diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 93022329fea4c5317dc14844838b520d955c7213..9a00bebd6f4931de9793a7ab5797c043214097ce 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": "20260531T060031Z", - "generated_at": "2026-05-31T06:08:06Z", + "snapshot_id": "20260531T120035Z", + "generated_at": "2026-05-31T12:08:28Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 3cd661806a6ca6a2b060afcdc84a4988dbd8746d..d4e25408a4254faad2e57a331a89b8a99f88004d 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": "20260531T060031Z", - "generated_at": "2026-05-31T06:08:06Z", + "snapshot_id": "20260531T120035Z", + "generated_at": "2026-05-31T12:08:28Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -10,13 +10,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 1, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:24643", "issue:30277", @@ -121,7 +121,7 @@ "issue:43827|issue:44512", "issue:43716|issue:44560", "issue:43828|issue:45685", - "issue:44560|issue:45685", + "issue:44918|issue:45685", "issue:43039|issue:43828", "issue:30277|issue:36331", "issue:34928|issue:36010", @@ -129,27 +129,27 @@ "issue:43039|issue:44805", "issue:45800|issue:46304", "issue:43296|issue:46304", - "issue:41762|issue:43596", "issue:39290|issue:41720", + "issue:41762|issue:43596", "issue:43901|issue:44512", "issue:44589|issue:46304", "issue:45823|issue:46294", "issue:45237|issue:45901" ], - "prepared_review_unit_hash": "8c18238b7e82d3ce7cbc219514d6739641f057b404485e9c8cb2c3f20b0759e4", + "prepared_review_unit_hash": "0b998587bd77bb9943d0978f3fced78f0b8e4663d5802295c37268c7f034d96f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 67, - "serialized_chars": 47822, + "serialized_chars": 47821, "estimated_input_tokens": 11956, "estimated_eval_tokens": 24168 }, @@ -158,384 +158,402 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are unrelated. The only clear duplicate cluster is the DeepSpeed ZeRO-3 index-out-of-bounds reports around Bert/Gemma3 model loading; the docs, dtype-mismatch, MoE, and API issues are distinct.", - "confidence": 0.82, - "canonical_issue_reason": "Issue 43638 is the clearest/broadest phrasing of the shared ZeRO-3 loading bug, with the same \"index 0 out of bounds for dimension 0 with size 0\" failure pattern as the other Bert/Gemma3 reports.", + "summary": "This cluster is mostly a loose thematic group of unrelated bug reports; only the two BERT + DeepSpeed ZeRO-3 index-out-of-bounds reports look like true duplicates.", + "confidence": 0.58, + "canonical_issue_reason": "issue:43638 is the clearest representative of the only true duplicate subgroup here: BERT + DeepSpeed ZeRO-3 failing with an index-0-out-of-bounds crash during model init/training.", "canonical_pr_reason": null, - "best_issue_reason": "43638 is the best representative issue because it states the concrete failure mode and scenario most clearly while matching the other ZeRO-3 index-error reports.", + "best_issue_reason": "There is no strong global canonical for the whole cluster because the items span many unrelated models, configs, and docs issues; issue:43638 is just the best anchor for the one genuinely duplicated bug family.", "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 API behavior." + "reason": "Different failure modes: missing Qwen3.5 config attribute vs streaming chat/completions API error." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Different bugs: AMD GPU inference failure vs MPS histogram op on Int tensors." + "reason": "Unrelated hardware/backend problems: AMD GPU support vs MPS histogram kernel on Int." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Both are model/runtime failures, but the triggers and error classes differ." + "reason": "Different models and error paths; image preprocessor dtype mismatch is unrelated to AMD GPU execution failure." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Different concrete dtype-mismatch path and model; not the same bug." + "reason": "Autocast dtype mismatch on Phi MoE is a different bug than GPT-OSS failing on AMD GPUs." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Unrelated model failures with no shared code-path indicated by the titles." + "reason": "Different model families and symptoms; AMD GPU issue vs invalid method call in GraniteMoEHybrid." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Different model-specific runtime problems; no evidence of the same underlying bug." + "reason": "Qwen-3 NVFP4 runtime error is a different code path from AMD GPU incompatibility." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch and ONNX export error are unrelated." + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ModernBERT ONNX export failure." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "ONNX export failure vs FP8 config failure are different issues." + "reason": "ONNX export failure and FineGrainedFP8Config failure are different model/backend issues." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "Different models and error modes; not the same bug." + "reason": "Export-to-ONNX problem is unrelated to Qwen3 auto device mapping cudaErrorAssert." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Both are issue-style reports, but one is a task crash and the other is outdated docs/examples." + "reason": "Task crash vs outdated RAG docs; same general area, not the same bug or change." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions, but they concern different removed features." + "reason": "Both are docs cleanup items, but they target different removed features and would not be one merged fix." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export error vs missing GenerationMixin import are unrelated." + "reason": "ModernBERT ONNX export and GenerationMixin import regression are unrelated." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "TRL SFT input-embedding unpacking and resume-training trust-remote-code are different problems." + "reason": "TRL SFT embedding unpacking vs resume-training with trust_remote_code are different training failures." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Mask-shape mismatch and device-mapping mismatch are different failures." + "reason": "Mask shape mismatch during indexing is unrelated to Gemma4 PLE device placement mismatch." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Different runtime errors: zero-size index error vs mask/tensor shape mismatch." + "reason": "Both involve indexing errors, but the reported symptoms and contexts differ too much to be confidently the same bug." }, { "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "Different problem domains and error signatures." + "reason": "ZeRO-3 BERT crash is unrelated to resuming training with trust_remote_code." }, { - "left": "issue:43638", + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Mask/index shape mismatch is unrelated to Qwen3.5 video feature/token mismatch." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Similar generic indexing symptom, but different trigger and message; not enough to treat as one bug." + }, + { + "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Zero-3 loading IndexError vs TP-only MoE issue are not the same bug." + "reason": "Indexing/mask mismatch is unrelated to GPT-OSS tensor-parallel-only failure." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video chat-template mismatch and training-resume issue are unrelated." + "reason": "Video chat-template StopIteration is unrelated to training resume behavior." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "ZeRO-3 load IndexError vs resume/trust-remote-code issue are different." + "reason": "ZeRO-3 initialization crash is unrelated to local model resume/trust_remote_code." }, { "left": "issue:43572", "right": "issue:46165", "accept": false, - "reason": "Config field regression vs training-resume issue are unrelated." + "reason": "Missing pad_token_idx after v5 update is unrelated to training resume failure." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice preset bug and ModernBERT ONNX export bug are unrelated." + "reason": "BarkProcessor voice_preset bug and ModernBERT export bug are unrelated." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both involve DeepSpeed, but one is collective mismatch and the other is a zero-size index error." + "reason": "Both touch DeepSpeed, but one is collective mismatch in Jamba-v01 and the other is BERT index-out-of-bounds; different bugs." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 loading IndexError ('size 0'/'index 0 out of bounds') on Bert models; same underlying bug." + "reason": "Near-identical ZeRO-3 BERT index-out-of-bounds reports; same underlying initialization/training bug and plausible single fix." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Incorrect detokenization and LoRA merge/save mismatch are different issues." + "reason": "Different model-specific tokenizer/ detokenization issue vs LoRA merge/save/reload mismatch." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger cross-entropy path and TP-only MoE issue are different code paths." + "reason": "Liger Kernel cross_entropy invocation is unrelated to GPT-OSS tensor-parallel failure." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Unrelated model/runtime failure vs LoRA merge/save inconsistency." + "reason": "Qwen-3 NVFP4 runtime error is unrelated to LoRA merge/save behavior." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Different DeepSpeed failures: weight shape error vs collective mismatch." + "reason": "Different DeepSpeed-related failures: 2-D weight error vs collective mismatch." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation-checkpointing/FSDP tensor recomputation and mask-shape mismatch are different bugs." + "reason": "Activation-checkpointing/FSDP tensor recomputation bug is unrelated to mask shape mismatch." }, { "left": "issue:45859", "right": "issue:45902", "accept": false, - "reason": "Same model component, but different concrete failures: CPU offload incompatibility vs uninitialized memory after materialization." + "reason": "CPU offload incompatibility in rotary embedding vs uninitialized memory after materialization are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "BERT ZeRO-3 index crash is unrelated to GPT-OSS only-TP failure." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "FSDP/activation checkpointing shape issue is unrelated to trainer custom compute_loss signature change." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 BERT init crash is unrelated to GPT-OSS tensor-parallel-only failure." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Mistral-3 image dtype mismatch is unrelated to GraniteMoEHybrid invalid method call." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Different Gemma4 input/embed behavior vs Qwen3.5 video template mismatch." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3 config attribute error and Mistral regex tokenizer crash are unrelated." + "reason": "Gemma3/vLLM config attribute error is unrelated to tokenizer regex patch crash." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Device-map forward mismatch and expensive reverse embedding lookup are distinct issues." + "reason": "Device-map forward mismatch and expensive reverse embedding lookup are different Gemma4 issues." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch/version incompatibility and Qwen-3-8B NVFP4 runtime failure are unrelated." + "reason": "Torch 2.10 compatibility report is unrelated to Qwen-3 NVFP4 runtime failure." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP checkpointing mismatch is unrelated to Qwen3 auto device mapping assertion failure." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Both mention model execution problems, but the dtype-mismatch autocast bug is not the TP-only MoE bug." + "reason": "Phi autocast dtype mismatch is unrelated to GPT-OSS tensor-parallel-only failure." }, { - "left": "issue:43541", - "right": "issue:43716", + "left": "issue:44918", + "right": "issue:45161", "accept": false, - "reason": "Grouped_mm tracing failure and image-preprocessor dtype mismatch are different problems." + "reason": "TRL SFT embedding issue is unrelated to GPT-OSS tensor-parallel-only failure." }, { - "left": "issue:43828", + "left": "issue:43716", "right": "issue:44292", "accept": false, - "reason": "Different models and error classes; not the same bug." + "reason": "Different models and symptoms: Mistral image dtype mismatch vs Qwen NVFP4 runtime error." }, { - "left": "issue:43716", - "right": "issue:45685", + "left": "issue:44292", + "right": "issue:44918", "accept": false, - "reason": "Dtype mismatch in multimodal inference vs MPS histogram op limitation are unrelated." + "reason": "Qwen NVFP4 runtime error is unrelated to TRL SFT embedding unpacking." }, { - "left": "issue:44918", - "right": "issue:45161", + "left": "issue:45507", + "right": "issue:45685", "accept": false, - "reason": "TRL SFT input-embedding unpacking and GPT-OSS TP-only failure are different issues." + "reason": "Invalid method call in GraniteMoEHybrid is unrelated to MPS histogram kernel limitation." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger kernel cross-entropy call and TRL trainer embedding handling are unrelated." + "reason": "Liger Kernel cross_entropy being called is unrelated to TRL SFT embedding unpacking." }, { "left": "issue:41762", "right": "issue:43638", - "accept": true, - "reason": "Same ZeRO-3 loading IndexError class with the same empty-tensor/out-of-bounds failure pattern." + "accept": false, + "reason": "Both are ZeRO-3 load/init failures, but the models and reported symptoms differ enough that they don't look like the same bug." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs regressions, but they refer to different removed v5 commands/tasks." + "reason": "Both are docs regressions, but they concern different removed features and likely separate cleanup edits." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Dtype mismatch and video StopIteration are different bugs." + "reason": "Mistral image dtype mismatch is unrelated to Qwen3-vl video StopIteration." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Autocast dtype mismatch and MPS histogram Int limitation are unrelated." + "reason": "Autocast dtype mismatch is unrelated to MPS histogram kernel support." }, { - "left": "issue:44560", + "left": "issue:44918", "right": "issue:45685", "accept": false, - "reason": "Video feature/token mismatch and MPS op support issue are unrelated." + "reason": "TRl SFT embedding unpacking is unrelated to MPS histogram kernel support." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Different model path and failure mode; both are not the same concrete bug." + "reason": "Liger Kernel cross_entropy invocation is unrelated to Phi autocast dtype mismatch." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "DeepSpeed collective mismatch and custom trainer kwarg error are unrelated." + "reason": "DeepSpeed collective mismatch and trainer compute_loss signature mismatch are unrelated." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "FSDP/checkpointing recomputation bug vs missing GenerationMixin import are unrelated." + "reason": "FSDP activation checkpointing issue is unrelated to GenerationMixin import regression." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "AMD GPU inference failure and outdated RAG examples are unrelated." + "reason": "AMD GPU runtime failure is unrelated to outdated RAG docs." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger kernel cross-entropy issue and mask/tensor shape mismatch are different bugs." + "reason": "Liger Kernel cross_entropy call is unrelated to mask shape mismatch." }, { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Torch/Transformers version incompatibility and USE_DISTRIBUTED=0 support regression are different issues." + "reason": "Version incompatibility report and USE_DISTRIBUTED=0 support regression are different compatibility problems." }, { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "PaddleOCR-VL vLLM load failure and distributed-build support regression are unrelated." - }, - { - "left": "issue:41762", - "right": "issue:43596", - "accept": true, - "reason": "Same ZeRO-3 empty-tensor IndexError during model loading; same underlying bug." + "reason": "PaddleOCR-VL/vLLM load failure is unrelated to pytorch distributed-disabled support breakage." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Gemma3 config attribute regression and Qwen3 auto device-map cuda assert are unrelated." + "reason": "Gemma3/vLLM config attribute error is unrelated to Qwen3 auto device mapping assertion failure." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Both are ZeRO-3 index-out-of-bounds reports, but they involve different models and likely separate load paths." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "TextClassificationPipeline docs issue and removed run-command docs issue are unrelated." + "reason": "TextClassificationPipeline docs wording and run-command docs removal are separate documentation issues." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Float8 storage lookup failure and USE_DISTRIBUTED=0 support regression are different problems." + "reason": "Float8 storage lookup failure is unrelated to distributed-disabled support breakage." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Gemma4 device mismatch and Qwen3.5 video chat-template mismatch are unrelated." + "reason": "Gemma4 device-map mismatch and Qwen3.5 video token mismatch are unrelated." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "GPT-OSS AMD GPU failure and table QA crash are unrelated." - }, - { - "left": "issue:44805", - "right": "issue:45874", - "accept": false, - "reason": "Mask-shape mismatch and reverse embedding lookup cost bug are different." - }, - { - "left": "issue:45507", - "right": "issue:45685", - "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid and MPS histogram Int limitation are unrelated." - }, - { - "left": "issue:45874", - "right": "issue:46294", - "accept": false, - "reason": "Reverse embedding lookup regression and video token/template mismatch are different bugs." - }, - { - "left": "issue:43596", - "right": "issue:43638", - "accept": true, - "reason": "Same ZeRO-3 empty-tensor IndexError during Bert model loading; same underlying bug." - }, - { - "left": "issue:44805", - "right": "issue:46294", - "accept": false, - "reason": "Mask shape mismatch and Qwen3.5 video token mismatch are unrelated." + "reason": "AMD GPU failure is unrelated to table-question-answering crash." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well-grounded in the packet, and the ZeRO-3 duplicate cluster is supported by the issue titles. The soft-edge verdicts are conservative overall, with the true positives limited to the near-identical IndexError/ZeRO-3 Bert/Gemma3 reports." + "feedback": "The summary is well grounded in the packet: this is mostly a broad thematic cluster with no single global canonical, and the only clearly duplicate subgroup is the ZeRO-3 BERT index-out-of-bounds pair (issue:43596 / issue:43638). The soft-edge verdicts are generally conservative and avoid overstating similarity." }, "evaluator_used": true, "retried": false, @@ -548,13 +566,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 2, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36010", "issue:39290", @@ -648,295 +666,301 @@ "issue:43541|issue:44589", "issue:41093|issue:43596" ], - "prepared_review_unit_hash": "9857738fa958e142a5b498275875ed6c3ba4725c422b88c5241711144cb060e2", + "prepared_review_unit_hash": "1db38b5924f5634c69bf9409335b40fb936f1b76f9e8015d8514ce9b90f0dd90", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 42, - "serialized_chars": 39562, - "estimated_input_tokens": 9891, - "estimated_eval_tokens": 20038 + "serialized_chars": 39559, + "estimated_input_tokens": 9890, + "estimated_eval_tokens": 20036 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly noisy and spans many unrelated regressions. The strongest duplicate-like subgroup is the AutoTokenizer v5 token-ID regression, and there is also a plausible Qwen video-template regression pair. The ZeRO-3/Bert loading pair is too weak to treat as the same underlying defect from the evidence here.", - "confidence": 0.57, - "canonical_issue_reason": "Issue 45920 is the best canonical issue because it is the broadest and most representative instance of the clearest real duplicate subgroup: a v5 AutoTokenizer token-ID regression affecting multiple model families.", + "summary": "Most items are unrelated model-specific regressions or docs/config issues. I only found two clear duplicate pairs: the v5 AutoTokenizer wrong-token-ID regression and the Qwen3.5 video/chat-template regression.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45920 is the best representative of the duplicate tokenizer-regression subgroup: it is broader than the Granite-specific report, remains open, and describes the same v5 wrong-token-ID failure across multiple model families.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45920 is the best overall issue candidate: open, broad, and directly representative of the clearest duplicate cluster in this packet.", + "best_issue_reason": "issue:45920 is the strongest global issue anchor among the accepted duplicates because it captures the widest version-wide tokenizer regression with multiple affected models.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Different failure modes and model families: Voxtral processor loading vs Gemma4 device-map mismatch." + "reason": "Different model families and different failures: processor loading vs device mismatch during forward." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Both are inference bugs, but the concrete paths differ: Liger Kernel cross-entropy dispatch vs Qwen3-VL video StopIteration." + "reason": "Unrelated subsystems: Liger kernel cross-entropy path vs Qwen video StopIteration." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Detokenization bug vs outdated RAG examples; unrelated symptoms and code paths." + "reason": "Different problems: detokenization bug vs RAG example documentation freshness." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Same underlying regression class: AutoTokenizer returns wrong token IDs in v5; 45920 is the broader version across more model families." + "reason": "Both report the same v5 AutoTokenizer regression producing wrong token IDs across model families; 45920 is the broader tracker." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "Checkpoint-loading failure for NemotronH vs documentation/examples issue; no shared bug." + "reason": "Checkpoint-loading bug vs outdated RAG examples; not the same underlying issue." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Both are version regressions, but one is remote_code breakage and the other is USE_DISTRIBUTED=0 PyTorch support; different code paths." + "reason": "Broad remote_code breakage vs a specific USE_DISTRIBUTED=0/pytorch compatibility regression." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Corrupted Ovis2 checkpoint vs NemotronH checkpoint-loading implementation issue; not the same defect." + "reason": "Both are load failures, but for different models and different root causes." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "Model-loading TypeError from init_empty_weights/_is_hf_initialized vs examples/RAG content issue." + "reason": "Model-loading error vs documentation issue; no shared code-path bug." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes Int8Params incompatibility vs distributed PyTorch support regression; unrelated." + "reason": "bitsandbytes Int8Params constructor change is unrelated to distributed build support." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding_window, but one is a missing config attribute in vLLM and the other is a Qwen3-MoE model issue; not clearly the same bug." + "reason": "Both mention sliding_window, but one is a missing config attribute in vLLM and the other is a model-specific sliding_window issue." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to the Gemma3 sliding_window_pattern problem." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "Different model-loading failures for different checkpoints; no common underlying change." + "reason": "Different models and different load failures; no evidence of the same bug." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are ZeRO-3/Bert IndexError reports, but the failure setups and concrete symptoms differ enough that the shared-bug claim is too speculative." + "reason": "Both are IndexErrors, but the reported shapes and contexts differ; not enough to treat as the same bug." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper model load/run failure vs table-question-answering task crash; unrelated." + "reason": "Whisper model failure vs table-question-answering crash; unrelated." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch compatibility complaint vs pydantic PretrainedConfig regression; different issues." + "reason": "Torch version incompatibility is not the same as the PretrainedConfig pydantic-field regression." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "Both mention versioned incompatibilities, but one is ZeRO-3/Bert IndexError and the other is torch/transformers incompatibility; not the same bug." + "reason": "Different failures: ZeRO-3/Bert index error vs torch 2.4.1 and transformers 5.8.0 incompatibility." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Detokenization regression vs table QA crash; no shared code path." + "reason": "Unrelated features and error paths." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper runtime failure vs PretrainedConfig/pydantic regression; unrelated." + "reason": "Different model/runtime bug vs config-model schema break." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Both are Qwen3.5 chat-related, but one is a crash with transformers chat and the other is streaming chat/completions API failure; too little evidence of the same bug." + "reason": "Both involve Qwen3.5, but the reported failures are different enough that they do not look like one concrete bug." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "Model initialization TypeError vs table-question-answering crash; different problems." + "reason": "Qwen-3-8B-NVFP4 loading error is unrelated to table QA crashes." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Sliding-window config issue vs Gemma4 device mismatch; not the same underlying defect." + "reason": "Gemma3 sliding_window_pattern issue and Gemma4 device mismatch are different bugs." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Tied-weights bug in Mistral3 vs corrupted Ovis2 checkpoint; unrelated." + "reason": "Weight-tying bug in Mistral3 is unrelated to a corrupted Ovis2 checkpoint." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 streaming API failure vs MPS histogram Int error in MoE; different code paths." + "reason": "Chat/completions streaming failure vs MPS histogram_mps limitation; different code paths." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "Grouped_mm tracing failure vs MPS histogram error; not the same bug." + "reason": "Torch dynamo grouped_mm tracing error is unrelated to the MPS histogram issue." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both are dtype-related MoE issues, but one is torchdynamo grouped_mm tracing and the other is autocast mismatch; not clearly duplicate." + "reason": "Both are MoE/dtype-adjacent, but the concrete failures differ and do not look like one bug." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure for Voxtral vs compiler/template node error; unrelated." + "reason": "Processor loading failure and template-node compilation error are unrelated." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "ImportError for GenerationMixin vs Gemma3 sliding_window attribute error; different issues." + "reason": "Same as above: import error vs config attribute regression." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch/transformers compatibility vs Gemma4 device mismatch; not the same bug." + "reason": "General version incompatibility vs a specific device-map forward mismatch." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are ZeRO-3 loading errors, but the reported failures differ materially and involve different models/code paths." + "reason": "Both are IndexErrors, but one is a mask/tensor shape mismatch and the other is a ZeRO-3 zero-size index failure." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "PaddleOCR-VL loading failure vs GGUF gpt-oss support request; unrelated." + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are separate issues." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig vs pydantic PretrainedConfig regression; different defects." + "reason": "Missing StableLmConfig pad_token_idx and PretrainedConfig pydantic break are distinct config regressions." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Grouped_mm tracing failure vs GPT-OSS MoE tensor-parallel issue; no concrete overlap." + "reason": "Grouped_mm tracing error vs TP-only GPT-OSS MoE support issue; not the same problem." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Int4 quantization OOM vs tied-weights translation-model regression; unrelated." + "reason": "Memory/OOM regression under int4 quantization vs tied-weights translation model bug." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Mask shape mismatch and SFT embedding unpacking both affect sequence handling, but the symptoms and likely code paths are too different to treat as the same bug." + "reason": "Mask-shape IndexError vs SFT trainer embedding-unpacking failure; different bugs." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Phi-tiny-MoE autocast dtype mismatch vs Qwen3.5 streaming API failure; unrelated." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to Qwen3.5 streaming chat failure." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "Unexpected _is_hf_initialized argument in model loading vs pydantic PretrainedConfig breakage; different changes." + "reason": "Unexpected _is_hf_initialized argument and pydantic PretrainedConfig break are different load/config issues." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF gpt-oss support request vs Qwen3.5 streaming inference bug; unrelated." + "reason": "gpt-oss support and Qwen3.5 chat streaming failure are unrelated." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Mistral-3 image preprocessor dtype mismatch vs Qwen3.5 video token/chat-template mismatch; different modalities and code paths." + "reason": "Different multimodal regressions: image-preprocessor dtype mismatch vs video token/feature mismatch." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding_window issue vs PretrainedConfig/pydantic regression; not the same bug." + "reason": "Sliding_window issue in Qwen3-MoE is not the same as the PretrainedConfig schema break." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote_code regression and torch 2.4.1 compatibility complaint are both version-related but distinct." + "reason": "Remote_code breakage and torch 2.4.1 compatibility problems are different versions of different failures." }, { "left": "issue:44479", "right": "issue:46294", "accept": true, - "reason": "Same Qwen video-input regression family: both report mismatches in video handling after v5.3.0, likely the same underlying template/token accounting bug." + "reason": "Both describe the same Qwen3.5 video-input regression where apply_chat_template/video tokens and features no longer align in v5.3.x." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "MoE grouped_mm tracing issue vs Float8 storage lookup error; unrelated." + "reason": "Torch dynamo grouped_mm tracing error is unrelated to the Float8 storage lookup failure." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are IndexError reports, but the error shapes and affected setups differ; not enough to treat them as duplicates." + "reason": "Different IndexError signatures and contexts; not the same underlying bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The summary is mostly grounded, and the tokenizer and Qwen video pairs look reasonable. However, the accepted 41093\u219443638 edge is too strong for the evidence in the packet: the titles share DeepSpeed ZeRO-3/Bert themes, but the failure modes and setups differ enough that calling them the same underlying defect overstates the evidence. Because the soft-edge verdicts are not all conservative, I can\u2019t accept the analysis." + "accept": true, + "feedback": "Grounded and conservative overall. The two accepted soft pairs are well supported by the packet: 45812/45920 are the same v5 AutoTokenizer wrong-token-ID regression, and 44479/46294 are the same Qwen3.5 video/apply_chat_template mismatch. The negative verdicts also stay cautious and do not overclaim shared root causes from superficial similarity." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -946,13 +970,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 3, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36331", "issue:38175", @@ -1037,232 +1061,232 @@ "issue:44910|issue:45859", "issue:42947|issue:44387" ], - "prepared_review_unit_hash": "644d96084aadd9b4dc5ed8e7dce4710423db2e829f34ae617f99c51e8c99dd96", + "prepared_review_unit_hash": "293895915564a67f82a9e16aff9bf79c5e8cb58299cc160a124c3c358e99e5a8", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37138, - "estimated_input_tokens": 9285, - "estimated_eval_tokens": 18826 + "serialized_chars": 37144, + "estimated_input_tokens": 9286, + "estimated_eval_tokens": 18828 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most candidate pairs are unrelated and should stay separate. The only clear duplicate cluster is the Transformers v5 non-persistent buffer regression, where several reports describe the same underlying buffer-initialization bug in different forms.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43950 is the strongest representative of the non-persistent-buffer regression: it clearly describes the v5 from_pretrained corruption path and is broad enough to subsume the narrower follow-up reports.", + "summary": "The cluster is mostly a set of unrelated Transformers issues with only one clear duplicate thread: non-persistent buffers being corrupted/junk-filled during `from_pretrained` in v5.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43950 is the clearest canonical issue: it states the regression precisely (`from_pretrained()` corrupts non-persistent buffers) and matches the closest duplicate reports about the same junk-filled buffer behavior.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43950 is the best overall issue candidate because it captures the core regression cleanly and with enough detail to serve as the canonical bug report.", + "best_issue_reason": "issue:43950 is the best representative issue because it is specific, regression-focused, and central to the only confirmed duplicate pair in this set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Both involve SigLIP2, but one is about degraded text embeddings and the other about bad AutoModel/pipeline outputs; not the same concrete bug." + "reason": "Both are about SigLIP2, but one is about worse text embeddings and the other about nonsensical pipeline/AutoModel results; the root cause is not clearly the same." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Different failures: composite multimodal config unwrapping vs uninitialized rotary embedding buffer after to_empty materialization." + "reason": "Both involve Qwen3.5/Qwen 3.6 internals, but one is about `text_config` unwrapping in `from_config` and the other about uninitialized rotary frequencies after materialization; different bugs." }, { "left": "issue:45542", "right": "issue:45800", "accept": false, - "reason": "Both are version/incompatibility complaints, but they point to different mechanisms and fixes." + "reason": "Different problems: TensorBoard/TensorFlow backend detection versus a torch/transformers version incompatibility." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Different models and paths: image preprocessor dtype mismatch vs Qwen3.5 embedding unpacking with TRL." + "reason": "Different model families and failure modes: image preprocessor dtype mismatch versus Qwen3.5 embedding unpacking with TRL SFT trainer." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP weight-shape error is unrelated to the Qwen3.5 embedding/TRL issue." + "reason": "FSDP `'weight' must be 2-D` and TRL embedding unpacking are unrelated failures." }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Both concern loading custom code, but one is cache-path collision and the other is wrong module resolution after save_pretrained." + "reason": "Both concern loading custom code, but one is a trust_remote_code cache-path collision and the other is loading the wrong custom module after save/load; too different to merge." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model initialization and flash-attn fallback logic are unrelated bugs." + "reason": "Notebook custom-model initialization and flash-attn fallback gating are unrelated code paths." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Both mention tied weights, but they affect different models and code paths." + "reason": "Both mention tied weights, but one is a specific Mistral3 lm_head tying bug while the other is a broader translation-model v5 issue; not the same concrete bug." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "High perplexity on gpt-oss and a GLM5 issue are unrelated." + "reason": "Perplexity on gpt-oss-20b and a GLM5 issue are unrelated." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache group-type failure is unrelated to tied-weights handling." + "reason": "PagedAttentionCache/Qwen3.5 crash is a separate cache-kernel issue, not the tied-weights regression." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "Qwen3Moe FP8 failure and flash-attn fallback gating are different bugs." + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn fallback checks." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to tied-weights regression." + "reason": "Quantized offloading behavior and tied-weights issues are different bugs." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "DSA indexer/ReLU discussion is unrelated to a docs-command removal issue." + "reason": "DSA ReLU discussion and a docs typo are unrelated." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading and translation-model tied weights are not the same underlying issue." + "reason": "NemotronH checkpoint loading and tied weights on translation models are different issues." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Qwen3.5 embedding unpacking with TRL is unrelated to GraniteMoEHybrid calling an invalid method." + "reason": "Qwen3.5 embedding unpacking with TRL and GraniteMoEHybrid invalid method calls are unrelated." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "Cache-strategy typo and removed run-command docs are separate documentation issues." + "reason": "A cache-strategy typo and a docs command removal note are not the same bug." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "DeepSeek detokenization and Qwen-3 NVFP4 runtime failure are different bugs." + "reason": "Incorrect detokenization and Qwen-3-8B-NVFP4 runtime errors are different failures." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "A model runtime error and an invalid-method call are unrelated." + "reason": "Different models and different failure modes." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "Both report the same Transformers v5 non-persistent-buffer corruption regression." + "reason": "Same underlying regression: v5 `from_pretrained` corrupts/fills non-persistent buffers with junk." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "SigLIP2 doc-example errors and low-quality SigLIP2 embeddings are related only by model family, not the same bug." + "reason": "SigLIP2 docs example errors versus general poor SigLIP2 text embeddings; not the same concrete issue." }, { "left": "issue:43644", "right": "issue:43950", "accept": true, - "reason": "Both describe the same v5 non-persistent-buffer junk/corruption problem." + "reason": "Direct duplicate of the non-persistent buffer corruption/junk-filled buffer regression." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Autocast dtype mismatch on Phi-tiny-MoE is not the same as Qwen3.5 SDPA/BF16 gradient-explosion behavior." + "reason": "Both mention dtype/BF16 behavior, but the actual failures are different model/training paths." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "CustomTrainer kwarg mismatch and ZeRO-3 loading index error are unrelated." + "reason": "CustomTrainer keyword-argument mismatch and DeepSpeed ZeRO-3 load-time index error are unrelated." }, { "left": "issue:43950", "right": "issue:45902", - "accept": true, - "reason": "Both are manifestations of the same non-persistent-buffer initialization bug after materialization/loading." + "accept": false, + "reason": "Both touch buffer/materialization behavior, but one is buffer corruption in `from_pretrained` and the other is uninitialized rotary frequency memory after `to_empty`; not the same bug." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Both mention bitsandbytes/storage compatibility, but they are different missing-kwarg vs missing-storage-object failures." + "reason": "Different quantization/storage incompatibilities: bitsandbytes `Int8Params` vs missing Float8 storage." }, { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "Checkpoint corruption and loading noise are not the same issue." + "reason": "Checkpoint corruption and loading noise are unrelated." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Both involve SigLIP2 behavior, but one is zero probabilities and the other is degraded text embedding quality; not a clear duplicate." + "reason": "Both concern SigLIP2, but the reported symptoms differ and no shared concrete bug is evident." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Bad AutoTokenizer error messaging and missing all_tied_weights_keys are unrelated." + "reason": "Bad AutoTokenizer error messaging and missing `all_tied_weights_keys` are unrelated." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding quality and tokenization changes across versions are different concerns." + "reason": "SigLIP2 embedding quality and tokenization changes across transformers versions are different issues." }, { "left": "issue:45663", "right": "issue:45902", "accept": false, - "reason": "Gemma4 FSDP2 KeyError and Qwen3_5Moe uninitialized buffer reads are unrelated." + "reason": "Gemma4 FSDP2 shared_kv_states bug and Qwen3.5 rotary buffer materialization bug are unrelated." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "FlashAttention crash from 3D position_ids is different from CPU-offload incompatibility." + "reason": "Qwen3.5 flash-attention position_ids crash and CPU-offload incompatibility in rotary embedding are different code paths." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "Gradient checkpointing/LoRA and int4 OOM are separate performance/memory issues." + "reason": "Gradient checkpointing/LoRA behavior and int4 quantization memory growth are different problems." } ] }, "evaluator_result": { "accept": true, - "feedback": "The analyst output is well grounded in the packet and stays conservative on the soft-edge pairs. The only accepted cluster is the Transformers v5 non-persistent-buffer regression, which is supported by the issue titles/descriptions. The other pair rejections are reasonably cautious and do not overclaim duplicates." + "feedback": "The summary is well grounded in the packet. The accepted duplicate thread around non-persistent buffer corruption in `from_pretrained` is supported by the issue titles and the soft-pair evidence, and the other soft-edge rejections read conservative rather than overstated. Minor note: the duplicate cluster is really a three-issue thread (`43644`, `43950`, `44534`), not just a pair, but the overall conclusion is still correct." }, "evaluator_used": true, "retried": false, @@ -1275,13 +1299,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 4, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36331", "issue:41093", @@ -1323,11 +1347,11 @@ "issue:44991", "issue:45020", "issue:45081", - "issue:45237", "issue:45305", "issue:45313", "issue:45356", "issue:45399", + "issue:45520", "issue:45685", "issue:45823", "issue:45923" @@ -1335,7 +1359,7 @@ "soft_pairs": [ "issue:41553|issue:43881", "issue:44403|issue:44451", - "issue:43975|issue:45237", + "issue:43975|issue:45520", "issue:43054|issue:43493", "issue:44403|issue:44488", "issue:42947|issue:43856", @@ -1362,245 +1386,251 @@ "issue:42947|issue:45923", "issue:43940|issue:45313", "issue:36331|issue:41093", - "issue:43854|issue:44863", "issue:41553|issue:44991", + "issue:43854|issue:44863", "issue:43257|issue:45305" ], - "prepared_review_unit_hash": "d97d41fdc1c461470d1ec12d3644fe9277aebc90f69b50668ee0622d6d762441", + "prepared_review_unit_hash": "0431bc28b08722a79abfa998fa381e1722cc01cfba4a64123ec7efd8414d380f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37399, - "estimated_input_tokens": 9350, - "estimated_eval_tokens": 18956 + "serialized_chars": 37434, + "estimated_input_tokens": 9359, + "estimated_eval_tokens": 18974 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is still mostly heterogeneous. Two duplicate pairs remain plausible and should be accepted: the v5 docs cleanup and the `_is_hf_initialized`/`init_empty_weights` loading TypeError. Issue 44403 is a reasonable umbrella representative because it is the most generic loading/noise report, but not because it has the strongest activity in this packet.", - "confidence": 0.72, - "canonical_issue_reason": "Issue 44403 is the broadest and most generic loading-related report here (`Unnecessary noise when loading a transformer`), so it works best as an umbrella representative for the cluster\u2019s shared theme. Its selection should be based on generality, not on having the highest discussion activity.", + "summary": "This set is mostly heterogeneous closed issues with only a few clear duplicate subclusters. The strongest matches are the v5 docs/pipeline wording issue, the bitsandbytes `_is_hf_initialized` incompatibility, and a likely SigLIP2 implementation discrepancy; the rest are too model- or symptom-specific to merge safely.", + "confidence": 0.78, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 44403 is the best overall issue representative because it is the least model-specific and most aligned with the cluster-wide pattern of load-time noise/errors. It is a better umbrella choice than the more specific model-failure reports, even though it is not the most active issue.", + "best_issue_reason": "issue:43872 is the clearest anchor because it names a concrete, reproducible `_is_hf_initialized` / Int8Params incompatibility and has a very strong duplicate counterpart in 44291.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Both are model-loading failures, but they appear to be different models and different code paths, not the same bug." + "reason": "Different models and failure modes; one is Voxtral tokenizer error messaging, the other is a GLM-4v load failure." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading-noise report vs a specific model loading failure; related theme, but not the same underlying issue." + "reason": "Generic loading noise vs a specific ScandiBERT load failure; not the same underlying bug." }, { "left": "issue:43975", - "right": "issue:45237", + "right": "issue:45520", "accept": false, - "reason": "Tokenizer/detokenization regression and AMD GPU runtime support are different problems." + "reason": "Tokenizer detokenization regression vs Python 3.13 flash_attn import-key error; unrelated code paths." }, { "left": "issue:43054", "right": "issue:43493", - "accept": false, - "reason": "Same model family area, but quality discrepancy vs implementation mismatch is not enough to treat as one bug." + "accept": true, + "reason": "Both point at the same SigLIP2 text-embedding discrepancy/implementation mismatch relative to expected behavior." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Both are loading-related, but they concern different models and specific failures." + "reason": "Both mention loading, but the concrete issues differ; this is too broad to treat as a duplicate." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Gradient checkpointing behavior and Qwen3 MoE memory usage are distinct bugs." + "reason": "LoRA gradient checkpointing behavior is unrelated to Qwen3 MoE training memory usage." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Both involve remote code, but one is missing config fields and the other is a broader loading regression." + "reason": "One is missing fields in AutoConfig with trust_remote_code, the other is a broader remote_code regression; same area but not the same bug." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Different model families and different failure modes." + "reason": "Different model families and different loading backends; no concrete shared failure path." }, { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "MPS histogram kernel issue vs Gemma4 device mismatch are unrelated concrete bugs." + "reason": "MPS histogram kernel limitation is unrelated to a Gemma4 device-mismatch forward bug." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "Both are Qwen/LoRA-related, but they describe different compatibility/warning issues." + "reason": "Both involve LoRA/training, but one is a hf4.x\u2192hf5.x Qwen3 MoE incompatibility and the other is a tie_word_embeddings warning." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Both are load-time errors, but they point to different models and different code paths." + "reason": "Different models and different errors; Molmo tie-weights attribute issue does not match a GLM-4v load failure." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression vs flash-attn fallback gating are unrelated fixes." + "reason": "Tokenizer loading for est-roberta is unrelated to flash-attn fallback gating." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Different failure classes: missing import vs `_is_hf_initialized` TypeError." + "reason": "ImportError for Qwen2.5-VL is unrelated to the bitsandbytes `_is_hf_initialized` constructor issue." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Remote-code regression vs Gemma4 device placement bug; not the same issue." + "reason": "Broad remote_code breakage does not match a specific Gemma4 PLE device placement bug." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping setup failure vs broader remote-code breakage are different bugs." + "reason": "add-new-model-like / tokenizer mapping logic is a different path from remote_code loading regressions." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Autocast dtype mismatch and device mismatch are different concrete failures." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is not the same as Gemma4 device mismatch." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer regex patch crash vs flash-attn fallback checks are unrelated code paths." + "reason": "Mistral tokenizer regex patch crash is unrelated to flash-attn2 fallback checks." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL loading error vs GLM5 issue are different failures." - }, - { - "left": "issue:45020", - "right": "issue:45356", - "accept": false, - "reason": "Broad remote-code regression vs Kimi tokenizer codec regression are different underlying bugs." - }, - { - "left": "issue:43742", - "right": "issue:45399", - "accept": false, - "reason": "MobileLLM key error vs flash-attn fallback gating have no concrete overlap." - }, - { - "left": "issue:43404", - "right": "issue:43950", - "accept": false, - "reason": "Tied-weight bug vs non-persistent buffer corruption are different code paths." - }, - { - "left": "issue:43827", - "right": "issue:44509", - "accept": true, - "reason": "Both report the same v5 docs problem: docs still reference removed pipeline tasks." - }, - { - "left": "issue:43792", - "right": "issue:44220", - "accept": false, - "reason": "Whisper runtime failure and `_torch_extract_fbank_features()` are related in theme, but not clearly the same bug." + "reason": "Qwen3VL weight_only loading error and GLM5 are unrelated model-specific failures." }, { "left": "issue:43872", "right": "issue:44291", "accept": true, - "reason": "Both describe the same `_is_hf_initialized` / `init_empty_weights` loading TypeError and share the same fix path." + "reason": "Same underlying `_is_hf_initialized` incompatibility during model init/loading with bitsandbytes/init_empty_weights." }, { "left": "issue:43257", "right": "issue:43866", "accept": false, - "reason": "Qwen3 MoE weight conversion issue vs corrupted Ovis2 checkpoint are different failures." + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to an Ovis2 checkpoint corruption report." }, { "left": "issue:42915", "right": "issue:43872", "accept": false, - "reason": "Qwen3 MoE FP8 config failure vs bitsandbytes initialization TypeError are unrelated." + "reason": "FineGrainedFP8Config failure and bitsandbytes `_is_hf_initialized` incompatibility are different bugs." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Related gradient-checkpointing theme, but different bugs: ineffective checkpointing vs missing support flag." + "reason": "Gradient checkpointing behavior is unrelated to a missing supports_gradient_checkpointing flag." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are ZeRO-3 load failures, but the titles indicate different model generations and different missing-weight symptoms." + "reason": "Both are ZeRO-3 weight-loading failures, but the model variants and missing-weight patterns are not specific enough to assert the same root cause." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Trainer API mismatch vs tensor mask shape mismatch are not the same bug." + "reason": "Unexpected compute_loss kwarg error and tensor mask shape mismatch are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer/model loading issues with no shared concrete failure path." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model-loading failures for different architectures." + "reason": "GLM-4.7-Flash unit-test load failure is unrelated to NemotronH checkpoint loading." }, { - "left": "issue:41553", - "right": "issue:44991", + "left": "issue:45020", + "right": "issue:45356", "accept": false, - "reason": "Model-specific loading error vs tokenizer-loading regression for a different model." + "reason": "Remote_code regressions and Kimi-K2.5 tokenizer codec/fix_mistral_regex problems are different issues." }, { - "left": "issue:43824", - "right": "issue:43872", + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "MobileLLM key error and flash-attn fallback gating are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 is unrelated to non-persistent buffer corruption in from_pretrained." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both report the same v5 docs regression: stale references to removed pipeline tasks." + }, + { + "left": "issue:43792", + "right": "issue:44220", "accept": false, - "reason": "Different failure classes and different loading mechanisms." + "reason": "Whisper cannot run is too broad to safely equate with a specific _torch_extract_fbank_features issue." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same `_is_hf_initialized` incompatibility manifesting through different loading entry points." }, { "left": "issue:43257", - "right": "issue:45305", + "right": "issue:43866", + "accept": false, + "reason": "Different models and different failure classes; no shared root cause." + }, + { + "left": "issue:42915", + "right": "issue:43872", "accept": false, - "reason": "Qwen3 MoE weight conversion and DeepSpeed gradient averaging are different problems." + "reason": "Qwen3 MoE FP8 training failure is not the same as the bitsandbytes/init_empty_weights constructor regression." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster summary is mostly conservative, but one key justification is overstated: issue 44403 is not supported as having the 'strongest discussion/inbound activity' in this packet (several others have higher discussion counts, e.g. 41553, 44451, 44488). Please revise the representative/canonical rationale to stay within the evidence. The two accepted duplicate pairs look plausible, but the reason for choosing 44403 needs to be more carefully grounded." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative about weaker links. The accepted soft-edge pairs are well supported by the issue titles/body themes, and the rejected pairs are cautious rather than overstated. No mergeability or same-bugfix claim appears inflated." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1610,13 +1640,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 5, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:24643", "issue:30064", @@ -1662,6 +1692,7 @@ "issue:45584", "issue:45588", "issue:45663", + "issue:45685", "issue:45823", "issue:45834", "issue:45902" @@ -1680,11 +1711,12 @@ "issue:43950|issue:45834", "issue:43550|issue:44910", "issue:43645|issue:43646", + "issue:44560|issue:45685", "issue:43366|issue:44589", "issue:42915|issue:43957", "issue:44560|issue:44918", - "issue:42617|issue:43366", "issue:43824|issue:43957", + "issue:42617|issue:43366", "issue:36010|issue:42915", "issue:44661|issue:45823", "issue:43632|issue:45834", @@ -1699,214 +1731,232 @@ "issue:38175|issue:39692", "issue:42491|issue:43720" ], - "prepared_review_unit_hash": "64773d9a0f10f2a8f0dce888b88656586884bfa8f5e12f96c6cc78ebbdc8b64b", + "prepared_review_unit_hash": "a0420541163aee96969bd7ccb5e627e301c2e99e713f13480fb45f92ca8648f7", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36612, - "estimated_input_tokens": 9153, - "estimated_eval_tokens": 18562 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37486, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous transformer issues with only a couple of plausible duplicate pairs. The strongest duplicate signal is the Transformers 5 custom-model initialization regression; a second likely pair is the SigLIP2 bad-output reports. Most other soft edges are same-subsystem but not the same bug.", - "confidence": 0.62, - "canonical_issue_reason": "issue:43646 is the cleanest representative of the main duplicate pair: it captures the broad Transformers 5 regression around custom model initialization, while issue:43645 is a narrower notebook-specific manifestation of the same breakage.", + "summary": "This cluster is mostly heterogeneous: the soft pairs span different models, symptoms, and code paths. Only the two Transformers 5.0 custom model initialization issues look like the same underlying bug.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43646 is the broader, cleaner description of the same Transformers 5.0 custom-model initialization regression; 43645 is the notebook-specific variant.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best single issue to anchor the cluster because it is broader and more general than the notebook-specific variant, and its title matches the underlying regression clearly.", + "best_issue_reason": "issue:43646 is the best representative because it captures the core regression without the extra Jupyter-specific framing.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "Both mention Qwen3.5/flash-attention-style failures, but one is a gradient explosion/SDPA fallback problem and the other is an AttributeError on s_aux=None; different code paths and symptoms." + "reason": "Both mention Qwen3.5/3D position_ids, but one is a flash-attention illegal-access crash and the other is SDPA/BF16 training instability; different concrete bugs and likely different fixes." }, { "left": "issue:41093", "right": "issue:44805", "accept": false, - "reason": "The titles are similarly generic IndexErrors, but there is not enough evidence they share the same underlying bug or repro; they could easily be unrelated mask-shape failures." + "reason": "Same generic IndexError wording, but no evidence they share the same failing code path or model; too little to treat as a duplicate." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both concern Qwen video inputs, but one is a broad v5.3.0 regression across several models while the other is a StopIteration in a specific Qwen3-VL embedding path." + "reason": "Both are video-related Qwen regressions, but one is a broad v5.3.0 input regression and the other is a StopIteration in Qwen3-vl-embedding; not clearly the same bug." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "These are unrelated: one is a bad tokenizer error message for Voxtral, the other is a model loading failure for cjvt/sleng-bert." + "reason": "Both concern loading/tokenizer errors, but they involve different models and different failure modes; not the same underlying issue." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Different models and different failure modes: PagedAttentionCache on Qwen3.5 versus NemotronH checkpoint loading." + "reason": "Different model families and different loader/cache failures; no shared concrete bug." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "DeepSpeed training weight-shape error and void segmentation map processing are unrelated." + "reason": "Completely different domains: DeepSpeed training weight shape error vs void segmentation map image-processor failure." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Both involve rotary/buffer internals, but they affect different models and different failure classes (temporal scaling vs uninitialized memory after meta/to_empty)." + "reason": "Both are low-level model internals, but one is Qwen2.5-VL rope scaling and the other is Qwen3.5-MoE uninitialized memory after meta materialization; unrelated fixes." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF gpt-oss support and a Phi-tiny-MoE autocast dtype mismatch are different problems." + "reason": "GGUF/gpt-oss support and a Phi-tiny-MoE autocast dtype mismatch are unrelated." }, { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Gemma4 FSDP2 KeyError from shared_kv_states rebuilding is unrelated to the Gemma4 PLE device-mismatch forward bug." + "reason": "Same Gemma4 family, but one is an FSDP2 shared_kv_states KeyError and the other is a device_map forward mismatch; different code paths." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "One is GGUF/gpt-oss support, the other is a Qwen3-MoE sliding_window issue." + "reason": "No overlap beyond broad Qwen/GGUF context; unrelated issues." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "from_pretrained corrupting non-persistent buffers is a different regression from a Kosmos2.5 long-OCR index error." + "reason": "Non-persistent buffer loading regression vs Kosmos2.5 OCR index error; unrelated." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different code paths: Bamba torch.compile/SDPA versus Qwen3.5 flash-attention crash from 3D position_ids." + "reason": "Different models and different attention paths; one is Bamba SDPA/compile, the other is Qwen3.5 Flash Attention with 3D position_ids." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Both report the same Transformers 5.0.0 regression breaking custom model initialization; 43645 is the notebook-specific variant of 43646's broader bug." + "reason": "These describe the same Transformers 5.0 regression breaking custom model initialization; the notebook mention in 43645 is just a usage context, not a distinct bug." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Video inference StopIteration and an MPS histogram op-not-implemented error are unrelated." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "GGUF support for gpt-oss and a Float8 storage lookup error are unrelated." + "reason": "GGUF model support and a Float8 storage lookup error are different problems." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and meta-device loading breakage are different issues." + "reason": "Qwen3Moe FP8 config failure vs meta-device loading issues; different failure modes and likely different fixes." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "One is a Qwen3-VL video StopIteration regression, the other is unpacking Qwen3.5 embeddings with TRL SFT trainer." + "reason": "Both involve Qwen3.5/video-ish workflows, but one is a video StopIteration and the other is embedding unpacking with TRL SFT; not the same bug." }, { - "left": "issue:42617", - "right": "issue:43366", + "left": "issue:43824", + "right": "issue:43957", "accept": false, - "reason": "3d_parallel.py runtime failure and GGUF gpt-oss support are unrelated." + "reason": "Missing class import and meta-device loading regression are unrelated." }, { - "left": "issue:43824", - "right": "issue:43957", + "left": "issue:42617", + "right": "issue:43366", "accept": false, - "reason": "Missing Qwen2.5-VL import and meta-device loading breakage are different problems." + "reason": "3d_parallel.py runtime failure and GGUF gpt-oss support are unrelated." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import failure and Qwen3Moe FP8 loading are unrelated." + "reason": "GenerationMixin import error vs Qwen3Moe FP8 failure; unrelated." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "Tokenizer mapping registration failure and Gemma4 PLE device mismatch are different bugs." + "reason": "Tokenizer mapping metadata bug and Gemma4 device mismatch are unrelated." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "_is_hf_initialized flag regression and Kosmos2.5 long-input index error do not share the same underlying cause." + "reason": "_is_hf_initialized flag regression and Kosmos2.5 OCR index error are different bugs." }, { "left": "issue:38175", "right": "issue:43994", - "accept": true, - "reason": "Both are about google/siglip2-base-patch16-224 producing invalid outputs through different entry points, which points to the same underlying SigLIP2 inference bug." + "accept": false, + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is a model/pipeline nonsense-results report; not enough evidence of the same root cause." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "AudioFlamingo3 batching token/embedding leakage and Whisper empty-transcription handling are unrelated." + "reason": "AudioFlamingo3 batched embedding/token leak vs Whisper empty-transcription generation failure; different models and mechanisms." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "A Qwen video StopIteration issue and a mask/tensor shape IndexError are different bugs." + "reason": "StopIteration in Qwen3-vl-embedding video input vs a generic mask/tensor shape mismatch; not the same failure." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both involve stricter config validation, but they affect different model families and do not clearly share the same failing code path." + "reason": "Strict config loading for granite_speech and tiny-random glm4v config validation are separate config issues." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "DeepSpeed zero3 Bert indexing failure and add-new-model-like TOKENIZER_MAPPING_NAMES failure are unrelated." + "reason": "DeepSpeed zero3/BertModel indexing error and tokenizer-mapping-related model addition failure are unrelated." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Same as above: a zero3 Bert init crash versus a tokenizer-mapping registration problem." + "reason": "Both touch initialization/loading, but one is a zero3 Bert index error and the other is add-new-model-like metadata handling; different bugs." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer mapping metadata bug vs generic index-mask mismatch; unrelated." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "device_map=auto dequantized-model offload failure and Strix Halo mmap OOM are different memory/loading issues." + "reason": "device_map/offload loading failure and Strix Halo mmap OOM are different memory-management issues." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Both involve SigLIP2, but 39692 is a docs/example problem with model-processor mismatch and quantization failure, not the same as zero probabilities." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Qwen3_MoE LoRA compatibility across HF versions and BitNet packed-weight unpacking are unrelated." + "reason": "Qwen3-MoE LoRA compatibility across versions and BitNet packed-weight loading are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The two accepted pairs are plausibly the same underlying regressions based on the titles, and the rejected pairs mostly differ by model, failure mode, or code path without overstating evidence." + "feedback": "The summary stays grounded in the packet and is appropriately conservative. The only accepted pair, issue:43645 / issue:43646, is well supported by the titles and descriptions as the same Transformers 5.0 custom-model initialization regression, with 43645 appearing as a notebook-specific variant. The other pairwise rejections are cautious and do not overclaim shared root causes." }, "evaluator_used": true, "retried": false, @@ -1919,13 +1969,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 6, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36010", "issue:39404", @@ -1978,8 +2028,8 @@ "soft_pairs": [ "issue:44960|issue:44977", "issue:41720|issue:42915", - "issue:44451|issue:44488", "issue:45313|issue:45357", + "issue:44451|issue:44488", "issue:42491|issue:45276", "issue:43708|issue:44075", "issue:43644|issue:44534", @@ -2008,172 +2058,172 @@ "issue:45663|issue:45859", "issue:43873|issue:44387" ], - "prepared_review_unit_hash": "f236d34ac82fc13a05066ad888810852d35fe9a536b395b5630c3ed224e624f0", + "prepared_review_unit_hash": "5127f3558a92c098a6e315a39557e1ee247013631ce56d92d1bc8ce4fbf89f46", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36445, - "estimated_input_tokens": 9112, - "estimated_eval_tokens": 18480 + "serialized_chars": 36443, + "estimated_input_tokens": 9111, + "estimated_eval_tokens": 18478 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5/model-specific regressions around loading, tokenizers, quantization, and multimodal generation. Only the non-persistent-buffer regression is a clear duplicate; most soft links are broad similarity rather than the same concrete bug.", - "confidence": 0.72, - "canonical_issue_reason": "issue:43644 is the cleanest anchor because it states the broad regression directly and has an explicit duplicate-style twin (issue:44534) with the same \"non-persistent buffers filled with junk\" symptom.", + "summary": "This cluster is mostly heterogeneous. The only strong duplicate pair is 43644/44534 about non-persistent buffers being filled with junk. The Gemma/T5Gemma resize_token_embeddings reports are similar in theme but are not a true duplicate pair here because they affect different model families and different submodules.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 43644 is the best canonical issue because it is the clearest, broader regression report and it has the strongest confirmed duplicate match in 44534.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest overall representative: it is open, has multiple inbound refs and active discussion, and covers a broader live regression area (offloading/quantization) than the narrow duplicate pair.", + "best_issue_reason": "Issue 43644 is the best representative issue for the cluster: it captures the only clearly duplicated bug and is more general than the other model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Both mention generation problems, but they are different models and different failure paths (GLM5 vs Qwen3.5 flash-attention)." + "reason": "Same broad area (generation) but different models and different failure modes." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Qwen3 auto device mapping cuda assert and Qwen3 MoE FP8 config failure are distinct bugs." + "reason": "CUDA auto-device-map assert and LoRA version-compatibility are unrelated bugs." }, { - "left": "issue:44451", - "right": "issue:44488", + "left": "issue:45313", + "right": "issue:45357", "accept": false, - "reason": "Both are load-failure reports for different BERT models, but there is no evidence they share the same root cause." + "reason": "Both are Qwen3.5 reports, but one is ZeRO-3 loading and the other is save_pretrained key serialization." }, { - "left": "issue:45313", - "right": "issue:45357", + "left": "issue:44451", + "right": "issue:44488", "accept": false, - "reason": "One is ZeRO-3 weight loading for language_model, the other is incorrect save_pretrained visual encoder keys; different code paths." + "reason": "Both are load failures, but they concern different models and no shared concrete code path is evident." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Different model families and different APIs; not the same underlying bug." + "reason": "Different bug classes: qwen3_moe hf4.x/hf5.x incompatibility vs Gemma4 resize_token_embeddings propagation." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Trainer checkpoint-resume max_steps math and SGD arg handling are unrelated." + "reason": "Trainer checkpoint math and SGD argument handling are unrelated." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Same regression: Transformers v5 fills non-persistent buffers with junk." + "reason": "Near-identical reports of the same non-persistent-buffer junk regression in Transformers 5." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Qwen3 MoE FP8 conversion failure and Strix Halo mmap OOM are unrelated." + "reason": "Qwen3 MoE loading/version issue is unrelated to Strix Halo mmap/OOM behavior." }, { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and different concrete mechanisms." + "reason": "Both involve tokenization, but they hit different tokenizer logic and different model families." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Both involve SigLIP2, but one is a docs/example mismatch and the other is bad inference output; not clearly one bug." + "reason": "SigLIP2 docs/example errors are not the same as a checkpoint-specific AutoModel/pipeline output regression." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different models and different loading failures." + "reason": "Different loading failures: unit-test load correctness vs weight-shape mismatch." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Qwen3 MoE weight conversion and GLM-4.7-Flash test loading are unrelated." + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to GLM-4.7-Flash loading tests." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Both are memory-related, but one is Qwen3 MoE training inefficiency and the other is quantization/offloading behavior." + "reason": "Both mention memory, but one is training memory usage and the other is quantization/offloading behavior." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different model families and failure modes." + "reason": "LoRA hf-version incompatibility is unrelated to Qwen3-VL weight-shape mismatch." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Non-persistent-buffer junk vs Jupyter custom-model initialization are different regressions." + "reason": "Different regressions: buffer initialization junk vs custom model initialization in notebooks." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Both mention memory pressure, but one is training inefficiency and the other is int4 quantization reserved-memory/OOM." + "reason": "Related to memory pressure, but not the same concrete bug or fix." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Unrelated bugs in different models and different embedding paths." + "reason": "LoRA/model-version compatibility is unrelated to the resize_token_embeddings output_embeddings bug." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both are flash-attention issues, but they hit different code paths and symptoms (packed 3D position_ids vs s_aux=None)." + "reason": "Different flash-attention crashes with different causes and affected paths." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading regressions, but on different model families with no clear shared fix." + "reason": "Tokenizer/load failures for different models; no shared root cause is clear enough." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Different product surfaces and different errors (chat crash vs Gemma4Processor missing _tokenizer)." + "reason": "Qwen3.5 chat crash and Gemma4Processor _tokenizer error are separate code paths." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Save/from_pretrained extra_state handling and init_empty_weights _is_hf_initialized are separate internals bugs." + "reason": "Extra_state save/load handling is distinct from init_empty_weights argument incompatibility." }, { "left": "issue:45276", "right": "issue:45335", "accept": false, - "reason": "Both are resize_token_embeddings regressions, but they are model-specific embedding plumbing issues, not clearly one bug." + "reason": "Both are resize_token_embeddings regressions, but they affect different model families and different submodules, so this is not a true duplicate." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both are lm_head tying bugs, but in different multimodal model classes and likely different code paths." + "reason": "Both mention lm_head tying, but the model classes and concrete failure paths differ." }, { "left": "issue:36010", @@ -2185,52 +2235,52 @@ "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Completely different model families and symptoms." + "reason": "Different models and different symptom types; not the same bug." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Torch version incompatibility and meta-device loading regressions are distinct." + "reason": "Torch version compatibility is unrelated to meta-device loading regressions." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/LoRA and apex RMSNorm memory leak are unrelated." + "reason": "Gradient checkpointing with LoRA is unrelated to the T5 apex RMSNorm memory leak." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Tokenizer vocab mismatch and model loading failure are not the same bug." + "reason": "Tokenizer vocab-size mismatch is different from a model-loading failure for sleng-bert." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Gemma-4 FSDP2 attention KeyError and Qwen multimodal text_config unwrapping are different bugs." + "reason": "Gemma4 FSDP2 attention-state rebuild and Qwen3.5/3.6 config unwrapping are unrelated." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Different model families and different execution paths." + "reason": "Different failures in different models: Gemma4 attention state vs Qwen3_5Moe CPU offload." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both touch quantization/memory, but one is offloading behavior and the other is CUDA reserved-memory regression." + "reason": "Quantization offloading and int4 reserved-memory OOM are related only broadly, not as one duplicate bug." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet overall, and the soft-edge verdicts are conservative. The one accepted pair, issue:43644 and issue:44534, is strongly supported by the identical regression title, and the other pairs are appropriately rejected with cautious reasoning. The only slightly strong phrasing is calling issue:44534 an \"explicit duplicate-style twin\" when the packet shows matching symptoms rather than an explicit duplicate link, but that does not materially undermine the evaluation." + "accept": false, + "feedback": "The cluster-level summary is mostly grounded, and the 43644/44534 duplicate call is well supported by the titles. However, the analyst overstates the Gemma/T5Gemma pair: 45276 and 45335 are similar resize_token_embeddings regressions, but they affect different model families and different submodules, so the packet does not justify treating them as a true duplicate with high confidence. The soft-edge verdicts should stay more conservative." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2240,13 +2290,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 7, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:39404", "issue:41720", @@ -2310,8 +2360,8 @@ "issue:43381|issue:44928", "issue:44164|issue:44756", "issue:45588|issue:45663", - "issue:41720|issue:44155", "issue:45684|issue:45698", + "issue:41720|issue:44155", "issue:44479|issue:44877", "issue:43957|issue:44756", "issue:43646|issue:43708", @@ -2333,244 +2383,232 @@ "issue:39404|issue:44912", "issue:44479|issue:44805" ], - "prepared_review_unit_hash": "c01565adab8a516876558a560333c49218add720d2737654f7ccca7a0ced3ae4", + "prepared_review_unit_hash": "e88e86c42f1168a91e3dc0df819d6c6715b48096eb644fc5cfff203398e1045c", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37914, - "estimated_input_tokens": 9479, - "estimated_eval_tokens": 19214 + "serialized_chars": 37909, + "estimated_input_tokens": 9478, + "estimated_eval_tokens": 19212 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated transformer issues. The only plausible duplicates are the resize_token_embeddings reports, where the same propagation bug shows up across different model families.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 45292 is the cleanest generic description of the shared resize_token_embeddings propagation bug, and it can subsume the more model-specific reports.", + "summary": "Most pairs are unrelated regressions across different subsystems. The only clear duplicate here is the Gemma4/resize_token_embeddings reports, where one issue is a broader version of the other.", + "confidence": 0.82, + "canonical_issue_reason": "issue:45276 is the most complete description of the resize_token_embeddings regression and subsumes the narrower 45292 report.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45292 is the best representative because it states the core bug in a model-agnostic way and matches the accepted duplicate variants.", + "best_issue_reason": "issue:45276 is the best representative issue: it is specific, reproducible, and covers the broader symptom set in the only clear duplicate pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "Different subsystems and failure modes: batched inference token leakage vs device-mapping device mismatch." + "reason": "Different failures: audio batched inference token leakage vs Gemma4 device-map forward mismatch." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "init_empty_weights argument error is unrelated to Strix Halo mmap/OOM behavior." + "reason": "Unrelated code paths: init_empty_weights argument handling vs mmap-related OOM behavior." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Unexpected _is_hf_initialized argument during meta init is unrelated to MXFP4 quantization fallback." + "reason": "Both involve 4-bit quantization, but they are different models and failures (NVFP4 load error vs MXFP4 fallback to bf16)." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "GLM5 is too broad and unrelated to a sinq save/load regression." + "reason": "GLM5 model-loading issue vs sinq save/load corruption; same broad area, different bug." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "Sparse-layer construction bug in SwitchTransformers is unrelated to Gemma4 FSDP2 shared_kv_states handling." + "reason": "Sparse-layer config creation bug and Gemma4 FSDP2 attention-state bug are distinct." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Both mention OOM, but one is int4 reserved-memory growth and the other is mmap on Strix Halo." + "reason": "Different OOM causes: int4 quantization reserved-memory growth vs mmap on Strix Halo." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Both involve gradient checkpointing, but one is ineffective under PEFT LoRA and the other forbids eval mode." + "reason": "Both mention gradient checkpointing, but one is LoRA effectiveness in training and the other is an eval-mode restriction." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same underlying bug: resize_token_embeddings does not propagate correctly to model embeddings/output layers." + "reason": "Same underlying resize_token_embeddings bug; 45292 is the narrower output_embeddings symptom of the broader failure reported in 45276." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "GLM-4.7-Flash loading failure is unrelated to Qwen3.5 save_pretrained regression." + "reason": "Different regressions: GLM-4.7-Flash loading failure vs Qwen3.5 save_pretrained checkpoint corruption." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction is unrelated to RLHF NaN explosion from 3D position_ids." + "reason": "Gradient-checkpointing eval-mode issue and RLHF NaN/SDPA-fallback issue 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 is unrelated to mmap/OOM on Strix Halo." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "flash_attention.py s_aux=None crash is unrelated to Gemma4 FSDP2 shared_kv_states rebuilding." + "reason": "Flash-attention sinkless-model crash vs Gemma4 shared_kv_states/FSDP2 KeyError are different code paths." }, { - "left": "issue:41720", - "right": "issue:44155", + "left": "issue:45684", + "right": "issue:45698", "accept": false, - "reason": "Qwen3 device-map cuda assert and AudioFlamingo3 token leak are different bugs." + "reason": "Custom-model save permission propagation and wrong-module loading are different serialization bugs." }, { - "left": "issue:45684", - "right": "issue:45698", + "left": "issue:41720", + "right": "issue:44155", "accept": false, - "reason": "Both concern save/load custom models, but one is permissions propagation and the other is wrong-module loading." + "reason": "Qwen3 device-mapping CUDA assert and AudioFlamingo3 batch leakage are unrelated." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Video-input regression for Qwen VL models is unrelated to granite_speech strict config loading." + "reason": "Video-input regression for Qwen models vs strict config preventing granite_speech loading." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "meta-device loading failures are unrelated to mmap/OOM behavior." + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are distinct issues." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization breakage is unrelated to resume_from_checkpoint max_steps calculation." + "reason": "Custom model initialization breakage and Trainer max_steps resume miscalculation are unrelated." }, { "left": "issue:45292", "right": "issue:45335", - "accept": true, - "reason": "Same underlying bug: resize_token_embeddings fails to update the relevant output/decoder embedding modules." + "accept": false, + "reason": "Similar API area, but different manifestations across different model classes; not enough evidence for the same bug." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Jupyter custom-model initialization and extra_state save/load are different issues." + "reason": "Notebook/custom-model init regression is unrelated to extra_state save/load handling." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "Tokenizer vocab-size mismatch for one model and a generic model-loading failure are not clearly the same bug." + "reason": "Tokenizer vocab mismatch for special tokens and a ScandiBERT load failure are not the same concrete bug." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Autocast dtype mismatch and meta-device loading are unrelated failure classes." + "reason": "Autocast dtype mismatch and meta-device loading failure are different problems." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "Torch 2.10 compatibility is too broad and unrelated to Phi-tiny-MoE autocast mismatch." + "reason": "Torch version incompatibility and autocast dtype mismatch are separate failure modes." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup error is unrelated to disabling mmap on Strix Halo." + "reason": "Float8 storage lookup error is unrelated to mmap/OOM on Strix Halo." }, { "left": "issue:43257", "right": "issue:44863", "accept": false, - "reason": "Different model families and different loading issues; not the same code-path problem." + "reason": "Qwen3 MOE conversion/loading issue and NemotronH checkpoint loading failure are different model-specific bugs." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention failure on torch>=2.9 is unrelated to Gemma4 device-mapping mismatch." + "reason": "Compiled flex_attention breakage on torch>=2.9 is unrelated to Gemma4 device-mapping mismatch." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper return_language pipeline regression is not the same as a broad tokenizer-version change report." + "reason": "Whisper pipeline return_language regression and tokenizer-version drift are different issues." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Both involve Transformers 5.0.0 and custom models, but the concrete breakages are different." - }, - { - "left": "issue:43425", - "right": "issue:45823", - "accept": false, - "reason": "Torch version incompatibility is unrelated to Gemma4 forward device mismatch." + "reason": "base_model_prefix loading breakage is distinct from the notebook custom-model initialization regression." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "Torch 2.10 compatibility and Mistral-3 dtype mismatch are unrelated." - }, - { - "left": "issue:44530", - "right": "issue:45084", - "accept": false, - "reason": "Qwen3.5 PagedAttentionCache crash is unrelated to non-template node compilation error." + "reason": "Torch 2.10 incompatibility and Mistral-3 dtype mismatch do not share the same underlying bug." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "Isolated reproduction of another issue and Qwen3.5 cache crash are not the same bug." + "reason": "Isolated reproduction of an old issue and Qwen3.5 PagedAttentionCache group-type crash are not the same bug." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export with torch_compilable_check and Perceiver interpolation failure are unrelated." + "reason": "torch.export/torch_compilable_check failure is unrelated to Perceiver resizing/interpolation behavior." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper return_language pipeline regression is unrelated to MXFP4 quantization fallback." + "reason": "Whisper pipeline return_language regression and MXFP4 quantization loading behavior are unrelated." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video-input regression and mask-shape IndexError are different failure modes." + "reason": "Video-input regression and mask-shape IndexError are different failures." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary stays within the packet evidence and is appropriately conservative. The only clearly supported duplicate cluster is the resize_token_embeddings trio (45276/45292/45335), and the chosen canonical issue 45292 is a reasonable generic representative. The negative soft-edge verdicts are also cautious and do not overclaim cross-issue equivalence." + "feedback": "The summary stays grounded in the packet and the soft-edge calls are conservative. The only accepted pair, issue:45276 vs issue:45292, is plausibly the same resize_token_embeddings regression with one report being a narrower symptom of the broader Gemma4 issue. No other pair is overstated." }, "evaluator_used": true, "retried": false, @@ -2583,13 +2621,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 8, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:32090", "issue:34567", @@ -2656,35 +2694,35 @@ "issue:44164|issue:44479", "issue:44488|issue:45081", "issue:43582|issue:43632", - "issue:43611|issue:43644", "issue:44661|issue:44683", + "issue:43611|issue:43644", "issue:40990|issue:43421", "issue:32090|issue:39404", "issue:43716|issue:45084", "issue:44488|issue:44987", "issue:43643|issue:44530", - "issue:34567|issue:40444", "issue:36032|issue:42915", + "issue:34567|issue:40444", "issue:44811|issue:45584", "issue:44530|issue:46304", "issue:43541|issue:45084", "issue:44451|issue:44987", "issue:43644|issue:43646" ], - "prepared_review_unit_hash": "6af9f12f65a94ee0b0a22bfde60730dfe962a376cae8d4f4f36f86ad55b6415f", + "prepared_review_unit_hash": "0d966ea9026f716fdcd339c6b642fecfdbc6c7496d3cbcc535686041fe7b8a86", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36036, + "serialized_chars": 36034, "estimated_input_tokens": 9009, "estimated_eval_tokens": 18274 }, @@ -2693,192 +2731,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous issue reports with a loose common theme of Transformers v5 regressions, but none of the soft-edge pairs look like the same underlying bug or change. No PRs were present.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43646 is the broadest, most representative regression report here: a v5 model-initialization breakage that plausibly spans multiple downstream loading failures.", + "summary": "This cluster is thematically similar but not duplicate: it mixes many unrelated Transformers v5 regressions across tokenizers, model loading, training, multimodal paths, docs, and internal runtime behavior.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best single representative issue because it describes a general Transformers v5 initialization regression rather than a model-specific symptom.", + "best_issue_reason": "No single issue is a safe global representative; the nearest broad regression issue is still too specific to model initialization and does not cover the rest of the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs issues, but they cover different doc mismatches: one about return_all_scores, the other about removed pipeline tasks." + "reason": "Both are docs-related, but one is about stale return_all_scores wording and the other about removed pipeline task docs; not the same bug." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Unrelated failures: a float8 storage lookup error versus a template-compilation error." + "reason": "Float8 storage lookup failure and template compilation failure are unrelated runtime errors with different code paths." }, { "left": "issue:43425", "right": "issue:44918", "accept": false, - "reason": "Different problems: Torch version incompatibility versus a TRL SFT trainer embedding/unpacking issue." + "reason": "Torch version incompatibility and Qwen3.5 embedding unpacking are different failures, not one underlying issue." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Separate tokenizer/loading bugs affecting different code paths and model formats." + "reason": "Tokenizer special-token name collision and gguf_file loading breakage are distinct tokenizer/loading bugs." }, { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "One is save/from_pretrained extra_state handling; the other is from_config text_config unwrapping for composite Qwen configs." + "reason": "save/from_pretrained extra_state handling and from_config text_config unwrapping affect different serialization paths." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Same family, but different underlying load failures: dev0 model-loading regression versus checkpoint weight-shape mismatch." + "reason": "Both involve Qwen3-VL loading, but one is a dev-version regression and the other is a weight-shape mismatch; not the same concrete bug." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "Different subsystems and errors: DeepSpeed zero3/BERT indexing versus torch>=2.9 flex_attention compilation." + "reason": "DeepSpeed zero3 BertModel init and compiled flex_attention on torch>=2.9 are unrelated." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer-loading problems, but for different models and different failing code paths." + "reason": "These are different tokenizer/model-loading failures for different models and different internal causes." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "A specific model producing bad outputs is not the same as a generic ModelOutput key-assignment bug." + "reason": "Siglip2 bad outputs and ModelOutput key assignment with prior None are unrelated behaviors." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "Masked LM prediction regression and tokenizer load failure are different bugs." + "reason": "CamemBERT masked-LM prediction regression and tokenizer loading failure for a different model are not the same issue." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Both mention memory, but one is Qwen3 MoE training inefficiency and the other is Strix Halo mmap OOM mitigation." + "reason": "Qwen3 MoE training memory inefficiency and Strix Halo mmap OOM are different memory problems with different fixes." }, { "left": "issue:39404", - "right": "issue:44661", + "right": "issue:44671", "accept": false, - "reason": "Whisper pipeline language return regression and add-new-model-like tokenizer mapping failure are unrelated." + "reason": "Whisper return_language pipeline regression and CamemBERT masked-LM regression are unrelated model-specific bugs." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Both are device/offload-related, but they affect different models and different forward paths." + "reason": "CPU device placement bug in XLNet relative positions and Qwen3.5 MoE CPU offload incompatibility are not the same code-path problem." }, { "left": "issue:44164", "right": "issue:44479", "accept": false, - "reason": "Serialization extra_state handling is unrelated to the Qwen VL video-input regression." + "reason": "Extra_state save/load failure and v5.3 video-input regression are unrelated." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Different tokenizer load failures with different triggers and model families." + "reason": "Two different tokenizer-loading failures for different models; no shared underlying defect." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "AppleSilicon allocator warmup TypeError and _is_hf_initialized regression are distinct runtime issues." + "reason": "AppleSilicon allocator warmup TypeError and _is_hf_initialized flag break are separate runtime regressions." }, { - "left": "issue:43611", - "right": "issue:43644", + "left": "issue:44661", + "right": "issue:44683", "accept": false, - "reason": "Different v5 regressions: base_model_prefix loading versus non-persistent buffer initialization." + "reason": "Model-like addition failing inside TOKENIZER_MAPPING_NAMES and compiled flex_attention failures are unrelated." }, { - "left": "issue:44661", - "right": "issue:44683", + "left": "issue:43611", + "right": "issue:43644", "accept": false, - "reason": "Token-mapping failure and flex_attention compilation failure are unrelated." + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; not a duplicate." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity regression on GPT-OSS is not the same as runtime special-token/post-processor update behavior." + "reason": "High perplexity on gpt-oss-20b and runtime post-processor updates in TokenizersBackend are unrelated." }, { "left": "issue:32090", "right": "issue:39404", "accept": false, - "reason": "Trainer GPU broadcast TypeError and Whisper return_language pipeline regression are unrelated." + "reason": "Trainer GPU broadcast NoneType error and Whisper return_language pipeline regression are different subsystems and bugs." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Different model-specific failures: dtype mismatch in Mistral-3 versus template compilation error." + "reason": "Image preprocessor/model dtype mismatch and template compilation failure are unrelated." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are load failures, but they involve different models and likely different tokenizer/config bugs." + "reason": "These are distinct model-loading regressions on different models and code paths." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "AutoConfig trust_remote_code field loss and Qwen3.5 paged-attention cache group-type crash are unrelated." + "reason": "AutoConfig trust_remote_code missing fields and PagedAttentionCache invalid group type are unrelated." }, { - "left": "issue:34567", - "right": "issue:40444", + "left": "issue:36032", + "right": "issue:42915", "accept": false, - "reason": "TrainerState token counter bug and Qwen2.5-VL iterable dataset failure are different issues." + "reason": "Tokenizer add_special_tokens conflict and FineGrainedFP8Config Qwen3Moe failure are different bugs." }, { - "left": "issue:36032", - "right": "issue:42915", + "left": "issue:34567", + "right": "issue:40444", "accept": false, - "reason": "T5 tokenizer method conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." + "reason": "TrainerState token counter not updating and Qwen2.5-VL iterable dataset multi-image finetuning failure are unrelated." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Both are Whisper/special-token related, but one is batch_decode skipping behavior and the other is generation failing on empty transcription." + "reason": "Both are Whisper issues, but one is batch_decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens; not the same bug." }, { "left": "issue:44530", "right": "issue:46304", "accept": false, - "reason": "Qwen3.5 cache/group-type crash and USE_DISTRIBUTED=0 support regression do not describe the same bug." + "reason": "Qwen3.5 PagedAttentionCache linear_attention crash and USE_DISTRIBUTED=0 support break are unrelated." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Torch dynamo tracing on Mixtral grouped_mm is unrelated to a template-node compilation error." + "reason": "Mixtral grouped_mm tracing failure and template compilation error are different runtime problems." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Two different model-loading failures for different checkpoints and likely different causes." + "reason": "Loading vesteinn/ScandiBERT and loading physical-intelligence/fast are separate model-loading failures." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Related to v5 initialization, but one concerns non-persistent buffers while the other is custom model initialization; not the same bug." + "reason": "Both are v5-related regressions, but one concerns non-persistent buffers and the other custom model initialization; too broad to merge as duplicates." } ] }, @@ -2894,13 +2932,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 9, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36032", "issue:42175", @@ -2979,245 +3017,242 @@ "issue:44220|issue:44683", "issue:36032|issue:43742", "issue:42915|issue:44291", - "issue:42915|issue:44589", "issue:43632|issue:44756", + "issue:42915|issue:44589", "issue:44062|issue:44987", "issue:43404|issue:45127", "issue:43611|issue:44534" ], - "prepared_review_unit_hash": "820eb4145c216429a5264d414cf9390e0e4522cdaba043dbf3241cef506cbf28", + "prepared_review_unit_hash": "83919fdb0df6aa307b336be5586a9c78a712e6a1bb100f07f083baf58649cfee", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 33, - "serialized_chars": 37080, - "estimated_input_tokens": 9270, - "estimated_eval_tokens": 18796 + "serialized_chars": 37081, + "estimated_input_tokens": 9271, + "estimated_eval_tokens": 18798 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a heterogeneous set of Transformers bugs/regressions, with a loose theme around v5 loading/initialization breakages. The soft pairs are mostly topical neighbors rather than true duplicates, so none should be merged.", - "confidence": 0.91, - "canonical_issue_reason": "Issue 43646 is the broadest representative of the recurring v5 custom-model initialization regression family, and several nearby reports (init_empty_weights, _is_hf_initialized, base_model_prefix, tied weights) are variations on that theme. It is still only a loose anchor, not a true duplicate target for the whole cluster.", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5 regressions. Only two soft edges look like true duplicates: the GGUF loading failures and the Qwen3.5 save_pretrained regressions.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43646 is the clearest broad representative: it captures the common v5 regression theme around broken custom-model initialization without being overly model-specific.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43646 is the best single issue to anchor this cluster because it is broad, clearly regression-oriented, and closest to multiple other initialization-related failures. It is more representative than the very model-specific or subsystem-specific issues.", + "best_issue_reason": "issue:43646 is the best standalone issue in the set because it is concise, broad, and directly describes a core Transformers v5 breakage affecting model initialization.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Both are audio feature extraction bugs, but one is about an unsupported `center` arg and the other about `_torch_extract_fbank_features`; same area, different failure mode." + "reason": "Both are audio-feature-extraction bugs, but the titles point to different failure surfaces; not enough evidence they are the same concrete code-path defect." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Both mention generation/cache behavior, but one is about compiled forward chunked generation inconsistency and the other about `StaticCache.get_seq_length()` type contract." + "reason": "Chunked generation with compiled forward is a different problem from StaticCache.get_seq_length returning a tensor." }, { "left": "issue:43452", "right": "issue:43482", - "accept": false, - "reason": "Both involve GGUF/tokenizer loading, but one is a `gguf_file` API regression and the other is a specific Qwen2.5-GGUF v5 loading failure." + "accept": true, + "reason": "Both describe GGUF/model loading breakage in Transformers v5 and plausibly stem from the same loading-path regression." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Both are loading failures after v5 changes, but the model-specific symptoms differ: missing `all_tied_weights_keys` vs a specific `cjvt/sleng-bert` load failure." + "reason": "One is a missing all_tied_weights_keys attribute in Molmo; the other is a specific model-load failure. Different bugs." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "One is dequantized model load with device_map offload, the other is GGUF/Qwen2.5 tokenizer/model loading; different code paths." + "reason": "gpu/cpu offload with dequantized models is unrelated to GGUF loading failures." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen-family regressions, but one is about `output_hidden_states` behavior and the other is a rope/position_ids scaling bug in Qwen2.5-VL." + "reason": "Different Qwen model families and different bugs: hidden-states output vs rope-index scaling." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Both concern v5 breaking model initialization/loading, but T5 tokenizer method conflict and notebook custom-model init are separate defects." + "reason": "T5 tokenizer special-token conflict is unrelated to custom model initialization in notebooks." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "The titles are unrelated: template compilation errors versus GraniteMoEHybrid calling an invalid method." + "reason": "Different subsystems and failures; no shared concrete code path." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "Both are backend/dependency mapping issues, but one is pip extras missing TensorFlow backend and the other is an unsupported `tensorflow_text` backend entry." + "reason": "Both involve backend resolution, but one is an install-extra omission and the other is an unsupported backend mapping entry." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Both are audio-related, but one is chat-template sampling-rate defaulting and the other is batched vs single-input encoder output mismatch from missing padding-mask support." + "reason": "Audio chat-template sampling-rate defaulting and MIMI padding-mask batching are different bugs." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are v5 load failures for specific models, but one is `physical-intelligence/fast` and the other is tokenizer loading for `EMBEDDIA/est-roberta`." + "reason": "Both are model/tokenizer load regressions, but they affect different models and likely different loader paths." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Both are v5 regressions, but custom model initialization and non-persistent buffer serialization are different underlying bugs." + "reason": "Custom model initialization and non-persistent buffer serialization are unrelated v5 regressions." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation performance overhead and cache-length return type are unrelated problems." + "reason": "Generation performance overhead is unrelated to cache sequence-length typing." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "One is a missing tied lm_head weight for Mistral3, the other is inconsistent lm_head serialization by device; related area, but not the same bug." - }, - { - "left": "issue:43883", - "right": "issue:44991", - "accept": false, - "reason": "Both are load-time regressions, but one is tied-weights attribute handling and the other is tokenizer loading for a specific model." + "reason": "Both mention tied weights, but one is a missing tie and the other is device-dependent serialization; not the same bug." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Segfault on Llava Next loading versus video-input validation failure in Qwen3OmniMoe are distinct multimodal bugs." + "reason": "Segfault on Llava Next loading and video-input processing in Qwen3OmniMoe are different issues." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Different multimodal models and different symptoms: nonsensical outputs vs processor/model image-size mismatch." + "reason": "Nonsensical model outputs and wrong processor image size are different problems." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Noise during transformer loading is not the same as the Mistral regex patch crash on tokenizer internals." + "reason": "Noise during loading and Mistral regex patch crashes are not the same defect." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "FineGrainedFP8Config failure and template-node compilation error are unrelated." + "reason": "Qwen3Moe FP8 config failure is unrelated to template-node compilation errors." }, { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Both are v5 regressions around model lifecycle, but notebook initialization and non-persistent buffer serialization are different code paths." + "reason": "Notebook/custom-model initialization and buffer serialization are different regressions." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "MIMI encoder padding-mask batching issue and DAC STE mismatch are separate model-specific correctness bugs." + "reason": "MIMI batching/padding-mask mismatch and DAC from_latents/STE mismatch are unrelated." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "`torch.export` with `torch_compilable_check` and OmDet-Turbo processor image sizing are unrelated." + "reason": "torch.export with torch_compilable_check is unrelated to OmDet-Turbo processor sizing." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "Generic model load failure for `vesteinn/ScandiBERT` is not the same as the Mistral tokenizer regex patch crash." + "reason": "Model-load failure for ScandiBERT is unrelated to Mistral regex patching." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Different models and different failure points: loading a model checkpoint versus crashing while patching Mistral tokenizer regexes." + "reason": "Different failure modes and different model families." }, { "left": "issue:45216", "right": "issue:45357", - "accept": false, - "reason": "Both are Qwen3.5 regressions around saving, but one is checkpoint correctness since 5.4.0 and the other is incorrect visual encoder keys in 5.5.3." + "accept": true, + "reason": "Both are Qwen3.5 save_pretrained regressions about incorrect checkpoint serialization, likely the same underlying bug." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio fbank feature extraction and compiled flex_attention are completely different subsystems." + "reason": "Audio fbank extraction and compiled flex_attention are unrelated." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "T5 tokenizer add_special_tokens conflict and MobileLLM key error are unrelated load-time failures." + "reason": "T5 tokenizer load conflict and MobileLLM key error are different loader issues." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Both are v5 regression reports, but Qwen3Moe FP8 config failure and `_is_hf_initialized` init_empty_weights argument error are distinct." + "reason": "FP8 config failure and init_empty_weights/_is_hf_initialized TypeError are different bugs." }, { - "left": "issue:42915", - "right": "issue:44589", + "left": "issue:43632", + "right": "issue:44756", "accept": false, - "reason": "FineGrainedFP8Config failure and missing Float8 storage are separate low-level float8 issues." + "reason": "Broken _is_hf_initialized flag and mmap/OOM on Strix Halo are unrelated." }, { - "left": "issue:43632", - "right": "issue:44756", + "left": "issue:42915", + "right": "issue:44589", "accept": false, - "reason": "`_is_hf_initialized` flag regression and Strix Halo mmap OOM are unrelated." + "reason": "FP8 config failure and Float8 storage deserialization failure are different issues." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "Tokenizer `AddedToken` keyword conflict and model-loading failure for `physical-intelligence/fast` are not the same bug." + "reason": "AddedToken special-argument duplication is unrelated to model loading regressions." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Tied lm_head weights in Mistral3 and LoRA merge collapse with extended vocabulary are different weight-tying problems with different triggers." + "reason": "Both touch tied embeddings, but one is a missing tie and the other is LoRA merge collapse with extended vocab; not clearly the same bug." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "Base model prefix loading breakage and non-persistent buffer junk fill are separate v5 regressions." + "reason": "base_model_prefix loading breakage is unrelated to non-persistent buffer junk filling." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays mostly conservative. The two accepted soft edges are reasonably supported by their shared GGUF-loading and Qwen3.5 save_pretrained themes, and the remaining rejections do not overclaim sameness based on the available titles. The cluster summary is also appropriately cautious about the set being mostly unrelated regressions." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3226,13 +3261,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 10, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:33666", "issue:36032", @@ -3283,14 +3318,14 @@ "issue:45923" ], "soft_pairs": [ - "issue:43296|issue:43531", "issue:43531|issue:44877", + "issue:43296|issue:43531", "issue:36032|issue:41720", "issue:44479|issue:45325", "issue:43425|issue:44403", "issue:40990|issue:44811", - "issue:43421|issue:44568", "issue:43792|issue:44589", + "issue:43421|issue:44568", "issue:33666|issue:40444", "issue:44403|issue:44991", "issue:43632|issue:44534", @@ -3299,8 +3334,8 @@ "issue:43381|issue:45923", "issue:44462|issue:45812", "issue:44829|issue:45479", - "issue:43482|issue:45538", "issue:40444|issue:45325", + "issue:43482|issue:45538", "issue:44291|issue:44589", "issue:43317|issue:43856", "issue:41553|issue:43582", @@ -3312,20 +3347,20 @@ "issue:43531|issue:44291", "issue:44265|issue:45072" ], - "prepared_review_unit_hash": "97fa0d00ad2343c037bcb253339f11e8871bda33064b0a45a5222740f3da06f5", + "prepared_review_unit_hash": "ea582e59583f99458a07f163d3a12f7b33b3d4a8c4237beb80491ae1aa546a6f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35722, + "serialized_chars": 35723, "estimated_input_tokens": 8931, "estimated_eval_tokens": 18118 }, @@ -3334,180 +3369,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pair looks like the same underlying bug/change. The cluster is heterogeneous, spanning unrelated tokenizer, model-loading, training, attention, and packaging issues; 45020 is the closest umbrella-style issue, but it is not a true duplicate target for the others.", - "confidence": 0.95, - "canonical_issue_reason": "issue:45020 is the broadest, most umbrella-like report in the set: it describes a cross-cutting regression affecting multiple recent models and has the strongest signal for representing the cluster, even though the rest are not duplicates of it.", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers/model-specific bug reports. The only common thread is broad regression-style loading/training failures, but the soft pairs mostly differ in model family, failure mode, and code path. No PRs are present.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45020 is the broadest regression-style issue here: it frames a recent-version breakage affecting remote_code models, which is the closest umbrella theme among otherwise disparate model/tokenizer bugs.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best representative issue because it is broad, well-scoped, and central to the recurring 'recent transformers versions break models' theme.", + "best_issue_reason": "issue:45020 is the best cluster anchor because it is the most generally phrased and most representative of the shared \u2018recent Transformers version regression\u2019 pattern, whereas the others are narrower, model-specific failures.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43296", - "right": "issue:43531", + "left": "issue:43531", + "right": "issue:44877", "accept": false, - "reason": "Different models and failures: PaddleOCR-VL loading in vLLM vs Qwen3-MoE sliding_window behavior." + "reason": "Both are model-loading/config issues, but one is a Qwen3-MoE sliding_window bug and the other is a Granite Speech config strictness problem; different models and code paths." }, { - "left": "issue:43531", - "right": "issue:44877", + "left": "issue:43296", + "right": "issue:43531", "accept": false, - "reason": "Both involve model/config loading, but the concrete bugs are unrelated and model-specific." + "reason": "PaddleOCR-VL load failure and Qwen3-MoE sliding_window failure are unrelated model-specific bugs." }, { "left": "issue:36032", "right": "issue:41720", "accept": false, - "reason": "Tokenizer method-name conflict vs Qwen3 device-map CUDA assert; no shared code path." + "reason": "T5 tokenizer special-token attribute conflict is unrelated to Qwen3 device-mapping cuda assertions." }, { "left": "issue:44479", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen-VL/video-related, but one is a broad video-input regression and the other is a specific still-image RoPE scaling bug." + "reason": "Both involve Qwen VL, but one is a broad video-input regression and the other is a specific still-image temporal position-id scaling bug; not the same failure." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Torch version compatibility vs noisy loading output; unrelated symptoms and fixes." + "reason": "Torch 2.10 compatibility and noisy transformer loading are different issues." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Perplexity anomaly on gpt-oss-20b vs Whisper batch_decode skip_special_tokens issue." + "reason": "GPT-OSS perplexity on WikiText and Whisper batch_decode skip_special_tokens handling are unrelated." }, { - "left": "issue:43421", - "right": "issue:44568", + "left": "issue:43792", + "right": "issue:44589", "accept": false, - "reason": "Runtime special-token/post-processor sync vs a tokenizer-specific BOS/EOS regression." + "reason": "Whisper-large-v2 runtime failure and missing Float8 storage object are different errors with no shared code path." }, { - "left": "issue:43792", - "right": "issue:44589", + "left": "issue:43421", + "right": "issue:44568", "accept": false, - "reason": "Whisper runtime failure vs missing Float8 storage type; different failure surfaces." + "reason": "Both concern tokenizer special tokens, but one is a backend post-processor update bug and the other is a tokenizer-specific BOS/EOS regression; too different to merge." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Both touch Qwen VL training, but one is multi-GPU training and the other is IterableDataset/multi-image finetuning." + "reason": "Both mention Qwen VL training, but multi-GPU training and iterable-dataset multi-image prompt failures are distinct problems." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Generic loading noise vs tokenizer loading failure for a specific model repo." + "reason": "General loading noise vs a specific tokenizer load failure for EMBEDDIA/est-roberta are different symptoms and fixes." }, { "left": "issue:43632", "right": "issue:44534", "accept": false, - "reason": "_is_hf_initialized flag breakage vs non-persistent buffer corruption; distinct regressions." + "reason": "Both are Transformers v5 regressions, but one is about _is_hf_initialized and the other about non-persistent buffers; different underlying bugs." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment bug vs Gemma-4 FSDP2 shared_kv_states KeyError." + "reason": "ModelOutput key assignment bug and Gemma-4 FSDP2 shared_kv_states KeyError are unrelated." }, { "left": "issue:44683", "right": "issue:45020", "accept": false, - "reason": "Torch 2.9 flex_attention compilation issue vs a transformers v5 remote_code regression; different layers." + "reason": "Compiled flex_attention on torch>=2.9 and remote_code breakage in recent Transformers are different compatibility issues." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Eval-mode gradient-checkpointing misuse vs missing supports_gradient_checkpointing flag." + "reason": "Gradient checkpointing in eval mode and missing supports_gradient_checkpointing flag are related only thematically; they are not the same bug." }, { "left": "issue:44462", "right": "issue:45812", "accept": false, - "reason": "Ignoring tokenizer.json in repo vs wrong token IDs for Granite models; not the same defect." + "reason": "Tokenizer.json being ignored and Granite token IDs being wrong are both tokenizer issues, but they point to different loading/processing failures." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Flash-attention training degeneration vs zero-loss single-label classification bug." + "reason": "Both involve degenerate classification behavior, but one is tied to flash_attention_3 and the other to problem_type/num_labels=1 loss logic." }, { - "left": "issue:43482", - "right": "issue:45538", + "left": "issue:40444", + "right": "issue:45325", "accept": false, - "reason": "Qwen2.5-GGUF loading failure vs CLIPTokenizer max-length constant; unrelated tokenizer/model issues." + "reason": "Qwen2.5-VL multi-image iterable-dataset training failure and Qwen2.5-VL temporal position-id scaling are different bugs." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:43482", + "right": "issue:45538", "accept": false, - "reason": "Multi-image prompt finetuning failure vs still-image temporal position scaling bug; different logic paths." + "reason": "GGUF loading failure in Transformers v5 and CLIPTokenizer model_max_length handling are unrelated." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized TypeError vs missing Float8 storage object; unrelated initialization/storage problems." + "reason": "init_empty_weights/_is_hf_initialized TypeError and missing Float8 storage object are distinct runtime issues." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Device-map offload loading failure vs Qwen3 MoE memory inefficiency; not the same bug." + "reason": "Device-map offload/load failure and Qwen3 MoE memory inefficiency are different training/inference problems." }, { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "Bad AutoTokenizer error message for Voxtral vs Apple Silicon TypeError in caching_allocator_warmup." + "reason": "Bad AutoTokenizer error message for Voxtral is unrelated to an Apple Silicon caching_allocator_warmup TypeError." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "GGUF loading regression vs missing pad_token_id on Llama4Config; different models and causes." + "reason": "Qwen2.5-GGUF loading failure and Llama4Config missing pad_token_id are unrelated model/config issues." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Video-LLaVA tower architecture issue vs Qwen3.5 save_pretrained key regression." + "reason": "Video-LLaVA tower/temporal-attention issue and Qwen3.5 save_pretrained key regression are different model families and code paths." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Missing TensorFlow in torch extra vs tensorboard-installed-without-TF backend error; packaging issues but not one bug." + "reason": "One is missing TensorFlow in a torch extra install, the other is tensorboard-only causing a TF backend error; related packaging area but not the same bug." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "MIMI padding-mask batch mismatch vs Gemma3 mixed-length NaNs from sliding-window/all-padding windows." + "reason": "Both are batching/padding-related model bugs, but they affect different models and mechanisms (padding mask support vs sliding-window all-padding NaNs)." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Qwen3 device-map CUDA assert vs Mistral regex patch AttributeError; different subsystems." + "reason": "Qwen3 auto device mapping cuda assert and Mistral tokenizer regex patch crash are unrelated." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window bug vs init_empty_weights TypeError; no shared concrete code path." + "reason": "Sliding_window Qwen3-MoE failure and init_empty_weights/_is_hf_initialized TypeError are different regressions." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export/torch_compilable_check failure vs bf16 dtype mismatch in inference; unrelated errors." + "reason": "torch.export/torch_compilable_check failure and bfloat16 dtype mismatches are unrelated torch compatibility bugs." } ] }, @@ -3523,13 +3558,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 11, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:35707", "issue:39692", @@ -3602,31 +3637,31 @@ "issue:44568|issue:45920", "issue:43701|issue:45325", "issue:43606|issue:43819", - "issue:44186|issue:44265", "issue:41720|issue:45823", + "issue:44186|issue:44265", "issue:43454|issue:43550", "issue:43262|issue:43746", "issue:43881|issue:44488", "issue:39692|issue:40444", - "issue:40990|issue:45584", "issue:43377|issue:44610", "issue:43454|issue:45072", + "issue:40990|issue:45584", "issue:43257|issue:45923" ], - "prepared_review_unit_hash": "62dfc06155965638f27ddcfa76b3f044d46c1af0cb299ce7a02664c06236f3b8", + "prepared_review_unit_hash": "c8e479840f8692cdbb194eba7ae24aa9aac02b19e2e87c03f7b1ae485e8a8452", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37137, + "serialized_chars": 37139, "estimated_input_tokens": 9285, "estimated_eval_tokens": 18826 }, @@ -3635,198 +3670,198 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: it mixes unrelated model-loading, tokenizer, vision, audio, and v5-regression reports. There is no clear duplicate set; the closest theme is generic Transformers v5 loading/init regressions, but the paired items still describe different bugs.", - "confidence": 0.84, - "canonical_issue_reason": "No single exact duplicate spans the cluster. If forced to anchor the cluster, issue 43611 is the broadest Transformers v5 model-loading regression and is thematically closest to several others.", + "summary": "This cluster is mostly a loose set of unrelated Transformers issues, with a few broad v5 regression themes around model loading, initialization, and tied weights. None of the soft pairs look close enough to merge as true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43611 is the broadest and most representative issue in the set: it describes a concrete v5 model-loading regression and overlaps conceptually with several nearby initialization/loading complaints more than the others do.", "canonical_pr_reason": null, - "best_issue_reason": "43611 is the best issue anchor because it is the most general, centrally relevant regression report in the set; the others are mostly model-specific or unrelated failures.", + "best_issue_reason": "issue:43611 is the best single representative issue because it is concrete, well-scoped, and captures one of the recurring regression themes in the cluster without being overly model-specific.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both involve v5 initialization/loading behavior, but one is about base_model_prefix loading and the other about custom model initialization; not the same concrete bug." + "reason": "Both are v5-related loading/init regressions, but one is about base_model_prefix loading and the other about custom model initialization; they are related theme-wise, not the same concrete bug." }, { "left": "issue:43632", "right": "issue:45072", "accept": false, - "reason": "_is_hf_initialized flag breakage and bfloat16 dtype mismatches are different failure modes in different code paths." + "reason": "Different failure modes: _is_hf_initialized/model init regression versus bf16 dtype mismatches in inference." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Different models and different bugs: SigLIP2 AutoModel/pipeline mismatch vs Perceiver interpolation on non-default resolution." + "reason": "Both are vision-model correctness issues, but the symptoms and code paths differ: SigLIP2 AutoModel/pipeline output corruption vs Perceiver resolution handling." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Packaging/backend install issue vs a model-loading failure in vLLM; unrelated." + "reason": "Completely unrelated: missing TensorFlow backend in an extra install versus PaddleOCR-VL load failure in vLLM." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Both are vision/OCR-adjacent, but one is a Qwen2.5-VL rope-index scaling bug and the other is a Kosmos2.5 long-input index error." + "reason": "Different models and bugs: Qwen2.5-VL rope/position-id scaling versus Kosmos2.5 OCR long-input index error." }, { "left": "issue:43792", "right": "issue:43957", "accept": false, - "reason": "Whisper runtime failure and meta-device loading regressions are unrelated." + "reason": "Whisper inference failure and meta-device loading regressions are not the same underlying bug." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Both mention tied lm_head weights, but one is a generation tie failure and the other is inconsistent serialization; different bugs." + "reason": "Both mention lm_head/tied weights, but one is a broken weight tie causing bad generation while the other is inconsistent serialization depending on device." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "ModelOutput key assignment and non-persistent buffer junk are unrelated regressions." + "reason": "Unrelated issues: ModelOutput key assignment versus non-persistent buffer initialization." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Different models and different loading/config failures; no shared concrete bug." + "reason": "Different areas: FineGrainedFP8Config support for Qwen3Moe versus GGUF loading failure for Qwen2.5." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen2.5-VL-related, but one is multi-image finetuning with IterableDataset and the other is video position-id handling." + "reason": "Both involve Qwen2.5-VL, but one is finetuning with multiple images per prompt and the other is video vision_position_ids regression." }, { "left": "issue:45325", "right": "issue:45381", "accept": false, - "reason": "Likely adjacent Qwen2.5-VL rope/position-id logic, but one is still-image temporal scaling and the other is video vision_position_ids; not clearly the same bug." + "reason": "Same model family, different bug: still-image temporal position scaling versus video-position-id computation." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Non-persistent buffer junk vs ModelOutput key handling are unrelated." + "reason": "Non-persistent buffer junk initialization is unrelated to ModelOutput key assignment." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL multi-image finetuning." + "reason": "Generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL finetuning with multiple images." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and ModelOutput key assignment are different issues." + "reason": "CLIPVisionModel hidden_states regression is unrelated to ModelOutput key assignment." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Both are model regressions, but one is Qwen VL video input behavior and the other is a Gemma4 PLE device mismatch." + "reason": "Different concrete failures: Qwen VL video-input regression versus Gemma4 PLE device mismatch under device_map auto." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU offload device mismatch and torch.export/export compatibility with torch_compilable_check are unrelated." + "reason": "CPU-offload device mismatch and torch.export export failures through torch_compilable_check are different problems." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "Unexpected _is_hf_initialized argument and strict config rejection are different loading failures." + "reason": "init_empty_weights/_is_hf_initialized argument error is distinct from strict config validation for granite_speech." }, { "left": "issue:43296", "right": "issue:45800", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM and a torch/transformers compatibility issue are not the same bug." + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to torch/Transformers version incompatibility." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is special-token insertion and the other is wrong token IDs across several models." + "reason": "Tokenizer special-token regression for mdeberta-v3-base is not the same as broader AutoTokenizer wrong-token-ID regressions across multiple models." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to Qwen2.5-VL rope-index scaling." + "reason": "Checkpoint resume key mismatch is unrelated to Qwen2.5-VL rope index scaling." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "CPU offload device mismatch and DAC.from_latents vs forward mismatch are unrelated." + "reason": "Both mention inference mismatches, but CPU offload device placement and DAC.from_latents vs forward STE mismatch are different bugs." }, { - "left": "issue:44186", - "right": "issue:44265", + "left": "issue:41720", + "right": "issue:45823", "accept": false, - "reason": "Tokenizer crashes on NER/padding and export failures with torch_compilable_check are different code paths." + "reason": "Qwen3 auto device map CUDA assertion and Gemma4 PLE device mismatch are separate model-specific device-placement bugs." }, { - "left": "issue:41720", - "right": "issue:45823", + "left": "issue:44186", + "right": "issue:44265", "accept": false, - "reason": "Different models and different device-mapping failures; not the same bug." + "reason": "Tokenizer NER/padding crash and torch.export with torch_compilable_check are unrelated code paths." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "AyaVision tied-weight loading and Bamba torch.compile/SDPA failure are unrelated." + "reason": "AyaVision weight tying/generation failure is unrelated to Bamba-9B-v2 torch.compile SDPA failure." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sample-rate defaulting and GraniteSpeech PEFT checkpoint loading are separate issues." + "reason": "Audio chat-template sample-rate defaulting and PEFT checkpoint loading are unrelated." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "Different models and different load failures; no shared underlying bug." + "reason": "Two different model-loading failures with no clear shared root cause." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "SigLIP2 docs example errors and Qwen2.5-VL multi-image finetuning failures are unrelated." - }, - { - "left": "issue:40990", - "right": "issue:45584", - "accept": false, - "reason": "High perplexity on a language model and Whisper empty-transcription failure are unrelated." + "reason": "Documentation example errors for SigLIP2 are not the same bug as Qwen2.5-VL finetuning with multiple images." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI padding-mask inconsistency and OmDet-Turbo processor/model size mismatch are different bugs." + "reason": "MIMI padding-mask batch inconsistency and OmDet-Turbo processor/model resolution mismatch are different bugs." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "Tied lm_head weight loading and dtype mismatches in inference are unrelated." + "reason": "Tied lm_head weights causing bad generation is unrelated to bf16 dtype mismatch issues." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on gpt-oss and Whisper empty-transcription generation failure are unrelated." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Qwen3 MOE weight conversion and missing supports_gradient_checkpointing are different issues." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and missing supports_gradient_checkpointing on a trust_remote_code model are different issues." } ] }, @@ -3842,13 +3877,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 12, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:34634", "issue:35707", @@ -3859,9 +3894,7 @@ "issue:42617", "issue:42915", "issue:42947", - "issue:43295", "issue:43317", - "issue:43322", "issue:43377", "issue:43404", "issue:43450", @@ -3892,11 +3925,12 @@ "issue:45072", "issue:45357", "issue:45406", + "issue:45632", + "issue:45684", "issue:45701", "issue:45704", "issue:45759", "issue:45823", - "issue:46050", "issue:46291" ], "soft_pairs": [ @@ -3925,235 +3959,234 @@ "issue:36331|issue:42617", "issue:44423|issue:45406", "issue:42175|issue:46291", - "issue:43873|issue:44756", "issue:44387|issue:45704", + "issue:43873|issue:44756", "issue:44610|issue:44898", - "issue:43749|issue:44534", "issue:43550|issue:43606", + "issue:43749|issue:44534", "issue:44186|issue:44610", - "issue:43295|issue:45701", - "issue:43322|issue:46050" + "issue:45632|issue:45684" ], - "prepared_review_unit_hash": "0ec081241a5b9915be8e3686a448d24b265d2373a5a71fd3a60f05ca5fdcc236", + "prepared_review_unit_hash": "33054856126c2f2a82126e9db8c4e3596ca79469e4ef759093c0a01a4a44c0a9", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37721, - "estimated_input_tokens": 9431, - "estimated_eval_tokens": 19118 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36883, + "estimated_input_tokens": 9221, + "estimated_eval_tokens": 18698 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a loose topical neighborhood of unrelated Transformers issues; none of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", - "confidence": 0.39, - "canonical_issue_reason": "issue:43873 is the best representative because it is broad, currently open, and sits at the center of the offloading/quantization-related subset with multiple inbound references.", + "summary": "This cluster is mostly a set of unrelated Transformer bugs with a few broad thematic overlaps (offloading/loading, tokenization, multimodal processors). I would not merge them as duplicates; only a couple look superficially similar, but the underlying failures differ.", + "confidence": 0.79, + "canonical_issue_reason": "If forced to pick one representative, issue 43873 is the broadest and most connected item in the offloading/quantization area, with multiple soft links and inbound references. It is still only a theme hub, not a true duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the strongest single issue to anchor the cluster: broad symptom framing, open status, and the most plausible overlap with nearby offloading/quantization reports.", + "best_issue_reason": "Issue 43873 is the best single issue to anchor the cluster because it has the strongest cross-links and the most general failure description among the offloading-related reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Both mention offloading/quantization, but one is a specific load failure for a dequantized model while the other is a broader offloading regression; not the same concrete bug." + "reason": "Both involve offloading/loading, but one is a dequantized-model load failure with device_map=auto and the other is a broader quantization offloading issue; not the same concrete bug." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "StableLmConfig pad_token_idx and Qwen multimodal text_config unwrapping are different config-path bugs." + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to unwrapping text_config for composite Qwen multimodal configs." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Perplexity quality on a model/dataset is unrelated to gradient checkpointing behavior with LoRA." + "reason": "High perplexity on a model is not the same bug as gradient checkpointing not being effective with PEFT LoRA." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "A padding-mask discrepancy in MIMI is unrelated to a Perceiver resolution/interpolate_pos_encoding failure." + "reason": "MIMI batched-vs-single output mismatch due to padding masks is unrelated to Perceiver image resolution failures." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "General offloading/quantization issues are not the same as a version-dependent tokenization change." + "reason": "Quantization/offloading behavior is unrelated to tokenization changes across transformer versions." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "Weight-tying serialization for one model family is unrelated to unpacking packed weights during accelerate loading." + "reason": "lm_head weight tying in a multimodal model is a different code path from packed-weight unpacking during accelerate loading." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Both involve Qwen, but one is a device-map CUDA assert and the other is a flash-attention generation problem; different failure modes and code paths." + "reason": "Qwen3 auto device mapping cuda asserts and Qwen3.5 flash-attention generation failures are different failure modes." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Incorrect lm_head serialization by device is separate from saving wrong visual encoder keys." + "reason": "Both touch model save/serialization, but one is about tied lm_head weight serialization and the other about visual encoder keys in Qwen3.5 save_pretrained." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer loading failure and a kernels package version/revision requirement are unrelated issues." + "reason": "Tokenizer loading for a specific model is unrelated to kernels v0.15 requiring version/revision metadata." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "BeitImageProcessorFast label reduction and missing all_tied_weights_keys are unrelated." + "reason": "BeitImageProcessorFast label reduction and Molmo missing all_tied_weights_keys are unrelated." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "A model-loading failure is not the same as a tokenizer-loading regression." + "reason": "A model loading failure is not the same as a tokenizer-loading regression for another model." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "A script execution problem and a Qwen3Moe FP8 load failure do not share the same underlying bug." + "reason": "Running 3d_parallel.py and Qwen3Moe FineGrainedFP8Config failures are unrelated." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to auto device mapping CUDA asserts." + "reason": "Progressive generation with inputs_embeds/past_key_values is a different generation-path bug from Qwen3 device-mapping cuda asserts." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "Bark voice preset behavior and generation with inputs_embeds/past_key_values are different features." + "reason": "BarkProcessor voice_preset not working is unrelated to generation with inputs_embeds and past_key_values." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Dtype propagation in BLIP2 is unrelated to Trainer resume/max_steps calculations." + "reason": "Blip2 dtype handling and Trainer resume_from_checkpoint max_steps calculation are unrelated training/loading bugs." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "A variable-name collision in Gemma3n forward is unrelated to a Gemma4 device_map offload mismatch." + "reason": "Both are Gemma multimodal issues, but one is a variable-name collision in forward and the other is a device mismatch in PLE; not the same bug." }, { "left": "issue:43450", "right": "issue:45701", "accept": false, - "reason": "Batched video processor shape handling is not the same as a tokenization-version regression." + "reason": "Video processor batch-shape bugs do not match tokenization behavior changes across versions." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "lm_head weight tying and visual encoder key serialization are different save/load bugs." + "reason": "Tied lm_head weight serialization and incorrect visual encoder keys are different model-specific save_pretrained issues." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "A tokenizer crash on NER/batched padding is unrelated to bfloat16 dtype mismatches in inference." + "reason": "LayoutLMv2 tokenizer batching issues are unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config loading for granite_speech is not the same as a tokenizer file-loading failure." + "reason": "Strict config blocking granite_speech loading is unrelated to tokenizer loading for EMBEDDIA/est-roberta." }, { "left": "issue:43550", - "right": "issue:43606", + "right": "issue:43720", "accept": false, - "reason": "torch.compile+SDPA on Bamba and Bark CPU-offload device mismatch are unrelated runtime issues." + "reason": "Bamba torch.compile SDPA failures and BitNet packed-weight unpacking failures are different code paths and root causes." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Trainer compute_loss API mismatch is unrelated to Qwen3Moe fine-grained FP8 loading." + "reason": "Trainer compute_loss argument incompatibility is unrelated to Qwen3Moe FineGrainedFP8Config failures." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "The same Trainer callback signature bug does not overlap with the 3d_parallel.py execution failure." + "reason": "Unexpected num_items_in_batch in compute_loss has nothing to do with 3d_parallel.py not running." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Continuous-batching multimodal crashes and Gemma4Processor missing _tokenizer are different processor/runtime bugs." + "reason": "Continuous-batching multimodal serve crash due to string handling is different from Gemma4Processor missing _tokenizer." }, { - "left": "issue:44756", - "right": "issue:43873", + "left": "issue:42175", + "right": "issue:46291", "accept": false, - "reason": "Disabling mmap to avoid OOM on Strix Halo is unrelated to quantization offloading behavior." + "reason": "TensorFlow backend packaging and kernels version/revision metadata are unrelated." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "Int4 quantization memory growth in Transformers 5.x is unrelated to the apex FusedRMSNorm memory leak." + "reason": "Int4 quantization memory growth/OOM and an apex RMSNorm memory leak are different memory issues with different causes." }, { - "left": "issue:44610", - "right": "issue:44898", + "left": "issue:43873", + "right": "issue:44756", "accept": false, - "reason": "An OmDet-Turbo processor size mismatch and a Perceiver interpolation failure are different model-specific shape issues." + "reason": "Both mention loading/offloading and memory pressure, but Strix Halo mmap OOM is a platform-specific fix, not the same as quantization offloading behavior." }, { - "left": "issue:43749", - "right": "issue:44534", + "left": "issue:44610", + "right": "issue:44898", "accept": false, - "reason": "FSDP CPU RAM efficient loading and non-persistent buffer junk serialization are separate loading/serialization bugs." + "reason": "Processor output size mismatch and Perceiver non-default-resolution failure are both vision preprocessing issues, but they affect different models and mechanisms." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Bamba torch.compile/SDPA failure does not match Bark CPU-offload device mismatch." + "reason": "Bamba torch.compile SDPA and Bark-small CPU offload device mismatch are unrelated." }, { - "left": "issue:44186", - "right": "issue:44610", + "left": "issue:43749", + "right": "issue:44534", "accept": false, - "reason": "LayoutLMv2Tokenizer input handling and OmDet-Turbo processor output resolution are unrelated." + "reason": "FSDP CPU RAM efficient loading broken is not the same as non-persistent buffers being filled with junk." }, { - "left": "issue:43295", - "right": "issue:45701", + "left": "issue:44186", + "right": "issue:44610", "accept": false, - "reason": "A processor.tokenizer regression is not the same as a broad tokenization-version change." + "reason": "LayoutLMv2 tokenizer NER/batch handling and OmDet processor image-size mismatch are different subsystems." }, { - "left": "issue:43322", - "right": "issue:46050", + "left": "issue:45632", + "right": "issue:45684", "accept": false, - "reason": "A Llava Next loading segfault and a sinq save/load quantization issue are different bugs." + "reason": "trust_remote_code cache-path collisions and save_pretrained permission propagation are both custom-model save/load topics, but they are distinct bugs." } ] }, @@ -4169,13 +4202,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 13, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:29159", "issue:30064", @@ -4188,6 +4221,7 @@ "issue:41720", "issue:42617", "issue:42915", + "issue:43295", "issue:43316", "issue:43317", "issue:43322", @@ -4218,23 +4252,23 @@ "issue:45161", "issue:45406", "issue:45464", - "issue:45632", - "issue:45684", "issue:45701", "issue:45759", "issue:45812", + "issue:46050", "issue:46291" ], "soft_pairs": [ - "issue:45632|issue:45684", + "issue:43322|issue:46050", + "issue:43295|issue:45701", "issue:42915|issue:43317", "issue:41720|issue:43322", "issue:40990|issue:45701", "issue:41720|issue:42617", "issue:41720|issue:45161", "issue:36010|issue:42617", - "issue:45081|issue:45812", "issue:43425|issue:46291", + "issue:45081|issue:45812", "issue:43525|issue:46291", "issue:43316|issue:44913", "issue:43531|issue:43572", @@ -4245,8 +4279,8 @@ "issue:43526|issue:43881", "issue:29159|issue:36032", "issue:45406|issue:45464", - "issue:30064|issue:34634", "issue:43550|issue:43653", + "issue:30064|issue:34634", "issue:43606|issue:43720", "issue:43421|issue:45701", "issue:44530|issue:46291", @@ -4257,214 +4291,220 @@ "issue:34567|issue:35707", "issue:43388|issue:45059" ], - "prepared_review_unit_hash": "45363074c40d6320617997af9918f1f2613ed94890ebb403dc89d335ad224aa7", + "prepared_review_unit_hash": "f446bc5262eb4bf413769d9181efc340b6bce67efbaf8d8971284cfa71c0c19e", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36021, - "estimated_input_tokens": 9006, - "estimated_eval_tokens": 18268 + "soft_pair_count": 31, + "serialized_chars": 36309, + "estimated_input_tokens": 9078, + "estimated_eval_tokens": 18412 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items are mostly superficially similar by subsystem or model family, but the soft pairs do not describe the same concrete bug or change. None of the proposed pairs look safe to merge as duplicates.", - "confidence": 0.86, - "canonical_issue_reason": "Issue 45812 is the clearest broad regression report in the set: it has a specific user-visible failure, spans all Granite models, and includes inbound references suggesting it is well-anchored as a representative issue.", + "summary": "The set is a loose collection of mostly unrelated issues across tokenizers, processors, model loading, distributed inference, and docs. None of the soft pairs look like the same underlying bug/change, so no duplicate merges are justified.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45812 is the strongest single issue candidate: it is concrete, clearly scoped, and describes a broad v4\u2192v5 regression with multiple references and an obvious failure mode.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45812 is the strongest representative issue for triage because it describes a concrete, high-impact tokenizer regression with a clearly scoped failure mode.", + "best_issue_reason": "issue:45812 is the most actionable and representative issue here because it names the affected component, the regression nature, and the concrete symptom very clearly.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45632", - "right": "issue:45684", + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Different failure classes: model loading segfault for Llava Next vs quantization/save-load breakage for sinq models." + }, + { + "left": "issue:43295", + "right": "issue:45701", "accept": false, - "reason": "One is a cache-path collision for local models; the other is permissions propagation during save_pretrained. Different bugs and code paths." + "reason": "Both mention tokenization, but one is a processor/tokenizer API regression and the other is a broad version-dependent tokenization change; not the same bug." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "Both involve model loading/offload, but one is a Qwen3MoE FP8 failure and the other is a dequantized-model gpu+cpu offload issue. Not the same bug." + "reason": "Different subsystems and symptoms: Qwen3MoE FP8 failure vs device-map/offload loading failure for a dequantized model." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Auto device mapping cudaErrorAssert on Qwen3 is unrelated to a Llava Next segmentation fault during model load." + "reason": "Unrelated model-specific failures: auto device mapping assert on Qwen3 vs segfault while loading Llava Next." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "One reports high perplexity on gpt-oss-20b; the other is a broad tokenization version-change complaint. Too different to merge." + "reason": "High perplexity is a model/eval quality issue; the other is a tokenizer regression report. No shared concrete code path." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Qwen3 auto device mapping crash is unrelated to failing to run 3d_parallel.py." + "reason": "One is a Qwen3 device-mapping assert, the other is a generic script execution problem; no overlap in bug cause." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "A800 cudaErrorAssert under auto device mapping is a different problem from GPT-OSS MoE tensor-parallel not working." + "reason": "Both are distributed-model issues, but they involve different models and different breakpoints (auto device map vs tensor-parallel-only on GPT-OSS MoE)." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "GenerationMixin import failure has no overlap with a 3d_parallel.py runtime failure." + "reason": "ImportError for GenerationMixin is unrelated to a failure running 3d_parallel.py." }, { - "left": "issue:45081", - "right": "issue:45812", + "left": "issue:43425", + "right": "issue:46291", "accept": false, - "reason": "Mistral regex patching crashes while loading a tokenizer; Granite produces wrong token IDs. Both tokenizer-related, but different failure modes and fixes." + "reason": "A torch version incompatibility and a kernels import requirement are different dependency problems." }, { - "left": "issue:43425", - "right": "issue:46291", + "left": "issue:45081", + "right": "issue:45812", "accept": false, - "reason": "Torch compatibility and kernels import requirements are unrelated environment issues." + "reason": "Both are tokenizer regressions, but one is a Mistral regex patch crash and the other is wrong token IDs for Granite; different code paths and fixes." }, { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "Missing pad_token_id in Llama4Config is unrelated to kernels requiring version/revision." + "reason": "Missing pad_token_id in a config is unrelated to the kernels version/revision import error." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Gemma3TextConfig API mismatch is not the same as GPTNeoX rotary_pct not persisting on reload." + "reason": "Different config issues: Gemma3 text-config API discrepancy vs GPTNeoX rotary_pct persistence bug." }, { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and StableLmConfig pad_token_idx are different configuration bugs." + "reason": "Qwen3-MoE sliding-window behavior and StableLm missing pad_token_idx are unrelated config/model bugs." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "CamemBERT masked-LM regression and Qwen3.5 recurrent-state reset are unrelated model-specific bugs." + "reason": "CamemBERT masked-LM regression and Qwen modular recurrent-state reset are different model behavior issues." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Video-input processing in Qwen3OmniMoe is unrelated to Qwen3VL from_pretrained weight_only loading errors." + "reason": "Both involve Qwen multimodal models, but one is video-input processing and the other is weight_only loading." }, { "left": "issue:43421", "right": "issue:45812", "accept": false, - "reason": "Runtime special-token/post-processor update bug is different from Granite token-ID regression." + "reason": "Runtime special-token/post-processor synchronization is a different tokenizer bug than Granite token-ID regression." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "LayoutLMv2 NER/padding crash and MLuke task AttributeError are distinct tokenizer/task failures." + "reason": "Different tokenizer implementations and failure modes: LayoutLMv2 NER/padding crash vs MLuke task AttributeError." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." + "reason": "BEiT image-processor label reduction and glm-4v loading failure do not share an underlying cause." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Slow-vs-fast tokenizer inconsistency is not the same as T5Tokenizer's add_special_tokens method conflict." + "reason": "General fast/slow tokenizer inconsistency is not the same as T5Tokenizer\u2019s add_special_tokens conflict on import/load." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to Qwen3.5 streaming chat/completions failure." + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming inference crash are distinct API paths and symptoms." }, { - "left": "issue:30064", - "right": "issue:34634", + "left": "issue:43550", + "right": "issue:43653", "accept": false, - "reason": "Void segmentation-map processing and Bark voice_preset handling are different processor bugs." + "reason": "torch.compile/SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token decode behavior." }, { - "left": "issue:43550", - "right": "issue:43653", + "left": "issue:30064", + "right": "issue:34634", "accept": false, - "reason": "Bamba torch.compile SDPA failure is unrelated to BigBirdTokenizer mask-token registration." + "reason": "Image segmentation-map handling and Bark voice_preset processing are different processor bugs." