diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 712c0e01e808b80d370d346fb28ac092b2a05d25..2c931c276370566b14a30ab66c73e22c68c078ab 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": "20260427T000017Z", - "generated_at": "2026-04-27T00:01:33Z", + "snapshot_id": "20260427T120026Z", + "generated_at": "2026-04-27T12:09:02Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1053,7 +1053,7 @@ "cluster_id": "cluster-41211-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #41211.", "status": "open", - "confidence": 0.75, + "confidence": 0.55, "canonical_issue_number": 41211, "canonical_pr_number": 44339, "issue_numbers": [ @@ -1064,8 +1064,7 @@ 44339 ], "evidence_types": [ - "closing_reference", - "shared_issue_target" + "closing_reference" ], "pr_comparisons": [ { @@ -1150,74 +1149,152 @@ ] }, { - "cluster_id": "cluster-43240-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #43240.", + "cluster_id": "cluster-43656-4", + "summary": "Cluster of 1 issues and 3 PRs centered on issue #43824.", "status": "open", - "confidence": 0.75, - "canonical_issue_number": 43240, - "canonical_pr_number": 43251, + "confidence": 0.8, + "canonical_issue_number": 43824, + "canonical_pr_number": 43656, "issue_numbers": [ - 43240 + 43824 ], "pr_numbers": [ - 43251, - 43254 + 43656, + 43836, + 43842 ], "evidence_types": [ "closing_reference", - "shared_issue_target" + "shared_issue_target", + "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 43251, - "right_pr_number": 43254, - "code_similarity": 0.794, - "size_similarity": 0.64, + "left_pr_number": 43656, + "right_pr_number": 43836, + "code_similarity": 0.36, + "size_similarity": 0.176, "file_overlap": 1.0, - "area_overlap": 0.667, - "patch_similarity": 0.882, + "area_overlap": 0.051, + "patch_similarity": 0.048, "shared_filenames": [ - "src/transformers/loss/loss_utils.py" + "src/transformers/cli/serve.py" ], "shared_file_areas": [ { - "filename": "src/transformers/loss/loss_utils.py", + "filename": "src/transformers/cli/serve.py", "left_ranges": [ + [ + 11, + 18 + ], [ 30, - 50 + 36 + ], + [ + 315, + 323 + ], + [ + 665, + 671 + ], + [ + 931, + 937 + ], + [ + 1843, + 1849 + ], + [ + 1868, + 1874 ] ], "right_ranges": [ + [ + 11, + 18 + ], + [ + 359, + 429 + ], + [ + 584, + 590 + ], + [ + 1892, + 1910 + ], + [ + 1917, + 1923 + ] + ] + } + ] + }, + { + "left_pr_number": 43656, + "right_pr_number": 43842, + "code_similarity": 0.405, + "size_similarity": 0.5, + "file_overlap": 1.0, + "area_overlap": 0.0, + "patch_similarity": 0.036, + "shared_filenames": [ + "src/transformers/cli/serve.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/serve.py", + "left_ranges": [ + [ + 11, + 18 + ], [ 30, - 43 + 36 + ], + [ + 315, + 323 + ], + [ + 665, + 671 + ], + [ + 931, + 937 + ], + [ + 1843, + 1849 + ], + [ + 1868, + 1874 + ] + ], + "right_ranges": [ + [ + 54, + 61 + ], + [ + 587, + 593 ] ] } ] - } - ] - }, - { - "cluster_id": "cluster-43824-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #43824.", - "status": "open", - "confidence": 0.75, - "canonical_issue_number": 43824, - "canonical_pr_number": 43836, - "issue_numbers": [ - 43824 - ], - "pr_numbers": [ - 43836, - 43842 - ], - "evidence_types": [ - "closing_reference", - "shared_issue_target" - ], - "pr_comparisons": [ + }, { "left_pr_number": 43836, "right_pr_number": 43842, @@ -1270,57 +1347,48 @@ ] }, { - "cluster_id": "cluster-45081-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", + "cluster_id": "cluster-43240-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43240.", "status": "open", - "confidence": 0.8, - "canonical_issue_number": 45081, - "canonical_pr_number": 45317, + "confidence": 0.75, + "canonical_issue_number": 43240, + "canonical_pr_number": 43251, "issue_numbers": [ - 45081 + 43240 ], "pr_numbers": [ - 45086, - 45317 + 43251, + 43254 ], "evidence_types": [ "closing_reference", - "shared_issue_target", - "soft_similarity" + "shared_issue_target" ], "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, + "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/tokenization_utils_tokenizers.py" + "src/transformers/loss/loss_utils.py" ], "shared_file_areas": [ { - "filename": "src/transformers/tokenization_utils_tokenizers.py", + "filename": "src/transformers/loss/loss_utils.py", "left_ranges": [ [ - 1360, - 1370 - ], - [ - 1374, - 1380 + 30, + 50 ] ], "right_ranges": [ [ - 1360, - 1370 - ], - [ - 1374, - 1380 + 30, + 43 ] ] } @@ -1329,71 +1397,57 @@ ] }, { - "cluster_id": "cluster-45520-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #45520.", + "cluster_id": "cluster-45081-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", "status": "open", - "confidence": 0.55, - "canonical_issue_number": 45520, - "canonical_pr_number": 45524, + "confidence": 0.8, + "canonical_issue_number": 45081, + "canonical_pr_number": 45317, "issue_numbers": [ - 45520 + 45081 ], "pr_numbers": [ - 45524, - 45650 + 45086, + 45317 ], "evidence_types": [ - "closing_reference" + "closing_reference", + "shared_issue_target", + "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 45524, - "right_pr_number": 45650, - "code_similarity": 0.901, - "size_similarity": 0.9, - "file_overlap": 1.0, - "area_overlap": 0.903, - "patch_similarity": 0.7, + "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/utils/import_utils.py" + "src/transformers/tokenization_utils_tokenizers.py" ], "shared_file_areas": [ { - "filename": "src/transformers/utils/import_utils.py", + "filename": "src/transformers/tokenization_utils_tokenizers.py", "left_ranges": [ [ - 948, - 954 - ], - [ - 967, - 973 - ], - [ - 979, - 985 + 1360, + 1370 ], [ - 990, - 996 + 1374, + 1380 ] ], "right_ranges": [ [ - 948, - 954 - ], - [ - 964, - 973 - ], - [ - 979, - 985 + 1360, + 1370 ], [ - 990, - 996 + 1374, + 1380 ] ] } @@ -1819,7 +1873,7 @@ 41356 ], "target_issue_number": 41211, - "reason": "PRs in cluster-41211-3 are treated as duplicates because they converge on issue #41211 with closing_reference, shared_issue_target evidence." + "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", @@ -1830,6 +1884,16 @@ "target_issue_number": 43240, "reason": "PRs in cluster-43240-3 are treated as duplicates because they converge on issue #43240 with closing_reference, shared_issue_target evidence." }, + { + "cluster_id": "cluster-43656-4", + "canonical_pr_number": 43656, + "duplicate_pr_numbers": [ + 43836, + 43842 + ], + "target_issue_number": 43824, + "reason": "PRs in cluster-43656-4 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target, soft_similarity evidence." + }, { "cluster_id": "cluster-43698-3", "canonical_pr_number": 43779, @@ -1839,15 +1903,6 @@ "target_issue_number": 43698, "reason": "PRs in cluster-43698-3 are treated as duplicates because they converge on issue #43698 with closing_reference, shared_issue_target evidence." }, - { - "cluster_id": "cluster-43824-3", - "canonical_pr_number": 43836, - "duplicate_pr_numbers": [ - 43842 - ], - "target_issue_number": 43824, - "reason": "PRs in cluster-43824-3 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target evidence." - }, { "cluster_id": "cluster-43979-11", "canonical_pr_number": 44007, @@ -1883,15 +1938,6 @@ "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-45520-3", - "canonical_pr_number": 45524, - "duplicate_pr_numbers": [ - 45650 - ], - "target_issue_number": 45520, - "reason": "PRs in cluster-45520-3 are treated as duplicates because they converge on issue #45520 with closing_reference evidence." - }, { "cluster_id": "cluster-45561-3", "canonical_pr_number": 45639, diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 242d9188c892281e347c8ef99781cb302ee6006a..4ee9598350f0caec03ca2629999ea587b8125993 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": "20260427T000017Z", - "generated_at": "2026-04-27T00:01:33Z", + "snapshot_id": "20260427T120026Z", + "generated_at": "2026-04-27T12:09:02Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -13,8 +13,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-13244-3", "nodes": [ @@ -26,7 +26,7 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "d9c5ad76e827e0e24d9f5e3ea3a8f9db1c96b0218027c794d5595dad72c09492", + "prepared_review_unit_hash": "9f08dd43c33773c61325575f36c0aa14b48447224fc818d2a894eb9e6d42ec60", "component_budget": { "node_count": 3, "item_count": 3, @@ -48,24 +48,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues do not look like duplicates of one another: one is about Tapas tokenization mismatch, one is about an auth/error-message problem when downloading a private model, and one is about a generic model-load failure for bert-base-uncased. The only commonality is superficial Hugging Face model-loading language.", - "confidence": 0.94, - "canonical_issue_reason": "Issue #16618 is the broadest and most central-sounding of the three, so it is the best single anchor if one must be chosen, but it is not a true duplicate of the others.", + "summary": "These three issues do not appear to be duplicates. #16618 is the broadest model-loading failure report, while #15980 is specifically about the error message shown when downloading a private model without authentication, and #13244 is about Tapas tokenization diverging from TensorFlow behavior.", + "confidence": 0.81, + "canonical_issue_reason": "#16618 is the best canonical issue because it is the most general user-facing failure report in this set and centers on model loading, which is the closest match to a core bug rather than a narrow UX or preprocessing discrepancy.", "canonical_pr_reason": null, - "best_issue_reason": "#16618 has the most general title and highest discussion activity, making it the most representative issue in this small cluster, even though the items appear unrelated.", + "best_issue_reason": "#16618 is the best overall issue candidate because it describes a concrete model-load failure and is broader than the other two, which are clearly narrower in scope.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tapas tokenization behavior is a different code path and bug class from a bert-base-uncased model-loading failure; no concrete overlap beyond general ML/model terminology." + "reason": "Tapas tokenization mismatch and failed loading of bert-base-uncased are different problems affecting different code paths." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "#15980 is specifically about a misleading error when downloading a private model while logged out, whereas #16618 is a separate model load failure for a public checkpoint; these are not the same underlying bug." + "reason": "Bad auth-related download messaging for private models is not the same underlying bug as an inability to load bert-base-uncased." } ] }, @@ -84,10 +84,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:24643", "issue:30277", @@ -203,391 +203,379 @@ "issue:41093|issue:41762", "issue:43296|issue:43366" ], - "prepared_review_unit_hash": "0dd0afefb097d22aa1742cd2e46b878bf3bbfefd0d0be02337bcc232d5112344", + "prepared_review_unit_hash": "1209ca2d249a0324b4088139e576eaa76d140574aed3ef7dbf500677feb27d16", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 63, - "serialized_chars": 46476, - "estimated_input_tokens": 11619, - "estimated_eval_tokens": 23494 + "serialized_chars": 46483, + "estimated_input_tokens": 11621, + "estimated_eval_tokens": 23498 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a grab-bag of unrelated model/runtime/doc issues. The strongest duplicate candidate is the DeepSpeed ZeRO-3 / Bert index-0-out-of-bounds pair (43596 vs 43638), but nearby ZeRO-3 index-error reports like 41762 keep the cluster from being perfectly certain from metadata alone.", - "confidence": 0.84, - "canonical_issue_reason": "issue:43596 is the best canonical issue because it cleanly names the shared DeepSpeed ZeRO-3 index-out-of-bounds failure without overfitting to one specific Bert variant, while still matching the closest duplicate candidate issue:43638.", + "summary": "The cluster is mostly a mix of unrelated model/runtime failures, but there is a small coherent subcluster around zero-3/indexing errors and a docs-cleanup subcluster around v5 pipeline removals.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44805 is the broadest representative of the repeated mask/index-shape failure family and best matches the closest duplicate-like reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43596 is the best global issue representative for this cluster because it is the clearest, most general description of the shared ZeRO-3 zero-sized-index crash among the provided issues.", + "best_issue_reason": "Issue 44805 is the most central and generic issue here, with the clearest overlap to other index/shape-mismatch reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different bugs: a missing config attribute vs a streaming chat/completions failure." + "reason": "Different Qwen3.5 failures: missing config field vs streaming chat API failure." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different models and symptoms: image-preprocessor dtype mismatch vs AMD GPU execution failure." + "reason": "Different models and different bugs; dtype mismatch in preprocessing is unrelated to AMD GPU execution failure." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Both are dtype/runtime problems, but on different models and code paths." + "reason": "Both are runtime dtype-related, but they affect different models and code paths." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Different model families and failures; only a broad subsystem overlap." + "reason": "Different model families and different failures; same broad MoE area is not enough." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Distinct model/runtime errors; no evidence they share the same underlying defect." + "reason": "Different inference stacks and error modes; not the same concrete bug." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch is unrelated to an ONNX export error." + "reason": "DeepSpeed collective mismatch is unrelated to ONNX export errors." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "Both are model-specific errors, but one is export-related and the other is FP8 inference." + "reason": "Export-to-ONNX failure and FP8 config failure are different problems." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "Different failure modes: ONNX export vs device-mapping CUDA assert." + "reason": "Different models and runtime environments; no shared code-path bug." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are docs cleanups for v5 removals, but they cover different removed entries and are not the same change." + "reason": "Both are v5 docs cleanups, but they target different removed features and are not the same doc change." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "Unrelated: ONNX export error vs import-path breakage for GenerationMixin." + "reason": "ONNX export issue vs import-path regression; unrelated." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Both mention index errors, but the model/setup and failure context are different enough to treat as separate bugs." + "reason": "Both are index errors, but one is ZeRO-3 loading on Bert while the other is a mask/tensor shape mismatch; not the same bug." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Same generic error class, but no clear evidence the Bert ZeRO-3 load bug matches this mask-shape mismatch report." + "reason": "Different error signatures and likely different root causes." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Different problem domains: mask/tensor shape mismatch vs GPT-OSS tensor-parallel behavior." - }, - { - "left": "issue:44805", - "right": "issue:44918", - "accept": false, - "reason": "Related wording around embeddings, but not enough to confirm the same underlying bug." + "reason": "Shape-mismatch IndexError vs GPT-OSS tensor-parallel issue; unrelated." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Model registration tooling issue vs runtime indexing mismatch." + "reason": "Model-mapping generation failure is not the same as the mask/tensor indexing error." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Collective mismatch and Bert ZeRO-3 index error are different DeepSpeed failures." + "reason": "Collective mismatch and ZeRO-3 index error are different failure classes." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice preset bug is unrelated to ONNX export." + "reason": "Processor voice preset bug is unrelated to ONNX export." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same underlying DeepSpeed ZeRO-3/Bert zero-sized index error; the latter is a slightly more specific variant of the same crash." + "reason": "Near-identical ZeRO-3 index-0-out-of-bounds failures on Bert models with the same underlying loading path." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Different symptoms and code paths: detokenization bug vs LoRA merge/save inconsistency." + "reason": "Both affect output correctness, but one is detokenization in v5 and the other is LoRA merge/reload inconsistency." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy call vs GPT-OSS tensor-parallel issue are not the same bug." + "reason": "Liger kernel cross-entropy dispatch and GPT-OSS tensor parallel issues are unrelated." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "NVFP4 runtime error and LoRA merge/reload mismatch are unrelated." + "reason": "NVFP4 runtime error and LoRA save/reload mismatch are different bugs." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Both involve tensor-shape issues, but the training/FSDP checkpointing bug is distinct from this mask-indexing error." + "reason": "FSDP/activation-checkpoint recomputation mismatch is a different training bug from the mask/tensor index error." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Training DeepSpeed weight-dimension error vs collective mismatch on ranks are different failures." + "reason": "DeepSpeed weight-dimension error and collective mismatch are different issues." }, { "left": "issue:43975", "right": "issue:45237", "accept": false, - "reason": "Detokenization bug is unrelated to AMD GPU model-loading/runtime failure." + "reason": "Different models and different failure mechanisms." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "Bert ZeRO-3 loading crash vs GPT-OSS tensor-parallel problem are different code paths." + "reason": "ZeRO-3 Bert loading error is unrelated to GPT-OSS tensor-parallel behavior." }, { "left": "issue:34928", "right": "issue:36331", "accept": false, - "reason": "Activation-checkpointing/FSDP recomputation bug is unrelated to CustomTrainer loss signature breakage." + "reason": "Activation-checkpoint/FSDP mismatch is unrelated to the custom-trainer argument regression." }, { "left": "issue:43596", "right": "issue:45161", "accept": false, - "reason": "Different model families and failure modes." + "reason": "Different model loading path and different failure mode." }, { "left": "issue:43716", "right": "issue:45507", "accept": false, - "reason": "Image dtype mismatch on Mistral-3 is unrelated to GraniteMoEHybrid method invocation." + "reason": "Image-preprocessor dtype mismatch is unrelated to invalid-method calls in GraniteMoEHybrid." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Different models and problems: missing config field vs tokenizer patch crash." + "reason": "Gemma3/vLLM config attribute issue and tokenizer regex patch crash are unrelated." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Autocast dtype mismatch on Phi-tiny-MoE does not match GPT-OSS tensor-parallel behavior." + "reason": "Different models and different runtime pathways; same MoE label is too broad." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch version compatibility issue vs model-specific NVFP4 runtime error." + "reason": "Torch version incompatibility and NVFP4 runtime error are unrelated." }, { "left": "issue:44918", "right": "issue:45161", "accept": false, - "reason": "TRL SFT embedding unpacking failure vs GPT-OSS tensor-parallel issue are separate bugs." - }, - { - "left": "issue:43039", - "right": "issue:44918", - "accept": false, - "reason": "Same broad training area, but cross-entropy/kernel routing and embedding unpacking are not the same defect." + "reason": "TRL embedding unpacking issue is not the same as GPT-OSS tensor-parallel failure." }, { - "left": "issue:41762", - "right": "issue:43638", + "left": "issue:34928", + "right": "issue:41720", "accept": false, - "reason": "Both are ZeRO-3 load crashes, but the model-specific details are not enough to confirm one shared bug." + "reason": "FSDP checkpointing mismatch and Qwen3 auto-device-map assert are different bugs." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs issues about removed v5 commands/tasks, but they target different removed references." + "reason": "Both are docs regressions in v5, but they concern different removed commands and are separate cleanup tasks." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Mistral image dtype mismatch is unrelated to Qwen3-vl video StopIteration." + "reason": "Different model-specific dtype/video-processing failures." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Different model and failure mode; no strong sign of shared root cause." + "reason": "Preprocessor dtype mismatch and TRL embedding unpacking are unrelated." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Both touch training/inference math, but not the same concrete code path or bug." + "reason": "Different model/runtime bugs despite both touching numeric kernels." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "DeepSpeed collective mismatch vs trainer method signature change are unrelated." + "reason": "Collective mismatch and trainer-signature regression are unrelated." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger Kernel cross-entropy issue is not the same as a mask/index shape mismatch." + "reason": "Liger-kernel cross-entropy dispatch is unrelated to indexing/shape mismatch errors." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "FSDP activation-checkpointing bug is unrelated to a GenerationMixin import regression." + "reason": "Training/checkpointing bug and import regression are unrelated." }, { "left": "issue:41762", "right": "issue:43596", "accept": false, - "reason": "Different models and failure details; same ZeRO-3 theme only at a high level." + "reason": "Both are ZeRO-3 index errors, but they hit different model families and likely different root causes." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Different docs pages and different removed or changed behavior." + "reason": "Docs mentioning return_all_scores is unrelated to the removed run command docs issue." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Config attribute missing in vLLM vs CUDA assert on A800 are unrelated." + "reason": "Different models and different runtime errors." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger Kernel cross-entropy issue is unrelated to Qwen3-vl video StopIteration." + "reason": "Liger kernel cross-entropy issue and Qwen3-vl video StopIteration are unrelated." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Corrupted checkpoint vs model implementation can't load checkpoints are different root causes." + "reason": "Checkpoint corruption and NemotronH loading failure are different issues." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "Unit-test load failure for GLM-4.7 vs corrupted Ovis2 checkpoint are separate issues." + "reason": "Different model-loading failures with no clear shared root cause." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Torch 2.10 compatibility and pydantic PretrainedConfig regression are different breakages." + "reason": "Torch compatibility and pydantic PretrainedConfig-field breakage are unrelated." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding-window behavior, but the models/configs and concrete failures differ." + "reason": "Gemma3/vLLM attribute issue and Qwen3-MoE sliding_window bug are different model regressions." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Mask-shape mismatch on one model/setup vs Bert ZeRO-3 zero-size index error on another." + "reason": "Same broad index-error family, but different concrete symptoms and code paths." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Mistral3 lm_head tying bug is unrelated to an Ovis2 checkpoint corruption report." + "reason": "lm_head tying bug is unrelated to checkpoint corruption." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Qwen3.5 chat crash and streaming inference failure are different manifestations with no clear shared root cause." + "reason": "Related model family, but the failure modes differ: chat generation crash vs streaming inference error." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper load failure is unrelated to the pydantic PretrainedConfig regression." + "reason": "Whisper model runtime failure and pydantic field regression are unrelated." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Different MoE dtype/runtime issues on different models and operations." + "reason": "Different MoE/runtime failures with distinct stack traces." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to Gemma3 sliding_window_pattern missing." + "reason": "ImportError for GenerationMixin is unrelated to Gemma3/vLLM attribute issues." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure and template compilation error are different subsystems." + "reason": "Processor failure and template-node compilation error are unrelated." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are indexing/runtime issues, but the model path and error context are different." + "reason": "Both are index errors, but the reported causes and affected paths differ." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Failed model load in vLLM is unrelated to GGUF architecture support." + "reason": "PaddleOCR-VL load failure and GGUF gpt-oss support request are different issues." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The verdicts are mostly conservative, but the canonical summary overstates certainty: calling issue:43596/issue:43638 the \"only true duplicate pair\" is stronger than the packet supports. The titles do make them a strong candidate duplicate, but there are other closely related ZeRO-3 index-out-of-bounds reports (for example issue:41762) that cannot be fully ruled out from the provided metadata alone." + "accept": true, + "feedback": "The summary is grounded in the packet: it correctly identifies a loose zero-3/indexing subcluster and a separate v5 docs-removal subcluster, while treating the rest as mostly unrelated runtime/model failures. The soft-edge verdicts are generally conservative, and the one accepted pair (43596/43638) is well-supported by the titles as a closely related ZeRO-3 Bert index-out-of-bounds family." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -600,10 +588,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36331", "issue:39186", @@ -664,8 +652,8 @@ "issue:43531|issue:45070", "issue:44661|issue:45464", "issue:39290|issue:43742", - "issue:43054|issue:43994", "issue:43541|issue:44589", + "issue:43054|issue:43994", "issue:41093|issue:43596", "issue:42915|issue:43541", "issue:40990|issue:44960", @@ -690,20 +678,20 @@ "issue:44560|issue:45507", "issue:36331|issue:41762" ], - "prepared_review_unit_hash": "ed436d5d67c936884101b41503e9ba2e03984c73f251951de611f070c02d9905", + "prepared_review_unit_hash": "3d3cfe81a9eaae410c2b37143b572c1a4a6fa3f93905260e1f0a076ba86fe67e", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37565, + "serialized_chars": 37566, "estimated_input_tokens": 9392, "estimated_eval_tokens": 19040 }, @@ -712,225 +700,225 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is still mostly heterogeneous, but the v5 non-persistent-buffer corruption reports form a broader duplicate subcluster than before: issues 43644, 43950, and 44534 all describe the same loading-time buffer corruption regression. Most other soft edges remain unrelated or only loosely similar subsystems.", - "confidence": 0.87, - "canonical_issue_reason": "issue:43950 is still the best canonical issue because it states the regression most explicitly (`from_pretrained()` silently corrupts `register_buffer(persistent=False)` buffers) and best anchors the wider buffer-corruption subcluster that also includes issues 43644 and 44534.", + "summary": "The set is mostly heterogeneous and does not form one large duplicate cluster. The only clear duplicate family is the non-persistent buffer regression in Transformers v5 (`43644`, `43950`, `44534`).", + "confidence": 0.93, + "canonical_issue_reason": "issue:43950 is the clearest canonical representative of the duplicate family because it most precisely identifies the root bug: `from_pretrained()` corrupting `register_buffer(persistent=False)` buffers in Transformers 5.x.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43950 is the best global issue representative because it is the clearest, most specific report of the underlying v5 buffer-loading bug while the rest of the cluster remains too mixed to serve as a broader canonical issue.", + "best_issue_reason": "issue:43950 is the best issue to anchor the duplicate set: it is the most specific, technically precise, and directly describes the underlying regression rather than just the symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Different bugs: missing `pad_token_idx` in one config vs a pydantic `PretrainedConfig` field regression in the other." + "reason": "Different bugs: one is a missing `pad_token_idx` config field in `StableLmConfig`, the other is a Pydantic `PretrainedConfig` field regression in v5." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Both mention v5 regressions, but one is int4 OOM/memory behavior and the other is tied-weights handling; not the same bug." + "reason": "Both involve v5 behavior changes, but one is an int4 memory/OOM issue and the other is tied-weight handling for translation models; not the same bug." }, { "left": "issue:43541", "right": "issue:45161", "accept": false, - "reason": "Different MoE problems: torch dynamo tracing failure vs tensor-parallel support not working." + "reason": "Different failure modes: torch dynamo grouped_mm tracing vs tensor-parallel support for GPT-OSS MoE." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Unrelated failures: autocast dtype mismatch vs streaming chat/completions API failure." + "reason": "Unrelated model/runtime problems: dtype mismatch under autocast for Phi-tiny-MoE vs streaming inference failure for Qwen3.5-0.8B." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "One is an `_is_hf_initialized` constructor argument error; the other is a pydantic `PretrainedConfig` regression." + "reason": "`init_empty_weights` unexpected `_is_hf_initialized` argument is a distinct constructor/API mismatch, not the Pydantic field regression." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "Different domains: GGUF architecture support for gpt-oss vs Qwen3.5 streaming inference failure." + "reason": "One is GGUF support for GPT-OSS architecture, the other is a streaming chat/completions failure for Qwen3.5; no shared code-path bug." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Sliding-window handling for Qwen3-MoE is unrelated to the `PretrainedConfig`/pydantic issue." + "reason": "Qwen3-MoE sliding-window handling is unrelated to the Pydantic `PretrainedConfig` field issue." }, { "left": "issue:44661", "right": "issue:45464", "accept": false, - "reason": "Model-registration failure in tokenizer mapping is unrelated to the Qwen3.5 streaming bug." + "reason": "Tokenizer mapping/add-new-model-like failure is unrelated to Qwen3.5 streaming inference." }, { "left": "issue:39290", "right": "issue:43742", "accept": false, - "reason": "Different model-loading failures: Gemma3 `sliding_window_pattern` missing vs MobileLLM key error." - }, - { - "left": "issue:43054", - "right": "issue:43994", - "accept": true, - "reason": "Both describe the same SigLIP2 text-embedding / output-quality regression; one reports degraded embeddings, the other nonsensical model/pipeline results." + "reason": "Gemma3 sliding-window attribute error and MobileLLM key error are different model-loading issues." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "Different code paths: Mixtral grouped_mm tracing failure vs missing Float8 storage type." + "reason": "Grouped_mm tracing failure and missing Float8 storage are distinct runtime errors." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Both concern SigLIP2, but one is degraded text embeddings while the other is bad AutoModel/pipeline outputs; not enough evidence of the same underlying bug." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are index-shape errors, but they arise in different setups and model-loading paths." + "reason": "Different shape/index errors in separate loading paths; no clear duplicate relation." }, { "left": "issue:42915", "right": "issue:43541", "accept": false, - "reason": "Different MoE failures: FineGrainedFP8Config support vs grouped_mm tracing on float32." + "reason": "FineGrainedFP8Config failure for Qwen3 MoE is unrelated to Mixtral dynamo tracing." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "Different symptoms and models: perplexity anomaly on gpt-oss vs a GLM5 issue." + "reason": "Perplexity regression on GPT-OSS and a generic GLM5 issue are not the same underlying defect." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP weight-shape failure is not the same as TRL SFT embedding unpacking failure." + "reason": "FSDP `'weight' must be 2-D` and Qwen3.5 SFT unpacking failure are separate training-time bugs." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Both involve tied weights, but they are different models and different manifestations; not the same concrete bug." + "reason": "Both mention tied weights, but one is a Mistral3 model-specific missing tie and the other is a broader v5 translation-model issue; not the same concrete bug." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model initialization breakage is unrelated to flash-attn fallback blocking." + "reason": "Jupyter custom-model initialization regression is unrelated to flash-attn fallback checks." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttention linear_attention crash is unrelated to tied-weight translation-model issues." + "reason": "PagedAttentionCache linear_attention crash is unrelated to tied-weight handling." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "NemotronH checkpoint loading failure is unrelated to tied-weight regressions." + "reason": "NemotronH checkpoint loading and translation-model tied weights are different problems." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is not the same bug as tied-weight handling." + "reason": "Quantization/offloading behavior is not the same as tied-weight regressions." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A ReLU joke/discussion about DSA indexer is unrelated to docs mentioning a removed command." + "reason": "DSA indexer/ReLU discussion and docs command removal are unrelated." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "Different areas: FP8 MoE support vs flash-attn2 fallback gating." + "reason": "FineGrainedFP8 Qwen3 MoE failure is unrelated to flash-attn fallback gating." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "TRL embedding unpacking and GraniteMoEHybrid invalid-method errors are different failures." + "reason": "TRL embedding unpacking failure and GraniteMoEHybrid invalid method are separate model/training bugs." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "A typo in cache strategies is unrelated to outdated docs for the removed run command." + "reason": "A typo in cache strategy docs is unrelated to docs still mentioning a removed command." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Detokenization bug and Qwen-3-8B-NVFP4 runtime error are unrelated." + "reason": "DeepSeek detokenization regression and Qwen-3-8B-NVFP4 runtime failure are unrelated." