44991", + "issue:43596|issue:44683", + "issue:44164|issue:45759", + "issue:44991|issue:45081", + "issue:43856|issue:44756", + "issue:44164|issue:44479", + "issue:42796|issue:43404", + "issue:44462|issue:46489", + "issue:39404|issue:44671", + "issue:44488|issue:45081", + "issue:44661|issue:44683", + "issue:44488|issue:44987", + "issue:32090|issue:39404", + "issue:43611|issue:43644", + "issue:44737|issue:45859", + "issue:46285|issue:46402", + "issue:43716|issue:45084", + "issue:43582|issue:43632", + "issue:36032|issue:42915", + "issue:40990|issue:43421", + "issue:43643|issue:44530", + "issue:44451|issue:44987", + "issue:34567|issue:40444", + "issue:42175|issue:46178", + "issue:43541|issue:45084", + "issue:42915|issue:44530", + "issue:44811|issue:45584", + "issue:43644|issue:43646", + "issue:40444|issue:46326" + ], + "prepared_review_unit_hash": "0d9bb9f820131d2b6c733a503af612a85eb33142b2c6ed6b2402708a80af24af", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37029, + "estimated_input_tokens": 9258, + "estimated_eval_tokens": 18772 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Loose cluster of unrelated closed issues; the soft links are mostly superficial topical overlaps (tokenizers, loading, Qwen, Whisper, v5 regressions) rather than the same bug/change.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out: the items span many unrelated subsystems and failure modes, and none of the soft pairs look like true duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "No good global best issue for dedup triage here; the cluster is too heterogeneous to anchor on one representative bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Both involve tokenizer/model loading, but one is a T5Tokenizer method-name conflict and the other is gguf_file / AutoTokenizer-from-pretrained handling." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Different Qwen model families and different failures: FineGrainedFP8Config loading vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and EMBEDDIA tokenizer loading are unrelated bugs." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Deepspeed ZeRO-3/BertModel index error and compiled flex_attention on torch>=2.9 are separate code paths." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "Extra_state save/load handling is distinct from AutoModelForCausalLM.from_config failing to unwrap text_config." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both touch tokenizer loading, but the causes differ: EMBEDDIA tokenizer load failure vs Mistral regex patch hitting a backend_tokenizer attribute error." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE memory inefficiency and disabling mmap on Strix Halo to avoid OOM are not the same bug." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "save/from_pretrained extra_state support and Qwen video-input regression are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Tokenizer BPE-marker handling and Mistral3 lm_head weight tying are different issues." + }, + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "One is AutoTokenizer ignoring tokenizer.json, the other is a DeepSeek-Coder v1 tokenization regression attributed to a prior fix gap." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper return_language pipeline behavior and CamemBERT masked-LM predictions do not share the same underlying change." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Loading cjvt/sleng-bert and Mistral regex patching fail in different places and for different reasons." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Model-like registration in TOKENIZER_MAPPING_NAMES and compiled flex_attention on torch>=2.9 are unrelated." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both are load failures, but they affect different models and mechanisms: cjvt/sleng-bert vs physical-intelligence/fast." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType and Whisper return_language pipeline failure are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "base_model_prefix loading breakage and non-persistent buffers being filled with junk are separate v5 regressions." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "XLNet CPU device placement and Qwen3_5Moe CPU offload incompatibility are different model-specific device bugs." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "CLIPTextModel structure inconsistency and Sam3 text encoder weights not loading are distinct load-path problems." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Mistral-3 dtype mismatch and a template compilation error are unrelated failures." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "AppleSilicon caching_allocator_warmup TypeError and _is_hf_initialized flag breakage are not the same issue." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "T5Tokenizer add_special_tokens conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and TokenizersBackend not updating its post-processor are separate bugs." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "trust_remote_code AutoConfig missing fields and PagedAttentionCache linear_attention crashes are different subsystems." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Different model loads fail for different reasons; there is no shared underlying defect." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "TrainerState token counting and Qwen2.5-VL iterable multi-image finetuning failures are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Missing tensorflow in a torch extra and an unknown tensorflow_text backend mapping error are distinct packaging/backend issues." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Mixtral grouped_mm tracing failure and non-template-node compilation error are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Qwen3.5 paged-attention linear_attention crash are different model/path problems." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Whisper batch_decode ignoring skip_special_tokens and empty-transcription failure after align_special_tokens are related only by model family, not the same bug." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Non-persistent buffer junk filling and custom model initialization breakage are separate v5 issues." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "Qwen2.5-VL multi-image iterable training and mistral_common continue_final_message behavior are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 10, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28530", + "issue:34567", + "issue:36032", + "issue:41720", + "issue:42915", + "issue:43089", + "issue:43262", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43452", + "issue:43482", + "issue:43540", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43746", + "issue:43819", + "issue:43883", + "issue:43994", + "issue:44062", + "issue:44206", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44849", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45507", + "issue:45987", + "issue:46315", + "issue:46326" + ], + "soft_pairs": [ + "issue:44206|issue:44220", + "issue:43883|issue:44488", + "issue:36032|issue:43645", + "issue:43452|issue:43482", + "issue:43317|issue:43482", + "issue:45084|issue:45507", + "issue:44987|issue:44991", + "issue:44849|issue:45325", + "issue:44464|issue:45987", + "issue:43262|issue:43377", + "issue:43746|issue:46326", + "issue:43645|issue:44534", + "issue:43646|issue:44534", + "issue:43404|issue:44466", + "issue:44403|issue:45081", + "issue:43883|issue:44991", + "issue:42915|issue:45084", + "issue:43994|issue:44610", + "issue:43377|issue:43819", + "issue:28530|issue:34567", + "issue:43322|issue:43540", + "issue:43089|issue:45987", + "issue:44987|issue:45081", + "issue:44451|issue:45081", + "issue:44265|issue:44610", + "issue:42915|issue:44291", + "issue:44220|issue:44683", + "issue:36032|issue:43742", + "issue:44062|issue:44987", + "issue:45216|issue:45357", + "issue:44991|issue:46315", + "issue:42915|issue:44589", + "issue:44479|issue:45325", + "issue:36032|issue:41720" + ], + "prepared_review_unit_hash": "49b91f710add350202ffaf5279d57c2a47e5dde37732246d1963a97765dbdcb9", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 38040, + "estimated_input_tokens": 9510, + "estimated_eval_tokens": 19276 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous with one clear duplicate pair around Qwen3.5 save_pretrained serialization; the other soft matches are mostly loose topical similarity, not the same bug.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45216 is the best canonical issue because it states the broader Qwen3.5 save_pretrained regression, while issue:45357 is a narrower follow-up about the same incorrect checkpoint serialization path.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the strongest global anchor in this set: it is the most general and representative report for the only true duplicate pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Both touch audio feature extraction, but one is a LasrFeatureExtractor center-arg regression and the other is a different fbank helper issue." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading failures, but they involve different models and different failure modes." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "T5 tokenizer method-name conflict is unrelated to custom model initialization breaking in notebooks." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "GGUF file handling problems are involved in both, but the concrete failures and models differ." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Different code paths: device_map/offload dequantized model loading versus GGUF loading under v5." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Unrelated failures: template compilation error versus GraniteMoE hybrid invalid method call." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are load failures in v5, but they are distinct model/tokenizer regressions." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Qwen3.5 output_hidden_states bug is not the same as the Qwen2.5-VL rope-index scaling regression." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Chunked generation inconsistency and StaticCache seq_length type mismatch are separate cache/generation bugs." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and MIMI padding-mask behavior are different bugs." + }, + { + "left": "issue:43746", + "right": "issue:46326", + "accept": false, + "reason": "Local PEFT checkpoint loading and mistral_common continue_final_message are unrelated tokenizer/model-loading paths." + }, + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Custom model init in notebooks is unrelated to non-persistent buffers being serialized with junk." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "General custom model initialization breakage does not match the buffer serialization bug." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied lm_head weights, but one is about model tying and the other about save_pretrained serialization behavior." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Noise when loading a transformer is not the same as the Mistral regex crash." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Different model-loading problems with different root causes." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to the non-template-node compile error." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Nonsensical SigLIP outputs and OmDet-Turbo processor input-size mismatch are different bugs." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "MIMI batching/padding issues are unrelated to DAC.from_latents missing STE behavior." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Early-stopping metric selection and num_input_tokens_seen updating are unrelated Trainer issues." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Segfault when loading Llava Next and video ValueError in Qwen3OmniMoe are different model-specific failures." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation sync overhead is a performance issue; StaticCache.get_seq_length is a contract/type bug." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Model loading failure and Mistral regex patch crash are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "ScandiBERT loading failure is not the same as the Mistral tokenizer regex crash." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export failure with torch_compilable_check is unrelated to the OmDet-Turbo processor size regression." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "FP8 model loading and init_empty_weights argument handling are separate issues." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "fbank feature extraction and compiled flex_attention on torch>=2.9 are different subsystems and bugs." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Tokenizer special-token conflict is unrelated to the MobileLLM loading key error." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken keyword conflict and a model-loading regression are not the same bug." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same Qwen3.5 save_pretrained regression; 45357 is a narrower symptom about incorrect visual encoder keys." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer loading failure and convert_tokens_to_ids slowdown are different problems." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "FP8 loading failure and missing Float8 storage lookup are not the same code-path bug." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen video/rope regressions, but they describe different concrete failures and should not be merged as one bug." + }, + { + "left": "issue:44534", + "right": "issue:45081", + "accept": false, + "reason": "Non-persistent buffer junk and Mistral regex crash are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted soft pair is well supported by the Qwen3.5 `save_pretrained` titles, and the remaining soft matches are appropriately rejected as broader topical similarity rather than the same bug." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 11, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43296", + "issue:43317", + "issue:43377", + "issue:43381", + "issue:43421", + "issue:43425", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43746", + "issue:43792", + "issue:43818", + "issue:43856", + "issue:44079", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44756", + "issue:44811", + "issue:44829", + "issue:44877", + "issue:44991", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45479", + "issue:45491", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:46402", + "issue:46491" + ], + "soft_pairs": [ + "issue:43531|issue:44877", + "issue:43632|issue:44756", + "issue:43296|issue:43531", + "issue:43425|issue:44403", + "issue:43611|issue:44534", + "issue:40444|issue:45325", + "issue:42796|issue:43746", + "issue:45812|issue:46491", + "issue:43792|issue:44589", + "issue:44403|issue:44991", + "issue:33666|issue:40444", + "issue:43632|issue:44534", + "issue:43381|issue:45923", + "issue:40990|issue:44811", + "issue:44079|issue:45663", + "issue:44291|issue:44589", + "issue:43421|issue:44568", + "issue:43482|issue:45538", + "issue:35707|issue:40444", + "issue:43482|issue:43525", + "issue:43317|issue:43856", + "issue:42175|issue:45542", + "issue:41553|issue:43582", + "issue:43818|issue:46402", + "issue:44079|issue:44534", + "issue:44462|issue:45812", + "issue:43531|issue:44291", + "issue:43818|issue:45357", + "issue:44829|issue:45479", + "issue:43377|issue:45491", + "issue:41720|issue:45081" + ], + "prepared_review_unit_hash": "f2094fe4cf51ec3ea651727d46e47a2394b83f99d052f602047b0640c07364e8", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37145, + "estimated_input_tokens": 9287, + "estimated_eval_tokens": 18830 + }, + "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/model-loading/tokenizer issues. There is a loose theme around v5 regressions and multimodal/model-loading bugs, but none of the soft pairs look like true duplicates or mergeable PR-equivalents.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43611 is the broadest, most representative loading-regression report in the set: it targets a core Transformers v5 model-loading path and is thematically closest to several other failures here.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single issue anchor because it is general, user-facing, and central to the cluster's repeated v5 loading regression theme.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Different models and different failure modes; same broad area only." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "One is an HF init flag regression, the other is an AppleSilicon mmap/OOM workaround." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "Both are Qwen-related, but the bugs are unrelated training/loading failures." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch version incompatibility is not the same as noisy loading output." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but they affect different loading internals and are not the same bug." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen2.5-VL issues: one is dataset/multi-image finetuning, the other is rope-index scaling." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to PEFT checkpoint loading." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both tokenizer bugs, but different models and different round-trip behavior." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper runtime failure is unrelated to a Float8 storage lookup error." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Generic loading noise is not the same as a specific tokenizer load failure." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both are VL training-related, but they are different underlying failures." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Different v5 regressions affecting different code paths." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Eval-mode checkpointing restriction is not the same as a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity regression and batch_decode special-token handling are unrelated." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 attention state are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Unexpected init argument on load is unrelated to missing Float8 storage support." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Runtime post-processor updates and a specific add_special_tokens regression are different bugs." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Qwen2.5-GGUF loading and CLIPTokenizer max_length are unrelated." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multi-image finetuning." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Both are load failures, but different models and different root causes." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "Device-map offload failure and MoE training memory inefficiency are different problems." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both involve TensorFlow packaging, but one is missing backend inclusion and the other is an undefined backend error." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Tokenizer loading error message and AppleSilicon warmup TypeError are unrelated." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Different multimodal model loading bugs with different affected components." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key handling is unrelated to non-persistent buffer initialization." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both are AutoTokenizer regressions, but not the same concrete tokenizer bug." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding-window issue and init_empty_weights argument error are unrelated." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Both are vision-language model issues, but they touch different loading/save paths and different root causes." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Different training pathologies: flash-attention-3 degeneracy vs single-label zero-loss behavior." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both are batching/padding-related, but the models and failure mechanisms differ." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Auto device mapping CUDA assert and Mistral regex patch crash are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 12, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43262", + "issue:43296", + "issue:43454", + "issue:43482", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43701", + "issue:43746", + "issue:43749", + "issue:43792", + "issue:43819", + "issue:43873", + "issue:43881", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45325", + "issue:45381", + "issue:45584", + "issue:45701", + "issue:45800", + "issue:45823", + "issue:45834", + "issue:46209", + "issue:46305", + "issue:46315" + ], + "soft_pairs": [ + "issue:43644|issue:44079", + "issue:40444|issue:45381", + "issue:43611|issue:43646", + "issue:44265|issue:45072", + "issue:43749|issue:44079", + "issue:45325|issue:45381", + "issue:42175|issue:43296", + "issue:43994|issue:44898", + "issue:43792|issue:43957", + "issue:42796|issue:45381", + "issue:43454|issue:44466", + "issue:45325|issue:45834", + "issue:42915|issue:43482", + "issue:43632|issue:45072", + "issue:43701|issue:45325", + "issue:44291|issue:44877", + "issue:44479|issue:45823", + "issue:43296|issue:45800", + "issue:43873|issue:45701", + "issue:43606|issue:44265", + "issue:43262|issue:43746", + "issue:40990|issue:45584", + "issue:34634|issue:35707", + "issue:42796|issue:46209", + "issue:35707|issue:41720", + "issue:43881|issue:44488", + "issue:44186|issue:44265", + "issue:39692|issue:40444", + "issue:43454|issue:43550", + "issue:43606|issue:43819", + "issue:41720|issue:45823", + "issue:46305|issue:46315" + ], + "prepared_review_unit_hash": "3ca0f4e2861c7c5455f72ef1c394bc2549fb6361d568067035843c359e4fb2de", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37195, + "estimated_input_tokens": 9299, + "estimated_eval_tokens": 18854 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous cluster of unrelated bug reports. A few issues are superficially close within the Transformers v5 loading regressions and Qwen2.5-VL position-id bugs, but none are close enough to treat as the same underlying fix target. No PRs are present.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43611 is the broadest representative of the Transformers 5.0 loading-regression subgroup; several nearby reports mention model-loading breakage, but this one is the clearest umbrella issue.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single issue to anchor the cluster because it describes a central v5 model-loading failure rather than a model-specific edge case.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer serialization junk is a different bug from ModelOutput key assignment." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Qwen2.5-VL multi-image finetuning failure is unrelated to video position-id regression." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 loading regressions, but one is about base_model_prefix and the other custom initialization; not the same code-path bug." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to bfloat16 dtype mismatch." + }, + { + "left": "issue:43749", + "right": "issue:44079", + "accept": false, + "reason": "FSDP CPU RAM efficient loading is not the same as ModelOutput key handling." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Both touch Qwen2.5-VL position ids, but one is still-image temporal scaling and the other is video vision_position_ids; too different to merge." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "TensorFlow backend packaging is unrelated to PaddleOCR-VL model loading." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "SigLIP2 nonsensical output and Perceiver resolution failure are different model-specific bugs." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "Whisper runtime failure is not the same as meta-device loading breakage." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE marker handling is unrelated to Qwen2.5-VL video position ids." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Tied lm_head weights not being set is different from inconsistent serialization behavior." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Qwen2.5-VL rope-index scaling and Kosmos2.5 OCR index errors are different bugs." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to Qwen2.5-GGUF loading in v5." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "_is_hf_initialized flag breakage is unrelated to torch.export dtype mismatches." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint resume key mismatch has no relation to Qwen2.5-VL rope indexing." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "init_empty_weights argument mismatch and strict config loading are separate issues." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Qwen VL video-input regression and Gemma4 PLE device mismatch are different failures." + }, + { + "left": "issue:43296", + "right": "issue:45800", + "accept": false, + "reason": "PaddleOCR-VL loading failure is unrelated to torch/transformers version compatibility." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantized offloading malfunction and tokenization slowdown are different subsystems and bugs." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to torch.export on compilable checks." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate default and PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on gpt-oss-20b is not the same as Whisper empty-transcription generation failure." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are different features and failures." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation inputs_embeds/past_key_values is unrelated to Qwen3 auto-device-map cuda asserts." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "glm-4v-9b loading failure and sleng-bert loading failure are different model-specific issues." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "LayoutLMv2Tokenizer crashes are unrelated to torch.export/torch_compilable_check." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "SigLIP2 doc-example errors and Qwen2.5-VL iterable-dataset multi-image finetuning failure are different problems." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "lm_head weight tying bug is unrelated to Bamba torch.compile SDPA failure." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Bark CPU-offload device mismatch is not the same as DAC.from_latents / missing STE mismatch." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Qwen3 auto device mapping cudaErrorAssert and Gemma4 PLE forward mismatch are different device issues." + }, + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "Sliding-window off-by-one is unrelated to convert_tokens_to_ids performance regression." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 13, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:26413", + "issue:28530", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42796", + "issue:42915", + "issue:42947", + "issue:43065", + "issue:43257", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43450", + "issue:43454", + "issue:43526", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43708", + "issue:43720", + "issue:43749", + "issue:43873", + "issue:43881", + "issue:43883", + "issue:44186", + "issue:44387", + "issue:44466", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45357", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45920", + "issue:45923", + "issue:46050", + "issue:46402" + ], + "soft_pairs": [ + "issue:43377|issue:44610", + "issue:42796|issue:44977", + "issue:44568|issue:45920", + "issue:43454|issue:45072", + "issue:43257|issue:45923", + "issue:43317|issue:43873", + "issue:43377|issue:44898", + "issue:43322|issue:46050", + "issue:40990|issue:42947", + "issue:43454|issue:43720", + "issue:43572|issue:45759", + "issue:43749|issue:44534", + "issue:41720|issue:43322", + "issue:42617|issue:42915", + "issue:44466|issue:45357", + "issue:43526|issue:43883", + "issue:41720|issue:44977", + "issue:26413|issue:28530", + "issue:43881|issue:44991", + "issue:44387|issue:45704", + "issue:36331|issue:42617", + "issue:40990|issue:45701", + "issue:36331|issue:42915", + "issue:44877|issue:44991", + "issue:43450|issue:45701", + "issue:44186|issue:45072", + "issue:43065|issue:46402", + "issue:43873|issue:44756", + "issue:43550|issue:43720", + "issue:43577|issue:43708", + "issue:43404|issue:45357" + ], + "prepared_review_unit_hash": "0115a2b4f30680f330e1e4ec56267e40bb24cbd2c0759dcf40cecc5fe013e4b6", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36499, + "estimated_input_tokens": 9125, + "estimated_eval_tokens": 18506 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is loosely centered on Transformers v5-era regressions (tokenization, loading, offloading, trainer behavior), but the soft links are mostly subsystem-level similarity rather than true duplicates. I would not merge any of the listed pairs as the same underlying bug.", + "confidence": 0.58, + "canonical_issue_reason": "issue:45920 is the broadest and most representative tokenizer-regression report in the set; it spans multiple model families and matches the cluster's strongest recurring theme. However, the cluster overall is still quite heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best single representative issue because it is multi-model, v5-specific, and broad enough to anchor the tokenizer-regression subset better than the narrower tickets.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Different bugs: batched-vs-single encoder output mismatch vs processor/model resolution mismatch." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen/Tokenizer-adjacent only at a very broad level; one is a tokenizer BPE-marker issue, the other is flash-attention generation failure." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both involve tokenization regressions, but one is missing BOS/EOS insertion for a specific tokenizer while the other is wrong token IDs across several models." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Unrelated: tied lm_head weights vs bfloat16 dtype mismatch in different model wrappers." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Different concerns: MoE weight conversion/loading vs a missing gradient-checkpointing support flag." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both touch offloading/quantization, but the concrete failure modes and code paths differ too much to treat as duplicates." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Different model families and symptoms: missing padding-mask support vs non-default-resolution image classification failure." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Segfault while loading Llava Next is not the same as quantized-model save/load corruption." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "Perplexity anomaly and gradient-checkpointing ineffectiveness are unrelated bugs." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "Tied-weight serialization is unrelated to packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Config missing pad_token_idx is different from failing to unwrap text_config for composite multimodal configs." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM-efficient loading and non-persistent buffer initialization are separate regressions." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Different loading failures on different models; no shared concrete bug path." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel execution failure is not the same as Qwen3MoE FP8 loading failure." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both are serialization-related, but one is lm_head tying/serialization and the other is visual-encoder key saving in a specific multimodal model." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "Image processor label reduction bug is unrelated to missing all_tied_weights_keys on a model class." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Both mention Qwen, but one is device-mapping assert and the other is flash-attention generation behavior." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Trainer checkpoint-resume sample iterator failure is not the same as early-stopping metric resolution." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Model loading failure and tokenizer loading failure affect different code paths and likely different root causes." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Both involve memory, but one is int4 quantization reservation/OOM and the other is an apex RMSNorm leak." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Unexpected compute_loss keyword argument is unrelated to the 3d_parallel runtime failure." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "A model-specific perplexity regression is not the same as a general tokenization-change report." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "Trainer API signature breakage is unrelated to Qwen3MoE FP8 loading." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config loading and tokenizer loading are separate problems." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape bug is unrelated to tokenization changes." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash on NER/padding is unrelated to dtype mismatch in inference." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "A dummy Conv2d in the pixel decoder is not the same issue as missing text-encoder weight loading." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Offloading/quantization behavior and Strix Halo mmap OOM are different root causes." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile/SDPA failure is unrelated to accelerate weight-unpacking failure." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Model dtype propagation and resume_from_checkpoint max_steps calculation are unrelated bugs." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 are not the same as incorrect visual encoder keys in Qwen3.5 save_pretrained." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 14, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28530", + "issue:29159", + "issue:30064", + "issue:33357", + "issue:33666", + "issue:34634", + "issue:35707", + "issue:36010", + "issue:36032", + "issue:36331", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43316", + "issue:43317", + "issue:43322", + "issue:43388", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43782", + "issue:43873", + "issue:43881", + "issue:44008", + "issue:44186", + "issue:44361", + "issue:44387", + "issue:44423", + "issue:44610", + "issue:44671", + "issue:44743", + "issue:44898", + "issue:44913", + "issue:45059", + "issue:45161", + "issue:45406", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45704", + "issue:45823", + "issue:46291" + ], + "soft_pairs": [ + "issue:44008|issue:45823", + "issue:41720|issue:42617", + "issue:44423|issue:45406", + "issue:44186|issue:44610", + "issue:44610|issue:44898", + "issue:43550|issue:43606", + "issue:43295|issue:45701", + "issue:42915|issue:43317", + "issue:45632|issue:45684", + "issue:36010|issue:42617", + "issue:41720|issue:45161", + "issue:43531|issue:43572", + "issue:44671|issue:44743", + "issue:42915|issue:46291", + "issue:28530|issue:33666", + "issue:44186|issue:44361", + "issue:36331|issue:41720", + "issue:43526|issue:43881", + "issue:43540|issue:43782", + "issue:42915|issue:43322", + "issue:43316|issue:44913", + "issue:28530|issue:30064", + "issue:29159|issue:36032", + "issue:33357|issue:35707", + "issue:30064|issue:34634", + "issue:43388|issue:45059", + "issue:43550|issue:43653", + "issue:43606|issue:43720", + "issue:43873|issue:45704", + "issue:44387|issue:45701" + ], + "prepared_review_unit_hash": "a548408dcafc7e21bc57994bee5c776716d3ba3b711b05eb3a4ae098cdb5822d", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36676, + "estimated_input_tokens": 9169, + "estimated_eval_tokens": 18594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most pairs share only loose topical similarity (tokenizers, model loading, offloading, multimodal inference, configs) but do not look like the same underlying bug or change. I would not merge any soft pairs; the best triage anchor is the broad offloading/quantization issue.", + "confidence": 0.34, + "canonical_issue_reason": "issue:43873 is the broadest and most central umbrella in the cluster, covering the recurring offloading/quantization/device-map failures better than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Among the issues here, 43873 is the most reusable representative for duplicate triage because it is broad, active, and closest to the repeated loading/offload failure pattern.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Both involve Gemma models, but one is a Gemma3n forward variable-collision bug and the other is a Gemma4 device-mapping mismatch; different code paths and failures." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Qwen3 auto-device-map CUDA assert and a 3d_parallel.py execution failure are unrelated problems." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Continuous-batching multimodal crash with a string `.to()` error is not the same as Gemma4Processor missing `_tokenizer`." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes on NER/padding, while OmDet-Turbo has an image-size mismatch; different processors and root causes." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "OmDet-Turbo input-size mismatch and Perceiver interpolation failure are unrelated model/processor bugs." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and Bark-small CPU offload device mismatch are different concrete issues." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Both mention version-related tokenizer behavior, but one is a regression in processor/tokenizer access and the other is a broad tokenization-change report; not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Qwen3MoE FineGrainedFP8Config failure is distinct from dequantized-model loading with gpu+cpu offload." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Cache-path collisions for trust_remote_code and permission propagation in save_pretrained are unrelated filesystem behaviors." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "GenerationMixin import failure and a 3d_parallel.py runtime issue are not the same underlying bug." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both are device/distributed-related, but one is a Qwen3 auto-device-map CUDA assert and the other is GPT-OSS MoE TP behavior; different failures." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and StableLmConfig pad_token_idx regression are unrelated." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM prediction errors and Qwen3.5 recurrent-state reset are different model-specific bugs." + }, + { + "left": "issue:42915", + "right": "issue:46291", + "accept": false, + "reason": "A Qwen3MoE FP8 config issue and a kernels import/version requirement problem are unrelated." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early-stopping metric selection and Qwen2-VL multi-GPU training are not the same issue." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Both are tokenizer failures, but the affected tokenizers, triggers, and failure modes differ." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Trainer `compute_loss` signature change is unrelated to Qwen3 device-mapping CUDA assertions." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and glm-4v-9b load failure are different bugs." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and Qwen3VL weight_only load error affect different modalities and paths." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3MoE FP8 config failure and Llava Next segmentation fault are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and GPTNeoXConfig rotary_pct reload behavior are different config regressions." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Early stopping metric lookup and void segmentation-map processing are unrelated issues." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency and T5Tokenizer `add_special_tokens` attribute conflict are different tokenizer bugs." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS CLIP bus error and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Void segmentation-map processing and Bark voice_preset handling are not the same bug." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Metric gathering label truncation and SAM3 text/bounding-box odd behavior are different code paths." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "Bamba SDPA/compile failure and BigBirdTokenizer mask-token registration are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both touch loading/offload, but one is a CPU-offload device mismatch and the other is packed-weight unpacking during accelerate loading; not the same concrete fix." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantization/offloading problems and a T5 apex RMSNorm memory leak are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:45701", + "accept": false, + "reason": "Int4 quantization OOM in Transformers 5.x and tokenization changes across versions are different regression classes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 15, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:34567", + "issue:34634", + "issue:35707", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42886", + "issue:43329", + "issue:43381", + "issue:43421", + "issue:43531", + "issue:43606", + "issue:43653", + "issue:43746", + "issue:43782", + "issue:43844", + "issue:43927", + "issue:44062", + "issue:44077", + "issue:44368", + "issue:44451", + "issue:44492", + "issue:44509", + "issue:44514", + "issue:44545", + "issue:44589", + "issue:44610", + "issue:44625", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45357", + "issue:45381", + "issue:45406", + "issue:45464", + "issue:45479", + "issue:45588", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:45923", + "issue:46050", + "issue:46402" + ], + "soft_pairs": [ + "issue:41720|issue:43746", + "issue:45406|issue:45464", + "issue:45357|issue:46402", + "issue:44492|issue:44509", + "issue:43531|issue:45759", + "issue:43606|issue:43653", + "issue:42490|issue:43329", + "issue:34634|issue:41720", + "issue:34567|issue:35707", + "issue:44451|issue:44991", + "issue:43531|issue:44589", + "issue:45759|issue:45859", + "issue:44077|issue:44625", + "issue:44625|issue:45479", + "issue:44898|issue:45072", + "issue:44514|issue:44545", + "issue:45588|issue:45759", + "issue:44062|issue:44451", + "issue:44987|issue:45042", + "issue:44368|issue:45357", + "issue:39692|issue:44610", + "issue:44918|issue:44936", + "issue:43421|issue:43927", + "issue:43381|issue:43844", + "issue:43746|issue:46050", + "issue:43782|issue:45381", + "issue:42886|issue:44843", + "issue:45704|issue:45923", + "issue:43421|issue:43653", + "issue:45704|issue:45859" + ], + "prepared_review_unit_hash": "9a8ae283047095734ed7efc91bd37660a089df23fca4ea3f29312de5cd8f2528", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37039, + "estimated_input_tokens": 9260, + "estimated_eval_tokens": 18776 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These candidates are mostly unrelated issue reports. The only clear duplicate pair is the two Qwen2_5_VLProcessor batched padding=False crash reports; everything else appears to be different bugs or different code paths within the same subsystem.", + "confidence": 0.97, + "canonical_issue_reason": "issue:44545 is the clearest canonical issue: it describes the exact Qwen2_5_VLProcessor.apply_chat_template batched padding=False crash, and it matches 44514 verbatim with slightly more reference activity.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44545 is the best issue representative because it is the exact duplicate target for the only accepted soft pair and has the strongest engagement signal among the two identical reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different bugs: Qwen3 auto device-mapping CUDA assert vs GraniteSpeech local checkpoint/PEFT loading." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Different model and failure mode: Gemma4Processor missing _tokenizer vs Qwen3.5 streaming chat/completions failure." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Both are model-loading regressions, but they target different models and different missing-weight paths." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategy docs is unrelated to outdated pipeline-task documentation." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Different config bugs: sliding_window handling in Qwen3-MoE vs text_config unwrapping for composite Qwen multimodal configs." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Unrelated: CPU offload device mismatch vs tokenizer mask-token special registration." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Different code paths: local-path trailing-slash import warning vs video token counting using undefined names." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset bug is unrelated to Qwen3 auto device mapping on A800." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token counters not updating is a different issue from progressive generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are load failures, but they involve different models and likely different tokenizer/model compatibility problems." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Different failures: Qwen3-MoE sliding_window handling vs Float8 storage lookup error." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Different bugs: multimodal config unwrapping vs rotary embedding CPU-offload compatibility." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "PatchTSMixer post_init policy and Qwen3.5 num_labels propagation are unrelated." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Both touch sequence-classification config, but one is num_labels propagation and the other is a degenerate-loss bug for num_labels=1." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Different model families and symptoms: Perceiver positional interpolation vs bf16 dtype mismatches." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same exact bug: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding=False." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "Different code paths: flash_attention sink handling vs Qwen composite config loading." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken constructor argument conflict is unrelated to ScandiBERT loading failure." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Different regressions: loading physical-intelligence/fast vs PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "A warning about tie_word_embeddings is not the same as incorrect visual-encoder key saving." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Docs-example errors for SigLIP2 are unrelated to OmDet-Turbo's processor/model image-size mismatch." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Different trainer issues: SFT embedding unpacking vs evaluate() failing after train()." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Runtime special-token/post-processor updates are unrelated to DiaConfig save/load token-ID loss." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction is unrelated to gradient blow-up under ZeRO-3." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Different persistence bugs: PEFT local checkpoint loading vs quantized save_pretrained/from_pretrained breakage." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen VL bugs: weight_only load error vs wrong video vision_position_ids." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": false, + "reason": "Both concern offline mode, but one is cache loading generally and the other is an unconditional model_info call in a specific regex patch." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "Apex RMSNorm memory leak and missing gradient-checkpointing support are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Special-token post-processor refresh and BigBirdTokenizer mask-token registration are different tokenizer bugs." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "Different issues: T5 apex RMSNorm leak vs Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted soft pair is the clear duplicate between issue:44514 and issue:44545, and choosing issue:44545 as canonical/best is supported by the identical title plus slightly higher reference activity. The other pairwise rejections are phrased cautiously enough from the packet metadata alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 16, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43295", + "issue:43322", + "issue:43335", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43618", + "issue:43643", + "issue:43645", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43746", + "issue:43756", + "issue:43819", + "issue:43873", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44488", + "issue:44661", + "issue:44671", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45081", + "issue:45183", + "issue:45356", + "issue:45479", + "issue:45663", + "issue:45685", + "issue:45704", + "issue:45834", + "issue:45907", + "issue:45923", + "issue:46032", + "issue:46294", + "issue:46463" + ], + "soft_pairs": [ + "issue:43531|issue:43645", + "issue:41720|issue:45704", + "issue:46032|issue:46463", + "issue:45663|issue:45923", + "issue:43421|issue:44186", + "issue:43994|issue:44671", + "issue:43746|issue:43873", + "issue:39692|issue:43873", + "issue:45081|issue:45356", + "issue:43596|issue:44220", + "issue:45356|issue:45907", + "issue:30064|issue:36010", + "issue:43540|issue:46294", + "issue:43720|issue:43819", + "issue:43653|issue:43720", + "issue:44220|issue:44661", + "issue:43295|issue:45834", + "issue:43550|issue:45072", + "issue:43618|issue:43697", + "issue:44062|issue:44488", + "issue:43550|issue:43819", + "issue:43540|issue:44008", + "issue:44186|issue:44898", + "issue:43653|issue:43819", + "issue:43531|issue:43643", + "issue:43322|issue:45685", + "issue:45183|issue:45479", + "issue:43643|issue:44877", + "issue:43335|issue:43756", + "issue:44843|issue:45081", + "issue:30064|issue:36331" + ], + "prepared_review_unit_hash": "419301f45f1954ac0c6e3d8befcb27aaab487f9de2bc271d04b0753cd8c3af0d", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37522, + "estimated_input_tokens": 9381, + "estimated_eval_tokens": 19018 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: tokenizer/loading regressions, multimodal processor bugs, model-specific inference issues, and unrelated trainer/config problems. No soft pair looks like a true duplicate or mergeable change, so I would not collapse any of them.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44843 is the closest cluster anchor because it\u2019s a framework-level tokenizer-loading regression with broad impact; the rest split into separate model- or subsystem-specific bugs, so there is no true canonical duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "44843 is the most reusable representative for this mixed set, but only as a loose anchor for the tokenizer/loading subtheme rather than a real duplicate canonical.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:43645", + "accept": false, + "reason": "Different bugs: Qwen3-MoE sliding-window handling vs Jupyter/custom-model initialization regression." + }, + { + "left": "issue:41720", + "right": "issue:45704", + "accept": false, + "reason": "Unrelated failures: Qwen3 auto-device-map CUDA assert vs T5 RMSNorm memory leak." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Different code paths and symptoms: Mamba2 cache correctness vs T5 decoder-input validation error text." + }, + { + "left": "issue:45663", + "right": "issue:45923", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states bug is unrelated to a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Both tokenizer-related, but one is post-processor update behavior and the other is a LayoutLMv2 NER/padding crash." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different model-specific regressions: SigLIP2 inference nonsense vs CamemBERT MLM predictions." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "Local checkpoint/PEFT loading failure is not the same as an offloading-with-quantization issue." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "A docs/example mismatch and quantization failure is distinct from the runtime offloading bug in 43873." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Same helper area, but different root causes: backend_tokenizer AttributeError vs codec/ warning regression in Kimi tokenizer." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Deepspeed/BertModel zero3 init IndexError is unrelated to fbank feature extraction." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer regression vs repo-file revision forwarding in CI are unrelated." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image processing for void segmentation maps has nothing to do with GenerationMixin import failures." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both involve video, but one is a Qwen3OmniMoe processing ValueError and the other is a Qwen3.5 chat-template token/feature mismatch." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "BitNet packed-weight loading and DAC latent reconstruction are different model-specific bugs." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and BitNet accelerate loading are unrelated." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "A torchaudio feature-extraction issue is not the same as add-new-model-like failing in TOKENIZER_MAPPING_NAMES." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor.tokenizer regression in custom models is different from Kosmos2.5 OCR long-input indexing." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba-9B-v2 is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Missing CLIPOutput attentions is not the same bug as RTDetrV2 producing different outputs." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken constructor argument collision is unrelated to sleng-bert loading failure." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "These are separate bugs in different models and execution paths: SDPA compile vs DAC latent conversion." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Video-input failure in Qwen3OmniMoe is unrelated to Gemma 3n's audio_mel_mask variable collision." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer crash in LayoutLMv2 is unrelated to Perceiver image-classification resolution handling." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and DAC latent mismatch do not share the same underlying bug." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and AutoConfig trust_remote_code field loss are unrelated." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next load segfault and MPS histogram_mps type support are different failures." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Input-validation XOR error messaging and single-label-classification zero loss are different problems." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both are config-loading related, but one is remote-code field loss and the other is strict config blocking granite_speech load." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are different model-architecture bugs." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Same tokenizer patch area, but offline model_info calls and backend_tokenizer AttributeError are distinct regressions." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Void segmentation-map processing and CustomTrainer.compute_loss kwargs are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 17, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36010", + "issue:39692", + "issue:41720", + "issue:43388", + "issue:43421", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43644", + "issue:43716", + "issue:43749", + "issue:43782", + "issue:43818", + "issue:43825", + "issue:43950", + "issue:43994", + "issue:44008", + "issue:44186", + "issue:44442", + "issue:44448", + "issue:44479", + "issue:44485", + "issue:44509", + "issue:44561", + "issue:44671", + "issue:44805", + "issue:44811", + "issue:44821", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45084", + "issue:45381", + "issue:45584", + "issue:45834", + "issue:45987", + "issue:46032", + "issue:46192", + "issue:46291", + "issue:46294", + "issue:46463", + "issue:46512" + ], + "soft_pairs": [ + "issue:41720|issue:44991", + "issue:43716|issue:44008", + "issue:43388|issue:46032", + "issue:44821|issue:45084", + "issue:43950|issue:45987", + "issue:44448|issue:44671", + "issue:39692|issue:45834", + "issue:43582|issue:46512", + "issue:45584|issue:45834", + "issue:44186|issue:44442", + "issue:43644|issue:43749", + "issue:43421|issue:44811", + "issue:43540|issue:45381", + "issue:43782|issue:44977", + "issue:43577|issue:43818", + "issue:43606|issue:45072", + "issue:44479|issue:45834", + "issue:44485|issue:44509", + "issue:34634|issue:36010", + "issue:44805|issue:44936", + "issue:43825|issue:43994", + "issue:46294|issue:46463", + "issue:44991|issue:46291", + "issue:44561|issue:45020", + "issue:45987|issue:46192", + "issue:28218|issue:29159" + ], + "prepared_review_unit_hash": "22ad27aaed99290bec0c6b3db069f15066558280973e2d6e3225c3ee334c86c5", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35644, + "estimated_input_tokens": 8911, + "estimated_eval_tokens": 18078 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: tokenizer, vision, audio, cache, dtype, and trainer issues mixed together. I do not see a true duplicate set, though 44561 and 45020 are the closest thematically around remote_code regressions.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the only near-related theme (remote_code breakages in recent transformers versions), so it is the best single representative even though the cluster is not a clean duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 has the widest scope and most general framing among the items here; it can absorb narrower remote_code breakage reports better than the more specific 44561. The rest of the cluster is unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "Both mention model loading/runtime failures, but one is a Qwen device-map CUDA assert and the other is a tokenizer load failure for a specific RoBERTa model; different bugs and code paths." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Different modalities and failures: Mistral-3 image preprocessor dtype mismatch vs Gemma3n variable-name collision causing an AttributeError." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "One is a metrics/gathering bug dropping tuple labels; the other is a Mamba2 cache correctness bug for seq_len > 1. Not the same underlying issue." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor URL loading and a template-node compilation TypeError are unrelated failures." + }, + { + "left": "issue:43950", + "right": "issue:45987", + "accept": false, + "reason": "from_pretrained corrupting non-persistent buffers is a model-loading regression; StaticCache.get_seq_length returning a tensor is a cache API contract bug." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Both are v4/v5 behavioral regressions, but they affect different models and different inference paths; not one shared bug." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "A documentation example error for SigLIP2 is unrelated to a Kosmos2.5 OCR index error." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon warmup TypeError and stale config dtype after from_config are different dtype-related problems with different code paths." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription generation after align_special_tokens is unrelated to Kosmos2.5 long OCR index errors." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes on NER/padding are unrelated to AutoTokenizer failing to load a FastSpeech2Conformer tokenizer." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer corruption in from_pretrained is not the same as FSDP CPU RAM efficient loading being broken." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Both concern tokenizer/processor behavior, but one is about post-processor updates after special token changes and the other is Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both are video-input bugs in Qwen-family models, but they describe different model variants and different failure mechanisms." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading failure and Qwen3.5 flash-attention generation failure are distinct issues." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "Blip2 dtype propagation and Video-LLaVA temporal-attention/weight-sharing defects are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "Bark CPU offload device mismatch and bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are different failures." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "A Qwen video-input regression is unrelated to a Kosmos2.5 OCR index error." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE implementation discussion and docs mentioning removed pipeline tasks do not describe the same bug." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset failure and GenerationMixin import breakage are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask shape IndexError in tensor indexing is unrelated to trainer.evaluate failing after train." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message wording and SigLIP2 nonsensical outputs with AutoModel/pipeline are different issues." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen3.5 video chat-template mismatch and a T5 decoder-input validation error are unrelated." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading for a specific model and a kernels dependency/version import failure are unrelated." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "These are the closest pair thematically, but 45020 is a broad umbrella report while 44561 is a specific root-cause report; the shared theme is not enough to confirm one concrete duplicate bug." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "A cache length API return-type bug is unrelated to return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Tokenizer extra-space-after-added-token and slow-vs-fast tokenizer inconsistency are related to tokenization behavior, but not the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 18, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:34634", + "issue:36331", + "issue:42175", + "issue:43262", + "issue:43329", + "issue:43452", + "issue:43526", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43638", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:44060", + "issue:44291", + "issue:44368", + "issue:44554", + "issue:44589", + "issue:44610", + "issue:44671", + "issue:44857", + "issue:44912", + "issue:44936", + "issue:44991", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45372", + "issue:45406", + "issue:45538", + "issue:45600", + "issue:45701", + "issue:46209", + "issue:46285", + "issue:46326" + ], + "soft_pairs": [ + "issue:44610|issue:45072", + "issue:43720|issue:45072", + "issue:43577|issue:43582", + "issue:44671|issue:45701", + "issue:43825|issue:43827", + "issue:43782|issue:45325", + "issue:43749|issue:44291", + "issue:42175|issue:45081", + "issue:43582|issue:44857", + "issue:43262|issue:43819", + "issue:43906|issue:46209", + "issue:45003|issue:46285", + "issue:34634|issue:36331", + "issue:43653|issue:44912", + "issue:43526|issue:43697", + "issue:45081|issue:45406", + "issue:44060|issue:44368", + "issue:43742|issue:45020", + "issue:45372|issue:46326", + "issue:45084|issue:45538", + "issue:44589|issue:45042", + "issue:43526|issue:44554", + "issue:43329|issue:45600", + "issue:43452|issue:44991", + "issue:43638|issue:44936", + "issue:43550|issue:44912" + ], + "prepared_review_unit_hash": "cbb1a42efaaa0bdf8de71b73d9da3aee52abce27a9bdd8065bab636d8d9303cb", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35295, + "estimated_input_tokens": 8824, + "estimated_eval_tokens": 17904 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a grab-bag of largely unrelated Transformers bugs; none of the soft pairs look like true duplicates. The closest umbrella issue is the broad remote_code/model-loading regression report, but it does not absorb the rest of the set.", + "confidence": 0.22, + "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the set and the best anchor for related loading/regression triage.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the most generally reusable triage target because it describes the widest failure class; the other issues are mostly narrow, model-specific bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Different failures: processor image-size mismatch vs dtype mismatch in inference." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight loading bug is unrelated to the dtype mismatch report." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Blip2 dtype initialization and Apple Silicon allocator warmup are separate code paths." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "One is a CamemBERT masked-LM regression; the other is a broad tokenization-change report." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Related theme, but one is an error-message bug and the other is docs referencing removed APIs." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen VL bugs: weight-only loading vs rope-index scaling." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Both are loading-related, but they hit different mechanisms and should not be merged." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "TensorFlow backend packaging and Mistral tokenizer regex crashing are unrelated." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Allocator warmup TypeError is not the same as the CUDA AMP crash in LwDetrImageLoss." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate default and DAC latent reconstruction are different bugs." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "Isolated reproduction of an older issue is unrelated to FP8 inference on B200." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access and CLIP text-model structure inconsistency are different defects." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Bark voice preset and CustomTrainer loss signature are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BEiT label reduction bug and RTDetrV2 output drift are distinct model issues." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Both involve tokenizer/processor loading, but the crash causes and affected components differ." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "These are similar warning reports, but they describe different tied-weights problems on different Qwen variants." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "MobileLLM key error is too specific to be the same as the broad remote_code regression." + }, + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "Both mention mistral_common, but one is an ImportError during Gemma 4 loading and the other is a continue_final_message bug." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template-node compilation failure and CLIPTokenizer max-length behavior are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage import failure and PIL/torchvision dependency handling are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "BEiT label reduction and an MPS attention correctness bug do not match." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Undefined video-token helper variables and removed auto-mapping config references are separate issues." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both involve model/tokenizer loading, but the failing paths and root causes differ." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed zero3 index error and trainer.evaluate() after train are different trainer bugs." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "SDPA torch.compile failure and MXFP4 quantization fallback are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 19, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:29159", + "issue:34567", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43450", + "issue:43454", + "issue:43531", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43645", + "issue:43650", + "issue:43696", + "issue:43720", + "issue:43742", + "issue:43746", + "issue:43749", + "issue:43819", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:43931", + "issue:44291", + "issue:44387", + "issue:44479", + "issue:44610", + "issue:44617", + "issue:44756", + "issue:44811", + "issue:44912", + "issue:44977", + "issue:45042", + "issue:45459", + "issue:45759", + "issue:45803", + "issue:45859", + "issue:46291", + "issue:46421", + "issue:46424" + ], + "soft_pairs": [ + "issue:43746|issue:44811", + "issue:43819|issue:43927", + "issue:36010|issue:43645", + "issue:43257|issue:43329", + "issue:42175|issue:46291", + "issue:44387|issue:44617", + "issue:43650|issue:43906", + "issue:38175|issue:44610", + "issue:43720|issue:44610", + "issue:36010|issue:36331", + "issue:43531|issue:45042", + "issue:43854|issue:44977", + "issue:39692|issue:44912", + "issue:46421|issue:46424", + "issue:45459|issue:45803", + "issue:42490|issue:43257", + "issue:43262|issue:44811", + "issue:43257|issue:43696", + "issue:43643|issue:45759", + "issue:43329|issue:44479", + "issue:43329|issue:43450", + "issue:43749|issue:44756", + "issue:43257|issue:43931", + "issue:43454|issue:43577", + "issue:34567|issue:40990", + "issue:43606|issue:45859", + "issue:42915|issue:43742", + "issue:44291|issue:45042", + "issue:29159|issue:34567" + ], + "prepared_review_unit_hash": "bf1abea57acbdb3e9bec7d8945c75af7db029eddfe0617d0fd7443996552c481", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36532, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This looks like a loose embedding cluster of unrelated Transformers issues. The soft pairs mostly share only broad themes (tokenization, loading, multimodal/video, OOM, CPU offload) but not the same concrete bug or code path, so none should be merged as duplicates. No PRs are present in the cluster.", + "confidence": 0.95, + "canonical_issue_reason": "If a single anchor is needed, issue 43257 is the best representative because it has the strongest discussion/reference activity and is a concrete, actionable model-loading regression. That said, the cluster itself is not a true duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43257 is the most suitable global issue anchor: it is concrete, well-discussed, and closer to the dominant loading/regression theme than the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "Different subsystems and symptoms: local checkpoint/PEFT loading vs Whisper batch_decode token filtering." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "Unrelated bugs: DAC latent reconstruction mismatch vs DiaConfig losing custom token IDs after save/load." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "Both touch model import/init behavior, but one is a GenerationMixin import error and the other is a Jupyter custom-model initialization regression." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Different concrete failures: Qwen3 MoE weight conversion under accelerate/deepspeed vs undefined variables in multimodal video token counting." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Different dependency/import problems: pip extras missing tensorflow backend vs pipelines import failing due to kernels package requirements." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both mention OOM, but the underlying code paths are different: int4 quantization memory growth vs Sam3Video runtime OOM." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "No clear bug overlap; one is a placeholder/data issue, the other is an isolated reproduction of a separate issue." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Different model-specific bugs: unexpected zero probabilities in SigLIP2 vs processor/model input-size mismatch in OmDet-Turbo." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight unpacking during accelerate loading is unrelated to OmDet-Turbo processor shape mismatch." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "Different APIs and failures: GenerationMixin import resolution vs Trainer.compute_loss signature change." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Different areas entirely: Qwen3-MoE sliding_window handling vs PIL image processor torchvision dependency regression." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Both are model-loading/generation issues, but they affect different models and failure modes; not the same underlying bug." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Unrelated: documentation example errors for SigLIP2 vs MXFP4 quantization fallback on gpt-oss-20b." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both involve cache behavior, but one is lazy-init recompilation with chunked prefill while the other is max_cache_length being ignored except for quantized cache." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Both discuss bad exception handling, but they are different code paths and different failure surfaces." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "No overlap: trailing-slash local-path warning vs Qwen3 MoE loading under accelerate/deepspeed." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Different audio processor issues: chat-template sample-rate default vs batch_decode ignoring skip_special_tokens." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3 MoE weight conversion failure is unrelated to GPT-oss-20b CUDA OOM." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "Both are config-loading issues, but one is missing fields with trust_remote_code and the other is unwrapping text_config for composite Qwen multimodal configs." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Related area (video multimodal), but not the same bug: undefined video-token helper variables vs a broader video input regression across models." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Different video bugs: token counting internals vs batched video processor output shape." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "Both concern loading/memory, but one is FSDP CPU RAM-efficient loading breakage and the other is mmap-related OOM on Strix Halo." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Different Qwen-family problems and different failure modes: MoE accelerate/deepspeed conversion vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "Different model internals: AyaVision lm_head tying bug vs Blip2 dtype propagation issue." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "Trainer token counting bug is unrelated to GPT-oss-20b perplexity regression." + }, + { + "left": "issue:43606", + "right": "issue:45859", + "accept": false, + "reason": "Both mention CPU offload, but they affect different models and code paths; not a mergeable duplicate." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "Different loading failures: Qwen3Moe FineGrainedFP8Config vs MobileLLM key error." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Unrelated: init_empty_weights compatibility issue vs image-processor torchvision dependency regression." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer fast/slow inconsistency is unrelated to TrainerState num_input_tokens_seen not updating." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 20, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43317", + "issue:43404", + "issue:43425", + "issue:43452", + "issue:43525", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43824", + "issue:43927", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44568", + "issue:44821", + "issue:44849", + "issue:44898", + "issue:44912", + "issue:44987", + "issue:44991", + "issue:45127", + "issue:45137", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45538", + "issue:45701", + "issue:45907", + "issue:46291", + "issue:46512" + ], + "soft_pairs": [ + "issue:43525|issue:46291", + "issue:42915|issue:44821", + "issue:44568|issue:45356", + "issue:43404|issue:45127", + "issue:44987|issue:45538", + "issue:43927|issue:44912", + "issue:43697|issue:45357", + "issue:44849|issue:45701", + "issue:42915|issue:45538", + "issue:43425|issue:46291", + "issue:43577|issue:46512", + "issue:43452|issue:43582", + "issue:42175|issue:43824", + "issue:43317|issue:45538", + "issue:33357|issue:34567", + "issue:28282|issue:34689", + "issue:45137|issue:45405", + "issue:41720|issue:43540", + "issue:44062|issue:44991", + "issue:44479|issue:45381", + "issue:43540|issue:44479", + "issue:43742|issue:44291", + "issue:43540|issue:44560", + "issue:44568|issue:45488", + "issue:28282|issue:29127", + "issue:44987|issue:45907", + "issue:24643|issue:29159", + "issue:42915|issue:43749", + "issue:38175|issue:44898", + "issue:42175|issue:44488" + ], + "prepared_review_unit_hash": "655cce5335cd01dae046c354f612cb14a0c3ef8f74a14f571c8361f69967e93d", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36787, + "estimated_input_tokens": 9197, + "estimated_eval_tokens": 18650 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly noisy and spans many unrelated bugs (tokenizers, model loading, dtype propagation, quantization, training, and video processing). The only clearly overlapping pair is the Qwen VL video-input regression reports.", + "confidence": 0.89, + "canonical_issue_reason": "issue:44479 is the broadest and most representative bug report: it describes a shared video-input regression across multiple Qwen VL model variants, which matches the only likely duplicate in the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best standalone issue because it is the most general and actionable description of the common underlying regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Unrelated failures: Llama4 config attribute vs kernels version/revision requirement." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Different subsystems and symptoms: FP8 config load error vs image processor URL loading." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different mechanics." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both involve tied weights, but one is a Mistral3 model bug and the other is LoRA merge with extended vocabulary." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Model-loading regression vs CLIPTokenizer max_length bug; no shared underlying code path." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Config save/load token IDs issue vs quantization fallback behavior." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different bugs: RTDetrV2 output mismatch vs Qwen3.5 save_pretrained key names." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-states bug is not the same as a broad tokenization-version change report." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to CLIPTokenizer model_max_length." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch compatibility issue vs kernels package version/revision requirement." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Related dtype theme, but one is Blip2 loading and the other is generic from_config config staleness." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "Tokenizer/model loading issue vs Apple Silicon allocator TypeError." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Packaging/backend inclusion bug vs missing class import error." + }, + { + "left": "issue:43317", + "right": "issue:45538", + "accept": false, + "reason": "Dequantized device_map offload bug is unrelated to CLIPTokenizer max_length." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "Mac bus error in model loading vs TrainerState counter not updating." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Generic missing-PyTorch import error vs a specific Llama 3.2 Vision model loading regression." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT version bump." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both mention Qwen and runtime failures, but one is auto device mapping CUDA assert and the other is video input processing." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "AddedToken keyword collision vs tokenizer loading for a specific model." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Both describe the same Qwen VL video-input regression in v5.3.0; 45381 is a specific symptom of the broader regression in 44479." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Different Qwen model families and different failures; too broad to be the same bug." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Loading MobileLLM-125M key error is unrelated to init_empty_weights argument handling." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-related Qwen issues, but the affected models and failure modes differ." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Tokenizer special-token bug vs LlamaTokenizer pre-tokenizer override regression." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Missing PyTorch import error vs LayoutLMv3 error-message clarity issue." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Different API regressions: model loading vs list_repo_files not forwarding revision." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed weight shape error vs tokenizer fast/slow inconsistency." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "FP8 config failure is unrelated to FSDP CPU RAM efficient loading." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Siglip2 zero probabilities and Perceiver non-default resolution failure are different model bugs." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Backend packaging issue vs model loading failure for a specific checkpoint." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported overlap is issue:44479 with issue:45381, and the analyst correctly treats the rest as unrelated or only superficially similar. No soft-edge verdict appears overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 21, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "issue:28282", + "issue:34567", + "issue:36331", + "issue:42796", + "issue:43257", + "issue:43404", + "issue:43452", + "issue:43493", + "issue:43526", + "issue:43531", + "issue:43577", + "issue:43720", + "issue:43742", + "issue:43756", + "issue:43906", + "issue:43994", + "issue:44062", + "issue:44291", + "issue:44488", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44737", + "issue:44792", + "issue:44821", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44964", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45405", + "issue:45479", + "issue:45588", + "issue:45600", + "issue:45823", + "issue:45920", + "issue:45987", + "issue:46121", + "issue:46399", + "issue:46463", + "issue:46468", + "issue:46491" + ], + "soft_pairs": [ + "issue:44737|issue:46463", + "issue:43493|issue:43994", + "issue:44291|issue:44821", + "issue:43257|issue:43756", + "issue:34567|issue:36331", + "issue:43577|issue:45072", + "issue:43906|issue:46468", + "issue:27957|issue:28282", + "issue:44062|issue:45081", + "issue:44568|issue:46491", + "issue:44792|issue:44964", + "issue:43404|issue:45987", + "issue:43526|issue:45479", + "issue:44737|issue:45588", + "issue:45042|issue:45081", + "issue:42796|issue:43906", + "issue:43742|issue:44589", + "issue:43756|issue:46399", + "issue:43756|issue:45325", + "issue:45823|issue:46463", + "issue:43452|issue:44488", + "issue:43531|issue:44821", + "issue:45405|issue:45600", + "issue:45357|issue:45920", + "issue:44913|issue:46121", + "issue:44625|issue:44877", + "issue:43720|issue:44898" + ], + "prepared_review_unit_hash": "266d99a667e865cbda1096fcafd69142cacca2b37853d0ae8bcacf4997a925aa", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35635, + "estimated_input_tokens": 8909, + "estimated_eval_tokens": 18074 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated bug reports; the only soft duplicate candidate I\u2019d accept is the SigLIP2 pair, where both issues point to the same model implementation producing wrong outputs. All other pairs look like different code paths, models, or failure modes.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43493 is the best canonical issue because it describes the underlying SigLIP2 implementation discrepancy against the original JAX model, while the paired report is a downstream symptom of the same problem.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43493 is the strongest representative of the cluster\u2019s only true duplicate thread: it is the most root-cause-focused and specific report for the SigLIP2 bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "Different bugs: XLNet CPU device placement in rotary encoding vs T5 decoder input validation error message." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Both report SigLIP2 producing incorrect results; 43493 frames the HF-vs-JAX implementation mismatch, and 43994 is the user-facing symptom in AutoModel/pipeline." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Unrelated failures: init_empty_weights/_is_hf_initialized loading error vs AutoImageProcessor URL loading." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Different models and failures: Qwen3 MOE loading/conversion vs Smollm3 RoPE-layer handling." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "Both trainer-related, but distinct issues: TrainerState token counting vs custom compute_loss signature mismatch." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Separate dtype bugs in different model families; no shared concrete code path." + }, + { + "left": "issue:43906", + "right": "issue:46468", + "accept": false, + "reason": "Different reports for different models/examples; too little evidence of the same underlying bug." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "XLMRoberta flash-attention issue is unrelated to AutoModel PyTorch import detection." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Both tokenizer-related, but one is AddedToken construction and the other is Mistral regex patching; distinct root causes." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions: special-token insertion vs CodeLlama encode/decode space preservation." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal model failures; no indication they share the same code path." + }, + { + "left": "issue:43404", + "right": "issue:45987", + "accept": false, + "reason": "Mistral3 lm_head tying bug is unrelated to StaticCache sequence-length return type." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "Different classification/processor issues with different affected components and symptoms." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU tensor creation is unrelated to flash_attention sink-less model handling." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "Image processor torchvision dependency regression is unrelated to Mistral tokenizer regex patching." + }, + { + "left": "issue:42796", + "right": "issue:43906", + "accept": false, + "reason": "Ministral tokenizer BPE-marker handling is unrelated to an isolated reproduction of another issue." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Model loading key error and Float8 storage lookup failure are different loading paths and different failures." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "Smollm3 RoPE-layer discrepancy and Qwen3.5 kernel registration crash are separate model-specific bugs." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Both mention RoPE, but they affect different models and different logic (layer selection vs temporal position scaling)." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 PLE device mismatch is unrelated to T5 decoder input validation." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "gguf_file loading regression and a specific model load failure are different issues without a clear shared root cause." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "PEFT version bump and stale auto_mappings references are unrelated maintenance/CI problems." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Qwen3.5 save_pretrained key regression is unrelated to tokenizer producing wrong token IDs." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "GPTNeoX rotary_pct reload bug and rope-parameter JSON list handling are distinct config/validation issues." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "num_labels propagation bug and strict granite_speech config loading are different configuration problems." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to Perceiver interpolation failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The only accepted soft edge, issue:43493 vs issue:43994, is reasonable from the titles alone because both concern SigLIP2 and the second appears to be a symptom/report of wrong outputs from the same model family. The other rejections are also cautious and consistent with the packet. One minor note: phrases like \"only true duplicate thread\" and \"same underlying bug\" are a bit stronger than the evidence strictly proves, but not enough to overturn the overall assessment." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 22, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:32090", + "issue:33357", + "issue:38175", + "issue:41762", + "issue:42898", + "issue:43295", + "issue:43452", + "issue:43531", + "issue:43577", + "issue:43643", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43819", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44466", + "issue:44484", + "issue:44493", + "issue:44514", + "issue:44568", + "issue:44704", + "issue:44743", + "issue:44811", + "issue:44843", + "issue:44936", + "issue:44991", + "issue:45005", + "issue:45200", + "issue:45216", + "issue:45356", + "issue:45372", + "issue:45375", + "issue:45381", + "issue:45542", + "issue:45823", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46463", + "issue:46491" + ], + "soft_pairs": [ + "issue:43697|issue:45216", + "issue:43643|issue:45375", + "issue:45200|issue:46463", + "issue:24643|issue:29127", + "issue:41762|issue:44936", + "issue:44493|issue:46050", + "issue:44466|issue:45005", + "issue:28282|issue:32090", + "issue:42898|issue:46491", + "issue:43819|issue:44811", + "issue:43452|issue:44843", + "issue:44514|issue:45381", + "issue:28282|issue:33357", + "issue:44936|issue:45823", + "issue:43688|issue:43749", + "issue:44361|issue:44442", + "issue:38175|issue:43295", + "issue:45372|issue:46170", + "issue:44484|issue:44704", + "issue:43653|issue:44361", + "issue:45542|issue:46178", + "issue:43295|issue:45356", + "issue:43452|issue:43577", + "issue:43577|issue:43697", + "issue:43653|issue:44568", + "issue:44464|issue:44743", + "issue:43531|issue:43742", + "issue:43646|issue:43749", + "issue:44843|issue:44991" + ], + "prepared_review_unit_hash": "5a4745431ab199b1c6b0901814381df2eeab557d9f60010565522aa28eae5822", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36729, + "estimated_input_tokens": 9183, + "estimated_eval_tokens": 18622 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: many separate v5 regressions, tokenizer/processor edge cases, and training/loading bugs. The only plausible duplicate-style pair is the tied-weights serialization reports; the rest look distinct.", + "confidence": 0.72, + "canonical_issue_reason": "issue:43646 is the broadest umbrella regression report in the set and the best single triage anchor, but it should not be treated as proof that the whole cluster is one duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the most general, high-signal issue here; it best represents the cluster\u2019s broad v5 regression theme without being overly specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "Different model families and different failure modes: output drift in RTDetrV2 vs checkpoint save corruption in Qwen3.5." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Both involve config/strict handling, but one is missing fields from trust_remote_code and the other drops a specific vision config field; not the same bug." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Gemma 4 multimodal token defaults and T5 decoder-input validation are unrelated code paths and symptoms." + }, + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "DeepSpeed training crash and a LayoutLMv3 error-message clarity issue are not the same underlying problem." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "ZeRO-3 load-time index error for Gemma3 is unrelated to trainer.evaluate() failing after train()." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Unexpected position-id keys and sinq save/load corruption are different serialization/model-state bugs." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": true, + "reason": "Both describe the same tied-weights serialization regression in v5, with 44466 as a concrete device-dependent instance and 45005 as the broader report." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Importing AutoModel without PyTorch and a NoneType GPU broadcast error in Trainer are unrelated failures." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "clean_up_tokenization_spaces behavior change and CodeLlama leading-space round-trip loss are different tokenizer regressions." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latent decoding mismatch and Whisper batch_decode skipping special tokens are separate model-specific bugs." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both are tokenizer loading regressions, but one is gguf_file support and the other is an offline hub call in _patch_mistral_regex." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Processor batched-template crashing and video vision_position_ids being wrong are distinct Qwen2.5-VL issues." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "Missing PyTorch import support and a MacOS bus error on CLIP are not the same failure." + }, + { + "left": "issue:44936", + "right": "issue:45823", + "accept": false, + "reason": "Trainer lifecycle failure and Gemma4 device-map mismatch are different execution paths and symptoms." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Auxiliary-loss normalization and FSDP CPU RAM-efficient loading are unrelated bugs." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "MLukeTokenizer AttributeError and AutoTokenizer failing to load FastSpeech2ConformerTokenizer are separate tokenizer issues." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and a processor.tokenizer regression are different model/processor problems." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Gemma 4 processor import breakage from mistral_common and Windows ROCm torch.distributed issues are unrelated dependency failures." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "A save_pretrained shard-size question and missing kwargs passthrough in AutoProcessor.from_pretrained are not duplicates." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "BigBird special-token registration and MLukeTokenizer task errors are different tokenizer bugs." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both concern backend detection, but they target different missing backend cases and are not the same underlying defect." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor/tokenizer regression in v4.57.5 and Kimi-K2.5 codec/warning issues are distinct." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading breakage and Blip2 dtype remaining float32 are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "A dtype propagation bug in Blip2 and RTDetrV2 output drift are different problems." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Special-token registration for BigBird and missing BOS/EOS insertion for mdeberta are separate tokenizer behaviors." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and Qwen3_5 recurrent-state reset are different generation/cache bugs." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and MobileLLM-125M load-time KeyError are not the same issue." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model initialization breakage and FSDP CPU RAM-efficient loading are different failures." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Offline-mode model_info() call in tokenizer patching and est-roberta tokenizer loading regression are different root causes." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The summary correctly treats the cluster as mostly heterogeneous and only accepts the tied-weights serialization pair, which is the one clearly plausible duplicate-style match from the titles. The rejections are cautious rather than overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 23, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:29159", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:43262", + "issue:43295", + "issue:43452", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43653", + "issue:43688", + "issue:43720", + "issue:43742", + "issue:43756", + "issue:43927", + "issue:44361", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44514", + "issue:44554", + "issue:44561", + "issue:44610", + "issue:44625", + "issue:44821", + "issue:44849", + "issue:44977", + "issue:45290", + "issue:45440", + "issue:45491", + "issue:45701", + "issue:45704", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46209", + "issue:46365", + "issue:46402" + ], + "soft_pairs": [ + "issue:29159|issue:33357", + "issue:43688|issue:45704", + "issue:43756|issue:46402", + "issue:43688|issue:45440", + "issue:43653|issue:43927", + "issue:44514|issue:45290", + "issue:43550|issue:43927", + "issue:43688|issue:45701", + "issue:44479|issue:44625", + "issue:45491|issue:46002", + "issue:43550|issue:44361", + "issue:44464|issue:44977", + "issue:43643|issue:44561", + "issue:41720|issue:44464", + "issue:33666|issue:36010", + "issue:43742|issue:44821", + "issue:44466|issue:46050", + "issue:43504|issue:43550", + "issue:32090|issue:33666", + "issue:43504|issue:43720", + "issue:43577|issue:44554", + "issue:44977|issue:46209", + "issue:43504|issue:44610", + "issue:44554|issue:46365", + "issue:43262|issue:43295", + "issue:43720|issue:43927", + "issue:43606|issue:43927", + "issue:43452|issue:44451", + "issue:45823|issue:46050", + "issue:33666|issue:36331", + "issue:43525|issue:43572", + "issue:44849|issue:46209" + ], + "prepared_review_unit_hash": "7aba7b1092b65d9cc20a92a2363bc35d4d1dad48c5c0bae8689a8b1d16512707", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36763, + "estimated_input_tokens": 9191, + "estimated_eval_tokens": 18638 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are distinct model-, tokenizer-, config-, or runtime-specific regressions that only share broad keywords like loading, generation, or chat templating. None of the soft-similarity pairs look like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue stands out because the set spans many unrelated code paths (tokenizers, config loading, generation, vision/video processors, trainer/runtime errors). The apparent overlaps are only surface-level.", + "canonical_pr_reason": null, + "best_issue_reason": "If forced to pick a representative anchor, issue:43643 is the most umbrella-like because it is a framework-level config/loading regression and closest to the recurring trust_remote_code/load-path theme, but it still is not a true duplicate hub.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer inconsistency vs MacOS bus error on a CLIP model; different failure modes and code paths." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Both mention model math bugs, but one is auxiliary-loss normalization and the other is a T5 RMSNorm memory leak." + }, + { + "left": "issue:43756", + "right": "issue:46402", + "accept": false, + "reason": "Smollm3 RoPE-layer truncation and Sam3 text-encoder weights not loading are unrelated model bugs." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Aux-loss normalization in MoE models is not the same as DeepSeekV3MoE diverging from the remote implementation." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "Tokenizer special-token registration vs config losing custom token IDs after save/load are different bugs." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both are apply_chat_template crashes, but the edge cases differ: batched padding=False vs assistant tool-call messages with no content." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile+SDPA failure for Bamba is unrelated to DiaConfig token-ID persistence." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Aux-loss normalization and tokenization-version changes are different subsystems and failure classes." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Qwen3.5 video-input regression and num_labels propagation are separate model/config issues." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both involve NaNs, but one is Gemma3 sliding-window attention and the other is classification-head init/checkout behavior." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "torch.compile/SDPA failure is unrelated to MLukeTokenizer AttributeError during tasks." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Chunked generation with compiled forward and Qwen3.5 flash-attention generation bugs are different concrete code paths." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both affect trust_remote_code flows, but one is missing fields from AutoConfig and the other is removal of is_torch_fx_available." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Qwen3 auto device-map cuda assert and compiled-forward chunked generation are unrelated runtime issues." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU training for Qwen2-VL and GenerationMixin import failure are not the same bug." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "MobileLLM key error while loading and AutoImageProcessor-from-URL failure are different loading paths." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "lm_head.weight serialization inconsistency and sinq quantized save/load breakage are distinct persistence bugs." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Beit legacy-field loading failure and Bamba torch.compile/SDPA failure are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer _gpu_broadcast_one NoneType error is not the same as Qwen2-VL multi-GPU training behavior." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Both are loading-time problems, but one is a legacy config field and the other is packed-weight unpacking for BitNet." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "BLIP2 dtype propagation and an MPS attention correctness bug are different model/runtime issues." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 flash-attention generation and FP8 inference on B200 are not the same failure." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Beit pretrained loading via legacy field and OmDet-Turbo processor/model input-size mismatch are separate problems." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "MPS attention correctness and importing Trainer being blocked by datasets are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Audio apply_chat_template sample-rate defaulting and processor.tokenizer/image-handling regression are different processor bugs." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "BitNet packed-weight loading and DiaConfig token-ID persistence do not share the same code path." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "Bark CPU offload device mismatch and DiaConfig save/load token-ID loss are unrelated." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and inability to load a specific pretrained model are different loading regressions." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Gemma4 PLE device mismatch and sinq quantized save/load breakage are distinct issues." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Qwen2-VL multi-GPU training and CustomTrainer.compute_loss signature mismatch are unrelated training bugs." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config and missing pad_token_idx in StableLmConfig are similar in theme but model-specific and not the same bug." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 output_hidden_states bug and FP8 inference failure on B200 are separate model/runtime regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 24, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28218", + "issue:33357", + "issue:33666", + "issue:39692", + "issue:43257", + "issue:43278", + "issue:43295", + "issue:43299", + "issue:43334", + "issue:43493", + "issue:43504", + "issue:43540", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43688", + "issue:43701", + "issue:43720", + "issue:43824", + "issue:43825", + "issue:43844", + "issue:43927", + "issue:43994", + "issue:44291", + "issue:44361", + "issue:44464", + "issue:44466", + "issue:44488", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44792", + "issue:44843", + "issue:44849", + "issue:44964", + "issue:44977", + "issue:45200", + "issue:45229", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45584", + "issue:45759", + "issue:45812", + "issue:46192", + "issue:46424", + "issue:46489" + ], + "soft_pairs": [ + "issue:43994|issue:46192", + "issue:43720|issue:44361", + "issue:43278|issue:43688", + "issue:43493|issue:43825", + "issue:44964|issue:45405", + "issue:39692|issue:43493", + "issue:28218|issue:33357", + "issue:44625|issue:45584", + "issue:44849|issue:45357", + "issue:39692|issue:43295", + "issue:44568|issue:46489", + "issue:43493|issue:43844", + "issue:43278|issue:43844", + "issue:44488|issue:44843", + "issue:43540|issue:44977", + "issue:43646|issue:44466", + "issue:44568|issue:45812", + "issue:43824|issue:44589", + "issue:33357|issue:33666", + "issue:45200|issue:45229", + "issue:44589|issue:45520", + "issue:44464|issue:46424", + "issue:43257|issue:43299", + "issue:43334|issue:43572", + "issue:43927|issue:44361", + "issue:43334|issue:45759", + "issue:43504|issue:43606", + "issue:43278|issue:43701", + "issue:44792|issue:45405", + "issue:43575|issue:45405", + "issue:43572|issue:44291" + ], + "prepared_review_unit_hash": "6e0ff0e526042443d30f4bdde563727047863a0c0f0958d0abf3aed4bfa8174e", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37103, + "estimated_input_tokens": 9276, + "estimated_eval_tokens": 18808 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are thematically unrelated overall; the soft-similarity edges mostly connect different model families or different failure modes, so none should be merged as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:28218 is the oldest, broadest tokenizer-regression report in the set, but even it is not a true duplicate representative for the rest of the items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "Different bugs: SigLIP2 inference/pipeline mismatch vs return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "BitNet packed-weight loading failure is unrelated to MLukeTokenizer\u2019s task-time AttributeError." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "BF16\u2192FP32 eval dtype drift is a different issue from auxiliary-loss normalization in MoE models." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 HF-vs-JAX discrepancy is unrelated to the pipeline translation error-message regression." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model loading failure and an unsupported PEFT version bump are not the same underlying bug." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Docs/example errors for SigLIP2 are separate from the SigLIP2 implementation discrepancy report." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer added-space bug does not match a MacOS bus error in a CLIP model." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription behavior are different code paths." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Hidden-states output bug is unrelated to incorrect visual-encoder key saving." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "SigLIP2 docs example problems are not the same as the processor.tokenizer regression in custom model code." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "These are distinct tokenizer regressions for different model families and different failure modes." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 discrepancy and gradient blow-up under ZeRO-3 do not share a concrete root cause." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Eval dtype change is unrelated to abnormal gradient growth with HfDeepSpeedConfig." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Model load failure for cjvt/sleng-bert is unrelated to the offline model_info call in mistral regex patching." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3OmniMoe video processing error and Qwen3.5 flash-attention generation bug are different problems." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model initialization breakage is separate from device-dependent lm_head serialization inconsistency." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "MDeBERTa special-token handling and Granite tokenizer regression are different tokenizer paths." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing a missing Qwen2.5-VL class is unrelated to a Float8 storage-object lookup failure." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "Mac bus error in CLIP inference is not the same as Qwen2-VL multi-GPU training support." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "mm_token_type_ids defaulting for text-only fine-tuning is unrelated to Gemma4 multi-GPU OOM." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage resolution and flash_attn import KeyError are separate backend issues." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency is different from cache-length handling in quantized cache." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Accelerate/DeepSpeed MOE conversion failure is not the same as the Qwen3VL MOE loading regression." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-VL missing pad_token_id and StableLmConfig missing pad_token_idx are unrelated config bugs." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig save/load token-ID loss is unrelated to MLukeTokenizer\u2019s AttributeError on tasks." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "These both involve multimodal Qwen configs, but one is a missing pad_token_id bug and the other is failure to unwrap text_config." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy-field load failure for Beit is unrelated to the CPU-offload device mismatch in bark-small." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "Eval dtype drift and resume_from_checkpoint key mismatch are different state-management bugs." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A failing Janus test and a bumped minimum PEFT version are not the same change." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Qwen2 TP OOM and PEFT version requirements are unrelated issues." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to init_empty_weights receiving an unexpected argument." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 25, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:34567", + "issue:37428", + "issue:38175", + "issue:42490", + "issue:43065", + "issue:43232", + "issue:43493", + "issue:43504", + "issue:43540", + "issue:43606", + "issue:43653", + "issue:43696", + "issue:43701", + "issue:43844", + "issue:43873", + "issue:43957", + "issue:43994", + "issue:44112", + "issue:44186", + "issue:44423", + "issue:44464", + "issue:44493", + "issue:44554", + "issue:44734", + "issue:44849", + "issue:44964", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45229", + "issue:45406", + "issue:45446", + "issue:45520", + "issue:45584", + "issue:45588", + "issue:45823", + "issue:46050", + "issue:46121", + "issue:46192", + "issue:46285", + "issue:46369" + ], + "soft_pairs": [ + "issue:44493|issue:46285", + "issue:45229|issue:45823", + "issue:43606|issue:44112", + "issue:43504|issue:43653", + "issue:44464|issue:45584", + "issue:43994|issue:44493", + "issue:43873|issue:46050", + "issue:29942|issue:30333", + "issue:45588|issue:46050", + "issue:44423|issue:44734", + "issue:43844|issue:44964", + "issue:43065|issue:46369", + "issue:42490|issue:43696", + "issue:29159|issue:32090", + "issue:45406|issue:46121", + "issue:28218|issue:34567", + "issue:45084|issue:45520", + "issue:44554|issue:45446", + "issue:15354|issue:38175", + "issue:43540|issue:45200", + "issue:43493|issue:43701", + "issue:45072|issue:45198", + "issue:37428|issue:43957", + "issue:45200|issue:45823", + "issue:44186|issue:46192", + "issue:43232|issue:44849" + ], + "prepared_review_unit_hash": "7ba4b9969eb8b59d9196424737acb0182b8ecb75616f0e7b0215426fdb216c51", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35586, + "estimated_input_tokens": 8897, + "estimated_eval_tokens": 18050 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: tokenizer quirks, serving crashes, model-specific regressions, quantization/offload issues, and CI failures are all mixed together. None of the soft pairs look like the same underlying bug/change, so I rejected them all.", + "confidence": 0.91, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated subsystems and model families, with no strong duplicate center to anchor the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the most broadly scoped and still-open bug report in the set, but it is not a duplicate representative for the rest of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Both mention key/position-related behavior, but one is a broad model-key regression and the other is a CLIP class structure inconsistency; not the same bug." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Same Gemma4 family, but one is CUDA OOM in multi-GPU inference and the other is a PLE device mismatch in forward; different failure modes and code paths." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Device/offload test wording overlaps, but one is a bark-small CPU offload mismatch and the other is a GraniteSpeech CI test issue; unrelated concrete problems." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Different components and symptoms: BEiT legacy-field loading vs BigBirdTokenizer special-token registration/decode output." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "One is chunked generation with compiled forward, the other is Whisper empty-transcription handling after align_special_tokens; not the same code path." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "Both are model-output correctness complaints, but one is a SigLIP2 inference regression and the other is a generic unexpected key/position-id issue." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Both involve quantization/offloading, but one is offload behavior with quantization and the other is save/load corruption for sinq quantization; different bugs." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures and MLflow reporting on job failure are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention sink handling and sinq save/load breakage are separate subsystems with different symptoms." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both are serve crashes, but one is multimodal continuous-batching string handling and the other is KV-cache continuation indexing; distinct defects." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading failure are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Same model family only in a very broad sense; dummy Conv2d setup and memory overflow are not the same underlying issue." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Import warning on trailing slash paths is unrelated to GPT-oss CUDA OOM." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency and Trainer NoneType broadcast errors are different bugs." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and rope-validation list typing are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer spacing after added tokens is not the same as TrainerState.num_input_tokens_seen not updating." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template compilation errors and flash_attn import KeyError are separate failures." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness and a PyTorch version check for AuxRequest are unrelated." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure and SigLIP2 zero probabilities are completely different issues." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Qwen3OmniMoe video processing ValueError and Gemma4 mm_token_type_ids defaults are unrelated model/input bugs." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 HF-vs-JAX discrepancy and resume_from_checkpoint key mismatch are different problems." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "dtype mismatches in inference and Wav2Vec2 save/tokenization failures do not share the same code path." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "flash attention import regression and meta-device loading breakage are unrelated." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Both touch Gemma4, but one is missing mm_token_type_ids defaults and the other is a PLE device mismatch during forward." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "LayoutLMv2 tokenizer NER/padding crash and return_tensors being ignored are different tokenizer issues." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and Qwen3.5 output_hidden_states behavior are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 26, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28218", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43450", + "issue:43452", + "issue:43493", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43653", + "issue:43824", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43950", + "issue:43957", + "issue:43975", + "issue:44112", + "issue:44466", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44792", + "issue:44898", + "issue:44987", + "issue:45127", + "issue:45200", + "issue:45290", + "issue:45325", + "issue:45356", + "issue:45381", + "issue:45412", + "issue:45520", + "issue:45563", + "issue:45823", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46285", + "issue:46512" + ], + "soft_pairs": [ + "issue:44521|issue:45290", + "issue:43295|issue:43450", + "issue:45823|issue:46192", + "issue:43295|issue:44610", + "issue:45381|issue:46209", + "issue:43295|issue:43873", + "issue:44554|issue:45356", + "issue:43572|issue:44987", + "issue:39692|issue:46192", + "issue:45520|issue:46178", + "issue:43550|issue:43844", + "issue:43572|issue:43957", + "issue:37428|issue:43824", + "issue:44466|issue:45127", + "issue:44898|issue:45412", + "issue:43450|issue:43575", + "issue:43950|issue:46285", + "issue:42490|issue:43493", + "issue:44792|issue:45563", + "issue:43334|issue:43525", + "issue:42915|issue:43824", + "issue:38175|issue:43493", + "issue:43452|issue:46512", + "issue:43824|issue:43975", + "issue:43653|issue:44112", + "issue:28218|issue:35141", + "issue:45200|issue:45325", + "issue:43278|issue:43856" + ], + "prepared_review_unit_hash": "377143ceb956ffb4d214101ea46bdf9f32128e999d2193f28ae3663d4bc35b2f", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "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 a broad semantic false-positive: the issues span unrelated bugs in tokenizers, multimodal processors, model loading, quantization, config attributes, and backend imports. None of the proposed soft pairs looks like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43295 is the broadest regression report in the set and touches the processor/tokenizer API surface, but the cluster is too heterogeneous for it to be a true canonical duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a strong global canonical for this cluster; 43295 is only the closest broad representative, not an actual duplicate umbrella.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both involve chat templates, but one is about assistant masks for multimodal inputs and the other is a crash with tool-call messages; different code paths." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor/tokenizer regression vs batched video shape bug; different failure modes and subsystems." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Device-mapping mismatch in Gemma4 vs return_tensors being ignored when text_kwargs is passed; unrelated behavior." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Custom model processor/tokenizer access issue vs OmDet processor image-size mismatch; not the same bug." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Wrong Qwen2.5-VL video position ids vs FP8 inference failure on B200; completely different problems." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression vs quantization offloading issue; no concrete overlap." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness bug vs tokenizer codec regression; unrelated." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig vs a model-loading failure for physical-intelligence/fast; different root causes." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 docs example errors vs ignored return_tensors; one is documentation/example drift, the other API handling." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "flash_attn import key mapping bug vs unsupported tensorflow_text backend; different import-system issues." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "Bamba SDPA compile failure vs gradient blow-up under DeepSpeed ZeRO-3; not the same code-path." + }, + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "StableLm config attribute regression vs meta-device loading bug; only broadly about loading, not duplicate." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Different missing-import errors for different symbols/models; same symptom class, not same bug." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Serialization inconsistency of lm_head.weight vs LoRA merge collapse with extended vocab; related to tied weights but distinct failures." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver interpolation bug vs RT-DETR memory leak; unrelated." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Batched video shape bug vs tensor-parallel OOM loading a large model; different layers and symptoms." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Non-persistent buffer corruption in from_pretrained vs CLIPTextModel structure inconsistency; not the same issue." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Trailing-slash import warning vs SigLIP2 implementation discrepancy; unrelated." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "Janus image-generation test failure vs stale generate() warning; one is a functional failure, the other a warning bug." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention missing pad_token_id, but they are model-specific config bugs for different architectures and likely different fixes." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "FP8 config failure for Qwen3Moe vs Qwen2.5-VL import error; no overlap beyond being model issues." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": false, + "reason": "Both are SigLIP2-related, but one reports zero probabilities and the other a broader HF-vs-JAX discrepancy; not clearly the same concrete bug." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file loading breakage vs stale model.config.dtype after from_config(dtype=...); separate loading APIs." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "ImportError for a missing Qwen class vs DeepSeek tokenizer detokenization regression; unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Tokenizer special-token registration bug vs CI failure in GraniteSpeech; different objects and causes." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer spacing after added token vs embedding reinitialization after resize; different tokenizer/model behaviors." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaulting issue vs Qwen2.5-VL rope/position-id scaling bug; unrelated multimodal details." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "dtype mismatch between train/eval vs inefficient memory usage in Qwen3 MoE training; not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 27, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:29942", + "issue:32090", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:43262", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43450", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43819", + "issue:43827", + "issue:43901", + "issue:44112", + "issue:44206", + "issue:44442", + "issue:44448", + "issue:44514", + "issue:44521", + "issue:44561", + "issue:44610", + "issue:44743", + "issue:44857", + "issue:45083", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45245", + "issue:45290", + "issue:45412", + "issue:45520", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:46170", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46512" + ], + "soft_pairs": [ + "issue:43526|issue:46192", + "issue:43388|issue:43526", + "issue:44442|issue:45198", + "issue:41720|issue:46209", + "issue:43262|issue:43450", + "issue:45520|issue:46170", + "issue:38175|issue:45412", + "issue:43819|issue:44112", + "issue:44561|issue:46234", + "issue:44206|issue:45584", + "issue:45759|issue:46285", + "issue:43827|issue:43901", + "issue:44610|issue:44857", + "issue:45290|issue:46326", + "issue:43819|issue:45083", + "issue:44514|issue:44521", + "issue:44857|issue:45593", + "issue:43335|issue:43550", + "issue:43611|issue:46402", + "issue:43606|issue:45083", + "issue:37428|issue:43425", + "issue:45759|issue:46512", + "issue:29942|issue:32090", + "issue:44448|issue:44743", + "issue:43335|issue:43653", + "issue:43550|issue:44112", + "issue:45127|issue:45245", + "issue:45200|issue:45588" + ], + "prepared_review_unit_hash": "a757eb708f730b8aa8a178720c929631c788133231a760017d7ac3671c9f7f08", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36373, + "estimated_input_tokens": 9094, + "estimated_eval_tokens": 18444 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose lexical grouping of unrelated issue reports. I would not merge any soft edge as a duplicate; the only reasonable representative is the active chat-template multimodal issue (44521), but even that is not a true hub for the others.", + "confidence": 0.69, + "canonical_issue_reason": "issue:44521 has the most discussion/inbound references and is the most representative of a recurring API failure theme, but the overall set is too heterogeneous to justify a strong canonical duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the best single representative by activity and breadth of impact; it is the clearest candidate to anchor triage, though not a duplicate of the rest.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "Both touch label/text processing, but one is a BeitImageProcessorFast label truncation bug and the other is a processor kwargs handling bug; different code paths and symptoms." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both involve labels in vision workflows, but one is gather_for_metrics dropping tuple labels and the other is reduce_labels returning only one label; not the same underlying defect." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Both are tokenizer-related failures, but one is AutoTokenizer loading FastSpeech2ConformerTokenizer and the other is Wav2Vec2 save/tokenization behavior; different models and failure modes." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Both are hardware/runtime compatibility problems, but one is a Qwen3 auto device-map cudaErrorAssert on A800 and the other is an FP8 inference issue on B200; not the same bug." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate handling and batched video shape handling are unrelated bugs in different processors." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "One is a flash_attn import key error on Python 3.13; the other is a Windows/ROCm torch.distributed dependency issue. Different platform failures." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Unexpected zero probabilities in a vision model and RT-DETR memory not being released are unrelated issues." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC latent/forward mismatch and GraniteSpeech stale device override CI failure do not share the same code path or symptom." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Both reference import_utils helpers, but they concern different missing/removed availability checks and different downstream breakages." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "A feature-extractor center-arg regression and a Whisper empty-transcription generation failure are distinct bugs." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Both concern model/config structure changes, but one is from_config unwrapping composite Qwen multimodal configs and the other is CLIP text model attribute inconsistency." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are documentation-related, but they cover different docs and different deprecated API references; not the same change." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "One is a processor producing the wrong image size; the other is an AMP/CUDA crash in LwDetrImageLoss." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both are chat-template/tokenizer issues, but one crashes on assistant tool-call messages and the other concerns continue_final_message for mistral_common tokenizers." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "DAC latent mismatch and a qwen3_omni_moe feature-length helper behaving unexpectedly are different models and different logic." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both involve apply_chat_template in multimodal settings, but one is a batching/padding crash and the other is all-zero assistant masks; related area, not the same bug." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "An AMP loss crash and missing auxiliary losses when denoising is off are separate detection/training issues." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and Bamba torch.compile SDPA failure are unrelated model/runtime bugs." + }, + { + "left": "issue:43611", + "right": "issue:46402", + "accept": false, + "reason": "Both are loading failures, but one is a broad base_model_prefix regression and the other is Sam3Model text encoder weights not loading; different root causes." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch in bark-small and qwen3_omni_moe feature-length behavior are unrelated." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Missing _flash_supports_window_size import and Torch 2.10 incompatibility are not the same concrete failure." + }, + { + "left": "issue:45759", + "right": "issue:46512", + "accept": false, + "reason": "from_config unwrapping composite multimodal configs and stale model.config.dtype after from_config are separate configuration issues." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 test failures and Trainer _gpu_broadcast_one NoneType errors are unrelated training/runtime problems." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Pegasus v4/v5 output differences and Qwen recurrent state reset with cache are distinct model-behavior bugs." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and BigBirdTokenizer special-token registration are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba and a GraniteSpeech CI device override test failure do not share the same underlying defect." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge/collapsed model behavior and the 2^24 category limit runtime error are unrelated." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 token-type defaults and flash_attention.py s_aux=None handling are different multimodal/integration bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 28, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:26413", + "issue:28218", + "issue:29127", + "issue:31501", + "issue:32090", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:40444", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43504", + "issue:43526", + "issue:43606", + "issue:43701", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:43854", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44360", + "issue:44448", + "issue:44493", + "issue:44561", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45229", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:45920", + "issue:46192", + "issue:46234" + ], + "soft_pairs": [ + "issue:44855|issue:45083", + "issue:37428|issue:44351", + "issue:43335|issue:43504", + "issue:43278|issue:43825", + "issue:43819|issue:44493", + "issue:39692|issue:43526", + "issue:29127|issue:32090", + "issue:44206|issue:45823", + "issue:43335|issue:45083", + "issue:45588|issue:45859", + "issue:26413|issue:31501", + "issue:43720|issue:44112", + "issue:43335|issue:44112", + "issue:43335|issue:45593", + "issue:45310|issue:45478", + "issue:39692|issue:44493", + "issue:45459|issue:46234", + "issue:45290|issue:45381", + "issue:34689|issue:35141", + "issue:44448|issue:44779", + "issue:43606|issue:44857", + "issue:42490|issue:43854", + "issue:44360|issue:44561", + "issue:44857|issue:45072", + "issue:38175|issue:44493", + "issue:28218|issue:32090", + "issue:45920|issue:46192", + "issue:43701|issue:45083", + "issue:40444|issue:43526", + "issue:44448|issue:44977", + "issue:44977|issue:45229" + ], + "prepared_review_unit_hash": "a52c5cf8d24243b5326ba3505407391a67c8f0d6c2732eb742ec467ba0ebdd3b", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37152, + "estimated_input_tokens": 9288, + "estimated_eval_tokens": 18832 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous and does not form a single duplicate set. The only clear duplicate pair is the Qwen3.5 MoE `from_pretrained` regression; the rest are related only at a broad subsystem level or are plainly unrelated.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45310 is the clearest canonical anchor because it directly names the shared Qwen3.5 MoE `from_pretrained` regression and is the earlier, slightly richer report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the best issue to serve as the cluster representative because it is the only unambiguous duplicate target in the set and describes a concrete, reusable bug report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Different problems: Python 3.13 import/parsing failure vs a helper length-calculation behavior change." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they are for different missing symbols and unrelated code paths." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "Config construction bug for SwitchTransformers vs BEiT pretrained-loading legacy-field issue." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Unrelated bugs: dtype drift during evaluation vs a pipeline error-message regression." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "Different model-path bugs: DAC latent reconstruction mismatch vs position-id key warnings across many models." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "Docs/example cleanup and quantization issues are not the same as a BEiT fast processor label bug." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "Error-message clarity for LayoutLMv3 boxes is unrelated to Trainer broadcasting `NoneType`." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Feature-extractor argument validation and Gemma4 device placement are different failures." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction is unrelated to qwen3_omni_moe output-length behavior." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Different Qwen/MoE issues: flash-attention auxiliary-state handling vs CPU-offload rotary embedding compatibility." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Checkpoint resume epoch-iterator failure is not the same as a training startup problem." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "Accelerate packed-weight loading and a GraniteSpeech CI test failure are different code paths." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "Different subsystems and different failure modes; only broad CI/device language overlaps." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Sparse-layer config bug does not match the D-FINE auxiliary-loss omission." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same underlying bug: identical Qwen3.5 MoE `from_pretrained` error on transformers>=5.4.0." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "Documentation/example issues are unrelated to the position-id key regression." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Tokenizer protobuf error masking and a missing import-utils symbol are unrelated." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template tool-call crash is a different bug from Qwen2.5-VL video position-id handling." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Model-loading breakage for Llama vision and embedding reinitialization after resize are different issues." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both affect outputs, but one is Pegasus generation drift and the other is DeepSeek tokenizer regression; not the same code path." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to float16 AMP loss crashes on CUDA." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "Local-path import warning and GLM-4.7-Flash loading failure are different model-loading problems." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA indexer discussion is unrelated to the removal of `is_torch_fx_available` breaking trust_remote_code models." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Different bugs in different models: LwDETR AMP crash vs dtype mismatch in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 is not the same as the position-id key regression." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-token spacing bug is unrelated to Trainer broadcast failures." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Broad AutoTokenizer ID regression across models is not the same as silently ignoring `return_tensors` when `text_kwargs` is passed." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Resume-from-checkpoint key mismatch and qwen3_omni_moe length helper behavior are unrelated." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "IterableDataset multi-image finetuning failure is different from BEiT `reduce_labels` returning only one label." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Pegasus v4/v5 output differences are not the same underlying bug as Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU OOM are different failure classes." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The evaluation is grounded in the packet: the only accepted soft pair is the Qwen3.5 MoE `from_pretrained` regression between issue:45310 and issue:45478, and the rest are conservatively rejected based on clearly different titles/code paths. The canonical/representative issue choice is supported by the earlier date and slightly fuller report for issue:45310." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 29, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:25251", + "issue:27957", + "issue:29942", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:41720", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43381", + "issue:43412", + "issue:43550", + "issue:43575", + "issue:43746", + "issue:43873", + "issue:43927", + "issue:43976", + "issue:44448", + "issue:44466", + "issue:44561", + "issue:44568", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:44987", + "issue:45083", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45440", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45698", + "issue:45803", + "issue:45823", + "issue:46209", + "issue:46399" + ], + "soft_pairs": [ + "issue:32090|issue:33357", + "issue:29942|issue:35141", + "issue:43550|issue:45083", + "issue:43746|issue:46209", + "issue:44855|issue:46399", + "issue:42490|issue:43575", + "issue:43927|issue:45698", + "issue:10000|issue:15354", + "issue:27957|issue:34689", + "issue:37428|issue:44987", + "issue:44561|issue:44779", + "issue:43278|issue:43329", + "issue:44857|issue:45588", + "issue:43873|issue:45229", + "issue:43278|issue:43381", + "issue:41720|issue:45229", + "issue:44977|issue:45563", + "issue:44568|issue:45491", + "issue:43976|issue:45440", + "issue:44448|issue:44466", + "issue:45588|issue:45823", + "issue:45372|issue:45588", + "issue:45198|issue:45803", + "issue:25251|issue:32090", + "issue:43412|issue:45588", + "issue:32090|issue:33290", + "issue:45372|issue:45406", + "issue:45245|issue:45563" + ], + "prepared_review_unit_hash": "ff3c8db3f2186ff49d70e67a7252d45b1b4e395b54944542fa5572d4fbf14264", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35735, + "estimated_input_tokens": 8934, + "estimated_eval_tokens": 18124 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated bug reports spanning different models, features, and failure modes. I found no soft-edge pair that looks like the same underlying bug or change, so all candidate links should be rejected.", + "confidence": 0.87, + "canonical_issue_reason": "issue:43873 is the broadest, highest-signal report in the set (quantization/offloading behavior) with the strongest discussion and inbound reference activity, so it is the best anchor for the cluster despite not matching the others as duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the most reusable representative issue here: it has the broadest scope, substantial follow-up interest, and a clear user-facing regression to anchor triage.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Different failures: Trainer GPU broadcast NoneType vs MacOS bus error in a community CLIP model." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to the token-embedding reinitialization bug in post_init." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "Different code paths: torch.compile/SDPA on Bamba vs feature-extract length helper in qwen3_omni_moe." + }, + { + "left": "issue:43746", + "right": "issue:46209", + "accept": false, + "reason": "PEFT local checkpoint loading and FP8 inference on B200 are different issues with different root causes." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Both mention decorators/JIT, but they describe different models and distinct parsing/registration failures." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Local-path trailing-slash warning is unrelated to the tensor-parallel OOM on Qwen2-57B-A14B-Instruct." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both involve save/load, but one loses custom token IDs while the other loads the wrong custom module." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "Different models and different problems: DistilBert raises an issue vs GeneratorExp unsupported by torch.jit.script." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Flash Attention 2 on XLMRoberta is unrelated to the Llama 3.2 90B Vision model-loading regression." + }, + { + "left": "issue:37428", + "right": "issue:44987", + "accept": false, + "reason": "ImportError for a flash-attention helper is a different failure from loading physical-intelligence/fast on v5.1.0." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Removal of a utility function breaking trust_remote_code is not the same bug as DeepSeek tokenizer output drift." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift in eval is unrelated to the undefined video-token helper names in multimodal token counting." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetr float16 AMP crash and flash_attention s_aux=None crash are different subsystems and different bugs." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Both are memory/offload themed, but one is quantization offloading behavior and the other is a Gemma4 multi-GPU OOM." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "A dtype regression and an eval-mode gradient-checkpointing restriction are not the same underlying issue." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Both involve distributed inference failures, but the models, errors, and code paths are distinct." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation corruption is unrelated to a stale warning in paged generate()." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer special-token handling and Gemma3 NaN embeddings are different bugs with different fixes." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python-version compatibility breakage is unrelated to the DeepSeekV3MoE parity divergence." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Both are v5 regressions, but one is output drift and the other is lm_head serialization/tied-weights handling." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Gemma-related, but one is a flash_attention null-aux crash and the other is a device-map mismatch in forward." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 processor import dependency failure is not the same as the flash_attention s_aux=None crash." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare-except error handling." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting behavior and Trainer GPU broadcast errors are unrelated." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export/training-guard failure is a different code path from the flash_attention crash." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast NoneType and deepspeed Adafactor OOM are different problems." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma 4 processor import failure and Gemma4Processor missing _tokenizer in serve are separate breakages." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-count runtime limit and stale paged-generate warning are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 30, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:29942", + "issue:30064", + "issue:33357", + "issue:34567", + "issue:35141", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43066", + "issue:43278", + "issue:43381", + "issue:43421", + "issue:43526", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43824", + "issue:43873", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44279", + "issue:44361", + "issue:44442", + "issue:44462", + "issue:44479", + "issue:44743", + "issue:44871", + "issue:44977", + "issue:45072", + "issue:45198", + "issue:45372", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45593", + "issue:45712", + "issue:45834", + "issue:46121", + "issue:46369" + ], + "soft_pairs": [ + "issue:43653|issue:44743", + "issue:42490|issue:43873", + "issue:33357|issue:39692", + "issue:43065|issue:45412", + "issue:29127|issue:34567", + "issue:44361|issue:45198", + "issue:44206|issue:44479", + "issue:44186|issue:45198", + "issue:43065|issue:43994", + "issue:43611|issue:46121", + "issue:44871|issue:44977", + "issue:42915|issue:42994", + "issue:43278|issue:43575", + "issue:45459|issue:45712", + "issue:45412|issue:46369", + "issue:29942|issue:30064", + "issue:43381|issue:43701", + "issue:43994|issue:45412", + "issue:45372|issue:45459", + "issue:43066|issue:44462", + "issue:44265|issue:45198", + "issue:43756|issue:45440", + "issue:43653|issue:44442", + "issue:43421|issue:45491", + "issue:27957|issue:29942", + "issue:29127|issue:33357", + "issue:45440|issue:45593", + "issue:43994|issue:46369", + "issue:43526|issue:45834", + "issue:29127|issue:35141", + "issue:43824|issue:44279", + "issue:43526|issue:44186", + "issue:43575|issue:43701", + "issue:44442|issue:45072", + "issue:42490|issue:43381" + ], + "prepared_review_unit_hash": "cd089768a08ee7335e57c35308530eed076886ae2c25194da8a7f99018610067", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 38119, + "estimated_input_tokens": 9530, + "estimated_eval_tokens": 19316 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose collection of unrelated Transformers bug reports spanning tokenizers, model loading, quantization, multimodal processors, and training/runtime regressions. None of the suggested soft edges look like true duplicates of the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Different problems: BigBird tokenizer special-token registration vs Qwen3.5 recurrent-state reset in cached generation." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Local path trailing-slash import warning is unrelated to quantization offloading behavior." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error in a community CLIP model is not the same as SigLIP2 doc example errors." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and RT-DETR memory leak are different model-specific bugs." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error messaging and TrainerState token counting are unrelated subsystems." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLukeTokenizer task failure is unrelated to Wav2Vec2 save/tokenization failures." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Unsupported center arg in LasrFeatureExtractor is not the same as the video-input regression." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Different components and failures: LayoutLMv2 tokenizer padding/truncation vs Wav2Vec2 save_pretrained/tokenization." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Sam3PixelDecoder dummy layer is unrelated to SigLIP2 wrong outputs through AutoModel/pipeline." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "base_model_prefix loading regression and RoPE-parameter JSON list handling are different code paths." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation issues." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both involve quantization, but one is FineGrainedFP8Config failure and the other is saving quantized models; not the same bug." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Embedding dtype drift in eval and TP-related OOM on Qwen2-57B are different failures." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Tokenizer error masking with missing protobuf is unrelated to leftover dummy classes and repo checks." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "RT-DETR GC/memory leak and sam3-litetext memory overflow are different model-specific memory issues." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures are not the same as void segmentation map processing failures." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Eval-mode gradient checkpointing limitation and resume_from_checkpoint key mismatch are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 wrong outputs and RT-DETR memory leak are different bugs in different models." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma 4 processor import failure from mistral_common is unrelated to protobuf-missing tokenizer error masking." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring repository tokenizer.json." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export failures with torch_compilable_check are unrelated to Wav2Vec2 tokenization/save issues." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Smollm3 RoPE-layer discrepancy and DeepseekV3MoE divergence are separate architecture implementation bugs." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBird mask-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime special-token post-processing and Gemma3 mixed-length NaN embeddings are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:29942", + "accept": false, + "reason": "XLMRoberta FA2 support and FA2 test failures are not the same concrete bug." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error clarity and CLIP MacOS bus error are unrelated." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3MoE implementation drift is unrelated to D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43994", + "right": "issue:46369", + "accept": false, + "reason": "SigLIP2 nonsensical outputs and sam3-litetext memory overflow are different issues." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to Kosmos2.5 OCR long-input index error." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "LayoutLMv3 error messaging and embedding reinitialization after resize are different bugs." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Specific Qwen2.5-VL import failure is too different from a generic dependency issue." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "Different tokenizer/image-processor problems with no shared code path." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "TP OOM on Qwen2-57B is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "AutoTokenizer loading failure and dtype mismatch in inference are unrelated." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Trailing-slash local model import warning is unrelated to eval-mode gradient checkpointing." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 31, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:33290", + "issue:34567", + "issue:34689", + "issue:35141", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43316", + "issue:43334", + "issue:43381", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43575", + "issue:43611", + "issue:43646", + "issue:43720", + "issue:43723", + "issue:43761", + "issue:43873", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44079", + "issue:44206", + "issue:44387", + "issue:44442", + "issue:44464", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44898", + "issue:44977", + "issue:45003", + "issue:45245", + "issue:45440", + "issue:45479", + "issue:45491", + "issue:45588", + "issue:45600", + "issue:45859", + "issue:45907", + "issue:46121", + "issue:46234", + "issue:46399" + ], + "soft_pairs": [ + "issue:43575|issue:44387", + "issue:42915|issue:43931", + "issue:43761|issue:45859", + "issue:46121|issue:46399", + "issue:43334|issue:43531", + "issue:43526|issue:44857", + "issue:44977|issue:45245", + "issue:43262|issue:44206", + "issue:44464|issue:45491", + "issue:27957|issue:29127", + "issue:27957|issue:33290", + "issue:45588|issue:45600", + "issue:43646|issue:44077", + "issue:33290|issue:34689", + "issue:44743|issue:44811", + "issue:33290|issue:35141", + "issue:33290|issue:34567", + "issue:43381|issue:43873", + "issue:44442|issue:44898", + "issue:44062|issue:45907", + "issue:45003|issue:45440", + "issue:42490|issue:44743", + "issue:43723|issue:45479", + "issue:43316|issue:43525", + "issue:42175|issue:46234", + "issue:34689|issue:43611", + "issue:43720|issue:44442", + "issue:43761|issue:44079" + ], + "prepared_review_unit_hash": "514fecb6b45c35431e57dfeb69e531d01671a3ab7033e546a3692322f6dd95cd", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36113, + "estimated_input_tokens": 9029, + "estimated_eval_tokens": 18314 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a heterogeneous similarity cluster: the soft pairs share only broad keywords, not the same underlying bug or fix path. No pair looks safe to merge as a duplicate; issue 43873 is the broadest open anchor if one must pick a representative issue.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43873 is the broadest, still-open report with inbound references, so it is the best cluster anchor; the others are narrower and cover unrelated failure modes.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the best global issue candidate because it is open, broadly framed, and already attracting references; it is the most plausible triage hub even though the set is not a true duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both mention OOM, but one is tensor-parallel model loading and the other is int4 quantization memory growth; different code paths." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Different models and failures: FineGrainedFP8Config crash vs weight-shape mismatch during Qwen3-VL loading." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "A CLIPVision hidden-states regression and a Qwen3 MoE CPU-offload incompatibility are unrelated bugs." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "Both are rope-related, but one is JSON/list validation in config conversion and the other is a missing decorator for rotary-pos-emb registration." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Pad-token-id API mismatch and sliding-window recurrent-state behavior are different model issues." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "BEiT label reduction and LwDetr float16 AMP crash affect different components and code paths." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation failure and category-cardinality limit are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and unsupported center-arg crash are different processor bugs." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward inconsistency and Gemma3 sliding-window NaNs are distinct symptoms with different causes." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "XLM-Roberta flash-attention support and LayoutLMv3 error messaging are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "Flash-attention model support and Adafactor/deepspeed OOM are not the same bug." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "A flash_attention sink-less-model crash and a removed auto-mapping reference in CI are different issues." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Both touch initialization semantics, but one is a broad v5 custom-model init regression and the other is a specific patchtsmixer post_init policy change." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "OOM with Adafactor/deepspeed and Llama 3.2 Vision load failure are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Recurrent-state reset in cache handling and Whisper batch_decode skip_special_tokens behavior are separate paths." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Training OOM and embedding reinitialization in post_init are different bugs." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "OOM under deepspeed is unrelated to TrainerState token counter not updating." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and quantization/offloading behavior are different subsystems." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver non-default-resolution failure are not the same issue." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "AddedToken special-argument duplication and list_repo_files revision forwarding are unrelated API bugs." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "Unsafe sys.modules access and DeepseekV3MoE divergence from remote implementation are different code-path problems." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Trailing-slash import warning and recurrent-state reset do not describe the same bug." + }, + { + "left": "issue:43723", + "right": "issue:45479", + "accept": false, + "reason": "AutoTokenizer loading regression and degenerate zero-loss classification are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Llama4Config missing pad_token_id are similar in theme only, not the same concrete bug." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Backend extras installation and a missing import_utils helper are unrelated packaging/API issues." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Llama 3.2 Vision loading breakage and base_model_prefix loading regression are different model-loading failures." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "Accelerate packed-weights handling and tokenizer loading are separate bugs." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVision hidden_states regression and ModelOutput key assignment bug affect different code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 32, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:25251", + "issue:27957", + "issue:28282", + "issue:32090", + "issue:42617", + "issue:42796", + "issue:42907", + "issue:43232", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43493", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43643", + "issue:43701", + "issue:43742", + "issue:43749", + "issue:43761", + "issue:43825", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44077", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44855", + "issue:44871", + "issue:44913", + "issue:44987", + "issue:45137", + "issue:45216", + "issue:45290", + "issue:45698", + "issue:45759", + "issue:45803", + "issue:46178", + "issue:46234", + "issue:46291", + "issue:46421", + "issue:46506" + ], + "soft_pairs": [ + "issue:43232|issue:43388", + "issue:44077|issue:44479", + "issue:43761|issue:44188", + "issue:43493|issue:43827", + "issue:43931|issue:45216", + "issue:44987|issue:46178", + "issue:42907|issue:45698", + "issue:44442|issue:45803", + "issue:43873|issue:44871", + "issue:42796|issue:45290", + "issue:43749|issue:43931", + "issue:43316|issue:45759", + "issue:43701|issue:46506", + "issue:44464|issue:46421", + "issue:44190|issue:44496", + "issue:43643|issue:44913", + "issue:44855|issue:45137", + "issue:46234|issue:46291", + "issue:25251|issue:28282", + "issue:43295|issue:43825", + "issue:43550|issue:43701", + "issue:42617|issue:43575", + "issue:43606|issue:43701", + "issue:43575|issue:43742", + "issue:27957|issue:32090", + "issue:16998|issue:22355" + ], + "prepared_review_unit_hash": "cc98b05da001c39d42f9d2aea6d103984c59f1ca775134a8776c867877bd1a93", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35526, + "estimated_input_tokens": 8882, + "estimated_eval_tokens": 18020 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: the items cover unrelated bugs across generation, loading, tokenization, config, compilation, and docs. I do not see any true duplicate pair among the soft edges, so all candidate merges should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: this is not a duplicate cluster, so there is no single issue that cleanly represents the whole set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43761 is the strongest standalone report here: it is a specific, reproducible regression with a clear symptom (`hidden_states=None` despite `output_hidden_states=True`). It is still not a duplicate hub for the rest of the items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Different areas: generation kwargs update vs metric gathering/last-batch label truncation." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "Unrelated bugs: patchtsmixer post_init validation vs Qwen video-input regression." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "Different code paths: CLIPVision hidden states vs attention-kernel branching under torch.compile." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "Model implementation parity vs docs still referencing removed pipelines; not the same bug." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Both are Qwen VL-related, but one is shape mismatch loading and the other is save_pretrained checkpoint corruption." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Model loading failure for a specific model vs backend-mapping error for tensorflow_text; unrelated causes." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Dequantized Ministral save failure vs wrong custom module loaded after save_pretrained; different defects." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer loading failure vs Fuyu tensor conversion swallowing exceptions; not the same underlying issue." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization/offloading behavior vs Gemma eos_token_id config mismatch; no shared bug." + }, + { + "left": "issue:42796", + "right": "issue:45290", + "accept": false, + "reason": "Tokenizer BPE marker handling vs chat-template tool-call crash; separate features and failures." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU-RAM-efficient loading vs Qwen3-VL shape mismatch; different loading problems." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "API discrepancy in Gemma3TextConfig vs missing text_config unwrap in multimodal AutoModelForCausalLM.from_config." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "Checkpoint-key mismatch vs CPU/CUDA output divergence; unrelated symptoms." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Both involve chunking/compile, but one is inconsistent generation outputs and the other is recompilation from cache init; not same bug." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading in a training example vs unrecognized model config loading; unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "trust_remote_code config-field loss vs rotary_pct reset on reload; different serialization/config issues." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 JIT parse/IndentationError vs DeepSpeed ZeRO3 deque underflow; unrelated runtime failures." + }, + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Missing import_utils helper vs kernels version/revision import failure in pipelines; different packaging problems." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline output nesting bug vs missing PyTorch dependency ImportError; not the same issue." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "Custom model/processor regression vs incorrect v5 translation error message; different problems." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile + SDPA failure vs resume_from_checkpoint key mismatch; unrelated code paths." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "3d_parallel script failure vs tensor-parallel OOM loading Qwen2-57B; different resource/launch issues." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch vs checkpoint key mismatch; no shared underlying defect." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "TP OOM loading a large model vs key error loading MobileLLM; unrelated model-loading failures." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "FlashAttention2 support question vs Trainer broadcast NoneType error; different subsystems." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length question vs missing transformers.onnx module; unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 33, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:41950", + "issue:42175", + "issue:43065", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43475", + "issue:43504", + "issue:43540", + "issue:43572", + "issue:43646", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:43716", + "issue:43761", + "issue:44077", + "issue:44162", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:45127", + "issue:45200", + "issue:45406", + "issue:45542", + "issue:45584", + "issue:45632", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46202", + "issue:46234", + "issue:46365", + "issue:46369", + "issue:46424" + ], + "soft_pairs": [ + "issue:43316|issue:43716", + "issue:43295|issue:44871", + "issue:45632|issue:46424", + "issue:44610|issue:46369", + "issue:42175|issue:43316", + "issue:45803|issue:46365", + "issue:44871|issue:45584", + "issue:44077|issue:44534", + "issue:43540|issue:44514", + "issue:43065|issue:44617", + "issue:43650|issue:46202", + "issue:44162|issue:44792", + "issue:45803|issue:46170", + "issue:44871|issue:46050", + "issue:41950|issue:43475", + "issue:43646|issue:44913", + "issue:43452|issue:44496", + "issue:43404|issue:43441", + "issue:43761|issue:44479", + "issue:45127|issue:45200", + "issue:43388|issue:43688", + "issue:43673|issue:44743", + "issue:43572|issue:46178", + "issue:43504|issue:45698", + "issue:45542|issue:46234", + "issue:44734|issue:45406", + "issue:44617|issue:45584" + ], + "prepared_review_unit_hash": "7bede5940b667dc993b307f45ed08e524764215b80afa9bbc2f21fdc5c023e83", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35847, + "estimated_input_tokens": 8962, + "estimated_eval_tokens": 18180 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a mixed bag of unrelated Transformers regressions and model-specific bugs; the soft pairs share broad subsystems at best, but not the same concrete failure mode. I would not merge any of the suggested pairs.", + "confidence": 0.43, + "canonical_issue_reason": "Issue 45698 is the strongest representative because it is open, active, and describes a broad save/load regression with custom modules that is likely to attract duplicate reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45698 is the best overall triage target for the cluster because it is the most active and generic failure mode among the issues listed.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Different bugs: config API discrepancy vs image-preprocessor dtype mismatch." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Processor.tokenizer regression and Gemma eos_token_id mismatch are unrelated." + }, + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "Cache-path collision and max_cache_length behavior are different cache issues." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "Different models and failure modes: input size mismatch vs OOM." + }, + { + "left": "issue:42175", + "right": "issue:43316", + "accept": false, + "reason": "Backend installation dependency issue vs config API discrepancy." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "Bare-except exception handling bug vs Trainer import conflict are unrelated." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure are different code paths." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5-related, but one is post_init policy and the other is non-persistent buffer initialization." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Both involve Qwen processors, but video-input validation and batched chat-template padding are separate bugs." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "SAM3 PixelDecoder dummy layer and SAM3 video OOM are not the same problem." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "The titles indicate unrelated issues with no shared failure mechanism." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage and janus image-generation test failure are different models/issues." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Different root causes: tensor conversion exception handling vs Windows/ROCm import dependency failure." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Gemma config mismatch and sinq save/load quantization failure are separate save/load bugs." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Both are video-related, but one misroutes processor selection and the other is a missing vision-encoder attribute." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "General custom model initialization regression vs GPTNeoX config reload defaulting are different issues." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "gguf loading failure and unrecognized Olmo model config are separate model-loading problems." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Same model family, but weight tying and FlashAttention failures are distinct code paths." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "CLIP vision hidden-state regression and Qwen video-input regression are unrelated." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA/extended-vocab collapse and Gemma4 mm_token_type_ids defaulting are different bugs." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Metric gathering label truncation and auxiliary-loss normalization are unrelated training issues." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Generic GenerationMixin cache issue and Qwen recurrent-state reset are not the same concrete bug." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "StableLm config field regression and tensorflow_text backend mapping error are unrelated." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy preset-field loading failure and wrong custom-module loading after save_pretrained are distinct serialization bugs." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "TensorFlow backend detection and missing import_utils symbol are different import/backend issues." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "Serve KV-cache indexing crash and Gemma4Processor missing _tokenizer are separate serve regressions." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "SAM3 video OOM and Whisper empty-transcription failure are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 34, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:35141", + "issue:41720", + "issue:42490", + "issue:43065", + "issue:43335", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43644", + "issue:43673", + "issue:43761", + "issue:43827", + "issue:43931", + "issue:43976", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44464", + "issue:44479", + "issue:44521", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44792", + "issue:44855", + "issue:44877", + "issue:44913", + "issue:44933", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45245", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:46369", + "issue:46421" + ], + "soft_pairs": [ + "issue:43673|issue:46421", + "issue:43525|issue:43531", + "issue:44792|issue:44977", + "issue:45381|issue:46369", + "issue:44792|issue:45600", + "issue:44534|issue:44913", + "issue:44190|issue:44964", + "issue:43931|issue:44060", + "issue:43761|issue:44242", + "issue:45563|issue:45584", + "issue:35141|issue:42490", + "issue:44855|issue:45588", + "issue:44479|issue:44734", + "issue:43827|issue:43994", + "issue:43065|issue:44610", + "issue:44933|issue:45362", + "issue:41720|issue:45245", + "issue:44617|issue:46369", + "issue:44464|issue:45563", + "issue:43644|issue:44913", + "issue:43550|issue:44190", + "issue:43976|issue:44855", + "issue:44521|issue:45381", + "issue:44077|issue:44877", + "issue:25251|issue:27957", + "issue:45127|issue:45357", + "issue:43540|issue:44734", + "issue:43335|issue:44190" + ], + "prepared_review_unit_hash": "42bad97681888888bc9f9bcdac8d5a32b29c659f5b4ebba3706a72799278df13", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35345, + "estimated_input_tokens": 8837, + "estimated_eval_tokens": 17930 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the soft pairs share only broad surface themes (cache/chunked prefill, multimodal/video, config/loading, Python/version issues) but not the same concrete bug or fix. I did not find any true duplicate issue pairs.", + "confidence": 0.93, + "canonical_issue_reason": "No clear canonical issue: the set spans many unrelated bugs/regressions, and the apparent similarities are superficial rather than the same underlying problem.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the strongest single representative only because it is open and has the most inbound references/discussion, but it is still just a weak stand-in for this heterogeneous set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43673", + "right": "issue:46421", + "accept": false, + "reason": "Both mention chunked prefill/cache, but one is about a missing GenerationMixin cache in v5.0.0 and the other about lazy cache init causing recompilation; different failure modes." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Different model/config bugs: missing pad_token_id on Llama4Config vs sliding_window handling in Qwen3-MoE." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "A failed janus test and a Qwen3.5 flash-attention generation issue are not the same code-path problem." + }, + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "Both involve multimodal/video-ish models, but one is a vision position-id regression and the other is a memory overflow on sam3-litetext." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "Test failure for janus vs broken auto_mappings references for removed Sam3LiteText configs; unrelated bugs." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk initialization and rotary_pct reloading are different config/state bugs." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "A local dataset loading failure and a model weight-loading mismatch are different issues." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Different Qwen model families and different symptoms: shape mismatch on load vs tied-weights warning logic." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "CLIP hidden_states regression and load-balancing loss omission are separate model behaviors." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "Stale num_return_sequences warning in paged generate is unrelated to Whisper failing on empty transcription." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Embedding resize/post_init reinit bug is unrelated to the trailing-slash local model import warning." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 decorator parsing failure and flash_attention s_aux=None crash are different problems." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Qwen VL video regression and serve KV-cache continuation indexing are distinct code paths." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing pipeline() after removals is not the same as SigLIP pipeline producing nonsense outputs." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and OmDet-Turbo 640px vs 224px mismatch are unrelated vision bugs." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "Nonexistent import from image_utils and Qwen3.5 chat crashes are different failures." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert and category-count overflow have no shared underlying fix." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Both mention memory overflow, but they affect different models and likely different causes." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Compiled chunked generation output inconsistency is unrelated to a stale generate warning." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer initialization junk and GPTNeoX rotary_pct reload default are different state/config bugs." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to local dataset loading in the training script." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "Python version support mismatch and a Python 3.13 decorator parsing error are not the same issue." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal, but all-zero assistant_masks and wrong vision_position_ids are different bugs in different processing steps." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "patchtsmixer post_init policy and granite_speech strict config loading are unrelated configuration issues." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "Pipeline top_k nesting bug and XLMRoberta Flash Attention 2 support are entirely different topics." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge with extended vocab collapse and incorrect visual encoder keys in save_pretrained are different regressions." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and serve KV cache continuation indexing are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "Sparse layer creation in SwitchTransformersConfig and local dataset loading are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 35, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36246", + "issue:38617", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:42796", + "issue:43116", + "issue:43295", + "issue:43404", + "issue:43643", + "issue:43650", + "issue:43673", + "issue:43854", + "issue:44016", + "issue:44112", + "issue:44186", + "issue:44190", + "issue:44220", + "issue:44246", + "issue:44279", + "issue:44295", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44734", + "issue:44737", + "issue:44779", + "issue:44938", + "issue:44964", + "issue:45003", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45446", + "issue:45542", + "issue:45584", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:45907", + "issue:46082", + "issue:46121", + "issue:46234", + "issue:46421", + "issue:46424", + "issue:46489", + "issue:46506" + ], + "soft_pairs": [ + "issue:44779|issue:46489", + "issue:44462|issue:45907", + "issue:44112|issue:45600", + "issue:44737|issue:45803", + "issue:42796|issue:44521", + "issue:43116|issue:44190", + "issue:44938|issue:45542", + "issue:43295|issue:45276", + "issue:45276|issue:45584", + "issue:45632|issue:46234", + "issue:36246|issue:41628", + "issue:45245|issue:45584", + "issue:46082|issue:46121", + "issue:36246|issue:38617", + "issue:43404|issue:45216", + "issue:42175|issue:44279", + "issue:44246|issue:45003", + "issue:45632|issue:46421", + "issue:44186|issue:44734", + "issue:44683|issue:45446", + "issue:41950|issue:44016", + "issue:43643|issue:44496", + "issue:44295|issue:46506", + "issue:43854|issue:44964", + "issue:44964|issue:45600", + "issue:43673|issue:46424", + "issue:44112|issue:44964", + "issue:43650|issue:44220", + "issue:44016|issue:44220" + ], + "prepared_review_unit_hash": "bbc0de35a1b8d2543d86afa9a89430d026b76fc1a6548d54056f0a93642cd5dd", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36439, + "estimated_input_tokens": 9110, + "estimated_eval_tokens": 18476 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated transformers issues; none of the soft pairs look like true duplicates. If forced to anchor the cluster, issue #44521 is the strongest representative because it has a concrete bug report, active discussion, inbound references, and a clear multimodal API path.", + "confidence": 0.78, + "canonical_issue_reason": "Issue #44521 is the best anchor: it describes a concrete, reproducible bug in `apply_chat_template` for multimodal inputs, has multiple references/discussion, and is specific enough to guide triage.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #44521 is the most suitable global representative because it is substantive, well-scoped, and directly actionable; the rest are mostly isolated model/import/config regressions with no clear duplicate relationship.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44779", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but one is DeepSeek generally and the other is DeepSeek-Coder v1 with a specific v5+ gap; not the same concrete bug." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "`AutoTokenizer` loading `tokenizer.json` and `list_repo_files` revision forwarding are unrelated code paths." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Both mention CI, but one is a GraniteSpeech test failure and the other is stale auto-mapping references; different bugs." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "CPU device placement in XLNet and a bare `except` in Fuyu tensor conversion are unrelated issues." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer BPE marker handling and multimodal assistant mask generation are different failure modes and code paths." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Example script classification output and local dataset loading in a different script are not the same bug." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import failure and missing TensorFlow backend wiring are separate environment/dependency issues." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Processor/tokenizer regression and Gemma embedding resizing are different model behaviors with no shared concrete fix." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Gemma embedding resizing and Whisper empty transcription generation are unrelated." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "A `trust_remote_code` cache-path collision and a missing availability helper are different utilities." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they involve different symbols and different modules." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Category-count overflow and Whisper empty-transcription failure are unrelated." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Llama head-dim validation and rope-validation list handling are distinct config-validation bugs." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Different missing imports in different modules; not the same underlying defect." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "One is a weight-tying bug in Mistral3 generation, the other is a save_pretrained regression in Qwen3.5." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "A missing TensorFlow extra and a generic dependency issue are too broad and not the same bug." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import latency and unsafe `sys.modules` access may be related at a high level, but they are not the same concrete problem." + }, + { + "left": "issue:45632", + "right": "issue:46421", + "accept": false, + "reason": "Both mention cache, but one is a remote-code path collision and the other is chunked-prefill recompilation from lazy cache init." + }, + { + "left": "issue:44186", + "right": "issue:44734", + "accept": false, + "reason": "Tokenizer crashes for NER padding/truncation and KV-cache continuation indexing in serve are different subsystems." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both touch flex_attention, but one is a torch>=2.9 runtime failure and the other is a bad version check for an import gate." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification pipeline behavior and a notebook syntax error are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Missing fields with `trust_remote_code=True` and an unrecognized model/config missing `model_type` are different loader failures." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "Reading registered `position_ids` and CPU/CUDA output divergence in PPDocLayoutV3 are unrelated." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different model-loading failures for different models; no evidence of a shared code-path bug." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Phi-4 multimodal loading and stale Sam3LiteText auto-mapping references are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Chunked-prefill cache initialization and `max_cache_length` honoring are both cache-related but not the same defect." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flakiness in GraniteSpeech and loading Phi-4 multimodal are unrelated." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "The placeholder/spam report and `_torch_extract_fbank_features()` are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error and feature-extraction bug are different issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 36, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "issue:28282", + "issue:32090", + "issue:35141", + "issue:42796", + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43526", + "issue:43531", + "issue:43577", + "issue:43650", + "issue:43673", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44279", + "issue:44336", + "issue:44368", + "issue:44496", + "issue:44792", + "issue:44933", + "issue:44987", + "issue:45042", + "issue:45083", + "issue:45216", + "issue:45335", + "issue:45375", + "issue:45405", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45600", + "issue:45759", + "issue:46082", + "issue:46123", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46506", + "issue:46515" + ], + "soft_pairs": [ + "issue:27957|issue:35141", + "issue:43479|issue:45375", + "issue:43531|issue:45600", + "issue:45083|issue:46506", + "issue:43295|issue:45335", + "issue:43976|issue:44987", + "issue:32090|issue:35141", + "issue:28282|issue:35141", + "issue:43526|issue:43901", + "issue:44496|issue:45042", + "issue:46123|issue:46369", + "issue:42907|issue:43441", + "issue:44016|issue:44062", + "issue:43441|issue:45588", + "issue:43673|issue:44336", + "issue:44792|issue:46424", + "issue:43381|issue:45593", + "issue:43824|issue:44933", + "issue:43976|issue:44279", + "issue:43901|issue:44496", + "issue:43299|issue:45216", + "issue:42796|issue:43650", + "issue:44496|issue:45759", + "issue:44279|issue:46515", + "issue:45335|issue:45584", + "issue:44075|issue:44368", + "issue:43577|issue:45216", + "issue:46082|issue:46399", + "issue:44279|issue:45405" + ], + "prepared_review_unit_hash": "c985b7a90d8c903ced6e04837aba8f9bdab7306e951f197a4afde8f2cb43ca46", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35834, + "estimated_input_tokens": 8959, + "estimated_eval_tokens": 18174 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the paired issues share broad vocabulary in a few cases (FlashAttention, model/config loading, tokenizer behavior), but they do not look like the same underlying bug/change. No soft edge is strong enough to merge.", + "confidence": 0.91, + "canonical_issue_reason": "No single issue cleanly represents the cluster because it contains many unrelated bugs. If forced to pick one, issue 45588 is the most concrete runtime code-path failure and one of the few directly tied to an implementation crash rather than a broad symptom.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45588 is the best standalone representative only in the weak sense that it is specific, reproducible, and points at a named failing integration path. It is still not a good global canonical for this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "Different subsystems and failure modes: FlashAttention support for XLMRoberta vs embedding reinitialization after resize_token_embeddings." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both are config-related, but they concern different models and different bugs (default sub-config initialization vs a missing field dropped by strict config handling)." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding window behavior is unrelated to stale auto_mappings references for removed Sam3LiteText configs." + }, + { + "left": "issue:45083", + "right": "issue:46506", + "accept": false, + "reason": "Different models and code paths: a feature-extraction length helper in qwen3_omni_moe vs CPU/CUDA numeric divergence in PPDocLayoutV3." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "One is a regression in processor/tokenizer image handling; the other is decoder embedding resize behavior in t5gemma." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Version/platform compatibility issue vs a model-loading failure for a specific repo; not the same bug." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer broadcast TypeError is unrelated to embedding reinitialization after resizing token embeddings." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "Missing PyTorch import error is unrelated to token-embedding reinitialization." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "ImageProcessor label reduction bug and pipeline docs mismatch are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Model config recognition failure is unrelated to PIL backend torchvision dependency handling." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "Partial-batch handling in MaskGenerationPipeline is unrelated to memory overflow in sam3-litetext." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Both involve Ministral-family models, but one is about saving dequantized weights and the other about FlashAttention loading/runtime failure." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error is unrelated to AddedToken keyword argument duplication." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both mention FlashAttention, but the reported failures are different and one is model-specific while the other is a generic sink-less-model AttributeError; insufficient evidence they are the same bug." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Generation cache missing during chunked prefill is unrelated to ANSI code emission in loading_report." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "Janus image-generation test failure is unrelated to cache config max length semantics." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and missing auxiliary losses in D-FINE are unrelated behaviors." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they concern different missing symbols/modules and different underlying causes." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python version support issue is unrelated to a generic dependency issue." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Documentation mismatch for TextClassificationPipeline is unrelated to model config recognition failure." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both involve Qwen multimodal models, but one is a load-time breakage and the other is incorrect save_pretrained checkpoint output." + }, + { + "left": "issue:42796", + "right": "issue:43650", + "accept": false, + "reason": "Tokenizer BPE marker handling is unrelated to the vague 'ADD THE DATA' issue." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Unrecognized model config and failing to unwrap text_config are different multimodal-config bugs." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "General dependency issue is not the same as torch 2.4.1 incompatibility." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize not propagating to decoder.embed_tokens is unrelated to Whisper empty-transcription generation after token alignment." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer argument handling is unrelated to a LoRA warning about tie_word_embeddings." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "BLIP2 dtype propagation and Qwen3.5 checkpoint saving are unrelated issues." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "LlamaConfig head_dim validation is unrelated to a Qwen3.5 kernel registration/decorator crash." + }, + { + "left": "issue:46082", + "right": "issue:46369", + "accept": false, + "reason": "LlamaConfig validation is unrelated to sam3-litetext memory overflow." + }, + { + "left": "issue:43577", + "right": "issue:45584", + "accept": false, + "reason": "Dtype loading inconsistency for BLIP2 is unrelated to Whisper generation on empty transcription." + }, + { + "left": "issue:44075", + "right": "issue:46515", + "accept": false, + "reason": "SGD argument usage and torch version incompatibility are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 37, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:33290", + "issue:38175", + "issue:39692", + "issue:41950", + "issue:42994", + "issue:43295", + "issue:43299", + "issue:43317", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43650", + "issue:43653", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44242", + "issue:44291", + "issue:44297", + "issue:44351", + "issue:44462", + "issue:44466", + "issue:44521", + "issue:44617", + "issue:44683", + "issue:44779", + "issue:45020", + "issue:45081", + "issue:45103", + "issue:45200", + "issue:45276", + "issue:45278", + "issue:45325", + "issue:45357", + "issue:45372", + "issue:45405", + "issue:45542", + "issue:45593", + "issue:45656", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45907", + "issue:46082", + "issue:46209", + "issue:46512" + ], + "soft_pairs": [ + "issue:45103|issue:45372", + "issue:41950|issue:45020", + "issue:44351|issue:45278", + "issue:38175|issue:44617", + "issue:43650|issue:46209", + "issue:44242|issue:45200", + "issue:43299|issue:45357", + "issue:43295|issue:44779", + "issue:39692|issue:43901", + "issue:44521|issue:45325", + "issue:41950|issue:45823", + "issue:44683|issue:45542", + "issue:43976|issue:45405", + "issue:44779|issue:45812", + "issue:46082|issue:46512", + "issue:43299|issue:43854", + "issue:43479|issue:43525", + "issue:44297|issue:44462", + "issue:44466|issue:45276", + "issue:44016|issue:44291", + "issue:45081|issue:45907", + "issue:43441|issue:45834", + "issue:43653|issue:45593", + "issue:42994|issue:43317", + "issue:33290|issue:45656" + ], + "prepared_review_unit_hash": "7f409c6712cd5a00e4f16551e622639afce881f370a84caba09ab0102004771d", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 25, + "serialized_chars": 34690, + "estimated_input_tokens": 8673, + "estimated_eval_tokens": 17602 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are distinct regressions or docs/issues touching broad Transformer subsystems, not duplicates of one another. None of the soft pairs look like the same concrete bug/change, so all are rejected.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:44779 is the closest broad anchor because it describes a clear v4\u2192v5 tokenizer regression, which is the most generally representative problem theme in this set; however, it is still not a true canonical duplicate for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Different failures: an auto-docstring annotation crash vs a Gemma 4 processor import/loading regression." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "One is a video-classification processor lookup bug; the other is a broad remote_code compatibility regression." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "Both mention imports, but one is missing HybridCache specifically while the other is a wide set of import errors after upgrade." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "Unrelated symptoms and code paths: zero probabilities in SigLIP2 vs CUDA OOM in Sam3Video." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Different problems entirely: vague data handling vs FP8 inference failure on B200." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Both involve training losses/config, but one is router load-balancing loss and the other is mm_token_type_ids defaults." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Same model family, but one is a load regression and the other is a save_pretrained key-serialization bug." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer/processor regressions, but one is processor.tokenizer / image-input API breakage and the other is wrong Deepseek tokenization." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Both are documentation-related, but they refer to different docs and different bugs." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal/template issues, but assistant mask generation and temporal RoPE scaling are different code paths." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Different subsystems: video pipeline processor lookup vs Gemma4 PLE device placement." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "No overlap beyond general framework/runtime failure: flex_attention compilation vs tensorboard/TensorFlow backend handling." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Separate dependency/version issues affecting unrelated packages (Python support vs PEFT minimum version)." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and do not clearly point to the same underlying bug." + }, + { + "left": "issue:46082", + "right": "issue:46512", + "accept": false, + "reason": "Different configuration/state issues: rejecting explicit head_dim vs stale config.dtype after from_config(dtype=...)." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different models and bugs: Qwen3VL MOE load regression vs GLM-4.7-Flash unit-test loading issue." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Both are config-related, but one is default subconfig initialization and the other is a missing pad_token_id attribute." + }, + { + "left": "issue:44297", + "right": "issue:44462", + "accept": false, + "reason": "Both are tokenizer-related, but one is save_pretrained metadata mismatch and the other is AutoTokenizer ignoring tokenizer.json." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Different serialization/resize bugs: tied lm_head weight saving vs resize_token_embeddings not propagating to Gemma4 embeddings." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "A notebook syntax error is unrelated to the init_empty_weights TypeError regression." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Different targets and failure modes: Mistral tokenizer regex patch crash vs list_repo_files not forwarding revision." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "Unrelated model-specific inference bugs: FlashAttention failure vs long OCR index error." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Different training/evaluation bugs: BigBirdTokenizer special-token registration vs D-FINE auxiliary-loss handling." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Both involve quantization/offload, but one is saving a quantized model and the other is loading a dequantized model with device_map=auto." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Both mention DeepSpeed and optimizers, but one is Adafactor OOM and the other is an optimizer.step() being invoked twice." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 38, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:41628", + "issue:42175", + "issue:42491", + "issue:43299", + "issue:43441", + "issue:43525", + "issue:43643", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", + "issue:44162", + "issue:44188", + "issue:44242", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44336", + "issue:44361", + "issue:44373", + "issue:44496", + "issue:44589", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:45103", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45561", + "issue:45563", + "issue:45593", + "issue:45600", + "issue:45712", + "issue:45800", + "issue:45823", + "issue:46291", + "issue:46399" + ], + "soft_pairs": [ + "issue:45600|issue:45712", + "issue:44623|issue:44749", + "issue:44261|issue:44263", + "issue:44623|issue:45397", + "issue:44361|issue:45593", + "issue:45542|issue:45800", + "issue:44016|issue:45823", + "issue:42175|issue:43976", + "issue:44671|issue:44779", + "issue:43441|issue:44683", + "issue:43901|issue:43994", + "issue:43723|issue:44162", + "issue:46291|issue:46399", + "issue:43673|issue:43761", + "issue:43525|issue:43881", + "issue:43824|issue:43976", + "issue:43901|issue:44246", + "issue:43525|issue:43643", + "issue:44297|issue:45397", + "issue:44589|issue:45362", + "issue:44188|issue:44336", + "issue:43784|issue:45542", + "issue:41628|issue:43976", + "issue:44336|issue:44373", + "issue:45561|issue:45563", + "issue:45103|issue:46399", + "issue:44246|issue:44496", + "issue:44295|issue:45397", + "issue:44242|issue:45593", + "issue:42491|issue:43299" + ], + "prepared_review_unit_hash": "de0ba74d655b4538c9c49494c893e6be0b2b933cc9e076e764e535a8e7a3a8d0", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36597, + "estimated_input_tokens": 9150, + "estimated_eval_tokens": 18556 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose soft-similarity bucket of mostly unrelated Transformers issues: packaging/backends, model loading, tokenizer regressions, docs, and kernel/runtime bugs. None of the soft pairs look like true duplicates, so all edges should be rejected.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43299 is the clearest concrete regression report and the closest to the only near-duplicate-looking subgroup (Qwen3 MoE loading), though the cluster overall is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43299 is the most representative standalone issue here: specific version, specific model family, and an actionable load failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Different file areas and failure modes: stale auto_mappings references vs leftover dummy classes leaking into dir(transformers)." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "processor.save_pretrained missing files is unrelated to the Chinese report about post-upgrade filtering slowdown." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Different models and code paths: MLA q_a_layernorm precision issue vs GlmMoeDsaIndexer torch.split behavior." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Save-pretrained missing files and gemma-4 zero3 from_pretrained are distinct load/save bugs on different models." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer AttributeError on tasks is unrelated to D-FINE missing auxiliary losses when denoising is off." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "Packaging/backend installation mismatch is not the same as a torch/transformers version incompatibility." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "A notebook syntax error is a different problem from a Gemma4 forward device-map mismatch." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Missing TensorFlow in the torch extra and Python-version support regression are different environment issues." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "CamemBERT masked-LM predictions and Deepseek tokenizer outputs are different components and bugs." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "Both involve attention kernels, but the triggers and code paths differ: FlashAttention on Ministral-3 vs compiled flex_attention on torch>=2.9." + }, + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Docs drift about return_all_scores is not the same as incorrect SigLIP2 model/pipeline outputs." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "AutoTokenizer loading failure and ESM2 breakage are different models and failure modes." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "Kernel package metadata import failure is unrelated to the apply_rotary_pos_emb decorator registration crash." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Generation cache missing during chunked_prefill is a different bug from CLIPVisionModel returning hidden_states=None." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Missing pad_token_id on Llama4Config and glm-4v-9b loading failure are not the same underlying issue." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Missing Qwen2.5-VL symbol import and Python version incompatibility are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "A documentation mismatch and intermittent import slowness are not duplicates." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Both mention missing config fields, but one is a specific Llama4Config attribute and the other is a trust_remote_code serialization gap." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "tokenizer_class mismatch in save_pretrained is unrelated to gemma-4 zero3 from_pretrained failures." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage lookup errors and Qwen3.5 chat crashes are different bugs with no shared code path." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence on torch.compile and ANSI code leakage from loading_report are unrelated." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "sentence-transformers import NameError and undefined tf backend from missing TensorFlow are different import-time failures." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "AutoImageProcessor import error is unrelated to the Python version support issue." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI output leakage and a wrong docstring for position_ids are different severity and different subsystems." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "A pytest-xdist race in testing utils is unrelated to a stale warning in paged generate()." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "auto_docstring crashes under future annotations and kernel decorator registration errors are different code paths." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Intermittent import slowness is not the same as an unrecognized model/config.json error." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer-reading error and gemma-4 zero3 from_pretrained are distinct bugs." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Both concern missing losses, but one is load-balancing loss with router logits and the other is D-FINE auxiliary losses with denoising; different models and logic." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both are Qwen3 MoE-related, but one is LoRA compatibility across hf4.x/hf5.x and the other is base Qwen3VL MoE loading on v5; not the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 39, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36246", + "issue:41950", + "issue:42757", + "issue:42886", + "issue:43257", + "issue:43262", + "issue:43352", + "issue:43475", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43673", + "issue:43901", + "issue:44038", + "issue:44162", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44485", + "issue:44496", + "issue:44561", + "issue:44734", + "issue:44749", + "issue:44779", + "issue:44857", + "issue:44871", + "issue:44877", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45250", + "issue:45292", + "issue:45362", + "issue:45375", + "issue:45561", + "issue:45584", + "issue:45701", + "issue:45759", + "issue:45859", + "issue:46132", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46506" + ], + "soft_pairs": [ + "issue:43531|issue:45362", + "issue:43526|issue:45003", + "issue:41950|issue:44220", + "issue:44246|issue:44871", + "issue:44336|issue:44485", + "issue:42886|issue:46132", + "issue:43550|issue:46506", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:44734|issue:46424", + "issue:44336|issue:46424", + "issue:45003|issue:45561", + "issue:44779|issue:45292", + "issue:45859|issue:46399", + "issue:36246|issue:43596", + "issue:43475|issue:46369", + "issue:44295|issue:44749", + "issue:44242|issue:44857", + "issue:43673|issue:44561", + "issue:43901|issue:45003", + "issue:43352|issue:45250", + "issue:44162|issue:45200", + "issue:43257|issue:45137", + "issue:44496|issue:45003", + "issue:45292|issue:45584", + "issue:43262|issue:44246", + "issue:45292|issue:45701", + "issue:45375|issue:45759", + "issue:44877|issue:45375", + "issue:44038|issue:46506", + "issue:44561|issue:46399" + ], + "prepared_review_unit_hash": "cc0b5c61d2798b2bde2ddb10037952f5b2be177d4b03d13afd0143f9db4c81b4", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36813, + "estimated_input_tokens": 9204, + "estimated_eval_tokens": 18664 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the soft links are mostly title-level similarity across unrelated bugs (Qwen, DeepSpeed, cache handling, docs, audio, vision, import errors). None of the proposed soft edges look like true duplicates, so all should be rejected. If forced to name a representative issue, 45003 is the broadest hub-like report, but it is not a real canonical duplicate anchor.", + "confidence": 0.92, + "canonical_issue_reason": "issue:45003 is the broadest, most central infrastructure-style report in the set and has the most superficial overlap with other items, but the cluster is too diverse for a true duplicate canonical.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45003 is the best overall representative only in the sense of being the broadest hub, not because it duplicates the others; it should not be treated as a merged target for this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Both mention Qwen, but one is a sliding_window bug in Qwen3-MoE and the other is a chat crash in Qwen3.5-35B; different failures and code paths." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction vs generic sys.modules access; unrelated components and symptoms." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification processor lookup vs audio fbank extraction; different pipeline areas." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import slowness report vs Gemma eos_token_id config mismatch; unrelated bugs." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI output in loading_report vs GLM-5 RoPE implementation; not the same issue." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache-loading failure vs AttentionInterface.register behavior change; different APIs and code paths." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure vs PPDocLayoutV3 CPU/CUDA divergence; different models and backend problems." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different missing symbols in different packages; not the same underlying bug." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Missing Qwen2.5-VL image processor vs missing HybridCache export; different symbols and fixes." + }, + { + "left": "issue:44734", + "right": "issue:46424", + "accept": false, + "reason": "KV-cache continuation tensor indexing crash vs max_cache_length behavior for Quantized Cache; different cache logic." + }, + { + "left": "issue:44336", + "right": "issue:46424", + "accept": false, + "reason": "Terminal-color formatting bug vs cache-length semantics; unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "sys.modules access bug vs pytest-xdist captured_info.txt race; different subsystems." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "DeepSeek tokenizer correctness regression vs resize_token_embeddings not updating output embeddings; distinct behaviors." + }, + { + "left": "issue:45859", + "right": "issue:46399", + "accept": false, + "reason": "CPU offload incompatibility vs missing decorator for kernel registration; both Qwen3.5-related but not the same defect." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Qwen2.5-VL import failure vs DeepSpeed ZeRO3 BertModel index error; unrelated." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "SAM3 video encoder attribute error vs SAM3-litetext memory overflow; different failure modes." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Position_ids buffer access error vs a transformers performance regression after upgrade; unrelated." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "MoE load-balancing loss omission vs float16 AMP crash in LwDetrImageLoss; different bugs." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Chunked-prefill cache regression vs v5 removal of is_torch_fx_available breaking trust_remote_code; different breakages." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mismatch for return_all_scores vs sys.modules access bug; unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Specific Nemotron Flash Attention 2.0 unsupported error vs a generic Flash Attention 2.0 topic; not the same concrete issue." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage vs Gemma 4 mm_token_type_ids defaulting; unrelated models and code paths." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Qwen3 MOE weight conversion issue vs ZeRO3 deque underflow; different DeepSpeed failures." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized model config vs unsafe sys.modules access; unrelated." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize/output-head sync bug vs Whisper empty-transcription generation failure; different subsystems." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio apply_chat_template sampling-rate bug vs import slowness; unrelated." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "resize_token_embeddings/output embedding mismatch vs tokenizer behavior changes across versions; distinct issues." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Missing Qwen3.5MoeVisionConfig field vs composite config not unwrapping text_config; similar family, but different config defects." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Strict granite_speech config loading vs Qwen vision config field loss; different models and symptoms." + }, + { + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3-VL-Moe v5 bug vs PPDocLayoutV3 CPU/CUDA divergence; unrelated." + }, + { + "left": "issue:44561", + "right": "issue:46399", + "accept": false, + "reason": "Removal of is_torch_fx_available breaking trust_remote_code vs Qwen3.5 apply_rotary_pos_emb registration problem; different regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 40, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:42886", + "issue:43065", + "issue:43232", + "issue:43408", + "issue:43475", + "issue:43540", + "issue:43646", + "issue:43653", + "issue:43673", + "issue:43701", + "issue:43906", + "issue:43931", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44188", + "issue:44297", + "issue:44361", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44483", + "issue:44623", + "issue:44749", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45335", + "issue:45397", + "issue:45412", + "issue:45459", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45656", + "issue:45685", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46170", + "issue:46202", + "issue:46421", + "issue:46431" + ], + "soft_pairs": [ + "issue:45412|issue:46123", + "issue:44466|issue:45335", + "issue:44987|issue:45459", + "issue:45137|issue:45600", + "issue:45685|issue:46431", + "issue:43475|issue:43540", + "issue:44995|issue:45200", + "issue:44060|issue:44077", + "issue:45397|issue:45656", + "issue:44479|issue:46170", + "issue:43653|issue:44162", + "issue:44368|issue:44977", + "issue:42886|issue:45229", + "issue:44483|issue:46202", + "issue:43906|issue:46202", + "issue:44995|issue:45588", + "issue:43065|issue:43408", + "issue:44368|issue:45563", + "issue:44188|issue:46202", + "issue:44877|issue:46170", + "issue:43931|issue:44368", + "issue:45987|issue:46082", + "issue:43065|issue:46123", + "issue:44188|issue:46421", + "issue:45584|issue:46170", + "issue:44297|issue:44623", + "issue:44361|issue:44749", + "issue:43646|issue:44162", + "issue:44749|issue:45656", + "issue:44623|issue:45656", + "issue:43701|issue:44749", + "issue:43232|issue:43673" + ], + "prepared_review_unit_hash": "51928517ff6df04990dc7001f8acd62f7a0ed8a50fb0b7b1bdac3d7f41e4c05b", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37276, + "estimated_input_tokens": 9319, + "estimated_eval_tokens": 18894 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated standalone issues that happen to share broad themes like generation, caching, tokenizers, or Gemma/Qwen/SAM subsystems. The soft pairs are not the same concrete bug or change, so none should be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44466 is the clearest standalone regression report here: it has a precise version boundary, a focused serialization/tied-weights symptom, and good reproducibility, so it is the best anchor among otherwise unrelated issues.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44466 is the strongest single triage candidate because it is specific, regression-oriented, and well-scoped; the rest of the set does not form a coherent duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "Different bugs: RT-DETR memory retention vs MaskGenerationPipeline final partial-batch handling." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both involve weights/embedding behavior, but one is serialization inconsistency and the other is resize_token_embeddings not propagating to decoder.embed_tokens." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Different tokenizer-loading failures: one breaks loading a specific model, the other hides real tokenizer errors when protobuf is missing." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "Unrelated: DeepSpeed ZeRO3 deque underflow vs stale Sam3LiteText references in auto_mappings.py." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "Different areas and failures: MPS histogram on Int vs chat-template absence in a processor." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "Both are video-related, but one is a missing attribute in SAM3 output and the other is a Qwen3OmniMoe input-processing ValueError." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "Different root causes: stale indexer cache on a second forward pass vs Gemma 4 mm_token_type_ids defaulting." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Only broad theme overlap; one is a tied-weights warning bug, the other is a deprecation/allowance issue for optional post_init." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Different bugs and code paths: gemma-4 zero3 from_pretrained vs optimizer step being called twice under deepspeed." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Different failure modes: video input regression in v5.3 vs Windows/ROCm import or torch.distributed dependency issues." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "Different tokenizer/model problems: BigBirdTokenizer special-token registration vs ESM2 being broken broadly." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "One is a tie_word_embeddings warning during LoRA fine-tuning, the other is a flash-attention generation regression." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Unrelated: offline cache tokenizer load failure vs Gemma4 multi-GPU OOM." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "Different APIs and symptoms: chat/completions request validation vs Gemma pipeline roundtrip on the \u2581 token." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "The latter is a Gemma pipeline roundtrip bug; the former is only an isolated reproduction of a different issue." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Different bugs: stale cache on repeated forward pass vs flash_attention.py crashing on s_aux=None." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 PixelDecoder dummy Conv2d vs sam3_video/sam3_tracker config mismatch are not the same defect." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "One is a tie-weights warning during fine-tuning; the other is a stale paged-generate warning about num_return_sequences." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "Different concerns: torch.compile branching divergence in attention kernels vs Gemma pipeline token roundtrip failure." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading vs Windows ROCm import/type errors are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Model weight-shape mismatch on load vs tied-weights warning during LoRA fine-tuning are different problems." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "StaticCache return-type contract bug vs LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "Different subsystems and failures: SAM3 decoder placeholder layer vs final partial-batch result dropping." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "Both mention cache/compile behavior, but one is attention-kernel divergence and the other is spurious recompilation from lazy cache init." + }, + { + "left": "issue:45584", + "right": "issue:46170", + "accept": false, + "reason": "Whisper empty transcription after align_special_tokens is unrelated to Windows ROCm import failures." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both are save_pretrained-related, but one is tokenizer_config metadata mismatch and the other is missing processor files." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "MLukeTokenizer AttributeError and a Chinese report of performance slowdown after upgrading are not the same bug." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Generic Transformers 5.0 custom init breakage does not match the separate ESM2 breakage report." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Performance regression after upgrading vs duplicate optimizer-step behavior under deepspeed are unrelated." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint key mismatch during resume is unrelated to a filtering performance regression." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both concern generation/caching, but the concrete failures are different: sync_gpus update timing vs cache missing during chunked_prefill." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 41, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:36010", + "issue:41628", + "issue:42175", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43704", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43873", + "issue:44038", + "issue:44188", + "issue:44222", + "issue:44351", + "issue:44623", + "issue:44792", + "issue:44869", + "issue:44871", + "issue:44933", + "issue:44945", + "issue:44964", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45406", + "issue:45478", + "issue:45712", + "issue:45753", + "issue:46132", + "issue:46202", + "issue:46326", + "issue:46399" + ], + "soft_pairs": [ + "issue:44871|issue:46132", + "issue:44869|issue:45356", + "issue:42994|issue:43482", + "issue:45230|issue:45397", + "issue:43475|issue:43643", + "issue:44038|issue:44623", + "issue:42371|issue:45092", + "issue:28282|issue:30990", + "issue:43475|issue:45381", + "issue:41628|issue:44933", + "issue:45375|issue:45478", + "issue:42994|issue:43525", + "issue:45310|issue:45375", + "issue:45406|issue:46202", + "issue:43723|issue:46326", + "issue:43784|issue:46399", + "issue:42673|issue:43704", + "issue:44964|issue:45137", + "issue:43873|issue:44945", + "issue:43232|issue:44188", + "issue:41628|issue:44351", + "issue:45081|issue:45103", + "issue:44945|issue:45753", + "issue:42994|issue:43824", + "issue:44222|issue:45310", + "issue:44038|issue:45310", + "issue:36010|issue:41628", + "issue:43232|issue:44792", + "issue:42831|issue:44222", + "issue:42175|issue:45712" + ], + "prepared_review_unit_hash": "886f0687ca4233881602d421cbc6e81795e49d716da5a1b9388a5a32cda94572", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36485, + "estimated_input_tokens": 9122, + "estimated_eval_tokens": 18500 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous transformer-related bugs. I only found two clear duplicate pairs: the Qwen3ForCausalLM VRAM leak reports and the Qwen3.5 MoE from_pretrained error reports. Everything else is too broad or only superficially similar.", + "confidence": 0.92, + "canonical_issue_reason": "issue:42673 is the clearest canonical issue: it has the fuller report for the exact duplicate pair with issue:43704, older age, and more discussion, making it the best representative of the duplicate subcluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the strongest issue representative overall because it is an exact duplicate target with substantial discussion and a specific reproducible symptom; the other apparent duplicate pair (issue:45310 / issue:45478) is separate and later.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Different bugs: Gemma eos-token configuration mismatch vs AttentionInterface behavior change." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both tokenizer/decoding-related, but they describe different failure modes and code paths." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Different model issues: quantized save_pretrained failure vs Qwen2.5-GGUF loading failure." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Too vague on the left and a specific ZeRO3/gemma-4 from_pretrained issue on the right; no clear same bug." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "One is a SAM 3 video attribute error, the other is missing fields from trust_remote_code config loading." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Different subsystems: Qwen3-VL-Moe bug vs processor.save_pretrained missing files." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API guidance is unrelated to InternVL2 meta-initialization incompatibility." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "Both mention loading, but one is a PyTorch import error and the other a sentence-transformers stall." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "Different multimodal models and distinct video/position-id failures." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they target different missing symbols and likely different export issues." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Same model family, but the reported missing field vs from_pretrained error are not clearly the same underlying bug." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Unrelated: quantized model saving failure vs Llama4Config missing pad_token_id." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Related Qwen3.5 MoE area, but one is a generic from_pretrained error and the other is a missing config field report; not enough evidence to merge." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Gemma4 processor attribute error and Gemma pipeline roundtrip failure are different code paths." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer loading issue in v5 vs continue_final_message behavior for mistral_common tokenizers are different bugs." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "Import-time NameError in sentence-transformers vs Qwen3.5 rotary registration crash; different causes." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact duplicate reports of the same Qwen3ForCausalLM VRAM leak with the same title and symptom." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Different failures: model loading vs DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization/offloading behavior vs incorrect output under pipeline parallelism are distinct problems." + }, + { + "left": "issue:43232", + "right": "issue:44188", + "accept": false, + "reason": "Generation kwargs update issue vs attention-kernel divergence under torch.compile; unrelated code paths." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing top-level symbols." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Tokenizer regex crash and auto_docstring annotation crash are unrelated." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline-parallelism output bug vs Qwen3_5 image-specific infinite loop; not the same issue." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized saving failure vs missing Qwen2_5_VLForConditionalGeneration import are different issues." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 save_pretrained on MoE vs Qwen3.5 MoE from_pretrained error; related area but not the same concrete bug." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Different Qwen/MoE reports with different triggers and symptoms." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "GenerationMixin import error and AutoImageProcessor import error are distinct missing-symbol regressions." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation helper bug vs janus image generation test failure; no clear common root cause." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "FP8 accuracy issue and FP8 save_pretrained issue are different failure modes." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Backend installation packaging issue vs dummy class cleanup; not the same bug." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same title and same Qwen3.5 MoE from_pretrained error; clear duplicate reports." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The two accepted duplicate pairs are well supported by matching titles/symptoms, and the other soft-edge rejections are cautious rather than overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 42, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:42175", + "issue:42222", + "issue:42898", + "issue:43278", + "issue:43295", + "issue:43377", + "issue:43479", + "issue:43504", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43976", + "issue:44016", + "issue:44060", + "issue:44075", + "issue:44079", + "issue:44222", + "issue:44242", + "issue:44295", + "issue:44297", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44869", + "issue:44933", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45137", + "issue:45200", + "issue:45310", + "issue:45356", + "issue:45405", + "issue:45561", + "issue:45600", + "issue:45656", + "issue:45712", + "issue:45812", + "issue:46077", + "issue:46123", + "issue:46231" + ], + "soft_pairs": [ + "issue:44496|issue:45561", + "issue:42175|issue:44933", + "issue:43504|issue:45561", + "issue:45561|issue:45600", + "issue:43976|issue:44933", + "issue:43278|issue:45137", + "issue:46077|issue:46123", + "issue:43377|issue:44242", + "issue:30990|issue:35141", + "issue:42898|issue:44568", + "issue:30333|issue:33290", + "issue:44623|issue:45310", + "issue:43479|issue:44869", + "issue:42222|issue:45200", + "issue:43479|issue:44079", + "issue:42898|issue:45812", + "issue:43723|issue:44568", + "issue:43874|issue:45092", + "issue:44016|issue:46231", + "issue:43784|issue:43824", + "issue:42898|issue:44779", + "issue:44734|issue:45103", + "issue:42898|issue:45356", + "issue:44295|issue:45656", + "issue:44075|issue:46231", + "issue:42898|issue:43295", + "issue:44060|issue:45125", + "issue:45405|issue:45712", + "issue:44222|issue:44297", + "issue:44933|issue:45712" + ], + "prepared_review_unit_hash": "038620d7d0dbccf517b358e866e579dec73ecd03ca001705e6d8bfd2edfdf75d", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36943, + "estimated_input_tokens": 9236, + "estimated_eval_tokens": 18728 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose cluster of mostly unrelated Transformers bug reports, concentrated around tokenizer/model-regression issues but spanning many separate code paths. None of the soft pairs look like the same underlying bug, so all proposed merges are rejected.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45812 is the clearest high-signal, user-facing regression here: a silent AutoTokenizer token-id break across Granite models with a well-scoped tokenizer path and broad impact.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the strongest standalone anchor in the set because it describes a concrete, broadly impactful regression with a specific reproducer and failure mode.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Different problems: a model-loading/config error versus a pytest-xdist captured-info race." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Unrelated: missing TensorFlow in an extra install versus a nonexistent import from image_utils." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "Legacy-field model preset loading failure does not match the xdist file-race bug." + }, + { + "left": "issue:45561", + "right": "issue:45600", + "accept": false, + "reason": "Test harness race on captured_info.txt versus stale auto_mappings references to removed configs." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python-version compatibility issue versus an import path bug in image_utils." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "Embedding dtype drift in eval is a different failure mode from a DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Gemma4 attention-mask construction bug versus a batch-iteration result-dropping bug." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Different model paths: MIMI padding-mask mismatch versus MoE load-balancing-loss gating." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Sentence-transformers loading hang is unrelated to token-embedding reinitialization in post_init." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer cleanup-space behavior change is not the same as missing BOS/EOS insertion on add_special_tokens." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow failure reporting and Adafactor OOM in DeepSpeed are distinct training issues." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor save_pretrained file omissions do not match the Qwen3.5-Moe from_pretrained error." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Default multimodal config initialization is unrelated to Whisper timestamp decoding on trailing replacement chars." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Broken vitpose presets and Gemma 4 mm_token_type_ids defaults are different model families and bugs." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Config-default initialization bug versus ModelOutput key assignment semantics." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but one is cleanup-space behavior and the other is wrong token IDs for Granite models." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "AutoTokenizer loading regression in v5 is not the same bug as missing special-token insertion for mdeberta-v3." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-patch helper is unrelated to old InternVL2 checkpoints breaking under meta initialization." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Notebook syntax error and documentation parameter-name mismatch are not duplicate code bugs." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Import NameError in sentence-transformers is unrelated to Qwen2.5-VL import availability." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Cleanup-space regression and Deepseek tokenizer incorrect results are separate tokenizer failures." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Serve KV-cache tensor indexing crash versus auto_docstring future-annotations crash are different subsystems." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Cleanup-space behavior change does not match the Kimi-K2.5 codec-handling regression." + }, + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "position_ids buffer access error is a different issue from DeepSpeed causing optimizer.step to run twice." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument handling bug and a docs typo in set_training are not the same code-path problem." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Tokenizer cleanup-space behavior is unrelated to processor.tokenizer / image-passing regression in custom model code." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Both mention Qwen3 variants, but the tied-weights warning and missing tensor-parallel plan are different defects." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "PEFT version pinning issue and leftover dummy-class cleanup are unrelated maintenance problems." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "FP8 save_pretrained behavior and tokenizer_class metadata mismatch affect different save paths." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Missing import from image_utils is unrelated to stale dummy classes leaking into dir(transformers)." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 43, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28282", + "issue:30333", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:41628", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:43122", + "issue:43232", + "issue:43475", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:44190", + "issue:44206", + "issue:44222", + "issue:44263", + "issue:44279", + "issue:44295", + "issue:44297", + "issue:44373", + "issue:44448", + "issue:44484", + "issue:44485", + "issue:44623", + "issue:44734", + "issue:44871", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45103", + "issue:45245", + "issue:45310", + "issue:45381", + "issue:45656", + "issue:45834", + "issue:46082", + "issue:46132", + "issue:46209", + "issue:46305", + "issue:46328", + "issue:46421", + "issue:46424", + "issue:46431" + ], + "soft_pairs": [ + "issue:27957|issue:30333", + "issue:30333|issue:34689", + "issue:44297|issue:45656", + "issue:44448|issue:46132", + "issue:44222|issue:44623", + "issue:30333|issue:36331", + "issue:43479|issue:43644", + "issue:42796|issue:43723", + "issue:42831|issue:44623", + "issue:30333|issue:35141", + "issue:22355|issue:28282", + "issue:45245|issue:46132", + "issue:46209|issue:46431", + "issue:41628|issue:42757", + "issue:43475|issue:46305", + "issue:44297|issue:45310", + "issue:43122|issue:46209", + "issue:43723|issue:44977", + "issue:45381|issue:46305", + "issue:45071|issue:46082", + "issue:44263|issue:45092", + "issue:44206|issue:45103", + "issue:43232|issue:46424", + "issue:44222|issue:44484", + "issue:44734|issue:46421", + "issue:44373|issue:44485", + "issue:43232|issue:45834", + "issue:44295|issue:44623", + "issue:44190|issue:46082", + "issue:43723|issue:44871", + "issue:44279|issue:46328" + ], + "prepared_review_unit_hash": "16055b342215787629e40ebc54d606e00a53d57d0f8d3d2f31c79b9800d4d0b3", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36724, + "estimated_input_tokens": 9181, + "estimated_eval_tokens": 18618 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the soft edges are mostly weak topical overlaps (imports, tokenizers, config, generation, cache, FP8, etc.) rather than the same underlying bug. No PRs are present, and there is no clear duplicate canonical issue.", + "confidence": 0.95, + "canonical_issue_reason": "No single issue is a defensible canonical: the items cover many unrelated failure modes, so the cluster should likely be split instead of deduplicated.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a strong global representative; the set is too broad to treat as one bug report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27957", + "right": "issue:30333", + "accept": false, + "reason": "Different problems: Flash Attention 2 model support vs MLflow reporting on failed training jobs." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Unrelated: MLflow job-status reporting vs Llama 3.2 Vision model-loading regression." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to optimizer stepping twice under deepspeed." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Different code paths: Pegasus output changes vs AttentionInterface.register altering callback behavior." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 MoE model serialization and the other is missing processor files." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "Training-job MLflow reporting is unrelated to CustomTrainer.compute_loss signature changes." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config default initialization bug vs non-persistent buffer initialization bug." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Tokenizer BPE-marker handling and AutoTokenizer loading regression are different bugs." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FP8 accuracy regression is unrelated to processor.save_pretrained missing files." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "MLflow failure reporting and embedding resize/post_init reinitialization are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Both are import errors, but for different missing dependencies and modules." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-cardinality runtime error is unrelated to AttentionInterface.register behavior changes." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "FP8 inference on B200 is unrelated to processor chat-template absence." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Different import failures in different packages/modules; not the same bug." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "SAM 3 Vision missing attribute and MoonshineStreaming mask off-by-one are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer-class metadata mismatch is not the same as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenization differences across versions are unrelated to FP8 inference failures on B200." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading regression and Qwen3.5 flash-attention generation failure are different code paths." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Both mention positional handling, but they are model/modality-specific and not the same defect." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "Config type-checking regression is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split return values bug is unrelated to InternVL2 meta-initialization incompatibility." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "Feature extractor argument handling and auto-docstring annotation crashes are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation kwargs update after sync_gpus is not the same bug as cache max_cache_length handling." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save_pretrained issue is unrelated to the default 50GB shard-size question." + }, + { + "left": "issue:44734", + "right": "issue:46421", + "accept": false, + "reason": "KV-cache continuation indexing crash and chunked-prefill cache recompilation are different cache bugs." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring typo/accuracy issue is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation sync_gpus bug is unrelated to Kosmos2.5 long-OCR index errors." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Position_ids buffer read error is unrelated to missing processor save_pretrained files." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading in an example script is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer loading regression and eos_token_id config inconsistency are different failures." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A vague dependency issue is not the same as a SequenceFeatureExtractor.pad performance regression." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 44, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28218", + "issue:30990", + "issue:38175", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42757", + "issue:42796", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43479", + "issue:43550", + "issue:43576", + "issue:43644", + "issue:43646", + "issue:43784", + "issue:43824", + "issue:44038", + "issue:44261", + "issue:44263", + "issue:44360", + "issue:44368", + "issue:44373", + "issue:44483", + "issue:44568", + "issue:44829", + "issue:44913", + "issue:45092", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45593", + "issue:45753", + "issue:46002", + "issue:46050", + "issue:46132", + "issue:46305", + "issue:46369", + "issue:46402" + ], + "soft_pairs": [ + "issue:44483|issue:45406", + "issue:44263|issue:44360", + "issue:42907|issue:46050", + "issue:45092|issue:45357", + "issue:44261|issue:45092", + "issue:30990|issue:38175", + "issue:43644|issue:46305", + "issue:42831|issue:45310", + "issue:44829|issue:46002", + "issue:44263|issue:44373", + "issue:41628|issue:43824", + "issue:42831|issue:44038", + "issue:43479|issue:44913", + "issue:43408|issue:46402", + "issue:22355|issue:27957", + "issue:42757|issue:45278", + "issue:39401|issue:44568", + "issue:42796|issue:43122", + "issue:42831|issue:45753", + "issue:42886|issue:43066", + "issue:22355|issue:28218", + "issue:44368|issue:46305", + "issue:27957|issue:30990", + "issue:42886|issue:42994", + "issue:43232|issue:45593", + "issue:43408|issue:46369", + "issue:43646|issue:46002", + "issue:43550|issue:46132", + "issue:43576|issue:43784", + "issue:42371|issue:44263" + ], + "prepared_review_unit_hash": "3b02d830f508d6e19d31df2c7ae0dc1e8cb83edeb578de6745923a41d930b682", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36250, + "estimated_input_tokens": 9063, + "estimated_eval_tokens": 18382 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are unrelated model, tokenizer, import, or config bugs, and there are no pull requests. I treated the broad import-regression report as the best issue anchor, but rejected all soft edges as non-duplicates.", + "confidence": 0.68, + "canonical_issue_reason": "Issue #45278 is the broadest umbrella item here: it captures multiple import failures after a version upgrade, making it the most representative anchor for this mixed set.", + "canonical_pr_reason": null, + "best_issue_reason": "Best general issue candidate for dedup triage because it summarizes a class of upgrade-related import breakages; the other issues are narrower, model- or tokenizer-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Different failures: chat/completions request validation vs Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Same subsystem (DSA indexer) but different bugs: torch.split behavior vs missing ReLU." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Different quantization save/load paths: dequantized Ministrals/Devstrals vs sinq models." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Different models and code paths: InternVL2 meta-init incompatibility vs Qwen3.5 visual encoder save keys." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Unrelated bugs: MLA RMS norm epsilon precision issue vs InternVL2 meta initialization failure." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Different symptoms and models: stuck loading sentence-transformers vs zero probabilities in SigLIP2." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Different code paths: non-persistent buffer initialization junk vs sliding-window mask off-by-one." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "Different failures: FineGrainedFP8 accuracy regression vs Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Different issues: Flash Attention 3 training degeneration vs NaNs in classification heads after init." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "One is a functional DSA-indexer bug; the other is only a docstring fix." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different registration fixes." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "Different bugs: FineGrainedFP8 accuracy issue vs Qwen3-VL-Moe failure in v5." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Different config bugs in different models: default multimodal configs vs GPTNeoX rotary_pct reload." + }, + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "Same model family, but different problems: wrong model-type warning vs text encoder weights not loading." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Unrelated: missing transformers.onnx module vs XLMRoberta Flash Attention 2 behavior." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Both are upgrade-related import issues, but not the same concrete missing symbol or failure path." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Different tokenizer bugs: wrong offset mapping vs BOS/EOS not added." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Both concern tokenization drift, but they describe different model-specific bugs and outputs." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "Unrelated: FP8 accuracy regression vs Qwen3_5 image-triggered infinite loop." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Different regressions: offline cache loading failure vs wrong tokenizer decoder type in v5." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Completely different topics: missing onnx import vs tokenizer spacing after added token." + }, + { + "left": "issue:44368", + "right": "issue:46305", + "accept": false, + "reason": "Different issues: tie_word_embeddings warning vs sliding-window mask off-by-one." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Different problems in different contexts: Flash Attention 2 support vs loading hang." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Unrelated failures: offline tokenizer cache vs quantized model save failure." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Different subsystems entirely: generation kwargs sync logic vs D-FINE denoising losses." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "Same family (sam3) but different bugs: tracker warning vs memory overflow in litetext." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Different regressions: custom model init breakage vs NaNs in classification heads." + }, + { + "left": "issue:43550", + "right": "issue:46132", + "accept": false, + "reason": "Different code paths: Bamba torch.compile/SDPA failure vs AttentionInterface.register behavior change." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Unrelated: transformers env CLI issue vs sentence-transformers import NameError." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "No meaningful overlap: TF32 settings warning vs DSA-indexer torch.split bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 45, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:22355", + "issue:29159", + "issue:30990", + "issue:33357", + "issue:36296", + "issue:39401", + "issue:41628", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44188", + "issue:44222", + "issue:44261", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44360", + "issue:44373", + "issue:44485", + "issue:44488", + "issue:44514", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45290", + "issue:45478", + "issue:45600", + "issue:46326" + ], + "soft_pairs": [ + "issue:44188|issue:44373", + "issue:44117|issue:44704", + "issue:44295|issue:44297", + "issue:30990|issue:33357", + "issue:44261|issue:44373", + "issue:42907|issue:44945", + "issue:43756|issue:44060", + "issue:41628|issue:45278", + "issue:42907|issue:43937", + "issue:43976|issue:44945", + "issue:39401|issue:44488", + "issue:43122|issue:44521", + "issue:43784|issue:45600", + "issue:44704|issue:45290", + "issue:42907|issue:44964", + "issue:43122|issue:46326", + "issue:39401|issue:44521", + "issue:44360|issue:44485", + "issue:44038|issue:44297", + "issue:43116|issue:45200", + "issue:44625|issue:44829", + "issue:44038|issue:44295", + "issue:43408|issue:43475", + "issue:42907|issue:43452", + "issue:43122|issue:44514", + "issue:44038|issue:44222", + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:22355|issue:29159", + "issue:42757|issue:44351", + "issue:36296|issue:43116", + "issue:43278|issue:44945" + ], + "prepared_review_unit_hash": "59344a7c77e1baf2993465226de7cb00ed408eb67c068d51d05f913125ef1547", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36393, + "estimated_input_tokens": 9099, + "estimated_eval_tokens": 18454 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate pairs are mostly superficial similarities across unrelated bugs. This set spans import errors, tokenizer behavior, multimodal chat templates, training, saving, and model-specific regressions, so no pair looks like the same underlying issue/change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Both mention position-related behavior, but one is a torch.compile attention-kernel divergence bug and the other is a docstring issue; not the same bug." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Both concern from_pretrained plumbing, but one is TOKENIZER_MAPPING_NAMES None handling and the other is cached_file kwargs passthrough; different code paths." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Both are about position_ids/tokenizer save behavior, but one is reading a registered buffer and the other is tokenizer_class metadata mismatch; distinct failures." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Different symptoms and subsystems: Sentence Transformers loading hang vs MacOS bus error with a CLIP model." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "LayerNorm epsilon precision mismatch is unrelated to a wrong docstring for position_ids." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Dequantized save_pretrained failure is unrelated to incorrect LLM output under pipeline parallelism." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Both involve model internals, but Smollm3 RoPE layer dropping and Qwen3-Next tied-weights warnings are different model-specific bugs." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but AutoImageProcessor import failure and many import errors after a major version upgrade are too broad and not the same issue." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Saving dequantized models and invalid GenerationConfig errors are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version compatibility breakage is unrelated to incorrect outputs under pipeline parallelism." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Tokenizer offset_mapping bug for Qwen3 is unrelated to a model loading failure for cjvt/sleng-bert." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Different tokenizer/template bugs: version-dependent tokenization changes vs multimodal assistant_masks all-zero output." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Importing sentence-transformers with latest transformers and Sam3LiteText auto_mappings CI breakage are different import/config issues." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and chat_template crashes on tool-call assistant messages are different bugs." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Model saving failure is unrelated to loading a specific multimodal model." + }, + { + "left": "issue:43122", + "right": "issue:46326", + "accept": false, + "reason": "Different tokenizer behavior regressions: version-to-version tokenization drift vs continue_final_message with mistral_common." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Qwen3 offset mapping and multimodal assistant mask generation are separate tokenizer/template problems." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to GLM-5 RoPE implementation concerns." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe bug and tokenizer.save_pretrained metadata mismatch affect different components." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Empty results in an example classification script and Gemma 4 mm_token_type_ids defaults are different issues." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 training degeneration are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe and position_ids buffer reading are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Model-type warning for sam3_video vs sam3_tracker is not the same as missing fpn_position_embeddings on Sam3VisionEncoderOutput." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Saving dequantized models and gguf_file failures in from_pretrained are different code paths." + }, + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Version-dependent tokenization differences are unrelated to batched apply_chat_template padding=False crashes." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Qwen3-VL-Moe loading issues are distinct from FP8 save_pretrained for moe." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import failures, but Qwen2_5_VLForConditionalGeneration missing and HybridCache missing are different symbols and regressions." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 save_pretrained behavior, but one is a saved-checkpoint regression and the other is a from_pretrained error; not clearly the same bug." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "A missing transformers.onnx module and tokenizer slow/fast inconsistency are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but is_offline_mode from huggingface_hub and HybridCache from transformers are different dependency issues." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor parallel training bug and empty multi-label classification example output are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype change during evaluate is unrelated to incorrect outputs under pipeline parallelism." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 46, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:33453", + "issue:38617", + "issue:39401", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42796", + "issue:43064", + "issue:43066", + "issue:43116", + "issue:43441", + "issue:43493", + "issue:43611", + "issue:43673", + "issue:43825", + "issue:43827", + "issue:43874", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44188", + "issue:44355", + "issue:44380", + "issue:44485", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:44938", + "issue:45072", + "issue:45083", + "issue:45278", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45563", + "issue:45588", + "issue:45753", + "issue:46002", + "issue:46077", + "issue:46231", + "issue:46328" + ], + "soft_pairs": [ + "issue:42175|issue:44938", + "issue:43493|issue:45753", + "issue:22355|issue:29127", + "issue:43992|issue:44704", + "issue:43066|issue:44779", + "issue:43937|issue:44485", + "issue:38617|issue:45278", + "issue:44188|issue:46328", + "issue:42673|issue:45072", + "issue:43611|issue:46002", + "issue:44355|issue:46231", + "issue:42371|issue:43874", + "issue:43116|issue:43493", + "issue:44038|issue:45478", + "issue:44625|issue:46002", + "issue:43976|issue:45446", + "issue:42222|issue:43441", + "issue:44075|issue:44936", + "issue:45357|issue:45478", + "issue:43116|issue:43825", + "issue:45083|issue:45446", + "issue:42673|issue:42796", + "issue:43064|issue:43673", + "issue:43937|issue:45563", + "issue:33453|issue:39401", + "issue:44380|issue:46077", + "issue:44117|issue:45588", + "issue:43827|issue:45446", + "issue:39401|issue:44779" + ], + "prepared_review_unit_hash": "fbcf486132484307c976f8a244ac7014896b73d96f5462ca4e614e50fb1ced09", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36051, + "estimated_input_tokens": 9013, + "estimated_eval_tokens": 18282 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly loosely similar regression reports, but none of the soft pairs look like the same underlying bug/change closely enough to merge. I rejected all candidate duplicate links.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a clear canonical for this set: the items span many unrelated bugs, regressions, and doc/API problems rather than one duplicate cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Both are import/loading failures, but one is a missing backend dependency and the other is a Python 3.14 startup failure; different causes and fixes." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "Both mention Qwen3.5, but one is a model discrepancy vs JAX and the other is an infinite-loop image bug; not the same defect." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Unrelated: missing `transformers.onnx` module import vs a layoutlmv3 validation error message issue." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Both involve config propagation/loading paths, but one is missing `embed_tokens.weight` in UMT5Encoder and the other is AutoProcessor kwargs not reaching `cached_file`." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but they affect different code paths and symptoms: decoder type vs incorrect Deepseek tokenization." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "Both concern GLM-5, but one is invalid GenerationConfig handling and the other is RoPE implementation discussion; not the same bug." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Both are import errors after version changes, but the former is a specific missing symbol and the latter is a broad upgrade breakage report." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "Different areas entirely: attention-kernel divergence under `torch.compile` vs a performance issue in `SequenceFeatureExtractor.pad`." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "Both are runtime issues, but one is VRAM leakage in `Qwen3ForCausalLM` and the other is dtype mismatches in CI inference for different models." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Base-model-prefix loading regression vs NaNs in classification heads; different failure modes and likely different fixes." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Compiled Python load errors are unrelated to a docs/API parameter-name mismatch in `set_training()`." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API settings warning is unrelated to a missing image-processor method breaking multimodal token counting." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script multi-label output bug is unrelated to the SigLIP2 implementation discrepancy." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5/Moe and Transformers v5, but one is a general bug report and the other is a specific from_pretrained error; not enough evidence of the same root cause." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "`num_labels` propagation in Qwen3.5 is unrelated to NaNs in classification heads." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version support breakage and a PyTorch version check bug are different compatibility issues." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Different model families and symptoms: vitpose model breakage vs Ministral FlashAttention failure." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer SGD argument handling is not the same as `trainer.evaluate()` failing after training; different code paths." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5 Moe regressions, but one is save_pretrained key serialization and the other is from_pretrained loading; distinct bugs." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example script behavior is unrelated to pipeline() task-support error messaging." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Different subsystems: qwen3_omni_moe feature-length helper vs flex_attention import/version-check logic." + }, + { + "left": "issue:42673", + "right": "issue:42796", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to Ministral tokenizer BPE marker handling." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state mismatch is unrelated to missing GenerationMixin cache during chunked prefill." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GenerationConfig validation and stale `num_return_sequences` warning are different generation issues." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression is not the same as Qwen3 wrong offset mapping; both are tokenizer-related but distinct." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "Both involve attention behavior, but GPT2 scaling under SDPA/FlashAttention is not the same as Gemma4 using causal masks for bidirectional attention." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping lookup failure is unrelated to a flash-attention crash on `s_aux=None`." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing removed pipeline() APIs is unrelated to a PyTorch version check bug." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer bugs, but one is offset mapping and the other is incorrect tokenization results in Deepseek; not the same issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 47, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:42490", + "issue:42617", + "issue:42673", + "issue:42831", + "issue:43066", + "issue:43122", + "issue:43408", + "issue:43450", + "issue:43673", + "issue:43742", + "issue:43906", + "issue:43937", + "issue:43986", + "issue:44117", + "issue:44220", + "issue:44261", + "issue:44291", + "issue:44410", + "issue:44466", + "issue:44484", + "issue:44496", + "issue:44829", + "issue:44877", + "issue:44933", + "issue:44977", + "issue:45071", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45362", + "issue:45405", + "issue:45446", + "issue:45478", + "issue:45561", + "issue:46002", + "issue:46178", + "issue:46209", + "issue:46231", + "issue:46298", + "issue:46328", + "issue:46512" + ], + "soft_pairs": [ + "issue:42673|issue:43906", + "issue:43673|issue:46328", + "issue:44291|issue:45071", + "issue:44466|issue:46002", + "issue:43450|issue:43986", + "issue:43937|issue:44977", + "issue:45405|issue:45446", + "issue:44410|issue:45310", + "issue:45341|issue:46328", + "issue:45341|issue:45561", + "issue:43122|issue:46298", + "issue:38617|issue:41628", + "issue:44117|issue:45081", + "issue:42222|issue:46178", + "issue:45071|issue:46512", + "issue:44829|issue:45216", + "issue:45362|issue:46328", + "issue:43408|issue:43742", + "issue:42617|issue:45341", + "issue:42490|issue:42673", + "issue:42371|issue:44261", + "issue:44877|issue:46231", + "issue:44484|issue:45310", + "issue:39401|issue:46209", + "issue:43937|issue:44496", + "issue:38617|issue:44220", + "issue:44933|issue:45278", + "issue:42371|issue:42831", + "issue:43066|issue:43122", + "issue:44410|issue:45478" + ], + "prepared_review_unit_hash": "25ba51af11ca9d0438db341ccc0ffae3e257e65d585c2c9b6feacb22bca606c2", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36439, + "estimated_input_tokens": 9110, + "estimated_eval_tokens": 18476 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: tokenizer, import/API, model-loading, training, and testing-utils reports that are not the same underlying bug. No soft-edge pair looks mergeable with high confidence; the clearest duplicate-like subcluster is the repeated Qwen3.5 MoE `from_pretrained` regression (45310/45478), though it is not one of the provided soft pairs.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45310 is the best canonical anchor for the only clear duplicate subcluster: it is the earlier, concrete report of the Qwen3.5 MoE `from_pretrained` error, and 45478 appears to be the later twin report with the same title.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45278 is the broadest triage anchor for the cluster as a whole because it captures the general import-regression theme after upgrading, even though it does not truly subsume the unrelated model- and test-specific bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "Both mention Qwen3 and threading/repro context, but one is a VRAM leak in dataloader threads and the other is an isolated reproduction of a different issue; not the same bug." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Chunked-prefill cache handling and feature-extractor padding performance are unrelated code paths and symptoms." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Both are v5-era regressions, but one is an init-empty-weights argument error and the other is a PreTrainedConfig type-checking break; different failures." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "One is tied-weights serialization/device-dependent state, the other is NaNs during checkout+init; not the same underlying defect." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Both involve video models, but one is batched video processor shape output and the other is AutoProcessor crashing without torchvision; distinct issues." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "GLM-5 invalid GenerationConfig and Qwen3.5 flash-attention generation failures are different model-specific bugs." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "A PEFT minimum-version bump and a PyTorch version guard for AuxRequest are separate dependency/version checks." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Qwen3Next missing projections is a model-architecture loading issue, while Qwen3.5 MoE `from_pretrained` is a different model family regression." + }, + { + "left": "issue:45341", + "right": "issue:46328", + "accept": false, + "reason": "Testing utility bug and feature-extractor padding performance bug are unrelated." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Both touch testing utils, but one is a generic bug report and the other is a specific xdist race on `captured_info.txt`; different problems." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Tokenizer behavior regression and noisy logging are unrelated." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols from different APIs and do not indicate the same bug." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer mapping fallback behavior and Mistral regex patching/backend_tokenizer crash are different tokenizer-loading issues." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "VitPose breakage and a tensorflow_text backend-mapping error are separate subsystem failures." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "One is a PreTrainedConfig type-checking regression, the other is stale `config.dtype` after `from_config(dtype=...)`; related theme, but not the same bug." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Flash-attention training degeneration and incorrect `save_pretrained` checkpoint serialization are different Qwen3.5-related problems." + }, + { + "left": "issue:45362", + "right": "issue:46328", + "accept": false, + "reason": "Chat/model crash for Qwen3.5-35B and a feature-extractor padding performance issue are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "SAM3 model-type warning and MobileLLM key error are different loading/configuration bugs." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 API guidance and MLA `rms_norm_eps` precision error are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Strict config loading for granite_speech and a documentation parameter-name mismatch are not the same issue." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "A missing import from `image_utils` may be part of broader upgrade fallout, but this is not the same concrete bug as the general 'many import errors' report." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "`max_shard_size` question and Qwen3.5 MoE `from_pretrained` failure are unrelated." + }, + { + "left": "issue:39401", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3 tokenizer offset mapping and FP8 inference failure on B200 are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 and unrecognized model/model_type config loading are different classes of loading errors." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for `layer_type_validation` and a feature-extractor helper bug are unrelated." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "General import errors after upgrade are broader than, and not identical to, the specific missing import from `image_utils`." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 settings API migration and FineGrainedFP8 accuracy regression are unrelated." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Wrong tokenizer decoder type and tokenization differences across versions are distinct tokenizer regressions." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3Next projection mismatch and Qwen3.5 MoE `from_pretrained` error are different model-loading failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 48, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "issue:38617", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42757", + "issue:43012", + "issue:43064", + "issue:43097", + "issue:43352", + "issue:43441", + "issue:43452", + "issue:43531", + "issue:43577", + "issue:43673", + "issue:43874", + "issue:43931", + "issue:43986", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44360", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44829", + "issue:44933", + "issue:44987", + "issue:45003", + "issue:45071", + "issue:45341", + "issue:45372", + "issue:45440", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45588", + "issue:45656", + "issue:45753", + "issue:46123", + "issue:46172", + "issue:46231", + "issue:46369" + ], + "soft_pairs": [ + "issue:42222|issue:45542", + "issue:44075|issue:45656", + "issue:42175|issue:42222", + "issue:44483|issue:45464", + "issue:43097|issue:44493", + "issue:43673|issue:45341", + "issue:43012|issue:43577", + "issue:43874|issue:44261", + "issue:44263|issue:44485", + "issue:41628|issue:45071", + "issue:45071|issue:45542", + "issue:27957|issue:43352", + "issue:44484|issue:45478", + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:43931|issue:44410", + "issue:43441|issue:44829", + "issue:43931|issue:44038", + "issue:43097|issue:43452", + "issue:45003|issue:45341", + "issue:43986|issue:45341", + "issue:45753|issue:46369", + "issue:44933|issue:46231", + "issue:43874|issue:44263", + "issue:43673|issue:46123", + "issue:43986|issue:44484", + "issue:44117|issue:44987", + "issue:44261|issue:44485", + "issue:43531|issue:45341", + "issue:45440|issue:46172", + "issue:43986|issue:45372", + "issue:43064|issue:44315", + "issue:38617|issue:42757", + "issue:43097|issue:45588" + ], + "prepared_review_unit_hash": "ccdb07e0fbc89448eaa3fb61eed4c3559a0155469f01c4e9376fc63ff8f503e6", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37602, + "estimated_input_tokens": 9401, + "estimated_eval_tokens": 19058 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose similarity cluster rather than a true duplicate set. The issues span unrelated import errors, model-specific loading bugs, optimizer/runtime regressions, and docs/test nits, so I rejected every soft edge as not the same underlying bug or change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43097 is the clearest representative of the set: it has a specific regression claim, clear version context, and broader downstream impact than the more model-specific or incidental reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43097 is the best single issue to anchor the cluster because it is well-scoped, user-facing, and described as a concrete v5 regression with some follow-up traffic.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "Both mention TensorFlow/backend behavior, but one is about broken vitpose models and the other about installing torch-only extras; not the same code-path bug." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Different optimizer problems: SGD args ignored vs deepspeed causing a double optimizer step." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "Package/backend installation issue versus model breakage; too broad and not the same failure." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both hit chat/completions, but one is a general API rejection and the other is a specific streaming inference failure on Qwen3.5-0.8B." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "One is a removed config method in v5.0.0; the other is unexpected position-id keys across many models. Different regressions." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Chunked-prefill cache handling and a testing_utils bug are unrelated." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Both involve dtype behavior, but one is a warning during bfloat16 compilation and the other is model components staying float32 when loading BLIP2." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Different GLM-family bugs: missing image-patch method vs missing rms_norm_eps in layernorm config." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "A torch.split return issue in a DSA indexer is not the same as a RoPE implementation bug." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "Importing AutoImageProcessor and PreTrainedConfig type checking are unrelated failures." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Type-checking breakage in config classes is unrelated to a TensorFlow backend installation problem." + }, + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "Both mention Flash Attention 2, but one is XLM-Roberta support and the other is NemotronHForCausalLM support; different model paths." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "Save-pretrained shard size behavior is unrelated to Qwen3.5 MoE from_pretrained loading." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different DeepSeek/GLM internals and different bugs; no shared code-path evidence." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to the deepseek-v4 dtype/transpose issue." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen-family loading issues, but one is a weight-shape mismatch and the other is missing projections in layer 0." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Different FlashAttention problems: one is model support in v5 RC, the other is degenerate training with flash_attention_3." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Different Qwen-VL/Qwen3-VL-MoE loading issues with no concrete overlap in failure mode." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Config method removal and gguf_file handling are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is not the same as a testing_utils bug." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Missing torchvision during AutoProcessor loading is unrelated to the testing_utils issue." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "An infinite loop on a specific image is not the same as a memory overflow in sam3-litetext." + }, + { + "left": "issue:44933", + "right": "issue:46231", + "accept": false, + "reason": "A bad import from image_utils and a docs/parameter-name mismatch are unrelated." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Different GLM bugs with different failing methods and symptoms." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "Generation cache missing in chunked_prefill and MaskGenerationPipeline dropping partial-batch results are different pipeline bugs." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "AutoProcessor without torchvision and max shard size in save_pretrained have no common underlying bug." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping returning None may be related in theme, but the reported failure modes are not clearly the same bug." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeekV3MoE divergence from remote impl and a deepseek-v4 comb dtype/transpose issue are different problems." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Missing torchvision for video AutoProcessor loading is unrelated to the ReasoningEffort import error in Gemma 4 processor loading." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Wrong optimizer states on nonzero ranks with FSDP2/PEFT is unrelated to Liger Kernel not being applied in model_init." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors from upstream packages, but they affect different missing symbols and different dependencies." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "Removed tie_embeddings method and flash_attention.py crashing on s_aux=None are unrelated failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 49, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:38617", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43502", + "issue:43550", + "issue:43824", + "issue:43825", + "issue:43873", + "issue:43906", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44077", + "issue:44117", + "issue:44162", + "issue:44222", + "issue:44351", + "issue:44368", + "issue:44380", + "issue:44410", + "issue:44484", + "issue:44655", + "issue:44704", + "issue:44936", + "issue:44938", + "issue:45103", + "issue:45375", + "issue:45478", + "issue:45479", + "issue:45584", + "issue:45712", + "issue:45753", + "issue:46050", + "issue:46132", + "issue:46172", + "issue:46178", + "issue:46298", + "issue:46365", + "issue:46421" + ], + "soft_pairs": [ + "issue:43097|issue:46050", + "issue:43502|issue:43992", + "issue:44410|issue:45375", + "issue:44038|issue:44410", + "issue:38617|issue:44351", + "issue:44062|issue:44117", + "issue:42890|issue:46298", + "issue:43097|issue:43873", + "issue:44936|issue:46365", + "issue:44704|issue:45103", + "issue:43986|issue:44117", + "issue:44380|issue:46421", + "issue:43824|issue:43986", + "issue:44938|issue:46178", + "issue:43992|issue:44484", + "issue:43208|issue:43550", + "issue:43066|issue:43906", + "issue:42907|issue:44655", + "issue:43012|issue:44368", + "issue:43012|issue:43550", + "issue:45712|issue:46298", + "issue:42907|issue:45753", + "issue:43825|issue:44655", + "issue:43097|issue:46298", + "issue:43352|issue:44380", + "issue:43352|issue:43976", + "issue:44162|issue:46172", + "issue:45584|issue:46365", + "issue:44380|issue:46132", + "issue:44077|issue:46365", + "issue:44380|issue:45479", + "issue:43064|issue:44380", + "issue:44222|issue:45478" + ], + "prepared_review_unit_hash": "49a1ec11dd89819d7c46d27d9edc917c4679d70c492407559534950f4a8ceb82", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36859, + "estimated_input_tokens": 9215, + "estimated_eval_tokens": 18686 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous: most pairs are only loosely similar by subsystem or model family, not the same underlying defect. I would not merge any soft edge as a duplicate based on the available evidence.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45478 is the most concrete, version-scoped runtime failure in the set; the rest are mostly distinct bugs or ancillary warnings, so it is the best single representative if one issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45478 is the most actionable and specific report here: it names a clear model family, a version boundary, and a failing from_pretrained path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Different bugs: a removed config flag warning versus a quantized model save/load regression." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Different failure modes: stray API requests with local_files_only versus a missing encoder weight on load." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Same broad Qwen family, but one is missing attention projections and the other is a config field being dropped; not the same bug." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen-related, but one is a generic v5/Qwen3-VL-Moe issue and the other is a specific missing-projection defect." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Separate import errors for different symbols; no shared code-path." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "Tokenizer construction conflict versus tokenizer mapping returning None; different root causes." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "Test flakiness from missing seeding is unrelated to a noisy logging message." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "A v5 removal warning is not the same as a quantization/offloading behavior issue." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "Trainer evaluation failure after training is unrelated to importing Trainer failing because of datasets." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "Different subsystems: cached_file kwargs forwarding versus auto_docstring annotation handling." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Video-model loading without torchvision is unrelated to tokenizer mapping assumptions." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "Backend attention scaling bug versus cache lazy-init recompilation issue; different code paths." + }, + { + "left": "issue:43824", + "right": "issue:43986", + "accept": false, + "reason": "Both are import/load problems, but they affect different objects and dependencies." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python version load failure and tensorflow_text backend validation error are unrelated." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing encoder weights versus max_shard_size semantics; not the same defect." + }, + { + "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:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type is not the same as an isolated reproduction report." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Save/load of dequantized models versus saving Pipeline objects; different object types and failure modes." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Both are warning-related, but they concern different config fields and different behaviors." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A precision warning is unrelated to a compile-time SDPA failure." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Repository dummy-class cleanup is unrelated to logging noise." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Model save/load failure and a specific-image infinite loop are not the same bug." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline task-support messaging and pipeline serialization are distinct issues." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed flag behavior and noisy logging do not share a concrete code-path problem." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Different attention-backend problems on different models; not mergeable as one fix." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Flash-Attention support on one model is unrelated to Python-version compatibility." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage and a DeepSeek tensor-transpose issue are separate model bugs." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper generation edge case is unrelated to importing Trainer." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "Attention scaling config ignored versus register() changing function behavior; different APIs." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "PatchTSMixer post_init validation and Trainer import failure have no shared root cause." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention backend scaling bug and degenerate classification loss are unrelated." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is not the same as attention scaling behavior under SDPA/FlashAttention." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both mention MoE/quantization context, but one is FP8 save_pretrained while the other is a Qwen3.5 from_pretrained error; not the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 50, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:36010", + "issue:36683", + "issue:38617", + "issue:42175", + "issue:42490", + "issue:42757", + "issue:42898", + "issue:42994", + "issue:43012", + "issue:43066", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43526", + "issue:43582", + "issue:43611", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43867", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44514", + "issue:44625", + "issue:44655", + "issue:44683", + "issue:44938", + "issue:45216", + "issue:45276", + "issue:45290", + "issue:45593", + "issue:45600", + "issue:45698", + "issue:46050", + "issue:46077", + "issue:46082", + "issue:46172", + "issue:46298", + "issue:46365" + ], + "soft_pairs": [ + "issue:36683|issue:43582", + "issue:42175|issue:46365", + "issue:42490|issue:43012", + "issue:38617|issue:43502", + "issue:43208|issue:43526", + "issue:43012|issue:44683", + "issue:44483|issue:44514", + "issue:42757|issue:43824", + "issue:45600|issue:46298", + "issue:45698|issue:46172", + "issue:44038|issue:44230", + "issue:43618|issue:44380", + "issue:43618|issue:43761", + "issue:44483|issue:45290", + "issue:43299|issue:43867", + "issue:42994|issue:44655", + "issue:42898|issue:43066", + "issue:44625|issue:46365", + "issue:36683|issue:43475", + "issue:43756|issue:44485", + "issue:45276|issue:46077", + "issue:36010|issue:42757", + "issue:43352|issue:43937", + "issue:43611|issue:44655", + "issue:44655|issue:46050", + "issue:36683|issue:44077", + "issue:43867|issue:45216", + "issue:43976|issue:44938", + "issue:44625|issue:46077", + "issue:43352|issue:45593", + "issue:46077|issue:46082" + ], + "prepared_review_unit_hash": "736308b63d9a42ca755562993f9e91399fb1bfd732380bc16df8071df455eff8", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36531, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a heterogeneous mix of unrelated issues; the soft-similarity links do not look like true duplicates, so I rejected all of them. The most representative issue is the high-activity save/load regression report.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45698 is the most active and broadly framed bug report in the set, with detailed repros and inbound references around a save_pretrained/from_pretrained regression.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is the best global representative because it has the strongest discussion signal and describes a concrete, reusable serialization/loading failure mode.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Different failures: Gemma3 config missing vocab_size vs Apple Silicon TypeError in caching_allocator_warmup." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "Both mention dependency/import problems, but one is torch backend packaging and the other is Trainer import conflict with datasets." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "Unrelated warnings: local path slash import warning vs bfloat16 compile warning." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Different code paths: missing configuration_utils symbol vs local_files_only still making API requests." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "Distinct model bugs in different components (xLSTM training issues vs BeitImageProcessorFast label reduction)." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "One is a PyTorch warning, the other is a flex_attention compile failure on newer torch." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Different surfaces: /v1/chat/completions request rejection vs Qwen2_5_VLProcessor batched chat template crash." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they are for different missing symbols from different modules." + }, + { + "left": "issue:45600", + "right": "issue:46298", + "accept": false, + "reason": "CI break from removed Sam3LiteText config refs is unrelated to a noisy logging message." + }, + { + "left": "issue:45698", + "right": "issue:46172", + "accept": false, + "reason": "Serialization/loading regression vs a DeepSeek-V4 dtype/transpose bug; not the same issue." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Same model family, but one is a general Qwen3-VL-Moe bug and the other is fp8 inference support." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIP attentions missing is unrelated to GPT2 attention scaling being ignored under SDPA/FlashAttention." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Related CLIP regressions, but different outputs and likely separate forward-path bugs." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "Chat-completions endpoint validation issue vs assistant tool-call template crash; different failure modes." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Both are load/save problems, but one is Qwen3.5 checkpoint regression and the other is a sorted-state_dict loading error." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model saving failure is not the same as Pipeline.save_pretrained failing." + }, + { + "left": "issue:42898", + "right": "issue:43066", + "accept": false, + "reason": "Tokenizer behavior change and wrong decoder type are both v5 tokenizer issues, but not the same bug." + }, + { + "left": "issue:44625", + "right": "issue:46365", + "accept": false, + "reason": "Qwen3.5 num_labels propagation bug is unrelated to the Trainer import conflict." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Different model-specific attribute errors in Gemma3 config and SAM 3 Video output classes." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Both mention RoPE, but one is a Smollm3 architecture mismatch and the other is a GLM-5 implementation issue." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Same family (Gemma4), but resize_token_embeddings and bidirectional attention are separate code paths." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Import errors in different modules: GenerationMixin from transformers.generation vs is_offline_mode from huggingface_hub." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Nemotron Flash Attention support and GLM-5 GenerationConfig validation are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Model loading with base_model_prefix is distinct from saving Pipeline objects." + }, + { + "left": "issue:44655", + "right": "issue:46050", + "accept": false, + "reason": "Both involve save_pretrained/from_pretrained, but one is Pipeline persistence and the other is quantized sinq round-trip." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "No meaningful overlap: Gemma3 config attribute error vs patchtsmixer post_init validation." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "Both are checkpoint/loading regressions, but they affect different mechanisms and models." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Separate Python-version compatibility problems for different releases." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Different Gemma4/Qwen3.5 bugs with no shared failing code path." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Nemotron Flash Attention support is unrelated to D-FINE auxiliary losses." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Different model/config issues: Gemma4 bidirectional attention vs LlamaConfig head_dim validation." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 51, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42890", + "issue:43010", + "issue:43064", + "issue:43065", + "issue:43502", + "issue:43525", + "issue:43575", + "issue:43618", + "issue:43723", + "issue:43867", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44075", + "issue:44077", + "issue:44112", + "issue:44222", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44387", + "issue:44492", + "issue:44556", + "issue:44625", + "issue:44855", + "issue:44857", + "issue:44938", + "issue:44964", + "issue:45071", + "issue:45125", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45600", + "issue:45850", + "issue:46202", + "issue:46285", + "issue:46365" + ], + "soft_pairs": [ + "issue:36683|issue:43525", + "issue:43618|issue:45216", + "issue:36683|issue:44625", + "issue:43992|issue:45310", + "issue:43064|issue:44492", + "issue:43010|issue:45278", + "issue:44075|issue:46365", + "issue:44355|issue:44855", + "issue:43867|issue:45375", + "issue:43867|issue:46202", + "issue:44938|issue:45071", + "issue:36683|issue:43064", + "issue:41669|issue:44246", + "issue:42890|issue:45561", + "issue:43010|issue:45850", + "issue:43992|issue:45478", + "issue:43867|issue:45357", + "issue:43010|issue:43976", + "issue:36296|issue:43575", + "issue:43618|issue:44077", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:43618|issue:44857", + "issue:44360|issue:45440", + "issue:43502|issue:44062", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:43723|issue:44556", + "issue:44387|issue:44556", + "issue:44222|issue:45125", + "issue:43010|issue:43065" + ], + "prepared_review_unit_hash": "077aa8ca16956a1997f861fb4a558384e1e59d43b904c3a602837026427254b0", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36821, + "estimated_input_tokens": 9206, + "estimated_eval_tokens": 18668 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a loose thematic cluster of unrelated Transformer regressions and CI issues. A few items share broad language around import/loading/version-upgrade problems, but none of the listed soft pairs look like true duplicates of the same concrete bug path.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 45278 is the broadest representative of the cluster\u2019s recurring theme: import/load regressions after upgrading Transformers. It is still not a true duplicate umbrella, but it is the most central summary issue among the set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45278 best fits as the global representative because it captures the widest version-upgrade regression surface and is the closest to the common thread across the issues, even though the cluster is not actually duplicate-heavy.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Different model configs and different missing attributes; both are config errors, but not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "One is about CLIPOutput attentions assignment, the other about Qwen3.5 save_pretrained checkpoint correctness; unrelated code paths." + }, + { + "left": "issue:36683", + "right": "issue:44625", + "accept": false, + "reason": "Both involve config propagation, but they affect different models and different fields." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Different model-loading failures for different architectures; no shared concrete bug." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "Trainer/FSDP/PEFT optimizer-state bug versus a typo in cache strategy text; unrelated." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Decorating cache update methods with no_grad is a specific runtime behavior bug; 45278 is a broad import-error regression report." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "SGD argument handling is unrelated to an import failure that disappears after uninstalling datasets." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both are import-related, but one is about compiled Python files and the other about a Python 3.13 torch.jit parsing issue." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "State-dict loading error versus a missing Qwen3.5 VisionConfig field; different failure mechanisms." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "Checkpoint loading sort-order bug versus Gemma pipeline roundtrip tokenization; not the same issue." + }, + { + "left": "issue:44938", + "right": "issue:45071", + "accept": false, + "reason": "Python 3.14 import failure and PretrainedConfig type-checking regression are distinct problems." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Different subsystems: config attribute loading versus training/optimizer state under FSDP2." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Both mention import performance, but one is about removing import * in models and the other is a general slow-import report." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "A flaky test from missing set_seed is not the same as an xdist race on captured_info.txt." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "One is a specific no_grad decorator request; the other is a vague broad complaint about v5 breaking many models." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Both concern model loading, but they are different models and different missing keys/errors." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "State-dict ordering/load failure is not the same as incorrect visual encoder keys being saved." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "No_grad decoration on cache update methods is unrelated to Python-version compatibility failures." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Both mention tensor parallelism, but one is a training bug and the other is an OOM during model load." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "CLIPOutput attentions regression and patchtsmixer post_init policy are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "AutoModel PyTorch dependency error and tokenizer-loading regression are different issues." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "CLIP attentions assignment and CLIP text-model structure inconsistency are related only at a broad CLIP level, not the same concrete bug." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "No connection between missing CLIP attentions and a float16 AMP crash in LwDetrImageLoss." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to a DeepSeekV3MoE implementation divergence." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Local-files-only network leakage and AddedToken keyword conflicts are different tokenizer/runtime bugs." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "A flaky integration test and a Phi-4 multimodal loading failure are unrelated." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Missing set_seed in a SamHQ test is not the same as a stale device override failure in GraniteSpeech CI." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Test seeding instability and auto_mappings referencing removed configs are different CI problems." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading in v5 and checkpoint reload compatibility after upgrading versions are both loading regressions, but not the same bug." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 CUDA memory regression and checkpoint reload failure are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained behavior and missing tensor-parallel plan metadata are different Qwen3.5 issues." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "Cache update no_grad semantics and a dummy Conv2d in Sam3PixelDecoder are unrelated code changes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 52, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:27957", + "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:43232", + "issue:43299", + "issue:43502", + "issue:43526", + "issue:43576", + "issue:43673", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43901", + "issue:44222", + "issue:44230", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44655", + "issue:44749", + "issue:44908", + "issue:44938", + "issue:44961", + "issue:45071", + "issue:45125", + "issue:45362", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:45850", + "issue:46132", + "issue:46143", + "issue:46144", + "issue:46209" + ], + "soft_pairs": [ + "issue:43576|issue:46132", + "issue:43502|issue:44485", + "issue:45125|issue:45478", + "issue:44355|issue:44749", + "issue:42548|issue:44351", + "issue:30333|issue:33453", + "issue:45125|issue:45375", + "issue:43010|issue:43232", + "issue:43502|issue:44222", + "issue:30990|issue:31795", + "issue:43299|issue:44556", + "issue:43502|issue:44961", + "issue:42831|issue:43874", + "issue:30333|issue:36296", + "issue:44355|issue:45071", + "issue:43827|issue:44961", + "issue:44355|issue:44655", + "issue:33453|issue:34689", + "issue:27957|issue:33453", + "issue:44315|issue:44360", + "issue:42757|issue:43576", + "issue:43576|issue:45071", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:43576|issue:44938", + "issue:44485|issue:44961", + "issue:44908|issue:44961", + "issue:44222|issue:44230", + "issue:46143|issue:46144", + "issue:36296|issue:43901", + "issue:41628|issue:42548", + "issue:36296|issue:43526", + "issue:44315|issue:46209", + "issue:43673|issue:44492", + "issue:45362|issue:45850" + ], + "prepared_review_unit_hash": "7d319c0a0acc624ca02e54c3f93f1459abb2b4f62d1d113b5edc16d9275d4682", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 37714, + "estimated_input_tokens": 9429, + "estimated_eval_tokens": 19114 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a true duplicate set: most items are unrelated regressions, import errors, docs issues, or model-specific bugs that only share Transformers/v5 similarity. No soft edge looks like the same underlying bug or change.", + "confidence": 0.77, + "canonical_issue_reason": "No single issue cleanly represents the whole cluster, but issue 43576 is the best anchor because it is a concrete, user-facing v5 regression with some discussion/inbound attention and broader reach than the model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43576 is the most generally representative and actionable item in the set; the others are mostly narrower, separate failures or docs problems.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Both touch v5-era behavior, but one is a broken env command and the other is AttentionInterface.register semantics; different code paths and symptoms." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "local_files_only network leakage is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5 MoE, but one is tensor-parallel plan support and the other is a from_pretrained loading failure; too different to merge as one bug." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled Python file errors and post-upgrade slowdown are different problems." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports for different symbols; likely separate export/version issues." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLflow reporting after failed training and tokenizer loading regression are unrelated." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3.5 MoE tensor-parallel support and missing vision config field are different defects." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Both involve generation/cache internals, but one is about no_grad decoration and the other about sync_gpus/update_model_kwargs behavior; not the same bug." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "local_files_only network calls and FP8 save_pretrained on MoE are separate issues." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Loading hang for Sentence Transformers is unrelated to documentation confusion in model.forward." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Both are version-regression complaints, but they hit different model families and failure modes." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No relation between offline-mode network requests and the unrelated 'racoon' issue." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "FineGrainedFP8 accuracy and a missing image-processor method are different problems." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow reporting and tensor-parallel training bugs are unrelated." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file loading errors and PretrainedConfig type-checking regressions are different issues." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline() and the unrelated 'racoon' issue do not match." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Both are loading/saving related at a high level, but compiled-file errors and save_pretrained failure are different code paths." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Both are loading regressions, but tokenizer loading and Llama 3.2 Vision model loading are distinct failures." + }, + { + "left": "issue:27957", + "right": "issue:33453", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to tokenizer loading regression." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel application and a DSA indexer ReLU discussion are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing huggingface_hub import and broken env command are different failures." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "A broken CLI env command and a PretrainedConfig typing regression are not the same bug." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI env breakage is unrelated to Qwen2.5-VL import errors." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and missing layer_type_validation import are separate issues." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "Unrelated placeholder issue versus Gemma-4 audio positional encoding bug." + }, + { + "left": "issue:43576", + "right": "issue:44938", + "accept": false, + "reason": "Broken env command and Python 3.14 load failure are different problems." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "GLM-5 RoPE implementation and 'racoon' are unrelated." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs bug and 'racoon' are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both are FP8/MoE-adjacent, but one is save_pretrained behavior and the other is inference support; not the same concrete fix." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Both concern RoFormer, but one is kwargs forwarding and the other is attention-interface integration; separate code-path problems." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor-parallel training bug and text-classification docs mention are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Different import symbols failing from transformers; likely separate export regressions." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor-parallel training and BeitImageProcessorFast reduce_labels behavior are unrelated." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger Kernel/model_init behavior and FP8 inference on B200 are different bugs." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Generation cache missing in v5 and a cache-strategy typo are not the same issue." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Both are broad v5 complaints, but one is a specific Qwen3.5 chat crash and the other is a generic 'too many models' report." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 53, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:31795", + "issue:33453", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43827", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44655", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45478", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46218", + "issue:46328" + ], + "soft_pairs": [ + "issue:44623|issue:45230", + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44393|issue:45071", + "issue:44273|issue:45003", + "issue:44393|issue:45753", + "issue:43576|issue:44556", + "issue:41669|issue:43673", + "issue:43525|issue:45850", + "issue:44230|issue:46095", + "issue:44556|issue:45092", + "issue:43572|issue:45850", + "issue:44492|issue:45850", + "issue:45278|issue:45850", + "issue:42371|issue:44492", + "issue:41669|issue:46218", + "issue:43408|issue:44315", + "issue:42890|issue:44908", + "issue:43352|issue:44393", + "issue:44246|issue:44273", + "issue:43576|issue:44908", + "issue:43976|issue:44393", + "issue:44295|issue:45230", + "issue:44297|issue:45230", + "issue:44908|issue:45071", + "issue:44704|issue:44908", + "issue:41669|issue:44351", + "issue:44492|issue:46328", + "issue:41669|issue:45341", + "issue:43976|issue:45850", + "issue:44393|issue:45478", + "issue:43976|issue:46218", + "issue:44077|issue:44861", + "issue:43827|issue:44908", + "issue:31795|issue:33453", + "issue:44315|issue:44829", + "issue:43576|issue:44273", + "issue:44360|issue:45468", + "issue:44655|issue:44861", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:43576|issue:46218" + ], + "prepared_review_unit_hash": "301aeb561b30cc1e517103e43297bc9079034010501c1fa4016145c25201d468", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 39694, + "estimated_input_tokens": 9924, + "estimated_eval_tokens": 20104 + }, + "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, with only a few weakly similar v5/import/lazy-loading complaints. I would not merge any of the soft edges; none are clearly the same underlying bug or change.", + "confidence": 0.87, + "canonical_issue_reason": "No single issue cleanly represents this set: the items span docs, model-specific bugs, import/loading regressions, and unrelated API breakages. If forced, issue:45850 is the broadest umbrella for the v5-breakage reports, but it is too vague to be a true canonical duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 is the closest umbrella representative because it captures the broad v5 regression theme seen across several reports, but it is still not specific enough to serve as a reliable duplicate anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "45230 is only a generic bug report; there is no evidence it refers to the same save_pretrained/file-missing problem." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Both mention runtime behavior, but one is lazy-loading/import behavior and the other is a TensorFlow backend/package issue; different bugs." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Both involve Qwen3/MoE, but one is a model-support regression and the other is missing tensor-parallel metadata; not the same defect." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "These are unrelated: a Qwen3-VL bbox-generation issue versus a PretrainedConfig type-checking regression." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Lazy loading/import behavior and unsafe sys.modules access are different code paths and different failure modes." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both are Qwen-related, but one is 2D bbox output error and the other is an infinite-loop case on a specific image; not the same bug." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Both are v5 upgrade regressions, but one is the env command and the other is checkpoint reload compatibility; different concrete problems." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-time slowdown from import * removal and missing GenerationMixin cache during chunked prefill are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "45850 is a vague umbrella complaint; it does not match the specific Llama4Config pad_token_id AttributeError." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 Qwen3-VL/MoE support and a DeepSeek weight-conversion question are different topics entirely." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both concern upgrade/reload compatibility, but one is general checkpoint reload after v4.57 and the other is remote-code/meta-init incompatibility for InternVL2; not clearly the same bug." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "The StableLm pad_token_idx regression is too specific to be the same as the broad, vague v5-breakage complaint." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "A typo in cache strategies is not the same underlying issue as a generic report that v5 breaks many models." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "Both are broad v5 regression reports, but one is specifically about import errors after upgrade and the other is a vague general complaint; not the same concrete bug." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API guidance and a cache-strategy typo are unrelated documentation/API issues." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import slowness from model imports and failure to import Qwen3_5ForCausalLM are distinct issues." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "A wrong model-type warning and Liger Kernel not being applied with model_init are separate behaviors." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Missing set_seed in tests and inverse_sqrt scheduler kwargs handling are unrelated test/stability and scheduler issues." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention 2.0 support for Nemotron and Qwen3-VL bbox hallucination are different model-specific problems." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Both touch import/lazy-loading performance, but one is a general intermittent import slowness report and the other is a specific lazy-loading bug; insufficient evidence they are the same defect." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "The env command regression and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility and Qwen3-VL bbox output are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "A position_ids buffer access error is not tied to the generic bug report." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch is a different issue from the generic bug report." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and inverse_sqrt scheduler kwargs are different APIs and different bugs." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Model import performance and missing HybridCache import are unrelated import-path issues." + }, + { + "left": "issue:44492", + "right": "issue:46328", + "accept": false, + "reason": "Cache-strategy typo and feature-extractor pad performance are not the same code-path problem." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import performance regression and a testing_utils bug do not match." + }, + { + "left": "issue:43976", + "right": "issue:45850", + "accept": false, + "reason": "Python support issues are not the same as a broad claim that v5 breaks too many models." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "They are both Qwen-related, but one is bbox output hallucination and the other is a Qwen3.5 MoE from_pretrained error; not the same underlying bug." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python compatibility and missing Qwen3_5ForCausalLM import are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init enforcement for patchtsmixer and _get_tied_weight_keys crashing on list are distinct implementation issues." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Pipeline docs referencing removed pipeline() and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion around forward arguments and tokenizer loading regression are different bugs." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger Kernel application with model_init and degenerate training under flash_attention_3 are different failure modes." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "The env command regression and lazy-loading issue are not the same concrete problem." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "A ReLU discussion and Gemma-4 audio positional encoding are unrelated." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline save_pretrained support and tied-weight key handling are separate save/serialization issues." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "A generic bug report does not establish the same Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading/import behavior and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command and missing Qwen3_5ForCausalLM import are different issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 54, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-463", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:37934", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:42936", + "issue:43066", + "issue:43240", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43606", + "issue:43618", + "issue:43668", + "issue:43704", + "issue:43717", + "issue:43813", + "issue:44031", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44351", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44861", + "issue:44908", + "issue:44938", + "issue:44998", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46144", + "issue:46172", + "issue:46218" + ], + "soft_pairs": [ + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:46018|issue:46172", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:31515|issue:31795", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:42371|issue:43704", + "issue:45850|issue:46218", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:43298|issue:46095", + "issue:43519|issue:43704", + "issue:44263|issue:45468", + "issue:43606|issue:45468", + "issue:44261|issue:45468", + "issue:44351|issue:46218", + "issue:43352|issue:46218", + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44492|issue:44998", + "issue:43618|issue:44998", + "issue:44998|issue:45850", + "issue:44938|issue:44998", + "issue:44861|issue:45290", + "issue:43525|issue:44861", + "issue:31515|issue:38617", + "issue:43572|issue:44861", + "issue:44230|issue:44623", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230", + "issue:43717|issue:46144", + "issue:42936|issue:46018", + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "c01201a6112e71ed5363463a43e39a92e7ad5ca503ee423b41b207f042350fcc", + "component_budget": { + "node_count": 463, + "item_count": 463, + "soft_pair_count": 1704, + "serialized_chars": 883384, + "estimated_input_tokens": 220846, + "estimated_eval_tokens": 441948 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 47, + "serialized_chars": 40307, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a loose grab-bag of unrelated issues, with one dominant theme around Transformers v5/model-loading regressions and tokenizer/config breakages. No soft-edge pair looks like the same concrete bug, so none are merged.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45850 is the broadest umbrella report in the cluster and best matches the dominant v5 regression / model-breakage theme.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 is the best single representative because it summarizes the widespread version-5 breakage pattern seen across several other issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Both are model-specific correctness bugs, but one is Qwen3VL timestamp calculation and the other is GLM-5 RoPE implementation; different code paths and symptoms." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Qwen3VL timestamp math and inverse_sqrt scheduler kwargs are unrelated subsystems." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek-V4 shared-expert gating and a comb.transpose dtype issue are different DeepSeek problems with different fixes." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star cleanup/performance issue is unrelated to a Qwen3ForCausalLM VRAM leak in dataloader threads." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion for forward args is not the same as a model-loading regression for Llama 3.2 Vision." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Slow checkpoint loading and forward-API documentation confusion are different issue types and different code paths." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in Qwen3 threading and GLM-5 RoPE implementation are unrelated bugs." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "GLM-5 RoPE math and Gemma-4 audio positional encoding are different model implementations." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 settings guidance and a Qwen3 memory leak do not share the same bug." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "A broad v5 breakage meta-issue is not the same as a Qwen3.5 import failure." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "local_files_only network requests and Qwen3VL timestamp calculation are unrelated." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Timestamp calculation in a processor and missing save_pretrained files are different defects." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Audio continuous batching support and DeepSeek-V4 weight conversion are different feature requests/bugs." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Qwen3VL timestamp math and Qwen3 threading VRAM leak are distinct model/runtime issues." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return handling in one model is unrelated to Gemma-4 audio positional encoding." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch for bark-small is a different failure mode than Gemma-4 audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "Missing rms_norm_eps precision issue in MLA q_a_layernorm is not the same as Gemma-4 audio positional encoding." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "Missing HybridCache import and missing Qwen3_5ForCausalLM import are different missing-symbol regressions." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention 2 support for NemotronH and Qwen3.5 import failure are unrelated." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Unexpected tokenizer behavior differences and wrong tokenizer decoder type are both tokenizer-adjacent, but not the same bug." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs propagation and the unrelated 'Unemployment' issue are not duplicates." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache-strategy typo and a nonsensical unrelated issue title do not match." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attentions regression and the unrelated 'Unemployment' issue are not the same bug." + }, + { + "left": "issue:44998", + "right": "issue:45850", + "accept": false, + "reason": "A spam/off-topic issue is not the same as a broad v5 breakage report." + }, + { + "left": "issue:44938", + "right": "issue:44998", + "accept": false, + "reason": "Python 3.14 load failure and the unrelated 'Unemployment' issue are unrelated." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "_get_tied_weight_keys crash and apply_chat_template tool-call crash affect different code paths." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Llama4 pad_token_id AttributeError and tied-weight-key handling are different configuration/model-loading bugs." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading and missing layer_type_validation import are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "StableLmConfig missing pad_token_idx and tied-weight-key crash are unrelated config issues." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 Qwen3-VL/MoE support and processor.save_pretrained missing files are different topics." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLFlow reporting on failed training jobs and slow checkpoint loading are unrelated." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Slow checkpoint loading and tensor-parallel training bug are different defects." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 support request and Qwen3.5 Moe from_pretrained error are different issues." + }, + { + "left": "issue:31515", + "right": "issue:33453", + "accept": false, + "reason": "Slow checkpoint loading and tokenizer loading regression are different problems." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching for audio models and FP8 inference support are unrelated feature gaps." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init_weights behavior and RoFormer attention interface usage are unrelated model implementations." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2Former ignore_value misuse and DeepSeek-V4 shared expert gating are different model bugs." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "ModernBERT config type hint and tokenizer regex warning are not the same bug." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Mask2Former ignore_value bug and missing kwargs in loss calculation are unrelated." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss-function kwargs propagation and RoFormer attention interface usage are different code paths." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former ignore_value and tokenizer regex warnings are unrelated." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "Mask2Former ignore_value bug and RoFormer attention interface issue are different models and fixes." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2Former ignore_value and a quantization_operation typo are unrelated." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Loss kwargs propagation and ModernBERT norm_eps type hint are unrelated." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "ModernBERT type-hint mismatch and quantization typo are unrelated." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Llama4TextL2Norm/RMS norm question and a ModernBERT type hint issue are unrelated." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former ignore_value and ModernBERT norm_eps typing are different bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10519-3", + "nodes": [ + "issue:10519", + "issue:11005", + "issue:11307" + ], + "soft_pairs": [ + "issue:10519|issue:11307", + "issue:10519|issue:11005", + "issue:11005|issue:11307" + ], + "prepared_review_unit_hash": "3ec353ecfd76e7191747003ad58b33c0e441f043e14593c3e79c78521b1323df", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues are not duplicates: they cover unrelated topics across tokenizer truncation, learning-rate scheduling, and Wav2Vec2 word timestamps. No single issue serves as a meaningful cluster canonical.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the issues describe different feature requests in different areas and do not share the same underlying problem.", + "canonical_pr_reason": null, + "best_issue_reason": "None of the issues is a good cluster representative; if forced, 10519 is the most self-contained, but it is still unrelated to the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Different subjects: tokenizer truncation behavior vs. Wav2Vec2 word boundary timestamps." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Unrelated requests: text truncation option vs. ReduceLROnPlateau-like scheduler functionality." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "No common underlying bug or change; one is about LR scheduling, the other about audio word offsets." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 3, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-13244-3", + "nodes": [ + "issue:13244", + "issue:15980", + "issue:16618" + ], + "soft_pairs": [ + "issue:13244|issue:16618", + "issue:15980|issue:16618" + ], + "prepared_review_unit_hash": "7ef7defb32fd41c2cc5f9c3ecd5f2c85ab87db28085daf81b5e84c01ea1da200", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2477, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2477, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues are not duplicates: one is about Tapas tokenization differences, one is about a misleading private-model auth error, and one is about failing to load bert-base-uncased. The only commonality is vague model/tokenization/HF-loading similarity, not the same underlying bug.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue clearly represents a shared bug here; the cluster appears to be a loose similarity grouping rather than true duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #16618 is the most general of the three and the only one directly about model loading, but it still does not cover the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization mismatch is a different problem from a model-loading failure for bert-base-uncased; they affect different code paths and symptoms." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "A bad private-model login error and a failure to load a public model are distinct issues with different root causes and user-facing behavior." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 4, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-26350-2", + "nodes": [ + "issue:26350", + "issue:33260" + ], + "soft_pairs": [ + "issue:26350|issue:33260" + ], + "prepared_review_unit_hash": "fcba337161f68748bd79fd7701cf80c90e5ae9e5d538ceb6fe8b66b7a749b5b4", + "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 are two distinct feature requests: one for Flash Attention 2 architecture support and one for GGUF architecture support. They share only a broad \u201ccommunity contribution\u201d framing, not the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue #33260 is the cleaner standalone request: it is more narrowly scoped and has more inbound references, so it is the better representative issue in this cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different features and code paths (Flash Attention 2 vs GGUF support); 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": 5, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-8", + "nodes": [ + "issue:27088", + "issue:28936", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370", + "issue:46222", + "issue:46226" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:27088|issue:38435", + "issue:27088|issue:28936", + "issue:28936|issue:43834", + "issue:28936|issue:44365", + "issue:28936|issue:38435", + "issue:28936|issue:44370", + "issue:38435|issue:44370", + "issue:38435|issue:44365", + "issue:38435|issue:43834", + "issue:38435|issue:46222", + "issue:38435|issue:46226", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:43834|issue:46222", + "issue:43834|issue:46226", + "issue:44365|issue:44370", + "issue:44365|issue:46222", + "issue:44365|issue:46226", + "issue:44370|issue:46222", + "issue:44370|issue:46226", + "issue:46222|issue:46226" + ], + "prepared_review_unit_hash": "6466810d1b85a4e72b8732a1bb3b40f6adffb2cb18904d37de8ff98180b60390", + "component_budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13531, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13531, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Mostly a set of separate i18n translation requests by language, not one true duplicate cluster. The only clear duplicate subgroups are the three identical placeholder-template issues and the two identical Thai-titled issues.", + "confidence": 0.94, + "canonical_issue_reason": "issue:43834 is the clearest representative of the duplicate placeholder-template subgroup; issue:44365 and issue:44370 match it closely, while the language-specific requests are distinct and should not be collapsed into the same issue.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43834 is the best overall issue to keep as the canonical representative because it is the most generic/central of the exact-duplicate subgroup and has the broadest overlap with the other template-like entries.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Different language requests; same translation theme but not the same underlying issue." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Different language requests; not a duplicate bug/change." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Turkish translation request vs generic template issue; related topic, but not the same issue." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Different language request (Turkish vs Romanian); not a duplicate." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Different language request (Turkish vs Spanish); not the same underlying issue." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish translation request is distinct from the generic template issue." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Different language request; only broadly similar by template." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Different language request (Spanish vs Romanian); not duplicates." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Different language request; same theme, different issue." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian translation request vs generic template issue; related but not the same issue." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian translation request vs generic template issue; not a duplicate." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian request is distinct from the generic placeholder issue." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "Different language/request content; not the same underlying issue." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "Different language/request content; not duplicates." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": true, + "reason": "Same placeholder title/body pattern; looks like the same template issue duplicated." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": true, + "reason": "Same generic template issue with identical title/body structure; likely duplicate." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-template issue vs unrelated Thai-titled issue; not the same content." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-template issue vs unrelated Thai-titled issue; not a duplicate." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": true, + "reason": "Identical placeholder-template issues; same underlying issue." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "Generic placeholder issue vs Thai-titled issue; similar only at a high level." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Generic placeholder issue vs Thai-titled issue; not the same issue." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "Generic placeholder issue vs Thai-titled issue; not a duplicate." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "Generic placeholder issue vs Thai-titled issue; not the same underlying issue." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Same exact Thai title/body pattern; these appear to be duplicate issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet: it separates the language-specific requests, and only groups the near-identical placeholder-template pairings and the Thai pair. The soft-edge accepts are conservative enough given the very high textual similarity and matching titles." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-16", + "nodes": [ + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36879|issue:41211", + "issue:36879|issue:43909", + "issue:36879|issue:44121", + "issue:36879|issue:43837", + "issue:36879|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:36879|issue:37015", + "issue:37015|issue:42503", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42738", + "issue:42738|issue:45027", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue: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:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133" + ], + "prepared_review_unit_hash": "08ee66004d9577990c1ac2896cab2a23192adc6315199d2f9a6f51eafff5836a", + "component_budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21663, + "estimated_input_tokens": 5416, + "estimated_eval_tokens": 11088 + }, + "budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21663, + "estimated_input_tokens": 5416, + "estimated_eval_tokens": 11088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a loose pile of unrelated model-request issues that only resemble each other at a very broad \u201cadd model support\u201d level. There are no convincing duplicates here, so all soft edges should be rejected.", + "confidence": 0.88, + "canonical_issue_reason": "No true duplicate cluster exists; if one issue must represent the set, 46133 is the most active and detailed generic model-request thread among otherwise heterogeneous requests.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46133 is the best triage representative because it has the strongest discussion/inbound activity, but it still does not unify the other requests as duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "Different model requests (EuroBert vs DEIMv2); same broad theme, not the same change." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config support vs Qwen3-TTS support are different models and code paths." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert config vs RF-DETR support are unrelated model additions." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are distinct models with different integration work." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "Object-detection model request vs audio model request; not the same underlying bug/change." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR support and weight-sparse transformer support are unrelated." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR vs Qwen3-ASR are different model families and implementations." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR and Qwen3-TTS are unrelated model support requests." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different model additions." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT support and sparse-transformer support are distinct changes." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "BERT-style model request vs audio model request; not duplicates." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are separate model requests." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are different models, even if both are BERT-like." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture vs Qwen3-ASR support are unrelated." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 architecture work is not the same as sparse-transformer support." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 and ModernVBERT are distinct model requests." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 architecture vs Voxtral TTS support are different code paths." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT models and Qwen3-ASR support are different additions." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT and Qwen3-TTS are separate model families." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT support and a generic BERT-with-RoPE request are not the same concrete change." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like models with RoPE vs Voxtral TTS support are unrelated." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are related by family name but are distinct audio modalities and separate support work." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral TTS are different model integrations." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "TTS support vs OCR model support are unrelated changes." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support is not the same as sparse-transformer support." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "TTS and audio-model request are different model-specific integrations." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are separate audio models and requests." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "ASR support and EUPE vision/perception encoder support are unrelated." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR and sparse-transformer support are different code-path changes." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are unrelated model requests." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer support and EUPE support are not the same underlying change." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral-4B-TTS are different speech-model integrations." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS are unrelated model additions." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "ASR support and DeepSeek-OCR2 support are different model requests." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 7, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-38740-2", + "nodes": [ + "issue:38740", + "issue:46443" + ], + "soft_pairs": [ + "issue:38740|issue:46443" + ], + "prepared_review_unit_hash": "36daff859e69e3fbee1771f7029c1e67aee48229f6387d9dc936e5f16d55f3ce", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1504, + "estimated_input_tokens": 376, + "estimated_eval_tokens": 1008 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1504, + "estimated_input_tokens": 376, + "estimated_eval_tokens": 1008 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are unrelated: one is a docs request for adding Pruna as an optimization framework, and the other is a Vision RoPE refactor. No duplicate/merge relationship is supported.", + "confidence": 0.98, + "canonical_issue_reason": "Issue 46443 is the better canonical representative because it is the substantive technical item; issue 38740 is a separate documentation request about a different topic.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46443 is the strongest issue in the set for canonical tracking since it describes an actual code refactor rather than a documentation addition.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Different topics and scopes: documentation about Pruna vs a Vision RoPE refactor. They do not describe the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 8, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41084-2", + "nodes": [ + "issue:41084", + "issue:43037" + ], + "soft_pairs": [ + "issue:41084|issue:43037" + ], + "prepared_review_unit_hash": "b109d2da563161a29894e068cf438e6bf170931298b1c7d903b240f244adecc8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1478, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1478, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are not duplicates: one requests Set Block Decoding support, while the other requests DeepSeek v3.2 support. They refer to different model/features with no shared targets or code-path evidence.", + "confidence": 0.98, + "canonical_issue_reason": "Issue #43037 is the more concrete, actionable request; #41084 is a broader/ambiguous decoding topic.", + "canonical_pr_reason": null, + "best_issue_reason": "#43037 is the best representative issue in this set because it is specific and clearly scoped to a single model-support request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different underlying requests: block decoding vs DeepSeek v3.2 support. No shared files, targets, or evidence of the same bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 9, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43598-2", + "nodes": [ + "issue:43598", + "issue:43600" + ], + "soft_pairs": [ + "issue:43598|issue:43600" + ], + "prepared_review_unit_hash": "3f7d0ebd870fcf6b5ffa56130a244e9f9b95dbe16d8f89da5e16e2295e806067", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues both mention custom trainers, but they target different Trainer behaviors: one is about `remove_unused_column` customizability, the other about FLOPs estimation. They do not look like the same underlying bug.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue 43598 is the more concrete and actionable report, but it addresses a different code path than FLOPs estimation, so there is no true canonical duplicate here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Shared mention of custom trainers is too broad; one issue concerns `remove_unused_column` behavior while the other concerns FLOPs counting, so they are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43684-2", + "nodes": [ + "issue:43684", + "issue:44405" + ], + "soft_pairs": [ + "issue:43684|issue:44405" + ], + "prepared_review_unit_hash": "c6f52a94191bbea14f6d73945f562802be2f94f512a4ffe92144a609cc91eb84", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related at a high level (both ask for Qwen family AutoModel support), but they target different model/config combinations and different auto classes, so they should not be merged as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43684 is the broader, more foundational request: register Qwen3-Omni with AutoModel/AutoModelForConditionalGeneration, which is the clearer canonical tracking item for auto-class registration work.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43684 is the best representative of the cluster because it is the more general auto-registration gap, while 44405 is a separate, narrower request for sequence-classification support on Qwen3.5.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying changes: one is Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration, the other is Qwen3.5 AutoModelForSequenceClassification support. They involve different model families/configs and different auto heads, so they are not the same bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 11, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43835-4", + "nodes": [ + "issue:43835", + "issue:43845", + "issue:43846", + "issue:43859" + ], + "soft_pairs": [ + "issue:43835|issue:43845", + "issue:43835|issue:43846", + "issue:43835|issue:43859", + "issue:43845|issue:43846", + "issue:43845|issue:43859", + "issue:43846|issue:43859" + ], + "prepared_review_unit_hash": "c0a00ce36dfba26aaf9ef10c63c055ebfb5ff6397b71c29516c00218ca479d05", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4266, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4266, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The three \"huggingface\" issues are likely duplicates of one another based on identical titles and high similarity, but the evidence is limited to metadata. \"Fett\" looks separate. I\u2019d anchor the cluster on the earliest huggingface issue.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43845 is the earliest of the repeated \"huggingface\" issues and is the best anchor for the likely duplicate set, given the identical titles and high similarity.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43845 is the cleanest representative of the likely duplicate cluster because it is earliest and shares the exact repeated title with the other huggingface issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different title and no concrete content overlap; soft similarity alone is not enough to treat them as duplicates." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different title and no concrete evidence of the same underlying report." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different title and only weak similarity evidence; not enough to accept as duplicates." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, though the packet does not expose the underlying issue text." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, with no evidence they are different reports." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Likely duplicate based on identical title and high similarity, but the available evidence is limited to metadata." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster anchor is reasonable, but several soft-edge justifications overstate the evidence. The packet only gives identical titles, body lengths, and soft-similarity scores; it does not provide concrete issue content showing the reports are the same. Phrases like \"clearly overlapping issue report\" are too strong here. Please keep the duplicate verdicts more conservative, e.g. \"likely duplicate based on identical title and high similarity.\"" + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 12, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44371-5", + "nodes": [ + "issue:44371", + "issue:44418", + "issue:44572", + "issue:44573", + "issue:44937" + ], + "soft_pairs": [ + "issue:44371|issue:44572", + "issue:44371|issue:44573", + "issue:44371|issue:44418", + "issue:44371|issue:44937", + "issue:44418|issue:44572", + "issue:44418|issue:44573", + "issue:44572|issue:44573", + "issue:44572|issue:44937", + "issue:44573|issue:44937" + ], + "prepared_review_unit_hash": "6e65ec30bc4253951e92f3965429250922f6ba2c2e9eb38cd178d4ddc7acffc0", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5785, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5785, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is dominated by low-information spam/placeholder issues. The only exact duplicate-looking items are the two \"\" issues, but there is still no evidence they describe the same underlying bug or change beyond identical placeholder titles, so I would not merge any pair on that basis alone.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44371 is the earliest exact \"\" placeholder entry and is as representative as any other low-information spam issue in this cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "44371 is the cleanest canonical representative because it is an exact placeholder title and the earliest of the identical spam-like issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are titled \"\", but that is not enough to establish the same underlying bug or change; these look like separate spam placeholders." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Identical placeholder titles alone are insufficient to confirm duplication of the same concrete problem." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "One is generic spam and the other is a vague documentation suggestion; no concrete shared issue is visible." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "The titles are unrelated; there is no indication they describe the same bug or requested change." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "A vague docs suggestion and a spam placeholder do not appear to be the same underlying item." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "A documentation enhancement request is not the same as an unrelated spam placeholder." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Although both are titled \"\", there is no evidence they refer to the same underlying report rather than two separate spam entries." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "The titles do not match and there is no shared concrete problem or change target." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "No meaningful overlap beyond being closed issues; they do not look like duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45145-2", + "nodes": [ + "issue:45145", + "issue:45151" + ], + "soft_pairs": [ + "issue:45145|issue:45151" + ], + "prepared_review_unit_hash": "ccc52389b5c8a5d1d1f9864401cbd918f5aa7bc7727bea700c4a80a3886c7a12", + "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 two issues have the same title and nearly identical metadata, so they are very likely duplicate reports of the same energy-optimization proposal. Issue #45145 is the better canonical issue because it was created first and has an inbound reference.", + "confidence": 0.98, + "canonical_issue_reason": "#45145 was opened first and appears to be the original report; #45151 is effectively the same issue with no distinguishing links or files.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the strongest representative because it is earlier and slightly better connected (it has an inbound reference).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "Same title, same topic, and no evidence of distinct scope; these look like duplicate copies of the same issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The conclusion is grounded in the packet: the issues have identical titles, close creation times, no distinct links/files, and one has only a slight inbound-reference advantage. The duplicate/ canonical-candidate call is conservative enough given the limited evidence, and the soft-edge accept does not overstate beyond the available metadata." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45295-2", + "nodes": [ + "issue:45295", + "issue:45373" + ], + "soft_pairs": [], + "prepared_review_unit_hash": "f5fd95fada2f530ab0999e343c17c16266cde0c7182bb22509972ca44827cd49", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues describe the same missing Gemma 4 sequence-classification support; the later, more specific open issue is the better cluster representative.", + "confidence": 0.94, + "canonical_issue_reason": "issue #45373 is the clearest canonical tracker: it is open, explicitly names the missing class `Gemma4ForSequenceClassification`, and states the module-level gap relative to Gemma 2/3.", + "canonical_pr_reason": null, + "best_issue_reason": "issue #45373 best captures the concrete request and likely remaining work, while #45295 is a shorter earlier duplicate with less detail and is already closed.", + "best_pr_reason": null, + "soft_edge_verdicts": [] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45672-2", + "nodes": [ + "issue:45672", + "issue:45674" + ], + "soft_pairs": [ + "issue:45672|issue:45674" + ], + "prepared_review_unit_hash": "7a3b6b2ef07228a7b3922690a274594ce9fbae157b9f41cb05b1f97e9c3a2bae", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1672, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate 4-bit inference bugs in the same general area, but they do not look like duplicates: one is a Gemma4 audio-module uint8/torch.finfo TypeError, the other is a BitsAndBytesConfig skip-modules regression that drops the default lm_head exclusion and triggers an AssertionError.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue is evident here because the issues describe different failure modes and different root causes.", + "canonical_pr_reason": null, + "best_issue_reason": "issue 45674 is the clearest representative of the cluster\u2019s broader 4-bit inference/BitsAndBytes regression theme, with a more specific and actionable root-cause description.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "The reports concern different code paths and symptoms: Gemma4 audio modules failing in torch.finfo on uint8 weights versus BitsAndBytesConfig clearing the default lm_head exclusion and asserting in 4-bit inference. Similar subsystem, but not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 16, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45762-2", + "nodes": [ + "issue:45762", + "issue:45763" + ], + "soft_pairs": [ + "issue:45762|issue:45763" + ], + "prepared_review_unit_hash": "cc53002243c7fafdf54318f45149fca1c891d85abf4cfade38755b4481ed1546", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are bare-name/date issues with no body or technical context, so they do not appear to describe the same bug or change and are not good canonical triage candidates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "Titles are different personal-name entries with different dates and no shared bug report content; likely unrelated/spam rather than the same underlying issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 17, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45764-2", + "nodes": [ + "issue:45764", + "issue:45765" + ], + "soft_pairs": [ + "issue:45764|issue:45765" + ], + "prepared_review_unit_hash": "9c051256cc1dca01ffc7e8a478879cf42bd3fc9cf8e3c9582ae8054c381e477a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues do not look like duplicates; they appear to be unrelated name/date entries rather than the same bug or request. No PRs are present.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a plausible canonical representative here; the items do not share a clear underlying problem or change.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "The titles are unrelated personal-name/date entries with no shared subject, code path, or explicit linkage; this does not look like the same underlying issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 18, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45810-2", + "nodes": [ + "issue:45810", + "issue:46295" + ], + "soft_pairs": [ + "issue:45810|issue:46295" + ], + "prepared_review_unit_hash": "dd379e76b76a6de5d7299e6a03345af513ec3f51f706e2aa32b25de36259d837", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1542, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1542, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated feature requests: one asks for a new Qwen3_5 token-classification/value-model class, the other asks for FP8 training support. They should not be clustered together.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different underlying requests and code changes: model/class support for token classification vs numerical precision/training support. No shared target or concrete overlapping bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 19, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45869-5", + "nodes": [ + "issue:45869", + "issue:45870", + "issue:45871", + "issue:45872", + "issue:45873" + ], + "soft_pairs": [ + "issue:45869|issue:45870", + "issue:45869|issue:45871", + "issue:45870|issue:45871", + "issue:45870|issue:45872", + "issue:45870|issue:45873", + "issue:45871|issue:45872", + "issue:45871|issue:45873", + "issue:45872|issue:45873" + ], + "prepared_review_unit_hash": "04242ae3784a557f29efbe57071946174e149fc4fc7387eb789313cb17d81950", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5725, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5725, + "estimated_input_tokens": 1432, + "estimated_eval_tokens": 3120 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "There are two separate exact-duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is about a dataset bundle and does not look like the same underlying request as the DataForge example/service issues.", + "confidence": 0.98, + "canonical_issue_reason": "45870 is a strong canonical representative for the 'Example: Financial time series prediction with DataForge' pair because it is an exact duplicate of 45871 and is the earlier issue; 45872/45873 form a separate duplicate pair, so there is no single canonical issue for all five items.", + "canonical_pr_reason": null, + "best_issue_reason": "45870 is the best overall issue to keep as the representative of its duplicate pair: it is earliest, fully duplicated by 45871, and its title is stable and specific. The other items are either a separate duplicate pair (45872/45873) or a different topic (45869).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Different topics: a financial data bundle promo versus a DataForge example issue; similar keywords but not the same underlying request." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Different topics: dataset bundle promotion versus DataForge example; not the same bug/change." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Exact duplicate title and matching metadata/body length indicate the same underlying issue." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both mention financial time series/DataForge, but one is an example issue and the other is a scraping-service request; not the same underlying item." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Same as 45870/45872: related theme, but different request and not the same issue." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Different concrete request: example prediction issue versus scraping-service issue." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Different concrete request: example prediction issue versus scraping-service issue." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Exact duplicate title and matching metadata/body length indicate the same underlying issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 have identical titles and matching body lengths, while 45869 is clearly a different topic. The soft-edge verdicts are conservative, with no obvious overstatement beyond the evidence available." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46439-2", + "nodes": [ + "issue:46439", + "issue:46441" + ], + "soft_pairs": [ + "issue:46439|issue:46441" + ], + "prepared_review_unit_hash": "366a0731ebf344bedb4ce1324517c0534f5993c8cbb193dd4b2cd1e1e6e61275", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both issues concern static-cache failures around linear-attention models, but they describe different failure modes: one is shape corruption during `early_initialization`, the other is a `KeyError` for `linear_attention` on Qwen3.5. They are related in subsystem, but not the same underlying bug.", + "confidence": 0.86, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it describes a more general and concrete static-cache code-path failure with a clear runtime consequence and root area (`early_initialization` / shape corruption).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46439 is the strongest representative of the cluster because it is more broadly applicable and closer to an actionable engine-level bug than the Qwen3.5-specific `KeyError` report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46439", + "right": "issue:46441", + "accept": false, + "reason": "Related static-cache/linear-attention breakage, but the symptoms differ materially: wrong tensor shapes in early initialization vs a missing `linear_attention` key on Qwen3.5. Not the same concrete bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-39895-14", + "nodes": [ + "pull_request:39895", + "pull_request:41855", + "pull_request:42668", + "pull_request:43251", + "pull_request:43254", + "pull_request:43424", + "pull_request:43823", + "pull_request:43995", + "pull_request:44827", + "pull_request:45170", + "pull_request:45213", + "pull_request:45655", + "pull_request:46048", + "pull_request:46267" + ], + "soft_pairs": [ + "pull_request:39895|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:41855|pull_request:43424", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:46267", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "fa1b7312c5d72dec8367e717985896f78bffc8022eaee6a6b9b2500cfd97cfad", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "One true duplicate pair stands out: the two `loss_utils.py` PRs both extend `fixed_cross_entropy` with the same new arguments on the same code path. The rest of the cluster is a grab bag of unrelated model additions, test fixes, docs, and refactors.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43251 is the clearest canonical: it is issue-linked to 43240, edits the target loss helper directly, and presents the fuller/primary implementation of the cross-entropy kwargs fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 43251 is the best overall representative because it is the most directly scoped fix in the cluster and the only one with an explicit issue target; it also subsumes the near-identical 43254 change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "Unrelated PRs: VideoPrism model/docs/weights conversion vs ExecuTorch dynamic-shape export tests. Different subsystems and code paths." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Different artifacts: Mistral tokenizer method addition vs adding the MobileLLM model. No shared bug or mergeable code-path overlap." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Tokenizer API change vs ExecuTorch test coverage; these do not fix the same underlying issue." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor auto-loading robustness changes vs adding a new model implementation. Same broad area of transformers, but not the same bug or change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Loss helper kwargs plumbing and a new model addition are unrelated changes." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Cross-entropy helper fix vs Mistral4 test/model refactor. No shared concrete code-path problem." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "ExecuTorch export test vs Mistral4 test/model changes; both are tests/infra-heavy, but they address different issues." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Dynamic-shape export test coverage and MobileLLM model addition are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "MobileLLM model addition vs fixing config resolution order in `modeling_utils.py`; different code paths and goals." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "New model implementation vs Falcon output-interface refactor; no common bug or combined PR candidate." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor vs a repo skill/documentation PR. Not the same underlying change." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon output-collection refactor and Mistral4 fixes are distinct model-specific changes." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test/model work vs a CLIP-family typo rename/mapping fix; different model families and bugs." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46267", + "accept": false, + "reason": "Mistral4-specific fixes vs chat parsing/serve updates; unrelated functionality." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 changes vs adding a TDT loss kernel integration; distinct code paths and concerns." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 model/test updates and config-resolution ordering fix are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Mistral4 fixes vs model-creation skill/docs; not the same bug/change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "Adding MobileLLM and adding a model-creation skill are different kinds of changes with no shared code-path issue." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Same file, same helper, same external behavior change: both extend `fixed_cross_entropy` to pass supported kwargs to `nn.functional.cross_entropy`. One is a near-duplicate variant of the same fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The only accepted pair, 43251/43254, is strongly supported by the packet: same file, same helper, same issue target, and near-identical kwargs plumbing. The other verdicts stay safely on the conservative side by rejecting pairs that are clearly different subsystems or bugfixes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-3", + "nodes": [ + "pull_request:40546", + "pull_request:42781", + "pull_request:46331" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781", + "pull_request:42781|pull_request:46331" + ], + "prepared_review_unit_hash": "c90d1d6a85fbfdba61b56a41ffab7bb9eeed402e89f9e3cdcd459b91bb018cf1", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10855, + "estimated_input_tokens": 2714, + "estimated_eval_tokens": 5684 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10855, + "estimated_input_tokens": 2714, + "estimated_eval_tokens": 5684 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is VibeVoice-centric, but the soft edges are not true duplicates: 40546 is the base VibeVoice implementation, 42781 is a later realtime/tokenizer extension, and 46331 is a separate Parakeet-RNNT PR that only overlaps on generic auto-registration files.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 40546 is the core VibeVoice implementation with the broadest base model/docs/auto wiring, so it is the cleanest cluster anchor.", + "best_issue_reason": null, + "best_pr_reason": "PR 40546 is the best representative of the cluster because it is the primary VibeVoice implementation; 42781 is a follow-on realtime feature set, and 46331 is unrelated.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "Related VibeVoice work, but not the same underlying change: 40546 adds the base model, while 42781 adds realtime/acoustic-tokenizer support and other extra plumbing, so they would not plausibly merge as one duplicate PR." + }, + { + "left": "pull_request:42781", + "right": "pull_request:46331", + "accept": false, + "reason": "Different model families and code paths: 42781 is VibeVoice realtime, while 46331 is Parakeet-RNNT; shared auto-file edits are generic and 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": 3, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40755-2", + "nodes": [ + "pull_request:40755", + "pull_request:41224" + ], + "soft_pairs": [ + "pull_request:40755|pull_request:41224" + ], + "prepared_review_unit_hash": "99d3314828221bd1f910a894967781fdeb84a6c2f837b397684f0fb3613856a6", + "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. No duplicate triage match here.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "pull_request:41224 is the more substantial standalone PR in this set, with a concrete model-support change and more review activity, but it is not a duplicate of pull_request:40755.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different models, different code paths, and different feature goals: TimesFM forecasting with covariates vs DINOv3 image classification support. They do not look 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": 4, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41356-2", + "nodes": [ + "pull_request:41356", + "pull_request:44339" + ], + "soft_pairs": [ + "pull_request:41356|pull_request:44339" + ], + "prepared_review_unit_hash": "aead1308c167b0bae6ff11bf9358340baa5e76f816618383418b505a10a9cd01", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two DEIMv2 pull requests target the same model family, but they are not the same change: PR 41356 is an early/basic model + image processor add, while PR 44339 is a much broader follow-up that adds loss code, auto-mappings, conversion tooling, modular files, and more integration work.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44339 is the more complete and representative DEIMv2 implementation; PR 41356 looks like an earlier partial version of the same model bring-up.", + "best_issue_reason": null, + "best_pr_reason": "PR 44339 is the strongest single artifact for the cluster because it subsumes the basic model work and includes the broader integration needed for a full Transformers addition.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41356", + "right": "pull_request:44339", + "accept": false, + "reason": "Same model area, but not the same concrete change: 44339 expands the implementation substantially and adds new code paths beyond the basic PR 41356. They look like related iterations, not mergeable duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41797-3", + "nodes": [ + "pull_request:41797", + "pull_request:45186", + "pull_request:45630" + ], + "soft_pairs": [ + "pull_request:41797|pull_request:45186", + "pull_request:45186|pull_request:45630" + ], + "prepared_review_unit_hash": "781cfc97896dc60a6f2df0187645d455c39f6b37b79a9fff0c900dd5fbfe0a38", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10434, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10434, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three unrelated model-addition pull requests that only overlap on the standard Transformers scaffolding for new models. They are not duplicates and should not be merged together.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR in this set: each pull request adds a different model family (DeepSeekOCR, Isaac, Kimi2-6) with distinct configs, processors, and tests.", + "best_issue_reason": null, + "best_pr_reason": "PR 45630 is the most active and broadest of the three, but it is not a better duplicate representative because it targets a different model than the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41797", + "right": "pull_request:45186", + "accept": false, + "reason": "Different model implementations and docs: DeepSeekOCR vs Isaac. Shared framework files are just normal new-model registration, not the same underlying change." + }, + { + "left": "pull_request:45186", + "right": "pull_request:45630", + "accept": false, + "reason": "Different model additions with different code paths and assets: Isaac vs Kimi2-6. Similar filenames reflect boilerplate for model onboarding, not a duplicate fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42975-2", + "nodes": [ + "pull_request:42975", + "pull_request:42976" + ], + "soft_pairs": [ + "pull_request:42975|pull_request:42976" + ], + "prepared_review_unit_hash": "7eca033038917baa9d7f82bf4618da6a5c187e5b764116f4c6f8b1d8fb4b9aa0", + "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": "Both items are GitHub Actions dependency bumps, but they target different action families and different scopes. PR 42975 is a broad workflow-tooling update for Node 24 compatibility, while PR 42976 only bumps docker/build-push-action in several Docker image workflows. They are related, but not the same underlying change.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the broader, more representative workflow-action upgrade and has the clearest stated purpose (Node 24 compatibility); PR 42976 is a narrower, separate action-version bump.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the best representative artifact for this cluster because it covers the wider GitHub Actions compatibility migration rather than a single action upgrade.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Reject: both are workflow maintenance, but they update different actions (general GitHub Actions/runtime compatibility vs docker/build-push-action) and do not appear to fix the same concrete problem or belong in one merged PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 7, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42978-4", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:43451", + "pull_request:45490" + ], + "soft_pairs": [ + "pull_request:42978|pull_request:43448", + "pull_request:42978|pull_request:43451", + "pull_request:42978|pull_request:45490", + "pull_request:43448|pull_request:43451", + "pull_request:43448|pull_request:45490", + "pull_request:43451|pull_request:45490" + ], + "prepared_review_unit_hash": "9dc2d25618202401aeec09219d72cff8990a5e1c675f2a2437d7405a222229d0", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14247, + "estimated_input_tokens": 3562, + "estimated_eval_tokens": 7380 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14247, + "estimated_input_tokens": 3562, + "estimated_eval_tokens": 7380 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are four separate model-addition PRs (ViT NEPA, Molmo, Molmo2, CTSM), not duplicates. The shared docs/auto-registry edits are expected for new model integrations and do not indicate the same underlying change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: each PR adds a different model family with distinct files, configs, and modeling code, so they are not the same change.", + "best_issue_reason": null, + "best_pr_reason": "No single best PR for deduplication; all are legitimate but unrelated integrations, so none should represent the cluster.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model families (ViT NEPA vs Molmo) with distinct modeling and test code; shared auto/docs updates are only boilerplate for registration." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "ViT NEPA and Molmo2 are unrelated model additions; they do not fix the same code-path problem." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "CTSM time-series integration is a different subsystem and architecture from ViT NEPA, so these are not duplicates." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are related by name but are separate model integrations with different files and functionality, not one mergeable duplicate PR." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM are unrelated model additions; overlapping registry/docs edits are not enough to treat them as the same change." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 and CTSM address different models and code paths, so they should not be merged as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 8, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-4", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815", + "pull_request:46380" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43291|pull_request:46380", + "pull_request:43665|pull_request:44815", + "pull_request:43665|pull_request:46380", + "pull_request:44815|pull_request:46380" + ], + "prepared_review_unit_hash": "6a10945238272666b9a87d5e57272e7010c85c7fdf77577fdb776d01844bcbfa", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 11145, + "estimated_input_tokens": 2787, + "estimated_eval_tokens": 5830 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 11145, + "estimated_input_tokens": 2787, + "estimated_eval_tokens": 5830 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are not duplicates: they touch different subsystems and fix different problems. 46380 and 44815 are the strongest real bug-fix candidates, while 43665 looks like test-only noise.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No issue links are present, and the PRs are unrelated enough that there is no true canonical duplicate. If one PR must represent the cluster, 46380 is the clearest concrete bug fix.", + "best_issue_reason": null, + "best_pr_reason": "46380 is the best representative overall: it fixes a specific runtime validation bug in generation config handling, with a coherent code-path and test coverage. 44815 is also substantive, but it targets a different FP8/dequantization path.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Whisper tokenizer decode/timestamp logic vs. CLIP/ViT test changes; different models, different bug classes, and 43665 appears to be test-only." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decoding behavior is unrelated to FP8 dequantization and model-loading changes." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Whisper tokenizer code vs. generation watermarking/config validation; distinct code paths and fixes." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "Test assertions in CLIP/ViT are unrelated to FP8 kernel/model-loading behavior." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "Test-only edits for CLIP/ViT do not match generation config validation in 46380." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "FP8 dequant/padding and model-loading changes are unrelated to generation config watermarking validation." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 9, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43488-2", + "nodes": [ + "pull_request:43488", + "pull_request:43532" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:43532" + ], + "prepared_review_unit_hash": "b3d449f812f0369e4dad745f57d716e1603376d008038cf26a6c3cd31682eedd", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4563, + "estimated_input_tokens": 1141, + "estimated_eval_tokens": 2538 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4563, + "estimated_input_tokens": 1141, + "estimated_eval_tokens": 2538 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two pull requests are not duplicates: one is a small repo-bot/test formatting probe, while the other changes CI workflows and related utility code. The only overlap is a shared test file, which is not enough to merge them as the same change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs target different code paths and purposes, with no evidence they fix the same concrete bug or implement the same change.", + "best_issue_reason": null, + "best_pr_reason": "If a representative is needed, PR 43532 is the better cluster anchor because it is the more substantial end-to-end workflow/utility change; PR 43488 is a narrow bot-check/test setup tweak.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Reject: these are unrelated PRs. 43488 only tweaks setup/versioning and a test hook, while 43532 rewrites CI workflow logic and touches multiple supporting files. Shared test filename alone does not indicate the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 10, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43612-2", + "nodes": [ + "pull_request:43612", + "pull_request:43613" + ], + "soft_pairs": [ + "pull_request:43612|pull_request:43613" + ], + "prepared_review_unit_hash": "a80accd323244a28022ca51524a27be0753c81477b0a4362023f3804552339ab", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs add similar new segmentation pipelines and share doc/registry/test plumbing, but they target different features and code paths: Promptable Concept Segmentation vs Promptable Visual Segmentation. They should not be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43613 is the broader and more mature change set: it covers Promptable Visual Segmentation and integrates more model families (SAM, SAM2, EdgeTAM, SAM3 tracker), so it is the better representative of this cluster if one must be chosen.", + "best_issue_reason": null, + "best_pr_reason": "PR #43613 is the better anchor for the cluster because it is the more general segmentation pipeline addition and has the larger supporting surface area. That said, it is still a distinct feature from #43612, not the same underlying change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "These are different pipeline additions with different intended behaviors and model integrations. Shared docs/registration/test scaffolding is not enough to treat them as the same concrete change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 11, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43651-2", + "nodes": [ + "pull_request:43651", + "pull_request:43663" + ], + "soft_pairs": [ + "pull_request:43651|pull_request:43663" + ], + "prepared_review_unit_hash": "9de4cac0da3dc8dc546304bd9478fb01821dce859f6c35eadba3bf083d6fc4b4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two Trainer.py extensibility PRs that look superficially similar but address different code paths: gradient-accumulation loss scaling vs. signature-column filtering.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43651 is the stronger representative because it makes a concrete runtime change to training_step and adds the new _loss_is_scaled_for_ga hook for custom trainers.", + "best_issue_reason": null, + "best_pr_reason": "PR 43651 is the best overall PR in this cluster as it introduces a clearer, more self-contained Trainer behavior change with the broader extensibility hook.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Reject: they touch different Trainer internals and solve different problems. 43651 changes loss normalization in training_step; 43663 factors out signature-column inspection for column filtering. Shared filename and similar extensibility framing are 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": 12, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43656-2", + "nodes": [ + "pull_request:43656", + "pull_request:43842" + ], + "soft_pairs": [ + "pull_request:43656|pull_request:43842" + ], + "prepared_review_unit_hash": "10181352bff315da093a668b6561586643ebfa37711b1f85ca91ec2ac760d2f8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3663, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch the same CLI file and both mention NameError-ish typing issues, but they fix different concrete problems: one is about annotation/runtime typing in `serve.py`, the other is specifically about `TypeAdapter` when `pydantic` is absent.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: 43656 targets a broader annotation/runtime typing cleanup, while 43842 specifically fixes the `TypeAdapter` import/type-guard issue when `pydantic` is not installed.", + "best_issue_reason": null, + "best_pr_reason": "PR 43842 is the more focused and explicit bug fix: it names the failing `TypeAdapter` path, adds the guarded import, and has a narrower, clearer issue target.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Reject: although both modify `src/transformers/cli/serve.py` and are related to typing/NameError problems, they address different runtime failures and are not the same underlying 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": 13, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43779-2", + "nodes": [ + "pull_request:43779", + "pull_request:43816" + ], + "soft_pairs": [ + "pull_request:43779|pull_request:43816" + ], + "prepared_review_unit_hash": "e7c674adcc031aa8883109a5dfc678b4407fb20b10764302092e384d031ba83a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests appear to implement the same SwanLab integration change: adding run id/resume support in `integration_utils.py` for issue #43698, with one variant also expanding the docs. They are strong duplicate PR candidates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43816 is the cleaner representative: it carries the same SwanLab id/resume support plus the accompanying documentation updates, making it the more complete version of the shared change.", + "best_issue_reason": null, + "best_pr_reason": "PR #43816 is the best global representative because it appears to include the full feature change and docs for SwanLab run id/resume support, while #43779 is the narrower implementation-focused variant.