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "CPU offload device mismatch and BitNet packed-weight unpacking during accelerate loading are different loading paths." + "reason": "Both are CI/loading issues, but they affect different models and mechanisms: CPU offload mismatch vs packed-weight unpacking." }, { "left": "issue:43421", "right": "issue:45701", "accept": false, - "reason": "One is a specific post-processor update bug; the other is a broad report that transformers version changes tokenization. Too vague to treat as the same bug." + "reason": "Special-token runtime updates and version-driven tokenization changes are not the same concrete bug." }, { "left": "issue:44530", "right": "issue:46291", "accept": false, - "reason": "PagedAttentionCache group-type crash and kernels version/revision import errors are unrelated." + "reason": "PagedAttentionCache group-type crash on Qwen3.5 has no overlap with the kernels import requirement issue." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Qwen3MoE FP8 loading failure and Llava Next segmentation fault are different model-loading issues." + "reason": "Different failures on different architectures/models; no shared code-path evidence." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Bark CPU offload device mismatch is unrelated to BigBirdTokenizer special-token registration." + "reason": "CPU offload device mismatch and token registration/decode failure are unrelated." }, { "left": "issue:43531", "right": "issue:45759", "accept": false, - "reason": "Sliding_window behavior in Qwen3-MoE and AutoModelForCausalLM.from_config text_config unwrapping are different Qwen-related bugs." + "reason": "Qwen3-MoE sliding_window behavior and AutoModel config unwrapping are different issues." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A cache-strategy typo and outdated docs for removed pipelines are unrelated." + "reason": "A typo in cache strategy docs/code and outdated pipeline-task docs are separate documentation problems." }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "TrainerState token count not updating is unrelated to progressive generation with inputs_embeds/past_key_values." + "reason": "TrainerState token-count tracking and progressive generation with inputs_embeds/past_key_values are unrelated." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "gather_for_metrics label truncation and SAM3 text/bounding-box behavior are different training/data-path bugs." + "reason": "Metric-gathering label truncation and SAM3 text/bounding-box behavior are different bugs in different components." } ] }, @@ -4480,13 +4520,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 14, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:33666", "issue:34634", @@ -4541,11 +4581,11 @@ "issue:44514|issue:44545", "issue:43873|issue:45704", "issue:42490|issue:43329", - "issue:43421|issue:43927", "issue:41720|issue:43746", + "issue:43421|issue:43927", "issue:44898|issue:45072", - "issue:45759|issue:45859", "issue:33666|issue:35707", + "issue:45759|issue:45859", "issue:44451|issue:44991", "issue:43531|issue:44589", "issue:45588|issue:45759", @@ -4570,226 +4610,226 @@ "issue:44625|issue:45479", "issue:43782|issue:45381" ], - "prepared_review_unit_hash": "e9422d73442396c708ec8cd98e03539ff6abb260f8ce66d0e4d0c3beb8ed54dc", + "prepared_review_unit_hash": "281cc107725bfb8c50ba977d6a57c9ec0769a4d16a89179a640fa6555bcfe63d", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37480, - "estimated_input_tokens": 9370, - "estimated_eval_tokens": 18996 + "serialized_chars": 37486, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Only one strong duplicate pair stands out: the two Qwen2_5_VLProcessor.apply_chat_template batched-input/padding=False issues. The rest are mostly separate model-specific bugs or unrelated tokenizer/offline/config regressions, so they should not be merged.", - "confidence": 0.91, - "canonical_issue_reason": "issue:44545 is the cleanest representative of the one clear duplicate pair: it states the exact batched-input/padding=False crash succinctly and matches issue:44514\u2019s underlying bug.", + "summary": "This cluster is mostly a set of unrelated issues; only one soft pair is a true duplicate. The Qwen2_5_VLProcessor batched apply_chat_template crash appears twice, while the other candidate pairs share only broad subsystem/topic similarity rather than the same bug.", + "confidence": 0.93, + "canonical_issue_reason": "issue:44545 is the clearest representative of the duplicated Qwen2_5_VLProcessor.apply_chat_template crash, with the same failure mode and slightly better linkage/wording.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44545 is the best overall issue to represent the cluster because it is the clearest formulation of the only accepted duplicate and is more concise than the near-identical variant.", + "best_issue_reason": "issue:44545 is the best issue representative because it is the exact bug report for the only confirmed duplicate pair and is the cleaner, more central report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Same API, same failure mode, and same trigger: Qwen2_5_VLProcessor.apply_chat_template crashes on batched input when padding is False." + "reason": "Same bug and same API path: Qwen2_5_VLProcessor.apply_chat_template crashing on batched input when padding=False." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Different problems: quantization/offloading behavior vs a T5 apex RMSNorm memory leak." + "reason": "Different problems: quantization/offloading behavior vs a T5 RMSNorm memory leak." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "Unrelated: local-path trailing-slash import warning vs a multimodal token counting bug in the video branch." + "reason": "Unrelated bugs: trailing-slash local path import warning vs a multimodal token counting video-branch NameError." }, { - "left": "issue:43421", - "right": "issue:43927", + "left": "issue:41720", + "right": "issue:43746", "accept": false, - "reason": "Different code paths: runtime special-token/post-processor sync vs DiaConfig custom token IDs being lost on save/load." + "reason": "Different models and failures: Qwen3 auto device mapping cuda assert vs GraniteSpeech PEFT local checkpoint loading." }, { - "left": "issue:41720", - "right": "issue:43746", + "left": "issue:43421", + "right": "issue:43927", "accept": false, - "reason": "Different models and failures: Qwen3 device-mapping cuda assert vs GraniteSpeech local-checkpoint loading with PEFT adapters." + "reason": "Different tokenizer/config issues: runtime special-token post-processor updates vs DiaConfig custom token IDs after save/load." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Different subsystems: image size/interpolation logic vs dtype mismatch during bfloat16 inference." + "reason": "Different code paths: Perceiver image resolution handling vs dtype mismatches in SwitchTransformers/TimmWrapperModel inference." }, { - "left": "issue:45759", - "right": "issue:45859", + "left": "issue:33666", + "right": "issue:35707", "accept": false, - "reason": "One is config unwrapping for composite multimodal models; the other is CPU-offload incompatibility in rotary embedding." + "reason": "Unrelated training/generation issues for different models and mechanisms." }, { - "left": "issue:33666", - "right": "issue:35707", + "left": "issue:45759", + "right": "issue:45859", "accept": false, - "reason": "Unrelated training issues: multi-GPU training vs progressive generation with inputs_embeds/past_key_values." + "reason": "Both involve Qwen 3.5 multimodal configs, but one is from_config text_config unwrapping and the other is CPU-offload rotary embedding compatibility; not the same bug." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are loading failures, but for different models and different tokenizer/model loading paths." + "reason": "Both are tokenizer-loading failures, but they affect different models and likely different root causes." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Unrelated: sliding_window behavior in Qwen3-MoE vs a missing Float8 storage type error." + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs missing Float8 storage type." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention sinkless-model AttributeError is unrelated to AutoModelForCausalLM text_config unwrapping." + "reason": "Different bugs: flash_attention sink handling vs composite config unwrapping in AutoModelForCausalLM.from_config." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "A warning about tie_word_embeddings is not the same as incorrect visual-encoder keys being saved." + "reason": "Same broad Qwen family, but one is a warning about tie_word_embeddings while the other is incorrect visual encoder keys on save_pretrained." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "AddedToken keyword collision is unrelated to a model loading failure." + "reason": "Different tokenizer issues: AddedToken keyword conflict vs loading a specific pretrained model." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Different failures: loading a specific model vs PIL image processors incorrectly requiring torchvision." + "reason": "Different subsystems: loading a specific model vs PIL backend image processor torchvision dependency regression." }, { "left": "issue:43421", "right": "issue:43653", "accept": false, - "reason": "Both involve tokenizers, but one is runtime post-processor updates and the other is BigBird mask-token registration/decoding." + "reason": "Different tokenizer bugs: runtime post-processor update vs BigBirdTokenizer mask token registration/empty decode." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Different documentation/model mismatch issues: SigLIP2 doc errors vs OmDet-Turbo processor output-size mismatch." + "reason": "Different documentation/model mismatch issue vs processor/model input-size mismatch." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "A T5 RMSNorm memory leak is unrelated to a missing supports_gradient_checkpointing flag." + "reason": "Different bugs: apex RMSNorm leak vs missing supports_gradient_checkpointing flag." }, { "left": "issue:45663", "right": "issue:45923", "accept": false, - "reason": "Gemma4 FSDP2 shared_kv_states reuse is unrelated to the Nemotron trust_remote_code gradient-checkpointing flag issue." + "reason": "Different failures: Gemma4 FSDP2 shared_kv_states bug vs missing gradient-checkpointing support flag." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Gradient checkpointing in eval mode is not the same as gradients increasing under HfDeepSpeedConfig ZeRO-3." + "reason": "Different training-mode behaviors: checkpointing disallowed in eval vs abnormal gradient growth under ZeRO-3." }, { "left": "issue:42886", "right": "issue:44843", "accept": false, - "reason": "Both involve offline mode, but the concrete causes differ: cache loading vs an unconditional model_info() call in _patch_mistral_regex." + "reason": "Both hit HF_HUB_OFFLINE, but one is a cache-load failure and the other is a specific Mistral regex/model_info offline regression." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Different trainer problems: unpacking Qwen3.5 input embeddings vs evaluate() failing after train()." + "reason": "Different trainer-related issues: Qwen3.5 embedding unpacking vs trainer.evaluate() after trainer.train()." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer-related, but the bugs differ: special-token post-processor updates vs LayoutLMv2 NER/batched padding crashes." + "reason": "Different tokenizer bugs affecting different models and failure modes." }, { "left": "issue:45704", "right": "issue:45859", "accept": false, - "reason": "A T5 fused RMSNorm leak and a Qwen3_5Moe rotary-embedding CPU-offload issue are unrelated." + "reason": "Different bugs: T5 memory leak vs Qwen3.5 MoE rotary embedding CPU-offload incompatibility." }, { "left": "issue:34634", "right": "issue:41720", "accept": false, - "reason": "Different models and causes: BarkProcessor voice preset vs Qwen3 auto device mapping cudaErrorAssert." + "reason": "Different model families and unrelated failures." }, { "left": "issue:43381", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/eval-mode behavior is unrelated to the T5 apex memory leak." + "reason": "Gradient checkpointing eval-mode restriction is unrelated to the T5 RMSNorm memory leak." }, { "left": "issue:43525", "right": "issue:43531", "accept": false, - "reason": "Different config/model bugs: missing pad_token_id on Llama4Config vs sliding_window handling for Qwen3-MoE." + "reason": "Different config/model issues: missing pad_token_id on Llama4Config vs Qwen3-MoE sliding_window behavior." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "They both mention tokenizer-regex behavior, but one is a backend_tokenizer AttributeError and the other is a Kimi-K2.5 codec/regression issue; not the same concrete bug." + "reason": "Related tokenizer area, but different concrete failures: backend_tokenizer attribute crash vs codec handling/incorrect warning regression." }, { "left": "issue:39692", "right": "issue:43873", "accept": false, - "reason": "Doc/example errors for SigLIP2 are unrelated to quantization/offloading behavior." + "reason": "Different problems: SigLIP2 docs/model mismatch vs quantization offloading behavior." }, { "left": "issue:43540", "right": "issue:46294", "accept": false, - "reason": "Both are video-input issues, but they involve different models and different failure points: processing ValueError vs apply_chat_template token/feature mismatch." + "reason": "Both concern Qwen3.5 video, but one is a ValueError in Qwen3OmniMoe video processing and the other is chat-template/video-token mismatch in Qwen3.5-9B." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Different classification issues: num_labels propagation in Qwen3.5 vs degenerate zero loss across sequence-classification models." + "reason": "Different classification/config bugs: num_labels propagation vs degenerate zero loss for single-label classification." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Different multimodal failures: Qwen3VL weight_only loading error vs qwen2.5-vl video vision_position_ids being wrong." + "reason": "Different multimodal model bugs: weight_only loading failure vs incorrect video vision_position_ids." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate pair, issue:44514 and issue:44545, is clearly supported by nearly identical titles. The other soft-edge rejections are phrased cautiously and do not overclaim mergeability." + "feedback": "The summary is well grounded in the packet: the only clearly duplicated pair is issue:44514/issue:44545, whose titles describe the same Qwen2_5_VLProcessor.apply_chat_template batched-padding crash. The remaining pair judgments stay conservative and do not overclaim duplicates. No pull-request mergeability concerns apply here because these are issue pairs, not PR pairs." }, "evaluator_used": true, "retried": false, @@ -4802,17 +4842,16 @@ "component_index": 1, "component_count": 18, "review_unit_index": 15, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:30064", "issue:36010", - "issue:36331", "issue:42915", "issue:43295", "issue:43335", @@ -4852,6 +4891,7 @@ "issue:45081", "issue:45183", "issue:45356", + "issue:45381", "issue:45479", "issue:45834", "issue:45907", @@ -4862,17 +4902,17 @@ "soft_pairs": [ "issue:45356|issue:45907", "issue:43994|issue:44671", - "issue:30064|issue:36010", "issue:43720|issue:43819", + "issue:30064|issue:36010", "issue:43746|issue:46050", - "issue:44077|issue:44625", "issue:43596|issue:44220", + "issue:44077|issue:44625", "issue:43653|issue:43720", "issue:42915|issue:43746", - "issue:43540|issue:44008", "issue:44220|issue:44661", - "issue:43550|issue:45072", + "issue:43540|issue:44008", "issue:43618|issue:43697", + "issue:43550|issue:45072", "issue:43746|issue:43873", "issue:43295|issue:45834", "issue:43550|issue:43819", @@ -4889,216 +4929,216 @@ "issue:43643|issue:44877", "issue:43421|issue:44811", "issue:44448|issue:44671", - "issue:30064|issue:36331" + "issue:43540|issue:45381" ], - "prepared_review_unit_hash": "fcc01433df960bd69a2cfc0f8f3a21587b7298db8021d7ec210cb1402588c40b", + "prepared_review_unit_hash": "2d46f29d098b03386e2fb72ecc58f5e86cf4202efac3f675d7bcc9a9e81ed331", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 37176, - "estimated_input_tokens": 9294, - "estimated_eval_tokens": 18844 + "serialized_chars": 37161, + "estimated_input_tokens": 9291, + "estimated_eval_tokens": 18838 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose bag of unrelated bug reports with a few weak tokenizer/loading similarities; there is no strong duplicate core. I would not merge any of the soft pairs.", - "confidence": 0.15, - "canonical_issue_reason": "No clear canonical duplicate exists in this set; issue 43873 is only the broadest/highest-activity open report, not a true representative of the others.", + "summary": "The cluster is mostly a grab bag of unrelated Transformers bug reports with only shallow keyword overlap around tokenizers, loading, and model regressions. I do not see any strong duplicate pairs; the soft matches are mostly same-subsysteem but different root causes.", + "confidence": 0.63, + "canonical_issue_reason": "Issue 44843 is the clearest, most concrete regression report with a single identifiable failing code path (`_patch_mistral_regex` / offline hub access), so it\u2019s the best anchor for this heterogeneous cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the best single representative by discussion/inbound references, but the cluster is too heterogeneous for a real canonical issue.", + "best_issue_reason": "Issue 44843 is the most actionable representative issue: specific symptom, clear repro surface, and it sits in a broader tokenizer-regression theme that appears multiple times in the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Different subsystems and failure modes: tokenizer codec/regression vs repo file listing forwarding." + "reason": "Both involve tokenizer-related behavior, but 45356 is a Mistral-regex/codec regression while 45907 is about forwarding `revision` in `list_repo_files`; different bugs." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Both are model-output regressions, but on different models and unrelated code paths." - }, - { - "left": "issue:30064", - "right": "issue:36010", - "accept": false, - "reason": "Image segmentation-map processing and a GenerationMixin import error are unrelated." + "reason": "Both are model-output quality regressions, but they affect different models and different code paths (`SigLIP2` vs `CamemBERT`)." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Packed-weight loading in BitNet and DAC latent decoding mismatch are separate bugs." + "reason": "One is a BitNet loading/unpacking failure, the other is a DAC latent/STE mismatch; unrelated functionality and failure modes." }, { - "left": "issue:43746", - "right": "issue:46050", + "left": "issue:30064", + "right": "issue:36010", "accept": false, - "reason": "Both involve checkpoint loading, but different models and different root causes." + "reason": "Image processor segmentation-map handling is unrelated to the `GenerationMixin` import error." }, { - "left": "issue:44077", - "right": "issue:44625", + "left": "issue:43746", + "right": "issue:46050", "accept": false, - "reason": "One is a post_init/config constraint issue; the other is num_labels propagation." + "reason": "Both mention loading/serialization, but one is PEFT adapter loading for GraniteSpeech and the other is quantized-model save/load corruption; different underlying problems." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "DeepSpeed/BERT init failure and fbank feature extraction are unrelated." + "reason": "Different failures in different paths: DeepSpeed ZeRO-3 init on BertModel vs `_torch_extract_fbank_features()` audio preprocessing." }, { - "left": "issue:43653", - "right": "issue:43720", + "left": "issue:44077", + "right": "issue:44625", "accept": false, - "reason": "Tokenizer special-token registration and BitNet packed-weight loading are unrelated." + "reason": "`patchtsmixer` post-init policy and Qwen3.5 `num_labels` propagation are unrelated config/model issues." }, { - "left": "issue:42915", - "right": "issue:43746", + "left": "issue:43653", + "right": "issue:43720", "accept": false, - "reason": "FP8 config failure for Qwen3Moe and GraniteSpeech PEFT checkpoint loading are different problems." + "reason": "BigBird special-token decode bug and BitNet packed-weight unpacking bug do not share the same root cause." }, { - "left": "issue:43540", - "right": "issue:44008", + "left": "issue:42915", + "right": "issue:43746", "accept": false, - "reason": "Both are multimodal-model bugs, but they affect different models and different tensors/inputs." + "reason": "Qwen3Moe FP8 config failure is not the same as GraniteSpeech local checkpoint loading with PEFT adapters." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio feature extraction and tokenizer mapping-name handling do not share the same bug." + "reason": "Audio feature extraction bug vs tokenizer-mapping-name failure when adding a model-like entry; different subsystems." }, { - "left": "issue:43550", - "right": "issue:45072", + "left": "issue:43540", + "right": "issue:44008", "accept": false, - "reason": "torch.compile+SDPA failure and bfloat16 dtype mismatches are different code-path issues." + "reason": "Video-input processing in Qwen3OmniMoe and Gemma 3n variable-name collision are distinct model-specific bugs." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "CLIPOutput attentions regression and RTDetrV2 output drift are separate model behaviors." + "reason": "CLIP attentions field regression and RTDetrV2 output differences are separate model API/behavior issues." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Both mention SDPA/dtype/compile issues, but one is a Bamba compile failure and the other is a dtype mismatch in two different models." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "Checkpoint loading for GraniteSpeech adapters vs quantization/offloading behavior are unrelated." + "reason": "Checkpoint/PEFT loading vs offloading-with-quantization are related only at a broad loading level, not the same bug." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Processor/tokenizer API regression and long OCR index failure are different issues." + "reason": "Processor/tokenizer regression and Kosmos2.5 OCR index error are different symptoms and code paths." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Different models and different failures: SDPA compile vs DAC latent conversion." + "reason": "Both are model/runtime bugs, but SDPA compile failure and DAC STE mismatch are unrelated." }, { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Input-validation messaging and classification-loss degeneracy are not the same bug." + "reason": "Validation error messaging across models is not the same as the single-label-classification loss degeneracy bug." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "Metric gathering truncation and Mamba2 cache-length correctness are unrelated." + "reason": "Metric-gathering last-batch behavior and Mamba2 cache correctness are different training/runtime issues." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "Tokenizer batching crash and Perceiver image-size handling are different subsystems." + "reason": "Tokenizer padding/truncation crash and Perceiver non-default-resolution failure are separate model-specific bugs." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and Smollm3 RoPE-layer dropping are different architecture/config bugs." + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are unrelated architecture/config regressions." }, { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Mistral image-preprocessor dtype mismatch and Gemma3n audio mask collision are different multimodal failures." + "reason": "Image preprocessor dtype mismatch in Mistral-3 is not the same as Gemma 3n audio tensor name collision." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "AddedToken construction error and sleng-bert loading failure do not appear to share a root cause." + "reason": "`AddedToken()` keyword collision and failure to load `cjvt/sleng-bert` are distinct tokenizer/load issues." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "Tokenizer special-token registration and DAC latent decoding are unrelated." + "reason": "Special-token registration bug and DAC latent mismatch do not share a code path or symptom." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Qwen3-MoE sliding-window handling and AutoConfig trust_remote_code field loss are separate issues." + "reason": "Qwen3-MoE sliding-window behavior and strict `trust_remote_code` config loading are unrelated." }, { "left": "issue:44843", "right": "issue:45081", "accept": false, - "reason": "Both touch _patch_mistral_regex, but one is offline model_info access and the other is a tokenizer attribute crash; not the same concrete bug." + "reason": "Both touch Mistral-regex handling, but 44843 is offline hub access from unconditional `model_info()` while 45081 is an AttributeError on tokenizer internals; close but not the same bug." }, { "left": "issue:43950", "right": "issue:45987", "accept": false, - "reason": "Non-persistent buffer corruption in from_pretrained and StaticCache seq-length return type are unrelated." + "reason": "Non-persistent buffer corruption in `from_pretrained()` is unrelated to `StaticCache.get_seq_length()` returning the wrong type." }, { "left": "issue:43643", "right": "issue:44877", "accept": false, - "reason": "Missing fields with trust_remote_code and strict granite_speech config loading are different config-path problems." + "reason": "`trust_remote_code` missing fields and strict config rejection for granite_speech are different config-loading failures." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Runtime special-token post-processing updates and Whisper batch_decode skip_special_tokens handling are different tokenizer behaviors." + "reason": "Runtime special-token post-processor update and Whisper `batch_decode()` ignoring `skip_special_tokens` are related to tokenization but not the same defect." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Different model regressions in v5, but no shared bug or code path is evident." + "reason": "Different model families and different regression surfaces: Pegasus generation output vs CamemBERT masked-LM predictions." }, { - "left": "issue:30064", - "right": "issue:36331", + "left": "issue:43540", + "right": "issue:45381", "accept": false, - "reason": "Segmentation-map processing and CustomTrainer compute_loss signature mismatch are unrelated." + "reason": "Qwen3OmniMoe video-processing error and Qwen2.5-VL video `vision_position_ids` issue are both multimodal, but the failing paths are distinct." } ] }, @@ -5114,24 +5154,25 @@ "component_index": 1, "component_count": 18, "review_unit_index": 16, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28218", "issue:29159", + "issue:30064", "issue:34634", "issue:36010", + "issue:36331", "issue:39692", "issue:42175", "issue:43322", "issue:43404", "issue:43452", - "issue:43540", "issue:43577", "issue:43582", "issue:43606", @@ -5160,7 +5201,6 @@ "issue:45072", "issue:45081", "issue:45084", - "issue:45381", "issue:45406", "issue:45584", "issue:45685", @@ -5171,17 +5211,17 @@ "issue:46209" ], "soft_pairs": [ - "issue:43540|issue:45381", - "issue:43577|issue:43818", + "issue:30064|issue:36331", "issue:43782|issue:44977", + "issue:43577|issue:43818", "issue:39692|issue:45834", "issue:45584|issue:45834", "issue:44821|issue:45084", "issue:43322|issue:45685", "issue:43606|issue:45072", "issue:44479|issue:45834", - "issue:44186|issue:44442", "issue:43825|issue:43994", + "issue:44186|issue:44442", "issue:34634|issue:36010", "issue:44485|issue:44509", "issue:28218|issue:29159", @@ -5200,212 +5240,209 @@ "issue:44671|issue:45701", "issue:42175|issue:45081" ], - "prepared_review_unit_hash": "c341d05d9797c6f4f596ffdb922124bf8bd15c545e6014aff2b7bc10dcf9a744", + "prepared_review_unit_hash": "406fb5e0e8454fed6a9c649241042ac7580872901fa1fe3f9d671b48af9a4054", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 35521, - "estimated_input_tokens": 8881, - "estimated_eval_tokens": 18018 + "serialized_chars": 35532, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous mix of unrelated tokenizer, multimodal, backend, and docs/runtime regression issues. Only two pairs look plausibly duplicate-like: the tokenizer added-token spacing bug and the remote_code/v5 breakage report.", - "confidence": 0.68, - "canonical_issue_reason": "issue:45020 is the broadest regression report in the set and the best anchor for the small v5/remote_code-related overlap; the rest of the cluster is too heterogeneous for a stronger canonical choice.", + "summary": "This set is mostly heterogeneous. The only recurring theme is tokenizer/version-regression reports, but none of the soft pairs clearly describe the same concrete bug or change, so I rejected all candidate merges.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45701 is the broadest live tokenizer-regression report and has the strongest cross-references; it is the best anchor for the small tokenizer-related theme, even though the overall item set is not a true duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the most useful overall issue anchor because it captures the broadest downstream breakage pattern among the few related items, even though most items are not true duplicates.", + "best_issue_reason": "issue:45701 is the most generally representative issue in this collection because it captures the broad tokenizer-version regression theme and is open with multiple inbound references.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43540", - "right": "issue:45381", + "left": "issue:30064", + "right": "issue:36331", "accept": false, - "reason": "Both are video-input bugs, but they affect different model families and likely different code paths." + "reason": "Image processing void-segmentation handling and a custom trainer signature error are unrelated code paths." }, { - "left": "issue:43577", - "right": "issue:43818", + "left": "issue:43782", + "right": "issue:44977", "accept": false, - "reason": "Different failures: Blip2 dtype loading vs Video-LLaVA missing temporal attention/weight sharing." + "reason": "Qwen3VL weight_only loading and Qwen3.5 flash-attention generation are different failures." }, { - "left": "issue:43782", - "right": "issue:44977", + "left": "issue:43577", + "right": "issue:43818", "accept": false, - "reason": "Different model and symptom: Qwen3VL load-time weight_only error vs Qwen3.5 flash-attention generation failure." + "reason": "BLIP2 dtype casting and Video-LLaVA missing temporal attention are distinct model bugs." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "Both mention vision/OCR examples, but they are different models and different failure modes." + "reason": "SigLIP2 doc example errors and Kosmos2.5 OCR index errors are unrelated." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty-transcription handling and Kosmos2.5 long OCR index errors are unrelated." + "reason": "Whisper empty-transcription generation and Kosmos2.5 OCR indexing are different problems." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor-from-URL loading and template compilation TypeError are unrelated." + "reason": "AutoImageProcessor-from-URL loading and template compilation errors do not share a concrete failure mode." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Llava Next segfault on model load is unrelated to MPS Int histogram support errors." + "reason": "Llava Next segfaults and an MPS histogram integer error are separate backend/model issues." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "CPU offload device mismatch and bfloat16 dtype mismatches are different bugs." + "reason": "CPU offload device mismatch and bfloat16 dtype mismatches are not the same bug." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Both are vision-related, but one is a video-input regression and the other is a long OCR index error." + "reason": "Qwen video regression and Kosmos2.5 OCR index errors are unrelated." }, { - "left": "issue:44186", - "right": "issue:44442", + "left": "issue:43825", + "right": "issue:43994", "accept": false, - "reason": "Tokenizer crash on NER/padding is not the same as AutoTokenizer failing to load a specific tokenizer class." + "reason": "A pipeline v5 error-message wording issue and a SigLIP2 model/pipeline result issue are different changes." }, { - "left": "issue:43825", - "right": "issue:43994", + "left": "issue:44186", + "right": "issue:44442", "accept": false, - "reason": "One is a v5 pipeline error-message issue; the other is a SigLIP2 model/pipeline correctness bug." + "reason": "LayoutLMv2 tokenizer crashes and AutoTokenizer loading FastSpeech2ConformerTokenizer are unrelated tokenizer paths." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset and GenerationMixin import errors are unrelated." + "reason": "BarkProcessor voice_preset handling and GenerationMixin import removal affect different subsystems." }, { "left": "issue:44485", "right": "issue:44509", "accept": false, - "reason": "GLM-5 RoPE discussion is unrelated to docs still mentioning removed pipeline tasks." + "reason": "GLM-5 RoPE implementation discussion and v5 pipeline-task doc cleanup are not the same bug." }, { "left": "issue:28218", "right": "issue:29159", - "accept": true, - "reason": "Same underlying tokenizer regression: added-token spacing behavior differs between tokenizer implementations." + "accept": false, + "reason": "Both are tokenizer-related, but one is a specific added-token spacing bug and the other is a broader slow/fast inconsistency report." }, { "left": "issue:44561", "right": "issue:45020", - "accept": true, - "reason": "44561 looks like a specific v5 remote_code breakage caused by the broader regression described in 45020." + "accept": false, + "reason": "Removal of is_torch_fx_available and remote_code model breakage are different API breakages." }, { "left": "issue:45987", "right": "issue:46192", "accept": false, - "reason": "StaticCache return type mismatch and return_tensors being ignored are different API bugs." + "reason": "StaticCache sequence-length return type and return_tensors being ignored are unrelated behavior changes." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "Both involve dtype/type issues, but on different subsystems and platforms." + "reason": "BLIP2 dtype propagation and Apple Silicon allocator warmup TypeError are unrelated." }, { "left": "issue:44610", "right": "issue:45072", "accept": false, - "reason": "Different models and different root causes: processor size mismatch vs dtype mismatch in inference." + "reason": "OmDet-Turbo processor size mismatch and dtype mismatch issues are different model/precision problems." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Mask-shape IndexError and trainer.evaluate-after-train failure are unrelated." + "reason": "Mask shape IndexError and trainer evaluate-after-train failure are distinct error scenarios." }, { "left": "issue:43720", "right": "issue:45072", "accept": false, - "reason": "Packed-weight loading in BitNet is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "BitNet packed-weight unpacking and dtype mismatch inference bugs do not share the same code path." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading failures, but for different models and likely different causes." + "reason": "gguf_file loading breakage and a model-specific tokenizer load failure in v5 are not the same underlying bug." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "Closely related v5 pipeline-removal theme, but one is runtime messaging and the other is stale documentation, not the same bug." + "reason": "Both mention removed pipeline tasks, but one is an error message issue and the other is stale documentation." }, { "left": "issue:43404", "right": "issue:43782", "accept": false, - "reason": "Different model families and different generation/loading failures." + "reason": "Mistral3 lm_head tying and Qwen3VL weight_only loading are different model-loading defects." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "AppleSilicon warmup TypeError and CUDA AMP loss crash are unrelated." + "reason": "Apple Silicon allocator warmup and float16 AMP CUDA loss crashes are unrelated runtime issues." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Mistral regex patch crash and Gemma4Processor serve crash are different code paths." + "reason": "Mistral regex patching and Gemma4Processor missing _tokenizer are separate tokenizer/processor bugs." }, { "left": "issue:43906", "right": "issue:46209", "accept": false, - "reason": "Isolated reproduction of one issue and FP8 B200 inference failure are unrelated." + "reason": "A reproduction of an earlier issue and an FP8 inference failure on B200 are not the same change." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "CamemBERT MLM prediction regression and a broad tokenization-version-change report are not clearly the same bug." + "reason": "CamemBERT masked-LM prediction regressions and general tokenization changes are different failures." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Missing TensorFlow backend in a torch install and Mistral tokenizer regex patching are unrelated." + "reason": "Missing TensorFlow in the torch extra and Mistral tokenizer regex patching are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays mostly conservative. The two accepted soft pairs are plausibly related, while the rest are reasonably rejected as only loosely similar. The canonical anchor choice is defensible for the remote_code/v5 subset, though 'broadest' is a bit strong given other broad issues in the set." - }, - "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 }, @@ -5414,13 +5451,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 17, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:34634", "issue:36010", @@ -5490,8 +5527,8 @@ "issue:44387|issue:44617", "issue:34634|issue:36331", "issue:43854|issue:44977", - "issue:36010|issue:43645", "issue:39692|issue:44912", + "issue:36010|issue:43645", "issue:38175|issue:44610", "issue:43720|issue:44610", "issue:45459|issue:45803", @@ -5503,214 +5540,214 @@ "issue:43531|issue:45042", "issue:43257|issue:43329" ], - "prepared_review_unit_hash": "c385fe9933a20854333e42e4a00a272ae914975f9bc13bceb5da65f38d27ddc2", + "prepared_review_unit_hash": "4c9c60b7b92bb8906755388c0500953e639f521c0e26f92e828352333c6a52b0", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36927, - "estimated_input_tokens": 9232, - "estimated_eval_tokens": 18720 + "serialized_chars": 36924, + "estimated_input_tokens": 9231, + "estimated_eval_tokens": 18718 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft pairs mostly connect unrelated bugs across different models, processors, loading paths, and training features. I found no pair that clearly represents the same underlying fix, so there is no strong canonical issue here.", + "summary": "These items are mostly unrelated model- or component-specific regressions; none of the soft pairs look like the same underlying bug or change, so no merges are justified.", "confidence": 0.92, - "canonical_issue_reason": "No single canonical issue stands out; the items are mostly separate bugs with only superficial theme overlap.", + "canonical_issue_reason": "issue:45020 is the broadest and most representative regression report in the set: it captures a general class of recent Transformers breakages affecting remote_code models, unlike the more narrowly scoped model/component-specific issues.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a good global representative because the set does not cohere around one concrete bug or code path.", + "best_issue_reason": "issue:45020 is the strongest standalone issue candidate because it is upstream-facing, broadly scoped, and more representative of the cluster than the highly specific bugs around single models, processors, or trainers.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio processor chat-template sampling-rate behavior is unrelated to DAC forward/from_latents mismatch and missing STE." + "reason": "Audio chat-template sampling-rate defaults and DAC latent reconstruction are unrelated code paths and failure modes." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "BigBirdTokenizer special-token decode bug is unrelated to MXFP4 quantization fallback/loading behavior." + "reason": "Tokenizer special-token registration and MXFP4 quantization loading are different bugs with no shared underlying fix." }, { "left": "issue:45003", "right": "issue:46285", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to CLIP text-model structure inconsistency." + "reason": "Unsafe sys.modules access and CLIP text model structure inconsistency are separate modeling-utils vs architecture issues." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "BEiT fast label reduction bug is unrelated to RTDetrV2 output drift in Transformers v5." + "reason": "Beit label reduction and RTDetr output drift are different processor/model problems; they would not be merged into one fix." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both mention Qwen tie-weight warnings, but they describe different model families and different symptoms; not clearly the same concrete bug." + "reason": "Incorrect weight tying in Qwen3-Next and a tied-embeddings warning during LoRA fine-tuning are not the same concrete bug." }, { "left": "issue:43650", "right": "issue:43906", "accept": false, - "reason": "A vague 'add the data' issue is not the same as an isolated reproduction of a specific upstream bug." + "reason": "A placeholder 'ADD THE DATA' issue and an isolated reproduction of another bug are not duplicates." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "MobileLLM load KeyError is a specific model-loading failure, while remote_code breakage is a broader recent-version regression." + "reason": "Loading a specific MobileLLM checkpoint failing with KeyError is distinct from the broad remote_code regression report." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template-node compile failure is unrelated to CLIPTokenizer's absurd model_max_length value." + "reason": "Template compilation errors and CLIPTokenizer max-length semantics are unrelated issues." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Undefined video-token helper variables are unrelated to stale auto_mappings references to removed configs." + "reason": "A multimodal token counting video-branch bug and stale auto_mappings references are different maintenance problems." }, { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "GraniteSpeech local-checkpoint/PEFT loading is unrelated to Whisper batch_decode ignoring skip_special_tokens." + "reason": "PEFT checkpoint loading for GraniteSpeech and Whisper batch_decode skip_special_tokens behavior are unrelated." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to PIL image processors requiring torchvision." + "reason": "Missing Float8 storage lookup and unnecessary torchvision dependency in PIL image processors affect different subsystems." }, { "left": "issue:43644", "right": "issue:43749", "accept": false, - "reason": "Non-persistent buffer junk in v5.0 and FSDP CPU RAM efficient loading are different loading-path regressions, not the same bug." + "reason": "Non-persistent buffer initialization corruption and FSDP CPU RAM efficient loading are both loading-related but not the same code-path bug." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "Zero3 index error with non-pretrained BERT is unrelated to trainer.evaluate() failing after train()." + "reason": "Zero3 empty-index failure on non-pretrained Bert and evaluate-after-train behavior are separate trainer/distributed-training issues." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "Torch.compile/SDPA failure in Bamba is unrelated to MXFP4 quantization fallback for git-oss-20b." + "reason": "torch.compile+SDPA failure in Bamba and MXFP4 fallback during load for git-oss are different model/runtime regressions." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "DAC STE mismatch and DiaConfig token-ID loss cause different failures in different code paths." + "reason": "DAC latent mismatch and DiaConfig token-ID persistence loss are unrelated implementation bugs." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Int4 quantization memory regression is unrelated to Sam3Video CUDA OOM." + "reason": "General int4 quantization memory regression and Sam3Video OOM are not the same concrete defect." }, { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "Bark voice_preset and CustomTrainer compute_loss signature errors are unrelated." + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated API issues." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "GLM-4.7-Flash load-test failure is unrelated to Qwen3.5 flash-attention generation issues." + "reason": "Model loading test failure for GLM-4.7-Flash and flash-attention generation issues in Qwen3.5 are different bugs." }, { - "left": "issue:36010", - "right": "issue:43645", + "left": "issue:39692", + "right": "issue:44912", "accept": false, - "reason": "GenerationMixin import error is unrelated to Jupyter custom-model initialization regression." + "reason": "SigLIP2 doc/example errors and MXFP4 load fallback are distinct; one is docs/model mismatch, the other quantization loading." }, { - "left": "issue:39692", - "right": "issue:44912", + "left": "issue:36010", + "right": "issue:43645", "accept": false, - "reason": "SigLIP2 docs/model mismatch and quantization failure is unrelated to MXFP4 loading fallback." + "reason": "GenerationMixin import breakage and Jupyter custom model initialization breakage are separate import/runtime issues." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Zero probabilities in SigLIP2 are unrelated to OmDet-Turbo processor/model input-size mismatch." + "reason": "Unexpected zero probabilities in SigLIP2 and OmDet-Turbo processor size mismatch are different model/processor regressions." }, { "left": "issue:43720", "right": "issue:44610", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to OmDet-Turbo image-size mismatch." + "reason": "BitNet packed-weight unpacking during accelerate loading and OmDet-Turbo input-size mismatch are unrelated." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Both are about exception handling, but they are different code paths and different failure modes; not the same bug." + "reason": "Tokenizer protobuf error masking and FuyuBatchFeature bare except swallowing interrupts are both error-handling smells but not the same fix or code path." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "BEiT label reduction bug is unrelated to the MPS attention correctness issue." + "reason": "Beit label reduction and MPS attention head-dimension correctness are unrelated bugs." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "AyaVision weight-tying failure is unrelated to BLIP2 dtype propagation bug." + "reason": "AyaVision weight tying and BLIP2 dtype propagation are distinct model initialization issues." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "ImportError for GenerationMixin and CustomTrainer.compute_loss signature mismatch are distinct issues." + "reason": "An import error for GenerationMixin and a custom trainer loss-signature error are unrelated failures." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio processor chat-template sampling rate is unrelated to Whisper decode token-skipping behavior." + "reason": "Audio apply_chat_template sampling-rate defaulting and Whisper batch_decode skip_special_tokens behavior are separate processor bugs." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "FSDP CPU RAM efficient loading and init_empty_weights unexpected argument are different loading regressions." + "reason": "FSDP CPU RAM loading and init_empty_weights _is_hf_initialized argument mismatch are different loading mechanisms and fixes." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior is unrelated to PIL backend torchvision dependency regression." + "reason": "Qwen3-MoE sliding_window and PIL backend torchvision dependency are unrelated regressions." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Qwen3 MOE loading/conversion and multimodal token counting are unrelated problems." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and multimodal token counting video-branch undefined names are unrelated." } ] }, @@ -5726,13 +5763,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 18, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:24643", "issue:29159", @@ -5796,8 +5833,8 @@ "issue:42915|issue:43742", "issue:44291|issue:45042", "issue:43257|issue:43931", - "issue:43927|issue:44912", "issue:43452|issue:43582", + "issue:43927|issue:44912", "issue:43697|issue:45357", "issue:44568|issue:45488", "issue:33357|issue:34567", @@ -5805,9 +5842,9 @@ "issue:41720|issue:43540", "issue:42915|issue:45538", "issue:44987|issue:45538", + "issue:44479|issue:45381", "issue:43540|issue:44479", "issue:43317|issue:45538", - "issue:44479|issue:45381", "issue:43540|issue:44560", "issue:44987|issue:45907", "issue:45137|issue:45405", @@ -5818,20 +5855,20 @@ "issue:38175|issue:44898", "issue:43257|issue:43696" ], - "prepared_review_unit_hash": "14d35e121b3b244f0082e2d9fc13b1c052fd5ea1235199bee5b1c2a35374b214", + "prepared_review_unit_hash": "7863ce51a072931d6ac40d058aec43ecb68585198b5a3d073426d05f61484e21", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37726, + "serialized_chars": 37725, "estimated_input_tokens": 9432, "estimated_eval_tokens": 19120 }, @@ -5840,220 +5877,208 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose set of unrelated Transformers bug reports, with only one likely duplicate-style pair around Qwen video-input regressions. The strongest umbrella issue is the Qwen video regression report, but it is still broader than most others in the set.", - "confidence": 0.78, - "canonical_issue_reason": "issue:44479 is the broadest and most central report in the only clear thematic pocket here (Qwen video-input regressions across multiple model families), so it best serves as the cluster anchor.", + "summary": "The cluster is mostly heterogeneous across tokenizer, loading, quantization, device-map, and training bugs. The only plausible subcluster is the Qwen video regressions, but the evidence is only soft title similarity; there are no explicit duplicate links or shared-target signals to confirm a merged duplicate set.", + "confidence": 0.67, + "canonical_issue_reason": "issue:44479 is the best available issue representative for the small Qwen video-regression subcluster because it broadly names the shared video-input failure mode. That said, the packet does not show explicit duplicate links or shared-target evidence, so this is only a cautious representative choice, not a confirmed duplicate hub.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best standalone representative because it describes a multi-model video-input regression in v5.3.0 and is the most umbrella-like issue among the candidates.", + "best_issue_reason": "issue:44479 is the strongest overall issue pick because it is the broadest, most central report among the only closely related video issues. The rest of the cluster is too mixed, and the related Qwen entries remain only tentatively similar.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Both are model-loading/configuration bugs, but one is about missing fields from trust_remote_code configs and the other about from_config not unwrapping text_config; different failure modes and code paths." + "reason": "Both involve config/model loading, but they describe different failures and code paths." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Trailing-slash path import warnings and Qwen3 MoE weight conversion under accelerate+deepspeed are unrelated problems." + "reason": "Trailing-slash local-path import warning is unrelated to Qwen3 MOE weight conversion." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Both involve video processing, but one is an internal multimodal token-count helper bug and the other is a batched-shape issue in video processors; not the same underlying defect." + "reason": "Both are video-related, but one is a missing helper in multimodal token counting and the other is a batched video processor shape bug." }, { "left": "issue:43606", "right": "issue:45859", "accept": false, - "reason": "Both mention CPU offload, but they affect different models and different code paths, so they are not the same bug." + "reason": "Both mention CPU offload, but they are different model-specific incompatibilities." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token accounting and GPT-oss perplexity are unrelated." + "reason": "TrainerState token accounting and GPT-oss perplexity are unrelated failures." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "These are separate tokenizer regressions with different symptoms and likely different root causes." + "reason": "These are separate tokenizer regressions affecting different models and behaviors." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to TrainerState's num_input_tokens_seen not updating." + "reason": "Slow/fast tokenizer inconsistency is unrelated to TrainerState token counting." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "Both concern video paths, but 43329 is a helper-function undefined-variable bug while 44479 is a broader video-input regression across Qwen models." + "reason": "They are both in the Qwen video area, but the concrete symptoms differ and the packet does not prove the same bug." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "Qwen3.5 output_hidden_states behavior and a generic tokenization-change report are not the same concrete bug." + "reason": "One is a specific Qwen3.5 output bug; the other is a broad tokenization regression report." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Qwen3Moe FP8 config failure and MobileLLM key errors are distinct model-loading issues." + "reason": "Qwen3Moe FineGrainedFP8Config failure and MobileLLM-125M key error are different loading problems." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized and PIL backend torchvision requirements are different failures." + "reason": "init_empty_weights/_is_hf_initialized and PIL-backend torchvision requirement are unrelated." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Qwen3 MoE weight-conversion/loading and Qwen3-VL weight-shape mismatch are different model-specific problems." + "reason": "Qwen3 MOE weight conversion and Qwen3-VL weight-shape mismatch are different loading bugs." }, { - "left": "issue:43927", - "right": "issue:44912", + "left": "issue:43452", + "right": "issue:43582", "accept": false, - "reason": "DiaConfig save/load token IDs and GPT-oss MXFP4 quantization loading are unrelated." + "reason": "gguf_file loading breakage and AppleSilicon TypeError are unrelated." }, { - "left": "issue:43452", - "right": "issue:43582", + "left": "issue:43927", + "right": "issue:44912", "accept": false, - "reason": "gguf_file / from_pretrained loader breakage is unrelated to an Apple Silicon allocator warmup TypeError." + "reason": "DiaConfig token-ID loss and MXFP4 quantization fallback are different bugs." }, { "left": "issue:43697", "right": "issue:45357", "accept": false, - "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained key issues are unrelated regressions." + "reason": "Different model families and different symptoms: RTDetrV2 output drift versus Qwen3.5 save_pretrained key handling." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Both are tokenizer-related, but one is BOS/EOS insertion behavior and the other is a LlamaTokenizer pre-tokenizer override affecting DeepSeek models." + "reason": "Both are tokenizer-related, but they affect different models and failure modes." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "A MacOS bus error in CLIP loading and a TrainerState counter bug are unrelated." + "reason": "MacOS bus error with CLIP and TrainerState token counting are unrelated." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "Qwen3Moe FP8 loading and AutoImageProcessor-from-URL are different features and failures." + "reason": "Qwen3Moe FP8 config failure and AutoImageProcessor URL loading are unrelated." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Qwen3 auto device-mapping CUDA assert and Qwen3OmniMoe video processing ValueError are not the same bug." + "reason": "Auto device mapping CUDA assert and Qwen3OmniMoe video ValueError are distinct failures." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "Qwen3Moe FP8 loading and CLIPTokenizer model_max_length are unrelated issues." + "reason": "Qwen3Moe FP8 config failure and CLIPTokenizer model_max_length bug are unrelated." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Loading physical-intelligence/fast and CLIPTokenizer max-length behavior are unrelated." - }, - { - "left": "issue:43540", - "right": "issue:44479", - "accept": false, - "reason": "Both are video-related, but one is a ValueError in Qwen3OmniMoe processing and the other is a broader multi-model video regression; not enough evidence they share the same root cause." - }, - { - "left": "issue:43317", - "right": "issue:45538", - "accept": false, - "reason": "Dequantized model offload loading and CLIPTokenizer max-length are unrelated." + "reason": "Loading physical-intelligence/fast and CLIPTokenizer model_max_length are unrelated." }, { "left": "issue:44479", "right": "issue:45381", - "accept": true, - "reason": "Both describe the same family of Qwen video-input regressions, centered on incorrect video-position handling in v5.3.0." + "accept": false, + "reason": "Both are Qwen video regressions, but 45381 looks like a specific vision_position_ids symptom and the packet does not confirm the same underlying code-path bug." }, { "left": "issue:43540", - "right": "issue:44560", + "right": "issue:44479", "accept": false, - "reason": "Qwen3OmniMoe video ValueError and Qwen3-vl-embedding StopIteration are different video bugs." + "reason": "Both involve video, but the Qwen3OmniMoe StopIteration report is not clearly the same regression as 44479." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "A model load failure and a list_repo_files revision-forwarding CI bug are unrelated." + "reason": "Model loading failure and list_repo_files revision forwarding are unrelated." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow and a PEFT minimum-version bump are unrelated." + "reason": "DeepSpeed ZeRO3 deque error and MIN_PEFT_VERSION packaging issue are unrelated." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed training weight-dimension error and tokenizer slow/fast inconsistency are unrelated." + "reason": "DeepSpeed weight-dimension error and tokenizer inconsistency are unrelated." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "Both are loading-related, but one is a missing key for MobileLLM and the other is an unexpected argument under init_empty_weights; different code paths." + "reason": "MobileLLM-125M key error and unexpected _is_hf_initialized are different loading failures." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Blip2 dtype persistence and SwitchTransformers/TimmWrapperModel bfloat16 mismatches are distinct dtype bugs." + "reason": "Both mention dtype, but the underlying bugs differ." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "AddedToken duplicate special-argument handling and EMBEDDIA/est-roberta tokenizer loading are unrelated tokenizer issues." + "reason": "AddedToken special-argument duplication and EMBEDDIA tokenizer loading are different tokenizer bugs." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and Perceiver interpolation failures are different vision-model bugs." + "reason": "Unexpected zero probabilities in SigLIP2 and Perceiver non-default-resolution failure are unrelated image-model issues." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MoE weight conversion/loading and GPT-oss CUDA OOM are not the same issue." + "reason": "Qwen3 MOE weight conversion and GPT-oss CUDA OOM are unrelated." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is well grounded in the packet: most items are indeed disparate bug reports, and the only clearly related pair is the Qwen video-input regression pocket centered on issue:44479 / issue:45381. The canonical-issue choice is conservative and supported by the breadth of issue:44479. I didn\u2019t see overstatement in the soft-edge verdicts." + "accept": false, + "feedback": "The overall clustering judgment is mostly conservative, but the canonical/best-issue reasoning overstates the evidence. In the packet there are no explicit duplicate links or shared-target evidence, so describing issue:44479 as a \u201cconfirmed duplicate-worthy subset\u201d and citing a \u201cdirect duplicate-style link\u201d is too strong. The Qwen video issues do look plausibly related by titles, but the packet only supports a tentative soft similarity, not confirmation. A more cautious summary would be acceptable." }, "evaluator_used": true, - "retried": false, - "accepted_nontrivial_soft_edge": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -6062,13 +6087,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 19, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28282", "issue:29127", @@ -6085,7 +6110,7 @@ "issue:43531", "issue:43643", "issue:43646", - "issue:43720", + "issue:43697", "issue:43742", "issue:43749", "issue:43756", @@ -6102,11 +6127,11 @@ "issue:44792", "issue:44821", "issue:44877", - "issue:44898", "issue:44913", "issue:44964", "issue:45042", "issue:45081", + "issue:45216", "issue:45325", "issue:45357", "issue:45375", @@ -6123,8 +6148,8 @@ "issue:44737|issue:45588", "issue:43404|issue:45987", "issue:34567|issue:36331", - "issue:43646|issue:43749", "issue:44062|issue:45081", + "issue:43646|issue:43749", "issue:44913|issue:46121", "issue:43493|issue:43994", "issue:43906|issue:45381", @@ -6146,204 +6171,204 @@ "issue:43756|issue:45325", "issue:43531|issue:44821", "issue:44625|issue:44877", - "issue:43720|issue:44898" + "issue:43697|issue:45216" ], - "prepared_review_unit_hash": "061885a9824ae469d6676f97ad49409d4157a70ed5c958d5da4c56e8dc4aad71", + "prepared_review_unit_hash": "d3a084ed0677bf37bb7fed8cbbb4dc0551a772033067fd8c0d2154c0e3808efe", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36152, - "estimated_input_tokens": 9038, - "estimated_eval_tokens": 18332 + "serialized_chars": 36146, + "estimated_input_tokens": 9037, + "estimated_eval_tokens": 18330 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated Transformers bug reports with a few superficial title similarities; I only see one plausible duplicate pair around SigLIP2 behavior.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43493 is the best representative because it describes the underlying SigLIP2 implementation mismatch directly, while issue:43994 looks like a downstream symptom of the same model-level problem.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers bugs and regressions, with one plausible duplicate SigLIP2 pair. No PRs are present.", + "confidence": 0.41, + "canonical_issue_reason": "issue:45920 is the broadest, most representative open report here: a v5 regression in AutoTokenizer producing wrong token IDs across multiple model families, with clear user impact and ongoing relevance.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43493 is the broadest and most diagnostic report in the cluster; it is the closest thing to a canonical issue, but the overall cluster is not a true duplicate set.", + "best_issue_reason": "issue:45920 has the strongest cross-cutting scope and is the best single issue to anchor this otherwise heterogeneous cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44737", "right": "issue:45588", "accept": false, - "reason": "Different code paths: XLNet CPU device placement vs flash-attention auxiliary-state handling." + "reason": "Different subsystems and symptoms: XLNet positional encoding device placement vs flash-attention sink handling." }, { "left": "issue:43404", "right": "issue:45987", "accept": false, - "reason": "lm_head tying bug and StaticCache seq-length return type are unrelated." + "reason": "Unrelated bugs: Mistral3 weight tying vs StaticCache sequence-length return type." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "Both are trainer-related, but one is token accounting and the other is a custom loss API break." + "reason": "Both involve Trainer APIs, but one is a stale counter and the other is a custom loss signature change; not the same bug." }, { - "left": "issue:43646", - "right": "issue:43749", + "left": "issue:44062", + "right": "issue:45081", "accept": false, - "reason": "Custom model initialization breakage and FSDP CPU RAM-efficient loading are distinct failures." + "reason": "Distinct tokenizer failures with different code paths: AddedToken kwargs conflict vs Mistral regex patching." }, { - "left": "issue:44062", - "right": "issue:45081", + "left": "issue:43646", + "right": "issue:43749", "accept": false, - "reason": "Tokenizer AddedToken kwarg conflict is not the same bug as the Mistral regex/backend_tokenizer crash." + "reason": "Different failures: config field loss under trust_remote_code vs FSDP CPU RAM efficient loading." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "Both involve rope params, but one is config reload persistence and the other is JSON-list validation/type handling." + "reason": "Both touch RoPE/config handling, but one is config reload persistence and the other is a validation TypeError on JSON-loaded lists." }, { "left": "issue:43493", "right": "issue:43994", "accept": true, - "reason": "Both point to the same SigLIP2 model-level discrepancy producing incorrect outputs, with 43994 likely a downstream manifestation of the implementation mismatch." + "reason": "Both point to the same SigLIP2 correctness problem: the HF implementation diverges from expected behavior, and the model/pipeline yields bad outputs for the same family/checkpoint." }, { "left": "issue:43906", "right": "issue:45381", "accept": false, - "reason": "A reproduction of another issue is not the same bug as the Qwen2.5-VL video position-id regression." + "reason": "Different model families and symptoms; one is an isolated reproduction of an older issue, the other is a Qwen2.5-VL video position-id bug." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "Both are memory/loading concerns, but the reported root causes and fixes are different." + "reason": "Both concern loading/performance, but one is FSDP RAM-efficient loading and the other is mmap/OOM on Strix Halo." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "Qwen3 MoE weight conversion and Smollm3 RoPE layer dropping are unrelated model-specific bugs." + "reason": "Different Qwen3-MoE vs Smollm3 model-specific bugs with no clear shared code path." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Different multimodal model-loading failures with no shared concrete code path." + "reason": "Separate multimodal model failures affecting different checkpoints and code paths." }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "init_empty_weights argument handling is unrelated to loading AutoImageProcessor from a URL." + "reason": "Different loading bugs: init_empty_weights / _is_hf_initialized vs loading AutoImageProcessor from URL." }, { "left": "issue:36010", "right": "issue:42175", "accept": false, - "reason": "GenerationMixin import failure and missing TensorFlow backend packaging are different issues." + "reason": "Unrelated issues: GenerationMixin import path vs missing TensorFlow backend in a torch install extra." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Generic missing-PyTorch import error is not the same as the Llama 3.2 vision model-loading regression." + "reason": "Generic missing-PyTorch ImportError vs a specific Llama 3.2 Vision model-loading regression." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Backend dependency packaging and missing Qwen2.5-VL import are separate problems." + "reason": "Different missing-dependency/import problems: backend packaging vs missing Qwen2.5-VL class export." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "Torchvision dependency enforcement is unrelated to the Mistral tokenizer regex crash." + "reason": "Different bugs: PIL backend now requiring torchvision vs Mistral regex patch crashing on tokenizer internals." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Incorrect visual encoder save keys and wrong tokenizer IDs are different subsystems and symptoms." + "reason": "Separate regressions: incorrect visual encoder keys on save_pretrained vs wrong token IDs on tokenization." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "GGUF loading breakage and a specific model-loading failure do not look like the same bug." + "reason": "Both are loading-related, but one is a gguf_file/from_pretrained breakage and the other is a model-specific load failure; not enough evidence of the same bug." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "A KeyError on one model and a Float8 storage lookup TypeError have different causes." + "reason": "Different model load failures: key error for MobileLLM-125M vs missing Float8 storage object." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "A version-pin problem and stale auto_mappings references are not duplicates." + "reason": "One is a dependency version bump issue; the other is stale auto_mappings CI breakage." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Different problems: PyTorch-not-found ImportError vs LayoutLMv3 box-content validation clarity." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both mention missing config fields, but the mechanisms differ: trust_remote_code object hydration vs strict-field dropping." + "reason": "Both involve config field handling, but one is missing fields from trust_remote_code and the other is a specific Qwen3_5MoeVisionConfig field dropped by strict config parsing." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "ImageProcessor label reduction and sequence-classification zero-loss behavior are unrelated." + "reason": "Different label/classification issues affecting different components and failure modes." }, { "left": "issue:43756", "right": "issue:45325", "accept": false, - "reason": "Smollm3 RoPE-layer omission and Qwen2.5-VL rope indexing are separate model bugs." + "reason": "Different Qwen/Smollm3 RoPE issues with distinct symptoms and code paths." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior is unrelated to AutoImageProcessor URL loading." + "reason": "Different domains entirely: Qwen3-MoE sliding_window behavior vs image processor URL loading." }, { "left": "issue:44625", "right": "issue:44877", "accept": false, - "reason": "num_labels propagation and granite_speech strict-config loading are different config bugs." - }, - { - "left": "issue:43720", - "right": "issue:44898", - "accept": false, - "reason": "Packed-weight unpacking during accelerate loading is not the same as Perceiver interpolation on non-default resolution." + "reason": "Separate config propagation bugs affecting different models and strictness logic." }, { - "left": "issue:28282", - "right": "issue:29127", + "left": "issue:43697", + "right": "issue:45216", "accept": false, - "reason": "PyTorch missing import and LayoutLMv3 box-validation messaging are unrelated." + "reason": "Different regressions: RTDetrV2 output changes in v5 vs Qwen3.5 save_pretrained checkpoint corruption." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. It correctly treats the cluster as mostly unrelated issues and only endorses the SigLIP2 pair as a plausible duplicate. The soft-edge verdicts are generally cautious and do not overstate evidence." + "feedback": "The summary is grounded in the packet: the cluster is indeed heterogeneous, issue:45920 is the broadest open report, and the only accepted soft pair (issue:43493 / issue:43994) is plausibly the same SigLIP2 bug family. The other pairwise rejections are phrased conservatively and do not overclaim evidence. No PR pairs are present." }, "evaluator_used": true, "retried": false, @@ -6356,13 +6381,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 20, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:24643", "issue:29127", @@ -6376,11 +6401,11 @@ "issue:43295", "issue:43452", "issue:43531", - "issue:43550", "issue:43577", "issue:43582", "issue:43653", "issue:43697", + "issue:43720", "issue:43742", "issue:43819", "issue:43906", @@ -6397,11 +6422,11 @@ "issue:44743", "issue:44811", "issue:44843", + "issue:44898", "issue:44936", "issue:44977", "issue:44991", "issue:45005", - "issue:45216", "issue:45290", "issue:45356", "issue:45372", @@ -6414,14 +6439,14 @@ "issue:46202" ], "soft_pairs": [ - "issue:43697|issue:45216", + "issue:43720|issue:44898", "issue:43582|issue:44977", "issue:44514|issue:45381", "issue:24643|issue:29127", "issue:41762|issue:44936", - "issue:44466|issue:45005", "issue:43452|issue:43577", "issue:43452|issue:44843", + "issue:44466|issue:45005", "issue:43906|issue:46202", "issue:45372|issue:46170", "issue:43819|issue:44811", @@ -6442,231 +6467,210 @@ "issue:44464|issue:44743", "issue:43653|issue:43927", "issue:43531|issue:43742", - "issue:41720|issue:43906", - "issue:43550|issue:43927", - "issue:44464|issue:44977", - "issue:43550|issue:44361" + "issue:41720|issue:43906" ], - "prepared_review_unit_hash": "5e8bf7eb657edee053ee0642383d93663b2c70541b40f50d4bbcf8beb0f93a3c", + "prepared_review_unit_hash": "44834958a2c1d5146b30be2d98f62cc6d4e720d3b0bc0260bc45f27e05923dec", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37505, - "estimated_input_tokens": 9377, - "estimated_eval_tokens": 19010 + "soft_pair_count": 29, + "serialized_chars": 36646, + "estimated_input_tokens": 9162, + "estimated_eval_tokens": 18580 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous: these items cover unrelated tokenizer, processor, backend, save/load, and model-execution bugs. None of the soft pairs looks like the same underlying issue, so I would not merge any of them as duplicates.", - "confidence": 0.95, - "canonical_issue_reason": "No single canonical issue is safe to choose: the cluster is a grab bag of unrelated bugs, so any representative would be arbitrary.", + "summary": "The cluster is highly heterogeneous: the soft pairs mostly share only broad themes like loading, tokenization, or generation, but not the same concrete bug path. I would not merge any of the suggested soft edges.", + "confidence": 0.35, + "canonical_issue_reason": "issue:43452 is the broadest concrete regression here: it breaks both AutoTokenizer and AutoModelForCausalLM loading via gguf_file, making it the best anchor if a canonical issue must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "No issue is a strong global canonical candidate; if forced, the closest is a broad save/load regression, but it still does not unify the cluster.", + "best_issue_reason": "issue:43452 is the most representative standalone loader regression in this set; it is broader and more central than the narrower model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43697", - "right": "issue:45216", + "left": "issue:43720", + "right": "issue:44898", "accept": false, - "reason": "Different failures: RTDetrV2 output drift vs Qwen3.5 checkpoint save/load regression." + "reason": "Different models and failures: BitNet packed-weight loading vs Perceiver image-size interpolation." }, { "left": "issue:43582", "right": "issue:44977", "accept": false, - "reason": "Apple Silicon allocator TypeError vs Qwen3.5 flash-attention generation bug." + "reason": "Unrelated code paths: Apple Silicon allocator warmup TypeError vs Qwen3.5 flash-attention generation." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Same model family, but different code paths: batched chat templating vs video position IDs." + "reason": "Both are Qwen2.5-VL, but one is batched chat templating with padding=False and the other is video vision_position_ids; different bugs." }, { "left": "issue:24643", "right": "issue:29127", "accept": false, - "reason": "DeepSpeed weight-dimension runtime error vs LayoutLMv3 error-message clarity issue." + "reason": "DeepSpeed weight-shape error is unrelated to LayoutLMv3 box-content validation and messaging." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "Gemma3 ZeRO-3 load crash vs trainer.evaluate() state bug; unrelated paths." - }, - { - "left": "issue:44466", - "right": "issue:45005", - "accept": false, - "reason": "Both mention tied weights, but one is serialization inconsistency and the other is a translation-model v5 regression." + "reason": "Gemma3 ZeRO-3 load crash is not the same as trainer.evaluate() failing after train()." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "gguf loading breakage vs Blip2 dtype staying float32; different bugs." + "reason": "Loader failure with gguf_file is unrelated to Blip2 dtype remaining float32." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Different loading bugs: gguf_file handling vs an unconditional model_info() call in a regex patch." + "reason": "Both concern model loading, but gguf_file breakage and offline model_info() calls in tokenizer patching are different bugs." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is device-dependent serialization of lm_head.weight and the other is a broader translation-model tied-weights issue." }, { "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "Isolated reproduction of one issue vs Gemma pipeline roundtrip failure on \u2581; no shared bug." + "reason": "A reproduction of another issue is not the same as the Gemma pipeline roundtrip failure on '\u2581'." }, { "left": "issue:45372", "right": "issue:46170", "accept": false, - "reason": "Gemma 4 processor import failure vs Windows ROCm torch.distributed dependency problem." + "reason": "Gemma 4 processor import failure from mistral_common is unrelated to Windows/ROCm torch.distributed import/type errors." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC forward/from_latents mismatch vs Whisper batch_decode skip_special_tokens handling." + "reason": "DAC latent/forward mismatch is unrelated to Whisper batch_decode ignoring skip_special_tokens." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Processor.tokenizer regression vs Kimi codec/regex warning regression." + "reason": "Different regressions: processor.tokenizer access/images passthrough vs Kimi-K2.5 codec handling and warning behavior." }, { "left": "issue:30064", "right": "issue:33666", "accept": false, - "reason": "Void segmentation map processing vs Qwen2-VL multi-GPU training; unrelated." + "reason": "Void segmentation-map processing and Qwen2-VL multi-GPU training are unrelated." }, { "left": "issue:44936", "right": "issue:45823", "accept": false, - "reason": "Trainer lifecycle bug vs Gemma4 PLE device mismatch under device_map=auto." + "reason": "Trainer lifecycle failure is not the same as Gemma4 device_map='auto' forward mismatch." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "BigBird special-token registration bug vs MLukeTokenizer AttributeError." + "reason": "Different tokenizer bugs affecting different models and code paths." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "MLukeTokenizer task failure vs AutoTokenizer loading a different tokenizer class." + "reason": "MLukeTokenizer task AttributeError is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "max_shard_size question vs kwargs not forwarded to cached_file; different concerns." + "reason": "save_pretrained shard-size behavior is unrelated to AutoProcessor kwargs not reaching cached_file." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Siglip zero probabilities vs processor.tokenizer regression; unrelated." + "reason": "Unexpected zero probabilities in SigLIP2 is unrelated to the processor/tokenizer regression." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Both hit apply_chat_template, but one is batching/padding and the other is tool-call/no-content handling." + "reason": "Both hit apply_chat_template, but one is batched padding=False and the other is assistant tool-call messages with no content." }, { "left": "issue:45542", "right": "issue:46178", "accept": false, - "reason": "Both are backend availability errors, but for different backends and different triggering conditions." + "reason": "Both are backend-installation related, but they fail in different backend-mapping scenarios and code paths." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "BigBird mask-token special registration vs mdeberta add_special_tokens not adding BOS/EOS." + "reason": "BigBird special-token registration and mdeberta BOS/EOS insertion are different tokenizer bugs." }, { "left": "issue:43577", "right": "issue:43697", "accept": false, - "reason": "Blip2 dtype loading issue vs RTDetrV2 output differences; unrelated." + "reason": "Blip2 dtype persistence and RTDetrV2 output differences are unrelated model-specific regressions." }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Tokenizer slow/fast inconsistency vs CLIP bus error on macOS." + "reason": "Slow/fast tokenizer inconsistency is unrelated to the MacOS CLIP bus error." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Offline mistral regex patch bug vs EMBEDDIA tokenizer load failure." + "reason": "Offline model_info() misuse in tokenizer patching is not the same as the est-roberta tokenizer loading failure." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Position-id warning regression vs quantized save/load breakage." + "reason": "Position-id warnings are unrelated to quantized-model save/load breakage." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Compiled chunked generation inconsistency vs Qwen3 recurrent-state reset bug." + "reason": "Chunked generation with compiled forward and recurrent-state reset with cache are different generation bugs." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "BigBird special-token bug vs DiaConfig custom token IDs lost on save/load." + "reason": "BigBird special-token registration is unrelated to DiaConfig losing custom token IDs on save/load." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Qwen3-MoE sliding_window issue vs MobileLLM loading key error." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to MobileLLM-125M key errors during load." }, { "left": "issue:41720", "right": "issue:43906", "accept": false, - "reason": "Qwen3 auto-device-map cudaErrorAssert vs a reproduction of a different issue." - }, - { - "left": "issue:43550", - "right": "issue:43927", - "accept": false, - "reason": "Bamba torch.compile SDPA bug vs DiaConfig token-ID persistence bug." - }, - { - "left": "issue:44464", - "right": "issue:44977", - "accept": false, - "reason": "Compiled chunked generation inconsistency vs Qwen3.5 flash-attention generation bug." - }, - { - "left": "issue:43550", - "right": "issue:44361", - "accept": false, - "reason": "Bamba compile/SDPA failure vs MLukeTokenizer AttributeError." + "reason": "Qwen3 device-mapping cudaErrorAssert is unrelated to an isolated reproduction of another issue." } ] }, @@ -6682,13 +6686,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 21, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28282", "issue:32090", @@ -6739,6 +6743,9 @@ "issue:46209" ], "soft_pairs": [ + "issue:43550|issue:43927", + "issue:44464|issue:44977", + "issue:43550|issue:44361", "issue:44977|issue:46209", "issue:43643|issue:44561", "issue:32090|issue:34567", @@ -6757,234 +6764,252 @@ "issue:44849|issue:46209", "issue:45823|issue:46050", "issue:28282|issue:32090", - "issue:43262|issue:43295", "issue:28282|issue:33357", + "issue:43262|issue:43295", "issue:43720|issue:43927", "issue:43452|issue:44451", "issue:43606|issue:43927", "issue:43994|issue:46192", "issue:33666|issue:36331", - "issue:43720|issue:44361", "issue:43756|issue:44913", + "issue:43720|issue:44361", "issue:43525|issue:43572", "issue:44849|issue:45357", "issue:43493|issue:43825", "issue:39692|issue:43493" ], - "prepared_review_unit_hash": "c8948d7f07ad962cb56c4d8493d52363f77032fe55f41771818c5934e8fa73cd", + "prepared_review_unit_hash": "3d68c933e896489d49e2ea134bca4a930ea286446b7da563a2f9ae55acfdf011", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36558, - "estimated_input_tokens": 9140, - "estimated_eval_tokens": 18536 + "soft_pair_count": 34, + "serialized_chars": 37451, + "estimated_input_tokens": 9363, + "estimated_eval_tokens": 18982 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No credible duplicate cluster here: the soft pairs are mostly superficial matches across unrelated bugs, regressions, config issues, and model-specific failures. All soft edges should be rejected.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue stands out because the items span many unrelated code paths (trainer, loading, generation, multimodal processors, configs, quantization, and device-mapping bugs).", + "summary": "This cluster is a grab bag of mostly unrelated transformer issues; the soft-similarity links look like false positives. The broadest and most canonical issue is the v5 regression breaking `trust_remote_code` models.", + "confidence": 0.71, + "canonical_issue_reason": "issue:44561 is the best representative because it describes a broad, version-level regression with a clear root cause and impact on multiple remote-code models.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue for deduplication: there is no coherent underlying bug/theme shared by the set, so picking one issue as representative would be misleading.", + "best_issue_reason": "issue:44561 has the clearest cross-cutting regression scope and is likely to be the most useful anchor for related reports.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "Different bugs: one is torch.compile/SDPA failure in Bamba, the other is config persistence losing token IDs." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Both are generation failures, but they involve different models and different failure modes (compiled forward vs flash-attention)." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Unrelated failures: torch.compile SDPA vs tokenizer task AttributeError." + }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Different failures: Qwen3.5 flash-attention generation bug vs FP8 inference on B200." + "reason": "Different root causes and subsystems: flash-attention generation vs FP8 inference on B200." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both touch trust_remote_code, but one is missing fields in AutoConfig and the other is a v5 removal breaking remote models." + "reason": "Both affect remote-code workflows, but one is missing fields in AutoConfig output while the other is a v5 API removal breakage." }, { "left": "issue:32090", "right": "issue:34567", "accept": false, - "reason": "Trainer broadcast NoneType error and TrainerState token counter not updating are unrelated trainer issues." + "reason": "Trainer broadcast type error and TrainerState token counter bug are unrelated." }, { "left": "issue:43577", "right": "issue:45588", "accept": false, - "reason": "Blip2 dtype-loading problem is unrelated to flash_attention.py crashing on s_aux=None." + "reason": "Blip2 dtype loading and flash_attention.py s_aux=None crash are different code paths." }, { "left": "issue:28282", "right": "issue:35141", "accept": false, - "reason": "ImportError for missing PyTorch vs token embedding resizing/post_init reinit are different bugs." + "reason": "ImportError for missing PyTorch and embedding reinitialization during post_init are not the same bug." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Legacy field in Beit semantic segmentation is unrelated to Bamba torch.compile SDPA failure." + "reason": "Legacy-field loading bug in BEiT is unrelated to Bamba torch.compile/SDPA failure." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Unsupported _gpu_broadcast_one types and Qwen2-VL multi-GPU training are different problems." + "reason": "Trainer broadcast error and multi-GPU training request are not the same issue." }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Different model-loading failures: Beit legacy field vs BitNet packed-weight unpacking." + "reason": "Different model-loading bugs: legacy preset field vs packed-weights unpacking during accelerate loading." }, { "left": "issue:45491", "right": "issue:46002", "accept": false, - "reason": "Both mention NaNs, but one is Gemma3 sliding-window attention and the other is classification-head init/checkpoint behavior." + "reason": "Both mention NaNs, but one is Gemma3 sliding-window attention and the other is classification head init/checkpoint behavior." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "Serialization inconsistency for tied lm_head weights is not the same as sinq quantized model save/load breakage." + "reason": "Both involve save/load behavior, but one is tied-weight serialization and the other is quantized sinq persistence." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Beit legacy preset loading and OmDet-Turbo processor size mismatch are unrelated." + "reason": "BEiT legacy field loading and OmDet-Turbo processor size mismatch are unrelated." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Qwen3 auto device-map CUDA assert is unrelated to chunked generation inconsistency under compiled forward." + "reason": "Qwen3 device-mapping assert and chunked generation with compiled forward are different concrete failures." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training issue vs GenerationMixin import error are not the same bug." + "reason": "Multi-GPU training and GenerationMixin import breakage are unrelated." }, { "left": "issue:43742", "right": "issue:44821", "accept": false, - "reason": "MobileLLM key error and AutoImageProcessor-from-URL failure affect different loading paths." + "reason": "KeyError on MobileLLM loading vs AutoImageProcessor URL loading are different components and bugs." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Both are MoE-related, but they concern different models and different correctness bugs." + "reason": "Both concern model correctness, but they are different architectures and different implementation mismatches." }, { "left": "issue:44849", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 hidden_states output bug is unrelated to FP8 inference on B200." + "reason": "Qwen3.5 hidden-states bug and FP8 inference on B200 are unrelated." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Gemma4 device-map mismatch and quantized save/load corruption are different issues." + "reason": "Device-mapped Gemma4 PLE mismatch and quantized save/load breakage are different problems." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Missing PyTorch ImportError and trainer _gpu_broadcast_one TypeError are unrelated." + "reason": "Missing PyTorch import error and Trainer GPU broadcast type error are unrelated." }, { - "left": "issue:43262", - "right": "issue:43295", + "left": "issue:28282", + "right": "issue:33357", "accept": false, - "reason": "Audio sampling-rate defaulting and processor.tokenizer/image-handling regression are different processor bugs." + "reason": "ImportError for missing PyTorch and Mac bus error with CLIP are not the same underlying bug." }, { - "left": "issue:28282", - "right": "issue:33357", + "left": "issue:43262", + "right": "issue:43295", "accept": false, - "reason": "PyTorch ImportError and MacOS CLIP bus error have no shared underlying cause." + "reason": "Both are regressions, but one is audio chat-template sampling-rate handling and the other is processor.tokenizer/image passthrough breakage." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "BitNet packed-weight loading and DiaConfig custom token ID loss are unrelated save/load bugs." + "reason": "Packed-weight loading failure and lost custom token IDs are separate loading bugs." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "gguf_file loading breakage and ScandiBERT load failure are distinct model-loading problems." + "reason": "gguf_file tokenizer/model loading and ScandiBERT load failure do not match on root cause." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch for bark-small is unrelated to DiaConfig save/load field loss." + "reason": "CPU offload device mismatch is unrelated to config save/load token ID loss." }, { "left": "issue:43994", "right": "issue:46192", "accept": false, - "reason": "SigLIP2 nonsensical outputs and return_tensors being ignored are different API behaviors." + "reason": "SigLIP2 inference quality bug and return_tensors being ignored are different issues." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Qwen2-VL multi-GPU training and custom compute_loss signature mismatch are unrelated Trainer issues." + "reason": "Multi-GPU training request and unexpected Trainer compute_loss kwarg are distinct Trainer issues." }, { - "left": "issue:43720", - "right": "issue:44361", + "left": "issue:43756", + "right": "issue:44913", "accept": false, - "reason": "BitNet accelerate-loading bug and MLukeTokenizer AttributeError are unrelated." + "reason": "Smollm3 RoPE layer mismatch and GPTNeoX rotary_pct reload regression are similar in theme but not the same bug." }, { - "left": "issue:43756", - "right": "issue:44913", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "Smollm3 dropping RoPE layers and GPTNeoX rotary_pct reset on reload are separate config/model bugs." + "reason": "Packed-weight unpacking failure and MLukeTokenizer task AttributeError are unrelated." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Both are missing padding config fields, but they affect different model configs and different attribute names." + "reason": "Both involve missing pad-token config fields after v5, but they affect different models/config attributes and are not one concrete bug." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Qwen3.5 hidden_states bug is not the same as incorrect visual encoder keys being saved." + "reason": "Hidden_states bug and incorrect visual-encoder key serialization are different Qwen3.5 issues." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 JAX/HF discrepancy and pipeline translation-task messaging are unrelated." + "reason": "SigLIP2 implementation discrepancy and pipeline translation-task messaging are unrelated." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "SigLIP2 documentation example errors are not the same as the broader SigLIP2 implementation discrepancy." + "reason": "Doc example errors for SigLIP2 and JAX-vs-HF implementation discrepancy are separate issues." } ] }, @@ -7000,13 +7025,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 22, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:33357", "issue:33666", @@ -7060,8 +7085,8 @@ "soft_pairs": [ "issue:44964|issue:45405", "issue:44479|issue:44625", - "issue:43577|issue:44554", "issue:43688|issue:43749", + "issue:43577|issue:44554", "issue:44568|issue:45812", "issue:43540|issue:44977", "issue:45200|issue:45229", @@ -7071,232 +7096,232 @@ "issue:44493|issue:46285", "issue:43756|issue:45083", "issue:43493|issue:43844", - "issue:43646|issue:44466", "issue:43334|issue:43572", - "issue:33357|issue:33666", + "issue:43646|issue:44466", "issue:45229|issue:45823", + "issue:33357|issue:33666", "issue:43334|issue:45759", "issue:43927|issue:44361", "issue:43824|issue:44589", "issue:43504|issue:43606", - "issue:44589|issue:45520", "issue:43278|issue:43844", + "issue:44589|issue:45520", "issue:44625|issue:45759", "issue:43257|issue:43299", "issue:43688|issue:45701", - "issue:43575|issue:45405", "issue:43572|issue:44291", + "issue:43575|issue:45405", "issue:43606|issue:44112", "issue:43504|issue:43653" ], - "prepared_review_unit_hash": "1937d8e6b8510164db301fc0f01411d8b97adb4ec75408e5085bea67e7a43be9", + "prepared_review_unit_hash": "1f891236216bc87b152740ad13992372c8691da729a147150612226dc7cc0798", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36899, - "estimated_input_tokens": 9225, - "estimated_eval_tokens": 18706 + "serialized_chars": 36896, + "estimated_input_tokens": 9224, + "estimated_eval_tokens": 18704 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is a loose mix of mostly unrelated model-loading, tokenizer, config, and training-regression reports; I did not find any soft pair that looks like the same underlying bug/change.", - "confidence": 0.78, - "canonical_issue_reason": "issue:45812 is the strongest representative because it describes a broad, silent tokenizer regression affecting multiple Granite models, which is a clearer and more user-facing failure mode than the other similarly themed reports.", + "summary": "The pairs are mostly adjacent regressions in similar subsystems, but none look like true duplicates: they differ in the concrete failure mode or code path. I\u2019d anchor the set on the Granite tokenizer regression issue as the cleanest concrete bug report.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45812 is the cleanest representative concrete regression here: it is user-facing, reproducible, and narrowly describes a wrong-tokenization failure with a clear scope.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45812 is the best issue candidate overall: it is concrete, high-impact, and framed as a cross-model regression with a well-defined symptom.", + "best_issue_reason": "issue:45812 is the strongest anchor because it is specific enough to triage and general enough to represent a real class of tokenizer regressions without being overly broad.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "One is a model-load failure for Phi-4 multimodal; the other is a dependency/version pin problem. Different root causes." + "reason": "Model-loading regression vs a dependency version bump; different underlying problems." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Video-input regression across Qwen VL variants vs num_labels propagation in Qwen3.5; different code paths and symptoms." + "reason": "Video-input regression vs num_labels propagation; not the same bug." }, { - "left": "issue:43577", - "right": "issue:44554", + "left": "issue:43688", + "right": "issue:43749", "accept": false, - "reason": "BLIP2 dtype propagation bug vs an attention correctness issue on MPS; not the same bug." + "reason": "Aux-loss normalization vs FSDP CPU-efficient loading; unrelated failure modes." }, { - "left": "issue:43688", - "right": "issue:43749", + "left": "issue:43577", + "right": "issue:44554", "accept": false, - "reason": "Aux-loss normalization bug vs FSDP CPU RAM efficient loading breakage; unrelated failure modes." + "reason": "dtype propagation issue vs MPS attention correctness bug." }, { "left": "issue:44568", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they target different models and behaviors (special-token handling vs Granite token IDs)." + "reason": "Tokenizer special-token handling for one model vs broad Granite token-ID regression; not the same code-path." }, { "left": "issue:43540", "right": "issue:44977", "accept": false, - "reason": "Qwen3OmniMoe video-processing ValueError vs Qwen3.5 flash-attention generation bug; different subsystems." + "reason": "Video processing ValueError vs flash-attention generation regression; different symptoms and fixes." }, { "left": "issue:45200", "right": "issue:45229", "accept": false, - "reason": "Text-only fine-tuning input defaults vs multi-GPU inference OOM; not the same underlying defect." + "reason": "Default token-type IDs vs multi-GPU inference OOM; not duplicate defects." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "A specific model failing to load vs unconditional hub metadata access breaking offline mode; different causes." + "reason": "Model loading failure vs offline model_info call in tokenizer patch; distinct issues." }, { "left": "issue:43688", "right": "issue:45704", "accept": false, - "reason": "Auxiliary-loss normalization bug vs T5/apex RMSNorm memory leak; unrelated." + "reason": "Aux-loss normalization bug vs T5 RMSNorm memory leak; unrelated." }, { "left": "issue:39692", "right": "issue:43295", "accept": false, - "reason": "Documentation example mistakes vs custom-model regression involving processor.tokenizer/image passing; not the same issue." + "reason": "Example/documentation errors vs processor.tokenizer API regression; related subsystem, different bug." }, { "left": "issue:44493", "right": "issue:46285", "accept": false, - "reason": "Unexpected position-id keys during loading vs CLIP text-model structure inconsistency; different bugs." + "reason": "Unexpected position-id keys vs CLIP class structure change; different concrete problems." }, { "left": "issue:43756", "right": "issue:45083", "accept": false, - "reason": "Smollm3 RoPE-layer discrepancy vs qwen3_omni_moe feature-length helper behavior; unrelated." + "reason": "RoPE-layer dropping bug vs feature-extract length helper behavior." }, { "left": "issue:43493", "right": "issue:43844", "accept": false, - "reason": "SigLIP2 implementation mismatch vs DeepSpeed/ZeRO gradient inflation; different failure modes." - }, - { - "left": "issue:43646", - "right": "issue:44466", - "accept": false, - "reason": "Custom model initialization breakage vs lm_head serialization inconsistency; not the same bug." + "reason": "SigLIP2 implementation mismatch vs ZeRO-3 gradient growth; unrelated." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Different config-attribute regressions in different model families (pad_token_id vs pad_token_idx)." + "reason": "Missing pad_token_id in Qwen3-VL vs missing pad_token_idx in StableLmConfig; same theme, different bug." }, { - "left": "issue:33357", - "right": "issue:33666", + "left": "issue:43646", + "right": "issue:44466", "accept": false, - "reason": "Mac CLIP bus error vs Qwen2-VL multi-GPU training request; unrelated." + "reason": "Custom model init breakage vs inconsistent lm_head serialization; not the same issue." }, { "left": "issue:45229", "right": "issue:45823", "accept": false, - "reason": "Inference OOM vs device mismatch in forward with device_map=\"auto\"; different concrete problems." + "reason": "CUDA OOM vs device mismatch during forward; different failure modes." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training; unrelated." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Missing pad_token_id on Qwen3-VL config vs failure to unwrap text_config for composite Qwen multimodal configs; different issues." + "reason": "Missing pad_token_id vs composite-config unwrapping in from_config; different code paths." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "DiaConfig save/load token ID loss vs MLukeTokenizer AttributeError; unrelated." + "reason": "Config save/load token ID loss vs tokenizer AttributeError on tasks; different bugs." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "ImportError for a model class vs missing Float8 storage type; different layers of the stack." + "reason": "ImportError for a model class vs float8 storage lookup failure; unrelated." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "BEiT legacy-field load failure vs bark-small CPU offload device mismatch; no shared root cause." + "reason": "Legacy-field load failure vs CPU offload device mismatch; different model paths." }, { - "left": "issue:44589", - "right": "issue:45520", + "left": "issue:43278", + "right": "issue:43844", "accept": false, - "reason": "Float8 storage deserialization issue vs flash_attn import key error on Python 3.13; unrelated." + "reason": "Embedding dtype drift vs abnormal gradient growth under ZeRO-3; not duplicates." }, { - "left": "issue:43278", - "right": "issue:43844", + "left": "issue:44589", + "right": "issue:45520", "accept": false, - "reason": "Embedding dtype drift between train/eval vs abnormal gradient growth under ZeRO-3; different problems." + "reason": "Float8 storage deserialization error vs flash_attn KeyError on Python 3.13; different causes." }, { "left": "issue:44625", "right": "issue:45759", "accept": false, - "reason": "num_labels propagation bug vs text_config unwrapping in from_config; different config handling paths." + "reason": "num_labels propagation vs text_config unwrapping; both config-related but not the same bug." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Qwen3 MoE weight-conversion issue under accelerate/deepspeed vs Qwen3VL MoE loading regression; related family, but not the same concrete bug." + "reason": "Accelerate/DeepSpeed conversion issue vs Qwen3VL MoE loading regression; different code paths." }, { "left": "issue:43688", "right": "issue:45701", "accept": false, - "reason": "Aux-loss normalization bug vs broad tokenizer-version behavior change; not the same underlying issue." + "reason": "Aux-loss normalization vs tokenization changes; unrelated." }, { - "left": "issue:43575", - "right": "issue:45405", + "left": "issue:43572", + "right": "issue:44291", "accept": false, - "reason": "Tensor-parallel OOM on Qwen2-57B vs an unreleased PEFT version bump; unrelated." + "reason": "Missing config field vs init_empty_weights argument mismatch; different regressions." }, { - "left": "issue:43572", - "right": "issue:44291", + "left": "issue:43575", + "right": "issue:45405", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig vs init_empty_weights TypeError from an unexpected argument; different regressions." + "reason": "Tensor-parallel OOM vs unreleased dependency version bump; not duplicate." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "bark-small CPU offload device mismatch vs GraniteSpeech CI device-override test staleness; different scopes." + "reason": "Runtime device mismatch vs stale CI test; one is a product bug, the other a test issue." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "BEiT pretrained-load failure vs BigBirdTokenizer special-token registration/empty decode bug; unrelated." + "reason": "Beit legacy-field load bug vs BigBirdTokenizer special-token decode bug; different models and failures." } ] }, @@ -7312,13 +7337,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 23, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:15354", "issue:28218", @@ -7380,8 +7405,8 @@ "issue:28218|issue:34567", "issue:43688|issue:43873", "issue:29159|issue:32090", - "issue:44554|issue:45446", "issue:44521|issue:45290", + "issue:44554|issue:45446", "issue:15354|issue:38175", "issue:45084|issue:45520", "issue:43540|issue:45200", @@ -7398,217 +7423,210 @@ "issue:43295|issue:43450", "issue:43295|issue:44610", "issue:39692|issue:46192", - "issue:44554|issue:45356", - "issue:43295|issue:43873" + "issue:44554|issue:45356" ], - "prepared_review_unit_hash": "b6ea5d45fc4ba95298d82d0de4846c7c6067b52803848c7fb8aebe0f77f01ac2", + "prepared_review_unit_hash": "53db50bab84aeef608aa72e5aba0b3f01150d64547d79eca33ca8120fb6bdb7c", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36786, - "estimated_input_tokens": 9197, - "estimated_eval_tokens": 18650 + "soft_pair_count": 29, + "serialized_chars": 36484, + "estimated_input_tokens": 9121, + "estimated_eval_tokens": 18498 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a collection of unrelated transformer bugs across tokenization, serving, multimodal, training, and quantization; none of the soft pairs look like true duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43873 is the strongest representative issue: open, well-discussed, and high-signal with inbound references around a concrete quantization/offloading failure.", + "summary": "These items are a loose thematic cluster, not a true duplicate set. Most soft pairs are only related at a broad subsystem level (tokenizers, processors, generation, quantization, serving) and do not describe the same concrete bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out: the cluster spans many unrelated failures. If one must be chosen as the closest representative, issue:43295 is the broadest regression-style report in core multimodal processor/tokenizer behavior.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the clearest standalone issue to anchor the cluster because it is active, specific, and broadly impactful.", + "best_issue_reason": "Issue:43295 is the best representative among an otherwise heterogeneous set because it describes a user-facing regression in core processor/tokenizer handling and is broader than the highly specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "Different bugs: janus image-generation test failure vs an unreleased PEFT version requirement." + "reason": "Janus image-generation test failure and a PEFT version pin issue are unrelated bugs." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Both are serve crashes, but one is continuous-batching multimodal input handling and the other is KV-cache continuation indexing." + "reason": "Both are serve-related crashes, but one is continuous batching multimodal input handling and the other is KV-cache continuation indexing in /v1/responses; different code paths." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Chunked generation inconsistencies are unrelated to Whisper failing on empty transcription after token alignment." + "reason": "Chunked generation with compiled forward and Whisper empty-transcription generation after token alignment are different generation bugs." }, { "left": "issue:45588", "right": "issue:46050", "accept": false, - "reason": "Flash-attention aux-argument AttributeError vs broken quantized save/load are different code paths and symptoms." + "reason": "Flash-attention sinkless-model None handling and sinq save/load quantization breakage are unrelated." }, { "left": "issue:43994", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 nonsensical outputs and unexpected position-id key errors are distinct failures." + "reason": "SigLIP2 bad outputs and position-id key errors are not the same underlying issue." }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "Flash Attention 2 test failures do not match MLflow not being reported after a failed training job." + "reason": "Flash Attention 2 test failures and MLFlow job state reporting are unrelated." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "ZeRO-3 gradient growth is unrelated to loading Phi-4 multimodal models." + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading are different training/loading problems." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Tokenizer added-space behavior is unrelated to TrainerState token-count updates." + "reason": "Tokenizer spacing after added tokens and TrainerState token counter updates are unrelated." }, { "left": "issue:43688", "right": "issue:43873", "accept": false, - "reason": "Aux-loss normalization and quantization/offloading are separate training bugs." + "reason": "Aux-loss normalization in MoE models and offloading with quantization are different bugs." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Slow/fast tokenizer inconsistency is not the same as NoneType broadcast errors in Trainer." + "reason": "Slow/fast tokenizer inconsistency and NoneType GPU broadcast errors are unrelated." }, { - "left": "issue:44554", - "right": "issue:45446", + "left": "issue:44521", + "right": "issue:45290", "accept": false, - "reason": "MPS attention correctness on head-dim mismatch is unrelated to a PyTorch version check bug." + "reason": "Both involve apply_chat_template, but one is zero assistant masks for multimodal inputs and the other is a crash on tool-call messages with no content." }, { - "left": "issue:44521", - "right": "issue:45290", + "left": "issue:44554", + "right": "issue:45446", "accept": false, - "reason": "Both touch chat templating, but all-zero assistant masks and crashes on tool-call messages are different failures." + "reason": "MPS attention correctness and a PyTorch version check in flex_attention are unrelated." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "JIT scripting GeneratorExp is unrelated to SigLIP2 zero-probability outputs." + "reason": "TorchScript export failure for ViT and SigLIP2 zero probabilities are unrelated." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Non-template-node compilation and flash_attn import KeyError are different problems." + "reason": "Template-node compilation TypeError and flash_attn import KeyError on Python 3.13 are unrelated." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Video-input processing in Qwen3OmniMoe is unrelated to Gemma4 token-type defaults." + "reason": "Qwen3OmniMoe video processing failure and Gemma4 text-only mm_token_type_ids handling are different multimodal issues." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Both mention quantization, but offloading failures and broken quantized persistence are not the same bug." + "reason": "Quantization offloading behavior and sinq save/load corruption are different quantization problems." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Text-only fine-tuning defaults and device-map forward mismatches are different Gemma4 issues." + "reason": "Gemma4 text-only token-type defaults and device-map forward device mismatch are separate bugs." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "BF16-to-FP32 eval drift is unrelated to resume_from_checkpoint key mismatches." + "reason": "BF16-to-FP32 embedding dtype drift and resume-from-checkpoint key mismatch are unrelated." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "Bfloat16 inference dtype mismatches and Wav2Vec2 save/tokenization failures are unrelated." + "reason": "bfloat16 dtype mismatches and Wav2Vec2 save/tokenization failures are unrelated." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Generation-kwargs update logic and output_hidden_states bugs are separate code paths." + "reason": "Generation kwargs after sync_gpus and Qwen3.5 hidden-states output are different generation bugs." }, { "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "LayoutLMv2 tokenizer crashes on NER/padding are unrelated to return_tensors being ignored with text_kwargs." + "reason": "LayoutLMv2 NER/padding crashes and return_tensors being ignored with text_kwargs are different tokenizer API bugs." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and RoPE validation TypeError are unrelated." + "reason": "Gemma4Processor _tokenizer attribute crash and rope-parameter list handling are unrelated." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer spacing behavior is unrelated to a MacOS bus error in a CLIP model." + "reason": "Tokenizer spacing behavior and a MacOS bus error on CLIP are unrelated." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device placement and return_tensors handling are different issues." + "reason": "Gemma4 device placement and return_tensors handling are unrelated." }, { "left": "issue:43493", "right": "issue:43701", "accept": false, - "reason": "SigLIP2 HF-vs-JAX discrepancy does not match checkpoint resume key mismatches." + "reason": "SigLIP2 implementation discrepancy and checkpoint resume key mismatch are unrelated." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Processor/tokenizer regression and batched video shape issues are different multimodal bugs." + "reason": "Processor.tokenizer regression and batched video shape errors are different multimodal processor bugs." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "Custom processor/tokenizer regression is unrelated to OmDet-Turbo image-size mismatch." + "reason": "Processor/tokenizer regression and OmDet-Turbo image-size mismatch are not the same bug." }, { "left": "issue:39692", "right": "issue:46192", "accept": false, - "reason": "SigLIP2 documentation/example errors are not the same as silent return_tensors ignoring." + "reason": "SigLIP2 docs/model mismatch and return_tensors handling are unrelated." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "Attention correctness on MPS is unrelated to Kimi-K2.5 tokenizer codec handling." - }, - { - "left": "issue:43295", - "right": "issue:43873", - "accept": false, - "reason": "Processor/tokenizer regression and quantization offloading are unrelated bugs." + "reason": "MPS attention correctness and Kimi-K2.5 tokenizer codec regression are unrelated." } ] }, @@ -7624,13 +7642,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 24, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:37428", "issue:38175", @@ -7653,6 +7671,7 @@ "issue:43824", "issue:43844", "issue:43856", + "issue:43873", "issue:43950", "issue:43957", "issue:43975", @@ -7662,8 +7681,6 @@ "issue:44514", "issue:44521", "issue:44561", - "issue:44610", - "issue:44857", "issue:44898", "issue:44987", "issue:45127", @@ -7683,6 +7700,7 @@ ], "soft_pairs": [ "issue:45381|issue:46209", + "issue:43295|issue:43873", "issue:43572|issue:44987", "issue:43950|issue:46285", "issue:43572|issue:43957", @@ -7701,251 +7719,231 @@ "issue:38175|issue:43493", "issue:43653|issue:44112", "issue:44442|issue:45198", - "issue:43295|issue:43493", "issue:42915|issue:43824", - "issue:43975|issue:45520", + "issue:43295|issue:43493", "issue:38175|issue:45412", "issue:43262|issue:43450", - "issue:45759|issue:46285", "issue:44514|issue:44521", + "issue:45759|issue:46285", "issue:37428|issue:43824", "issue:45520|issue:46170", "issue:41720|issue:46209", - "issue:44561|issue:46234", - "issue:44610|issue:44857" + "issue:44561|issue:46234" ], - "prepared_review_unit_hash": "636f6f2b9671ed1ae21eb3cea0926c00525011cbae873392ecef5f7d408145fb", + "prepared_review_unit_hash": "cff56ff462b9df5133188fc6dd6ed52d3350c4acc7783af1083faadbad1a7f0f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37332, - "estimated_input_tokens": 9333, - "estimated_eval_tokens": 18922 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36452, + "estimated_input_tokens": 9113, + "estimated_eval_tokens": 18482 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous, with one clear duplicate-like pair in the SigLIP2 reports. The rest of the soft edges look like broad lexical similarity rather than the same underlying bug/change.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43493 is the strongest canonical issue because it describes the underlying SigLIP2 implementation discrepancy at the model level; issue:38175 reads like a symptom report of that same problem (zero probabilities on the SigLIP2 model).", + "summary": "Mostly unrelated Transformers regression reports. The only likely duplicate-like pair is the Qwen3-VL / Llama4 missing `pad_token_id` config failure; the rest are separate bugs across different models/subsystems. No PRs are present.", + "confidence": 0.72, + "canonical_issue_reason": "issue:43334 is the best anchor for the only plausible duplicate pair: it matches 43525 in the same `pad_token_id`-missing model-load failure pattern on new config classes.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43493 is the best overall issue to anchor the only accepted duplicate pair, and it is more root-cause-oriented than the symptom report in issue:38175.", + "best_issue_reason": "issue:43334 is the clearest representative of the one near-duplicate theme and is described as a concrete load failure with a specific exception path.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Both are transformer model bugs, but one is Qwen2.5-VL position IDs and the other is FP8 inference on B200; different code paths and symptoms." + "reason": "Different failures: Qwen2.5-VL vision position ids vs FP8 inference OOM on B200." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression vs quantization offloading behavior; unrelated code paths." }, { "left": "issue:43572", "right": "issue:44987", "accept": false, - "reason": "Both involve loading/regression after v5, but one is missing pad_token_idx in StableLmConfig and the other is a load failure for a specific model; not the same bug." + "reason": "StableLm `pad_token_idx` issue is unrelated to a separate model-loading failure for `physical-intelligence/fast`." }, { "left": "issue:43950", "right": "issue:46285", "accept": false, - "reason": "Non-persistent buffer corruption in from_pretrained is unrelated to the CLIP text model structural mismatch." + "reason": "Non-persistent buffer corruption on load is unrelated to CLIP text model structure inconsistency." }, { "left": "issue:43572", "right": "issue:43957", "accept": false, - "reason": "Both concern loading regressions, but one is a missing config attribute and the other is meta-device loading breakage; too broad to be the same issue." + "reason": "Different regressions: missing pad token field vs meta-device loading breakage." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Perceiver resolution handling and RT-DETR memory release are unrelated." + "reason": "Perceiver interpolation failure and RT-DETR memory release bug are unrelated." }, { "left": "issue:43550", "right": "issue:43844", "accept": false, - "reason": "torch.compile/SDPA on Bamba and ZeRO-3 gradient growth are different training/runtime problems." + "reason": "torch.compile SDPA failure for Bamba and DeepSpeed gradient growth are distinct issues." }, { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Serialization of tied lm_head weights is not the same as LoRA merging causing model collapse." + "reason": "Serialization of tied weights vs LoRA merge collapse; similar theme but not the same bug." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Both are multimodal-model issues, but one is mm_token_type_ids defaulting wrong and the other is rope/position-id scaling; distinct bugs." + "reason": "Different multimodal bugs: missing default mm token ids vs wrong temporal rope scaling." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Video processor batch shape and tensor-parallel OOM are unrelated." + "reason": "Video processor batch-shape bug and tensor-parallel OOM are unrelated." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "flash_attn import-utils KeyError on Python 3.13 is unrelated to tensorflow_text backend mapping." + "reason": "Both are import/backend errors, but for different optional dependencies and different failure modes." }, { "left": "issue:43334", "right": "issue:43525", - "accept": false, - "reason": "Both mention missing pad_token_id, but they affect different configs/models and do not clearly share the same concrete failure path." + "accept": true, + "reason": "Both report the same `AttributeError`/missing `pad_token_id` regression on new config classes during model loading." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Missing flash support helper import is unrelated to meta-device loading regression." + "reason": "Flash-attention import error and meta-device loading breakage are separate regressions." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "Local-path import warning with trailing slash is unrelated to GPT-oss OOM." + "reason": "Trailing-slash import warning is unrelated to GPT-oss OOM." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Missing Qwen2_5_VL symbol import and DeepSeek detokenization are separate issues." + "reason": "ImportError for `Qwen2_5_VLForConditionalGeneration` is unrelated to detokenization behavior in deepseek-coder." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype drift in evaluate and Qwen3 MoE memory efficiency are unrelated." + "reason": "Embedding dtype mismatch and Qwen3 MoE memory inefficiency are different problems." }, { "left": "issue:43526", "right": "issue:46192", "accept": false, - "reason": "BEiT reduce_labels truncation and text_kwargs/return_tensors handling are different processor bugs." + "reason": "Beit `reduce_labels` bug is unrelated to `return_tensors` being ignored with `text_kwargs`." }, { "left": "issue:38175", "right": "issue:43493", - "accept": true, - "reason": "Both report SigLIP2 output/behavior problems, and 43493 frames the likely underlying HF-vs-JAX implementation discrepancy that explains the zero-probability symptom." + "accept": false, + "reason": "Siglip2 zero-probabilities issue is not the same as the JAX/HF implementation discrepancy." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "BigBirdTokenizer special-token registration and GraniteSpeech CI device override are unrelated." + "reason": "Tokenizer special-token registration and GraniteSpeech CI test failure are unrelated." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "FastSpeech2ConformerTokenizer loading and Wav2Vec2 save_pretrained/tokenization failures are different tokenizer/model issues." - }, - { - "left": "issue:43295", - "right": "issue:43493", - "accept": false, - "reason": "Processor.tokenizer regression and SigLIP2 implementation discrepancy are unrelated." + "reason": "AutoTokenizer load failure and Wav2Vec2 save/tokenization failure affect different tokenization paths." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe and Qwen2_5_VL import error are unrelated." + "reason": "Qwen3Moe FP8 failure and Qwen2.5-VL import error are unrelated." }, { - "left": "issue:43975", - "right": "issue:45520", + "left": "issue:43295", + "right": "issue:43493", "accept": false, - "reason": "DeepSeek detokenization regression and flash_attn import error are different failure modes." + "reason": "Processor/tokenizer regression and Siglip2 implementation discrepancy are unrelated." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 zero probabilities and RT-DETR memory leak are unrelated." + "reason": "Siglip2 output issue is unrelated to RT-DETR memory cleanup." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate defaults and video processor batched-shape bugs are different modality-specific issues." - }, - { - "left": "issue:45759", - "right": "issue:46285", - "accept": false, - "reason": "Composite Qwen text_config unwrapping and CLIP text-model structure inconsistency are not the same bug." + "reason": "Audio chat-template sampling-rate default and video batch-shape bug are unrelated." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both involve Qwen2.5-VL chat_template, but one is a batching/padding crash and the other is all-zero assistant masks; related area, not the same concrete bug." - }, - { - "left": "issue:44561", - "right": "issue:46234", - "accept": false, - "reason": "Missing is_torch_fx_available and missing is_falcon_mamba_ssm_available are different import_utils symbol regressions." + "reason": "Same model family, but one is a batching crash and the other is wrong assistant masks; not the same concrete bug." }, { - "left": "issue:44610", - "right": "issue:44857", + "left": "issue:45759", + "right": "issue:46285", "accept": false, - "reason": "OmDet-Turbo image size mismatch and LwDetrImageLoss AMP crash are unrelated." + "reason": "Composite Qwen config unwrapping and CLIP structure inconsistency are different model-loading changes." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "A flash-support import error and a Qwen2.5-VL class import error are different missing-symbol issues." + "reason": "Missing flash-attention helper import and missing Qwen2.5-VL export are unrelated import errors." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Python 3.13 flash_attn KeyError and Windows ROCm torch.distributed import/type errors are unrelated platform issues." + "reason": "Python 3.13 flash_attn backend lookup and Windows ROCm torch.distributed import failure are separate platform-specific issues." }, { "left": "issue:41720", "right": "issue:46209", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert and FP8 inference on B200 are different runtime failures." + "reason": "Qwen3 auto device map assert and FP8 inference OOM are different runtime failures." }, { "left": "issue:44561", "right": "issue:46234", "accept": false, - "reason": "These are both missing helper exports, but they concern different symbols and different affected model paths." - }, - { - "left": "issue:44610", - "right": "issue:44857", - "accept": false, - "reason": "Same as above: processor input-size mismatch versus AMP loss crash are unrelated." + "reason": "Missing `is_torch_fx_available` and missing `is_falcon_mamba_ssm_available` are different import-utils symbols and regressions." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded in the packet and appropriately conservative. The only accepted pair is well-supported by the titles as a symptom/root-cause SigLIP2 match, and the remaining soft-edge rejections stay within the evidence available from the issue summaries." + "feedback": "The summary is grounded in the packet and stays conservative. The one accepted duplicate pair (issue:43334 / issue:43525) is well supported by the shared missing-`pad_token_id` load failure pattern, and the rejected soft edges are generally reasoned from the issue titles without overclaiming. No PRs are present, so that part is also consistent." }, "evaluator_used": true, "retried": false, @@ -7958,13 +7956,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 25, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28218", "issue:29942", @@ -7993,8 +7991,8 @@ "issue:44351", "issue:44448", "issue:44493", + "issue:44610", "issue:44743", - "issue:44779", "issue:44792", "issue:44855", "issue:44857", @@ -8016,6 +8014,7 @@ "issue:46234" ], "soft_pairs": [ + "issue:44610|issue:44857", "issue:43819|issue:44112", "issue:43827|issue:43901", "issue:43819|issue:45083", @@ -8025,16 +8024,16 @@ "issue:43606|issue:45083", "issue:44792|issue:45563", "issue:37428|issue:43425", - "issue:44448|issue:44743", "issue:45127|issue:45245", + "issue:44448|issue:44743", "issue:45200|issue:45588", "issue:43335|issue:43653", "issue:43611|issue:43931", "issue:45290|issue:45381", "issue:44855|issue:45083", "issue:43550|issue:44112", - "issue:43819|issue:44493", "issue:44206|issue:45823", + "issue:43819|issue:44493", "issue:43335|issue:43504", "issue:29942|issue:32090", "issue:45588|issue:45859", @@ -8048,221 +8047,249 @@ "issue:39692|issue:44493", "issue:28218|issue:35141", "issue:43720|issue:44112", - "issue:43335|issue:44112", - "issue:44448|issue:44779", - "issue:43335|issue:45593" + "issue:43335|issue:44112" ], - "prepared_review_unit_hash": "05cb03414e2e127f1891acede16a4a369efa6adcaea7ba43b140b95faf22495e", + "prepared_review_unit_hash": "887832fce16cddb7411f7665c4bf498212b0a7e084f1fb30b352111a682354ca", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 38411, - "estimated_input_tokens": 9603, - "estimated_eval_tokens": 19462 + "soft_pair_count": 34, + "serialized_chars": 38127, + "estimated_input_tokens": 9532, + "estimated_eval_tokens": 19320 }, "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. The only clear duplicate pair is the repeated Qwen3.5 MoE `from_pretrained` error report; the rest are distinct bugs or doc/test regressions in different code paths.", - "confidence": 0.93, - "canonical_issue_reason": "issue:45310 is the earlier report of the exact same Qwen3.5 MoE `from_pretrained` failure and already has an inbound reference, so it is the better canonical issue.", + "summary": "This cluster is mostly heterogeneous; the only clear duplicate pair is the Qwen3.5 Moe `from_pretrained` error. Everything else looks like separate bugs, regressions, or docs issues in different code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45310 is the strongest canonical candidate because it matches 45478 on the same Qwen3.5 Moe `from_pretrained` error title, has an inbound reference, and has a slightly longer body with more context.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the strongest representative issue in this cluster because it is the original, clearly scoped report for the one true duplicate pair.", + "best_issue_reason": "Issue 45310 is the best representative issue in this cluster because it is the confirmed duplicate target for 45478, while also providing slightly more context and an inbound reference.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different models and failures: OmDet-Turbo processor input-size mismatch vs LwDetrImageLoss AMP crash." + }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "Different failures: DAC `from_latents` vs GraniteSpeech CI device-override test. Same broad runtime area, but not the same bug." + "reason": "Both are model/runtime bugs, but the DAC STE mismatch and GraniteSpeech stale device override are unrelated code paths." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs-related, but one is about pipeline removals for summarization/translation while the other is about `return_all_scores` behavior in text classification docs." + "reason": "Both are docs/pipeline related, but one is about removed `pipeline()` usage and the other about `return_all_scores` docs for text classification." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "Unrelated components and symptoms: DAC latent conversion vs Qwen3 Omni MoE feature-length helper behavior." + "reason": "Unrelated subsystems: DAC latent reconstruction vs qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "Different model paths and errors: unsupported `center` arg in `LasrFeatureExtractor` vs Whisper empty-transcription failure after `align_special_tokens`." + "reason": "Both involve audio-ish workflows, but they describe distinct regressions: unsupported `center` arg vs empty transcription after special-token alignment." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "Distinct bugs: SwitchTransformers config creates a sparse layer incorrectly, while Bamba fails under `torch.compile` with SDPA." + "reason": "Different models and failures: SwitchTransformers sparse-layer construction vs Bamba torch.compile SDPA behavior." }, { "left": "issue:44857", "right": "issue:45593", "accept": false, - "reason": "Different loss/model logic: LwDetr AMP crash vs D-FINE not using auxiliary losses when denoising is off." + "reason": "Different detection/training behaviors: AMP crash in LwDetrImageLoss vs missing auxiliary losses in D-FINE." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "A CPU offload device-mismatch in Bark-small is not the same as the Qwen3 Omni MoE helper-function issue." + "reason": "CPU offload device mismatch in Bark is unrelated to qwen3_omni_moe feature-length calculations." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "One is a failing Janus image-generation test, the other is a stale warning in paged generation; no shared code-path bug." + "reason": "A failing Janus image-generation test is not the same bug as a stale `num_return_sequences` warning in paged generate()." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "An import error for `_flash_supports_window_size` is unrelated to Torch 2.10 incompatibility in general." + "reason": "ImportError for a flash-attention helper and Torch 2.10 incompatibility are different release/runtime issues." }, { - "left": "issue:44448", - "right": "issue:44743", + "left": "issue:45127", + "right": "issue:45245", "accept": false, - "reason": "Pegasus output differences across v4/v5 and Qwen recurrent-state resets are separate model-behavior regressions." + "reason": "LoRA merge/collapsed outputs and the 2^24 category limit are unrelated model/data constraints." }, { - "left": "issue:45127", - "right": "issue:45245", + "left": "issue:44448", + "right": "issue:44743", "accept": false, - "reason": "LoRA merge collapse with extended vocab is unrelated to the `2^24` category limit runtime error." + "reason": "Different code paths: Pegasus v4/v5 output drift vs recurrent state reset in modular_qwen3_5." }, { "left": "issue:45200", "right": "issue:45588", "accept": false, - "reason": "Gemma4 text-only fine-tuning input handling and flash-attention `s_aux=None` crashes are different bugs." + "reason": "Gemma4 text-only token-type defaults and flash_attention `s_aux=None` handling are separate bugs." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and BigBird tokenizer special-token registration are unrelated issues." + "reason": "SwitchTransformers sparse-layer creation and BigBirdTokenizer special-token registration are unrelated." }, { "left": "issue:43611", "right": "issue:43931", "accept": false, - "reason": "Both are loading errors, but one is about `base_model_prefix` compatibility and the other is a Qwen3-VL weight-shape mismatch." + "reason": "Base-model-prefix loading regression and Qwen3-VL weight-shape mismatch are different model-loading failures." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat-template crashes on tool-call messages and Qwen2.5-VL video position IDs are not the same underlying bug." + "reason": "Chat-template tool-call handling and Qwen2.5-VL vision position IDs are separate issues." }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "Python 3.13 `IndentationError` in DebertaV2 import is unrelated to the Qwen3 Omni MoE helper behavior." + "reason": "Python 3.13 torch.jit indentation parsing and qwen3_omni_moe feature-length helper behavior are unrelated." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "Bamba `torch.compile` SDPA failure and a stale GraniteSpeech CI test are not the same code-path problem." + "reason": "Bamba torch.compile SDPA failure is not the same as a stale GraniteSpeech CI device override test." }, { - "left": "issue:43819", - "right": "issue:44493", + "left": "issue:44206", + "right": "issue:45823", "accept": false, - "reason": "`from_latents`/STE mismatch in DAC is unrelated to the unexpected position-id key issue." + "reason": "Both are device/feature regressions, but one is a feature-extractor arg issue and the other is a Gemma4 PLE forward device mismatch." }, { - "left": "issue:44206", - "right": "issue:45823", + "left": "issue:43819", + "right": "issue:44493", "accept": false, - "reason": "Different models and failures: unsupported `center` arg vs Gemma4 PLE device mismatch during forward." + "reason": "DAC latent mismatch and position-id key warnings are unrelated model internals." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "SwitchTransformers sparse-layer creation and BEiT pretrained preset loading via a legacy field are distinct bugs." + "reason": "SwitchTransformers sparse-layer bug and Beit pretrained preset loading bug are different model families and failures." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and `_gpu_broadcast_one` rejecting `NoneType` are unrelated CI/runtime issues." + "reason": "Flash Attention 2 test failures and Trainer `_gpu_broadcast_one` NoneType errors do not describe the same underlying bug." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Both touch flash/offload areas, but one crashes on `s_aux=None` in flash attention while the other is a Qwen3.5 MoE rotary-embedding CPU-offload incompatibility." + "reason": "Flash-attention sinkless-model crash and Qwen3.5 Moe CPU-offload incompatibility are different runtime problems." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Missing `_flash_supports_window_size` and missing `HybridCache` are separate import regressions." + "reason": "Missing `_flash_supports_window_size` import and missing `HybridCache` import are separate API/export regressions." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Embedding dtype drift in evaluate and a stale translation-task error message in pipeline are unrelated." + "reason": "Embedding dtype drift between train/eval and a pipeline v5 translation message regression are unrelated." }, { "left": "issue:45459", "right": "issue:46234", "accept": false, - "reason": "Tokenizer protobuf error handling and a missing import-utils symbol are different import/compatibility issues." + "reason": "Tokenizer protobuf error masking and a missing `is_falcon_mamba_ssm_available` utility are separate import-utils issues." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "SigLIP2 documentation problems and BEiT fast image processor label reduction are unrelated." + "reason": "SigLIP2 docs errors and BeitImageProcessorFast `reduce_labels` behavior are different concerns." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers sparse-layer creation is unrelated to the Qwen3 Omni MoE feature-length helper behavior." + "reason": "SwitchTransformers sparse-layer creation is unrelated to qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "`gather_for_metrics` truncating tuple labels and `reduce_labels` returning one label are different label-processing bugs." + "reason": "Metric gathering label truncation and BeitImageProcessorFast label reduction are not the same bug." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "These are the same Qwen3.5 MoE `from_pretrained` error reported twice, with the same version-specific failure and no meaningful distinction in code path." + "reason": "Exact same Qwen3.5 Moe `from_pretrained` error on transformers>=5.4.0; 45478 is a duplicate report." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 doc issues and position-id key warnings are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer added-space behavior and embedding reinitialization on resize are separate tokenizer/model-init bugs." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading failure and GraniteSpeech stale device test failure are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformers sparse-layer configuration and GraniteSpeech CI device override do not share the same underlying defect." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The clustering is well grounded in the packet. The only accepted duplicate pair, issue:45310 vs issue:45478, is a strong match: identical title, same failure mode, and the earlier issue is a sensible canonical choice given its earlier timestamp and inbound reference. The rejected soft edges are generally conservative and based on distinct titles/code paths, without overstating duplicate evidence." + "accept": false, + "feedback": "The duplicate-pair call on 45310/45478 looks correct and the rest of the soft-edge rejections are conservative. However, the justification overreaches when it says 45310 has the \"richest thread\" among the pair: the packet shows 45478 has more discussion activity (4 vs 2), so that comparative claim is not grounded. Tighten the representative/canonical rationale to only the supported facts (same title, inbound reference, slightly longer body) and avoid unsupported thread-quality claims." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -8272,13 +8299,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 26, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:25251", "issue:29127", @@ -8290,6 +8317,7 @@ "issue:40444", "issue:42490", "issue:43278", + "issue:43335", "issue:43381", "issue:43412", "issue:43526", @@ -8297,7 +8325,6 @@ "issue:43575", "issue:43606", "issue:43611", - "issue:43653", "issue:43701", "issue:43746", "issue:43854", @@ -8308,7 +8335,6 @@ "issue:44493", "issue:44561", "issue:44568", - "issue:44743", "issue:44779", "issue:44857", "issue:44977", @@ -8321,6 +8347,7 @@ "issue:45491", "issue:45563", "issue:45588", + "issue:45593", "issue:45698", "issue:45803", "issue:45823", @@ -8330,11 +8357,13 @@ "issue:46285" ], "soft_pairs": [ + "issue:44448|issue:44779", + "issue:43335|issue:45593", "issue:42490|issue:43854", "issue:43606|issue:44857", "issue:45920|issue:46192", - "issue:38175|issue:44493", "issue:44360|issue:44561", + "issue:38175|issue:44493", "issue:44857|issue:45072", "issue:44977|issue:45229", "issue:44448|issue:44977", @@ -8358,224 +8387,229 @@ "issue:45198|issue:45803", "issue:45588|issue:45823", "issue:44448|issue:44466", - "issue:43653|issue:44743", "issue:25251|issue:32090", "issue:45245|issue:45563" ], - "prepared_review_unit_hash": "0129e108b81707004d8ddc72ab8795edb795cd3493437fe58506ff3525f3fd69", + "prepared_review_unit_hash": "2b5a0ae93dd148847c34cbf4db58d4db58676dd16f47d07b895170d00ef58e5e", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37199, - "estimated_input_tokens": 9300, - "estimated_eval_tokens": 18856 + "soft_pair_count": 32, + "serialized_chars": 37485, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are only loosely related by broad themes like v5 regressions, tokenization, serialization, and inference edge cases. I don\u2019t see any soft pair that looks like the same underlying bug/change, so all candidate edges are rejected.", - "confidence": 0.61, - "canonical_issue_reason": "issue:45920 is the broadest and most representative report in the set: it is open, affects multiple models/tokenizers, and describes a concrete v5 regression with clear user impact.", + "summary": "Mostly unrelated issue set with a few superficial v5/regression and tokenizer/load-save themes; no true duplicate pair stands out.", + "confidence": 0.32, + "canonical_issue_reason": "Issue 44448 is the broadest representative of the v4/v5 behavior-regression theme and is one of the more centrally connected items in the soft-pair graph.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the best issue candidate for triage because it is the most general, well-described, and still actionable report among the provided issues.", + "best_issue_reason": "44448 is the best single issue anchor for this cluster because it reflects an observable version-regression in model behavior, while most other issues are narrower component-specific bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both mention v4/v5 regressions, but one is Pegasus generation drift and the other is DeepSeek tokenizer IDs; different code paths." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Sparse-layer config bug vs D-FINE auxiliary-loss omission; unrelated models and failures." + }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "Both concern loading/import behavior, but one is a local-path slash warning and the other is a model-load failure in tests; different code paths and symptoms." + "reason": "Local-path trailing-slash import warning vs GLM-4.7 test load failure; different symptoms." }, { "left": "issue:43606", "right": "issue:44857", "accept": false, - "reason": "Different failures: CPU offload device mismatch for bark-small versus float16 AMP crash in LwDetrImageLoss." + "reason": "CPU-offload device mismatch vs FP16 CUDA crash in a loss module; not the same bug." }, { "left": "issue:45920", "right": "issue:46192", "accept": false, - "reason": "One is a tokenizer regression across several models; the other is `return_tensors` being ignored when `text_kwargs` is passed. Related area, not the same bug." + "reason": "Both are tokenization-related, but one is wrong IDs across models and the other is ignored return_tensors kwargs." }, { - "left": "issue:38175", - "right": "issue:44493", + "left": "issue:44360", + "right": "issue:44561", "accept": false, - "reason": "One is zero probabilities in SigLIP2; the other is position-id key warnings across models. Different underlying issues." + "reason": "DSA indexer activation omission vs v5 removal of is_torch_fx_available; no shared failure mode." }, { - "left": "issue:44360", - "right": "issue:44561", + "left": "issue:38175", + "right": "issue:44493", "accept": false, - "reason": "DSA indexer/ReLU discussion is unrelated to removing `is_torch_fx_available` and trust_remote_code breakage." + "reason": "SigLIP2 zero probabilities and unexpected position-id keys are unrelated inference issues." }, { "left": "issue:44857", "right": "issue:45072", "accept": false, - "reason": "Both are dtype-related crashes, but they affect different models and different inference paths; not the same concrete bug." + "reason": "Both involve dtype problems, but different models, ops, and failure points." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation failure and multi-GPU CUDA OOM are distinct problems." + "reason": "Flash-attention generation bug vs multi-GPU OOM; distinct runtime issues." }, { "left": "issue:44448", "right": "issue:44977", "accept": false, - "reason": "Pegasus v4/v5 output differences are a model-regression issue, but unrelated to Qwen3.5 flash-attention generation." + "reason": "Both are v5 generation regressions, but Pegasus output drift and Qwen flash-attention failure are different paths." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Model-loading breakage for Llama Vision is not the same as post-init reinitialization of `lm_head` after resizing embeddings." + "reason": "Model loading breakage for Llama 3.2 vs embedding reinit after resize; unrelated." }, { "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "Checkpoint key mismatch and Qwen3 Omni helper-length behavior are different failure modes." + "reason": "Checkpoint key mismatch and a feature-length helper bug are unrelated." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "LayoutLMV3 error-message clarity is unrelated to the trainer `_gpu_broadcast_one` NoneType TypeError." + "reason": "LayoutLMv3 error messaging vs Trainer broadcast NoneType; no overlap." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "Bamba torch.compile/SDPA failure is not the same as the Qwen3 Omni feature-length helper bug." + "reason": "torch.compile/SDPA failure in Bamba is separate from qwen3_omni_moe length calculation." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Both involve save/load behavior, but one is custom token IDs in config and the other is wrong custom module resolution; different root causes." + "reason": "Both concern save/load, but one loses custom token IDs and the other loads the wrong custom module." }, { "left": "issue:40444", "right": "issue:43526", "accept": false, - "reason": "IterableDataset multi-image finetuning failure and BEiT fast processor label reduction are unrelated." + "reason": "IterableDataset multi-image finetuning failure vs BEiT label reduction bug." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures are not the same bug as embedding reinitialization after resizing." + "reason": "Flash Attention 2 test failures vs output embedding reinit after resize." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "Local-path import warning and tensor-parallel OOM are unrelated issues." + "reason": "Trailing-slash import warning vs tp OOM on Qwen2-57B; unrelated." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "Removal of `is_torch_fx_available` breaking remote-code models is unrelated to DeepSeek tokenizer regressions." + "reason": "Trust_remote_code breakage from API removal vs DeepSeek tokenizer regression; different root causes." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Qwen3.5 flash-attention generation and stale `num_return_sequences` warning are not the same underlying bug." + "reason": "Flash-attention generation issue vs stale warning for num_return_sequences." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Tokenizer special-token insertion failure is unrelated to Gemma3 NaN embeddings from sliding-window attention." + "reason": "Special tokens omission in tokenizer vs Gemma3 NaNs from sliding-window attention." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "Both mention runtime failures, but one is AMP dtype instability and the other is a flash-attention attribute error on sink-less models." + "reason": "Different crash sites: LwDetr AMP loss vs flash_attention.py s_aux=None." }, { "left": "issue:43746", "right": "issue:46209", "accept": false, - "reason": "Local checkpoint loading with PEFT adapters and FP8 inference on B200 are separate issues." + "reason": "PEFT checkpoint loading vs FP8 inference on B200; unrelated." }, { "left": "issue:45372", "right": "issue:45588", "accept": false, - "reason": "Gemma4 processor loading import failure is unrelated to the flash_attention `s_aux=None` crash." + "reason": "Gemma4 processor import failure vs flash_attention.py None aux crash." }, { "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Export-time training-guard/erfinv problems are different from the flash_attention sink-less model crash." + "reason": "Executorch export guard/erfinv issue vs flash_attention None aux crash." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "A dtype change between train/eval is not the same as gradient checkpointing being disallowed in eval mode." + "reason": "Eval/train dtype drift vs gradient checkpointing disallowed in eval; not the same bug." }, { "left": "issue:43611", "right": "issue:46285", "accept": false, - "reason": "Both are model-loading regressions, but the `base_model_prefix` breakage and CLIP structure inconsistency are distinct API issues." + "reason": "Base_model_prefix loading regression vs CLIPTextModel structure mismatch; distinct model-loading issues." }, { "left": "issue:45198", "right": "issue:45803", "accept": false, - "reason": "Wav2Vec2 save/tokenization failure and Fuyu batch feature exception handling are unrelated." + "reason": "Wav2Vec2 save/tokenization failure vs Fuyu bare-except bug." }, { "left": "issue:45588", "right": "issue:45823", "accept": false, - "reason": "Flash-attention `s_aux=None` crash and Gemma4 PLE device mismatch during forward are different code-path failures." + "reason": "flash_attention s_aux=None crash vs Gemma4 device_map mismatch; different components." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "Both are v5 regressions around model behavior/serialization, but one is output divergence and the other is device-dependent `lm_head.weight` serialization." - }, - { - "left": "issue:43653", - "right": "issue:44743", - "accept": false, - "reason": "BigBirdTokenizer special-token registration and Qwen3.5 recurrent-state reset are unrelated." + "reason": "One is output drift between v4/v5, the other is lm_head serialization depending on device." }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "Pipeline top_k nesting behavior and trainer GPU broadcast NoneType errors are completely different bugs." + "reason": "Top_k pipeline output nesting bug vs GPU broadcast NoneType error." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "Category-cardinality runtime error and stale warning emission are unrelated." + "reason": "Category limit error vs stale generate warning." } ] }, @@ -8591,20 +8625,19 @@ "component_index": 1, "component_count": 18, "review_unit_index": 27, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:10000", "issue:15354", "issue:29127", "issue:29942", "issue:30064", - "issue:32090", "issue:33357", "issue:34567", "issue:36010", @@ -8635,6 +8668,7 @@ "issue:44479", "issue:44610", "issue:44734", + "issue:44743", "issue:44871", "issue:44977", "issue:45198", @@ -8649,10 +8683,11 @@ "issue:46121" ], "soft_pairs": [ + "issue:43653|issue:44743", "issue:36010|issue:37428", - "issue:44186|issue:45198", "issue:45459|issue:45712", "issue:43065|issue:45412", + "issue:44186|issue:45198", "issue:43611|issue:46121", "issue:44361|issue:45198", "issue:33357|issue:39692", @@ -8661,226 +8696,225 @@ "issue:45372|issue:45459", "issue:44206|issue:44479", "issue:29942|issue:30064", - "issue:43976|issue:45440", "issue:42915|issue:42994", + "issue:43976|issue:45440", "issue:43994|issue:45412", "issue:43278|issue:43575", - "issue:44265|issue:45198", "issue:43066|issue:44462", + "issue:44265|issue:45198", "issue:10000|issue:15354", "issue:42490|issue:43873", "issue:43873|issue:45229", "issue:43421|issue:45491", "issue:43381|issue:43701", "issue:44734|issue:45406", - "issue:43653|issue:44442", "issue:43065|issue:43994", + "issue:43653|issue:44442", "issue:44610|issue:45412", - "issue:41720|issue:45229", - "issue:32090|issue:33357" + "issue:41720|issue:45229" ], - "prepared_review_unit_hash": "7d3ac21e2297fd39986412580f3de28853dcd97940c1395da50fda0fa6d21382", + "prepared_review_unit_hash": "4b1ccfadf8ee22363058be0af29cee1d90a6b90d40a5e442d4558b0bb25a8278", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36447, - "estimated_input_tokens": 9112, - "estimated_eval_tokens": 18480 + "serialized_chars": 36437, + "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 not a duplicate set around one bug; it mixes many unrelated Transformers issues across tokenizers, quantization, serving, model loading, and model-specific regressions. None of the soft pairs look like the same underlying defect closely enough to merge.", - "confidence": 0.92, - "canonical_issue_reason": "No clear canonical issue: the items span distinct subsystems and different failure modes, so there is no single bug that safely represents the cluster.", + "summary": "This cluster is highly heterogeneous: most items are unrelated model-, tokenizer-, quantization-, serving-, or environment-specific bugs rather than duplicates of one underlying defect. No strong duplicate canonical emerges.", + "confidence": 0.97, + "canonical_issue_reason": "No single issue is a clear canonical target; the cluster spans many unrelated failure modes across different models and subsystems.", "canonical_pr_reason": null, - "best_issue_reason": "No best issue candidate is suitable as a cluster representative because the reports are too heterogeneous and would mislead triage.", + "best_issue_reason": "No issue is a strong global representative. If forced, 43873 is the broadest/general discussion, but it is still not a duplicate hub.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:36010", - "right": "issue:37428", + "left": "issue:43653", + "right": "issue:44743", "accept": false, - "reason": "Both are import errors, but for different symbols in different modules and likely unrelated regressions." + "reason": "Tokenizer special-token registration/empty decode is unrelated to Qwen recurrent-state cache reset behavior." }, { - "left": "issue:44186", - "right": "issue:45198", + "left": "issue:36010", + "right": "issue:37428", "accept": false, - "reason": "Different tokenizer/model areas and different failures: LayoutLMv2 NER padding crash vs Wav2Vec2 save/tokenization issues." + "reason": "These are different import errors in different modules with no shared underlying code-path." }, { "left": "issue:45459", "right": "issue:45712", "accept": false, - "reason": "One is a protobuf-related tokenizer error-handling bug; the other is leftover dummy classes leaking into module listings." + "reason": "Hiding tokenizer errors when protobuf is missing is unrelated to leftover dummy classes leaking into dir(transformers)." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Different models and different problems: Sam3 pixel decoder dummy conv vs RT-DETR memory release on deletion." + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory not being released on deletion." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Both involve tokenization, but they affect different models and different failure modes; not the same bug." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "One is a base_model_prefix loading regression; the other is a rope-validation type error with JSON-loaded lists." + "reason": "base_model_prefix loading breaks are unrelated to rope-validation TypeError on a JSON-loaded list." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "Both involve tokenization, but one is an AttributeError in MLukeTokenizer while the other is save_pretrained/tokenization failure for Wav2Vec2." + "reason": "MLukeTokenizer AttributeError is a different tokenizer issue from Wav2Vec2 save/tokenization failure." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "Both mention CLIP/SigLIP-like vision-language use cases, but one is a MacOS bus error and the other is doc-example/model-processor mismatch plus quantization failure." + "reason": "Both mention vision models, but the concrete bugs differ: MacOS bus error vs docs/model-processor mismatch and quantization failure." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Both concern Gemma family generation behavior, but one is eos_token_id config mismatch and the other is flash-attention generation failure." + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation failure." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error-message clarity and TrainerState token counting are unrelated." + "reason": "LayoutLMv3 error handling for boxes and TrainerState token counting are unrelated subsystems." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "Different root causes: Gemma 4 processor import failure vs tokenizer protobuf error masking." + "reason": "Gemma 4 processor loading blocked by a mistral_common import is not the same as tokenizer error masking without protobuf." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Both are version regressions, but one is a LaserFeatureExtractor arg crash and the other is a video input regression across Qwen VL variants." + "reason": "Unsupported feature-extractor argument regression and video-input regression are different code paths and models." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Different attention/image-processing issues: Flash Attention 2 test failures vs void segmentation map handling." + "reason": "Flash Attention 2 test failures are unrelated to image processor void-segmentation-map handling." }, { - "left": "issue:43976", - "right": "issue:45440", + "left": "issue:42915", + "right": "issue:42994", "accept": false, - "reason": "Different failures: Python version incompatibility in 5.1.0 vs DeepseekV3MoE divergence from remote implementation." + "reason": "FP8 quantization config failure on load and quantized model saving failure are distinct bugs." }, { - "left": "issue:42915", - "right": "issue:42994", + "left": "issue:43976", + "right": "issue:45440", "accept": false, - "reason": "Both are quantization-related Qwen3Moe problems, but one is FineGrainedFP8Config failure and the other is quantized model saving failure." + "reason": "Python version compatibility regression is unrelated to DeepSeekV3 implementation divergence." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "Different subsystems: SigLIP2 bad outputs with AutoModel/pipeline vs RT-DETR memory cleanup." + "reason": "SigLIP2 nonsense outputs and RT-DETR memory leaks are unrelated issues." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "One is a dtype mismatch between train/eval; the other is tensor-parallel OOM on a large Qwen2 model." + "reason": "Embedding dtype drift in eval and Qwen2 OOM under tensor parallelism are different problems." }, { - "left": "issue:44265", - "right": "issue:45198", + "left": "issue:43066", + "right": "issue:44462", "accept": false, - "reason": "Export failure due to torch_compilable_check is unrelated to Wav2Vec2 save/tokenization issues." + "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring repository tokenizer.json." }, { - "left": "issue:43066", - "right": "issue:44462", + "left": "issue:44265", + "right": "issue:45198", "accept": false, - "reason": "Tokenizer decoder type in v5 and AutoTokenizer ignoring tokenizer.json are distinct loading/serialization bugs." + "reason": "torch.export failures from torch_compilable_check are unrelated to Wav2Vec2 save/tokenization failures." }, { "left": "issue:10000", "right": "issue:15354", "accept": false, - "reason": "German DistilBertModel issue and torch.jit.script GeneratorExp export failure are unrelated model/export problems." + "reason": "German DistilBertModel and GeneratorExp/TorchScript export are unrelated model/framework issues." }, { "left": "issue:42490", "right": "issue:43873", "accept": false, - "reason": "Local-path trailing-slash import warning is unrelated to quantization offloading behavior." + "reason": "A trailing-slash import warning is unrelated to quantization offloading not working." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Both involve memory/quantization concerns, but one is offloading not working while the other is multi-GPU CUDA OOM on Gemma4." + "reason": "Offloading/quantization behavior is not the same as multi-GPU CUDA OOM for Gemma4." }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Runtime special-token post-processing update bug is unrelated to Gemma3 NaN embeddings from sliding-window attention." + "reason": "Runtime post-processor updates in TokenizersBackend are unrelated to Gemma3 NaN embeddings with mixed-length batching." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Eval-mode gradient checkpointing limitation and resume_from_checkpoint key mismatch are separate training issues." + "reason": "Eval-mode gradient checkpointing and resume-from-checkpoint key mismatch are different training-flow bugs." }, { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "Different serve-time failures: KV-cache continuation tensor indexing vs Gemma4Processor missing _tokenizer." + "reason": "KV-cache continuation tensor indexing in serve is unrelated to Gemma4Processor missing _tokenizer." }, { - "left": "issue:43653", - "right": "issue:44442", + "left": "issue:43065", + "right": "issue:43994", "accept": false, - "reason": "BigBirdTokenizer special-token registration/empty decode is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + "reason": "A dummy layer in Sam3PixelDecoder is unrelated to SigLIP2 wrong/nonsensical outputs." }, { - "left": "issue:43065", - "right": "issue:43994", + "left": "issue:43653", + "right": "issue:44442", "accept": false, - "reason": "Different model families and different defects: dummy Conv2d in Sam3PixelDecoder vs SigLIP2 nonsensical outputs." + "reason": "BigBirdTokenizer special-token registration is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." }, { "left": "issue:44610", "right": "issue:45412", "accept": false, - "reason": "Processor/model input-size mismatch for OmDet-Turbo is unrelated to RT-DETR memory release." + "reason": "Processor image-size mismatch and RT-DETR memory leaks are different bugs." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Different inference failures: Qwen3 auto device mapping cudaErrorAssert vs Gemma4 multi-GPU CUDA OOM." - }, - { - "left": "issue:32090", - "right": "issue:33357", - "accept": false, - "reason": "Trainer _gpu_broadcast_one NoneType error is unrelated to MacOS bus error with community CLIP model." + "reason": "Qwen3 auto device mapping cudaErrorAssert and Gemma4 multi-GPU OOM are different failure modes." } ] }, @@ -8896,16 +8930,17 @@ "component_index": 1, "component_count": 18, "review_unit_index": 28, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:29127", "issue:29942", + "issue:32090", "issue:33357", "issue:35141", "issue:36010", @@ -8928,7 +8963,6 @@ "issue:43761", "issue:43824", "issue:43931", - "issue:44062", "issue:44077", "issue:44186", "issue:44206", @@ -8950,12 +8984,12 @@ "issue:45588", "issue:45593", "issue:45834", - "issue:45859", - "issue:45907" + "issue:45859" ], "soft_pairs": [ - "issue:43278|issue:43329", + "issue:32090|issue:33357", "issue:43756|issue:45440", + "issue:43278|issue:43329", "issue:45372|issue:45406", "issue:43526|issue:45834", "issue:29942|issue:36010", @@ -8967,211 +9001,210 @@ "issue:43575|issue:43701", "issue:43575|issue:44387", "issue:42490|issue:43381", - "issue:43526|issue:44857", "issue:44977|issue:45245", + "issue:43526|issue:44857", "issue:43334|issue:43531", "issue:43824|issue:44279", "issue:44442|issue:45072", - "issue:43646|issue:44077", "issue:29127|issue:35141", + "issue:43646|issue:44077", "issue:44743|issue:44811", "issue:42915|issue:43931", "issue:43262|issue:44206", "issue:43452|issue:45588", "issue:44977|issue:45584", - "issue:33357|issue:38175", - "issue:44062|issue:45907" + "issue:33357|issue:38175" ], - "prepared_review_unit_hash": "b2396327881f87b52b041188ec68acee268a6f1de3f956ff2e864e12e966d93b", + "prepared_review_unit_hash": "472958a73cff785d13966d21fb63f26b556bcbc72c8fc144b9c2a05ddcf878a0", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35869, - "estimated_input_tokens": 8968, - "estimated_eval_tokens": 18192 + "serialized_chars": 35308, + "estimated_input_tokens": 8827, + "estimated_eval_tokens": 17910 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a noisy mix of unrelated Transformers issues across many models and subsystems. I found no credible duplicate center; all proposed soft edges are false positives.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the items span unrelated failures (tokenizers, audio, vision, quantization, generation, import errors) with different root causes.", + "summary": "This cluster is heterogeneous: the candidate pairs mostly share only broad subsystem vocabulary, not the same underlying bug. I would not merge any of the soft edges. If a single representative issue is needed, issue 43761 is the cleanest, most concrete regression report in the set.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 43761 is the most self-contained and clearly scoped report here, with a specific regression, concrete expected behavior, and an explicit failure condition/version context.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a good representative of the cluster because there is no common underlying bug or change linking these reports.", + "best_issue_reason": "Issue 43761 is a strong representative because it describes one precise code-path failure rather than a broad or speculative symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43278", - "right": "issue:43329", + "left": "issue:32090", + "right": "issue:33357", "accept": false, - "reason": "Different subsystems and failures: dtype drift in evaluation vs undefined multimodal token helper names in video token counting." + "reason": "Different failures: Trainer GPU broadcast type error vs MacOS bus error loading a CLIP model." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Unrelated models and behaviors: Smollm3 RoPE-layer placement vs DeepSeekV3 implementation divergence." + "reason": "Both concern model implementation correctness, but they target different models and different bugs." }, { - "left": "issue:45372", - "right": "issue:45406", + "left": "issue:43278", + "right": "issue:43329", "accept": false, - "reason": "Both mention Gemma 4 loading, but one is a missing mistral_common import and the other is a serve-side missing _tokenizer attribute; different code paths." + "reason": "Unrelated areas: embedding dtype drift vs multimodal token counting video-branch NameError-style bug." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma 4 loading problems, but one is a dependency import failure and the other is a missing processor attribute in serve." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug vs Kosmos2.5 OCR index error are separate image-processing issues." + "reason": "Both are vision/OCR-adjacent, but the concrete failures are different and do not share a code path." }, { "left": "issue:29942", "right": "issue:36010", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to a GenerationMixin import error." + "reason": "Flash Attention 2 test failures are not the same as a GenerationMixin import error." }, { "left": "issue:43723", "right": "issue:44554", "accept": false, - "reason": "Tokenizer loading regression in v5 is unrelated to an MPS attention correctness bug." + "reason": "Tokenizer-loading issue and MPS attention correctness bug are unrelated." }, { "left": "issue:43761", "right": "issue:45859", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and Qwen3 MoE CPU-offload incompatibility are different bugs." + "reason": "CLIP vision hidden-state regression is unrelated to Qwen rotary embedding CPU offload incompatibility." }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "DeepSeekV3 divergence is not the same as D-FINE missing auxiliary losses." + "reason": "Different models and different defects: remote implementation divergence vs missing auxiliary losses." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "auto_docstring annotation handling is unrelated to Gemma 4 processor import problems." + "reason": "Auto-docstring annotation handling crash is unrelated to the mistral_common ReasoningEffort import error." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer/image-preprocessing related at a high level, but they affect different models and distinct failure modes." + "reason": "Different components and failures: BEiT image processor label reduction vs LayoutLMv2 tokenizer padding/NER crash." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "Tensor-parallel OOM on model load is unrelated to resume_from_checkpoint key mismatches." + "reason": "Tensor-parallel OOM and resume checkpoint key mismatch are unrelated issues." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Both involve OOM wording, but one is model-loading TP memory pressure and the other is int4 quantization reserved-memory growth; not the same bug." + "reason": "Both mention OOM, but one is model-load TP memory pressure and the other is int4 quantization reserved-memory behavior." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Import warning on local path parsing is unrelated to gradient checkpointing being disallowed in eval mode." + "reason": "Local-model slash-path import warning is unrelated to gradient checkpointing in eval mode." }, { - "left": "issue:43526", - "right": "issue:44857", + "left": "issue:44977", + "right": "issue:45245", "accept": false, - "reason": "BeitImageProcessorFast label reduction and LwDetrImageLoss AMP crash are separate image bugs." + "reason": "Flash-attention generation bug and category-cardinality runtime error are unrelated." }, { - "left": "issue:44977", - "right": "issue:45245", + "left": "issue:43526", + "right": "issue:44857", "accept": false, - "reason": "Qwen3.5 flash-attention generation issue is unrelated to a categories cardinality limit error." + "reason": "Different image-loss/processor bugs with no shared concrete code path." }, { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Qwen3-VL missing pad_token_id and Qwen3-MoE sliding_window behavior are different configuration/load issues." + "reason": "Both involve Qwen3/Qwen3-MoE, but one is a missing pad_token_id on load and the other is a sliding-window runtime issue." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "A specific missing class import is not the same as a general dependency issue with transformers." + "reason": "ImportError for a specific model class is not the same as a general dependency issue." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "FastSpeech2ConformerTokenizer loading failure and bf16 dtype mismatches are unrelated." + "reason": "Tokenizer loading failure and dtype mismatch in inference are unrelated." }, { - "left": "issue:43646", - "right": "issue:44077", + "left": "issue:29127", + "right": "issue:35141", "accept": false, - "reason": "General custom model initialization breakage is too broad to equate with a specific PatchTSMixer post_init constraint." + "reason": "Different model subsystems and different symptoms: clearer error messaging vs output embedding reinitialization." }, { - "left": "issue:29127", - "right": "issue:35141", + "left": "issue:43646", + "right": "issue:44077", "accept": false, - "reason": "LayoutLMv3 error messaging and token-embedding reinitialization are different code paths and concerns." + "reason": "Transformers 5.0 custom initialization breakage is broader and distinct from patchtsmixer's optional post_init restriction." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Qwen3.5 recurrent-state reset under cache is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "Qwen recurrent-state caching bug is unrelated to Whisper batch_decode special-token handling." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe is unrelated to Qwen3-VL weight-shape mismatch." + "reason": "FineGrainedFP8Config failure and Qwen3-VL weight-shape mismatch are different loading problems." }, { "left": "issue:43262", "right": "issue:44206", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and unsupported center-arg propagation are separate audio regressions." + "reason": "Audio chat-template sampling-rate default mismatch is unrelated to an unsupported center argument crash." }, { "left": "issue:43452", "right": "issue:45588", "accept": false, - "reason": "gguf_file loading regressions and flash_attention s_aux=None crashes are unrelated." + "reason": "Tokenizer/model loading with gguf_file has no concrete overlap with flash_attention s_aux=None handling." }, { "left": "issue:44977", "right": "issue:45584", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug is unrelated to Whisper empty-transcription failures." + "reason": "Flash-attention generation issue is unrelated to Whisper empty-transcription failure." }, { "left": "issue:33357", "right": "issue:38175", "accept": false, - "reason": "Both involve vision models, but a MacOS bus error and zero-probability outputs are not the same defect." - }, - { - "left": "issue:44062", - "right": "issue:45907", - "accept": false, - "reason": "AddedToken argument duplication and list_repo_files revision forwarding are unrelated API bugs." + "reason": "Both are vision-model symptoms, but one is a MacOS bus error and the other is unexpected zero probabilities." } ] }, @@ -9187,13 +9220,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 29, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28218", "issue:32090", @@ -9215,6 +9248,7 @@ "issue:43827", "issue:43873", "issue:43931", + "issue:44062", "issue:44077", "issue:44188", "issue:44190", @@ -9229,7 +9263,6 @@ "issue:44898", "issue:44913", "issue:44964", - "issue:44987", "issue:45137", "issue:45216", "issue:45290", @@ -9239,13 +9272,14 @@ "issue:45698", "issue:45759", "issue:45803", - "issue:46178", + "issue:45907", "issue:46209", "issue:46234" ], "soft_pairs": [ - "issue:28218|issue:32090", + "issue:44062|issue:45907", "issue:44464|issue:45491", + "issue:28218|issue:32090", "issue:33290|issue:34567", "issue:44442|issue:44898", "issue:43381|issue:43873", @@ -9255,39 +9289,38 @@ "issue:32090|issue:33290", "issue:43720|issue:44442", "issue:44479|issue:45290", - "issue:34689|issue:36010", "issue:43931|issue:45216", + "issue:34689|issue:36010", "issue:44442|issue:45803", "issue:43550|issue:44743", "issue:43493|issue:43827", "issue:43873|issue:44871", "issue:42907|issue:45698", "issue:42175|issue:46234", + "issue:43761|issue:44188", "issue:43643|issue:44913", "issue:43316|issue:45759", - "issue:43761|issue:44188", "issue:45290|issue:46209", "issue:44855|issue:45137", "issue:33290|issue:34689", - "issue:44568|issue:45479", "issue:44190|issue:44496", - "issue:44077|issue:44479", - "issue:44987|issue:46178" + "issue:44568|issue:45479", + "issue:44077|issue:44479" ], - "prepared_review_unit_hash": "55f68b09cfb97bc0c5dd2b76fcb1618c43d9823868c1d6711a17de7c9bb08448", + "prepared_review_unit_hash": "3e4d3e965a1bc3d1f47fd0042240d080f1bfd199076b5a3936b8d7fb1cea13ab", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36147, + "serialized_chars": 36145, "estimated_input_tokens": 9037, "estimated_eval_tokens": 18330 }, @@ -9296,186 +9329,192 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The soft links are mostly false positives: they span unrelated transformers bugs across tokenization, training, quantization/offloading, multimodal model loading, and docs. I would not merge any of these pairs as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue fits this cluster; the items are heterogeneous and do not share one underlying bug or change.", + "summary": "This cluster is mostly a set of unrelated Transformers bug reports that only share broad topical similarity (loading, configs, generation, quantization). I found no soft pair that looks like the same underlying issue, so none should be merged as duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "No clear duplicate cluster exists here. If one issue must represent the set, issue:45698 is the broadest and most active loading/serialization regression report, with multiple inbound refs and high discussion activity.", "canonical_pr_reason": null, - "best_issue_reason": "If a representative is required, issue:28218 is the cleanest standalone report, but it only covers tokenizer spacing and does not represent the rest of the cluster.", + "best_issue_reason": "Issue:45698 is the best single representative because it is active, well-discussed, and concerns a core save/load path that is plausibly adjacent to several other configuration/loading regressions in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:28218", - "right": "issue:32090", + "left": "issue:44062", + "right": "issue:45907", "accept": false, - "reason": "Tokenizer added-space bug vs Trainer `_gpu_broadcast_one` TypeError; unrelated code paths." + "reason": "Completely different code paths: AddedToken construction vs Hub repo file listing/revision forwarding." }, { "left": "issue:44464", "right": "issue:45491", "accept": false, - "reason": "Compiled-forward chunked generation mismatch vs Gemma3 NaN embeddings from sliding-window padding; different failures." + "reason": "Both are generation/model-behavior bugs, but the concrete failures differ: compiled forward inconsistency vs Gemma3 sliding-window NaNs on mixed-length batches." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-space behavior is unrelated to the Trainer/NoneType GPU broadcast error." }, { "left": "issue:33290", "right": "issue:34567", "accept": false, - "reason": "DeepSpeed Adafactor OOM vs TrainerState token counter not updating; not the same bug." + "reason": "DeepSpeed/Adafactor OOM and num_input_tokens_seen not updating are unrelated trainer issues." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "AutoTokenizer load failure vs Perceiver image-size/interpolation failure; different subsystems." + "reason": "Tokenizer loading failure and Perceiver interpolation failure are unrelated model-specific bugs." }, { "left": "issue:43381", "right": "issue:43873", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction vs quantization offloading bug; separate behaviors." + "reason": "Gradient checkpointing eval-mode restriction and quantization offloading failure are different mechanisms." }, { "left": "issue:44964", "right": "issue:45440", "accept": false, - "reason": "Phi-4 multimodal load regression vs DeepSeekV3 implementation divergence; not a shared code-path." + "reason": "Separate multimodal model-loading regression vs DeepSeekV3MoE implementation divergence." }, { "left": "issue:43316", "right": "issue:43525", "accept": false, - "reason": "Gemma3TextConfig API mismatch vs Llama4Config missing `pad_token_id`; different model/config issues." + "reason": "Both are config/API issues, but for different models and different missing fields; not the same bug." }, { "left": "issue:34689", "right": "issue:43611", "accept": false, - "reason": "Llama 3.2 model loading regression vs `base_model_prefix` loading regression; too broad and different causes." + "reason": "Both involve version-related loading breakages, but one is a specific model loader regression and the other is a base_model_prefix handling change." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "GPU broadcast NoneType error vs OOM with Adafactor/DeepSpeed; unrelated training failures." + "reason": "Trainer broadcast NoneType error and DeepSpeed Adafactor OOM do not share a concrete failure mode." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "BitNet packed-weight loading bug vs tokenizer loading bug; different components and fixes." + "reason": "BitNet packed-weight loading and FastSpeech2 tokenizer loading are unrelated." }, { "left": "issue:44479", "right": "issue:45290", "accept": false, - "reason": "Video-input regression for Qwen VL models vs chat-template tool-call crash; not the same bug." + "reason": "Video-input regression across Qwen VL models is unrelated to chat-template tokenization crashing on tool-call messages." }, { - "left": "issue:34689", - "right": "issue:36010", + "left": "issue:43931", + "right": "issue:45216", "accept": false, - "reason": "Model loading regression vs missing `GenerationMixin` import; unrelated issues." + "reason": "Model shape mismatch on load and incorrect checkpoint saving are separate serialization/load-path problems." }, { - "left": "issue:43931", - "right": "issue:45216", + "left": "issue:34689", + "right": "issue:36010", "accept": false, - "reason": "Weight-shape mismatch on model load vs incorrect checkpoint save_pretrained regression; different directions of failure." + "reason": "Both mention loading/import breakage, but one is a model-load regression and the other is a GenerationMixin import path issue." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer load failure vs Fuyu batch-feature exception handling bug; no shared underlying defect." + "reason": "Tokenizer load failure and Fuyu tensor conversion exception swallowing are unrelated." }, { "left": "issue:43550", "right": "issue:44743", "accept": false, - "reason": "Bamba torch.compile/SDPA failure vs Qwen recurrent-state reset bug; different models and mechanics." + "reason": "torch.compile SDPA failure in Bamba and recurrent-state reset in modular_qwen3_5 are different model/runtime bugs." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation discrepancy vs docs still referencing removed `pipeline()`; one is code, the other documentation." + "reason": "SigLIP2 implementation mismatch and docs still mentioning pipeline() are unrelated." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading issue vs Gemma3 eos token config mismatch; not duplicates." + "reason": "Quantization offloading behavior and Gemma3 eos_token_id mismatch are not the same underlying issue." }, { "left": "issue:42907", "right": "issue:45698", "accept": false, - "reason": "Dequantized save failure vs wrong custom module loaded after save_pretrained; both serialization-related but distinct bugs." + "reason": "Both touch save/load flows, but one is about saving dequantized Ministrals/Devstrals and the other about loading the wrong custom module after save_pretrained; too different to merge." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Packaging/backends issue with TensorFlow not included vs missing import utility symbol; unrelated." + "reason": "Backend dependency installation and a missing import_utils helper are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and attention-kernel divergence under torch.compile are different bugs." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "`trust_remote_code` dropping fields vs GPTNeoX `rotary_pct` not persisting on reload; different config paths." + "reason": "Missing fields with trust_remote_code and rotary_pct not persisting on reload are distinct config-handling problems." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Gemma3TextConfig API discrepancy vs Qwen multimodal `text_config` unwrapping bug; not the same model behavior." - }, - { - "left": "issue:43761", - "right": "issue:44188", - "accept": false, - "reason": "CLIPVision hidden-states regression vs attention-kernel divergence under `torch.compile`; different code paths." + "reason": "Gemma3TextConfig API discrepancy and Qwen multimodal text_config unwrapping are related only at a very high level; they are not the same concrete defect." }, { "left": "issue:45290", "right": "issue:46209", "accept": false, - "reason": "Tool-call chat-template crash vs FP8 inference failure on B200; unrelated runtime issues." + "reason": "Chat-template tool-call crash is unrelated to FP8 inference failure on B200." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Python 3.13 decorator parsing issue vs DeepSpeed ZeRO3 deque underflow; no common root cause." + "reason": "Python 3.13 Deberta import parsing error and DeepSpeed ZeRO3 deque IndexError are unrelated." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "Adafactor/DeepSpeed OOM vs model-loading regression for Llama 3.2; different layers of the stack." + "reason": "Adafactor/DeepSpeed OOM and Llama 3.2 loading regression are unrelated." }, { - "left": "issue:44568", - "right": "issue:45479", + "left": "issue:44190", + "right": "issue:44496", "accept": false, - "reason": "Tokenizer special-token addition bug vs degenerate zero-loss sequence-classification bug; not duplicates." + "reason": "Local dataset loading in an example script and unrecognized model type in config.json are different failures." }, { - "left": "issue:44190", - "right": "issue:44496", + "left": "issue:44568", + "right": "issue:45479", "accept": false, - "reason": "Local dataset script issue vs unrecognized model/config issue; unrelated." + "reason": "Tokenizer add_special_tokens regression and sequence-classification zero-loss bug are unrelated." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "Optional `post_init` policy issue vs video-input regression; different concerns." + "reason": "PatchTSMixer post_init validation and Qwen VL video-input regression are unrelated." }, { - "left": "issue:44987", - "right": "issue:46178", + "left": "issue:43873", + "right": "issue:44871", "accept": false, - "reason": "Model loading failure for a specific repo vs missing backend mapping for `tensorflow_text`; separate bugs." + "reason": "Quantization offloading behavior and Gemma3 token-ID inconsistency are different issues." } ] }, @@ -9491,13 +9530,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 30, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:25251", "issue:28282", @@ -9516,7 +9555,6 @@ "issue:43441", "issue:43452", "issue:43475", - "issue:43504", "issue:43540", "issue:43550", "issue:43575", @@ -9537,33 +9575,35 @@ "issue:44871", "issue:44877", "issue:44913", + "issue:44987", "issue:45127", "issue:45200", "issue:45584", "issue:45656", - "issue:45698", "issue:45803", "issue:46050", "issue:46170", + "issue:46178", "issue:46202" ], "soft_pairs": [ "issue:43650|issue:46202", + "issue:44987|issue:46178", "issue:29127|issue:33357", "issue:25251|issue:28282", "issue:43295|issue:43825", - "issue:42617|issue:43575", "issue:43550|issue:43701", + "issue:42617|issue:43575", "issue:43606|issue:43701", "issue:33290|issue:35141", "issue:43575|issue:43742", "issue:43540|issue:44514", - "issue:43316|issue:43716", "issue:33290|issue:45656", + "issue:43316|issue:43716", "issue:44871|issue:45584", - "issue:43316|issue:44877", - "issue:43404|issue:43441", "issue:44162|issue:44792", + "issue:43404|issue:43441", + "issue:43316|issue:44877", "issue:43295|issue:44871", "issue:45803|issue:46170", "issue:43065|issue:44617", @@ -9573,199 +9613,198 @@ "issue:43452|issue:44496", "issue:45127|issue:45200", "issue:43761|issue:44479", - "issue:43232|issue:43388", - "issue:43504|issue:45698" + "issue:43232|issue:43388" ], - "prepared_review_unit_hash": "c2053b6ddbed0b8af52dcee3dce3bdd0e8533f9dab1fb7188d780c65fecb3a96", + "prepared_review_unit_hash": "68d17b440e64b82f630c97bfa68be107d12449be8bc1803168485d46ea548e43", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35126, - "estimated_input_tokens": 8782, - "estimated_eval_tokens": 17820 + "serialized_chars": 35132, + "estimated_input_tokens": 8783, + "estimated_eval_tokens": 17822 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it spans unrelated bugs in loading/saving, multimodal processing, generation, optimizer behavior, and model-specific regressions. None of the suggested soft pairs look like the same underlying bug or change, so I reject all of them. If one issue must anchor the cluster, the custom save/load regression in #45698 is the broadest and most active representative.", - "confidence": 0.93, - "canonical_issue_reason": "If forced to choose one anchor, #45698 is the broadest active issue in the group: it concerns save_pretrained/from_pretrained custom-module loading, has multiple inbound references, and overlaps with several other load/serialization-themed reports at a high level.", + "summary": "This cluster is a grab bag of unrelated Transformers issues; the soft links are mostly title-level similarity only, so none should be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": "If a canonical has to be chosen, issue:44871 is the least-bad anchor because it is concrete, reproducible, and appears in multiple soft-link candidates, but the overall cluster is too heterogeneous to support a strong canonical.", "canonical_pr_reason": null, - "best_issue_reason": "#45698 is the strongest standalone representative because it is open, actively discussed, and describes a concrete, user-facing save/load regression with broad impact.", + "best_issue_reason": "Issue:44871 is the most actionable representative here: a specific Gemma config mismatch with clear symptoms, unlike the broader or more ambiguous neighboring issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43650", "right": "issue:46202", "accept": false, - "reason": "Unrelated: one is a vague \"add the data\" issue, the other is a Gemma pipeline roundtrip/tokenization bug." + "reason": "Unrelated: a placeholder/open-ended issue versus a Gemma pipeline roundtrip bug around a special token." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Different failure modes and code paths: model loading for one repo/model versus backend registration for tensorflow_text." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "Different failures: LayoutLMv3 error-message clarity versus a MacOS bus error in a CLIP community model." + "reason": "Distinct bugs: LayoutLMv3 box-content error versus a MacOS bus error in CLIP community model loading." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Different subsystems and symptoms: pipeline top_k nesting bug versus missing PyTorch ImportError." + "reason": "Not the same bug: pipeline top_k output nesting versus missing PyTorch import for AutoModel." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Both involve v5 regressions, but one is processor/tokenizer custom-code breakage and the other is a pipeline task-support error message; not the same bug." + "reason": "Both are regressions, but one is custom model code accessing processor.tokenizer while the other is a pipeline error-message/translation-support issue." }, { - "left": "issue:42617", - "right": "issue:43575", + "left": "issue:43550", + "right": "issue:43701", "accept": false, - "reason": "Different root causes: failure to run a script versus OOM while loading a large model with tensor parallelism." + "reason": "Completely different problems: torch.compile/SDPA OOM versus resume_from_checkpoint key mismatch." }, { - "left": "issue:43550", - "right": "issue:43701", + "left": "issue:42617", + "right": "issue:43575", "accept": false, - "reason": "Unrelated: torch.compile/SDPA failure versus resume_from_checkpoint key mismatch." + "reason": "Different symptoms and likely different paths: a 3d_parallel.py runtime issue versus tensor-parallel OOM on a specific Qwen model." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "Device-mismatch with CPU offload is unrelated to checkpoint key mismatches." + "reason": "Device mismatch with CPU offload is unrelated to checkpoint key mismatches." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "Different areas: DeepSpeed/Adafactor OOM versus embedding reinitialization during post_init." + "reason": "Different bugs: DeepSpeed Adafactor OOM versus token embedding reinitialization during post_init." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "Both are loading issues, but one is tensor-parallel OOM and the other is a key error loading a specific model; different bugs." + "reason": "Both are load-time issues, but one is tensor-parallel OOM and the other is a KeyError while loading MobileLLM." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Different code paths: video-input processing in Qwen3OmniMoe versus batched chat-template handling." + "reason": "Both involve multimodal/video processing, but the failures are different processor/code-path bugs and not clearly mergeable." }, { - "left": "issue:43316", - "right": "issue:43716", + "left": "issue:33290", + "right": "issue:45656", "accept": false, - "reason": "Unrelated model/config API discrepancy versus image-preprocessor dtype mismatch." + "reason": "Related to DeepSpeed, but one is an OOM and the other is double optimizer stepping; not the same underlying defect." }, { - "left": "issue:33290", - "right": "issue:45656", + "left": "issue:43316", + "right": "issue:43716", "accept": false, - "reason": "Both mention DeepSpeed, but one is an optimizer OOM and the other is double optimizer stepping; not the same fix." + "reason": "Different models and different issues: Gemma3TextConfig API discrepancy versus image-preprocessor dtype mismatch in Mistral-3." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Different models and bugs: Gemma EOS token config inconsistency versus Whisper empty-transcription generation failure." + "reason": "Both are tokenization/generation edge cases, but one is Gemma eos_token_id inconsistency and the other is Whisper empty-transcription handling after align_special_tokens." }, { - "left": "issue:43316", - "right": "issue:44877", + "left": "issue:44162", + "right": "issue:44792", "accept": false, - "reason": "Different config issues: Gemma3TextConfig API mismatch versus granite_speech strict-config loading." + "reason": "Unrelated: ESM2 breakage versus a Janus image-generation test failure." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Distinct Mistral-family regressions: lm_head weight tying versus FlashAttention failure." + "reason": "Different model-specific regressions: lm_head weight tying versus FlashAttention failure in Ministral-3." }, { - "left": "issue:44162", - "right": "issue:44792", + "left": "issue:43316", + "right": "issue:44877", "accept": false, - "reason": "Unrelated: ESM2 model breakage versus a failing janus image-generation test." + "reason": "Config-related issues, but not the same defect: API mismatch in Gemma3TextConfig versus strict config validation blocking granite_speech." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "One is processor/tokenizer regression; the other is Gemma EOS token list/single-value inconsistency." + "reason": "Both are regression-style model/processor bugs, but they affect different mechanisms: processor.tokenizer access versus Gemma eos_token_id configuration." }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Different concerns: bare except swallowing interrupts versus Windows/ROCm torch.distributed dependency issues." + "reason": "Unrelated: a bare except swallowing errors versus a Windows ROCm torch.distributed import/dependency problem." }, { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "Same SAM3 family, but different problems: dummy convolution implementation versus CUDA OOM." + "reason": "Same product area (SAM3) but not the same issue: dummy Conv2d placeholder versus CUDA OOM." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Both are multimodal/video-related, but one is pipeline processor selection and the other is a missing attribute in Sam3VisionEncoderOutput." + "reason": "Different subsystems and symptoms: video-classification pipeline processor selection versus missing fpn_position_embeddings on a SAM3 vision encoder output." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "General Transformers 5.0 custom-init breakage is too broad to merge with a GPTNeoX config reload default bug." + "reason": "Both involve model init/config behavior, but one is a broad Transformers 5 custom-init breakage and the other is a GPTNeoX config field not persisting." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "Different serialization/config regressions: Gemma EOS token config versus quantized-model save/load breakage." + "reason": "Both concern save/load or config persistence, but one is Gemma eos_token_id inconsistency and the other is quantized model round-tripping after save_pretrained." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "Both are loading-related, but one is GGUF-specific tokenizer/model loading and the other is missing model_type in config.json." + "reason": "Both are model-loading failures, but one is gguf_file handling and the other is a missing model_type in config.json." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "Different training-time bugs: LoRA merge collapse after vocab extension versus missing mm_token_type_ids defaulting." + "reason": "Different feature areas: LoRA merge with extended vocabulary versus Gemma 4 multimodal token type IDs defaulting." }, { "left": "issue:43761", "right": "issue:44479", "accept": false, - "reason": "Different multimodal regressions: CLIPVisionModel hidden_states omission versus Qwen video-input regression." + "reason": "Different regressions in different models: CLIPVisionModel hidden_states behavior versus Qwen video-input handling." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation kwargs update after sync_gpus is unrelated to gather_for_metrics dropping tuple labels." - }, - { - "left": "issue:43504", - "right": "issue:45698", - "accept": false, - "reason": "Both affect loading, but one is a legacy-field preset load failure and the other is wrong custom-module selection after save/load; different root causes." + "reason": "Completely different code paths: generation kwargs sync handling versus gather_for_metrics dropping tuple labels." } ] }, @@ -9781,19 +9820,20 @@ "component_index": 1, "component_count": 18, "review_unit_index": 31, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:35141", "issue:41720", "issue:42490", "issue:43065", "issue:43335", + "issue:43504", "issue:43540", "issue:43550", "issue:43572", @@ -9810,12 +9850,10 @@ "issue:44242", "issue:44462", "issue:44479", - "issue:44521", "issue:44534", "issue:44610", "issue:44617", "issue:44734", - "issue:44737", "issue:44743", "issue:44792", "issue:44855", @@ -9826,12 +9864,12 @@ "issue:45127", "issue:45245", "issue:45357", - "issue:45381", "issue:45542", "issue:45584", "issue:45588", "issue:45600", "issue:45632", + "issue:45698", "issue:45803", "issue:45907", "issue:45987", @@ -9839,6 +9877,7 @@ "issue:46234" ], "soft_pairs": [ + "issue:43504|issue:45698", "issue:43572|issue:46178", "issue:44077|issue:44534", "issue:45542|issue:46234", @@ -9854,230 +9893,215 @@ "issue:43827|issue:43994", "issue:41720|issue:45245", "issue:43065|issue:44610", - "issue:45803|issue:45987", "issue:43931|issue:44060", + "issue:45803|issue:45987", "issue:35141|issue:42490", - "issue:43644|issue:44913", "issue:43335|issue:44855", + "issue:43644|issue:44913", "issue:43976|issue:44855", "issue:43550|issue:44190", "issue:45632|issue:46234", "issue:44077|issue:44877", "issue:43540|issue:44734", "issue:44462|issue:45907", - "issue:45127|issue:45357", - "issue:44521|issue:45381", - "issue:43335|issue:44190", - "issue:44737|issue:45803" + "issue:45127|issue:45357" ], - "prepared_review_unit_hash": "fc1ef8210b8fa777be54378565419aeee54a839dbf60c74ab6fd266749eec46e", + "prepared_review_unit_hash": "975b58e3002b250fea734bef48467125f5152ce12b479fa9c3bb0134d0365cbe", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36805, - "estimated_input_tokens": 9202, - "estimated_eval_tokens": 18660 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35603, + "estimated_input_tokens": 8901, + "estimated_eval_tokens": 18058 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous; the only clear duplicate pair is the Transformers v5 non-persistent-buffer initialization regression. Everything else points to distinct models, configs, or failure modes.", - "confidence": 0.88, - "canonical_issue_reason": "issue:43644 is the earliest and most specific report of the non-persistent-buffer junk initialization bug; issue:44534 is the same regression reworded later.", + "summary": "No soft pair looks like a true duplicate; the items mostly span unrelated bugs across different models, configs, loaders, and runtime paths. The strongest issue candidate is the open, heavily discussed custom-module loading regression.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45698 is the strongest representative issue: it is open, has the most discussion activity, and describes a concrete, user-facing loading regression with clear reproduction potential.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the best anchor because it is the earliest report, directly states the regression, and matches the one confirmed duplicate in the set.", + "best_issue_reason": "issue:45698 is the best standalone issue to keep as canonical because it has broad impact, active discussion, and a focused bug report about incorrect module loading after save/load.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Different bugs: one is a legacy-field preset loading failure for BEiT semantic segmentation, the other is wrong custom-module resolution after save/load." + }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "Different backend/config issues: StableLm pad_token_idx after v5 update vs undefined tensorflow_text backend registration." + "reason": "Unrelated regressions: StableLm pad token config metadata vs an import-utils backend registration error for tensorflow_text." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "Unrelated bugs: patchtsmixer post_init policy vs v5 buffer initialization corruption." + "reason": "Different code paths: a restriction on optional post_init in patchtsmixer versus a v5 buffer initialization corruption issue." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "TensorBoard/TensorFlow backend mismatch is unrelated to missing import_utils helper for falcon_mamba_ssm." + "reason": "Not the same problem: TensorBoard/TensorFlow backend setup versus a missing import_utils availability helper." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both involve cache/state, but different code paths and symptoms: missing GenerationMixin cache in chunked_prefill vs recurrent states resetting with seq_len>1." + "reason": "Both involve cache behavior, but they target different mechanisms and models: chunked_prefill cache presence vs recurrent state reset in modular_qwen3_5." }, { "left": "issue:43761", "right": "issue:44242", "accept": false, - "reason": "Different models and bugs: CLIP hidden_states regression vs load balancing loss omission." + "reason": "Different model behaviors: CLIP hidden_states output regression versus missing load-balancing loss when router logits are disabled." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus image-generation test failure is not the same as Qwen3.5 flash-attention generation corruption." + "reason": "Separate issues: a failing janus image-generation test versus Qwen3.5 flash-attention generation malfunction." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after align_special_tokens." + "reason": "No shared underlying bug: Sam3Video CUDA OOM is unrelated to Whisper empty-transcription generation failure after token alignment." }, { "left": "issue:44534", "right": "issue:44913", "accept": false, - "reason": "Same general v5 era, but different bugs: buffer initialization corruption vs GPTNeoX rotary_pct not persisting on reload." + "reason": "Different root causes: junk-filled non-persistent buffers versus GPTNeoX rotary_pct not persisting on reload." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Dataset-loading script failure is unrelated to Phi-4 multimodal model loading shape mismatch." + "reason": "Unrelated failures: local dataset loading in a training script versus loading a multimodal model with latest transformers." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "Janus test failure and auto_mappings CI breakage are separate issues with different affected files and failure modes." + "reason": "One is a model test failure, the other is a stale auto-mapping reference to removed Sam3LiteText configs breaking CI." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Different Qwen-related regressions: video input preprocessing vs /v1/responses KV-cache continuation indexing." + "reason": "Different failure modes and code paths: video-input regression in several Qwen variants versus KV-cache continuation indexing in transformers serve." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 import/parse error in DebertaV2Model is unrelated to flash_attention sink-less-model crash." + "reason": "Unrelated issues: Python 3.13 import parsing failure in DebertaV2Model versus flash_attention sink-less model AttributeError." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs still referencing pipeline() is not the same as SigLIP2 producing bad outputs through AutoModel/pipeline." + "reason": "Docs still referencing removed pipeline() calls is not the same as incorrect inference results for SigLIP2 AutoModel/pipeline usage." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Qwen3 auto device mapping cuda assert is unrelated to category-count overflow in preprocessing." + "reason": "Different symptoms and subsystems: Qwen3 auto device mapping CUDA assert versus a category-count RuntimeError." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is a model-implementation detail, not the same as OmDet-Turbo processor image-size mismatch." + "reason": "Not duplicates: a dummy Conv2d in Sam3PixelDecoder versus a processor/model input-size mismatch in OmDet-Turbo." }, { - "left": "issue:45803", - "right": "issue:45987", + "left": "issue:43931", + "right": "issue:44060", "accept": false, - "reason": "Bare except swallowing errors in FuyuBatchFeature is unrelated to StaticCache.get_seq_length returning the wrong type." + "reason": "Different regressions: Qwen3-VL weight shape mismatch on load versus an incorrect tied-weights warning involving dt_bias." }, { - "left": "issue:43931", - "right": "issue:44060", + "left": "issue:45803", + "right": "issue:45987", "accept": false, - "reason": "Model load shape mismatch for Qwen3-VL is a different bug from the false tied-weights warning in Qwen3-Next." + "reason": "One concerns exception handling in tensor conversion, the other a return-type contract bug in StaticCache.get_seq_length()." }, { "left": "issue:35141", "right": "issue:42490", "accept": false, - "reason": "Token embedding resize/reinit bug is unrelated to local-path trailing-slash import warning." + "reason": "Different areas entirely: token embedding reinitialization after resize versus a trailing-slash path import warning." }, { - "left": "issue:43644", - "right": "issue:44913", + "left": "issue:43335", + "right": "issue:44855", "accept": false, - "reason": "Same broad release window, but different concrete defects: junk-filled buffers vs rotary_pct not preserved on reload." + "reason": "Unrelated model/config bug versus a Python 3.13 parser/import problem caused by @torch.jit.script comment placement." }, { - "left": "issue:43335", - "right": "issue:44855", + "left": "issue:43644", + "right": "issue:44913", "accept": false, - "reason": "SwitchTransformers sparse-layer creation bug is unrelated to DebertaV2 Python 3.13 import parsing failure." + "reason": "Separate v5 regression classes: non-persistent buffer junk versus GPTNeoX config serialization losing rotary_pct." }, { "left": "issue:43976", "right": "issue:44855", "accept": false, - "reason": "Python version support issue is unrelated to the decorator/comment parsing error in DebertaV2Model." + "reason": "No overlap: Python-version compatibility regression versus a DebertaV2 import parsing error." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "Bamba torch.compile/SDPA failure is unrelated to local dataset loading in the training script." + "reason": "Different subsystems: SDPA/torch.compile failure in Bamba versus dataset loading in a training script." }, { "left": "issue:45632", "right": "issue:46234", "accept": false, - "reason": "Cache-path collision for trust_remote_code local models is unrelated to the missing import_utils function." + "reason": "Cache-path collision for trust_remote_code local models is unrelated to a missing import_utils availability symbol." }, { "left": "issue:44077", "right": "issue:44877", "accept": false, - "reason": "Optional post_init policy for patchtsmixer is unrelated to strict config blocking granite_speech loading." + "reason": "Different strictness bugs: patchtsmixer post_init policy versus granite_speech config loading under strict config checks." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError and serve KV-cache continuation indexing are different failure points." + "reason": "Different code paths: Qwen3OmniMoe video-input processing error versus transformers serve KV-cache continuation indexing." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json is unrelated to list_repo_files not forwarding revision." + "reason": "Not the same bug: tokenizer.json selection logic versus list_repo_files failing to forward revision." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "LoRA merge collapse with tied embeddings is unrelated to incorrect visual encoder keys in save_pretrained." - }, - { - "left": "issue:44521", - "right": "issue:45381", - "accept": false, - "reason": "All-zero assistant masks in multimodal chat templates is not the same bug as wrong vision_position_ids in qwen2.5-vl video input." - }, - { - "left": "issue:43335", - "right": "issue:44190", - "accept": false, - "reason": "Single-layer sparse-layer creation bug is unrelated to the local dataset loading script failure." - }, - { - "left": "issue:44737", - "right": "issue:45803", - "accept": false, - "reason": "CPU arange in XLNet relative position encoding is unrelated to the Fuyu bare-except error." + "reason": "Distinct issues: LoRA merging with extended vocabulary on tied embeddings versus incorrect visual encoder keys during save_pretrained." } ] }, @@ -10093,13 +10117,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 32, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36246", "issue:38175", @@ -10109,17 +10133,14 @@ "issue:43065", "issue:43116", "issue:43295", - "issue:43388", + "issue:43335", "issue:43404", "issue:43441", "issue:43479", - "issue:43526", "issue:43531", "issue:43643", "issue:43650", - "issue:43688", "issue:43854", - "issue:43901", "issue:44112", "issue:44186", "issue:44190", @@ -10127,37 +10148,42 @@ "issue:44246", "issue:44464", "issue:44496", + "issue:44521", "issue:44561", "issue:44683", "issue:44734", + "issue:44737", "issue:44933", "issue:44938", "issue:44964", "issue:45003", - "issue:45042", "issue:45216", "issue:45245", "issue:45276", - "issue:45335", "issue:45362", "issue:45375", + "issue:45381", "issue:45446", "issue:45542", "issue:45563", "issue:45584", "issue:45600", "issue:45632", + "issue:45803", "issue:46082", "issue:46121" ], "soft_pairs": [ + "issue:44521|issue:45381", + "issue:43335|issue:44190", + "issue:44737|issue:45803", "issue:44112|issue:45600", "issue:43116|issue:44190", "issue:46082|issue:46121", - "issue:45276|issue:45584", "issue:43479|issue:45375", - "issue:43295|issue:45276", + "issue:45276|issue:45584", "issue:44938|issue:45542", + "issue:43295|issue:45276", "issue:45245|issue:45584", "issue:44464|issue:45563", "issue:43650|issue:44220", @@ -10177,226 +10203,216 @@ "issue:36246|issue:38617", "issue:44561|issue:45632", "issue:42907|issue:43441", - "issue:43531|issue:45600", - "issue:43295|issue:45335", - "issue:43526|issue:43901", - "issue:44496|issue:45042", - "issue:43388|issue:43688" + "issue:43531|issue:45600" ], - "prepared_review_unit_hash": "e6afe2ae56d4c5bc150ec3ffa8756e8235f503bd1648b18a3666554efb2aa5a9", + "prepared_review_unit_hash": "812d09aafe246ff508aaaa37560f0acb15e07d014ede82964cac2d2990ec1913", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36992, - "estimated_input_tokens": 9248, - "estimated_eval_tokens": 18752 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36142, + "estimated_input_tokens": 9036, + "estimated_eval_tokens": 18328 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pair looks like the same underlying bug/change; most are only loosely related by subsystem (imports, configs, generation, or model families) and would not be merged as duplicates.", - "confidence": 0.93, - "canonical_issue_reason": "issue:43295 is the most concrete, actionable regression report in the set: it names the affected version, describes a specific broken user-facing path, and gives a clear failure mode.", + "summary": "The candidate pairs are mostly broad Transformers issues that share only subsystem-level similarity, not the same concrete bug. I would not merge any of the soft edges, and there is no clear single canonical issue in this heterogeneous set.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43295 is the strongest standalone issue for follow-up because it is a crisp regression with a specific reproduction surface and broad downstream impact.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal/Qwen-related, but one is about assistant masks in chat templating and the other about video vision position ids; different failure modes and code paths." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "Completely unrelated bugs: SwitchTransformers config creation vs an image-classification example script failing to load a local dataset." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet device placement bug and a Fuyu exception-handling bug are unrelated code paths." + }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "Both mention CI, but one is a stale GraniteSpeech device override test and the other is a broken auto-mapping reference; different failures and code paths." + "reason": "Both mention CI, but one is a stale device-override test and the other is broken auto-mapping references to removed configs." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Both involve example scripts, but one is multi-label classification returning empty results while the other cannot load a local dataset for image classification; unrelated bugs." + "reason": "Both touch example scripts, but one is a multi-label classification output bug and the other is a dataset-loading problem in a different script." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "Both are config/validation issues, but one is about LlamaConfig head_dim constraints and the other is a TypeError in rope-parameter dict conversion." - }, - { - "left": "issue:45276", - "right": "issue:45584", - "accept": false, - "reason": "Different models and failures: Gemma4 embedding resize propagation vs Whisper empty transcription after special-token alignment." + "reason": "Different config-validation bugs: head_dim rejection in LlamaConfig vs rope validation type handling." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Both concern multimodal config fields, but they target different model families and different missing/defaulted attributes." + "reason": "Both are multimodal config issues, but they involve different model families and different initialization/serialization bugs." }, { - "left": "issue:43295", - "right": "issue:45276", + "left": "issue:45276", + "right": "issue:45584", "accept": false, - "reason": "One is a processor/tokenizer regression; the other is token-embedding resize not propagating in Gemma4. Same broad area, different bug." + "reason": "Unrelated issues: embedding resize propagation in gemma4 versus Whisper generation on empty transcription." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Both are import/startup failures, but one is Python 3.14 load failure and the other is a TensorBoard/TensorFlow backend mismatch." + "reason": "Python 3.14 import failure and TensorBoard/TensorFlow backend detection are distinct problems." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "A processor/tokenizer regression is not the same as token-embedding resize not propagating in gemma4." }, { "left": "issue:45245", "right": "issue:45584", "accept": false, - "reason": "Different domains: category-count overflow in categorical handling versus Whisper transcription behavior after token alignment." + "reason": "Different bugs: category-cardinality runtime error vs Whisper empty-transcription generation failure." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Both touch generation, but one is inconsistent outputs with compiled forward while the other is a stale warning about num_return_sequences." + "reason": "Both are generation-related, but one is incorrect outputs from compiled forward and the other is a stale warning about num_return_sequences." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "No clear overlap: a placeholder 'ADD THE DATA' issue versus a bug in _torch_extract_fbank_features()." + "reason": "The first is a vague data request; the second is a specific feature-extraction bug." }, { "left": "issue:44186", "right": "issue:44734", "accept": false, - "reason": "Both crash, but one is LayoutLMv2Tokenizer on NER/padding and the other is serving KV-cache continuation with wrong tensor indexing." + "reason": "Different crash sites and models: LayoutLMv2Tokenizer padding/NER vs server KV-cache continuation indexing." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Both are ImportError reports, but they are for different missing symbols in different modules and are not the same export bug." + "reason": "Both are import errors, but for different symbols and different APIs; not the same missing-export bug." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "Different model behaviors: tied-weight issue in Mistral3 vs incorrect save_pretrained output for Qwen3.5." + "reason": "Weight-tying in Mistral3 and checkpoint saving regression in Qwen3.5 are different model-specific issues." }, { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Both are import-time concerns, but one is slow import and the other is unsafe sys.modules access; not the same underlying defect." + "reason": "Import-performance regression and unsafe sys.modules access are not the same underlying defect." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both involve flex_attention/versioning, but one is compiled flex_attention failing on torch>=2.9 and the other is a wrong version check for AuxRequest import." + "reason": "Both involve flex_attention/versioning, but one is a torch>=2.9 compiled failure and the other is an incorrect version check for AuxRequest import." }, { "left": "issue:38175", "right": "issue:43065", "accept": false, - "reason": "Unrelated: SigLIP2 zero probabilities versus a dummy Conv2d in Sam3PixelDecoder." + "reason": "Different models and unrelated symptoms: zero probabilities in SigLIP2 vs a dummy Conv2d in Sam3PixelDecoder." }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "Different problems: a non-existent import from image_utils versus a chat crash with Qwen3.5-35B." + "reason": "A missing import from image_utils and a Qwen3.5 chat crash are unrelated." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "Both relate to config loading, but one loses fields under trust_remote_code and the other lacks a model_type in config.json." + "reason": "AutoConfig missing fields under trust_remote_code is unrelated to an unrecognized-model config.json error." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both are model loading failures, but they concern different models and likely different root causes." + "reason": "Different model-loading failures for different models; same general area but not the same bug." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "One is a runtime model-load regression; the other is CI breakage from removed Sam3LiteText config references." + "reason": "A model-loading regression and broken auto-mapping references are unrelated." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "A stale GraniteSpeech CI test and a Phi-4 multimodal load failure are unrelated beyond both appearing in CI." + "reason": "CI test flakiness and a Phi-4-multimodal loading issue are different problems." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "Both are generation-related, but one is a warning regression and the other is a concrete Whisper failure on empty transcription." + "reason": "A stale generate warning and Whisper empty-transcription failure are not the same code-path issue." }, { "left": "issue:44734", "right": "issue:45632", "accept": false, - "reason": "Serving KV-cache indexing and trust_remote_code cache-path collisions are different code paths and bugs." + "reason": "KV-cache continuation indexing in the server is unrelated to trust_remote_code cache-path collisions." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Both are missing-import reports, but they involve different modules and symbols with no evidence of a shared removal." + "reason": "Different import failures for different names; not a duplicate export regression." }, { "left": "issue:44561", "right": "issue:45632", "accept": false, - "reason": "One breaks trust_remote_code models by removing is_torch_fx_available; the other is a cache path collision for local models." + "reason": "Both involve trust_remote_code, but one breaks models due to a removed helper and the other is a local cache-path collision." }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Same model family, but saving dequantized Ministrals/Devstrals and FlashAttention failures are separate concrete bugs." + "reason": "Ministral/Devstral save_pretrained behavior and FlashAttention failures are separate bugs." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping references are unrelated." - }, - { - "left": "issue:43295", - "right": "issue:45335", - "accept": false, - "reason": "Both involve model component wiring, but one is processor.tokenizer/image handling and the other is t5gemma decoder embedding updates." - }, - { - "left": "issue:43526", - "right": "issue:43901", - "accept": false, - "reason": "Different layers entirely: BeitImageProcessorFast label reduction versus TextClassificationPipeline documentation mismatch." - }, - { - "left": "issue:44496", - "right": "issue:45042", - "accept": false, - "reason": "An unrecognized model/config error is unrelated to PIL image processors requiring torchvision." - }, - { - "left": "issue:43388", - "right": "issue:43688", - "accept": false, - "reason": "Both touch training/evaluation losses, but one is metric gathering dropping labels and the other is auxiliary-loss normalization." + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup are unrelated." } ] }, @@ -10412,13 +10428,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 33, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:32090", "issue:35141", @@ -10429,18 +10445,18 @@ "issue:43295", "issue:43299", "issue:43381", + "issue:43388", "issue:43441", "issue:43479", "issue:43525", + "issue:43526", "issue:43577", "issue:43650", "issue:43673", + "issue:43688", "issue:43824", - "issue:43854", "issue:43901", - "issue:43976", "issue:44016", - "issue:44062", "issue:44075", "issue:44220", "issue:44242", @@ -10453,6 +10469,7 @@ "issue:44779", "issue:44933", "issue:45020", + "issue:45042", "issue:45200", "issue:45216", "issue:45278", @@ -10465,11 +10482,14 @@ "issue:45593", "issue:45759", "issue:45812", - "issue:45823", "issue:46209", "issue:46234" ], "soft_pairs": [ + "issue:43526|issue:43901", + "issue:43295|issue:45335", + "issue:43388|issue:43688", + "issue:44496|issue:45042", "issue:43441|issue:45588", "issue:43673|issue:44336", "issue:41950|issue:44016", @@ -10480,9 +10500,9 @@ "issue:43901|issue:44496", "issue:44016|issue:44220", "issue:44496|issue:45759", - "issue:45335|issue:45584", - "issue:45372|issue:46234", "issue:43650|issue:46209", + "issue:45372|issue:46234", + "issue:45335|issue:45584", "issue:32090|issue:35141", "issue:43479|issue:43525", "issue:44242|issue:45200", @@ -10491,224 +10511,220 @@ "issue:43299|issue:45357", "issue:44779|issue:45812", "issue:43295|issue:44779", - "issue:39692|issue:43901", "issue:43824|issue:44933", + "issue:39692|issue:43901", "issue:44279|issue:45405", "issue:44351|issue:45278", - "issue:41950|issue:45020", - "issue:44016|issue:44062", - "issue:41950|issue:45823", - "issue:43976|issue:45405", - "issue:43299|issue:43854" + "issue:41950|issue:45020" ], - "prepared_review_unit_hash": "1431513e362b1001967aea3e74d8f118a23b273cc3006c6d3c758f7efa0b7066", + "prepared_review_unit_hash": "d13d94114697c6c868db115248ff328c6768884b8b85d4c536e0e171d331c768", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36682, - "estimated_input_tokens": 9171, - "estimated_eval_tokens": 18598 + "serialized_chars": 36767, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose set of Transformers regression/loading/import issues, not a single duplicate bug. I found no soft-edge pair that clearly shares the same concrete root cause, so all candidate links are rejected.", - "confidence": 0.93, - "canonical_issue_reason": "issue:45278 is the broadest umbrella report in the set: it captures the general post-upgrade breakage pattern and can serve as a cluster anchor better than the many model-specific or narrow bug reports.", + "summary": "No soft pair looks like a true duplicate: the items span unrelated bugs across tokenizers, processors, configs, imports, training utilities, and docs. A broad version-regression import issue is the closest umbrella representative, but it does not unify the set.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like report here because it summarizes multiple import breakages after a Transformers version upgrade, making it the closest representative among otherwise unrelated issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best representative issue because it is the most general, user-facing regression summary in the group; the other issues are narrower manifestations or unrelated bugs.", + "best_issue_reason": "issue:45278 is the most suitable cluster representative because it describes a wide regression class rather than a single narrow symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "Different problems: image-processor label reduction bug vs docs still mentioning return_all_scores." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Different code paths: multimodal processor.tokenizer regression vs t5gemma embedding resize not propagating." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Different training bugs: last-batch metric gathering vs auxiliary-loss normalization." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Different bugs: model config recognition failure vs PIL image processor requiring torchvision." + }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both involve FlashAttention, but they describe different failures: a model-specific FlashAttention breakage vs an AttributeError on `s_aux=None` for sink-less models. Different code paths and fixes." + "reason": "Both mention flash attention, but they hit different failures/models and different code paths." }, { "left": "issue:43673", "right": "issue:44336", "accept": false, - "reason": "Generation cache handling in chunked prefill and ANSI color output in loading reports are unrelated bugs." + "reason": "Unrelated: generation cache/chunked prefill vs ANSI formatting in loading_report." }, { "left": "issue:41950", "right": "issue:44016", "accept": false, - "reason": "Video-classification pipeline processor lookup and a notebook syntax error are completely different issues." + "reason": "Different domains: video-classification pipeline processor lookup vs a notebook syntax error." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "SGD optimizer arguments being ignored is unrelated to a warning about `tie_word_embeddings=False`." + "reason": "Different issues: SGD optimizer args ignored vs a configuration warning about tie_word_embeddings." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "A missing TensorFlow backend dependency and a generic dependency issue are too vague and not the same concrete problem." + "reason": "Different dependency complaints: missing tensorflow in extras vs a generic transformers dependency issue." }, { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Both are Qwen regression reports, but one is about MoE model loading and the other about incorrect `save_pretrained` output. Different code paths." + "reason": "Different regressions: Qwen3VL MoE model loading vs Qwen3.5 save_pretrained checkpoint corruption." }, { "left": "issue:43381", "right": "issue:45593", "accept": false, - "reason": "Gradient checkpointing in eval mode and missing auxiliary losses when denoising is off are unrelated modeling behaviors." + "reason": "Different bugs: gradient checkpointing disallowed in eval vs D-FINE auxiliary losses omitted when denoising is off." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Docs mentioning `return_all_scores` and an unrecognized model/config loading error do not share a root cause." + "reason": "Docs-only mismatch vs a model loading/config error; not the same bug." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "A notebook syntax error and `_torch_extract_fbank_features()` behavior are unrelated." + "reason": "Notebook syntax error vs audio feature extraction bug; unrelated." }, { "left": "issue:44496", "right": "issue:45759", "accept": false, - "reason": "Both concern model/config loading, but one is an unknown model type error while the other is failure to unwrap `text_config` for composite Qwen configs. Not the same bug." + "reason": "Different model-instantiation failures: generic unrecognized model vs missing text_config unwrapping for composite Qwen multimodal configs." }, { - "left": "issue:45335", - "right": "issue:45584", + "left": "issue:43650", + "right": "issue:46209", "accept": false, - "reason": "Tokenizer embedding resize behavior for t5gemma is unrelated to Whisper failing on empty transcription." + "reason": "No clear relation: placeholder title vs FP8 inference failure on B200." }, { "left": "issue:45372", "right": "issue:46234", "accept": false, - "reason": "Both are import-related, but one is a Gemma processor dependency break and the other is a missing utility symbol; different missing imports and fixes." + "reason": "Different missing symbols/import surfaces: ReasoningEffort import error vs falcon-mamba availability helper missing." }, { - "left": "issue:43650", - "right": "issue:46209", + "left": "issue:45335", + "right": "issue:45584", "accept": false, - "reason": "Placeholder/crude issue title and FP8 inference failure on B200 are unrelated." + "reason": "Different bugs: token embedding resize propagation vs Whisper empty-transcription generation failure." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer GPU broadcast TypeError and embedding reinitialization after resize are unrelated training/model initialization bugs." + "reason": "Both involve None/tie_word_embedding themes, but one is Trainer GPU broadcast of None and the other is post_init reinitialization." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Default multimodal config initialization and missing `pad_token_id` on `Llama4Config` are different config bugs." + "reason": "Different config problems: default vision/audio init from None vs missing pad_token_id attribute." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Missing load-balancing loss and `mm_token_type_ids` defaults are different modeling issues." + "reason": "Different mechanisms: router-logit load balancing loss vs mm_token_type_ids defaulting for text-only fine-tuning." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Precision casting on BLIP2 loading and incorrect Qwen3.5 checkpoint saving are unrelated." + "reason": "Different regressions: dtype not applied to Blip2 submodules vs incorrect Qwen3.5 save_pretrained output." }, { "left": "issue:38175", "right": "issue:44617", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and CUDA OOM in Sam3Video are clearly unrelated." + "reason": "Different symptoms and subsystems: unexpected zero probabilities vs CUDA out-of-memory." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Both are Qwen3.5-related but cover different save/load regressions, so they are not the same underlying bug." + "reason": "Same model family but different concrete bugs: loading failure vs save_pretrained writing wrong visual-encoder keys." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Tokenizer regression in DeepSeek and tokenizer regression in Granite are different model-specific failures." + "reason": "Both are tokenizer regressions, but they affect different model families and likely different fixes." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Processor/tokenizer access breakage in custom multimodal code is unrelated to DeepSeek tokenizer output corruption." + "reason": "Processor/tokenizer-image handling regression vs DeepSeek tokenizer output bug; not the same issue." }, { - "left": "issue:39692", - "right": "issue:43901", + "left": "issue:43824", + "right": "issue:44933", "accept": false, - "reason": "Documentation example errors and pipeline docs mentioning `return_all_scores` are not the same issue." + "reason": "Different import failures: missing Qwen2.5-VL class vs nonexistent image_utils import." }, { - "left": "issue:43824", - "right": "issue:44933", + "left": "issue:39692", + "right": "issue:43901", "accept": false, - "reason": "Missing model export and missing image_utils import are both import problems but not the same missing symbol or code path." + "reason": "Different docs issues: SigLIP2 example errors vs TextClassificationPipeline return_all_scores documentation drift." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "A general dependency issue and a specific unreleased PEFT version bump are different problems." + "reason": "Generic dependency problem vs a specific unreleased minimum PEFT version bump." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "HybridCache import failure may be one item under the broad import-breakage umbrella, but the titles do not show the same concrete root cause; too broad to merge." + "reason": "Related by imports, but 45278 is a broad report while this is a specific HybridCache export failure; not clearly one bug." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Video-classification processor lookup and remote-code model breakage are unrelated regression classes." - }, - { - "left": "issue:44016", - "right": "issue:44062", - "accept": false, - "reason": "Notebook syntax error and AddedToken keyword conflict are unrelated." - }, - { - "left": "issue:41950", - "right": "issue:45823", - "accept": false, - "reason": "Pipeline processor lookup and Gemma4 PLE device mismatch are unrelated." - }, - { - "left": "issue:43976", - "right": "issue:45405", - "accept": false, - "reason": "Python version compatibility and PEFT version pinning are different dependency issues." - }, - { - "left": "issue:43299", - "right": "issue:43854", - "accept": false, - "reason": "Qwen3VL MoE loading and GLM-4.7-Flash test loading failures are model-specific and distinct." + "reason": "Different problems: video-classification pipeline processor lookup vs remote_code model breakage." } ] }, @@ -10724,43 +10740,42 @@ "component_index": 1, "component_count": 18, "review_unit_index": 34, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ + "issue:41950", "issue:42994", "issue:43065", + "issue:43299", "issue:43317", "issue:43441", "issue:43450", "issue:43525", - "issue:43643", "issue:43650", "issue:43653", "issue:43723", - "issue:43784", + "issue:43854", "issue:43881", "issue:43901", "issue:43976", "issue:43994", "issue:44016", + "issue:44062", "issue:44162", - "issue:44188", "issue:44246", "issue:44261", "issue:44263", "issue:44279", "issue:44297", - "issue:44336", "issue:44361", "issue:44462", "issue:44466", "issue:44521", - "issue:44617", "issue:44623", "issue:44671", "issue:44683", @@ -10772,6 +10787,7 @@ "issue:45276", "issue:45325", "issue:45397", + "issue:45405", "issue:45542", "issue:45593", "issue:45600", @@ -10782,19 +10798,23 @@ "issue:46132" ], "soft_pairs": [ + "issue:44016|issue:44062", + "issue:41950|issue:45823", + "issue:43976|issue:45405", + "issue:43299|issue:43854", "issue:44683|issue:45542", "issue:44462|issue:45081", "issue:43650|issue:46132", "issue:43976|issue:44987", "issue:44297|issue:44462", "issue:43976|issue:44279", - "issue:43450|issue:44279", "issue:45081|issue:45907", - "issue:44466|issue:45276", + "issue:43450|issue:44279", "issue:44521|issue:45325", + "issue:44466|issue:45276", "issue:43441|issue:45834", - "issue:43653|issue:45593", "issue:42994|issue:43317", + "issue:43653|issue:45593", "issue:45600|issue:45712", "issue:44623|issue:44749", "issue:44623|issue:45397", @@ -10806,228 +10826,224 @@ "issue:43441|issue:44683", "issue:43723|issue:44162", "issue:43901|issue:43994", - "issue:43901|issue:44246", "issue:44297|issue:45397", "issue:43525|issue:43881", - "issue:43450|issue:44617", - "issue:43525|issue:43643", - "issue:43784|issue:45542", - "issue:44188|issue:44336" + "issue:43901|issue:44246" ], - "prepared_review_unit_hash": "a262405d73d14b8f23d2a3356bade94282bc637f5e34e3d5428780789354c2b0", + "prepared_review_unit_hash": "8ded950a8648a019489007aedacc52603b58cd7de229e0c5ab9be8d2b6d1ec0f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36789, - "estimated_input_tokens": 9198, - "estimated_eval_tokens": 18652 + "serialized_chars": 36740, + "estimated_input_tokens": 9185, + "estimated_eval_tokens": 18626 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: the soft pairs share superficial theme words (tokenizers, loading, v5, model bugs) but do not describe the same underlying defect, so none should be merged as duplicates.", - "confidence": 0.88, - "canonical_issue_reason": "No single issue clearly represents a duplicate cluster here; if one must be chosen, issue 44462 is the clearest, most specific regression report, but it is not a true duplicate of the others.", + "summary": "This cluster is heterogeneous: the issues span unrelated tokenizer, vision, loading, docs, and performance bugs. I found no soft pair that looks like the same underlying change or defect, so all soft edges are rejected.", + "confidence": 0.36, + "canonical_issue_reason": "Issue 41950 is the strongest standalone report in the set: it has the highest discussion activity and inbound references, and is a clear, concrete bug. It is only a representative choice, not evidence that the cluster is a true duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44462 is the strongest standalone bug report in the set: it has a precise symptom, narrow scope, and an obvious code path (AutoTokenizer loading repository tokenizer.json).", + "best_issue_reason": "Issue 41950 is the best overall issue to represent this group because it is active, well-specified, and externally referenced, even though the cluster itself is not cohesive.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken keyword conflict are unrelated failures in different subsystems." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Video-classification processor lookup and Gemma4 device-mapping mismatch are different bugs." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version compatibility and a PEFT version bump are unrelated dependency problems." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different model-loading regressions affecting different models and likely different code paths." + }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Unrelated failures: flex_attention compilation on new torch vs missing TensorFlow backend when only tensorboard is installed." + "reason": "FlexAttention on torch 2.9 and missing TensorFlow backend for tensorboard are unrelated." }, { "left": "issue:44462", "right": "issue:45081", "accept": false, - "reason": "Both involve tokenizer loading, but one is AutoTokenizer ignoring tokenizer.json and the other is a Mistral regex patch crash on a tokenizer object; different code paths." + "reason": "Both are tokenizer-loading issues, but they involve different mechanisms and different failure modes." }, { "left": "issue:43650", "right": "issue:46132", "accept": false, - "reason": "No shared bug: a vague data request versus a behavioral change in AttentionInterface.register." + "reason": "Generic placeholder request and AttentionInterface behavior change are not the same bug." }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Different problems: Python version compatibility for Transformers 5.1.0 versus loading a specific model package." + "reason": "Python compatibility issue and a model-loading failure for a specific repo are unrelated." }, { "left": "issue:44297", "right": "issue:44462", "accept": false, - "reason": "Tokenizer serialization mismatch is a save_pretrained bug, while 44462 is about loading tokenizer.json; not the same defect." + "reason": "Tokenizer save metadata mismatch is distinct from AutoTokenizer ignoring tokenizer.json on load." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "Python compatibility regression and a generic dependency issue are not the same underlying bug." - }, - { - "left": "issue:43450", - "right": "issue:44279", - "accept": false, - "reason": "Video processor batch-shape bug is unrelated to an unspecified dependency issue." + "reason": "Dependency/version support issue and a generic dependency issue are not the same concrete defect." }, { "left": "issue:45081", "right": "issue:45907", "accept": false, - "reason": "Mistral regex patching crash and list_repo_files not forwarding revision are separate loading vs hub API bugs." + "reason": "Mistral tokenizer regex patch crash and list_repo_files revision forwarding are unrelated." }, { - "left": "issue:44466", - "right": "issue:45276", + "left": "issue:43450", + "right": "issue:44279", "accept": false, - "reason": "Both are Gemma4-related but the code paths differ: lm_head.weight serialization versus resize_token_embeddings not updating embeddings." + "reason": "Video processor batch-shape bug and a generic dependency issue do not match." }, { "left": "issue:44521", "right": "issue:45325", "accept": false, - "reason": "Different multimodal bugs: assistant mask generation versus Qwen2.5-VL rope index scaling." + "reason": "Assistant mask generation and temporal rope indexing are different multimodal bugs." }, { - "left": "issue:43441", - "right": "issue:45834", + "left": "issue:44466", + "right": "issue:45276", "accept": false, - "reason": "FlashAttention failure in Ministral-3 is unrelated to Kosmos2.5 OCR long-input indexing." + "reason": "Serialization of lm_head.weight and resize_token_embeddings propagation are different issues." }, { - "left": "issue:43653", - "right": "issue:45593", + "left": "issue:43441", + "right": "issue:45834", "accept": false, - "reason": "Tokenizer special-token registration bug and D-FINE auxiliary-loss behavior are unrelated." + "reason": "FlashAttention failure for Ministral-3 and OCR index error for Kosmos2.5 are unrelated model bugs." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Both concern quantization, but one is saving quantized models and the other is loading dequantized models with device_map/offload; not the same bug." + "reason": "Saving quantized models and loading dequantized models with offload are different code paths and failures." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer special-token registration and D-FINE auxiliary losses are unrelated." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Both are repo maintenance issues around missing symbols, but one is auto_mappings referencing removed configs and the other is dummy classes leaking into dir/check_repo." + "reason": "Removed config references in auto_mappings and leftover dummy classes are separate CI cleanups." }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "Processor save_pretrained missing files is unrelated to a performance regression after upgrading Transformers." + "reason": "Missing processor save files and tokenizer/data filtering slowdown are unrelated." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Different failures: missing processor save files versus Gemma-4 zero3 from_pretrained." + "reason": "Processor save_pretrained file omission and gemma-4 zero3 loading are unrelated." }, { "left": "issue:43065", "right": "issue:44898", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is a model-definition placeholder issue; Perceiver non-default resolution failure is a separate interpolation bug." + "reason": "Dummy Conv2d in Sam3PixelDecoder and Perceiver resize/interpolation failure are different vision bugs." }, { "left": "issue:44261", "right": "issue:44263", "accept": false, - "reason": "Different model internals: missing rms_norm_eps in MLA q_a_layernorm versus torch.split return-value handling in GlmMoeDsaIndexer." + "reason": "LayerNorm epsilon config issue and torch.split return-value handling are unrelated." }, { "left": "issue:44361", "right": "issue:45593", "accept": false, - "reason": "MLukeTokenizer AttributeError and D-FINE auxiliary-loss behavior are unrelated." + "reason": "MLukeTokenizer AttributeError and D-FINE denoising-loss behavior are unrelated." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Both are v5 regressions affecting text models, but CamemBERT masked-LM predictions and Deepseek tokenizer outputs are different code paths and symptoms." + "reason": "CamemBERT masked-LM regression and Deepseek tokenizer regression are different model-specific bugs." }, { "left": "issue:44016", "right": "issue:45823", "accept": false, - "reason": "Notebook syntax error is unrelated to Gemma4 PLE device-mismatch during forward." + "reason": "Notebook syntax error and Gemma4 device mismatch are unrelated." }, { "left": "issue:43441", "right": "issue:44683", "accept": false, - "reason": "FlashAttention model-loading failure and compiled flex_attention torch>=2.9 failure are separate attention-related bugs." + "reason": "Both mention attention/flash attention, but they are distinct failures on different operators and versions." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer loading issue in v5 and ESM2 model breakage are not the same underlying defect." + "reason": "Tokenizer loading regression and ESM2 model breakage are not the same underlying defect." }, { "left": "issue:43901", "right": "issue:43994", "accept": false, - "reason": "Docs mismatch for return_all_scores versus incorrect SigLIP2 pipeline outputs are different issues." - }, - { - "left": "issue:43901", - "right": "issue:44246", - "accept": false, - "reason": "Pipeline docs confusion is unrelated to import-time slowness." + "reason": "Docs mismatch for pipeline output and incorrect model inference results are different issues." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer class serialization mismatch and Gemma-4 zero3 loading failure do not share a code path." + "reason": "Tokenizer serialization metadata and gemma-4 zero3 loading are unrelated." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Missing pad_token_id on Llama4Config is distinct from glm-4v-9b loading failure." - }, - { - "left": "issue:43450", - "right": "issue:44617", - "accept": false, - "reason": "Batched video processor shape bug and Sam3Video CUDA OOM are different problems." - }, - { - "left": "issue:43525", - "right": "issue:43643", - "accept": false, - "reason": "A missing config attribute on Llama4Config is not the same as trust_remote_code returning incomplete config fields." - }, - { - "left": "issue:43784", - "right": "issue:45542", - "accept": false, - "reason": "Import-time NameError in sentence-transformers integration and TensorFlow backend selection error are unrelated." + "reason": "Missing pad_token_id on Llama4Config and glm-4v-9b loading failure are different model/config bugs." }, { - "left": "issue:44188", - "right": "issue:44336", + "left": "issue:43901", + "right": "issue:44246", "accept": false, - "reason": "Attention kernel divergence under torch.compile is unrelated to ANSI codes emitted by loading_report." + "reason": "Pipeline docs mismatch and import-time slowdown are unrelated." } ] }, @@ -11043,13 +11059,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 35, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:16998", "issue:22355", @@ -11062,14 +11078,16 @@ "issue:43262", "issue:43299", "issue:43352", - "issue:43526", + "issue:43450", + "issue:43525", + "issue:43643", "issue:43673", "issue:43761", + "issue:43784", "issue:43824", - "issue:43901", "issue:43976", "issue:44016", - "issue:44162", + "issue:44188", "issue:44220", "issue:44242", "issue:44246", @@ -11083,24 +11101,26 @@ "issue:44521", "issue:44561", "issue:44589", + "issue:44617", "issue:44749", "issue:44779", "issue:44857", "issue:44871", "issue:44964", - "issue:45003", "issue:45081", - "issue:45200", "issue:45250", "issue:45292", "issue:45362", "issue:45397", - "issue:45412", + "issue:45542", "issue:45593", - "issue:46123", "issue:46132" ], "soft_pairs": [ + "issue:43450|issue:44617", + "issue:43525|issue:43643", + "issue:44188|issue:44336", + "issue:43784|issue:45542", "issue:44016|issue:44291", "issue:43262|issue:44521", "issue:43673|issue:43761", @@ -11112,234 +11132,223 @@ "issue:44246|issue:44496", "issue:44295|issue:45397", "issue:44242|issue:45593", - "issue:44246|issue:44871", "issue:44336|issue:44485", + "issue:44246|issue:44871", "issue:43824|issue:43976", - "issue:43976|issue:44964", "issue:41950|issue:44220", + "issue:43976|issue:44964", "issue:36246|issue:42757", - "issue:36246|issue:44351", "issue:42886|issue:46132", + "issue:36246|issue:44351", "issue:44779|issue:45292", "issue:44242|issue:44857", - "issue:45081|issue:45362", "issue:43673|issue:44561", + "issue:45081|issue:45362", "issue:44295|issue:44749", - "issue:43352|issue:45250", - "issue:44162|issue:45200", - "issue:43901|issue:45003", - "issue:43526|issue:45003", - "issue:44496|issue:45003", - "issue:45412|issue:46123" + "issue:43352|issue:45250" ], - "prepared_review_unit_hash": "3abe23da715bce8c7d24ed7e708f49b5f33c1671453c45029f20b5a03f2f56d2", + "prepared_review_unit_hash": "c1dd519f264f0534cb8b1cb40e0352f0b2e3a0eaf0103e65154765f2ce5faad3", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36636, - "estimated_input_tokens": 9159, - "estimated_eval_tokens": 18574 + "soft_pair_count": 29, + "serialized_chars": 36333, + "estimated_input_tokens": 9084, + "estimated_eval_tokens": 18424 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a set of unrelated, model-specific or environment-specific transformer bugs. None of the soft pairs look like clear same-bug duplicates under a strict triage standard, so I rejected all of them.", - "confidence": 0.89, - "canonical_issue_reason": "Issue 45003 is the best canonical representative: it is a concrete core-runtime bug in modeling_utils with a clear code path, rather than a vague environment or model-specific symptom.", + "summary": "This cluster is mostly a loose topical grouping of unrelated Transformers issues. The pairs share broad themes like imports, multimodal chat, or Qwen/MoE, but they do not describe the same concrete bug or mergeable fix.", + "confidence": 0.82, + "canonical_issue_reason": "issue:44521 is the strongest representative: it is a concrete, well-scoped multimodal `apply_chat_template` bug with active discussion and inbound references, making it the best anchor among these heterogeneous reports.", "canonical_pr_reason": null, - "best_issue_reason": "45003 is the strongest standalone issue for routing/triage because it describes a specific unsafe core access pattern and is easier to act on than the more ambiguous or narrowly model-bound reports.", + "best_issue_reason": "issue:44521 is the clearest and most actionable issue in the set; it has the most direct bug description and the strongest evidence of user impact.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43450", + "right": "issue:44617", + "accept": false, + "reason": "Different failures: batched video processor shape mismatch vs video-model OOM. Same broad area, not the same bug." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "One is a missing `pad_token_id` on `Llama4Config`; the other is `trust_remote_code` losing fields in `AutoConfig`. Related only at a high level." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI codes in `loading_report`." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both mention import-time breakage, but one is `nn` missing for sentence-transformers and the other is TensorFlow backend handling with only tensorboard installed." + }, { "left": "issue:44016", "right": "issue:44291", "accept": false, - "reason": "One is a notebook syntax error; the other is a TypeError in init_empty_weights. Different failure modes and code paths." + "reason": "A notebook syntax error and an `init_empty_weights` TypeError are different problems with no shared fix path." }, { "left": "issue:43262", "right": "issue:44521", "accept": false, - "reason": "Both mention chat templating/multimodal processing, but one is about a default sampling rate and the other about assistant masks. Distinct bugs." + "reason": "Both touch `apply_chat_template`, but one is wrong audio sampling-rate default and the other is all-zero assistant masks; different code paths and symptoms." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Chunked prefill cache handling and CLIPVisionModel hidden_states are unrelated model/runtime issues." + "reason": "Chunked-prefill cache handling and CLIP vision hidden-states propagation are unrelated." }, { "left": "issue:41628", "right": "issue:43976", "accept": false, - "reason": "Importing AutoImageProcessor and Python-version compatibility are different problems." + "reason": "Missing `AutoImageProcessor` export and a Python-version compatibility issue are not the same bug." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage lookup and Qwen3.5 chat crashes are unrelated regressions." + "reason": "Float8 storage deserialization failure is unrelated to Qwen3.5 chat crashes." }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "ANSI output in loading_report and a wrong docstring for position_ids are not the same bug." + "reason": "Documentation/ANSI output issue versus a wrong docstring for `position_ids`; not the same underlying change." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "A model_max_length question for DeBERTa-V3 and a missing transformers.onnx module are unrelated." + "reason": "`model_max_length` on DeBERTa-V3 and missing `transformers.onnx` are unrelated." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both involve Qwen3/MoE regressions, but one is LoRA compatibility across hf4.x/hf5.x and the other is model loading in Qwen3VL-MoE. Not the same concrete issue." + "reason": "Both involve Qwen3 MoE families, but one is LoRA compatibility across hf4.x/hf5.x and the other is a Qwen3VL MoE loading regression; too different to merge." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Slow import sometimes and an unrecognized model/config error are different root causes." + "reason": "Import slowness and an unrecognized model/config error are unrelated." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "position_ids buffer access and Gemma-4 zero3 from_pretrained are unrelated." + "reason": "Buffer registration/`position_ids` access and gemma-4 ZeRO-3 loading are different issues." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Both involve missing auxiliary losses, but they affect different models and different flags/code paths." - }, - { - "left": "issue:44246", - "right": "issue:44871", - "accept": false, - "reason": "Import slowness and Gemma eos_token_id mismatch are unrelated." + "reason": "Both concern losses omitted under flags, but for different models and different toggles (`output_router_logits` vs denoising)." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "Loading-report ANSI formatting and GLM-5 RoPE implementation are different areas with no shared bug." + "reason": "ANSI code output and GLM-5 RoPE implementation are unrelated." }, { - "left": "issue:43824", - "right": "issue:43976", + "left": "issue:44246", + "right": "issue:44871", "accept": false, - "reason": "A missing export for Qwen2.5-VL generation and a Python-version compatibility problem are unrelated." + "reason": "Import-time performance and Gemma EOS token config mismatch are not the same bug." }, { - "left": "issue:43976", - "right": "issue:44964", + "left": "issue:43824", + "right": "issue:43976", "accept": false, - "reason": "A runtime support matrix issue and a specific multimodal model loading failure are not the same underlying bug." + "reason": "A missing Qwen2.5 VL export and a Python-version compatibility break are unrelated." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification processor lookup and _torch_extract_fbank_features() are different pipeline/model bugs." + "reason": "Video-classification pipeline image-processor lookup and `_torch_extract_fbank_features()` are separate subsystems." }, { - "left": "issue:36246", - "right": "issue:42757", + "left": "issue:43976", + "right": "issue:44964", "accept": false, - "reason": "Both are import errors, but one is a missing transformers symbol and the other is a missing huggingface_hub symbol. Different packages and fixes." + "reason": "Python 3.9+ compatibility and Phi-4 multimodal model loading are not the same failure." }, { "left": "issue:36246", - "right": "issue:44351", + "right": "issue:42757", "accept": false, - "reason": "Missing Qwen2.5-VLImageProcessor and missing HybridCache exports are separate module export problems." + "reason": "Missing Qwen2.5 VL symbol and missing `is_offline_mode` from huggingface_hub are different import errors." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline cache loading for tokenizers and AttentionInterface.register behavior changes are unrelated." + "reason": "Offline tokenizer cache loading and `AttentionInterface.register` behavior changes are unrelated." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing exports and different code paths." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "Deepseek tokenizer correctness and resize_token_embeddings not propagating to output embeddings are different bugs." + "reason": "Deepseek tokenizer regression and `resize_token_embeddings` not updating output embeddings are different model behavior bugs." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Load-balancing loss omission and an AMP/CUDA crash in LwDetrImageLoss are distinct issues." + "reason": "Different model-specific loss/AMP failures; no common concrete fix." }, { - "left": "issue:45081", - "right": "issue:45362", + "left": "issue:43673", + "right": "issue:44561", "accept": false, - "reason": "A Mistral tokenizer regex patch crash and a Qwen3.5 chat crash are unrelated." + "reason": "Cache behavior in chunked prefill and removal of `is_torch_fx_available` affecting trust-remote-code models are unrelated." }, { - "left": "issue:43673", - "right": "issue:44561", + "left": "issue:45081", + "right": "issue:45362", "accept": false, - "reason": "Chunked prefill cache handling and removal of is_torch_fx_available breaking trust_remote_code are different regressions." + "reason": "Mistral tokenizer regex patch crash and Qwen3.5 chat crash are different model/tokenizer paths." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "An error reading position_ids after buffer registration and slow filtering after upgrade are unrelated." + "reason": "`position_ids` buffer access and Chinese-reported performance regression after upgrade are unrelated." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "The Nemotron issue is explicit lack of Flash Attention 2 support; the other is only a vague Flash Attention 2 title and does not match the same concrete bug." - }, - { - "left": "issue:44162", - "right": "issue:45200", - "accept": false, - "reason": "ESM2 breakage and Gemma-4 mm_token_type_ids defaulting are unrelated." - }, - { - "left": "issue:43901", - "right": "issue:45003", - "accept": false, - "reason": "A docs mismatch for return_all_scores and unsafe sys.modules access are not the same issue." - }, - { - "left": "issue:43526", - "right": "issue:45003", - "accept": false, - "reason": "BeitImageProcessorFast reduce_labels behavior and modeling_utils sys.modules access are unrelated." - }, - { - "left": "issue:44496", - "right": "issue:45003", - "accept": false, - "reason": "Model config recognition failure and unsafe sys.modules access are different bugs." - }, - { - "left": "issue:45412", - "right": "issue:46123", - "accept": false, - "reason": "Memory not released after RT-DETR deletion and MaskGenerationPipeline dropping final partial batch are unrelated." + "reason": "Specific Nemotron FA2 support error versus a generic Flash Attention 2.0 topic; not the same concrete issue." } ] }, @@ -11355,29 +11364,27 @@ "component_index": 1, "component_count": 18, "review_unit_index": 36, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36246", - "issue:42886", "issue:43065", "issue:43232", "issue:43257", "issue:43262", "issue:43408", "issue:43475", + "issue:43526", "issue:43540", "issue:43596", "issue:43653", - "issue:43673", + "issue:43901", "issue:43931", - "issue:44060", - "issue:44077", "issue:44162", "issue:44246", "issue:44336", @@ -11385,7 +11392,7 @@ "issue:44466", "issue:44479", "issue:44483", - "issue:44734", + "issue:44496", "issue:44877", "issue:44977", "issue:44987", @@ -11393,11 +11400,11 @@ "issue:45003", "issue:45137", "issue:45200", - "issue:45229", "issue:45292", "issue:45335", "issue:45375", "issue:45397", + "issue:45412", "issue:45459", "issue:45561", "issue:45563", @@ -11409,15 +11416,21 @@ "issue:45759", "issue:45987", "issue:46082", + "issue:46123", "issue:46170", "issue:46202" ], "soft_pairs": [ - "issue:45292|issue:45584", + "issue:44162|issue:45200", + "issue:43901|issue:45003", + "issue:43526|issue:45003", + "issue:44496|issue:45003", + "issue:45412|issue:46123", "issue:43262|issue:44246", + "issue:45292|issue:45584", "issue:43257|issue:45137", - "issue:45292|issue:45701", "issue:45375|issue:45759", + "issue:45292|issue:45701", "issue:36246|issue:43596", "issue:44877|issue:45375", "issue:45561|issue:45563", @@ -11437,221 +11450,216 @@ "issue:44368|issue:44977", "issue:44368|issue:45563", "issue:44877|issue:46170", - "issue:43931|issue:44368", - "issue:42886|issue:45229", - "issue:44060|issue:44077", - "issue:43673|issue:44734", - "issue:45003|issue:45561", - "issue:45584|issue:46170" + "issue:43931|issue:44368" ], - "prepared_review_unit_hash": "3da1118b4e9a2b5dc7d9e20afaad4b811bd403fd50d9301a6d871b99116a88fd", + "prepared_review_unit_hash": "af200d3d10a89cf949aeb43f59bc30b82585f5266f323c95cf638beacdda8f1f", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36799, - "estimated_input_tokens": 9200, - "estimated_eval_tokens": 18656 + "serialized_chars": 36323, + "estimated_input_tokens": 9081, + "estimated_eval_tokens": 18418 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: most items are distinct bugs spanning tokenizer loading, generation, config strictness, DeepSpeed, SAM3, Qwen, Gemma, and unrelated utilities. I don\u2019t see any soft pair that meets the bar for a true duplicate or mergeable same-fix match.", - "confidence": 0.22, - "canonical_issue_reason": "issue:44479 is the broadest, most user-impacting regression report in the set and has concrete reproduction scope across several Qwen model variants, making it the best single anchor even though the cluster is not cohesive.", + "summary": "The cluster is mostly a set of unrelated bug reports that only share broad subsystem terms (loading, generation, config, deepspeed). I would not deduplicate any of the soft pairs.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45003 is the strongest issue candidate to represent the cluster only because it is the broadest infrastructure-level bug here and appears in multiple soft-similarity links; however, it still does not match the other issues as a duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best standalone representative because it describes a concrete, multi-model regression with clear user impact; the other issues are narrower or more specialized.", + "best_issue_reason": "issue:45003 is the best single issue to keep as the cluster anchor if one must be chosen, since it is a concrete modeling_utils bug with broader applicability than the more model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45292", - "right": "issue:45584", + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 corruption/broken workflow is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mentioning return_all_scores do not match a runtime sys.modules access bug." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to modeling_utils module lookup." + }, + { + "left": "issue:44496", + "right": "issue:45003", "accept": false, - "reason": "Different failures: token embedding resizing vs Whisper empty-transcription generation after special-token alignment." + "reason": "Unrecognized-model config loading and unsafe sys.modules access are different failure modes." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "RT-DETR memory cleanup and MaskGenerationPipeline final-batch handling are unrelated." }, { "left": "issue:43262", "right": "issue:44246", "accept": false, - "reason": "Audio chat-template sampling-rate bug is unrelated to intermittent slow import behavior." + "reason": "Audio chat_template sampling-rate default bug is unrelated to slow import timing." }, { - "left": "issue:43257", - "right": "issue:45137", + "left": "issue:45292", + "right": "issue:45584", "accept": false, - "reason": "Both mention DeepSpeed, but one is Qwen3 MoE weight conversion and the other is a ZeRO3 deque/index error; different code paths." + "reason": "resize_token_embeddings/output_embeddings mismatch is not the same as Whisper empty transcription after token alignment." }, { - "left": "issue:45292", - "right": "issue:45701", + "left": "issue:43257", + "right": "issue:45137", "accept": false, - "reason": "Resize-token-embeddings bug and tokenization-change report are not the same underlying issue." + "reason": "Both involve DeepSpeed, but one is Qwen3 MoE weight conversion and the other is a zero3 deque/index error." }, { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Both touch Qwen multimodal config handling, but one is a missing strict-field drop and the other is AutoModelForCausalLM config unwrapping." + "reason": "Both are Qwen multimodal config issues, but one drops a missing field and the other fails to unwrap text_config; not the same bug." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Embedding resize behavior is unrelated to tokenization changes across versions." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "ImportError for a missing processor class is unrelated to a ZeRO3 BertModel index error." + "reason": "Qwen2_5_VL import error and DeepSpeed BertModel index error are unrelated." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Both involve config loading/strictness, but they affect different models and different failure modes." + "reason": "Granite speech strict-config loading and Qwen config field dropping are different models and code paths." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "xdist file-race in testing utils is unrelated to a stale generate warning." + "reason": "A pytest-xdist file race is unrelated to a stale generation warning." }, { "left": "issue:43232", "right": "issue:44336", "accept": false, - "reason": "Generation kwargs/cache bug and ANSI-color output bug are unrelated." + "reason": "Generation kwargs update after sync_gpus has nothing to do with ANSI output in loading_report." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Model load failure and protobuf-error masking are different tokenizer/loading paths." + "reason": "Model-loading failure and tokenizer error masking are separate issues in different code paths." }, { "left": "issue:43065", "right": "issue:43408", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d in pixel decoder and the other is a model-type warning for tracker instantiation." + "reason": "SAM3 pixel decoder dummy conv and SAM3 tracker model-type warning are not the same underlying defect." }, { "left": "issue:45987", "right": "issue:46082", "accept": false, - "reason": "Cache length return-type contract issue is unrelated to LlamaConfig head_dim validation." + "reason": "StaticCache seq-length return type and LlamaConfig head_dim validation are unrelated." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "Stale indexer cache crash in GlmMoeDsa is not the same as Gemma 4 mm_token_type_ids defaulting." + "reason": "GlmMoeDsa stale cache crash is unrelated to Gemma 4 token-type defaults." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Both discuss embedding/weight behavior, but one is tied-weight serialization and the other is resize_token_embeddings not updating decoder embeddings." + "reason": "Tied-weight serialization inconsistency is unrelated to t5gemma resize_token_embeddings propagation." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "SAM3 vision encoder output attribute error and Qwen3Omni video-input ValueError are separate multimodal bugs." + "reason": "SAM3 vision encoder output attribute loss and Qwen3OmniMoe video-input validation are different bugs." }, { "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Qwen video-input regression is unrelated to the Windows/ROCm torch.distributed import issue." + "reason": "Qwen video-input regression and Windows ROCm torch.distributed import failure are unrelated." }, { "left": "issue:45397", "right": "issue:45656", "accept": false, - "reason": "Gemma-4 ZeRO3 loading and duplicate optimizer.step under DeepSpeed are different training/loading problems." + "reason": "Gemma-4 zero3 loading and duplicate optimizer.step calls under deepspeed are different problems." }, { "left": "issue:44483", "right": "issue:46202", "accept": false, - "reason": "API request rejection in v5.3 and Gemma pipeline roundtrip failure on \u2581 are unrelated." + "reason": "chat/completions request rejection is unrelated to Gemma pipeline roundtrip token normalization." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "ZeRO3 deque error and stale auto_mappings reference to removed Sam3LiteText configs are different issues." + "reason": "DeepSpeed deque failure and stale auto_mappings references to removed SAM3 Lite configs do not share a code path." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "Tokenizer special-token registration bug is unrelated to the broader ESM2 breakage report." + "reason": "BigBirdTokenizer special-token registration bug is unrelated to the ESM2 breakage report." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "GlmMoeDsa second-pass cache crash and flash_attention s_aux=None AttributeError are different code paths." + "reason": "GlmMoeDsa stale cache crash and flash_attention s_aux=None AttributeError are different failures." }, { "left": "issue:44368", "right": "issue:44977", "accept": false, - "reason": "A tied-weights warning in Qwen3.5 LoRA fine-tuning is not the same as a flash-attention generation regression." + "reason": "A tied-word-embeddings warning is not the same as a flash-attention generation regression." }, { "left": "issue:44368", "right": "issue:45563", "accept": false, - "reason": "Tied-weights warning and stale paged-generate warning are unrelated generation issues." + "reason": "Stale num_return_sequences warning and tie_word_embeddings warning are unrelated." }, { "left": "issue:44877", "right": "issue:46170", "accept": false, - "reason": "Strict config loading for granite_speech is unrelated to the Windows ROCm torch.distributed import/type error." + "reason": "Strict granite_speech config loading and Windows ROCm dependency issues are unrelated." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Weight-shape mismatch when loading Qwen3-VL is unrelated to tie_word_embeddings warnings in Qwen3.5." - }, - { - "left": "issue:42886", - "right": "issue:45229", - "accept": false, - "reason": "Offline tokenizer cache loading failure is unrelated to Gemma 4 multi-GPU CUDA OOM." - }, - { - "left": "issue:44060", - "right": "issue:44077", - "accept": false, - "reason": "Incorrect tied-weights warning in Qwen3-Next and patchtsmixer post_init allowance are not the same bug." - }, - { - "left": "issue:43673", - "right": "issue:44734", - "accept": false, - "reason": "Chunked-prefill cache missing in GenerationMixin and /v1/responses KV-cache continuation indexing are different generation issues." - }, - { - "left": "issue:45003", - "right": "issue:45561", - "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils is unrelated to the xdist captured_info.txt race." - }, - { - "left": "issue:45584", - "right": "issue:46170", - "accept": false, - "reason": "Whisper empty-transcription failure after token alignment is unrelated to the Windows ROCm import/type dependency issue." + "reason": "Qwen3-VL weight-shape mismatch is unrelated to a tie_word_embeddings warning." } ] }, @@ -11667,19 +11675,19 @@ "component_index": 1, "component_count": 18, "review_unit_index": 37, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28282", "issue:30990", "issue:41628", "issue:42371", - "issue:42673", + "issue:42886", "issue:42994", "issue:43232", "issue:43475", @@ -11689,298 +11697,280 @@ "issue:43646", "issue:43673", "issue:43701", - "issue:43704", - "issue:43824", "issue:43827", "issue:44038", + "issue:44060", + "issue:44077", "issue:44162", "issue:44188", - "issue:44222", "issue:44297", - "issue:44351", "issue:44361", "issue:44623", "issue:44661", + "issue:44734", "issue:44749", "issue:44869", "issue:44871", "issue:44933", - "issue:44964", - "issue:45081", + "issue:45003", "issue:45092", - "issue:45103", - "issue:45137", + "issue:45229", "issue:45230", "issue:45310", "issue:45356", "issue:45375", "issue:45381", "issue:45397", - "issue:45406", "issue:45478", + "issue:45561", + "issue:45584", "issue:45656", "issue:46077", "issue:46123", "issue:46132", + "issue:46170", "issue:46202" ], "soft_pairs": [ + "issue:42886|issue:45229", + "issue:44060|issue:44077", + "issue:43673|issue:44734", + "issue:45003|issue:45561", + "issue:45584|issue:46170", "issue:44297|issue:44623", "issue:44869|issue:45356", "issue:44361|issue:44749", "issue:43646|issue:44162", "issue:44749|issue:45656", "issue:44623|issue:45656", - "issue:44871|issue:46132", "issue:43673|issue:43827", + "issue:44871|issue:46132", "issue:43701|issue:44749", "issue:44188|issue:46202", "issue:45230|issue:45397", "issue:42994|issue:43482", "issue:43475|issue:45381", - "issue:28282|issue:30990", - "issue:43475|issue:43643", "issue:44038|issue:44623", + "issue:43475|issue:43643", + "issue:28282|issue:30990", "issue:42371|issue:45092", "issue:43232|issue:43673", "issue:41628|issue:44933", - "issue:43475|issue:44661", "issue:45375|issue:45478", + "issue:43475|issue:44661", "issue:42994|issue:43525", "issue:46077|issue:46123", - "issue:45081|issue:45103", - "issue:45310|issue:45375", - "issue:45406|issue:46202", - "issue:42673|issue:43704", - "issue:41628|issue:44351", - "issue:44964|issue:45137", - "issue:44222|issue:45310", - "issue:42994|issue:43824" + "issue:45310|issue:45375" ], - "prepared_review_unit_hash": "11c7f16869a4fe869c1ba60c1d1f91f257bde7975a750ab85ab40ec4aba12093", + "prepared_review_unit_hash": "878631b08c10809bbb978baa77d5b5e73d93b6cf23844298d16d549ca18ac947", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36810, - "estimated_input_tokens": 9203, - "estimated_eval_tokens": 18662 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35546, + "estimated_input_tokens": 8887, + "estimated_eval_tokens": 18030 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a mixed bag of Transformers bug reports across unrelated models/features. Only two soft-edge pairs are true duplicates: the Qwen3.5-Moe from_pretrained regression and the Qwen3ForCausalLM VRAM leak.", - "confidence": 0.62, - "canonical_issue_reason": "issue:43646 is the broadest, most representative regression ticket here: Transformers 5.0.0 breaking custom model initialization, which matches the cluster\u2019s many v5 compatibility failures.", + "summary": "Loose cluster around Transformers v5 regressions and adjacent model/tokenizer bugs, but the soft pairs mostly point to different failure modes and subsystems rather than true duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43646 is the broadest and most representative v5 breakage report in the set, centered on custom/remote model initialization, which overlaps thematically with several other regression reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best single anchor for the cluster because it is high-signal, broadly scoped, and not tied to one niche model or documentation-only symptom.", + "best_issue_reason": "issue:43646 is the best single issue to anchor the cluster because it is concrete, high-signal, and spans a common v5 compatibility path without being overly narrow.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline tokenizer cache loading and multi-GPU CUDA OOM are unrelated failure modes; same library, different code paths." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "One is a tied-weights warning bug in Qwen3-Next, the other is a model API constraint for patchtsmixer; not the same underlying change." + }, + { + "left": "issue:43673", + "right": "issue:44734", + "accept": false, + "reason": "Both involve generation/KV cache behavior, but one is a missing cache object in v5.0.0 during chunked prefill and the other is a serve-time indexing crash in /v1/responses." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to a pytest-xdist captured_info.txt race in testing utilities." + }, + { + "left": "issue:45584", + "right": "issue:46170", + "accept": false, + "reason": "Empty transcription generation after token alignment is unrelated to Windows ROCm import/type errors around torch.distributed." + }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Related save_pretrained area, but one is tokenizer.save_pretrained and the other is processor.save_pretrained; different code paths and failure modes." + "reason": "Tokenizer save_pretrained metadata mismatch and processor.save_pretrained missing files are both serialization-related, but they affect different save paths and artifacts." }, { "left": "issue:44869", "right": "issue:45356", "accept": false, - "reason": "Whisper timestamp decoding crash vs Kimi tokenizer codec regression; different models and different underlying bugs." + "reason": "Whisper timestamp decode crash and Kimi-K2.5 tokenizer codec regression are distinct tokenizer/decoder bugs with different triggers." }, { "left": "issue:44361", "right": "issue:44749", "accept": false, - "reason": "Tokenizer AttributeError on tasks versus a post-upgrade performance slowdown; not the same bug." + "reason": "MLukeTokenizer task AttributeError and Chinese performance regression after upgrade are not the same bug." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "Both are regressions, but one is custom model initialization and the other is ESM2 breakage; different concrete failures." + "reason": "Custom model initialization failure and ESM2 breakage may both be v5 regressions, but they are separate model-specific compatibility issues." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "Performance regression after upgrade vs DeepSpeed optimizer step duplication; different symptoms and code paths." + "reason": "Data filtering slowdown after upgrade and deepspeed optimizer-step duplication are unrelated performance/ \u0627\u0644\u062a\u062f\u0631\u064a\u0628 issues." }, { "left": "issue:44623", "right": "issue:45656", "accept": false, - "reason": "Missing files on save_pretrained is unrelated to optimizer stepping under DeepSpeed." + "reason": "Missing processor save artifacts versus double optimizer stepping are different subsystems and symptoms." }, { - "left": "issue:44871", - "right": "issue:46132", + "left": "issue:43673", + "right": "issue:43827", "accept": false, - "reason": "Gemma eos_token configuration mismatch is unrelated to AttentionInterface.register changing function behavior." + "reason": "Generation cache bug and docs still referencing pipeline() are not the same underlying issue." }, { - "left": "issue:43673", - "right": "issue:43827", + "left": "issue:44871", + "right": "issue:46132", "accept": false, - "reason": "Generation cache bug versus docs still referencing pipeline(); code defect vs documentation issue." + "reason": "Gemma eos_token_id mismatch and AttentionInterface.register behavior change are separate configuration/API problems." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "resume_from_checkpoint key mismatch is a checkpoint/state issue, not the same as a slowdown regression." + "reason": "resume_from_checkpoint key mismatch is unrelated to the reported data-filtering slowdown." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "Attention kernel divergence under torch.compile is unrelated to Gemma pipeline roundtrip behavior." + "reason": "torch.compile divergence in attention kernels and Gemma pipeline roundtrip failure on a special character are unrelated." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "Placeholder bug report vs Gemma-4 ZeRO3 from_pretrained failure; no shared concrete bug." + "reason": "Generic bug report and gemma-4 zero3 from_pretrained error do not indicate the same concrete problem." }, { "left": "issue:42994", "right": "issue:43482", "accept": false, - "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are different operations and symptoms." + "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are different directions of the save/load lifecycle." }, { "left": "issue:43475", "right": "issue:45381", "accept": false, - "reason": "SAM3 video encoder attribute error and qwen2.5-vl video position-id bug are different model-specific issues." + "reason": "SAM 3 Video missing attribute and Qwen2.5-VL video position-id error are separate multimodal model bugs." }, { - "left": "issue:28282", - "right": "issue:30990", + "left": "issue:44038", + "right": "issue:44623", "accept": false, - "reason": "ImportError about missing PyTorch and a loading hang in Sentence Transformers are different failure modes." + "reason": "Qwen3-VL-Moe breakage and missing processor save files are unrelated." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "SAM3 encoder attribute error is unrelated to trust_remote_code returning missing fields from AutoConfig." + "reason": "SAM 3 Video encoder output attribute error and trust_remote_code missing fields are both initialization-related but on different model families and causes." }, { - "left": "issue:44038", - "right": "issue:44623", + "left": "issue:28282", + "right": "issue:30990", "accept": false, - "reason": "Qwen3-VL-Moe bug and processor.save_pretrained missing files are different concrete problems." + "reason": "Missing PyTorch import and SentenceTransformers loading hang are unrelated user-environment/runtime issues." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 API/settings warning is unrelated to InternVL2 meta-initialization incompatibility." + "reason": "TF32 settings docs/API issue and remote-code checkpoint meta-initialization incompatibility are not the same bug." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Both involve generation internals, but update_model_kwargs_for_generation after sync_gpus is not the same as missing GenerationMixin cache in chunked_prefill." + "reason": "GenerationMixin key update after sync_gpus and missing cache during chunked prefill are both generation-related but different concrete failures." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both are import-related, but the missing AutoImageProcessor import and the image_utils import issue are not clearly the same symbol/path." + "reason": "AutoImageProcessor import failure and nonexistent import from image_utils are distinct API/export regressions." }, { - "left": "issue:43475", - "right": "issue:44661", + "left": "issue:45375", + "right": "issue:45478", "accept": false, - "reason": "SAM3VisionEncoderOutput missing attribute is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES." + "reason": "Missing config field silently dropped by strict handling and Qwen3.5 MoE from_pretrained error are related to the same model family but not the same concrete issue." }, { - "left": "issue:45375", - "right": "issue:45478", - "accept": true, - "reason": "Same title and same Qwen3.5-Moe from_pretrained regression; these are duplicates." + "left": "issue:43475", + "right": "issue:44661", + "accept": false, + "reason": "SAM 3 Video attribute error and add-new-model-like failure in TOKENIZER_MAPPING_NAMES are different code paths and model infra concerns." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized save failure and Llama4 pad_token_id AttributeError are different models and different operations." + "reason": "Quantized model saving and Llama4Config missing pad_token_id are separate bugs." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Gemma4 causal-mask bug and MaskGenerationPipeline final-batch dropping bug are unrelated pipeline/attention issues." - }, - { - "left": "issue:45081", - "right": "issue:45103", - "accept": false, - "reason": "Mistral tokenizer regex crash versus auto_docstring annotations crash; different subsystems and triggers." + "reason": "Incorrect attention mask construction and final-partial-batch dropping in MaskGenerationPipeline are unrelated logic errors." }, { "left": "issue:45310", "right": "issue:45375", "accept": false, - "reason": "Same model family, but one is a from_pretrained load error and the other is a missing config field silently dropped; not clearly the same bug." - }, - { - "left": "issue:45406", - "right": "issue:46202", - "accept": false, - "reason": "Gemma4 processor missing _tokenizer and Gemma pipeline roundtrip on \u2581 are different bugs." - }, - { - "left": "issue:42673", - "right": "issue:43704", - "accept": true, - "reason": "Exact same title describing the same Qwen3ForCausalLM VRAM leak in multiple dataloader threads." - }, - { - "left": "issue:41628", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but they concern different missing names: AutoImageProcessor versus HybridCache." - }, - { - "left": "issue:44964", - "right": "issue:45137", - "accept": false, - "reason": "Loading a multimodal model and a DeepSpeed ZeRO3 deque IndexError are unrelated." - }, - { - "left": "issue:44222", - "right": "issue:45310", - "accept": false, - "reason": "FP8 save_pretrained on MoE versus Qwen3.5-Moe from_pretrained load failure; different directions and different bugs." - }, - { - "left": "issue:42994", - "right": "issue:43824", - "accept": false, - "reason": "Quantized save failure is unrelated to the missing Qwen2_5_VLForConditionalGeneration import." + "reason": "Qwen3.5 MoE from_pretrained error and missing deepstack_visual_indexes in Qwen3_5MoeVisionConfig are different model-specific regressions." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The cluster summary is grounded in the packet and the pair verdicts are conservative. The two accepted soft-edge pairs are exact-title duplicates, and the rejected pairs are generally limited to clearly different subsystems, models, or failure modes. The canonical anchor choice is also reasonable given the broad Transformers v5 compatibility regression." - }, - "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 }, @@ -11989,13 +11979,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 38, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:30333", "issue:30990", @@ -12005,49 +11995,57 @@ "issue:41628", "issue:42175", "issue:42222", + "issue:42673", "issue:42831", "issue:42898", + "issue:42994", "issue:43065", "issue:43232", "issue:43278", "issue:43377", + "issue:43704", "issue:43723", "issue:43784", "issue:43824", "issue:43874", "issue:43901", - "issue:44016", "issue:44038", - "issue:44060", - "issue:44075", "issue:44188", "issue:44222", "issue:44242", + "issue:44351", "issue:44373", - "issue:44448", "issue:44496", "issue:44568", "issue:44623", - "issue:44734", "issue:44779", "issue:44792", + "issue:44964", + "issue:45081", "issue:45092", "issue:45103", - "issue:45125", "issue:45137", "issue:45200", "issue:45310", "issue:45356", + "issue:45406", "issue:45561", "issue:45600", "issue:45712", "issue:45812", "issue:46123", - "issue:46231" + "issue:46202" ], "soft_pairs": [ - "issue:36010|issue:41628", + "issue:45081|issue:45103", + "issue:45406|issue:46202", + "issue:42673|issue:43704", + "issue:41628|issue:44351", + "issue:44964|issue:45137", + "issue:44222|issue:45310", + "issue:42994|issue:43824", "issue:44038|issue:45310", + "issue:36010|issue:41628", "issue:42831|issue:44222", "issue:42175|issue:45712", "issue:44496|issue:45561", @@ -12061,230 +12059,240 @@ "issue:42222|issue:45200", "issue:44623|issue:45310", "issue:43065|issue:46123", - "issue:30333|issue:33290", "issue:44188|issue:44373", + "issue:30333|issue:33290", "issue:43232|issue:44188", "issue:42898|issue:45812", "issue:43784|issue:43824", "issue:43723|issue:44568", - "issue:42898|issue:44779", - "issue:42898|issue:45356", "issue:43874|issue:45092", - "issue:44734|issue:45103", - "issue:44016|issue:46231", - "issue:44075|issue:46231", - "issue:42898|issue:44448", - "issue:44060|issue:45125" + "issue:42898|issue:45356", + "issue:42898|issue:44779" ], - "prepared_review_unit_hash": "7646c36fea5ec559f7a77f47efc9ac82d483090e714f6a4611a45bddac341c93", + "prepared_review_unit_hash": "0e2b75db758eff492a45213710f122583209a1fe03b9fefcf2996c53566b2f26", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35897, - "estimated_input_tokens": 8975, - "estimated_eval_tokens": 18206 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37043, + "estimated_input_tokens": 9261, + "estimated_eval_tokens": 18778 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is largely heterogeneous: most items are distinct bugs across unrelated subsystems. The soft-paired items share broad themes in a few cases (imports, tokenizers, generation), but none are strong enough to treat as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "issue:45812 is the strongest representative standalone bug here: a broad, silent tokenizer regression affecting all Granite models with clear user impact.", + "summary": "This cluster is mostly heterogeneous. Only one soft pair looks like a true duplicate (the two identical Qwen3ForCausalLM VRAM-leak issues); the rest are at best loosely similar by model family or transformer-v5 regression themes and should stay separate.", + "confidence": 0.9, + "canonical_issue_reason": "42673 is the clearest anchor because 42673/43704 are the only exact duplicate pair in the set (same title and same underlying Qwen3ForCausalLM VRAM leak).", "canonical_pr_reason": null, - "best_issue_reason": "issue:45812 has the clearest severity and scope among the issues, making it the best single issue to anchor the cluster.", + "best_issue_reason": "42673 is the best representative issue here: it has the strongest duplicate evidence and the most concrete, self-contained bug report among the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:36010", - "right": "issue:41628", + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Both are crashes, but they are different code paths: Mistral tokenizer regex patching vs auto-docstring parsing with future annotations." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Both touch Gemma, but one is a processor attribute crash and the other is a tokenizer roundtrip failure; not the same bug." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same reported Qwen3ForCausalLM VRAM leak in multiple dataloader threads; this is a duplicate." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and unrelated code paths." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading failure and DeepSpeed ZeRO3 deque error are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45310", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols and different code paths (`GenerationMixin` vs `AutoImageProcessor`)." + "reason": "Both mention MoE/FP8, but one is FP8 save_pretrained and the other is a Qwen3.5 Moe from_pretrained error; different failures." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized model saving and a Qwen2.5-VL import error are not the same underlying problem." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Same model family and release era, but one is a Qwen3-VL-Moe bug and the other is a Qwen3.5 MoE from_pretrained failure; likely related theme, not the same bug." + "reason": "Both concern Qwen3-family models, but the concrete failures differ enough that they would not be one merged fix." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors in transformers v5, but they involve different APIs and separate breakages." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "One reports accuracy drift with FineGrainedFP8, the other save_pretrained behavior for FP8 MoE; different symptoms and fixes." + "reason": "FP8 accuracy regression and FP8 save_pretrained bug are different issues." }, { "left": "issue:42175", "right": "issue:45712", "accept": false, - "reason": "Packaging/backend dependency issue versus leftover dummy PT classes leaking into dir/check_repo; not the same underlying problem." + "reason": "Package-dependency/backend inclusion problem vs dummy-class cleanup/check-repo failure are unrelated." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "These are unrelated: model loading/config recognition versus a pytest-xdist race in testing utils." + "reason": "Model config loading failure and pytest-xdist file race are unrelated." }, { "left": "issue:43377", "right": "issue:44242", "accept": false, - "reason": "Different model and failure mode: missing padding-mask support in MIMI encoder versus load-balancing loss omission in MoE routing." + "reason": "Missing padding-mask support in MIMI encoder and missing load-balancing loss in MoE are different code paths." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Both involve tokenization regressions, but one is about `clean_up_tokenization_spaces` behavior and the other about missing BOS/EOS insertion for a specific tokenizer." + "reason": "Both are tokenizer regressions in v5, but one is clean_up_tokenization_spaces behavior and the other is add_special_tokens/BOS-EOS handling." }, { "left": "issue:45561", "right": "issue:45600", "accept": false, - "reason": "A race condition in test workers is unrelated to auto_mappings referencing removed Sam3LiteText configs." + "reason": "Test-worker file race and broken auto_mappings references are unrelated." }, { "left": "issue:43901", "right": "issue:45561", "accept": false, - "reason": "Docs/behavior mismatch for TextClassificationPipeline is unrelated to a captured_info.txt race under xdist." + "reason": "Docstring mismatch for TextClassificationPipeline and pytest-xdist file race are unrelated." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation KV-cache continuation bug versus a Janus image-generation test failure; no shared code-path." + "reason": "Generation kwargs/update bug and a Janus image-generation test failure do not share the same failure mode." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Loading hangs in Sentence Transformers and token embedding reinitialization are different model initialization issues." + "reason": "Sentence-transformers loading stall and embedding reinitialization after resize are unrelated." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Embedding dtype drift in eval is unrelated to the DeepSpeed ZeRO3 deque underflow." + "reason": "BF16-to-FP32 dtype drift during evaluation and ZeRO3 deque underflow are unrelated." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "VitPose breakage and Gemma 4 multimodal token defaults are different model-specific bugs." + "reason": "VitPose breakage and Gemma 4 mm_token_type_ids defaulting are different multimodal issues." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Missing processor save_pretrained artifacts is a serialization issue, not the same as Qwen3.5 MoE loading failure." + "reason": "Missing files in processor.save_pretrained and Qwen3.5 Moe from_pretrained error are different problems." }, { "left": "issue:43065", "right": "issue:46123", "accept": false, - "reason": "Dummy Conv2d placeholder in Sam3PixelDecoder is unrelated to MaskGenerationPipeline dropping final partial-batch results." + "reason": "Dummy Conv2d in Sam3PixelDecoder and MaskGenerationPipeline partial-batch dropping are unrelated." }, { - "left": "issue:30333", - "right": "issue:33290", + "left": "issue:44188", + "right": "issue:44373", "accept": false, - "reason": "MLFlow reporting on failed training jobs and DeepSpeed OOM with Adafactor are unrelated problems." + "reason": "Torch.compile attention branching bug and a wrong docstring are not duplicates." }, { - "left": "issue:44188", - "right": "issue:44373", + "left": "issue:30333", + "right": "issue:33290", "accept": false, - "reason": "Attention-kernel divergence under torch.compile is unrelated to a wrong docstring for position_ids." + "reason": "MLFlow reporting on failed jobs and DeepSpeed OOM with Adafactor are unrelated." }, { "left": "issue:43232", "right": "issue:44188", "accept": false, - "reason": "Both touch generation/attention internals, but one is a KV-cache continuation bug and the other is compile-time branching in attention kernels." + "reason": "Both involve generation/runtime behavior, but they are different concrete bugs and code paths." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions in v5, but they affect different behaviors and models; not the same concrete bug." + "reason": "Both are tokenizer regressions across v4/v5, but they affect different models and different tokenization behavior." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but for different missing classes and different affected packages/models." + "reason": "Both are import errors, but for different symbols and different downstream packages." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Tokenizer loading in v5 and add_special_tokens behavior are distinct tokenizer issues with different failure modes." - }, - { - "left": "issue:42898", - "right": "issue:44779", - "accept": false, - "reason": "Both are v5 tokenizer regressions, but one is about cleanup-space behavior and the other about DeepSeek codec handling." - }, - { - "left": "issue:42898", - "right": "issue:45356", - "accept": false, - "reason": "Related only at the broad tokenizer-regression level; Kimi-K2.5 codec handling is a separate bug from clean_up_tokenization_spaces behavior." + "reason": "Both concern tokenizer behavior in v5, but one is load-time tokenizer issues and the other is special-token insertion semantics." }, { "left": "issue:43874", "right": "issue:45092", "accept": false, - "reason": "Image processor API breakage and remote-code meta-init incompatibility are different multimodal loading issues." - }, - { - "left": "issue:44734", - "right": "issue:45103", - "accept": false, - "reason": "Server KV-cache tensor indexing crash and auto_docstring annotation parsing crash are unrelated." - }, - { - "left": "issue:44016", - "right": "issue:46231", - "accept": false, - "reason": "Notebook syntax error and a docs parameter-name mismatch are not the same issue." - }, - { - "left": "issue:44075", - "right": "issue:46231", - "accept": false, - "reason": "Optimizer argument handling and documentation typo are unrelated." + "reason": "Both are multimodal model-loading issues, but the missing image-patch method and meta-initialization incompatibility are distinct bugs." }, { - "left": "issue:44060", - "right": "issue:45125", + "left": "issue:42898", + "right": "issue:45356", "accept": false, - "reason": "Wrong tied-weights warning in Qwen3-Next and missing `_tp_plan` in Qwen3.5 MoE are different model-specific bugs." + "reason": "Both are tokenizer regressions around v5, but the codec/fix_mistral_regex problem is separate from clean_up_tokenization_spaces." }, { - "left": "issue:44222", - "right": "issue:45200", + "left": "issue:42898", + "right": "issue:44779", "accept": false, - "reason": "FP8 save_pretrained behavior for MoE does not match the Gemma 4 multimodal-token defaulting issue." + "reason": "Granite token ID regression and DeepSeek tokenizer regression are separate model-specific bugs." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays conservative. The only accepted soft pair, issue:42673 vs issue:43704, is clearly supported by identical titles. The other soft-edge rejections do not overclaim duplicate status based on the provided titles and issue metadata, and the overall summary correctly treats the cluster as mostly heterogeneous." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -12293,13 +12301,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 39, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28282", "issue:30333", @@ -12310,14 +12318,16 @@ "issue:42175", "issue:42757", "issue:42831", + "issue:42898", "issue:43122", - "issue:43232", + "issue:43475", "issue:43479", - "issue:43644", "issue:43723", - "issue:43761", "issue:43873", "issue:43976", + "issue:44016", + "issue:44060", + "issue:44075", "issue:44079", "issue:44190", "issue:44206", @@ -12325,32 +12335,35 @@ "issue:44263", "issue:44295", "issue:44297", - "issue:44373", "issue:44448", - "issue:44483", - "issue:44484", - "issue:44485", "issue:44623", + "issue:44734", "issue:44869", - "issue:44871", "issue:44933", "issue:44945", "issue:44977", "issue:45071", "issue:45092", "issue:45103", + "issue:45125", "issue:45245", "issue:45310", + "issue:45381", "issue:45405", - "issue:45406", "issue:45656", "issue:45712", - "issue:45834", "issue:46082", "issue:46132", - "issue:46209" + "issue:46209", + "issue:46231", + "issue:46305" ], "soft_pairs": [ + "issue:44734|issue:45103", + "issue:44016|issue:46231", + "issue:44075|issue:46231", + "issue:44060|issue:45125", + "issue:42898|issue:44448", "issue:44295|issue:45656", "issue:45405|issue:45712", "issue:44448|issue:46132", @@ -12360,192 +12373,210 @@ "issue:43479|issue:44079", "issue:43976|issue:44933", "issue:44297|issue:45656", - "issue:30333|issue:34689", "issue:44222|issue:44623", + "issue:30333|issue:34689", "issue:42831|issue:44623", "issue:45245|issue:46132", - "issue:30333|issue:36331", "issue:44190|issue:46082", + "issue:30333|issue:36331", "issue:43479|issue:44869", "issue:30333|issue:35141", "issue:28282|issue:30333", "issue:45071|issue:46082", - "issue:44206|issue:45103", + "issue:43475|issue:46305", "issue:44297|issue:45310", "issue:41628|issue:42757", + "issue:44206|issue:45103", "issue:43723|issue:44977", + "issue:45381|issue:46305", "issue:44263|issue:45092", "issue:43122|issue:46209", - "issue:43873|issue:44945", - "issue:43232|issue:45834", - "issue:44222|issue:44484", - "issue:44373|issue:44485", - "issue:44295|issue:44623", - "issue:43479|issue:43644", - "issue:44483|issue:45406", - "issue:43761|issue:46082", - "issue:43723|issue:44871" + "issue:43873|issue:44945" ], - "prepared_review_unit_hash": "93ec876320bee82da94cec875c762842d62e67c0e5456fcfac99b3e5426dca4f", + "prepared_review_unit_hash": "fb45d91805ef96d863added34f725d25ad84de582a7dfd72dfb330c240cbfe15", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37639, - "estimated_input_tokens": 9410, - "estimated_eval_tokens": 19076 + "soft_pair_count": 33, + "serialized_chars": 37489, + "estimated_input_tokens": 9373, + "estimated_eval_tokens": 19002 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated Transformers issues. None of the soft pairs look like the same underlying bug or change, so all should be rejected as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "issue:43644 is the broadest and most representative regression in the set, centered on a core v5 initialization/buffer behavior rather than a narrow model-specific or tooling-only problem.", + "summary": "These items are a loose grab bag of Transformers regressions and compatibility issues, but the soft pairs are not true duplicates; they mostly share only broad themes like \"v5 breakage\" or \"save_pretrained\".", + "confidence": 0.34, + "canonical_issue_reason": "issue:45071 is the most representative closed item: a concrete Transformers v5 regression that affects a core API contract. It best matches the cluster\u2019s general pattern of version-related breakages.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 best represents the cluster because it is a general v5 regression and is the closest fit to the few initialization/buffer-adjacent reports, even though the cluster is not truly cohesive.", + "best_issue_reason": "issue:45071 is the best anchor because it is specific, user-facing, and broadly aligned with the cluster\u2019s recurring v5 compatibility complaints.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Different subsystems and failure modes: serving KV-cache continuation indexing vs auto-docstring handling of future annotations." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Both touch documentation/examples, but one is a notebook syntax error and the other is a parameter-name mismatch; not the same bug." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument handling and a docs typo are unrelated changes." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Different Qwen bugs: a wrong tied-weights warning versus a missing tensor-parallel plan." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "Both involve v5 output differences, but one is tokenization cleanup behavior and the other is Pegasus generation output; not one underlying defect." + }, { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "Different subsystems: a buffer/position_ids read issue versus a deepspeed optimizer-step count problem." + "reason": "Position-ids buffer access and double optimizer stepping with deepspeed are separate runtime issues." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "Unrelated: PEFT version pinning/release status versus leftover dummy classes leaking into the package API." + "reason": "A version pin/release mismatch is unrelated to leftover dummy classes leaking into imports." }, { "left": "issue:44448", "right": "issue:46132", "accept": false, - "reason": "Different bugs: Pegasus v4/v5 output drift versus AttentionInterface.register changing registered function behavior." + "reason": "Different code paths: Pegasus output regression versus AttentionInterface.register behavior changes." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "One is backend extras packaging, the other is a missing image_utils import; no shared code-path bug." + "reason": "Different import/export problems in different areas; one is backend packaging, the other an image_utils symbol export issue." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "Both mention save_pretrained, but one is FP8/MoE serialization and the other is tokenizer metadata mismatch; not the same fix." + "reason": "Both mention save_pretrained, but one is FP8 MOE serialization and the other is tokenizer_config metadata; not the same bug." }, { "left": "issue:44933", "right": "issue:45712", "accept": false, - "reason": "Different import/package issues: a nonexistent image_utils import versus dummy class cleanup in the no-torch path." + "reason": "Both are import-surface issues, but one is a missing import from image_utils and the other is dummy classes leaking from the torch-less shim." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Phi4MultimodalConfig default handling is unrelated to ModelOutput key assignment semantics." + "reason": "Different objects and failures: multimodal config default initialization vs ModelOutput key assignment behavior." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version compatibility regression versus a missing import symbol; no shared underlying bug." + "reason": "Python version compatibility and a missing image_utils import are unrelated." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer save_pretrained metadata bug is unrelated to deepspeed optimizer stepping." + "reason": "Tokenizer save metadata and double optimizer stepping have no shared code-path." }, { - "left": "issue:30333", - "right": "issue:34689", + "left": "issue:44222", + "right": "issue:44623", "accept": false, - "reason": "MLFlow reporting on failed training is unrelated to model loading regression for Llama 3.2 Vision." + "reason": "Both involve saving artifacts, but FP8 model serialization and processor.save_pretrained missing files are different implementations and bugs." }, { - "left": "issue:44222", - "right": "issue:44623", + "left": "issue:30333", + "right": "issue:34689", "accept": false, - "reason": "Both are save_pretrained-related, but they affect different artifacts and code paths: FP8 model weights versus processor files." + "reason": "Training-job MLflow reporting and Llama 3.2 Vision model loading are unrelated." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FP8 accuracy regression is not the same issue as missing processor save_pretrained files." + "reason": "An FP8 accuracy regression is unrelated to processor save_pretrained missing files." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "Category-count runtime limit is unrelated to AttentionInterface.register behavior." - }, - { - "left": "issue:30333", - "right": "issue:36331", - "accept": false, - "reason": "Failed-job MLFlow reporting and CustomTrainer.compute_loss signature mismatch are unrelated." + "reason": "Category-count limits and AttentionInterface.register behavior are separate issues." }, { "left": "issue:44190", "right": "issue:46082", "accept": false, - "reason": "Dataset loading in the example script is unrelated to LlamaConfig head_dim validation." - }, - { - "left": "issue:43479", - "right": "issue:44869", - "accept": false, - "reason": "Multimodal config initialization and Whisper timestamp decode crash are different code paths." + "reason": "Local dataset loading and LlamaConfig head_dim validation are unrelated." }, { "left": "issue:30333", - "right": "issue:35141", + "right": "issue:36331", "accept": false, - "reason": "MLFlow job-status reporting is unrelated to embedding resize/post_init reinitialization." + "reason": "MLflow reporting after failed training and Trainer compute_loss signature mismatch are different trainer bugs." }, { - "left": "issue:28282", - "right": "issue:30333", + "left": "issue:43479", + "right": "issue:44869", "accept": false, - "reason": "ImportError for PyTorch backend is unrelated to MLFlow reporting on failed training jobs." + "reason": "Multimodal config defaults and Whisper timestamp decoding are unrelated." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "Both involve config/type validation, but they are different concrete failures and not one fixable code-path bug." + "reason": "A general PretrainedConfig type-checking regression is not the same as LlamaConfig head_dim validation." }, { - "left": "issue:44206", - "right": "issue:45103", + "left": "issue:43475", + "right": "issue:46305", "accept": false, - "reason": "Feature extractor argument support and auto-docstring future-annotations crash are unrelated." + "reason": "Different model families and bugs: missing SAM3 attribute vs MoonshineStreaming off-by-one masking." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to Qwen3.5 MoE from_pretrained failure." + "reason": "Tokenizer save metadata mismatch and Qwen3.5 MoE from_pretrained errors are different failure modes." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Different missing-symbol import errors from different modules; not the same underlying regression." + "reason": "Both are import errors, but for unrelated symbols and dependencies." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "LaserFeatureExtractor argument handling and auto-docstring annotation parsing are unrelated." }, { "left": "issue:43723", @@ -12553,65 +12584,29 @@ "accept": false, "reason": "Tokenizer loading in v5 and Qwen3.5 flash-attention generation failures are unrelated." }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Different sequence/streaming bugs in different models; no shared code-path." + }, { "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "torch.split return handling in a MoE indexer is unrelated to meta initialization incompatibility in old InternVL2 checkpoints." + "reason": "A torch.split return-value bug and meta-init incompatibility in old InternVL2 checkpoints are unrelated." }, { "left": "issue:43122", "right": "issue:46209", "accept": false, - "reason": "Tokenizer output changes and FP8 inference failure on B200 are unrelated." + "reason": "Tokenization differences across versions and FP8 inference on B200 are unrelated." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to incorrect output under pipeline parallelism." - }, - { - "left": "issue:43232", - "right": "issue:45834", - "accept": false, - "reason": "Generation kwargs update after sync_gpus and Kosmos2.5 OCR index errors are different bugs." - }, - { - "left": "issue:44222", - "right": "issue:44484", - "accept": false, - "reason": "FP8 MoE save_pretrained and max_shard_size default selection are not the same issue." - }, - { - "left": "issue:44373", - "right": "issue:44485", - "accept": false, - "reason": "Docstring correction for position_ids is unrelated to the GLM-5 RoPE implementation discussion." - }, - { - "left": "issue:44295", - "right": "issue:44623", - "accept": false, - "reason": "Position_ids buffer read error and processor save_pretrained file omission are unrelated." - }, - { - "left": "issue:43479", - "right": "issue:43644", - "accept": false, - "reason": "Both touch initialization, but one is Phi4MultimodalConfig None handling while the other is non-persistent buffer contents; not the same bug." - }, - { - "left": "issue:44483", - "right": "issue:45406", - "accept": false, - "reason": "Chat/completions request rejection and Gemma4Processor _tokenizer attribute crash are different server-side failures." - }, - { - "left": "issue:43761", - "right": "issue:46082", - "accept": false, - "reason": "CLIPVisionModel hidden_states regression and LlamaConfig head_dim validation are unrelated." + "reason": "Quantization offloading and pipeline-parallel output correctness are different issues." } ] }, @@ -12627,23 +12622,20 @@ "component_index": 1, "component_count": 18, "review_unit_index": 40, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:22355", "issue:28282", "issue:30990", - "issue:33357", "issue:38175", "issue:39401", "issue:41628", - "issue:42371", - "issue:42757", "issue:42831", "issue:42886", "issue:42907", @@ -12651,45 +12643,56 @@ "issue:43066", "issue:43232", "issue:43479", - "issue:43550", - "issue:43576", - "issue:43646", - "issue:43784", + "issue:43644", + "issue:43723", + "issue:43761", "issue:43824", + "issue:43950", "issue:44038", - "issue:44117", + "issue:44222", "issue:44261", "issue:44263", "issue:44295", - "issue:44297", "issue:44360", + "issue:44368", "issue:44373", - "issue:44464", + "issue:44483", + "issue:44484", + "issue:44485", "issue:44514", "issue:44568", - "issue:44704", + "issue:44623", "issue:44829", + "issue:44871", "issue:44913", "issue:44945", "issue:45092", "issue:45103", - "issue:45278", - "issue:45290", "issue:45310", "issue:45357", - "issue:45593", + "issue:45406", "issue:45753", + "issue:45834", "issue:46002", "issue:46050", "issue:46082", - "issue:46132" + "issue:46305" ], "soft_pairs": [ + "issue:43232|issue:45834", + "issue:44222|issue:44484", + "issue:44373|issue:44485", + "issue:44295|issue:44623", + "issue:43479|issue:43644", + "issue:44483|issue:45406", + "issue:43761|issue:46082", + "issue:43723|issue:44871", "issue:45092|issue:45357", "issue:44263|issue:44360", "issue:30990|issue:38175", "issue:44261|issue:45092", "issue:42907|issue:46050", + "issue:43644|issue:46305", "issue:44514|issue:45103", "issue:44829|issue:46002", "issue:43479|issue:44913", @@ -12699,224 +12702,208 @@ "issue:41628|issue:43824", "issue:42831|issue:44038", "issue:39401|issue:44568", + "issue:44368|issue:46305", "issue:42886|issue:43066", "issue:42886|issue:42994", - "issue:44945|issue:45753", - "issue:42757|issue:45278", - "issue:44464|issue:46082", - "issue:44117|issue:44704", - "issue:43646|issue:46002", - "issue:43576|issue:43784", - "issue:42371|issue:44263", - "issue:43550|issue:46132", - "issue:30990|issue:33357", - "issue:43232|issue:45593", - "issue:44295|issue:44297", - "issue:44704|issue:45290", - "issue:44261|issue:44373" + "issue:43950|issue:46305", + "issue:44945|issue:45753" ], - "prepared_review_unit_hash": "630a025509b1bf85c044f52b5ea1dfa547a1e683c82446b15581ededee29259b", + "prepared_review_unit_hash": "f1acb502fbe1a4df5c107cf2f5dfd79099238911cbbb455e6bae3077a0a8a4e9", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36027, - "estimated_input_tokens": 9007, - "estimated_eval_tokens": 18270 + "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": "Mostly heterogeneous issue reports with only loose thematic overlap around Transformers v5 regressions, imports, tokenization, and serialization. I do not see any strong duplicate pairs; the closest umbrella-style issue is the broad import-regression report, but the set is not a true duplicate cluster.", - "confidence": 0.78, - "canonical_issue_reason": "Issue 45278 is the broadest umbrella-style report in this set and best represents the small import-regression theme, though the overall cluster is still heterogeneous.", + "summary": "The cluster is highly heterogeneous: these are mostly unrelated issues that only share broad surface-level themes like imports, tokenizers, saving/loading, or multimodal regressions. I do not see a true duplicate set among the soft pairs.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue is a true canonical duplicate target here, but issue:43950 is the clearest standalone regression report: it names the concrete failure (`from_pretrained()` corrupting non-persistent buffers), the transformers 5.x regression scope, and the behavior is specific enough to anchor related reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45278 is the most representative high-level issue here because it describes a broad post-upgrade regression rather than a narrow, model-specific symptom.", + "best_issue_reason": "issue:43950 is the strongest general issue in the set for triage purposes because it is precise, reproducible, and describes a concrete code-path regression rather than a broad symptom or discussion-only problem.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Different subsystems and failures: generation kwargs handling vs long OCR input index error." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "Quantized save_pretrained/moe failure is unrelated to max_shard_size behavior." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring typo/report about position_ids is not the same as GLM-5 RoPE implementation." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Buffer registration/read error is unrelated to processor save_pretrained missing files." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config None-handling for multimodal defaults is a different bug from corrupted non-persistent buffers." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat completions request handling and Gemma4Processor _tokenizer crash are different code paths." + }, + { + "left": "issue:43761", + "right": "issue:46082", + "accept": false, + "reason": "CLIP hidden_states regression and LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer load failure and eos_token_id config mismatch are separate tokenizer issues." + }, { "left": "issue:45092", "right": "issue:45357", "accept": false, - "reason": "Different models and different failures: meta-initialization incompatibility vs incorrect saved visual encoder keys." + "reason": "Meta-init incompatibility for old InternVL2 checkpoints is different from bad visual encoder keys on save_pretrained." }, { "left": "issue:44263", "right": "issue:44360", "accept": false, - "reason": "Same broad subsystem name, but the bugs differ: torch.split return handling vs missing ReLU in the DSA indexer." + "reason": "Both mention DSA indexer, but one is about torch.split outputs and the other about a missing ReLU; not the same bug." }, { "left": "issue:30990", "right": "issue:38175", "accept": false, - "reason": "Unrelated symptoms and model families; one is a loading hang, the other is a zero-probability accuracy issue." + "reason": "Sentence Transformers loading hang and SIGLIP2 zero probabilities are unrelated symptoms." }, { "left": "issue:44261", "right": "issue:45092", "accept": false, - "reason": "Different code paths and models; a missing rms_norm_eps config field is not the same as meta-init incompatibility." + "reason": "MLA rms_norm_eps precision issue is not the same as remote-code/meta-init incompatibility." }, { "left": "issue:42907", "right": "issue:46050", "accept": false, - "reason": "Both touch quantized saving/loading, but they describe different quantization flows and failure modes." + "reason": "Both involve quantization/save-load, but they point to different models and different failure modes; too broad to merge." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer corruption is unrelated to MoonshineStreaming sliding-window off-by-one." }, { "left": "issue:44514", "right": "issue:45103", "accept": false, - "reason": "Processor batched chat-template crash and auto-docstring AttributeError are unrelated bugs." + "reason": "Batched chat template crash and auto_docstring annotation crash are unrelated." }, { "left": "issue:44829", "right": "issue:46002", "accept": false, - "reason": "Both involve training instability, but the triggers and root causes are different and not mergeable into one fix." + "reason": "Flash-attention training degeneration and NaNs after checkout/init are different concrete problems." }, { "left": "issue:43479", "right": "issue:44913", "accept": false, - "reason": "Both are config-related, but one is multimodal default initialization and the other is a GPTNeoX reload regression." + "reason": "Different configs and lifecycle bugs: default subconfig init vs rotary_pct persistence on reload." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "Different areas entirely: FP8 accuracy regression vs Qwen3.5 MoE from_pretrained failure." + "reason": "FineGrainedFP8 accuracy issue and Qwen3.5 MoE from_pretrained error are not the same bug." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "A code bug in the DSA indexer and a wrong docstring are not the same underlying issue." + "reason": "Indexer split behavior and a wrong docstring are not the same underlying change." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Both are import errors, but they involve different missing modules and different dependency problems." + "reason": "Both are import errors, but they stem from different missing components/modules and should not be merged." }, { "left": "issue:41628", "right": "issue:43824", "accept": false, - "reason": "Different missing symbols with no evidence of a single shared code-path regression." + "reason": "Missing AutoImageProcessor export and missing Qwen2_5_VLForConditionalGeneration export are distinct symbol-import failures." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "One is an accuracy issue, the other is a model-specific from_pretrained error." + "reason": "Generic FP8 accuracy regression is not the same as the Qwen3-VL-Moe bug." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer bugs, but offset_mapping and special-token insertion are distinct behaviors." + "reason": "Qwen3 offset_mapping bug and mdeberta BOS/EOS add_special_tokens bug affect different tokenizers and code paths." }, { - "left": "issue:42886", - "right": "issue:43066", + "left": "issue:44368", + "right": "issue:46305", "accept": false, - "reason": "Cache/offline loading failure and tokenizer decoder-type regression are separate issues." + "reason": "A tied-embeddings warning in fine-tuning is unrelated to a streaming mask off-by-one." }, { "left": "issue:42886", - "right": "issue:42994", - "accept": false, - "reason": "Offline cache loading and quantized model saving are unrelated failure modes." - }, - { - "left": "issue:44945", - "right": "issue:45753", - "accept": false, - "reason": "Pipeline-parallelism output corruption and a Qwen3_5 infinite loop on one image are different bugs." - }, - { - "left": "issue:42757", - "right": "issue:45278", - "accept": false, - "reason": "The latter is a broad post-upgrade import-regression report, but the specific missing symbol issue is not clearly the same underlying bug." - }, - { - "left": "issue:44464", - "right": "issue:46082", - "accept": false, - "reason": "Compiled-forward chunked generation inconsistency and a LlamaConfig validation error are unrelated." - }, - { - "left": "issue:44117", - "right": "issue:44704", - "accept": false, - "reason": "From_pretrained handling of tokenizer mappings and kwargs passing to cached_file are different internal problems." - }, - { - "left": "issue:43646", - "right": "issue:46002", - "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are not the same concrete bug." - }, - { - "left": "issue:43576", - "right": "issue:43784", - "accept": false, - "reason": "An env CLI regression and a sentence-transformers import NameError are unrelated." - }, - { - "left": "issue:42371", - "right": "issue:44263", - "accept": false, - "reason": "TF32 settings guidance and a DSA indexer torch.split issue have no shared root cause." - }, - { - "left": "issue:43550", - "right": "issue:46132", - "accept": false, - "reason": "torch.compile/SDPA failure in Bamba and AttentionInterface.register behavior change are different attention bugs." - }, - { - "left": "issue:30990", - "right": "issue:33357", - "accept": false, - "reason": "Both involve loading problems, but sentence-transformers loading hang and CLIP bus error are distinct symptoms." - }, - { - "left": "issue:43232", - "right": "issue:45593", + "right": "issue:43066", "accept": false, - "reason": "Generation kwargs handling and D-FINE auxiliary-loss behavior are unrelated." + "reason": "Offline cache loading failure and wrong decoder type in v5 are different tokenizer regressions." }, { - "left": "issue:44295", - "right": "issue:44297", + "left": "issue:42886", + "right": "issue:42994", "accept": false, - "reason": "Registering position_ids as a buffer and tokenizer_config class mismatch are separate serialization bugs." + "reason": "Tokenizer offline cache loading and quantized model saving failure are unrelated." }, { - "left": "issue:44704", - "right": "issue:45290", + "left": "issue:43950", + "right": "issue:46305", "accept": false, - "reason": "Both involve processor APIs, but kwargs propagation to cached_file and assistant-tool-call chat-template crashes are different code paths." + "reason": "Non-persistent buffer corruption is unrelated to the sliding-window mask bug." }, { - "left": "issue:44261", - "right": "issue:44373", + "left": "issue:44945", + "right": "issue:45753", "accept": false, - "reason": "A precision/config bug and a docstring issue are not duplicates." + "reason": "Pipeline parallelism output errors and an image-triggered infinite loop are different failure modes." } ] }, @@ -12932,47 +12919,45 @@ "component_index": 1, "component_count": 18, "review_unit_index": 41, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ - "issue:22355", "issue:28218", - "issue:29159", "issue:30990", - "issue:36296", - "issue:39401", + "issue:33357", "issue:41628", + "issue:42371", "issue:42757", "issue:42907", - "issue:43066", "issue:43116", "issue:43122", "issue:43232", "issue:43408", - "issue:43452", "issue:43475", + "issue:43550", + "issue:43576", + "issue:43646", "issue:43723", "issue:43756", "issue:43784", - "issue:43824", - "issue:43906", "issue:43937", - "issue:43992", "issue:44038", "issue:44060", - "issue:44222", + "issue:44117", + "issue:44261", + "issue:44263", "issue:44295", "issue:44297", - "issue:44351", "issue:44360", + "issue:44373", "issue:44442", + "issue:44464", "issue:44485", - "issue:44488", "issue:44514", "issue:44521", "issue:44625", @@ -12982,274 +12967,244 @@ "issue:44945", "issue:44964", "issue:45200", - "issue:45216", "issue:45245", "issue:45278", - "issue:45446", - "issue:45478", - "issue:45600" + "issue:45290", + "issue:45593", + "issue:45600", + "issue:46002", + "issue:46082", + "issue:46132" ], "soft_pairs": [ + "issue:42757|issue:45278", + "issue:44464|issue:46082", + "issue:44117|issue:44704", + "issue:43576|issue:43784", + "issue:43646|issue:46002", + "issue:42371|issue:44263", + "issue:43550|issue:46132", + "issue:30990|issue:33357", + "issue:43232|issue:45593", + "issue:44295|issue:44297", + "issue:44704|issue:45290", + "issue:44261|issue:44373", "issue:43723|issue:44442", "issue:42907|issue:44945", "issue:43723|issue:44779", - "issue:41628|issue:45278", "issue:42907|issue:43937", - "issue:43784|issue:45600", + "issue:41628|issue:45278", "issue:43756|issue:44060", + "issue:43784|issue:45600", "issue:43408|issue:43475", "issue:43122|issue:44514", - "issue:43116|issue:45200", "issue:42907|issue:44964", - "issue:44625|issue:44829", + "issue:43116|issue:45200", "issue:44360|issue:44485", + "issue:44625|issue:44829", "issue:43122|issue:44521", - "issue:43232|issue:45245", "issue:44038|issue:44297", + "issue:43232|issue:45245", "issue:28218|issue:30990", - "issue:43122|issue:44779", "issue:44038|issue:44295", - "issue:42907|issue:43452", - "issue:36296|issue:43116", - "issue:44038|issue:44222", - "issue:39401|issue:44488", - "issue:45216|issue:45478", - "issue:39401|issue:44521", - "issue:43824|issue:44351", - "issue:22355|issue:28218", - "issue:43784|issue:45446", - "issue:22355|issue:29159", - "issue:42757|issue:44351", - "issue:43122|issue:43906", - "issue:43992|issue:44704", - "issue:43066|issue:44779" + "issue:43122|issue:44779" ], - "prepared_review_unit_hash": "694b3ea5798556437b4f8ab0d3d4aab7e2b4ead77fafadb30b18f6e6eddeef1e", + "prepared_review_unit_hash": "dd506ebcb330fbbd1f004fc26d5eed99361e656266eb9179761680e066c11990", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37141, - "estimated_input_tokens": 9286, - "estimated_eval_tokens": 18828 + "soft_pair_count": 31, + "serialized_chars": 36606, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a bag of unrelated transformer regressions, many in v5 and many around tokenizers/imports, but the listed pairs do not look like true duplicates under a strict same-bug standard.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43122 is the broadest and most representative tokenizer-regression report in the set, with a clear v4-to-v5 behavior change and enough generality to anchor the cluster better than the model-specific or import-specific issues.", + "summary": "This cluster is mostly heterogeneous: the soft pairs share broad subsystems (tokenizers, imports, generation, multimodal chat templating) but not the same concrete bug or fix. I would not merge any of the candidate pairs as duplicates.", + "confidence": 0.71, + "canonical_issue_reason": "If forced to pick one reference issue, issue:43122 is the broadest and most representative tokenizer-regression report in the set; however, the cluster as a whole does not have a true canonical duplicate issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43122 best captures the recurring tokenizer-behavior regression theme, but it should be treated as representative rather than a confirmed duplicate hub.", + "best_issue_reason": "issue:43122 is the best single issue to anchor nearby tokenizer-regression discussions because it is relatively broad, clearly reproducible, and has inbound references; still, it does not represent the rest of this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43723", - "right": "issue:44442", + "left": "issue:42757", + "right": "issue:45278", "accept": false, - "reason": "Both involve AutoTokenizer loading, but one is a broad v5 loading regression and the other is a specific FastSpeech2ConformerTokenizer failure; not enough evidence they are the same concrete bug." + "reason": "Both are import-related, but 42757 is one missing symbol from huggingface_hub while 45278 is a broad upgrade breakage with many import errors; not the same underlying bug." }, { - "left": "issue:42907", - "right": "issue:44945", + "left": "issue:44464", + "right": "issue:46082", "accept": false, - "reason": "Different subsystems and symptoms: dequantized model saving versus incorrect output under pipeline parallelism." + "reason": "Chunked generation with compiled forward is unrelated to LlamaConfig head_dim validation." }, { - "left": "issue:43723", - "right": "issue:44779", + "left": "issue:44117", + "right": "issue:44704", "accept": false, - "reason": "Both are tokenizer-related v5 regressions, but one is loading failure and the other is incorrect tokenization output for Deepseek; not the same bug." + "reason": "One is TOKENIZER_MAPPING_NAMES returning None; the other is AutoProcessor not forwarding kwargs to cached_file. Different code paths and failures." }, { - "left": "issue:41628", - "right": "issue:45278", + "left": "issue:43576", + "right": "issue:43784", "accept": false, - "reason": "The second is a broad import-error report after upgrade; the first is a specific missing AutoImageProcessor import. Too general to merge as duplicates." + "reason": "A broken CLI env command and a NameError during sentence-transformers import are unrelated issues." }, { - "left": "issue:42907", - "right": "issue:43937", - "accept": false, - "reason": "Unrelated issues: saving dequantized models versus invalid GenerationConfig for GLM-5." - }, - { - "left": "issue:43784", - "right": "issue:45600", - "accept": false, - "reason": "Both are import-related breakages, but they point to different missing symbols and different root causes." - }, - { - "left": "issue:43756", - "right": "issue:44060", - "accept": false, - "reason": "Different model families and failure modes; no clear shared code-path bug." - }, - { - "left": "issue:43408", - "right": "issue:43475", + "left": "issue:43646", + "right": "issue:46002", "accept": false, - "reason": "Both mention SAM 3, but one is a model-type/config warning and the other is a missing attribute at runtime; not the same underlying defect." + "reason": "Custom model initialization breakage and NaNs in classification heads are different problems with different symptoms and likely different fixes." }, { - "left": "issue:43122", - "right": "issue:44514", + "left": "issue:42371", + "right": "issue:44263", "accept": false, - "reason": "Tokenizer regression versus processor chat-template crash; different code paths and symptoms." + "reason": "TF32 settings guidance and a torch.split return-value issue are unrelated." }, { - "left": "issue:43116", - "right": "issue:45200", + "left": "issue:43550", + "right": "issue:46132", "accept": false, - "reason": "Multi-label classification example failure is unrelated to Gemma 4 multimodal token-type defaults." + "reason": "torch.compile with SDPA on Bamba and AttentionInterface.register changing function behavior are distinct bugs." }, { - "left": "issue:42907", - "right": "issue:44964", + "left": "issue:30990", + "right": "issue:33357", "accept": false, - "reason": "Different problems: dequantized save failure versus loading a multimodal model." + "reason": "Sentence Transformers loading hangs and a MacOS bus error on a CLIP model are not the same defect." }, { - "left": "issue:44625", - "right": "issue:44829", + "left": "issue:43232", + "right": "issue:45593", "accept": false, - "reason": "Unrelated model/config propagation bug versus flash-attention training degradation." + "reason": "Generation kwargs handling after sync_gpus is unrelated to D-FINE auxiliary losses when denoising is off." }, { - "left": "issue:44360", - "right": "issue:44485", + "left": "issue:44295", + "right": "issue:44297", "accept": false, - "reason": "No shared bug; DSA indexer discussion is unrelated to GLM-5 RoPE behavior." + "reason": "Reading position_ids after buffer registration and tokenizer_class mismatch in save_pretrained are different tokenizer/model serialization issues." }, { - "left": "issue:43122", - "right": "issue:44521", + "left": "issue:44704", + "right": "issue:45290", "accept": false, - "reason": "Both involve tokenization/templating, but one is a generic tokenizer regression and the other is multimodal assistant-mask generation; different defects." + "reason": "Missing kwargs passthrough in AutoProcessor.from_pretrained and chat-template crashes on tool-call assistant messages are separate code paths." }, { - "left": "issue:43232", - "right": "issue:45245", + "left": "issue:44261", + "right": "issue:44373", "accept": false, - "reason": "Generation kwargs handling and category-count limits are unrelated." + "reason": "A config epsilon precision issue and a wrong docstring are not duplicates." }, { - "left": "issue:44038", - "right": "issue:44297", + "left": "issue:43723", + "right": "issue:44442", "accept": false, - "reason": "Different issues: Qwen3-VL-Moe compatibility versus tokenizer_class metadata saved in tokenizer_config.json." + "reason": "Both concern tokenizer loading, but one is a general v5 loading issue and the other is a specific FastSpeech2ConformerTokenizer failure; not enough evidence of the same bug." }, { - "left": "issue:28218", - "right": "issue:30990", + "left": "issue:42907", + "right": "issue:44945", "accept": false, - "reason": "Tokenizer spacing bug and Sentence Transformers load hang are unrelated." + "reason": "Saving dequantized Ministrals/Devstrals and incorrect output under pipeline parallelism are unrelated." }, { - "left": "issue:43122", + "left": "issue:43723", "right": "issue:44779", "accept": false, - "reason": "Both are v5 tokenizer regressions, but they concern different models and failure modes; not safely mergeable as one bug." - }, - { - "left": "issue:44038", - "right": "issue:44295", - "accept": false, - "reason": "Same broad family, but one is a model-loading issue and the other is a position_ids buffer access error." + "reason": "Both mention tokenizer regressions in v5, but they affect different tokenizer paths and symptoms; too broad to call the same bug." }, { "left": "issue:42907", - "right": "issue:43452", - "accept": false, - "reason": "Different affected APIs and different symptoms; no concrete same-bug evidence." - }, - { - "left": "issue:36296", - "right": "issue:43116", + "right": "issue:43937", "accept": false, - "reason": "Tensor-parallel training bug is unrelated to the example-script multi-label classification issue." + "reason": "Model-saving of dequantized weights is unrelated to GLM-5 GenerationConfig validation." }, { - "left": "issue:44038", - "right": "issue:44222", + "left": "issue:41628", + "right": "issue:45278", "accept": false, - "reason": "Both touch Qwen3-VL-Moe, but save_pretrained FP8 behavior and general v5 loading errors are distinct code paths." + "reason": "AutoImageProcessor import failure may be one example of import breakage, but 45278 is a broad collection of import errors after upgrade, not the same specific issue." }, { - "left": "issue:39401", - "right": "issue:44488", + "left": "issue:43756", + "right": "issue:44060", "accept": false, - "reason": "Different models and defects: wrong offset mapping versus inability to load a specific model." + "reason": "Smollm3 RoPE layer loss and Qwen3-Next tied-weights warnings are unrelated model-specific bugs." }, { - "left": "issue:45216", - "right": "issue:45478", + "left": "issue:43784", + "right": "issue:45600", "accept": false, - "reason": "Both concern Qwen3.5 and v5.4+, but one is save_pretrained checkpoint corruption and the other is from_pretrained loading error for MoE; not the same concrete bug." + "reason": "Missing nn during import and stale auto_mappings references to removed Sam3LiteText configs are different import/config cleanup problems." }, { - "left": "issue:39401", - "right": "issue:44521", + "left": "issue:43408", + "right": "issue:43475", "accept": false, - "reason": "Qwen3 tokenizer offset mapping and multimodal assistant_mask generation are different code paths." + "reason": "A model-type mismatch warning and a missing attribute on Sam3VisionEncoderOutput are separate SAM 3 issues." }, { - "left": "issue:43824", - "right": "issue:44351", + "left": "issue:43122", + "right": "issue:44514", "accept": false, - "reason": "Both are import errors, but for different symbols and different root causes." + "reason": "Tokenization drift across versions and a batched apply_chat_template crash with padding=False are different failures." }, { - "left": "issue:22355", - "right": "issue:28218", + "left": "issue:44625", + "right": "issue:44829", "accept": false, - "reason": "No meaningful overlap: missing transformers.onnx module versus extra-space tokenization after added token." + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 training degeneration are unrelated." }, { - "left": "issue:43784", - "right": "issue:45446", + "left": "issue:44704", + "right": "issue:44521", "accept": false, - "reason": "Different import failures in different libraries/code paths; not duplicates." + "reason": "kwargs forwarding to cached_file and zero assistant_masks for multimodal inputs are distinct chat/template behaviors." }, { - "left": "issue:22355", - "right": "issue:29159", + "left": "issue:44038", + "right": "issue:44297", "accept": false, - "reason": "Missing module import and slow/fast tokenizer inconsistency are unrelated." + "reason": "A Qwen3-VL-Moe bug and tokenizer save_pretrained metadata mismatch are not the same underlying problem." }, { - "left": "issue:42757", - "right": "issue:44351", + "left": "issue:43232", + "right": "issue:45245", "accept": false, - "reason": "Different packages and missing names: huggingface_hub offline mode versus transformers HybridCache." + "reason": "Generation config updates after sync_gpus and a categories-cardinality runtime error are unrelated." }, { - "left": "issue:43122", - "right": "issue:43906", + "left": "issue:28218", + "right": "issue:30990", "accept": false, - "reason": "The latter is a reproduction of a specific earlier issue; the former is a tokenizer regression across versions. Not the same artifact." + "reason": "Added-token spacing behavior and Sentence Transformers loading hangs are unrelated." }, { - "left": "issue:43992", - "right": "issue:44704", + "left": "issue:44038", + "right": "issue:44295", "accept": false, - "reason": "Different APIs and bugs: UMT5Encoder weight loading versus AutoProcessor kwargs forwarding." + "reason": "Qwen3-VL-Moe behavior and position_ids buffer access errors are different bugs." }, { - "left": "issue:43066", + "left": "issue:43122", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions in v5, but one is decoder-type metadata and the other is incorrect Deepseek tokenization; not enough to call them the same bug." + "reason": "Both are tokenizer regressions, but they concern different models and different observable failures; too weak to merge as the same bug." } ] }, @@ -13265,72 +13220,86 @@ "component_index": 1, "component_count": 18, "review_unit_index": 42, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:22355", + "issue:28218", "issue:29127", + "issue:29159", "issue:33453", - "issue:38617", + "issue:36296", "issue:39401", - "issue:42175", "issue:42222", "issue:42371", + "issue:42757", "issue:42831", - "issue:43064", + "issue:42907", + "issue:43066", "issue:43116", - "issue:43381", + "issue:43122", "issue:43441", - "issue:43450", + "issue:43452", "issue:43493", "issue:43611", - "issue:43673", + "issue:43784", + "issue:43824", "issue:43825", - "issue:43827", "issue:43874", + "issue:43906", "issue:43937", "issue:43976", - "issue:43986", + "issue:43992", "issue:44038", "issue:44075", - "issue:44117", + "issue:44222", "issue:44297", + "issue:44351", "issue:44355", - "issue:44380", - "issue:44466", "issue:44485", - "issue:44625", + "issue:44488", + "issue:44521", + "issue:44704", "issue:44779", "issue:44936", - "issue:44938", - "issue:44945", - "issue:44977", "issue:45083", - "issue:45278", + "issue:45216", "issue:45357", "issue:45446", "issue:45478", "issue:45563", - "issue:45588", "issue:45753", "issue:46002", - "issue:46077", "issue:46231" ], "soft_pairs": [ + "issue:42907|issue:43452", + "issue:36296|issue:43116", + "issue:44038|issue:44222", + "issue:39401|issue:44488", + "issue:45216|issue:45478", + "issue:39401|issue:44521", + "issue:43824|issue:44351", + "issue:22355|issue:28218", + "issue:43784|issue:45446", + "issue:22355|issue:29159", + "issue:42757|issue:44351", + "issue:43122|issue:43906", + "issue:43992|issue:44704", + "issue:43066|issue:44779", "issue:44355|issue:46231", "issue:43937|issue:44485", "issue:43611|issue:46002", "issue:42371|issue:43874", "issue:42222|issue:43441", "issue:44075|issue:44936", - "issue:43976|issue:45446", "issue:42831|issue:44297", + "issue:43976|issue:45446", "issue:45357|issue:45478", "issue:44038|issue:45478", "issue:43116|issue:43825", @@ -13339,229 +13308,232 @@ "issue:43116|issue:43493", "issue:43937|issue:45563", "issue:22355|issue:29127", - "issue:33453|issue:39401", - "issue:44625|issue:46002", - "issue:38617|issue:45278", - "issue:39401|issue:44779", - "issue:44380|issue:46077", - "issue:42175|issue:44938", - "issue:44117|issue:45588", - "issue:43064|issue:43673", - "issue:43381|issue:44945", - "issue:43827|issue:45446", - "issue:43450|issue:43986", - "issue:43937|issue:44977", - "issue:44466|issue:46002" + "issue:33453|issue:39401" ], - "prepared_review_unit_hash": "328be6552c0bc8b7cbe68e2f604e96de12ec9eab6b5a05a4157d1007bf41bf31", + "prepared_review_unit_hash": "733e2a3a8ac1d1a6d9d3ba728380828e2ca30bfae54a3f6dd4837789ebca2f06", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36047, - "estimated_input_tokens": 9012, - "estimated_eval_tokens": 18280 + "soft_pair_count": 31, + "serialized_chars": 36429, + "estimated_input_tokens": 9108, + "estimated_eval_tokens": 18472 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: import-path breakages, tokenizer regressions, model-specific failures, training/eval bugs, docs issues, and config problems. There is no single underlying bug or change that ties these issues together, so it should not be collapsed into one duplicate set.", - "confidence": 0.95, - "canonical_issue_reason": "No single canonical issue fits this cluster; the reports span unrelated subsystems and failure modes. If a representative is needed, issue 45278 is the broadest umbrella report, but it still does not cover the rest.", + "summary": "This set is mostly heterogeneous: tokenizer regressions, model loading/saving bugs, import errors, and unrelated training/runtime issues. The only clear duplicate-like matches are the Qwen3.5 save/load serialization regressions.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45216 is the best anchor for the Qwen3.5 checkpoint/save_pretrained regression family: it is broad, regression-tagged, and directly explains the related from_pretrained failures.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45278 is the most umbrella-like report among the set, but it is still too generic to serve as a true canonical for these unrelated bugs.", + "best_issue_reason": "issue:45216 is the strongest global issue candidate because it captures the underlying serialization bug more generally than the later, narrower symptom reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44355", - "right": "issue:46231", + "left": "issue:42907", + "right": "issue:43452", "accept": false, - "reason": "Compiled-Python import/runtime errors are unrelated to the set_training() doc parameter mismatch." + "reason": "Different code paths: dequantized model saving vs gguf loading failure." }, { - "left": "issue:43937", - "right": "issue:44485", + "left": "issue:36296", + "right": "issue:43116", "accept": false, - "reason": "GenerationConfig validation failure and GLM-5 RoPE design/discussion are different problems." + "reason": "Tensor-parallel training bug and multi-label example-script behavior are unrelated." }, { - "left": "issue:43611", - "right": "issue:46002", + "left": "issue:44038", + "right": "issue:44222", "accept": false, - "reason": "base_model_prefix loading regression is unrelated to NaNs in classification heads after checkout/init." + "reason": "Both mention MoE, but one is a Qwen3-VL-Moe runtime issue and the other is FP8 save_pretrained; not the same bug." }, { - "left": "issue:42371", - "right": "issue:43874", + "left": "issue:39401", + "right": "issue:44488", "accept": false, - "reason": "TF32 behavior warnings are unrelated to a missing image-processor method causing multimodal token errors." + "reason": "Wrong offset_mapping in Qwen3 tokenizer vs a separate model-loading failure for cjvt/sleng-bert." }, { - "left": "issue:42222", - "right": "issue:43441", + "left": "issue:45216", + "right": "issue:45478", + "accept": true, + "reason": "Same Qwen3.5 save/load serialization regression: incorrect saved checkpoint content leads to from_pretrained failure." + }, + { + "left": "issue:39401", + "right": "issue:44521", "accept": false, - "reason": "Broken ViTPose model files and Ministral FlashAttention failure are different model-specific issues." + "reason": "Offset mapping bug and all-zero assistant masks are different tokenizer/template problems." }, { - "left": "issue:44075", - "right": "issue:44936", + "left": "issue:43824", + "right": "issue:44351", "accept": false, - "reason": "Unused SGD args in optimizer setup is unrelated to trainer.evaluate() failing after trainer.train()." + "reason": "Different missing symbols: Qwen2_5_VLForConditionalGeneration vs HybridCache." }, { - "left": "issue:43976", - "right": "issue:45446", + "left": "issue:22355", + "right": "issue:28218", "accept": false, - "reason": "Python version compatibility in 5.1.0 is unrelated to the AuxRequest version-check bug in flex_attention." + "reason": "ONNX import error and extra-space-after-added-token tokenizer bug are unrelated." }, { - "left": "issue:42831", - "right": "issue:44297", + "left": "issue:43784", + "right": "issue:45446", "accept": false, - "reason": "FineGrainedFP8 accuracy issues and tokenizer save_pretrained metadata mismatch are unrelated." + "reason": "Sentence-transformers import NameError is not the same as the flex_attention PyTorch version check bug." }, { - "left": "issue:45357", - "right": "issue:45478", + "left": "issue:22355", + "right": "issue:29159", "accept": false, - "reason": "Qwen3.5 visual encoder key serialization and Qwen3.5 Moe from_pretrained loading are distinct bugs." + "reason": "ONNX module import failure and slow/fast tokenizer inconsistency are unrelated." }, { - "left": "issue:44038", - "right": "issue:45478", + "left": "issue:42757", + "right": "issue:44351", "accept": false, - "reason": "Qwen3-VL-Moe generation bug and Qwen3.5 Moe from_pretrained error are different code paths." + "reason": "Both are import errors, but they concern different missing APIs and different failure causes." }, { - "left": "issue:43116", - "right": "issue:43825", + "left": "issue:43122", + "right": "issue:43906", "accept": false, - "reason": "Multi-label classification script behavior and pipeline() translation-task messaging are unrelated." + "reason": "An isolated reproduction can be related, but there is not enough evidence that it is the same concrete tokenizer bug." }, { - "left": "issue:43493", - "right": "issue:45753", + "left": "issue:43992", + "right": "issue:44704", "accept": false, - "reason": "SigLIP2 implementation discrepancy and Qwen3_5 infinite-loop behavior are different model regressions." + "reason": "Missing embed_tokens.weight in UMT5Encoder and AutoProcessor kwargs forwarding are different loading issues." }, { - "left": "issue:45083", - "right": "issue:45446", + "left": "issue:43066", + "right": "issue:44779", "accept": false, - "reason": "qwen3_omni_moe feature-length helper behavior is unrelated to the flex_attention AuxRequest import check." + "reason": "Wrong tokenizer decoder type and Deepseek tokenization regression are distinct symptoms/code paths." }, { - "left": "issue:43116", - "right": "issue:43493", + "left": "issue:44355", + "right": "issue:46231", "accept": false, - "reason": "Example-script classification output and SigLIP2 JAX-vs-HF discrepancy are unrelated." + "reason": "Compiled-Python runtime errors and a documentation parameter-name mismatch are unrelated." }, { "left": "issue:43937", - "right": "issue:45563", + "right": "issue:44485", "accept": false, - "reason": "Invalid GenerationConfig and stale paged-generate warning for num_return_sequences are different issues." + "reason": "GLM-5 generation config validation and GLM-5 RoPE discussion are not the same bug." }, { - "left": "issue:22355", - "right": "issue:29127", + "left": "issue:43611", + "right": "issue:46002", "accept": false, - "reason": "Missing transformers.onnx module import and LayoutLMv3 error-message clarity are unrelated." + "reason": "base_model_prefix loading breakage and NaNs in classification heads are unrelated." }, { - "left": "issue:33453", - "right": "issue:39401", + "left": "issue:42371", + "right": "issue:43874", "accept": false, - "reason": "Tokenizer loading regression and Qwen3 offset_mapping bug are separate tokenizer issues with different symptoms." + "reason": "TF32 API guidance issue and missing get_number_of_image_patches AttributeError are unrelated." }, { - "left": "issue:44625", - "right": "issue:46002", + "left": "issue:42222", + "right": "issue:43441", "accept": false, - "reason": "Qwen3.5 num_labels propagation and NaNs in classification heads are not the same underlying bug." + "reason": "Vitpose import breakage and Ministral-3 FlashAttention failure are unrelated." }, { - "left": "issue:38617", - "right": "issue:45278", + "left": "issue:44075", + "right": "issue:44936", "accept": false, - "reason": "Missing layer_type_validation import and broad import errors after upgrade are related only at a very high level, not the same concrete bug." + "reason": "SGD args not used and trainer.evaluate() failing after training are different trainer/optimizer bugs." }, { - "left": "issue:39401", - "right": "issue:44779", + "left": "issue:42831", + "right": "issue:44297", "accept": false, - "reason": "Qwen3 offset_mapping errors and Deepseek tokenizer incorrect results are different tokenizer regressions." + "reason": "FineGrainedFP8 accuracy issue and tokenizer_class mismatch on save_pretrained are unrelated." }, { - "left": "issue:44380", - "right": "issue:46077", + "left": "issue:43976", + "right": "issue:45446", "accept": false, - "reason": "GPT2 attention scaling with SDPA/FlashAttention and Gemma4 bidirectional mask construction are unrelated attention bugs." + "reason": "Python version compatibility regression and flex_attention version-check bug are unrelated." }, { - "left": "issue:42175", - "right": "issue:44938", - "accept": false, - "reason": "Missing TensorFlow backend in pip extra and Python 3.14 load failure are different packaging/runtime problems." + "left": "issue:45357", + "right": "issue:45478", + "accept": true, + "reason": "Same Qwen3.5 save/load serialization family: incorrect visual-encoder keys in save_pretrained plausibly cause the from_pretrained error." }, { - "left": "issue:44117", - "right": "issue:45588", + "left": "issue:44038", + "right": "issue:45478", "accept": false, - "reason": "Tokenizer mapping None handling and flash_attention s_aux=None crash are unrelated code paths." + "reason": "Both involve Qwen3.5/Qwen3.5-VL-Moe, but one is a transformer/model bug and the other is a save/load serialization failure." }, { - "left": "issue:43064", - "right": "issue:43673", + "left": "issue:43116", + "right": "issue:43825", "accept": false, - "reason": "FSDP2/PEFT optimizer-state mismatch and GenerationMixin cache missing during chunked_prefill are different training vs generation bugs." + "reason": "Multi-label example script output and pipeline translation-task messaging are unrelated." }, { - "left": "issue:43381", - "right": "issue:44945", + "left": "issue:43493", + "right": "issue:45753", "accept": false, - "reason": "Gradient checkpointing eval-mode limitation and pipeline parallelism output corruption are unrelated execution-mode issues." + "reason": "SigLIP2 implementation discrepancy and Qwen3_5 infinite-loop-on-image are unrelated multimodal bugs." }, { - "left": "issue:43827", + "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing pipeline() and flex_attention AuxRequest version check are not the same bug." + "reason": "Qwen3_omni_moe feature-length helper behavior and flex_attention PyTorch version check are unrelated." }, { - "left": "issue:43450", - "right": "issue:43986", + "left": "issue:43116", + "right": "issue:43493", "accept": false, - "reason": "Batched video processor shape bug and AutoProcessor crash without torchvision are different video-processing problems." + "reason": "Multi-label classification example behavior and SigLIP2 implementation discrepancy are unrelated." }, { "left": "issue:43937", - "right": "issue:44977", + "right": "issue:45563", "accept": false, - "reason": "GenerationConfig validation and Qwen3.5 flash-attention generation failure are unrelated." + "reason": "GLM-5 generation-config validation and paged generate stale warning are unrelated generation issues." }, { - "left": "issue:44466", - "right": "issue:46002", + "left": "issue:22355", + "right": "issue:29127", "accept": false, - "reason": "lm_head.weight serialization inconsistency and NaNs in classification heads are different model-state issues." + "reason": "ONNX import failure and layoutlmv3 error-message clarity are unrelated." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression and Qwen3 tokenizer offset_mapping bug are different tokenizer failures." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analyst output is grounded in the packet and the negative pairings stay appropriately conservative. The only accepted matches are the Qwen3.5 save/load regression family, which is well-supported by the issue titles, and the overall heterogeneity summary fits the set." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -13570,46 +13542,46 @@ "component_index": 1, "component_count": 18, "review_unit_index": 43, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:38617", "issue:39401", "issue:41628", "issue:42175", "issue:42222", - "issue:42371", "issue:42673", - "issue:43012", + "issue:43064", "issue:43066", - "issue:43097", "issue:43122", + "issue:43381", "issue:43408", "issue:43421", - "issue:43577", + "issue:43450", + "issue:43673", "issue:43742", - "issue:43874", + "issue:43827", "issue:43937", - "issue:44075", + "issue:43986", "issue:44117", - "issue:44220", - "issue:44261", - "issue:44263", "issue:44291", + "issue:44380", "issue:44410", + "issue:44466", "issue:44483", "issue:44484", - "issue:44485", - "issue:44493", "issue:44496", + "issue:44625", + "issue:44779", "issue:44829", "issue:44877", - "issue:44964", + "issue:44938", + "issue:44945", "issue:44977", "issue:45071", "issue:45072", @@ -13620,14 +13592,26 @@ "issue:45405", "issue:45446", "issue:45464", - "issue:45478", - "issue:45542", - "issue:45656", + "issue:45588", + "issue:46002", + "issue:46077", "issue:46178", "issue:46231", "issue:46298" ], "soft_pairs": [ + "issue:44625|issue:46002", + "issue:38617|issue:45278", + "issue:44380|issue:46077", + "issue:39401|issue:44779", + "issue:42175|issue:44938", + "issue:44117|issue:45588", + "issue:43064|issue:43673", + "issue:43381|issue:44945", + "issue:43827|issue:45446", + "issue:43450|issue:43986", + "issue:43937|issue:44977", + "issue:44466|issue:46002", "issue:45405|issue:45446", "issue:44117|issue:45081", "issue:43122|issue:46298", @@ -13636,237 +13620,211 @@ "issue:44291|issue:45071", "issue:44829|issue:45216", "issue:39401|issue:43421", - "issue:42222|issue:46178", "issue:38617|issue:41628", + "issue:42222|issue:46178", "issue:43408|issue:43742", "issue:44877|issue:46231", "issue:44484|issue:45310", "issue:43937|issue:44496", "issue:43066|issue:43122", - "issue:44483|issue:45464", - "issue:44964|issue:45278", - "issue:44075|issue:45656", - "issue:42371|issue:44261", - "issue:42222|issue:45542", - "issue:43012|issue:43577", - "issue:44410|issue:45478", - "issue:38617|issue:44220", - "issue:42175|issue:42222", - "issue:43097|issue:44493", - "issue:43874|issue:44261", - "issue:41628|issue:45071", - "issue:44263|issue:44485", - "issue:45071|issue:45542", - "issue:42673|issue:44977" + "issue:44483|issue:45464" ], - "prepared_review_unit_hash": "114c6ab7eff91703b45fa5e2eddebd2e2d393961f53d8b4c63db77aa24e335d3", + "prepared_review_unit_hash": "244ca988404866e3980c46494b9f40167bf662ef513d9f9be2153b5c4eb9749d", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36590, - "estimated_input_tokens": 9148, - "estimated_eval_tokens": 18552 + "soft_pair_count": 28, + "serialized_chars": 36088, + "estimated_input_tokens": 9022, + "estimated_eval_tokens": 18300 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a loose collection of transformer regressions and model-specific bugs rather than true duplicates. I did not find any soft pair that clearly describes the same underlying bug/change.", - "confidence": 0.58, - "canonical_issue_reason": "issue:45071 is the most central, broadly scoped regression in the set: a v5.4.0 PreTrainedConfig/type-checking break that can affect multiple load paths, making it the best anchor among mostly model-specific reports.", + "summary": "This cluster is a loose mix of Transformers v5 regressions (imports, tokenization, loading, training, and API behavior), but the individual reports are mostly distinct rather than true duplicates. The best anchor issue is the broad v5 upgrade regression report.", + "confidence": 0.36, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report in the set, covering multiple import failures after upgrading to v5.5.0, which makes it the closest representative of the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45071 has the clearest cross-cutting technical impact and is more representative than the many narrow model-specific or documentation-only reports.", + "best_issue_reason": "issue:45278 is the most general and cluster-like issue; it can summarize a subset of the v5 regression/import breakage reports better than the highly specific model- or feature-level bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45405", - "right": "issue:45446", - "accept": false, - "reason": "Different failure modes: one is a packaging/version floor bump for PEFT, the other is a PyTorch version check bug in flex_attention import logic." - }, - { - "left": "issue:44117", - "right": "issue:45081", - "accept": false, - "reason": "Both involve tokenizer internals, but one is TOKENIZER_MAPPING_NAMES returning None while the other is a Mistral regex patch crash on a Tokenizer object; different code paths." - }, - { - "left": "issue:43122", - "right": "issue:46298", + "left": "issue:44625", + "right": "issue:46002", "accept": false, - "reason": "Version-dependent tokenization regression vs a noisy logging message; no shared bug." + "reason": "Different symptoms and code paths: Qwen3.5 config propagation vs NaNs in classification heads after checkout/init." }, { - "left": "issue:42673", - "right": "issue:45072", + "left": "issue:38617", + "right": "issue:45278", "accept": false, - "reason": "Unrelated problems: VRAM leak in Qwen3ForCausalLM versus dtype mismatch in SwitchTransformers/TimmWrapperModel inference." + "reason": "Both are import-related, but one is a specific missing symbol and the other is a broad upgrade-regression report; not clearly the same underlying bug." }, { - "left": "issue:44410", - "right": "issue:45310", + "left": "issue:44380", + "right": "issue:46077", "accept": false, - "reason": "Both are Qwen-family issues, but one is missing projections in qwen3next and the other is a Qwen3.5 MoE from_pretrained load error; not the same defect." + "reason": "Both involve attention behavior, but the model, backend, and failure mode differ: scaling ignored vs causal masks still built." }, { - "left": "issue:44291", - "right": "issue:45071", + "left": "issue:39401", + "right": "issue:44779", "accept": false, - "reason": "Both are v5/loading regressions, but one is an init_empty_weights argument mismatch and the other is PretrainedConfig type checking; different call sites and fixes." + "reason": "Both are tokenizer regressions, but they concern different models and different observable failures (offset mapping vs tokenization output)." }, { - "left": "issue:44829", - "right": "issue:45216", + "left": "issue:42175", + "right": "issue:44938", "accept": false, - "reason": "Flash-attention training degeneration and save_pretrained checkpoint corruption are separate behaviors." + "reason": "Unrelated environment/dependency problems: missing TensorFlow in an install extra vs Python 3.14 load failure." }, { - "left": "issue:39401", - "right": "issue:43421", + "left": "issue:44117", + "right": "issue:45588", "accept": false, - "reason": "Wrong offset mapping in Qwen3 tokenizer is unrelated to runtime special-token/post-processor update behavior." + "reason": "Different subsystems and failures: tokenizer mapping lookup vs flash_attention crash on missing s_aux." }, { - "left": "issue:42222", - "right": "issue:46178", + "left": "issue:43064", + "right": "issue:43673", "accept": false, - "reason": "Different backend/model failures: vitpose model breakage versus tensorflow_text not being registered in BACKENDS_MAPPING." + "reason": "Training/FSDP optimizer-state bug vs generation cache missing during chunked prefill; no shared concrete code-path." }, { - "left": "issue:38617", - "right": "issue:41628", + "left": "issue:43381", + "right": "issue:44945", "accept": false, - "reason": "Both are import errors, but they refer to different missing symbols from different modules and likely distinct regressions." + "reason": "Gradient checkpointing eval-mode issue is unrelated to pipeline-parallel incorrect output." }, { - "left": "issue:43408", - "right": "issue:43742", + "left": "issue:43827", + "right": "issue:45446", "accept": false, - "reason": "Model-type warning for sam3_video/sam3_tracker is unrelated to a KeyError loading MobileLLM-125M." + "reason": "Docs cleanup vs a PyTorch version-check bug in flex_attention; not the same change or bug." }, { - "left": "issue:44877", - "right": "issue:46231", + "left": "issue:43450", + "right": "issue:43986", "accept": false, - "reason": "Strict config loading failure is not the same as a documentation typo in set_training." + "reason": "Both are video-related, but one is a batched-shape bug and the other is a crash due to missing torchvision." }, { - "left": "issue:44484", - "right": "issue:45310", + "left": "issue:43937", + "right": "issue:44977", "accept": false, - "reason": "save_pretrained shard-size behavior is unrelated to the Qwen3.5 MoE from_pretrained error." + "reason": "Different model-specific generation failures: invalid GenerationConfig vs flash-attention generation breakage." }, { - "left": "issue:43937", - "right": "issue:44496", + "left": "issue:44466", + "right": "issue:46002", "accept": false, - "reason": "Invalid GenerationConfig for GLM-5 and unrecognized-model config.json errors are different load-time failures." + "reason": "Serialization of tied weights depending on device is unrelated to NaNs in classification heads." }, { - "left": "issue:43066", - "right": "issue:43122", + "left": "issue:45405", + "right": "issue:45446", "accept": false, - "reason": "Tokenizer decoder-type mismatch and differing tokenization output across versions are related only broadly; the concrete bug and fix path are not the same." + "reason": "PEFT version bump/release issue is unrelated to the flex_attention import-version check." }, { - "left": "issue:44483", - "right": "issue:45464", + "left": "issue:44117", + "right": "issue:45081", "accept": false, - "reason": "Both mention chat/completions, but one is a request acceptance issue in v5.3 and the other is streaming inference failure on Qwen3.5-0.8B." + "reason": "Both touch tokenizer loading, but the actual defects differ: mapping names None vs backend_tokenizer attribute access in Mistral regex patching." }, { - "left": "issue:44964", - "right": "issue:45278", + "left": "issue:43122", + "right": "issue:46298", "accept": false, - "reason": "A specific Phi-4 multimodal load problem is not the same as generalized import errors after upgrading 4.57.0 to 5.5.0." + "reason": "Tokenization regression vs noisy logging message; clearly unrelated." }, { - "left": "issue:44075", - "right": "issue:45656", + "left": "issue:42673", + "right": "issue:45072", "accept": false, - "reason": "SGD argument handling and double optimizer.step() under deepspeed are distinct optimizer bugs." + "reason": "VRAM leak in multi-threaded dataloader use is unrelated to dtype mismatches in inference." }, { - "left": "issue:42371", - "right": "issue:44261", + "left": "issue:44410", + "right": "issue:45310", "accept": false, - "reason": "TF32 API guidance and MLA q_a_layernorm rms_norm_eps precision issues do not share a concrete code path." + "reason": "Different Qwen variants and different failures: missing projections vs from_pretrained loading error." }, { - "left": "issue:42222", - "right": "issue:45542", + "left": "issue:44291", + "right": "issue:45071", "accept": false, - "reason": "Model breakage in vitpose is unrelated to backend detection when TensorFlow is absent but tensorboard is installed." + "reason": "Unexpected _is_hf_initialized argument in init_empty_weights is a different problem from PretrainedConfig type-checking breakage." }, { - "left": "issue:43012", - "right": "issue:43577", + "left": "issue:44829", + "right": "issue:45216", "accept": false, - "reason": "A PyTorch compile warning on bfloat16 models and dtype fields staying float32 when loading Blip2 are different precision issues." + "reason": "Training degenerate loss with flash_attention_3 is unrelated to checkpoint serialization regression." }, { - "left": "issue:44410", - "right": "issue:45478", + "left": "issue:39401", + "right": "issue:43421", "accept": false, - "reason": "Different Qwen3-related bugs: missing projections in qwen3next versus Qwen3.5 MoE from_pretrained failure." + "reason": "Offset-mapping bug and runtime special-token/post-processor update bug are related to tokenizers generally but not the same underlying defect." }, { "left": "issue:38617", - "right": "issue:44220", + "right": "issue:41628", "accept": false, - "reason": "ImportError for layer_type_validation and a feature extraction bug in _torch_extract_fbank_features() are unrelated." + "reason": "Both are import errors, but they refer to different missing symbols and likely different fixes." }, { - "left": "issue:42175", - "right": "issue:42222", + "left": "issue:42222", + "right": "issue:46178", "accept": false, - "reason": "Backend installation metadata and vitpose model breakage are different problems." + "reason": "vitpose model breakage and tensorflow_text backend mapping failure are unrelated." }, { - "left": "issue:43097", - "right": "issue:44493", + "left": "issue:43408", + "right": "issue:43742", "accept": false, - "reason": "Removal of tie_embeddings_and_encoder_decoder and unexpected position-id keys are both version-regression flavored, but the concrete breakage differs." + "reason": "Model-type mismatch warning for sam3_tracker is unrelated to a MobileLLM key error." }, { - "left": "issue:43874", - "right": "issue:44261", + "left": "issue:44877", + "right": "issue:46231", "accept": false, - "reason": "Missing get_number_of_image_patches in Glm46VImageProcessorFast and missing rms_norm_eps in MLA q_a_layernorm are unrelated multimodal/model issues." + "reason": "Strict config loading and a docs parameter-name mismatch are not the same bug." }, { - "left": "issue:41628", - "right": "issue:45071", + "left": "issue:44484", + "right": "issue:45310", "accept": false, - "reason": "An AutoImageProcessor import failure is not the same as a PreTrainedConfig type-checking regression." + "reason": "Save_pretrained shard-size question is unrelated to Qwen3.5 from_pretrained loading failures." }, { - "left": "issue:44263", - "right": "issue:44485", + "left": "issue:43937", + "right": "issue:44496", "accept": false, - "reason": "A torch.split return-value issue in GlmMoeDsaIndexer is unrelated to GLM-5 RoPE implementation." + "reason": "Both are load/generation-related, but one is a GenerationConfig validity error and the other is an unrecognized-model/config.json problem." }, { - "left": "issue:45071", - "right": "issue:45542", + "left": "issue:43066", + "right": "issue:43122", "accept": false, - "reason": "PretrainedConfig type-checking regression and TensorFlow backend detection are separate subsystems." + "reason": "Both are v5 tokenizer regressions, but one is decoder type selection and the other is changed tokenization output; not the same concrete bug." }, { - "left": "issue:42673", - "right": "issue:44977", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "Qwen3ForCausalLM VRAM leak and Qwen3.5 flash-attention generation failure are different bugs in different execution paths." + "reason": "Both mention chat/completions, but one is a general request rejection and the other is a model-specific streaming inference failure." } ] }, @@ -13882,63 +13840,77 @@ "component_index": 1, "component_count": 18, "review_unit_index": 44, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:38617", + "issue:41628", + "issue:42175", + "issue:42222", "issue:42371", "issue:42673", - "issue:42757", "issue:42831", - "issue:42890", "issue:43012", - "issue:43064", "issue:43097", "issue:43335", "issue:43408", "issue:43441", "issue:43452", - "issue:43502", - "issue:43873", + "issue:43577", "issue:43874", "issue:43906", "issue:43931", "issue:43976", "issue:43986", - "issue:43992", "issue:44038", - "issue:44062", + "issue:44075", "issue:44117", + "issue:44220", + "issue:44261", "issue:44263", - "issue:44315", - "issue:44351", "issue:44360", "issue:44368", "issue:44410", "issue:44484", + "issue:44485", + "issue:44493", "issue:44829", "issue:44933", "issue:44945", + "issue:44964", + "issue:44977", "issue:44987", "issue:45071", - "issue:45200", - "issue:45341", - "issue:45375", + "issue:45278", "issue:45440", "issue:45478", + "issue:45542", "issue:45588", - "issue:45859", + "issue:45656", "issue:46050", "issue:46172", - "issue:46231", - "issue:46298" + "issue:46231" ], "soft_pairs": [ + "issue:44075|issue:45656", + "issue:44964|issue:45278", + "issue:42371|issue:44261", + "issue:42222|issue:45542", + "issue:43012|issue:43577", + "issue:44410|issue:45478", + "issue:38617|issue:44220", + "issue:42175|issue:42222", + "issue:43874|issue:44261", + "issue:43097|issue:44493", + "issue:41628|issue:45071", + "issue:44263|issue:44485", + "issue:45071|issue:45542", + "issue:42673|issue:44977", "issue:42673|issue:43906", "issue:44263|issue:46172", "issue:43012|issue:43408", @@ -13957,234 +13929,234 @@ "issue:43986|issue:44484", "issue:45440|issue:46172", "issue:43097|issue:46050", - "issue:43097|issue:45588", - "issue:43064|issue:44315", - "issue:42673|issue:45200", - "issue:38617|issue:42757", - "issue:44062|issue:44117", - "issue:44410|issue:45375", - "issue:42890|issue:46298", - "issue:44038|issue:44410", - "issue:43931|issue:44410", - "issue:43502|issue:43992", - "issue:45341|issue:45859", - "issue:43097|issue:43873", - "issue:38617|issue:44351" + "issue:43097|issue:45588" ], - "prepared_review_unit_hash": "a45f8c669aa12fc792c5943c03dd5efa33f9e031cc19ff21b366364620399288", + "prepared_review_unit_hash": "5d05464b4f90a55d125641f12af2268869f06cd700d584333ffcf10df06d7723", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 31, - "serialized_chars": 36305, - "estimated_input_tokens": 9077, - "estimated_eval_tokens": 18410 + "soft_pair_count": 33, + "serialized_chars": 36829, + "estimated_input_tokens": 9208, + "estimated_eval_tokens": 18672 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The batch is highly heterogeneous: import/API regressions, model-specific loading bugs, warnings, config quirks, and docs/test issues. I don\u2019t see any convincing duplicate pairs; all soft edges should be rejected.", - "confidence": 0.97, - "canonical_issue_reason": "No true canonical duplicate cluster exists here. If forced to pick an anchor, issue 43097 is the best representative because it has inbound references and describes a concrete, broadly impactful v5 regression with downstream discussion.", + "summary": "The cluster is heterogeneous, with mostly unrelated transformer regressions and API/behavior issues. None of the soft pairs look like the same underlying bug or fix, so I would not merge any of them as duplicates.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45278 is the broadest and most umbrella-like issue in the set, centered on widespread import breakage after upgrading Transformers, which makes it the best representative target for this mixed cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43097 is the strongest single issue to anchor this set: it is discussion-backed, has inbound references, and is a concrete regression rather than a narrow model-only or test-only problem.", + "best_issue_reason": "issue:45278 best captures the general user-facing regression theme in the cluster and is broad enough to subsume related import/load failures better than the narrower, model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42673", - "right": "issue:43906", + "left": "issue:44075", + "right": "issue:45656", "accept": false, - "reason": "VRAM leak in multi-threaded Qwen3ForCausalLM vs an isolated reproduction of another issue; no shared code-path or symptom match." + "reason": "Both concern optimizers, but one says SGD args are ignored while the other says deepspeed calls optimizer.step twice; different failure modes and fixes." }, { - "left": "issue:44263", - "right": "issue:46172", + "left": "issue:44964", + "right": "issue:45278", "accept": false, - "reason": "A torch.split/indexer bug and a DeepSeek-V4 transpose/cast bug are unrelated model internals." + "reason": "One is a specific Phi-4 multimodal loading failure; the other is a broad post-upgrade import-error report. Same area, not the same bug." }, { - "left": "issue:43012", - "right": "issue:43408", + "left": "issue:42371", + "right": "issue:44261", "accept": false, - "reason": "Both mention warnings, but one is a bfloat16/PyTorch precision warning and the other is a model-type mismatch warning." + "reason": "TF32 API guidance and an MLA rms_norm_eps precision issue are unrelated problems." }, { - "left": "issue:44360", - "right": "issue:46172", + "left": "issue:42222", + "right": "issue:45542", "accept": false, - "reason": "DSA indexer ReLU discussion is unrelated to DeepSeek-V4 tensor transpose/casting." + "reason": "VitPose breakage and missing TensorFlow backend from a tensorboard-only install are different issues." }, { - "left": "issue:44484", - "right": "issue:45478", + "left": "issue:43012", + "right": "issue:43577", "accept": false, - "reason": "Save_pretrained shard-size policy question vs Qwen3.5-Moe from_pretrained error; different behavior and code paths." + "reason": "Both mention dtype/precision, but one is a compile warning on bfloat16 models and the other is a Blip2 loading bug where dtype fields stay float32." }, { - "left": "issue:43441", - "right": "issue:44829", + "left": "issue:44410", + "right": "issue:45478", "accept": false, - "reason": "Both involve flash attention, but one is an inference failure on Ministral-3 and the other is degenerate training with flash_attention_3." + "reason": "Both involve Qwen3-family models, but one is a missing projection in qwen3next and the other is a Qwen3.5 MoE from_pretrained error; not the same code path." }, { - "left": "issue:42371", - "right": "issue:42831", + "left": "issue:38617", + "right": "issue:44220", "accept": false, - "reason": "TF32 configuration guidance and FineGrainedFP8 accuracy regression are distinct issues." + "reason": "An import error from configuration_utils and a feature-extraction fbank issue are unrelated." }, { - "left": "issue:43931", - "right": "issue:44038", + "left": "issue:42175", + "right": "issue:42222", "accept": false, - "reason": "Different Qwen3-VL failures with different model families and failure modes; not the same bug." + "reason": "TensorFlow backend packaging and broken VitPose models are different bugs." }, { - "left": "issue:43335", - "right": "issue:45071", + "left": "issue:43874", + "right": "issue:44261", "accept": false, - "reason": "Sparse-layer config bug vs PretrainedConfig type-checking breakage; unrelated regressions." + "reason": "A missing image-patch helper in a fast image processor and an MLA layernorm epsilon precision issue do not share the same underlying defect." }, { "left": "issue:43097", - "right": "issue:43452", + "right": "issue:44493", "accept": false, - "reason": "Embedding-tie config removal is unrelated to gguf_file loading failures in tokenizer/model from_pretrained." + "reason": "One is a config removal/regression around tied embeddings, the other is unexpected position-id key warnings; different symptoms and fixes." }, { - "left": "issue:43874", - "right": "issue:44263", + "left": "issue:41628", + "right": "issue:45071", "accept": false, - "reason": "Missing image-patch helper in GLM46V is unrelated to a torch.split return-value issue in GlmMoeDsaIndexer." + "reason": "Missing AutoImageProcessor import and PretrainedConfig type-checking regressions are unrelated API breakages." }, { - "left": "issue:44933", - "right": "issue:46231", + "left": "issue:44263", + "right": "issue:44485", "accept": false, - "reason": "Missing import from image_utils and a documentation parameter-name mismatch are different classes of issues." + "reason": "Both mention GLM/DSA, but one is about torch.split return values and the other is a RoPE implementation bug." }, { - "left": "issue:44117", - "right": "issue:44987", + "left": "issue:45071", + "right": "issue:45542", "accept": false, - "reason": "Tokenizer mapping None-handling and a model loading failure for physical-intelligence/fast are not the same underlying defect." + "reason": "PretrainedConfig type checking and an undefined tf backend error from tensorboard-only installs are separate problems." }, { - "left": "issue:43408", - "right": "issue:44368", + "left": "issue:42673", + "right": "issue:44977", "accept": false, - "reason": "Model-type warning for sam3_video/sam3_tracker vs a tie_word_embeddings warning in Qwen3.5 LoRA fine-tuning." + "reason": "VRAM leakage in multi-threaded dataloaders and Qwen3.5 flash-attention generation failure are different bugs." }, { - "left": "issue:43976", - "right": "issue:44945", + "left": "issue:42673", + "right": "issue:43906", "accept": false, - "reason": "Python version compatibility breakage and incorrect pipeline-parallel outputs are unrelated." + "reason": "The VRAM leak report is about Qwen3ForCausalLM threading; the linked reproduction issue is not the same concrete failure." }, { - "left": "issue:43986", - "right": "issue:44484", + "left": "issue:44263", + "right": "issue:46172", "accept": false, - "reason": "AutoProcessor crash without torchvision and shard-size policy in save_pretrained do not share a concrete bug path." + "reason": "Different model families and different code paths: a DSA indexer split issue versus a DeepSeek V4 dtype/transpose bug." }, { - "left": "issue:45440", + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "A bfloat16 precision warning during compile and a SAM3 model-type mismatch warning are unrelated." + }, + { + "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "Both mention DeepSeek-family models, but one is a divergence from remote implementation and the other is a tensor operation bug." + "reason": "A missing ReLU in the DSA indexer and a DeepSeek V4 comb dtype/transpose issue are not the same bug." }, { - "left": "issue:43097", - "right": "issue:46050", + "left": "issue:44484", + "right": "issue:45478", "accept": false, - "reason": "Config removal warning and quantized-model save/load corruption are separate problems." + "reason": "A save_pretrained shard-size question and a Qwen3.5 MoE loading error have no shared underlying fix." }, { - "left": "issue:43097", - "right": "issue:45588", + "left": "issue:43441", + "right": "issue:44829", "accept": false, - "reason": "Tie-embeddings deprecation/removal is unrelated to flash_attention.py crashing on s_aux=None." + "reason": "Both involve FlashAttention, but one is a Ministral-3 failure and the other is degenerate training with flash_attention_3; not mergeable as one PR." }, { - "left": "issue:43064", - "right": "issue:44315", + "left": "issue:42371", + "right": "issue:42831", "accept": false, - "reason": "Wrong optimizer state under FSDP2/PEFT/cpu_ram_efficient_loading is unrelated to Liger Kernel not being applied with model_init." + "reason": "TF32 settings guidance and FineGrainedFP8 accuracy loss are separate numeric-behavior issues." }, { - "left": "issue:42673", - "right": "issue:45200", + "left": "issue:43931", + "right": "issue:44038", "accept": false, - "reason": "VRAM leak in threaded Qwen3ForCausalLM is unrelated to mm_token_type_ids defaults for Gemma 4 fine-tuning." + "reason": "A Qwen3-VL checkpoint shape mismatch and a Qwen3-VL-Moe issue on Transformers 5.0 are related only by model family, not by the same concrete bug." }, { - "left": "issue:38617", - "right": "issue:42757", + "left": "issue:43335", + "right": "issue:45071", "accept": false, - "reason": "Different missing imports from different modules and APIs; not the same import regression." + "reason": "SwitchTransformers sparse-layer config bug and PretrainedConfig type-checking regression are unrelated." }, { - "left": "issue:44062", - "right": "issue:44117", + "left": "issue:43097", + "right": "issue:43452", "accept": false, - "reason": "AddedToken special-argument conflict and tokenizer mapping None handling are different tokenizer-loading failures." + "reason": "Tied-embedding config removal and gguf_file loading failures are different loading regressions." }, { - "left": "issue:44410", - "right": "issue:45375", + "left": "issue:43874", + "right": "issue:44263", "accept": false, - "reason": "Missing attention projections in qwen3next and a dropped config field in Qwen3_5MoeVisionConfig are different defects." + "reason": "Missing get_number_of_image_patches and torch.split return handling are different implementation bugs." }, { - "left": "issue:42890", - "right": "issue:46298", + "left": "issue:44933", + "right": "issue:46231", "accept": false, - "reason": "A flaky test due to missing set_seed and a noisy logging message are unrelated." + "reason": "A missing image_utils import and a documentation typo in set_training are unrelated." }, { - "left": "issue:44038", - "right": "issue:44410", + "left": "issue:44117", + "right": "issue:44987", "accept": false, - "reason": "Qwen3-VL-Moe loading bug and qwen3next missing projections are different model issues with different failure modes." + "reason": "Tokenizer mapping returning None and a physical-intelligence/fast load failure are different issues." }, { - "left": "issue:43931", - "right": "issue:44410", + "left": "issue:43408", + "right": "issue:44368", "accept": false, - "reason": "Weight-shape mismatch during Qwen3-VL-30B load is not the same as missing layer projections in qwen3next." + "reason": "Both are warnings, but one is a model-type mismatch and the other is a tied-embeddings config warning; not the same bug." }, { - "left": "issue:43502", - "right": "issue:43992", + "left": "issue:43976", + "right": "issue:44945", "accept": false, - "reason": "local_files_only network leakage and missing embed_tokens.weight in UMT5Encoder are unrelated loading bugs." + "reason": "Python compatibility breakage and incorrect output under pipeline parallelism are unrelated." }, { - "left": "issue:45341", - "right": "issue:45859", + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "A crash without torchvision and a save_pretrained shard-size question do not share the same underlying defect." + }, + { + "left": "issue:45440", + "right": "issue:46172", "accept": false, - "reason": "A testing_utils bug and a CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding are unrelated." + "reason": "Both mention DeepSeek, but one is a native-vs-remote implementation divergence and the other is a dtype/transpose bug in DeepSeek V4." }, { "left": "issue:43097", - "right": "issue:43873", + "right": "issue:46050", "accept": false, - "reason": "Embedding-tie removal and quantization/offloading behavior are not the same issue." + "reason": "Tie-embedding config removal and sinq quantization save/load breakage are unrelated regressions." }, { - "left": "issue:38617", - "right": "issue:44351", + "left": "issue:43097", + "right": "issue:45588", "accept": false, - "reason": "Both are import errors, but from different symbols and different modules; no shared root cause is evident." + "reason": "Tied-embedding config removal and a flash_attention crash on s_aux=None are different code-path failures." } ] }, @@ -14200,63 +14172,76 @@ "component_index": 1, "component_count": 18, "review_unit_index": 45, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:38617", - "issue:42490", "issue:42617", + "issue:42673", "issue:42757", + "issue:42890", "issue:42907", "issue:43012", + "issue:43064", "issue:43066", "issue:43097", "issue:43208", "issue:43502", - "issue:43526", "issue:43550", "issue:43792", - "issue:43824", "issue:43825", + "issue:43873", "issue:43906", + "issue:43931", "issue:43986", "issue:43992", + "issue:44038", + "issue:44062", "issue:44117", "issue:44162", - "issue:44222", "issue:44261", + "issue:44315", + "issue:44351", "issue:44368", - "issue:44483", + "issue:44410", "issue:44484", "issue:44485", - "issue:44514", "issue:44655", - "issue:44683", "issue:44704", - "issue:44938", "issue:44964", "issue:45103", + "issue:45200", "issue:45290", "issue:45341", "issue:45372", - "issue:45478", + "issue:45375", "issue:45561", "issue:45588", - "issue:45600", - "issue:45698", "issue:45712", "issue:45753", + "issue:45859", "issue:46123", "issue:46172", - "issue:46178", "issue:46298" ], "soft_pairs": [ + "issue:43064|issue:44315", + "issue:42673|issue:45200", + "issue:38617|issue:42757", + "issue:44062|issue:44117", + "issue:44410|issue:45375", + "issue:42890|issue:46298", + "issue:44038|issue:44410", + "issue:43931|issue:44410", + "issue:43502|issue:43992", + "issue:45341|issue:45859", + "issue:43097|issue:43873", + "issue:38617|issue:44351", "issue:44261|issue:44485", "issue:44964|issue:45341", "issue:44704|issue:45103", @@ -14272,244 +14257,234 @@ "issue:43992|issue:44484", "issue:43012|issue:43550", "issue:43825|issue:44655", - "issue:45712|issue:46298", "issue:43012|issue:44368", + "issue:45712|issue:46298", "issue:42907|issue:44655", "issue:42907|issue:45753", "issue:43097|issue:46298", "issue:44162|issue:46172", - "issue:43792|issue:46123", - "issue:44938|issue:46178", - "issue:44483|issue:44514", - "issue:44222|issue:45478", - "issue:43208|issue:43526", - "issue:43012|issue:44683", - "issue:42490|issue:43012", - "issue:42757|issue:43824", - "issue:45600|issue:46298", - "issue:38617|issue:43502", - "issue:45698|issue:46172" + "issue:43792|issue:46123" ], - "prepared_review_unit_hash": "ed664e5d83d9d457a09518e7b7030ba3f1930ab27cbabb625d54ff0abfca455e", + "prepared_review_unit_hash": "75b1e382fdd06905e6fe6659335b624cf31206a63cca7da6c5c2fcaad2a40758", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 32, - "serialized_chars": 36475, - "estimated_input_tokens": 9119, - "estimated_eval_tokens": 18494 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37751, + "estimated_input_tokens": 9438, + "estimated_eval_tokens": 19132 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly loosely related Transformers bugs, not true duplicates. The only small coherent subcluster is around processor/model loading import failures, but even there the underlying causes differ; no PRs are present.", - "confidence": 0.63, - "canonical_issue_reason": "issue:43986 is the most representative concrete bug in the loading/import-failure subset and has the strongest overlap with other soft-similarity candidates.", + "summary": "These soft links are mostly superficial matches across unrelated issues; they span different models, APIs, warnings, tests, and dependency/import failures rather than one underlying bug.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical duplicate stands out: the items are heterogeneous and do not collapse to one shared code-path failure.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43986 is the best triage target because it describes a specific failure mode, a clear affected code path, and sits near several related processor-loading reports.", + "best_issue_reason": "issue:43986 is the clearest self-contained, actionable bug report in the set, but it is only a representative issue\u2014not a duplicate target for the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44261", - "right": "issue:44485", + "left": "issue:43064", + "right": "issue:44315", "accept": false, - "reason": "Different model-specific bugs: MLA q_a_layernorm precision/config handling vs GLM-5 RoPE implementation." + "reason": "Different failures: FSDP2/PEFT optimizer-state corruption vs Liger kernel not applied under model_init." }, { - "left": "issue:44964", - "right": "issue:45341", + "left": "issue:42673", + "right": "issue:45200", "accept": false, - "reason": "Unrelated areas: model loading for Phi-4 multimodal vs a testing_utils bug." + "reason": "VRAM leak across dataloader threads is unrelated to Gemma 4 mm_token_type_ids defaults for text-only fine-tuning." }, { - "left": "issue:44704", - "right": "issue:45103", + "left": "issue:38617", + "right": "issue:42757", "accept": false, - "reason": "Different layers entirely: AutoProcessor kwargs plumbing vs auto_docstring annotation handling." + "reason": "Both are import errors, but for different missing symbols from different packages and code paths." }, { - "left": "issue:43986", + "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Both involve loading, but one is a missing torchvision dependency during video-model loading and the other is tokenizer mapping assumptions." + "reason": "AddedToken keyword conflict is a constructor bug; TOKENIZER_MAPPING_NAMES returning None is a separate lookup/dispatch bug." }, { - "left": "issue:42617", - "right": "issue:45341", + "left": "issue:44410", + "right": "issue:45375", "accept": false, - "reason": "3d_parallel.py training/runtime failure is unrelated to the testing_utils issue." + "reason": "One is missing Qwen3Next attention projections; the other is a missing vision config field silently dropped by strict config handling." }, { - "left": "issue:43066", - "right": "issue:43906", + "left": "issue:42890", + "right": "issue:46298", "accept": false, - "reason": "Tokenizer decoder/type mismatch and isolated reproduction of a different issue are not the same concrete bug." + "reason": "A flaky test seed issue is unrelated to a noisy logging message." }, { - "left": "issue:43986", - "right": "issue:45372", + "left": "issue:44038", + "right": "issue:44410", "accept": false, - "reason": "Both are processor-loading errors, but they stem from different missing symbols/dependencies and different model families." + "reason": "Qwen3-VL-Moe loading failure and Qwen3Next missing projections are different model-specific defects." }, { - "left": "issue:43986", - "right": "issue:45290", + "left": "issue:43931", + "right": "issue:44410", "accept": false, - "reason": "Processor loading without torchvision is a different failure path from chat templating with tool-call messages." + "reason": "Shape-mismatch loading for Qwen3-VL-30B is not the same bug as missing attn_qkv/attn_gate projections." }, { - "left": "issue:45341", - "right": "issue:45588", + "left": "issue:43502", + "right": "issue:43992", "accept": false, - "reason": "Testing utility bug vs flash-attention crash are unrelated code paths." + "reason": "local_files_only still triggering API requests is unrelated to UMT5Encoder missing embed_tokens.weight." }, { "left": "issue:45341", - "right": "issue:45561", + "right": "issue:45859", "accept": false, - "reason": "Both touch testing utilities, but one is a simple bug and the other is an xdist race on captured_info.txt." + "reason": "testing_utils bug and Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are unrelated." }, { - "left": "issue:43208", - "right": "issue:43550", + "left": "issue:43097", + "right": "issue:43873", "accept": false, - "reason": "Distinct model/backend failures: xLSTM training blockers vs Bamba torch.compile SDPA failure." + "reason": "Removed tie_embeddings_and_encoder_decoder in v5 is a breaking API change, not the same as offloading/quantization behavior." }, { - "left": "issue:43986", - "right": "issue:45341", + "left": "issue:38617", + "right": "issue:44351", "accept": false, - "reason": "Processor loading/import failure and testing_utils bug are unrelated." + "reason": "Different missing imports from different namespaces; no shared underlying code-path." }, { - "left": "issue:43992", - "right": "issue:44484", + "left": "issue:44261", + "right": "issue:44485", "accept": false, - "reason": "Missing embed_tokens.weight during UMT5Encoder load is unrelated to max_shard_size default behavior." + "reason": "MLA q_a_layernorm epsilon precision and GLM-5 RoPE implementation are separate model math issues." }, { - "left": "issue:43012", - "right": "issue:43550", + "left": "issue:44964", + "right": "issue:45341", "accept": false, - "reason": "One is a PyTorch warning on bfloat16 compilation; the other is a model-specific compile failure with SDPA." + "reason": "Phi-4 multimodal loading failure is unrelated to the testing_utils bug." }, { - "left": "issue:43825", - "right": "issue:44655", + "left": "issue:44704", + "right": "issue:45103", "accept": false, - "reason": "Pipeline translation support messaging and pipeline save_pretrained are different bugs." + "reason": "AutoProcessor kwargs forwarding to cached_file is a different failure from auto_docstring crashing on postponed annotations." }, { - "left": "issue:45712", - "right": "issue:46298", + "left": "issue:43986", + "right": "issue:44117", "accept": false, - "reason": "Left is leftover dummy classes leaking into imports; right is a noisy logging message." + "reason": "Missing torchvision during AutoProcessor load is unrelated to TOKENIZER_MAPPING_NAMES sometimes returning None." }, { - "left": "issue:43012", - "right": "issue:44368", + "left": "issue:42617", + "right": "issue:45341", "accept": false, - "reason": "Both mention warnings, but the model/compiler conditions and root causes differ." + "reason": "Cannot run 3d_parallel.py and a testing_utils bug do not describe the same defect." }, { - "left": "issue:42907", - "right": "issue:44655", + "left": "issue:43066", + "right": "issue:43906", "accept": false, - "reason": "Dequantized model save failure is not the same as pipeline save_pretrained failure." + "reason": "Wrong tokenizer decoder type is a concrete tokenizer regression; the other item is only an isolated reproduction of a different issue." }, { - "left": "issue:42907", - "right": "issue:45753", + "left": "issue:43986", + "right": "issue:45372", "accept": false, - "reason": "Saving dequantized models and a Qwen3.5 infinite-loop repro are unrelated." + "reason": "Both are processor-loading errors, but one is missing torchvision and the other is a mistral_common import failure; different dependencies and paths." }, { - "left": "issue:43097", - "right": "issue:46298", + "left": "issue:43986", + "right": "issue:45290", "accept": false, - "reason": "Removed config field behavior is unrelated to a noisy logging message." + "reason": "Processor loading without torchvision and chat-template tool-call crashes are separate bugs." }, { - "left": "issue:44162", - "right": "issue:46172", + "left": "issue:45341", + "right": "issue:45588", "accept": false, - "reason": "ESM2 breakage and deepseek-v4 transpose/dtype issue are separate model bugs." + "reason": "A testing utility bug is not the same as flash_attention crashing on sink-less models." }, { - "left": "issue:43792", - "right": "issue:46123", + "left": "issue:45341", + "right": "issue:45561", "accept": false, - "reason": "Whisper model load/run failure and MaskGenerationPipeline batch bookkeeping bug are unrelated." + "reason": "Testing utility bug vs pytest-xdist race on captured_info.txt are different concurrency paths." }, { - "left": "issue:44938", - "right": "issue:46178", + "left": "issue:43208", + "right": "issue:43550", "accept": false, - "reason": "Python 3.14 import/load failure is unrelated to a BACKENDS_MAPPING validation error." + "reason": "xLSTM training blockers and Bamba torch.compile+SDPA failure are different model/runtime problems." }, { - "left": "issue:44483", - "right": "issue:44514", + "left": "issue:43992", + "right": "issue:44484", "accept": false, - "reason": "Chat/completions API incompatibility and batched chat-template crash are different concrete failures." + "reason": "Missing checkpoint weight on load is unrelated to the max_shard_size default question in save_pretrained()." }, { - "left": "issue:44222", - "right": "issue:45478", + "left": "issue:43012", + "right": "issue:43550", "accept": false, - "reason": "Both mention MoE/save/from_pretrained, but they affect different model families and different code paths." + "reason": "A bfloat16 warning during compile is not the same as a model-specific torch.compile SDPA failure." }, { - "left": "issue:43208", - "right": "issue:43526", + "left": "issue:43825", + "right": "issue:44655", "accept": false, - "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated." + "reason": "pipeline() translation support messaging and saving Pipeline objects are different pipeline API issues." }, { "left": "issue:43012", - "right": "issue:44683", + "right": "issue:44368", "accept": false, - "reason": "A PyTorch precision warning is not the same as compiled flex_attention failing on torch>=2.9." + "reason": "Both mention warnings, but they concern different config fields and different model paths." }, { - "left": "issue:42490", - "right": "issue:43012", + "left": "issue:45712", + "right": "issue:46298", "accept": false, - "reason": "Trailing-slash local model import warning and bfloat16 compile warning are unrelated." + "reason": "Left is leftover dummy classes/check_repo leakage; right is just a noisy logging message." }, { - "left": "issue:42757", - "right": "issue:43824", + "left": "issue:42907", + "right": "issue:44655", "accept": false, - "reason": "Missing is_offline_mode import and missing Qwen2_5_VLForConditionalGeneration are different import errors." + "reason": "Failing to save dequantized Ministrals/Devstrals is unrelated to saving generic Pipeline objects." }, { - "left": "issue:45600", + "left": "issue:43097", "right": "issue:46298", "accept": false, - "reason": "Broken auto_mappings references and a noisy logging message are unrelated." + "reason": "A removed API warning is not the same as a noisy logging message." }, { - "left": "issue:38617", - "right": "issue:43502", + "left": "issue:44162", + "right": "issue:46172", "accept": false, - "reason": "Missing layer_type_validation import and unwanted API calls despite local_files_only are different failures." + "reason": "ESM2 breakage and deepseek-v4 tensor-transpose bug are unrelated model-specific failures." }, { - "left": "issue:45698", - "right": "issue:46172", + "left": "issue:43792", + "right": "issue:46123", "accept": false, - "reason": "Wrong custom-module loading after save_pretrained is unrelated to the deepseek-v4 tensor operation bug." + "reason": "Whisper load failure and MaskGenerationPipeline dropping the final partial batch are different bugs in different components." } ] }, @@ -14525,64 +14500,74 @@ "component_index": 1, "component_count": 18, "review_unit_index": 46, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:36010", "issue:36683", + "issue:38617", + "issue:42490", "issue:42757", "issue:42994", + "issue:43012", "issue:43064", "issue:43066", + "issue:43208", "issue:43299", "issue:43352", - "issue:43441", "issue:43475", "issue:43502", "issue:43525", + "issue:43526", "issue:43611", "issue:43618", "issue:43673", "issue:43761", + "issue:43824", "issue:43867", - "issue:43881", "issue:43901", - "issue:43906", "issue:43937", - "issue:43976", - "issue:43992", "issue:44038", - "issue:44117", + "issue:44222", "issue:44230", - "issue:44242", - "issue:44355", "issue:44380", "issue:44483", "issue:44485", + "issue:44514", "issue:44625", "issue:44655", - "issue:44855", + "issue:44683", + "issue:44938", "issue:45216", "issue:45276", "issue:45290", - "issue:45310", - "issue:45375", "issue:45478", - "issue:45479", - "issue:45753", + "issue:45600", + "issue:45698", "issue:46077", "issue:46082", "issue:46123", - "issue:46132", - "issue:46202", - "issue:46209" + "issue:46172", + "issue:46178", + "issue:46209", + "issue:46298" ], "soft_pairs": [ + "issue:44938|issue:46178", + "issue:44483|issue:44514", + "issue:44222|issue:45478", + "issue:43208|issue:43526", + "issue:43012|issue:44683", + "issue:42490|issue:43012", + "issue:42757|issue:43824", + "issue:45600|issue:46298", + "issue:38617|issue:43502", + "issue:45698|issue:46172", "issue:44038|issue:44230", "issue:43673|issue:46123", "issue:44483|issue:45290", @@ -14595,268 +14580,217 @@ "issue:42994|issue:44655", "issue:43901|issue:44655", "issue:44625|issue:46077", - "issue:43611|issue:44655", "issue:36010|issue:42757", "issue:43352|issue:43937", + "issue:43611|issue:44655", "issue:43867|issue:45216", "issue:46077|issue:46082", "issue:43502|issue:44485", "issue:36683|issue:43525", - "issue:43618|issue:45216", - "issue:36683|issue:44242", - "issue:44380|issue:46132", - "issue:36683|issue:43881", - "issue:43906|issue:45753", - "issue:43299|issue:45753", - "issue:43352|issue:43976", - "issue:43992|issue:45310", - "issue:43441|issue:44380", - "issue:43352|issue:45478", - "issue:44355|issue:44855", - "issue:43867|issue:46202", - "issue:43867|issue:45375", - "issue:44380|issue:45479", - "issue:43352|issue:44380", - "issue:36683|issue:44117" + "issue:43618|issue:45216" ], - "prepared_review_unit_hash": "d500ea418c2cabcf19e1f7c97cd96e6d24f1214919074b7c3a0ac06133778931", + "prepared_review_unit_hash": "ccacc19be3c17fcf089101d1f4982c3e3f9ef3b87f7c07eb6ed256cb8fd3705c", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 35, - "serialized_chars": 37861, - "estimated_input_tokens": 9466, - "estimated_eval_tokens": 19188 + "soft_pair_count": 30, + "serialized_chars": 36257, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a loose thematic cluster of distinct Transformers v5 regressions (loading, config propagation, attention backends, save/pretrained, and pipeline behavior). The soft pairs are similar in wording but do not describe the same underlying bug/change, so none should be merged as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": "issue:43611 is the most generic and representative report in the set: a broad v5 model-loading regression around `base_model_prefix` that plausibly anchors the wider cluster of loading-related breakages.", + "summary": "This cluster is mostly heterogeneous: nearly all soft pairs describe different bugs, models, or code paths. I would not merge any of the suggested issue pairs as duplicates.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45698 is the strongest anchor in this set: it is open, has the most discussion/inbound references, and describes a broad save/load regression with clear user impact.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43611 is also the strongest standalone triage candidate: it is concrete, reproducible, and broad enough to serve as the clearest reference point for the v5 loading regression family.", + "best_issue_reason": "issue:45698 looks like the best overall issue to keep as the representative item because it is active, broadly impactful, and directly about core from_pretrained/save_pretrained behavior.", "best_pr_reason": null, - "soft_edge_verdicts": [ - { - "left": "issue:44038", - "right": "issue:44230", - "accept": false, - "reason": "Both mention Qwen3 VL/MoE and v5, but one is a loading regression and the other is an fp8 inference-support request; different bugs and code paths." - }, - { - "left": "issue:43673", - "right": "issue:46123", - "accept": false, - "reason": "GenerationMixin cache/chunked_prefill failure is unrelated to MaskGenerationPipeline dropping the final partial batch." - }, - { - "left": "issue:44483", - "right": "issue:45290", - "accept": false, - "reason": "One is a /v1/chat/completions API validation failure; the other is apply_chat_template crashing on tool-call assistant messages. Different layers." - }, - { - "left": "issue:43064", - "right": "issue:44380", - "accept": false, - "reason": "Trainer/FSDP2 optimizer-state corruption is unrelated to GPT2 attention scaling being ignored under SDPA/FlashAttention." - }, - { - "left": "issue:43618", - "right": "issue:43761", - "accept": false, - "reason": "Both are CLIP-related regressions, but one concerns missing attentions assignment and the other hidden_states being None; not the same bug." - }, + "soft_edge_verdicts": [ { - "left": "issue:43299", - "right": "issue:43867", + "left": "issue:44938", + "right": "issue:46178", "accept": false, - "reason": "Qwen3VL-MoE loading breakage and state_dict-sorted loading errors are different loading failures with different triggers." + "reason": "Both are import-related, but one is a Python 3.14 load failure and the other is a backend mapping error for tensorflow_text; different failure modes and code paths." }, { - "left": "issue:43066", - "right": "issue:46209", + "left": "issue:44483", + "right": "issue:44514", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 and FP8 inference failing on B200 are unrelated issues." + "reason": "One is an API request/acceptance problem on /v1/chat/completions, the other is a chat templating crash on batched input with padding=False; not the same bug." }, { - "left": "issue:36683", - "right": "issue:43475", + "left": "issue:44222", + "right": "issue:45478", "accept": false, - "reason": "Missing `vocab_size` in Gemma3Config and missing `fpn_position_embeddings` in SAM 3 Video outputs are different attribute errors on different models." + "reason": "Both mention Qwen3.5/MoE and FP8/save/load, but one is about saving quantized models and the other about loading from_pretrained on v5.4.0; related area, not the same concrete defect." }, { - "left": "issue:45276", - "right": "issue:46077", + "left": "issue:43208", + "right": "issue:43526", "accept": false, - "reason": "Gemma4 token-embedding resize behavior and Gemma4 bidirectional-attention masking are separate model-behavior bugs." + "reason": "Completely different subsystems: xLSTM training bugs versus BeitImageProcessorFast label reduction behavior." }, { - "left": "issue:42994", - "right": "issue:44655", + "left": "issue:43012", + "right": "issue:44683", "accept": false, - "reason": "Quantized model saving failure and pipeline `save_pretrained` failure are distinct save-path problems." + "reason": "A PyTorch precision warning from bfloat16 compilation is unrelated to compiled flex_attention failures on torch>=2.9." }, { - "left": "issue:43901", - "right": "issue:44655", + "left": "issue:42490", + "right": "issue:43012", "accept": false, - "reason": "Docs for `return_all_scores` versus pipeline serialization behavior are not the same issue." + "reason": "Local-path import warnings are unrelated to bfloat16/float32 precision warnings." }, { - "left": "issue:44625", - "right": "issue:46077", + "left": "issue:42757", + "right": "issue:43824", "accept": false, - "reason": "Qwen3.5 `num_labels` propagation and Gemma4 attention-mask construction are unrelated model/config regressions." + "reason": "Both are import errors, but they concern different missing symbols from different packages and different breakages." }, { - "left": "issue:43611", - "right": "issue:44655", + "left": "issue:45600", + "right": "issue:46298", "accept": false, - "reason": "Base-model-prefix loading regression and pipeline serialization failure are different APIs and different failure modes." + "reason": "CI breakage from removed Sam3LiteText references is unrelated to a noisy logging message." }, { - "left": "issue:36010", - "right": "issue:42757", + "left": "issue:38617", + "right": "issue:43502", "accept": false, - "reason": "These are both import errors, but for different symbols from different modules (`transformers.generation` vs `huggingface_hub`)." + "reason": "One is an import error from configuration_utils; the other is network requests occurring despite local_files_only=True. Different bugs." }, { - "left": "issue:43352", - "right": "issue:43937", + "left": "issue:45698", + "right": "issue:46172", "accept": false, - "reason": "FlashAttention support for Nemotron and GLM-5 generation-config validation are unrelated bugs." + "reason": "Save/load custom-module mismatch is unrelated to a DeepSeek tensor transpose/dtype issue." }, { - "left": "issue:43867", - "right": "issue:45216", + "left": "issue:44038", + "right": "issue:44230", "accept": false, - "reason": "Sorted-state_dict loading failure and Qwen3.5 save_pretrained regression are different directions of model I/O failure." + "reason": "Both are Qwen3-VL/MoE-related, but one is a general v5 bug and the other is specifically FP8 inference support; not clearly the same underlying fix." }, { - "left": "issue:46077", - "right": "issue:46082", + "left": "issue:43673", + "right": "issue:46123", "accept": false, - "reason": "Gemma4 bidirectional-attention masking and LlamaConfig head_dim validation are unrelated configuration behaviors." + "reason": "GenerationMixin cache missing in chunked_prefill and MaskGenerationPipeline dropping the final partial batch are different functional bugs." }, { - "left": "issue:43502", - "right": "issue:44485", + "left": "issue:44483", + "right": "issue:45290", "accept": false, - "reason": "Offline-mode network leakage and GLM-5 RoPE implementation are unrelated." + "reason": "One is a chat-completions API compatibility failure; the other is a template/tokenization crash with tool calls and no content." }, { - "left": "issue:36683", - "right": "issue:43525", + "left": "issue:43064", + "right": "issue:44380", "accept": false, - "reason": "Missing `vocab_size` in Gemma3Config and missing `pad_token_id` in Llama4Config are different missing-field regressions." + "reason": "Trainer/FSDP2 optimizer-state corruption is unrelated to GPT2 attention scaling being ignored under SDPA/FlashAttention." }, { "left": "issue:43618", - "right": "issue:45216", + "right": "issue:43761", "accept": false, - "reason": "CLIP attention outputs and Qwen3.5 checkpoint saving are unrelated model families and code paths." + "reason": "Both are CLIP v5 regressions, but one is missing attentions and the other is missing hidden_states; they look like separate output-path issues." }, { - "left": "issue:36683", - "right": "issue:44242", + "left": "issue:43299", + "right": "issue:43867", "accept": false, - "reason": "Gemma3 config attribute loss and MoE load-balancing loss computation are different bugs." + "reason": "Qwen3VL MoE loading failures and sorted-state_dict load errors are different loading problems." }, { - "left": "issue:44380", - "right": "issue:46132", + "left": "issue:43066", + "right": "issue:46209", "accept": false, - "reason": "GPT2 attention scaling under backends and AttentionInterface.register behavior changes are different attention-stack issues." + "reason": "Tokenizer decoder type mismatch and FP8 inference failure on B200 are unrelated." }, { "left": "issue:36683", - "right": "issue:43881", + "right": "issue:43475", "accept": false, - "reason": "Gemma3 missing `vocab_size` and GLM-4V-9B loading failure are unrelated model-specific regressions." + "reason": "Gemma3Config missing vocab_size and SAM 3 Video missing fpn_position_embeddings are different attribute errors in different models." }, { - "left": "issue:43906", - "right": "issue:45753", + "left": "issue:45276", + "right": "issue:46077", "accept": false, - "reason": "An isolated reproduction of an earlier issue is not the same as a Qwen3_5 infinite-loop-on-image bug." + "reason": "Gemma4 embedding-resize propagation and Gemma4 bidirectional-attention masking are distinct model bugs." }, { - "left": "issue:43299", - "right": "issue:45753", + "left": "issue:42994", + "right": "issue:44655", "accept": false, - "reason": "Qwen3VL-MoE loading breakage and Qwen3_5 infinite loop on a specific image are different problems." + "reason": "Quantized model save failure and pipeline save_pretrained failure are different save paths with different symptoms." }, { - "left": "issue:43352", - "right": "issue:43976", + "left": "issue:43901", + "right": "issue:44655", "accept": false, - "reason": "FlashAttention support for Nemotron and Python-version compatibility in Transformers 5.1 are unrelated." + "reason": "Documentation mismatch for return_all_scores is unrelated to pipeline serialization failing." }, { - "left": "issue:43992", - "right": "issue:45310", + "left": "issue:44625", + "right": "issue:46077", "accept": false, - "reason": "UMT5Encoder missing weights during load and Qwen3.5 MoE from_pretrained failure are different model-loading regressions." + "reason": "Qwen3.5 num_labels propagation and Gemma4 attention-mask construction are unrelated." }, { - "left": "issue:43441", - "right": "issue:44380", + "left": "issue:36010", + "right": "issue:42757", "accept": false, - "reason": "Ministral FlashAttention failure and GPT2 attention scaling being ignored are both attention-related but not the same concrete bug." + "reason": "Different missing imports from different modules; no shared underlying bug." }, { "left": "issue:43352", - "right": "issue:45478", - "accept": false, - "reason": "Nemotron FlashAttention support and Qwen3.5 MoE from_pretrained errors are unrelated." - }, - { - "left": "issue:44355", - "right": "issue:44855", + "right": "issue:43937", "accept": false, - "reason": "Compiled-Python-file import errors and a Python 3.13 `@torch.jit.script` parsing/IndentationError are different import-time failures." + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig validation are unrelated." }, { - "left": "issue:43867", - "right": "issue:46202", + "left": "issue:43611", + "right": "issue:44655", "accept": false, - "reason": "State_dict-sorted loading failure and Gemma pipeline roundtrip failure on `\u2581` are unrelated I/O bugs." + "reason": "Base_model_prefix loading issue and pipeline save_pretrained failure are different code paths." }, { "left": "issue:43867", - "right": "issue:45375", + "right": "issue:45216", "accept": false, - "reason": "Loading error with sorted state_dict and silently dropped `deepstack_visual_indexes` are different config/load issues." + "reason": "A sorted state_dict load error is not the same as Qwen3.5 checkpoint corruption introduced by save_pretrained in 5.4.0." }, { - "left": "issue:44380", - "right": "issue:45479", + "left": "issue:46077", + "right": "issue:46082", "accept": false, - "reason": "Attention scaling under backends and degenerate zero-loss classification with `num_labels=1` are unrelated." + "reason": "Gemma4 causal-mask regression and LlamaConfig head_dim validation are unrelated." }, { - "left": "issue:43352", - "right": "issue:44380", + "left": "issue:43502", + "right": "issue:44485", "accept": false, - "reason": "Both touch attention backends, but one is unsupported FlashAttention on Nemotron and the other is GPT2 scaling being ignored; different fixes." + "reason": "Offline-mode network access is unrelated to GLM-5 RoPE implementation issues." }, { "left": "issue:36683", - "right": "issue:44117", + "right": "issue:43525", "accept": false, - "reason": "Gemma3 missing `vocab_size` and tokenizer-mapping `None` handling are different failure modes in different subsystems." + "reason": "Different configs and different missing attributes: vocab_size versus pad_token_id." } ] }, @@ -14872,66 +14806,81 @@ "component_index": 1, "component_count": 18, "review_unit_index": 47, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:28282", - "issue:30333", - "issue:30990", - "issue:31795", "issue:33453", "issue:36296", + "issue:36683", "issue:41669", - "issue:42548", "issue:42890", - "issue:43010", "issue:43064", "issue:43097", + "issue:43299", "issue:43352", + "issue:43441", "issue:43502", "issue:43575", - "issue:43576", "issue:43618", - "issue:43723", "issue:43867", + "issue:43881", + "issue:43906", "issue:43976", "issue:43992", "issue:44062", "issue:44112", + "issue:44117", "issue:44222", + "issue:44242", "issue:44246", - "issue:44351", "issue:44355", "issue:44360", "issue:44380", - "issue:44387", "issue:44492", - "issue:44556", "issue:44655", - "issue:44749", + "issue:44855", "issue:44857", "issue:44938", "issue:44964", "issue:44977", "issue:45125", - "issue:45278", + "issue:45310", "issue:45357", "issue:45375", "issue:45440", "issue:45478", + "issue:45479", "issue:45561", "issue:45600", + "issue:45753", "issue:46132", + "issue:46202", "issue:46285" ], "soft_pairs": [ - "issue:41669|issue:44246", + "issue:36683|issue:44242", + "issue:44380|issue:46132", + "issue:36683|issue:43881", + "issue:43906|issue:45753", + "issue:43299|issue:45753", + "issue:43352|issue:43976", + "issue:43992|issue:45310", + "issue:43441|issue:44380", + "issue:43352|issue:45478", + "issue:44355|issue:44855", + "issue:43867|issue:46202", + "issue:43867|issue:45375", + "issue:44380|issue:45479", + "issue:43352|issue:44380", + "issue:36683|issue:44117", "issue:43618|issue:44380", + "issue:41669|issue:44246", "issue:43097|issue:43352", "issue:42890|issue:45561", "issue:43064|issue:44492", @@ -14950,239 +14899,240 @@ "issue:42890|issue:44964", "issue:42890|issue:44112", "issue:42890|issue:45600", - "issue:44222|issue:45125", - "issue:43576|issue:46132", - "issue:44355|issue:44749", - "issue:42548|issue:44351", - "issue:45125|issue:45478", - "issue:44387|issue:44556", - "issue:30333|issue:33453", - "issue:45125|issue:45375", - "issue:43010|issue:45278", - "issue:30990|issue:31795", - "issue:43723|issue:44556", - "issue:43502|issue:44222" + "issue:44222|issue:45125" ], - "prepared_review_unit_hash": "fc6c56f3d36befb3467046d44fd2ee8b87393f32731acc6f3784401210be1943", + "prepared_review_unit_hash": "8aa57892d8fae62adcd8ec080c7f31aa6975323616e01658b48a46ae5d9c6dda", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 36870, - "estimated_input_tokens": 9218, - "estimated_eval_tokens": 18692 + "soft_pair_count": 36, + "serialized_chars": 38087, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly separate Transformers bugs spanning import/perf, loading, tensor parallelism, cache behavior, docs, and test flakiness. None of the soft pairs look like the same concrete underlying bug/change closely enough to merge as duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "No clear single duplicate cluster emerges from the titles alone; if one issue must represent the set, issue:44246 is the broadest user-facing report (general import slowness) and could serve as the umbrella triage item.", + "summary": "This cluster is a heterogeneous mix of Transformers issues (model loading, attention backends, tensor parallelism, CI flakiness, and import-time regressions). None of the soft-linked pairs look like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.46, + "canonical_issue_reason": "Issue 36683 is the most active open item with a concrete runtime failure and broad model-loading relevance; among this mixed set it is the best anchor issue, even though the cluster is not a true duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44246 is the most general and representative-sounding issue in this list, with a symptom broad enough to relate to the import/performance subgroup better than the narrower reports.", + "best_issue_reason": "Issue 36683 is the strongest representative issue here: active discussion, clear failure mode, and enough specificity to be actionable without being just a test or minor cleanup report.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:41669", - "right": "issue:44246", + "left": "issue:36683", + "right": "issue:44242", "accept": false, - "reason": "Related to import performance, but 41669 proposes a specific root cause/fix (import * in models) while 44246 is a broader intermittent slowness report; not clearly the same bug." + "reason": "Different bugs: Gemma3 config missing vocab_size vs MoE load-balancing loss handling." }, { - "left": "issue:43618", - "right": "issue:44380", + "left": "issue:44380", + "right": "issue:46132", "accept": false, - "reason": "Both concern attention-related behavior, but they are different models and different failures (missing CLIP attentions vs GPT2 scaling ignored)." + "reason": "Both touch attention code, but one is GPT-2 scaling under SDPA/FlashAttention and the other changes AttentionInterface.register behavior." }, { - "left": "issue:43097", - "right": "issue:43352", + "left": "issue:36683", + "right": "issue:43881", "accept": false, - "reason": "Different problems: a removed config API in v5 versus Flash Attention support missing for NemotronHForCausalLM." + "reason": "Both are load-time failures, but they affect different models and stem from different missing attributes/code paths." }, { - "left": "issue:42890", - "right": "issue:45561", + "left": "issue:43906", + "right": "issue:45753", "accept": false, - "reason": "Both are test/CI related, but one is seed-related flakiness and the other is an xdist race on a captured-info file; not the same bug." + "reason": "An isolated reproduction of an earlier issue is not the same as a specific Qwen3.5 image-loop regression." }, { - "left": "issue:43064", - "right": "issue:44492", + "left": "issue:43299", + "right": "issue:45753", "accept": false, - "reason": "Completely unrelated: distributed training optimizer-state corruption versus a cache-strategy typo." + "reason": "Both mention Qwen3.5-era models, but one is MoE loading and the other is multimodal infinite-loop generation." }, { - "left": "issue:43976", - "right": "issue:44938", + "left": "issue:43352", + "right": "issue:43976", "accept": false, - "reason": "Both mention Python version compatibility, but they describe different version ranges and failure modes; not the same underlying issue." + "reason": "Nemotron FlashAttention 2 support gating is unrelated to Python version compatibility." }, { - "left": "issue:43867", - "right": "issue:45357", + "left": "issue:43992", + "right": "issue:45310", "accept": false, - "reason": "Both involve model loading/saving regressions, but one is a sorted state_dict load error and the other is incorrect visual encoder keys for Qwen3.5 save_pretrained." + "reason": "UMT5 missing embed_tokens weights and Qwen3.5 MoE from_pretrained errors are different model-specific load failures." }, { - "left": "issue:43992", + "left": "issue:43441", + "right": "issue:44380", + "accept": false, + "reason": "Both involve attention, but one is unsupported FlashAttention on Ministral-3 and the other is ignored attention scaling on GPT-2 backends." + }, + { + "left": "issue:43352", "right": "issue:45478", "accept": false, - "reason": "Both are from_pretrained/load failures, but for different models and different causes; too broad to treat as one bug." + "reason": "Same broad topic family, but 45478 is a different Qwen3.5 MoE loading regression, not the Nemotron FlashAttention issue." }, { - "left": "issue:36296", - "right": "issue:43575", + "left": "issue:44355", + "right": "issue:44855", "accept": false, - "reason": "Both touch tensor parallelism, but one is a training bug and the other is an OOM when loading a specific model; not the same code-path problem." + "reason": "Both are import errors, but one is about compiled Python files and the other is a torch.jit parsing problem in DebertaV2Model." }, { - "left": "issue:43064", - "right": "issue:44655", + "left": "issue:43867", + "right": "issue:46202", "accept": false, - "reason": "Distributed optimizer-state issue versus saving Pipeline objects; no shared underlying defect." + "reason": "State-dict ordering load failures and Gemma pipeline roundtrip failures are unrelated." }, { - "left": "issue:28282", - "right": "issue:33453", + "left": "issue:43867", + "right": "issue:45375", "accept": false, - "reason": "ImportError for missing PyTorch dependency versus tokenizer loading regression; unrelated." + "reason": "Generic state_dict loading errors are not the same as a missing Qwen3_5MoeVisionConfig field." }, { - "left": "issue:43618", - "right": "issue:46285", + "left": "issue:44380", + "right": "issue:45479", "accept": false, - "reason": "Different CLIP-related regressions: missing attentions assignment versus structural inconsistency between two CLIP text model classes." + "reason": "Attention backend config bug vs degenerate zero-loss classification behavior." }, { - "left": "issue:43618", - "right": "issue:44857", + "left": "issue:43352", + "right": "issue:44380", "accept": false, - "reason": "Attention-output regression versus AMP/CUDA crash in a loss; no shared bug." + "reason": "Both are attention-related, but they affect different models and different backend behavior." }, { - "left": "issue:44655", - "right": "issue:44938", + "left": "issue:36683", + "right": "issue:44117", "accept": false, - "reason": "Saving Pipeline objects is unrelated to Python 3.14 import/load failures." + "reason": "Different missing-attribute regressions in different parts of the loading stack." }, { "left": "issue:43618", - "right": "issue:44977", + "right": "issue:44380", "accept": false, - "reason": "Different model families and different failure modes; one is CLIP attentions, the other is Qwen3.5 flash-attention generation issues." + "reason": "CLIPOutput attentions regression is not the same bug as GPT-2 attention scaling being ignored." }, { - "left": "issue:44360", - "right": "issue:45440", + "left": "issue:41669", + "right": "issue:44246", "accept": false, - "reason": "The DSA indexer ReLU discussion is unrelated to DeepSeekV3 implementation divergence." + "reason": "Import-time slowdown from wildcard imports and intermittent import slowness are related only at a very high level, not the same defect." }, { - "left": "issue:43502", - "right": "issue:44062", + "left": "issue:43097", + "right": "issue:43352", "accept": false, - "reason": "Offline-loading API requests versus tokenizer AddedToken argument duplication are unrelated." + "reason": "Tie-embeddings API removal and FlashAttention support are unrelated regressions." }, { "left": "issue:42890", - "right": "issue:44964", + "right": "issue:45561", "accept": false, - "reason": "Test seeding flakiness versus a specific multimodal model load failure; not the same issue." + "reason": "Both are test flakiness, but one is missing set_seed and the other is an xdist file-race issue." }, { - "left": "issue:42890", - "right": "issue:44112", + "left": "issue:43064", + "right": "issue:44492", "accept": false, - "reason": "Both are CI/test failures, but one is seed hygiene and the other is a stale device-override expectation." + "reason": "Trainer optimizer-state corruption and a cache-strategy typo are unrelated." }, { - "left": "issue:42890", - "right": "issue:45600", + "left": "issue:43976", + "right": "issue:44938", "accept": false, - "reason": "Test flakiness versus auto_mappings referencing removed configs; unrelated." + "reason": "Both are Python compatibility reports, but for different versions and likely different root causes." }, { - "left": "issue:44222", - "right": "issue:45125", + "left": "issue:43867", + "right": "issue:45357", "accept": false, - "reason": "Both are MOE/tensor-parallel themed, but one is FP8 save_pretrained and the other is missing _tp_plan for loading; different bugs." + "reason": "Different model save/load regressions: state_dict ordering versus incorrect visual encoder keys." }, { - "left": "issue:43576", - "right": "issue:46132", + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training bug and tp-related OOM when loading a large model are not the same underlying issue." + }, + { + "left": "issue:43064", + "right": "issue:44655", "accept": false, - "reason": "A broken env CLI command is unrelated to AttentionInterface.register changing registered-function behavior." + "reason": "Wrong optimizer states during training is unrelated to saving Pipeline objects." }, { - "left": "issue:44355", - "right": "issue:44749", + "left": "issue:28282", + "right": "issue:33453", "accept": false, - "reason": "Compiled Python-file errors versus a slowdown after upgrading transformers; unrelated." + "reason": "Missing PyTorch ImportError and tokenizer-loading regression are unrelated." }, { - "left": "issue:42548", - "right": "issue:44351", + "left": "issue:43618", + "right": "issue:46285", "accept": false, - "reason": "Different missing top-level imports (PreTrainedModel vs HybridCache); same broad symptom, different underlying API breakage." + "reason": "Both are CLIP-related, but one is missing attentions output and the other is structural divergence between two CLIP text models." }, { - "left": "issue:45125", - "right": "issue:45478", + "left": "issue:43618", + "right": "issue:44857", "accept": false, - "reason": "Same broad Qwen3.5/MoE area, but one is tensor-parallel plan metadata and the other is from_pretrained failure; not clearly one fix." + "reason": "Attention-output regression and AMP/CUDA loss crash are different code paths and symptoms." }, { - "left": "issue:44387", - "right": "issue:44556", + "left": "issue:43618", + "right": "issue:44977", "accept": false, - "reason": "Both are upgrade/load regressions, but one is int4 CUDA reserved-memory/OOM and the other is checkpoint reload compatibility; too different." + "reason": "CLIP attentions regression is unrelated to Qwen3.5 flash-attention generation failure." }, { - "left": "issue:30333", - "right": "issue:33453", + "left": "issue:44360", + "right": "issue:45440", "accept": false, - "reason": "MLFlow reporting on failed training jobs is unrelated to tokenizer loading regression." + "reason": "DSA indexer implementation details are unrelated to DeepseekV3MoE diverging from the remote implementation." }, { - "left": "issue:45125", - "right": "issue:45375", + "left": "issue:43502", + "right": "issue:44062", "accept": false, - "reason": "Both are Qwen3.5/MoE issues, but one is tensor parallel planning and the other is a missing vision config field silently dropped by strict parsing." + "reason": "local_files_only network leakage and AddedToken keyword conflicts are different bugs." }, { - "left": "issue:43010", - "right": "issue:45278", + "left": "issue:42890", + "right": "issue:44964", "accept": false, - "reason": "Cache update no_grad behavior versus many import errors after version update are unrelated." + "reason": "Seed-related CI flakiness is unrelated to a specific multimodal model load failure." }, { - "left": "issue:30990", - "right": "issue:31795", + "left": "issue:42890", + "right": "issue:44112", "accept": false, - "reason": "Sentence Transformers loading hang versus model.forward documentation confusion; different domains entirely." + "reason": "Missing set_seed in tests and stale device override behavior are different CI issues." }, { - "left": "issue:43723", - "right": "issue:44556", + "left": "issue:42890", + "right": "issue:45600", "accept": false, - "reason": "Both involve loading after v5 upgrade, but one is tokenizer loading and the other is checkpoint reload compatibility; not the same bug." + "reason": "A flaky test seed issue is not the same as broken auto_mappings references causing CI failures." }, { - "left": "issue:43502", - "right": "issue:44222", + "left": "issue:44222", + "right": "issue:45125", "accept": false, - "reason": "Offline download behavior versus FP8 MOE save_pretrained; unrelated." + "reason": "FP8 save_pretrained behavior and a missing tensor-parallel plan are separate model-export/parallelism bugs." } ] }, @@ -15198,15 +15148,16 @@ "component_index": 1, "component_count": 18, "review_unit_index": 48, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:30333", + "issue:30990", "issue:31795", "issue:33453", "issue:34689", @@ -15221,40 +15172,51 @@ "issue:43502", "issue:43526", "issue:43576", + "issue:43723", "issue:43824", "issue:43827", "issue:43874", "issue:43901", "issue:43976", - "issue:44038", "issue:44222", "issue:44230", - "issue:44273", "issue:44315", + "issue:44351", "issue:44355", "issue:44360", "issue:44368", - "issue:44373", - "issue:44393", + "issue:44387", "issue:44556", "issue:44623", "issue:44655", + "issue:44749", "issue:44908", "issue:44938", "issue:44961", - "issue:45003", "issue:45071", "issue:45125", "issue:45230", + "issue:45278", + "issue:45375", "issue:45468", - "issue:45542", + "issue:45478", "issue:45656", - "issue:45753", "issue:45850", - "issue:46095", + "issue:46132", "issue:46209" ], "soft_pairs": [ + "issue:43576|issue:46132", + "issue:44355|issue:44749", + "issue:45125|issue:45478", + "issue:42548|issue:44351", + "issue:44387|issue:44556", + "issue:30333|issue:33453", + "issue:45125|issue:45375", + "issue:43010|issue:45278", + "issue:30990|issue:31795", + "issue:43723|issue:44556", + "issue:43502|issue:44222", "issue:42831|issue:43874", "issue:43010|issue:45850", "issue:43502|issue:44961", @@ -15266,9 +15228,9 @@ "issue:44315|issue:44360", "issue:33453|issue:34689", "issue:43299|issue:44556", + "issue:42757|issue:43576", "issue:43576|issue:45071", "issue:33453|issue:36296", - "issue:42757|issue:43576", "issue:43010|issue:45656", "issue:43576|issue:43824", "issue:31795|issue:38617", @@ -15281,242 +15243,264 @@ "issue:41628|issue:42548", "issue:36296|issue:43901", "issue:36296|issue:43526", - "issue:44623|issue:45230", - "issue:44273|issue:45542", - "issue:44038|issue:45125", - "issue:44273|issue:45003", - "issue:44393|issue:45071", - "issue:44230|issue:46095", - "issue:44373|issue:44961", - "issue:43576|issue:44556", - "issue:44393|issue:45753" + "issue:44623|issue:45230" ], - "prepared_review_unit_hash": "fb8fa068ed88d6b702cdb2d0faf3ccc017dad9bde0cc6a507c8388e4060afdc9", + "prepared_review_unit_hash": "e2b7280b697fa473078d4a3a5b5dd904c396eb613d827832bb7c0eac9f13d811", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 35, - "serialized_chars": 36938, - "estimated_input_tokens": 9235, - "estimated_eval_tokens": 18726 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38487, + "estimated_input_tokens": 9622, + "estimated_eval_tokens": 19500 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of unrelated Transformers issues, with a recurring theme of v5 regressions/import/loading breakages. None of the soft pairs look like true duplicates; they are adjacent symptom reports at best.", - "confidence": 0.64, - "canonical_issue_reason": "issue:45850 is the broadest umbrella report in the set and best matches the cluster\u2019s overall theme of Transformers v5 breakage, even though the underlying items are not true duplicates.", + "summary": "These issues are mostly a grab bag of unrelated Transformers regressions, import errors, docs bugs, and model-specific failures. The soft pairs are only superficially similar by wording; none look like the same concrete defect. The broadest representative issue is the v5 regression umbrella.", + "confidence": 0.63, + "canonical_issue_reason": "issue:45850 is the broadest umbrella here: it captures the general v5 breakage theme that many of the other regression-style reports superficially resemble.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45850 is the most representative anchor for this mixed cluster because it is the most general user-facing regression report; it is suitable only as a loose umbrella, not as a precise duplicate target.", + "best_issue_reason": "issue:45850 best represents the cluster because it is the most general, current, and central user-facing breakage report among otherwise unrelated issues.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Different bugs: broken `transformers env` CLI vs `AttentionInterface.register` behavior change." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated symptoms: compiled Python file errors vs a data-filtering slowdown after upgrade." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5 Moe-related, but one is missing tensor-parallel plan data and the other is a `from_pretrained` loading failure; not the same code path." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Different missing exports: `PreTrainedModel` vs `HybridCache`." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Memory/OOM regression under int4 quantization is not the same as failing to reload a checkpoint after upgrading." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow reporting on failed training jobs is unrelated to tokenizer loading regression." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Tensor-parallel plan omission is separate from `deepstack_visual_indexes` being dropped by strict config handling." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "`@torch.no_grad` cache update issue is unrelated to broad import failures after upgrade." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "A loading hang in Sentence Transformers is not the same as docs confusion around `input_ids` and `past_key_values`." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Both mention loading, but one is tokenizer loading in v5 and the other is checkpoint reloading after upgrade; different failures." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Unexpected hub API access with `local_files_only=True` is unrelated to FP8 `save_pretrained` behavior." + }, { "left": "issue:42831", "right": "issue:43874", "accept": false, - "reason": "Different failures: FP8 accuracy regressions vs a missing image-patch method causing an AttributeError." + "reason": "FP8 accuracy regression and a missing image-patch helper causing an AttributeError are different bugs." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "One is a no_grad decorator issue on cache/layer updates; the other is a vague broad v5 breakage complaint." + "reason": "A specific cache-update decorator bug is not the same as a broad claim that v5 breaks many models." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "local_files_only network leakage is unrelated to the unrelated 'racoon' report." + "reason": "`local_files_only` network access bug is unrelated to the placeholder issue titled `racoon`." }, { "left": "issue:43010", "right": "issue:44368", "accept": false, - "reason": "Different code paths: autograd/no_grad behavior vs a model config warning about tie_word_embeddings." + "reason": "Cache update no-grad behavior has nothing to do with a warning about `tie_word_embeddings`." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLFlow reporting on failed jobs is unrelated to a tensor-parallel training bug." + "reason": "MLFlow status reporting and tensor-parallel training behavior are distinct problems." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs still mentioning pipeline() after removals is unrelated to the 'racoon' issue." + "reason": "Docs still referencing `pipeline()` after removals is unrelated to the placeholder `racoon` issue." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled Python file loading errors and PreTrainedConfig type-checking are different regressions." + "reason": "Compiled-file errors and `PretrainedConfig` type-checking breakage are different failure modes." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "One is about running compiled Python files; the other is saving Pipeline objects." + "reason": "Compiled Python file handling is unrelated to saving Pipeline objects." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger kernel application in model_init and a DSA indexer ReLU omission are unrelated." + "reason": "Liger Kernel not applied with `model_init` is unrelated to a missing ReLU in the DSA indexer." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Both are loading-related, but one is tokenizer loading regression and the other is a specific Llama 3.2 vision model load breakage." + "reason": "Tokenizer-loading regression and Llama 3.2 Vision model-loading regression are different model/code-path issues." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Both concern model loading after version changes, but they describe different models and different breakpoints." + "reason": "A Qwen3VL MoE loading breakage is not the same as a general checkpoint reload regression after upgrade." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing `is_offline_mode` export and broken `transformers env` CLI are unrelated." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "Broken env CLI and PreTrainedConfig type checking are unrelated failures." + "reason": "Broken environment command and `PretrainedConfig` type-checking regression do not describe the same defect." }, { - "left": "issue:42757", - "right": "issue:43576", + "left": "issue:33453", + "right": "issue:36296", "accept": false, - "reason": "Missing is_offline_mode import from huggingface_hub is not the same as the v5 env command breakage." + "reason": "Tokenizer loading regression and tensor-parallel training bug are different issues." }, { - "left": "issue:43824", - "right": "issue:43576", + "left": "issue:43010", + "right": "issue:45656", + "accept": false, + "reason": "Cache update no-grad behavior is unrelated to DeepSpeed calling optimizer.step twice." + }, + { + "left": "issue:43576", + "right": "issue:43824", "accept": false, - "reason": "ImportError for a specific Qwen class is a different import-path issue than the env command bug." + "reason": "CLI `transformers env` breakage is unrelated to missing Qwen2.5-VL import symbols." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Docstring confusion for model.forward is unrelated to a missing configuration_utils symbol import." + "reason": "Documentation confusion and an `ImportError` for `layer_type_validation` are not the same bug." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "The 'racoon' issue and Gemma audio positional encoding bug are unrelated." + "reason": "Placeholder issue `racoon` has no clear relation to Gemma-4 audio positional encoding." }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "FP8 save_pretrained for MoE and FP8 inference support are related only by topic, not the same concrete bug." + "reason": "Both touch FP8/MoE, but one is a save/load bug and the other is an inference-support request; not mergeable as one fix." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "no_grad update behavior is unrelated to the Python 3.9+ compatibility failure." + "reason": "Cache update decorator behavior is unrelated to the Python version compatibility report." }, { "left": "issue:43576", "right": "issue:44938", "accept": false, - "reason": "The env CLI breakage and Python 3.14 load failure are different regressions." + "reason": "Broken CLI command and Python 3.14 load failure are separate problems." }, { "left": "issue:44908", "right": "issue:44961", "accept": false, - "reason": "Scheduler kwargs being ignored has nothing to do with the 'racoon' report." + "reason": "Scheduler kwargs bug and placeholder `racoon` issue do not match." }, { "left": "issue:44315", "right": "issue:46209", "accept": false, - "reason": "Liger kernel application and FP8 inference on B200 are separate bugs in different paths." + "reason": "Liger Kernel application with `model_init` is unrelated to FP8 inference failing on B200 GPUs." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "Import errors for AutoImageProcessor and PreTrainedModel are both import failures, but they involve different symbols and likely different fixes." + "reason": "Different import errors for different symbols from `transformers`." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor parallel training and text-classification docs mentioning return_all_scores are unrelated." + "reason": "Tensor-parallel training bug and pipeline docs mismatch are unrelated." }, { "left": "issue:36296", "right": "issue:43526", "accept": false, - "reason": "Tensor-parallel training bug and BeitImageProcessorFast reduce_labels behavior are different issues." + "reason": "Tensor-parallel training bug is unrelated to `reduce_labels` returning the wrong shape/value in BeitImageProcessorFast." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "Missing processor.save_pretrained files is a concrete bug; 'Bug report' is too vague and not the same issue." - }, - { - "left": "issue:44273", - "right": "issue:45542", - "accept": false, - "reason": "Lazy loading behavior and an undefined TensorFlow backend error are unrelated." - }, - { - "left": "issue:44038", - "right": "issue:45125", - "accept": false, - "reason": "Both involve Qwen3/Qwen3.5 MoE, but one is a broad v5 breakage and the other is a missing tensor-parallel plan." - }, - { - "left": "issue:44273", - "right": "issue:45003", - "accept": false, - "reason": "Lazy loading malfunction is not the same as unsafe sys.modules access in modeling_utils." - }, - { - "left": "issue:44393", - "right": "issue:45071", - "accept": false, - "reason": "Qwen3-VL bounding-box output errors are unrelated to PreTrainedConfig type-checking." - }, - { - "left": "issue:44230", - "right": "issue:46095", - "accept": false, - "reason": "FP8 inference support and deepseekv4 weight-conversion help are not the same bug or change." - }, - { - "left": "issue:44373", - "right": "issue:44961", - "accept": false, - "reason": "A wrong docstring for position_ids is unrelated to the 'racoon' report." - }, - { - "left": "issue:43576", - "right": "issue:44556", - "accept": false, - "reason": "The env command bug and checkpoint reload failures after upgrading versions are different concrete problems." - }, - { - "left": "issue:44393", - "right": "issue:45753", - "accept": false, - "reason": "2D bounding-box output errors and an infinite loop on a specific image are different Qwen3-VL failures." + "reason": "Missing files from `processor.save_pretrained(...)` is specific; the other item is an empty generic bug report." } ] }, @@ -15532,13 +15516,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 49, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:31795", "issue:33453", @@ -15547,49 +15531,57 @@ "issue:42890", "issue:43352", "issue:43408", - "issue:43519", "issue:43525", "issue:43572", "issue:43576", "issue:43673", - "issue:43704", "issue:43827", "issue:43874", "issue:43976", + "issue:44038", "issue:44077", + "issue:44230", "issue:44246", "issue:44273", "issue:44295", "issue:44297", "issue:44315", "issue:44351", - "issue:44360", + "issue:44373", "issue:44393", - "issue:44485", "issue:44492", "issue:44556", - "issue:44655", "issue:44704", "issue:44829", "issue:44861", "issue:44908", + "issue:44961", + "issue:45003", "issue:45071", "issue:45092", + "issue:45125", "issue:45230", "issue:45278", - "issue:45310", "issue:45341", "issue:45362", - "issue:45468", "issue:45478", + "issue:45542", + "issue:45753", "issue:45850", - "issue:46018", + "issue:46095", "issue:46143", "issue:46144", - "issue:46172", "issue:46218" ], "soft_pairs": [ + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44273|issue:45003", + "issue:44393|issue:45071", + "issue:44230|issue:46095", + "issue:44373|issue:44961", + "issue:43576|issue:44556", + "issue:44393|issue:45753", "issue:41669|issue:43673", "issue:43525|issue:45850", "issue:44556|issue:45092", @@ -15620,292 +15612,270 @@ "issue:41669|issue:44492", "issue:44315|issue:44829", "issue:42371|issue:44492", - "issue:43576|issue:44273", - "issue:44655|issue:44861", - "issue:44360|issue:45468", - "issue:45230|issue:45310", - "issue:44273|issue:44908", - "issue:43576|issue:46218", - "issue:43827|issue:44492", - "issue:46018|issue:46172", - "issue:43519|issue:44485", - "issue:43519|issue:44908", - "issue:41669|issue:43704" + "issue:43576|issue:44273" ], - "prepared_review_unit_hash": "464655914bbdeed4bb051b991e83d18544b4cbc84f25baf83a41b8d265547bba", + "prepared_review_unit_hash": "44c8d08d29d64678ff928112b2f1aab3bdb582492d4a4b4add479835883b1be8", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 41, - "serialized_chars": 39264, - "estimated_input_tokens": 9816, - "estimated_eval_tokens": 19888 + "soft_pair_count": 39, + "serialized_chars": 38785, + "estimated_input_tokens": 9697, + "estimated_eval_tokens": 19650 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of unrelated Transformers v5 regressions and model-specific bugs. The soft pairs share broad themes like import/loading breakage, but they are not the same concrete underlying issue, so none should be merged as duplicates.", - "confidence": 0.71, - "canonical_issue_reason": "issue:45850 is the broadest current umbrella for the cluster\u2019s dominant theme (v5 breaking multiple models), so it is the best issue to anchor triage, even though many other items are distinct bugs.", + "summary": "Mostly a noisy cluster of unrelated issues. The only coherent theme is import/lazy-loading/v5 import regressions, but none of the soft pairs are close enough to treat as duplicates or the same fix.", + "confidence": 0.86, + "canonical_issue_reason": "issue:44273 is the broadest umbrella for the import/lazy-loading regression theme and best matches the shared symptom family across the relevant issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45850 is open, broad, and representative of the recurring v5 compatibility complaints; it is the best single issue for organizing related reports, but not a precise duplicate target for the others.", + "best_issue_reason": "issue:41669 is the most concrete root-cause report in the cluster: it identifies import * usage as the source of the import-time slowdown and is the most actionable issue to anchor the duplicate set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Both are runtime/import-related at a high level, but one is about lazy loading and the other is a TensorFlow backend warning path; different bugs and fixes." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Different model families and different failures: Qwen3-VL-Moe inference bug versus missing tensor-parallel plan in Qwen3.5 MoE." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "One is a broad lazy-loading complaint; the other is a specific sys.modules access bug in modeling_utils. Related area, not the same underlying defect." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Completely different surfaces: Qwen3-VL 2D bounding-box output bug versus PretrainedConfig type-checking breakage." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "Different features and code paths: FP8 inference support for Qwen models versus a DeepSeek weight-conversion question." + }, + { + "left": "issue:44373", + "right": "issue:44961", + "accept": false, + "reason": "A docstring typo about position_ids is unrelated to the placeholder 'racoon' issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Both mention v5, but one is the transformers env command and the other is checkpoint reload compatibility; not the same bug." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen3-VL/Qwen3.5, but the symptoms differ: 2D box output failure versus an infinite-loop case on a specific image." + }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Different problems: import-star performance/file explosion vs missing GenerationMixin cache during chunked prefill." + "reason": "Import-performance regression versus missing GenerationMixin cache during chunked prefill; different root causes and fixes." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "Specific Llama4Config pad_token_id AttributeError is not the same as a broad v5 models-broken report." + "reason": "Specific Llama4Config pad_token_id error versus a broad complaint that v5 breaks many models." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both are v5 compatibility regressions, but they affect different model/load paths and are not the same bug." + "reason": "Both are upgrade-compatibility problems, but they affect different model-loading paths and are not clearly the same defect." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "StableLmConfig missing pad_token_idx is a specific config regression, not the same as a general breakage tracker." + "reason": "StableLmConfig pad_token_idx regression is a specific config issue, not the same as the broad v5 breakage complaint." }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Qwen3.5 chat crash is model/chat specific and not the same underlying issue as a general v5 breakage report." + "reason": "A Qwen3.5 chat crash is a model-specific failure, not the same as the generic 'v5 breaks too many models' report." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import-star slowdown/file explosion is unrelated to a Qwen3.5 import failure." + "reason": "Import * slowdown and a missing Qwen3_5ForCausalLM import error may both involve imports, but the underlying problems are different." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Long import time and lazy loading not working are related only at a high level; they describe different failures." + "reason": "Both concern import behavior, but one reports intermittent slow imports while the other claims lazy loading is broken; not enough evidence of the same bug." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Missing set_seed in tests is unrelated to inverse_sqrt scheduler kwargs handling." + "reason": "Test seed omissions in SAM HQ tests have nothing to do with the inverse_sqrt scheduler ignoring kwargs." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model type mismatch warning is unrelated to Liger Kernel not being applied with model_init." + "reason": "Model type mismatch warning for SAM3 versus Liger Kernel not applied in model_init are unrelated." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor kwargs forwarding and scheduler kwargs are separate code paths." + "reason": "AutoProcessor cached_file kwargs forwarding and scheduler kwargs handling are separate code paths." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention 2 support error for Nemotron and Qwen3-VL bbox output issues are unrelated." + "reason": "Flash Attention 2 support error in Nemotron is unrelated to Qwen3-VL 2D bounding-box output behavior." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer_class mismatch on save_pretrained is not the same as a generic bug report." + "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to a generic bug report placeholder." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version compatibility failure is unrelated to Qwen3-VL bounding-box output errors." + "reason": "Python version compatibility regression is unrelated to the Qwen3-VL bounding-box bug." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "position_ids buffer access error is unrelated to the generic bug report." + "reason": "Position_ids buffer access error is unrelated to a generic bug report placeholder." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "Broken transformers env command and scheduler kwargs handling are different features." + "reason": "Broken env command and inverse_sqrt scheduler kwargs are separate issues." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler kwargs bug and PretrainedConfig type-checking breakage are unrelated regressions." + "reason": "Scheduler keyword handling and PretrainedConfig type-checking are unrelated." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-star slowdown is unrelated to HybridCache import failure." + "reason": "Import * cleanup and missing HybridCache export may both touch imports, but they are different missing-symbol/import surfaces." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import-star performance issue is unrelated to a testing_utils bug." + "reason": "Import-time slowdown is unrelated to a small testing_utils bug." }, { "left": "issue:45278", "right": "issue:45850", "accept": false, - "reason": "Both are broad v5 breakage reports, but one is about import errors and the other is a generic model-breakage complaint." + "reason": "General import errors after upgrade and a broad claim that v5 breaks many models are too vague to be the same underlying issue." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Missing GenerationMixin cache and a typo in cache strategy docs/code are not the same defect." + "reason": "Chunked-prefill cache regression and a typo in cache strategies are unrelated." }, { "left": "issue:46143", "right": "issue:46144", "accept": false, - "reason": "RoFormer kwargs passthrough and attention-interface usage are adjacent but distinct implementation issues." + "reason": "Missing kwargs passthrough and attention-interface noncompliance in RoFormer are related to the same model family but not the same concrete defect." }, { "left": "issue:43874", "right": "issue:45850", "accept": false, - "reason": "GLM46V image processor method missing is a specific multimodal token bug, not a general v5 breakage report." + "reason": "A specific GLM46V image-processor method bug is not the same as the broad v5 breakage complaint." }, { "left": "issue:44393", "right": "issue:45478", "accept": false, - "reason": "Qwen3-VL bbox hallucination and Qwen3.5 MoE from_pretrained error are different model bugs." + "reason": "Different Qwen-related failures with different symptoms and code paths." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init policy and _tied_weights_keys list handling are unrelated model/config issues." + "reason": "Optional post_init validation and tied-weight-key AttributeError are distinct issues." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "Python version support issue is unrelated to Qwen3.5 import failure." + "reason": "Python version incompatibility and a missing Qwen3_5 import are separate failures." }, { "left": "issue:43827", "right": "issue:44908", "accept": false, - "reason": "Docs still referencing pipeline() and scheduler kwargs are unrelated." + "reason": "Documentation still referencing pipeline() is unrelated to scheduler kwargs handling." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion around model.forward arguments is unrelated to tokenizer loading regression." + "reason": "Model.forward documentation confusion and tokenizer loading regression are unrelated." }, { "left": "issue:41669", "right": "issue:44492", "accept": false, - "reason": "Import-star slowdown/file explosion is unrelated to a cache-strategy typo." + "reason": "Import * performance regression and a cache-strategy typo are not the same bug." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "Liger Kernel application with model_init and flash_attention_3 training degeneration are different bugs." + "reason": "Liger Kernel not applied in model_init and flash_attention_3 causing degenerate training are different changes/issues." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 settings migration and a cache-strategy typo do not share an underlying code-path bug." + "reason": "TF32 API guidance and a cache-strategy typo are unrelated." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "Broken env command and lazy loading malfunction are separate issues." - }, - { - "left": "issue:44655", - "right": "issue:44861", - "accept": false, - "reason": "Pipeline save_pretrained failure is unrelated to tied-weight key handling." - }, - { - "left": "issue:44360", - "right": "issue:45468", - "accept": false, - "reason": "DSA indexer/ReLU discussion is unrelated to Gemma4 audio relative positional encoding." - }, - { - "left": "issue:45230", - "right": "issue:45310", - "accept": false, - "reason": "Generic bug report is not the same concrete issue as Qwen3.5 MoE from_pretrained failure." - }, - { - "left": "issue:44273", - "right": "issue:44908", - "accept": false, - "reason": "Lazy loading malfunction and inverse_sqrt scheduler kwargs are unrelated." - }, - { - "left": "issue:43576", - "right": "issue:46218", - "accept": false, - "reason": "Broken env command is unrelated to Qwen3.5 import failure." - }, - { - "left": "issue:43827", - "right": "issue:44492", - "accept": false, - "reason": "Docs referencing pipeline() and a cache-strategy typo are not the same bug." - }, - { - "left": "issue:46018", - "right": "issue:46172", - "accept": false, - "reason": "DeepSeek-V4 gating logic and a dtype/transpose bug in DeepSeek-V4 are distinct code-path problems." - }, - { - "left": "issue:43519", - "right": "issue:44485", - "accept": false, - "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are unrelated." - }, - { - "left": "issue:43519", - "right": "issue:44908", - "accept": false, - "reason": "Timestamp calculation bug and scheduler kwargs handling are unrelated." - }, - { - "left": "issue:41669", - "right": "issue:43704", - "accept": false, - "reason": "Import-star slowdown/file explosion is unrelated to a VRAM leak in Qwen3ForCausalLM." + "reason": "Broken transformers env command and lazy loading issues overlap on version context only; they do not look like the same concrete bug." } ] }, @@ -15921,13 +15891,13 @@ "component_index": 1, "component_count": 18, "review_unit_index": 50, - "review_unit_count": 50, + "review_unit_count": 51, "status": "reviewed", "reason": null, "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-10000-438", + "cluster_id": "cluster-10000-439", "nodes": [ "issue:30333", "issue:31515", @@ -15935,35 +15905,39 @@ "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:43576", "issue:43606", "issue:43668", "issue:43704", "issue:43717", - "issue:43813", + "issue:43827", "issue:44031", "issue:44230", "issue:44261", "issue:44263", + "issue:44273", "issue:44351", + "issue:44360", "issue:44485", + "issue:44492", "issue:44623", + "issue:44655", "issue:44861", "issue:44908", "issue:44998", + "issue:45230", "issue:45290", "issue:45310", "issue:45468", @@ -15971,9 +15945,20 @@ "issue:46018", "issue:46095", "issue:46144", + "issue:46172", "issue:46218" ], "soft_pairs": [ + "issue:44360|issue:45468", + "issue:44655|issue:44861", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:43576|issue:46218", + "issue:43827|issue:44492", + "issue:46018|issue:46172", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:41669|issue:43704", "issue:31795|issue:34689", "issue:31515|issue:31795", "issue:44485|issue:45468", @@ -16007,297 +15992,418 @@ "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" + "issue:43668|issue:44031" ], - "prepared_review_unit_hash": "c3fb829db32280e187e2388d5dbc53ec07a75656b4088cbbcbe962055b9396cc", + "prepared_review_unit_hash": "614ce861ade4edddad97d1c3049cda05ce228da564bbd3fc7eafdf10e05eb78c", "component_budget": { - "node_count": 438, - "item_count": 438, - "soft_pair_count": 1611, - "serialized_chars": 835229, - "estimated_input_tokens": 208808, - "estimated_eval_tokens": 417872 + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 }, "budget": { - "node_count": 43, - "item_count": 43, - "soft_pair_count": 43, - "serialized_chars": 36958, - "estimated_input_tokens": 9240, - "estimated_eval_tokens": 18736 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 44, + "serialized_chars": 39956, + "estimated_input_tokens": 9989, + "estimated_eval_tokens": 20234 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is very heterogeneous: loading/regression, tokenizer, config, training, multimodal, and scheduler issues are mixed together. None of the soft-similarity pairs look like the same concrete bug/change, so I would not merge any of them as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "Issue 45850 is the broadest umbrella report in the set (\u201cSince 5.0 version it breaks too many models\u201d), so it is the least-specific representative of the loose theme here.", + "summary": "This cluster is heterogeneous: most items are separate regressions, API changes, docs issues, or model-specific bugs that only share a broad Transformers/v5 similarity. I found no soft-edge pair that clearly describes the same underlying defect/change, so all candidates should be rejected.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45850 is the broadest representative of the cluster\u2019s recurring v5 breakage theme, even though it is more of an umbrella complaint than a precise duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "45850 is the best single representative only because it is the most generic version-regression report; the rest are too specific and largely unrelated.", + "best_issue_reason": "issue:45850 best matches the overall cluster topic because it captures the general \u201cv5 breaks many models\u201d regression pattern seen across several items.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "Different model families and failures: DSA indexer/ReLU vs Gemma audio positional encoding." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "One is save_pretrained for Pipeline objects; the other is _get_tied_weight_keys crashing on a list." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report vs a specific Qwen3.5 MoE from_pretrained error; no concrete match." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading behavior and inverse_sqrt scheduler kwargs are unrelated code paths." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command vs import failure for Qwen3_5ForCausalLM are different issues." + }, + { + "left": "issue:43827", + "right": "issue:44492", + "accept": false, + "reason": "Docs referencing pipeline() and a cache-strategy typo are not the same defect." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek-V4 shared expert gating vs a deepseek-v4 tensor transpose issue are distinct." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are separate model bugs." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp math and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star cleanup in models does not match a VRAM leak in Qwen3ForCausalLM." + }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Documentation confusion vs a specific model-loading regression; same area, not the same bug." + "reason": "Documentation confusion and a Llama 3.2 model-loading regression are different problems." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Slow checkpoint loading and forward-API documentation are unrelated." + "reason": "Slow checkpoint loading is not the same as confusing forward docs." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "Different models and different math paths: RoPE implementation vs audio positional encoding." + "reason": "RoPE implementation and Gemma4AudioRelPositionalEncoding are distinct model-specific changes." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak in dataloader-thread use is unrelated to GLM RoPE behavior." + "reason": "A VRAM leak in multi-threaded loading and a RoPE bug do not share the same failure mode." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Continuous batching audio support and DeepSeek weight conversion are different problems." + "reason": "Continuous batching support for audio models is unrelated to a DeepSeek-V4 weight conversion script request." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "Generic v5 regression complaint vs a specific import failure for one model." + "reason": "Broad v5 breakage complaint vs a specific Qwen3_5 import failure; too coarse to merge." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Local-files-only network leakage and Qwen3VL timestamp math are unrelated." + "reason": "local_files_only network requests and Qwen3VL timestamp calculation are unrelated." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 settings guidance vs a VRAM leak; not the same bug." + "reason": "TF32 settings API guidance vs VRAM leakage are different issues." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Timestamp calculation bug vs missing saved processor files; different code paths." + "reason": "Processor timestamp logic and missing save_pretrained files are separate bugs." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Qwen3VL timestamp issue and Qwen3 VRAM leak are unrelated." + "reason": "Different components: Qwen3VL processor timestamps vs Qwen3ForCausalLM memory leak." }, { "left": "issue:43606", "right": "issue:45468", "accept": false, - "reason": "CPU offload device mismatch vs Gemma4 audio positional encoding; different failures." + "reason": "CPU offload device mismatch in bark-small is unrelated to Gemma audio positional encoding." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "Indexer split-return handling vs Gemma4 audio encoding; no concrete overlap." + "reason": "torch.split return values in GlmMoeDsaIndexer do not match Gemma audio positional encoding." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "Missing rms_norm_eps in MLA q_a_layernorm is not the same as Gemma4 audio encoding." + "reason": "MLA q_a_layernorm epsilon precision issue is a different model-specific bug." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "Different import errors for different symbols/models." + "reason": "HybridCache import failure and Qwen3_5ForCausalLM import failure are different symbols and code paths." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Flash-Attn2 unsupported error vs module import failure; separate issues." + "reason": "Flash Attention support gap in NemotronHForCausalLM is not the same as Qwen3_5 import failure." }, { "left": "issue:42913", "right": "issue:43066", "accept": false, - "reason": "Both mention tokenizer v5, but one is behavioral drift and the other is decoder-type mismatch; too broad to treat as one bug." + "reason": "Both are tokenizer-related but describe different v5 symptoms and likely different code paths." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Scheduler kwargs handling is unrelated to the placeholder 'Unemployment' issue." + "reason": "Scheduler kwargs bug and an unrelated non-technical issue are not duplicates." }, { "left": "issue:44861", "right": "issue:45290", "accept": false, - "reason": "Tied-weight-key attribute crash vs chat-template tool-call crash; different paths." + "reason": "_tied_weights_keys list crash and apply_chat_template tool-call crash are unrelated." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Missing pad_token_id on Llama4Config is unrelated to tied-weight-key list handling." + "reason": "Missing pad_token_id in Llama4Config is not the same as tied-weights key handling." }, { - "left": "issue:43572", - "right": "issue:44861", + "left": "issue:43704", + "right": "issue:44998", "accept": false, - "reason": "StableLm config pad_token_idx regression and tied-weight-key crash are separate." + "reason": "VRAM leak vs unrelated non-technical issue." }, { "left": "issue:41669", "right": "issue:44998", "accept": false, - "reason": "Import-star cleanup and the unrelated placeholder issue do not match." + "reason": "Import-star cleanup issue vs unrelated non-technical issue." }, { "left": "issue:43066", "right": "issue:44998", "accept": false, - "reason": "Tokenizer decoder-type regression and the placeholder issue are unrelated." - }, - { - "left": "issue:30333", - "right": "issue:31515", - "accept": false, - "reason": "MLflow status reporting on failed training jobs is unrelated to slow from_pretrained checkpoint loading." + "reason": "Tokenizer decoder type bug vs unrelated non-technical issue." }, { "left": "issue:31515", - "right": "issue:36296", + "right": "issue:38617", "accept": false, - "reason": "Checkpoint load performance and tensor-parallel training bug are different issues." + "reason": "Slow from_pretrained checkpoint loading and missing layer_type_validation import are different." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 support request and missing processor save files do not share the same bug." + "reason": "fp8 support request for qwen3-vl/qwen3.5 MoE and missing processor save files are unrelated." }, { "left": "issue:44998", "right": "issue:46218", "accept": false, - "reason": "The placeholder issue is unrelated to a model import failure." + "reason": "Unrelated non-technical issue vs Qwen3_5 import failure." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "Failed-job MLflow reporting and slow checkpoint loading are unrelated." + "reason": "MLFlow job-status reporting and checkpoint load slowness are unrelated." }, { "left": "issue:31515", - "right": "issue:33453", + "right": "issue:36296", "accept": false, - "reason": "Both touch loading, but one is performance and the other is a tokenizer-loading regression." + "reason": "Checkpoint loading slowness and tensor-parallel training bug are different concerns." }, { - "left": "issue:43298", - "right": "issue:44230", + "left": "issue:44230", + "right": "issue:45310", "accept": false, - "reason": "Continuous batching audio support and FP8 inference support are different feature requests." + "reason": "fp8 inference support request and Qwen3.5 MoE from_pretrained failure are not the same bug." }, { "left": "issue:31515", "right": "issue:33453", "accept": false, - "reason": "Checkpoint load slowness and tokenizer regression are not the same underlying defect." + "reason": "Checkpoint loading performance regression and tokenizer loading regression are distinct." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Audio-model continuous batching support and qwen3-vl/qwen3.5 fp8 inference support are different feature gaps." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba init-weights behavior and RoFormer attention interface are unrelated model-specific changes." + "reason": "Mamba init_weights usage and RoFormer attention interface are unrelated model implementations." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former ignore_value usage and DeepSeek-V4 gating are different model bugs." + "reason": "Mask2former ignore_value handling is not the same as DeepSeek-V4 shared-expert gating." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "Type-hint typo and tokenizer regex warning are unrelated." - }, + "reason": "Incorrect type hint in ModernBERTConfig and tokenizer regex 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": 18, + "review_unit_index": 51, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46144" + ], + "soft_pairs": [ + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "2b10f80c507cdcef3e1eb7c83b3f0fac260952bb51775e571bc950351fc6d7b7", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 9, + "serialized_chars": 6639, + "estimated_input_tokens": 1660, + "estimated_eval_tokens": 3576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are not duplicates; they span unrelated areas including a norm-layer question, Mask2Former ignore_value handling, loss kwargs propagation, a config type hint, a typo, tokenizer regex warnings, and RoFormer attention interface usage.", + "confidence": 0.96, + "canonical_issue_reason": "No true canonical duplicate exists in this set. If one issue must anchor the cluster, issue 43240 is the broadest/general framework bug, but it is still unrelated to the other reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43240 is the closest thing to a cluster anchor because it concerns generic loss-call plumbing rather than a single model-specific typo or configuration detail.", + "best_pr_reason": null, + "soft_edge_verdicts": [ { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Mask2Former loss handling and kwargs propagation to loss are different code paths." + "reason": "Mask2Former ignore_value handling and loss-kwargs propagation are different bugs in different code paths." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss-kwargs propagation and RoFormer attention interface are unrelated." + "reason": "Loss kwargs forwarding and RoFormer attention interface usage are unrelated implementation issues." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former ignore_value bug and tokenizer regex warning do not match." + "reason": "A Mask2Former loss-parameter bug is not the same as tokenizer regex warning behavior." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "No concrete overlap between Mask2Former ignore_value handling and RoFormer attention." + "reason": "These concern different subsystems: segmentation loss handling vs attention interface usage in RoFormer." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Mask2Former ignore_value and a quantization-operation typo are unrelated." + "reason": "Mask2Former ignore_value logic and a typo in a quantization attribute are unrelated." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Passing kwargs into loss and a config type-hint issue are different bugs." + "reason": "Passing kwargs to a loss function and an incorrect type hint in ModernBERTConfig are not the same underlying bug." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Type-hint correction and quantization typo are unrelated." + "reason": "A config type-hint error is unrelated to a typo in a quantization field name." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Llama4 norm implementation question and ModernBERT norm_eps typing are different concerns." + "reason": "A question about Llama4TextL2Norm vs RMSNorm is unrelated to a ModernBERTConfig annotation problem." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value bug and ModernBERT type hint are unrelated." + "reason": "Mask2Former loss-parameter handling and a ModernBERT config typing issue do not share the same code path or fix." } ] }, @@ -16330,12 +16436,12 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "8ac245787a8f2cf04fdc37283afd78fe219e2f317792e6efd443897721d26be9", + "prepared_review_unit_hash": "5ff925bcca17d8cae1050ee7f139f04be2298d337c4cdda928b13aeb67b7837d", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2875, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -16343,7 +16449,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2875, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -16352,30 +16458,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three unrelated issues: one about tokenizer truncation behavior, one about scheduler functionality, and one about Wav2Vec2 word boundary timestamps. The soft similarity signal is not enough to treat any pair as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No single issue is a canonical duplicate target because the three tickets cover different features/subsystems and different user problems.", + "summary": "These three issues are unrelated: one is about truncation behavior in tokenization, one about ReduceLROnPlateau-like scheduler functionality, and one about Wav2Vec2 word time offsets. No duplicate cluster emerges.", + "confidence": 0.98, + "canonical_issue_reason": "Issue #10519 is the clearest concrete product request and best represents the only coherent topic in the set; the other two issues are unrelated feature requests in different areas.", "canonical_pr_reason": null, - "best_issue_reason": "None of the issues is suitable as a global representative for the cluster since they are not the same underlying problem.", + "best_issue_reason": "#10519 is the most specific and actionable issue; it is the best standalone representative, but not a duplicate of the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Tokenizer truncation-from-beginning request is unrelated to Wav2Vec2 word start/end timestamp extraction." + "reason": "Token truncation options and Wav2Vec2 word boundary timestamps are different features with no shared code-path or concrete bug overlap." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "Truncation strategy behavior and ReduceLROnPlateau-like scheduler functionality are different features with no shared code-path problem." + "reason": "Truncation-from-beginning is unrelated to ReduceLROnPlateau-like scheduler behavior; they concern different subsystems and problems." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "A learning-rate scheduler feature request is unrelated to Wav2Vec2 word timing offsets." + "reason": "A learning-rate scheduler feature request and Wav2Vec2 time-offset extraction are unrelated issues with no plausible duplication." } ] }, @@ -16407,7 +16513,7 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "25a83c19d695655ee8395804412cda0170386575b99adf22250968812138bd12", + "prepared_review_unit_hash": "abe46e12df50bd2b8975547c25fe78a4b98496b57276d0e9899e0c931253d464", "component_budget": { "node_count": 3, "item_count": 3, @@ -16429,24 +16535,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues do not appear to describe the same bug: one is about Tapas tokenization differences, one is about a misleading login/auth error when downloading a private model, and one is a generic model-loading failure for bert-base-uncased.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster looks mixed: one Tapas tokenization issue and two model-loading issues. The two model-loading tickets are superficially similar, but the available evidence is not strong enough to treat them as the same concrete bug, so I would not merge any soft edges.", + "confidence": 0.84, + "canonical_issue_reason": "Issue #16618 is the broadest and most central model-loading report in the set, with the most discussion and the clearest user-facing failure around loading a pretrained model.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "#16618 is the best representative of the cluster because it captures the main model-loading problem theme, whereas #13244 is unrelated and #15980 is narrower and more specific to private-model login/error messaging.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tapas tokenization mismatch is a different code path and symptom than failing to load bert-base-uncased; they do not look like the same underlying bug." + "reason": "Different bug class and code path: Tapas tokenization behavior versus inability to load a pretrained model." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "This is specifically about the error message shown when downloading a private model without being logged in, which is distinct from a general model-loading failure." + "reason": "Both concern model access/loading, but one is specifically about a private-model download/auth error and the other is a generic failure to load 'bert-base-uncased'; not enough evidence they are the same bug." } ] }, @@ -16476,12 +16582,12 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "34add8b1dc85a7b39be5e2efd36a51c4321c8a2fa5b94b3b6ac3783a477c2fe1", + "prepared_review_unit_hash": "89dea94304e87b13a1de33358b66b54f86d807632925567de5212ed71dd7aa1e", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -16489,7 +16595,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -16498,18 +16604,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are unrelated feature requests: one for Flash Attention 2 support across more architectures, the other for GGUF support across more architectures. They do not describe the same underlying bug or change, so they should not be clustered as duplicates.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "These are two separate community contribution feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They are only loosely similar at a high level and should not be treated as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 26350 is the better anchor only because it is older and has substantially more discussion activity, but it is not a duplicate of 33260.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 26350 is the strongest representative issue in this pair due to age and higher engagement, though the topics are distinct.", "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); shared wording about \"more architectures\" is too broad to indicate the same change." + "reason": "Different requested features in different inference/runtime paths: Flash Attention 2 vs GGUF support. The similarity is only broad subsystem-level language about architectures, not the same underlying bug or change." } ] }, @@ -16544,7 +16650,6 @@ ], "soft_pairs": [ "issue:27088|issue:44370", - "issue:27088|issue:46222", "issue:27088|issue:44365", "issue:27088|issue:43834", "issue:27088|issue:38435", @@ -16554,13 +16659,13 @@ "issue:28936|issue:38435", "issue:28936|issue:44370", "issue:38435|issue:44370", - "issue:38435|issue:46222", "issue:38435|issue:44365", "issue:38435|issue:43834", + "issue:38435|issue:46222", "issue:38435|issue:46226", "issue:43834|issue:44370", - "issue:43834|issue:46222", "issue:43834|issue:44365", + "issue:43834|issue:46222", "issue:43834|issue:46226", "issue:44365|issue:44370", "issue:44365|issue:46222", @@ -16569,191 +16674,188 @@ "issue:44370|issue:46226", "issue:46222|issue:46226" ], - "prepared_review_unit_hash": "5629030469b56ab6d8f2fcb8d4503dbcb60ab0c11d6d6da377c77abe19db07cc", + "prepared_review_unit_hash": "252ce62d8d69026efe4b1c8607c6e7193f685d2145e253ea524fc363b335a77d", "component_budget": { "node_count": 8, "item_count": 8, - "soft_pair_count": 25, - "serialized_chars": 13680, - "estimated_input_tokens": 3420, - "estimated_eval_tokens": 7096 + "soft_pair_count": 24, + "serialized_chars": 13529, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 }, "budget": { "node_count": 8, "item_count": 8, - "soft_pair_count": 25, - "serialized_chars": 13680, - "estimated_input_tokens": 3420, - "estimated_eval_tokens": 7096 + "soft_pair_count": 24, + "serialized_chars": 13529, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are mostly i18n translation-request issues, but they look like separate language-specific requests or repeated templates rather than true duplicates. No PRs are present.", - "confidence": 0.79, - "canonical_issue_reason": "issue:46222 is the best cluster anchor: it is open, has inbound references, and sits at the center of the near-identical template-style issues.", + "summary": "This cluster is mostly boilerplate-heavy i18n translation requests with strong textual overlap. The most defensible duplicate is the Thai pair 46222/46226, which has identical visible titles and a jaccard score of 1.0; the packet does not expose the body text, so it should be treated as a likely duplicate rather than an exact one. The other pairs look very similar, but the available evidence is still too template-driven to merge confidently.", + "confidence": 0.81, + "canonical_issue_reason": "issue:46222 is the best canonical issue for the one highly grounded duplicate pair: it is the earlier of 46222/46226 and the packet shows identical visible titles plus jaccard 1.0. Since the body text is not exposed, this is best described as a likely duplicate rather than a proven exact duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46222 is the strongest representative by activity/centrality, though the cluster itself does not look like one concrete duplicate issue.", + "best_issue_reason": "issue:46222 is the strongest canonical issue candidate because the 46222/46226 pair has the clearest duplicate signal in the packet: same visible title, identical body length, and jaccard 1.0, with 46222 appearing first.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Both are translation-request issues, but there is no evidence they are the same language or same underlying request; the similarity is only template-level." - }, - { - "left": "issue:27088", - "right": "issue:46222", - "accept": false, - "reason": "Turkish vs generic placeholder translation request: same broad i18n theme, but not the same concrete issue." + "reason": "Very similar translation-template issues, but the packet only shows normalized placeholder-style metadata and no body text; I can't confirm they are the same concrete request." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different language-specific request versus generic template issue; too little evidence of a shared underlying bug/change." + "reason": "High similarity appears template-driven, but there is not enough exposed content to prove the underlying request is identical." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "High textual similarity likely comes from shared translation-template wording, not the same specific request." + "reason": "These look like related translation-request templates, but the available evidence is still too indirect to treat them as duplicates." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Turkish and Romanian translation requests are distinct issues, not duplicates." + "reason": "Strong shared i18n wording, but the packet does not expose enough request-specific content to merge these issues confidently." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish and Spanish translation requests are separate language tasks, so they should not be merged." + "reason": "Same broad docs-translation pattern, but the evidence shown is template-level similarity rather than the same underlying request." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish-specific request vs generic placeholder issue; no concrete match on the underlying request." + "reason": "Both are highly similar translation issues, but the packet lacks the underlying body text needed to confirm they are the same request." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "The overlap is only in the translation template, not in a shared language-specific issue." + "reason": "The overlap looks like shared boilerplate; without the actual body text, I would not merge them as duplicates." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Spanish and Romanian translation requests are different tasks and not duplicates." + "reason": "Very similar i18n translation-template issues, but the exposed data is not specific enough to establish the same underlying task." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Generic template similarity is not enough to show the same underlying issue." + "reason": "High similarity alone is not enough here; the packet does not expose the concrete request text needed to confirm duplication." }, { "left": "issue:38435", - "right": "issue:43834", + "right": "issue:44370", "accept": false, - "reason": "Romanian-specific request and a generic placeholder issue are not the same concrete issue." + "reason": "These are closely related translation-request issues, but the similarity appears to come from shared boilerplate rather than a confirmed duplicate." }, { "left": "issue:38435", - "right": "issue:44370", + "right": "issue:44365", "accept": false, - "reason": "These look like separate translation requests sharing boilerplate text only." + "reason": "The packet suggests a template match, but it does not show enough underlying content to safely merge them." }, { "left": "issue:38435", - "right": "issue:46222", + "right": "issue:43834", "accept": false, - "reason": "No evidence they refer to the same language or same translation task; reject as template-only similarity." + "reason": "Strong textual overlap, but the available evidence is still not sufficient to prove identical underlying content." }, { "left": "issue:38435", - "right": "issue:44365", + "right": "issue:46222", "accept": false, - "reason": "The match is superficial; these are not the same concrete request." + "reason": "These may be related i18n template issues, but the exposed metadata is not enough to confirm they are the same request." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "The Thai text issue is unrelated to the translation-request template cluster." + "reason": "High similarity, but the packet does not expose the body text required to verify a duplicate relationship." }, { "left": "issue:43834", "right": "issue:44370", "accept": false, - "reason": "Both use the same placeholder template, but that does not show they are the same issue instance." + "reason": "The titles are placeholder-like and the similarity is high, but the packet does not provide enough concrete content to confidently merge them." }, { "left": "issue:43834", - "right": "issue:46222", + "right": "issue:44365", "accept": false, - "reason": "Shared template text is insufficient; no concrete evidence of a single underlying bug/change." + "reason": "Despite near-identical similarity scores, the available evidence is still too template-driven to call this a duplicate with confidence." }, { "left": "issue:43834", - "right": "issue:44365", + "right": "issue:46222", "accept": false, - "reason": "These look like separate template-generated translation requests rather than duplicates." + "reason": "Very close text match, but the visible titles differ and the packet does not expose the body text, so I would not merge them confidently." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "The Thai-language issue is not the same underlying translation request as the placeholder template issues." + "reason": "High similarity alone is not enough here; the issue bodies are not visible, so the duplicate claim is not well grounded." }, { "left": "issue:44365", "right": "issue:44370", "accept": false, - "reason": "Near-identical wording appears templated, but there is no proof they target the same concrete request." + "reason": "These are extremely similar, but without the underlying body text I can't be sure they are the same concrete issue." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Same template family, but not enough to establish a true duplicate." + "reason": "The similarity looks template-based, but the packet does not expose enough request-specific content to confirm duplication." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Different content and likely different request; reject." + "reason": "Very close similarity, but the evidence shown is still not sufficient to merge them as duplicates." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Identical placeholder titles likely reflect a reused template, not the same underlying issue." + "reason": "The overlap is strong, but the packet only provides similarity metrics and titles, not the body text needed for a confident merge." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "The Thai issue is unrelated to the generic translation-template issues." + "reason": "High similarity alone does not establish the same underlying request when the body text is not visible." }, { "left": "issue:46222", "right": "issue:46226", - "accept": false, - "reason": "These are not the same request; one is a placeholder translation template and the other is a distinct Thai-language issue." + "accept": true, + "reason": "Highly likely duplicate pair: same visible Thai title, identical body length, and jaccard 1.0. The body text is not shown, so this is best treated as a very strong duplicate rather than an exact one." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The conclusion is a bit too strong for the packet. The Thai pair 46222/46226 looks very likely duplicate from identical titles, identical body length, and jaccard 1.0, but the packet does not expose the body text, so calling it an \"exact duplicate\" / \"same Thai title/body\" is not fully grounded. Also several soft-edge rejections rely on language-specific distinctions that are not directly visible in the packet for the placeholder-title issues. Please soften the canonical verdict to match the available evidence." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -16825,12 +16927,12 @@ "issue:43837|issue:45177", "issue:45322|issue:46133" ], - "prepared_review_unit_hash": "6ce165aecd499449833d045f7950298564362bbdefd3c96e93c196af68b66095", + "prepared_review_unit_hash": "5f2cb292ac0290508727adf178fe662622e495b7c2786b2dcc97cb7efe0d6096", "component_budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21657, + "serialized_chars": 21658, "estimated_input_tokens": 5415, "estimated_eval_tokens": 11086 }, @@ -16838,7 +16940,7 @@ "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21657, + "serialized_chars": 21658, "estimated_input_tokens": 5415, "estimated_eval_tokens": 11086 }, @@ -16847,216 +16949,228 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are mostly unrelated model-request issues that only share a generic \"add support/model\" pattern. I would not collapse them as duplicates; they cover distinct architectures and modalities (vision, audio, OCR, TTS/ASR, text).", - "confidence": 0.34, - "canonical_issue_reason": "If one issue must stand in for the cluster, 46133 is the clearest representative because it is a concrete, well-scoped model request with a specific model name and descriptive title. That said, the set is not a true duplicate cluster.", + "summary": "These are broad model-request issues with only template-level similarity. They cover different model families and modalities, so they should not be collapsed as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue fits: the set mixes unrelated requests (text, vision-language, audio, OCR, detection, configs) rather than the same underlying bug or feature.", "canonical_pr_reason": null, - "best_issue_reason": "46133 is the strongest single representative among these because it is specific, recent, and clearly framed as a model-support request. The rest are distinct requests rather than the same underlying bug/change.", + "best_issue_reason": "No issue cleanly subsumes the others; keeping any one as canonical would be misleading because the requested changes are distinct.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "Different requested models (EuroBert config vs DEIMv2); only the generic \"add model\" framing overlaps." + "reason": "EuroBert config request vs DEIMv2 model request; different models and code paths." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert config and Qwen3-TTS are different model families and modalities." + "reason": "Model config addition vs Qwen3-TTS support; different modality and implementation scope." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert and RF-DETR are unrelated model requests." + "reason": "EuroBert config vs RF-DETR model; unrelated architectures." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are separate architectures with no shared concrete code path." + "reason": "RF-DETR and DEIMv2 are different model families with different integration work." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "RF-DETR is a detection model; LFM2.5 Audio 1.5B is an audio model." + "reason": "RF-DETR object detection request vs LFM2.5 Audio model; different underlying feature." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR and the weight-sparse transformer request concern different model changes." + "reason": "RF-DETR model request vs weight-sparse transformer request; not the same change." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR and Qwen3-ASR are different model-support requests." + "reason": "RF-DETR detection model vs Qwen3-ASR support; different modalities." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR and Qwen3-TTS are unrelated models and code paths." + "reason": "RF-DETR model request vs Qwen3-TTS support; unrelated model additions." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are different model requests." + "reason": "NeoBERT addition vs Qwen3-ASR support; different models and functionality." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and the weight-sparse transformer request are not the same change." + "reason": "NeoBERT model request vs sparse transformer request; different code changes." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT and LFM2.5 Audio 1.5B target different model families." + "reason": "NeoBERT vs audio model support; no shared underlying bug/change." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "RF-DETR and NeoBERT are distinct model additions." + "reason": "RF-DETR and NeoBERT are distinct model integrations." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are different BERT-like models, not one underlying fix." + "reason": "NeoBERT request vs ModernVBERT models; different model families." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture support and Qwen3-ASR are unrelated model requests." + "reason": "MiniCPM3 architecture request vs Qwen3-ASR support; different modalities." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 and the weight-sparse transformer request do not describe the same change." + "reason": "MiniCPM3 architecture addition vs sparse transformer proposal; unrelated work." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are separate model-support tasks." + "reason": "MiniCPM3 model architecture vs ModernVBERT models; not the same change." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 and Voxtral-4B-TTS-2603 are different model families and modalities." + "reason": "MiniCPM3 architecture request vs Voxtral TTS support; different model paths." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." + "reason": "DEIMv2 and Qwen3-TTS are different model requests with no shared concrete fix." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT models and Qwen3-ASR are different concrete additions." + "reason": "ModernVBERT models vs Qwen3-ASR support; separate model integrations." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are separate model requests." + "reason": "ModernVBERT vs Qwen3-TTS; different model families and modalities." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models and BERT-like models with RoPE are related only at a very high level, not the same change." + "reason": "ModernVBERT models vs generic BERT-like RoPE request; only broad semantic overlap." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like models with RoPE and Voxtral TTS are unrelated model additions." + "reason": "BERT-like RoPE request vs Voxtral TTS support; unrelated feature additions." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are different tasks (speech synthesis vs speech recognition)." + "reason": "Qwen3-TTS support vs Qwen3-ASR support; related branding but distinct concrete capabilities." }, { "left": "issue:43671", - "right": "issue:43909", + "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and LFM2.5 Audio 1.5B are distinct audio-model requests." + "reason": "Qwen3-TTS support vs Voxtral TTS support; different models and implementations." }, { "left": "issue:43671", - "right": "issue:44121", + "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and the weight-sparse transformer request are not the same underlying bug/change." + "reason": "Qwen3-TTS support vs DeepSeek-OCR2; different modality and code path." }, { "left": "issue:43671", - "right": "issue:45027", + "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS and Voxtral TTS are different TTS model requests, not a single duplicate." + "reason": "Qwen3-TTS support vs sparse transformer request; not the same underlying change." }, { - "left": "issue:43671", - "right": "issue:45177", + "left": "issue:43837", + "right": "issue:43909", "accept": false, - "reason": "Qwen3-TTS and DeepSeek-OCR2 target different modalities and code paths." + "reason": "Qwen3-ASR support vs LFM2.5 Audio; both audio-related but different model requests." }, { "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR support vs EUPE encoder; different model families and tasks." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio vs SAM3-LiteText; different model domains." + }, + { + "left": "issue:43671", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio 1.5B are different speech/audio model requests." + "reason": "Qwen3-TTS support vs LFM2.5 Audio; both audio-themed but not the same model or change." }, { - "left": "issue:43837", - "right": "issue:44121", + "left": "issue:44121", + "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR and the weight-sparse transformer request are unrelated." + "reason": "Sparse transformer request vs EUPE encoder; distinct architectural changes." }, { "left": "issue:43837", - "right": "issue:45027", + "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR and Voxtral TTS are different speech-model changes." + "reason": "Qwen3-ASR support vs sparse transformer proposal; different feature areas." }, { "left": "issue:43837", - "right": "issue:45322", + "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and EUPE are unrelated model additions." + "reason": "Qwen3-ASR support vs Voxtral TTS support; ASR and TTS are separate additions." }, { - "left": "issue:43909", - "right": "issue:44205", + "left": "issue:37015", + "right": "issue:45027", "accept": false, - "reason": "LFM2.5 Audio 1.5B and SAM3-LiteText are different modalities and model families." + "reason": "NeoBERT model request vs Voxtral TTS support; unrelated model work." }, { - "left": "issue:44121", - "right": "issue:45322", + "left": "issue:43837", + "right": "issue:45177", "accept": false, - "reason": "The weight-sparse transformer and EUPE requests are not the same concrete model change." + "reason": "Qwen3-ASR support vs DeepSeek-OCR2; different modality and implementation." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are different model requests, even though both are vision-language related." + "reason": "EUPE encoder request vs TIPSv2 vision-language pretraining; different model proposals." } ] }, @@ -17086,12 +17200,12 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "515a01f620689f87cfc235a95f4924ef43b8f63b8ff5b1e8a7008b22c490b935", + "prepared_review_unit_hash": "06ab3184c07df60a20cebabd2fa10ca4cb48c3dfa69200c2d2426d734e32e7d9", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -17099,7 +17213,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -17108,18 +17222,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one is a block decoding request, the other is a model-specific support request for DeepSeek v3.2.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "The two issues are unrelated: one is about block decoding, the other is a DeepSeek v3.2 support request. No duplicate/canonical relationship is evident.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the titles and likely problem domains do not match.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Neither issue is a true duplicate, but 43037 is the more specific user-facing feature request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different underlying requests with no shared target or code path: block decoding vs DeepSeek v3.2 support." + "reason": "Different topics and no shared target or code-path evidence; this does not look like the same underlying bug or change." } ] }, @@ -17149,12 +17263,12 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "9e66cc55f179a047b406972de5e05af8df22df04dbdfef6a18753066456799e4", + "prepared_review_unit_hash": "7f149f6076aacee460fc076144ce5b59c3eae407aee954396870ea2ac2a99267", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1584, + "serialized_chars": 1583, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -17162,7 +17276,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1584, + "serialized_chars": 1583, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -17171,18 +17285,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues both mention custom trainers, but they describe different problems: one is about `remove_unused_column` behavior in `Trainer`, the other about FLOPS estimation for custom trainers. They are related at a high level but do not look like the same underlying bug.", - "confidence": 0.91, - "canonical_issue_reason": "Issue 43598 is the more concrete and actionable report, centered on `remove_unused_column` and Trainer customizability; however, there is no strong evidence it is a duplicate of the FLOPS-estimation issue.", + "summary": "These are related to Trainer customizability, but they describe different problems: one is about `remove_unused_column`, the other about FLOPS estimation for custom trainers. They should not be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #43598 is the better canonical issue only if the cluster is centered on Trainer customizability, because it is a concrete request to revisit a specific Trainer behavior (`remove_unused_column`).", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43598 is the better canonical representative for this mini-cluster because it names a specific Trainer behavior and likely a narrower code-path, while 43600 is about a separate estimation concern.", + "best_issue_reason": "Neither issue is a strong global canonical for the pair; if forced to pick one, #43598 is slightly more concrete and narrowly scoped than the FLOPS-estimation request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "They overlap only on the broad theme of custom trainers. One concerns `remove_unused_column` customizability; the other concerns FLOPS estimation. Different symptoms and likely different fixes." + "reason": "They are adjacent Trainer customization topics, but not the same underlying bug or change. One concerns column removal behavior, the other concerns FLOPS counting for custom trainers." } ] }, @@ -17212,7 +17326,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "28eefe11b0c809d00fae4f9fdfc0de7920cf946b40c65343b760c027197152ec", + "prepared_review_unit_hash": "a5037a0e4d9d2af8461aa0dfb211a502ae01d0fc29824daccf816c7691b989b3", "component_budget": { "node_count": 2, "item_count": 2, @@ -17234,18 +17348,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one requests Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration, while the other requests AutoModelForSequenceClassification support for Qwen3.5. They target different model families and different heads/APIs.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the two issues are about different models and different model-head support, so they do not describe the same underlying request.", + "summary": "These are not duplicates: one requests general AutoModel/AutoModelForConditionalGeneration registration for Qwen3-Omni, while the other requests AutoModelForSequenceClassification support for Qwen3.5. They involve different model families and different auto classes/capabilities.", + "confidence": 0.94, + "canonical_issue_reason": "No true canonical duplicate here; if one must be chosen as the representative issue, #43684 is broader and more central because it concerns model registration in core auto classes.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the broader registration request, but it should not be treated as a stand-in for 44405 because the latter is a separate Qwen3.5 sequence-classification feature request.", + "best_issue_reason": "#43684 is the better representative issue because it describes the broader auto-model registration gap, whereas #44405 is a separate, more specific sequence-classification support request for a different config/model line.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying feature requests: Qwen3-Omni AutoModel/AutoModelForConditionalGeneration registration vs Qwen3.5 AutoModelForSequenceClassification support. They are related only at a broad taxonomy level, not the same change." + "reason": "Different underlying requests: Qwen3-Omni auto-registration vs Qwen3.5 sequence-classification support. Different model/config targets and different auto classes, so they should not be merged as duplicates." } ] }, @@ -17282,7 +17396,7 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "6e9d6bf26bc2ea46e31e6bfb3e89fb7320a17b40a841ae417b189a85b7eb3aae", + "prepared_review_unit_hash": "549bef0f323a46daa752b8ecd0190a9ad53433bbb35a421f48d9f50bcdafcbc5", "component_budget": { "node_count": 4, "item_count": 4, @@ -17304,58 +17418,58 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Issue 43835 looks unrelated. The three \"huggingface \" issues are title-similar, but the packet provides no body, link, target, or diff evidence, so the pairwise duplicate calls should stay conservative and rejected. If a representative is needed, 43845 is the earliest of the similar titles, but not proven duplicate.", - "confidence": 0.63, - "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface \" titles and is the best provisional anchor for the cluster, but the evidence is too thin to establish true duplication.", + "summary": "Three issues share the same title/body pattern (\u201chuggingface \u201d) and look like duplicate placeholders; the unrelated issue \u201cFett\u201d does not belong in that cluster.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43845 is the earliest of the three near-identical \u201chuggingface \u201d issues and is the best representative of that duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43845 is the most suitable cluster representative by chronology and title pattern, though the available metadata is insufficient to confirm it as the same underlying report as 43846 or 43859.", + "best_issue_reason": "issue:43845 best fits the cluster because it matches the repeated title and is the earliest of the matching issues; issue:43835 appears unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles and no concrete evidence that they describe the same bug or request." + "reason": "Different titles with no shared content beyond generic similarity; no evidence they describe the same issue." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no concrete evidence that they describe the same bug or request." + "reason": "Different titles with no concrete overlap in description or targets." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles and no concrete evidence that they describe the same bug or request." + "reason": "Different titles and no shared issue details; not the same underlying bug." }, { "left": "issue:43845", "right": "issue:43846", - "accept": false, - "reason": "Only weak title-level similarity is available; without bodies, links, or shared targets, this is not enough to treat them as duplicates." + "accept": true, + "reason": "Same title (\u201chuggingface \u201d) and essentially identical sparse content, so they look like duplicates." }, { "left": "issue:43845", "right": "issue:43859", - "accept": false, - "reason": "Only weak title-level similarity is available; without bodies, links, or shared targets, this is not enough to treat them as duplicates." + "accept": true, + "reason": "Same title and no differentiating details; likely the same duplicate issue." }, { "left": "issue:43846", "right": "issue:43859", - "accept": false, - "reason": "Only weak title-level similarity is available; without bodies, links, or shared targets, this is not enough to treat them as duplicates." + "accept": true, + "reason": "Same title and matching minimal content, indicating the same underlying issue." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The canonical selection is grounded, but the pairwise duplicate verdicts are too strong for the packet. All three \"huggingface \" items only share title-level soft_similarity, and the packet explicitly marks deterministic_accept as false for every pair. Without body content or explicit links, calling them duplicates or 'strongly suggesting the same underlying report' overstates the evidence. Keep the 43835 pairs rejected, but the 43845/43846/43859 pairs should be more conservative." + "accept": true, + "feedback": "The clustering is grounded in the packet: issues 43845/43846/43859 share the same title and have strong soft similarity, while 43835 is distinct and only weakly similar. The chosen canonical issue is the earliest among the matching set, and the soft-edge verdicts are conservative." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -17389,7 +17503,7 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "767c9d7ced90a8f05943d0f2882eced33eed3cbb69c23f79a919a73955ee7faa", + "prepared_review_unit_hash": "d6d401e5dfe42ffe0846e5094c05d1bf062e3e8e89c2b9fbd597572392c2a7f9", "component_budget": { "node_count": 5, "item_count": 5, @@ -17411,66 +17525,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly low-signal spam placeholders, plus one documentation-suggestion issue and one unrelated named issue. There is not enough evidence to merge any pair as true duplicates.", - "confidence": 0.89, - "canonical_issue_reason": "issue:44371 is the earliest generic spam placeholder and is as representative as the other spam-only issues.", + "summary": "This cluster looks like unrelated low-signal/spam issues rather than true duplicates. None of the soft pairs should be merged.", + "confidence": 0.93, + "canonical_issue_reason": "issue:44418 is the only item with a non-spam, vaguely substantive title, so it is the best representative if one must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44371 is the best issue representative for the spam subgroup because it is earliest and has the same minimal signal as the other spam entries.", + "best_issue_reason": "issue:44418 is the most suitable issue to anchor the cluster because it is at least a recognizable documentation request, unlike the placeholder spam titles.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are titled '', but there is no substantive body or target evidence to prove the same underlying issue." + "reason": "Both are titled '', but that is not enough to show the same underlying issue; they read as separate spam submissions with no concrete shared bug or change." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Identical spam titles are too weak to confirm a true duplicate without concrete bug/change details." + "reason": "Same placeholder spam title, but no evidence of one underlying report; treat as distinct spam issues." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "Spam placeholder and documentation enhancement are clearly different issue types and topics." + "reason": "'' vs a documentation suggestion are clearly different topics." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder does not match the unrelated named issue 'Check out \"Google LLC\"'." + "reason": "'' and 'Check out \"Google LLC\"' are unrelated and look like separate spam-like submissions." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Documentation suggestion is unrelated to a spam placeholder issue." + "reason": "A documentation enhancement request is not the same underlying issue as a spam placeholder." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "Documentation suggestion is unrelated to a spam placeholder issue." + "reason": "These titles point to different kinds of reports; no duplicate evidence beyond generic similarity." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are spam placeholders, but the record is too thin to confirm they are the same underlying report rather than separate spam submissions." + "reason": "Identical spam titles alone do not establish the same bug or change; they could be independent junk submissions." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder and named issue are not the same underlying report." + "reason": "Placeholder spam and a branded promotional title are not the same underlying issue." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Spam placeholder and named issue are not the same underlying report." + "reason": "These look like different spam artifacts, not one fixable problem." } ] }, @@ -17500,12 +17614,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "79182dcb309c28dcf80f685d11a74c23d5b15ecdec59161f7398916c89bab94a", + "prepared_review_unit_hash": "1e5819a01de7f68854674bdae2d138309ce027dc7110d6bd478612817547dc0a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1646, + "serialized_chars": 1645, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -17513,7 +17627,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1646, + "serialized_chars": 1645, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -17522,28 +17636,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues look closely related based on identical titles and soft text similarity, but the packet does not firmly establish that they are the same report. #45145 is the better representative of the pair because it was created first and has the only inbound reference.", - "confidence": 0.73, - "canonical_issue_reason": "#45145 is the earlier issue and the only one with inbound reference activity, so it is the better canonical representative if this cluster is treated as one topic.", + "summary": "The two closed issues share an identical title and soft similarity signals, so they likely describe the same report; however, the available packet does not expose the issue bodies, so the duplicate conclusion should be kept conservative.", + "confidence": 0.86, + "canonical_issue_reason": "#45145 is the better canonical issue because it was created first and already has an inbound reference, which makes it the more likely original thread.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the strongest overall representative because it predates #45151 and has the most external attention in the packet, while #45151 has only limited additional discussion.", + "best_issue_reason": "#45145 is the best issue candidate because it is the earliest of the two and has the only inbound reference, while the evidence available here does not distinguish a different scope.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", - "accept": false, - "reason": "The issues have identical titles and moderate soft similarity, so they are closely related, but the packet does not directly show that their bodies are the same. Given the conservative evidence, I would not confirm them as the same underlying request." + "accept": true, + "reason": "The issues have the same title and a soft similarity signal; that is enough to treat them as likely duplicates, but the packet does not show the bodies, so the match is based on title/similarity rather than any confirmed identical proposal." } ] }, "evaluator_result": { "accept": false, - "feedback": "The canonical choice is mostly grounded (earlier creation time and one inbound reference), but the duplicate/\u201csame underlying request\u201d language is a bit too strong for the evidence shown. The packet only provides a soft-similarity signal (jaccard ~0.63) plus identical titles; it does not directly establish that the bodies are the same, and #45151 also has its own discussion activity. Please phrase the verdict more conservatively and avoid implying exact duplication or stronger engagement differences than the packet supports." + "feedback": "The packet supports that these two closed issues are highly similar and share the same title, but the analyst overstates the evidence in a few places. In particular, the packet does not show the bodies themselves, only similar body lengths, so claims like \u201cnearly identical body size,\u201d \u201csame broad proposal,\u201d and \u201cno evidence of different scope\u201d go beyond what is grounded. The duplicate-issue conclusion may be plausible, but the soft-edge verdict should be stated more conservatively based on title match and soft similarity only." }, "evaluator_used": true, "retried": true, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -17620,12 +17734,12 @@ "soft_pairs": [ "issue:45672|issue:45674" ], - "prepared_review_unit_hash": "e6edc30a09831dbb5fe436cf4de6087f4a0a5c71ac118688aa94d875e630a4e2", + "prepared_review_unit_hash": "842e6ab531aaabd33816066e9a3abce215014ae236d6614d858d3fa166350a5c", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1671, + "serialized_chars": 1670, "estimated_input_tokens": 418, "estimated_eval_tokens": 1092 }, @@ -17633,7 +17747,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1671, + "serialized_chars": 1670, "estimated_input_tokens": 418, "estimated_eval_tokens": 1092 }, @@ -17642,18 +17756,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two separate 4-bit inference bugs are grouped only by broad quantization context: one is a Gemma4 audio-module uint8/torch.finfo dtype failure, the other is a BitsAndBytesConfig lm_head exclusion bug causing an AssertionError. They do not look like the same underlying issue.", - "confidence": 0.96, - "canonical_issue_reason": "Issue #45674 is the better representative because it states a clearer, more general configuration bug with a concrete failure mode and is still open for triage.", + "summary": "These two issues are related only at a high level (4-bit inference in the same stack) but describe different failures and code paths: one is a Gemma4 audio-module `torch.finfo()` TypeError on `uint8` weights, the other is a BitsAndBytesConfig `llm_int8_skip_modules` regression that removes the default `lm_head` exclusion and triggers an AssertionError.", + "confidence": 0.95, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue #45674 is the most useful cluster anchor: it is specific enough to identify the bug class, but broader than the Gemma4-only symptom in #45672.", + "best_issue_reason": "Issue 45674 is the more general and actionable report: it names a concrete BitsAndBytesConfig regression, states the observed assertion failure, and is framed in a way that is easier to map to a fix than the narrower Gemma4-specific audio-module TypeError.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Both involve 4-bit inference, but the failure mechanisms differ: Gemma4 audio modules hit a uint8/torch.finfo TypeError, while BitsAndBytesConfig loses the default lm_head exclusion and triggers an AssertionError. These are not the same underlying bug." + "reason": "Different underlying bugs: 45672 is a Gemma4 audio-module `torch.finfo()` TypeError on `uint8` weights, while 45674 is a BitsAndBytesConfig skip-modules regression that clears `lm_head` exclusion and causes an AssertionError. Same general area, not the same fix." } ] }, @@ -17683,7 +17797,7 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "49ab16da32407d942a4842c77854263a5929295010838cc173f273b76fda4777", + "prepared_review_unit_hash": "cde04dc22fae4edbfc9b7902a316cb3c1d7aa61e8411e532d6e578c81d402acf", "component_budget": { "node_count": 2, "item_count": 2, @@ -17705,18 +17819,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are unrelated empty-name/date entries and do not appear to describe the same bug or request. Treat this cluster as non-duplicate noise/spam rather than a real issue family.", + "summary": "The two issues are unrelated placeholder/spam-like reports with different personal names and dates; there is no evidence they describe the same bug or task.", "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Neither issue is a good canonical representative because neither contains a real problem description or actionable context.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Different personal names and dates with no body, no shared technical context, and no indication of the same underlying issue." + "reason": "Different named entities and dates with no shared technical context, code path, or problem statement; they do not appear to be the same underlying issue." } ] }, @@ -17746,7 +17860,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "f8529b4803b423bc9052328c8e8bad0195e78e5663407f9621b8d72a9bdd9b3f", + "prepared_review_unit_hash": "c396abe2d662c192ee161454d9b074c3fd11fe3167013d7c72a97b3ce7ecb29e", "component_budget": { "node_count": 2, "item_count": 2, @@ -17768,18 +17882,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues appear unrelated; both are just personal-name/date titles with no body or technical context, so the soft similarity is not enough to treat them as duplicates.", + "summary": "The two issues are just title-only personal-name/date entries and do not describe the same underlying problem; the similarity is superficial only.", "confidence": 0.98, - "canonical_issue_reason": "Neither issue has any substantive description, links, or code-related context; there is no basis to prefer one as a canonical bug report.", + "canonical_issue_reason": "No clear canonical issue: both are unrelated, low-context entries with no substantive problem description.", "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a strong canonical candidate because both lack actionable content and appear to be non-technical placeholder-style entries.", + "best_issue_reason": "45764 is only marginally preferable as the earliest created issue, but it is not a meaningful representative of the pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "They do not describe the same underlying bug or change; the titles are different personal names and dates with no shared issue context." + "reason": "The titles share only a name/date pattern; there is no evidence they refer to the same bug, request, or code path." } ] }, @@ -17809,7 +17923,7 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "553fe329449f104e45a089d8d88dfe11b05e7d27317dabe04e9aa13e50def247", + "prepared_review_unit_hash": "67356b72308b09767f276b3499a0f60c2a086f4d8ec0be629a5ac0e0cf0a579e", "component_budget": { "node_count": 2, "item_count": 2, @@ -17831,18 +17945,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are unrelated: one requests a new Qwen3_5 token-classification model class, while the other asks for FP8 training support. They do not describe the same bug or feature and should not be clustered as duplicates.", - "confidence": 0.98, + "summary": "The two issues are not duplicates: one requests a new token-classification head for Qwen3_5, while the other asks for FP8 training support. They touch different features and code paths.", + "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 45810 is the more concrete and narrowly scoped request, so it is the better cluster representative, but it is not the same underlying problem as issue 46295.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different requests affecting different parts of the library: adding a model class for token classification versus enabling FP8 training support. No shared code path, target, or underlying issue." + "reason": "Different feature requests with no shared concrete bug or implementation target: model head addition for token classification versus FP8 training support." } ] }, @@ -17882,12 +17996,12 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "04be3d7142d777c0c3b82692c9d5d4027130793db9f8427961c0c22c7c32ff88", + "prepared_review_unit_hash": "ca896772fffd2448fe667557ae093990a4a0554627487050933b3e2b4d627449", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, + "serialized_chars": 5725, "estimated_input_tokens": 1432, "estimated_eval_tokens": 3120 }, @@ -17895,7 +18009,7 @@ "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, + "serialized_chars": 5725, "estimated_input_tokens": 1432, "estimated_eval_tokens": 3120 }, @@ -17904,66 +18018,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is heterogeneous: it contains two exact-duplicate issue pairs (45870/45871 and 45872/45873) plus one unrelated dataset post (45869). There is no single underlying bug/change across all five items.", - "confidence": 0.97, - "canonical_issue_reason": "45872 is the best canonical representative among the substantive DataForge issues because it is the clearest standalone topic and has an exact duplicate in 45873; 45870/45871 form a separate duplicate pair, and 45869 is unrelated.", + "summary": "The set splits into two exact-duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is related only at a broad financial-data theme level and should not be merged with the others.", + "confidence": 0.95, + "canonical_issue_reason": "45870 is the best canonical representative to keep for the first duplicate pair because 45870 and 45871 are identical, and 45870 is the earlier of the two.", "canonical_pr_reason": null, - "best_issue_reason": "45872 is the strongest issue to keep as the representative record: it is more specific than the repeated example-title issues, has a clear duplicate partner, and is not just a dataset promo post.", + "best_issue_reason": "45870 is the strongest single issue to retain overall: it is an exact duplicate pair member, clearly representative of one concrete topic, and there is no PR evidence in the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Different issue types: a dataset promotion versus an example about time series prediction. Related finance wording, but not the same underlying request." + "reason": "Related financial-domain wording, but one is a dataset bundle listing and the other is an example request; not the same underlying issue." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Different issue types: a dataset promotion versus an example about time series prediction. Related finance wording, but not the same underlying request." + "reason": "Same as 45869/45870: broad theme overlap only, with different concrete topic and no evidence of an exact duplicate." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Same title and matching content footprint; this looks like a direct duplicate issue." + "reason": "Exact title/body duplicate pair for the same example request." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Both mention finance/DataForge, but one is an example and the other is a service/scraping request. Same broad theme, not the same concrete issue." + "reason": "Both mention financial time series, but 45872 adds a scraping service component and is a different concrete request." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Different concrete topic: example notebook/content versus a DataForge service/scraping issue." + "reason": "Same as 45870/45872: overlapping domain, but different underlying change/request." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Same broad theme, but different underlying ask; this is an example issue while the other is about a DataForge service/scraping workflow." + "reason": "Broad thematic similarity only; not the same issue or change." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Same broad theme, but different underlying ask; this is an example issue while the other is about a DataForge service/scraping workflow." + "reason": "Broad thematic similarity only; not the same issue or change." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Same title and matching content footprint; this looks like a direct duplicate issue." + "reason": "Exact duplicate pair with matching title and body length, describing the same request." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 have matching titles and identical body lengths, so treating them as duplicate pairs is reasonable, and 45869 is clearly a different dataset-promo item by title. The soft-edge rejections are conservative, and there is no evidence of a single unified change across all five items." + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 are the only clear duplicate pairs, while 45869 is only soft-related by theme. The canonical choice of the earlier duplicate is conservative, and the soft-edge rejects do not overstate the evidence." }, "evaluator_used": true, "retried": false, @@ -18022,160 +18136,160 @@ "pull_request:46048|pull_request:46268", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "df10614854ec97c5080aaf9fa92bd01f08973f3771e221efc605a7c087767e62", + "prepared_review_unit_hash": "55973c6b1cd8e83e75b75c5a002e5d702118e53064d71ec73f13f3d07878cfa4", "component_budget": { "node_count": 15, "item_count": 15, "soft_pair_count": 20, - "serialized_chars": 40312, - "estimated_input_tokens": 10078, - "estimated_eval_tokens": 20412 + "serialized_chars": 40306, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 }, "budget": { "node_count": 15, "item_count": 15, "soft_pair_count": 20, - "serialized_chars": 40312, - "estimated_input_tokens": 10078, - "estimated_eval_tokens": 20412 + "serialized_chars": 40306, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly a set of unrelated PRs that only share broad keyword similarity. The only true duplicate-like pair is the two fixed_cross_entropy patches targeting the same issue; the rest should remain separate.", + "summary": "This cluster is mostly unrelated PRs. The only genuine duplicate-looking pair is the two `fixed_cross_entropy` patches (43251 and 43254); the rest target different models, docs, or kernel-routing changes and should stay separate.", "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is the clearest canonical representative for the shared cross_entropy kwargs fix: it directly targets issue 43240, touches the exact implementation site, and is the more specific patch of the near-duplicate pair.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 43251 is the best overall PR candidate in the cluster because it has an explicit issue target and the most direct, concrete implementation of the shared loss fix.", + "best_pr_reason": "PR 43251 is the best representative of the only true duplicate subgroup: it has the explicit issue link, touches the same `fixed_cross_entropy` code path as 43254, and is the more complete version of that fix.", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "Both are large model/infra PRs, but they fix unrelated things: VideoPrism model integration vs executorch export dynamic-shape testing." + "reason": "Different features: VideoPrism model addition vs an ExecuTorch dynamic-shape export test; no shared code-path bug." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Tokenizer helper methods in MistralCommon are unrelated to executorch dynamic-shape export tests." + "reason": "Unrelated work: Mistral tokenizer methods vs ExecuTorch export testing." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Missing tokenizer methods and adding MobileLLM are different model/code-path changes with no shared bug." + "reason": "Different subsystems: tokenizer API additions vs a new MobileLLM model implementation." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness and adding MobileLLM are separate changes in different subsystems." + "reason": "Processor auto-loading robustness and MobileLLM model scaffolding are distinct changes." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss-function kwargs fix is unrelated to adding a new MobileLLM model." + "reason": "`fixed_cross_entropy` kwargs fix is unrelated to adding MobileLLM." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Cross-entropy kwargs plumbing and Mistral4 test fixes are different code paths and bugs." + "reason": "Cross-entropy loss kwargs fix and Mistral4 test fixes are different code paths." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Executorch export dynamic-shape tests do not describe the same underlying change as Mistral4 test fixes." + "reason": "ExecuTorch dynamic-shape export tests are unrelated to Mistral4 test/model changes." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Exportability tests and MobileLLM model addition are unrelated." + "reason": "Different areas: export testing vs new model implementation." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "Adding MobileLLM is unrelated to fixing config resolution order in modeling_utils." + "reason": "Adding MobileLLM and fixing `cls.config` resolution are unrelated changes." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "New model addition and Falcon output-interface refactoring are different changes." + "reason": "MobileLLM model addition and Falcon output-interface refactor do not fix the same bug." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon model refactoring is unrelated to the model-creation skill PR." + "reason": "Falcon refactor and the model-creation skill are unrelated." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon standardized-output refactor and Mistral4 test fixes are not the same bug/change." + "reason": "Falcon standardized-output refactor is unrelated to Mistral4 test fixes." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test fixes are unrelated to the layrnorm-to-layernorm renaming pass." + "reason": "Mistral4 tests/model work and `layrnorm` renaming are separate issues." }, { "left": "pull_request:44827", "right": "pull_request:46267", "accept": false, - "reason": "Mistral4 test fixes and chat parsing updates do not fix the same concrete problem." + "reason": "Mistral4 tests and chat-response parsing updates are different features." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 test fixes and TDT loss kernel integration are different areas and problems." + "reason": "Mistral4 test fixes and TDT loss-kernel integration are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test fixes are unrelated to the config-resolution order change." + "reason": "Mistral4 test/model work is unrelated to `cls.config` resolution order." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "A model test-fix PR is not the same underlying change as a skill/documentation PR." + "reason": "No shared bug: Mistral4 fixes vs model-creation skill content." }, { "left": "pull_request:43823", "right": "pull_request:45213", "accept": false, - "reason": "Adding MobileLLM and adding a model-creation skill are unrelated." + "reason": "A new MobileLLM model and a model-creation skill are not the same underlying change." }, { "left": "pull_request:46048", "right": "pull_request:46268", "accept": false, - "reason": "Both touch hub kernel routing, but they add different kernels for different purposes and could not plausibly be merged as one fix." + "reason": "Both mention hub kernels, but they add different kernel entries for different purposes and are not the same concrete fix." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "These are two near-duplicate patches for the same fixed_cross_entropy kwargs support in the same file, both addressing issue 43240 and fixing the same code path." + "reason": "Both patches fix the same `fixed_cross_entropy` implementation in `loss_utils.py`, add the same missing kwargs support, and target the same issue (`43240`); they could be merged into one PR." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The only clearly duplicate-like pair is the two `fixed_cross_entropy` PRs sharing the same issue target and file, and the other soft-edge rejections are reasonably supported by the diffs." + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted pair, 43251/43254, is well supported by the shared file, shared issue target, and near-identical `fixed_cross_entropy` kwargs fix. The other soft-edge rejections appropriately treat the PRs as distinct models, docs, tests, or routing changes without overstating duplicate evidence." }, "evaluator_used": true, "retried": false, @@ -18202,7 +18316,7 @@ "soft_pairs": [ "pull_request:40546|pull_request:42781" ], - "prepared_review_unit_hash": "0e4033e0ba1425b4660cdbf2fee3bbc2a5a2149506f8907ae13917598306d78b", + "prepared_review_unit_hash": "a711540e673acf560b1e24eb3d5bb78aa73b75933ecede5013439458715c3964", "component_budget": { "node_count": 2, "item_count": 2, @@ -18224,18 +18338,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about VibeVoice, but they target different scope: one implements the base VibeVoice model, the other adds a realtime variant and extra acoustic-tokenizer/generation plumbing. They are related, but not the same underlying change and not good merge candidates into one PR.", - "confidence": 0.87, + "summary": "These are related but distinct VibeVoice pull requests: one adds the base VibeVoice implementation, while the other extends VibeVoice with realtime/acoustic-tokenizer work. They overlap in shared model scaffolding, but they do not look like the same concrete fix or mergeable duplicate PR.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #40546 is the broader foundational VibeVoice implementation and predates the realtime follow-up; it is the more canonical representative of this cluster.", + "canonical_pr_reason": "PR #42781 is the broader, more up-to-date VibeVoice change set and includes the realtime-specific pieces plus the acoustic tokenizer/model work, making it the better representative artifact for this cluster.", "best_issue_reason": null, - "best_pr_reason": "PR #40546 is the best single representative because it covers the core VibeVoice model integration, Auto registrations, docs, and conversion/modeling code, whereas #42781 is a narrower realtime extension.", + "best_pr_reason": "PR #42781 is the best representative because it covers the larger VibeVoice surface area and the later, more specialized integration work; PR #40546 is the earlier base implementation and not a duplicate of it.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Related by model family, but the scope differs: #40546 adds the base VibeVoice integration, while #42781 adds VibeVoice Realtime plus additional acoustic tokenizer and generation changes. They are not the same concrete code-path fix/change and would not naturally be merged as one PR." + "reason": "Shared VibeVoice scaffolding alone is not enough: #40546 is a base implementation, while #42781 adds realtime and acoustic-tokenizer functionality plus extra generation/feature-extraction changes. They are related but not the same concrete code-path fix." } ] }, @@ -18265,12 +18379,12 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "2b5b54900cf10a49f7351da74b3b79e60ca7037367f29227026298f6927746ff", + "prepared_review_unit_hash": "603981ac9dcb8fe3da87008f5af56435759eef98c7219fd98d29727e1db8b453", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4674, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -18278,7 +18392,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4674, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -18287,18 +18401,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are unrelated feature additions in different model families: TimesFM covariate forecasting vs DINOv3 image-classification support. They should not be clustered as duplicates.", + "summary": "These pull requests are on unrelated features: one adds covariate-based forecasting for TimesFM, the other adds DINOv3 image-classification support. They do not appear to fix the same bug or implement the same change, so the soft duplicate edge should be rejected.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No clear canonical PR for a duplicate cluster because these changes target different code paths and different models; if forced to pick one representative, PR #41224 is slightly more mature and review-heavy, but it is not a duplicate of #40755.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #41224 is the better representative of the cluster only by activity/review maturity, not by semantic similarity. PR #40755 is a separate TimesFM feature and should stand alone.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different subsystems and different user-facing capabilities: TimesFM covariate forecasting vs DINOv3 image classification. No shared bug or mergeable fix path." + "reason": "Different models, different code paths, and different user-facing features. The shared 'model support' theme is too broad to indicate the same underlying change, so they should not be merged or treated as duplicates." } ] }, @@ -18393,12 +18507,12 @@ "pull_request:41797|pull_request:45186", "pull_request:45186|pull_request:45630" ], - "prepared_review_unit_hash": "a12f1c0c2950bac8f1b29dd56fd7dc61a82e2bb233da35e41ad8c2ac0d356af3", + "prepared_review_unit_hash": "d598fb53bff9bbea76fbec3957355952937483d29af5906aac7a618744047fed", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10435, + "serialized_chars": 10434, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -18406,7 +18520,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10435, + "serialized_chars": 10434, "estimated_input_tokens": 2609, "estimated_eval_tokens": 5474 }, @@ -18415,24 +18529,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three unrelated model-addition pull requests (DeepSeek OCR, Isaac, Kimi2-6). They overlap on Transformers integration scaffolding, but they do not fix the same underlying code path, so the soft duplicate links should be rejected.", + "summary": "All three items are separate new-model integration PRs for different models (DeepSeek OCR, Isaac, Kimi2-6). They share boilerplate files and repo-wide mapping edits, but not the same underlying change.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the cluster contains three distinct model-onboarding PRs for different models, not duplicates of one change.", + "canonical_pr_reason": "No canonical PR: each pull request adds support for a different model and different codepaths.", "best_issue_reason": null, - "best_pr_reason": "None of these PRs are duplicate candidates; if a representative is needed, #45630 is the broadest and most mature model-addition PR, but it is still unrelated to the others.", + "best_pr_reason": "PR 45630 is the most substantial and review-active of the set, but it is still a distinct model addition rather than a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "Different model integrations (DeepSeek OCR vs Isaac). Shared auto-registry and docs wiring are generic onboarding boilerplate, not the same bug or change." + "reason": "Different model integrations (DeepSeek OCR vs Isaac); overlap is only in shared infrastructure and documentation/mapping scaffolding." }, { "left": "pull_request:45186", "right": "pull_request:45630", "accept": false, - "reason": "Different model additions (Isaac vs Kimi2-6) with different files and code paths. Similar Transformers plumbing is too broad to treat as one mergeable PR." + "reason": "Different model integrations (Isaac vs Kimi2-6); similar auto-registration changes do not make them the same bug/change." } ] }, @@ -18462,7 +18576,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "72113ba88d837a217c6755e52a03ac5050d00020c7e6d506bbca75970307725a", + "prepared_review_unit_hash": "5d69c09c47b8dab151e9100501ea1d944e6e39a446c9ac7f416e5c1fac4144af", "component_budget": { "node_count": 2, "item_count": 2, @@ -18484,18 +18598,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both GitHub Actions maintenance PRs, but they target different action families and different motivations: PR 42975 is a broad Node 24 compatibility upgrade for core workflow actions, while PR 42976 is a separate action-version refresh centered on Docker build workflows. They should not be merged as duplicates.", - "confidence": 0.94, + "summary": "These PRs overlap in GitHub Actions workflow maintenance, but they are not the same change set: one is an Actions runtime compatibility sweep, the other is a broader dependency/version bump with concrete docker build action upgrades. Do not dedupe them as duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the better anchor because it has the clearer, broader platform-maintenance scope (core workflow actions updated for Node 24 compatibility), while PR 42976 is a separate Docker/build-action upgrade.", + "canonical_pr_reason": "No single canonical PR: 42975 is specifically about GitHub Actions Node 24 compatibility across many workflows, while 42976 is a separate upgrade pass that prominently changes docker/build-push-action and related workflow versions.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the strongest representative of the cluster\u2019s workflow-maintenance theme; PR 42976 is narrower and focused on different actions and workflows.", + "best_pr_reason": "PR 42975 is the better representative of the Node 24 compatibility effort because its title and diff align with a focused workflow-runtime upgrade across many files.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Do not treat as duplicates: 42975 upgrades checkout/cache/upload-artifact for Node 24 compatibility across many workflows, whereas 42976 updates docker/build-push-action and related workflow actions. They touch overlapping YAML files but not the same underlying change or code path." + "reason": "They share workflow files and both upgrade actions, but the concrete changes differ: 42975 is a Node 24 compatibility sweep for GitHub Actions, whereas 42976 mainly bumps docker/build-push-action and other workflow versions. They do not look like the same underlying fix and should not be merged into one PR." } ] }, @@ -18532,70 +18646,70 @@ "pull_request:43448|pull_request:45490", "pull_request:43451|pull_request:45490" ], - "prepared_review_unit_hash": "ab9b6807bf6d04df0fb14009849dfe7b5b099ebab0dc7ad6c1ccb6af408e2fe3", + "prepared_review_unit_hash": "d14ba206381b317bef975b23431b6831ff7d5d5dd77d837ad5d5e56fb253b2b1", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 14184, - "estimated_input_tokens": 3546, - "estimated_eval_tokens": 7348 + "serialized_chars": 14179, + "estimated_input_tokens": 3545, + "estimated_eval_tokens": 7346 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 14184, - "estimated_input_tokens": 3546, - "estimated_eval_tokens": 7348 + "serialized_chars": 14179, + "estimated_input_tokens": 3545, + "estimated_eval_tokens": 7346 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are four unrelated model-addition PRs. They share Hugging Face boilerplate updates (docs nav, auto mappings, tests), but each introduces a different model family and code path: ViT NEPA, Molmo, Molmo2, and CTSM.", + "summary": "These PRs are separate model-addition efforts with only superficial overlap in docs and auto-registry files. They do not look like duplicates or mergeable variants of the same bug/fix/change.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR for the cluster: the PRs are distinct feature additions for different models, not duplicates of the same change.", + "canonical_pr_reason": "No issue in the cluster. Among the PRs, #43451 is the strongest representative because it is the most substantial, has the highest review activity, and contains a complete model integration across modeling, processing, conversion, tests, and auto-mappings.", "best_issue_reason": null, - "best_pr_reason": "If one PR must be picked as the strongest representative, 43451 is the most substantial and review-heavy Molmo2 addition, but it still is not a cluster-level duplicate representative.", + "best_pr_reason": "#43451 is the best overall representative of the cluster for triage purposes: it is a full-featured model addition with broad coverage and substantial review, whereas the others are different model additions with less basis for comparison.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model additions: ViT NEPA vs Molmo. Shared auto/docs files are boilerplate, not the same underlying change." + "reason": "ViT NEPA and Molmo are unrelated model additions. Shared docs/auto files reflect framework registration, not the same underlying code-path or fix." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "Different model additions with separate configs, modeling, and tests: ViT NEPA vs Molmo2." + "reason": "ViT NEPA and Molmo2 are different model integrations with different architectures and codepaths; overlap is only in common registry/docs plumbing." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Different model families and implementations: ViT NEPA vs CTSM; only shared registration/docs scaffolding." + "reason": "ViT NEPA and CTSM are unrelated model additions in different domains, so they are not the same change." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo and Molmo2 are related in name but they add different model versions with distinct files and code paths, so not the same change." + "reason": "Molmo and Molmo2 are distinct model families with separate implementations, tests, and processors; same-name similarity is not enough to treat them as duplicates." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo vs CTSM are unrelated model additions; overlap is limited to generic docs/auto-mapping plumbing." + "reason": "Molmo and CTSM address different model families and codepaths; shared registry files are incidental." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM are separate model integrations with different implementations; shared files reflect repository plumbing only." + "reason": "Molmo2 and CTSM are unrelated additions spanning different subsystems, so they should not be merged as duplicates." } ] }, @@ -18628,7 +18742,7 @@ "pull_request:43291|pull_request:44815", "pull_request:43665|pull_request:44815" ], - "prepared_review_unit_hash": "53ce151febbe97ea03bd0e86640eee79e68e62a322b09331ad2228280f1fe64f", + "prepared_review_unit_hash": "cf83c761cd1008c4a478ac9a8fa4b042a2d801a84b965a7638d08c22e51dd648", "component_budget": { "node_count": 3, "item_count": 3, @@ -18650,30 +18764,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three unrelated pull requests: one Whisper test/tokenizer fix, one clearly bogus test-only change, and one substantive FP8/dequant loading fix. None should be merged as duplicates; PR #44815 is the strongest canonical PR for the cluster.", - "confidence": 0.96, + "summary": "All three items are pull requests, but they do not describe the same underlying change. PR 44815 is a real dequant/loading fix; PR 43291 is a Whisper tokenizer/test-related fix; PR 43665 only edits tests and appears unrelated to the others. No soft pair should be merged.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #44815 is the only substantive change set with a concrete implementation focus (FP8/dequant logic across loading, quantization, and Mistral support). The other PRs are either narrow Whisper test/tokenizer work or obviously unrelated test edits.", + "canonical_pr_reason": "PR 44815 is the strongest canonical PR because it contains the clearest substantive code fix across multiple implementation files, with a concrete change in model loading plus FP8/quantization logic. It is more representative of an actual bug fix than the test-only or narrow Whisper PRs.", "best_issue_reason": null, - "best_pr_reason": "PR #44815 has the clearest real bug-fix scope and the broadest concrete code-path coverage. It is the best representative of the cluster, even though the other PRs are not duplicates.", + "best_pr_reason": "PR 44815 is the best overall PR to represent this cluster because it addresses a specific runtime/code-path issue and has broader technical impact than the other two. PR 43291 is narrower and centered on Whisper decode/tests, while PR 43665 appears to be only test modifications and is not a comparable code fix.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Whisper tokenizer/test changes and unrelated CLIP/ViT test edits do not fix the same underlying bug or code path." + "reason": "Both touch tests, but they do not fix the same bug or code path. 43291 is Whisper decode/tokenization related, while 43665 only alters CLIP/ViT tests with no shared implementation change." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper decoding/tokenizer behavior is unrelated to the FP8/dequant loading and quantization fix in the other PR." + "reason": "These PRs target different subsystems and different problems: Whisper decode behavior vs model loading/FP8 dequantization. They are not mergeable as one underlying fix." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "The CLIP/ViT test-only patch is unrelated to the FP8/dequant implementation fix; they do not plausibly merge into one PR." + "reason": "No shared failing code path or concrete change exists between them. 43665 is test-only and unrelated to 44815's loading/quantization fix." } ] }, @@ -18710,64 +18824,64 @@ "pull_request:45932|pull_request:46100", "pull_request:46100|pull_request:46114" ], - "prepared_review_unit_hash": "965d4fd2eb121864f4da181556d950829cf4186ea005742a72dd71e06e9441a0", + "prepared_review_unit_hash": "ad7615b67e2d3aa0063fa67dd44adb7fe8a82e6c968ad8a9f5011da31956f062", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 12260, - "estimated_input_tokens": 3065, - "estimated_eval_tokens": 6386 + "serialized_chars": 12262, + "estimated_input_tokens": 3066, + "estimated_eval_tokens": 6388 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 5, - "serialized_chars": 12260, - "estimated_input_tokens": 3065, - "estimated_eval_tokens": 6386 + "serialized_chars": 12262, + "estimated_input_tokens": 3066, + "estimated_eval_tokens": 6388 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are mostly unrelated dummy/test-infra changes; the shared ViT test file overlap is incidental, and there is no safe duplicate cluster here.", - "confidence": 0.94, + "summary": "These are mostly unrelated synthetic/test PRs with weak filename overlap, mainly around ViT tests and CI scaffolding. The apparent similarities are title/style-driven rather than the same underlying bug or change, so there is no true duplicate PR pair here.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: the items split into several different synthetic changes (workflow/config, model code, and failing-test injections) rather than one shared fix.", + "canonical_pr_reason": "pull_request:45932 is the broadest representative of the synthetic test PRs, spanning multiple model and workflow files rather than a single tiny test edit.", "best_issue_reason": null, - "best_pr_reason": "PR 46114 is the broadest representative of the test-migration / CI-theme changes, but it still is not a reliable duplicate anchor because it does not match the other PRs\u2019 concrete edits.", + "best_pr_reason": "pull_request:45932 is the best representative PR in the cluster because it has the widest change surface and looks like the central 'testing grafana' synthetic change set.", "soft_edge_verdicts": [ { "left": "pull_request:43488", - "right": "pull_request:46114", + "right": "pull_request:43532", "accept": false, - "reason": "Both are synthetic test-related PRs, but 43488 is a setup/version tweak while 46114 is a multi-file test/CI migration; not the same change or bug." + "reason": "Both are dummy '[don't merge]' PRs, but one is a setup/version test and the other is a workflow/model instrumentation change; the shared ViT filename is not enough to make them the same concrete change." }, { "left": "pull_request:43488", - "right": "pull_request:43532", + "right": "pull_request:46114", "accept": false, - "reason": "Shared \"don't merge\" intent is not enough; one changes packaging/versioning, the other rewrites workflow and testing utilities." + "reason": "The overlap is only a shared ViT test file and test-harness context; the actual edits are different synthetic changes with different scopes and goals." }, { "left": "pull_request:43532", "right": "pull_request:46114", "accept": false, - "reason": "They both touch test infrastructure, but the concrete edits differ substantially across workflows, utilities, and unrelated model/test files." + "reason": "They both touch test infrastructure, but the modified files and patches are different enough that they do not look like one mergeable fix." }, { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "45932 adds OTEL/workflow plumbing and model code changes, while 46100 only injects failing assertions into ViT tests; different underlying changes." + "reason": "Both are test-related synthetic PRs, but 45932 is a broad grafana/testing change across workflows and model code while 46100 is just a ViT test migration; not the same underlying change." }, { "left": "pull_request:46100", "right": "pull_request:46114", "accept": false, - "reason": "Both mention test migration, but 46100 is a small ViT test sabotage and 46114 is a broader multi-test/CI change; not mergeable as one PR." + "reason": "The shared ViT test file is incidental; 46100 is a tiny failing-test migration and 46114 is a broader forked PR with CI and multiple test files, so they are not the same code-path problem." } ] }, @@ -18797,7 +18911,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "c5f19abbf8de2f2cd29b625037c120f819e1f3cc4b737ae85d2d68b26622cc43", + "prepared_review_unit_hash": "66e432627a47cded58e4a015026a44e5335ab76983e52c4a1ff2e92623ad0d25", "component_budget": { "node_count": 2, "item_count": 2, @@ -18819,18 +18933,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two separate pipeline feature PRs in the same segmentation family: one for promptable concept segmentation and one for promptable visual segmentation. They share scaffolding, but they target different concrete pipelines and model/test coverage, so they should not be merged as duplicates.", - "confidence": 0.95, + "summary": "Two open PRs add different segmentation pipelines with similar scaffolding, but they target different features and code paths: promptable concept segmentation vs promptable visual segmentation.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No duplicate PR here; if one must be used as the cluster representative, PR #43613 is the broader and more concrete implementation (visual segmentation, multiple model integrations, and dedicated tests/docs).", + "canonical_pr_reason": "No true duplicate PR exists in this cluster. If a representative is needed, 43613 is slightly more central due to broader model coverage and more review activity, but it is still a different pipeline from 43612.", "best_issue_reason": null, - "best_pr_reason": "PR #43613 is the best representative because it is the more complete feature PR with wider code and model coverage. PR #43612 is a distinct concept-segmentation pipeline and not the same underlying change.", + "best_pr_reason": "43613 is the better representative for the cluster only because it is the broader/more active PR, not because it duplicates 43612; the underlying functionality differs.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different features: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. Shared docs/pipeline scaffolding and adjacent filenames are not enough to treat them as the same fix/change." + "reason": "Do not merge: they implement different pipelines with different targets and model integrations. Shared docs/auto-registration/test scaffolding is generic and not enough to treat them as the same underlying change." } ] }, @@ -18860,12 +18974,12 @@ "soft_pairs": [ "pull_request:43649|pull_request:46301" ], - "prepared_review_unit_hash": "48b04474e099c854a3b832ba6105c08d3668f1b052f326ddfcdb3928dd1d4954", + "prepared_review_unit_hash": "2387052f69f7b5368c6357b031e26e85c117eab5ad7cfb97c24c885186f912e7", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4550, + "serialized_chars": 4549, "estimated_input_tokens": 1138, "estimated_eval_tokens": 2532 }, @@ -18873,7 +18987,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4550, + "serialized_chars": 4549, "estimated_input_tokens": 1138, "estimated_eval_tokens": 2532 }, @@ -18882,18 +18996,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests are not duplicates: #43649 is a broad CI/workflow and failure-reporting change, while #46301 is a separate CircleCI/test-marker update for tensor-parallel coverage.", + "summary": "Both items are pull requests, but they address different CI/test infrastructure changes and do not look like the same underlying bug or change.", "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "#46301 is the tighter, more self-contained PR, but it is not a duplicate of #43649; the two changes target different CI/test paths and would not naturally merge into one fix.", + "canonical_pr_reason": "PR 43649 is the better canonical PR because it has the more specific title and a clearer, broader CI/reporting context than the generic 'check' PR 46301.", "best_issue_reason": null, - "best_pr_reason": "#46301 is the better representative of the cluster only in the sense that it is more focused; #43649 is broader and touches several unrelated workflow/test files.", + "best_pr_reason": "PR 43649 is the strongest representative of the cluster: it is more descriptive, has more evidence of a concrete workflow/reporting change, and is less ambiguous than PR 46301.", "soft_edge_verdicts": [ { "left": "pull_request:43649", "right": "pull_request:46301", "accept": false, - "reason": "Different underlying changes: #43649 updates self-comment CI and failure-reporting workflows/tests, while #46301 changes CircleCI test selection and tensor-parallel mixin tests. They do not fix the same concrete bug or code path." + "reason": "Reject: the diffs target different code paths. 43649 modifies self-comment CI, scheduled workflow behavior, and test/reporting utilities for failure reporting, while 46301 changes CircleCI test selection and tensor-parallel mixin tests. They are not the same fix and would not plausibly be merged into one PR." } ] }, @@ -18923,40 +19037,40 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "cdc9ec597edd2b276a343014c7f2108853da706165b01e9c104ea08fb7d76c21", + "prepared_review_unit_hash": "8941adea8e8b83723543430feefe68bc0264fbd68878d6b3e7881575137e59e8", "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": "These are two unrelated trainer.py PRs: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for custom trainers. They share a file and a broad subsystem, but not the same underlying bug or change.", + "summary": "Both PRs touch `src/transformers/trainer.py` and add extensibility hooks for custom trainers, but they address different code paths: gradient-accumulation loss scaling vs. signature-column filtering. They are not duplicates.", "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43651 is the better representative of the cluster because it makes a focused behavioral change in training_step with a clear bug-fix motivation, whereas #43663 is a separate extensibility refactor around input-column filtering.", + "canonical_pr_reason": "PR 43663 is the cleaner representative of this cluster because it introduces a reusable helper for column-signature inspection, but it is not the same change as 43651.", "best_issue_reason": null, - "best_pr_reason": "PR #43651 is the stronger standalone candidate: it targets a concrete training-time loss-scaling path and is easier to treat as the cluster's representative change.", + "best_pr_reason": "PR 43663 is the best single PR to represent the cluster\u2019s theme of Trainer extensibility, though it should not be merged with 43651 as a duplicate.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Reject: these modify different Trainer code paths and solve different problems\u2014gradient accumulation loss scaling vs. signature-column filtering/override hooks. They are not the same underlying fix and would not plausibly be merged as one PR." + "reason": "Different Trainer behaviors: 43651 changes loss normalization during gradient accumulation, while 43663 refactors how allowed input columns are derived for PEFT/custom trainers. Same file and broad subsystem, but not the same underlying fix." } ] }, @@ -18986,12 +19100,12 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "04458fae4bb3f6a9efbb45a28b570702bc4ee74d47b8b6e5cbe5eb811b13f12e", + "prepared_review_unit_hash": "11519020dfaecfaeb4e96b8a4b731740edef7ff2b6ab3b5baf759a5e5ea05eec", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -18999,7 +19113,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 3663, + "serialized_chars": 3664, "estimated_input_tokens": 916, "estimated_eval_tokens": 2088 }, @@ -19008,24 +19122,27 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two different CLI typing/name-resolution fixes in the same file, not the same underlying change: one targets a pydantic/TypeAdapter import annotation issue, the other is a broader forward-reference typing cleanup.", - "confidence": 0.89, + "summary": "PR 43842 is the most concrete standalone fix for the `TypeAdapter`/`pydantic`-absent failure in `src/transformers/cli/serve.py`. PR 43656 touches the same file but looks like a broader typing/__future__ cleanup and is not clearly the same bugfix, so the pair should not be merged as duplicates.", + "confidence": 0.86, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:43842 is the more canonical PR because it directly addresses the reported TypeAdapter NameError when pydantic is not installed, with a minimal targeted fix in the validator annotation/import path.", + "canonical_pr_reason": "PR 43842 is the clearest direct fix: it explicitly addresses the `TypeAdapter NameError when pydantic is not installed` by moving the import under `TYPE_CHECKING` and quoting the annotation.", "best_issue_reason": null, - "best_pr_reason": "pull_request:43842 is the best representative because its diff and title align tightly with the concrete runtime error it claims to fix; pull_request:43656 is a separate typing/annotation adjustment in the same module.", + "best_pr_reason": "PR 43842 is the best standalone PR for this cluster because it targets the specific runtime/import failure with a minimal, tightly scoped change.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Shared file alone is not enough here: 43656 changes general typing/forward-reference annotations, while 43842 specifically fixes TypeAdapter handling when pydantic is absent. They do not appear to be the same underlying bug or mergeable as one fix." + "reason": "Reject: 43656\u2019s diff is a broader typing/`__future__` cleanup in the same file, while 43842 specifically fixes `TypeAdapter` when `pydantic` is missing. Shared filename alone is not enough to treat them as the same bugfix or mergeable change." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The packet does not strongly support treating these as the same bugfix or as clearly mergeable. PR 43842 is explicitly about `TypeAdapter` when `pydantic` is not installed, while PR 43656\u2019s diff excerpt shows a broader `__future__`/typing cleanup in the same file but does not clearly establish the same fix. The soft edge verdict is too strong given only shared-file similarity and low Jaccard overlap." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -19211,388 +19328,304 @@ "pull_request:44044|pull_request:44129", "pull_request:44044|pull_request:44154" ], - "prepared_review_unit_hash": "978f12ff6f3e9e2e7575b23bed15c6b487f00be33cc01c13b8d09f09b6da2db5", + "prepared_review_unit_hash": "c8723e15ab65b85c70c9892164b1bcdd280d41d69123536580d929ab1c4858f7", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84895, - "estimated_input_tokens": 21224, - "estimated_eval_tokens": 42704 + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 }, "budget": { "node_count": 27, "item_count": 27, "soft_pair_count": 72, - "serialized_chars": 75091, - "estimated_input_tokens": 18773, - "estimated_eval_tokens": 37802 + "serialized_chars": 75096, + "estimated_input_tokens": 18774, + "estimated_eval_tokens": 37804 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of separate, model-specific refactors to standardized output tracing. The strongest duplicate subgroup is GPT-Neo, and the GPT-J/CodeGen subgroup is also plausible, but acceptance should stay limited to clearly overlapping concrete refactors.", - "confidence": 0.84, + "summary": "This cluster is an umbrella of output-tracing refactors across many Transformer model families. The only true duplicate-style overlaps are within the GPT-Neo line and the GPT-J/CodeGen line; cross-family similarities are mostly theme-level, not the same concrete change. I corrected the GPT-J/CodeGen overlap to accept as a soft edge.", + "confidence": 0.74, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44722 is the best canonical anchor for the GPT-J/CodeGen subgroup because it is the most complete and recent iteration, spans both modeling_gptj.py and modeling_codegen.py, and directly targets #43979.", + "canonical_pr_reason": "PR 44722 is the best representative because it is the most complete GPT-J/CodeGen refactor in the cluster, explicitly targets the standardized output-tracing change, and overlaps the same GPT-J code path as the earlier GPT-J-only PRs.", "best_issue_reason": null, - "best_pr_reason": "44722 is the best single PR representative for the real duplicate cluster: it captures the broad GPT-J/CodeGen output-tracing refactor more completely than the earlier GPT-J-only iterations, while the rest of the cluster is mostly other model-specific refactors.", + "best_pr_reason": "PR 44722 is the strongest single PR to keep as the cluster representative: it is the latest, broadest GPT-J/CodeGen variant and best captures the recurring standardized-tracing refactor pattern.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model-specific refactors; only the generic output-tracing theme overlaps." + "reason": "Different model families (CVT/FNet vs GPT-J); only the refactor theme matches." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Different model-specific refactors; the shared tracing theme is too broad." + "reason": "Different model implementations; shared issue target does not make this the same change." }, { "left": "pull_request:44066", "right": "pull_request:44085", - "accept": false, - "reason": "Same GPT-J file, but 44066 is the broader GPT-J/CodeGen refactor while 44085 is a narrower GPT-J-only variant; the overlap is only a shared slice of the file, not one mergeable change." + "accept": true, + "reason": "Same GPT-J output-tracing refactor line, with overlapping GPT-J code and the same standardized decorator migration." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "Different image-model code paths; same output-tracing theme only." + "reason": "RegNet/ResNet versus EfficientNet are separate model refactors, not one duplicate change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "Different model families and concrete code paths; only the standardized-output pattern overlaps." + "reason": "EfficientNet and GPT-J/CodeGen are unrelated concrete code paths." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen vs EfficientNet are unrelated concrete refactors." + "reason": "GPT-J/CodeGen versus EfficientNet are different model-specific refactors." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "Different model families and concrete code paths." + "reason": "MobileNetV2 and DeBERTa are distinct implementations with unrelated output code." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen vs MGP-STR are unrelated concrete changes." + "reason": "GPT-J/CodeGen and MGP-STR are different model code paths." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen vs MPT are unrelated concrete code paths." + "reason": "GPT-J/CodeGen and MPT are separate model refactors." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Both are GPT-Neo output-tracing migrations; 44068 looks like the same concrete refactor path with later cleanup/tests and could plausibly be merged into one PR." + "reason": "Same GPT-Neo output-tracing refactor on the same model/file family." }, { "left": "pull_request:44018", "right": "pull_request:44066", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen are different concrete refactors." + "reason": "GPT-Neo and GPT-J/CodeGen are different models." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen vs GPT-Neo are different model-specific refactors." + "reason": "GPT-J/CodeGen versus GPT-Neo are different concrete model paths." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet and GPT-J/CodeGen are unrelated concrete changes." + "reason": "ResNet and GPT-J/CodeGen are unrelated model implementations." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "GPT-J/CodeGen vs MGP-STR are unrelated concrete changes." + "reason": "MGP-STR and GPT-J/CodeGen are different model-specific refactors." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet vs GPT-J are unrelated concrete changes." + "reason": "ResNet versus GPT-J are different code paths." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo vs ResNet are unrelated model-specific refactors." + "reason": "GPT-Neo and ResNet are separate model changes." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet vs MPT are unrelated concrete changes." + "reason": "ResNet and MPT are unrelated implementations." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert are unrelated concrete refactors." + "reason": "GPT-Neo and Flaubert are different model refactors." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert are unrelated concrete refactors." + "reason": "GPT-Neo and Flaubert are different code paths." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen are different concrete code paths." + "reason": "GPT-Neo and GPT-J/CodeGen are separate model implementations." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT vs GPT-J/CodeGen are unrelated concrete changes." + "reason": "MPT and GPT-J/CodeGen are different implementations." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo vs GPT-J are different model-specific refactors." + "reason": "GPT-Neo and GPT-J are different models." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo vs MGP-STR are unrelated concrete changes." + "reason": "GPT-Neo and MGP-STR are different model-specific changes." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo vs MGP-STR are unrelated concrete changes." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44085", - "accept": false, - "reason": "MPT vs GPT-J are unrelated concrete changes." + "reason": "GPT-Neo and MGP-STR are different model families." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT2 vs MGP-STR are unrelated concrete refactors." + "reason": "GPT-2 and MGP-STR are unrelated implementations." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J vs MGP-STR are unrelated concrete changes." + "reason": "GPT-J and MGP-STR are different model code paths." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT vs Flaubert/XLM are unrelated concrete changes." - }, - { - "left": "pull_request:44059", - "right": "pull_request:44722", - "accept": false, - "reason": "GPT2 vs GPT-J/CodeGen are unrelated concrete refactors." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-Neo vs MPT are different model code paths." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44722", - "accept": false, - "reason": "SegFormer vs GPT-J/CodeGen are unrelated concrete changes." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44161", - "accept": false, - "reason": "SegFormer vs LongT5 are unrelated concrete changes." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44073", - "accept": false, - "reason": "SegFormer vs VisualBert are unrelated concrete changes." - }, - { - "left": "pull_request:44068", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-Neo vs MPT are unrelated model-specific refactors." + "reason": "ViLT and Flaubert are separate model implementations; only the style of refactor overlaps." }, { "left": "pull_request:43997", "right": "pull_request:44073", "accept": false, - "reason": "RegNet vs VisualBert are unrelated concrete changes." + "reason": "RegNet and VisualBert are different concrete changes." }, { - "left": "pull_request:44071", - "right": "pull_request:44086", + "left": "pull_request:44056", + "right": "pull_request:44101", "accept": false, - "reason": "MPT vs MGP-STR are unrelated concrete changes." + "reason": "MPNet and Flaubert are unrelated model refactors." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet vs Flaubert/XLM are unrelated concrete changes." + "reason": "TextNet and Flaubert are different model code paths." }, { - "left": "pull_request:44056", - "right": "pull_request:44101", + "left": "pull_request:43996", + "right": "pull_request:44161", "accept": false, - "reason": "MPNet vs Flaubert/XLM are unrelated concrete changes." + "reason": "CVT/FNet and LongT5 are separate model implementations." }, { - "left": "pull_request:44059", - "right": "pull_request:44068", + "left": "pull_request:44017", + "right": "pull_request:44722", "accept": false, - "reason": "GPT2 vs GPT-Neo are different concrete refactors." + "reason": "Segformer and GPT-J/CodeGen are unrelated concrete changes." }, { - "left": "pull_request:44101", + "left": "pull_request:44017", "right": "pull_request:44161", "accept": false, - "reason": "Flaubert/XLM vs LongT5 are unrelated concrete changes." + "reason": "Segformer and LongT5 are different models." }, { "left": "pull_request:44017", - "right": "pull_request:44018", - "accept": false, - "reason": "SegFormer vs GPT-Neo are unrelated model-specific refactors." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44056", + "right": "pull_request:44073", "accept": false, - "reason": "RegNet vs MPNet are unrelated concrete changes." + "reason": "Segformer and VisualBert are separate model refactors." }, { - "left": "pull_request:44071", - "right": "pull_request:44076", + "left": "pull_request:44059", + "right": "pull_request:44071", "accept": false, - "reason": "MPT vs ImageGPT are unrelated concrete code paths." + "reason": "GPT-2 and MPT are different model code paths." }, { "left": "pull_request:43997", - "right": "pull_request:44018", - "accept": false, - "reason": "RegNet vs GPT-Neo are unrelated model-specific refactors." - }, - { - "left": "pull_request:44056", - "right": "pull_request:44161", + "right": "pull_request:44074", "accept": false, - "reason": "MPNet vs LongT5 are unrelated concrete changes." + "reason": "RegNet and TextNet are unrelated concrete changes." }, { - "left": "pull_request:43997", - "right": "pull_request:44161", + "left": "pull_request:44071", + "right": "pull_request:44129", "accept": false, - "reason": "RegNet vs LongT5 are unrelated concrete changes." + "reason": "MPT and SpeechT5 are different implementations." }, { - "left": "pull_request:44074", - "right": "pull_request:44161", + "left": "pull_request:44010", + "right": "pull_request:44071", "accept": false, - "reason": "TextNet vs LongT5 are unrelated concrete changes." + "reason": "SqueezeBert and MPT are separate model refactors." }, { "left": "pull_request:44018", - "right": "pull_request:44161", + "right": "pull_request:44129", "accept": false, - "reason": "GPT-Neo vs LongT5 are unrelated concrete changes." + "reason": "GPT-Neo and SpeechT5 are different model code paths." }, { "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "SqueezeBert vs VisualBert are unrelated concrete code paths." + "reason": "SqueezeBert and VisualBert are unrelated model changes." }, { "left": "pull_request:44010", - "right": "pull_request:44044", - "accept": false, - "reason": "SqueezeBert vs DeBERTa are unrelated concrete refactors." - }, - { - "left": "pull_request:44044", - "right": "pull_request:44129", + "right": "pull_request:44018", "accept": false, - "reason": "DeBERTa vs SpeechT5 are unrelated concrete changes." + "reason": "SqueezeBert and GPT-Neo are different models." }, { "left": "pull_request:44044", - "right": "pull_request:44154", + "right": "pull_request:44074", "accept": false, - "reason": "DeBERTa vs ViLT are unrelated concrete changes." + "reason": "DeBERTa and TextNet are distinct implementations." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor family; 44722 looks like the fuller iteration and can plausibly absorb the earlier GPT-J slice." + "reason": "Same GPT-J/CodeGen standardized output-tracing refactor line, with overlapping GPT-J/CodeGen files and the same issue target." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J output-tracing refactor family; both target the same concrete code path in modeling_gptj.py." - }, - { - "left": "pull_request:44085", - "right": "pull_request:44129", - "accept": false, - "reason": "GPT-J vs SpeechT5 are unrelated concrete changes." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44007", - "accept": false, - "reason": "Different model families and code paths." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44129", - "accept": false, - "reason": "Different model families and code paths." - }, - { - "left": "pull_request:44044", - "right": "pull_request:44129", - "accept": false, - "reason": "Different model families and code paths." + "reason": "Same GPT-J output-tracing refactor; 44722 extends the same code-path change with CodeGen." } ] }, "evaluator_result": { "accept": false, - "feedback": "The overall cluster read is mostly grounded, but at least one soft-edge rationale is too strong. In particular, `44066` and `44085` both touch `src/transformers/models/gptj/modeling_gptj.py`, so describing them as having \u201cdifferent models and code paths\u201d overstates the separation. If you want to keep the pair rejected, the reason should be narrower (e.g. scope/iteration differences), not a full model-family mismatch. The GPT-Neo subgroup and the GPT-J/CodeGen subgroup are plausible, but the verdicts need to stay more conservative." + "feedback": "Reject: one soft-edge verdict is not grounded in the packet. The pair pull_request:44066 vs pull_request:44722 was marked false with a reason about GPT-Neo, but both PRs are in the GPT-J/CodeGen output-tracing refactor line, share the same issue target (#43979), and touch overlapping GPT-J/CodeGen files. That makes the rejection too strong for a soft edge. The rest is mostly conservative, but this misclassification is enough to reject." }, "evaluator_used": true, "retried": true, @@ -19630,14 +19663,14 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "44a953abd41f81c17278a6cd6181464db4dd8a42b780fc0aa484a5742caf6aac", + "prepared_review_unit_hash": "948d2037b7caef8eb0ca2dccd93f6a19cea1602f469db697609817963047a62b", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84895, - "estimated_input_tokens": 21224, - "estimated_eval_tokens": 42704 + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 }, "budget": { "node_count": 8, @@ -19652,48 +19685,48 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are all model-specific PRs for the same broad refactor pattern: migrating output tracing to standardized capture/decorator helpers. They share theme and terminology, but each targets a different model/code path, so they should not be deduped as the same change.", - "confidence": 0.88, + "summary": "A set of model-specific PRs refactoring output tracing to the new capture_outputs/can_return_tuple pattern. They share a common migration theme, but each targets a different architecture and code path, so they are not duplicates of one another.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44129 is the clearest representative of the set: it has an explicit issue link, a substantive output-tracing refactor, and good traceability compared with the others.", + "canonical_pr_reason": "44129 is the strongest representative PR: it has the clearest title, an explicit issue link, and a substantial concrete refactor in a single model implementation. That said, it is still model-specific rather than a cluster-wide duplicate fix.", "best_issue_reason": null, - "best_pr_reason": "PR 44129 is the strongest single representative of this cluster because it is explicitly tied to issue 43979 and concretely implements the standardized output capture migration.", + "best_pr_reason": "44129 is the best single PR candidate because it is well-scoped, explicitly tracked to an issue, and shows a concrete standardized output-capture change rather than just a naming cleanup. It is not merge-equivalent to the others, but it is the most complete exemplar of the pattern.", "soft_edge_verdicts": [ { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "Both are standardized-output refactors, but they apply to different model implementations (FocalNet vs ImageGPT) and different forward paths, so they are not the same concrete change." + "reason": "Both are output-tracing refactors, but they touch different models (FocalNet vs ImageGPT) and different internal hooks. Same migration pattern, not the same underlying bug or change." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "These are separate model migrations (VisualBert vs TextNet). Similar refactor style, but not the same underlying bug or mergeable single PR." + "reason": "VisualBert and TextNet are unrelated model implementations. These are separate standardized-output refactors, not one concrete code-path fix." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "MPNet and VisualBert are unrelated model-specific output-tracing updates; they share a theme, not a duplicate code-path fix." + "reason": "MPNet and VisualBert each update their own output-capture plumbing. They share only the refactor theme, not a mergeable shared defect." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "MPNet and TextNet touch different model code paths, so this is not the same underlying change." + "reason": "Different models and different forward/output handling. This is a broad migration pattern, not a duplicate PR pair." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are distinct model refactors; same helper migration pattern, but not the same concrete fix." + "reason": "RegNet and SqueezeBert are separate model-specific implementations. The similarity is only the standardized output tracing migration." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 are different model-specific output-capture migrations; they are too broad to treat as duplicates." + "reason": "SqueezeBert and SpeechT5 modify different code paths and model classes. They are not the same concrete fix and would not plausibly be merged into one PR." } ] }, @@ -19772,12 +19805,12 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "68d838d362ec6fea6954389f416bd371d3224a9f86ce4e7b2ca27cf2bb2ea961", + "prepared_review_unit_hash": "bd357bf09aa2aca7f7f87c0149a8b9c071e312e6c850bb71168573345a30a03f", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 39, - "serialized_chars": 39777, + "serialized_chars": 39778, "estimated_input_tokens": 9945, "estimated_eval_tokens": 20146 }, @@ -19785,7 +19818,7 @@ "node_count": 13, "item_count": 13, "soft_pair_count": 39, - "serialized_chars": 39777, + "serialized_chars": 39778, "estimated_input_tokens": 9945, "estimated_eval_tokens": 20146 }, @@ -19794,246 +19827,246 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs all share a superficial 'refactor output tracing' theme, but they target different model families and distinct forward-path signatures; they are not duplicates of the same underlying fix.", - "confidence": 0.95, + "summary": "This cluster is a theme cluster of unrelated model-specific output-tracing refactors, not a duplicate cluster. The PRs share similar mechanics and titles, but they touch different model implementations and code paths, so they should not be deduped together.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44003 is the broadest and most representative of the cluster\u2019s theme, with substantive output-tracing changes across closely related Mamba-family files.", + "canonical_pr_reason": "PR 44003 is the broadest, most code-heavy refactor in the set and the clearest representative of the shared output-tracing theme.", "best_issue_reason": null, - "best_pr_reason": "PR 44003 is the best single representative if one needs a canonical item for this cluster, but it still should not be merged as a duplicate with the others.", + "best_pr_reason": "PR 44003 is the best single representative to inspect first because it shows the pattern clearly and spans two closely related Mamba implementations.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different model backends and forward paths; same refactor motif only." + "reason": "Same refactor theme, but timm_backbone and upernet are different model code paths with no evidence of a shared underlying bug." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Timm backbone vs vision-text dual encoder are unrelated code paths." + "reason": "Both are output-tracing refactors, but they target different architectures and forward paths; not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Different models; both tweak output handling but not the same bug." + "reason": "Different models and outputs; this is a parallel cleanup, not a duplicate fix." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Timm backbone refactor is separate from Mamba-family tracing changes." + "reason": "timm_backbone and mamba/falcon_mamba are unrelated implementations; similar decorator/import changes are not enough for duplication." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Different architectures; no shared concrete defect." + "reason": "Different model families and different forward implementations; no shared concrete bug or change." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "MobileNetV1 and UnivNet have distinct forward implementations and outputs." + "reason": "MobileNetV1 and UnivNet are unrelated model paths; similar output-tracing cleanup does not make them duplicates." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "MobileNetV1 and UperNet are unrelated model paths." + "reason": "Different architectures and files, with no indication of the same bug or mergeable patch." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Different model families; no single shared code-path problem." + "reason": "These are separate model-specific refactors; they share a pattern but not an underlying code-path issue." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "MobileNetV1 output tracing is unrelated to encoder-decoder refactors." + "reason": "MobileNetV1 and vision_encoder_decoder are unrelated forward paths; too broad to treat as the same fix." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Distinct models with different signatures and outputs." + "reason": "Different encoder-decoder wrapper and different modality; not the same concrete change." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Vision-text dual encoder and UnivNet are different code paths." + "reason": "Vision-text dual encoder and UnivNet are separate model implementations; only the refactor style matches." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different model classes; same refactor pattern only." + "reason": "Different model families and code paths; no shared defect or patch." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "Vision-text dual encoder is unrelated to Mamba-family output tracing." + "reason": "CLIP-like dual encoder vs Mamba models are unrelated; similarity is only at the refactor pattern level." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "Different subsystems and forward signatures." + "reason": "Different models and different forward semantics; these are not duplicate fixes." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "UnivNet and UperNet do not share a concrete implementation bug." + "reason": "UnivNet and UperNet are unrelated; this is only a similar output-tracing cleanup." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "UnivNet output refactor is separate from Mamba-family changes." + "reason": "Different code paths and architectures; no evidence they fix the same underlying issue." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Different models and output contracts." + "reason": "These are distinct model-specific refactors, not a single mergable PR." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "UperNet segmentation and RWKV language modeling are unrelated." + "reason": "UperNet and RWKV are separate models with different output handling; not the same change." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "UperNet and Mamba-family refactors affect different code paths." + "reason": "Different subsystems and implementations; similar output-tracing edits are insufficient for duplication." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Mamba-family tracing changes and CodeGen cache/output changes are different fixes." + "reason": "Mamba/Falcon-Mamba and CodeGen are different model families; the output-refactor implementations are not the same concrete fix." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Timm backbone and CodeGen touch unrelated model stacks." + "reason": "timm_backbone and CodeGen are unrelated code paths; no duplicate-level evidence." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "UnivNet and CodeGen have different output semantics and bugs." + "reason": "UnivNet and CodeGen do not share the same underlying bug or change." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "UperNet segmentation refactor is not the same as CodeGen cache/output refactor." + "reason": "UperNet and CodeGen are unrelated beyond the generic refactor pattern." }, { "left": "pull_request:44000", "right": "pull_request:44004", "accept": false, - "reason": "Vision-text dual encoder and CodeGen are distinct code paths." + "reason": "Different model families and different forward logic; not a shared concrete patch." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Depth Anything and Vision-Encoder-Decoder are separate model families." + "reason": "DepthAnything and vision_encoder_decoder are separate model paths; same maintenance theme, not the same fix." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Depth Anything and Speech-Encoder-Decoder do not share the same bug." + "reason": "Different models and different forward implementations; not duplicates." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Depth estimation and SuperPoint feature extraction are unrelated." + "reason": "DepthAnything and SuperPoint are unrelated code paths, so they should stay separate." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "Depth Anything and RWKV are unrelated architectures." + "reason": "Different model families and output semantics; no evidence of one underlying bug." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Depth Anything and DPR have different forward contracts." + "reason": "DepthAnything and DPR are unrelated models; similar API cleanup is not enough." }, { "left": "pull_request:44026", - "right": "pull_request:44027", + "right": "pull_request:44029", "accept": false, - "reason": "Encoder-decoder wrappers for vision and speech are different implementations." + "reason": "Vision-encoder-decoder and RWKV are unrelated implementations with different output-tracing changes." }, { "left": "pull_request:44026", - "right": "pull_request:44028", + "right": "pull_request:44027", "accept": false, - "reason": "Vision encoder-decoder and SuperPoint are unrelated code paths." + "reason": "These are both encoder-decoder wrappers, but they target different modalities and files; too broad to merge as one duplicate fix." }, { "left": "pull_request:44026", - "right": "pull_request:44029", + "right": "pull_request:44028", "accept": false, - "reason": "Different model families with no shared concrete fix." + "reason": "Different model families and code paths; no shared concrete bug." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "Vision encoder-decoder and DPR differ in both purpose and outputs." + "reason": "Vision-encoder-decoder and DPR are unrelated; only the refactor motif is shared." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "Speech encoder-decoder and SuperPoint are unrelated subsystems." + "reason": "Speech-encoder-decoder and SuperPoint are unrelated model paths; not the same change." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Speech encoder-decoder and RWKV do not share a mergeable fix." + "reason": "Different architectures and forward code; no duplicate-level overlap." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Different model wrappers and output paths." + "reason": "Speech-encoder-decoder and DPR are different model families; this is only stylistic similarity." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "SuperPoint and RWKV are entirely different models." + "reason": "SuperPoint and RWKV are unrelated implementations; not the same underlying fix." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "SuperPoint feature extraction is unrelated to DPR retrieval encoding." + "reason": "SuperPoint and DPR are different models and code paths; keep separate." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "RWKV and DPR are different architectures with different output issues." + "reason": "RWKV and DPR have different forward logic and output handling; no shared concrete bug or mergeable change." } ] }, @@ -20066,52 +20099,52 @@ "pull_request:44662|pull_request:45075", "pull_request:45075|pull_request:45401" ], - "prepared_review_unit_hash": "a716170a19e1284f09657545d173940580fa5ecf4923bdb1e289ea4abfae76bb", + "prepared_review_unit_hash": "fad38a6253374a85944eccf3881fa9ae6d5bf513ed2bf1aa1fae305f104b0000", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10912, - "estimated_input_tokens": 2728, - "estimated_eval_tokens": 5712 + "serialized_chars": 10913, + "estimated_input_tokens": 2729, + "estimated_eval_tokens": 5714 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 10912, - "estimated_input_tokens": 2728, - "estimated_eval_tokens": 5712 + "serialized_chars": 10913, + "estimated_input_tokens": 2729, + "estimated_eval_tokens": 5714 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three PRs are separate model-integration additions (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). The overlap is limited to shared Transformer boilerplate such as auto-registry/docs updates, not the same bug or code path.", - "confidence": 0.97, + "summary": "These are three separate model-addition PRs for unrelated models (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). They share some Auto/docs scaffolding, but the concrete model code, filenames, and target capabilities are different enough that they should not be merged as duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: these are distinct model-addition PRs for different architectures and do not appear to be duplicates.", + "canonical_pr_reason": "No single canonical PR: each PR implements a different model integration with its own configuration, modeling, processing, tests, and docs.", "best_issue_reason": null, - "best_pr_reason": "45075 is the strongest standalone representative only in the sense that it is a full model integration with substantial implementation, conversion, and test coverage; however, it is still not a duplicate of the others.", + "best_pr_reason": "If a representative is needed, PR 45401 is the most explicitly scoped (linked to issue 45027) and self-contained, but it is still not a duplicate of the other two PRs.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Different model implementations for PenguinVL vs Voxtral TTS. Shared files are only registry/docs boilerplate, so they do not fix the same concrete code-path problem." + "reason": "Both add new model integrations, but for different models and codepaths (PenguinVL vs Voxtral TTS). Shared Auto/doc files are generic scaffold, not the same underlying change." }, { "left": "pull_request:44662", "right": "pull_request:45075", "accept": false, - "reason": "PenguinVL and DeepSeek-OCR-2 are unrelated model additions. Similarities are limited to common auto-mapping and documentation plumbing, not the same underlying change." + "reason": "PenguinVL and DeepSeek-OCR-2 are distinct model additions with different model families, processors, and tests; overlap is only in common registration/docs infrastructure." }, { "left": "pull_request:45075", "right": "pull_request:45401", "accept": false, - "reason": "DeepSeek-OCR-2 and Voxtral TTS add different model families with separate configs, modeling, and processors. The shared touchpoints are generic integration scaffolding only." + "reason": "DeepSeek-OCR-2 and Voxtral TTS are unrelated feature additions for different models and tasks; they do not appear to fix or implement the same concrete change." } ] }, @@ -20141,7 +20174,7 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "8ccb753ab6fe5e69b80e15449e88547fdebfd49d92b38f707c65eb950ef8b4c3", + "prepared_review_unit_hash": "6b4248c86a5fd7616cfab8c71c6aadc0bc7198b6cca89dc9647b3ef3cd8ccde0", "component_budget": { "node_count": 2, "item_count": 2, @@ -20163,18 +20196,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are thematically related to module/kernel fusion, but they target different layers and likely different changes: one introduces a generic Module Fusion API, the other wires up n-to-1 kernel fusion through KernelConfig and model loading. They do not look like the same concrete bugfix or mergeable duplicate PR.", - "confidence": 0.82, + "summary": "These PRs are related by the general idea of fusion, but they target different concrete changes: one adds a new module-fusion API, the other wires up n-to-1 kernel fusion through `KernelConfig`. They are not duplicates.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45363 is the better canonical representative because it implements the more concrete end-user feature path (KernelConfig-driven kernel fusion) across the actual integration and model-loading code, rather than only adding a standalone API layer.", + "canonical_pr_reason": "PR 44979 is the more foundational change: it introduces the standalone `Module Fusion API` and the new fusion abstraction. PR 45363 is a separate, more specific integration built around kernel configs and fused kernels.", "best_issue_reason": null, - "best_pr_reason": "PR 45363 is the strongest standalone candidate: it has a clearer scoped feature, touches the runtime integration path, and includes an example demonstrating the fused-vs-original behavior.", + "best_pr_reason": "PR 44979 is the better cluster anchor because it defines the core fusion mechanism; PR 45363 depends on a different code path and is better treated as a follow-on feature rather than the same change.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Related fusion work, but not the same underlying change. 44979 adds a generic Module Fusion API and tests; 45363 implements KernelConfig-based n-to-1 kernel fusion across integrations/model loading. Different scopes and code paths, so not a duplicate PR pair." + "reason": "Both mention fusion, but 44979 adds a general module-fusion API while 45363 implements n-to-1 kernel fusion via `KernelConfig` and integration hooks. Different underlying change, not one PR." } ] }, @@ -20204,12 +20237,12 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "b35bf2fbded94da0c544ac47797f2f99de1e6a13c6c7d3703f77dd42e50d952b", + "prepared_review_unit_hash": "a5bf13985e99c331bf45424b997bd4d6ccb24739f70993a61422e3a84aa390c6", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -20217,7 +20250,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4414, + "serialized_chars": 4413, "estimated_input_tokens": 1104, "estimated_eval_tokens": 2464 }, @@ -20226,18 +20259,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests address the same Mistral regex patch AttributeError in the same file and on the same code path; 45317 is the more complete version because it adds tests.", - "confidence": 0.96, + "summary": "Both PRs patch the same Mistral regex pre-tokenizer AttributeError in the same code path; 45317 is the fuller version with added test coverage, so it is the better canonical PR.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "45317 is the better canonical PR because it contains the same fix as 45086 plus test coverage, making it the more complete representative of the change.", + "canonical_pr_reason": "45317 makes the same fix as 45086 in the same function, but also adds a test file, so it is the more complete representative of this change.", "best_issue_reason": null, - "best_pr_reason": "45317 is the best PR candidate since it fixes the same concrete bug as 45086 and also includes accompanying tests.", + "best_pr_reason": "45317 is the best standalone PR because it addresses the same concrete bug and includes regression tests, making it more merge-ready than the minimal patch in 45086.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same AttributeError fix in _patch_mistral_regex, same target issue, same code path, and the second PR is just the tested/expanded version of the first." + "reason": "Same underlying AttributeError fix in _patch_mistral_regex, same file, and the second PR is a superset with tests; they could plausibly be merged as one change." } ] }, @@ -20267,7 +20300,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "0effe5c08ab4ce6a6bf7e453bd2226144828b07127568b45efc3c8ddcbe4591f", + "prepared_review_unit_hash": "893db24b5af07f76928b9e914b3273d611023953fdb9df1e35a027d7ed9e9604", "component_budget": { "node_count": 2, "item_count": 2, @@ -20289,18 +20322,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs touch Nemotron H internals, but they are not duplicates: #45115 is a narrow conversion/mapping and inheritance refactor, while #45569 is a broader model-family split introducing dedicated dense/sparse variants, docs, auto-mapping updates, and tests.", - "confidence": 0.89, + "summary": "Two NemotronH PRs in the same area, but they implement different changes: one refactors conversion/mapping and inheritance, while the other splits NemotronH into dense/sparse model families with docs and tests. Not a duplicate pair.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the more comprehensive and representative change for the Nemotron H family work: it adds the dense/sparse model variants, updates docs and auto-mapping, and includes tests, making it the better canonical PR for this area.", + "canonical_pr_reason": "PR #45569 is the better representative of the cluster because it is the more complete user-facing implementation change: it adds the dense/sparse model split plus docs and tests, whereas #45115 is a narrower refactor/conversion-mapping change.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 is the best single representative because it covers the full user-visible model restructuring, not just the underlying conversion/mapping cleanup. PR #45115 is a narrower implementation step and not a duplicate of the larger split.", + "best_pr_reason": "PR #45569 has the broader and more finished scope, touching the actual model packaging and documentation rather than only conversion/mapping internals.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Shared Nemotron H files and broad theme, but the concrete work differs: #45115 adds checkpoint conversion mappings and inheritance-related refactoring, while #45569 introduces the actual NemotronHDense/NemotronHSparse split plus docs, auto mappings, and tests. Not the same underlying change and not safely mergeable as one PR." + "reason": "Related NemotronH work, but the concrete changes differ: #45115 focuses on inheritance/conversion mapping and modular refactoring, while #45569 introduces separate NemotronHDense/NemotronHSparse implementations, docs, and tests. They do not look like the same underlying fix and would not plausibly merge into one PR." } ] }, @@ -20330,7 +20363,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "328b2aaa3e32b3613bffc06bed7c221f390aae34a3270800a8c0e42515f50da1", + "prepared_review_unit_hash": "848f5af2402577187bda8ec8b2b4751f4334197f21372db0e9ae003da3b090d8", "component_budget": { "node_count": 2, "item_count": 2, @@ -20352,18 +20385,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related but not duplicate PRs: one introduces heterogeneous per-layer config primitives, while the other wires heterogeneous configs into model/cache/masking behavior. They look like consecutive parts of the same initiative, not the same concrete change.", - "confidence": 0.88, + "summary": "Two related but distinct PRs: one introduces heterogeneous per-layer config plumbing, the other adds broader heterogeneous modeling/runtime support on top of it. They share files and theme, but they are not the same concrete change.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "45332 is the broader end-to-end feature PR: it integrates heterogeneous layer handling into cache/modeling/masking paths and includes the more complete user-facing behavior. 45333 is the narrower config-only foundation.", + "canonical_pr_reason": "PR #45332 is the better canonical PR because it is the broader, more complete change: it adds heterogeneous model support across cache, masking, modeling, and tests, and appears to build on the config plumbing introduced separately.", "best_issue_reason": null, - "best_pr_reason": "45332 is the best representative PR for the cluster because it covers the substantive runtime support for heterogeneous models, not just the config plumbing.", + "best_pr_reason": "PR #45332 is the strongest representative of the cluster because it addresses the full heterogeneous-model feature rather than only the configuration layer, making it the more comprehensive standalone change.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "45333 adds per-layer configuration support and helper plumbing, while 45332 adds separate model/cache/masking integration. They share the same initiative and files, but not the same concrete code-path fix/change, so they should not be merged as duplicates." + "reason": "Related feature work, but not the same underlying change. #45333 is focused on per-layer heterogeneous config support, while #45332 extends that into cache/modeling/runtime behavior. They could be separate PRs and are not strict duplicates." } ] }, @@ -20393,7 +20426,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "e0c41e4520cfa97aec46a354914cecec868e814619d1204d71a3a17cc28e41ea", + "prepared_review_unit_hash": "172a4ddc0585ee2628b454c365992ec01ca8fbcd56f08f14fd6f932440bb1cfa", "component_budget": { "node_count": 2, "item_count": 2, @@ -20415,18 +20448,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related only at a broad typing-cleanup level, not the same concrete change. PR 45415 is a wide type-checking sweep across many files, while PR 45425 is a focused typing protocol/update for modeling_utils and related integrations.", - "confidence": 0.88, + "summary": "Both pull requests are typing-related, but they address different changes: PR 45415 is a broad repository-wide type-checking sweep with many ignores, while PR 45425 adds concrete typing protocol/helpers in `_typing.py` and adjusts a few consumers. They should not be merged as duplicates.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45425 is the more focused, concrete change: it adds specific typing protocols and updates the affected code paths around modeling_utils/PEFT integration.", + "canonical_pr_reason": "PR 45425 is the more specific change, centered on shared typing abstractions and their consumers rather than a broad type-checking pass.", "best_issue_reason": null, - "best_pr_reason": "PR 45425 is the best representative because it targets a single typing compatibility problem, whereas 45415 is a broad repo-wide type-checking pass.", + "best_pr_reason": "PR 45425 is the best representative of the cluster because it makes the more concrete, code-level typing change; PR 45415 is mostly a wide lint/type-check configuration sweep.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Different scope and likely different underlying work: 45415 is a broad type-checking sweep, while 45425 is a targeted typing/API compatibility change. Shared files alone are not enough to merge them into one PR." + "reason": "They are related only at a broad typing/annotation level. 45415 is a large type-checking cleanup across many files, while 45425 introduces new typing protocols and utilities for specific consumers; they do not fix the same concrete problem and would not plausibly be one merged PR." } ] }, @@ -20456,12 +20489,12 @@ "soft_pairs": [ "pull_request:45513|pull_request:46084" ], - "prepared_review_unit_hash": "70a07e9b46a353dbfcbb3ed486a2fed087f72cd0c65aee70bab0feda2ea6e85f", + "prepared_review_unit_hash": "a8a453f507e7c867a2ba885d7d0a4a90742fac93115e67bacdb8a93cfffad20d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4858, + "serialized_chars": 4857, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -20469,7 +20502,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4858, + "serialized_chars": 4857, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -20478,18 +20511,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs both touch cached-forward edge cases with seq_len>1, but they target different model families and different code paths: Qwen/Olmo GDN linear attention vs. Mamba2 CUDA kernel forwarding. They are not duplicates.", - "confidence": 0.96, + "summary": "These are two unrelated pull requests fixing cached-forward behavior in different model families: Qwen3.5/GDN linear attention versus Mamba2 use_cache handling. They do not appear to be the same underlying bug or change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46084 is the clearest canonical PR: it has an explicit issue target (#46032), a narrowly scoped fix in one model, and the diff directly addresses the seq_len>1 cache path.", + "canonical_pr_reason": "PR #46084 is the better canonical PR because it is tightly scoped to a single concrete bug, has an explicit linked issue (#46032), and the diff targets one model path with a clear seq_len>1 cache fix.", "best_issue_reason": null, - "best_pr_reason": "PR #46084 is the best representative of a concrete bug fix because it is tightly scoped, explicitly tied to an issue, and focused on a single cache-guard condition in Mamba2.", + "best_pr_reason": "PR #46084 is the best representative of its own bug fix: it is narrowly focused, issue-linked, and more self-contained than the broader multi-model PR #45513.", "soft_edge_verdicts": [ { "left": "pull_request:45513", "right": "pull_request:46084", "accept": false, - "reason": "Both discuss cached multi-token forwarding, but they fix different implementations in different architectures. #45513 is about GDN linear attention in Qwen3.5/Qwen3-Next/Olmo Hybrid, while #46084 is about Mamba2 CUDA-kernel cache handling. The code paths and bugs are not the same, so they should not be merged as duplicates." + "reason": "Different model architectures and different code paths: Qwen3.5 GDN linear attention cached continuation versus Mamba2 cache handling for seq_len>1. They are not plausibly mergeable into one PR for the same underlying bug." } ] }, @@ -20519,7 +20552,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "a71b194d9c58fa1eaf364e01fad75e51038e944a515d3873ba911fb23f2451a1", + "prepared_review_unit_hash": "79d63d464c9b2fb505d74bf845fdc450a68adbbbaa839874a2d645afd0beeb6a", "component_budget": { "node_count": 2, "item_count": 2, @@ -20541,18 +20574,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related EP-loading fixes, but they target different failures and code paths: one is about DeepSpeed ZeRO-3/accelerate launch serialization and detecting EP, the other about FSDP2 rank-0 broadcast overwriting expert parameters during model loading. They are not duplicates.", - "confidence": 0.93, + "summary": "Both pull requests are about expert parallelism in Transformers loading, but they fix different failure modes and should not be merged as duplicates: one is about preserving distributed_config/EP detection for ZeRO-3 loading via accelerate, the other is about preventing expert weights from being overwritten during FSDP2/rank-0 broadcast.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45662 is the broader and more concrete EP fix: it addresses a specific parameter materialization bug during distributed loading and includes the core loader change in addition to supporting EP-related adjustments.", + "canonical_pr_reason": "PR 45662 is the better canonical PR for the cluster because it addresses a concrete correctness bug in expert loading and spans the main affected code paths, whereas PR 45548 is narrower and more infrastructural.", "best_issue_reason": null, - "best_pr_reason": "PR 45662 is the best representative of the cluster because it fixes an explicit, reproducible code-path problem with expert parameters being overwritten during distributed loading; the other PR is a separate EP/DeepSpeed compatibility fix.", + "best_pr_reason": "PR 45662 is the strongest standalone representative of the topic: it fixes a specific silent data corruption issue in EP loading, with clear code-path changes across model loading and MoE handling.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "They share EP-related infrastructure, but the underlying bugs differ: 45548 is about EP + DeepSpeed ZeRO-3 loading via accelerate launch and config serialization, while 45662 fixes FSDP2 rank-0 broadcast overwriting experts during loading. Different failure mode, different fix path." + "reason": "They share the general EP/distributed-loading theme, but the bugs are different: 45548 adds distributed_config serialization and an EP detection property for ZeRO-3/accelerate loading, while 45662 fixes expert parameters being overwritten after sharding/broadcast in FSDP2. Different root causes, different fixes." } ] }, @@ -20582,12 +20615,12 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "c9e9eb7ad4b10cb064783dfbe8d3d89a69cccc03c609eddb0dd6f45837277a42", + "prepared_review_unit_hash": "fd57e2126a655340e20c4b711d24fbaaa96bbd705b627625f2af9f4d889810e9", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4589, + "serialized_chars": 4590, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -20595,7 +20628,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4589, + "serialized_chars": 4590, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -20604,18 +20637,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs address the same xdist collision problem around captured_info logs; PR #45645 is the broader, more complete fix and subsumes the targeted testing_utils change in #45639.", - "confidence": 0.82, + "summary": "Both PRs target the same xdist-safe captured_info logging problem in testing_utils, but 45645 is the broader follow-up that also updates CI log collection to handle the new per-worker files. It looks like 45639 is a narrower subset of 45645.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45645 is the best canonical PR because it includes the core xdist-safe change in src/transformers/testing_utils.py plus the CI/workflow updates needed to consume multiple captured_info files, making it the most complete representation of the fix.", + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the core fix in src/transformers/testing_utils.py plus the CI-side adjustment needed to consume multiple captured_info files, making it the more complete representative of the change.", "best_issue_reason": null, - "best_pr_reason": "PR #45645 is the strongest overall candidate: it covers the same underlying bug as #45639 and adds the surrounding CI handling needed to preserve/debug the logs end-to-end.", + "best_pr_reason": "PR 45645 is the best standalone representative since it addresses the same underlying xdist collision bug and includes the additional workflow handling required for end-to-end correctness. PR 45639 appears to be a narrower partial fix.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Same underlying xdist-safe captured_info logging fix; #45645 clearly subsumes #45639 by including the testing_utils change and adding the CI-side artifact handling for multiple captured_info files." + "reason": "Both PRs fix the same concrete captured_info/xdist collision path in testing_utils. 45645 is a superset that also updates CI log collection, so they are close enough to be treated as the same underlying change cluster." } ] }, @@ -20647,12 +20680,12 @@ "pull_request:45720|pull_request:46259", "pull_request:46023|pull_request:46259" ], - "prepared_review_unit_hash": "521f8f14bb2c7e5ff116ac31a78d029ce2dcfdb06fd1c393e058bac68a80e8b3", + "prepared_review_unit_hash": "65b6e816b85758b2d7298adf478b173553c8fd9421ac59306b8303e6d4c892ac", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10910, + "serialized_chars": 10911, "estimated_input_tokens": 2728, "estimated_eval_tokens": 5712 }, @@ -20660,7 +20693,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 10910, + "serialized_chars": 10911, "estimated_input_tokens": 2728, "estimated_eval_tokens": 5712 }, @@ -20669,24 +20702,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three PRs are separate linter/modeling-rule changes, not duplicates: 45720 targets TRF003 for model forward tuple handling, 46023 targets TRF016 for image/video processor flag consumption, and 46259 targets a different modeling rule (TRF018). The overlap is mostly shared docs/rule infrastructure, not the same underlying fix.", - "confidence": 0.97, + "summary": "All three items are pull requests in the linter/modeling-rules area, but they address different rule implementations/sweeps rather than the same underlying fix. The soft similarity is mostly due to shared docs and some overlapping model files, not a mergeable duplicate change.", + "confidence": 0.83, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR #46259 is the strongest representative of the cluster: it is the broadest modeling-rule change, with a documented rule addition and a substantial sweep across model files. The other PRs are narrower, rule-specific follow-ups.", "best_issue_reason": null, - "best_pr_reason": "46259 is the best standalone representative of the cluster because it is a complete rule-introduction PR with broad rule-doc and model updates, but it is not duplicate-equivalent to the others.", + "best_pr_reason": "PR #46259 is the best single PR to represent this cluster because it has the broadest scope and clearest rule-definition intent. PR #46023 is a narrower fixes-only pass, and PR #45720 targets a different linter rule entirely.", "soft_edge_verdicts": [ { "left": "pull_request:45720", "right": "pull_request:46259", "accept": false, - "reason": "Different linter rules and different code paths: TRF003 for model forward signatures/return tuples vs TRF018 for processor flag usage. Shared files are only broad modeling-rule infrastructure." + "reason": "Different concrete changes: #45720 implements part of TRF003 for model forward tuple/return_dict behavior, while #46259 is about a different modeling-rule sweep for image/video processor boolean flags. Shared linter/doc context is not enough to treat them as the same fix." }, { "left": "pull_request:46023", "right": "pull_request:46259", "accept": false, - "reason": "Both touch modeling_rules.md and rule scaffolding, but they fix different checks: TRF016 for image/video processor do_* flags vs TRF018 for another rule. Not the same concrete change." + "reason": "Both touch modeling-rule docs and model files, but they target different rule work: #46023 is fixes for TRF016/rule 16, while #46259 is a separate rule addition/sweep. The overlap is broad subsystem similarity, not the same concrete bug or change." } ] }, @@ -20785,7 +20818,7 @@ "pull_request:46073|pull_request:46196", "pull_request:46154|pull_request:46196" ], - "prepared_review_unit_hash": "fb83ebda670d8adcb3902ceea5c9034e53e12c74e2f2b40c980a6383212a2fcd", + "prepared_review_unit_hash": "df96d2b9cf84e7bd31e595eedf1ed80a04757e509a50aba395ec51576cf755e9", "component_budget": { "node_count": 3, "item_count": 3, @@ -20807,30 +20840,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three separate model-specific PRs adding SDPA dispatch support, not duplicates of the same concrete change. They share a feature theme, but each targets different code paths and test/docs surfaces.", + "summary": "Three open PRs all add SDPA dispatch support, but they target different model families and code paths (CTRL, T5, and SeamlessM4T/v2 + wav2vec2 variants). They are related in theme, not duplicates of the same underlying change.", "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:46196 is the broadest and most representative of the SDPA rollout, covering multiple model families and associated tests, so it is the best canonical PR if one must be chosen.", + "canonical_pr_reason": "PR 46196 is the broadest and most representative of the cluster, covering multiple related speech/seq2seq model implementations rather than a single model.", "best_issue_reason": null, - "best_pr_reason": "pull_request:46196 is the strongest representative because it spans more than one model implementation and includes the most evidence of a completed feature rollout; the others are narrower, model-specific subsets.", + "best_pr_reason": "PR 46196 is the strongest cluster representative because it spans several models and includes the most comprehensive SDPA dispatch work.", "soft_edge_verdicts": [ { "left": "pull_request:46073", "right": "pull_request:46154", "accept": false, - "reason": "Both are SDPA-dispatch PRs, but they patch different model implementations (CTRL vs T5) with distinct code paths; same feature theme is not enough to treat them as duplicates." + "reason": "Both are SDPA dispatch additions, but they modify different model implementations (CTRL vs T5) and are separate code-path changes, not one mergeable fix." }, { "left": "pull_request:46073", "right": "pull_request:46196", "accept": false, - "reason": "CTRL-only SDPA support and the wav2vec2/SeamlessM4T SDPA support touch unrelated model code paths; they are separate feature PRs, not one underlying fix." + "reason": "These target different families and different attention implementations; the overlap is only the broad SDPA theme, which is too general to treat as duplicates." }, { "left": "pull_request:46154", "right": "pull_request:46196", "accept": false, - "reason": "T5 SDPA dispatch and wav2vec2/SeamlessM4T SDPA dispatch affect different architectures and tests, so they should not be merged as duplicates." + "reason": "T5 SDPA support is a distinct model-specific change from the multi-model SeamlessM4T/wav2vec2 dispatch work, so they should not be merged as the same PR." } ] }, @@ -20860,12 +20893,12 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "252ae9f53a78c89b23bbbd8423e869adc11f3c9f9b79f6a03232490e9af2d0b9", + "prepared_review_unit_hash": "06758fc1430bf820ffd0d0761d9376c5265672b5e67b2b0448fdb48e637aef04", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5571, + "serialized_chars": 5570, "estimated_input_tokens": 1393, "estimated_eval_tokens": 3042 }, @@ -20873,7 +20906,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5571, + "serialized_chars": 5570, "estimated_input_tokens": 1393, "estimated_eval_tokens": 3042 }, @@ -20882,18 +20915,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs overlap on distributed-code files, but they are not the same concrete change: #46126 is a focused EP/Trainer integration patch, while #46269 is a broad distributed branch base with many unrelated edits.", - "confidence": 0.93, + "summary": "Two open PRs touch overlapping distributed/TP files, but they are different scopes: one is a focused EP + trainer integration, the other is a broad distributed branch-base umbrella touching many unrelated areas. They are not the same underlying change.