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43779", + "right": "pull_request:43816", + "accept": true, + "reason": "Same underlying SwanLab callback enhancement for id/resume support in the same file and same issue target; the doc differences do not change the core code-path change, so these are duplicate PRs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: both PRs target issue #43698, touch the same file, and the previews indicate the same SwanLab id/resume support with #43816 adding docs. The duplicate/mergeable verdict is conservative enough given the shared target and overlapping code change." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43996", + "pull_request:43997", + "pull_request:44007", + "pull_request:44010", + "pull_request:44013", + "pull_request:44017", + "pull_request:44018", + "pull_request:44019", + "pull_request:44044", + "pull_request:44056", + "pull_request:44059", + "pull_request:44066", + "pull_request:44068", + "pull_request:44071", + "pull_request:44072", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44085", + "pull_request:44086", + "pull_request:44098", + "pull_request:44101", + "pull_request:44116", + "pull_request:44129", + "pull_request:44154", + "pull_request:44161", + "pull_request:44722" + ], + "soft_pairs": [ + "pull_request:43996|pull_request:44085", + "pull_request:43996|pull_request:44044", + "pull_request:44066|pull_request:44085", + "pull_request:44007|pull_request:44072", + "pull_request:44072|pull_request:44722", + "pull_request:44066|pull_request:44072", + "pull_request:44013|pull_request:44044", + "pull_request:44066|pull_request:44086", + "pull_request:44066|pull_request:44071", + "pull_request:44018|pull_request:44068", + "pull_request:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", + "pull_request:44019|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44086|pull_request:44722", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44068|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44059|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44098|pull_request:44101", + "pull_request:44059|pull_request:44722", + "pull_request:44018|pull_request:44071", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44722", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44071|pull_request:44086", + "pull_request:44073|pull_request:44101", + "pull_request:44017|pull_request:44071", + "pull_request:44073|pull_request:44161", + "pull_request:44059|pull_request:44071", + "pull_request:44056|pull_request:44101", + "pull_request:44018|pull_request:44059", + "pull_request:44059|pull_request:44068", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44017", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44018", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request:44018|pull_request:44161", + "pull_request:43997|pull_request:44071", + "pull_request:44010|pull_request:44071", + "pull_request:44010|pull_request:44073", + "pull_request:44018|pull_request:44129", + "pull_request:43997|pull_request:44074", + "pull_request:43997|pull_request:44129", + "pull_request:44071|pull_request:44129", + "pull_request:44010|pull_request:44018", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154" + ], + "prepared_review_unit_hash": "158a1d3a071672acfe5c5dc8239d8f01ccd24567f936b7e429c0766ff8c218ef", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 + }, + "budget": { + "node_count": 27, + "item_count": 27, + "soft_pair_count": 72, + "serialized_chars": 75090, + "estimated_input_tokens": 18773, + "estimated_eval_tokens": 37802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly standardized output-tracing refactors spread across different model families. The clearest duplicate-like grouping is the GPT-J/CodeGen line around 44066/44085/44722; most other PRs are separate model-specific slices under umbrella issue #43979 rather than the same concrete change.", + "confidence": 0.79, + "canonical_issue_reason": "Issue #43979 is the shared umbrella tracker referenced by most PRs and best represents the underlying standardized output-tracing effort.", + "canonical_pr_reason": "PR #44722 is the strongest representative of the GPT-J/CodeGen subcluster: it is explicit, issue-linked, and spans the shared GPT-J code path plus the copied CodeGen block.", + "best_issue_reason": "As the common tracking issue, #43979 is the best issue-level anchor for the cluster; the individual PRs are implementation slices rather than separate problem statements.", + "best_pr_reason": "PR #44722 is the best single PR exemplar in this set because it is the most complete GPT-J/CodeGen-family refactor and most clearly aligns with the umbrella issue.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Shared umbrella issue is not enough; FNet/CVT and GPT-J are different model code paths." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "FNet/CVT refactor is unrelated to DeBERTa; same issue target does not make them duplicates." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Both touch GPT-J output-tracing in the same file, so they overlap on the same concrete code path and are plausibly mergeable." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "RegNet and EfficientNet are separate model implementations, not the same bug/change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet output tracing and GPT-J/CodeGen output tracing are unrelated model-specific refactors." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen and EfficientNet touch different code paths and are not mergeable duplicates." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are unrelated implementations despite the shared umbrella issue." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen and MGP-STR are different models with different output paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen and MPT are separate model families; same refactor theme is too broad to merge." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Both are GPT-Neo output-tracing refactors on the same code path and are close enough to be duplicate change proposals." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen and GPT-Neo are different concrete implementations." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet and GPT-J/CodeGen are unrelated model code paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are not the same underlying change." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are separate models with no shared code path." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are unrelated refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are different model implementations." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert/XLM are unrelated concrete code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are distinct model families." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert/XLM are unrelated refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are unrelated implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo and GPT-J are different model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated model-specific changes." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR do not share the same concrete bug/change." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and XLM/Flaubert are different models; same refactor style is too broad." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-2 and MGP-STR are unrelated model implementations." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are different concrete code paths." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and XLM/Flaubert are separate model families." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "SegFormer and LongT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "SegFormer and GPT-J/CodeGen are unrelated concrete changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "SegFormer and VisualBert are different model code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo and MPT are unrelated models." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44073", + "accept": false, + "reason": "RegNet and VisualBert are unrelated model implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT and MGP-STR are separate model families." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44101", + "accept": false, + "reason": "VisualBert and XLM/Flaubert are unrelated code paths." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44071", + "accept": false, + "reason": "SegFormer and MPT are different models." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-2 and MPT are unrelated refactors." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44101", + "accept": false, + "reason": "MPNet and XLM/Flaubert are separate model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44059", + "accept": false, + "reason": "GPT-Neo and GPT-2 are different model families." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-2 and GPT-Neo are not the same concrete change." + }, + { + "left": "pull_request:44101", + "right": "pull_request:44161", + "accept": false, + "reason": "XLM/Flaubert and LongT5 are unrelated model implementations." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and SegFormer are different model paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are unrelated refactors." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBert are separate model code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are unrelated model changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "FNet/CVT and LongT5 are unrelated implementations." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "FNet/CVT and SegFormer are different model families." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBert and DeBERTa are unrelated concrete changes." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are separate model-specific refactors." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing refactor family; 44722 is a later, compatible version of the same change." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both edits are GPT-J output-tracing refactors on the same file and same underlying code path." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated model implementations." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "FNet/CVT and RegNet are unrelated model code paths." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "FNet/CVT and SpeechT5 are not the same underlying bug/change." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa and SpeechT5 are different models." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa and VITS are unrelated concrete changes." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the soft-edge call on 44066 vs 44085 is too strong. Both PRs touch the same GPT-J file (`src/transformers/models/gptj/modeling_gptj.py`) and share the umbrella issue target, so describing them as \u201cdifferent model families and different code paths\u201d overstates the separation. At minimum, this pair looks plausibly mergeable or duplicate-like enough that a conservative verdict would not rule it out so firmly. The summary is mostly grounded, but the exclusivity claim (\u201conly true duplicate-like groups\u201d) should also be softened." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 35, + "review_unit_index": 2, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43997", + "pull_request:44010", + "pull_request:44024", + "pull_request:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44129" + ], + "soft_pairs": [ + "pull_request:44024|pull_request:44076", + "pull_request:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "c37762392b7782bbcf0d8640b918d5640f8eff8cfff7939a38c6d0ed780fb6fe", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84894, + "estimated_input_tokens": 21224, + "estimated_eval_tokens": 42704 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 6, + "serialized_chars": 17125, + "estimated_input_tokens": 4282, + "estimated_eval_tokens": 8820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they touch different architectures and code paths. They are related thematically, not duplicates of the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:44129 is the clearest representative: it has an explicit issue target, a focused title, and a concrete output-tracing refactor in one model implementation.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:44129 is the best single representative for the cluster because it is narrowly scoped and explicitly tied to the standardized output capture work, but it is not a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "Both are output-capture refactors, but they target different models (FocalNet vs ImageGPT) and different code paths, so they are not the same change." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBert and TextNet are separate model implementations; this is the same refactor pattern, not a shared underlying bug or mergeable duplicate PR." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert are unrelated model-specific output-tracing updates; they do not fix the same concrete defect." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "These are analogous refactors in different model files, not duplicate fixes to one code-path problem." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert each change their own model forward/output handling, so they are not the same underlying change." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 both involve output handling, but they modify different model internals and are not the same concrete fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-13", + "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": "06b9e3e1609830c6189c1d7130612d00c47831fd663a6368f549830fd4371a14", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41333, + "estimated_input_tokens": 10334, + "estimated_eval_tokens": 20924 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41333, + "estimated_input_tokens": 10334, + "estimated_eval_tokens": 20924 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a family of similar-looking but distinct model-specific output-tracing refactors. They share naming/style and utility usage, but each PR targets a different code path and should not be merged as duplicates.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44025 is the clearest representative of the cluster: it applies the refactor to a primary model plus its paired prompt variant, making it a good exemplar of the shared pattern.", + "best_issue_reason": null, + "best_pr_reason": "PR 44025 is the best standalone representative of the theme because it is explicit, self-contained, and spans a closely related model pair without being overly broad.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Same refactor pattern, but Timm Backbone and UperNet are different model code paths." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Different model families; same tracing cleanup theme is not enough for a duplicate PR." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Timm Backbone vs UnivNet are unrelated implementations." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Different architectures and files; not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Timm Backbone and RWKV are separate model paths." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "MobileNetV1 and UnivNet are different code paths." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "MobileNetV1 vs UperNet is only a broad thematic match." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "MobileNetV1 and DepthAnything target different model families." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "MobileNetV1 vs VisionEncoderDecoder are distinct implementations." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "MobileNetV1 vs SpeechEncoderDecoder are separate code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Vision-Text Dual Encoder and UnivNet are unrelated model changes." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families; not the same fix." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Vision-Text Dual Encoder and DepthAnything are separate code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44026", + "accept": false, + "reason": "Different architectures and files; same output-tracing refactor is too broad." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "UnivNet and UperNet are unrelated models." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "UnivNet vs DepthAnything are distinct implementations." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44026", + "accept": false, + "reason": "Different code paths; not mergeable as one PR." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "UperNet and RWKV are different model families." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44025", + "accept": false, + "reason": "UperNet and DepthAnything are separate changes." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "UnivNet vs Mamba/Falcon-Mamba are unrelated implementations." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "UperNet vs Mamba/Falcon-Mamba are different code paths." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Vision-Text Dual Encoder and Mamba/Falcon-Mamba are unrelated." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Mamba/Falcon-Mamba and SuperPoint are separate model families." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Mamba/Falcon-Mamba vs CodeGen is not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Timm Backbone and CodeGen are unrelated code paths." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "UnivNet vs CodeGen share only the refactor theme." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "UperNet vs CodeGen are different implementations." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Vision-Text Dual Encoder vs CodeGen are distinct model paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "DepthAnything and VisionEncoderDecoder are different code paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "DepthAnything and SpeechEncoderDecoder are separate implementations." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "DepthAnything vs SuperPoint is only a broad pattern match." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "DepthAnything and RWKV are unrelated model families." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "DepthAnything and DPR are different code paths." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "VisionEncoderDecoder vs RWKV are unrelated implementations." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "VisionEncoderDecoder and SpeechEncoderDecoder are separate wrappers, not the same fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "VisionEncoderDecoder vs SuperPoint are different model paths." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "VisionEncoderDecoder and DPR are unrelated code paths." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "SpeechEncoderDecoder vs SuperPoint are separate implementations." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "SpeechEncoderDecoder and RWKV are unrelated model families." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "SpeechEncoderDecoder and DPR are distinct code paths." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "SuperPoint and RWKV are different model families." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "SuperPoint and DPR are separate implementations." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "RWKV and DPR are unrelated model code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 16, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44662-2", + "nodes": [ + "pull_request:44662", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401" + ], + "prepared_review_unit_hash": "6c9621482385977d79ffe69e483c88eb31dd67dc617f93ebddbf9bbf06710bbf", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated model-addition PRs: one implements PenguinVL, the other adds Voxtral TTS support. They share only common Transformers integration scaffolding, not the same underlying change or code path.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45401 is the better canonical representative only because it is explicitly tied to issue 45027 and has a tighter, well-scoped model addition; PR 44662 is a separate unrelated implementation for a different model.", + "best_issue_reason": null, + "best_pr_reason": "PR 45401 is the better overall PR representative for the cluster because it has an explicit tracking issue and a narrower, more clearly defined feature scope. PR 44662 is a distinct PenguinVL implementation and should not be merged with it.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different models and different feature work: PenguinVL implementation vs Voxtral TTS support. Shared auto/model registration files are generic integration overlap, not evidence of the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 17, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44979-2", + "nodes": [ + "pull_request:44979", + "pull_request:45363" + ], + "soft_pairs": [ + "pull_request:44979|pull_request:45363" + ], + "prepared_review_unit_hash": "065146b3ef00eeaf2b853102a8ba0d028b1f00f78d09a30b95e284dcd3224716", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4406, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4406, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs are about fusion, but they are not duplicates: PR 44979 introduces a generic Module Fusion API, while PR 45363 implements KernelConfig-based n-to-1 kernel fusion integration and an example. Same broad area, different layers/code paths.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45363 is the more concrete, end-user-facing change: it wires fusion through KernelConfig and modeling hooks rather than just adding the abstraction layer.", + "best_issue_reason": null, + "best_pr_reason": "PR 45363 is the best representative of the cluster because it captures the actual runtime integration path and practical usage example for fusion.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "They overlap on fusion concepts but target different implementations: 44979 adds a generic Module Fusion API, while 45363 adds KernelConfig-driven n-to-1 kernel fusion integration. They are related, not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 18, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45086-2", + "nodes": [ + "pull_request:45086", + "pull_request:45317" + ], + "soft_pairs": [ + "pull_request:45086|pull_request:45317" + ], + "prepared_review_unit_hash": "5256a53f486842f1b25657abfee6c5fd24498960bf8fcdcece93743fa09ccb1d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests fix the same `_patch_mistral_regex` AttributeError in `tokenization_utils_tokenizers.py` for the same tracked issue; the second PR is a slightly more complete version with tests.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45317 is the stronger canonical PR: it fixes the same bug as #45086 and also adds test coverage, making it the more complete representative of the change.", + "best_issue_reason": null, + "best_pr_reason": "PR #45317 is the best overall PR because it addresses the same concrete code-path bug while including validation tests; #45086 is effectively an earlier, narrower variant of the same fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying `_patch_mistral_regex` AttributeError fix in the same function/file, same explicit issue target, and the second PR only adds tests rather than changing the problem being fixed." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45115-2", + "nodes": [ + "pull_request:45115", + "pull_request:45569" + ], + "soft_pairs": [ + "pull_request:45115|pull_request:45569" + ], + "prepared_review_unit_hash": "aa9aa39d1dc95db112c3bd0d9af1c5ff9276c01b643e7ec369f5c32c472f3d44", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are about NemotronH, but they address different changes: #45115 is a narrow conversion/mapping and inheritance refactor, while #45569 is a broader model split into dense/sparse variants with docs and auto-mapping updates. They are related, but not the same underlying fix/change set.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45569 is the broader, more central NemotronH rework and better represents the cluster topic; #45115 is a narrower supporting refactor focused on conversion mappings and inheritance cleanup.", + "best_issue_reason": null, + "best_pr_reason": "#45569 is the best single PR to keep because it covers the main architectural change and the new dense/sparse model support, whereas #45115 only adjusts conversion naming and model wiring.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Related NemotronH work, but not the same concrete change: #45115 refactors conversion mapping/inheritance compatibility, while #45569 introduces dedicated dense/sparse model packages, docs, and auto-mapping updates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 20, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45332-2", + "nodes": [ + "pull_request:45332", + "pull_request:45333" + ], + "soft_pairs": [ + "pull_request:45332|pull_request:45333" + ], + "prepared_review_unit_hash": "e41fe80370b097da6b802665eecf683f575f96bf9c4a434717910a8e77a2b508", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5465, + "estimated_input_tokens": 1367, + "estimated_eval_tokens": 2990 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5465, + "estimated_input_tokens": 1367, + "estimated_eval_tokens": 2990 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two related but distinct pull requests: one adds heterogeneous config support, the other adds broader heterogeneous modeling/cache support. They share implementation files, but they are not the same concrete change and are better treated as separate work items.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45332 is the broader end-to-end feature PR and better represents the cluster; PR #45333 is a narrower config-only piece that looks like a prerequisite or companion change rather than the same PR.", + "best_issue_reason": null, + "best_pr_reason": "PR #45332 is the best representative PR because it covers the larger heterogeneous model support effort across config, modeling, cache, and tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related feature work, but not the same underlying bug/change: one is config support and the other is modeling/cache support. Shared files alone are not enough to treat them as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45415-2", + "nodes": [ + "pull_request:45415", + "pull_request:45425" + ], + "soft_pairs": [ + "pull_request:45415|pull_request:45425" + ], + "prepared_review_unit_hash": "56b29307a2e2f3161c4a4347551bc2c607d622c537ebdefa059a03844f13d7f3", + "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 PRs are typing-related, but they address different scopes: one is a broad type-checking sweep across many transformers modules, while the other adds shared typing protocols/aliases and a smaller set of typing refinements. They are related, but not the same concrete change.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45415 is the broader, more representative typing cleanup: it adds type checking coverage/config and touches many modules in one pass. PR 45425 is a narrower follow-up focused on shared typing helpers and protocol definitions.", + "best_issue_reason": null, + "best_pr_reason": "PR 45415 is the best representative artifact for the cluster because it captures the main end-to-end type-checking initiative and has the widest surface area.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Both are typing improvements, but they do not fix the same concrete bug or change. 45415 is a broad type-checking sweep across many files, while 45425 adds typing aliases/protocols and related fixes in a smaller set of files. They could be separate PRs and are 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": 22, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45513-2", + "nodes": [ + "pull_request:45513", + "pull_request:46084" + ], + "soft_pairs": [ + "pull_request:45513|pull_request:46084" + ], + "prepared_review_unit_hash": "2b0e9c24acf751d11367a2187195f5e7812164fe7c643c66d126d311ad74d979", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4858, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4858, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs about cache-aware multi-token forward paths, but they fix different model families and different code paths: Qwen/Olmo linear attention vs Mamba2 cache state handling. They are superficially similar, not duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45513 is the more central representative of the cluster: it has broader scope across several related Qwen/Olmo model files, more discussion/inbound references, and the main cache-forward fix theme the cluster appears to revolve around.", + "best_issue_reason": null, + "best_pr_reason": "PR #45513 is the best standalone representative because it is the broader, higher-activity cache-forward fix. PR #46084 is a separate Mamba2 bug fix and should not be merged into this cluster despite the similar cache/seq_len wording.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45513", + "right": "pull_request:46084", + "accept": false, + "reason": "Both mention cached multi-token forward paths, but they target different architectures and concrete implementations: Qwen/Olmo GDN linear attention vs Mamba2 CUDA kernel cache states. The diffs do not indicate the same underlying bug or a mergeable single PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 23, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45548-2", + "nodes": [ + "pull_request:45548", + "pull_request:45662" + ], + "soft_pairs": [ + "pull_request:45548|pull_request:45662" + ], + "prepared_review_unit_hash": "d05ddc9be153b248a3e8aec9d1d65c6ff423a13610e2a8e4bfbeec0eafe1a8b1", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs touch expert-parallel loading, but they address different bugs and different code paths: 45548 is about distributed_config serialization and accelerate/ZeRO-3 loading, while 45662 is about expert parameters being overwritten during FSDP2 rank-0 broadcast.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: the two patches solve related EP loading problems, but on different mechanisms and paths, so they should not be merged as one duplicate fix.", + "best_issue_reason": null, + "best_pr_reason": "45662 is the stronger representative of the cluster because it targets a concrete, reproducible parameter-overwrite bug with a broader fix across loading/integration code, whereas 45548 is narrower and focused on serialization/launch behavior.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Different underlying bugs. 45548 removes distributed_config from serialization and adds has_ep for accelerate/ZeRO-3 loading; 45662 wraps sharded params to prevent expert weights being overwritten during FSDP2 rank-0 broadcast. Shared EP context is not enough to treat them as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45639-2", + "nodes": [ + "pull_request:45639", + "pull_request:45645" + ], + "soft_pairs": [ + "pull_request:45639|pull_request:45645" + ], + "prepared_review_unit_hash": "651095e3ac14c7915455b92951552488d47b90970383e1f68f78f44de3fa26e4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same xdist-related captured_info logging collision in testing utilities; PR 45645 is the broader, more complete fix and includes CI log handling, while PR 45639 is a narrower subset.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:45645 is the best canonical PR because it covers the core testing_utils xdist-safe file change plus the CI workflow update to aggregate multiple captured_info files, making it the most complete representation of the fix.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:45645 is the strongest representative of the cluster: it fixes the concrete xdist collision in the code path and also preserves CI debug log collection, whereas 45639 only addresses the utility-side portion.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Both PRs fix the same underlying captured_info/xdist collision in testing_utils, and 45645 extends the same fix to CI log collection; 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": 25, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45670-3", + "nodes": [ + "pull_request:45670", + "pull_request:46320", + "pull_request:46401" + ], + "soft_pairs": [ + "pull_request:45670|pull_request:46320", + "pull_request:45670|pull_request:46401", + "pull_request:46320|pull_request:46401" + ], + "prepared_review_unit_hash": "e5eea6e37be9354b6a818b6aa4e174e902d5adf5d4c95b9538580091a0e8e4b0", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7576, + "estimated_input_tokens": 1894, + "estimated_eval_tokens": 4044 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7576, + "estimated_input_tokens": 1894, + "estimated_eval_tokens": 4044 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are related only at a high level: two add attn_implementation dispatch support for different model families, while one is a separate AutoModel mapping fix. They do not look like duplicates or mergeable into one change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true duplicate cluster here. If a representative PR is needed, PR 46401 is the broadest and most developed attn-dispatch change across multiple BLIP-family models.", + "best_issue_reason": null, + "best_pr_reason": "PR 46401 is the strongest representative of the cluster because it has the broadest scope, touches multiple model implementations/tests, and has the most discussion/review activity.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45670", + "right": "pull_request:46320", + "accept": false, + "reason": "Different changes: AutoModel registry wiring for GLMASR versus RecurrentGemma attention-dispatch support. Not the same code-path bug or fix." + }, + { + "left": "pull_request:45670", + "right": "pull_request:46401", + "accept": false, + "reason": "Unrelated fixes: AutoModel mapping for GLMASR versus BLIP-family attn_implementation dispatch support." + }, + { + "left": "pull_request:46320", + "right": "pull_request:46401", + "accept": false, + "reason": "Both concern attention dispatch, but for different model families and different implementation paths; they are not the same concrete fix and would not plausibly be merged as one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 26, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45720-2", + "nodes": [ + "pull_request:45720", + "pull_request:46259" + ], + "soft_pairs": [ + "pull_request:45720|pull_request:46259" + ], + "prepared_review_unit_hash": "c08547e1f95f9ad4450748d63c460d23ef9b8f46051f1246c73b9bc368724b0e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7188, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7188, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two unrelated linter-rule PRs were grouped together because they both modify Transformers modeling code, but they target different concrete problems: TRF003 for return_dict/can_return_tuple in model forwards vs. TRF018 for dead do_* flags in image/video processors. They should not be deduplicated.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: 45720 and 46259 address different rules, different code paths, and different model families, so they are not duplicates.", + "best_issue_reason": null, + "best_pr_reason": "PR 45720 is the better representative of the cluster\u2019s linter-rule work because it has broader impact across many model forwards and introduces the reusable can_return_tuple decorator; 46259 is a separate processor-specific rule.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45720", + "right": "pull_request:46259", + "accept": false, + "reason": "Both are linter-related, but they fix different issues: forward return handling in text/audio models vs. unused do_* preprocessing flags in image/video processors. Not the same change and not mergeable as one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 27, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45737-2", + "nodes": [ + "pull_request:45737", + "pull_request:45738" + ], + "soft_pairs": [ + "pull_request:45737|pull_request:45738" + ], + "prepared_review_unit_hash": "cc2ade73a49ec6a8883a7a24c433988645588bcd049e8aad2bce56b1b24daa1c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same MusicgenMelody cache initialization bug in `forward()`: replacing `EncoderDecoderCache(DynamicCache, DynamicCache)` with a single `DynamicCache`. PR #45737 is the more complete duplicate because it bundles the fix with a regression test, while #45738 is the fix-only variant.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45737 is the best canonical PR because it contains the same code fix as #45738 plus a regression test, making it the most complete representation of the bug fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #45737 is the strongest overall candidate: it fixes the concrete cache bug and adds a targeted regression test, whereas #45738 only changes the implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45737", + "right": "pull_request:45738", + "accept": true, + "reason": "Same underlying code-path bug in `musicgen_melody` cache setup; the diffs match on the core fix and the test-only addition in #45737 could plausibly live in the same PR as #45738." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet: both PRs make the same `musicgen_melody` cache initialization change, and #45737 only adds a regression test on top of that fix. The soft-edge accept is conservative enough because the diffs are mergeable into one PR for the same bugfix." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 28, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45932-2", + "nodes": [ + "pull_request:45932", + "pull_request:46100" + ], + "soft_pairs": [ + "pull_request:45932|pull_request:46100" + ], + "prepared_review_unit_hash": "2e1556eeaaa84146eefb95b6c51edcaccba5504a488b1a80ce7733394bc733c8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are not duplicates: one is a broad testing/infra and model-change PR, while the other is a small ViT test-edit PR with intentional failing assertions. The similarity is superficial.", + "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:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "Reject: they do not address the same code-path or bug. PR 45932 touches CI/workflow plumbing and several model files, while PR 46100 only edits a ViT test file and adds failing assertions for migration. Shared model/test surface is too broad and not a plausible single merged fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 29, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46073-3", + "nodes": [ + "pull_request:46073", + "pull_request:46154", + "pull_request:46196" + ], + "soft_pairs": [ + "pull_request:46073|pull_request:46154", + "pull_request:46073|pull_request:46196", + "pull_request:46154|pull_request:46196" + ], + "prepared_review_unit_hash": "9df4f97523bdf2b2ef90866a130bfb7224a30675f218ea0df4dd801627bc3e32", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8083, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8083, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three PRs are separate model-specific SDPA dispatch enablement changes. They share a common theme, but they do not look like the same concrete bug/fix and would not plausibly be merged as one PR.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46196 is the broadest representative of the shared SDPA-dispatch theme, spanning multiple model families and their tests, so it is the best cluster anchor among the three.", + "best_issue_reason": null, + "best_pr_reason": "PR 46196 is the most comprehensive and central example of the shared feature work, covering several affected model implementations rather than a single model.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46073", + "right": "pull_request:46154", + "accept": false, + "reason": "Both add SDPA dispatch support, but for different model implementations (CTRL vs T5) with different code paths and tests; same theme, not the same change." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46196", + "accept": false, + "reason": "These target different model families and modules; 46196 is a broader multi-model change, not the same concrete fix as the CTRL-only PR." + }, + { + "left": "pull_request:46154", + "right": "pull_request:46196", + "accept": false, + "reason": "T5-specific SDPA dispatch work is distinct from the broader wav2vec2/SeamlessM4T patch; they share a feature goal but not a mergeable single code-path 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": 30, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46126-2", + "nodes": [ + "pull_request:46126", + "pull_request:46269" + ], + "soft_pairs": [ + "pull_request:46126|pull_request:46269" + ], + "prepared_review_unit_hash": "cec03688df403c6cd08df80f1572f22f0d74c3a16374c0bad5fd2780bbe55309", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5571, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5571, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related distributed-training PRs, but they are not duplicates: one is a focused EP + Trainer integration, while the other is a much broader distributed branch base touching many subsystems. They share some files because the larger PR likely builds the foundation the smaller one uses, not because they fix the same concrete change.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46269 is the better canonical representative because it is the broader foundational distributed branch, spans many more affected components, and appears to subsume the shared infrastructure that #46126 builds on.", + "best_issue_reason": null, + "best_pr_reason": "PR #46269 is the best single representative for the cluster since it is the larger umbrella change and the shared-file overlap is better explained by dependency on the same distributed infrastructure than by duplicate scope.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Reject: #46126 is a targeted EP + Trainer integration PR, while #46269 is a broad distributed branch base spanning FSDP, tensor parallel, integrations, and model changes. They overlap in infrastructure files but do not appear to be the same concrete change or a mergeable single PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 31, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46138-6", + "nodes": [ + "pull_request:46138", + "pull_request:46432", + "pull_request:46501", + "pull_request:46502", + "pull_request:46503", + "pull_request:46504" + ], + "soft_pairs": [ + "pull_request:46138|pull_request:46504", + "pull_request:46138|pull_request:46432", + "pull_request:46138|pull_request:46503", + "pull_request:46138|pull_request:46501", + "pull_request:46138|pull_request:46502", + "pull_request:46432|pull_request:46503", + "pull_request:46432|pull_request:46501", + "pull_request:46432|pull_request:46502", + "pull_request:46432|pull_request:46504", + "pull_request:46501|pull_request:46503", + "pull_request:46501|pull_request:46502", + "pull_request:46501|pull_request:46504", + "pull_request:46502|pull_request:46503", + "pull_request:46502|pull_request:46504", + "pull_request:46503|pull_request:46504" + ], + "prepared_review_unit_hash": "4f2b0604cdb15736b4404874325136cce6a36f1ed6a29a12f33a1613f4f474a6", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14226, + "estimated_input_tokens": 3557, + "estimated_eval_tokens": 7370 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14226, + "estimated_input_tokens": 3557, + "estimated_eval_tokens": 7370 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are small, unrelated CI/workflow maintenance PRs. They share a superficial \u201cchore: update workflow\u201d pattern, but they do not fix the same concrete bug or change, so there is no duplicate cluster here.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Both harden shell input handling in workflows, but they affect different jobs/files and different failure modes; not the same underlying change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46432", + "accept": false, + "reason": "One updates stdin handling for a Python helper; the other changes SSH runner setup and token handling. Different code paths and unrelated fixes." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46503", + "accept": false, + "reason": "Benchmark trigger logic is unrelated to the self-comment CI message-passing change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46501", + "accept": false, + "reason": "Permissions restructuring for anti-slop has no overlap with PR comment piping." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46502", + "accept": false, + "reason": "Reviewer checkout ref selection is a different workflow concern than handling PR comment input." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "SSH runner security/setup changes and benchmark dispatch support are separate workflow changes." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "Different workflow files and unrelated permission changes; not the same bug." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Checkout base ref behavior and SSH runner hardening are unrelated." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46504", + "accept": false, + "reason": "Both touch workflow scripts, but one is SSH runner security/setup and the other is commit-message parsing in a Docker-image workflow." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Anti-slop permissions and benchmark workflow dispatch are unrelated changes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "They modify different workflow behaviors with no shared bug or code path." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "No shared underlying issue; one is permissions scoping, the other is shell-safe commit message inspection." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Reviewer assignment checkout ref and benchmark dispatch are unrelated workflow updates." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Different CI jobs and different behavioral fixes; not duplicates." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark dispatch support and commit-message parsing in build-image automation are separate 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": 32, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46152-2", + "nodes": [ + "pull_request:46152", + "pull_request:46207" + ], + "soft_pairs": [ + "pull_request:46152|pull_request:46207" + ], + "prepared_review_unit_hash": "7540af6984132a967ef84606ef296b44a8618f79389a899c45ae632e80a2c917", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4275, + "estimated_input_tokens": 1069, + "estimated_eval_tokens": 2394 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4275, + "estimated_input_tokens": 1069, + "estimated_eval_tokens": 2394 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two unrelated source-reading refactors in different code paths; they share the same general theme but not the same underlying bug/change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46207 is the better canonical PR because it is more concrete and narrowly scoped: it updates the _can_set_attn_implementation source-detection path and includes a test change, making it easier to triage than the broader import-utils refactor.", + "best_issue_reason": null, + "best_pr_reason": "PR 46207 is the strongest overall representative of the cluster because it has a clear, specific behavior change in one code path and a matching test update; PR 46152 is a separate implementation cleanup in another module.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46152", + "right": "pull_request:46207", + "accept": false, + "reason": "Both avoid reading source via open(), but they change different modules and different logic paths. They are similar refactors, 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": 33, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46173-3", + "nodes": [ + "pull_request:46173", + "pull_request:46436", + "pull_request:46474" + ], + "soft_pairs": [ + "pull_request:46436|pull_request:46474", + "pull_request:46173|pull_request:46474" + ], + "prepared_review_unit_hash": "f21f88628ca440bf95d7ecd8614236e429fa557368f6b97d17d2b1104a059a70", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs (46173 and 46474) are effectively the same StaticLayer.get_seq_length typing fix, while 46436 is a separate behavioral change in static-cache setup. I treat 46173/46474 as duplicates and keep 46173 as the canonical representative; 46436 is the strongest standalone PR in the cluster.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46173 is the earlier, slightly more active version of the identical StaticLayer.get_seq_length return-type fix, and it directly targets the linked issue #45987.", + "best_issue_reason": null, + "best_pr_reason": "PR 46436 is the only substantive behavior change in the cluster: it makes generation respect the configured static cache max length instead of only the computed maximum.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46436", + "right": "pull_request:46474", + "accept": false, + "reason": "Different fixes and code paths: 46436 changes generation cache initialization to honor max_cache_len, while 46474 only relaxes a type hint in cache_utils. Same subsystem, but not the same underlying bug/change." + }, + { + "left": "pull_request:46173", + "right": "pull_request:46474", + "accept": true, + "reason": "These are the same concrete change: both update StaticLayer.get_seq_length's type annotation and implementation in cache_utils, with the same linked issue #45987." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet. 46173 and 46474 share the same file, issue target, and identical diff preview, so treating them as duplicates is conservative. 46436 is a separate change in generation cache handling, with different files and a different issue target, so keeping it separate is also well supported." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 34, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46293-2", + "nodes": [ + "pull_request:46293", + "pull_request:46478" + ], + "soft_pairs": [ + "pull_request:46293|pull_request:46478" + ], + "prepared_review_unit_hash": "7bd39ae48d659fdee338d62b65f52e0c5fa39368b02b874770cb97e7816d0b13", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4443, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4443, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs target the same hub-kernels area and share issue #46291, but they implement different fixes: one updates kernel repo/version metadata for kernels>=0.15.0, while the other pins LayerRepository revisions to main and adds broader config/test changes. They are related, but not the same duplicate change.", + "confidence": 0.82, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46478 is the better canonical PR because it is broader and more complete: it updates multiple hub-kernel references, touches kernel config, and includes test coverage, making it the stronger representative of the fix set.", + "best_issue_reason": null, + "best_pr_reason": "PR #46478 is the best overall PR in the cluster because it addresses the problem more comprehensively and is more suitable as the merge target for the shared hub-kernel compatibility work.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46293", + "right": "pull_request:46478", + "accept": false, + "reason": "Both mention the same issue and file, but the concrete changes differ: #46293 adjusts repo/version compatibility metadata, while #46478 pins revisions to main and extends config/tests. They are related fixes, not the same PR-worthy code change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 35, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46376-2", + "nodes": [ + "pull_request:46376", + "pull_request:46417" + ], + "soft_pairs": [ + "pull_request:46376|pull_request:46417" + ], + "prepared_review_unit_hash": "9cf1417a8ad023e3fbc849f514b594650edca756cbdef0664d1f19d2beb4f090", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5938, + "estimated_input_tokens": 1485, + "estimated_eval_tokens": 3226 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5938, + "estimated_input_tokens": 1485, + "estimated_eval_tokens": 3226 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are unrelated: one adds multimodal support to continuous batching infrastructure, while the other introduces the HunYuan VL model, docs, processors, and tests. They share only a broad ML/transformers context, not the same concrete change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46417 is the stronger representative of the cluster because it is a complete, self-contained model addition with code, docs, and tests; PR 46376 is a draft infrastructure feature that is not comparable as a duplicate candidate.", + "best_issue_reason": null, + "best_pr_reason": "PR 46417 is the best standalone PR in this set: it has a clear deliverable, broader implementation coverage, and appears more complete than the draft multimodal batching PR.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46376", + "right": "pull_request:46417", + "accept": false, + "reason": "Different changes and code paths: continuous batching multimodal support vs. adding the HunYuan VL model. They are not the same underlying bug or change, and would not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/manifest.json b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..051eb3d3fdd9a57e869120bcd21a6f7c9639f84c --- /dev/null +++ b/snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260610t120053z", + "artifacts": { + "hybrid": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-06-10T12:12:03Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260610T120053Z", + "variant": "hybrid" +} diff --git a/snapshots/20260610T120053Z/manifest.json b/snapshots/20260610T120053Z/manifest.json index 799dc6ea4e5316b967b9b412198e75cf35433cd8..27a7e686c7aa1e6fa61f828705faea45b7411ab2 100644 --- a/snapshots/20260610T120053Z/manifest.json +++ b/snapshots/20260610T120053Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 452 }, "extracted_at": "2026-06-10T12:00:53Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260610t120053z", + "runs": { + "hybrid-model-20260610t120053z": { + "analysis_id": "hybrid-model-20260610t120053z", + "artifacts": { + "hybrid": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260610T120053Z/analysis-runs/hybrid-model-20260610t120053z/manifest.json", + "model": null, + "published_at": "2026-06-10T12:12:03Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260610T120053Z", "watermark": {