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "#46126 is the cleaner representative of the cluster because it is a narrow, self-contained feature patch with a specific EP/Trainer integration goal; #46269 is a much broader umbrella branch touching many unrelated distributed areas.", + "canonical_pr_reason": "PR #46126 is the more focused and concrete change, centered on EP/trainer integration and the specific distributed code paths it affects. PR #46269 is much broader and looks like a branch-base umbrella rather than a single duplicate fix.", "best_issue_reason": null, - "best_pr_reason": "#46126 is the best candidate for a single concrete change because its scope is focused and internally coherent, whereas #46269 is too broad to serve as a good duplicate/merge target for the same underlying change.", + "best_pr_reason": "PR #46126 is the best representative artifact for this cluster because it has a clear, narrow purpose and the overlap with the other PR is confined to shared distributed files rather than the same full change.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Shared distributed filenames are not enough here; the titles and diffs show different scopes. #46126 targets EP + Trainer integration, while #46269 is a large distributed branch base spanning FSDP, MoE, loading, generation, CI, and more. They do not look mergeable as one PR for the same bug/change." + "reason": "Shared distributed filenames are not enough here: #46126 is a targeted EP/trainer integration PR, while #46269 is a broad distributed branch-base PR spanning many extra modules and CI changes. They do not appear to fix the same concrete code-path problem and would not plausibly merge into one PR." } ] }, @@ -20923,12 +20956,12 @@ "soft_pairs": [ "pull_request:46152|pull_request:46207" ], - "prepared_review_unit_hash": "cea62eff1047af9d67d781aa39ebb59d3fd0b0800528e99511633b22307c7fac", + "prepared_review_unit_hash": "a2aa2a995cb0b786e400108dc47953484e1320b4979426436316e9660187e833", "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 }, @@ -20936,7 +20969,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4276, + "serialized_chars": 4275, "estimated_input_tokens": 1069, "estimated_eval_tokens": 2394 }, @@ -20945,18 +20978,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs share the broad theme of replacing direct file-open source inspection, but they fix different code paths: one in `import_utils.py` for resolving package resources, the other in `modeling_utils.py` for attention-implementation support checks. They are related conceptually, but not the same underlying change.", + "summary": "Both PRs address replacing file-based source inspection with a safer source-access mechanism, but they touch different code paths and are not the same mergeable fix.", "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46207 is the cleaner representative of its own change: it makes the `_can_set_attn_implementation` source check safer and includes tests, while staying focused on one concrete behavior in `modeling_utils.py`.", + "canonical_pr_reason": "PR 46207 is the stronger representative because it targets a concrete source-reading helper in modeling_utils and includes tests, making the change more self-contained and clearly validated.", "best_issue_reason": null, - "best_pr_reason": "PR #46207 is the best fit for the cluster as a representative PR because its intent is explicit, scoped to a single code path, and backed by tests; PR #46152 is a separate infrastructure change in a different module.", + "best_pr_reason": "PR 46207 is the best standalone PR candidate since it fixes a specific implementation path with test coverage; PR 46152 is related in theme but addresses a different helper in import_utils.", "soft_edge_verdicts": [ { "left": "pull_request:46152", "right": "pull_request:46207", "accept": false, - "reason": "They both avoid `open()` for source-related logic, but they modify different modules and solve different problems: resource traversal in `import_utils.py` versus attention-implementation detection in `modeling_utils.py`. Shared theme is not enough to treat them as the same PR." + "reason": "Shared theme only: both avoid open()-based source reading, but one is in import_utils and the other in modeling_utils, with different functions and no clear same underlying bug or single mergeable code-path fix." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index eb17c38b2ad89c41eba0bc4a31958028bda220b2..4d8a40229406075614c0af19d39ef871f187c964 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260531t060031z", + "analysis_id": "hybrid-model-20260531t120035z", "archived_artifacts": { - "hybrid": "snapshots/20260531T060031Z/analysis-runs/hybrid-model-20260531t060031z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260531T060031Z/analysis-runs/hybrid-model-20260531t060031z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-05-31T06:08:08Z", + "published_at": "2026-05-31T12:08:30Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260531T060031Z", + "snapshot_id": "20260531T120035Z", "variant": "hybrid" } diff --git a/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..9a00bebd6f4931de9793a7ab5797c043214097ce --- /dev/null +++ b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json @@ -0,0 +1,1997 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260531T120035Z", + "generated_at": "2026-05-31T12:08:28Z", + "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-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." + } + ], + "best_issue": { + "cluster_id": "cluster-45647-3", + "issue_number": 45647, + "reason": "Issue #45647 is the strongest global issue candidate because it is open, belongs to a cluster with 3 artifacts, and carries 13 discussion comments plus 8 inbound references.", + "score": 164.0 + }, + "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/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..d4e25408a4254faad2e57a331a89b8a99f88004d --- /dev/null +++ b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,21004 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260531T120035Z", + "generated_at": "2026-05-31T12:08:28Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 1, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43296", + "issue:43425", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44589", + "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" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:45507", + "issue:44292|issue:45237", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue: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:44805|issue:46294", + "issue:43596|issue:44805", + "issue:44805|issue:45161", + "issue:44560|issue:46165", + "issue:43596|issue:46165", + "issue:43572|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43975|issue:44457", + "issue:43039|issue:45161", + "issue:44292|issue:44457", + "issue:24643|issue:30277", + "issue:34928|issue:44805", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45874|issue:46294", + "issue:39290|issue:45081", + "issue:45823|issue:45874", + "issue:43425|issue:44292", + "issue:34928|issue:41720", + "issue:43828|issue:45161", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:45685", + "issue:44918|issue:45161", + "issue:43716|issue:44292", + "issue:44292|issue:44918", + "issue:45507|issue:45685", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43716|issue:44560", + "issue:43828|issue:45685", + "issue:44918|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:34928|issue:36010", + "issue:45237|issue:46015", + "issue:43039|issue:44805", + "issue:45800|issue:46304", + "issue:43296|issue:46304", + "issue:39290|issue:41720", + "issue:41762|issue:43596", + "issue:43901|issue:44512", + "issue:44589|issue:46304", + "issue:45823|issue:46294", + "issue:45237|issue:45901" + ], + "prepared_review_unit_hash": "0b998587bd77bb9943d0978f3fced78f0b8e4663d5802295c37268c7f034d96f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 67, + "serialized_chars": 47821, + "estimated_input_tokens": 11956, + "estimated_eval_tokens": 24168 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose thematic group of unrelated bug reports; only the two BERT + DeepSpeed ZeRO-3 index-out-of-bounds reports look like true duplicates.", + "confidence": 0.58, + "canonical_issue_reason": "issue:43638 is the clearest representative of the only true duplicate subgroup here: BERT + DeepSpeed ZeRO-3 failing with an index-0-out-of-bounds crash during model init/training.", + "canonical_pr_reason": null, + "best_issue_reason": "There is no strong global canonical for the whole cluster because the items span many unrelated models, configs, and docs issues; issue:43638 is just the best anchor for the one genuinely duplicated bug family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failure modes: missing Qwen3.5 config attribute vs streaming chat/completions API error." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Unrelated hardware/backend problems: AMD GPU support vs MPS histogram kernel on Int." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different models and error paths; image preprocessor dtype mismatch is unrelated to AMD GPU execution failure." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Autocast dtype mismatch on Phi MoE is a different bug than GPT-OSS failing on AMD GPUs." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different model families and symptoms; AMD GPU issue vs invalid method call in GraniteMoEHybrid." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error is a different code path from AMD GPU incompatibility." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed ZeRO-3 collective mismatch is unrelated to ModernBERT ONNX export failure." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export failure and FineGrainedFP8Config failure are different model/backend issues." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Export-to-ONNX problem is unrelated to Qwen3 auto device mapping cudaErrorAssert." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Task crash vs outdated RAG docs; same general area, not the same bug or change." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs cleanup items, but they target different removed features and would not be one merged fix." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ModernBERT ONNX export and GenerationMixin import regression are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "TRL SFT embedding unpacking vs resume-training with trust_remote_code are different training failures." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Mask shape mismatch during indexing is unrelated to Gemma4 PLE device placement mismatch." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both involve indexing errors, but the reported symptoms and contexts differ too much to be confidently the same bug." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 BERT crash is unrelated to resuming training with trust_remote_code." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Mask/index shape mismatch is unrelated to Qwen3.5 video feature/token mismatch." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Similar generic indexing symptom, but different trigger and message; not enough to treat as one bug." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Indexing/mask mismatch is unrelated to GPT-OSS tensor-parallel-only failure." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Video chat-template StopIteration is unrelated to training resume behavior." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 initialization crash is unrelated to local model resume/trust_remote_code." + }, + { + "left": "issue:43572", + "right": "issue:46165", + "accept": false, + "reason": "Missing pad_token_idx after v5 update is unrelated to training resume failure." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "BarkProcessor voice_preset bug and ModernBERT export bug are unrelated." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Both touch DeepSpeed, but one is collective mismatch in Jamba-v01 and the other is BERT index-out-of-bounds; different bugs." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Near-identical ZeRO-3 BERT index-out-of-bounds reports; same underlying initialization/training bug and plausible single fix." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Different model-specific tokenizer/ detokenization issue vs LoRA merge/save/reload mismatch." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross_entropy invocation is unrelated to GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Qwen-3 NVFP4 runtime error is unrelated to LoRA merge/save behavior." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "Different DeepSpeed-related failures: 2-D weight error vs collective mismatch." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpointing/FSDP tensor recomputation bug is unrelated to mask shape mismatch." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": false, + "reason": "CPU offload incompatibility in rotary embedding vs uninitialized memory after materialization are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "BERT ZeRO-3 index crash is unrelated to GPT-OSS only-TP failure." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "FSDP/activation checkpointing shape issue is unrelated to trainer custom compute_loss signature change." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 BERT init crash is unrelated to GPT-OSS tensor-parallel-only failure." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Mistral-3 image dtype mismatch is unrelated to GraniteMoEHybrid invalid method call." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Different Gemma4 input/embed behavior vs Qwen3.5 video template mismatch." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3/vLLM config attribute error is unrelated to tokenizer regex patch crash." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Device-map forward mismatch and expensive reverse embedding lookup are different Gemma4 issues." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Torch 2.10 compatibility report is unrelated to Qwen-3 NVFP4 runtime failure." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP checkpointing mismatch is unrelated to Qwen3 auto device mapping assertion failure." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Phi autocast dtype mismatch is unrelated to GPT-OSS tensor-parallel-only failure." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRL SFT embedding issue is unrelated to GPT-OSS tensor-parallel-only failure." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different models and symptoms: Mistral image dtype mismatch vs Qwen NVFP4 runtime error." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Qwen NVFP4 runtime error is unrelated to TRL SFT embedding unpacking." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Invalid method call in GraniteMoEHybrid is unrelated to MPS histogram kernel limitation." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger Kernel cross_entropy being called is unrelated to TRL SFT embedding unpacking." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": false, + "reason": "Both are ZeRO-3 load/init failures, but the models and reported symptoms differ enough that they don't look like the same bug." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs regressions, but they concern different removed features and likely separate cleanup edits." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Mistral image dtype mismatch is unrelated to Qwen3-vl video StopIteration." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch is unrelated to MPS histogram kernel support." + }, + { + "left": "issue:44918", + "right": "issue:45685", + "accept": false, + "reason": "TRl SFT embedding unpacking is unrelated to MPS histogram kernel support." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Liger Kernel cross_entropy invocation is unrelated to Phi autocast dtype mismatch." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "DeepSpeed collective mismatch and trainer compute_loss signature mismatch are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "FSDP activation checkpointing issue is unrelated to GenerationMixin import regression." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "AMD GPU runtime failure is unrelated to outdated RAG docs." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger Kernel cross_entropy call is unrelated to mask shape mismatch." + }, + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Version incompatibility report and USE_DISTRIBUTED=0 support regression are different compatibility problems." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "PaddleOCR-VL/vLLM load failure is unrelated to pytorch distributed-disabled support breakage." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Gemma3/vLLM config attribute error is unrelated to Qwen3 auto device mapping assertion failure." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Both are ZeRO-3 index-out-of-bounds reports, but they involve different models and likely separate load paths." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "TextClassificationPipeline docs wording and run-command docs removal are separate documentation issues." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Float8 storage lookup failure is unrelated to distributed-disabled support breakage." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Gemma4 device-map mismatch and Qwen3.5 video token mismatch are unrelated." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "AMD GPU failure is unrelated to table-question-answering crash." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: this is mostly a broad thematic cluster with no single global canonical, and the only clearly duplicate subgroup is the ZeRO-3 BERT index-out-of-bounds pair (issue:43596 / issue:43638). The soft-edge verdicts are generally conservative and avoid overstating similarity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 2, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36010", + "issue:39290", + "issue:41093", + "issue:41762", + "issue:43039", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43792", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43975", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44479", + "issue:44560", + "issue:44589", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45161", + "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:44841|issue:45823", + "issue:43039|issue:44560", + "issue:43975|issue:46015", + "issue:45812|issue:45920", + "issue:44863|issue:46015", + "issue:45020|issue:46304", + "issue:43866|issue:44863", + "issue:44292|issue:46015", + "issue:43872|issue:46304", + "issue:39290|issue:43531", + "issue:43854|issue:43866", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43425|issue:45070", + "issue:43596|issue:45800", + "issue:43975|issue:45901", + "issue:43792|issue:45070", + "issue:45362|issue:45464", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:43404|issue:43866", + "issue:45464|issue:45685", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:44841|issue:45084", + "issue:36010|issue:39290", + "issue:45800|issue:45823", + "issue:41093|issue:41762", + "issue:43296|issue:43366", + "issue:43572|issue:45070", + "issue:43541|issue:45161", + "issue:44387|issue:45005", + "issue:44805|issue:44918", + "issue:43828|issue:45464", + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:43716|issue:46294", + "issue:43531|issue:45070", + "issue:45020|issue:45800", + "issue:44479|issue:46294", + "issue:43541|issue:44589", + "issue:41093|issue:43596" + ], + "prepared_review_unit_hash": "1db38b5924f5634c69bf9409335b40fb936f1b76f9e8015d8514ce9b90f0dd90", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 42, + "serialized_chars": 39559, + "estimated_input_tokens": 9890, + "estimated_eval_tokens": 20036 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are unrelated model-specific regressions or docs/config issues. I only found two clear duplicate pairs: the v5 AutoTokenizer wrong-token-ID regression and the Qwen3.5 video/chat-template regression.", + "confidence": 0.78, + "canonical_issue_reason": "issue:45920 is the best representative of the duplicate tokenizer-regression subgroup: it is broader than the Granite-specific report, remains open, and describes the same v5 wrong-token-ID failure across multiple model families.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the strongest global issue anchor among the accepted duplicates because it captures the widest version-wide tokenizer regression with multiple affected models.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Different model families and different failures: processor loading vs device mismatch during forward." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Unrelated subsystems: Liger kernel cross-entropy path vs Qwen video StopIteration." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Different problems: detokenization bug vs RAG example documentation freshness." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Both report the same v5 AutoTokenizer regression producing wrong token IDs across model families; 45920 is the broader tracker." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "Checkpoint-loading bug vs outdated RAG examples; not the same underlying issue." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Broad remote_code breakage vs a specific USE_DISTRIBUTED=0/pytorch compatibility regression." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Both are load failures, but for different models and different root causes." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Model-loading error vs documentation issue; no shared code-path bug." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes Int8Params constructor change is unrelated to distributed build support." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but one is a missing config attribute in vLLM and the other is a model-specific sliding_window issue." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to the Gemma3 sliding_window_pattern problem." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Different models and different load failures; no evidence of the same bug." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both are IndexErrors, but the reported shapes and contexts differ; not enough to treat as the same bug." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper model failure vs table-question-answering crash; unrelated." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch version incompatibility is not the same as the PretrainedConfig pydantic-field regression." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "Different failures: ZeRO-3/Bert index error vs torch 2.4.1 and transformers 5.8.0 incompatibility." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Unrelated features and error paths." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Different model/runtime bug vs config-model schema break." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Both involve Qwen3.5, but the reported failures are different enough that they do not look like one concrete bug." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 loading error is unrelated to table QA crashes." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Gemma3 sliding_window_pattern issue and Gemma4 device mismatch are different bugs." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "Weight-tying bug in Mistral3 is unrelated to a corrupted Ovis2 checkpoint." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Chat/completions streaming failure vs MPS histogram_mps limitation; different code paths." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Torch dynamo grouped_mm tracing error is unrelated to the MPS histogram issue." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Both are MoE/dtype-adjacent, but the concrete failures differ and do not look like one bug." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Processor loading failure and template-node compilation error are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "Same as above: import error vs config attribute regression." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "General version incompatibility vs a specific device-map forward mismatch." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are IndexErrors, but one is a mask/tensor shape mismatch and the other is a ZeRO-3 zero-size index failure." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are separate issues." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Missing StableLmConfig pad_token_idx and PretrainedConfig pydantic break are distinct config regressions." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Grouped_mm tracing error vs TP-only GPT-OSS MoE support issue; not the same problem." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "Memory/OOM regression under int4 quantization vs tied-weights translation model bug." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Mask-shape IndexError vs SFT trainer embedding-unpacking failure; different bugs." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to Qwen3.5 streaming chat failure." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument and pydantic PretrainedConfig break are different load/config issues." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "gpt-oss support and Qwen3.5 chat streaming failure are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal regressions: image-preprocessor dtype mismatch vs video token/feature mismatch." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Sliding_window issue in Qwen3-MoE is not the same as the PretrainedConfig schema break." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote_code breakage and torch 2.4.1 compatibility problems are different versions of different failures." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": true, + "reason": "Both describe the same Qwen3.5 video-input regression where apply_chat_template/video tokens and features no longer align in v5.3.x." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Torch dynamo grouped_mm tracing error is unrelated to the Float8 storage lookup failure." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Different IndexError signatures and contexts; not the same underlying bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The two accepted soft pairs are well supported by the packet: 45812/45920 are the same v5 AutoTokenizer wrong-token-ID regression, and 44479/46294 are the same Qwen3.5 video/apply_chat_template mismatch. The negative verdicts also stay cautious and do not overclaim shared root causes from superficial similarity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 3, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36331", + "issue:38175", + "issue:39186", + "issue:39692", + "issue:40990", + "issue:41553", + "issue:41762", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43404", + "issue:43644", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43866", + "issue:43872", + "issue:43873", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:43994", + "issue:44292", + "issue:44360", + "issue:44387", + "issue:44403", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:44960", + "issue:45005", + "issue:45399", + "issue:45507", + "issue:45542", + "issue:45632", + "issue:45663", + "issue:45698", + "issue:45701", + "issue:45759", + "issue:45800", + "issue:45859", + "issue:45902" + ], + "soft_pairs": [ + "issue:43054|issue:43994", + "issue:45759|issue:45902", + "issue:45542|issue:45800", + "issue:43716|issue:44918", + "issue:39186|issue:44918", + "issue:45632|issue:45698", + "issue:43645|issue:45399", + "issue:43404|issue:45005", + "issue:40990|issue:44960", + "issue:44530|issue:45005", + "issue:42915|issue:45399", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:44863|issue:45005", + "issue:44918|issue:45507", + "issue:44492|issue:44512", + "issue:43975|issue:44292", + "issue:44292|issue:45507", + "issue:43950|issue:44534", + "issue:39692|issue:43054", + "issue:43644|issue:43950", + "issue:43828|issue:44928", + "issue:36331|issue:41762", + "issue:43950|issue:45902", + "issue:43872|issue:44589", + "issue:43866|issue:44403", + "issue:38175|issue:43054", + "issue:41553|issue:43883", + "issue:43054|issue:45701", + "issue:45663|issue:45902", + "issue:44910|issue:45859", + "issue:42947|issue:44387" + ], + "prepared_review_unit_hash": "293895915564a67f82a9e16aff9bf79c5e8cb58299cc160a124c3c358e99e5a8", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37144, + "estimated_input_tokens": 9286, + "estimated_eval_tokens": 18828 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated Transformers issues with only one clear duplicate thread: non-persistent buffers being corrupted/junk-filled during `from_pretrained` in v5.", + "confidence": 0.92, + "canonical_issue_reason": "issue:43950 is the clearest canonical issue: it states the regression precisely (`from_pretrained()` corrupts non-persistent buffers) and matches the closest duplicate reports about the same junk-filled buffer behavior.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43950 is the best representative issue because it is specific, regression-focused, and central to the only confirmed duplicate pair in this set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Both are about SigLIP2, but one is about worse text embeddings and the other about nonsensical pipeline/AutoModel results; the root cause is not clearly the same." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Both involve Qwen3.5/Qwen 3.6 internals, but one is about `text_config` unwrapping in `from_config` and the other about uninitialized rotary frequencies after materialization; different bugs." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "Different problems: TensorBoard/TensorFlow backend detection versus a torch/transformers version incompatibility." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Different model families and failure modes: image preprocessor dtype mismatch versus Qwen3.5 embedding unpacking with TRL SFT trainer." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP `'weight' must be 2-D` and TRL embedding unpacking are unrelated failures." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both concern loading custom code, but one is a trust_remote_code cache-path collision and the other is loading the wrong custom module after save/load; too different to merge." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Notebook custom-model initialization and flash-attn fallback gating are unrelated code paths." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a specific Mistral3 lm_head tying bug while the other is a broader translation-model v5 issue; not the same concrete bug." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "Perplexity on gpt-oss-20b and a GLM5 issue are unrelated." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache/Qwen3.5 crash is a separate cache-kernel issue, not the tied-weights regression." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn fallback checks." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantized offloading behavior and tied-weights issues are different bugs." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA ReLU discussion and a docs typo are unrelated." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading and tied weights on translation models are different issues." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "Qwen3.5 embedding unpacking with TRL and GraniteMoEHybrid invalid method calls are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A cache-strategy typo and a docs command removal note are not the same bug." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Incorrect detokenization and Qwen-3-8B-NVFP4 runtime errors are different failures." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "Different models and different failure modes." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same underlying regression: v5 `from_pretrained` corrupts/fills non-persistent buffers with junk." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "SigLIP2 docs example errors versus general poor SigLIP2 text embeddings; not the same concrete issue." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Direct duplicate of the non-persistent buffer corruption/junk-filled buffer regression." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Both mention dtype/BF16 behavior, but the actual failures are different model/training paths." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "CustomTrainer keyword-argument mismatch and DeepSpeed ZeRO-3 load-time index error are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "Both touch buffer/materialization behavior, but one is buffer corruption in `from_pretrained` and the other is uninitialized rotary frequency memory after `to_empty`; not the same bug." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Different quantization/storage incompatibilities: bitsandbytes `Int8Params` vs missing Float8 storage." + }, + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "Checkpoint corruption and loading noise are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Both concern SigLIP2, but the reported symptoms differ and no shared concrete bug is evident." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad AutoTokenizer error messaging and missing `all_tied_weights_keys` are unrelated." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "SigLIP2 embedding quality and tokenization changes across transformers versions are different issues." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states bug and Qwen3.5 rotary buffer materialization bug are unrelated." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Qwen3.5 flash-attention position_ids crash and CPU-offload incompatibility in rotary embedding are different code paths." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Gradient checkpointing/LoRA behavior and int4 quantization memory growth are different problems." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet. The accepted duplicate thread around non-persistent buffer corruption in `from_pretrained` is supported by the issue titles and the soft-pair evidence, and the other soft-edge rejections read conservative rather than overstated. Minor note: the duplicate cluster is really a three-issue thread (`43644`, `43950`, `44534`), not just a pair, but the overall conclusion is still correct." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 4, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36331", + "issue:41093", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43257", + "issue:43404", + "issue:43493", + "issue:43643", + "issue:43742", + "issue:43782", + "issue:43792", + "issue:43824", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43940", + "issue:43950", + "issue:43975", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44509", + "issue:44661", + "issue:44863", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45305", + "issue:45313", + "issue:45356", + "issue:45399", + "issue:45520", + "issue:45685", + "issue:45823", + "issue:45923" + ], + "soft_pairs": [ + "issue:41553|issue:43881", + "issue:44403|issue:44451", + "issue:43975|issue:45520", + "issue:43054|issue:43493", + "issue:44403|issue:44488", + "issue:42947|issue:43856", + "issue:43643|issue:45020", + "issue:41762|issue:42915", + "issue:45685|issue:45823", + "issue:42491|issue:44368", + "issue:43881|issue:43883", + "issue:44991|issue:45399", + "issue:43824|issue:43872", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:45081|issue:45399", + "issue:43782|issue:44960", + "issue:45020|issue:45356", + "issue:43742|issue:45399", + "issue:43404|issue:43950", + "issue:43827|issue:44509", + "issue:43792|issue:44220", + "issue:43872|issue:44291", + "issue:43257|issue:43866", + "issue:42915|issue:43872", + "issue:42947|issue:45923", + "issue:43940|issue:45313", + "issue:36331|issue:41093", + "issue:41553|issue:44991", + "issue:43854|issue:44863", + "issue:43257|issue:45305" + ], + "prepared_review_unit_hash": "0431bc28b08722a79abfa998fa381e1722cc01cfba4a64123ec7efd8414d380f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37434, + "estimated_input_tokens": 9359, + "estimated_eval_tokens": 18974 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous closed issues with only a few clear duplicate subclusters. The strongest matches are the v5 docs/pipeline wording issue, the bitsandbytes `_is_hf_initialized` incompatibility, and a likely SigLIP2 implementation discrepancy; the rest are too model- or symptom-specific to merge safely.", + "confidence": 0.78, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:43872 is the clearest anchor because it names a concrete, reproducible `_is_hf_initialized` / Int8Params incompatibility and has a very strong duplicate counterpart in 44291.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different models and failure modes; one is Voxtral tokenizer error messaging, the other is a GLM-4v load failure." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic loading noise vs a specific ScandiBERT load failure; not the same underlying bug." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer detokenization regression vs Python 3.13 flash_attn import-key error; unrelated code paths." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": true, + "reason": "Both point at the same SigLIP2 text-embedding discrepancy/implementation mismatch relative to expected behavior." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Both mention loading, but the concrete issues differ; this is too broad to treat as a duplicate." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "LoRA gradient checkpointing behavior is unrelated to Qwen3 MoE training memory usage." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "One is missing fields in AutoConfig with trust_remote_code, the other is a broader remote_code regression; same area but not the same bug." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Different model families and different loading backends; no concrete shared failure path." + }, + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "MPS histogram kernel limitation is unrelated to a Gemma4 device-mismatch forward bug." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "Both involve LoRA/training, but one is a hf4.x\u2192hf5.x Qwen3 MoE incompatibility and the other is a tie_word_embeddings warning." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different models and different errors; Molmo tie-weights attribute issue does not match a GLM-4v load failure." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading for est-roberta is unrelated to flash-attn fallback gating." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "ImportError for Qwen2.5-VL is unrelated to the bitsandbytes `_is_hf_initialized` constructor issue." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Broad remote_code breakage does not match a specific Gemma4 PLE device placement bug." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "add-new-model-like / tokenizer mapping logic is a different path from remote_code loading regressions." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE is not the same as Gemma4 device mismatch." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Mistral tokenizer regex patch crash is unrelated to flash-attn2 fallback checks." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight_only loading error and GLM5 are unrelated model-specific failures." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same underlying `_is_hf_initialized` incompatibility during model init/loading with bitsandbytes/init_empty_weights." + }, + { + "left": "issue:43257", + "right": "issue:43866", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to an Ovis2 checkpoint corruption report." + }, + { + "left": "issue:42915", + "right": "issue:43872", + "accept": false, + "reason": "FineGrainedFP8Config failure and bitsandbytes `_is_hf_initialized` incompatibility are different bugs." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "Gradient checkpointing behavior is unrelated to a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are ZeRO-3 weight-loading failures, but the model variants and missing-weight patterns are not specific enough to assert the same root cause." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unexpected compute_loss kwarg error and tensor mask shape mismatch are unrelated." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer/model loading issues with no shared concrete failure path." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "GLM-4.7-Flash unit-test load failure is unrelated to NemotronH checkpoint loading." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Remote_code regressions and Kimi-K2.5 tokenizer codec/fix_mistral_regex problems are different issues." + }, + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "MobileLLM key error and flash-attn fallback gating are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 is unrelated to non-persistent buffer corruption in from_pretrained." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both report the same v5 docs regression: stale references to removed pipeline tasks." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Whisper cannot run is too broad to safely equate with a specific _torch_extract_fbank_features issue." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same `_is_hf_initialized` incompatibility manifesting through different loading entry points." + }, + { + "left": "issue:43257", + "right": "issue:43866", + "accept": false, + "reason": "Different models and different failure classes; no shared root cause." + }, + { + "left": "issue:42915", + "right": "issue:43872", + "accept": false, + "reason": "Qwen3 MoE FP8 training failure is not the same as the bitsandbytes/init_empty_weights constructor regression." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative about weaker links. The accepted soft-edge pairs are well supported by the issue titles/body themes, and the rejected pairs are cautious rather than overstated. No mergeability or same-bugfix claim appears inflated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 5, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36010", + "issue:38175", + "issue:39692", + "issue:41093", + "issue:41553", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43317", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43645", + "issue:43646", + "issue:43720", + "issue:43824", + "issue:43828", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44479", + "issue:44488", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:45030", + "issue:45325", + "issue:45584", + "issue:45588", + "issue:45663", + "issue:45685", + "issue:45823", + "issue:45834", + "issue:45902" + ], + "soft_pairs": [ + "issue:44928|issue:45588", + "issue:41093|issue:44805", + "issue:44479|issue:44560", + "issue:41553|issue:44488", + "issue:44530|issue:44863", + "issue:24643|issue:30064", + "issue:45325|issue:45902", + "issue:43366|issue:43828", + "issue:45663|issue:45823", + "issue:43366|issue:43531", + "issue:43950|issue:45834", + "issue:43550|issue:44910", + "issue:43645|issue:43646", + "issue:44560|issue:45685", + "issue:43366|issue:44589", + "issue:42915|issue:43957", + "issue:44560|issue:44918", + "issue:43824|issue:43957", + "issue:42617|issue:43366", + "issue:36010|issue:42915", + "issue:44661|issue:45823", + "issue:43632|issue:45834", + "issue:38175|issue:43994", + "issue:44155|issue:45584", + "issue:44560|issue:44805", + "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:42491|issue:43720" + ], + "prepared_review_unit_hash": "a0420541163aee96969bd7ccb5e627e301c2e99e713f13480fb45f92ca8648f7", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37486, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the soft pairs span different models, symptoms, and code paths. Only the two Transformers 5.0 custom model initialization issues look like the same underlying bug.", + "confidence": 0.93, + "canonical_issue_reason": "issue:43646 is the broader, cleaner description of the same Transformers 5.0 custom-model initialization regression; 43645 is the notebook-specific variant.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best representative because it captures the core regression without the extra Jupyter-specific framing.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Both mention Qwen3.5/3D position_ids, but one is a flash-attention illegal-access crash and the other is SDPA/BF16 training instability; different concrete bugs and likely different fixes." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": false, + "reason": "Same generic IndexError wording, but no evidence they share the same failing code path or model; too little to treat as a duplicate." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-related Qwen regressions, but one is a broad v5.3.0 input regression and the other is a StopIteration in Qwen3-vl-embedding; not clearly the same bug." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Both concern loading/tokenizer errors, but they involve different models and different failure modes; not the same underlying issue." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different model families and different loader/cache failures; no shared concrete bug." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Completely different domains: DeepSpeed training weight shape error vs void segmentation map image-processor failure." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Both are low-level model internals, but one is Qwen2.5-VL rope scaling and the other is Qwen3.5-MoE uninitialized memory after meta materialization; unrelated fixes." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF/gpt-oss support and a Phi-tiny-MoE autocast dtype mismatch are unrelated." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Same Gemma4 family, but one is an FSDP2 shared_kv_states KeyError and the other is a device_map forward mismatch; different code paths." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "No overlap beyond broad Qwen/GGUF context; unrelated issues." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer loading regression vs Kosmos2.5 OCR index error; unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different models and different attention paths; one is Bamba SDPA/compile, the other is Qwen3.5 Flash Attention with 3D position_ids." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "These describe the same Transformers 5.0 regression breaking custom model initialization; the notebook mention in 43645 is just a usage context, not a distinct bug." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Video inference StopIteration and an MPS histogram op-not-implemented error are unrelated." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "GGUF model support and a Float8 storage lookup error are different problems." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Qwen3Moe FP8 config failure vs meta-device loading issues; different failure modes and likely different fixes." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Both involve Qwen3.5/video-ish workflows, but one is a video StopIteration and the other is embedding unpacking with TRL SFT; not the same bug." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Missing class import and meta-device loading regression are unrelated." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py runtime failure and GGUF gpt-oss support are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "GenerationMixin import error vs Qwen3Moe FP8 failure; unrelated." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Tokenizer mapping metadata bug and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "_is_hf_initialized flag regression and Kosmos2.5 OCR index error are different bugs." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": false, + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is a model/pipeline nonsense-results report; not enough evidence of the same root cause." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "AudioFlamingo3 batched embedding/token leak vs Whisper empty-transcription generation failure; different models and mechanisms." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "StopIteration in Qwen3-vl-embedding video input vs a generic mask/tensor shape mismatch; not the same failure." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Strict config loading for granite_speech and tiny-random glm4v config validation are separate config issues." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "DeepSpeed zero3/BertModel indexing error and tokenizer-mapping-related model addition failure are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Both touch initialization/loading, but one is a zero3 Bert index error and the other is add-new-model-like metadata handling; different bugs." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer mapping metadata bug vs generic index-mask mismatch; unrelated." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "device_map/offload loading failure and Strix Halo mmap OOM are different memory-management issues." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Both involve SigLIP2, but 39692 is a docs/example problem with model-processor mismatch and quantization failure, not the same as zero probabilities." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3-MoE LoRA compatibility across versions and BitNet packed-weight loading are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet and is appropriately conservative. The only accepted pair, issue:43645 / issue:43646, is well supported by the titles and descriptions as the same Transformers 5.0 custom-model initialization regression, with 43645 appearing as a notebook-specific variant. The other pairwise rejections are cautious and do not overclaim shared root causes." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 6, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44756", + "issue:44910", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45276", + "issue:45292", + "issue:45313", + "issue:45335", + "issue:45357", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45859" + ], + "soft_pairs": [ + "issue:44960|issue:44977", + "issue:41720|issue:42915", + "issue:45313|issue:45357", + "issue:44451|issue:44488", + "issue:42491|issue:45276", + "issue:43708|issue:44075", + "issue:43644|issue:44534", + "issue:42915|issue:44756", + "issue:45488|issue:45812", + "issue:39692|issue:43994", + "issue:43854|issue:43931", + "issue:43257|issue:43854", + "issue:43856|issue:43873", + "issue:42491|issue:43931", + "issue:43644|issue:43645", + "issue:43856|issue:44387", + "issue:42491|issue:45292", + "issue:44910|issue:45588", + "issue:44488|issue:44991", + "issue:45362|issue:45406", + "issue:44164|issue:44291", + "issue:45276|issue:45335", + "issue:43404|issue:43454", + "issue:36010|issue:39404", + "issue:39692|issue:44960", + "issue:43425|issue:43957", + "issue:42947|issue:45704", + "issue:44402|issue:44488", + "issue:45663|issue:45759", + "issue:45663|issue:45859", + "issue:43873|issue:44387" + ], + "prepared_review_unit_hash": "5127f3558a92c098a6e315a39557e1ee247013631ce56d92d1bc8ce4fbf89f46", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36443, + "estimated_input_tokens": 9111, + "estimated_eval_tokens": 18478 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. The only strong duplicate pair is 43644/44534 about non-persistent buffers being filled with junk. The Gemma/T5Gemma resize_token_embeddings reports are similar in theme but are not a true duplicate pair here because they affect different model families and different submodules.", + "confidence": 0.79, + "canonical_issue_reason": "Issue 43644 is the best canonical issue because it is the clearest, broader regression report and it has the strongest confirmed duplicate match in 44534.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43644 is the best representative issue for the cluster: it captures the only clearly duplicated bug and is more general than the other model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Same broad area (generation) but different models and different failure modes." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "CUDA auto-device-map assert and LoRA version-compatibility are unrelated bugs." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5 reports, but one is ZeRO-3 loading and the other is save_pretrained key serialization." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are load failures, but they concern different models and no shared concrete code path is evident." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Different bug classes: qwen3_moe hf4.x/hf5.x incompatibility vs Gemma4 resize_token_embeddings propagation." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint math and SGD argument handling are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Near-identical reports of the same non-persistent-buffer junk regression in Transformers 5." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE loading/version issue is unrelated to Strix Halo mmap/OOM behavior." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both involve tokenization, but they hit different tokenizer logic and different model families." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "SigLIP2 docs/example errors are not the same as a checkpoint-specific AutoModel/pipeline output regression." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different loading failures: unit-test load correctness vs weight-shape mismatch." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to GLM-4.7-Flash loading tests." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory, but one is training memory usage and the other is quantization/offloading behavior." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "LoRA hf-version incompatibility is unrelated to Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Different regressions: buffer initialization junk vs custom model initialization in notebooks." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Related to memory pressure, but not the same concrete bug or fix." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "LoRA/model-version compatibility is unrelated to the resize_token_embeddings output_embeddings bug." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Different flash-attention crashes with different causes and affected paths." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Tokenizer/load failures for different models; no shared root cause is clear enough." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Qwen3.5 chat crash and Gemma4Processor _tokenizer error are separate code paths." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Extra_state save/load handling is distinct from init_empty_weights argument incompatibility." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Both are resize_token_embeddings regressions, but they affect different model families and different submodules, so this is not a true duplicate." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both mention lm_head tying, but the model classes and concrete failure paths differ." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin and Whisper return_language pipeline behavior are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "Different models and different symptom types; not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch version compatibility is unrelated to meta-device loading regressions." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing with LoRA is unrelated to the T5 apex RMSNorm memory leak." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Tokenizer vocab-size mismatch is different from a model-loading failure for sleng-bert." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Gemma4 FSDP2 attention-state rebuild and Qwen3.5/3.6 config unwrapping are unrelated." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Different failures in different models: Gemma4 attention state vs Qwen3_5Moe CPU offload." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Quantization offloading and int4 reserved-memory OOM are related only broadly, not as one duplicate bug." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster-level summary is mostly grounded, and the 43644/44534 duplicate call is well supported by the titles. However, the analyst overstates the Gemma/T5Gemma pair: 45276 and 45335 are similar resize_token_embeddings regressions, but they affect different model families and different submodules, so the packet does not justify treating them as a true duplicate with high confidence. The soft-edge verdicts should stay more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 7, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:39404", + "issue:41720", + "issue:42947", + "issue:43257", + "issue:43335", + "issue:43381", + "issue:43425", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43716", + "issue:43828", + "issue:43854", + "issue:43906", + "issue:43957", + "issue:44155", + "issue:44164", + "issue:44265", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44530", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44898", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:45084", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45701", + "issue:45823", + "issue:46050" + ], + "soft_pairs": [ + "issue:44155|issue:45823", + "issue:44291|issue:44756", + "issue:44292|issue:44912", + "issue:44960|issue:46050", + "issue:43335|issue:45663", + "issue:44387|issue:44756", + "issue:42947|issue:43381", + "issue:45276|issue:45292", + "issue:43854|issue:45216", + "issue:43381|issue:44928", + "issue:44164|issue:44756", + "issue:45588|issue:45663", + "issue:45684|issue:45698", + "issue:41720|issue:44155", + "issue:44479|issue:44877", + "issue:43957|issue:44756", + "issue:43646|issue:43708", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:44402|issue:44451", + "issue:43828|issue:43957", + "issue:43425|issue:43828", + "issue:44589|issue:44756", + "issue:43257|issue:44863", + "issue:44683|issue:45823", + "issue:39404|issue:45701", + "issue:43611|issue:43645", + "issue:43425|issue:45823", + "issue:44530|issue:45084", + "issue:43425|issue:43716", + "issue:43906|issue:44530", + "issue:44265|issue:44898", + "issue:39404|issue:44912", + "issue:44479|issue:44805" + ], + "prepared_review_unit_hash": "e88e86c42f1168a91e3dc0df819d6c6715b48096eb644fc5cfff203398e1045c", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37909, + "estimated_input_tokens": 9478, + "estimated_eval_tokens": 19212 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are unrelated regressions across different subsystems. The only clear duplicate here is the Gemma4/resize_token_embeddings reports, where one issue is a broader version of the other.", + "confidence": 0.82, + "canonical_issue_reason": "issue:45276 is the most complete description of the resize_token_embeddings regression and subsumes the narrower 45292 report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45276 is the best representative issue: it is specific, reproducible, and covers the broader symptom set in the only clear duplicate pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "Different failures: audio batched inference token leakage vs Gemma4 device-map forward mismatch." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "Unrelated code paths: init_empty_weights argument handling vs mmap-related OOM behavior." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both involve 4-bit quantization, but they are different models and failures (NVFP4 load error vs MXFP4 fallback to bf16)." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "GLM5 model-loading issue vs sinq save/load corruption; same broad area, different bug." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "Sparse-layer config creation bug and Gemma4 FSDP2 attention-state bug are distinct." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Different OOM causes: int4 quantization reserved-memory growth vs mmap on Strix Halo." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both mention gradient checkpointing, but one is LoRA effectiveness in training and the other is an eval-mode restriction." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "Same underlying resize_token_embeddings bug; 45292 is the narrower output_embeddings symptom of the broader failure reported in 45276." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different regressions: GLM-4.7-Flash loading failure vs Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Gradient-checkpointing eval-mode issue and RLHF NaN/SDPA-fallback issue are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "save/from_pretrained extra_state handling is unrelated to mmap/OOM on Strix Halo." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Flash-attention sinkless-model crash vs Gemma4 shared_kv_states/FSDP2 KeyError are different code paths." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Custom-model save permission propagation and wrong-module loading are different serialization bugs." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Qwen3 device-mapping CUDA assert and AudioFlamingo3 batch leakage are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video-input regression for Qwen models vs strict config preventing granite_speech loading." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are distinct issues." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization breakage and Trainer max_steps resume miscalculation are unrelated." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": false, + "reason": "Similar API area, but different manifestations across different model classes; not enough evidence for the same bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Notebook/custom-model init regression is unrelated to extra_state save/load handling." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Tokenizer vocab mismatch for special tokens and a ScandiBERT load failure are not the same concrete bug." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Autocast dtype mismatch and meta-device loading failure are different problems." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch version incompatibility and autocast dtype mismatch are separate failure modes." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup error is unrelated to mmap/OOM on Strix Halo." + }, + { + "left": "issue:43257", + "right": "issue:44863", + "accept": false, + "reason": "Qwen3 MOE conversion/loading issue and NemotronH checkpoint loading failure are different model-specific bugs." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention breakage on torch>=2.9 is unrelated to Gemma4 device-mapping mismatch." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline return_language regression and tokenizer-version drift are different issues." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "base_model_prefix loading breakage is distinct from the notebook custom-model initialization regression." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch 2.10 incompatibility and Mistral-3 dtype mismatch do not share the same underlying bug." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Isolated reproduction of an old issue and Qwen3.5 PagedAttentionCache group-type crash are not the same bug." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to Perceiver resizing/interpolation behavior." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline return_language regression and MXFP4 quantization loading behavior are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression and mask-shape IndexError are different failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays grounded in the packet and the soft-edge calls are conservative. The only accepted pair, issue:45276 vs issue:45292, is plausibly the same resize_token_embeddings regression with one report being a narrower symptom of the broader Gemma4 issue. No other pair is overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 8, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:32090", + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42915", + "issue:43299", + "issue:43421", + "issue:43425", + "issue:43452", + "issue:43541", + "issue:43582", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43646", + "issue:43716", + "issue:43856", + "issue:43901", + "issue:43931", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44451", + "issue:44479", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44756", + "issue:44811", + "issue:44918", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45584", + "issue:45759", + "issue:45859", + "issue:46304" + ], + "soft_pairs": [ + "issue:43901|issue:44509", + "issue:44589|issue:45084", + "issue:43425|issue:44918", + "issue:36032|issue:43452", + "issue:44164|issue:45759", + "issue:43299|issue:43931", + "issue:43596|issue:44683", + "issue:44991|issue:45081", + "issue:43994|issue:44079", + "issue:44671|issue:44991", + "issue:43856|issue:44756", + "issue:39404|issue:44671", + "issue:44737|issue:45859", + "issue:44164|issue:44479", + "issue:44488|issue:45081", + "issue:43582|issue:43632", + "issue:44661|issue:44683", + "issue:43611|issue:43644", + "issue:40990|issue:43421", + "issue:32090|issue:39404", + "issue:43716|issue:45084", + "issue:44488|issue:44987", + "issue:43643|issue:44530", + "issue:36032|issue:42915", + "issue:34567|issue:40444", + "issue:44811|issue:45584", + "issue:44530|issue:46304", + "issue:43541|issue:45084", + "issue:44451|issue:44987", + "issue:43644|issue:43646" + ], + "prepared_review_unit_hash": "0d966ea9026f716fdcd339c6b642fecfdbc6c7496d3cbcc535686041fe7b8a86", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36034, + "estimated_input_tokens": 9009, + "estimated_eval_tokens": 18274 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is thematically similar but not duplicate: it mixes many unrelated Transformers v5 regressions across tokenizers, model loading, training, multimodal paths, docs, and internal runtime behavior.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a safe global representative; the nearest broad regression issue is still too specific to model initialization and does not cover the rest of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but one is about stale return_all_scores wording and the other about removed pipeline task docs; not the same bug." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup failure and template compilation failure are unrelated runtime errors with different code paths." + }, + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch version incompatibility and Qwen3.5 embedding unpacking are different failures, not one underlying issue." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Tokenizer special-token name collision and gguf_file loading breakage are distinct tokenizer/loading bugs." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "save/from_pretrained extra_state handling and from_config text_config unwrapping affect different serialization paths." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Both involve Qwen3-VL loading, but one is a dev-version regression and the other is a weight-shape mismatch; not the same concrete bug." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "DeepSpeed zero3 BertModel init and compiled flex_attention on torch>=2.9 are unrelated." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "These are different tokenizer/model-loading failures for different models and different internal causes." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Siglip2 bad outputs and ModelOutput key assignment with prior None are unrelated behaviors." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and tokenizer loading failure for a different model are not the same issue." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory inefficiency and Strix Halo mmap OOM are different memory problems with different fixes." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper return_language pipeline regression and CamemBERT masked-LM regression are unrelated model-specific bugs." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "CPU device placement bug in XLNet relative positions and Qwen3.5 MoE CPU offload incompatibility are not the same code-path problem." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "Extra_state save/load failure and v5.3 video-input regression are unrelated." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Two different tokenizer-loading failures for different models; no shared underlying defect." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "AppleSilicon allocator warmup TypeError and _is_hf_initialized flag break are separate runtime regressions." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Model-like addition failing inside TOKENIZER_MAPPING_NAMES and compiled flex_attention failures 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 a duplicate." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and runtime post-processor updates in TokenizersBackend are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error and Whisper return_language pipeline regression are different subsystems and bugs." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Image preprocessor/model dtype mismatch and template compilation failure are unrelated." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "These are distinct model-loading regressions on different models and code paths." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "AutoConfig trust_remote_code missing fields and PagedAttentionCache invalid group type are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "Tokenizer add_special_tokens conflict and FineGrainedFP8Config Qwen3Moe failure are different bugs." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "TrainerState token counter not updating and Qwen2.5-VL iterable dataset multi-image finetuning failure are unrelated." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both are Whisper issues, but one is batch_decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens; not the same bug." + }, + { + "left": "issue:44530", + "right": "issue:46304", + "accept": false, + "reason": "Qwen3.5 PagedAttentionCache linear_attention crash and USE_DISTRIBUTED=0 support break are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Mixtral grouped_mm tracing failure and template compilation error are different runtime problems." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Loading vesteinn/ScandiBERT and loading physical-intelligence/fast are separate model-loading failures." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5-related regressions, but one concerns non-persistent buffers and the other custom model initialization; too broad to merge as duplicates." + } + ] + }, + "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": 18, + "review_unit_index": 9, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36032", + "issue:42175", + "issue:42915", + "issue:43089", + "issue:43262", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43452", + "issue:43482", + "issue:43540", + "issue:43611", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "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:44488", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44756", + "issue:44849", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45127", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45507", + "issue:45987", + "issue:46178" + ], + "soft_pairs": [ + "issue:44206|issue:44220", + "issue:44464|issue:45987", + "issue:43452|issue:43482", + "issue:43883|issue:44488", + "issue:43317|issue:43482", + "issue:44849|issue:45325", + "issue:36032|issue:43645", + "issue:45084|issue:45507", + "issue:42175|issue:46178", + "issue:43262|issue:43377", + "issue:44987|issue:44991", + "issue:43646|issue:44534", + "issue:43089|issue:45987", + "issue:43404|issue:44466", + "issue:43883|issue:44991", + "issue:43322|issue:43540", + "issue:43994|issue:44610", + "issue:44403|issue:45081", + "issue:42915|issue:45084", + "issue:43645|issue:44534", + "issue:43377|issue:43819", + "issue:44265|issue:44610", + "issue:44451|issue:45081", + "issue:44987|issue:45081", + "issue:45216|issue:45357", + "issue:44220|issue:44683", + "issue:36032|issue:43742", + "issue:42915|issue:44291", + "issue:43632|issue:44756", + "issue:42915|issue:44589", + "issue:44062|issue:44987", + "issue:43404|issue:45127", + "issue:43611|issue:44534" + ], + "prepared_review_unit_hash": "83919fdb0df6aa307b336be5586a9c78a712e6a1bb100f07f083baf58649cfee", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 37081, + "estimated_input_tokens": 9271, + "estimated_eval_tokens": 18798 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5 regressions. Only two soft edges look like true duplicates: the GGUF loading failures and the Qwen3.5 save_pretrained regressions.", + "confidence": 0.67, + "canonical_issue_reason": "issue:43646 is the clearest broad representative: it captures the common v5 regression theme around broken custom-model initialization without being overly model-specific.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best standalone issue in the set because it is concise, broad, and directly describes a core Transformers v5 breakage affecting model initialization.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Both are audio-feature-extraction bugs, but the titles point to different failure surfaces; not enough evidence they are the same concrete code-path defect." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Chunked generation with compiled forward is a different problem from StaticCache.get_seq_length returning a tensor." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": true, + "reason": "Both describe GGUF/model loading breakage in Transformers v5 and plausibly stem from the same loading-path regression." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "One is a missing all_tied_weights_keys attribute in Molmo; the other is a specific model-load failure. Different bugs." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "gpu/cpu offload with dequantized models is unrelated to GGUF loading failures." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen model families and different bugs: hidden-states output vs rope-index scaling." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "T5 tokenizer special-token conflict is unrelated to custom model initialization in notebooks." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Different subsystems and failures; no shared concrete code path." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Both involve backend resolution, but one is an install-extra omission and the other is an unsupported backend mapping entry." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and MIMI padding-mask batching are different bugs." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are model/tokenizer load regressions, but they affect different models and likely different loader paths." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Custom model initialization and non-persistent buffer serialization are unrelated v5 regressions." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation performance overhead is unrelated to cache sequence-length typing." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied weights, but one is a missing tie and the other is device-dependent serialization; not the same bug." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Segfault on Llava Next loading and video-input processing in Qwen3OmniMoe are different issues." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Nonsensical model outputs and wrong processor image size are different problems." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Noise during loading and Mistral regex patch crashes are not the same defect." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to template-node compilation errors." + }, + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Notebook/custom-model initialization and buffer serialization are different regressions." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "MIMI batching/padding-mask mismatch and DAC from_latents/STE mismatch are unrelated." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export with torch_compilable_check is unrelated to OmDet-Turbo processor sizing." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "Model-load failure for ScandiBERT is unrelated to Mistral regex patching." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Different failure modes and different model families." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Both are Qwen3.5 save_pretrained regressions about incorrect checkpoint serialization, likely the same underlying bug." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Audio fbank extraction and compiled flex_attention are unrelated." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "T5 tokenizer load conflict and MobileLLM key error are different loader issues." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "FP8 config failure and init_empty_weights/_is_hf_initialized TypeError are different bugs." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Broken _is_hf_initialized flag and mmap/OOM on Strix Halo are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "FP8 config failure and Float8 storage deserialization failure are different issues." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken special-argument duplication is unrelated to model loading regressions." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both touch tied embeddings, but one is a missing tie and the other is LoRA merge collapse with extended vocab; not clearly the same bug." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "base_model_prefix loading breakage is unrelated to non-persistent buffer junk filling." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays mostly conservative. The two accepted soft edges are reasonably supported by their shared GGUF-loading and Qwen3.5 save_pretrained themes, and the remaining rejections do not overclaim sameness based on the available titles. The cluster summary is also appropriately cautious about the set being mostly unrelated regressions." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 10, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:33666", + "issue:36032", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:43296", + "issue:43317", + "issue:43377", + "issue:43381", + "issue:43421", + "issue:43425", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43582", + "issue:43632", + "issue:43792", + "issue:43818", + "issue:43856", + "issue:44079", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44479", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44683", + "issue:44811", + "issue:44829", + "issue:44877", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45479", + "issue:45491", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923" + ], + "soft_pairs": [ + "issue:43531|issue:44877", + "issue:43296|issue:43531", + "issue:36032|issue:41720", + "issue:44479|issue:45325", + "issue:43425|issue:44403", + "issue:40990|issue:44811", + "issue:43792|issue:44589", + "issue:43421|issue:44568", + "issue:33666|issue:40444", + "issue:44403|issue:44991", + "issue:43632|issue:44534", + "issue:44079|issue:45663", + "issue:44683|issue:45020", + "issue:43381|issue:45923", + "issue:44462|issue:45812", + "issue:44829|issue:45479", + "issue:40444|issue:45325", + "issue:43482|issue:45538", + "issue:44291|issue:44589", + "issue:43317|issue:43856", + "issue:41553|issue:43582", + "issue:43482|issue:43525", + "issue:43818|issue:45357", + "issue:42175|issue:45542", + "issue:43377|issue:45491", + "issue:41720|issue:45081", + "issue:43531|issue:44291", + "issue:44265|issue:45072" + ], + "prepared_review_unit_hash": "ea582e59583f99458a07f163d3a12f7b33b3d4a8c4237beb80491ae1aa546a6f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35723, + "estimated_input_tokens": 8931, + "estimated_eval_tokens": 18118 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated Transformers/model-specific bug reports. The only common thread is broad regression-style loading/training failures, but the soft pairs mostly differ in model family, failure mode, and code path. No PRs are present.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45020 is the broadest regression-style issue here: it frames a recent-version breakage affecting remote_code models, which is the closest umbrella theme among otherwise disparate model/tokenizer bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the best cluster anchor because it is the most generally phrased and most representative of the shared \u2018recent Transformers version regression\u2019 pattern, whereas the others are narrower, model-specific failures.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Both are model-loading/config issues, but one is a Qwen3-MoE sliding_window bug and the other is a Granite Speech config strictness problem; different models and code paths." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL load failure and Qwen3-MoE sliding_window failure are unrelated model-specific bugs." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "T5 tokenizer special-token attribute conflict is unrelated to Qwen3 device-mapping cuda assertions." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Both involve Qwen VL, but one is a broad video-input regression and the other is a specific still-image temporal position-id scaling bug; not the same failure." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch 2.10 compatibility and noisy transformer loading are different issues." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "GPT-OSS perplexity on WikiText and Whisper batch_decode skip_special_tokens handling are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper-large-v2 runtime failure and missing Float8 storage object are different errors with no shared code path." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both concern tokenizer special tokens, but one is a backend post-processor update bug and the other is a tokenizer-specific BOS/EOS regression; too different to merge." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both mention Qwen VL training, but multi-GPU training and iterable-dataset multi-image prompt failures are distinct problems." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "General loading noise vs a specific tokenizer load failure for EMBEDDIA/est-roberta are different symptoms and fixes." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is about _is_hf_initialized and the other about non-persistent buffers; different underlying bugs." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment bug and Gemma-4 FSDP2 shared_kv_states KeyError are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45020", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and remote_code breakage in recent Transformers are different compatibility issues." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Gradient checkpointing in eval mode and missing supports_gradient_checkpointing flag are related only thematically; they are not the same bug." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Tokenizer.json being ignored and Granite token IDs being wrong are both tokenizer issues, but they point to different loading/processing failures." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both involve degenerate classification behavior, but one is tied to flash_attention_3 and the other to problem_type/num_labels=1 loss logic." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Qwen2.5-VL multi-image iterable-dataset training failure and Qwen2.5-VL temporal position-id scaling are different bugs." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "GGUF loading failure in Transformers v5 and CLIPTokenizer model_max_length handling are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized TypeError and missing Float8 storage object are distinct runtime issues." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "Device-map offload/load failure and Qwen3 MoE memory inefficiency are different training/inference problems." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Bad AutoTokenizer error message for Voxtral is unrelated to an Apple Silicon caching_allocator_warmup TypeError." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Qwen2.5-GGUF loading failure and Llama4Config missing pad_token_id are unrelated model/config issues." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA tower/temporal-attention issue and Qwen3.5 save_pretrained key regression are different model families and code paths." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "One is missing TensorFlow in a torch extra install, the other is tensorboard-only causing a TF backend error; related packaging area but not the same bug." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both are batching/padding-related model bugs, but they affect different models and mechanisms (padding mask support vs sliding-window all-padding NaNs)." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Qwen3 auto device mapping cuda assert and Mistral tokenizer regex patch crash are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Sliding_window Qwen3-MoE failure and init_empty_weights/_is_hf_initialized TypeError are different regressions." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export/torch_compilable_check failure and bfloat16 dtype mismatches are unrelated torch compatibility 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": 18, + "review_unit_index": 11, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:35707", + "issue:39692", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43296", + "issue:43377", + "issue:43454", + "issue:43482", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43701", + "issue:43746", + "issue:43761", + "issue:43792", + "issue:43819", + "issue:43881", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45325", + "issue:45381", + "issue:45584", + "issue:45800", + "issue:45823", + "issue:45834", + "issue:45920", + "issue:45923" + ], + "soft_pairs": [ + "issue:43611|issue:43646", + "issue:43632|issue:45072", + "issue:43994|issue:44898", + "issue:42175|issue:43296", + "issue:45325|issue:45834", + "issue:43792|issue:43957", + "issue:43454|issue:44466", + "issue:44079|issue:44534", + "issue:42915|issue:43482", + "issue:40444|issue:45381", + "issue:45325|issue:45381", + "issue:43644|issue:44079", + "issue:35707|issue:40444", + "issue:43761|issue:44079", + "issue:44479|issue:45823", + "issue:43606|issue:44265", + "issue:44291|issue:44877", + "issue:43296|issue:45800", + "issue:44568|issue:45920", + "issue:43701|issue:45325", + "issue:43606|issue:43819", + "issue:41720|issue:45823", + "issue:44186|issue:44265", + "issue:43454|issue:43550", + "issue:43262|issue:43746", + "issue:43881|issue:44488", + "issue:39692|issue:40444", + "issue:43377|issue:44610", + "issue:43454|issue:45072", + "issue:40990|issue:45584", + "issue:43257|issue:45923" + ], + "prepared_review_unit_hash": "c8e479840f8692cdbb194eba7ae24aa9aac02b19e2e87c03f7b1ae485e8a8452", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37139, + "estimated_input_tokens": 9285, + "estimated_eval_tokens": 18826 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose set of unrelated Transformers issues, with a few broad v5 regression themes around model loading, initialization, and tied weights. None of the soft pairs look close enough to merge as true duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43611 is the broadest and most representative issue in the set: it describes a concrete v5 model-loading regression and overlaps conceptually with several nearby initialization/loading complaints more than the others do.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single representative issue because it is concrete, well-scoped, and captures one of the recurring regression themes in the cluster without being overly model-specific.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5-related loading/init regressions, but one is about base_model_prefix loading and the other about custom model initialization; they are related theme-wise, not the same concrete bug." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "Different failure modes: _is_hf_initialized/model init regression versus bf16 dtype mismatches in inference." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Both are vision-model correctness issues, but the symptoms and code paths differ: SigLIP2 AutoModel/pipeline output corruption vs Perceiver resolution handling." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Completely unrelated: missing TensorFlow backend in an extra install versus PaddleOCR-VL load failure in vLLM." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Different models and bugs: Qwen2.5-VL rope/position-id scaling versus Kosmos2.5 OCR long-input index error." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "Whisper inference failure and meta-device loading regressions are not the same underlying bug." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both mention lm_head/tied weights, but one is a broken weight tie causing bad generation while the other is inconsistent serialization depending on device." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "Unrelated issues: ModelOutput key assignment versus non-persistent buffer initialization." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Different areas: FineGrainedFP8Config support for Qwen3Moe versus GGUF loading failure for Qwen2.5." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Both involve Qwen2.5-VL, but one is finetuning with multiple images per prompt and the other is video vision_position_ids regression." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Same model family, different bug: still-image temporal position scaling versus video-position-id computation." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer junk initialization is unrelated to ModelOutput key assignment." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL finetuning with multiple images." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression is unrelated to ModelOutput key assignment." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Different concrete failures: Qwen VL video-input regression versus Gemma4 PLE device mismatch under device_map auto." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU-offload device mismatch and torch.export export failures through torch_compilable_check are different problems." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized argument error is distinct from strict config validation for granite_speech." + }, + { + "left": "issue:43296", + "right": "issue:45800", + "accept": false, + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to torch/Transformers version incompatibility." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Tokenizer special-token regression for mdeberta-v3-base is not the same as broader AutoTokenizer wrong-token-ID regressions across multiple models." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to Qwen2.5-VL rope index scaling." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Both mention inference mismatches, but CPU offload device placement and DAC.from_latents vs forward STE mismatch are different bugs." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Qwen3 auto device map CUDA assertion and Gemma4 PLE device mismatch are separate model-specific device-placement bugs." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Tokenizer NER/padding crash and torch.export with torch_compilable_check are unrelated code paths." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "AyaVision weight tying/generation failure is unrelated to Bamba-9B-v2 torch.compile SDPA failure." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting and PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Two different model-loading failures with no clear shared root cause." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "Documentation example errors for SigLIP2 are not the same bug as Qwen2.5-VL finetuning with multiple images." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "MIMI padding-mask batch inconsistency and OmDet-Turbo processor/model resolution mismatch are different bugs." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Tied lm_head weights causing bad generation is unrelated to bf16 dtype mismatch issues." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on gpt-oss and Whisper empty-transcription generation failure are unrelated." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and missing supports_gradient_checkpointing on a trust_remote_code model 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": 18, + "review_unit_index": 12, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42617", + "issue:42915", + "issue:42947", + "issue:43317", + "issue:43377", + "issue:43404", + "issue:43450", + "issue:43454", + "issue:43526", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43708", + "issue:43720", + "issue:43749", + "issue:43873", + "issue:43881", + "issue:43883", + "issue:44008", + "issue:44186", + "issue:44387", + "issue:44423", + "issue:44466", + "issue:44534", + "issue:44610", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45357", + "issue:45406", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:46291" + ], + "soft_pairs": [ + "issue:43317|issue:43873", + "issue:43572|issue:45759", + "issue:40990|issue:42947", + "issue:43377|issue:44898", + "issue:43873|issue:45701", + "issue:43454|issue:43720", + "issue:41720|issue:44977", + "issue:44466|issue:45357", + "issue:44991|issue:46291", + "issue:43526|issue:43883", + "issue:43881|issue:44991", + "issue:42617|issue:42915", + "issue:35707|issue:41720", + "issue:34634|issue:35707", + "issue:43577|issue:43708", + "issue:44008|issue:45823", + "issue:43450|issue:45701", + "issue:43404|issue:45357", + "issue:44186|issue:45072", + "issue:44877|issue:44991", + "issue:43550|issue:43720", + "issue:36331|issue:42915", + "issue:36331|issue:42617", + "issue:44423|issue:45406", + "issue:42175|issue:46291", + "issue:44387|issue:45704", + "issue:43873|issue:44756", + "issue:44610|issue:44898", + "issue:43550|issue:43606", + "issue:43749|issue:44534", + "issue:44186|issue:44610", + "issue:45632|issue:45684" + ], + "prepared_review_unit_hash": "33054856126c2f2a82126e9db8c4e3596ca79469e4ef759093c0a01a4a44c0a9", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36883, + "estimated_input_tokens": 9221, + "estimated_eval_tokens": 18698 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated Transformer bugs with a few broad thematic overlaps (offloading/loading, tokenization, multimodal processors). I would not merge them as duplicates; only a couple look superficially similar, but the underlying failures differ.", + "confidence": 0.79, + "canonical_issue_reason": "If forced to pick one representative, issue 43873 is the broadest and most connected item in the offloading/quantization area, with multiple soft links and inbound references. It is still only a theme hub, not a true duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 is the best single issue to anchor the cluster because it has the strongest cross-links and the most general failure description among the offloading-related reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both involve offloading/loading, but one is a dequantized-model load failure with device_map=auto and the other is a broader quantization offloading issue; not the same concrete bug." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is unrelated to unwrapping text_config for composite Qwen multimodal configs." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "High perplexity on a model is not the same bug as gradient checkpointing not being effective with PEFT LoRA." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "MIMI batched-vs-single output mismatch due to padding masks is unrelated to Perceiver image resolution failures." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to tokenization changes across transformer versions." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "lm_head weight tying in a multimodal model is a different code path from packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3 auto device mapping cuda asserts and Qwen3.5 flash-attention generation failures are different failure modes." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both touch model save/serialization, but one is about tied lm_head weight serialization and the other about visual encoder keys in Qwen3.5 save_pretrained." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading for a specific model is unrelated to kernels v0.15 requiring version/revision metadata." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and Molmo missing all_tied_weights_keys are unrelated." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "A model loading failure is not the same as a tokenizer-loading regression for another model." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "Running 3d_parallel.py and Qwen3Moe FineGrainedFP8Config failures are unrelated." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is a different generation-path bug from Qwen3 device-mapping cuda asserts." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset not working is unrelated to generation with inputs_embeds and past_key_values." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Blip2 dtype handling and Trainer resume_from_checkpoint max_steps calculation are unrelated training/loading bugs." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Both are Gemma multimodal issues, but one is a variable-name collision in forward and the other is a device mismatch in PLE; not the same bug." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape bugs do not match tokenization behavior changes across versions." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Tied lm_head weight serialization and incorrect visual encoder keys are different model-specific save_pretrained issues." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "LayoutLMv2 tokenizer batching issues are unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config blocking granite_speech loading is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "Bamba torch.compile SDPA failures and BitNet packed-weight unpacking failures are different code paths and root causes." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "Trainer compute_loss argument incompatibility is unrelated to Qwen3Moe FineGrainedFP8Config failures." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Unexpected num_items_in_batch in compute_loss has nothing to do with 3d_parallel.py not running." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Continuous-batching multimodal serve crash due to string handling is different from Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "TensorFlow backend packaging and kernels version/revision metadata are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Int4 quantization memory growth/OOM and an apex RMSNorm memory leak are different memory issues with different causes." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both mention loading/offloading and memory pressure, but Strix Halo mmap OOM is a platform-specific fix, not the same as quantization offloading behavior." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Processor output size mismatch and Perceiver non-default-resolution failure are both vision preprocessing issues, but they affect different models and mechanisms." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Bamba torch.compile SDPA and Bark-small CPU offload device mismatch are unrelated." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM efficient loading broken is not the same as non-persistent buffers being filled with junk." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "LayoutLMv2 tokenizer NER/batch handling and OmDet processor image-size mismatch are different subsystems." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "trust_remote_code cache-path collisions and save_pretrained permission propagation are both custom-model save/load topics, but they are distinct 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": 18, + "review_unit_index": 13, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:29159", + "issue:30064", + "issue:34567", + "issue:34634", + "issue:35707", + "issue:36010", + "issue:36032", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43316", + "issue:43317", + "issue:43322", + "issue:43388", + "issue:43421", + "issue:43425", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43782", + "issue:43881", + "issue:44186", + "issue:44361", + "issue:44492", + "issue:44509", + "issue:44530", + "issue:44671", + "issue:44743", + "issue:44913", + "issue:45059", + "issue:45081", + "issue:45161", + "issue:45406", + "issue:45464", + "issue:45701", + "issue:45759", + "issue:45812", + "issue:46050", + "issue:46291" + ], + "soft_pairs": [ + "issue:43322|issue:46050", + "issue:43295|issue:45701", + "issue:42915|issue:43317", + "issue:41720|issue:43322", + "issue:40990|issue:45701", + "issue:41720|issue:42617", + "issue:41720|issue:45161", + "issue:36010|issue:42617", + "issue:43425|issue:46291", + "issue:45081|issue:45812", + "issue:43525|issue:46291", + "issue:43316|issue:44913", + "issue:43531|issue:43572", + "issue:44671|issue:44743", + "issue:43540|issue:43782", + "issue:43421|issue:45812", + "issue:44186|issue:44361", + "issue:43526|issue:43881", + "issue:29159|issue:36032", + "issue:45406|issue:45464", + "issue:43550|issue:43653", + "issue:30064|issue:34634", + "issue:43606|issue:43720", + "issue:43421|issue:45701", + "issue:44530|issue:46291", + "issue:42915|issue:43322", + "issue:43606|issue:43653", + "issue:43531|issue:45759", + "issue:44492|issue:44509", + "issue:34567|issue:35707", + "issue:43388|issue:45059" + ], + "prepared_review_unit_hash": "f446bc5262eb4bf413769d9181efc340b6bce67efbaf8d8971284cfa71c0c19e", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36309, + "estimated_input_tokens": 9078, + "estimated_eval_tokens": 18412 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is a loose collection of mostly unrelated issues across tokenizers, processors, model loading, distributed inference, and docs. None of the soft pairs look like the same underlying bug/change, so no duplicate merges are justified.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45812 is the strongest single issue candidate: it is concrete, clearly scoped, and describes a broad v4\u2192v5 regression with multiple references and an obvious failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the most actionable and representative issue here because it names the affected component, the regression nature, and the concrete symptom very clearly.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Different failure classes: model loading segfault for Llava Next vs quantization/save-load breakage for sinq models." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Both mention tokenization, but one is a processor/tokenizer API regression and the other is a broad version-dependent tokenization change; not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Different subsystems and symptoms: Qwen3MoE FP8 failure vs device-map/offload loading failure for a dequantized model." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Unrelated model-specific failures: auto device mapping assert on Qwen3 vs segfault while loading Llava Next." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "High perplexity is a model/eval quality issue; the other is a tokenizer regression report. No shared concrete code path." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "One is a Qwen3 device-mapping assert, the other is a generic script execution problem; no overlap in bug cause." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both are distributed-model issues, but they involve different models and different breakpoints (auto device map vs tensor-parallel-only on GPT-OSS MoE)." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to a failure running 3d_parallel.py." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "A torch version incompatibility and a kernels import requirement are different dependency problems." + }, + { + "left": "issue:45081", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but one is a Mistral regex patch crash and the other is wrong token IDs for Granite; different code paths and fixes." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Missing pad_token_id in a config is unrelated to the kernels version/revision import error." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Different config issues: Gemma3 text-config API discrepancy vs GPTNeoX rotary_pct persistence bug." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and StableLm missing pad_token_idx are unrelated config/model bugs." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM regression and Qwen modular recurrent-state reset are different model behavior issues." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Both involve Qwen multimodal models, but one is video-input processing and the other is weight_only loading." + }, + { + "left": "issue:43421", + "right": "issue:45812", + "accept": false, + "reason": "Runtime special-token/post-processor synchronization is a different tokenizer bug than Granite token-ID regression." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer implementations and failure modes: LayoutLMv2 NER/padding crash vs MLuke task AttributeError." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BEiT image-processor label reduction and glm-4v loading failure do not share an underlying cause." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "General fast/slow tokenizer inconsistency is not the same as T5Tokenizer\u2019s add_special_tokens conflict on import/load." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer and Qwen3.5 streaming inference crash are distinct API paths and symptoms." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "torch.compile/SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token decode behavior." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Image segmentation-map handling and Bark voice_preset processing are different processor bugs." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both are CI/loading issues, but they affect different models and mechanisms: CPU offload mismatch vs packed-weight unpacking." + }, + { + "left": "issue:43421", + "right": "issue:45701", + "accept": false, + "reason": "Special-token runtime updates and version-driven tokenization changes are not the same concrete bug." + }, + { + "left": "issue:44530", + "right": "issue:46291", + "accept": false, + "reason": "PagedAttentionCache group-type crash on Qwen3.5 has no overlap with the kernels import requirement issue." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Different failures on different architectures/models; no shared code-path evidence." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "CPU offload device mismatch and token registration/decode failure are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and AutoModel config unwrapping are different issues." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategy docs/code and outdated pipeline-task docs are separate documentation problems." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "TrainerState token-count tracking and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Metric-gathering label truncation and SAM3 text/bounding-box behavior are different bugs in different components." + } + ] + }, + "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": 18, + "review_unit_index": 14, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:33666", + "issue:34634", + "issue:35707", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42886", + "issue:43329", + "issue:43381", + "issue:43421", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43653", + "issue:43746", + "issue:43782", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:44062", + "issue:44186", + "issue:44368", + "issue:44451", + "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:45081", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:45923", + "issue:46294" + ], + "soft_pairs": [ + "issue:44514|issue:44545", + "issue:43873|issue:45704", + "issue:42490|issue:43329", + "issue:41720|issue:43746", + "issue:43421|issue:43927", + "issue:44898|issue:45072", + "issue:33666|issue:35707", + "issue:45759|issue:45859", + "issue:44451|issue:44991", + "issue:43531|issue:44589", + "issue:45588|issue:45759", + "issue:44368|issue:45357", + "issue:44062|issue:44451", + "issue:44987|issue:45042", + "issue:43421|issue:43653", + "issue:39692|issue:44610", + "issue:45704|issue:45923", + "issue:45663|issue:45923", + "issue:43381|issue:43844", + "issue:42886|issue:44843", + "issue:44918|issue:44936", + "issue:43421|issue:44186", + "issue:45704|issue:45859", + "issue:34634|issue:41720", + "issue:43381|issue:45704", + "issue:43525|issue:43531", + "issue:45081|issue:45356", + "issue:39692|issue:43873", + "issue:43540|issue:46294", + "issue:44625|issue:45479", + "issue:43782|issue:45381" + ], + "prepared_review_unit_hash": "281cc107725bfb8c50ba977d6a57c9ec0769a4d16a89179a640fa6555bcfe63d", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37486, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 + }, + "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 one soft pair is a true duplicate. The Qwen2_5_VLProcessor batched apply_chat_template crash appears twice, while the other candidate pairs share only broad subsystem/topic similarity rather than the same bug.", + "confidence": 0.93, + "canonical_issue_reason": "issue:44545 is the clearest representative of the duplicated Qwen2_5_VLProcessor.apply_chat_template crash, with the same failure mode and slightly better linkage/wording.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44545 is the best issue representative because it is the exact bug report for the only confirmed duplicate pair and is the cleaner, more central report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same bug and same API path: Qwen2_5_VLProcessor.apply_chat_template crashing on batched input when padding=False." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Different problems: quantization/offloading behavior vs a T5 RMSNorm memory leak." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Unrelated bugs: trailing-slash local path import warning vs a multimodal token counting video-branch NameError." + }, + { + "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." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Different tokenizer/config issues: runtime special-token post-processor updates vs DiaConfig custom token IDs after save/load." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Different code paths: Perceiver image resolution handling vs dtype mismatches in SwitchTransformers/TimmWrapperModel inference." + }, + { + "left": "issue:33666", + "right": "issue:35707", + "accept": false, + "reason": "Unrelated training/generation issues for different models and mechanisms." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Both involve Qwen 3.5 multimodal configs, but one is from_config text_config unwrapping and the other is CPU-offload rotary embedding compatibility; not the same bug." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading failures, but they affect different models and likely different root causes." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Different failures: Qwen3-MoE sliding_window behavior vs missing Float8 storage type." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "Different bugs: flash_attention sink handling vs composite config unwrapping in AutoModelForCausalLM.from_config." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "Same broad Qwen family, but one is a warning about tie_word_embeddings while the other is incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "Different tokenizer issues: AddedToken keyword conflict vs loading a specific pretrained model." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Different subsystems: loading a specific model vs PIL backend image processor torchvision dependency regression." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Different tokenizer bugs: runtime post-processor update vs BigBirdTokenizer mask token registration/empty decode." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Different documentation/model mismatch issue vs processor/model input-size mismatch." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "Different bugs: apex RMSNorm leak vs missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:45663", + "right": "issue:45923", + "accept": false, + "reason": "Different failures: Gemma4 FSDP2 shared_kv_states bug vs missing gradient-checkpointing support flag." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Different training-mode behaviors: checkpointing disallowed in eval vs abnormal gradient growth under ZeRO-3." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": false, + "reason": "Both hit HF_HUB_OFFLINE, but one is a cache-load failure and the other is a specific Mistral regex/model_info offline regression." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Different trainer-related issues: Qwen3.5 embedding unpacking vs trainer.evaluate() after trainer.train()." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Different tokenizer bugs affecting different models and failure modes." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "Different bugs: T5 memory leak vs Qwen3.5 MoE rotary embedding CPU-offload incompatibility." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "Different model families and unrelated failures." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction is unrelated to the T5 RMSNorm memory leak." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Different config/model issues: missing pad_token_id on Llama4Config vs Qwen3-MoE sliding_window behavior." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Related tokenizer area, but different concrete failures: backend_tokenizer attribute crash vs codec handling/incorrect warning regression." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "Different problems: SigLIP2 docs/model mismatch vs quantization offloading behavior." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both concern Qwen3.5 video, but one is a ValueError in Qwen3OmniMoe video processing and the other is chat-template/video-token mismatch in Qwen3.5-9B." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Different classification/config bugs: num_labels propagation vs degenerate zero loss for single-label classification." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different multimodal model bugs: weight_only loading failure vs incorrect video vision_position_ids." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: the only clearly duplicated pair is issue:44514/issue:44545, whose titles describe the same Qwen2_5_VLProcessor.apply_chat_template batched-padding crash. The remaining pair judgments stay conservative and do not overclaim duplicates. No pull-request mergeability concerns apply here because these are issue pairs, not PR pairs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 15, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:42915", + "issue:43295", + "issue:43335", + "issue:43388", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43618", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43716", + "issue:43720", + "issue:43746", + "issue:43756", + "issue:43819", + "issue:43873", + "issue:43950", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44077", + "issue:44186", + "issue:44220", + "issue:44448", + "issue:44488", + "issue:44625", + "issue:44661", + "issue:44671", + "issue:44811", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45081", + "issue:45183", + "issue:45356", + "issue:45381", + "issue:45479", + "issue:45834", + "issue:45907", + "issue:45987", + "issue:46032", + "issue:46050" + ], + "soft_pairs": [ + "issue:45356|issue:45907", + "issue:43994|issue:44671", + "issue:43720|issue:43819", + "issue:30064|issue:36010", + "issue:43746|issue:46050", + "issue:43596|issue:44220", + "issue:44077|issue:44625", + "issue:43653|issue:43720", + "issue:42915|issue:43746", + "issue:44220|issue:44661", + "issue:43540|issue:44008", + "issue:43618|issue:43697", + "issue:43550|issue:45072", + "issue:43746|issue:43873", + "issue:43295|issue:45834", + "issue:43550|issue:43819", + "issue:45183|issue:45479", + "issue:43388|issue:46032", + "issue:44186|issue:44898", + "issue:43335|issue:43756", + "issue:43716|issue:44008", + "issue:44062|issue:44488", + "issue:43653|issue:43819", + "issue:43531|issue:43643", + "issue:44843|issue:45081", + "issue:43950|issue:45987", + "issue:43643|issue:44877", + "issue:43421|issue:44811", + "issue:44448|issue:44671", + "issue:43540|issue:45381" + ], + "prepared_review_unit_hash": "2d46f29d098b03386e2fb72ecc58f5e86cf4202efac3f675d7bcc9a9e81ed331", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37161, + "estimated_input_tokens": 9291, + "estimated_eval_tokens": 18838 + }, + "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 bug reports with only shallow keyword overlap around tokenizers, loading, and model regressions. I do not see any strong duplicate pairs; the soft matches are mostly same-subsysteem but different root causes.", + "confidence": 0.63, + "canonical_issue_reason": "Issue 44843 is the clearest, most concrete regression report with a single identifiable failing code path (`_patch_mistral_regex` / offline hub access), so it\u2019s the best anchor for this heterogeneous cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44843 is the most actionable representative issue: specific symptom, clear repro surface, and it sits in a broader tokenizer-regression theme that appears multiple times in the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Both involve tokenizer-related behavior, but 45356 is a Mistral-regex/codec regression while 45907 is about forwarding `revision` in `list_repo_files`; different bugs." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Both are model-output quality regressions, but they affect different models and different code paths (`SigLIP2` vs `CamemBERT`)." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "One is a BitNet loading/unpacking failure, the other is a DAC latent/STE mismatch; unrelated functionality and failure modes." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image processor segmentation-map handling is unrelated to the `GenerationMixin` import error." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Both mention loading/serialization, but one is PEFT adapter loading for GraniteSpeech and the other is quantized-model save/load corruption; different underlying problems." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Different failures in different paths: DeepSpeed ZeRO-3 init on BertModel vs `_torch_extract_fbank_features()` audio preprocessing." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "`patchtsmixer` post-init policy and Qwen3.5 `num_labels` propagation are unrelated config/model issues." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBird special-token decode bug and BitNet packed-weight unpacking bug do not share the same root cause." + }, + { + "left": "issue:42915", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is not the same as GraniteSpeech local checkpoint loading with PEFT adapters." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction bug vs tokenizer-mapping-name failure when adding a model-like entry; different subsystems." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Video-input processing in Qwen3OmniMoe and Gemma 3n variable-name collision are distinct model-specific bugs." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "CLIP attentions field regression and RTDetrV2 output differences are separate model API/behavior issues." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Both mention SDPA/dtype/compile issues, but one is a Bamba compile failure and the other is a dtype mismatch in two different models." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "Checkpoint/PEFT loading vs offloading-with-quantization are related only at a broad loading level, not the same bug." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor/tokenizer regression and Kosmos2.5 OCR index error are different symptoms and code paths." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Both are model/runtime bugs, but SDPA compile failure and DAC STE mismatch are unrelated." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Validation error messaging across models is not the same as the single-label-classification loss degeneracy bug." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Metric-gathering last-batch behavior and Mamba2 cache correctness are different training/runtime issues." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer padding/truncation crash and Perceiver non-default-resolution failure are separate model-specific bugs." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and Smollm3 RoPE-layer dropping are unrelated architecture/config regressions." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Image preprocessor dtype mismatch in Mistral-3 is not the same as Gemma 3n audio tensor name collision." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "`AddedToken()` keyword collision and failure to load `cjvt/sleng-bert` are distinct tokenizer/load issues." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "Special-token registration bug and DAC latent mismatch do not share a code path or symptom." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and strict `trust_remote_code` config loading are unrelated." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Both touch Mistral-regex handling, but 44843 is offline hub access from unconditional `model_info()` while 45081 is an AttributeError on tokenizer internals; close but not the same bug." + }, + { + "left": "issue:43950", + "right": "issue:45987", + "accept": false, + "reason": "Non-persistent buffer corruption in `from_pretrained()` is unrelated to `StaticCache.get_seq_length()` returning the wrong type." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "`trust_remote_code` missing fields and strict config rejection for granite_speech are different config-loading failures." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Runtime special-token post-processor update and Whisper `batch_decode()` ignoring `skip_special_tokens` are related to tokenization but not the same defect." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Different model families and different regression surfaces: Pegasus generation output vs CamemBERT masked-LM predictions." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Qwen3OmniMoe video-processing error and Qwen2.5-VL video `vision_position_ids` issue are both multimodal, but the failing paths are distinct." + } + ] + }, + "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": 18, + "review_unit_index": 16, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:30064", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:42175", + "issue:43322", + "issue:43404", + "issue:43452", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43720", + "issue:43782", + "issue:43818", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:43994", + "issue:44186", + "issue:44442", + "issue:44479", + "issue:44485", + "issue:44509", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44821", + "issue:44857", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45406", + "issue:45584", + "issue:45685", + "issue:45701", + "issue:45834", + "issue:45987", + "issue:46192", + "issue:46209" + ], + "soft_pairs": [ + "issue:30064|issue:36331", + "issue:43782|issue:44977", + "issue:43577|issue:43818", + "issue:39692|issue:45834", + "issue:45584|issue:45834", + "issue:44821|issue:45084", + "issue:43322|issue:45685", + "issue:43606|issue:45072", + "issue:44479|issue:45834", + "issue:43825|issue:43994", + "issue:44186|issue:44442", + "issue:34634|issue:36010", + "issue:44485|issue:44509", + "issue:28218|issue:29159", + "issue:44561|issue:45020", + "issue:45987|issue:46192", + "issue:43577|issue:43582", + "issue:44610|issue:45072", + "issue:44805|issue:44936", + "issue:43720|issue:45072", + "issue:43452|issue:44991", + "issue:43825|issue:43827", + "issue:43404|issue:43782", + "issue:43582|issue:44857", + "issue:45081|issue:45406", + "issue:43906|issue:46209", + "issue:44671|issue:45701", + "issue:42175|issue:45081" + ], + "prepared_review_unit_hash": "406fb5e0e8454fed6a9c649241042ac7580872901fa1fe3f9d671b48af9a4054", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35532, + "estimated_input_tokens": 8883, + "estimated_eval_tokens": 18022 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous. The only recurring theme is tokenizer/version-regression reports, but none of the soft pairs clearly describe the same concrete bug or change, so I rejected all candidate merges.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45701 is the broadest live tokenizer-regression report and has the strongest cross-references; it is the best anchor for the small tokenizer-related theme, even though the overall item set is not a true duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45701 is the most generally representative issue in this collection because it captures the broad tokenizer-version regression theme and is open with multiple inbound references.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image processing void-segmentation handling and a custom trainer signature error are unrelated code paths." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading and Qwen3.5 flash-attention generation are different failures." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "BLIP2 dtype casting and Video-LLaVA missing temporal attention are distinct model bugs." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "SigLIP2 doc example errors and Kosmos2.5 OCR index errors are unrelated." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription generation and Kosmos2.5 OCR indexing are different problems." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor-from-URL loading and template compilation errors do not share a concrete failure mode." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next segfaults and an MPS histogram integer error are separate backend/model issues." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch and bfloat16 dtype mismatches are not the same bug." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Qwen video regression and Kosmos2.5 OCR index errors are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "A pipeline v5 error-message wording issue and a SigLIP2 model/pipeline result issue are different changes." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 tokenizer crashes and AutoTokenizer loading FastSpeech2ConformerTokenizer are unrelated tokenizer paths." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset handling and GenerationMixin import removal affect different subsystems." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE implementation discussion and v5 pipeline-task doc cleanup are not the same bug." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Both are tokenizer-related, but one is a specific added-token spacing bug and the other is a broader slow/fast inconsistency report." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Removal of is_torch_fx_available and remote_code model breakage are different API breakages." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "StaticCache sequence-length return type and return_tensors being ignored are unrelated behavior changes." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "BLIP2 dtype propagation and Apple Silicon allocator warmup TypeError are unrelated." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "OmDet-Turbo processor size mismatch and dtype mismatch issues are different model/precision problems." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask shape IndexError and trainer evaluate-after-train failure are distinct error scenarios." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight unpacking and dtype mismatch inference bugs do not share the same code path." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "gguf_file loading breakage and a model-specific tokenizer load failure in v5 are not the same underlying bug." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Both mention removed pipeline tasks, but one is an error message issue and the other is stale documentation." + }, + { + "left": "issue:43404", + "right": "issue:43782", + "accept": false, + "reason": "Mistral3 lm_head tying and Qwen3VL weight_only loading are different model-loading defects." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Apple Silicon allocator warmup and float16 AMP CUDA loss crashes are unrelated runtime issues." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Mistral regex patching and Gemma4Processor missing _tokenizer are separate tokenizer/processor bugs." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "A reproduction of an earlier issue and an FP8 inference failure on B200 are not the same change." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM prediction regressions and general tokenization changes are different failures." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "Missing TensorFlow in the torch extra and Mistral tokenizer regex patching 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": 18, + "review_unit_index": 17, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:34634", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43454", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43650", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43746", + "issue:43749", + "issue:43819", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:44060", + "issue:44291", + "issue:44368", + "issue:44387", + "issue:44554", + "issue:44589", + "issue:44610", + "issue:44617", + "issue:44811", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45084", + "issue:45459", + "issue:45538", + "issue:45600", + "issue:45803", + "issue:46285" + ], + "soft_pairs": [ + "issue:43262|issue:43819", + "issue:43653|issue:44912", + "issue:45003|issue:46285", + "issue:43526|issue:43697", + "issue:44060|issue:44368", + "issue:43650|issue:43906", + "issue:43742|issue:45020", + "issue:45084|issue:45538", + "issue:43329|issue:45600", + "issue:43746|issue:44811", + "issue:44589|issue:45042", + "issue:43644|issue:43749", + "issue:43638|issue:44936", + "issue:43550|issue:44912", + "issue:43819|issue:43927", + "issue:44387|issue:44617", + "issue:34634|issue:36331", + "issue:43854|issue:44977", + "issue:39692|issue:44912", + "issue:36010|issue:43645", + "issue:38175|issue:44610", + "issue:43720|issue:44610", + "issue:45459|issue:45803", + "issue:43526|issue:44554", + "issue:43454|issue:43577", + "issue:36010|issue:36331", + "issue:43262|issue:44811", + "issue:43749|issue:44291", + "issue:43531|issue:45042", + "issue:43257|issue:43329" + ], + "prepared_review_unit_hash": "4c9c60b7b92bb8906755388c0500953e639f521c0e26f92e828352333c6a52b0", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36924, + "estimated_input_tokens": 9231, + "estimated_eval_tokens": 18718 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated model- or component-specific regressions; none of the soft pairs look like the same underlying bug or change, so no merges are justified.", + "confidence": 0.92, + "canonical_issue_reason": "issue:45020 is the broadest and most representative regression report in the set: it captures a general class of recent Transformers breakages affecting remote_code models, unlike the more narrowly scoped model/component-specific issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the strongest standalone issue candidate because it is upstream-facing, broadly scoped, and more representative of the cluster than the highly specific bugs around single models, processors, or trainers.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate defaults and DAC latent reconstruction are unrelated code paths and failure modes." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "Tokenizer special-token registration and MXFP4 quantization loading are different bugs with no shared underlying fix." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access and CLIP text model structure inconsistency are separate modeling-utils vs architecture issues." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "Beit label reduction and RTDetr output drift are different processor/model problems; they would not be merged into one fix." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Incorrect weight tying in Qwen3-Next and a tied-embeddings warning during LoRA fine-tuning are not the same concrete bug." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "A placeholder 'ADD THE DATA' issue and an isolated reproduction of another bug are not duplicates." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "Loading a specific MobileLLM checkpoint failing with KeyError is distinct from the broad remote_code regression report." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template compilation errors and CLIPTokenizer max-length semantics are unrelated issues." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "A multimodal token counting video-branch bug and stale auto_mappings references are different maintenance problems." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT checkpoint loading for GraniteSpeech and Whisper batch_decode skip_special_tokens behavior are unrelated." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Missing Float8 storage lookup and unnecessary torchvision dependency in PIL image processors affect different subsystems." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer initialization corruption and FSDP CPU RAM efficient loading are both loading-related but not the same code-path bug." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Zero3 empty-index failure on non-pretrained Bert and evaluate-after-train behavior are separate trainer/distributed-training issues." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba and MXFP4 fallback during load for git-oss are different model/runtime regressions." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC latent mismatch and DiaConfig token-ID persistence loss are unrelated implementation bugs." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "General int4 quantization memory regression and Sam3Video OOM are not the same concrete defect." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature mismatch are unrelated API issues." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Model loading test failure for GLM-4.7-Flash and flash-attention generation issues in Qwen3.5 are different bugs." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "SigLIP2 doc/example errors and MXFP4 load fallback are distinct; one is docs/model mismatch, the other quantization loading." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import breakage and Jupyter custom model initialization breakage are separate import/runtime issues." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and OmDet-Turbo processor size mismatch are different model/processor regressions." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading and OmDet-Turbo input-size mismatch are unrelated." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer protobuf error masking and FuyuBatchFeature bare except swallowing interrupts are both error-handling smells but not the same fix or code path." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Beit label reduction and MPS attention head-dimension correctness are unrelated bugs." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "AyaVision weight tying and BLIP2 dtype propagation are distinct model initialization issues." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "An import error for GenerationMixin and a custom trainer loss-signature error are unrelated failures." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio apply_chat_template sampling-rate defaulting and Whisper batch_decode skip_special_tokens behavior are separate processor bugs." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "FSDP CPU RAM loading and init_empty_weights _is_hf_initialized argument mismatch are different loading mechanisms and fixes." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding_window and PIL backend torchvision dependency are unrelated regressions." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and multimodal token counting video-branch undefined names 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": 18, + "review_unit_index": 18, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:24643", + "issue:29159", + "issue:33357", + "issue:34567", + "issue:38175", + "issue:40990", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43329", + "issue:43450", + "issue:43452", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43643", + "issue:43696", + "issue:43697", + "issue:43742", + "issue:43927", + "issue:43931", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44560", + "issue:44568", + "issue:44821", + "issue:44849", + "issue:44898", + "issue:44912", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45137", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45538", + "issue:45701", + "issue:45759", + "issue:45859", + "issue:45907" + ], + "soft_pairs": [ + "issue:43643|issue:45759", + "issue:42490|issue:43257", + "issue:43329|issue:43450", + "issue:43606|issue:45859", + "issue:34567|issue:40990", + "issue:44568|issue:45356", + "issue:29159|issue:34567", + "issue:43329|issue:44479", + "issue:44849|issue:45701", + "issue:42915|issue:43742", + "issue:44291|issue:45042", + "issue:43257|issue:43931", + "issue:43452|issue:43582", + "issue:43927|issue:44912", + "issue:43697|issue:45357", + "issue:44568|issue:45488", + "issue:33357|issue:34567", + "issue:42915|issue:44821", + "issue:41720|issue:43540", + "issue:42915|issue:45538", + "issue:44987|issue:45538", + "issue:44479|issue:45381", + "issue:43540|issue:44479", + "issue:43317|issue:45538", + "issue:43540|issue:44560", + "issue:44987|issue:45907", + "issue:45137|issue:45405", + "issue:24643|issue:29159", + "issue:43742|issue:44291", + "issue:43577|issue:45072", + "issue:44062|issue:44991", + "issue:38175|issue:44898", + "issue:43257|issue:43696" + ], + "prepared_review_unit_hash": "7863ce51a072931d6ac40d058aec43ecb68585198b5a3d073426d05f61484e21", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37725, + "estimated_input_tokens": 9432, + "estimated_eval_tokens": 19120 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous across tokenizer, loading, quantization, device-map, and training bugs. The only plausible subcluster is the Qwen video regressions, but the evidence is only soft title similarity; there are no explicit duplicate links or shared-target signals to confirm a merged duplicate set.", + "confidence": 0.67, + "canonical_issue_reason": "issue:44479 is the best available issue representative for the small Qwen video-regression subcluster because it broadly names the shared video-input failure mode. That said, the packet does not show explicit duplicate links or shared-target evidence, so this is only a cautious representative choice, not a confirmed duplicate hub.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the strongest overall issue pick because it is the broadest, most central report among the only closely related video issues. The rest of the cluster is too mixed, and the related Qwen entries remain only tentatively similar.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "Both involve config/model loading, but they describe different failures and code paths." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Trailing-slash local-path import warning is unrelated to Qwen3 MOE weight conversion." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Both are video-related, but one is a missing helper in multimodal token counting and the other is a batched video processor shape bug." + }, + { + "left": "issue:43606", + "right": "issue:45859", + "accept": false, + "reason": "Both mention CPU offload, but they are different model-specific incompatibilities." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token accounting and GPT-oss perplexity are unrelated failures." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "These are separate tokenizer regressions affecting different models and behaviors." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency is unrelated to TrainerState token counting." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "They are both in the Qwen video area, but the concrete symptoms differ and the packet does not prove the same bug." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "One is a specific Qwen3.5 output bug; the other is a broad tokenization regression report." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3Moe FineGrainedFP8Config failure and MobileLLM-125M key error are different loading problems." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized and PIL-backend torchvision requirement are unrelated." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3 MOE weight conversion and Qwen3-VL weight-shape mismatch are different loading bugs." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "gguf_file loading breakage and AppleSilicon TypeError are unrelated." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "DiaConfig token-ID loss and MXFP4 quantization fallback are different bugs." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different model families and different symptoms: RTDetrV2 output drift versus Qwen3.5 save_pretrained key handling." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer-related, but they affect different models and failure modes." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS bus error with CLIP and TrainerState token counting are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and AutoImageProcessor URL loading are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Auto device mapping CUDA assert and Qwen3OmniMoe video ValueError are distinct failures." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and CLIPTokenizer model_max_length bug are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Loading physical-intelligence/fast and CLIPTokenizer model_max_length are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen video regressions, but 45381 looks like a specific vision_position_ids symptom and the packet does not confirm the same underlying code-path bug." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Both involve video, but the Qwen3OmniMoe StopIteration report is not clearly the same regression as 44479." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Model loading failure and list_repo_files revision forwarding are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error and MIN_PEFT_VERSION packaging issue are unrelated." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed weight-dimension error and tokenizer inconsistency are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "MobileLLM-125M key error and unexpected _is_hf_initialized are different loading failures." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype, but the underlying bugs differ." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "AddedToken special-argument duplication and EMBEDDIA tokenizer loading are different tokenizer bugs." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and Perceiver non-default-resolution failure are unrelated image-model issues." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3 MOE weight conversion and GPT-oss CUDA OOM are unrelated." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall clustering judgment is mostly conservative, but the canonical/best-issue reasoning overstates the evidence. In the packet there are no explicit duplicate links or shared-target evidence, so describing issue:44479 as a \u201cconfirmed duplicate-worthy subset\u201d and citing a \u201cdirect duplicate-style link\u201d is too strong. The Qwen video issues do look plausibly related by titles, but the packet only supports a tentative soft similarity, not confirmation. A more cautious summary would be acceptable." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 19, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:34567", + "issue:34689", + "issue:36010", + "issue:36331", + "issue:42175", + "issue:43257", + "issue:43404", + "issue:43452", + "issue:43493", + "issue:43526", + "issue:43531", + "issue:43643", + "issue:43646", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43824", + "issue:43906", + "issue:43994", + "issue:44062", + "issue:44291", + "issue:44488", + "issue:44589", + "issue:44625", + "issue:44737", + "issue:44756", + "issue:44792", + "issue:44821", + "issue:44877", + "issue:44913", + "issue:44964", + "issue:45042", + "issue:45081", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45375", + "issue:45381", + "issue:45405", + "issue:45479", + "issue:45588", + "issue:45600", + "issue:45920", + "issue:45987", + "issue:46121" + ], + "soft_pairs": [ + "issue:44737|issue:45588", + "issue:43404|issue:45987", + "issue:34567|issue:36331", + "issue:44062|issue:45081", + "issue:43646|issue:43749", + "issue:44913|issue:46121", + "issue:43493|issue:43994", + "issue:43906|issue:45381", + "issue:43749|issue:44756", + "issue:43257|issue:43756", + "issue:44792|issue:44964", + "issue:44291|issue:44821", + "issue:36010|issue:42175", + "issue:28282|issue:34689", + "issue:42175|issue:43824", + "issue:45042|issue:45081", + "issue:45357|issue:45920", + "issue:43452|issue:44488", + "issue:43742|issue:44589", + "issue:45405|issue:45600", + "issue:28282|issue:29127", + "issue:43643|issue:45375", + "issue:43526|issue:45479", + "issue:43756|issue:45325", + "issue:43531|issue:44821", + "issue:44625|issue:44877", + "issue:43697|issue:45216" + ], + "prepared_review_unit_hash": "d3a084ed0677bf37bb7fed8cbbb4dc0551a772033067fd8c0d2154c0e3808efe", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36146, + "estimated_input_tokens": 9037, + "estimated_eval_tokens": 18330 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated Transformers bugs and regressions, with one plausible duplicate SigLIP2 pair. No PRs are present.", + "confidence": 0.41, + "canonical_issue_reason": "issue:45920 is the broadest, most representative open report here: a v5 regression in AutoTokenizer producing wrong token IDs across multiple model families, with clear user impact and ongoing relevance.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 has the strongest cross-cutting scope and is the best single issue to anchor this otherwise heterogeneous cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "Different subsystems and symptoms: XLNet positional encoding device placement vs flash-attention sink handling." + }, + { + "left": "issue:43404", + "right": "issue:45987", + "accept": false, + "reason": "Unrelated bugs: Mistral3 weight tying vs StaticCache sequence-length return type." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "Both involve Trainer APIs, but one is a stale counter and the other is a custom loss signature change; not the same bug." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Distinct tokenizer failures with different code paths: AddedToken kwargs conflict vs Mistral regex patching." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Different failures: config field loss under trust_remote_code vs FSDP CPU RAM efficient loading." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Both touch RoPE/config handling, but one is config reload persistence and the other is a validation TypeError on JSON-loaded lists." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Both point to the same SigLIP2 correctness problem: the HF implementation diverges from expected behavior, and the model/pipeline yields bad outputs for the same family/checkpoint." + }, + { + "left": "issue:43906", + "right": "issue:45381", + "accept": false, + "reason": "Different model families and symptoms; one is an isolated reproduction of an older issue, the other is a Qwen2.5-VL video position-id bug." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "Both concern loading/performance, but one is FSDP RAM-efficient loading and the other is mmap/OOM on Strix Halo." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Different Qwen3-MoE vs Smollm3 model-specific bugs with no clear shared code path." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Separate multimodal model failures affecting different checkpoints and code paths." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Different loading bugs: init_empty_weights / _is_hf_initialized vs loading AutoImageProcessor from URL." + }, + { + "left": "issue:36010", + "right": "issue:42175", + "accept": false, + "reason": "Unrelated issues: GenerationMixin import path vs missing TensorFlow backend in a torch install extra." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Generic missing-PyTorch ImportError vs a specific Llama 3.2 Vision model-loading regression." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Different missing-dependency/import problems: backend packaging vs missing Qwen2.5-VL class export." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "Different bugs: PIL backend now requiring torchvision vs Mistral regex patch crashing on tokenizer internals." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Separate regressions: incorrect visual encoder keys on save_pretrained vs wrong token IDs on tokenization." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but one is a gguf_file/from_pretrained breakage and the other is a model-specific load failure; not enough evidence of the same bug." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Different model load failures: key error for MobileLLM-125M vs missing Float8 storage object." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "One is a dependency version bump issue; the other is stale auto_mappings CI breakage." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Different problems: PyTorch-not-found ImportError vs LayoutLMv3 box-content validation clarity." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Both involve config field handling, but one is missing fields from trust_remote_code and the other is a specific Qwen3_5MoeVisionConfig field dropped by strict config parsing." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "Different label/classification issues affecting different components and failure modes." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen/Smollm3 RoPE issues with distinct symptoms and code paths." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Different domains entirely: Qwen3-MoE sliding_window behavior vs image processor URL loading." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "Separate config propagation bugs affecting different models and strictness logic." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "Different regressions: RTDetrV2 output changes in v5 vs Qwen3.5 save_pretrained checkpoint corruption." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: the cluster is indeed heterogeneous, issue:45920 is the broadest open report, and the only accepted soft pair (issue:43493 / issue:43994) is plausibly the same SigLIP2 bug family. The other pairwise rejections are phrased conservatively and do not overclaim evidence. No PR pairs are present." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 20, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:24643", + "issue:29127", + "issue:29159", + "issue:30064", + "issue:33357", + "issue:33666", + "issue:38175", + "issue:41720", + "issue:41762", + "issue:43295", + "issue:43452", + "issue:43531", + "issue:43577", + "issue:43582", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43819", + "issue:43906", + "issue:43927", + "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:44898", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45005", + "issue:45290", + "issue:45356", + "issue:45372", + "issue:45381", + "issue:45542", + "issue:45823", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46202" + ], + "soft_pairs": [ + "issue:43720|issue:44898", + "issue:43582|issue:44977", + "issue:44514|issue:45381", + "issue:24643|issue:29127", + "issue:41762|issue:44936", + "issue:43452|issue:43577", + "issue:43452|issue:44843", + "issue:44466|issue:45005", + "issue:43906|issue:46202", + "issue:45372|issue:46170", + "issue:43819|issue:44811", + "issue:43295|issue:45356", + "issue:30064|issue:33666", + "issue:44936|issue:45823", + "issue:43653|issue:44361", + "issue:44361|issue:44442", + "issue:44484|issue:44704", + "issue:38175|issue:43295", + "issue:44514|issue:45290", + "issue:45542|issue:46178", + "issue:43653|issue:44568", + "issue:43577|issue:43697", + "issue:29159|issue:33357", + "issue:44843|issue:44991", + "issue:44493|issue:46050", + "issue:44464|issue:44743", + "issue:43653|issue:43927", + "issue:43531|issue:43742", + "issue:41720|issue:43906" + ], + "prepared_review_unit_hash": "44834958a2c1d5146b30be2d98f62cc6d4e720d3b0bc0260bc45f27e05923dec", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36646, + "estimated_input_tokens": 9162, + "estimated_eval_tokens": 18580 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous: the soft pairs mostly share only broad themes like loading, tokenization, or generation, but not the same concrete bug path. I would not merge any of the suggested soft edges.", + "confidence": 0.35, + "canonical_issue_reason": "issue:43452 is the broadest concrete regression here: it breaks both AutoTokenizer and AutoModelForCausalLM loading via gguf_file, making it the best anchor if a canonical issue must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43452 is the most representative standalone loader regression in this set; it is broader and more central than the narrower model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "Different models and failures: BitNet packed-weight loading vs Perceiver image-size interpolation." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "Unrelated code paths: Apple Silicon allocator warmup TypeError vs Qwen3.5 flash-attention generation." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL, but one is batched chat templating with padding=False and the other is video vision_position_ids; different bugs." + }, + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "DeepSpeed weight-shape error is unrelated to LayoutLMv3 box-content validation and messaging." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3 ZeRO-3 load crash is not the same as trainer.evaluate() failing after train()." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "Loader failure with gguf_file is unrelated to Blip2 dtype remaining float32." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both concern model loading, but gguf_file breakage and offline model_info() calls in tokenizer patching are different bugs." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is device-dependent serialization of lm_head.weight and the other is a broader translation-model tied-weights issue." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "A reproduction of another issue is not the same as the Gemma pipeline roundtrip failure on '\u2581'." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Gemma 4 processor import failure from mistral_common is unrelated to Windows/ROCm torch.distributed import/type errors." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latent/forward mismatch is unrelated to Whisper batch_decode ignoring skip_special_tokens." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Different regressions: processor.tokenizer access/images passthrough vs Kimi-K2.5 codec handling and warning behavior." + }, + { + "left": "issue:30064", + "right": "issue:33666", + "accept": false, + "reason": "Void segmentation-map processing and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:44936", + "right": "issue:45823", + "accept": false, + "reason": "Trainer lifecycle failure is not the same as Gemma4 device_map='auto' forward mismatch." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer bugs affecting different models and code paths." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "MLukeTokenizer task AttributeError is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "save_pretrained shard-size behavior is unrelated to AutoProcessor kwargs not reaching cached_file." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 is unrelated to the processor/tokenizer regression." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both hit apply_chat_template, but one is batched padding=False and the other is assistant tool-call messages with no content." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both are backend-installation related, but they fail in different backend-mapping scenarios and code paths." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "BigBird special-token registration and mdeberta BOS/EOS insertion are different tokenizer bugs." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "Blip2 dtype persistence and RTDetrV2 output differences are unrelated model-specific regressions." + }, + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency is unrelated to the MacOS CLIP bus error." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Offline model_info() misuse in tokenizer patching is not the same as the est-roberta tokenizer loading failure." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Position-id warnings are unrelated to quantized-model save/load breakage." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and recurrent-state reset with cache are different generation bugs." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token registration is unrelated to DiaConfig losing custom token IDs on save/load." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to MobileLLM-125M key errors during load." + }, + { + "left": "issue:41720", + "right": "issue:43906", + "accept": false, + "reason": "Qwen3 device-mapping cudaErrorAssert is unrelated to an isolated reproduction of another 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": 18, + "review_unit_index": 21, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28282", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:34567", + "issue:35141", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43262", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43688", + "issue:43720", + "issue:43742", + "issue:43756", + "issue:43825", + "issue:43927", + "issue:43994", + "issue:44361", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44561", + "issue:44610", + "issue:44821", + "issue:44849", + "issue:44913", + "issue:44977", + "issue:45357", + "issue:45440", + "issue:45491", + "issue:45588", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46192", + "issue:46209" + ], + "soft_pairs": [ + "issue:43550|issue:43927", + "issue:44464|issue:44977", + "issue:43550|issue:44361", + "issue:44977|issue:46209", + "issue:43643|issue:44561", + "issue:32090|issue:34567", + "issue:43577|issue:45588", + "issue:28282|issue:35141", + "issue:43504|issue:43550", + "issue:32090|issue:33666", + "issue:43504|issue:43720", + "issue:45491|issue:46002", + "issue:44466|issue:46050", + "issue:43504|issue:44610", + "issue:41720|issue:44464", + "issue:33666|issue:36010", + "issue:43742|issue:44821", + "issue:43688|issue:45440", + "issue:44849|issue:46209", + "issue:45823|issue:46050", + "issue:28282|issue:32090", + "issue:28282|issue:33357", + "issue:43262|issue:43295", + "issue:43720|issue:43927", + "issue:43452|issue:44451", + "issue:43606|issue:43927", + "issue:43994|issue:46192", + "issue:33666|issue:36331", + "issue:43756|issue:44913", + "issue:43720|issue:44361", + "issue:43525|issue:43572", + "issue:44849|issue:45357", + "issue:43493|issue:43825", + "issue:39692|issue:43493" + ], + "prepared_review_unit_hash": "3d68c933e896489d49e2ea134bca4a930ea286446b7da563a2f9ae55acfdf011", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 34, + "serialized_chars": 37451, + "estimated_input_tokens": 9363, + "estimated_eval_tokens": 18982 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a grab bag of mostly unrelated transformer issues; the soft-similarity links look like false positives. The broadest and most canonical issue is the v5 regression breaking `trust_remote_code` models.", + "confidence": 0.71, + "canonical_issue_reason": "issue:44561 is the best representative because it describes a broad, version-level regression with a clear root cause and impact on multiple remote-code models.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44561 has the clearest cross-cutting regression scope and is likely to be the most useful anchor for related reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "Different bugs: one is torch.compile/SDPA failure in Bamba, the other is config persistence losing token IDs." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Both are generation failures, but they involve different models and different failure modes (compiled forward vs flash-attention)." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Unrelated failures: torch.compile SDPA vs tokenizer task AttributeError." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Different root causes and subsystems: flash-attention generation vs FP8 inference on B200." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both affect remote-code workflows, but one is missing fields in AutoConfig output while the other is a v5 API removal breakage." + }, + { + "left": "issue:32090", + "right": "issue:34567", + "accept": false, + "reason": "Trainer broadcast type error and TrainerState token counter bug are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:45588", + "accept": false, + "reason": "Blip2 dtype loading and flash_attention.py s_aux=None crash are different code paths." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "ImportError for missing PyTorch and embedding reinitialization during post_init are not the same bug." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Legacy-field loading bug in BEiT is unrelated to Bamba torch.compile/SDPA failure." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer broadcast error and multi-GPU training request are not the same issue." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Different model-loading bugs: legacy preset field vs packed-weights unpacking during accelerate loading." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both mention NaNs, but one is Gemma3 sliding-window attention and the other is classification head init/checkpoint behavior." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both involve save/load behavior, but one is tied-weight serialization and the other is quantized sinq persistence." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "BEiT legacy field loading and OmDet-Turbo processor size mismatch are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Qwen3 device-mapping assert and chunked generation with compiled forward are different concrete failures." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU training and GenerationMixin import breakage are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "KeyError on MobileLLM loading vs AutoImageProcessor URL loading are different components and bugs." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both concern model correctness, but they are different architectures and different implementation mismatches." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Device-mapped Gemma4 PLE mismatch and quantized save/load breakage are different problems." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Missing PyTorch import error and Trainer GPU broadcast type error are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "ImportError for missing PyTorch and Mac bus error with CLIP are not the same underlying bug." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Both are regressions, but one is audio chat-template sampling-rate handling and the other is processor.tokenizer/image passthrough breakage." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Packed-weight loading failure and lost custom token IDs are separate loading bugs." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "gguf_file tokenizer/model loading and ScandiBERT load failure do not match on root cause." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to config save/load token ID loss." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 inference quality bug and return_tensors being ignored are different issues." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Multi-GPU training request and unexpected Trainer compute_loss kwarg are distinct Trainer issues." + }, + { + "left": "issue:43756", + "right": "issue:44913", + "accept": false, + "reason": "Smollm3 RoPE layer mismatch and GPTNeoX rotary_pct reload regression are similar in theme but not the same bug." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Packed-weight unpacking failure and MLukeTokenizer task AttributeError are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Both involve missing pad-token config fields after v5, but they affect different models/config attributes and are not one concrete bug." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Hidden_states bug and incorrect visual-encoder key serialization are different Qwen3.5 issues." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and pipeline translation-task messaging are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Doc example errors for SigLIP2 and JAX-vs-HF implementation discrepancy are separate issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 22, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "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:43577", + "issue:43606", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43749", + "issue:43756", + "issue:43824", + "issue:43844", + "issue:43927", + "issue:44112", + "issue:44291", + "issue:44361", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44493", + "issue:44554", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44843", + "issue:44964", + "issue:44977", + "issue:45083", + "issue:45200", + "issue:45229", + "issue:45405", + "issue:45520", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46285" + ], + "soft_pairs": [ + "issue:44964|issue:45405", + "issue:44479|issue:44625", + "issue:43688|issue:43749", + "issue:43577|issue:44554", + "issue:44568|issue:45812", + "issue:43540|issue:44977", + "issue:45200|issue:45229", + "issue:44488|issue:44843", + "issue:43688|issue:45704", + "issue:39692|issue:43295", + "issue:44493|issue:46285", + "issue:43756|issue:45083", + "issue:43493|issue:43844", + "issue:43334|issue:43572", + "issue:43646|issue:44466", + "issue:45229|issue:45823", + "issue:33357|issue:33666", + "issue:43334|issue:45759", + "issue:43927|issue:44361", + "issue:43824|issue:44589", + "issue:43504|issue:43606", + "issue:43278|issue:43844", + "issue:44589|issue:45520", + "issue:44625|issue:45759", + "issue:43257|issue:43299", + "issue:43688|issue:45701", + "issue:43572|issue:44291", + "issue:43575|issue:45405", + "issue:43606|issue:44112", + "issue:43504|issue:43653" + ], + "prepared_review_unit_hash": "1f891236216bc87b152740ad13992372c8691da729a147150612226dc7cc0798", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36896, + "estimated_input_tokens": 9224, + "estimated_eval_tokens": 18704 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The pairs are mostly adjacent regressions in similar subsystems, but none look like true duplicates: they differ in the concrete failure mode or code path. I\u2019d anchor the set on the Granite tokenizer regression issue as the cleanest concrete bug report.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45812 is the cleanest representative concrete regression here: it is user-facing, reproducible, and narrowly describes a wrong-tokenization failure with a clear scope.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the strongest anchor because it is specific enough to triage and general enough to represent a real class of tokenizer regressions without being overly broad.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model-loading regression vs a dependency version bump; different underlying problems." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video-input regression vs num_labels propagation; not the same bug." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization vs FSDP CPU-efficient loading; unrelated failure modes." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "dtype propagation issue vs MPS attention correctness bug." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Tokenizer special-token handling for one model vs broad Granite token-ID regression; not the same code-path." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Video processing ValueError vs flash-attention generation regression; different symptoms and fixes." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Default token-type IDs vs multi-GPU inference OOM; not duplicate defects." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Model loading failure vs offline model_info call in tokenizer patch; distinct issues." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Aux-loss normalization bug vs T5 RMSNorm memory leak; unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Example/documentation errors vs processor.tokenizer API regression; related subsystem, different bug." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id keys vs CLIP class structure change; different concrete problems." + }, + { + "left": "issue:43756", + "right": "issue:45083", + "accept": false, + "reason": "RoPE-layer dropping bug vs feature-extract length helper behavior." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 implementation mismatch vs ZeRO-3 gradient growth; unrelated." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Missing pad_token_id in Qwen3-VL vs missing pad_token_idx in StableLmConfig; same theme, different bug." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model init breakage vs inconsistent lm_head serialization; not the same issue." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "CUDA OOM vs device mismatch during forward; different failure modes." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training; unrelated." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_id vs composite-config unwrapping in from_config; different code paths." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "Config save/load token ID loss vs tokenizer AttributeError on tasks; different bugs." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "ImportError for a model class vs float8 storage lookup failure; unrelated." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy-field load failure vs CPU offload device mismatch; different model paths." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Embedding dtype drift vs abnormal gradient growth under ZeRO-3; not duplicates." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage deserialization error vs flash_attn KeyError on Python 3.13; different causes." + }, + { + "left": "issue:44625", + "right": "issue:45759", + "accept": false, + "reason": "num_labels propagation vs text_config unwrapping; both config-related but not the same bug." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Accelerate/DeepSpeed conversion issue vs Qwen3VL MoE loading regression; different code paths." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Aux-loss normalization vs tokenization changes; unrelated." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing config field vs init_empty_weights argument mismatch; different regressions." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Tensor-parallel OOM vs unreleased dependency version bump; not duplicate." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "Runtime device mismatch vs stale CI test; one is a product bug, the other a test issue." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Beit legacy-field load bug vs BigBirdTokenizer special-token decode bug; different models and failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 23, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:33357", + "issue:34567", + "issue:38175", + "issue:39692", + "issue:43232", + "issue:43278", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43688", + "issue:43701", + "issue:43844", + "issue:43873", + "issue:43994", + "issue:44186", + "issue:44423", + "issue:44464", + "issue:44493", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44734", + "issue:44792", + "issue:44849", + "issue:44964", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45356", + "issue:45405", + "issue:45406", + "issue:45446", + "issue:45520", + "issue:45584", + "issue:45588", + "issue:45823", + "issue:46050", + "issue:46121", + "issue:46192" + ], + "soft_pairs": [ + "issue:44792|issue:45405", + "issue:44423|issue:44734", + "issue:44464|issue:45584", + "issue:45588|issue:46050", + "issue:43994|issue:44493", + "issue:29942|issue:30333", + "issue:43844|issue:44964", + "issue:28218|issue:34567", + "issue:43688|issue:43873", + "issue:29159|issue:32090", + "issue:44521|issue:45290", + "issue:44554|issue:45446", + "issue:15354|issue:38175", + "issue:45084|issue:45520", + "issue:43540|issue:45200", + "issue:43873|issue:46050", + "issue:45200|issue:45823", + "issue:43278|issue:43701", + "issue:45072|issue:45198", + "issue:43232|issue:44849", + "issue:44186|issue:46192", + "issue:45406|issue:46121", + "issue:28218|issue:33357", + "issue:45823|issue:46192", + "issue:43493|issue:43701", + "issue:43295|issue:43450", + "issue:43295|issue:44610", + "issue:39692|issue:46192", + "issue:44554|issue:45356" + ], + "prepared_review_unit_hash": "53db50bab84aeef608aa72e5aba0b3f01150d64547d79eca33ca8120fb6bdb7c", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36484, + "estimated_input_tokens": 9121, + "estimated_eval_tokens": 18498 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a loose thematic cluster, not a true duplicate set. Most soft pairs are only related at a broad subsystem level (tokenizers, processors, generation, quantization, serving) and do not describe the same concrete bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out: the cluster spans many unrelated failures. If one must be chosen as the closest representative, issue:43295 is the broadest regression-style report in core multimodal processor/tokenizer behavior.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:43295 is the best representative among an otherwise heterogeneous set because it describes a user-facing regression in core processor/tokenizer handling and is broader than the highly specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "Janus image-generation test failure and a PEFT version pin issue are unrelated bugs." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both are serve-related crashes, but one is continuous batching multimodal input handling and the other is KV-cache continuation indexing in /v1/responses; different code paths." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation with compiled forward and Whisper empty-transcription generation after token alignment are different generation bugs." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention sinkless-model None handling and sinq save/load quantization breakage are unrelated." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 bad outputs and position-id key errors are not the same underlying issue." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures and MLFlow job state reporting are unrelated." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "ZeRO-3 gradient growth and Phi-4 multimodal loading are different training/loading problems." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer spacing after added tokens and TrainerState token counter updates are unrelated." + }, + { + "left": "issue:43688", + "right": "issue:43873", + "accept": false, + "reason": "Aux-loss normalization in MoE models and offloading with quantization are different bugs." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency and NoneType GPU broadcast errors are unrelated." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both involve apply_chat_template, but one is zero assistant masks for multimodal inputs and the other is a crash on tool-call messages with no content." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness and a PyTorch version check in flex_attention are unrelated." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript export failure for ViT and SigLIP2 zero probabilities are unrelated." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template-node compilation TypeError and flash_attn import KeyError on Python 3.13 are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Qwen3OmniMoe video processing failure and Gemma4 text-only mm_token_type_ids handling are different multimodal issues." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Quantization offloading behavior and sinq save/load corruption are different quantization problems." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 text-only token-type defaults and device-map forward device mismatch are separate bugs." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "BF16-to-FP32 embedding dtype drift and resume-from-checkpoint key mismatch are unrelated." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "bfloat16 dtype mismatches and Wav2Vec2 save/tokenization failures are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs after sync_gpus and Qwen3.5 hidden-states output are different generation bugs." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "LayoutLMv2 NER/padding crashes and return_tensors being ignored with text_kwargs are different tokenizer API bugs." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor _tokenizer attribute crash and rope-parameter list handling are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer spacing behavior and a MacOS bus error on CLIP are unrelated." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device placement and return_tensors handling are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and checkpoint resume key mismatch are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression and batched video shape errors are different multimodal processor bugs." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Processor/tokenizer regression and OmDet-Turbo image-size mismatch are not the same bug." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and return_tensors handling are unrelated." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness and Kimi-K2.5 tokenizer codec regression are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 24, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:37428", + "issue:38175", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43450", + "issue:43493", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43653", + "issue:43696", + "issue:43824", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43950", + "issue:43957", + "issue:43975", + "issue:44112", + "issue:44442", + "issue:44466", + "issue:44514", + "issue:44521", + "issue:44561", + "issue:44898", + "issue:44987", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45325", + "issue:45381", + "issue:45412", + "issue:45520", + "issue:45759", + "issue:46170", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285" + ], + "soft_pairs": [ + "issue:45381|issue:46209", + "issue:43295|issue:43873", + "issue:43572|issue:44987", + "issue:43950|issue:46285", + "issue:43572|issue:43957", + "issue:44898|issue:45412", + "issue:43550|issue:43844", + "issue:44466|issue:45127", + "issue:45200|issue:45325", + "issue:43450|issue:43575", + "issue:45520|issue:46178", + "issue:43334|issue:43525", + "issue:37428|issue:43957", + "issue:42490|issue:43696", + "issue:43824|issue:43975", + "issue:43278|issue:43856", + "issue:43526|issue:46192", + "issue:38175|issue:43493", + "issue:43653|issue:44112", + "issue:44442|issue:45198", + "issue:42915|issue:43824", + "issue:43295|issue:43493", + "issue:38175|issue:45412", + "issue:43262|issue:43450", + "issue:44514|issue:44521", + "issue:45759|issue:46285", + "issue:37428|issue:43824", + "issue:45520|issue:46170", + "issue:41720|issue:46209", + "issue:44561|issue:46234" + ], + "prepared_review_unit_hash": "cff56ff462b9df5133188fc6dd6ed52d3350c4acc7783af1083faadbad1a7f0f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36452, + "estimated_input_tokens": 9113, + "estimated_eval_tokens": 18482 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated Transformers regression reports. The only likely duplicate-like pair is the Qwen3-VL / Llama4 missing `pad_token_id` config failure; the rest are separate bugs across different models/subsystems. No PRs are present.", + "confidence": 0.72, + "canonical_issue_reason": "issue:43334 is the best anchor for the only plausible duplicate pair: it matches 43525 in the same `pad_token_id`-missing model-load failure pattern on new config classes.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43334 is the clearest representative of the one near-duplicate theme and is described as a concrete load failure with a specific exception path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Different failures: Qwen2.5-VL vision position ids vs FP8 inference OOM on B200." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression vs quantization offloading behavior; unrelated code paths." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "StableLm `pad_token_idx` issue is unrelated to a separate model-loading failure for `physical-intelligence/fast`." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Non-persistent buffer corruption on load is unrelated to CLIP text model structure inconsistency." + }, + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "Different regressions: missing pad token field vs meta-device loading breakage." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver interpolation failure and RT-DETR memory release bug are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "torch.compile SDPA failure for Bamba and DeepSpeed gradient growth are distinct issues." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Serialization of tied weights vs LoRA merge collapse; similar theme but not the same bug." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Different multimodal bugs: missing default mm token ids vs wrong temporal rope scaling." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Video processor batch-shape bug and tensor-parallel OOM are unrelated." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Both are import/backend errors, but for different optional dependencies and different failure modes." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": true, + "reason": "Both report the same `AttributeError`/missing `pad_token_id` regression on new config classes during model loading." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash-attention import error and meta-device loading breakage are separate regressions." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Trailing-slash import warning is unrelated to GPT-oss OOM." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "ImportError for `Qwen2_5_VLForConditionalGeneration` is unrelated to detokenization behavior in deepseek-coder." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Embedding dtype mismatch and Qwen3 MoE memory inefficiency are different problems." + }, + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "Beit `reduce_labels` bug is unrelated to `return_tensors` being ignored with `text_kwargs`." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": false, + "reason": "Siglip2 zero-probabilities issue is not the same as the JAX/HF implementation discrepancy." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Tokenizer special-token registration and GraniteSpeech CI test failure are unrelated." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "AutoTokenizer load failure and Wav2Vec2 save/tokenization failure affect different tokenization paths." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3Moe FP8 failure and Qwen2.5-VL import error are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "Processor/tokenizer regression and Siglip2 implementation discrepancy are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Siglip2 output issue is unrelated to RT-DETR memory cleanup." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate default and video batch-shape bug are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Same model family, but one is a batching crash and the other is wrong assistant masks; not the same concrete bug." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Composite Qwen config unwrapping and CLIP structure inconsistency are different model-loading changes." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Missing flash-attention helper import and missing Qwen2.5-VL export are unrelated import errors." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Python 3.13 flash_attn backend lookup and Windows ROCm torch.distributed import failure are separate platform-specific issues." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3 auto device map assert and FP8 inference OOM are different runtime failures." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Missing `is_torch_fx_available` and missing `is_falcon_mamba_ssm_available` are different import-utils symbols and regressions." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The one accepted duplicate pair (issue:43334 / issue:43525) is well supported by the shared missing-`pad_token_id` load failure pattern, and the rejected soft edges are generally reasoned from the issue titles without overclaiming. No PRs are present, so that part is also consistent." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 25, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28218", + "issue:29942", + "issue:32090", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:43278", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43901", + "issue:43931", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44448", + "issue:44493", + "issue:44610", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44857", + "issue:45083", + "issue:45127", + "issue:45200", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:46234" + ], + "soft_pairs": [ + "issue:44610|issue:44857", + "issue:43819|issue:44112", + "issue:43827|issue:43901", + "issue:43819|issue:45083", + "issue:44206|issue:45584", + "issue:43335|issue:43550", + "issue:44857|issue:45593", + "issue:43606|issue:45083", + "issue:44792|issue:45563", + "issue:37428|issue:43425", + "issue:45127|issue:45245", + "issue:44448|issue:44743", + "issue:45200|issue:45588", + "issue:43335|issue:43653", + "issue:43611|issue:43931", + "issue:45290|issue:45381", + "issue:44855|issue:45083", + "issue:43550|issue:44112", + "issue:44206|issue:45823", + "issue:43819|issue:44493", + "issue:43335|issue:43504", + "issue:29942|issue:32090", + "issue:45588|issue:45859", + "issue:37428|issue:44351", + "issue:43278|issue:43825", + "issue:45459|issue:46234", + "issue:39692|issue:43526", + "issue:43335|issue:45083", + "issue:43388|issue:43526", + "issue:45310|issue:45478", + "issue:39692|issue:44493", + "issue:28218|issue:35141", + "issue:43720|issue:44112", + "issue:43335|issue:44112" + ], + "prepared_review_unit_hash": "887832fce16cddb7411f7665c4bf498212b0a7e084f1fb30b352111a682354ca", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 38127, + "estimated_input_tokens": 9532, + "estimated_eval_tokens": 19320 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous; the only clear duplicate pair is the Qwen3.5 Moe `from_pretrained` error. Everything else looks like separate bugs, regressions, or docs issues in different code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 45310 is the strongest canonical candidate because it matches 45478 on the same Qwen3.5 Moe `from_pretrained` error title, has an inbound reference, and has a slightly longer body with more context.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45310 is the best representative issue in this cluster because it is the confirmed duplicate target for 45478, while also providing slightly more context and an inbound reference.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different models and failures: OmDet-Turbo processor input-size mismatch vs LwDetrImageLoss AMP crash." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "Both are model/runtime bugs, but the DAC STE mismatch and GraniteSpeech stale device override are unrelated code paths." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs/pipeline related, but one is about removed `pipeline()` usage and the other about `return_all_scores` docs for text classification." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "Unrelated subsystems: DAC latent reconstruction vs qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Both involve audio-ish workflows, but they describe distinct regressions: unsupported `center` arg vs empty transcription after special-token alignment." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "Different models and failures: SwitchTransformers sparse-layer construction vs Bamba torch.compile SDPA behavior." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different detection/training behaviors: AMP crash in LwDetrImageLoss vs missing auxiliary losses in D-FINE." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch in Bark is unrelated to qwen3_omni_moe feature-length calculations." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "A failing Janus image-generation test is not the same bug as a stale `num_return_sequences` warning in paged generate()." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "ImportError for a flash-attention helper and Torch 2.10 incompatibility are different release/runtime issues." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge/collapsed outputs and the 2^24 category limit are unrelated model/data constraints." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Different code paths: Pegasus v4/v5 output drift vs recurrent state reset in modular_qwen3_5." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 text-only token-type defaults and flash_attention `s_aux=None` handling are separate bugs." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and BigBirdTokenizer special-token registration are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43931", + "accept": false, + "reason": "Base-model-prefix loading regression and Qwen3-VL weight-shape mismatch are different model-loading failures." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template tool-call handling and Qwen2.5-VL vision position IDs are separate issues." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 torch.jit indentation parsing and qwen3_omni_moe feature-length helper behavior are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Bamba torch.compile SDPA failure is not the same as a stale GraniteSpeech CI device override test." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Both are device/feature regressions, but one is a feature-extractor arg issue and the other is a Gemma4 PLE forward device mismatch." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC latent mismatch and position-id key warnings are unrelated model internals." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "SwitchTransformers sparse-layer bug and Beit pretrained preset loading bug are different model families and failures." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 test failures and Trainer `_gpu_broadcast_one` NoneType errors do not describe the same underlying bug." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Flash-attention sinkless-model crash and Qwen3.5 Moe CPU-offload incompatibility are different runtime problems." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Missing `_flash_supports_window_size` import and missing `HybridCache` import are separate API/export regressions." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Embedding dtype drift between train/eval and a pipeline v5 translation message regression are unrelated." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Tokenizer protobuf error masking and a missing `is_falcon_mamba_ssm_available` utility are separate import-utils issues." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "SigLIP2 docs errors and BeitImageProcessorFast `reduce_labels` behavior are different concerns." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Metric gathering label truncation and BeitImageProcessorFast label reduction are not the same bug." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Exact same Qwen3.5 Moe `from_pretrained` error on transformers>=5.4.0; 45478 is a duplicate report." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 doc issues and position-id key warnings are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer added-space behavior and embedding reinitialization on resize are separate tokenizer/model-init bugs." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading failure and GraniteSpeech stale device test failure are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformers sparse-layer configuration and GraniteSpeech CI device override do not share the same underlying defect." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The duplicate-pair call on 45310/45478 looks correct and the rest of the soft-edge rejections are conservative. However, the justification overreaches when it says 45310 has the \"richest thread\" among the pair: the packet shows 45478 has more discussion activity (4 vs 2), so that comparative claim is not grounded. Tighten the representative/canonical rationale to only the supported facts (same title, inbound reference, slightly longer body) and avoid unsupported thread-quality claims." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 26, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:25251", + "issue:29127", + "issue:29942", + "issue:32090", + "issue:34689", + "issue:35141", + "issue:38175", + "issue:40444", + "issue:42490", + "issue:43278", + "issue:43335", + "issue:43381", + "issue:43412", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43611", + "issue:43701", + "issue:43746", + "issue:43854", + "issue:43927", + "issue:44360", + "issue:44448", + "issue:44466", + "issue:44493", + "issue:44561", + "issue:44568", + "issue:44779", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45593", + "issue:45698", + "issue:45803", + "issue:45823", + "issue:45920", + "issue:46192", + "issue:46209", + "issue:46285" + ], + "soft_pairs": [ + "issue:44448|issue:44779", + "issue:43335|issue:45593", + "issue:42490|issue:43854", + "issue:43606|issue:44857", + "issue:45920|issue:46192", + "issue:44360|issue:44561", + "issue:38175|issue:44493", + "issue:44857|issue:45072", + "issue:44977|issue:45229", + "issue:44448|issue:44977", + "issue:34689|issue:35141", + "issue:43701|issue:45083", + "issue:29127|issue:32090", + "issue:43550|issue:45083", + "issue:43927|issue:45698", + "issue:40444|issue:43526", + "issue:29942|issue:35141", + "issue:42490|issue:43575", + "issue:44561|issue:44779", + "issue:44977|issue:45563", + "issue:44568|issue:45491", + "issue:44857|issue:45588", + "issue:43746|issue:46209", + "issue:45372|issue:45588", + "issue:43412|issue:45588", + "issue:43278|issue:43381", + "issue:43611|issue:46285", + "issue:45198|issue:45803", + "issue:45588|issue:45823", + "issue:44448|issue:44466", + "issue:25251|issue:32090", + "issue:45245|issue:45563" + ], + "prepared_review_unit_hash": "2b5a0ae93dd148847c34cbf4db58d4db58676dd16f47d07b895170d00ef58e5e", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37485, + "estimated_input_tokens": 9372, + "estimated_eval_tokens": 19000 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue set with a few superficial v5/regression and tokenizer/load-save themes; no true duplicate pair stands out.", + "confidence": 0.32, + "canonical_issue_reason": "Issue 44448 is the broadest representative of the v4/v5 behavior-regression theme and is one of the more centrally connected items in the soft-pair graph.", + "canonical_pr_reason": null, + "best_issue_reason": "44448 is the best single issue anchor for this cluster because it reflects an observable version-regression in model behavior, while most other issues are narrower component-specific bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both mention v4/v5 regressions, but one is Pegasus generation drift and the other is DeepSeek tokenizer IDs; different code paths." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Sparse-layer config bug vs D-FINE auxiliary-loss omission; unrelated models and failures." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "Local-path trailing-slash import warning vs GLM-4.7 test load failure; different symptoms." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU-offload device mismatch vs FP16 CUDA crash in a loss module; not the same bug." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Both are tokenization-related, but one is wrong IDs across models and the other is ignored return_tensors kwargs." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA indexer activation omission vs v5 removal of is_torch_fx_available; no shared failure mode." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 zero probabilities and unexpected position-id keys are unrelated inference issues." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype problems, but different models, ops, and failure points." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Flash-attention generation bug vs multi-GPU OOM; distinct runtime issues." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Both are v5 generation regressions, but Pegasus output drift and Qwen flash-attention failure are different paths." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Model loading breakage for Llama 3.2 vs embedding reinit after resize; unrelated." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch and a feature-length helper bug are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error messaging vs Trainer broadcast NoneType; no overlap." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is separate from qwen3_omni_moe length calculation." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Both concern save/load, but one loses custom token IDs and the other loads the wrong custom module." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "IterableDataset multi-image finetuning failure vs BEiT label reduction bug." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures vs output embedding reinit after resize." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Trailing-slash import warning vs tp OOM on Qwen2-57B; unrelated." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Trust_remote_code breakage from API removal vs DeepSeek tokenizer regression; different root causes." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation issue vs stale warning for num_return_sequences." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Special tokens omission in tokenizer vs Gemma3 NaNs from sliding-window attention." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "Different crash sites: LwDetr AMP loss vs flash_attention.py s_aux=None." + }, + { + "left": "issue:43746", + "right": "issue:46209", + "accept": false, + "reason": "PEFT checkpoint loading vs FP8 inference on B200; unrelated." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 processor import failure vs flash_attention.py None aux crash." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export guard/erfinv issue vs flash_attention None aux crash." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Eval/train dtype drift vs gradient checkpointing disallowed in eval; not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:46285", + "accept": false, + "reason": "Base_model_prefix loading regression vs CLIPTextModel structure mismatch; distinct model-loading issues." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure vs Fuyu bare-except bug." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "flash_attention s_aux=None crash vs Gemma4 device_map mismatch; different components." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "One is output drift between v4/v5, the other is lm_head serialization depending on device." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Top_k pipeline output nesting bug vs GPU broadcast NoneType error." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category limit error vs stale generate warning." + } + ] + }, + "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": 18, + "review_unit_index": 27, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:29127", + "issue:29942", + "issue:30064", + "issue:33357", + "issue:34567", + "issue:36010", + "issue:37428", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43066", + "issue:43278", + "issue:43381", + "issue:43421", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43873", + "issue:43976", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44361", + "issue:44442", + "issue:44462", + "issue:44479", + "issue:44610", + "issue:44734", + "issue:44743", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45229", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45491", + "issue:45712", + "issue:46121" + ], + "soft_pairs": [ + "issue:43653|issue:44743", + "issue:36010|issue:37428", + "issue:45459|issue:45712", + "issue:43065|issue:45412", + "issue:44186|issue:45198", + "issue:43611|issue:46121", + "issue:44361|issue:45198", + "issue:33357|issue:39692", + "issue:44871|issue:44977", + "issue:29127|issue:34567", + "issue:45372|issue:45459", + "issue:44206|issue:44479", + "issue:29942|issue:30064", + "issue:42915|issue:42994", + "issue:43976|issue:45440", + "issue:43994|issue:45412", + "issue:43278|issue:43575", + "issue:43066|issue:44462", + "issue:44265|issue:45198", + "issue:10000|issue:15354", + "issue:42490|issue:43873", + "issue:43873|issue:45229", + "issue:43421|issue:45491", + "issue:43381|issue:43701", + "issue:44734|issue:45406", + "issue:43065|issue:43994", + "issue:43653|issue:44442", + "issue:44610|issue:45412", + "issue:41720|issue:45229" + ], + "prepared_review_unit_hash": "4b1ccfadf8ee22363058be0af29cee1d90a6b90d40a5e442d4558b0bb25a8278", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36437, + "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 highly heterogeneous: most items are unrelated model-, tokenizer-, quantization-, serving-, or environment-specific bugs rather than duplicates of one underlying defect. No strong duplicate canonical emerges.", + "confidence": 0.97, + "canonical_issue_reason": "No single issue is a clear canonical target; the cluster spans many unrelated failure modes across different models and subsystems.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a strong global representative. If forced, 43873 is the broadest/general discussion, but it is still not a duplicate hub.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Tokenizer special-token registration/empty decode is unrelated to Qwen recurrent-state cache reset behavior." + }, + { + "left": "issue:36010", + "right": "issue:37428", + "accept": false, + "reason": "These are different import errors in different modules with no shared underlying code-path." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Hiding tokenizer errors when protobuf is missing is unrelated to leftover dummy classes leaking into dir(transformers)." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory not being released on deletion." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Both involve tokenization, but they affect different models and different failure modes; not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "base_model_prefix loading breaks are unrelated to rope-validation TypeError on a JSON-loaded list." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLukeTokenizer AttributeError is a different tokenizer issue from Wav2Vec2 save/tokenization failure." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "Both mention vision models, but the concrete bugs differ: MacOS bus error vs docs/model-processor mismatch and quantization failure." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error handling for boxes and TrainerState token counting are unrelated subsystems." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma 4 processor loading blocked by a mistral_common import is not the same as tokenizer error masking without protobuf." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Unsupported feature-extractor argument regression and video-input regression are different code paths and models." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to image processor void-segmentation-map handling." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "FP8 quantization config failure on load and quantized model saving failure are distinct bugs." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python version compatibility regression is unrelated to DeepSeekV3 implementation divergence." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 nonsense outputs and RT-DETR memory leaks are unrelated issues." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Embedding dtype drift in eval and Qwen2 OOM under tensor parallelism are different problems." + }, + { + "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 from torch_compilable_check are unrelated to Wav2Vec2 save/tokenization failures." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "German DistilBertModel and GeneratorExp/TorchScript export are unrelated model/framework issues." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "A trailing-slash import warning is unrelated to quantization offloading not working." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Offloading/quantization behavior is not the same as multi-GPU CUDA OOM for Gemma4." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime post-processor updates in TokenizersBackend are unrelated to Gemma3 NaN embeddings with mixed-length batching." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Eval-mode gradient checkpointing and resume-from-checkpoint key mismatch are different training-flow bugs." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation tensor indexing in serve is unrelated to Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "A dummy layer in Sam3PixelDecoder is unrelated to SigLIP2 wrong/nonsensical outputs." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer special-token registration is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + }, + { + "left": "issue:44610", + "right": "issue:45412", + "accept": false, + "reason": "Processor image-size mismatch and RT-DETR memory leaks are different bugs." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3 auto device mapping cudaErrorAssert and Gemma4 multi-GPU OOM are different failure modes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 28, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:35141", + "issue:36010", + "issue:38175", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43278", + "issue:43329", + "issue:43334", + "issue:43381", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43575", + "issue:43646", + "issue:43701", + "issue:43723", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43931", + "issue:44077", + "issue:44186", + "issue:44206", + "issue:44279", + "issue:44387", + "issue:44442", + "issue:44554", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45103", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45440", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45834", + "issue:45859" + ], + "soft_pairs": [ + "issue:32090|issue:33357", + "issue:43756|issue:45440", + "issue:43278|issue:43329", + "issue:45372|issue:45406", + "issue:43526|issue:45834", + "issue:29942|issue:36010", + "issue:43723|issue:44554", + "issue:43761|issue:45859", + "issue:45440|issue:45593", + "issue:45103|issue:45372", + "issue:43526|issue:44186", + "issue:43575|issue:43701", + "issue:43575|issue:44387", + "issue:42490|issue:43381", + "issue:44977|issue:45245", + "issue:43526|issue:44857", + "issue:43334|issue:43531", + "issue:43824|issue:44279", + "issue:44442|issue:45072", + "issue:29127|issue:35141", + "issue:43646|issue:44077", + "issue:44743|issue:44811", + "issue:42915|issue:43931", + "issue:43262|issue:44206", + "issue:43452|issue:45588", + "issue:44977|issue:45584", + "issue:33357|issue:38175" + ], + "prepared_review_unit_hash": "472958a73cff785d13966d21fb63f26b556bcbc72c8fc144b9c2a05ddcf878a0", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35308, + "estimated_input_tokens": 8827, + "estimated_eval_tokens": 17910 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the candidate pairs mostly share only broad subsystem vocabulary, not the same underlying bug. I would not merge any of the soft edges. If a single representative issue is needed, issue 43761 is the cleanest, most concrete regression report in the set.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 43761 is the most self-contained and clearly scoped report here, with a specific regression, concrete expected behavior, and an explicit failure condition/version context.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43761 is a strong representative because it describes one precise code-path failure rather than a broad or speculative symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Different failures: Trainer GPU broadcast type error vs MacOS bus error loading a CLIP model." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Both concern model implementation correctness, but they target different models and different bugs." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Unrelated areas: embedding dtype drift vs multimodal token counting video-branch NameError-style bug." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma 4 loading problems, but one is a dependency import failure and the other is a missing processor attribute in serve." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "Both are vision/OCR-adjacent, but the concrete failures are different and do not share a code path." + }, + { + "left": "issue:29942", + "right": "issue:36010", + "accept": false, + "reason": "Flash Attention 2 test failures are not the same as a GenerationMixin import error." + }, + { + "left": "issue:43723", + "right": "issue:44554", + "accept": false, + "reason": "Tokenizer-loading issue and MPS attention correctness bug are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIP vision hidden-state regression is unrelated to Qwen rotary embedding CPU offload incompatibility." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "Different models and different defects: remote implementation divergence vs missing auxiliary losses." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "Auto-docstring annotation handling crash is unrelated to the mistral_common ReasoningEffort import error." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "Different components and failures: BEiT image processor label reduction vs LayoutLMv2 tokenizer padding/NER crash." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "Tensor-parallel OOM and resume checkpoint key mismatch are unrelated issues." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both mention OOM, but one is model-load TP memory pressure and the other is int4 quantization reserved-memory behavior." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Local-model slash-path import warning is unrelated to gradient checkpointing in eval mode." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation bug and category-cardinality runtime error are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "Different image-loss/processor bugs with no shared concrete code path." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Both involve Qwen3/Qwen3-MoE, but one is a missing pad_token_id on load and the other is a sliding-window runtime issue." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "ImportError for a specific model class is not the same as a general dependency issue." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer loading failure and dtype mismatch in inference are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Different model subsystems and different symptoms: clearer error messaging vs output embedding reinitialization." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Transformers 5.0 custom initialization breakage is broader and distinct from patchtsmixer's optional post_init restriction." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Qwen recurrent-state caching bug is unrelated to Whisper batch_decode special-token handling." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "FineGrainedFP8Config failure and Qwen3-VL weight-shape mismatch are different loading problems." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sampling-rate default mismatch is unrelated to an unsupported center argument crash." + }, + { + "left": "issue:43452", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer/model loading with gguf_file has no concrete overlap with flash_attention s_aux=None handling." + }, + { + "left": "issue:44977", + "right": "issue:45584", + "accept": false, + "reason": "Flash-attention generation issue is unrelated to Whisper empty-transcription failure." + }, + { + "left": "issue:33357", + "right": "issue:38175", + "accept": false, + "reason": "Both are vision-model symptoms, but one is a MacOS bus error and the other is unexpected zero probabilities." + } + ] + }, + "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": 18, + "review_unit_index": 29, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28218", + "issue:32090", + "issue:33290", + "issue:34567", + "issue:34689", + "issue:36010", + "issue:42175", + "issue:42907", + "issue:43316", + "issue:43381", + "issue:43493", + "issue:43525", + "issue:43550", + "issue:43611", + "issue:43643", + "issue:43720", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44568", + "issue:44743", + "issue:44855", + "issue:44871", + "issue:44898", + "issue:44913", + "issue:44964", + "issue:45137", + "issue:45216", + "issue:45290", + "issue:45440", + "issue:45479", + "issue:45491", + "issue:45698", + "issue:45759", + "issue:45803", + "issue:45907", + "issue:46209", + "issue:46234" + ], + "soft_pairs": [ + "issue:44062|issue:45907", + "issue:44464|issue:45491", + "issue:28218|issue:32090", + "issue:33290|issue:34567", + "issue:44442|issue:44898", + "issue:43381|issue:43873", + "issue:44964|issue:45440", + "issue:43316|issue:43525", + "issue:34689|issue:43611", + "issue:32090|issue:33290", + "issue:43720|issue:44442", + "issue:44479|issue:45290", + "issue:43931|issue:45216", + "issue:34689|issue:36010", + "issue:44442|issue:45803", + "issue:43550|issue:44743", + "issue:43493|issue:43827", + "issue:43873|issue:44871", + "issue:42907|issue:45698", + "issue:42175|issue:46234", + "issue:43761|issue:44188", + "issue:43643|issue:44913", + "issue:43316|issue:45759", + "issue:45290|issue:46209", + "issue:44855|issue:45137", + "issue:33290|issue:34689", + "issue:44190|issue:44496", + "issue:44568|issue:45479", + "issue:44077|issue:44479" + ], + "prepared_review_unit_hash": "3e4d3e965a1bc3d1f47fd0042240d080f1bfd199076b5a3936b8d7fb1cea13ab", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 36145, + "estimated_input_tokens": 9037, + "estimated_eval_tokens": 18330 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated Transformers bug reports that only share broad topical similarity (loading, configs, generation, quantization). I found no soft pair that looks like the same underlying issue, so none should be merged as duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "No clear duplicate cluster exists here. If one issue must represent the set, issue:45698 is the broadest and most active loading/serialization regression report, with multiple inbound refs and high discussion activity.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:45698 is the best single representative because it is active, well-discussed, and concerns a core save/load path that is plausibly adjacent to several other configuration/loading regressions in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "Completely different code paths: AddedToken construction vs Hub repo file listing/revision forwarding." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Both are generation/model-behavior bugs, but the concrete failures differ: compiled forward inconsistency vs Gemma3 sliding-window NaNs on mixed-length batches." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-space behavior is unrelated to the Trainer/NoneType GPU broadcast error." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "DeepSpeed/Adafactor OOM and num_input_tokens_seen not updating are unrelated trainer issues." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver interpolation failure are unrelated model-specific bugs." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction and quantization offloading failure are different mechanisms." + }, + { + "left": "issue:44964", + "right": "issue:45440", + "accept": false, + "reason": "Separate multimodal model-loading regression vs DeepSeekV3MoE implementation divergence." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Both are config/API issues, but for different models and different missing fields; not the same bug." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Both involve version-related loading breakages, but one is a specific model loader regression and the other is a base_model_prefix handling change." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast NoneType error and DeepSpeed Adafactor OOM do not share a concrete failure mode." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight loading and FastSpeech2 tokenizer loading are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45290", + "accept": false, + "reason": "Video-input regression across Qwen VL models is unrelated to chat-template tokenization crashing on tool-call messages." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Model shape mismatch on load and incorrect checkpoint saving are separate serialization/load-path problems." + }, + { + "left": "issue:34689", + "right": "issue:36010", + "accept": false, + "reason": "Both mention loading/import breakage, but one is a model-load regression and the other is a GenerationMixin import path issue." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer load failure and Fuyu tensor conversion exception swallowing are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44743", + "accept": false, + "reason": "torch.compile SDPA failure in Bamba and recurrent-state reset in modular_qwen3_5 are different model/runtime bugs." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation mismatch and docs still mentioning pipeline() are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading behavior and Gemma3 eos_token_id mismatch are not the same underlying issue." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Both touch save/load flows, but one is about saving dequantized Ministrals/Devstrals and the other about loading the wrong custom module after save_pretrained; too different to merge." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Backend dependency installation and a missing import_utils helper are unrelated." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and attention-kernel divergence under torch.compile are different bugs." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "Missing fields with trust_remote_code and rotary_pct not persisting on reload are distinct config-handling problems." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Qwen multimodal text_config unwrapping are related only at a very high level; they are not the same concrete defect." + }, + { + "left": "issue:45290", + "right": "issue:46209", + "accept": false, + "reason": "Chat-template tool-call crash is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 Deberta import parsing error and DeepSpeed ZeRO3 deque IndexError are unrelated." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Adafactor/DeepSpeed OOM and Llama 3.2 loading regression are unrelated." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading in an example script and unrecognized model type in config.json are different failures." + }, + { + "left": "issue:44568", + "right": "issue:45479", + "accept": false, + "reason": "Tokenizer add_special_tokens regression and sequence-classification zero-loss bug are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "PatchTSMixer post_init validation and Qwen VL video-input regression are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading behavior and Gemma3 token-ID inconsistency 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": 18, + "review_unit_index": 30, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:25251", + "issue:28282", + "issue:29127", + "issue:33290", + "issue:33357", + "issue:35141", + "issue:41950", + "issue:42617", + "issue:43065", + "issue:43232", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43475", + "issue:43540", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43650", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43761", + "issue:43825", + "issue:44162", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44617", + "issue:44792", + "issue:44871", + "issue:44877", + "issue:44913", + "issue:44987", + "issue:45127", + "issue:45200", + "issue:45584", + "issue:45656", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46202" + ], + "soft_pairs": [ + "issue:43650|issue:46202", + "issue:44987|issue:46178", + "issue:29127|issue:33357", + "issue:25251|issue:28282", + "issue:43295|issue:43825", + "issue:43550|issue:43701", + "issue:42617|issue:43575", + "issue:43606|issue:43701", + "issue:33290|issue:35141", + "issue:43575|issue:43742", + "issue:43540|issue:44514", + "issue:33290|issue:45656", + "issue:43316|issue:43716", + "issue:44871|issue:45584", + "issue:44162|issue:44792", + "issue:43404|issue:43441", + "issue:43316|issue:44877", + "issue:43295|issue:44871", + "issue:45803|issue:46170", + "issue:43065|issue:44617", + "issue:41950|issue:43475", + "issue:43646|issue:44913", + "issue:44871|issue:46050", + "issue:43452|issue:44496", + "issue:45127|issue:45200", + "issue:43761|issue:44479", + "issue:43232|issue:43388" + ], + "prepared_review_unit_hash": "68d17b440e64b82f630c97bfa68be107d12449be8bc1803168485d46ea548e43", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35132, + "estimated_input_tokens": 8783, + "estimated_eval_tokens": 17822 + }, + "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; the soft links are mostly title-level similarity only, so none should be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": "If a canonical has to be chosen, issue:44871 is the least-bad anchor because it is concrete, reproducible, and appears in multiple soft-link candidates, but the overall cluster is too heterogeneous to support a strong canonical.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:44871 is the most actionable representative here: a specific Gemma config mismatch with clear symptoms, unlike the broader or more ambiguous neighboring issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "Unrelated: a placeholder/open-ended issue versus a Gemma pipeline roundtrip bug around a special token." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Different failure modes and code paths: model loading for one repo/model versus backend registration for tensorflow_text." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "Distinct bugs: LayoutLMv3 box-content error versus a MacOS bus error in CLIP community model loading." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Not the same bug: pipeline top_k output nesting versus missing PyTorch import for AutoModel." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "Both are regressions, but one is custom model code accessing processor.tokenizer while the other is a pipeline error-message/translation-support issue." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "Completely different problems: torch.compile/SDPA OOM versus resume_from_checkpoint key mismatch." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "Different symptoms and likely different paths: a 3d_parallel.py runtime issue versus tensor-parallel OOM on a specific Qwen model." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "Device mismatch with CPU offload is unrelated to checkpoint key mismatches." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Different bugs: DeepSpeed Adafactor OOM versus token embedding reinitialization during post_init." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Both are load-time issues, but one is tensor-parallel OOM and the other is a KeyError while loading MobileLLM." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Both involve multimodal/video processing, but the failures are different processor/code-path bugs and not clearly mergeable." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Related to DeepSpeed, but one is an OOM and the other is double optimizer stepping; not the same underlying defect." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Different models and different issues: Gemma3TextConfig API discrepancy versus image-preprocessor dtype mismatch in Mistral-3." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Both are tokenization/generation edge cases, but one is Gemma eos_token_id inconsistency and the other is Whisper empty-transcription handling after align_special_tokens." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "Unrelated: ESM2 breakage versus a Janus image-generation test failure." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Different model-specific regressions: lm_head weight tying versus FlashAttention failure in Ministral-3." + }, + { + "left": "issue:43316", + "right": "issue:44877", + "accept": false, + "reason": "Config-related issues, but not the same defect: API mismatch in Gemma3TextConfig versus strict config validation blocking granite_speech." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Both are regression-style model/processor bugs, but they affect different mechanisms: processor.tokenizer access versus Gemma eos_token_id configuration." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Unrelated: a bare except swallowing errors versus a Windows ROCm torch.distributed import/dependency problem." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Same product area (SAM3) but not the same issue: dummy Conv2d placeholder versus CUDA OOM." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Different subsystems and symptoms: video-classification pipeline processor selection versus missing fpn_position_embeddings on a SAM3 vision encoder output." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Both involve model init/config behavior, but one is a broad Transformers 5 custom-init breakage and the other is a GPTNeoX config field not persisting." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Both concern save/load or config persistence, but one is Gemma eos_token_id inconsistency and the other is quantized model round-tripping after save_pretrained." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Both are model-loading failures, but one is gguf_file handling and the other is a missing model_type in config.json." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "Different feature areas: LoRA merge with extended vocabulary versus Gemma 4 multimodal token type IDs defaulting." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "Different regressions in different models: CLIPVisionModel hidden_states behavior versus Qwen video-input handling." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Completely different code paths: generation kwargs sync handling versus gather_for_metrics dropping tuple labels." + } + ] + }, + "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": 18, + "review_unit_index": 31, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:35141", + "issue:41720", + "issue:42490", + "issue:43065", + "issue:43335", + "issue:43504", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43644", + "issue:43673", + "issue:43761", + "issue:43827", + "issue:43931", + "issue:43976", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44462", + "issue:44479", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44877", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45245", + "issue:45357", + "issue:45542", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45632", + "issue:45698", + "issue:45803", + "issue:45907", + "issue:45987", + "issue:46178", + "issue:46234" + ], + "soft_pairs": [ + "issue:43504|issue:45698", + "issue:43572|issue:46178", + "issue:44077|issue:44534", + "issue:45542|issue:46234", + "issue:43673|issue:44743", + "issue:43761|issue:44242", + "issue:44792|issue:44977", + "issue:44617|issue:45584", + "issue:44534|issue:44913", + "issue:44190|issue:44964", + "issue:44792|issue:45600", + "issue:44479|issue:44734", + "issue:44855|issue:45588", + "issue:43827|issue:43994", + "issue:41720|issue:45245", + "issue:43065|issue:44610", + "issue:43931|issue:44060", + "issue:45803|issue:45987", + "issue:35141|issue:42490", + "issue:43335|issue:44855", + "issue:43644|issue:44913", + "issue:43976|issue:44855", + "issue:43550|issue:44190", + "issue:45632|issue:46234", + "issue:44077|issue:44877", + "issue:43540|issue:44734", + "issue:44462|issue:45907", + "issue:45127|issue:45357" + ], + "prepared_review_unit_hash": "975b58e3002b250fea734bef48467125f5152ce12b479fa9c3bb0134d0365cbe", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35603, + "estimated_input_tokens": 8901, + "estimated_eval_tokens": 18058 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like a true duplicate; the items mostly span unrelated bugs across different models, configs, loaders, and runtime paths. The strongest issue candidate is the open, heavily discussed custom-module loading regression.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45698 is the strongest representative issue: it is open, has the most discussion activity, and describes a concrete, user-facing loading regression with clear reproduction potential.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is the best standalone issue to keep as canonical because it has broad impact, active discussion, and a focused bug report about incorrect module loading after save/load.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Different bugs: one is a legacy-field preset loading failure for BEiT semantic segmentation, the other is wrong custom-module resolution after save/load." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "Unrelated regressions: StableLm pad token config metadata vs an import-utils backend registration error for tensorflow_text." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Different code paths: a restriction on optional post_init in patchtsmixer versus a v5 buffer initialization corruption issue." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "Not the same problem: TensorBoard/TensorFlow backend setup versus a missing import_utils availability helper." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Both involve cache behavior, but they target different mechanisms and models: chunked_prefill cache presence vs recurrent state reset in modular_qwen3_5." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "Different model behaviors: CLIP hidden_states output regression versus missing load-balancing loss when router logits are disabled." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Separate issues: a failing janus image-generation test versus Qwen3.5 flash-attention generation malfunction." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "No shared underlying bug: Sam3Video CUDA OOM is unrelated to Whisper empty-transcription generation failure after token alignment." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Different root causes: junk-filled non-persistent buffers versus GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Unrelated failures: local dataset loading in a training script versus loading a multimodal model with latest transformers." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "One is a model test failure, the other is a stale auto-mapping reference to removed Sam3LiteText configs breaking CI." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Different failure modes and code paths: video-input regression in several Qwen variants versus KV-cache continuation indexing in transformers serve." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Unrelated issues: Python 3.13 import parsing failure in DebertaV2Model versus flash_attention sink-less model AttributeError." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing removed pipeline() calls is not the same as incorrect inference results for SigLIP2 AutoModel/pipeline usage." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Different symptoms and subsystems: Qwen3 auto device mapping CUDA assert versus a category-count RuntimeError." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Not duplicates: a dummy Conv2d in Sam3PixelDecoder versus a processor/model input-size mismatch in OmDet-Turbo." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Different regressions: Qwen3-VL weight shape mismatch on load versus an incorrect tied-weights warning involving dt_bias." + }, + { + "left": "issue:45803", + "right": "issue:45987", + "accept": false, + "reason": "One concerns exception handling in tensor conversion, the other a return-type contract bug in StaticCache.get_seq_length()." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Different areas entirely: token embedding reinitialization after resize versus a trailing-slash path import warning." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "Unrelated model/config bug versus a Python 3.13 parser/import problem caused by @torch.jit.script comment placement." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Separate v5 regression classes: non-persistent buffer junk versus GPTNeoX config serialization losing rotary_pct." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "No overlap: Python-version compatibility regression versus a DebertaV2 import parsing error." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "Different subsystems: SDPA/torch.compile failure in Bamba versus dataset loading in a training script." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "Cache-path collision for trust_remote_code local models is unrelated to a missing import_utils availability symbol." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "Different strictness bugs: patchtsmixer post_init policy versus granite_speech config loading under strict config checks." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Different code paths: Qwen3OmniMoe video-input processing error versus transformers serve KV-cache continuation indexing." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "Not the same bug: tokenizer.json selection logic versus list_repo_files failing to forward revision." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "Distinct issues: LoRA merging with extended vocabulary on tied embeddings versus incorrect visual encoder keys during 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": 18, + "review_unit_index": 32, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36246", + "issue:38175", + "issue:38617", + "issue:41628", + "issue:42907", + "issue:43065", + "issue:43116", + "issue:43295", + "issue:43335", + "issue:43404", + "issue:43441", + "issue:43479", + "issue:43531", + "issue:43643", + "issue:43650", + "issue:43854", + "issue:44112", + "issue:44186", + "issue:44190", + "issue:44220", + "issue:44246", + "issue:44464", + "issue:44496", + "issue:44521", + "issue:44561", + "issue:44683", + "issue:44734", + "issue:44737", + "issue:44933", + "issue:44938", + "issue:44964", + "issue:45003", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45362", + "issue:45375", + "issue:45381", + "issue:45446", + "issue:45542", + "issue:45563", + "issue:45584", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:46082", + "issue:46121" + ], + "soft_pairs": [ + "issue:44521|issue:45381", + "issue:43335|issue:44190", + "issue:44737|issue:45803", + "issue:44112|issue:45600", + "issue:43116|issue:44190", + "issue:46082|issue:46121", + "issue:43479|issue:45375", + "issue:45276|issue:45584", + "issue:44938|issue:45542", + "issue:43295|issue:45276", + "issue:45245|issue:45584", + "issue:44464|issue:45563", + "issue:43650|issue:44220", + "issue:44186|issue:44734", + "issue:36246|issue:41628", + "issue:43404|issue:45216", + "issue:44246|issue:45003", + "issue:44683|issue:45446", + "issue:38175|issue:43065", + "issue:44933|issue:45362", + "issue:43643|issue:44496", + "issue:43854|issue:44964", + "issue:44964|issue:45600", + "issue:44112|issue:44964", + "issue:45563|issue:45584", + "issue:44734|issue:45632", + "issue:36246|issue:38617", + "issue:44561|issue:45632", + "issue:42907|issue:43441", + "issue:43531|issue:45600" + ], + "prepared_review_unit_hash": "812d09aafe246ff508aaaa37560f0acb15e07d014ede82964cac2d2990ec1913", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36142, + "estimated_input_tokens": 9036, + "estimated_eval_tokens": 18328 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate pairs are mostly broad Transformers issues that share only subsystem-level similarity, not the same concrete bug. I would not merge any of the soft edges, and there is no clear single canonical issue in this heterogeneous set.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal/Qwen-related, but one is about assistant masks in chat templating and the other about video vision position ids; different failure modes and code paths." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "Completely unrelated bugs: SwitchTransformers config creation vs an image-classification example script failing to load a local dataset." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet device placement bug and a Fuyu exception-handling bug are unrelated code paths." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Both mention CI, but one is a stale device-override test and the other is broken auto-mapping references to removed configs." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Both touch example scripts, but one is a multi-label classification output bug and the other is a dataset-loading problem in a different script." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Different config-validation bugs: head_dim rejection in LlamaConfig vs rope validation type handling." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both are multimodal config issues, but they involve different model families and different initialization/serialization bugs." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Unrelated issues: embedding resize propagation in gemma4 versus Whisper generation on empty transcription." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import failure and TensorBoard/TensorFlow backend detection are distinct problems." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "A processor/tokenizer regression is not the same as token-embedding resize not propagating in gemma4." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Different bugs: category-cardinality runtime error vs Whisper empty-transcription generation failure." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Both are generation-related, but one is incorrect outputs from compiled forward and the other is a stale warning about num_return_sequences." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "The first is a vague data request; the second is a specific feature-extraction bug." + }, + { + "left": "issue:44186", + "right": "issue:44734", + "accept": false, + "reason": "Different crash sites and models: LayoutLMv2Tokenizer padding/NER vs server KV-cache continuation indexing." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different symbols and different APIs; not the same missing-export bug." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Weight-tying in Mistral3 and checkpoint saving regression in Qwen3.5 are different model-specific issues." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import-performance regression and unsafe sys.modules access are not the same underlying defect." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both involve flex_attention/versioning, but one is a torch>=2.9 compiled failure and the other is an incorrect version check for AuxRequest import." + }, + { + "left": "issue:38175", + "right": "issue:43065", + "accept": false, + "reason": "Different models and unrelated symptoms: zero probabilities in SigLIP2 vs a dummy Conv2d in Sam3PixelDecoder." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "A missing import from image_utils and a Qwen3.5 chat crash are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "AutoConfig missing fields under trust_remote_code is unrelated to an unrecognized-model config.json error." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different model-loading failures for different models; same general area but not the same bug." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "A model-loading regression and broken auto-mapping references are unrelated." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flakiness and a Phi-4-multimodal loading issue are different problems." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale generate warning and Whisper empty-transcription failure are not the same code-path issue." + }, + { + "left": "issue:44734", + "right": "issue:45632", + "accept": false, + "reason": "KV-cache continuation indexing in the server is unrelated to trust_remote_code cache-path collisions." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Different import failures for different names; not a duplicate export regression." + }, + { + "left": "issue:44561", + "right": "issue:45632", + "accept": false, + "reason": "Both involve trust_remote_code, but one breaks models due to a removed helper and the other is a local cache-path collision." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Ministral/Devstral save_pretrained behavior and FlashAttention failures are separate bugs." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup 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": 18, + "review_unit_index": 33, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:32090", + "issue:35141", + "issue:38175", + "issue:39692", + "issue:41950", + "issue:42175", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43388", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43526", + "issue:43577", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:43824", + "issue:43901", + "issue:44016", + "issue:44075", + "issue:44220", + "issue:44242", + "issue:44279", + "issue:44336", + "issue:44351", + "issue:44368", + "issue:44496", + "issue:44617", + "issue:44779", + "issue:44933", + "issue:45020", + "issue:45042", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45335", + "issue:45357", + "issue:45372", + "issue:45405", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45759", + "issue:45812", + "issue:46209", + "issue:46234" + ], + "soft_pairs": [ + "issue:43526|issue:43901", + "issue:43295|issue:45335", + "issue:43388|issue:43688", + "issue:44496|issue:45042", + "issue:43441|issue:45588", + "issue:43673|issue:44336", + "issue:41950|issue:44016", + "issue:44075|issue:44368", + "issue:42175|issue:44279", + "issue:43299|issue:45216", + "issue:43381|issue:45593", + "issue:43901|issue:44496", + "issue:44016|issue:44220", + "issue:44496|issue:45759", + "issue:43650|issue:46209", + "issue:45372|issue:46234", + "issue:45335|issue:45584", + "issue:32090|issue:35141", + "issue:43479|issue:43525", + "issue:44242|issue:45200", + "issue:43577|issue:45216", + "issue:38175|issue:44617", + "issue:43299|issue:45357", + "issue:44779|issue:45812", + "issue:43295|issue:44779", + "issue:43824|issue:44933", + "issue:39692|issue:43901", + "issue:44279|issue:45405", + "issue:44351|issue:45278", + "issue:41950|issue:45020" + ], + "prepared_review_unit_hash": "d13d94114697c6c868db115248ff328c6768884b8b85d4c536e0e171d331c768", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36767, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair looks like a true duplicate: the items span unrelated bugs across tokenizers, processors, configs, imports, training utilities, and docs. A broad version-regression import issue is the closest umbrella representative, but it does not unify the set.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like report here because it summarizes multiple import breakages after a Transformers version upgrade, making it the closest representative among otherwise unrelated issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most suitable cluster representative because it describes a wide regression class rather than a single narrow symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "Different problems: image-processor label reduction bug vs docs still mentioning return_all_scores." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Different code paths: multimodal processor.tokenizer regression vs t5gemma embedding resize not propagating." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Different training bugs: last-batch metric gathering vs auxiliary-loss normalization." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Different bugs: model config recognition failure vs PIL image processor requiring torchvision." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both mention flash attention, but they hit different failures/models and different code paths." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Unrelated: generation cache/chunked prefill vs ANSI formatting in loading_report." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Different domains: video-classification pipeline processor lookup vs a notebook syntax error." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Different issues: SGD optimizer args ignored vs a configuration warning about tie_word_embeddings." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "Different dependency complaints: missing tensorflow in extras vs a generic transformers dependency issue." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Different regressions: Qwen3VL MoE model loading vs Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Different bugs: gradient checkpointing disallowed in eval vs D-FINE auxiliary losses omitted when denoising is off." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Docs-only mismatch vs a model loading/config error; not the same bug." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error vs audio feature extraction bug; unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Different model-instantiation failures: generic unrecognized model vs missing text_config unwrapping for composite Qwen multimodal configs." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "No clear relation: placeholder title vs FP8 inference failure on B200." + }, + { + "left": "issue:45372", + "right": "issue:46234", + "accept": false, + "reason": "Different missing symbols/import surfaces: ReasoningEffort import error vs falcon-mamba availability helper missing." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Different bugs: token embedding resize propagation vs Whisper empty-transcription generation failure." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Both involve None/tie_word_embedding themes, but one is Trainer GPU broadcast of None and the other is post_init reinitialization." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Different config problems: default vision/audio init from None vs missing pad_token_id attribute." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Different mechanisms: router-logit load balancing loss vs mm_token_type_ids defaulting for text-only fine-tuning." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Different regressions: dtype not applied to Blip2 submodules vs incorrect Qwen3.5 save_pretrained output." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "Different symptoms and subsystems: unexpected zero probabilities vs CUDA out-of-memory." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Same model family but different concrete bugs: loading failure vs save_pretrained writing wrong visual-encoder keys." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and likely different fixes." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Processor/tokenizer-image handling regression vs DeepSeek tokenizer output bug; not the same issue." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Different import failures: missing Qwen2.5-VL class vs nonexistent image_utils import." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Different docs issues: SigLIP2 example errors vs TextClassificationPipeline return_all_scores documentation drift." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Generic dependency problem vs a specific unreleased minimum PEFT version bump." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "Related by imports, but 45278 is a broad report while this is a specific HybridCache export failure; not clearly one bug." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Different problems: video-classification pipeline processor lookup vs remote_code model breakage." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 34, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:41950", + "issue:42994", + "issue:43065", + "issue:43299", + "issue:43317", + "issue:43441", + "issue:43450", + "issue:43525", + "issue:43650", + "issue:43653", + "issue:43723", + "issue:43854", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", + "issue:44062", + "issue:44162", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44297", + "issue:44361", + "issue:44462", + "issue:44466", + "issue:44521", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:44898", + "issue:44987", + "issue:45081", + "issue:45276", + "issue:45325", + "issue:45397", + "issue:45405", + "issue:45542", + "issue:45593", + "issue:45600", + "issue:45712", + "issue:45823", + "issue:45834", + "issue:45907", + "issue:46132" + ], + "soft_pairs": [ + "issue:44016|issue:44062", + "issue:41950|issue:45823", + "issue:43976|issue:45405", + "issue:43299|issue:43854", + "issue:44683|issue:45542", + "issue:44462|issue:45081", + "issue:43650|issue:46132", + "issue:43976|issue:44987", + "issue:44297|issue:44462", + "issue:43976|issue:44279", + "issue:45081|issue:45907", + "issue:43450|issue:44279", + "issue:44521|issue:45325", + "issue:44466|issue:45276", + "issue:43441|issue:45834", + "issue:42994|issue:43317", + "issue:43653|issue:45593", + "issue:45600|issue:45712", + "issue:44623|issue:44749", + "issue:44623|issue:45397", + "issue:43065|issue:44898", + "issue:44261|issue:44263", + "issue:44361|issue:45593", + "issue:44671|issue:44779", + "issue:44016|issue:45823", + "issue:43441|issue:44683", + "issue:43723|issue:44162", + "issue:43901|issue:43994", + "issue:44297|issue:45397", + "issue:43525|issue:43881", + "issue:43901|issue:44246" + ], + "prepared_review_unit_hash": "8ded950a8648a019489007aedacc52603b58cd7de229e0c5ab9be8d2b6d1ec0f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36740, + "estimated_input_tokens": 9185, + "estimated_eval_tokens": 18626 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the issues span unrelated tokenizer, vision, loading, docs, and performance bugs. I found no soft pair that looks like the same underlying change or defect, so all soft edges are rejected.", + "confidence": 0.36, + "canonical_issue_reason": "Issue 41950 is the strongest standalone report in the set: it has the highest discussion activity and inbound references, and is a clear, concrete bug. It is only a representative choice, not evidence that the cluster is a true duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 41950 is the best overall issue to represent this group because it is active, well-specified, and externally referenced, even though the cluster itself is not cohesive.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken keyword conflict are unrelated failures in different subsystems." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Video-classification processor lookup and Gemma4 device-mapping mismatch are different bugs." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version compatibility and a PEFT version bump are unrelated dependency problems." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different model-loading regressions affecting different models and likely different code paths." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "FlexAttention on torch 2.9 and missing TensorFlow backend for tensorboard are unrelated." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-loading issues, but they involve different mechanisms and different failure modes." + }, + { + "left": "issue:43650", + "right": "issue:46132", + "accept": false, + "reason": "Generic placeholder request and AttentionInterface behavior change are not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Python compatibility issue and a model-loading failure for a specific repo are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:44462", + "accept": false, + "reason": "Tokenizer save metadata mismatch is distinct from AutoTokenizer ignoring tokenizer.json on load." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Dependency/version support issue and a generic dependency issue are not the same concrete defect." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Mistral tokenizer regex patch crash and list_repo_files revision forwarding are unrelated." + }, + { + "left": "issue:43450", + "right": "issue:44279", + "accept": false, + "reason": "Video processor batch-shape bug and a generic dependency issue do not match." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Assistant mask generation and temporal rope indexing are different multimodal bugs." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Serialization of lm_head.weight and resize_token_embeddings propagation are different issues." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "FlashAttention failure for Ministral-3 and OCR index error for Kosmos2.5 are unrelated model bugs." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Saving quantized models and loading dequantized models with offload are different code paths and failures." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer special-token registration and D-FINE auxiliary losses are unrelated." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Removed config references in auto_mappings and leftover dummy classes are separate CI cleanups." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "Missing processor save files and tokenizer/data filtering slowdown are unrelated." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Processor save_pretrained file omission and gemma-4 zero3 loading are unrelated." + }, + { + "left": "issue:43065", + "right": "issue:44898", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and Perceiver resize/interpolation failure are different vision bugs." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "LayerNorm epsilon config issue and torch.split return-value handling are unrelated." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "MLukeTokenizer AttributeError and D-FINE denoising-loss behavior are unrelated." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "CamemBERT masked-LM regression and Deepseek tokenizer regression are different model-specific bugs." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "Notebook syntax error and Gemma4 device mismatch are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "Both mention attention/flash attention, but they are distinct failures on different operators and versions." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "Tokenizer loading regression and ESM2 model breakage are not the same underlying defect." + }, + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Docs mismatch for pipeline output and incorrect model inference results are different issues." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer serialization metadata and gemma-4 zero3 loading are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Missing pad_token_id on Llama4Config and glm-4v-9b loading failure are different model/config bugs." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Pipeline docs mismatch and import-time slowdown 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": 18, + "review_unit_index": 35, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43262", + "issue:43299", + "issue:43352", + "issue:43450", + "issue:43525", + "issue:43643", + "issue:43673", + "issue:43761", + "issue:43784", + "issue:43824", + "issue:43976", + "issue:44016", + "issue:44188", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44291", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44373", + "issue:44485", + "issue:44496", + "issue:44521", + "issue:44561", + "issue:44589", + "issue:44617", + "issue:44749", + "issue:44779", + "issue:44857", + "issue:44871", + "issue:44964", + "issue:45081", + "issue:45250", + "issue:45292", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45593", + "issue:46132" + ], + "soft_pairs": [ + "issue:43450|issue:44617", + "issue:43525|issue:43643", + "issue:44188|issue:44336", + "issue:43784|issue:45542", + "issue:44016|issue:44291", + "issue:43262|issue:44521", + "issue:43673|issue:43761", + "issue:41628|issue:43976", + "issue:44589|issue:45362", + "issue:44336|issue:44373", + "issue:16998|issue:22355", + "issue:42491|issue:43299", + "issue:44246|issue:44496", + "issue:44295|issue:45397", + "issue:44242|issue:45593", + "issue:44336|issue:44485", + "issue:44246|issue:44871", + "issue:43824|issue:43976", + "issue:41950|issue:44220", + "issue:43976|issue:44964", + "issue:36246|issue:42757", + "issue:42886|issue:46132", + "issue:36246|issue:44351", + "issue:44779|issue:45292", + "issue:44242|issue:44857", + "issue:43673|issue:44561", + "issue:45081|issue:45362", + "issue:44295|issue:44749", + "issue:43352|issue:45250" + ], + "prepared_review_unit_hash": "c1dd519f264f0534cb8b1cb40e0352f0b2e3a0eaf0103e65154765f2ce5faad3", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36333, + "estimated_input_tokens": 9084, + "estimated_eval_tokens": 18424 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose topical grouping of unrelated Transformers issues. The pairs share broad themes like imports, multimodal chat, or Qwen/MoE, but they do not describe the same concrete bug or mergeable fix.", + "confidence": 0.82, + "canonical_issue_reason": "issue:44521 is the strongest representative: it is a concrete, well-scoped multimodal `apply_chat_template` bug with active discussion and inbound references, making it the best anchor among these heterogeneous reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the clearest and most actionable issue in the set; it has the most direct bug description and the strongest evidence of user impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43450", + "right": "issue:44617", + "accept": false, + "reason": "Different failures: batched video processor shape mismatch vs video-model OOM. Same broad area, not the same bug." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "One is a missing `pad_token_id` on `Llama4Config`; the other is `trust_remote_code` losing fields in `AutoConfig`. Related only at a high level." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence under `torch.compile` is unrelated to ANSI codes in `loading_report`." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both mention import-time breakage, but one is `nn` missing for sentence-transformers and the other is TensorFlow backend handling with only tensorboard installed." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "A notebook syntax error and an `init_empty_weights` TypeError are different problems with no shared fix path." + }, + { + "left": "issue:43262", + "right": "issue:44521", + "accept": false, + "reason": "Both touch `apply_chat_template`, but one is wrong audio sampling-rate default and the other is all-zero assistant masks; different code paths and symptoms." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Chunked-prefill cache handling and CLIP vision hidden-states propagation are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "Missing `AutoImageProcessor` export and a Python-version compatibility issue are not the same bug." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage deserialization failure is unrelated to Qwen3.5 chat crashes." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "Documentation/ANSI output issue versus a wrong docstring for `position_ids`; not the same underlying change." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "`model_max_length` on DeBERTa-V3 and missing `transformers.onnx` are unrelated." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both involve Qwen3 MoE families, but one is LoRA compatibility across hf4.x/hf5.x and the other is a Qwen3VL MoE loading regression; too different to merge." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import slowness and an unrecognized model/config error are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "Buffer registration/`position_ids` access and gemma-4 ZeRO-3 loading are different issues." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Both concern losses omitted under flags, but for different models and different toggles (`output_router_logits` vs denoising)." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI code output and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import-time performance and Gemma EOS token config mismatch are not the same bug." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "A missing Qwen2.5 VL export and a Python-version compatibility break are unrelated." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification pipeline image-processor lookup and `_torch_extract_fbank_features()` are separate subsystems." + }, + { + "left": "issue:43976", + "right": "issue:44964", + "accept": false, + "reason": "Python 3.9+ compatibility and Phi-4 multimodal model loading are not the same failure." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Missing Qwen2.5 VL symbol and missing `is_offline_mode` from huggingface_hub are different import errors." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline tokenizer cache loading and `AttentionInterface.register` behavior changes are unrelated." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing exports and different code paths." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Deepseek tokenizer regression and `resize_token_embeddings` not updating output embeddings are different model behavior bugs." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Different model-specific loss/AMP failures; no common concrete fix." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Cache behavior in chunked prefill and removal of `is_torch_fx_available` affecting trust-remote-code models are unrelated." + }, + { + "left": "issue:45081", + "right": "issue:45362", + "accept": false, + "reason": "Mistral tokenizer regex patch crash and Qwen3.5 chat crash are different model/tokenizer paths." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "`position_ids` buffer access and Chinese-reported performance regression after upgrade are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Specific Nemotron FA2 support error versus a generic Flash Attention 2.0 topic; not the same concrete issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 36, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36246", + "issue:43065", + "issue:43232", + "issue:43257", + "issue:43262", + "issue:43408", + "issue:43475", + "issue:43526", + "issue:43540", + "issue:43596", + "issue:43653", + "issue:43901", + "issue:43931", + "issue:44162", + "issue:44246", + "issue:44336", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44483", + "issue:44496", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45412", + "issue:45459", + "issue:45561", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45656", + "issue:45701", + "issue:45759", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46170", + "issue:46202" + ], + "soft_pairs": [ + "issue:44162|issue:45200", + "issue:43901|issue:45003", + "issue:43526|issue:45003", + "issue:44496|issue:45003", + "issue:45412|issue:46123", + "issue:43262|issue:44246", + "issue:45292|issue:45584", + "issue:43257|issue:45137", + "issue:45375|issue:45759", + "issue:45292|issue:45701", + "issue:36246|issue:43596", + "issue:44877|issue:45375", + "issue:45561|issue:45563", + "issue:43232|issue:44336", + "issue:44987|issue:45459", + "issue:43065|issue:43408", + "issue:45987|issue:46082", + "issue:44995|issue:45200", + "issue:44466|issue:45335", + "issue:43475|issue:43540", + "issue:44479|issue:46170", + "issue:45397|issue:45656", + "issue:44483|issue:46202", + "issue:45137|issue:45600", + "issue:43653|issue:44162", + "issue:44995|issue:45588", + "issue:44368|issue:44977", + "issue:44368|issue:45563", + "issue:44877|issue:46170", + "issue:43931|issue:44368" + ], + "prepared_review_unit_hash": "af200d3d10a89cf949aeb43f59bc30b82585f5266f323c95cf638beacdda8f1f", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36323, + "estimated_input_tokens": 9081, + "estimated_eval_tokens": 18418 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated bug reports that only share broad subsystem terms (loading, generation, config, deepspeed). I would not deduplicate any of the soft pairs.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45003 is the strongest issue candidate to represent the cluster only because it is the broadest infrastructure-level bug here and appears in multiple soft-similarity links; however, it still does not match the other issues as a duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45003 is the best single issue to keep as the cluster anchor if one must be chosen, since it is a concrete modeling_utils bug with broader applicability than the more model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 corruption/broken workflow is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mentioning return_all_scores do not match a runtime sys.modules access bug." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to modeling_utils module lookup." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized-model config loading and unsafe sys.modules access are different failure modes." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "RT-DETR memory cleanup and MaskGenerationPipeline final-batch handling are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio chat_template sampling-rate default bug is unrelated to slow import timing." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "resize_token_embeddings/output_embeddings mismatch is not the same as Whisper empty transcription after token alignment." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both involve DeepSpeed, but one is Qwen3 MoE weight conversion and the other is a zero3 deque/index error." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Both are Qwen multimodal config issues, but one drops a missing field and the other fails to unwrap text_config; not the same bug." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Embedding resize behavior is unrelated to tokenization changes across versions." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Qwen2_5_VL import error and DeepSpeed BertModel index error are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Granite speech strict-config loading and Qwen config field dropping are different models and code paths." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "A pytest-xdist file race is unrelated to a stale generation warning." + }, + { + "left": "issue:43232", + "right": "issue:44336", + "accept": false, + "reason": "Generation kwargs update after sync_gpus has nothing to do with ANSI output in loading_report." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model-loading failure and tokenizer error masking are separate issues in different code paths." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 pixel decoder dummy conv and SAM3 tracker model-type warning are not the same underlying defect." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "StaticCache seq-length return type and LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "GlmMoeDsa stale cache crash is unrelated to Gemma 4 token-type defaults." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Tied-weight serialization inconsistency is unrelated to t5gemma resize_token_embeddings propagation." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "SAM3 vision encoder output attribute loss and Qwen3OmniMoe video-input validation are different bugs." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Qwen video-input regression and Windows ROCm torch.distributed import failure are unrelated." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma-4 zero3 loading and duplicate optimizer.step calls under deepspeed are different problems." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "chat/completions request rejection is unrelated to Gemma pipeline roundtrip token normalization." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed deque failure and stale auto_mappings references to removed SAM3 Lite configs do not share a code path." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug is unrelated to the ESM2 breakage report." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "GlmMoeDsa stale cache crash and flash_attention s_aux=None AttributeError are different failures." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "A tied-word-embeddings warning is not the same as a flash-attention generation regression." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "Stale num_return_sequences warning and tie_word_embeddings warning are unrelated." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict granite_speech config loading and Windows ROCm dependency issues are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch is unrelated to a tie_word_embeddings warning." + } + ] + }, + "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": 18, + "review_unit_index": 37, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:41628", + "issue:42371", + "issue:42886", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43701", + "issue:43827", + "issue:44038", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44188", + "issue:44297", + "issue:44361", + "issue:44623", + "issue:44661", + "issue:44734", + "issue:44749", + "issue:44869", + "issue:44871", + "issue:44933", + "issue:45003", + "issue:45092", + "issue:45229", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45478", + "issue:45561", + "issue:45584", + "issue:45656", + "issue:46077", + "issue:46123", + "issue:46132", + "issue:46170", + "issue:46202" + ], + "soft_pairs": [ + "issue:42886|issue:45229", + "issue:44060|issue:44077", + "issue:43673|issue:44734", + "issue:45003|issue:45561", + "issue:45584|issue:46170", + "issue:44297|issue:44623", + "issue:44869|issue:45356", + "issue:44361|issue:44749", + "issue:43646|issue:44162", + "issue:44749|issue:45656", + "issue:44623|issue:45656", + "issue:43673|issue:43827", + "issue:44871|issue:46132", + "issue:43701|issue:44749", + "issue:44188|issue:46202", + "issue:45230|issue:45397", + "issue:42994|issue:43482", + "issue:43475|issue:45381", + "issue:44038|issue:44623", + "issue:43475|issue:43643", + "issue:28282|issue:30990", + "issue:42371|issue:45092", + "issue:43232|issue:43673", + "issue:41628|issue:44933", + "issue:45375|issue:45478", + "issue:43475|issue:44661", + "issue:42994|issue:43525", + "issue:46077|issue:46123", + "issue:45310|issue:45375" + ], + "prepared_review_unit_hash": "878631b08c10809bbb978baa77d5b5e73d93b6cf23844298d16d549ca18ac947", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35546, + "estimated_input_tokens": 8887, + "estimated_eval_tokens": 18030 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Loose cluster around Transformers v5 regressions and adjacent model/tokenizer bugs, but the soft pairs mostly point to different failure modes and subsystems rather than true duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43646 is the broadest and most representative v5 breakage report in the set, centered on custom/remote model initialization, which overlaps thematically with several other regression reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best single issue to anchor the cluster because it is concrete, high-signal, and spans a common v5 compatibility path without being overly narrow.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline tokenizer cache loading and multi-GPU CUDA OOM are unrelated failure modes; same library, different code paths." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "One is a tied-weights warning bug in Qwen3-Next, the other is a model API constraint for patchtsmixer; not the same underlying change." + }, + { + "left": "issue:43673", + "right": "issue:44734", + "accept": false, + "reason": "Both involve generation/KV cache behavior, but one is a missing cache object in v5.0.0 during chunked prefill and the other is a serve-time indexing crash in /v1/responses." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to a pytest-xdist captured_info.txt race in testing utilities." + }, + { + "left": "issue:45584", + "right": "issue:46170", + "accept": false, + "reason": "Empty transcription generation after token alignment is unrelated to Windows ROCm import/type errors around torch.distributed." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch and processor.save_pretrained missing files are both serialization-related, but they affect different save paths and artifacts." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Whisper timestamp decode crash and Kimi-K2.5 tokenizer codec regression are distinct tokenizer/decoder bugs with different triggers." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "MLukeTokenizer task AttributeError and Chinese performance regression after upgrade are not the same bug." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Custom model initialization failure and ESM2 breakage may both be v5 regressions, but they are separate model-specific compatibility issues." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Data filtering slowdown after upgrade and deepspeed optimizer-step duplication are unrelated performance/ \u0627\u0644\u062a\u062f\u0631\u064a\u0628 issues." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Missing processor save artifacts versus double optimizer stepping are different subsystems and symptoms." + }, + { + "left": "issue:43673", + "right": "issue:43827", + "accept": false, + "reason": "Generation cache bug and docs still referencing pipeline() are not the same underlying issue." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id mismatch and AttentionInterface.register behavior change are separate configuration/API problems." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "resume_from_checkpoint key mismatch is unrelated to the reported data-filtering slowdown." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "torch.compile divergence in attention kernels and Gemma pipeline roundtrip failure on a special character are unrelated." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Generic bug report and gemma-4 zero3 from_pretrained error do not indicate the same concrete problem." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are different directions of the save/load lifecycle." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "SAM 3 Video missing attribute and Qwen2.5-VL video position-id error are separate multimodal model bugs." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe breakage and missing processor save files are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "SAM 3 Video encoder output attribute error and trust_remote_code missing fields are both initialization-related but on different model families and causes." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "Missing PyTorch import and SentenceTransformers loading hang are unrelated user-environment/runtime issues." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 settings docs/API issue and remote-code checkpoint meta-initialization incompatibility are not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "GenerationMixin key update after sync_gpus and missing cache during chunked prefill are both generation-related but different concrete failures." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "AutoImageProcessor import failure and nonexistent import from image_utils are distinct API/export regressions." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Missing config field silently dropped by strict handling and Qwen3.5 MoE from_pretrained error are related to the same model family but not the same concrete issue." + }, + { + "left": "issue:43475", + "right": "issue:44661", + "accept": false, + "reason": "SAM 3 Video attribute error and add-new-model-like failure in TOKENIZER_MAPPING_NAMES are different code paths and model infra concerns." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized model saving and Llama4Config missing pad_token_id are separate bugs." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Incorrect attention mask construction and final-partial-batch dropping in MaskGenerationPipeline are unrelated logic errors." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3.5 MoE from_pretrained error and missing deepstack_visual_indexes in Qwen3_5MoeVisionConfig are different model-specific 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": 18, + "review_unit_index": 38, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:36010", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42673", + "issue:42831", + "issue:42898", + "issue:42994", + "issue:43065", + "issue:43232", + "issue:43278", + "issue:43377", + "issue:43704", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43901", + "issue:44038", + "issue:44188", + "issue:44222", + "issue:44242", + "issue:44351", + "issue:44373", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44779", + "issue:44792", + "issue:44964", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45200", + "issue:45310", + "issue:45356", + "issue:45406", + "issue:45561", + "issue:45600", + "issue:45712", + "issue:45812", + "issue:46123", + "issue:46202" + ], + "soft_pairs": [ + "issue:45081|issue:45103", + "issue:45406|issue:46202", + "issue:42673|issue:43704", + "issue:41628|issue:44351", + "issue:44964|issue:45137", + "issue:44222|issue:45310", + "issue:42994|issue:43824", + "issue:44038|issue:45310", + "issue:36010|issue:41628", + "issue:42831|issue:44222", + "issue:42175|issue:45712", + "issue:44496|issue:45561", + "issue:43377|issue:44242", + "issue:42898|issue:44568", + "issue:45561|issue:45600", + "issue:43901|issue:45561", + "issue:43232|issue:44792", + "issue:30990|issue:35141", + "issue:43278|issue:45137", + "issue:42222|issue:45200", + "issue:44623|issue:45310", + "issue:43065|issue:46123", + "issue:44188|issue:44373", + "issue:30333|issue:33290", + "issue:43232|issue:44188", + "issue:42898|issue:45812", + "issue:43784|issue:43824", + "issue:43723|issue:44568", + "issue:43874|issue:45092", + "issue:42898|issue:45356", + "issue:42898|issue:44779" + ], + "prepared_review_unit_hash": "0e2b75db758eff492a45213710f122583209a1fe03b9fefcf2996c53566b2f26", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37043, + "estimated_input_tokens": 9261, + "estimated_eval_tokens": 18778 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. Only one soft pair looks like a true duplicate (the two identical Qwen3ForCausalLM VRAM-leak issues); the rest are at best loosely similar by model family or transformer-v5 regression themes and should stay separate.", + "confidence": 0.9, + "canonical_issue_reason": "42673 is the clearest anchor because 42673/43704 are the only exact duplicate pair in the set (same title and same underlying Qwen3ForCausalLM VRAM leak).", + "canonical_pr_reason": null, + "best_issue_reason": "42673 is the best representative issue here: it has the strongest duplicate evidence and the most concrete, self-contained bug report among the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Both are crashes, but they are different code paths: Mistral tokenizer regex patching vs auto-docstring parsing with future annotations." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Both touch Gemma, but one is a processor attribute crash and the other is a tokenizer roundtrip failure; not the same bug." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same reported Qwen3ForCausalLM VRAM leak in multiple dataloader threads; this is a duplicate." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and unrelated code paths." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading failure and DeepSpeed ZeRO3 deque error are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "Both mention MoE/FP8, but one is FP8 save_pretrained and the other is a Qwen3.5 Moe from_pretrained error; different failures." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Quantized model saving and a Qwen2.5-VL import error are not the same underlying problem." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both concern Qwen3-family models, but the concrete failures differ enough that they would not be one merged fix." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors in transformers v5, but they involve different APIs and separate breakages." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "FP8 accuracy regression and FP8 save_pretrained bug are different issues." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Package-dependency/backend inclusion problem vs dummy-class cleanup/check-repo failure are unrelated." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Model config loading failure and pytest-xdist file race are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Missing padding-mask support in MIMI encoder and missing load-balancing loss in MoE are different code paths." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is clean_up_tokenization_spaces behavior and the other is add_special_tokens/BOS-EOS handling." + }, + { + "left": "issue:45561", + "right": "issue:45600", + "accept": false, + "reason": "Test-worker file race and broken auto_mappings references are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:45561", + "accept": false, + "reason": "Docstring mismatch for TextClassificationPipeline and pytest-xdist file race are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs/update bug and a Janus image-generation test failure do not share the same failure mode." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Sentence-transformers loading stall and embedding reinitialization after resize are unrelated." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "BF16-to-FP32 dtype drift during evaluation and ZeRO3 deque underflow are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "VitPose breakage and Gemma 4 mm_token_type_ids defaulting are different multimodal issues." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Missing files in processor.save_pretrained and Qwen3.5 Moe from_pretrained error are different problems." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder and MaskGenerationPipeline partial-batch dropping are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Torch.compile attention branching bug and a wrong docstring are not duplicates." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLFlow reporting on failed jobs and DeepSpeed OOM with Adafactor are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44188", + "accept": false, + "reason": "Both involve generation/runtime behavior, but they are different concrete bugs and code paths." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions across v4/v5, but they affect different models and different tokenization behavior." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but for different symbols and different downstream packages." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Both concern tokenizer behavior in v5, but one is load-time tokenizer issues and the other is special-token insertion semantics." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Both are multimodal model-loading issues, but the missing image-patch method and meta-initialization incompatibility are distinct bugs." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions around v5, but the codec/fix_mistral_regex problem is separate from clean_up_tokenization_spaces." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Granite token ID regression and DeepSeek tokenizer regression are separate model-specific bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays conservative. The only accepted soft pair, issue:42673 vs issue:43704, is clearly supported by identical titles. The other soft-edge rejections do not overclaim duplicate status based on the provided titles and issue metadata, and the overall summary correctly treats the cluster as mostly heterogeneous." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 39, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28282", + "issue:30333", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:41628", + "issue:42175", + "issue:42757", + "issue:42831", + "issue:42898", + "issue:43122", + "issue:43475", + "issue:43479", + "issue:43723", + "issue:43873", + "issue:43976", + "issue:44016", + "issue:44060", + "issue:44075", + "issue:44079", + "issue:44190", + "issue:44206", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44448", + "issue:44623", + "issue:44734", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45245", + "issue:45310", + "issue:45381", + "issue:45405", + "issue:45656", + "issue:45712", + "issue:46082", + "issue:46132", + "issue:46209", + "issue:46231", + "issue:46305" + ], + "soft_pairs": [ + "issue:44734|issue:45103", + "issue:44016|issue:46231", + "issue:44075|issue:46231", + "issue:44060|issue:45125", + "issue:42898|issue:44448", + "issue:44295|issue:45656", + "issue:45405|issue:45712", + "issue:44448|issue:46132", + "issue:42175|issue:44933", + "issue:44222|issue:44297", + "issue:44933|issue:45712", + "issue:43479|issue:44079", + "issue:43976|issue:44933", + "issue:44297|issue:45656", + "issue:44222|issue:44623", + "issue:30333|issue:34689", + "issue:42831|issue:44623", + "issue:45245|issue:46132", + "issue:44190|issue:46082", + "issue:30333|issue:36331", + "issue:43479|issue:44869", + "issue:30333|issue:35141", + "issue:28282|issue:30333", + "issue:45071|issue:46082", + "issue:43475|issue:46305", + "issue:44297|issue:45310", + "issue:41628|issue:42757", + "issue:44206|issue:45103", + "issue:43723|issue:44977", + "issue:45381|issue:46305", + "issue:44263|issue:45092", + "issue:43122|issue:46209", + "issue:43873|issue:44945" + ], + "prepared_review_unit_hash": "fb45d91805ef96d863added34f725d25ad84de582a7dfd72dfb330c240cbfe15", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37489, + "estimated_input_tokens": 9373, + "estimated_eval_tokens": 19002 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a loose grab bag of Transformers regressions and compatibility issues, but the soft pairs are not true duplicates; they mostly share only broad themes like \"v5 breakage\" or \"save_pretrained\".", + "confidence": 0.34, + "canonical_issue_reason": "issue:45071 is the most representative closed item: a concrete Transformers v5 regression that affects a core API contract. It best matches the cluster\u2019s general pattern of version-related breakages.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45071 is the best anchor because it is specific, user-facing, and broadly aligned with the cluster\u2019s recurring v5 compatibility complaints.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Different subsystems and failure modes: serving KV-cache continuation indexing vs auto-docstring handling of future annotations." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Both touch documentation/examples, but one is a notebook syntax error and the other is a parameter-name mismatch; not the same bug." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument handling and a docs typo are unrelated changes." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Different Qwen bugs: a wrong tied-weights warning versus a missing tensor-parallel plan." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "Both involve v5 output differences, but one is tokenization cleanup behavior and the other is Pegasus generation output; not one underlying defect." + }, + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Position-ids buffer access and double optimizer stepping with deepspeed are separate runtime issues." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "A version pin/release mismatch is unrelated to leftover dummy classes leaking into imports." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Different code paths: Pegasus output regression versus AttentionInterface.register behavior changes." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Different import/export problems in different areas; one is backend packaging, the other an image_utils symbol export issue." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 MOE serialization and the other is tokenizer_config metadata; not the same bug." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Both are import-surface issues, but one is a missing import from image_utils and the other is dummy classes leaking from the torch-less shim." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Different objects and failures: multimodal config default initialization vs ModelOutput key assignment behavior." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version compatibility and a missing image_utils import are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata and double optimizer stepping have no shared code-path." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both involve saving artifacts, but FP8 model serialization and processor.save_pretrained missing files are different implementations and bugs." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Training-job MLflow reporting and Llama 3.2 Vision model loading are unrelated." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "An FP8 accuracy regression is unrelated to processor save_pretrained missing files." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-count limits and AttentionInterface.register behavior are separate issues." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading and LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting after failed training and Trainer compute_loss signature mismatch are different trainer bugs." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Multimodal config defaults and Whisper timestamp decoding are unrelated." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "A general PretrainedConfig type-checking regression is not the same as LlamaConfig head_dim validation." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "Different model families and bugs: missing SAM3 attribute vs MoonshineStreaming off-by-one masking." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save metadata mismatch and Qwen3.5 MoE from_pretrained errors are different failure modes." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for unrelated symbols and dependencies." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "LaserFeatureExtractor argument handling and auto-docstring annotation parsing are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading in v5 and Qwen3.5 flash-attention generation failures are unrelated." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Different sequence/streaming bugs in different models; no shared code-path." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "A torch.split return-value bug and meta-init incompatibility in old InternVL2 checkpoints are unrelated." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenization differences across versions and FP8 inference on B200 are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading and pipeline-parallel output correctness 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": 18, + "review_unit_index": 40, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30990", + "issue:38175", + "issue:39401", + "issue:41628", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43232", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43761", + "issue:43824", + "issue:43950", + "issue:44038", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44360", + "issue:44368", + "issue:44373", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44514", + "issue:44568", + "issue:44623", + "issue:44829", + "issue:44871", + "issue:44913", + "issue:44945", + "issue:45092", + "issue:45103", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45753", + "issue:45834", + "issue:46002", + "issue:46050", + "issue:46082", + "issue:46305" + ], + "soft_pairs": [ + "issue:43232|issue:45834", + "issue:44222|issue:44484", + "issue:44373|issue:44485", + "issue:44295|issue:44623", + "issue:43479|issue:43644", + "issue:44483|issue:45406", + "issue:43761|issue:46082", + "issue:43723|issue:44871", + "issue:45092|issue:45357", + "issue:44263|issue:44360", + "issue:30990|issue:38175", + "issue:44261|issue:45092", + "issue:42907|issue:46050", + "issue:43644|issue:46305", + "issue:44514|issue:45103", + "issue:44829|issue:46002", + "issue:43479|issue:44913", + "issue:42831|issue:45310", + "issue:44263|issue:44373", + "issue:22355|issue:28282", + "issue:41628|issue:43824", + "issue:42831|issue:44038", + "issue:39401|issue:44568", + "issue:44368|issue:46305", + "issue:42886|issue:43066", + "issue:42886|issue:42994", + "issue:43950|issue:46305", + "issue:44945|issue:45753" + ], + "prepared_review_unit_hash": "f1acb502fbe1a4df5c107cf2f5dfd79099238911cbbb455e6bae3077a0a8a4e9", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "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": "The cluster is highly heterogeneous: these are mostly unrelated issues that only share broad surface-level themes like imports, tokenizers, saving/loading, or multimodal regressions. I do not see a true duplicate set among the soft pairs.", + "confidence": 0.93, + "canonical_issue_reason": "No single issue is a true canonical duplicate target here, but issue:43950 is the clearest standalone regression report: it names the concrete failure (`from_pretrained()` corrupting non-persistent buffers), the transformers 5.x regression scope, and the behavior is specific enough to anchor related reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43950 is the strongest general issue in the set for triage purposes because it is precise, reproducible, and describes a concrete code-path regression rather than a broad symptom or discussion-only problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Different subsystems and failures: generation kwargs handling vs long OCR input index error." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "Quantized save_pretrained/moe failure is unrelated to max_shard_size behavior." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring typo/report about position_ids is not the same as GLM-5 RoPE implementation." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Buffer registration/read error is unrelated to processor save_pretrained missing files." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config None-handling for multimodal defaults is a different bug from corrupted non-persistent buffers." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat completions request handling and Gemma4Processor _tokenizer crash are different code paths." + }, + { + "left": "issue:43761", + "right": "issue:46082", + "accept": false, + "reason": "CLIP hidden_states regression and LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer load failure and eos_token_id config mismatch are separate tokenizer issues." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Meta-init incompatibility for old InternVL2 checkpoints is different from bad visual encoder keys on save_pretrained." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Both mention DSA indexer, but one is about torch.split outputs and the other about a missing ReLU; not the same bug." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading hang and SIGLIP2 zero probabilities are unrelated symptoms." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "MLA rms_norm_eps precision issue is not the same as remote-code/meta-init incompatibility." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Both involve quantization/save-load, but they point to different models and different failure modes; too broad to merge." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer corruption is unrelated to MoonshineStreaming sliding-window off-by-one." + }, + { + "left": "issue:44514", + "right": "issue:45103", + "accept": false, + "reason": "Batched chat template crash and auto_docstring annotation crash are unrelated." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "Flash-attention training degeneration and NaNs after checkout/init are different concrete problems." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Different configs and lifecycle bugs: default subconfig init vs rotary_pct persistence on reload." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy issue and Qwen3.5 MoE from_pretrained error are not the same bug." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "Indexer split behavior and a wrong docstring are not the same underlying change." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Both are import errors, but they stem from different missing components/modules and should not be merged." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Missing AutoImageProcessor export and missing Qwen2_5_VLForConditionalGeneration export are distinct symbol-import failures." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "Generic FP8 accuracy regression is not the same as the Qwen3-VL-Moe bug." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Qwen3 offset_mapping bug and mdeberta BOS/EOS add_special_tokens bug affect different tokenizers and code paths." + }, + { + "left": "issue:44368", + "right": "issue:46305", + "accept": false, + "reason": "A tied-embeddings warning in fine-tuning is unrelated to a streaming mask off-by-one." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong decoder type in v5 are different tokenizer regressions." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Tokenizer offline cache loading and quantized model saving failure are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer corruption is unrelated to the sliding-window mask bug." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline parallelism output errors and an image-triggered infinite loop are different failure modes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 41, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28218", + "issue:30990", + "issue:33357", + "issue:41628", + "issue:42371", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43475", + "issue:43550", + "issue:43576", + "issue:43646", + "issue:43723", + "issue:43756", + "issue:43784", + "issue:43937", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44442", + "issue:44464", + "issue:44485", + "issue:44514", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44779", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45200", + "issue:45245", + "issue:45278", + "issue:45290", + "issue:45593", + "issue:45600", + "issue:46002", + "issue:46082", + "issue:46132" + ], + "soft_pairs": [ + "issue:42757|issue:45278", + "issue:44464|issue:46082", + "issue:44117|issue:44704", + "issue:43576|issue:43784", + "issue:43646|issue:46002", + "issue:42371|issue:44263", + "issue:43550|issue:46132", + "issue:30990|issue:33357", + "issue:43232|issue:45593", + "issue:44295|issue:44297", + "issue:44704|issue:45290", + "issue:44261|issue:44373", + "issue:43723|issue:44442", + "issue:42907|issue:44945", + "issue:43723|issue:44779", + "issue:42907|issue:43937", + "issue:41628|issue:45278", + "issue:43756|issue:44060", + "issue:43784|issue:45600", + "issue:43408|issue:43475", + "issue:43122|issue:44514", + "issue:42907|issue:44964", + "issue:43116|issue:45200", + "issue:44360|issue:44485", + "issue:44625|issue:44829", + "issue:43122|issue:44521", + "issue:44038|issue:44297", + "issue:43232|issue:45245", + "issue:28218|issue:30990", + "issue:44038|issue:44295", + "issue:43122|issue:44779" + ], + "prepared_review_unit_hash": "dd506ebcb330fbbd1f004fc26d5eed99361e656266eb9179761680e066c11990", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36606, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: the soft pairs share broad subsystems (tokenizers, imports, generation, multimodal chat templating) but not the same concrete bug or fix. I would not merge any of the candidate pairs as duplicates.", + "confidence": 0.71, + "canonical_issue_reason": "If forced to pick one reference issue, issue:43122 is the broadest and most representative tokenizer-regression report in the set; however, the cluster as a whole does not have a true canonical duplicate issue.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43122 is the best single issue to anchor nearby tokenizer-regression discussions because it is relatively broad, clearly reproducible, and has inbound references; still, it does not represent the rest of this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but 42757 is one missing symbol from huggingface_hub while 45278 is a broad upgrade breakage with many import errors; not the same underlying bug." + }, + { + "left": "issue:44464", + "right": "issue:46082", + "accept": false, + "reason": "Chunked generation with compiled forward is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "One is TOKENIZER_MAPPING_NAMES returning None; the other is AutoProcessor not forwarding kwargs to cached_file. Different code paths and failures." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "A broken CLI env command and a NameError during sentence-transformers import are unrelated issues." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Custom model initialization breakage and NaNs in classification heads are different problems with different symptoms and likely different fixes." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 settings guidance and a torch.split return-value issue are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:46132", + "accept": false, + "reason": "torch.compile with SDPA on Bamba and AttentionInterface.register changing function behavior are distinct bugs." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Sentence Transformers loading hangs and a MacOS bus error on a CLIP model are not the same defect." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation kwargs handling after sync_gpus is unrelated to D-FINE auxiliary losses when denoising is off." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Reading position_ids after buffer registration and tokenizer_class mismatch in save_pretrained are different tokenizer/model serialization issues." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "Missing kwargs passthrough in AutoProcessor.from_pretrained and chat-template crashes on tool-call assistant messages are separate code paths." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "A config epsilon precision issue and a wrong docstring are not duplicates." + }, + { + "left": "issue:43723", + "right": "issue:44442", + "accept": false, + "reason": "Both concern tokenizer loading, but one is a general v5 loading issue and the other is a specific FastSpeech2ConformerTokenizer failure; not enough evidence of the same bug." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Saving dequantized Ministrals/Devstrals and incorrect output under pipeline parallelism are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44779", + "accept": false, + "reason": "Both mention tokenizer regressions in v5, but they affect different tokenizer paths and symptoms; too broad to call the same bug." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Model-saving of dequantized weights is unrelated to GLM-5 GenerationConfig validation." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "AutoImageProcessor import failure may be one example of import breakage, but 45278 is a broad collection of import errors after upgrade, not the same specific issue." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE layer loss and Qwen3-Next tied-weights warnings are unrelated model-specific bugs." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Missing nn during import and stale auto_mappings references to removed Sam3LiteText configs are different import/config cleanup problems." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "A model-type mismatch warning and a missing attribute on Sam3VisionEncoderOutput are separate SAM 3 issues." + }, + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Tokenization drift across versions and a batched apply_chat_template crash with padding=False are different failures." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and flash_attention_3 training degeneration are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:44521", + "accept": false, + "reason": "kwargs forwarding to cached_file and zero assistant_masks for multimodal inputs are distinct chat/template behaviors." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "A Qwen3-VL-Moe bug and tokenizer save_pretrained metadata mismatch are not the same underlying problem." + }, + { + "left": "issue:43232", + "right": "issue:45245", + "accept": false, + "reason": "Generation config updates after sync_gpus and a categories-cardinality runtime error are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:30990", + "accept": false, + "reason": "Added-token spacing behavior and Sentence Transformers loading hangs are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe behavior and position_ids buffer access errors are different bugs." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they concern different models and different observable failures; too weak to merge as 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": 18, + "review_unit_index": 42, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:22355", + "issue:28218", + "issue:29127", + "issue:29159", + "issue:33453", + "issue:36296", + "issue:39401", + "issue:42222", + "issue:42371", + "issue:42757", + "issue:42831", + "issue:42907", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43441", + "issue:43452", + "issue:43493", + "issue:43611", + "issue:43784", + "issue:43824", + "issue:43825", + "issue:43874", + "issue:43906", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44222", + "issue:44297", + "issue:44351", + "issue:44355", + "issue:44485", + "issue:44488", + "issue:44521", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:45083", + "issue:45216", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45563", + "issue:45753", + "issue:46002", + "issue:46231" + ], + "soft_pairs": [ + "issue:42907|issue:43452", + "issue:36296|issue:43116", + "issue:44038|issue:44222", + "issue:39401|issue:44488", + "issue:45216|issue:45478", + "issue:39401|issue:44521", + "issue:43824|issue:44351", + "issue:22355|issue:28218", + "issue:43784|issue:45446", + "issue:22355|issue:29159", + "issue:42757|issue:44351", + "issue:43122|issue:43906", + "issue:43992|issue:44704", + "issue:43066|issue:44779", + "issue:44355|issue:46231", + "issue:43937|issue:44485", + "issue:43611|issue:46002", + "issue:42371|issue:43874", + "issue:42222|issue:43441", + "issue:44075|issue:44936", + "issue:42831|issue:44297", + "issue:43976|issue:45446", + "issue:45357|issue:45478", + "issue:44038|issue:45478", + "issue:43116|issue:43825", + "issue:43493|issue:45753", + "issue:45083|issue:45446", + "issue:43116|issue:43493", + "issue:43937|issue:45563", + "issue:22355|issue:29127", + "issue:33453|issue:39401" + ], + "prepared_review_unit_hash": "733e2a3a8ac1d1a6d9d3ba728380828e2ca30bfae54a3f6dd4837789ebca2f06", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36429, + "estimated_input_tokens": 9108, + "estimated_eval_tokens": 18472 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous: tokenizer regressions, model loading/saving bugs, import errors, and unrelated training/runtime issues. The only clear duplicate-like matches are the Qwen3.5 save/load serialization regressions.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45216 is the best anchor for the Qwen3.5 checkpoint/save_pretrained regression family: it is broad, regression-tagged, and directly explains the related from_pretrained failures.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 is the strongest global issue candidate because it captures the underlying serialization bug more generally than the later, narrower symptom reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Different code paths: dequantized model saving vs gguf loading failure." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor-parallel training bug and multi-label example-script behavior are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Both mention MoE, but one is a Qwen3-VL-Moe runtime issue and the other is FP8 save_pretrained; not the same bug." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Wrong offset_mapping in Qwen3 tokenizer vs a separate model-loading failure for cjvt/sleng-bert." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": true, + "reason": "Same Qwen3.5 save/load serialization regression: incorrect saved checkpoint content leads to from_pretrained failure." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Offset mapping bug and all-zero assistant masks are different tokenizer/template problems." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Different missing symbols: Qwen2_5_VLForConditionalGeneration vs HybridCache." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "ONNX import error and extra-space-after-added-token tokenizer bug are unrelated." + }, + { + "left": "issue:43784", + "right": "issue:45446", + "accept": false, + "reason": "Sentence-transformers import NameError is not the same as the flex_attention PyTorch version check bug." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "ONNX module import failure and slow/fast tokenizer inconsistency are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing APIs and different failure causes." + }, + { + "left": "issue:43122", + "right": "issue:43906", + "accept": false, + "reason": "An isolated reproduction can be related, but there is not enough evidence that it is the same concrete tokenizer bug." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Missing embed_tokens.weight in UMT5Encoder and AutoProcessor kwargs forwarding are different loading issues." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Wrong tokenizer decoder type and Deepseek tokenization regression are distinct symptoms/code paths." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Compiled-Python runtime errors and a documentation parameter-name mismatch are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GLM-5 generation config validation and GLM-5 RoPE discussion are not the same bug." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "base_model_prefix loading breakage and NaNs in classification heads are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance issue and missing get_number_of_image_patches AttributeError are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "Vitpose import breakage and Ministral-3 FlashAttention failure are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "SGD args not used and trainer.evaluate() failing after training are different trainer/optimizer bugs." + }, + { + "left": "issue:42831", + "right": "issue:44297", + "accept": false, + "reason": "FineGrainedFP8 accuracy issue and tokenizer_class mismatch on save_pretrained are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version compatibility regression and flex_attention version-check bug are unrelated." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": true, + "reason": "Same Qwen3.5 save/load serialization family: incorrect visual-encoder keys in save_pretrained plausibly cause the from_pretrained error." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5/Qwen3.5-VL-Moe, but one is a transformer/model bug and the other is a save/load serialization failure." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Multi-label example script output and pipeline translation-task messaging are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and Qwen3_5 infinite-loop-on-image are unrelated multimodal bugs." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Qwen3_omni_moe feature-length helper behavior and flex_attention PyTorch version check are unrelated." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Multi-label classification example behavior and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GLM-5 generation-config validation and paged generate stale warning are unrelated generation issues." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "ONNX import failure and layoutlmv3 error-message clarity are unrelated." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression and Qwen3 tokenizer offset_mapping bug are different tokenizer failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analyst output is grounded in the packet and the negative pairings stay appropriately conservative. The only accepted matches are the Qwen3.5 save/load regression family, which is well-supported by the issue titles, and the overall heterogeneity summary fits the set." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 43, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42673", + "issue:43064", + "issue:43066", + "issue:43122", + "issue:43381", + "issue:43408", + "issue:43421", + "issue:43450", + "issue:43673", + "issue:43742", + "issue:43827", + "issue:43937", + "issue:43986", + "issue:44117", + "issue:44291", + "issue:44380", + "issue:44410", + "issue:44466", + "issue:44483", + "issue:44484", + "issue:44496", + "issue:44625", + "issue:44779", + "issue:44829", + "issue:44877", + "issue:44938", + "issue:44945", + "issue:44977", + "issue:45071", + "issue:45072", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45405", + "issue:45446", + "issue:45464", + "issue:45588", + "issue:46002", + "issue:46077", + "issue:46178", + "issue:46231", + "issue:46298" + ], + "soft_pairs": [ + "issue:44625|issue:46002", + "issue:38617|issue:45278", + "issue:44380|issue:46077", + "issue:39401|issue:44779", + "issue:42175|issue:44938", + "issue:44117|issue:45588", + "issue:43064|issue:43673", + "issue:43381|issue:44945", + "issue:43827|issue:45446", + "issue:43450|issue:43986", + "issue:43937|issue:44977", + "issue:44466|issue:46002", + "issue:45405|issue:45446", + "issue:44117|issue:45081", + "issue:43122|issue:46298", + "issue:42673|issue:45072", + "issue:44410|issue:45310", + "issue:44291|issue:45071", + "issue:44829|issue:45216", + "issue:39401|issue:43421", + "issue:38617|issue:41628", + "issue:42222|issue:46178", + "issue:43408|issue:43742", + "issue:44877|issue:46231", + "issue:44484|issue:45310", + "issue:43937|issue:44496", + "issue:43066|issue:43122", + "issue:44483|issue:45464" + ], + "prepared_review_unit_hash": "244ca988404866e3980c46494b9f40167bf662ef513d9f9be2153b5c4eb9749d", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36088, + "estimated_input_tokens": 9022, + "estimated_eval_tokens": 18300 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of Transformers v5 regressions (imports, tokenization, loading, training, and API behavior), but the individual reports are mostly distinct rather than true duplicates. The best anchor issue is the broad v5 upgrade regression report.", + "confidence": 0.36, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report in the set, covering multiple import failures after upgrading to v5.5.0, which makes it the closest representative of the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most general and cluster-like issue; it can summarize a subset of the v5 regression/import breakage reports better than the highly specific model- or feature-level bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "Different symptoms and code paths: Qwen3.5 config propagation vs NaNs in classification heads after checkout/init." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but one is a specific missing symbol and the other is a broad upgrade-regression report; not clearly the same underlying bug." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "Both involve attention behavior, but the model, backend, and failure mode differ: scaling ignored vs causal masks still built." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they concern different models and different observable failures (offset mapping vs tokenization output)." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Unrelated environment/dependency problems: missing TensorFlow in an install extra vs Python 3.14 load failure." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Different subsystems and failures: tokenizer mapping lookup vs flash_attention crash on missing s_aux." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Training/FSDP optimizer-state bug vs generation cache missing during chunked prefill; no shared concrete code-path." + }, + { + "left": "issue:43381", + "right": "issue:44945", + "accept": false, + "reason": "Gradient checkpointing eval-mode issue is unrelated to pipeline-parallel incorrect output." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs cleanup vs a PyTorch version-check bug in flex_attention; not the same change or bug." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Both are video-related, but one is a batched-shape bug and the other is a crash due to missing torchvision." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Different model-specific generation failures: invalid GenerationConfig vs flash-attention generation breakage." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of tied weights depending on device is unrelated to NaNs in classification heads." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "PEFT version bump/release issue is unrelated to the flex_attention import-version check." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Both touch tokenizer loading, but the actual defects differ: mapping names None vs backend_tokenizer attribute access in Mistral regex patching." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Tokenization regression vs noisy logging message; clearly unrelated." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in multi-threaded dataloader use is unrelated to dtype mismatches in inference." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Different Qwen variants and different failures: missing projections vs from_pretrained loading error." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument in init_empty_weights is a different problem from PretrainedConfig type-checking breakage." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Training degenerate loss with flash_attention_3 is unrelated to checkpoint serialization regression." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Offset-mapping bug and runtime special-token/post-processor update bug are related to tokenizers generally but not the same underlying defect." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they refer to different missing symbols and likely different fixes." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "vitpose model breakage and tensorflow_text backend mapping failure are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Model-type mismatch warning for sam3_tracker is unrelated to a MobileLLM key error." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Strict config loading and a docs parameter-name mismatch are not the same bug." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Save_pretrained shard-size question is unrelated to Qwen3.5 from_pretrained loading failures." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Both are load/generation-related, but one is a GenerationConfig validity error and the other is an unrecognized-model/config.json problem." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both are v5 tokenizer regressions, but one is decoder type selection and the other is changed tokenization output; not the same concrete bug." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both mention chat/completions, but one is a general request rejection and the other is a model-specific streaming inference failure." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 44, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:38617", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:43012", + "issue:43097", + "issue:43335", + "issue:43408", + "issue:43441", + "issue:43452", + "issue:43577", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43986", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44360", + "issue:44368", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44829", + "issue:44933", + "issue:44945", + "issue:44964", + "issue:44977", + "issue:44987", + "issue:45071", + "issue:45278", + "issue:45440", + "issue:45478", + "issue:45542", + "issue:45588", + "issue:45656", + "issue:46050", + "issue:46172", + "issue:46231" + ], + "soft_pairs": [ + "issue:44075|issue:45656", + "issue:44964|issue:45278", + "issue:42371|issue:44261", + "issue:42222|issue:45542", + "issue:43012|issue:43577", + "issue:44410|issue:45478", + "issue:38617|issue:44220", + "issue:42175|issue:42222", + "issue:43874|issue:44261", + "issue:43097|issue:44493", + "issue:41628|issue:45071", + "issue:44263|issue:44485", + "issue:45071|issue:45542", + "issue:42673|issue:44977", + "issue:42673|issue:43906", + "issue:44263|issue:46172", + "issue:43012|issue:43408", + "issue:44360|issue:46172", + "issue:44484|issue:45478", + "issue:43441|issue:44829", + "issue:42371|issue:42831", + "issue:43931|issue:44038", + "issue:43335|issue:45071", + "issue:43097|issue:43452", + "issue:43874|issue:44263", + "issue:44933|issue:46231", + "issue:44117|issue:44987", + "issue:43408|issue:44368", + "issue:43976|issue:44945", + "issue:43986|issue:44484", + "issue:45440|issue:46172", + "issue:43097|issue:46050", + "issue:43097|issue:45588" + ], + "prepared_review_unit_hash": "5d05464b4f90a55d125641f12af2268869f06cd700d584333ffcf10df06d7723", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36829, + "estimated_input_tokens": 9208, + "estimated_eval_tokens": 18672 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous, with mostly unrelated transformer regressions and API/behavior issues. None of the soft pairs look like the same underlying bug or fix, so I would not merge any of them as duplicates.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45278 is the broadest and most umbrella-like issue in the set, centered on widespread import breakage after upgrading Transformers, which makes it the best representative target for this mixed cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 best captures the general user-facing regression theme in the cluster and is broad enough to subsume related import/load failures better than the narrower, model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Both concern optimizers, but one says SGD args are ignored while the other says deepspeed calls optimizer.step twice; different failure modes and fixes." + }, + { + "left": "issue:44964", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific Phi-4 multimodal loading failure; the other is a broad post-upgrade import-error report. Same area, not the same bug." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 API guidance and an MLA rms_norm_eps precision issue are unrelated problems." + }, + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "VitPose breakage and missing TensorFlow backend from a tensorboard-only install are different issues." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "Both mention dtype/precision, but one is a compile warning on bfloat16 models and the other is a Blip2 loading bug where dtype fields stay float32." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3-family models, but one is a missing projection in qwen3next and the other is a Qwen3.5 MoE from_pretrained error; not the same code path." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "An import error from configuration_utils and a feature-extraction fbank issue are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "TensorFlow backend packaging and broken VitPose models are different bugs." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "A missing image-patch helper in a fast image processor and an MLA layernorm epsilon precision issue do not share the same underlying defect." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "One is a config removal/regression around tied embeddings, the other is unexpected position-id key warnings; different symptoms and fixes." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "Missing AutoImageProcessor import and PretrainedConfig type-checking regressions are unrelated API breakages." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "Both mention GLM/DSA, but one is about torch.split return values and the other is a RoPE implementation bug." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "PretrainedConfig type checking and an undefined tf backend error from tensorboard-only installs are separate problems." + }, + { + "left": "issue:42673", + "right": "issue:44977", + "accept": false, + "reason": "VRAM leakage in multi-threaded dataloaders and Qwen3.5 flash-attention generation failure are different bugs." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "The VRAM leak report is about Qwen3ForCausalLM threading; the linked reproduction issue is not the same concrete failure." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different model families and different code paths: a DSA indexer split issue versus a DeepSeek V4 dtype/transpose bug." + }, + { + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "A bfloat16 precision warning during compile and a SAM3 model-type mismatch warning are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "A missing ReLU in the DSA indexer and a DeepSeek V4 comb dtype/transpose issue are not the same bug." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "A save_pretrained shard-size question and a Qwen3.5 MoE loading error have no shared underlying fix." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both involve FlashAttention, but one is a Ministral-3 failure and the other is degenerate training with flash_attention_3; not mergeable as one PR." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 settings guidance and FineGrainedFP8 accuracy loss are separate numeric-behavior issues." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "A Qwen3-VL checkpoint shape mismatch and a Qwen3-VL-Moe issue on Transformers 5.0 are related only by model family, not by the same concrete bug." + }, + { + "left": "issue:43335", + "right": "issue:45071", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug and PretrainedConfig type-checking regression are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Tied-embedding config removal and gguf_file loading failures are different loading regressions." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Missing get_number_of_image_patches and torch.split return handling are different implementation bugs." + }, + { + "left": "issue:44933", + "right": "issue:46231", + "accept": false, + "reason": "A missing image_utils import and a documentation typo in set_training are unrelated." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping returning None and a physical-intelligence/fast load failure are different issues." + }, + { + "left": "issue:43408", + "right": "issue:44368", + "accept": false, + "reason": "Both are warnings, but one is a model-type mismatch and the other is a tied-embeddings config warning; not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python compatibility breakage and incorrect output under pipeline parallelism are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "A crash without torchvision and a save_pretrained shard-size question do not share the same underlying defect." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Both mention DeepSeek, but one is a native-vs-remote implementation divergence and the other is a dtype/transpose bug in DeepSeek V4." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Tie-embedding config removal and sinq quantization save/load breakage are unrelated regressions." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "Tied-embedding config removal and a flash_attention crash on s_aux=None are different code-path 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": 18, + "review_unit_index": 45, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:38617", + "issue:42617", + "issue:42673", + "issue:42757", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43502", + "issue:43550", + "issue:43792", + "issue:43825", + "issue:43873", + "issue:43906", + "issue:43931", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44162", + "issue:44261", + "issue:44315", + "issue:44351", + "issue:44368", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44704", + "issue:44964", + "issue:45103", + "issue:45200", + "issue:45290", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45561", + "issue:45588", + "issue:45712", + "issue:45753", + "issue:45859", + "issue:46123", + "issue:46172", + "issue:46298" + ], + "soft_pairs": [ + "issue:43064|issue:44315", + "issue:42673|issue:45200", + "issue:38617|issue:42757", + "issue:44062|issue:44117", + "issue:44410|issue:45375", + "issue:42890|issue:46298", + "issue:44038|issue:44410", + "issue:43931|issue:44410", + "issue:43502|issue:43992", + "issue:45341|issue:45859", + "issue:43097|issue:43873", + "issue:38617|issue:44351", + "issue:44261|issue:44485", + "issue:44964|issue:45341", + "issue:44704|issue:45103", + "issue:43986|issue:44117", + "issue:42617|issue:45341", + "issue:43066|issue:43906", + "issue:43986|issue:45372", + "issue:43986|issue:45290", + "issue:45341|issue:45588", + "issue:45341|issue:45561", + "issue:43208|issue:43550", + "issue:43986|issue:45341", + "issue:43992|issue:44484", + "issue:43012|issue:43550", + "issue:43825|issue:44655", + "issue:43012|issue:44368", + "issue:45712|issue:46298", + "issue:42907|issue:44655", + "issue:42907|issue:45753", + "issue:43097|issue:46298", + "issue:44162|issue:46172", + "issue:43792|issue:46123" + ], + "prepared_review_unit_hash": "75b1e382fdd06905e6fe6659335b624cf31206a63cca7da6c5c2fcaad2a40758", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37751, + "estimated_input_tokens": 9438, + "estimated_eval_tokens": 19132 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These soft links are mostly superficial matches across unrelated issues; they span different models, APIs, warnings, tests, and dependency/import failures rather than one underlying bug.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical duplicate stands out: the items are heterogeneous and do not collapse to one shared code-path failure.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43986 is the clearest self-contained, actionable bug report in the set, but it is only a representative issue\u2014not a duplicate target for the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Different failures: FSDP2/PEFT optimizer-state corruption vs Liger kernel not applied under model_init." + }, + { + "left": "issue:42673", + "right": "issue:45200", + "accept": false, + "reason": "VRAM leak across dataloader threads is unrelated to Gemma 4 mm_token_type_ids defaults for text-only fine-tuning." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different missing symbols from different packages and code paths." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken keyword conflict is a constructor bug; TOKENIZER_MAPPING_NAMES returning None is a separate lookup/dispatch bug." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "One is missing Qwen3Next attention projections; the other is a missing vision config field silently dropped by strict config handling." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "A flaky test seed issue is unrelated to a noisy logging message." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Qwen3-VL-Moe loading failure and Qwen3Next missing projections are different model-specific defects." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Shape-mismatch loading for Qwen3-VL-30B is not the same bug as missing attn_qkv/attn_gate projections." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "local_files_only still triggering API requests is unrelated to UMT5Encoder missing embed_tokens.weight." + }, + { + "left": "issue:45341", + "right": "issue:45859", + "accept": false, + "reason": "testing_utils bug and Qwen3_5MoeTextRotaryEmbedding CPU-offload incompatibility are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder in v5 is a breaking API change, not the same as offloading/quantization behavior." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different namespaces; no shared underlying code-path." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "MLA q_a_layernorm epsilon precision and GLM-5 RoPE implementation are separate model math issues." + }, + { + "left": "issue:44964", + "right": "issue:45341", + "accept": false, + "reason": "Phi-4 multimodal loading failure is unrelated to the testing_utils bug." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "AutoProcessor kwargs forwarding to cached_file is a different failure from auto_docstring crashing on postponed annotations." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Missing torchvision during AutoProcessor load is unrelated to TOKENIZER_MAPPING_NAMES sometimes returning None." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "Cannot run 3d_parallel.py and a testing_utils bug do not describe the same defect." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type is a concrete tokenizer regression; the other item is only an isolated reproduction of a different issue." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Both are processor-loading errors, but one is missing torchvision and the other is a mistral_common import failure; different dependencies and paths." + }, + { + "left": "issue:43986", + "right": "issue:45290", + "accept": false, + "reason": "Processor loading without torchvision and chat-template tool-call crashes are separate bugs." + }, + { + "left": "issue:45341", + "right": "issue:45588", + "accept": false, + "reason": "A testing utility bug is not the same as flash_attention crashing on sink-less models." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "Testing utility bug vs pytest-xdist race on captured_info.txt are different concurrency paths." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training blockers and Bamba torch.compile+SDPA failure are different model/runtime problems." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing checkpoint weight on load is unrelated to the max_shard_size default question in save_pretrained()." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A bfloat16 warning during compile is not the same as a model-specific torch.compile SDPA failure." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "pipeline() translation support messaging and saving Pipeline objects are different pipeline API issues." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "Both mention warnings, but they concern different config fields and different model paths." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Left is leftover dummy classes/check_repo leakage; right is just a noisy logging message." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Failing to save dequantized Ministrals/Devstrals is unrelated to saving generic Pipeline objects." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "A removed API warning is not the same as a noisy logging message." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 breakage and deepseek-v4 tensor-transpose bug are unrelated model-specific failures." + }, + { + "left": "issue:43792", + "right": "issue:46123", + "accept": false, + "reason": "Whisper load failure and MaskGenerationPipeline dropping the final partial batch are different bugs in different components." + } + ] + }, + "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": 18, + "review_unit_index": 46, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:36010", + "issue:36683", + "issue:38617", + "issue:42490", + "issue:42757", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43525", + "issue:43526", + "issue:43611", + "issue:43618", + "issue:43673", + "issue:43761", + "issue:43824", + "issue:43867", + "issue:43901", + "issue:43937", + "issue:44038", + "issue:44222", + "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:45478", + "issue:45600", + "issue:45698", + "issue:46077", + "issue:46082", + "issue:46123", + "issue:46172", + "issue:46178", + "issue:46209", + "issue:46298" + ], + "soft_pairs": [ + "issue:44938|issue:46178", + "issue:44483|issue:44514", + "issue:44222|issue:45478", + "issue:43208|issue:43526", + "issue:43012|issue:44683", + "issue:42490|issue:43012", + "issue:42757|issue:43824", + "issue:45600|issue:46298", + "issue:38617|issue:43502", + "issue:45698|issue:46172", + "issue:44038|issue:44230", + "issue:43673|issue:46123", + "issue:44483|issue:45290", + "issue:43064|issue:44380", + "issue:43618|issue:43761", + "issue:43299|issue:43867", + "issue:43066|issue:46209", + "issue:36683|issue:43475", + "issue:45276|issue:46077", + "issue:42994|issue:44655", + "issue:43901|issue:44655", + "issue:44625|issue:46077", + "issue:36010|issue:42757", + "issue:43352|issue:43937", + "issue:43611|issue:44655", + "issue:43867|issue:45216", + "issue:46077|issue:46082", + "issue:43502|issue:44485", + "issue:36683|issue:43525", + "issue:43618|issue:45216" + ], + "prepared_review_unit_hash": "ccacc19be3c17fcf089101d1f4982c3e3f9ef3b87f7c07eb6ed256cb8fd3705c", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36257, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: nearly all soft pairs describe different bugs, models, or code paths. I would not merge any of the suggested issue pairs as duplicates.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45698 is the strongest anchor in this set: it is open, has the most discussion/inbound references, and describes a broad save/load regression with clear user impact.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 looks like the best overall issue to keep as the representative item because it is active, broadly impactful, and directly about core from_pretrained/save_pretrained behavior.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Both are import-related, but one is a Python 3.14 load failure and the other is a backend mapping error for tensorflow_text; different failure modes and code paths." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "One is an API request/acceptance problem on /v1/chat/completions, the other is a chat templating crash on batched input with padding=False; not the same bug." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5/MoE and FP8/save/load, but one is about saving quantized models and the other about loading from_pretrained on v5.4.0; related area, not the same concrete defect." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "Completely different subsystems: xLSTM training bugs versus BeitImageProcessorFast label reduction behavior." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "A PyTorch precision warning from bfloat16 compilation is unrelated to compiled flex_attention failures on torch>=2.9." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "Local-path import warnings are unrelated to bfloat16/float32 precision warnings." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols from different packages and different breakages." + }, + { + "left": "issue:45600", + "right": "issue:46298", + "accept": false, + "reason": "CI breakage from removed Sam3LiteText references is unrelated to a noisy logging message." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "One is an import error from configuration_utils; the other is network requests occurring despite local_files_only=True. Different bugs." + }, + { + "left": "issue:45698", + "right": "issue:46172", + "accept": false, + "reason": "Save/load custom-module mismatch is unrelated to a DeepSeek tensor transpose/dtype issue." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both are Qwen3-VL/MoE-related, but one is a general v5 bug and the other is specifically FP8 inference support; not clearly the same underlying fix." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "GenerationMixin cache missing in chunked_prefill and MaskGenerationPipeline dropping the final partial batch are different functional bugs." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "One is a chat-completions API compatibility failure; the other is a template/tokenization crash with tool calls and no content." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "Trainer/FSDP2 optimizer-state corruption is unrelated to GPT2 attention scaling being ignored under SDPA/FlashAttention." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Both are CLIP v5 regressions, but one is missing attentions and the other is missing hidden_states; they look like separate output-path issues." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Qwen3VL MoE loading failures and sorted-state_dict load errors are different loading problems." + }, + { + "left": "issue:43066", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer decoder type mismatch and FP8 inference failure on B200 are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3Config missing vocab_size and SAM 3 Video missing fpn_position_embeddings are different attribute errors in different models." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Gemma4 embedding-resize propagation and Gemma4 bidirectional-attention masking are distinct model bugs." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model save failure and pipeline save_pretrained failure are different save paths with different symptoms." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "Documentation mismatch for return_all_scores is unrelated to pipeline serialization failing." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Gemma4 attention-mask construction are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Different missing imports from different modules; no shared underlying bug." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig validation are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Base_model_prefix loading issue and pipeline save_pretrained failure are different code paths." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "A sorted state_dict load error is not the same as Qwen3.5 checkpoint corruption introduced by save_pretrained in 5.4.0." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Gemma4 causal-mask regression and LlamaConfig head_dim validation are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Offline-mode network access is unrelated to GLM-5 RoPE implementation issues." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Different configs and different missing attributes: vocab_size versus pad_token_id." + } + ] + }, + "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": 18, + "review_unit_index": 47, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42890", + "issue:43064", + "issue:43097", + "issue:43299", + "issue:43352", + "issue:43441", + "issue:43502", + "issue:43575", + "issue:43618", + "issue:43867", + "issue:43881", + "issue:43906", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44112", + "issue:44117", + "issue:44222", + "issue:44242", + "issue:44246", + "issue:44355", + "issue:44360", + "issue:44380", + "issue:44492", + "issue:44655", + "issue:44855", + "issue:44857", + "issue:44938", + "issue:44964", + "issue:44977", + "issue:45125", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45479", + "issue:45561", + "issue:45600", + "issue:45753", + "issue:46132", + "issue:46202", + "issue:46285" + ], + "soft_pairs": [ + "issue:36683|issue:44242", + "issue:44380|issue:46132", + "issue:36683|issue:43881", + "issue:43906|issue:45753", + "issue:43299|issue:45753", + "issue:43352|issue:43976", + "issue:43992|issue:45310", + "issue:43441|issue:44380", + "issue:43352|issue:45478", + "issue:44355|issue:44855", + "issue:43867|issue:46202", + "issue:43867|issue:45375", + "issue:44380|issue:45479", + "issue:43352|issue:44380", + "issue:36683|issue:44117", + "issue:43618|issue:44380", + "issue:41669|issue:44246", + "issue:43097|issue:43352", + "issue:42890|issue:45561", + "issue:43064|issue:44492", + "issue:43976|issue:44938", + "issue:43867|issue:45357", + "issue:43992|issue:45478", + "issue:36296|issue:43575", + "issue:43064|issue:44655", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:43618|issue:44857", + "issue:44655|issue:44938", + "issue:43618|issue:44977", + "issue:44360|issue:45440", + "issue:43502|issue:44062", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:44222|issue:45125" + ], + "prepared_review_unit_hash": "8aa57892d8fae62adcd8ec080c7f31aa6975323616e01658b48a46ae5d9c6dda", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38087, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a heterogeneous mix of Transformers issues (model loading, attention backends, tensor parallelism, CI flakiness, and import-time regressions). None of the soft-linked pairs look like the same underlying bug/change, so I would not merge any of them as duplicates.", + "confidence": 0.46, + "canonical_issue_reason": "Issue 36683 is the most active open item with a concrete runtime failure and broad model-loading relevance; among this mixed set it is the best anchor issue, even though the cluster is not a true duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 36683 is the strongest representative issue here: active discussion, clear failure mode, and enough specificity to be actionable without being just a test or minor cleanup report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "Different bugs: Gemma3 config missing vocab_size vs MoE load-balancing loss handling." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "Both touch attention code, but one is GPT-2 scaling under SDPA/FlashAttention and the other changes AttentionInterface.register behavior." + }, + { + "left": "issue:36683", + "right": "issue:43881", + "accept": false, + "reason": "Both are load-time failures, but they affect different models and stem from different missing attributes/code paths." + }, + { + "left": "issue:43906", + "right": "issue:45753", + "accept": false, + "reason": "An isolated reproduction of an earlier issue is not the same as a specific Qwen3.5 image-loop regression." + }, + { + "left": "issue:43299", + "right": "issue:45753", + "accept": false, + "reason": "Both mention Qwen3.5-era models, but one is MoE loading and the other is multimodal infinite-loop generation." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron FlashAttention 2 support gating is unrelated to Python version compatibility." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "UMT5 missing embed_tokens weights and Qwen3.5 MoE from_pretrained errors are different model-specific load failures." + }, + { + "left": "issue:43441", + "right": "issue:44380", + "accept": false, + "reason": "Both involve attention, but one is unsupported FlashAttention on Ministral-3 and the other is ignored attention scaling on GPT-2 backends." + }, + { + "left": "issue:43352", + "right": "issue:45478", + "accept": false, + "reason": "Same broad topic family, but 45478 is a different Qwen3.5 MoE loading regression, not the Nemotron FlashAttention issue." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both are import errors, but one is about compiled Python files and the other is a torch.jit parsing problem in DebertaV2Model." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "State-dict ordering load failures and Gemma pipeline roundtrip failures are unrelated." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "Generic state_dict loading errors are not the same as a missing Qwen3_5MoeVisionConfig field." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention backend config bug vs degenerate zero-loss classification behavior." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "Both are attention-related, but they affect different models and different backend behavior." + }, + { + "left": "issue:36683", + "right": "issue:44117", + "accept": false, + "reason": "Different missing-attribute regressions in different parts of the loading stack." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIPOutput attentions regression is not the same bug as GPT-2 attention scaling being ignored." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Import-time slowdown from wildcard imports and intermittent import slowness are related only at a very high level, not the same defect." + }, + { + "left": "issue:43097", + "right": "issue:43352", + "accept": false, + "reason": "Tie-embeddings API removal and FlashAttention support are unrelated regressions." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Both are test flakiness, but one is missing set_seed and the other is an xdist file-race issue." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "Trainer optimizer-state corruption and a cache-strategy typo are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Both are Python compatibility reports, but for different versions and likely different root causes." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Different model save/load regressions: state_dict ordering versus incorrect visual encoder keys." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training bug and tp-related OOM when loading a large model are not the same underlying issue." + }, + { + "left": "issue:43064", + "right": "issue:44655", + "accept": false, + "reason": "Wrong optimizer states during training is unrelated to saving Pipeline objects." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch ImportError and tokenizer-loading regression are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "Both are CLIP-related, but one is missing attentions output and the other is structural divergence between two CLIP text models." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "Attention-output regression and AMP/CUDA loss crash are different code paths and symptoms." + }, + { + "left": "issue:43618", + "right": "issue:44977", + "accept": false, + "reason": "CLIP attentions regression is unrelated to Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer implementation details are unrelated to DeepseekV3MoE diverging from the remote implementation." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "local_files_only network leakage and AddedToken keyword conflicts are different bugs." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Seed-related CI flakiness is unrelated to a specific multimodal model load failure." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Missing set_seed in tests and stale device override behavior are different CI issues." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "A flaky test seed issue is not the same as broken auto_mappings references causing CI failures." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained behavior and a missing tensor-parallel plan are separate model-export/parallelism 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": 18, + "review_unit_index": 48, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41628", + "issue:42548", + "issue:42757", + "issue:42831", + "issue:43010", + "issue:43299", + "issue:43502", + "issue:43526", + "issue:43576", + "issue:43723", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43901", + "issue:43976", + "issue:44222", + "issue:44230", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44368", + "issue:44387", + "issue:44556", + "issue:44623", + "issue:44655", + "issue:44749", + "issue:44908", + "issue:44938", + "issue:44961", + "issue:45071", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45375", + "issue:45468", + "issue:45478", + "issue:45656", + "issue:45850", + "issue:46132", + "issue:46209" + ], + "soft_pairs": [ + "issue:43576|issue:46132", + "issue:44355|issue:44749", + "issue:45125|issue:45478", + "issue:42548|issue:44351", + "issue:44387|issue:44556", + "issue:30333|issue:33453", + "issue:45125|issue:45375", + "issue:43010|issue:45278", + "issue:30990|issue:31795", + "issue:43723|issue:44556", + "issue:43502|issue:44222", + "issue:42831|issue:43874", + "issue:43010|issue:45850", + "issue:43502|issue:44961", + "issue:43010|issue:44368", + "issue:30333|issue:36296", + "issue:43827|issue:44961", + "issue:44355|issue:45071", + "issue:44355|issue:44655", + "issue:44315|issue:44360", + "issue:33453|issue:34689", + "issue:43299|issue:44556", + "issue:42757|issue:43576", + "issue:43576|issue:45071", + "issue:33453|issue:36296", + "issue:43010|issue:45656", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:44222|issue:44230", + "issue:43010|issue:43976", + "issue:43576|issue:44938", + "issue:44908|issue:44961", + "issue:44315|issue:46209", + "issue:41628|issue:42548", + "issue:36296|issue:43901", + "issue:36296|issue:43526", + "issue:44623|issue:45230" + ], + "prepared_review_unit_hash": "e2b7280b697fa473078d4a3a5b5dd904c396eb613d827832bb7c0eac9f13d811", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38487, + "estimated_input_tokens": 9622, + "estimated_eval_tokens": 19500 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly a grab bag of unrelated Transformers regressions, import errors, docs bugs, and model-specific failures. The soft pairs are only superficially similar by wording; none look like the same concrete defect. The broadest representative issue is the v5 regression umbrella.", + "confidence": 0.63, + "canonical_issue_reason": "issue:45850 is the broadest umbrella here: it captures the general v5 breakage theme that many of the other regression-style reports superficially resemble.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 best represents the cluster because it is the most general, current, and central user-facing breakage report among otherwise unrelated issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "Different bugs: broken `transformers env` CLI vs `AttentionInterface.register` behavior change." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated symptoms: compiled Python file errors vs a data-filtering slowdown after upgrade." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5 Moe-related, but one is missing tensor-parallel plan data and the other is a `from_pretrained` loading failure; not the same code path." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Different missing exports: `PreTrainedModel` vs `HybridCache`." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Memory/OOM regression under int4 quantization is not the same as failing to reload a checkpoint after upgrading." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLFlow reporting on failed training jobs is unrelated to tokenizer loading regression." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Tensor-parallel plan omission is separate from `deepstack_visual_indexes` being dropped by strict config handling." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "`@torch.no_grad` cache update issue is unrelated to broad import failures after upgrade." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "A loading hang in Sentence Transformers is not the same as docs confusion around `input_ids` and `past_key_values`." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Both mention loading, but one is tokenizer loading in v5 and the other is checkpoint reloading after upgrade; different failures." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Unexpected hub API access with `local_files_only=True` is unrelated to FP8 `save_pretrained` behavior." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "FP8 accuracy regression and a missing image-patch helper causing an AttributeError are different bugs." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "A specific cache-update decorator bug is not the same as a broad claim that v5 breaks many models." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "`local_files_only` network access bug is unrelated to the placeholder issue titled `racoon`." + }, + { + "left": "issue:43010", + "right": "issue:44368", + "accept": false, + "reason": "Cache update no-grad behavior has nothing to do with a warning about `tie_word_embeddings`." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLFlow status reporting and tensor-parallel training behavior are distinct problems." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing `pipeline()` after removals is unrelated to the placeholder `racoon` issue." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file errors and `PretrainedConfig` type-checking breakage are different failure modes." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Compiled Python file handling is unrelated to saving Pipeline objects." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel not applied with `model_init` is unrelated to a missing ReLU in the DSA indexer." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer-loading regression and Llama 3.2 Vision model-loading regression are different model/code-path issues." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "A Qwen3VL MoE loading breakage is not the same as a general checkpoint reload regression after upgrade." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing `is_offline_mode` export and broken `transformers env` CLI are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "Broken environment command and `PretrainedConfig` type-checking regression do not describe the same defect." + }, + { + "left": "issue:33453", + "right": "issue:36296", + "accept": false, + "reason": "Tokenizer loading regression and tensor-parallel training bug are different issues." + }, + { + "left": "issue:43010", + "right": "issue:45656", + "accept": false, + "reason": "Cache update no-grad behavior is unrelated to DeepSpeed calling optimizer.step twice." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI `transformers env` breakage is unrelated to missing Qwen2.5-VL import symbols." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and an `ImportError` for `layer_type_validation` are not the same bug." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "Placeholder issue `racoon` has no clear relation to Gemma-4 audio positional encoding." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both touch FP8/MoE, but one is a save/load bug and the other is an inference-support request; not mergeable as one fix." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update decorator behavior is unrelated to the Python version compatibility report." + }, + { + "left": "issue:43576", + "right": "issue:44938", + "accept": false, + "reason": "Broken CLI command and Python 3.14 load failure are separate problems." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs bug and placeholder `racoon` issue do not match." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger Kernel application with `model_init` is unrelated to FP8 inference failing on B200 GPUs." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Different import errors for different symbols from `transformers`." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor-parallel training bug and pipeline docs mismatch are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor-parallel training bug is unrelated to `reduce_labels` returning the wrong shape/value in BeitImageProcessorFast." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Missing files from `processor.save_pretrained(...)` is specific; the other item is an empty generic bug report." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 18, + "review_unit_index": 49, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "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:43874", + "issue:43976", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44373", + "issue:44393", + "issue:44492", + "issue:44556", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45341", + "issue:45362", + "issue:45478", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46143", + "issue:46144", + "issue:46218" + ], + "soft_pairs": [ + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44273|issue:45003", + "issue:44393|issue:45071", + "issue:44230|issue:46095", + "issue:44373|issue:44961", + "issue:43576|issue:44556", + "issue:44393|issue:45753", + "issue:41669|issue:43673", + "issue:43525|issue:45850", + "issue:44556|issue:45092", + "issue:43572|issue:45850", + "issue:45362|issue:45850", + "issue:41669|issue:46218", + "issue:44246|issue:44273", + "issue:42890|issue:44908", + "issue:43408|issue:44315", + "issue:44704|issue:44908", + "issue:43352|issue:44393", + "issue:44297|issue:45230", + "issue:43976|issue:44393", + "issue:44295|issue:45230", + "issue:43576|issue:44908", + "issue:44908|issue:45071", + "issue:41669|issue:44351", + "issue:41669|issue:45341", + "issue:45278|issue:45850", + "issue:43673|issue:44492", + "issue:46143|issue:46144", + "issue:43874|issue:45850", + "issue:44393|issue:45478", + "issue:44077|issue:44861", + "issue:43976|issue:46218", + "issue:43827|issue:44908", + "issue:31795|issue:33453", + "issue:41669|issue:44492", + "issue:44315|issue:44829", + "issue:42371|issue:44492", + "issue:43576|issue:44273" + ], + "prepared_review_unit_hash": "44c8d08d29d64678ff928112b2f1aab3bdb582492d4a4b4add479835883b1be8", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 38785, + "estimated_input_tokens": 9697, + "estimated_eval_tokens": 19650 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a noisy cluster of unrelated issues. The only coherent theme is import/lazy-loading/v5 import regressions, but none of the soft pairs are close enough to treat as duplicates or the same fix.", + "confidence": 0.86, + "canonical_issue_reason": "issue:44273 is the broadest umbrella for the import/lazy-loading regression theme and best matches the shared symptom family across the relevant issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:41669 is the most concrete root-cause report in the cluster: it identifies import * usage as the source of the import-time slowdown and is the most actionable issue to anchor the duplicate set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Both are runtime/import-related at a high level, but one is about lazy loading and the other is a TensorFlow backend warning path; different bugs and fixes." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Different model families and different failures: Qwen3-VL-Moe inference bug versus missing tensor-parallel plan in Qwen3.5 MoE." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "One is a broad lazy-loading complaint; the other is a specific sys.modules access bug in modeling_utils. Related area, not the same underlying defect." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Completely different surfaces: Qwen3-VL 2D bounding-box output bug versus PretrainedConfig type-checking breakage." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "Different features and code paths: FP8 inference support for Qwen models versus a DeepSeek weight-conversion question." + }, + { + "left": "issue:44373", + "right": "issue:44961", + "accept": false, + "reason": "A docstring typo about position_ids is unrelated to the placeholder 'racoon' issue." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Both mention v5, but one is the transformers env command and the other is checkpoint reload compatibility; not the same bug." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen3-VL/Qwen3.5, but the symptoms differ: 2D box output failure versus an infinite-loop case on a specific image." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-performance regression versus missing GenerationMixin cache during chunked prefill; different root causes and fixes." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "Specific Llama4Config pad_token_id error versus a broad complaint that v5 breaks many models." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are upgrade-compatibility problems, but they affect different model-loading paths and are not clearly the same defect." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression is a specific config issue, not the same as the broad v5 breakage complaint." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "A Qwen3.5 chat crash is a model-specific failure, not the same as the generic 'v5 breaks too many models' report." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import * slowdown and a missing Qwen3_5ForCausalLM import error may both involve imports, but the underlying problems are different." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Both concern import behavior, but one reports intermittent slow imports while the other claims lazy loading is broken; not enough evidence of the same bug." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Test seed omissions in SAM HQ tests have nothing to do with the inverse_sqrt scheduler ignoring kwargs." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Model type mismatch warning for SAM3 versus Liger Kernel not applied in model_init are unrelated." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor cached_file kwargs forwarding and scheduler kwargs handling are separate code paths." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention 2 support error in Nemotron is unrelated to Qwen3-VL 2D bounding-box output behavior." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to a generic bug report placeholder." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility regression is unrelated to the Qwen3-VL bounding-box bug." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "Position_ids buffer access error is unrelated to a generic bug report placeholder." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken env command and inverse_sqrt scheduler kwargs are separate issues." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler keyword handling and PretrainedConfig type-checking are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import * cleanup and missing HybridCache export may both touch imports, but they are different missing-symbol/import surfaces." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import-time slowdown is unrelated to a small testing_utils bug." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "General import errors after upgrade and a broad claim that v5 breaks many models are too vague to be the same underlying issue." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked-prefill cache regression and a typo in cache strategies are unrelated." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Missing kwargs passthrough and attention-interface noncompliance in RoFormer are related to the same model family but not the same concrete defect." + }, + { + "left": "issue:43874", + "right": "issue:45850", + "accept": false, + "reason": "A specific GLM46V image-processor method bug is not the same as the broad v5 breakage complaint." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen-related failures with different symptoms and code paths." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init validation and tied-weight-key AttributeError are distinct issues." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python version incompatibility and a missing Qwen3_5 import are separate failures." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Documentation still referencing pipeline() is unrelated to scheduler kwargs handling." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Model.forward documentation confusion and tokenizer loading regression are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44492", + "accept": false, + "reason": "Import * performance regression and a cache-strategy typo are not the same bug." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Liger Kernel not applied in model_init and flash_attention_3 causing degenerate training are different changes/issues." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API guidance and a cache-strategy typo are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken transformers env command and lazy loading issues overlap on version context only; they do not look like 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": 18, + "review_unit_index": 50, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:42936", + "issue:43066", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43606", + "issue:43668", + "issue:43704", + "issue:43717", + "issue:43827", + "issue:44031", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45230", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46144", + "issue:46172", + "issue:46218" + ], + "soft_pairs": [ + "issue:44360|issue:45468", + "issue:44655|issue:44861", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:43576|issue:46218", + "issue:43827|issue:44492", + "issue:46018|issue:46172", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:31515|issue:31795", + "issue:44485|issue:45468", + "issue:43704|issue:44485", + "issue:43298|issue:46095", + "issue:45850|issue:46218", + "issue:43502|issue:43519", + "issue:42371|issue:43704", + "issue:43519|issue:44623", + "issue:43519|issue:43704", + "issue:43606|issue:45468", + "issue:44263|issue:45468", + "issue:44261|issue:45468", + "issue:44351|issue:46218", + "issue:43352|issue:46218", + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44861|issue:45290", + "issue:43525|issue:44861", + "issue:43704|issue:44998", + "issue:43572|issue:44861", + "issue:41669|issue:44998", + "issue:43066|issue:44998", + "issue:31515|issue:38617", + "issue:44230|issue:44623", + "issue:44998|issue:46218", + "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" + ], + "prepared_review_unit_hash": "614ce861ade4edddad97d1c3049cda05ce228da564bbd3fc7eafdf10e05eb78c", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 44, + "serialized_chars": 39956, + "estimated_input_tokens": 9989, + "estimated_eval_tokens": 20234 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are separate regressions, API changes, docs issues, or model-specific bugs that only share a broad Transformers/v5 similarity. I found no soft-edge pair that clearly describes the same underlying defect/change, so all candidates should be rejected.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45850 is the broadest representative of the cluster\u2019s recurring v5 breakage theme, even though it is more of an umbrella complaint than a precise duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45850 best matches the overall cluster topic because it captures the general \u201cv5 breaks many models\u201d regression pattern seen across several items.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "Different model families and failures: DSA indexer/ReLU vs Gemma audio positional encoding." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "One is save_pretrained for Pipeline objects; the other is _get_tied_weight_keys crashing on a list." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report vs a specific Qwen3.5 MoE from_pretrained error; no concrete match." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading behavior and inverse_sqrt scheduler kwargs are unrelated code paths." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command vs import failure for Qwen3_5ForCausalLM are different issues." + }, + { + "left": "issue:43827", + "right": "issue:44492", + "accept": false, + "reason": "Docs referencing pipeline() and a cache-strategy typo are not the same defect." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek-V4 shared expert gating vs a deepseek-v4 tensor transpose issue are distinct." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are separate model bugs." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp math and scheduler kwargs handling are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star cleanup in models does not match a VRAM leak in Qwen3ForCausalLM." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Documentation confusion and a Llama 3.2 model-loading regression are different problems." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Slow checkpoint loading is not the same as confusing forward docs." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "RoPE implementation and Gemma4AudioRelPositionalEncoding are distinct model-specific changes." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "A VRAM leak in multi-threaded loading and a RoPE bug do not share the same failure mode." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching support for audio models is unrelated to a DeepSeek-V4 weight conversion script request." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "Broad v5 breakage complaint vs a specific Qwen3_5 import failure; too coarse to merge." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "local_files_only network requests and Qwen3VL timestamp calculation are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 settings API guidance vs VRAM leakage are different issues." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp logic and missing save_pretrained files are separate bugs." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Different components: Qwen3VL processor timestamps vs Qwen3ForCausalLM memory leak." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch in bark-small is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return values in GlmMoeDsaIndexer do not match Gemma audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "MLA q_a_layernorm epsilon precision issue is a different model-specific bug." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "HybridCache import failure and Qwen3_5ForCausalLM import failure are different symbols and code paths." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention support gap in NemotronHForCausalLM is not the same as Qwen3_5 import failure." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both are tokenizer-related but describe different v5 symptoms and likely different code paths." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug and an unrelated non-technical issue are not duplicates." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "_tied_weights_keys list crash and apply_chat_template tool-call crash are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is not the same as tied-weights key handling." + }, + { + "left": "issue:43704", + "right": "issue:44998", + "accept": false, + "reason": "VRAM leak vs unrelated non-technical issue." + }, + { + "left": "issue:41669", + "right": "issue:44998", + "accept": false, + "reason": "Import-star cleanup issue vs unrelated non-technical issue." + }, + { + "left": "issue:43066", + "right": "issue:44998", + "accept": false, + "reason": "Tokenizer decoder type bug vs unrelated non-technical issue." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading and missing layer_type_validation import are different." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "fp8 support request for qwen3-vl/qwen3.5 MoE and missing processor save files are unrelated." + }, + { + "left": "issue:44998", + "right": "issue:46218", + "accept": false, + "reason": "Unrelated non-technical issue vs Qwen3_5 import failure." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLFlow job-status reporting and checkpoint load slowness are unrelated." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint loading slowness and tensor-parallel training bug are different concerns." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "fp8 inference support request and Qwen3.5 MoE from_pretrained failure are not the same bug." + }, + { + "left": "issue:31515", + "right": "issue:33453", + "accept": false, + "reason": "Checkpoint loading performance regression and tokenizer loading regression are distinct." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Audio-model continuous batching support and qwen3-vl/qwen3.5 fp8 inference support are different feature gaps." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba init_weights usage and RoFormer attention interface are unrelated model implementations." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2former ignore_value handling is not the same as DeepSeek-V4 shared-expert gating." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "Incorrect type hint in ModernBERTConfig and tokenizer regex 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": 18, + "review_unit_index": 51, + "review_unit_count": 51, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-439", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031", + "issue:46144" + ], + "soft_pairs": [ + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "2b10f80c507cdcef3e1eb7c83b3f0fac260952bb51775e571bc950351fc6d7b7", + "component_budget": { + "node_count": 439, + "item_count": 439, + "soft_pair_count": 1616, + "serialized_chars": 837663, + "estimated_input_tokens": 209416, + "estimated_eval_tokens": 419088 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 9, + "serialized_chars": 6639, + "estimated_input_tokens": 1660, + "estimated_eval_tokens": 3576 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are not duplicates; they span unrelated areas including a norm-layer question, Mask2Former ignore_value handling, loss kwargs propagation, a config type hint, a typo, tokenizer regex warnings, and RoFormer attention interface usage.", + "confidence": 0.96, + "canonical_issue_reason": "No true canonical duplicate exists in this set. If one issue must anchor the cluster, issue 43240 is the broadest/general framework bug, but it is still unrelated to the other reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43240 is the closest thing to a cluster anchor because it concerns generic loss-call plumbing rather than a single model-specific typo or configuration detail.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Mask2Former ignore_value handling and loss-kwargs propagation are different bugs in different code paths." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss kwargs forwarding and RoFormer attention interface usage are unrelated implementation issues." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "A Mask2Former loss-parameter bug is not the same as tokenizer regex warning behavior." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "These concern different subsystems: segmentation loss handling vs attention interface usage in RoFormer." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2Former ignore_value logic and a typo in a quantization attribute are unrelated." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Passing kwargs to a loss function and an incorrect type hint in ModernBERTConfig are not the same underlying bug." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "A config type-hint error is unrelated to a typo in a quantization field name." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "A question about Llama4TextL2Norm vs RMSNorm is unrelated to a ModernBERTConfig annotation problem." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former loss-parameter handling and a ModernBERT config typing issue do not share the same code path or fix." + } + ] + }, + "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": 18, + "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": "5ff925bcca17d8cae1050ee7f139f04be2298d337c4cdda928b13aeb67b7837d", + "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 unrelated: one is about truncation behavior in tokenization, one about ReduceLROnPlateau-like scheduler functionality, and one about Wav2Vec2 word time offsets. No duplicate cluster emerges.", + "confidence": 0.98, + "canonical_issue_reason": "Issue #10519 is the clearest concrete product request and best represents the only coherent topic in the set; the other two issues are unrelated feature requests in different areas.", + "canonical_pr_reason": null, + "best_issue_reason": "#10519 is the most specific and actionable issue; it is the best standalone representative, but not a duplicate of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Token truncation options and Wav2Vec2 word boundary timestamps are different features with no shared code-path or concrete bug overlap." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Truncation-from-beginning is unrelated to ReduceLROnPlateau-like scheduler behavior; they concern different subsystems and problems." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "A learning-rate scheduler feature request and Wav2Vec2 time-offset extraction are unrelated issues with no plausible duplication." + } + ] + }, + "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": 18, + "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": "abe46e12df50bd2b8975547c25fe78a4b98496b57276d0e9899e0c931253d464", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster looks mixed: one Tapas tokenization issue and two model-loading issues. The two model-loading tickets are superficially similar, but the available evidence is not strong enough to treat them as the same concrete bug, so I would not merge any soft edges.", + "confidence": 0.84, + "canonical_issue_reason": "Issue #16618 is the broadest and most central model-loading report in the set, with the most discussion and the clearest user-facing failure around loading a pretrained model.", + "canonical_pr_reason": null, + "best_issue_reason": "#16618 is the best representative of the cluster because it captures the main model-loading problem theme, whereas #13244 is unrelated and #15980 is narrower and more specific to private-model login/error messaging.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Different bug class and code path: Tapas tokenization behavior versus inability to load a pretrained model." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Both concern model access/loading, but one is specifically about a private-model download/auth error and the other is a generic failure to load 'bert-base-uncased'; not enough evidence they are the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 4, + "component_count": 18, + "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": "89dea94304e87b13a1de33358b66b54f86d807632925567de5212ed71dd7aa1e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate community contribution feature requests: one for Flash Attention 2 support on more architectures, the other for GGUF support on more architectures. They are only loosely similar at a high level and should not be treated as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 26350 is the better anchor only because it is older and has substantially more discussion activity, but it is not a duplicate of 33260.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 26350 is the strongest representative issue in this pair due to age and higher engagement, though the topics are distinct.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different requested features in different inference/runtime paths: Flash Attention 2 vs GGUF support. The similarity is only broad subsystem-level language about architectures, not the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 5, + "component_count": 18, + "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": "252ce62d8d69026efe4b1c8607c6e7193f685d2145e253ea524fc363b335a77d", + "component_budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13529, + "estimated_input_tokens": 3383, + "estimated_eval_tokens": 7022 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13529, + "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 mostly boilerplate-heavy i18n translation requests with strong textual overlap. The most defensible duplicate is the Thai pair 46222/46226, which has identical visible titles and a jaccard score of 1.0; the packet does not expose the body text, so it should be treated as a likely duplicate rather than an exact one. The other pairs look very similar, but the available evidence is still too template-driven to merge confidently.", + "confidence": 0.81, + "canonical_issue_reason": "issue:46222 is the best canonical issue for the one highly grounded duplicate pair: it is the earlier of 46222/46226 and the packet shows identical visible titles plus jaccard 1.0. Since the body text is not exposed, this is best described as a likely duplicate rather than a proven exact duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46222 is the strongest canonical issue candidate because the 46222/46226 pair has the clearest duplicate signal in the packet: same visible title, identical body length, and jaccard 1.0, with 46222 appearing first.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Very similar translation-template issues, but the packet only shows normalized placeholder-style metadata and no body text; I can't confirm they are the same concrete request." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "High similarity appears template-driven, but there is not enough exposed content to prove the underlying request is identical." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "These look like related translation-request templates, but the available evidence is still too indirect to treat them as duplicates." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Strong shared i18n wording, but the packet does not expose enough request-specific content to merge these issues confidently." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Same broad docs-translation pattern, but the evidence shown is template-level similarity rather than the same underlying request." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Both are highly similar translation issues, but the packet lacks the underlying body text needed to confirm they are the same request." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "The overlap looks like shared boilerplate; without the actual body text, I would not merge them as duplicates." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Very similar i18n translation-template issues, but the exposed data is not specific enough to establish the same underlying task." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "High similarity alone is not enough here; the packet does not expose the concrete request text needed to confirm duplication." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "These are closely related translation-request issues, but the similarity appears to come from shared boilerplate rather than a confirmed duplicate." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "The packet suggests a template match, but it does not show enough underlying content to safely merge them." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Strong textual overlap, but the available evidence is still not sufficient to prove identical underlying content." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "These may be related i18n template issues, but the exposed metadata is not enough to confirm they are the same request." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "High similarity, but the packet does not expose the body text required to verify a duplicate relationship." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "The titles are placeholder-like and the similarity is high, but the packet does not provide enough concrete content to confidently merge them." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "Despite near-identical similarity scores, the available evidence is still too template-driven to call this a duplicate with confidence." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Very close text match, but the visible titles differ and the packet does not expose the body text, so I would not merge them confidently." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "High similarity alone is not enough here; the issue bodies are not visible, so the duplicate claim is not well grounded." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "These are extremely similar, but without the underlying body text I can't be sure they are the same concrete issue." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "The similarity looks template-based, but the packet does not expose enough request-specific content to confirm duplication." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Very close similarity, but the evidence shown is still not sufficient to merge them as duplicates." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "The overlap is strong, but the packet only provides similarity metrics and titles, not the body text needed for a confident merge." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "High similarity alone does not establish the same underlying request when the body text is not visible." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Highly likely duplicate pair: same visible Thai title, identical body length, and jaccard 1.0. The body text is not shown, so this is best treated as a very strong duplicate rather than an exact one." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The conclusion is a bit too strong for the packet. The Thai pair 46222/46226 looks very likely duplicate from identical titles, identical body length, and jaccard 1.0, but the packet does not expose the body text, so calling it an \"exact duplicate\" / \"same Thai title/body\" is not fully grounded. Also several soft-edge rejections rely on language-specific distinctions that are not directly visible in the packet for the placeholder-title issues. Please soften the canonical verdict to match the available evidence." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 18, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-16", + "nodes": [ + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36879|issue:41211", + "issue:36879|issue:43909", + "issue:36879|issue:44121", + "issue:36879|issue:43837", + "issue:36879|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:36879|issue:37015", + "issue:37015|issue:42503", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42738", + "issue:42738|issue:45027", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43909|issue:44205", + "issue:43671|issue:43909", + "issue:44121|issue:45322", + "issue:43837|issue:44121", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133" + ], + "prepared_review_unit_hash": "5f2cb292ac0290508727adf178fe662622e495b7c2786b2dcc97cb7efe0d6096", + "component_budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21658, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 + }, + "budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21658, + "estimated_input_tokens": 5415, + "estimated_eval_tokens": 11086 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are broad model-request issues with only template-level similarity. They cover different model families and modalities, so they should not be collapsed as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue fits: the set mixes unrelated requests (text, vision-language, audio, OCR, detection, configs) rather than the same underlying bug or feature.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue cleanly subsumes the others; keeping any one as canonical would be misleading because the requested changes are distinct.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "EuroBert config request vs DEIMv2 model request; different models and code paths." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "Model config addition vs Qwen3-TTS support; different modality and implementation scope." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert config vs RF-DETR model; unrelated architectures." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are different model families with different integration work." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR object detection request vs LFM2.5 Audio model; different underlying feature." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR model request vs weight-sparse transformer request; not the same change." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR detection model vs Qwen3-ASR support; different modalities." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR model request vs Qwen3-TTS support; unrelated model additions." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT addition vs Qwen3-ASR support; different models and functionality." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT model request vs sparse transformer request; different code changes." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT vs audio model support; no shared underlying bug/change." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are distinct model integrations." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT request vs ModernVBERT models; different model families." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture request vs Qwen3-ASR support; different modalities." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 architecture addition vs sparse transformer proposal; unrelated work." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 model architecture vs ModernVBERT models; not the same change." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 architecture request vs Voxtral TTS support; different model paths." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are different model requests with no shared concrete fix." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT models vs Qwen3-ASR support; separate model integrations." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT vs Qwen3-TTS; different model families and modalities." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT models vs generic BERT-like RoPE request; only broad semantic overlap." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like RoPE request vs Voxtral TTS support; unrelated feature additions." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS support vs Qwen3-ASR support; related branding but distinct concrete capabilities." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS support vs Voxtral TTS support; different models and implementations." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS support vs DeepSeek-OCR2; different modality and code path." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support vs sparse transformer request; not the same underlying change." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR support vs LFM2.5 Audio; both audio-related but different model requests." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR support vs EUPE encoder; different model families and tasks." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio vs SAM3-LiteText; different model domains." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-TTS support vs LFM2.5 Audio; both audio-themed but not the same model or change." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Sparse transformer request vs EUPE encoder; distinct architectural changes." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support vs sparse transformer proposal; different feature areas." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR support vs Voxtral TTS support; ASR and TTS are separate additions." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT model request vs Voxtral TTS support; unrelated model work." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR support vs DeepSeek-OCR2; different modality and implementation." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE encoder request vs TIPSv2 vision-language pretraining; different model proposals." + } + ] + }, + "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": 18, + "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": "06ab3184c07df60a20cebabd2fa10ca4cb48c3dfa69200c2d2426d734e32e7d9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are unrelated: one is about block decoding, the other is a DeepSeek v3.2 support request. No duplicate/canonical relationship is evident.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the titles and likely problem domains do not match.", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a true duplicate, but 43037 is the more specific user-facing feature request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different topics and no shared target or code-path evidence; this does not look like 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": 18, + "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": "7f149f6076aacee460fc076144ce5b59c3eae407aee954396870ea2ac2a99267", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1583, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1583, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related to Trainer customizability, but they describe different problems: one is about `remove_unused_column`, the other about FLOPS estimation for custom trainers. They should not be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #43598 is the better canonical issue only if the cluster is centered on Trainer customizability, because it is a concrete request to revisit a specific Trainer behavior (`remove_unused_column`).", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a strong global canonical for the pair; if forced to pick one, #43598 is slightly more concrete and narrowly scoped than the FLOPS-estimation request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "They are adjacent Trainer customization topics, but not the same underlying bug or change. One concerns column removal behavior, the other concerns FLOPS counting for custom trainers." + } + ] + }, + "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": 18, + "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": "a5037a0e4d9d2af8461aa0dfb211a502ae01d0fc29824daccf816c7691b989b3", + "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 not duplicates: one requests general AutoModel/AutoModelForConditionalGeneration registration for Qwen3-Omni, while the other requests AutoModelForSequenceClassification support for Qwen3.5. They involve different model families and different auto classes/capabilities.", + "confidence": 0.94, + "canonical_issue_reason": "No true canonical duplicate here; if one must be chosen as the representative issue, #43684 is broader and more central because it concerns model registration in core auto classes.", + "canonical_pr_reason": null, + "best_issue_reason": "#43684 is the better representative issue because it describes the broader auto-model registration gap, whereas #44405 is a separate, more specific sequence-classification support request for a different config/model line.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying requests: Qwen3-Omni auto-registration vs Qwen3.5 sequence-classification support. Different model/config targets and different auto classes, 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": "issue", + "component_index": 10, + "component_count": 18, + "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": "549bef0f323a46daa752b8ecd0190a9ad53433bbb35a421f48d9f50bcdafcbc5", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three issues share the same title/body pattern (\u201chuggingface \u201d) and look like duplicate placeholders; the unrelated issue \u201cFett\u201d does not belong in that cluster.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43845 is the earliest of the three near-identical \u201chuggingface \u201d issues and is the best representative of that duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43845 best fits the cluster because it matches the repeated title and is the earliest of the matching issues; issue:43835 appears unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles with no shared content beyond generic similarity; no evidence they describe the same issue." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles with no concrete overlap in description or targets." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different titles and no shared issue details; not the same underlying bug." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "Same title (\u201chuggingface \u201d) and essentially identical sparse content, so they look like duplicates." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "Same title and no differentiating details; likely the same duplicate issue." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Same title and matching minimal content, indicating the same underlying issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering is grounded in the packet: issues 43845/43846/43859 share the same title and have strong soft similarity, while 43835 is distinct and only weakly similar. The chosen canonical issue is the earliest among the matching set, and the soft-edge verdicts are conservative." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 11, + "component_count": 18, + "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": "d6d401e5dfe42ffe0846e5094c05d1bf062e3e8e89c2b9fbd597572392c2a7f9", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster looks like unrelated low-signal/spam issues rather than true duplicates. None of the soft pairs should be merged.", + "confidence": 0.93, + "canonical_issue_reason": "issue:44418 is the only item with a non-spam, vaguely substantive title, so it is the best representative if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44418 is the most suitable issue to anchor the cluster because it is at least a recognizable documentation request, unlike the placeholder spam titles.", + "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 show the same underlying issue; they read as separate spam submissions with no concrete shared bug or change." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Same placeholder spam title, but no evidence of one underlying report; treat as distinct spam issues." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "'' vs a documentation suggestion are clearly different topics." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "'' and 'Check out \"Google LLC\"' are unrelated and look like separate spam-like submissions." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "A documentation enhancement request is not the same underlying issue as a spam placeholder." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "These titles point to different kinds of reports; no duplicate evidence beyond generic similarity." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Identical spam titles alone do not establish the same bug or change; they could be independent junk submissions." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Placeholder spam and a branded promotional title are not the same underlying issue." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "These look like different spam artifacts, not one fixable problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 12, + "component_count": 18, + "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": "1e5819a01de7f68854674bdae2d138309ce027dc7110d6bd478612817547dc0a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues share an identical title and soft similarity signals, so they likely describe the same report; however, the available packet does not expose the issue bodies, so the duplicate conclusion should be kept conservative.", + "confidence": 0.86, + "canonical_issue_reason": "#45145 is the better canonical issue because it was created first and already has an inbound reference, which makes it the more likely original thread.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the best issue candidate because it is the earliest of the two and has the only inbound reference, while the evidence available here does not distinguish a different scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "The issues have the same title and a soft similarity signal; that is enough to treat them as likely duplicates, but the packet does not show the bodies, so the match is based on title/similarity rather than any confirmed identical proposal." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The packet supports that these two closed issues are highly similar and share the same title, but the analyst overstates the evidence in a few places. In particular, the packet does not show the bodies themselves, only similar body lengths, so claims like \u201cnearly identical body size,\u201d \u201csame broad proposal,\u201d and \u201cno evidence of different scope\u201d go beyond what is grounded. The duplicate-issue conclusion may be plausible, but the soft-edge verdict should be stated more conservatively based on title match and soft similarity only." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 18, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45295-2", + "nodes": [ + "issue:45295", + "issue:45373" + ], + "soft_pairs": [], + "prepared_review_unit_hash": "f5fd95fada2f530ab0999e343c17c16266cde0c7182bb22509972ca44827cd49", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 0, + "serialized_chars": 1286, + "estimated_input_tokens": 322, + "estimated_eval_tokens": 900 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues describe the same missing Gemma 4 sequence-classification support; the later, more specific open issue is the better cluster representative.", + "confidence": 0.94, + "canonical_issue_reason": "issue #45373 is the clearest canonical tracker: it is open, explicitly names the missing class `Gemma4ForSequenceClassification`, and states the module-level gap relative to Gemma 2/3.", + "canonical_pr_reason": null, + "best_issue_reason": "issue #45373 best captures the concrete request and likely remaining work, while #45295 is a shorter earlier duplicate with less detail and is already closed.", + "best_pr_reason": null, + "soft_edge_verdicts": [] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 18, + "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": "842e6ab531aaabd33816066e9a3abce215014ae236d6614d858d3fa166350a5c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1670, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1670, + "estimated_input_tokens": 418, + "estimated_eval_tokens": 1092 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are related only at a high level (4-bit inference in the same stack) but describe different failures and code paths: one is a Gemma4 audio-module `torch.finfo()` TypeError on `uint8` weights, the other is a BitsAndBytesConfig `llm_int8_skip_modules` regression that removes the default `lm_head` exclusion and triggers an AssertionError.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45674 is the more general and actionable report: it names a concrete BitsAndBytesConfig regression, states the observed assertion failure, and is framed in a way that is easier to map to a fix than the narrower Gemma4-specific audio-module TypeError.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Different underlying bugs: 45672 is a Gemma4 audio-module `torch.finfo()` TypeError on `uint8` weights, while 45674 is a BitsAndBytesConfig skip-modules regression that clears `lm_head` exclusion and causes an AssertionError. Same general area, not the same fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 18, + "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": "cde04dc22fae4edbfc9b7902a316cb3c1d7aa61e8411e532d6e578c81d402acf", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1476, + "estimated_input_tokens": 369, + "estimated_eval_tokens": 994 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1476, + "estimated_input_tokens": 369, + "estimated_eval_tokens": 994 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are unrelated placeholder/spam-like reports with different personal names and dates; there is no evidence they describe the same bug or task.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good canonical representative because neither contains a real problem description or actionable context.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "Different named entities and dates with no shared technical context, code path, or problem statement; they do not appear to be 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": 16, + "component_count": 18, + "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": "c396abe2d662c192ee161454d9b074c3fd11fe3167013d7c72a97b3ce7ecb29e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are just title-only personal-name/date entries and do not describe the same underlying problem; the similarity is superficial only.", + "confidence": 0.98, + "canonical_issue_reason": "No clear canonical issue: both are unrelated, low-context entries with no substantive problem description.", + "canonical_pr_reason": null, + "best_issue_reason": "45764 is only marginally preferable as the earliest created issue, but it is not a meaningful representative of the pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "The titles share only a name/date pattern; there is no evidence they refer to the same bug, request, or code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 17, + "component_count": 18, + "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": "67356b72308b09767f276b3499a0f60c2a086f4d8ec0be629a5ac0e0cf0a579e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are not duplicates: one requests a new token-classification head for Qwen3_5, while the other asks for FP8 training support. They touch different features and code paths.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45810 is the more concrete and narrowly scoped request, so it is the better cluster representative, but it is not the same underlying problem as issue 46295.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different feature requests with no shared concrete bug or implementation target: model head addition for token classification versus FP8 training support." + } + ] + }, + "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": 18, + "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": "ca896772fffd2448fe667557ae093990a4a0554627487050933b3e2b4d627449", + "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": "The set splits into two exact-duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is related only at a broad financial-data theme level and should not be merged with the others.", + "confidence": 0.95, + "canonical_issue_reason": "45870 is the best canonical representative to keep for the first duplicate pair because 45870 and 45871 are identical, and 45870 is the earlier of the two.", + "canonical_pr_reason": null, + "best_issue_reason": "45870 is the strongest single issue to retain overall: it is an exact duplicate pair member, clearly representative of one concrete topic, and there is no PR evidence in the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Related financial-domain wording, but one is a dataset bundle listing and the other is an example request; not the same underlying issue." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Same as 45869/45870: broad theme overlap only, with different concrete topic and no evidence of an exact duplicate." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Exact title/body duplicate pair for the same example request." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both mention financial time series, but 45872 adds a scraping service component and is a different concrete request." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Same as 45870/45872: overlapping domain, but different underlying change/request." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Broad thematic similarity only; not the same issue or change." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Broad thematic similarity only; not the same issue or change." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Exact duplicate pair with matching title and body length, describing the same request." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 are the only clear duplicate pairs, while 45869 is only soft-related by theme. The canonical choice of the earlier duplicate is conservative, and the soft-edge rejects do not overstate the evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 30, + "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-39895-15", + "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", + "pull_request:46268" + ], + "soft_pairs": [ + "pull_request:39895|pull_request:43424", + "pull_request:41855|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:46267", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:46048|pull_request:46268", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "55973c6b1cd8e83e75b75c5a002e5d702118e53064d71ec73f13f3d07878cfa4", + "component_budget": { + "node_count": 15, + "item_count": 15, + "soft_pair_count": 20, + "serialized_chars": 40306, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 + }, + "budget": { + "node_count": 15, + "item_count": 15, + "soft_pair_count": 20, + "serialized_chars": 40306, + "estimated_input_tokens": 10077, + "estimated_eval_tokens": 20410 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly unrelated PRs. The only genuine duplicate-looking pair is the two `fixed_cross_entropy` patches (43251 and 43254); the rest target different models, docs, or kernel-routing changes and should stay separate.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 43251 is the best representative of the only true duplicate subgroup: it has the explicit issue link, touches the same `fixed_cross_entropy` code path as 43254, and is the more complete version of that fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "Different features: VideoPrism model addition vs an ExecuTorch dynamic-shape export test; no shared code-path bug." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Unrelated work: Mistral tokenizer methods vs ExecuTorch export testing." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Different subsystems: tokenizer API additions vs a new MobileLLM model implementation." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor auto-loading robustness and MobileLLM model scaffolding are distinct changes." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "`fixed_cross_entropy` kwargs fix is unrelated to adding MobileLLM." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Cross-entropy loss kwargs fix and Mistral4 test fixes are different code paths." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "ExecuTorch dynamic-shape export tests are unrelated to Mistral4 test/model changes." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Different areas: export testing vs new model implementation." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "Adding MobileLLM and fixing `cls.config` resolution are unrelated changes." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "MobileLLM model addition and Falcon output-interface refactor do not fix the same bug." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor and the model-creation skill are unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon standardized-output refactor is unrelated to Mistral4 test fixes." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 tests/model work and `layrnorm` renaming are separate issues." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46267", + "accept": false, + "reason": "Mistral4 tests and chat-response parsing updates are different features." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 test fixes and TDT loss-kernel integration are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 test/model work is unrelated to `cls.config` resolution order." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "No shared bug: Mistral4 fixes vs model-creation skill content." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "A new MobileLLM model and a model-creation skill are not the same underlying change." + }, + { + "left": "pull_request:46048", + "right": "pull_request:46268", + "accept": false, + "reason": "Both mention hub kernels, but they add different kernel entries for different purposes and are not the same concrete fix." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both patches fix the same `fixed_cross_entropy` implementation in `loss_utils.py`, add the same missing kwargs support, and target the same issue (`43240`); they could be merged into one PR." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted pair, 43251/43254, is well supported by the shared file, shared issue target, and near-identical `fixed_cross_entropy` kwargs fix. The other soft-edge rejections appropriately treat the PRs as distinct models, docs, tests, or routing changes without overstating duplicate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 30, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-2", + "nodes": [ + "pull_request:40546", + "pull_request:42781" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781" + ], + "prepared_review_unit_hash": "a711540e673acf560b1e24eb3d5bb78aa73b75933ecede5013439458715c3964", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7226, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7226, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related but distinct VibeVoice pull requests: one adds the base VibeVoice implementation, while the other extends VibeVoice with realtime/acoustic-tokenizer work. They overlap in shared model scaffolding, but they do not look like the same concrete fix or mergeable duplicate PR.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #42781 is the broader, more up-to-date VibeVoice change set and includes the realtime-specific pieces plus the acoustic tokenizer/model work, making it the better representative artifact for this cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR #42781 is the best representative because it covers the larger VibeVoice surface area and the later, more specialized integration work; PR #40546 is the earlier base implementation and not a duplicate of it.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "Shared VibeVoice scaffolding alone is not enough: #40546 is a base implementation, while #42781 adds realtime and acoustic-tokenizer functionality plus extra generation/feature-extraction changes. They are related but not the same concrete code-path fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 3, + "component_count": 30, + "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": "603981ac9dcb8fe3da87008f5af56435759eef98c7219fd98d29727e1db8b453", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4674, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4674, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These pull requests are on unrelated features: one adds covariate-based forecasting for TimesFM, the other adds DINOv3 image-classification support. They do not appear to fix the same bug or implement the same change, so the soft duplicate edge should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different models, different code paths, and different user-facing features. The shared 'model support' theme is too broad to indicate the same underlying change, so they should not be merged or treated 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": 4, + "component_count": 30, + "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": 30, + "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": "d598fb53bff9bbea76fbec3957355952937483d29af5906aac7a618744047fed", + "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": "All three items are separate new-model integration PRs for different models (DeepSeek OCR, Isaac, Kimi2-6). They share boilerplate files and repo-wide mapping edits, but not the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: each pull request adds support for a different model and different codepaths.", + "best_issue_reason": null, + "best_pr_reason": "PR 45630 is the most substantial and review-active of the set, but it is still a distinct model addition rather than a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41797", + "right": "pull_request:45186", + "accept": false, + "reason": "Different model integrations (DeepSeek OCR vs Isaac); overlap is only in shared infrastructure and documentation/mapping scaffolding." + }, + { + "left": "pull_request:45186", + "right": "pull_request:45630", + "accept": false, + "reason": "Different model integrations (Isaac vs Kimi2-6); similar auto-registration changes do not make them the same bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 30, + "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": "5d69c09c47b8dab151e9100501ea1d944e6e39a446c9ac7f416e5c1fac4144af", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs overlap in GitHub Actions workflow maintenance, but they are not the same change set: one is an Actions runtime compatibility sweep, the other is a broader dependency/version bump with concrete docker build action upgrades. Do not dedupe them as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: 42975 is specifically about GitHub Actions Node 24 compatibility across many workflows, while 42976 is a separate upgrade pass that prominently changes docker/build-push-action and related workflow versions.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the better representative of the Node 24 compatibility effort because its title and diff align with a focused workflow-runtime upgrade across many files.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "They share workflow files and both upgrade actions, but the concrete changes differ: 42975 is a Node 24 compatibility sweep for GitHub Actions, whereas 42976 mainly bumps docker/build-push-action and other workflow versions. They do not look like the same underlying fix and should not 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": 7, + "component_count": 30, + "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": "d14ba206381b317bef975b23431b6831ff7d5d5dd77d837ad5d5e56fb253b2b1", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14179, + "estimated_input_tokens": 3545, + "estimated_eval_tokens": 7346 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14179, + "estimated_input_tokens": 3545, + "estimated_eval_tokens": 7346 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are separate model-addition efforts with only superficial overlap in docs and auto-registry files. They do not look like duplicates or mergeable variants of the same bug/fix/change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No issue in the cluster. Among the PRs, #43451 is the strongest representative because it is the most substantial, has the highest review activity, and contains a complete model integration across modeling, processing, conversion, tests, and auto-mappings.", + "best_issue_reason": null, + "best_pr_reason": "#43451 is the best overall representative of the cluster for triage purposes: it is a full-featured model addition with broad coverage and substantial review, whereas the others are different model additions with less basis for comparison.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "ViT NEPA and Molmo are unrelated model additions. Shared docs/auto files reflect framework registration, not the same underlying code-path or fix." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "ViT NEPA and Molmo2 are different model integrations with different architectures and codepaths; overlap is only in common registry/docs plumbing." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "ViT NEPA and CTSM are unrelated model additions in different domains, so they are not the same change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are distinct model families with separate implementations, tests, and processors; same-name similarity is not enough to treat them as duplicates." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM address different model families and codepaths; shared registry files are incidental." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 and CTSM are unrelated additions spanning different subsystems, 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": 30, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-3", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43665|pull_request:44815" + ], + "prepared_review_unit_hash": "cf83c761cd1008c4a478ac9a8fa4b042a2d801a84b965a7638d08c22e51dd648", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7093, + "estimated_input_tokens": 1774, + "estimated_eval_tokens": 3804 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7093, + "estimated_input_tokens": 1774, + "estimated_eval_tokens": 3804 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three items are pull requests, but they do not describe the same underlying change. PR 44815 is a real dequant/loading fix; PR 43291 is a Whisper tokenizer/test-related fix; PR 43665 only edits tests and appears unrelated to the others. No soft pair should be merged.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44815 is the strongest canonical PR because it contains the clearest substantive code fix across multiple implementation files, with a concrete change in model loading plus FP8/quantization logic. It is more representative of an actual bug fix than the test-only or narrow Whisper PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 44815 is the best overall PR to represent this cluster because it addresses a specific runtime/code-path issue and has broader technical impact than the other two. PR 43291 is narrower and centered on Whisper decode/tests, while PR 43665 appears to be only test modifications and is not a comparable code fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Both touch tests, but they do not fix the same bug or code path. 43291 is Whisper decode/tokenization related, while 43665 only alters CLIP/ViT tests with no shared implementation change." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "These PRs target different subsystems and different problems: Whisper decode behavior vs model loading/FP8 dequantization. They are not mergeable as one underlying fix." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "No shared failing code path or concrete change exists between them. 43665 is test-only and unrelated to 44815's loading/quantization 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": 9, + "component_count": 30, + "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-43488-5", + "nodes": [ + "pull_request:43488", + "pull_request:43532", + "pull_request:45932", + "pull_request:46100", + "pull_request:46114" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:46114", + "pull_request:43488|pull_request:43532", + "pull_request:43532|pull_request:46114", + "pull_request:45932|pull_request:46100", + "pull_request:46100|pull_request:46114" + ], + "prepared_review_unit_hash": "ad7615b67e2d3aa0063fa67dd44adb7fe8a82e6c968ad8a9f5011da31956f062", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 12262, + "estimated_input_tokens": 3066, + "estimated_eval_tokens": 6388 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 5, + "serialized_chars": 12262, + "estimated_input_tokens": 3066, + "estimated_eval_tokens": 6388 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are mostly unrelated synthetic/test PRs with weak filename overlap, mainly around ViT tests and CI scaffolding. The apparent similarities are title/style-driven rather than the same underlying bug or change, so there is no true duplicate PR pair here.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:45932 is the broadest representative of the synthetic test PRs, spanning multiple model and workflow files rather than a single tiny test edit.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:45932 is the best representative PR in the cluster because it has the widest change surface and looks like the central 'testing grafana' synthetic change set.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Both are dummy '[don't merge]' PRs, but one is a setup/version test and the other is a workflow/model instrumentation change; the shared ViT filename is not enough to make them the same concrete change." + }, + { + "left": "pull_request:43488", + "right": "pull_request:46114", + "accept": false, + "reason": "The overlap is only a shared ViT test file and test-harness context; the actual edits are different synthetic changes with different scopes and goals." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46114", + "accept": false, + "reason": "They both touch test infrastructure, but the modified files and patches are different enough that they do not look like one mergeable fix." + }, + { + "left": "pull_request:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "Both are test-related synthetic PRs, but 45932 is a broad grafana/testing change across workflows and model code while 46100 is just a ViT test migration; not the same underlying change." + }, + { + "left": "pull_request:46100", + "right": "pull_request:46114", + "accept": false, + "reason": "The shared ViT test file is incidental; 46100 is a tiny failing-test migration and 46114 is a broader forked PR with CI and multiple test files, so they are not the same code-path problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 10, + "component_count": 30, + "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": "66e432627a47cded58e4a015026a44e5335ab76983e52c4a1ff2e92623ad0d25", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs add different segmentation pipelines with similar scaffolding, but they target different features and code paths: promptable concept segmentation vs promptable visual segmentation.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true duplicate PR exists in this cluster. If a representative is needed, 43613 is slightly more central due to broader model coverage and more review activity, but it is still a different pipeline from 43612.", + "best_issue_reason": null, + "best_pr_reason": "43613 is the better representative for the cluster only because it is the broader/more active PR, not because it duplicates 43612; the underlying functionality differs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Do not merge: they implement different pipelines with different targets and model integrations. Shared docs/auto-registration/test scaffolding is generic and not enough to treat them as 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": 11, + "component_count": 30, + "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-43649-2", + "nodes": [ + "pull_request:43649", + "pull_request:46301" + ], + "soft_pairs": [ + "pull_request:43649|pull_request:46301" + ], + "prepared_review_unit_hash": "2387052f69f7b5368c6357b031e26e85c117eab5ad7cfb97c24c885186f912e7", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4549, + "estimated_input_tokens": 1138, + "estimated_eval_tokens": 2532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4549, + "estimated_input_tokens": 1138, + "estimated_eval_tokens": 2532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are pull requests, but they address different CI/test infrastructure changes and do not look like the same underlying bug or change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43649 is the better canonical PR because it has the more specific title and a clearer, broader CI/reporting context than the generic 'check' PR 46301.", + "best_issue_reason": null, + "best_pr_reason": "PR 43649 is the strongest representative of the cluster: it is more descriptive, has more evidence of a concrete workflow/reporting change, and is less ambiguous than PR 46301.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43649", + "right": "pull_request:46301", + "accept": false, + "reason": "Reject: the diffs target different code paths. 43649 modifies self-comment CI, scheduled workflow behavior, and test/reporting utilities for failure reporting, while 46301 changes CircleCI test selection and tensor-parallel mixin tests. They are not the same fix and would not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 12, + "component_count": 30, + "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": "8941adea8e8b83723543430feefe68bc0264fbd68878d6b3e7881575137e59e8", + "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": "Both PRs touch `src/transformers/trainer.py` and add extensibility hooks for custom trainers, but they address different code paths: gradient-accumulation loss scaling vs. signature-column filtering. They are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43663 is the cleaner representative of this cluster because it introduces a reusable helper for column-signature inspection, but it is not the same change as 43651.", + "best_issue_reason": null, + "best_pr_reason": "PR 43663 is the best single PR to represent the cluster\u2019s theme of Trainer extensibility, though it should not be merged with 43651 as a duplicate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Different Trainer behaviors: 43651 changes loss normalization during gradient accumulation, while 43663 refactors how allowed input columns are derived for PEFT/custom trainers. Same file and broad subsystem, but not the same underlying 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": 13, + "component_count": 30, + "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": "11519020dfaecfaeb4e96b8a4b731740edef7ff2b6ab3b5baf759a5e5ea05eec", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 3664, + "estimated_input_tokens": 916, + "estimated_eval_tokens": 2088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "PR 43842 is the most concrete standalone fix for the `TypeAdapter`/`pydantic`-absent failure in `src/transformers/cli/serve.py`. PR 43656 touches the same file but looks like a broader typing/__future__ cleanup and is not clearly the same bugfix, so the pair should not be merged as duplicates.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43842 is the clearest direct fix: it explicitly addresses the `TypeAdapter NameError when pydantic is not installed` by moving the import under `TYPE_CHECKING` and quoting the annotation.", + "best_issue_reason": null, + "best_pr_reason": "PR 43842 is the best standalone PR for this cluster because it targets the specific runtime/import failure with a minimal, tightly scoped change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Reject: 43656\u2019s diff is a broader typing/`__future__` cleanup in the same file, while 43842 specifically fixes `TypeAdapter` when `pydantic` is missing. Shared filename alone is not enough to treat them as the same bugfix or mergeable change." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The packet does not strongly support treating these as the same bugfix or as clearly mergeable. PR 43842 is explicitly about `TypeAdapter` when `pydantic` is not installed, while PR 43656\u2019s diff excerpt shows a broader `__future__`/typing cleanup in the same file but does not clearly establish the same fix. The soft edge verdict is too strong given only shared-file similarity and low Jaccard overlap." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 30, + "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": 15, + "component_count": 30, + "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:44018|pull_request:44066", + "pull_request:44066|pull_request:44068", + "pull_request:44019|pull_request:44722", + "pull_request:44086|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44018|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request: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:44722", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44071|pull_request:44086", + "pull_request:44017|pull_request:44071", + "pull_request:44073|pull_request:44101", + "pull_request:44059|pull_request:44071", + "pull_request:44018|pull_request:44059", + "pull_request:44073|pull_request:44161", + "pull_request:44056|pull_request:44101", + "pull_request:44059|pull_request:44068", + "pull_request:44101|pull_request:44161", + "pull_request:44017|pull_request:44018", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:43997|pull_request:44018", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request:44018|pull_request:44161", + "pull_request:43997|pull_request:44071", + "pull_request:44010|pull_request:44071", + "pull_request:44018|pull_request:44129", + "pull_request:44010|pull_request:44073", + "pull_request:43997|pull_request:44074", + "pull_request:44071|pull_request:44129", + "pull_request:44010|pull_request:44018", + "pull_request:43997|pull_request:44129", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "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": "c8723e15ab65b85c70c9892164b1bcdd280d41d69123536580d929ab1c4858f7", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 + }, + "budget": { + "node_count": 27, + "item_count": 27, + "soft_pair_count": 72, + "serialized_chars": 75096, + "estimated_input_tokens": 18774, + "estimated_eval_tokens": 37804 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is an umbrella of output-tracing refactors across many Transformer model families. The only true duplicate-style overlaps are within the GPT-Neo line and the GPT-J/CodeGen line; cross-family similarities are mostly theme-level, not the same concrete change. I corrected the GPT-J/CodeGen overlap to accept as a soft edge.", + "confidence": 0.74, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44722 is the best representative because it is the most complete GPT-J/CodeGen refactor in the cluster, explicitly targets the standardized output-tracing change, and overlaps the same GPT-J code path as the earlier GPT-J-only PRs.", + "best_issue_reason": null, + "best_pr_reason": "PR 44722 is the strongest single PR to keep as the cluster representative: it is the latest, broadest GPT-J/CodeGen variant and best captures the recurring standardized-tracing refactor pattern.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model families (CVT/FNet vs GPT-J); only the refactor theme matches." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "Different model implementations; shared issue target does not make this the same change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Same GPT-J output-tracing refactor line, with overlapping GPT-J code and the same standardized decorator migration." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "RegNet/ResNet versus EfficientNet are separate model refactors, not one duplicate change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet and GPT-J/CodeGen are unrelated concrete code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen versus EfficientNet are different model-specific refactors." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are distinct implementations with unrelated output code." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen and MGP-STR are different model code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen and MPT are separate model refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo output-tracing refactor on the same model/file family." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are different models." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen versus GPT-Neo are different concrete model paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet and GPT-J/CodeGen are unrelated model implementations." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are different model-specific refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet versus GPT-J are different code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are separate model changes." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are unrelated implementations." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are different model refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are different code paths." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are separate model implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT and GPT-J/CodeGen are different implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo and GPT-J are different models." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are different model-specific changes." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are different model families." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-2 and MGP-STR are unrelated implementations." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are different model code paths." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert are separate model implementations; only the style of refactor overlaps." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44073", + "accept": false, + "reason": "RegNet and VisualBert are different concrete changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44101", + "accept": false, + "reason": "MPNet and Flaubert are unrelated model refactors." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert are different model code paths." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet and LongT5 are separate model implementations." + }, + { + "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:44161", + "accept": false, + "reason": "Segformer and LongT5 are different models." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "Segformer and VisualBert are separate model refactors." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-2 and MPT are different model code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are unrelated concrete changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are different implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are separate model refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are different model code paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBert are unrelated model changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44018", + "accept": false, + "reason": "SqueezeBert and GPT-Neo are different models." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are distinct implementations." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen standardized output-tracing refactor line, with overlapping GPT-J/CodeGen files and the same issue target." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J output-tracing refactor; 44722 extends the same code-path change with CodeGen." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: one soft-edge verdict is not grounded in the packet. The pair pull_request:44066 vs pull_request:44722 was marked false with a reason about GPT-Neo, but both PRs are in the GPT-J/CodeGen output-tracing refactor line, share the same issue target (#43979), and touch overlapping GPT-J/CodeGen files. That makes the rejection too strong for a soft edge. The rest is mostly conservative, but this misclassification is enough to reject." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 30, + "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": "948d2037b7caef8eb0ca2dccd93f6a19cea1602f469db697609817963047a62b", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 6, + "serialized_chars": 17125, + "estimated_input_tokens": 4282, + "estimated_eval_tokens": 8820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "A set of model-specific PRs refactoring output tracing to the new capture_outputs/can_return_tuple pattern. They share a common migration theme, but each targets a different architecture and code path, so they are not duplicates of one another.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "44129 is the strongest representative PR: it has the clearest title, an explicit issue link, and a substantial concrete refactor in a single model implementation. That said, it is still model-specific rather than a cluster-wide duplicate fix.", + "best_issue_reason": null, + "best_pr_reason": "44129 is the best single PR candidate because it is well-scoped, explicitly tracked to an issue, and shows a concrete standardized output-capture change rather than just a naming cleanup. It is not merge-equivalent to the others, but it is the most complete exemplar of the pattern.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "Both are output-tracing refactors, but they touch different models (FocalNet vs ImageGPT) and different internal hooks. Same migration pattern, not the same underlying bug or change." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "VisualBert and TextNet are unrelated model implementations. These are separate standardized-output refactors, not one concrete code-path fix." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert each update their own output-capture plumbing. They share only the refactor theme, not a mergeable shared defect." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "Different models and different forward/output handling. This is a broad migration pattern, not a duplicate PR pair." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are separate model-specific implementations. The similarity is only the standardized output tracing migration." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 modify different code paths and model classes. They are not the same concrete fix and would not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 16, + "component_count": 30, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-13", + "nodes": [ + "pull_request:43998", + "pull_request:43999", + "pull_request:44000", + "pull_request:44001", + "pull_request:44002", + "pull_request:44003", + "pull_request:44004", + "pull_request:44025", + "pull_request:44026", + "pull_request:44027", + "pull_request:44028", + "pull_request:44029", + "pull_request:44030" + ], + "soft_pairs": [ + "pull_request:43998|pull_request:44002", + "pull_request:43998|pull_request:44000", + "pull_request:43998|pull_request:44001", + "pull_request:43998|pull_request:44003", + "pull_request:43998|pull_request:44029", + "pull_request:43999|pull_request:44001", + "pull_request:43999|pull_request:44002", + "pull_request:43999|pull_request:44025", + "pull_request:43999|pull_request:44026", + "pull_request:43999|pull_request:44027", + "pull_request:44000|pull_request:44001", + "pull_request:44000|pull_request:44002", + "pull_request:44000|pull_request:44003", + "pull_request:44000|pull_request:44025", + "pull_request:44001|pull_request:44002", + "pull_request:44001|pull_request:44003", + "pull_request:44001|pull_request:44025", + "pull_request:44002|pull_request:44029", + "pull_request:44002|pull_request:44003", + "pull_request:44003|pull_request: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": "bd357bf09aa2aca7f7f87c0149a8b9c071e312e6c850bb71168573345a30a03f", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 39, + "serialized_chars": 39778, + "estimated_input_tokens": 9945, + "estimated_eval_tokens": 20146 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 39, + "serialized_chars": 39778, + "estimated_input_tokens": 9945, + "estimated_eval_tokens": 20146 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a theme cluster of unrelated model-specific output-tracing refactors, not a duplicate cluster. The PRs share similar mechanics and titles, but they touch different model implementations and code paths, so they should not be deduped together.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44003 is the broadest, most code-heavy refactor in the set and the clearest representative of the shared output-tracing theme.", + "best_issue_reason": null, + "best_pr_reason": "PR 44003 is the best single representative to inspect first because it shows the pattern clearly and spans two closely related Mamba implementations.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Same refactor theme, but timm_backbone and upernet are different model code paths with no evidence of a shared underlying bug." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Both are output-tracing refactors, but they target different architectures and forward paths; not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different models and outputs; this is a parallel cleanup, not a duplicate fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "timm_backbone and mamba/falcon_mamba are unrelated implementations; similar decorator/import changes are not enough for duplication." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and different forward implementations; no shared concrete bug or change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "MobileNetV1 and UnivNet are unrelated model paths; similar output-tracing cleanup does not make them duplicates." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different architectures and files, with no indication of the same bug or mergeable patch." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "These are separate model-specific refactors; they share a pattern but not an underlying code-path issue." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "MobileNetV1 and vision_encoder_decoder are unrelated forward paths; too broad to treat as the same fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different encoder-decoder wrapper and different modality; not the same concrete change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Vision-text dual encoder and UnivNet are separate model implementations; only the refactor style matches." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families and code paths; no shared defect or patch." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "CLIP-like dual encoder vs Mamba models are unrelated; similarity is only at the refactor pattern level." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Different models and different forward semantics; these are not duplicate fixes." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "UnivNet and UperNet are unrelated; this is only a similar output-tracing cleanup." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different code paths and architectures; no evidence they fix the same underlying issue." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "These are distinct model-specific refactors, not a single mergable PR." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "UperNet and RWKV are separate models with different output handling; not the same change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Different subsystems and implementations; similar output-tracing edits are insufficient for duplication." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Mamba/Falcon-Mamba and CodeGen are different model families; the output-refactor implementations are not the same concrete fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "timm_backbone and CodeGen are unrelated code paths; no duplicate-level evidence." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "UnivNet and CodeGen do not share the same underlying bug or change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "UperNet and CodeGen are unrelated beyond the generic refactor pattern." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model families and different forward logic; not a shared concrete patch." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "DepthAnything and vision_encoder_decoder are separate model paths; same maintenance theme, not the same fix." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different models and different forward implementations; not duplicates." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "DepthAnything and SuperPoint are unrelated code paths, so they should stay separate." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model families and output semantics; no evidence of one underlying bug." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "DepthAnything and DPR are unrelated models; similar API cleanup is not enough." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Vision-encoder-decoder and RWKV are unrelated implementations with different output-tracing changes." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "These are both encoder-decoder wrappers, but they target different modalities and files; too broad to merge as one duplicate fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families and code paths; no shared concrete bug." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Vision-encoder-decoder and DPR are unrelated; only the refactor motif is shared." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Speech-encoder-decoder and SuperPoint are unrelated model paths; not the same change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures and forward code; no duplicate-level overlap." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Speech-encoder-decoder and DPR are different model families; this is only stylistic similarity." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "SuperPoint and RWKV are unrelated implementations; not the same underlying fix." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "SuperPoint and DPR are different models and code paths; keep separate." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "RWKV and DPR have different forward logic and output handling; no shared 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": 17, + "component_count": 30, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44662-3", + "nodes": [ + "pull_request:44662", + "pull_request:45075", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401", + "pull_request:44662|pull_request:45075", + "pull_request:45075|pull_request:45401" + ], + "prepared_review_unit_hash": "fad38a6253374a85944eccf3881fa9ae6d5bf513ed2bf1aa1fae305f104b0000", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10913, + "estimated_input_tokens": 2729, + "estimated_eval_tokens": 5714 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10913, + "estimated_input_tokens": 2729, + "estimated_eval_tokens": 5714 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three separate model-addition PRs for unrelated models (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). They share some Auto/docs scaffolding, but the concrete model code, filenames, and target capabilities are different enough that they should not be merged as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: each PR implements a different model integration with its own configuration, modeling, processing, tests, and docs.", + "best_issue_reason": null, + "best_pr_reason": "If a representative is needed, PR 45401 is the most explicitly scoped (linked to issue 45027) and self-contained, but it is still not a duplicate of the other two PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Both add new model integrations, but for different models and codepaths (PenguinVL vs Voxtral TTS). Shared Auto/doc files are generic scaffold, not the same underlying change." + }, + { + "left": "pull_request:44662", + "right": "pull_request:45075", + "accept": false, + "reason": "PenguinVL and DeepSeek-OCR-2 are distinct model additions with different model families, processors, and tests; overlap is only in common registration/docs infrastructure." + }, + { + "left": "pull_request:45075", + "right": "pull_request:45401", + "accept": false, + "reason": "DeepSeek-OCR-2 and Voxtral TTS are unrelated feature additions for different models and tasks; they do not appear to fix or implement 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": 18, + "component_count": 30, + "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": "6b4248c86a5fd7616cfab8c71c6aadc0bc7198b6cca89dc9647b3ef3cd8ccde0", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4408, + "estimated_input_tokens": 1102, + "estimated_eval_tokens": 2460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are related by the general idea of fusion, but they target different concrete changes: one adds a new module-fusion API, the other wires up n-to-1 kernel fusion through `KernelConfig`. They are not duplicates.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44979 is the more foundational change: it introduces the standalone `Module Fusion API` and the new fusion abstraction. PR 45363 is a separate, more specific integration built around kernel configs and fused kernels.", + "best_issue_reason": null, + "best_pr_reason": "PR 44979 is the better cluster anchor because it defines the core fusion mechanism; PR 45363 depends on a different code path and is better treated as a follow-on feature rather than the same change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Both mention fusion, but 44979 adds a general module-fusion API while 45363 implements n-to-1 kernel fusion via `KernelConfig` and integration hooks. Different underlying change, not one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 30, + "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": "a5bf13985e99c331bf45424b997bd4d6ccb24739f70993a61422e3a84aa390c6", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4413, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs patch the same Mistral regex pre-tokenizer AttributeError in the same code path; 45317 is the fuller version with added test coverage, so it is the better canonical PR.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "45317 makes the same fix as 45086 in the same function, but also adds a test file, so it is the more complete representative of this change.", + "best_issue_reason": null, + "best_pr_reason": "45317 is the best standalone PR because it addresses the same concrete bug and includes regression tests, making it more merge-ready than the minimal patch in 45086.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying AttributeError fix in _patch_mistral_regex, same file, and the second PR is a superset with tests; they could plausibly be merged as one change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 20, + "component_count": 30, + "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": "893db24b5af07f76928b9e914b3273d611023953fdb9df1e35a027d7ed9e9604", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two NemotronH PRs in the same area, but they implement different changes: one refactors conversion/mapping and inheritance, while the other splits NemotronH into dense/sparse model families with docs and tests. Not a duplicate pair.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45569 is the better representative of the cluster because it is the more complete user-facing implementation change: it adds the dense/sparse model split plus docs and tests, whereas #45115 is a narrower refactor/conversion-mapping change.", + "best_issue_reason": null, + "best_pr_reason": "PR #45569 has the broader and more finished scope, touching the actual model packaging and documentation rather than only conversion/mapping internals.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Related NemotronH work, but the concrete changes differ: #45115 focuses on inheritance/conversion mapping and modular refactoring, while #45569 introduces separate NemotronHDense/NemotronHSparse implementations, docs, and tests. They do not look like the same underlying fix and would not plausibly merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 30, + "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": "848f5af2402577187bda8ec8b2b4751f4334197f21372db0e9ae003da3b090d8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5397, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5397, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two related but distinct PRs: one introduces heterogeneous per-layer config plumbing, the other adds broader heterogeneous modeling/runtime support on top of it. They share files and theme, but they are not the same concrete change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45332 is the better canonical PR because it is the broader, more complete change: it adds heterogeneous model support across cache, masking, modeling, and tests, and appears to build on the config plumbing introduced separately.", + "best_issue_reason": null, + "best_pr_reason": "PR #45332 is the strongest representative of the cluster because it addresses the full heterogeneous-model feature rather than only the configuration layer, making it the more comprehensive standalone change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related feature work, but not the same underlying change. #45333 is focused on per-layer heterogeneous config support, while #45332 extends that into cache/modeling/runtime behavior. They could be separate PRs and are not strict 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": 30, + "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": "172a4ddc0585ee2628b454c365992ec01ca8fbcd56f08f14fd6f932440bb1cfa", + "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 pull requests are typing-related, but they address different changes: PR 45415 is a broad repository-wide type-checking sweep with many ignores, while PR 45425 adds concrete typing protocol/helpers in `_typing.py` and adjusts a few consumers. They should not be merged as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45425 is the more specific change, centered on shared typing abstractions and their consumers rather than a broad type-checking pass.", + "best_issue_reason": null, + "best_pr_reason": "PR 45425 is the best representative of the cluster because it makes the more concrete, code-level typing change; PR 45415 is mostly a wide lint/type-check configuration sweep.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "They are related only at a broad typing/annotation level. 45415 is a large type-checking cleanup across many files, while 45425 introduces new typing protocols and utilities for specific consumers; they do not fix the same concrete problem and would not plausibly be 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": 23, + "component_count": 30, + "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": "a8a453f507e7c867a2ba885d7d0a4a90742fac93115e67bacdb8a93cfffad20d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4857, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4857, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated pull requests fixing cached-forward behavior in different model families: Qwen3.5/GDN linear attention versus Mamba2 use_cache handling. They do not appear to be the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46084 is the better canonical PR because it is tightly scoped to a single concrete bug, has an explicit linked issue (#46032), and the diff targets one model path with a clear seq_len>1 cache fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #46084 is the best representative of its own bug fix: it is narrowly focused, issue-linked, and more self-contained than the broader multi-model PR #45513.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45513", + "right": "pull_request:46084", + "accept": false, + "reason": "Different model architectures and different code paths: Qwen3.5 GDN linear attention cached continuation versus Mamba2 cache handling for seq_len>1. They are not plausibly mergeable into one PR for the same underlying bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 30, + "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": "79d63d464c9b2fb505d74bf845fdc450a68adbbbaa839874a2d645afd0beeb6a", + "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 pull requests are about expert parallelism in Transformers loading, but they fix different failure modes and should not be merged as duplicates: one is about preserving distributed_config/EP detection for ZeRO-3 loading via accelerate, the other is about preventing expert weights from being overwritten during FSDP2/rank-0 broadcast.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45662 is the better canonical PR for the cluster because it addresses a concrete correctness bug in expert loading and spans the main affected code paths, whereas PR 45548 is narrower and more infrastructural.", + "best_issue_reason": null, + "best_pr_reason": "PR 45662 is the strongest standalone representative of the topic: it fixes a specific silent data corruption issue in EP loading, with clear code-path changes across model loading and MoE handling.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "They share the general EP/distributed-loading theme, but the bugs are different: 45548 adds distributed_config serialization and an EP detection property for ZeRO-3/accelerate loading, while 45662 fixes expert parameters being overwritten after sharding/broadcast in FSDP2. Different root causes, different fixes." + } + ] + }, + "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": 30, + "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": "fd57e2126a655340e20c4b711d24fbaaa96bbd705b627625f2af9f4d889810e9", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4590, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4590, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs target the same xdist-safe captured_info logging problem in testing_utils, but 45645 is the broader follow-up that also updates CI log collection to handle the new per-worker files. It looks like 45639 is a narrower subset of 45645.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the core fix in src/transformers/testing_utils.py plus the CI-side adjustment needed to consume multiple captured_info files, making it the more complete representative of the change.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the best standalone representative since it addresses the same underlying xdist collision bug and includes the additional workflow handling required for end-to-end correctness. PR 45639 appears to be a narrower partial fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Both PRs fix the same concrete captured_info/xdist collision path in testing_utils. 45645 is a superset that also updates CI log collection, so they are close enough to be treated as the same underlying change cluster." + } + ] + }, + "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": 30, + "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-45720-3", + "nodes": [ + "pull_request:45720", + "pull_request:46023", + "pull_request:46259" + ], + "soft_pairs": [ + "pull_request:45720|pull_request:46259", + "pull_request:46023|pull_request:46259" + ], + "prepared_review_unit_hash": "65b6e816b85758b2d7298adf478b173553c8fd9421ac59306b8303e6d4c892ac", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10911, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10911, + "estimated_input_tokens": 2728, + "estimated_eval_tokens": 5712 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three items are pull requests in the linter/modeling-rules area, but they address different rule implementations/sweeps rather than the same underlying fix. The soft similarity is mostly due to shared docs and some overlapping model files, not a mergeable duplicate change.", + "confidence": 0.83, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46259 is the strongest representative of the cluster: it is the broadest modeling-rule change, with a documented rule addition and a substantial sweep across model files. The other PRs are narrower, rule-specific follow-ups.", + "best_issue_reason": null, + "best_pr_reason": "PR #46259 is the best single PR to represent this cluster because it has the broadest scope and clearest rule-definition intent. PR #46023 is a narrower fixes-only pass, and PR #45720 targets a different linter rule entirely.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45720", + "right": "pull_request:46259", + "accept": false, + "reason": "Different concrete changes: #45720 implements part of TRF003 for model forward tuple/return_dict behavior, while #46259 is about a different modeling-rule sweep for image/video processor boolean flags. Shared linter/doc context is not enough to treat them as the same fix." + }, + { + "left": "pull_request:46023", + "right": "pull_request:46259", + "accept": false, + "reason": "Both touch modeling-rule docs and model files, but they target different rule work: #46023 is fixes for TRF016/rule 16, while #46259 is a separate rule addition/sweep. The overlap is broad subsystem similarity, not the same concrete bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 27, + "component_count": 30, + "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": 30, + "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": "df96d2b9cf84e7bd31e595eedf1ed80a04757e509a50aba395ec51576cf755e9", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8082, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8082, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three open PRs all add SDPA dispatch support, but they target different model families and code paths (CTRL, T5, and SeamlessM4T/v2 + wav2vec2 variants). They are related in theme, not duplicates of the same underlying change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46196 is the broadest and most representative of the cluster, covering multiple related speech/seq2seq model implementations rather than a single model.", + "best_issue_reason": null, + "best_pr_reason": "PR 46196 is the strongest cluster representative because it spans several models and includes the most comprehensive SDPA dispatch work.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46073", + "right": "pull_request:46154", + "accept": false, + "reason": "Both are SDPA dispatch additions, but they modify different model implementations (CTRL vs T5) and are separate code-path changes, not one mergeable fix." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46196", + "accept": false, + "reason": "These target different families and different attention implementations; the overlap is only the broad SDPA theme, which is too general to treat as duplicates." + }, + { + "left": "pull_request:46154", + "right": "pull_request:46196", + "accept": false, + "reason": "T5 SDPA support is a distinct model-specific change from the multi-model SeamlessM4T/wav2vec2 dispatch work, so they should not be merged as the same PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 29, + "component_count": 30, + "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": "06758fc1430bf820ffd0d0761d9376c5265672b5e67b2b0448fdb48e637aef04", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5570, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5570, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch overlapping distributed/TP files, but they are different scopes: one is a focused EP + trainer integration, the other is a broad distributed branch-base umbrella touching many unrelated areas. They are not the same underlying change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46126 is the more focused and concrete change, centered on EP/trainer integration and the specific distributed code paths it affects. PR #46269 is much broader and looks like a branch-base umbrella rather than a single duplicate fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #46126 is the best representative artifact for this cluster because it has a clear, narrow purpose and the overlap with the other PR is confined to shared distributed files rather than the same full change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Shared distributed filenames are not enough here: #46126 is a targeted EP/trainer integration PR, while #46269 is a broad distributed branch-base PR spanning many extra modules and CI changes. They do not appear to fix the same concrete code-path problem and would not plausibly merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 30, + "component_count": 30, + "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": "a2aa2a995cb0b786e400108dc47953484e1320b4979426436316e9660187e833", + "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": "Both PRs address replacing file-based source inspection with a safer source-access mechanism, but they touch different code paths and are not the same mergeable fix.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46207 is the stronger representative because it targets a concrete source-reading helper in modeling_utils and includes tests, making the change more self-contained and clearly validated.", + "best_issue_reason": null, + "best_pr_reason": "PR 46207 is the best standalone PR candidate since it fixes a specific implementation path with test coverage; PR 46152 is related in theme but addresses a different helper in import_utils.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46152", + "right": "pull_request:46207", + "accept": false, + "reason": "Shared theme only: both avoid open()-based source reading, but one is in import_utils and the other in modeling_utils, with different functions and no clear same underlying bug or single mergeable code-path fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/manifest.json b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..baf290d5e246f0f19a957b1b6d468d28b88b0128 --- /dev/null +++ b/snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260531t120035z", + "artifacts": { + "hybrid": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-05-31T12:08:30Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260531T120035Z", + "variant": "hybrid" +} diff --git a/snapshots/20260531T120035Z/manifest.json b/snapshots/20260531T120035Z/manifest.json index 3fe50964267bb5f89b6e10ffb9a349aa12f4edfc..88a2da31bfb34c8e895ddc01886674527d333d6d 100644 --- a/snapshots/20260531T120035Z/manifest.json +++ b/snapshots/20260531T120035Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 201 }, "extracted_at": "2026-05-31T12:00:35Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260531t120035z", + "runs": { + "hybrid-model-20260531t120035z": { + "analysis_id": "hybrid-model-20260531t120035z", + "artifacts": { + "hybrid": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260531T120035Z/analysis-runs/hybrid-model-20260531t120035z/manifest.json", + "model": null, + "published_at": "2026-05-31T12:08:30Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260531T120035Z", "watermark": {