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "These are the same bug: v5 corrupts or fills non-persistent buffers with junk during loading / `from_pretrained()`." + "reason": "Same underlying v5 regression: non-persistent buffers registered with `persistent=False` are getting corrupted/junk values during loading." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Clear duplicate wording and same bug: Transformers 5.x fills/corrupts non-persistent buffers during `from_pretrained()`." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime issues and GraniteMoEHybrid invalid-method calls are unrelated." + "reason": "Different model-specific failures with no shared bug or code path." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Both involve SigLIP2, but one is a docs/example mismatch and quantization failure while the other is model-quality degradation; not the same concrete bug." + "reason": "SigLIP2 docs example errors and SigLIP2 embedding quality regression are related only at model family level, not the same concrete bug." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Different failure modes: autocast dtype mismatch vs RLHF NaN/gradient explosion from 3D position IDs." + "reason": "Different Qwen3.5-related failures: autocast dtype mismatch vs RLHF NaN explosion from position IDs/SDPA fallback." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Bad AutoTokenizer error messaging and missing `all_tied_weights_keys` are unrelated bugs." + "reason": "Bad AutoTokenizer error messaging for Voxtral is unrelated to missing `all_tied_weights_keys` in Molmo." }, { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "Corrupted Ovis2 checkpoint and generic loading noise are different problems." + "reason": "Corrupted Ovis2 checkpoint and noisy load logging are not the same bug." }, { "left": "issue:44560", "right": "issue:45507", "accept": false, - "reason": "Qwen3-vl video StopIteration and GraniteMoEHybrid invalid method are unrelated." + "reason": "Qwen3-vl-embedding video `StopIteration` and GraniteMoEHybrid invalid method are unrelated." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Different training/API errors: unexpected `num_items_in_batch` kwarg vs DeepSpeed ZeRO-3 zero-sized index error." - }, - { - "left": "issue:43644", - "right": "issue:43950", - "accept": true, - "reason": "Same underlying regression: Transformers 5.0 corrupts non-persistent buffers during model loading / initialization." + "reason": "Trainer `compute_loss` signature mismatch and Gemma3/ZeRO-3 index error are separate issues." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The summary is mostly grounded, but one soft-edge verdict is not conservative enough: issue:43644 and issue:43950 look like the same v5 non-persistent-buffer corruption regression from their titles, so rejecting that pair is hard to justify. Because the buffer-corruption subcluster is therefore a bit broader than the summary implies, I would not accept this cluster evaluation as-is." + "accept": true, + "feedback": "The summary is grounded in the packet: the only clearly supported duplicate family is the non-persistent buffer regression (`43644` / `43950` / `44534`), and the accepted links are consistent with the issue titles. The soft-edge rejections are generally conservative and do not overclaim duplicate status from weak similarity alone." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -943,10 +931,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:38175", "issue:41553", @@ -1006,8 +994,8 @@ "issue:43881|issue:43883", "issue:44403|issue:44451", "issue:43975|issue:45520", - "issue:42947|issue:43856", "issue:43643|issue:45020", + "issue:42947|issue:43856", "issue:44403|issue:44488", "issue:43782|issue:44960", "issue:42491|issue:44368", @@ -1035,253 +1023,253 @@ "issue:42915|issue:43872", "issue:43257|issue:45305" ], - "prepared_review_unit_hash": "8904ab1fd950f36182f52f65e52956abf757a7993d3c6a67a8cb5eb1001f9b2a", + "prepared_review_unit_hash": "e38c6858ba3425ef79e36fe70ca05af31300dee0d9cfb35e5e54ee2f89d9a4c8", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 36, - "serialized_chars": 38602, - "estimated_input_tokens": 9651, - "estimated_eval_tokens": 19558 + "serialized_chars": 38600, + "estimated_input_tokens": 9650, + "estimated_eval_tokens": 19556 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly heterogeneous. Only a small subset is well-supported as same-bug duplicates: the SigLIP2 discrepancy pair, the remote_code/AutoConfig regression pair, and the stale v5 docs pair. The Qwen3/Whisper/ZeRO-3 and other same-area candidates are too weakly evidenced for a strict duplicate accept.", - "confidence": 0.78, - "canonical_issue_reason": "Issue 45020 is the broadest open umbrella in the cluster and clearly captures the remote_code loading-regression family.", + "summary": "This cluster is mostly heterogeneous loading/regression reports; only two soft pairs look like true duplicates. The strongest duplicate family is the `_is_hf_initialized`/`init_empty_weights` TypeError, and the docs-removal reports are also the same issue.", + "confidence": 0.71, + "canonical_issue_reason": "issue:44291 is the cleanest representative bug report: it states the exact `_is_hf_initialized` root cause, the `init_empty_weights` context, and generalizes beyond one model-specific symptom.", "canonical_pr_reason": null, - "best_issue_reason": "45020 is the strongest anchor for the cluster because it is open, broad, and directly covers a concrete related duplicate candidate (43643).", + "best_issue_reason": "issue:44291 is the best single issue to keep because it has the clearest repro and root-cause description; the other items are either narrower variants or unrelated regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Different models and failure modes: Voxtral error messaging vs GLM-4v-9b loading failure." + "reason": "Different failures: bad AutoTokenizer error-message behavior vs a model-specific loading crash." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Both are loading-related, but the specific bugs differ: bitsandbytes kwarg handling vs missing Float8 storage." + "reason": "Same broad incompatibility area, but not the same concrete bug or code path." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Both involve SigLIP2, but zero probabilities and worse embeddings are not enough to establish the same underlying bug." + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is degraded text embeddings; different symptoms." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "Gradient checkpointing/LoRA ineffective and int4 quantization CUDA reserved-memory OOM are different problems." + "reason": "Different problems: LoRA gradient checkpointing vs int4 memory reservation/OOM." }, { "left": "issue:43054", "right": "issue:43493", - "accept": true, - "reason": "These look like the same SigLIP2 implementation discrepancy; 43493 names the HF-vs-JAX mismatch and 43054 reports a likely symptom." + "accept": false, + "reason": "Same model family, but one is quality discrepancy and the other is HF-vs-JAX implementation mismatch; not enough to treat as the same bug." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Different model families and unrelated loader failures." + "reason": "Unrelated model-loading crash vs missing attribute on a different model class." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise is not the same bug as failure to load ScandiBERT." + "reason": "One is loading noise/logging, the other is a specific model load failure." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Detokenization regression and flash_attn import failure are unrelated code paths." + "reason": "Different subsystems: tokenizer detokenization vs flash_attn import/key handling on Python 3.13." }, { - "left": "issue:42947", - "right": "issue:43856", + "left": "issue:43643", + "right": "issue:45020", "accept": false, - "reason": "Both mention memory, but the model/setup and failure modes differ." + "reason": "Both involve remote code, but one is missing fields from AutoConfig and the other is a broader recent-version breakage report." }, { - "left": "issue:43643", - "right": "issue:45020", - "accept": true, - "reason": "43643 is a concrete instance of the broader remote_code regression described by 45020: AutoConfig returns missing fields." + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Different training-performance issues: gradient checkpointing vs Qwen3 MoE memory usage." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Loading noise is not the same underlying bug as failing to load a specific BERT checkpoint." + "reason": "Generic load-noise complaint vs a specific tokenizer/model loading failure." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Different model families and different errors; these do not look like the same bug." + "reason": "Different model families and different failure modes." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "A Qwen3 MoE compatibility failure is not the same as a tie_word_embeddings warning during fine-tuning." + "reason": "LoRA/Qwen3-MoE incompatibility vs a warning about tie_word_embeddings." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Both are special-config load failures, but Gemma3 ZeRO-3 and Qwen3 MoE FP8 issues are different bugs." + "reason": "Both are Qwen/Gemma loading issues, but they affect different models and different error paths." }, { "left": "issue:43975", "right": "issue:44451", "accept": false, - "reason": "Same as above: detokenization regression vs checkpoint load failure are unrelated." + "reason": "Different symptoms and model targets." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Remote_code loading regressions and Kimi-K2.5 tokenizer codec/fix_mistral_regex behavior are different tokenizer issues." + "reason": "Both are version regressions, but one is broad remote_code breakage and the other is a Kimi tokenizer codec/regex warning bug." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer-mapping failure in add-new-model-like is a different code path from remote_code loading regressions." + "reason": "Different areas: tokenizer mapping helper failure vs remote_code loading regressions." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "Importing a missing class and bitsandbytes storage keyword handling are unrelated." + "reason": "ImportError for a VL class is unrelated to the bitsandbytes/_is_hf_initialized TypeError." }, { "left": "issue:43054", "right": "issue:43278", "accept": false, - "reason": "Both mention embeddings, but SigLIP2 quality/discrepancy is not the same as a BF16-to-FP32 dtype change." + "reason": "Different embedding issues: SigLIP2 text quality vs dtype drift between train and eval." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression and flash-attn fallback blocking are separate problems." + "reason": "Tokenizer loading failure vs flash-attn fallback gating logic." }, { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "These are the same stale-doc issue: both still reference pipeline tasks removed in v5." + "reason": "Same docs bug: stale pipeline-task references after v5 removals." }, { "left": "issue:43792", "right": "issue:44220", "accept": false, - "reason": "The audio feature-extraction path is plausible, but the packet does not establish the same root cause under a strict accept standard." + "reason": "Likely related audio path, but the reports are not specific enough to prove the same concrete bug." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer regex crash is unrelated to the flash-attn fallback gating problem." + "reason": "Tokenizer regex crash vs flash-attn2 fallback gating are different issues." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Untied lm_head weights and silent non-persistent buffer corruption are different regressions." + "reason": "Different state-management bugs: untied lm_head weights vs non-persistent buffer corruption." }, { "left": "issue:43257", "right": "issue:43866", "accept": false, - "reason": "Qwen3 MoE weight conversion and Ovis2 checkpoint corruption are not the same bug." + "reason": "Different model/domain and different failure mode." }, { "left": "issue:42915", "right": "issue:45020", "accept": false, - "reason": "Qwen3 MoE FP8 loading failure is unrelated to the remote_code regression in 45020." + "reason": "Different Qwen3-MoE FP8 training bug vs a broad remote_code regression report." }, { "left": "issue:43742", "right": "issue:45399", "accept": false, - "reason": "MobileLLM key error and flash-attn fallback blocking are different failures." + "reason": "KeyError while loading a model vs flash-attn fallback selection logic." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Different models and different tokenizer/loading symptoms." + "reason": "Tokenizer loading problems, but different models and different error behavior." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "Missing-key load failure for MobileLLM does not look like the remote_code missing-fields regression." + "reason": "Different bug classes; both are loading issues only at a very broad level." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are Qwen3-family ZeRO-3 load failures, but the packet does not establish a shared root cause; too weak for strict duplicate acceptance." + "reason": "Both mention ZeRO-3 and Qwen, but they are different model versions with different load failures." }, { "left": "issue:43872", "right": "issue:44291", - "accept": false, - "reason": "Similar initialization context, but the specific failures differ: Int8Params kwarg vs unexpected _is_hf_initialized in init_empty_weights." + "accept": true, + "reason": "Same exact `_is_hf_initialized` / `init_empty_weights` TypeError family." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model families and different checkpoint-loading failures." + "reason": "Model-loading test failure for GLM-4.7-Flash vs NemotronH checkpoint compatibility issue." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "RLHF NaN/SDPA fallback and flash_attention s_aux=None crash are unrelated." + "reason": "Different training/runtime failures: RLHF gradient explosion vs flash_attention auxiliary-state crash." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Different models and different failure characteristics." + "reason": "Different tokenizer/model load reports." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Qwen3 MoE weight conversion and gradient averaging under DeepSpeed are different training/loading bugs." + "reason": "Different Qwen3 MoE training/loading bugs; no shared concrete code path is evident." } ] }, "evaluator_result": { - "accept": false, - "feedback": "Mostly grounded, but not conservative enough for a strict accept. A few positive soft-edge calls go beyond what the packet supports: 43792\u219444220 and 43940\u219445313 are plausible same-area regressions, but the packet only gives titles/body lengths and does not establish a shared root cause. The remote_code / docs / SigLIP2 links are better grounded, but the accepted pairs are not all equally well-evidenced." + "accept": true, + "feedback": "The clustering is grounded in the packet and the verdicts stay conservative. The two accepted pairs are well-supported: 43872/44291 share the same `_is_hf_initialized` / `init_empty_weights` TypeError family, and 43827/44509 are both the v5 docs stale-pipeline-task issue. The rest are reasonably treated as heterogeneous or only loosely related." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1294,10 +1282,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:24643", "issue:30064", @@ -1379,20 +1367,20 @@ "issue:42491|issue:43720", "issue:44960|issue:44977" ], - "prepared_review_unit_hash": "707f357bf46b62670a7e9c71fb071223dd325ef79575fa9a72e614b786287967", + "prepared_review_unit_hash": "b898a8eee5d636182263cdf90a11ba17a159618c37a89709713edbad6149c79e", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36153, + "serialized_chars": 36154, "estimated_input_tokens": 9039, "estimated_eval_tokens": 18334 }, @@ -1401,198 +1389,198 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous. I only accepted the near-verbatim duplicate pairs around mask-shape IndexError, custom model initialization in Transformers 5.0.0, and non-persistent buffer corruption; the rest look like distinct bugs.", - "confidence": 0.86, - "canonical_issue_reason": "issue:43644 is the clearest representative of the non-persistent-buffer corruption regression; it matches 44534 almost verbatim and is the most concise canonical report.", + "summary": "This cluster is mostly heterogeneous transformer bug reports. Only two soft duplicate pairs look strong enough to accept: the non-persistent-buffer regression and the SigLIP2 wrong-output reports. The rest are different failure modes or different model/code paths.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 43644 is the best cluster anchor: it is the clearest report of the non-persistent-buffer corruption regression and is earlier than the near-exact duplicate 44534.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the strongest global issue candidate because it states the bug clearly and has an almost exact duplicate in 44534.", + "best_issue_reason": "43644 is the strongest issue to keep as representative because it is specific, early, and directly describes the underlying regression rather than a symptom variant.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Different failures: trainer signature mismatch vs mask/tensor shape IndexError." + "reason": "Different bugs: custom trainer signature mismatch vs a mask/tensor shape IndexError." }, { "left": "issue:41093", "right": "issue:44805", - "accept": true, - "reason": "Same mask/tensor shape mismatch error; the two reports describe the same underlying IndexError." + "accept": false, + "reason": "Both are IndexError-shaped reports, but they appear to involve different examples/models and are not clearly the same failure." }, { "left": "issue:43872", "right": "issue:45538", "accept": false, - "reason": "Unrelated bugs: bitsandbytes Int8Params constructor error vs tokenizer max_length default." + "reason": "Unrelated: bitsandbytes constructor incompatibility vs CLIPTokenizer max-length behavior." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both are Qwen video-related, but one is a broad regression and the other is a specific StopIteration in a different code path." + "reason": "Both concern Qwen video paths, but one is a broad v5.3.0 regression and the other is a StopIteration in the embedding model; not the same concrete bug." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "Different subsystems and failure modes: PagedAttentionCache group-type crash vs checkpoint loading failure." + "reason": "Different model families and failures: Qwen3.5 PagedAttentionCache crash vs NemotronH checkpoint loading." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "Unrelated model-support issue vs dtype mismatch under autocast." + "reason": "GGUF/gpt-oss support is unrelated to the Phi-tiny-MoE autocast dtype mismatch." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "Completely different areas: DeepSpeed training error vs image processor segmentation-map handling." + "reason": "DeepSpeed training weight-shape error vs void segmentation map processing are unrelated." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "Different bugs in different pipelines: AudioFlamingo3 batched inference leak vs Whisper empty-transcription handling." + "reason": "Different audio/text workflows and different bugs: embedding/token leak across tracks vs empty-transcription generation failure." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "Both mention Qwen-family models, but the bugs are different: GGUF support vs sliding_window behavior." + "reason": "Different Qwen-related issues: GGUF architecture support vs sliding_window behavior." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different attention/backend crashes with different triggers and model families." + "reason": "Different attention/compile failures on different models and code paths." }, { "left": "issue:43645", "right": "issue:43646", - "accept": true, - "reason": "Same Transformers 5.0.0 custom-model initialization regression; one is just the Jupyter-specific manifestation." + "accept": false, + "reason": "Close in theme, but one is notebook-specific and the other is a broader custom-init report; not enough to confirm the same bug." }, { "left": "issue:43950", "right": "issue:45325", "accept": false, - "reason": "Different regressions: non-persistent buffer corruption vs Qwen2.5-VL rope-index scaling." + "reason": "Non-persistent buffer corruption is unrelated to the Qwen2.5-VL rope-index scaling bug." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Different Qwen3.5 video/embedding failures with distinct symptoms and code paths." + "reason": "Different Qwen3.5 video/training failures: StopIteration in video embedding vs TRL SFT embedding unpacking." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "GGUF architecture support is unrelated to Float8 storage lookup failure." + "reason": "GGUF model support is unrelated to the Float8 storage TypeError." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Different loading problems: missing import for a class vs meta-device loading regression." + "reason": "Import error for a missing model class vs meta-device loading regression." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Different bugs with different models and triggers." + "reason": "Qwen3Moe FP8 failure and meta-device loading are different bugs." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "Unrelated: 3d_parallel.py runtime failure vs GGUF model-architecture support." + "reason": "3d_parallel.py runtime issue vs GGUF gpt-oss support are unrelated." }, { "left": "issue:38175", "right": "issue:43994", - "accept": false, - "reason": "SigLIP2 output-quality issue vs pipeline/AutoModel mismatch; not the same bug." + "accept": true, + "reason": "Same underlying SigLIP2 wrong-output bug for the same checkpoint; one report says zero probabilities, the other says nonsensical outputs." }, { "left": "issue:43646", "right": "issue:43950", "accept": false, - "reason": "Different Transformers 5.x regressions: custom model init vs non-persistent buffer corruption." + "reason": "Custom model initialization regression vs non-persistent buffer corruption are different failures." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Different failures: Qwen video StopIteration vs mask-shape IndexError." + "reason": "Different errors and likely different models/code paths; not the same bug." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "Different errors in different areas: GenerationMixin import vs FineGrainedFP8Config failure." + "reason": "GenerationMixin import failure is unrelated to Qwen3Moe FineGrainedFP8Config errors." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "Different edge cases: ZeRO-3 Bert init crash vs TOKENIZER_MAPPING_NAMES lookup failure." + "reason": "DeepSpeed zero3 Bert index error vs tokenizer-mapping add-new-model-like failure are different bugs." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both are config-loading regressions, but the titles point to different models and different validation paths; not enough to merge confidently." + "reason": "Both are config-validation problems, but they affect different models and are not clearly the same regression." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Different loading bugs in unrelated code paths." + "reason": "Different load-time failures: zero-sized tensor index error vs TOKENIZER_MAPPING_NAMES edge case." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Different problems: gpu/cpu offload loading regression vs mmap OOM workaround." + "reason": "device_map/offload loading failure is unrelated to mmap-on-Strix-Halo OOM avoidance." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Near-verbatim duplicate: both report the same non-persistent-buffer junk/corruption regression in v5." + "reason": "Exact same regression: Transformers v5 fills non-persistent buffers with junk." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "Different Qwen3.5 bugs: ZeRO-3 weight loading vs save_pretrained visual-encoder key regression." + "reason": "One is ZeRO-3 weight loading for language_model; the other is save_pretrained writing incorrect visual encoder keys." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Different SigLIP2 issues: nonsensical outputs vs documentation/example failures." + "reason": "Same model family, but one is an output correctness bug and the other is a documentation example/config/quantization issue." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Different model-loading/compatibility failures with no clear shared code path." + "reason": "Different loading/backward-compatibility problems: Qwen3_moe LoRA compatibility vs BitNet packed-weight unpacking." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "Different model families and different flash-attention/generation issues." + "reason": "Different model families and symptoms: GLM5 vs Qwen3.5 flash-attention generation bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays mostly conservative. The three accepted pairs are well supported by near-identical titles: the mask-shape IndexError reports, the Transformers 5.0.0 custom-model initialization reports, and the non-persistent-buffer junk reports. The rejected pairs are generally justified by clearly different titles/code paths, and I don\u2019t see an overstatement that would warrant rejection." + "feedback": "The summary is grounded in the packet and stays conservative. The representative issue choice for 43644 over 44534 is well supported by the titles/timestamps, and the two accepted soft pairs are plausible same-bug duplicates. The rejected pairs are consistently framed as different bugs or different code paths without overclaiming." }, "evaluator_used": true, "retried": false, @@ -1608,10 +1596,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36010", "issue:39404", @@ -1690,9 +1678,9 @@ "issue:43425|issue:43957", "issue:44387|issue:44756", "issue:43854|issue:45216", - "issue:42947|issue:43381", "issue:44291|issue:44756", "issue:44292|issue:44912", + "issue:42947|issue:43381", "issue:45276|issue:45292", "issue:44743|issue:44960", "issue:43381|issue:44928", @@ -1700,262 +1688,262 @@ "issue:44164|issue:44756", "issue:43646|issue:43708" ], - "prepared_review_unit_hash": "c68f4d016d0401b63105ccc951a630c4261dc9ac7b0b287cc58ccbc0fc4bdd4a", + "prepared_review_unit_hash": "bf7f871294ff767ca3a302ab9e9626ce5155d2422a3611d7b3e1beb0628cab13", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 37, - "serialized_chars": 38100, - "estimated_input_tokens": 9525, - "estimated_eval_tokens": 19306 + "serialized_chars": 38104, + "estimated_input_tokens": 9526, + "estimated_eval_tokens": 19308 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are unrelated or only broadly similar. The only clear duplicate cluster is the resize_token_embeddings regression around output/decoder embeddings, with 45292 as the best umbrella issue.", - "confidence": 0.73, - "canonical_issue_reason": "Issue 45292 is the most general, model-agnostic statement of the resize_token_embeddings failure, so it best serves as the duplicate anchor for the embedding-resize cluster.", + "summary": "This set is mostly a grab-bag of unrelated Transformers bug reports, with a few small duplicate families. The clearest duplicate set is around `resize_token_embeddings` not updating all tied embeddings, and there is also a likely shared lm_head weight-tying regression in multimodal generation models. No PRs are present.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45276 is the best canonical issue for the strongest duplicate family: it states the broader `resize_token_embeddings` propagation bug and subsumes the more specific follow-ups.", "canonical_pr_reason": null, - "best_issue_reason": "45292 is the broadest and cleanest representation of the underlying bug; the other accepted issues are model-specific manifestations of the same resize_token_embeddings regression.", + "best_issue_reason": "issue:45276 is the most representative issue in the cluster because it is concrete, well-scoped, and has clear duplicate support from closely related reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Both are model-loading failures, but they name different models and don\u2019t clearly point to the same concrete bug." + "reason": "Both are tokenizer/model load failures, but they target different checkpoints and describe separate report details; not enough evidence of the same underlying bug." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Completely different symptoms: checkpoint resumption math vs SGD optimizer arguments not being used." + "reason": "Trainer resume/max_steps logic and SGD optimizer args are unrelated subsystems and failure modes." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "Different models and different failure modes; both are loading issues, but not the same bug." + "reason": "Different models and different loading errors; no shared concrete bug path is evident." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "One is Qwen3 MoE training memory use, the other is quantization/offloading behavior; too far apart." + "reason": "Both mention memory/efficiency, but one is MoE training memory usage and the other is quantization offloading behavior." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "Different model families and different bugs; no shared concrete code-path problem." + "reason": "These concern unrelated problems: Qwen3 MoE version compatibility vs Gemma4 token embedding resizing." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "Both involve Qwen3 variants, but one is a device-mapping crash and the other is FP8 config failure; not the same issue." + "reason": "One is an auto device-mapping CUDA assert, the other is a Qwen3 MoE LoRA version-compatibility issue." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Different models and distinct loading problems; not a duplicate." + "reason": "Different code paths and symptoms; no duplicate relationship is supported." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Both mention SigLIP2, but one is a broken docs example and the other is bad inference output; different underlying issues." + "reason": "Same model family, but one is a docs/example mismatch and quantization issue while the other is bad inference output from AutoModel/pipeline." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/model-loading complaints, but they affect different models and don\u2019t clearly share one root cause." + "reason": "Different models and different tokenizer-loading failures; too broad to treat as the same bug." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "Both relate to memory, but one is training efficiency and the other is int4 quantization OOM; not the same bug." + "reason": "Both are memory-related, but the causes differ: training inefficiency vs increased reserved memory under int4 quantization." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Completely different code paths: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer." + "reason": "Different model families and different crash points (chat flow vs serve processor attribute error)." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Different problems: FP8 config failure vs mmap/OOM on Strix Halo." + "reason": "Qwen3 MoE LoRA compatibility and Strix Halo mmap OOM are unrelated issues." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Both are Transformers 5.0 regressions, but one is junk buffers and the other is notebook/custom-model initialization." + "reason": "Both are Transformers 5.0 regressions, but one is junk non-persistent buffers and the other is notebook custom-model initialization." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Unrelated topics: tokenizer-mapping registration vs compiled flex_attention on torch>=2.9." + "reason": "Tokenizer-mapping registration failure and torch>=2.9 flex_attention compilation are unrelated." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different model-specific loading failures with no clear shared code path." + "reason": "Version incompatibility for a Qwen3 MoE LoRA checkpoint is not the same bug as a Qwen3-VL shape mismatch." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both touch flash attention, but the crashes come from different conditions and different bugs." + "reason": "Both touch flash attention, but the root causes differ: misread 3D position_ids vs `s_aux=None` attribute handling." }, { "left": "issue:43404", "right": "issue:43454", - "accept": false, - "reason": "Both are untied lm_head reports, but they involve different model classes and are not clearly the same bug." + "accept": true, + "reason": "Both report missing lm_head weight tying causing broken generation in multimodal model classes; they appear to share the same fix path." }, { "left": "issue:42491", "right": "issue:43856", "accept": false, - "reason": "Different Qwen3-related issues: LoRA compatibility vs training memory usage." + "reason": "These are unrelated Qwen3 MoE issues with different symptoms and code paths." }, { "left": "issue:45276", "right": "issue:45335", "accept": true, - "reason": "Same resize_token_embeddings regression pattern: both say model-specific embeddings are not updated after resizing, plausibly fixed by one PR." + "reason": "Both describe `resize_token_embeddings` failing to propagate to nested/tied embeddings in related model families, which looks like the same underlying regression." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to Whisper pipeline return_language behavior." + "reason": "Importing `GenerationMixin` and Whisper `return_language` pipeline behavior are unrelated." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Different save/load failures: extra_state handling vs init_empty_weights argument mismatch." + "reason": "Save/load extra_state handling and `init_empty_weights` constructor argument handling are different loading-path bugs." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Both are loading-related, but one is a vocab-size mismatch and the other is a model load failure for a different BERT model." + "reason": "One is tokenizer vocab-size mismatch, the other is a specific model loading failure; not the same bug." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Unrelated problems in different domains: Qwen3 device mapping vs AudioFlamingo3 batched inference leakage." + "reason": "CUDA assert in Qwen3 auto device mapping and AudioFlamingo3 token/embedding leakage are unrelated." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both mention quantization/memory, but the underlying issues differ: offloading behavior vs reserved-memory growth." + "reason": "Both involve quantization and memory, but one is offloading behavior and the other is int4 reserved-memory OOM." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "SigLIP2 docs/model issue is unrelated to GLM5." + "reason": "SigLIP2 example/model issues and GLM5 are entirely different reports." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Torch version incompatibility is not the same as meta-device loading failures." + "reason": "Torch 2.10 incompatibility and meta-device model loading failures are different compatibility problems." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Both are OOM-related, but the cause and code path are different." + "reason": "Both mention OOM, but one is int4 quantization reserved memory and the other is mmap on Strix Halo." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different models and different regressions: unit-test loading vs save_pretrained checkpoint correctness." - }, - { - "left": "issue:42947", - "right": "issue:43381", - "accept": false, - "reason": "Both concern gradient checkpointing, but one is ineffective with PEFT LoRA and the other is eval-mode misuse." + "reason": "Different model-loading regressions: GLM-4.7 unit tests vs Qwen3.5 save_pretrained checkpoint correctness." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "TypeError in init_empty_weights is unrelated to mmap/OOM on Strix Halo." + "reason": "Unexpected `_is_hf_initialized` during meta init and Strix Halo mmap OOM are unrelated." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Different quantization formats and different failure modes; not the same bug." + "reason": "Both are quantization-related, but they affect different models and different quantization paths, so they are not clearly the same bug." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Gradient checkpointing in PEFT/LoRA and gradient checkpointing in eval mode are different issues." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same underlying resize_token_embeddings bug; 45292 is the generic output_embeddings case and 45276 is a model-specific manifestation." + "reason": "These are effectively the same `resize_token_embeddings` regression, with 45276 phrased more broadly." }, { "left": "issue:44743", "right": "issue:44960", "accept": false, - "reason": "Different models and unrelated behavior: recurrent state reset vs GLM5." + "reason": "Qwen3.5 recurrent-state reset and GLM5 are unrelated." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to the Qwen3.5 3D position_ids/SDPA issue." + "reason": "Gradient checkpointing eval-mode behavior and 3D position_ids causing SDPA fallback are unrelated." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Different memory/loading problems with different root causes." + "reason": "Meta-device loading breakage and disabling mmap on Strix Halo are different problems." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "Save/from_pretrained extra_state handling is unrelated to mmap-induced OOM." + "reason": "Save/from_pretrained extra_state handling is unrelated to mmap-related OOM." }, { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Custom model initialization in notebooks is unrelated to resume_from_checkpoint max_steps calculation." + "reason": "Custom model initialization regressions and trainer checkpoint step math are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate cluster is the resize_token_embeddings regression, which is strongly supported by the issue titles and the high-similarity pairings. The rejected soft edges are generally justified by distinct models, symptoms, or code paths." + "feedback": "Grounded and conservative overall. The summary stays within the packet, the resize_token_embeddings family is well supported by the strongest soft pairs, and the other verdicts avoid over-claiming duplicates where evidence is weak. No PRs are present in the packet." }, "evaluator_used": true, "retried": false, @@ -1971,10 +1959,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:32090", "issue:36032", @@ -2046,266 +2034,266 @@ "issue:44479|issue:44805", "issue:43994|issue:44079", "issue:44671|issue:44991", - "issue:43856|issue:44756", "issue:43425|issue:44918", + "issue:43856|issue:44756", "issue:43596|issue:44683", "issue:44589|issue:45084", "issue:39404|issue:44671", "issue:40990|issue:42947", "issue:36032|issue:43452", "issue:44991|issue:45081", - "issue:32090|issue:39404", "issue:43582|issue:43632", + "issue:32090|issue:39404", "issue:43957|issue:44164", "issue:40990|issue:43421", "issue:44164|issue:44479", "issue:43883|issue:44488", "issue:44488|issue:45081" ], - "prepared_review_unit_hash": "0ef6b833a51d465ead549fc8f21b8848bf84024bfb42390ce5b780332695f9a7", + "prepared_review_unit_hash": "f48db81fd8dfa6cf288828f0a5c95cc80c3976bcf58b4726d4c67c6a6ae5dbc7", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 35, - "serialized_chars": 38224, - "estimated_input_tokens": 9556, - "estimated_eval_tokens": 19368 + "serialized_chars": 38227, + "estimated_input_tokens": 9557, + "estimated_eval_tokens": 19370 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a noisy grouping of Transformers v5 loading/regression reports across many unrelated models and code paths. One soft edge is a true duplicate around `resize_token_embeddings` not propagating to tied/output embeddings in T5Gemma; the rest are too distinct to merge.", - "confidence": 0.55, - "canonical_issue_reason": "Issue 43611 is the broadest and most central representative: a v5 model-loading regression. Most other items are narrower variants of loading/tokenizer/config breakage rather than the same exact bug.", + "summary": "The cluster is mostly a set of unrelated Transformers regressions; only one pair looks like a true duplicate/same bug. Most soft matches are too broad or target different code paths/models.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45292 is the broader, earlier report of resize_token_embeddings not propagating to output embeddings; issue:45335 looks like a specific t5gemma manifestation of the same bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43611 best anchors the cluster because it captures the dominant theme of v5 breaking model loading without being overly model-specific.", + "best_issue_reason": "issue:45292 is the best representative issue because it states the general failure mode without narrowing to one model subclass.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Both are v5 regressions, but they hit different subsystems and models: video input handling vs strict config loading." + "reason": "Different regressions: video-input handling for Qwen VL models vs strict config blocking granite_speech loading." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same underlying bug: `resize_token_embeddings` is not propagating to the tied/output embedding path, including `decoder.embed_tokens` in T5Gemma." + "reason": "Same underlying bug: resize_token_embeddings fails to update downstream embeddings; 45335 is a t5gemma-specific instance." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "Different failures on different model/tokenizer assets; one is vocab-size/token mismatch, the other is a general load failure." + "reason": "Both concern tokenizer loading, but the model symptoms and likely root causes differ; not clearly the same bug." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Custom model initialization in notebooks and `extra_state` handling are separate code paths and symptoms." + "reason": "Custom model init in notebooks vs save/from_pretrained extra_state handling are different code paths." }, { "left": "issue:43257", "right": "issue:44863", "accept": false, - "reason": "Different architectures and load paths: Qwen3 MoE conversion with accelerate/deepspeed vs NemotronH checkpoint loading." + "reason": "Different model-loading failures for different architectures; only broad subsystem overlap." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "One is a version compatibility complaint; the other is a specific autocast dtype mismatch on Phi-tiny-MoE." + "reason": "Torch version incompatibility is not the same as a Phi autocast dtype mismatch." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Autocast dtype mismatch and meta-device loading are different regressions with different triggers." + "reason": "One is autocast dtype mismatch; the other is meta-device loading breakage. Different failures." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup failure and Strix Halo mmap/OOM are unrelated runtime issues." + "reason": "Float8 storage lookup error vs Apple Silicon mmap OOM are unrelated." }, { "left": "issue:43931", "right": "issue:44863", "accept": false, - "reason": "Both are loading errors, but for different models and failure modes: Qwen3-VL shape mismatch vs NemotronH checkpoint incompatibility." + "reason": "Different checkpoint-loading problems for different models; no clear shared bug." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "Isolated reproduction of another issue and a PagedAttentionCache linear_attention crash are not the same bug." + "reason": "Isolated reproduction of an older issue vs PagedAttentionCache group-type crash are different issues." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Base model loading with `base_model_prefix` and Jupyter custom-model initialization are distinct regressions." + "reason": "Base_model_prefix loading regression and Jupyter custom-model initialization regression are separate bugs." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "Torch version incompatibility is not the same as the Mistral-3 image-preprocessor dtype mismatch." + "reason": "Torch compatibility issue vs Mistral-3 image dtype mismatch; not the same defect." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper `return_language` pipeline regression and MXFP4 fallback behavior are unrelated." + "reason": "Whisper pipeline language return regression vs quantization fallback issue are unrelated." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "PagedAttentionCache group-type crash and template-node compilation error are different failures." + "reason": "PagedAttentionCache linear_attention crash and template-node compilation error are different code paths." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs-related, but they concern different removed/changed pipeline APIs and are not the same issue." + "reason": "Both are docs-related, but they cover different API/task deprecations." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Different Qwen3-VL loading problems with different symptoms; not enough evidence of one shared defect." + "reason": "Qwen3VL MoE loading regression and Qwen3-VL weight-shape mismatch are not the same concrete bug." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export/compiler-check failure and Perceiver interpolation failure are separate bugs." + "reason": "torch.export/torch_compilable_check failure is unrelated to Perceiver interpolate_pos_encoding behavior." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video input regression and mask shape mismatch are unrelated." + "reason": "Video input regression and mask-shape IndexError are different problems." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Nonsensical SigLIP outputs and `ModelOutput` key assignment are different issues." + "reason": "Incorrect SigLIP2 outputs vs ModelOutput key assignment bug are distinct." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT prediction regression and tokenizer loading failure for a different model are not the same bug." + "reason": "CamemBERT masked-LM prediction issue and tokenizer-loading failure are unrelated." }, { - "left": "issue:43856", - "right": "issue:44756", + "left": "issue:43425", + "right": "issue:44918", "accept": false, - "reason": "Training memory usage and Strix Halo mmap/OOM are different performance/runtime problems." + "reason": "Torch version compatibility and Qwen3.5 embedding unpacking/trainer issue are different." }, { - "left": "issue:43425", - "right": "issue:44918", + "left": "issue:43856", + "right": "issue:44756", "accept": false, - "reason": "Torch compatibility and Qwen3.5 unpacking with TRL SFT trainer are unrelated." + "reason": "Training memory efficiency and mmap OOM avoidance are not the same bug." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "DeepSpeed Zero3/BertModel indexing and compiled flex_attention on torch >=2.9 are different failures." + "reason": "Deepspeed zero3/BertModel index error is unrelated to compiled flex_attention on newer torch." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Float8 storage resolution and non-template-node compilation errors are unrelated." + "reason": "Different failure surfaces: storage type lookup vs template compilation." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper pipeline behavior and CamemBERT masked LM outputs are unrelated model-specific regressions." + "reason": "Whisper pipeline regression and CamemBERT prediction regression are unrelated." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Perplexity quality on gpt-oss and gradient checkpointing with LoRA are different concerns." + "reason": "High perplexity on a model is not the same as ineffective gradient checkpointing with LoRA." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Tokenizer method-name conflict and gguf_file loading breakage are different tokenizer/model-loading bugs." + "reason": "T5 tokenizer special-token conflict and gguf_file loading breakage are different tokenizer/model-loading bugs." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Tokenizer loading regression for one model and Mistral regex patch crash are separate issues." + "reason": "Both are tokenizer load failures, but they involve different models and different failure mechanisms." }, { - "left": "issue:32090", - "right": "issue:39404", + "left": "issue:43582", + "right": "issue:43632", "accept": false, - "reason": "Trainer GPU broadcast TypeError and Whisper return_language regression are unrelated." + "reason": "Apple Silicon allocator TypeError and _is_hf_initialized flag regression are unrelated." }, { - "left": "issue:43582", - "right": "issue:43632", + "left": "issue:32090", + "right": "issue:39404", "accept": false, - "reason": "Apple Silicon allocator TypeError and `_is_hf_initialized` regression are distinct." + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper return_language pipeline regression." }, { "left": "issue:43957", "right": "issue:44164", "accept": false, - "reason": "Meta-device loading and `extra_state` save/load handling are separate code paths." + "reason": "Meta-device loading breakage and extra_state save/from_pretrained handling are different issues." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity issue and runtime special-token/post-processor update bug are unrelated." + "reason": "Perplexity quality issue and runtime special-token post-processor update feature are unrelated." }, { "left": "issue:44164", "right": "issue:44479", "accept": false, - "reason": "`extra_state` serialization and video input regression are unrelated." + "reason": "extra_state serialization and Qwen video-input regression are different bugs." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Missing `all_tied_weights_keys` attribute and failure to load a specific BERT model are different bugs." + "reason": "Missing tied-weights metadata in Molmo and tokenizer loading failure for sleng-bert are unrelated." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Model loading failure for `cjvt/sleng-bert` and the Mistral regex patch crash are unrelated." + "reason": "Different tokenizer-loading failures affecting different models and code paths." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The chosen anchor issue and canonical issue are reasonable, and the only accepted soft edge (45292/45335) is well-supported as the same resize_token_embeddings bug. The remaining rejections are appropriately cautious given the differing models, subsystems, and failure modes." + "feedback": "The summary is well-grounded in the packet and stays conservative. The one accepted duplicate pair (45292/45335) is plausibly the same resize_token_embeddings issue with a model-specific manifestation, and the other soft pairs are appropriately treated as distinct or only superficially related." }, "evaluator_used": true, "retried": false, @@ -2321,10 +2309,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:34567", "issue:36032", @@ -2348,7 +2336,6 @@ "issue:43645", "issue:43646", "issue:43716", - "issue:43819", "issue:43825", "issue:43883", "issue:43994", @@ -2389,9 +2376,9 @@ "issue:43643|issue:44530", "issue:44451|issue:44987", "issue:43452|issue:43482", + "issue:43541|issue:45084", "issue:36032|issue:42915", "issue:43317|issue:43482", - "issue:43541|issue:45084", "issue:42915|issue:44530", "issue:39404|issue:43825", "issue:43404|issue:44466", @@ -2406,232 +2393,225 @@ "issue:45216|issue:45357", "issue:43322|issue:43540", "issue:44403|issue:45081", - "issue:44451|issue:45081", - "issue:43377|issue:43819" + "issue:44451|issue:45081" ], - "prepared_review_unit_hash": "2c05125888aecc6ae72328128e350ab83ef117affd4a3d4d3cf3d28af226ff0c", + "prepared_review_unit_hash": "c205bbb24062553b876015592e5dca510175526c0c1b7562ba861f4cf91a1b7c", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37197, - "estimated_input_tokens": 9300, - "estimated_eval_tokens": 18856 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36325, + "estimated_input_tokens": 9082, + "estimated_eval_tokens": 18420 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous v5/model-loading regressions. Only one duplicate pair is high-confidence; the rest look like adjacent but distinct bugs. No PRs are present.", - "confidence": 0.63, - "canonical_issue_reason": "issue:45216 is the clearest representative duplicate: it directly describes the Qwen3.5 save_pretrained regression, and issue:45357 is the same failure resurfacing later with the incorrect visual-encoder keys wording.", + "summary": "The cluster is mostly unrelated bug reports with a few near-duplicates. The Qwen3.5 `save_pretrained` reports (45216/45357) are the strongest duplicate pair; the GGUF reports (43452/43482) are too weakly connected from titles alone to call duplicates confidently.", + "confidence": 0.66, + "canonical_issue_reason": "Issue 45216 is the best canonical issue because it describes the broader Qwen3.5 `save_pretrained` regression, while 45357 is a narrower follow-up about the same broken serialization path.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45216 is the best global issue candidate because it is the most concrete and clearly duplicated report among the set, with a later issue pointing to the same underlying save_pretrained bug.", + "best_issue_reason": "45216 is the strongest anchor for the cluster: it states the general failure mode more broadly than the follow-up report and best represents the shared regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43425", "right": "issue:45161", "accept": false, - "reason": "Both are broad runtime/model-compatibility complaints, but one is about Torch 2.10 incompatibility and the other about TP-only behavior in GPT-OSS MoE; different bugs." + "reason": "Different problems: Torch version incompatibility vs MoE tensor-parallel loading failure." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Different failure modes: loading models with base_model_prefix vs filling non-persistent buffers with junk." + "reason": "Both are v5 regressions, but one is model loading with `base_model_prefix` and the other is non-persistent buffer serialization; not the same bug." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Both involve Whisper, but one is batch_decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens; not the same bug." + "reason": "Both involve Whisper special tokens, but one is `batch_decode()` behavior and the other is empty-transcription generation after `align_special_tokens`; too different to treat as duplicates." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Different models and different load failures; only a broad \"cannot load model\" theme is shared." + "reason": "Separate model-loading failures for different checkpoints; same broad subsystem only." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Both are audio-feature related, but one is an unsupported center argument crash and the other is a generic _torch_extract_fbank_features issue; insufficient evidence they are the same bug." + "reason": "Both are audio-feature related, but the titles point to different failure points (`LasrFeatureExtractor` center arg vs `_torch_extract_fbank_features()`), so the duplication is not clear enough." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Image preprocessor/model dtype mismatch vs template compilation TypeError; unrelated code paths." + "reason": "Unrelated: image preprocessor/model dtype mismatch vs template compilation error." }, { "left": "issue:42947", "right": "issue:44756", "accept": false, - "reason": "Gradient checkpointing with PEFT LoRA is unrelated to disabling mmap for Strix Halo OOM." + "reason": "Different bugs: ineffective gradient checkpointing vs mmap-induced OOM on Strix Halo." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "Trainer token accounting not updating is unrelated to Qwen2.5-VL iterable dataset failures with multiple images." + "reason": "Trainer token-count tracking and multi-image iterable finetuning are unrelated." }, { "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Different bugs: missing all_tied_weights_keys on Molmo vs tokenizer loading failure for EMBEDDIA/est-roberta." + "reason": "Different models and different loading failures; no concrete shared code-path indicated." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Both are v5 regressions, but one is about non-persistent buffers and the other about custom model initialization; not the same concrete failure." + "reason": "Both are v5 regressions around model handling, but one is junk buffers and the other is custom model initialization; not the same defect." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "AutoConfig missing fields with trust_remote_code is unrelated to a PagedAttentionCache linear_attention crash." + "reason": "Completely different areas: remote-code config fields vs Qwen3.5 paged-attention group parsing." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading complaints, but they affect different models and different failure paths." + "reason": "Both are model load failures, but for different models and likely different root causes." }, { "left": "issue:43452", "right": "issue:43482", "accept": false, - "reason": "The first is a generic gguf_file/from_pretrained regression; the second is a specific Qwen2.5-GGUF load failure. Related area, but not clearly the same underlying bug." + "reason": "The titles both mention GGUF loading, but the packet does not establish the same regression or concrete code-path failure, so this is too weak to merge as duplicates." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Different failures: torch dynamo grouped_mm tracing vs template-node compilation." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "T5 tokenizer special-token method conflict is unrelated to Qwen3Moe FineGrainedFP8Config failure." + "reason": "Tokenizer method-name conflict vs Qwen3Moe FP8 config failure; unrelated." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "device_map/offload dequantization failure is a different loading path from GGUF model loading." - }, - { - "left": "issue:43541", - "right": "issue:45084", - "accept": false, - "reason": "Torch dynamo grouped_mm tracing error is unrelated to the template-node compilation error." + "reason": "Different loading paths: device-map/offload dequantized model loading vs GGUF model loading." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "FP8 config failure and PagedAttentionCache linear_attention crash are different bugs." + "reason": "Both mention Qwen MoE models, but the actual bugs are different (FP8 config vs invalid linear_attention group type)." }, { "left": "issue:39404", "right": "issue:43825", "accept": false, - "reason": "Whisper return_language pipeline regression is not the same as an error-message wording issue about translation tasks." + "reason": "Whisper `return_language` regression and generic pipeline translation-task messaging are not the same concrete bug." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "Untied lm_head weights in Mistral3ForConditionalGeneration is distinct from device-dependent serialization of lm_head.weight." + "reason": "Both involve tied weights, but one is a missing tie in Mistral3 and the other is device-dependent serialization behavior." }, { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Qwen3.5 output_hidden_states bug is unrelated to Qwen2.5-VL rope index scaling for still images." + "reason": "Different Qwen-family bugs on different models/code paths." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "T5 tokenizer add_special_tokens conflict is unrelated to custom model initialization in notebooks." + "reason": "Tokenizer method conflict vs custom model initialization breakage; unrelated." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Different model/tokenizer loading failures with no concrete shared code-path evidence." + "reason": "Separate model-loading failures; insufficient evidence of a shared underlying defect." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio processor sampling-rate default is unrelated to MIMI batched-vs-single padding-mask behavior." + "reason": "Audio chat-template sampling-rate default vs MIMI batching/padding-mask outputs; different bugs." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Template compilation error and GraniteMoEHybrid invalid method are different issues." + "reason": "Template compilation error vs GraniteMoEHybrid invalid method call; unrelated." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization breakage is not the same as non-persistent buffer junk filling." + "reason": "Both are Transformers v5 issues, but one is custom model initialization and the other is non-persistent buffer corruption; not clearly the same bug." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Siglip2 nonsensical outputs and OmDet-Turbo input-size mismatch are separate preprocessing/model alignment bugs." + "reason": "Both are vision-related mismatches, but one is nonsensical model outputs and the other is a concrete processor size mismatch." }, { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Same Qwen3.5 save_pretrained regression: the later issue is just a newer wording/version of the same incorrect checkpoint-saving bug." + "reason": "Same Qwen3.5 `save_pretrained` serialization regression; 45357 is a narrower report of the same broken checkpoint save behavior." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Llava Next segfault and Qwen3OmniMoe video-input ValueError are different model-loading/processing failures." + "reason": "Different models and failures: Llava Next segfault vs Qwen3OmniMoe video-processing ValueError." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Generic loading-noise complaint vs a specific Mistral tokenizer regex patch crash; not the same bug." + "reason": "Generic loading noise vs Mistral regex patch crash; not the same bug." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT loading failure and Mistral tokenizer patch crash affect different components and failure modes." - }, - { - "left": "issue:43377", - "right": "issue:43819", - "accept": false, - "reason": "MIMI missing padding-mask support and DAC.from_latents missing STE mismatch are distinct audio-model correctness bugs." + "reason": "Different model-loading failure and tokenizer-patch crash; no shared concrete issue." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the one accepted soft pair is a conservative duplicate call based on closely aligned Qwen3.5 save_pretrained titles. The rest of the verdicts appropriately stay negative or cautious, and there are no PRs to reconcile." + "accept": false, + "feedback": "The summary is mostly grounded, but one soft-edge accept looks too strong for the evidence available. The Qwen3.5 `save_pretrained` pair (45216/45357) is well supported, but the GGUF pair (43452/43482) is only linked by broad title similarity and a shared subsystem; the packet does not establish the same regression, so accepting it overstates the evidence. The rest of the rejects are conservatively phrased, and there are no PRs." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2644,12 +2624,11 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ - "issue:33666", "issue:36032", "issue:40444", "issue:40990", @@ -2659,6 +2638,7 @@ "issue:42915", "issue:43296", "issue:43317", + "issue:43377", "issue:43404", "issue:43421", "issue:43425", @@ -2668,8 +2648,8 @@ "issue:43611", "issue:43632", "issue:43742", - "issue:43792", "issue:43818", + "issue:43819", "issue:43856", "issue:44062", "issue:44220", @@ -2700,10 +2680,11 @@ ], "soft_pairs": [ "issue:41553|issue:43582", + "issue:43377|issue:43819", "issue:44220|issue:44683", - "issue:44265|issue:44610", "issue:43632|issue:44756", "issue:43404|issue:45127", + "issue:44265|issue:44610", "issue:42915|issue:45084", "issue:44987|issue:45081", "issue:43296|issue:43531", @@ -2713,10 +2694,10 @@ "issue:40990|issue:44811", "issue:43611|issue:44534", "issue:36032|issue:43742", - "issue:42175|issue:45542", "issue:43317|issue:43856", - "issue:40444|issue:45325", + "issue:42175|issue:45542", "issue:43632|issue:44534", + "issue:40444|issue:45325", "issue:43531|issue:44877", "issue:42915|issue:44589", "issue:43421|issue:44568", @@ -2726,228 +2707,217 @@ "issue:44683|issue:45020", "issue:43482|issue:45538", "issue:44829|issue:45479", - "issue:33666|issue:40444", - "issue:44403|issue:44991", - "issue:43792|issue:44589" + "issue:44403|issue:44991" ], - "prepared_review_unit_hash": "dac511e84c3cb8a621e42c3a70f397f104d724e19281a6d9de483739f648bcd2", + "prepared_review_unit_hash": "0dfae2517b3208f1f971e1b4fb26f2c7085ba959a420ec3aafa9bc6d2c1b857d", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36662, - "estimated_input_tokens": 9166, - "estimated_eval_tokens": 18588 + "soft_pair_count": 29, + "serialized_chars": 36489, + "estimated_input_tokens": 9123, + "estimated_eval_tokens": 18502 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly heterogeneous. I only see two likely duplicate pairs: the TensorFlow backend packaging bug (42175/45542) and the Qwen2.5-VL video/temporal-position regression (44479/45325). The rest look like similarity-based false positives rather than the same underlying issue.", - "confidence": 0.69, - "canonical_issue_reason": "No single canonical issue across the full set; the items split into unrelated topics, with only two small duplicate-like subgroups.", + "summary": "This cluster is mostly a grab-bag of unrelated issues, with a few loosely similar transformer/tokenizer regressions but no clear duplicate pair. I would not merge any of the soft-edge candidates into one bug/change.", + "confidence": 0.91, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No global best issue for the whole set; 42175 best represents the TensorFlow packaging subgroup and 44479 best represents the Qwen video regression subgroup, but neither covers the cluster as a whole.", + "best_issue_reason": "No single issue cleanly represents the cluster; the reports span many distinct code paths and model families.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "Different bugs: tokenizer loading error vs Apple Silicon allocator TypeError." + "reason": "Different failures in different subsystems: tokenizer loading error vs Apple Silicon allocator TypeError." }, { - "left": "issue:44220", - "right": "issue:44683", + "left": "issue:43377", + "right": "issue:43819", "accept": false, - "reason": "Unrelated code paths: fbank extraction vs compiled flex attention on newer torch." + "reason": "Both are model correctness bugs, but they affect unrelated components and code paths (MIMI padding mask vs DAC STE behavior)." }, { - "left": "issue:44265", - "right": "issue:44610", + "left": "issue:44220", + "right": "issue:44683", "accept": false, - "reason": "Different problems: export failure with torch_compilable_check vs processor/model input-size mismatch." + "reason": "Audio feature extraction bug vs compiled flex_attention regression; no shared underlying fix." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "Unrelated: `_is_hf_initialized` loading regression vs mmap/OOM on Strix Halo." + "reason": "Separate regressions: `_is_hf_initialized`/v5 init behavior vs mmap OOM on Strix Halo." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Both involve models, but the concrete bugs differ: lm_head tying vs LoRA merge collapse." + "reason": "Both involve weight tying behavior, but one is missing tying in Mistral3 while the other is LoRA merge collapse with extended vocab; not the same bug." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to the OmDet-Turbo processor image-size mismatch." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "Different failure modes and code paths: FP8 config handling vs template compilation error." + "reason": "Qwen3Moe FP8 config issue and a template-node compilation error are unrelated." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Both are load-time regressions, but one is remote-code model loading and the other is a Mistral regex patch crash." + "reason": "Both are tokenizer/model loading problems, but they target different models and distinct code paths." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "Different subsystems: PaddleOCR-VL load in vLLM vs Qwen3-MoE sliding-window behavior." + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to Qwen3-MoE sliding-window behavior." }, { "left": "issue:44479", "right": "issue:45325", - "accept": true, - "reason": "Same Qwen2.5-VL video/temporal-position regression in 5.3.0; 45325 looks like a narrower manifestation of the broader breakage in 44479." + "accept": false, + "reason": "Both concern Qwen VL temporal handling, but one is a broad video-input regression and the other is a still-image position-id scaling bug." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "Different tokenizer/model-loading failures with no shared underlying bug." + "reason": "AddedToken keyword conflict is unrelated to loading a specific model repo with newer transformers." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Different regressions: FineGrainedFP8Config vs unexpected `_is_hf_initialized` argument." + "reason": "FP8 model config failure and `_is_hf_initialized` init_empty_weights error are different regression classes." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "Perplexity anomaly and Whisper decode behavior are unrelated." + "reason": "Perplexity evaluation issue vs Whisper batch_decode special-token handling; no shared code path." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "Different v5 regressions: `base_model_prefix` loading vs non-persistent buffer initialization." + "reason": "Base-model-prefix loading regression is unrelated to non-persistent buffer initialization." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "Both are loading issues, but they describe different tokenizer/model failures." - }, - { - "left": "issue:42175", - "right": "issue:45542", - "accept": true, - "reason": "Both report the same packaging problem: TensorFlow backend is missing/misinstalled when using the torch-focused install path." + "reason": "Tokenizer method-name conflict vs MobileLLM key error; no same underlying bug." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "Different problems: device-map offload load failure vs MoE training memory usage." + "reason": "Dequantized model offload failure and Qwen3 MoE training memory usage are separate problems." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:42175", + "right": "issue:45542", "accept": false, - "reason": "Both are Qwen2.5-VL-related, but multiple images per prompt is not the same bug as the temporal position-id regression." + "reason": "Both mention TensorFlow packaging, but one is missing TF in the torch extra and the other is an undefined backend when only tensorboard is installed; not the same concrete bug." }, { "left": "issue:43632", "right": "issue:44534", "accept": false, - "reason": "Different v5 issues affecting separate initialization paths." + "reason": "Different transformer v5 regressions affecting different initialization/storage behaviors." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "IterableDataset multi-image prompt failure is unrelated to Qwen2.5-VL still-image temporal position scaling." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Unrelated model/config bugs: Qwen3-MoE sliding_window vs granite_speech strict config loading." + "reason": "Qwen3-MoE sliding_window bug and granite_speech config strictness are unrelated." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "Different failures: FP8 config handling vs missing Float8 storage type." + "reason": "FP8 config failure vs missing Float8 storage type; related to float8 broadly but not the same bug." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Both involve special tokens, but one is runtime backend/post-processor sync and the other is a tokenizer-specific BOS/EOS regression." + "reason": "Both involve special tokens, but one is a backend update feature request and the other is a tokenizer regression; too different to be the same change." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Different multimodal models and different bugs: Video-LLaVA weights/temporal attention vs Qwen3.5 save_pretrained key mapping." + "reason": "Video-LLaVA tower/temporal-attention problem is unrelated to Qwen3.5 visual-encoder save_pretrained key regression." }, { "left": "issue:36032", "right": "issue:41720", "accept": false, - "reason": "Tokenizer load conflict vs Qwen3 auto-device-map CUDA assert are unrelated." + "reason": "Tokenizer name conflict vs CUDA assert during device-mapped Qwen3 loading; unrelated." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Torch version compatibility and log-noise loading behavior are not the same bug." + "reason": "Torch version incompatibility and loading noise are not the same bug." }, { "left": "issue:44683", "right": "issue:45020", "accept": false, - "reason": "Different scopes: torch flex_attention failure vs remote_code loading regressions." + "reason": "Flex attention compilation regression is unrelated to remote_code loading breakage." }, { "left": "issue:43482", "right": "issue:45538", "accept": false, - "reason": "Unrelated issues: GGUF loading under v5 vs CLIPTokenizer max_length constant." + "reason": "Qwen2.5-GGUF/v5 loading issue and CLIPTokenizer model_max_length are unrelated." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Both affect classification training, but the concrete symptoms and causes differ: flash_attention_3 degenerate training vs num_labels=1 zero-loss bug." - }, - { - "left": "issue:33666", - "right": "issue:40444", - "accept": false, - "reason": "Different Qwen2-VL training scenarios; no shared underlying defect is evident." + "reason": "Both mention degenerate classification behavior, but one is tied to flash_attention_3 while the other is a num_labels=1/problem_type loss issue." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "General loading noise vs tokenizer loading failure for a specific model are unrelated." - }, - { - "left": "issue:43792", - "right": "issue:44589", - "accept": false, - "reason": "Whisper runtime failure and Float8 storage lookup error are different bugs." + "reason": "Generic loading noise is unrelated to the est-roberta tokenizer loading failure." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the two accepted soft-edge pairs are conservative and supported by the titles: the TensorFlow backend packaging issues align closely, and the Qwen2.5-VL temporal/video-position issues are plausibly the same regression family. The remaining rejections are appropriately cautious and based on distinct bug descriptions." - }, - "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 }, @@ -2959,11 +2929,12 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ + "issue:33666", "issue:35707", "issue:40444", "issue:40990", @@ -2972,7 +2943,6 @@ "issue:42915", "issue:43262", "issue:43296", - "issue:43317", "issue:43377", "issue:43454", "issue:43482", @@ -2988,7 +2958,6 @@ "issue:43749", "issue:43792", "issue:43828", - "issue:43873", "issue:43881", "issue:43883", "issue:43957", @@ -3014,21 +2983,23 @@ "issue:45584" ], "soft_pairs": [ + "issue:33666|issue:40444", + "issue:43792|issue:44589", "issue:44079|issue:44534", "issue:35707|issue:40444", "issue:43377|issue:45491", "issue:43482|issue:43525", "issue:44291|issue:44589", "issue:43749|issue:44079", - "issue:44079|issue:45072", "issue:43644|issue:44079", + "issue:44079|issue:45072", "issue:40444|issue:45381", "issue:43632|issue:45072", - "issue:44265|issue:45072", "issue:43881|issue:44488", + "issue:44265|issue:45072", "issue:43531|issue:44291", - "issue:45325|issue:45381", "issue:41720|issue:45081", + "issue:45325|issue:45381", "issue:43611|issue:43646", "issue:42175|issue:43296", "issue:43454|issue:44466", @@ -3042,217 +3013,222 @@ "issue:43262|issue:43746", "issue:43632|issue:45563", "issue:43881|issue:44991", - "issue:43317|issue:43873", "issue:40990|issue:45584" ], - "prepared_review_unit_hash": "765a9db262bface973899821bcf6ac8b5d5635c7f092969d7ca224cf46b2e0d7", + "prepared_review_unit_hash": "cfea4018848479f06d5a136282f5ab5742ef8e6ff1a2d13a933374cda3a11ede", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36713, - "estimated_input_tokens": 9179, - "estimated_eval_tokens": 18614 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36409, + "estimated_input_tokens": 9103, + "estimated_eval_tokens": 18462 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of mostly unrelated Transformers regression reports, with a broad theme around v5-era loading/init/serialization breakages and a few multimodal or quantization-specific bugs. I did not find any soft pair that clearly describes the same underlying fix.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 43611 is the best representative of the cluster\u2019s dominant theme: broad Transformers v5 model-loading regressions. It is clearer and more central than the highly model-specific reports.", + "summary": "Mostly a heterogeneous set of unrelated issue reports, with only one clear duplicate-style pair around the v5 non-persistent-buffer regression. The rest share broad Transformer/v5 themes at best, but not the same underlying bug.", + "confidence": 0.57, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43611 is the most suitable cluster representative because it captures the recurring v5 loading/init breakage pattern seen across several nearby issues, while staying relatively general.", + "best_issue_reason": "issue:43644 is the strongest representative of the set: it states a concrete, version-specific regression and also matches the near-identical follow-up issue:44534.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Different models and failure modes; both are training/finetuning issues, not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Both are loading/runtime failures, but the errors and model contexts are different." + }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "Different bugs: ModelOutput key handling vs non-persistent buffer serialization." + "reason": "Both are v5 regressions, but one is ModelOutput key handling and the other is buffer initialization." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Unrelated failures: generation with inputs_embeds/past_key_values vs Qwen2.5-VL finetuning on iterable multi-image prompts." + "reason": "Progressive generation with caches is unrelated to multi-image iterable finetuning." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both involve batching/padding, but they affect different models and different code paths; not the same bug." + "reason": "Both involve batching/padding, but they are distinct model-specific bugs with different symptoms." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "Separate loading issues: GGUF compatibility vs missing pad_token_id on Llama4Config." + "reason": "Separate model-loading failures with different root causes." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "Both are v5 initialization/loading regressions, but one is an unexpected init arg and the other is a float8 storage lookup failure." + "reason": "Init-empty-weights argument error is unrelated to the Float8 storage lookup failure." }, { "left": "issue:43749", "right": "issue:44079", "accept": false, - "reason": "FSDP CPU RAM efficient loading and ModelOutput key assignment are unrelated." + "reason": "FSDP CPU RAM-efficient loading is a different path than ModelOutput key assignment." }, { - "left": "issue:44079", - "right": "issue:45072", + "left": "issue:43644", + "right": "issue:44079", "accept": false, - "reason": "Different problems: output container key assignment vs bfloat16 dtype mismatch in inference." + "reason": "Non-persistent buffer junk and ModelOutput key handling are unrelated regressions." }, { - "left": "issue:43644", - "right": "issue:44079", + "left": "issue:44079", + "right": "issue:45072", "accept": false, - "reason": "Non-persistent buffer junk fills and ModelOutput key handling are distinct bugs." + "reason": "Different subsystems: output container semantics vs bfloat16 dtype mismatch." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Same model family, but different symptoms and modalities: multi-image finetuning vs video position-id bug." + "reason": "Different Qwen2.5-VL multimodal bugs, one training data handling and one video position ids." }, { "left": "issue:43632", "right": "issue:45072", "accept": false, - "reason": "Unrelated regressions: _is_hf_initialized flag behavior vs bfloat16 dtype mismatch." + "reason": "_is_hf_initialized/init-loading regression is unrelated to dtype mismatches in inference." }, { - "left": "issue:44265", - "right": "issue:45072", + "left": "issue:43881", + "right": "issue:44488", "accept": false, - "reason": "torch.export/torch_compilable_check failure is unrelated to the dtype mismatch issue." + "reason": "Two model-loading failures, but no evidence they share the same underlying cause." }, { - "left": "issue:43881", - "right": "issue:44488", + "left": "issue:44265", + "right": "issue:45072", "accept": false, - "reason": "Both are model loading failures, but for different models and likely different root causes." + "reason": "torch.export/torch_compilable_check is unrelated to bfloat16 inference dtype issues." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights argument handling are not the same bug." + "reason": "Qwen3-MoE sliding-window behavior is unrelated to init_empty_weights argument handling." }, { - "left": "issue:45325", - "right": "issue:45381", + "left": "issue:41720", + "right": "issue:45081", "accept": false, - "reason": "Closely related Qwen2.5-VL position-id issues, but one is still-image temporal scaling and the other is video vision_position_ids; not clearly one fix." + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to Mistral tokenizer patching." }, { - "left": "issue:41720", - "right": "issue:45081", + "left": "issue:45325", + "right": "issue:45381", "accept": false, - "reason": "Completely unrelated: CUDA assert in Qwen3 auto device mapping vs Mistral tokenizer regex patch crash." + "reason": "Both are Qwen2.5-VL position-id bugs, but one is still-image temporal scaling and the other is video vision positions; not clearly the same defect." }, { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both are v5 model initialization regressions, but the specific failures differ: base_model_prefix loading vs custom model init." + "reason": "Both concern v5 custom-model compatibility, but the titles indicate different failure surfaces and not the same concrete bug." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Backend packaging issue vs a vLLM model-loading failure; no shared bug." + "reason": "Packaging/backend selection problem versus a model-load failure." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "Both mention tied lm_head weights, but one is a missing tie causing bad generation and the other is device-dependent serialization inconsistency." + "reason": "Tied-weight generation failure and serialization inconsistency are related thematically, but not the same underlying bug." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Different vision-model behavior issues with different architectures and failure modes." + "reason": "Different vision models and different positional/interpolation issues." }, { "left": "issue:43828", "right": "issue:44683", "accept": false, - "reason": "Autocast dtype mismatch and compiled flex_attention failure are unrelated." + "reason": "Autocast dtype mismatch and compiled flex_attention failure are separate issues." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Quantization-related CUDA reserved-memory regression is not the same as a model-specific OOM report." + "reason": "Both mention OOM, but one is an int4 quantization memory regression and the other is a model-specific runtime OOM." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Different problems: Qwen3MoE with FineGrainedFP8Config vs Qwen2.5-GGUF loading on v5." + "reason": "Different model families and unrelated loading failures." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to Molmo tied-weights attribute failure." + "reason": "Image-processor label reduction bug versus missing tied-weights attribute." }, { "left": "issue:43792", "right": "issue:43957", "accept": false, - "reason": "Whisper runtime failure and meta-device model-loading breakage are different issues." + "reason": "Generic run/load failure versus meta-device loading regression; not the same bug." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "Checkpoint key mismatch is unrelated to Qwen2.5-VL rope/position-id scaling." + "reason": "Checkpoint-resume key mismatch is unrelated to Qwen2.5-VL position-id scaling." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sample-rate defaulting and GraniteSpeech PEFT checkpoint loading are separate bugs." + "reason": "Audio chat-template sample-rate defaulting and GraniteSpeech PEFT checkpoint loading are unrelated." }, { "left": "issue:43632", "right": "issue:45563", "accept": false, - "reason": "_is_hf_initialized regression and stale num_return_sequences warning are unrelated." + "reason": "_is_hf_initialized regression is unrelated to a stale warning in paged generate()." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "Different loading failures on different models; no shared concrete code-path." - }, - { - "left": "issue:43317", - "right": "issue:43873", - "accept": false, - "reason": "Both involve offloading/quantization, but one is a dequantized-model load failure on gpu+cpu offload and the other is a broader offloading complaint." + "reason": "Different model-loading issues for different models and tokenizers." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Completely unrelated: high perplexity on WikiText-2 vs Whisper empty-transcription generation failure." + "reason": "Perplexity anomaly and Whisper empty-transcription generation failure are unrelated." } ] }, @@ -3271,10 +3247,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:34634", "issue:35707", @@ -3326,13 +3302,14 @@ "issue:45406" ], "soft_pairs": [ - "issue:43606|issue:44265", + "issue:43317|issue:43873", "issue:34634|issue:35707", "issue:44291|issue:44877", + "issue:43606|issue:44265", "issue:44186|issue:44265", "issue:43606|issue:43819", - "issue:44610|issue:45325", "issue:35707|issue:41720", + "issue:44610|issue:45325", "issue:43454|issue:43550", "issue:43377|issue:44898", "issue:43454|issue:45072", @@ -3354,32 +3331,31 @@ "issue:43577|issue:43708", "issue:42617|issue:42915", "issue:44877|issue:44991", - "issue:43550|issue:43720", - "issue:44610|issue:44898", "issue:44479|issue:44918", - "issue:44186|issue:44610", + "issue:44610|issue:44898", + "issue:43550|issue:43720", "issue:36331|issue:42915", + "issue:44186|issue:44610", "issue:36331|issue:42617", "issue:43316|issue:44913", "issue:44671|issue:44743", "issue:42915|issue:43317", - "issue:43550|issue:43606", - "issue:41720|issue:42617" + "issue:43550|issue:43606" ], - "prepared_review_unit_hash": "d603dc8e93b51392bc00dc3dcf2c9cbe3e273de007545e63a8e963cb6b7a8d04", + "prepared_review_unit_hash": "d0542c4053d39ed37e52d81f28902ed72d8757045509a2b332feda2af72eb6f1", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 39, - "serialized_chars": 39474, + "serialized_chars": 39476, "estimated_input_tokens": 9869, "estimated_eval_tokens": 19994 }, @@ -3388,246 +3364,246 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is very heterogeneous: most items are model- or feature-specific bugs, and the soft links are mostly superficial (shared subsystems like loading, offload, generation, or multimodal support) rather than true duplicates. No PRs are present.", - "confidence": 0.23, - "canonical_issue_reason": "Issue 43606 is the closest thing to a hub: it is a concrete loading/offload device-mismatch bug and sits near several other offload/quantization-related reports, but the overall cluster is still too mixed for a strong canonical choice.", + "summary": "This cluster is very heterogeneous: most items are distinct model-specific loading, training, or inference bugs, and none of the soft pairs look like true duplicates. The best available anchor is a broad loading/config issue, but only as a loose representative, not a merged duplicate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:44877 is the broadest loading/configuration bug in the set and has the most nearby soft links, so it is the least specific representative of this heterogeneous cluster.", "canonical_pr_reason": null, - "best_issue_reason": "43606 is the most representative issue for a duplicate cluster because it describes a specific, reproducible runtime failure in model loading/offload, which is the most recurring concrete failure mode among the soft-linked items.", + "best_issue_reason": "issue:44877 is the best cluster anchor for triage because it is broad and central among the loading/configuration failures, though it is not a strong duplicate canon.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43606", - "right": "issue:44265", + "left": "issue:43317", + "right": "issue:43873", "accept": false, - "reason": "Different bugs: CPU offload device mismatch vs torch.export failure with torch_compilable_check." + "reason": "Both involve offloading/quantization, but one is a specific dequantized-model load failure and the other is a general offloading regression; not the same bug." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "Unrelated models/features: Bark voice_preset handling vs inputs_embeds/past_key_values generation." + "reason": "BarkProcessor voice preset and progressive generation with inputs_embeds/past_key_values are unrelated code paths." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "Both are loading-related, but one is init_empty_weights/_is_hf_initialized and the other is strict config validation; different root causes." + "reason": "One is init_empty_weights/_is_hf_initialized argument handling; the other is strict config loading for granite_speech. Different failures." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch in Bark-small is unrelated to torch.export failing on torch_compilable_check." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer crash on NER/padding differs from export-time failure in model compilation checks." + "reason": "Tokenizer crashes on NER/padding are unrelated to export-time compilation failures." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "Both mention model execution issues, but the failures are distinct: offload device mismatch vs missing STE in DAC.from_latents." + "reason": "Device mismatch during offload and DAC.from_latents mismatch with missing STE are different bugs." }, { - "left": "issue:44610", - "right": "issue:45325", + "left": "issue:35707", + "right": "issue:41720", "accept": false, - "reason": "Different image/video preprocessing bugs on different code paths and models." + "reason": "Progressive generation with embeddings/cache and Qwen3 auto device-map cudaErrorAssert are different problems." }, { - "left": "issue:35707", - "right": "issue:41720", + "left": "issue:44610", + "right": "issue:45325", "accept": false, - "reason": "Generation cache/input embedding bug is unrelated to auto device-map cuda assertion." + "reason": "Processor output size mismatch and Qwen2.5-VL rope index temporal scaling are distinct multimodal preprocessing bugs." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "Both involve model inference behavior, but one is tied lm_head weights and the other is torch.compile with SDPA; not the same defect." + "reason": "lm_head weight tying in AyaVision and Bamba torch.compile SDPA failure are not the same issue." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "MIMI padding-mask batching bug is unrelated to Perceiver non-default resolution failures." + "reason": "MIMI batching/padding-mask inconsistency and Perceiver non-default resolution failure are different model behaviors." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "Different dtype/serialization concerns: tied lm_head weights vs bfloat16 dtype mismatches in unrelated models." + "reason": "Weight tying bug and bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel are unrelated." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI encoder batching issue and OmDet-Turbo image-size mismatch are different model-specific preprocessing bugs." + "reason": "MIMI batched-vs-single output mismatch and OmDet-Turbo processor/model size mismatch are different subsystems." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "Documentation example errors for SigLIP2 are not the same as Qwen2.5-VL iterable dataset finetuning failures." + "reason": "Documentation example errors for SigLIP2 and Qwen2.5-VL multi-image IterableDataset failure are separate issues." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Both concern saving/loading state, but one is lm_head serialization inconsistency and the other is incorrect visual encoder keys." + "reason": "Serialization of tied lm_head weights depends on device, while the other is incorrect visual encoder key saving for Qwen3.5." }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "FSDP CPU RAM-efficient loading is a loader path issue; non-persistent buffer junk is a separate v5 buffer initialization bug." + "reason": "FSDP CPU RAM-efficient loading and non-persistent buffer initialization junk are different failures." }, { "left": "issue:43792", "right": "issue:44683", "accept": false, - "reason": "Whisper model runtime failure and compiled flex_attention on torch>=2.9 are unrelated failures." + "reason": "Whisper-large-v2 runtime failure and compiled flex_attention on torch>=2.9 are unrelated." }, { "left": "issue:43296", "right": "issue:44991", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM and tokenizer load failure for a different model are distinct problems." + "reason": "PaddleOCR-VL loading in vLLM and tokenizer loading for EMBEDDIA/est-roberta are different model/load paths." }, { "left": "issue:43296", "right": "issue:44877", "accept": false, - "reason": "Model load failure vs strict config rejection; no shared underlying bug." + "reason": "vLLM PaddleOCR-VL load failure and granite_speech strict config loading are both load-related but not the same concrete bug." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "Different code paths: tied-weight serialization vs packed-weight unpacking during accelerate loading." + "reason": "AyaVision lm_head tying and BitNet packed-weight unpacking during accelerate loading are different code paths." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Auto device mapping cuda assert and Llava Next segmentation fault are different runtime failures." + "reason": "Qwen3 auto device mapping cudaErrorAssert and Llava Next segmentation fault are separate model-loading failures." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "Weight tying in Mistral3 is unrelated to Qwen3.5 visual encoder key serialization." + "reason": "Mistral3 lm_head tying and Qwen3.5 visual encoder save_pretrained regression are unrelated serialization/tie bugs." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "Auto device-map cuda assert and Qwen3.5 flash-attention generation regression are different bugs." + "reason": "Qwen3 auto device mapping crash and Qwen3.5 flash-attention generation failure are distinct regressions." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BeitImageProcessorFast label reduction bug is unrelated to glm-4v-9b loading failure." + "reason": "BEiT Fast label reduction bug and glm-4v-9b loading failure are unrelated." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Continuous-batching multimodal crash and Gemma4Processor missing _tokenizer are different serve-time errors." + "reason": "Continuous-batching multimodal crash due to .to() on strings and Gemma4Processor missing _tokenizer are different serve bugs." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Tokenizer padding crash and bfloat16 dtype mismatch are unrelated." + "reason": "NER/padding tokenizer crash and dtype mismatch in inference are unrelated." }, { "left": "issue:43873", "right": "issue:44756", "accept": false, - "reason": "Quantization/offloading behavior problem is not the same as disabling mmap to avoid OOM." + "reason": "General offloading/quantization problem and Strix Halo mmap OOM mitigation are not the same bug." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Blip2 dtype not propagated and Trainer max_steps resume math are unrelated training/loading bugs." + "reason": "Blip2 dtype propagation and Trainer resume_from_checkpoint max_steps calculation are unrelated." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "3d_parallel.py failure and Qwen3Moe FineGrainedFP8Config failure are different model/runtime issues." + "reason": "3d_parallel.py startup failure and Qwen3Moe FineGrainedFP8Config failure are different issues." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config load rejection differs from tokenizer loading failure." + "reason": "Strict config blocking granite_speech and tokenizer loading failure for est-roberta are different load-time problems." }, { - "left": "issue:43550", - "right": "issue:43720", + "left": "issue:44479", + "right": "issue:44918", "accept": false, - "reason": "torch.compile+SDPA failure is not the same as packed-weight unpacking during accelerate loading." + "reason": "Video-input regression in Qwen models and TRL unpacking of Qwen3.5 input embeddings are different failures." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both involve image preprocessing, but the exact mismatch is different: processor output size vs model expected size." + "reason": "OmDet-Turbo processor/model resolution mismatch and Perceiver interpolation failure are separate vision preprocessing bugs." }, { - "left": "issue:44479", - "right": "issue:44918", + "left": "issue:43550", + "right": "issue:43720", "accept": false, - "reason": "Video-input regression and input-embedding unpacking in TRL are separate multimodal/training bugs." + "reason": "torch.compile+SDPA failure in Bamba and packed-weight unpacking failure in BitNet are not the same concrete bug." }, { - "left": "issue:44186", - "right": "issue:44610", + "left": "issue:36331", + "right": "issue:42915", "accept": false, - "reason": "Tokenizer crashes and OmDet-Turbo input-size mismatch do not share a common root cause." + "reason": "CustomTrainer compute_loss signature change and Qwen3Moe FP8 loading failure are unrelated." }, { - "left": "issue:36331", - "right": "issue:42915", + "left": "issue:44186", + "right": "issue:44610", "accept": false, - "reason": "CustomTrainer signature change and Qwen3Moe FP8 config failure are unrelated." + "reason": "Tokenizer padding/truncation crash and processor output-size mismatch are different model/preprocessor bugs." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Unexpected num_items_in_batch argument is unrelated to 3d_parallel.py execution failure." + "reason": "Trainer compute_loss keyword regression and 3d_parallel.py execution failure are unrelated." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Gemma3TextConfig API discrepancy and GPTNeoX rotary_pct reload bug are different config-serialization issues." + "reason": "Gemma3TextConfig API discrepancy and GPTNeoXConfig rotary_pct reload regression are different config issues." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "CamemBERT masked-LM prediction regression is unrelated to Qwen3.5 recurrent-state reset with cache." + "reason": "CamemBERT masked-LM regression and Qwen3.5 recurrent-state cache reset bug are unrelated model behaviors." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "FineGrainedFP8Config failure and gpu+cpu offload dequantized model loading failure are different bugs." + "reason": "Qwen3Moe FP8 loading failure and dequantized model offload failure are different problems." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Compile/SDPA failure and Bark CPU-offload device mismatch are not the same code-path problem." - }, - { - "left": "issue:41720", - "right": "issue:42617", - "accept": false, - "reason": "Auto device-map cuda assert and 3d_parallel.py failure are different runtime issues." + "reason": "Bamba torch.compile SDPA failure and Bark-small CPU-offload device mismatch are unrelated." } ] }, @@ -3646,10 +3622,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30064", "issue:33357", @@ -3702,6 +3678,7 @@ ], "soft_pairs": [ "issue:43531|issue:43572", + "issue:41720|issue:42617", "issue:44292|issue:44479", "issue:41720|issue:45161", "issue:43540|issue:43782", @@ -3715,237 +3692,237 @@ "issue:44514|issue:44545", "issue:33357|issue:35707", "issue:41720|issue:43746", - "issue:43606|issue:43720", "issue:30064|issue:34634", "issue:44492|issue:44509", + "issue:43606|issue:43720", "issue:43550|issue:43653", "issue:44368|issue:45357", "issue:44625|issue:45479", "issue:43421|issue:43927", "issue:43606|issue:43653", - "issue:34567|issue:35707", "issue:43782|issue:45381", + "issue:34567|issue:35707", "issue:34634|issue:41720", - "issue:44898|issue:45072", "issue:42886|issue:44843", + "issue:44898|issue:45072", "issue:43421|issue:43653", "issue:44451|issue:44991", "issue:44062|issue:44451" ], - "prepared_review_unit_hash": "c4b068ce52954c9a3804895b04e3015af7b4d9e5d5749da7f64dd684d2ec2a19", + "prepared_review_unit_hash": "45fb029a16afca766af9c5707ea2328dc3e33fe137caedf1988a31d4facd0a5d", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36736, - "estimated_input_tokens": 9184, - "estimated_eval_tokens": 18624 + "soft_pair_count": 31, + "serialized_chars": 37035, + "estimated_input_tokens": 9259, + "estimated_eval_tokens": 18774 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of unrelated Transformers bug reports. The only clear duplicate pair is the two identical Qwen2_5_VLProcessor.apply_chat_template crash reports (44514/44545).", - "confidence": 0.95, - "canonical_issue_reason": "Issue 44514 is the earliest report of the exact same batched apply_chat_template padding=False crash; 44545 has the same title/symptom and reads like a duplicate follow-up.", + "summary": "Mostly unrelated issue reports. The only clear duplicate pairs here are the batched `Qwen2_5_VLProcessor.apply_chat_template(..., padding=False)` crash and the HF_HUB_OFFLINE tokenizer-load regression; the rest look like different bugs in different code paths or model families.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 42886 is the cleanest canonical report for the offline-tokenizer regression: it states the user-facing failure, while 44843 provides the root-cause detail (`model_info()` being called in offline mode).", "canonical_pr_reason": null, - "best_issue_reason": "44514 is the best representative because it is the original, most direct report of the one clear duplicate bug in the cluster.", + "best_issue_reason": "Issue 42886 is the best representative issue in this set because it is a clear, broadly scoped bug report with a concrete failure mode and version/environment context.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Different bugs: Qwen3-MoE sliding_window behavior vs StableLmConfig pad_token_idx regression." + "reason": "Different models and different failures: Qwen3-MoE `sliding_window` vs StableLmConfig `pad_token_idx`." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Unrelated symptoms and subsystems; one is Qwen3 auto-device-map CUDA assert, the other is a 3d_parallel.py runtime failure." }, { "left": "issue:44292", "right": "issue:44479", "accept": false, - "reason": "Different failures: Qwen-3-8B-NVFP4 runtime error vs a broader video-input regression across Qwen VL/MoE models." + "reason": "Different bug classes: Qwen-3-8B-NVFP4 runtime error vs Qwen VL video-input regression." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Different code paths and symptoms: CUDA assert in Qwen3 auto device mapping vs TP-only failure for GPT-OSS MoE." + "reason": "Both involve model execution, but they are different concrete failures and code paths (CUDA assert vs TP-only MoE issue)." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Different underlying problems: video input ValueError in Qwen3OmniMoe vs from_pretrained weight_only loading error in Qwen3VL." + "reason": "Both are Qwen multimodal reports, but one is a video-processing ValueError and the other is a `from_pretrained` weight-loading error." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Unrelated functionality: gather_for_metrics label trimming vs SAM3 text/bbox processing odd behavior." + "reason": "Different pipelines: Trainer metric gathering/label dropping vs SAM3 text+box prompt behavior." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Both tokenizer-related, but different models and failures: LayoutLMv2 NER/padding crash vs MLuke AttributeError on tasks." + "reason": "Different tokenizer bugs in different models; no shared failing code path is evident." }, { "left": "issue:41720", "right": "issue:43873", "accept": false, - "reason": "One is a Qwen3 device-mapping CUDA assert; the other is a quantization/offloading behavior issue. Not the same bug." + "reason": "Qwen3 device-mapping CUDA assert is unrelated to quantization offloading behavior." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Distinct issues: FineGrainedFP8Config failure in Qwen3Moe vs Llava Next loading segfault." + "reason": "Different failures on different models: FineGrainedFP8Config vs Llava Next loading segfault." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to failing to run 3d_parallel.py." + "reason": "ImportError for `GenerationMixin` is unrelated to a 3d_parallel.py execution problem." }, { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Different surfaces: Gemma4Processor missing _tokenizer in transformers serve vs streaming chat/completions failure on Qwen3.5-0.8B." + "reason": "Different components and symptoms: Gemma4Processor server crash vs Qwen3.5 streaming inference API failure." }, { "left": "issue:36331", "right": "issue:41720", "accept": false, - "reason": "Trainer compute_loss signature regression is unrelated to Qwen3 auto device mapping CUDA asserts." + "reason": "Trainer API signature change vs model-device-mapping CUDA assert; not the same bug." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Identical title and same batched apply_chat_template crash with padding=False; 44545 is a duplicate of 44514." + "reason": "Same `Qwen2_5_VLProcessor.apply_chat_template(..., padding=False)` crash with effectively identical title and symptom." }, { "left": "issue:33357", "right": "issue:35707", "accept": false, - "reason": "Different problems: MacOS bus error with a CLIP model vs progressive generation with inputs_embeds/past_key_values." + "reason": "MacOS CLIP bus error and progressive generation with `inputs_embeds`/`past_key_values` are unrelated." }, { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Different models and failure modes: Qwen3 device mapping crash vs GraniteSpeech PEFT local-checkpoint loading issue." - }, - { - "left": "issue:43606", - "right": "issue:43720", - "accept": false, - "reason": "Both involve loading/offloading, but one is CPU offload device mismatch and the other is packed-weight unpacking during accelerate loading." + "reason": "Different loading/runtime failures: Qwen3 CUDA assert vs GraniteSpeech local checkpoint/PEFT loading." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Completely different areas: image processor segmentation maps vs BarkProcessor voice_preset handling." + "reason": "Image segmentation-map processing and BarkProcessor voice preset handling are unrelated." }, { "left": "issue:44492", "right": "issue:44509", "accept": false, - "reason": "A typo in cache strategies is not the same as stale docs mentioning removed pipeline tasks." + "reason": "A typo in cache strategies is not the same as outdated docs for removed pipeline tasks." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Different model-specific loading/offload failures with different root causes." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "Different subsystems: torch.compile/SDPA failure in Bamba vs BigBirdTokenizer special-token registration bug." + "reason": "torch.compile/SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token registration." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "One is a warning about tie_word_embeddings; the other is incorrect visual encoder keys saved in Qwen3.5." + "reason": "Warning about `tie_word_embeddings` during LoRA fine-tuning is not the same as incorrect visual-encoder keys in `save_pretrained`." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Config propagation for num_labels is not the same as zero-loss behavior for single_label_classification with num_labels=1." + "reason": "Both mention classification config, but one is Qwen3.5 config propagation and the other is a broader zero-loss training bug with `num_labels=1`." }, { "left": "issue:43421", "right": "issue:43927", "accept": false, - "reason": "Runtime special-token post-processor updates vs save/load loss of custom token IDs are related but not the same bug." + "reason": "Runtime special-token/post-processor update issue is different from config token-ID loss after save/load." }, { "left": "issue:43606", "right": "issue:43653", "accept": false, - "reason": "Device mismatch during CPU offload is unrelated to tokenizer mask-token registration." - }, - { - "left": "issue:34567", - "right": "issue:35707", - "accept": false, - "reason": "TrainerState token counting is unrelated to progressive generation with inputs_embeds/past_key_values." + "reason": "Different bug families: CPU offload device mismatch vs tokenizer decode/special-token registration." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Both Qwen VL-related, but one is weight_only loading and the other is a video-input position-id regression." + "reason": "Different Qwen VL issues: `weight_only=True` loading error vs video `vision_position_ids` regression." }, { - "left": "issue:34634", - "right": "issue:41720", + "left": "issue:34567", + "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice_preset bug does not match a Qwen3 CUDA device-mapping crash." + "reason": "Trainer state token-count tracking is unrelated to progressive generation cache behavior." }, { - "left": "issue:44898", - "right": "issue:45072", + "left": "issue:34634", + "right": "issue:41720", "accept": false, - "reason": "Perceiver interpolation failure and bfloat16 dtype mismatches are different model/runtime bugs." + "reason": "BarkProcessor voice preset bug is unrelated to Qwen3 auto device mapping." }, { "left": "issue:42886", "right": "issue:44843", - "accept": false, - "reason": "Both mention offline mode, but one is tokenizer cache loading and the other is unconditional model_info() in _patch_mistral_regex; not clearly the same root cause." + "accept": true, + "reason": "Same HF_HUB_OFFLINE tokenizer-load regression; 44843 pinpoints the unconditional `model_info()` call in `_patch_mistral_regex`." }, { - "left": "issue:43421", - "right": "issue:43653", + "left": "issue:44898", + "right": "issue:45072", "accept": false, - "reason": "Different tokenizer issues: runtime post-processor refresh vs mask token registration." + "reason": "Different failures in different models: Perceiver positional interpolation vs dtype mismatch in bfloat16 inference." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Specific tokenizer loading failure for one model vs a broader transformers>=5 tokenizer regression are not the same underlying bug." + "reason": "Both are tokenizer-loading complaints, but they involve different model repos and likely different compatibility issues." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "AddedToken duplicate-special-argument TypeError is unrelated to loading EMBEDDIA/est-roberta." + "reason": "`AddedToken(..., special=...)` argument duplication is unrelated to loading ScandiBERT." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate pair, 44514/44545, is supported by identical titles and the same reported crash surface. The other soft-edge rejections are reasonable given the limited evidence in the packet." + "feedback": "The summary is well grounded in the packet and stays conservative. The two accepted pairs are supported by near-identical or directly complementary titles, and the rejected pairs are phrased cautiously without overclaiming shared root cause. No clear evidence of unsupported duplication claims." }, "evaluator_used": true, "retried": false, @@ -3961,10 +3938,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30064", "issue:36010", @@ -3988,7 +3965,6 @@ "issue:43819", "issue:43844", "issue:43873", - "issue:43927", "issue:43994", "issue:44008", "issue:44062", @@ -4016,20 +3992,20 @@ "issue:45479" ], "soft_pairs": [ - "issue:43381|issue:43844", - "issue:44918|issue:44936", "issue:44077|issue:44625", "issue:44987|issue:45042", + "issue:44918|issue:44936", "issue:43531|issue:44589", + "issue:43381|issue:43844", "issue:43994|issue:44671", "issue:43421|issue:44186", "issue:39692|issue:44610", - "issue:45081|issue:45356", "issue:43525|issue:43531", + "issue:45081|issue:45356", "issue:43596|issue:44220", "issue:45183|issue:45479", - "issue:39692|issue:43873", "issue:43746|issue:45381", + "issue:39692|issue:43873", "issue:43540|issue:44008", "issue:43618|issue:43697", "issue:40444|issue:43746", @@ -4041,227 +4017,220 @@ "issue:43653|issue:43720", "issue:30064|issue:36010", "issue:43421|issue:44811", - "issue:44843|issue:45081", "issue:44062|issue:44488", + "issue:44843|issue:45081", "issue:44186|issue:44898", - "issue:43550|issue:45072", "issue:43540|issue:45381", - "issue:43550|issue:43927" + "issue:43550|issue:45072" ], - "prepared_review_unit_hash": "2ef909c90530ca3be9e144fcde58fc2db0debc8a95319e91482587be3c800b69", + "prepared_review_unit_hash": "c8f89df0e08377584f068d2578d779bd1b23025eaa8b680682c3d6a7cfe5ccff", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37303, - "estimated_input_tokens": 9326, - "estimated_eval_tokens": 18908 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36406, + "estimated_input_tokens": 9102, + "estimated_eval_tokens": 18460 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a grab-bag of mostly unrelated Transformers bug reports across tokenizers, multimodal processors, training/runtime, and model-specific regressions. None of the soft pairs look like true duplicates of the same underlying bug or change.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue cleanly fits this mixed cluster; if forced, issue 44843 is the broadest open tokenizer-regression report, but the items are largely unrelated.", + "summary": "This cluster is a mixed bag of Transformers bug reports across tokenizers, multimodal processors, model configs, and training/runtime regressions; the soft pairs mostly share superficial subsystem overlap rather than the same underlying defect.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 43421 is the cleanest central representative of the tokenizer/runtime-regression subset, but the overall cluster is too heterogeneous to treat as one true duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44843 is the best representative only in the narrow sense that it is open, concrete, and in the most recurring subtheme here (tokenizer/mistral-regex behavior), though this is not a true duplicate cluster.", + "best_issue_reason": "Issue 43421 is the strongest representative because it is concrete and reproducible, and several nearby items are tokenizer-adjacent even though they are not duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43381", - "right": "issue:43844", - "accept": false, - "reason": "Both involve training/gradient behavior, but one is eval-mode checkpointing and the other is ZeRO-3 gradient explosion; different bugs." - }, - { - "left": "issue:44918", - "right": "issue:44936", - "accept": false, - "reason": "Both mention trainer workflows, but one is SFT embedding unpacking and the other is evaluate-after-train state handling; not the same failure." - }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "Unrelated models and problems: patchtsmixer post_init policy vs Qwen3.5 num_labels propagation." + "reason": "Different models and failures: patchtsmixer post-init policy versus Qwen3.5 label propagation." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Different issues: model loading regression for physical-intelligence/fast vs PIL image processors incorrectly requiring torchvision." + "reason": "Unrelated problems: loading a specific model checkpoint versus PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Different trainer issues: unpacking Qwen3.5 embeddings with TRL versus evaluate() failing after train()." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Separate model bugs in different areas: Qwen3-MoE sliding_window handling vs Float8 storage lookup failure." + "reason": "Separate bugs in different paths: Qwen3-MoE sliding_window handling versus float8 storage lookup." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Both are training-related, but one is eval-mode gradient checkpointing and the other is ZeRO-3 gradient growth." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Both are model-output correctness bugs, but they affect different models and code paths." + "reason": "Different model-specific inference regressions: SigLIP2 nonsense outputs versus CamemBERT MLM predictions." }, { "left": "issue:43421", "right": "issue:44186", "accept": false, - "reason": "Both are tokenizer-related, but one is runtime special-token/post-processor sync and the other is a LayoutLMv2 NER padding crash." + "reason": "Tokenizer-related, but not the same bug: runtime special-token post-processing versus LayoutLMv2 NER/padding crashes." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "Different processor/model mismatches in different models; one is a SigLIP2 example/quantization issue, the other an OmDet-Turbo size mismatch." + "reason": "Both concern vision docs/processors, but one is a SigLIP2 example and the other is OmDet-Turbo input sizing." }, { - "left": "issue:45081", - "right": "issue:45356", + "left": "issue:43525", + "right": "issue:43531", "accept": false, - "reason": "Both touch _patch_mistral_regex/tokenizer loading, but the reported failures are different (missing backend_tokenizer vs codec/warning regression)." + "reason": "Config attribute error in Llama4 versus sliding-window behavior in Qwen3-MoE." }, { - "left": "issue:43525", - "right": "issue:43531", + "left": "issue:45081", + "right": "issue:45356", "accept": false, - "reason": "Unrelated config bug vs sliding_window bug in different models." + "reason": "Related tokenizer area, but different defects: backend_tokenizer crash versus Kimi-K2.5 codec/regression warning issues." }, { "left": "issue:43596", "right": "issue:44220", "accept": false, - "reason": "Different subsystems: ZeRO-3/BertModel init failure vs feature extraction helper issue." + "reason": "Distinct initialization paths: ZeRO-3/BertModel indexing versus audio fbank feature extraction." }, { "left": "issue:45183", "right": "issue:45479", "accept": false, - "reason": "Both are broad validation/classification bugs, but one is input_ids/inputs_embeds validation and the other is degenerate loss with num_labels=1." + "reason": "Validation bug versus loss formulation bug; they affect different code paths and symptoms." }, { - "left": "issue:39692", - "right": "issue:43873", + "left": "issue:43746", + "right": "issue:45381", "accept": false, - "reason": "Unrelated: SigLIP2 doc/example mismatch vs quantization offloading behavior." + "reason": "Different domains: PEFT checkpoint loading for GraniteSpeech versus Qwen2.5-VL video position IDs." }, { - "left": "issue:43746", - "right": "issue:45381", + "left": "issue:39692", + "right": "issue:43873", "accept": false, - "reason": "Different model-specific bugs: GraniteSpeech checkpoint loading with PEFT vs Qwen2.5-VL video vision_position_ids." + "reason": "Doc/example issue versus a general offloading-with-quantization runtime issue." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Both are multimodal model bugs, but they affect different models and distinct code paths." + "reason": "Separate multimodal bugs: Qwen3OmniMoe video processing versus Gemma 3n audio_mel_mask name collision." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "Both concern model outputs, but the CLIP attentions regression and RTDetrV2 output drift are separate bugs." + "reason": "Different model regressions with different outputs and surfaces; not the same underlying defect." }, { "left": "issue:40444", "right": "issue:43746", "accept": false, - "reason": "Different areas: Qwen2.5-VL iterable dataset/multi-image failure vs GraniteSpeech PEFT checkpoint loading." + "reason": "Different loading/training problems affecting unrelated models and workflows." }, { "left": "issue:43746", "right": "issue:43873", "accept": false, - "reason": "No overlap beyond being runtime issues; PEFT checkpoint loading and quantization offloading are unrelated." + "reason": "PEFT checkpoint loading is unrelated to quantization/offloading behavior." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio feature extraction helper vs tokenizer-mapping/model registration logic; unrelated." + "reason": "Audio feature extraction issue versus tokenizer-mapping/model registration failure." }, { "left": "issue:43716", "right": "issue:44008", "accept": false, - "reason": "Different multimodal bugs in different models: image preprocessor dtype mismatch vs Gemma 3n variable collision." + "reason": "Different multimodal mismatch bugs: image preprocessor dtype versus audio tensor attribute collision." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "Both are model configuration issues, but they concern different architectures and different parameter logic." + "reason": "Different model config/content bugs: sparse-layer creation versus Smollm3 RoPE-layer omission." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Different bugs: accelerate loading of packed weights vs DAC latent decoding/STE mismatch." + "reason": "Different quantization issues: accelerate unpacking of packed weights versus DAC STE mismatch." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Tokenizer special-token registration vs BitNet packed-weight loading; unrelated." + "reason": "Tokenizer special-token registration versus BitNet packed-weight loading are unrelated failures." }, { "left": "issue:30064", "right": "issue:36010", "accept": false, - "reason": "Completely different problems: image processor segmentation maps vs import failure for GenerationMixin." + "reason": "Image segmentation-map processing has nothing to do with the GenerationMixin import error." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Both are tokenizer/processor-adjacent, but one is post-processor updating and the other is Whisper batch_decode skip_special_tokens behavior." - }, - { - "left": "issue:44843", - "right": "issue:45081", - "accept": false, - "reason": "Same helper area, but different concrete failures: offline model_info call vs backend_tokenizer AttributeError." + "reason": "Both touch tokenization, but one is runtime special-token updates and the other is Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "Different tokenizer bugs: AddedToken construction error vs failure loading a specific model/tokenizer." + "reason": "AddedToken construction error versus model loading failure for cjvt/sleng-bert." }, { - "left": "issue:44186", - "right": "issue:44898", + "left": "issue:44843", + "right": "issue:45081", "accept": false, - "reason": "Unrelated models and paths: LayoutLMv2 tokenizer crash vs Perceiver image-classification resolution handling." + "reason": "Same patch family, but one is offline hub access in model_info() and the other is a backend_tokenizer attribute crash." }, { - "left": "issue:43550", - "right": "issue:45072", + "left": "issue:44186", + "right": "issue:44898", "accept": false, - "reason": "Different runtime issues in different models: torch.compile+SDPA failure vs bf16 dtype mismatch." + "reason": "Different vision/tokenizer bugs: LayoutLMv2 NER padding versus Perceiver image-size handling." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Both involve video/multimodal inputs, but they are different models with different bugs and fixes." + "reason": "Both are multimodal, but video-processing failures in different Qwen variants are not the same bug." }, { "left": "issue:43550", - "right": "issue:43927", + "right": "issue:45072", "accept": false, - "reason": "No overlap: torch.compile/SDPA model failure vs config token-ID persistence bug." + "reason": "Different low-level issues: torch.compile with SDPA versus bfloat16 dtype mismatches." } ] }, @@ -4280,10 +4249,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30064", "issue:34634", @@ -4308,6 +4277,7 @@ "issue:43825", "issue:43827", "issue:43906", + "issue:43927", "issue:43994", "issue:44060", "issue:44186", @@ -4336,13 +4306,14 @@ "soft_pairs": [ "issue:43577|issue:43818", "issue:43643|issue:44877", + "issue:43550|issue:43927", "issue:43531|issue:43643", "issue:43782|issue:44977", "issue:43644|issue:43749", "issue:44448|issue:44671", "issue:43577|issue:43582", - "issue:43653|issue:43819", "issue:44561|issue:45020", + "issue:43653|issue:43819", "issue:43825|issue:43994", "issue:43550|issue:43819", "issue:44821|issue:45084", @@ -4351,12 +4322,12 @@ "issue:43650|issue:43906", "issue:41720|issue:44991", "issue:44805|issue:44936", - "issue:30064|issue:36331", "issue:44485|issue:44509", + "issue:30064|issue:36331", "issue:43606|issue:45072", "issue:42175|issue:45081", - "issue:44610|issue:45072", "issue:43782|issue:45325", + "issue:44610|issue:45072", "issue:34634|issue:36010", "issue:44060|issue:44368", "issue:45081|issue:45406", @@ -4364,208 +4335,214 @@ "issue:43653|issue:44671", "issue:43749|issue:44291" ], - "prepared_review_unit_hash": "61a2cc16c4f14525d498d795fe325810d6dd893d295f66394ca324fde0a996c7", + "prepared_review_unit_hash": "cc8f72d4fb4f7e60c29ff24d8495872414bb925574834cdc5e2f2b1e854f49c4", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35985, - "estimated_input_tokens": 8997, - "estimated_eval_tokens": 18250 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36878, + "estimated_input_tokens": 9220, + "estimated_eval_tokens": 18696 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a heterogeneous mix of Transformers regressions and model-specific bugs (tokenizers, multimodal processors, remote_code loading, v5 behavior changes). I do not see any soft pair that is a true duplicate; several are thematically related, but they are different concrete bugs or different artifacts (runtime vs docs).", - "confidence": 0.79, - "canonical_issue_reason": "issue:45020 is the broadest representative here: it captures the recurring recent-transformers/remote_code regression theme better than the more specific reports.", + "summary": "This set is highly heterogeneous: it spans unrelated bugs in processors, tokenizers, model loading, dtype/device mismatches, docs, and training behavior. None of the soft pairs look like the same concrete underlying bug or change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 45020 is the broadest umbrella-like report here (recent remote_code regressions), but the overall cluster is too diverse to collapse into a single duplicate issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best single issue to keep as the cluster representative because it is open, recent, and broadly covers the shared failure mode seen in several related reports.", + "best_issue_reason": "45020 is the best representative issue by scope, but it is still only a loose umbrella and not a true duplicate target for most of the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43577", "right": "issue:43818", "accept": false, - "reason": "Different models and failure modes: BLIP2 dtype handling vs Video-LLaVA missing temporal attention/shared weights." + "reason": "Different failures: BLIP2 dtype casting vs Video-LLaVA missing temporal attention/weight sharing." }, { "left": "issue:43643", "right": "issue:44877", "accept": false, - "reason": "Both involve config loading, but one is missing fields with trust_remote_code while the other is strict config rejection for granite_speech; not the same bug." + "reason": "Both involve config loading, but one is trust_remote_code missing fields and the other is strict config blocking granite_speech; different bugs." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to DiaConfig save/load token ID loss." }, { "left": "issue:43531", "right": "issue:43643", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and trust_remote_code config field loss are unrelated bugs." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to trust_remote_code config field loss." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "Same model family, but different concrete issues: from_pretrained weight_only failure vs flash-attention generation failure." + "reason": "Different Qwen model/runtime issues: weight_only load error vs flash-attention generation failure." }, { "left": "issue:43644", "right": "issue:43749", "accept": false, - "reason": "Both are loading-related regressions, but one is non-persistent buffers being junk-filled and the other is FSDP CPU RAM efficient loading breakage." + "reason": "Non-persistent buffer junk in v5 is a different problem from FSDP CPU RAM efficient loading." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Both are output regressions in v5, but they affect different models/tasks and appear to be distinct bugs." + "reason": "Both are v5 regressions, but Pegasus output drift and CamemBERT masked-LM predictions are separate model-specific bugs." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "Unrelated issues: BLIP2 dtype values vs Apple Silicon caching_allocator_warmup TypeError." + "reason": "BLIP2 dtype mismatch is not the same as Apple Silicon caching_allocator_warmup TypeError." }, { - "left": "issue:43653", - "right": "issue:43819", + "left": "issue:44561", + "right": "issue:45020", "accept": false, - "reason": "Tokenizer special-token registration bug vs DAC forward/from_latents mismatch; different code paths." + "reason": "Removal of is_torch_fx_available is one specific remote_code breakage; 45020 is a broader umbrella report." }, { - "left": "issue:44561", - "right": "issue:45020", + "left": "issue:43653", + "right": "issue:43819", "accept": false, - "reason": "44561 is a specific root cause for remote_code breakage; 45020 is a broad umbrella issue that may include multiple causes, so they are not the same concrete bug." + "reason": "BigBirdTokenizer special-token decode bug is unrelated to DAC.from_latents/STE mismatch." }, { "left": "issue:43825", "right": "issue:43994", "accept": false, - "reason": "Both mention pipeline behavior, but one is an incorrect error message for removed translation tasks and the other is nonsensical model outputs for siglip2." + "reason": "Pipeline error-message text in v5 is unrelated to SigLIP2 nonsensical outputs." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Different failures: torch.compile/SDPA on Bamba-9B-v2 vs DAC from_latents/STE mismatch." + "reason": "Bamba torch.compile with SDPA and DAC latent reconstruction are different code paths and symptoms." }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor-from-URL loading and template compilation errors are unrelated." + "reason": "AutoImageProcessor URL loading and non-template node compilation are unrelated." }, { "left": "issue:43825", "right": "issue:43827", "accept": false, - "reason": "Related to the v5 pipeline removals, but one is a runtime error message and the other is stale documentation; not the same artifact or fix." + "reason": "Both touch v5 pipeline removals, but one is runtime error messaging and the other is stale docs; not the same concrete bug." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "Different tokenizer bugs affecting different models and input shapes." + "reason": "LayoutLMv2 NER/padding crash and FastSpeech2ConformerTokenizer loading failure are separate tokenizer issues." }, { "left": "issue:43650", "right": "issue:43906", "accept": false, - "reason": "The first is a vague/low-signal issue; the second is an isolated reproduction of a specific prior bug." + "reason": "The placeholder 'ADD THE DATA' issue is unrelated to the isolated reproduction of issue 38071." }, { "left": "issue:41720", "right": "issue:44991", "accept": false, - "reason": "Different subsystems and failures: device mapping cudaErrorAssert vs tokenizer loading regression." + "reason": "Qwen3 auto device mapping cuda assert and EMBEDDIA tokenizer loading failure are unrelated." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "A mask/index shape mismatch is not the same as trainer.evaluate failing after train." + "reason": "Mask/index shape mismatch is not the same as trainer.evaluate() failing after train()." }, { - "left": "issue:30064", - "right": "issue:36331", + "left": "issue:44485", + "right": "issue:44509", "accept": false, - "reason": "Image processor segmentation-map handling and CustomTrainer.compute_loss signature mismatch are unrelated." + "reason": "GLM-5 RoPE discussion is unrelated to docs still mentioning removed pipeline tasks." }, { - "left": "issue:44485", - "right": "issue:44509", + "left": "issue:30064", + "right": "issue:36331", "accept": false, - "reason": "GLM-5 RoPE implementation discussion vs docs still mentioning removed pipeline tasks; different concerns." + "reason": "ImageProcessor void segmentation maps and CustomTrainer.compute_loss signature mismatch are different bugs." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "CPU offload device mismatch and bfloat16 dtype mismatches are different bug classes." + "reason": "CPU offload device mismatch in bark-small is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Packaging/backend dependency issue vs Mistral regex patch crash; unrelated." + "reason": "Packaging/backend dependency issue is unrelated to Mistral tokenizer regex patch crash." }, { - "left": "issue:44610", - "right": "issue:45072", + "left": "issue:43782", + "right": "issue:45325", "accept": false, - "reason": "Processor resolution mismatch vs dtype mismatch in inference are different concrete problems." + "reason": "Different VL bugs: Qwen3VL weight_only load error versus Qwen2.5-VL RoPE position scaling." }, { - "left": "issue:43782", - "right": "issue:45325", + "left": "issue:44610", + "right": "issue:45072", "accept": false, - "reason": "Both are Qwen vision-language issues, but one is a load-time weight_only error and the other is a RoPE/position-id scaling bug." + "reason": "OmDet-Turbo processor size mismatch and bfloat16 dtype mismatches are unrelated." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset behavior and GenerationMixin import failure are unrelated." + "reason": "BarkProcessor voice_preset and GenerationMixin import error are unrelated issues." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both are tie-weights warnings in Qwen variants, but the reported mis-tie bug and the tie_word_embeddings=False warning are different issues." + "reason": "Qwen3-Next tied-weight warning and Qwen3.5 tie_word_embeddings warning are similar only at a high level, but not the same bug." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Different processors and different missing attributes: backend_tokenizer vs _tokenizer." + "reason": "Mistral tokenizer regex crash and Gemma4Processor _tokenizer missing are different processor/tokenizer failures." }, { "left": "issue:43720", "right": "issue:45072", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to bfloat16 dtype mismatches." + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to dtype mismatches in inference." }, { "left": "issue:43653", "right": "issue:44671", "accept": false, - "reason": "BigBirdTokenizer special-token registration and CamemBERT masked-LM predictions are different bugs." + "reason": "BigBirdTokenizer special-token registration and CamemBERT masked-LM output issues are different tokenizer/model bugs." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Both touch loading/init paths, but FSDP CPU RAM efficient loading and init_empty_weights _is_hf_initialized signature mismatch are not the same concrete problem." + "reason": "FSDP CPU RAM efficient loading and init_empty_weights unexpected _is_hf_initialized argument are different loading-path failures." } ] }, @@ -4584,10 +4561,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:34567", "issue:34634", @@ -4649,20 +4626,20 @@ "issue:34634|issue:36331", "issue:44568|issue:45488", "issue:43454|issue:43577", - "issue:43638|issue:44936", "issue:43452|issue:44991", + "issue:43638|issue:44936", "issue:43257|issue:43329", "issue:43746|issue:44811", "issue:43854|issue:44977", - "issue:38175|issue:44610", "issue:44589|issue:45042", - "issue:43257|issue:43696", + "issue:38175|issue:44610", "issue:39692|issue:44912", - "issue:43819|issue:43927", + "issue:43257|issue:43696", "issue:36010|issue:43645", - "issue:43720|issue:44610", "issue:44568|issue:45356", + "issue:43819|issue:43927", "issue:43550|issue:44912", + "issue:43720|issue:44610", "issue:43749|issue:44756", "issue:43262|issue:44811", "issue:43531|issue:45042", @@ -4672,226 +4649,226 @@ "issue:43257|issue:43756", "issue:34567|issue:40990" ], - "prepared_review_unit_hash": "8d5934c31a105aa84e50677c9fa64f4dc088117526299fb551985c1005e9e25f", + "prepared_review_unit_hash": "14f853e63b62b864b1a931307eff15bbb05e40759ae9d8c8e59b886a96891cd0", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37475, - "estimated_input_tokens": 9369, - "estimated_eval_tokens": 18994 + "serialized_chars": 37477, + "estimated_input_tokens": 9370, + "estimated_eval_tokens": 18996 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the items span unrelated trainer, tokenizer, model-loading, vision, audio, and quantization bugs. None of the soft edge pairs look like the same underlying defect, so there is no safe duplicate canonical here.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue is appropriate; the cluster is not a duplicate set and the soft-paired issues address different code paths/models/platforms.", + "summary": "Mostly unrelated closed/open issue reports across different Transformers subsystems; no PRs are present, and none of the suggested soft pairs look like true duplicates of the same underlying bug.", + "confidence": 0.24, + "canonical_issue_reason": "There is no strong single canonical because the set is heterogeneous. If a fallback anchor is needed, the broad tokenizer regression in issue 45488 is the clearest, most reusable bug report.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue for deduping: every candidate is a distinct bug report rather than a representative duplicate anchor.", + "best_issue_reason": "Issue 45488 is the most triage-useful fallback: it describes a concrete, user-visible regression with a clear failure mode and broad impact on a major tokenizer family.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "Different failures in different vision models: BEiTImageProcessorFast reduce_labels vs RTDetrV2 output divergence." + "reason": "Both are vision-model regressions, but one is a BeitImageProcessor label-reduction bug and the other is an RTDetrV2 output-drift report; different code paths." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "AppleSilicon allocator warmup TypeError is unrelated to the CUDA AMP crash in LwDetrImageLoss." + "reason": "Unrelated failures: Apple Silicon allocator warmup TypeError vs. CUDA AMP loss crash in LwDetrImageLoss." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and DAC.from_latents vs forward mismatch are separate audio bugs." + "reason": "Audio processor chat-template sampling-rate bug vs. DAC latent/STE mismatch; different components and symptoms." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "BEiT label reduction bug and MPS attention value-head correctness issue are unrelated." + "reason": "BeitImageProcessorFast label handling and an MPS attention correctness issue are not the same underlying bug." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Video token counting undefined symbols and auto_mappings CI breakage are different problems." + "reason": "Video multimodal token counting bug vs. auto_mappings CI breakage; different files and failure modes." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template-node compile error does not match the CLIPTokenizer model_max_length bug." + "reason": "Compile/template-node error is unrelated to CLIPTokenizer's model_max_length behavior." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "BigBirdTokenizer special-token registration and GPT-oss MXFP4 loading/fallback are unrelated." + "reason": "BigBird special-token registration bug vs. gpt-oss MXFP4 load fallback; no shared code path." }, { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss kwargs are different API regressions." + "reason": "BarkProcessor voice_preset issue and CustomTrainer.compute_loss signature breakage are unrelated." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Both are tokenizer-related, but one is add_special_tokens BOS/EOS behavior and the other is LlamaTokenizer overriding pre-tokenization." + "reason": "Both are tokenizer regressions, but one is missing BOS/EOS insertion and the other overrides tokenizer.json pre-tokenizer settings; not the same concrete bug." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "AyaVision lm_head tying and Blip2 dtype propagation are distinct model-loading bugs." + "reason": "AyaVision lm_head tying bug and BLIP2 dtype propagation bug affect different model-loading paths." }, { - "left": "issue:43638", - "right": "issue:44936", + "left": "issue:43452", + "right": "issue:44991", "accept": false, - "reason": "DeepSpeed ZeRO3 empty-index failure is not the same as evaluate() failing after train()." + "reason": "GGUF/from_pretrained breakage and est-roberta tokenizer loading failure are distinct issues." }, { - "left": "issue:43452", - "right": "issue:44991", + "left": "issue:43638", + "right": "issue:44936", "accept": false, - "reason": "Both involve loading, but gguf_file handling and EMBEDDIA/est-roberta tokenizer loading have different causes and code paths." + "reason": "DeepSpeed ZeRO3 index error for a non-pretrained Bert model is unrelated to evaluate() failing after train()." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed and multimodal video token counting are unrelated." + "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed is unrelated to the multimodal video token-counting bug." }, { "left": "issue:43746", "right": "issue:44811", "accept": false, - "reason": "PEFT adapter checkpoint loading and Whisper batch_decode skip_special_tokens are separate issues." + "reason": "PEFT/local-checkpoint loading problem and Whisper batch_decode skip_special_tokens behavior are different subsystems." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "GLM-4.7-Flash unit-test loading failure and Qwen3.5 flash-attention generation bug are different model-specific problems." - }, - { - "left": "issue:38175", - "right": "issue:44610", - "accept": false, - "reason": "SigLIP2 zero probabilities and OmDet-Turbo 640px/224px processor mismatch are not the same bug." + "reason": "GLM-4.7-Flash load failure in tests vs. Qwen3.5 flash-attention generation regression; separate model-specific bugs." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup failure and PIL-backend torchvision dependency regression are unrelated." + "reason": "Float8 storage lookup error and PIL image-processor torchvision dependency regression are unrelated." }, { - "left": "issue:43257", - "right": "issue:43696", + "left": "issue:38175", + "right": "issue:44610", "accept": false, - "reason": "Qwen3 MoE conversion/loading and GPT-oss CUDA OOM are different failure modes." + "reason": "SigLIP2 zero-probabilities and OmDet-Turbo input-size mismatch are both vision issues but not the same bug." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 doc-example errors and GPT-oss MXFP4 quantization fallback are unrelated." + "reason": "Documentation example/model mismatch and gpt-oss MXFP4 loading are different failures." }, { - "left": "issue:43819", - "right": "issue:43927", + "left": "issue:43257", + "right": "issue:43696", "accept": false, - "reason": "DAC.from_latents STE mismatch and DiaConfig token-ID loss after save/load are different audio/config bugs." + "reason": "Qwen3 MoE weight conversion and CUDA OOM on GPT-oss-20b are unrelated." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "GenerationMixin import regression and Jupyter custom-model initialization issue are not the same defect." + "reason": "ImportError for GenerationMixin and Jupyter custom-model initialization breakage are not the same issue." }, { - "left": "issue:43720", - "right": "issue:44610", + "left": "issue:44568", + "right": "issue:45356", "accept": false, - "reason": "Packed-weight unpacking during accelerate loading is unrelated to OmDet-Turbo input-size mismatch." + "reason": "Both are tokenizer regressions, but the failure modes are different: missing special tokens vs. codec/regex handling regression." }, { - "left": "issue:44568", - "right": "issue:45356", + "left": "issue:43819", + "right": "issue:43927", "accept": false, - "reason": "Tokenizer special-token omission and Kimi-K2.5 codec/warning regression are different tokenizer paths." + "reason": "DAC forward/from_latents mismatch and DiaConfig token-id persistence loss are separate model/config bugs." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "Bamba torch.compile SDPA failure and GPT-oss MXFP4 loading issue are unrelated." + "reason": "SDPA/torch.compile failure in Bamba is unrelated to quantization fallback in gpt-oss-20b." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight loading bug in BitNet and OmDet-Turbo processor size mismatch are different code paths." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "FSDP CPU RAM efficient loading and Strix Halo mmap OOM are distinct memory/loading bugs." + "reason": "FSDP CPU RAM efficient loading breakage and Strix Halo mmap OOM are both memory/loading themed but not the same concrete bug." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio chat-template sampling-rate defaults and Whisper batch_decode skip_special_tokens are unrelated processor bugs." + "reason": "Both touch audio processors, but one is chat-template sampling rate and the other is batch_decode special-token handling." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and PIL-backend torchvision dependency are unrelated." + "reason": "Qwen3-MoE sliding-window issue and PIL backend torchvision dependency regression are unrelated." }, { "left": "issue:43329", "right": "issue:44479", "accept": false, - "reason": "Both are video-related, but one is an internal token-counting undefined-symbol bug and the other is a broader input regression across models." + "reason": "Both mention video/multimodal regressions, but one is undefined helper usage in token counting and the other is a broader model-specific video-input regression; not enough to treat as the same bug." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Qwen3 MoE weight conversion and Qwen3-VL shape mismatch are different model-loading problems." + "reason": "Qwen3 MoE weight conversion bug and Qwen3-VL-30B-A3B weight-shape mismatch are distinct model-loading failures." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "GenerationMixin import failure and CustomTrainer.compute_loss kwarg error are unrelated." + "reason": "GenerationMixin import failure and CustomTrainer compute_loss signature incompatibility are unrelated API breaks." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "Qwen3 MoE loading/conversion is unrelated to Smollm3 RoPE-layer dropping." + "reason": "Qwen3 MoE loading/conversion bug and Smollm3 RoPE-layer mismatch concern different models and mechanisms." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token-count tracking and GPT-oss perplexity are different training/eval issues." + "reason": "TrainerState token-count tracking and GPT-oss perplexity are not the same underlying problem." } ] }, @@ -4910,10 +4887,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:28282", "issue:33357", @@ -4979,16 +4956,16 @@ "issue:33357|issue:34567", "issue:42915|issue:43742", "issue:43540|issue:44560", - "issue:43906|issue:45381", "issue:43540|issue:44479", + "issue:43906|issue:45381", "issue:42915|issue:45538", "issue:42175|issue:43824", "issue:42915|issue:44821", "issue:43526|issue:45479", "issue:44062|issue:44991", - "issue:45137|issue:45405", "issue:42175|issue:44488", "issue:44737|issue:45588", + "issue:45137|issue:45405", "issue:38175|issue:44898", "issue:28282|issue:34689", "issue:43646|issue:43749", @@ -4997,227 +4974,224 @@ "issue:43582|issue:43782", "issue:43742|issue:44291" ], - "prepared_review_unit_hash": "c72ff5b101cb6d4f8a0bf572c044b7493d6b46ce418ec76ee5b0cc11ff577ca2", + "prepared_review_unit_hash": "f0302386fd268a03a070db91126c50832351480e62d8bd79021dbb840e7f3b06", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37131, - "estimated_input_tokens": 9283, - "estimated_eval_tokens": 18822 + "serialized_chars": 37136, + "estimated_input_tokens": 9284, + "estimated_eval_tokens": 18824 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the paired items share broad keywords (video, loading, dtype, config, tokenizer) but not the same underlying bug or change. I would reject all soft edges as duplicates and avoid picking a single canonical issue/PR for the cluster.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is mostly heterogeneous, with one clear duplicate-like pair around the Qwen video-input regression. Most other soft pairs share only broad subsystem similarity and should stay separate.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44479 is the broadest and most central report in the nearest video-regression subcluster, covering the 5.3.0 Qwen video-input breakage across multiple models.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 44479 best represents the cluster\u2019s strongest duplicate candidate because it states the shared regression at a higher level than the narrower follow-up reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Both involve video/multimodal processing, but 43329 is an undefined-function bug in token counting while 43450 is a batched-shape bug in video processors." + "reason": "Different bugs: undefined video token-count helpers vs batched video processor shape handling." }, { "left": "issue:43697", "right": "issue:45357", "accept": false, - "reason": "Same Qwen/vision ecosystem, but 43697 reports differing inference outputs in RTDetrV2 and 45357 is a save_pretrained key serialization regression." + "reason": "Different failures: RTDetrV2 output drift vs save/load visual-encoder key serialization." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Both mention dtype, but they affect different models and code paths: Blip2 loading stays float32 versus SwitchTransformers/TimmWrapperModel inference mismatches." + "reason": "Both mention dtype, but they affect different models and code paths." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Both are Qwen-related failures, but one is an auto device-map CUDA assert and the other is a video-input ValueError in Qwen3OmniMoe." + "reason": "Qwen3 auto device-map CUDA assert is not the same as Qwen3OmniMoe video-input ValueError." }, { "left": "issue:44479", "right": "issue:45381", - "accept": false, - "reason": "The second looks like a specific manifestation of a broader video regression, but there is not enough evidence that it is the same concrete bug/code path." + "accept": true, + "reason": "Both describe the same Qwen video-input regression in v5.3.0; 45381 looks like a specific symptom of the broader bug." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Different problems: a model load failure for physical-intelligence/fast versus an abnormal CLIPTokenizer max_length default." + "reason": "Generic model-loading regression vs a tokenizer model_max_length bug." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "gguf_file/from_pretrained breakage is unrelated to the Apple Silicon caching_allocator_warmup TypeError." + "reason": "GGUF loading failure and Apple Silicon warmup TypeError are unrelated." }, { "left": "issue:42175", "right": "issue:44938", "accept": false, - "reason": "One is a packaging extra missing TensorFlow, the other is a Python 3.14 load failure; different failure modes and fixes." + "reason": "Packaging/backend inclusion issue vs Python 3.14 load failure." }, { "left": "issue:44291", "right": "issue:45042", "accept": false, - "reason": "Unexpected _is_hf_initialized handling during init_empty_weights is unrelated to PIL image processors requiring torchvision." + "reason": "Unexpected init_empty_weights argument and torchvision requirement regression are different bugs." }, { "left": "issue:43317", "right": "issue:45538", "accept": false, - "reason": "Device-map/offload loading failure and CLIPTokenizer max_length are unrelated issues." + "reason": "Dequantized model offload failure is unrelated to CLIPTokenizer max_length." }, { "left": "issue:43927", "right": "issue:44912", "accept": false, - "reason": "One is config token IDs being lost on save/load; the other is MXFP4 quantization falling back to bf16." + "reason": "Config field loss on save/load and MXFP4 quantization fallback are different issues." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "A MacOS bus error while loading CLIP models is unrelated to TrainerState token-count tracking." + "reason": "MacOS bus error on CLIP loading is unrelated to TrainerState token counter updates." }, { "left": "issue:42915", "right": "issue:43742", "accept": false, - "reason": "Different model-loading failures with different symptoms: FineGrainedFP8Config versus a KeyError on MobileLLM-125M." + "reason": "FineGrainedFP8Config failure and MobileLLM key error are distinct loading problems." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Both are video-related, but they target different Qwen model families and different failures (ValueError vs StopIteration)." - }, - { - "left": "issue:43906", - "right": "issue:45381", - "accept": false, - "reason": "43906 is just a reproduction of another issue; it does not match the specific qwen2.5-vl video position-id regression." + "reason": "Both are video-related, but the models and failure modes differ (ValueError vs StopIteration)." }, { "left": "issue:43540", "right": "issue:44479", "accept": false, - "reason": "Related area, but not the same concrete bug: one is a Qwen3OmniMoe video processing error, the other is a broader multi-model video regression." + "reason": "Qwen3OmniMoe video-input ValueError is not the same concrete regression as the broader Qwen 5.3.0 video-input bug." }, { - "left": "issue:42915", - "right": "issue:45538", + "left": "issue:43906", + "right": "issue:45381", "accept": false, - "reason": "Unrelated: FP8 config failure versus tokenizer model_max_length behavior." + "reason": "A reproduction of another issue is not the same bug as the Qwen video regression." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Missing TensorFlow in the torch extra is unrelated to an import error for Qwen2_5_VLForConditionalGeneration." + "reason": "TensorFlow backend packaging and missing Qwen2_5_VL export are unrelated." }, { "left": "issue:42915", "right": "issue:44821", "accept": false, - "reason": "Different subsystems entirely: Qwen3Moe FP8 config versus AutoImageProcessor URL loading." + "reason": "FP8 config failure is unrelated to loading AutoImageProcessor from URL." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "BeitImageProcessorFast label reduction and sequence-classification zero loss are different bugs in different pipelines." + "reason": "BeitImageProcessorFast label reduction bug and sequence-classification zero-loss bug are different code paths." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "AddedToken argument duplication is unrelated to tokenizer loading for EMBEDDIA/est-roberta." - }, - { - "left": "issue:45137", - "right": "issue:45405", - "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow is unrelated to a PEFT minimum-version bump." + "reason": "AddedToken special-arg conflict is unrelated to tokenizer loading for est-roberta." }, { "left": "issue:42175", "right": "issue:44488", "accept": false, - "reason": "One is packaging metadata, the other is a model-loading failure for a specific checkpoint." + "reason": "Backend install issue vs model loading failure for cjvt/sleng-bert." }, { "left": "issue:44737", "right": "issue:45588", "accept": false, - "reason": "XLNet CPU-placement inefficiency is unrelated to the flash_attention s_aux=None AttributeError." + "reason": "CPU arange placement bug and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT version bump." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Different image-model bugs with different symptoms and code paths." + "reason": "Unexpected zero probabilities and Perceiver non-default-resolution failure are different vision bugs." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Generic PyTorch-missing import error versus a specific Llama 3.2 Vision loading regression." + "reason": "Missing PyTorch ImportError and Llama 3.2 Vision model loading break are not the same issue." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Custom model initialization breakage is unrelated to FSDP_CPU_RAM_EFFICIENT_LOADING." + "reason": "Custom model init breakage and FSDP CPU RAM efficient loading are distinct failures." }, { "left": "issue:36010", "right": "issue:42175", "accept": false, - "reason": "GenerationMixin import failure and missing TensorFlow backend dependency are unrelated." + "reason": "GenerationMixin import failure is unrelated to TensorFlow backend packaging." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both are config-field preservation issues, but one is a generic trust_remote_code load problem and the other is a specific Qwen3_5MoeVisionConfig field drop." + "reason": "Missing fields after trust_remote_code load and dropped deepstack_visual_indexes are different config-serialization bugs." }, { "left": "issue:43582", "right": "issue:43782", "accept": false, - "reason": "Apple Silicon warmup TypeError and Qwen3VL weight_only loading error are unrelated." + "reason": "Apple Silicon warmup TypeError and Qwen3VL weight_only load error are unrelated." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "KeyError loading MobileLLM-125M is unrelated to init_empty_weights passing an unexpected argument." + "reason": "MobileLLM key error and init_empty_weights argument error are different loading issues." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: most pairs are clearly cross-subsystem or cross-code-path matches, and the one accepted pair (44479/45381) is a conservative duplicate call within the same Qwen video-input regression family. The canonical/best-issue rationale is also appropriately cautious given the titles and cluster structure." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -5229,10 +5203,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:28282", "issue:29127", @@ -5286,8 +5260,8 @@ "soft_pairs": [ "issue:43756|issue:45325", "issue:28282|issue:29127", - "issue:43493|issue:43994", "issue:44484|issue:44704", + "issue:43493|issue:43994", "issue:34567|issue:36331", "issue:44291|issue:44821", "issue:44062|issue:45081", @@ -5312,20 +5286,20 @@ "issue:43295|issue:45356", "issue:43819|issue:44811" ], - "prepared_review_unit_hash": "0216faf31c17d37657588dee09b292bf9892b5401c152b5b5d8a381b5a9ef85a", + "prepared_review_unit_hash": "00d8efedb29f76e158422f41b3ed228efe7c09dd2cd2e67981a2bddb8e9666c9", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36183, + "serialized_chars": 36184, "estimated_input_tokens": 9046, "estimated_eval_tokens": 18348 }, @@ -5334,181 +5308,184 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it contains many unrelated Transformers bug reports, mostly around different model-loading, tokenizer, processor, and training regressions. None of the soft pairs look like true duplicates or the same underlying change.", - "confidence": 0.23, - "canonical_issue_reason": "issue:43452 is the broadest, most concrete loader-regression candidate here, affecting both AutoTokenizer.from_pretrained and AutoModelForCausalLM.from_pretrained via gguf_file.", + "summary": "Only one pair looks like a true duplicate: the two SigLIP2 issues describing the same model-implementation mismatch. The rest are either different models, different code paths, or only loosely related symptoms.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43493 is the broader root-cause report about SigLIP2 diverging from the original JAX implementation; issue:43994 is a downstream symptom on a specific checkpoint/pipeline path.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43452 is the strongest representative issue in the set because it has a clear, user-facing failure mode and spans multiple entry points, but it is still not a good global duplicate anchor for the rest of the cluster.", + "best_issue_reason": "issue:43493 is the best representative issue because it states the underlying bug more generally and is more useful as the canonical tracker.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43756", "right": "issue:45325", "accept": false, - "reason": "Different bugs in different models: Smollm3 RoPE-layer dropping vs Qwen2.5-VL still-image temporal position scaling." + "reason": "Both mention RoPE/position handling, but they concern different model families and different mechanisms." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "One is a missing-PyTorch ImportError for AutoModel; the other is a LayoutLMv3 error-message clarity issue." - }, - { - "left": "issue:43493", - "right": "issue:43994", - "accept": false, - "reason": "Both mention SigLIP2, but one is an implementation-vs-JAX discrepancy and the other is nonsensical outputs in AutoModel/pipeline; not clearly the same defect." + "reason": "One is a missing PyTorch dependency error; the other is a LayoutLMv3 validation/error-message clarity issue." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "Unrelated: save_pretrained max_shard_size question vs AutoProcessor kwargs not reaching cached_file." + "reason": "Unrelated: save_pretrained shard sizing vs AutoProcessor kwargs forwarding." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Same underlying SigLIP2 bug: HF outputs diverge from the reference implementation and manifest as bad results on the named checkpoint/pipeline." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "Different trainer issues: token counting not updating vs CustomTrainer.compute_loss signature breakage." + "reason": "TrainerState token counting and CustomTrainer.compute_loss signature are separate trainer issues." }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "Different loading paths and failures: init_empty_weights/_is_hf_initialized vs loading AutoImageProcessor from URL." + "reason": "Different loading failures: init_empty_weights/_is_hf_initialized versus AutoImageProcessor URL loading." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Tokenizers AddedToken special-kwarg conflict is unrelated to the Mistral regex patch crash." + "reason": "Different tokenizer bugs with different failure modes and code paths." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "Both are model-loading failures, but one is gguf_file regression across tokenizer/model classes while the other is a specific model that fails to load; no clear shared root cause." + "reason": "Both are loading-related, but one is gguf_file/AutoTokenizer-AutoModelForCausalLM behavior and the other is a specific model that fails to load for a different reason." }, { "left": "issue:44792", "right": "issue:44964", "accept": false, - "reason": "Different multimodal model failures: a janus test case vs loading Phi-4 multimodal instruct." + "reason": "Different multimodal models and different failure points; one is a failed test, the other is model loading." }, { "left": "issue:44514", "right": "issue:45290", "accept": false, - "reason": "Same API family, but different crash conditions in apply_chat_template: batched padding=False vs assistant tool-call messages with no content." + "reason": "Both are chat-template crashes, but under different inputs and triggers, so not the same concrete bug." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "A dependency version floor problem is unrelated to stale auto_mappings references causing CI breakage." + "reason": "Release/versioning issue vs CI break from removed config references; unrelated." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both touch trust_remote_code, but one is missing fields in AutoConfig output while the other is removal of is_torch_fx_available breaking remote-code models." + "reason": "Different trust_remote_code regressions affecting different APIs and failure modes." }, { "left": "issue:43688", "right": "issue:43749", "accept": false, - "reason": "Different subsystems: auxiliary-loss normalization vs FSDP CPU RAM-efficient loading." + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are unrelated." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Both involve Qwen2.5-VL, but one is chat-template batching and the other is wrong video vision_position_ids." + "reason": "Different Qwen2.5-VL problems: chat templating crash vs incorrect vision_position_ids." }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "Different models and regressions: RTDetrV2 output drift vs Qwen3.5 save_pretrained checkpoint corruption." + "reason": "Different models and different regressions: RTDetrV2 output drift vs Qwen3.5 save_pretrained checkpoint corruption." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "Generic load failure for MobileLLM is unrelated to missing Float8 storage object handling." + "reason": "KeyError when loading MobileLLM-125M is unrelated to missing Float8 storage type." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "PIL image-processor torchvision dependency regression is unrelated to the Mistral regex AttributeError." + "reason": "PIL backend torchvision dependency and Mistral regex backend_tokenizer crash are unrelated." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding_window bug is unrelated to AutoImageProcessor URL loading." + "reason": "Sliding-window behavior in Qwen3-MoE has nothing to do with loading AutoImageProcessor from URL." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "gguf_file loading regression is unrelated to Blip2 dtype propagation staying float32." + "reason": "gguf_file loading breakage and BLIP2 dtype propagation are different bugs." }, { "left": "issue:44625", "right": "issue:44877", "accept": false, - "reason": "Qwen3.5 num_labels propagation and granite_speech strict config loading are different config bugs." + "reason": "num_labels propagation and strict config loading are not the same underlying issue." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Both concern loading flows, but one is gguf_file breakage and the other is Mistral regex patch forcing model_info offline; different root causes." + "reason": "Different tokenizer/processor loading paths; offline model_info bug is unrelated to gguf_file breakage." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Auxiliary-loss normalization bug is unrelated to native DeepseekV3MoE diverging from remote implementation." + "reason": "Both are model correctness issues, but they affect different architectures and subsystems." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "Different tokenizer special-token regressions in different models; not the same bug." + "reason": "Different tokenizer special-token problems with different tokenizers and symptoms." }, { "left": "issue:43720", "right": "issue:44898", "accept": false, - "reason": "BitNet packed-weight loading and Perceiver interpolation are unrelated code paths." + "reason": "BitNet accelerate-loading unpacking bug and Perceiver image classification failure are unrelated." }, { "left": "issue:41762", "right": "issue:44936", "accept": false, - "reason": "DeepSpeed ZeRO-3 loading crash is unrelated to trainer.evaluate() failing after train()." + "reason": "DeepSpeed ZeRO-3 loading failure and trainer.evaluate() after train() failure are separate bugs." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Both are regressions, but one is processor.tokenizer/image-tokenizer behavior and the other is Kimi-K2.5 tokenizer codec handling and warning text." + "reason": "Processor/tokenizer API regression and Kimi-K2.5 codec/mistral-regex regression are different issues." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC from_latents/forward mismatch is unrelated to Whisper batch_decode skip_special_tokens handling." + "reason": "DAC latent decoding mismatch and Whisper batch_decode skip_special_tokens behavior are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted pair, issue:43493 vs issue:43994, is plausibly the same SigLIP2 bug at different levels of specificity, and the other verdicts do not overclaim beyond the titles/summaries provided." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -5520,10 +5497,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:24643", "issue:28282", @@ -5540,11 +5517,13 @@ "issue:43262", "issue:43295", "issue:43329", + "issue:43334", "issue:43452", "issue:43504", "issue:43525", "issue:43531", "issue:43550", + "issue:43572", "issue:43577", "issue:43643", "issue:43653", @@ -5560,7 +5539,6 @@ "issue:44466", "issue:44479", "issue:44554", - "issue:44610", "issue:44625", "issue:44743", "issue:44843", @@ -5579,11 +5557,11 @@ "issue:24643|issue:29127", "issue:43525|issue:43643", "issue:41720|issue:43906", - "issue:38175|issue:43295", "issue:43577|issue:43697", + "issue:38175|issue:43295", "issue:44361|issue:44442", - "issue:43653|issue:44361", "issue:44843|issue:44991", + "issue:43653|issue:44361", "issue:44479|issue:44936", "issue:43122|issue:43906", "issue:44977|issue:45356", @@ -5594,246 +5572,240 @@ "issue:43577|issue:45588", "issue:43531|issue:43742", "issue:43577|issue:44554", - "issue:43653|issue:43927", "issue:44464|issue:44977", + "issue:43653|issue:43927", "issue:43329|issue:45325", - "issue:43550|issue:44361", "issue:44479|issue:44625", - "issue:43504|issue:43720", - "issue:43504|issue:43550", + "issue:43550|issue:44361", "issue:43452|issue:44451", "issue:41720|issue:44464", - "issue:32090|issue:34567", + "issue:43504|issue:43720", + "issue:43504|issue:43550", "issue:32090|issue:33666", - "issue:43504|issue:44610", + "issue:32090|issue:34567", + "issue:43334|issue:43572", "issue:43262|issue:45084" ], - "prepared_review_unit_hash": "8dc3b0cc816ca625e6ea418ab15b92107d1d2d8d7cf32ed83a2a7a64907cbf92", + "prepared_review_unit_hash": "e63e0114647d63a010600a30b7caaf2c92f304c113def96e5f22a6fc518c378e", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 33, - "serialized_chars": 37303, - "estimated_input_tokens": 9326, - "estimated_eval_tokens": 18908 + "serialized_chars": 37914, + "estimated_input_tokens": 9479, + "estimated_eval_tokens": 19214 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose similarity bucket of mostly unrelated Transformers bug reports and regressions across tokenization, config loading, multimodal processing, training, and generation. The items share broad subsystem themes but do not look like duplicate reports of the same underlying defect.", - "confidence": 0.93, - "canonical_issue_reason": "No single canonical issue stands out because the issues are heterogeneous and mostly describe different failure modes, code paths, and model families.", + "summary": "These items are a heterogeneous set of unrelated bug reports across tokenizer loading, config propagation, vision/multimodal preprocessing, trainer, and generation. The soft edges mostly share only broad subsystem similarity, not the same concrete underlying bug, so none should be merged.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out: the set spans many distinct failure modes and architectures, and the apparent overlaps are mostly broad-topic false positives rather than duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44843 is the clearest standalone report: it has a specific regression, a narrow code path (`_patch_mistral_regex` / `AutoTokenizer.from_pretrained`), and an actionable offline-mode failure. It is a good representative issue, but not a true cluster canonical.", + "best_issue_reason": "No global best issue for deduplication; any one choice would misrepresent the rest because the reported bugs are materially different.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44466", "right": "issue:45005", "accept": false, - "reason": "Both concern tied-weight serialization, but they describe different manifestations and model classes; not the same concrete bug." + "reason": "Both mention tied weights, but one is a serialization/device-dependent lm_head.weight bug and the other is a translation-model loading issue in v5; different failure modes." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Both mention generation/caching in Qwen3.5 code, but one is compiled-forward inconsistency and the other is recurrent-state reset logic; different bugs." + "reason": "Both concern generation/caching, but one is inconsistent chunked generation under compiled forward and the other is recurrent states resetting with cache; not the same bug." }, { "left": "issue:24643", "right": "issue:29127", "accept": false, - "reason": "Training/DeepSpeed weight-shape runtime error vs LayoutLMv3 box validation error; unrelated failure modes." + "reason": "DeepSpeed training weight-dimension RuntimeError and LayoutLMv3 box-validation messaging are unrelated." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Both involve missing config fields, but one is a Llama4Config attribute bug and the other is `trust_remote_code` returning incomplete configs; not the same underlying defect." + "reason": "Both involve missing config fields, but one is Llama4 pad_token_id access and the other is trust_remote_code returning incomplete AutoConfig; different code paths and symptoms." }, { "left": "issue:41720", "right": "issue:43906", "accept": false, - "reason": "Both are about model loading/runtime issues, but one is a Qwen3 auto device-mapping CUDA assert and the other is a reproduction of a different issue; no shared code-path bug." + "reason": "Qwen3 auto device mapping CUDA assert and an isolated reproduction of another issue are not enough to establish the same bug." }, { - "left": "issue:38175", - "right": "issue:43295", + "left": "issue:43577", + "right": "issue:43697", "accept": false, - "reason": "Siglip2 zero-probabilities vs processor/tokenizer access regression; different behavior and different fix surface." + "reason": "Blip2 dtype remaining float32 and RTDetrV2 output drift in v5 are separate model behavior bugs." }, { - "left": "issue:43577", - "right": "issue:43697", + "left": "issue:38175", + "right": "issue:43295", "accept": false, - "reason": "Blip2 dtype propagation bug vs RTDetrV2 output mismatch in v5; different models and different correctness issues." + "reason": "Siglip2 zero probabilities and processor.tokenizer regression are unrelated multimodal model issues." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Both are tokenizer-related load/task failures, but they affect different tokenizers and different error paths." + "reason": "Both are tokenizer-related load failures, but for different tokenizer classes and different error shapes." }, { - "left": "issue:43653", - "right": "issue:44361", + "left": "issue:44843", + "right": "issue:44991", "accept": false, - "reason": "BigBird mask-token special-token registration bug vs MLukeTokenizer task AttributeError; unrelated tokenizer defects." + "reason": "Both affect tokenizer loading, but one is an offline model_info call regression and the other is a specific tokenizer load failure for EMBEDDIA/est-roberta." }, { - "left": "issue:44843", - "right": "issue:44991", + "left": "issue:43653", + "right": "issue:44361", "accept": false, - "reason": "Offline `model_info()` call in tokenizer patching vs tokenizer loading failure for a specific model; not the same regression." + "reason": "BigBirdTokenizer special-token registration and MLukeTokenizer AttributeError are unrelated tokenizer defects." }, { "left": "issue:44479", "right": "issue:44936", "accept": false, - "reason": "Video-input regression for Qwen VL models vs trainer evaluate/train interaction; different subsystems and symptoms." + "reason": "Video-input regression in Qwen VL models and trainer.evaluate() failing after train are different subsystems and bugs." }, { "left": "issue:43122", "right": "issue:43906", "accept": false, - "reason": "Tokenizer output changed across versions vs isolated reproduction of a different issue; no same underlying bug." + "reason": "Different tokenization across versions and an isolated reproduction of another issue do not indicate the same concrete defect." }, { "left": "issue:44977", "right": "issue:45356", "accept": false, - "reason": "Both are Qwen3.5 tokenizer/generation regressions, but one is flash-attention generation and the other is codec/warning behavior; not the same bug." + "reason": "Qwen3.5 flash-attention generation failure and Kimi-K2.5 tokenizer regression are unrelated." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "PyTorch missing import error vs Trainer `_gpu_broadcast_one` NoneType error; completely different causes." + "reason": "Missing PyTorch import and a NoneType broadcast error in Trainer are distinct errors." }, { "left": "issue:28282", "right": "issue:33357", "accept": false, - "reason": "Missing PyTorch import vs Mac bus error on a community CLIP model; unrelated." + "reason": "AutoModel import failure without PyTorch and MacOS bus error in CLIP are different problems." }, { "left": "issue:28282", "right": "issue:35141", "accept": false, - "reason": "Missing PyTorch import vs tied embedding reinitialization after resize; different code paths." + "reason": "ImportError for missing torch and output embedding reinitialization after resize are unrelated." }, { "left": "issue:30064", "right": "issue:33666", "accept": false, - "reason": "Void segmentation-map processing vs Qwen2-VL multi-GPU training; only broad vision-model similarity." + "reason": "Void segmentation map processing and Qwen2-VL multi-GPU training are different areas and failure modes." }, { "left": "issue:43577", "right": "issue:45588", "accept": false, - "reason": "Both involve model loading/runtime behavior, but one is dtype propagation and the other is flash-attention crashing on `s_aux=None`; different bugs." + "reason": "Blip2 dtype handling and flash_attention crash on s_aux=None are not the same bug." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE vs key error loading MobileLLM-125M; unrelated issues." + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM load key error are unrelated." }, { "left": "issue:43577", "right": "issue:44554", "accept": false, - "reason": "Blip2 dtype bug vs MPS attention correctness issue; different models and runtime backends." - }, - { - "left": "issue:43653", - "right": "issue:43927", - "accept": false, - "reason": "BigBird tokenizer special-token registration vs DiaConfig custom token IDs lost on save/load; different tokenizer/config problems." + "reason": "Blip2 precision loading and an MPS attention correctness issue with mismatched head dims are separate defects." }, { "left": "issue:44464", "right": "issue:44977", "accept": false, - "reason": "Both are generation regressions, but one is chunked generation under compiled forward and the other is flash-attention on Qwen3.5; not the same bug." + "reason": "Chunked generation under compiled forward and Qwen3.5 flash-attention generation are different generation bugs." }, { - "left": "issue:43329", - "right": "issue:45325", + "left": "issue:43653", + "right": "issue:43927", "accept": false, - "reason": "Undefined video-token helper/merge_size in multimodal token counting vs incorrect temporal scaling in rope index; both video-related but different code paths." + "reason": "BigBirdTokenizer decode output issue and DiaConfig custom token ID loss after save/load are different tokenizer/config bugs." }, { - "left": "issue:43550", - "right": "issue:44361", + "left": "issue:43329", + "right": "issue:45325", "accept": false, - "reason": "torch.compile SDPA failure in Bamba vs MLukeTokenizer AttributeError; unrelated." + "reason": "Both are multimodal token-indexing issues, but one is an undefined video-branch implementation bug and the other is still-image temporal scaling in rope index; not the same root cause." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Video-input regression in Qwen VL models vs num_labels propagation in Qwen3.5 config; different issues." - }, - { - "left": "issue:43504", - "right": "issue:43720", - "accept": false, - "reason": "Legacy-field load failure for Beit semantic segmentation vs packed-weight unpacking failure for BitNet; different loading bugs." + "reason": "Video regression in several Qwen VL models and Qwen3.5 num_labels propagation are unrelated." }, { - "left": "issue:43504", - "right": "issue:43550", + "left": "issue:43550", + "right": "issue:44361", "accept": false, - "reason": "Both are model-loading/runtime regressions, but they affect different architectures and failure mechanisms." + "reason": "torch.compile+SDPA failure in Bamba-9B-v2 and tokenizer AttributeError in MLukeTokenizer are different subsystems." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "gguf_file support failure for tokenizer/model loading vs inability to load a specific model checkpoint; not the same defect." + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and inability to load a specific model are both loading problems, but not the same concrete bug." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "CUDA assert under auto device mapping vs chunked-generation inconsistency; different runtime problems." + "reason": "CUDA assert on Qwen3 auto device mapping and compiled chunked generation inconsistency are unrelated runtime failures." }, { - "left": "issue:32090", - "right": "issue:34567", + "left": "issue:43504", + "right": "issue:43720", "accept": false, - "reason": "NoneType broadcast in Trainer vs TrainerState token counter not updating; separate training bugs." + "reason": "Legacy field handling in Beit semantic segmentation and BitNet packed-weight unpacking during accelerate loading are distinct model-loading issues." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Trainer broadcast error vs Qwen2-VL multi-GPU training request; not the same failure." + "reason": "Trainer broadcast NoneType error and Qwen2-VL multi-GPU training request are not the same bug." }, { - "left": "issue:43504", - "right": "issue:44610", + "left": "issue:32090", + "right": "issue:34567", + "accept": false, + "reason": "Trainer GPU broadcast error and TrainerState.num_input_tokens_seen not updating are different trainer issues." + }, + { + "left": "issue:43334", + "right": "issue:43572", "accept": false, - "reason": "Legacy-field load issue for Beit vs OmDet-Turbo processor/model input-size mismatch; different model-specific bugs." + "reason": "Both mention missing pad-token-related fields, but they affect different configs and models; no shared code-path evidence." }, { "left": "issue:43262", "right": "issue:45084", "accept": false, - "reason": "Audio chat-template sampling-rate default bug vs template-node compilation type error; unrelated processor/compiler issues." + "reason": "Audio processor chat-template sample-rate default and template compilation TypeError are unrelated." } ] }, @@ -5852,10 +5824,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:33666", "issue:36010", @@ -5863,10 +5835,9 @@ "issue:39692", "issue:43257", "issue:43278", - "issue:43295", "issue:43299", - "issue:43334", "issue:43493", + "issue:43504", "issue:43525", "issue:43540", "issue:43572", @@ -5890,6 +5861,7 @@ "issue:44488", "issue:44521", "issue:44589", + "issue:44610", "issue:44625", "issue:44734", "issue:44821", @@ -5907,28 +5879,28 @@ "issue:45584" ], "soft_pairs": [ + "issue:43504|issue:44610", "issue:43278|issue:43688", - "issue:43334|issue:43572", "issue:44849|issue:45357", "issue:43756|issue:44913", "issue:43525|issue:43572", "issue:43540|issue:43716", - "issue:43493|issue:43825", "issue:43742|issue:44821", + "issue:43493|issue:43825", "issue:33666|issue:36010", "issue:45200|issue:45229", + "issue:43688|issue:43873", "issue:43720|issue:43927", "issue:43606|issue:43927", - "issue:43688|issue:43873", "issue:43257|issue:43299", "issue:43278|issue:43844", - "issue:43720|issue:44361", "issue:39692|issue:43493", "issue:44488|issue:44843", "issue:43540|issue:44977", + "issue:43720|issue:44361", "issue:44964|issue:45405", - "issue:43493|issue:43844", "issue:33666|issue:36331", + "issue:43493|issue:43844", "issue:43756|issue:44060", "issue:44625|issue:45584", "issue:44589|issue:45520", @@ -5936,223 +5908,216 @@ "issue:44521|issue:45290", "issue:44423|issue:44734", "issue:43824|issue:44589", - "issue:43688|issue:43746", - "issue:39692|issue:43295" + "issue:43688|issue:43746" ], - "prepared_review_unit_hash": "7f989cd5fe6a50068bf2a952a26e9beee04c8702f9e6285c4277584333876afa", + "prepared_review_unit_hash": "bb1b6cd3da4e2bede23b035b6c1d01b64aae75136ee2e23a10346537d78235cb", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37119, - "estimated_input_tokens": 9280, - "estimated_eval_tokens": 18816 + "soft_pair_count": 30, + "serialized_chars": 36766, + "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 a loose collection of mostly unrelated Transformers bug reports spanning different models, regressions, loading paths, and training/runtime issues. None of the soft pairs look like true duplicates or the same underlying change.", + "summary": "These items are a loose similarity cluster, not a duplicate set: they cover many unrelated models, configs, and failure modes. No soft pair looks like the same underlying bug/change, so all candidate edges are rejected.", "confidence": 0.96, - "canonical_issue_reason": "No clear canonical issue: the items mix unrelated model-specific bugs, config regressions, training issues, and docs/runtime failures rather than one duplicate report.", + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No issue cleanly represents the whole cluster; 43646 is the broadest regression-style report, but the set is too heterogeneous for a strong canonical representative.", + "best_issue_reason": "No single canonical issue is appropriate because the cluster is heterogeneous; the closest-to-broad regression report is 43646, but it still does not subsume the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43278", - "right": "issue:43688", + "left": "issue:43504", + "right": "issue:44610", "accept": false, - "reason": "Different bugs: BF16/FP32 drift in eval vs auxiliary-loss normalization for MoE models." + "reason": "Different models and bugs: Beit legacy-field loading vs OmDet processor size mismatch." }, { - "left": "issue:43334", - "right": "issue:43572", + "left": "issue:43278", + "right": "issue:43688", "accept": false, - "reason": "Both mention missing pad-token fields, but they affect different config classes and different attributes (`pad_token_id` vs `pad_token_idx`)." + "reason": "Unrelated problems: dtype drift between train/eval vs auxiliary loss normalization." }, { "left": "issue:44849", "right": "issue:45357", "accept": false, - "reason": "Qwen3.5 hidden-states generation bug is unrelated to incorrect visual-encoder keys on save." + "reason": "Same family (Qwen3.5) but different failures: hidden-states output vs save_pretrained visual-encoder keys." }, { "left": "issue:43756", "right": "issue:44913", "accept": false, - "reason": "Different config issues: RoPE-layer dropping in Smollm3 vs `rotary_pct` not persisting in GPTNeoX." + "reason": "Both involve rotary/RoPE config, but one is a layer-drop mismatch and the other is reload/default persistence." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Similar symptom class, but different models and different missing config fields; not the same concrete bug." + "reason": "Similar symptom class, but different configs/fields and models: pad_token_id vs pad_token_idx." }, { "left": "issue:43540", "right": "issue:43716", "accept": false, - "reason": "Qwen3OmniMoe video-input validation is unrelated to Mistral-3 image-preprocessor dtype mismatch." + "reason": "Different multimodal code paths and modalities: Qwen3OmniMoe video processing vs Mistral-3 image dtype mismatch." }, { - "left": "issue:43493", - "right": "issue:43825", + "left": "issue:43742", + "right": "issue:44821", "accept": false, - "reason": "SigLIP2 implementation drift is unrelated to a pipeline error-message regression." + "reason": "Different loading failures: MobileLLM key error vs AutoImageProcessor URL loading." }, { - "left": "issue:43742", - "right": "issue:44821", + "left": "issue:43493", + "right": "issue:43825", "accept": false, - "reason": "Different loading paths: MobileLLM key lookup failure vs loading `AutoImageProcessor` from URL." + "reason": "SigLIP2 implementation discrepancy is unrelated to a pipeline translation error-message regression." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training issue and `GenerationMixin` import failure are unrelated." + "reason": "Unrelated: multi-GPU training for Qwen2-VL vs GenerationMixin import error." }, { "left": "issue:45200", "right": "issue:45229", "accept": false, - "reason": "One is a missing default for `mm_token_type_ids`; the other is a multi-GPU inference OOM." + "reason": "Different issues in Gemma 4: text-only mm_token_type_ids defaults vs multi-GPU inference CUDA OOM." }, { - "left": "issue:43720", - "right": "issue:43927", + "left": "issue:43688", + "right": "issue:43873", "accept": false, - "reason": "Packed-weight unpacking during accelerate loading is unrelated to config token IDs being lost on save/load." + "reason": "Different subsystems: auxiliary-loss normalization vs quantization/offloading behavior." }, { - "left": "issue:43606", + "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "CPU offload device-mismatch bug is not the same as custom token IDs disappearing after serialization." + "reason": "Different models and failure modes: packed-weight accelerate loading vs config token-ID persistence." }, { - "left": "issue:43688", - "right": "issue:43873", + "left": "issue:43606", + "right": "issue:43927", "accept": false, - "reason": "Aux-loss normalization and quantization/offloading behavior are different bugs." + "reason": "CPU offload device mismatch is unrelated to DiaConfig save/load token-ID loss." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Both concern Qwen3 MoE loading, but the concrete failure modes and model variants differ enough to avoid merging." + "reason": "Both are Qwen3 MoE-related, but one is deepspeed conversion/loading and the other is a v5 model-load break." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "Training/eval dtype mismatch is unrelated to the ZeRO-3 gradient growth report." - }, - { - "left": "issue:43720", - "right": "issue:44361", - "accept": false, - "reason": "Packed-weight loading failure is unrelated to an `AttributeError` in `MLukeTokenizer` task handling." + "reason": "Different bugs: BF16/FP32 dtype drift vs DeepSpeed gradient growth." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "Documentation example errors are not the same bug as the SigLIP2 HF-vs-JAX implementation discrepancy." + "reason": "Documentation example errors are not the same bug as SigLIP2 implementation mismatch." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "Model loading failure and offline-mode breakage from an unconditional hub call are different root causes." + "reason": "Different causes: a specific model failing to load vs offline-mode break from unconditional hub metadata lookup." }, { "left": "issue:43540", "right": "issue:44977", "accept": false, - "reason": "Video-input processing failure is unrelated to Qwen3.5 flash-attention generation issues." + "reason": "Both are Qwen3.5-related, but one is video-input processing and the other is flash-attention generation behavior." }, { - "left": "issue:44964", - "right": "issue:45405", + "left": "issue:43720", + "right": "issue:44361", "accept": false, - "reason": "Model-load regression and a PEFT version-pin problem are not the same underlying bug." + "reason": "Different components and failures: BitNet packed weights vs MLukeTokenizer AttributeError." }, { - "left": "issue:43493", - "right": "issue:43844", + "left": "issue:44964", + "right": "issue:45405", "accept": false, - "reason": "SigLIP2 implementation mismatch does not overlap with the DeepSpeed gradient anomaly." + "reason": "Model loading regression vs dependency version bump; not the same underlying issue." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Multi-GPU training support and the `compute_loss()` signature change are distinct issues." + "reason": "Unrelated training and trainer API issues." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 discrepancy is unrelated to DeepSpeed/gradient magnitude behavior." }, { "left": "issue:43756", "right": "issue:44060", "accept": false, - "reason": "Smollm3 RoPE-layer handling is unrelated to the Qwen3-Next tied-weights warning." + "reason": "Both mention model internals, but one is RoPE-layer dropping and the other is tied-weights warning corruption." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "`num_labels` propagation in Qwen3.5 is unrelated to Whisper failing on empty transcription after token alignment." + "reason": "Different models and symptoms: Qwen3.5 num_labels propagation vs Whisper empty-transcription generation failure." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage lookup failure and a missing `flash_attn` import key on Python 3.13 are different errors." + "reason": "Different incompatibilities: missing Float8 storage type vs flash_attn import-utils KeyError on Python 3.13." }, { "left": "issue:43646", "right": "issue:44466", "accept": false, - "reason": "General custom-model initialization regression and device-dependent `lm_head.weight` serialization are not the same fix." + "reason": "Both are v5 regressions, but one is custom model initialization and the other is lm_head serialization depending on device." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "Both are `apply_chat_template`-adjacent, but one is multimodal assistant-mask generation and the other is a tool-call crash." + "reason": "Same API area, but different bugs: multimodal assistant masks vs tool-call/no-content crash." }, { "left": "issue:44423", "right": "issue:44734", "accept": false, - "reason": "Serve crashes in different code paths: multimodal continuous batching vs KV-cache continuation indexing." + "reason": "Both are serving crashes, but the code paths differ: multimodal continuous batching vs KV-cache continuation indexing." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "Importing a specific Qwen2.5-VL class is unrelated to the Float8 storage-object error." + "reason": "Import error for a specific model class is unrelated to the Float8 storage lookup failure." }, { "left": "issue:43688", "right": "issue:43746", "accept": false, - "reason": "Aux-loss normalization and PEFT-adapter checkpoint loading are unrelated bugs." - }, - { - "left": "issue:39692", - "right": "issue:43295", - "accept": false, - "reason": "SigLIP2 docs mistakes are not the same as the processor/tokenizer regression in v4.57.5." + "reason": "Different model families and issues: loss normalization vs PEFT/local-checkpoint loading." } ] }, @@ -6171,10 +6136,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:15354", "issue:29942", @@ -6184,6 +6149,7 @@ "issue:37428", "issue:38175", "issue:39401", + "issue:39692", "issue:43232", "issue:43278", "issue:43295", @@ -6218,7 +6184,6 @@ "issue:45084", "issue:45198", "issue:45200", - "issue:45325", "issue:45356", "issue:45405", "issue:45446", @@ -6226,14 +6191,15 @@ "issue:45584" ], "soft_pairs": [ + "issue:39692|issue:43295", "issue:33357|issue:33666", "issue:43572|issue:44291", - "issue:43504|issue:43606", "issue:43994|issue:44493", + "issue:43504|issue:43606", + "issue:44464|issue:45584", "issue:43575|issue:45405", "issue:43927|issue:44361", "issue:43504|issue:43653", - "issue:44464|issue:45584", "issue:43278|issue:43701", "issue:44792|issue:45405", "issue:43606|issue:44112", @@ -6251,199 +6217,198 @@ "issue:44315|issue:44849", "issue:44554|issue:45356", "issue:43334|issue:43525", - "issue:43295|issue:44610", - "issue:45200|issue:45325" + "issue:43295|issue:44610" ], - "prepared_review_unit_hash": "d9b0981fd619ba28f960adb4962cda8ec2a8f1024590c7cc42109d2a4a4207b6", + "prepared_review_unit_hash": "5bac5679595785c57eb405b2946199a2f8dcab5f010d59530ed496d861ed24a0", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35867, - "estimated_input_tokens": 8967, - "estimated_eval_tokens": 18190 + "serialized_chars": 35856, + "estimated_input_tokens": 8964, + "estimated_eval_tokens": 18184 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a mixed bag of unrelated transformer bugs and regressions (tokenizers, multimodal processors, loading, dtype, generation, CI, and environment-specific failures). I don\u2019t see a true duplicate center, so all soft edges should be rejected.", - "confidence": 0.97, - "canonical_issue_reason": "Issue 43957 is the broadest and most representative loading-regression report in the set, but the cluster is too heterogeneous for a true duplicate canonical.", + "summary": "The set is heterogeneous: most items are distinct bug reports across unrelated models/features, and none of the soft pairs look like the same underlying issue. No clear canonical issue/PR emerges.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 43957 is the best single representative because it describes a generic model-loading regression affecting multiple models and versions, which is the closest thing to a central theme here.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:33357", - "right": "issue:33666", + "left": "issue:39692", + "right": "issue:43295", "accept": false, - "reason": "Different models, platforms, and bug classes: MacOS bus error on CLIP vs multi-GPU training support for Qwen2-VL." + "reason": "Both involve processor/tokenizer behavior, but one is a SigLIP2 docs-example failure and the other is a v4.57.5 regression in custom code; different root causes." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training; unrelated models and failure modes." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "Both touch loading/config, but one is a missing pad token field after a version update and the other is an unexpected _is_hf_initialized argument in init_empty_weights; not the same bug." + "reason": "Both are 5.x compatibility/loading problems, but they hit different APIs and errors (`pad_token_idx` vs `_is_hf_initialized`); not the same bug." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 wrong outputs and a generic position-id key warning are different symptoms with no shared concrete code path." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Unrelated failures: BEiT legacy field load problem vs suno/bark CPU offload device mismatch." + "reason": "BEiT legacy-field loading failure and Bark CPU-offload device mismatch are unrelated." }, { - "left": "issue:43994", - "right": "issue:44493", + "left": "issue:44464", + "right": "issue:45584", "accept": false, - "reason": "Different symptoms and code paths: SigLIP2 nonsensical inference vs unexpected position-id keys in many models." + "reason": "Chunked generation with compiled forward vs Whisper empty-transcription handling after token alignment; different generation paths." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "OOM during tensor-parallel loading is unrelated to a PyPI dependency version bump." + "reason": "Tensor-parallel OOM for a large model vs an unreleased dependency version bump; no overlap in bug/change." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "Both mention token-related failures, but one is a DiaConfig save/load custom-token-ID bug and the other is an MLukeTokenizer AttributeError on tasks." + "reason": "DiaConfig save/load token-ID loss and MLukeTokenizer task-time AttributeError are separate tokenizer/config problems." }, { "left": "issue:43504", "right": "issue:43653", "accept": false, - "reason": "Different tokenizer/model problems: BEiT legacy-field loading vs BigBirdTokenizer mask token registration and empty decode output." - }, - { - "left": "issue:44464", - "right": "issue:45584", - "accept": false, - "reason": "Different generation bugs: compiled-forward chunked generation inconsistency vs Whisper empty-transcription handling after align_special_tokens." + "reason": "Different models and failures: BEiT legacy-field load issue vs BigBird mask-token registration/decoding bug." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "One is a dtype change between train and eval, the other is a resume_from_checkpoint key mismatch; not the same underlying issue." + "reason": "Embedding dtype drift in evaluate vs checkpoint resume key mismatch are unrelated." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "A failed Janus image-generation test is unrelated to a package-version availability issue." + "reason": "Janus image generation test failure vs PEFT version pin issue; no common underlying defect." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "Both involve devices in CI, but the concrete failures differ: CPU offload mismatch vs a stale device-override test in GraniteSpeech." + "reason": "CPU offload device mismatch in Bark vs a stale GraniteSpeech CI device-override test; not the same bug." }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to MLFlow reporting when a training job fails." + "reason": "Flash Attention 2 test failures and MLFlow-run state reporting are completely different subsystems." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript export/generator-expression support is unrelated to SigLIP2 zero-probability inference." + "reason": "TorchScript export limitation for GeneratorExp vs SigLIP2 zero probabilities; different models and issues." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "An MPS attention correctness issue is unrelated to a PyTorch version-check bug for AuxRequest import." + "reason": "MPS attention correctness bug vs a PyTorch version-check/import bug in flex_attention; unrelated code paths." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Both are multimodal, but one is video-input validation in Qwen3OmniMoe and the other is Gemma 4 missing mm_token_type_ids defaults." + "reason": "Video input processing in Qwen3OmniMoe vs Gemma 4 token-type-id defaulting; different features and failures." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "TypeError compiling non-template nodes is unrelated to a flash_attn import KeyError on Python 3.13." + "reason": "Template-node compilation error vs Python 3.13 flash_attn import KeyError; unrelated runtime problems." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "ZeRO-3 gradient inflation during training is unrelated to a latest-transformers model-loading failure for Phi-4 multimodal." + "reason": "DeepSpeed ZeRO-3 gradient growth vs Phi-4 multimodal load failure; different models and symptoms." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Both are tokenizer issues, but they affect different models and different behaviors: offset_mapping vs add_special_tokens not adding BOS/EOS." + "reason": "Qwen3 offset_mapping bug vs mdeberta BOS/EOS insertion regression; both tokenizer issues but not the same concrete defect." }, { "left": "issue:43582", "right": "issue:44977", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError is unrelated to Qwen3.5 flash-attention generation problems." + "reason": "Apple Silicon allocator warmup TypeError and Qwen3.5 flash-attention generation failure are unrelated." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Generation key handling after sync_gpus is a different bug from Qwen3.5 output_hidden_states behavior." + "reason": "Generation kwarg update after sync_gpus vs output_hidden_states bug in Qwen3.5; different generation code paths." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "dtype mismatch in bfloat16 inference is not the same as Wav2Vec2 save_pretrained/tokenization failure." + "reason": "bfloat16 dtype mismatches in inference vs Wav2Vec2 save/tokenization failure; no shared underlying bug." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Both involve loading/import regressions, but one is a missing flash_supports_window_size import and the other is a meta-device model-loading breakage; different root causes and fixes." + "reason": "Flash-attention import error vs meta-device model loading breakage; distinct failures." }, { "left": "issue:44315", "right": "issue:44849", "accept": false, - "reason": "Liger Kernel application with model_init is unrelated to Qwen3.5 output_hidden_states." + "reason": "Liger Kernel not applied during model_init and Qwen3.5 hidden-states generation bug are unrelated." }, { "left": "issue:44554", "right": "issue:45356", "accept": false, - "reason": "MPS attention correctness is unrelated to Kimi-K2.5 tokenizer codec handling and warning behavior." + "reason": "MPS attention correctness issue vs Kimi-K2.5 tokenizer codec regression; different components and root causes." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Both are pad_token_id-related attribute errors, but they affect different configs/models and are not clearly the same bug." + "reason": "Both mention missing pad_token_id, but they affect different configs/models and appear to be separate loading bugs." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "A processor.tokenizer regression is unrelated to OmDet-Turbo producing 640px inputs instead of the expected 224px." - }, - { - "left": "issue:45200", - "right": "issue:45325", - "accept": false, - "reason": "Gemma 4 mm_token_type_ids defaults and Qwen2.5-VL rope-index temporal scaling are different multimodal bugs with different code paths." + "reason": "Processor/tokenizer API regression vs OmDet-Turbo image-size mismatch; unrelated symptoms and fixes." } ] }, @@ -6462,10 +6427,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:37428", "issue:38175", @@ -6491,7 +6456,6 @@ "issue:44206", "issue:44360", "issue:44442", - "issue:44462", "issue:44464", "issue:44466", "issue:44514", @@ -6508,16 +6472,19 @@ "issue:45003", "issue:45127", "issue:45198", + "issue:45200", "issue:45245", "issue:45290", + "issue:45325", "issue:45381", "issue:45412", "issue:45563", "issue:45584" ], "soft_pairs": [ - "issue:43493|issue:43701", + "issue:45200|issue:45325", "issue:44466|issue:45127", + "issue:43493|issue:43701", "issue:44898|issue:45412", "issue:43873|issue:44464", "issue:44792|issue:45563", @@ -6530,241 +6497,237 @@ "issue:43824|issue:43975", "issue:43295|issue:43450", "issue:43295|issue:43873", - "issue:44442|issue:45198", "issue:44360|issue:44561", + "issue:44442|issue:45198", "issue:38175|issue:45412", "issue:43295|issue:43994", "issue:42915|issue:43824", - "issue:43278|issue:43856", "issue:38175|issue:43493", + "issue:43278|issue:43856", "issue:43827|issue:43901", "issue:44849|issue:45003", "issue:44568|issue:44991", - "issue:43653|issue:44112", "issue:44206|issue:45584", + "issue:43653|issue:44112", "issue:43450|issue:43493", "issue:45290|issue:45381", - "issue:45127|issue:45245", - "issue:44462|issue:44561" + "issue:45127|issue:45245" ], - "prepared_review_unit_hash": "39fa84fb9727b15d9c094587b611168e0587a3af21acca98f3e40db3b128581f", + "prepared_review_unit_hash": "879606e3e90ea6998c436e4063c4feca00ef48c27909f404f50a0e9731b56873", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36152, - "estimated_input_tokens": 9038, - "estimated_eval_tokens": 18332 + "serialized_chars": 36805, + "estimated_input_tokens": 9202, + "estimated_eval_tokens": 18660 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a mix of unrelated bugs; the only strong duplicate-like pair is the SigLIP2 implementation/output regression. All other soft edges are same-subsystem or same-feature, but not the same concrete bug/change.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43994 is the broadest SigLIP2 report: it names the exact model, covers bad outputs through both AutoModel and pipeline, and best subsumes the narrower SigLIP2 symptom reports.", + "summary": "This cluster is highly heterogeneous: most soft pairs only share a subsystem or wording, not the same underlying defect. I would not merge any of the candidate pairs, and there is no safe single canonical issue for the whole set.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue fits: the items span unrelated bugs across multimodal processing, tokenizers, training, generation, and import/runtime failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43994 is the most representative issue in the only apparent duplicate family, with the clearest end-user failure description and broadest symptom coverage.", + "best_issue_reason": "No issue is a safe representative because the soft pairs do not describe one shared bug/change; they are separate regressions in different code paths.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43493", - "right": "issue:43701", + "left": "issue:45200", + "right": "issue:45325", "accept": false, - "reason": "SigLIP2 model regression vs. a checkpoint resume key mismatch; different code paths and failure modes." + "reason": "Different defects: Gemma 4 token-type defaults vs Qwen2.5-VL RoPE/temporal position scaling." }, { "left": "issue:44466", "right": "issue:45127", "accept": false, - "reason": "Serialization of tied lm_head weights vs. LoRA merge collapse with extended vocab; not the same underlying bug." + "reason": "One is lm_head.weight serialization varying by device; the other is LoRA merge collapse with extended vocab and tied weights." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 implementation mismatch and resume-from-checkpoint key mismatch are unrelated failure modes." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Perceiver interpolation failure vs. RT-DETR memory not being released; unrelated problems." + "reason": "Perceiver interpolation failure and RT-DETR memory not being released are distinct bugs." }, { "left": "issue:43873", "right": "issue:44464", "accept": false, - "reason": "Quantization/offloading behavior vs. chunked generation with compiled forward; different features and symptoms." + "reason": "Quantization/offloading behavior is unrelated to chunked generation inconsistencies under compiled forward." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "Janus image-generation test failure vs. a stale paging warning; not the same concrete issue." + "reason": "Janus model test failure and a stale paged-generate warning are different issues." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but they refer to different missing symbols and different model families." + "reason": "Both are import errors, but for different symbols and different code paths." }, { "left": "issue:43388", "right": "issue:43526", "accept": false, - "reason": "Both involve label handling, but one is Trainer metric gathering and the other is BeitImageProcessorFast label reduction." + "reason": "Both involve labels, but one is gather_for_metrics dropping tuple elements and the other is BeitImageProcessorFast reduce_labels collapsing arrays." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both touch apply_chat_template, but one is a batched padding=False crash and the other is zero assistant masks for multimodal inputs." + "reason": "Both mention apply_chat_template, but one is a batching crash with padding=False and the other is incorrect assistant masks for multimodal inputs." }, { "left": "issue:43572", "right": "issue:44987", "accept": false, - "reason": "StableLm config regression vs. loading a specific model repository on v5.1.0; not the same bug." + "reason": "StableLm config pad_token_idx regression is unrelated to loading physical-intelligence/fast." }, { "left": "issue:43844", "right": "issue:44936", "accept": false, - "reason": "ZeRO-3 gradient growth vs. trainer.evaluate failing after train(); different failure modes." + "reason": "ZeRO-3 gradient growth and trainer.evaluate() failing after train() are different training bugs." }, { "left": "issue:43572", "right": "issue:44589", "accept": false, - "reason": "Missing pad_token_idx in a config vs. missing Float8 storage type; unrelated." + "reason": "Missing pad_token_idx in config and Float8 storage lookup errors are unrelated." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Different model-loading/import issues: Qwen VL class import vs. deepseek detokenization regression." + "reason": "Qwen2_5_VL import failure and DeepSeek detokenization bug are different model-specific issues." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Processor/tokenizer regression vs. batched video processor shape bug; same area but not the same underlying change." + "reason": "Processor/tokenizer regression and batched video shape error are separate multimodal bugs." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor API regression vs. quantization/offloading behavior; unrelated." + "reason": "Processor/tokenizer regression is unrelated to quantization/offloading behavior." }, { - "left": "issue:44442", - "right": "issue:45198", + "left": "issue:44360", + "right": "issue:44561", "accept": false, - "reason": "Different tokenizer/model save-load failures in different architectures." + "reason": "A missing ReLU in the DSA indexer is unrelated to removal of is_torch_fx_available." }, { - "left": "issue:44360", - "right": "issue:44561", + "left": "issue:44442", + "right": "issue:45198", "accept": false, - "reason": "DSA indexer architecture discussion vs. removal of is_torch_fx_available breaking trust_remote_code models; not the same bug." + "reason": "FastSpeech2Conformer tokenizer loading and Wav2Vec2 save/tokenization failures are not the same bug." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 probability issue vs. RT-DETR memory reclamation; unrelated." + "reason": "SigLIP2 zero-probabilities and RT-DETR GC/memory release issues are unrelated." }, { "left": "issue:43295", "right": "issue:43994", "accept": false, - "reason": "Custom processor/tokenizer regression vs. SigLIP2 bad-output regression; different code paths." + "reason": "Custom processor/tokenizer regression does not match the SigLIP2 nonsensical output issue." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3 MoE FP8 config failure vs. Qwen2.5-VL import error; different problems." + "reason": "Qwen3Moe FP8 config failure and Qwen2.5-VL import failure are unrelated." }, { - "left": "issue:43278", - "right": "issue:43856", + "left": "issue:38175", + "right": "issue:43493", "accept": false, - "reason": "Training/evaluate embedding dtype mismatch vs. Qwen3 MoE memory usage; not duplicates." + "reason": "Both involve SigLIP2 but one is zero probabilities while the other is HF vs JAX discrepancy; not the same concrete bug." }, { - "left": "issue:38175", - "right": "issue:43493", - "accept": true, - "reason": "Both point to the same SigLIP2 regression: incorrect model behavior in the HF implementation versus the expected outputs/implementation." + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Embedding dtype mismatch on evaluate and Qwen3 MoE memory usage are different training issues." }, { "left": "issue:43827", "right": "issue:43901", "accept": false, - "reason": "Both are docs regressions, but they mention different obsolete pipeline APIs and are separate documentation fixes." + "reason": "Both are docs cleanup around pipeline changes, but they cover different docs and different API mentions." }, { "left": "issue:44849", "right": "issue:45003", "accept": false, - "reason": "Qwen3.5 hidden-states bug vs. unsafe sys.modules access in modeling_utils; different issues." + "reason": "Qwen3.5 hidden_states bug and unsafe sys.modules access are unrelated." }, { "left": "issue:44568", "right": "issue:44991", "accept": false, - "reason": "Tokenizer special-token handling for mdeberta-v3 vs. a different tokenizer load failure; not the same bug." + "reason": "Different tokenizer regressions for different models; not one shared bug." }, { - "left": "issue:43653", - "right": "issue:44112", + "left": "issue:44206", + "right": "issue:45584", "accept": false, - "reason": "BigBirdTokenizer special-token registration vs. a GraniteSpeech CI test issue; unrelated." + "reason": "Unsupported center arg in LasrFeatureExtractor and Whisper empty-transcription failure are separate regressions." }, { - "left": "issue:44206", - "right": "issue:45584", + "left": "issue:43653", + "right": "issue:44112", "accept": false, - "reason": "FeatureExtractor center-arg regression vs. Whisper empty transcription failure; different code paths." + "reason": "BigBirdTokenizer special-token registration and GraniteSpeech stale CI test are unrelated." }, { "left": "issue:43450", "right": "issue:43493", "accept": false, - "reason": "Video processor batch shape bug vs. SigLIP2 implementation discrepancy; unrelated." + "reason": "Batched video shape bug and SigLIP2 model discrepancy are different issues." }, { "left": "issue:45290", "right": "issue:45381", "accept": false, - "reason": "Chat-template tool-call crash vs. Qwen2.5-VL video vision_position_ids bug; different failures." + "reason": "Chat-template crash with tool-call messages and Qwen2.5-VL video vision_position_ids are distinct bugs." }, { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "LoRA/vocab weight-tying collapse vs. category-count limit runtime error; unrelated." - }, - { - "left": "issue:44462", - "right": "issue:44561", - "accept": false, - "reason": "AutoTokenizer ignoring tokenizer.json vs. removal of is_torch_fx_available breaking remote-code models; different issues." + "reason": "LoRA merge/collapsed weights and the 2^24 categories runtime error are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The one accepted SigLIP2 pair is plausibly duplicate-like from the titles, and the rest are reasonably distinguished as different bugs or failure modes rather than asserted duplicates." - }, - "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 }, @@ -6776,10 +6739,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:29127", "issue:29942", @@ -6787,7 +6750,6 @@ "issue:37428", "issue:38175", "issue:39692", - "issue:40444", "issue:43262", "issue:43278", "issue:43335", @@ -6808,6 +6770,7 @@ "issue:44292", "issue:44351", "issue:44448", + "issue:44462", "issue:44493", "issue:44554", "issue:44561", @@ -6831,13 +6794,14 @@ "issue:45593" ], "soft_pairs": [ + "issue:44462|issue:44561", "issue:43819|issue:45083", "issue:44561|issue:44779", "issue:43450|issue:43575", "issue:45200|issue:45588", "issue:37428|issue:43425", - "issue:44857|issue:45593", "issue:44610|issue:44857", + "issue:44857|issue:45593", "issue:43262|issue:43450", "issue:43819|issue:44112", "issue:44448|issue:44743", @@ -6852,11 +6816,11 @@ "issue:43819|issue:44493", "issue:45200|issue:45381", "issue:43450|issue:43526", - "issue:29127|issue:32090", "issue:39692|issue:44493", + "issue:29127|issue:32090", "issue:43335|issue:43653", - "issue:44448|issue:44779", "issue:44855|issue:45083", + "issue:44448|issue:44779", "issue:38175|issue:44493", "issue:44292|issue:45520", "issue:44977|issue:45229", @@ -6865,273 +6829,258 @@ "issue:43550|issue:44112", "issue:44568|issue:45491", "issue:43335|issue:45072", - "issue:43723|issue:44554", - "issue:40444|issue:43526", - "issue:44448|issue:44977", - "issue:43335|issue:45593" + "issue:43723|issue:44554" ], - "prepared_review_unit_hash": "753d223b2229f07556d20779e5619a3bff235f76bf4f49bfc78886986aa71477", + "prepared_review_unit_hash": "426c12fcf93507ac587fb1a1074b1ed3a39d9643245429443df2de339371d397", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 38, - "serialized_chars": 39152, - "estimated_input_tokens": 9788, - "estimated_eval_tokens": 19832 + "soft_pair_count": 36, + "serialized_chars": 38521, + "estimated_input_tokens": 9631, + "estimated_eval_tokens": 19518 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of unrelated model-specific bugs and v5 regressions; only one soft pair looks like a true duplicate: the two identical Qwen3.5 MoE from_pretrained reports.", - "confidence": 0.93, - "canonical_issue_reason": "issue:45310 is the better anchor because it is the earlier, fuller report of the same Qwen3.5 MoE from_pretrained error as issue:45478.", + "summary": "This set is mostly a grab-bag of unrelated bug reports. The only clear duplicate pair is the Qwen3.5 MoE from_pretrained error (issues 45310 and 45478); everything else looks like separate regressions or different subsystems.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45310 is the earlier and slightly better-documented report for the same Qwen3.5 MoE from_pretrained error also filed as 45478.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the strongest issue candidate overall: it has an exact duplicate counterpart, a narrowly scoped concrete failure mode, and enough detail to serve as the canonical thread.", + "best_issue_reason": "Issue 45310 is the best representative because it is the earliest clear duplicate target in the set and has the same concrete failure as 45478.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44462", + "right": "issue:44561", + "accept": false, + "reason": "Both are v5-era loading regressions, but one is AutoTokenizer/tokenizer.json and the other is removal of is_torch_fx_available for trust_remote_code; different bugs." + }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "Different models and failures: DAC latent/STE mismatch vs Qwen3 Omni feature-length helper behavior." + "reason": "Different code paths: DAC.from_latents parity with forward pass versus a Qwen3 Omni MoE helper length calculation issue." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "Both are v5-related, but one is about removed utility APIs and the other about tokenizer output correctness for Deepseek." + "reason": "Both mention tokenizer breakage in v5, but one is about a removed utility and remote-code loading; the other is DeepSeek tokenization behavior." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Unrelated: video processor batch shape bug vs tensor-parallel OOM when loading a large Qwen model." + "reason": "Video processor shape mismatch and Qwen2 model loading OOM are unrelated." }, { "left": "issue:45200", "right": "issue:45588", "accept": false, - "reason": "Gemma 4 mm_token_type_ids defaulting issue is unrelated to flash_attention sink-less crash with s_aux=None." + "reason": "Gemma4 text-only mm_token_type_ids defaulting and flash_attention crashing on s_aux=None are different failures." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "ImportError for a flash-attention helper is not the same underlying problem as Torch 2.10 incompatibility." + "reason": "ImportError for a flash-attention helper and a Torch 2.10 compatibility issue are not the same bug." }, { - "left": "issue:44857", - "right": "issue:45593", + "left": "issue:44610", + "right": "issue:44857", "accept": false, - "reason": "Different losses and code paths: float16 AMP crash in LwDetrImageLoss vs missing auxiliary losses in D-FINE." + "reason": "Processor output size mismatch versus AMP/CUDA loss crash are unrelated model-specific issues." }, { - "left": "issue:44610", - "right": "issue:44857", + "left": "issue:44857", + "right": "issue:45593", "accept": false, - "reason": "Processor output/input size mismatch is unrelated to an AMP/CUDA loss crash." + "reason": "Both involve training losses, but one is a float16 CUDA crash and the other is missing auxiliary losses when denoising is off." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sample-rate default bug is unrelated to batched video processor shapes." + "reason": "Audio chat-template sampling-rate default and batched video processor shape are different subsystems." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "Different failure classes: DAC latent mismatch vs a stale CI override test in GraniteSpeech." + "reason": "DAC latent/forward mismatch is unrelated to a stale CI device override test." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Pegasus output drift between v4/v5 is not the same as recurrent state reset in qwen3_5 cache handling." + "reason": "Pegasus v4/v5 output differences and Qwen3.5 recurrent state resetting are distinct model behaviors." }, { "left": "issue:45310", "right": "issue:45478", "accept": true, - "reason": "Same exact title and same concrete Qwen3.5 MoE from_pretrained regression; these look like duplicate reports of the same bug." + "reason": "Same title and same Qwen3.5 MoE from_pretrained error; 45478 is a later duplicate of 45310." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "CPU offload device mismatch in Bark-small is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "CPU offload device mismatch for bark-small versus a Qwen3 Omni MoE feature-length helper issue." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Embedding dtype regression during evaluate is unrelated to a pipeline task-support error message." + "reason": "Embedding dtype drift in train/eval is unrelated to pipeline task support error messaging." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 test failures and Trainer _gpu_broadcast_one NoneType errors are different bugs." + "reason": "Flash Attention test failures and Trainer _gpu_broadcast_one NoneType errors are different problems." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they concern different symbols and code paths (_flash_supports_window_size vs HybridCache)." + "reason": "Both are import errors, but they concern different missing symbols in different modules." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "SigLIP2 docs example issues are unrelated to BeitImageProcessorFast reduce_labels behavior." + "reason": "SigLIP2 doc example errors and BEiT fast processor label reduction bug are unrelated." }, { "left": "issue:43611", "right": "issue:43931", "accept": false, - "reason": "base_model_prefix loading breakage is different from a Qwen3-VL weight shape mismatch." + "reason": "base_model_prefix loading breakage in v5 and Qwen3-VL weight-shape mismatch are different model-loading failures." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug is not the same as Bamba torch.compile/SDPA failure." + "reason": "SwitchTransformers sparse-layer configuration bug and Bamba torch.compile SDPA failure are unrelated." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC latent mismatch and unexpected position_id keys are unrelated model bugs." + "reason": "DAC latent mismatch and unexpected position-id key warnings are different issues." }, { "left": "issue:45200", "right": "issue:45381", "accept": false, - "reason": "Gemma4 mm_token_type_ids defaulting is unrelated to Qwen2.5-VL video vision_position_ids being wrong." + "reason": "Gemma4 mm_token_type_ids defaults and Qwen2.5-VL video position ids are separate input-prep bugs." }, { "left": "issue:43450", "right": "issue:43526", "accept": false, - "reason": "Video batch shape handling and Beit label reduction are different processor bugs." + "reason": "Batched video output shape bug versus BEiT fast reduce_labels behavior are unrelated." }, { - "left": "issue:29127", - "right": "issue:32090", + "left": "issue:39692", + "right": "issue:44493", "accept": false, - "reason": "LayoutLMv3 box-content error messaging and Trainer broadcast NoneType errors do not share the same root cause." + "reason": "Documentation/quantization issues for SigLIP2 do not match the position-id warning regression." }, { - "left": "issue:39692", - "right": "issue:44493", + "left": "issue:29127", + "right": "issue:32090", "accept": false, - "reason": "SigLIP2 docs example failures are unrelated to position_id key warnings across models." + "reason": "LayoutLMv3 error-message clarity and Trainer GPU broadcast type error are unrelated." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "SwitchTransformers sparse-layer config and BigBirdTokenizer special-token registration are separate issues." + "reason": "SwitchTransformers sparse-layer creation bug and BigBirdTokenizer special-token registration bug are not the same defect." }, { - "left": "issue:44448", - "right": "issue:44779", + "left": "issue:44855", + "right": "issue:45083", "accept": false, - "reason": "Pegasus v4/v5 output drift is unrelated to Deepseek tokenizer regressions." + "reason": "Python 3.13 IndentationError in DebertaV2 import is unrelated to Qwen3 Omni MoE helper behavior." }, { - "left": "issue:44855", - "right": "issue:45083", + "left": "issue:44448", + "right": "issue:44779", "accept": false, - "reason": "Python 3.13 IndentationError in DebertaV2 import is unrelated to qwen3_omni_moe feature-length computation." + "reason": "Pegasus v4/v5 output change and DeepSeek tokenizer regression are separate tokenizer/model issues." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 zero probabilities and unexpected position_id keys are different symptoms in different model families." + "reason": "Unexpected zero probabilities in SigLIP2 and unexpected position-id keys are different symptoms with different causes." }, { "left": "issue:44292", "right": "issue:45520", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure is unrelated to Python 3.13 flash_attn import KeyError." + "reason": "Qwen-3-8B-NVFP4 runtime failure and flash_attn import KeyError under Python 3.13 are not the same bug." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Flash-attention generation bug in Qwen3.5 is unrelated to Gemma4 multi-GPU CUDA OOM." + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU CUDA OOM are unrelated." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug does not match qwen3_omni_moe feature-length helper behavior." + "reason": "SwitchTransformers sparse-layer misconfiguration and Qwen3 Omni MoE helper length behavior are different code paths." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "Sparse-layer creation in SwitchTransformers is unrelated to Beit semantic segmentation preset loading." + "reason": "Sparse-layer creation bug and BEiT semantic segmentation legacy-field loading bug are unrelated." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "Bamba compile/SDPA failure and GraniteSpeech stale CI override are separate problems." + "reason": "torch.compile SDPA failure and stale CI device override test failure are not the same underlying bug." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Tokenizer add_special_tokens regression is unrelated to Gemma3 NaN embeddings from sliding-window attention." + "reason": "mDeBERTa add_special_tokens behavior and Gemma3 NaN embeddings from sliding-window attention are different tokenizer/model issues." }, { "left": "issue:43335", "right": "issue:45072", "accept": false, - "reason": "SwitchTransformers config bug is not the same as dtype mismatches in bfloat16 inference." + "reason": "SwitchTransformers sparse layer creation and bfloat16 dtype mismatch in inference are distinct." }, { "left": "issue:43723", "right": "issue:44554", "accept": false, - "reason": "AutoTokenizer loading in v5 and MPS attention correctness when value head dim differs are unrelated." - }, - { - "left": "issue:40444", - "right": "issue:43526", - "accept": false, - "reason": "Qwen2.5-VL iterable dataset multi-image finetuning failure is unrelated to BeitImageProcessorFast label reduction." - }, - { - "left": "issue:44448", - "right": "issue:44977", - "accept": false, - "reason": "Pegasus output drift and Qwen3.5 flash-attention generation issues affect different models and code paths." - }, - { - "left": "issue:43335", - "right": "issue:45593", - "accept": false, - "reason": "SwitchTransformers sparse-layer config bug is unrelated to D-FINE auxiliary-loss behavior." + "reason": "AutoTokenizer loading regression in v5 and MPS attention correctness on differing head dims are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: the only clearly supported duplicate/merge candidate is issue:45310 \u2194 issue:45478, which share the exact title and highly similar metadata. The remaining soft-pair rejections are conservative and consistent with the titles shown. No PR-pair mergeability concerns apply here." + "feedback": "The assessment is grounded in the packet, and the lone accepted duplicate pair (45310/45478) is well-supported by identical titles and chronology. The other pairwise rejections are conservative and do not overclaim duplication. The only minor caution is that 'slightly better-documented' is a bit stronger than the packet directly shows, but it does not materially affect the verdict." }, "evaluator_used": true, "retried": false, @@ -7147,10 +7096,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:25251", "issue:29942", @@ -7159,14 +7108,15 @@ "issue:34689", "issue:35141", "issue:37428", + "issue:40444", "issue:41720", "issue:43065", "issue:43066", "issue:43278", "issue:43329", - "issue:43381", + "issue:43335", "issue:43412", - "issue:43421", + "issue:43526", "issue:43550", "issue:43606", "issue:43653", @@ -7176,7 +7126,6 @@ "issue:43976", "issue:43994", "issue:44112", - "issue:44186", "issue:44448", "issue:44451", "issue:44462", @@ -7188,7 +7137,6 @@ "issue:44987", "issue:45072", "issue:45083", - "issue:45198", "issue:45229", "issue:45245", "issue:45372", @@ -7196,29 +7144,32 @@ "issue:45412", "issue:45440", "issue:45459", - "issue:45491", "issue:45520", "issue:45563", - "issue:45588" + "issue:45588", + "issue:45593" ], "soft_pairs": [ + "issue:40444|issue:43526", + "issue:44448|issue:44977", + "issue:43335|issue:45593", "issue:43720|issue:44112", "issue:43066|issue:44462", - "issue:34689|issue:35141", "issue:45372|issue:45588", + "issue:34689|issue:35141", "issue:44977|issue:45563", - "issue:43606|issue:44857", "issue:44857|issue:45072", + "issue:43606|issue:44857", "issue:43412|issue:45588", "issue:37428|issue:44987", "issue:43873|issue:44493", - "issue:43701|issue:45083", - "issue:45372|issue:45406", "issue:43873|issue:45229", + "issue:45372|issue:45406", "issue:29942|issue:35141", - "issue:43550|issue:45083", + "issue:43701|issue:45083", "issue:43976|issue:45440", "issue:45245|issue:45563", + "issue:43550|issue:45083", "issue:43278|issue:43329", "issue:32090|issue:33357", "issue:41720|issue:45229", @@ -7229,225 +7180,216 @@ "issue:25251|issue:32090", "issue:43065|issue:45412", "issue:43653|issue:44743", - "issue:44857|issue:45588", - "issue:44186|issue:45198", - "issue:43278|issue:43381", - "issue:43421|issue:45491" + "issue:44857|issue:45588" ], - "prepared_review_unit_hash": "acbe7f45d48ed2629880a382df1404db17ad6dcab2da98fa8b9a65bea0af28d3", + "prepared_review_unit_hash": "c27b3e69374e7ce3587541174fdf24a16d5043249d5db43cc8b96801fb2a99eb", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 37059, - "estimated_input_tokens": 9265, - "estimated_eval_tokens": 18786 + "serialized_chars": 36493, + "estimated_input_tokens": 9124, + "estimated_eval_tokens": 18504 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No duplicate-worthy cluster emerged: the items span many unrelated Transformers bugs and regressions, and none of the soft edges are strong enough to merge.", - "confidence": 0.93, - "canonical_issue_reason": "No single canonical issue fits this cluster; the issues cover distinct code paths (tokenizers, flash-attn, device mapping, serialization, export, and model-specific regressions).", + "summary": "No soft pair looks like the same underlying bug/change; these are mostly unrelated model-, tokenizer-, device-, or CI-specific issues. There is no clear canonical issue or PR for this mixed set.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "No global best issue is suitable as a canonical anchor because the set is not one underlying bug/change.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Different bugs: Qwen2.5-VL multi-image IterableDataset failure vs BeitImageProcessorFast label reduction." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failure modes: Pegasus v4/v5 output drift vs Qwen3.5 flash-attention generation issue." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model logic: SwitchTransformers sparse-layer config bug vs D-FINE auxiliary-loss behavior." + }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "Different failures: BitNet packed-weight loading vs GraniteSpeech stale device-override CI test." + "reason": "Different subsystems: BitNet accelerate loading failure vs GraniteSpeech CI stale device-override test." }, { "left": "issue:43066", "right": "issue:44462", "accept": false, - "reason": "Both tokenizer-related, but one is decoder type in v5 and the other is AutoTokenizer ignoring repo tokenizer.json." + "reason": "Both tokenizer-related, but one is wrong decoder type in v5 and the other ignores tokenizer.json; not the same bug." }, { - "left": "issue:34689", - "right": "issue:35141", + "left": "issue:45372", + "right": "issue:45588", "accept": false, - "reason": "Llama 3.2 Vision loading regression and embedding resize/post_init reinit are unrelated bugs." + "reason": "Gemma4 processor import issue vs flash_attention.py AttributeError on s_aux=None are distinct code paths." }, { - "left": "issue:45372", - "right": "issue:45588", + "left": "issue:34689", + "right": "issue:35141", "accept": false, - "reason": "Gemma 4 processor import failure vs flash_attention.py crashing on s_aux=None are different code paths." + "reason": "Different failure surfaces: Llama 3.2 Vision load regression vs embedding reinit after resize/post_init." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Flash-attention generation bug and stale num_return_sequences warning are not the same issue." + "reason": "Qwen3.5 flash-attention generation bug is unrelated to the stale paged-generate warning." }, { - "left": "issue:43606", - "right": "issue:44857", + "left": "issue:44857", + "right": "issue:45072", "accept": false, - "reason": "CPU offload device mismatch and LwDetr AMP float16 crash are distinct runtime errors." + "reason": "Different loss/model paths: LwDetrImageLoss AMP crash vs dtype mismatches in SwitchTransformers/TimmWrapperModel inference." }, { - "left": "issue:44857", - "right": "issue:45072", + "left": "issue:43606", + "right": "issue:44857", "accept": false, - "reason": "Different dtype problems in different models and execution modes; not the same bug." + "reason": "Both runtime/device related, but CPU offload mismatch for bark-small is not the same as LwDetr AMP/CUDA crash." }, { "left": "issue:43412", "right": "issue:45588", "accept": false, - "reason": "Executorch export guard/erfinv issue is unrelated to sink-less flash_attention AttributeError." + "reason": "Executorch export/training-guard and erfinv support issue is unrelated to flash_attention.py s_aux=None crash." }, { "left": "issue:37428", "right": "issue:44987", "accept": false, - "reason": "Both loading-related, but one is an import error in flash_attention_utils and the other is a model-loading regression for physical-intelligence/fast." + "reason": "Both mention loading/flash-attention, but one is a missing internal import and the other is a model-loading failure for a specific repo." }, { "left": "issue:43873", "right": "issue:44493", "accept": false, - "reason": "Quantization offload failure and unexpected position-id keys are different symptoms and likely different code paths." + "reason": "Quantization offload behavior is unrelated to unexpected position-id keys in model outputs." }, { - "left": "issue:43701", - "right": "issue:45083", + "left": "issue:43873", + "right": "issue:45229", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "Quantized offloading bug is not the same as Gemma4 multi-GPU CUDA OOM." }, { "left": "issue:45372", "right": "issue:45406", "accept": false, - "reason": "Both involve Gemma4, but one is a dependency import problem and the other is a serve-time missing _tokenizer attribute." - }, - { - "left": "issue:43873", - "right": "issue:45229", - "accept": false, - "reason": "Generic offloading/quantization issue vs Gemma4 multi-GPU OOM are different problems." + "reason": "Gemma4 processor import error and Gemma4Processor missing _tokenizer in serve are different bugs." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures are not the same as token-embedding resize reinitialization." + "reason": "Flash Attention 2 test failures are unrelated to token-embedding resize reinitialization." }, { - "left": "issue:43550", + "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "torch.compile+SDPA failure in Bamba is unrelated to the qwen3_omni_moe length-helper bug." + "reason": "Checkpoint resume key mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python version compatibility failure and DeepseekV3MoE divergence are different regressions." + "reason": "Python-version compatibility issue is unrelated to DeepseekV3MoE divergence from remote implementation." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "Category-cardinality runtime error is unrelated to the stale generate() warning." + "reason": "Category-count limit error is unrelated to the stale num_return_sequences warning." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure is unrelated to qwen3_omni_moe length helper behavior." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "BF16/FP32 dtype drift and undefined multimodal-token helper variables are separate bugs." + "reason": "Embedding dtype drift in eval is unrelated to the multimodal token-count video-branch bug." }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Trainer NoneType broadcast error and MacOS bus error on CLIP are unrelated." + "reason": "Trainer broadcast TypeError and MacOS CLIP bus error are different failures." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Both are GPU inference failures, but on different models and with different root causes (cudaErrorAssert vs OOM)." + "reason": "Qwen3 auto device-map cudaErrorAssert is distinct from Gemma4 multi-GPU OOM." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "Gemma4 processor import issue and tokenizer error masking when protobuf is absent are different failures." + "reason": "Gemma4 processor import breakage is unrelated to tokenizer-error masking when protobuf is missing." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "Both are v5 behavior regressions, but one is Pegasus output mismatch and the other is lm_head serialization inconsistency." + "reason": "Pegasus output differences and lm_head.weight serialization inconsistencies are different regressions." }, { "left": "issue:44451", "right": "issue:45520", "accept": false, - "reason": "ScandiBERT loading failure and Python 3.13 flash_attn KeyError are unrelated." + "reason": "ScandiBERT loading failure is unrelated to flash_attn import key errors on Python 3.13." }, { "left": "issue:43065", "right": "issue:43994", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as SigLIP2 producing nonsensical AutoModel/pipeline output." - }, - { - "left": "issue:25251", - "right": "issue:32090", - "accept": false, - "reason": "Pipeline top_k nesting bug and Trainer GPU broadcast TypeError are unrelated." + "reason": "Sam3PixelDecoder dummy Conv2d issue is unrelated to SigLIP2 producing nonsensical outputs." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder and RT-DETR memory not being released are different issues." + "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as RT-DETR memory-not-released behavior." }, { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug is unrelated to Qwen recurrent-state reset in cache mode." + "reason": "BigBirdTokenizer special-token registration bug is unrelated to Qwen3 recurrent-state reset with cache." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "LwDetr float16 AMP crash and sink-less flash_attention AttributeError are different runtime paths." - }, - { - "left": "issue:44186", - "right": "issue:45198", - "accept": false, - "reason": "LayoutLMv2 tokenizer crash on NER/padding is not the same as Wav2Vec2 save_pretrained/tokenization failure." - }, - { - "left": "issue:43278", - "right": "issue:43381", - "accept": false, - "reason": "Embedding dtype drift during eval and gradient checkpointing disallowed in eval mode are separate behaviors." - }, - { - "left": "issue:43421", - "right": "issue:45491", - "accept": false, - "reason": "Runtime special-token/post-processor update bug is unrelated to Gemma3 NaN embeddings with sliding-window attention." + "reason": "Different areas: LwDetr float16 AMP crash vs flash_attention.py AttributeError on sink-less models." } ] }, @@ -7466,10 +7408,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:29127", "issue:29942", @@ -7488,6 +7430,7 @@ "issue:43278", "issue:43334", "issue:43381", + "issue:43421", "issue:43526", "issue:43531", "issue:43653", @@ -7495,7 +7438,6 @@ "issue:43756", "issue:43761", "issue:43824", - "issue:43931", "issue:43994", "issue:44079", "issue:44186", @@ -7521,8 +7463,11 @@ "issue:45593" ], "soft_pairs": [ - "issue:44361|issue:45198", + "issue:44186|issue:45198", + "issue:43278|issue:43381", + "issue:43421|issue:45491", "issue:42915|issue:42994", + "issue:44361|issue:45198", "issue:43994|issue:45412", "issue:44265|issue:45198", "issue:44206|issue:44479", @@ -7535,8 +7480,8 @@ "issue:43824|issue:44279", "issue:33357|issue:39692", "issue:29127|issue:34567", - "issue:32090|issue:33290", "issue:45440|issue:45593", + "issue:32090|issue:33290", "issue:45072|issue:45229", "issue:43381|issue:43701", "issue:43653|issue:44442", @@ -7548,211 +7493,222 @@ "issue:29127|issue:33357", "issue:29127|issue:35141", "issue:43761|issue:44079", - "issue:43262|issue:44206", - "issue:42915|issue:43931" + "issue:43262|issue:44206" ], - "prepared_review_unit_hash": "0c7135ccb38daa86fed32adba9840ed6fc1888016a0e4b08b4cfd022593eaeef", + "prepared_review_unit_hash": "385a1812659d0d9ee762cb3117e2b3ac62b993b9f33c9c73222eb12aa2e9b89a", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36445, - "estimated_input_tokens": 9112, - "estimated_eval_tokens": 18480 + "soft_pair_count": 31, + "serialized_chars": 37066, + "estimated_input_tokens": 9267, + "estimated_eval_tokens": 18790 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of unrelated regression reports across tokenizers, vision/audio processors, generation, and training utilities. The soft links are similarity-only and do not look like true duplicates.", - "confidence": 0.18, - "canonical_issue_reason": "issue:45103 is the strongest representative issue here: it is open, narrowly scoped, has inbound references, and describes a concrete crash in _process_kwargs_parameters with a clear reproduction path.", + "summary": "This cluster is heterogeneous: most items are distinct bugs across tokenizers, vision/audio processors, training, generation, and model-loading paths. None of the soft-similarity pairs look like the same underlying fix, so all should be rejected. No PRs are present in the cluster.", + "confidence": 0.82, + "canonical_issue_reason": "issue:45198 is the strongest hub in the soft-pair graph and describes a concrete, user-facing tokenizer/save_pretrained failure. That said, the surrounding items are still too diverse for a true duplicate cluster, so this is only a weak canonical choice.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45103 is the best issue candidate because it is actionable, specific, and central enough to stand in for a real bug report; the rest of the cluster is too heterogeneous to serve as a stronger canonical.", + "best_issue_reason": "issue:45198 is the best representative of the small tokenizer-failure subgroup because it is concrete and appears in multiple soft-similarity links. It is still not a good global duplicate anchor for the whole set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44361", + "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "Different tokenizer stacks and different failures: MLukeTokenizer task-time AttributeError vs Wav2Vec2 save_pretrained/tokenization problems." + "reason": "Different models and failure modes: LayoutLMv2Tokenizer NER/padding crash vs Wav2Vec2 save_pretrained/tokenization failure." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Unrelated bugs: BF16->FP32 dtype drift in evaluate vs gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Different subsystems: runtime post-processor/special-token updates vs Gemma3 NaN embeddings from sliding-window padding." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "Both involve quantization, but one is a FineGrainedFP8Config runtime failure and the other is a quantized model saving failure; not the same bug." + "reason": "Both are quantization-related but not the same bug: FineGrainedFP8Config failure vs quantized model saving failure." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer bugs on different code paths: MLuke task AttributeError vs Wav2Vec2 tokenization/save failure." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 wrong outputs and an RT-DETR memory leak are unrelated failure modes in different subsystems." + "reason": "No shared underlying defect: SigLIP2 bad outputs vs RT-DETR memory not released on deletion." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "torch.export/torch_compilable_check failure is unrelated to Wav2Vec2 tokenization and save_pretrained behavior." + "reason": "torch.export failure from torch_compilable_check is unrelated to Wav2Vec2 tokenization/save_pretrained errors." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Unsupported feature-extractor arg for Lasr is unrelated to the Qwen video-input regression." + "reason": "Different regressions in different models: unsupported center arg in LASR extractor vs Qwen video input regression." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma eos_token_id mismatch and Qwen3.5 flash-attention generation issues are different model/config bugs." + "reason": "Distinct issues: Gemma eos_token_id config mismatch vs Qwen3.5 flash-attention generation failure." }, { "left": "issue:43257", "right": "issue:43278", "accept": false, - "reason": "Loading/conversion failure for Qwen3 MoE weights is not the same as BF16-to-FP32 dtype drift during evaluation." + "reason": "Different problem classes: Qwen3 MoE weight conversion with accelerate/deepspeed vs training/eval dtype change." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Both mention MOE-related model discrepancies, but they concern different models and different concrete defects." + "reason": "Separate model-specific correctness issues: Smollm3 RoPE layer count vs DeepseekV3 implementation divergence." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Failing Flash Attention 2 tests and void segmentation-map processing are unrelated." + "reason": "Flash Attention 2 test failures and void segmentation map processing are unrelated bugs." }, { "left": "issue:44610", "right": "issue:45412", "accept": false, - "reason": "Processor output size mismatch is not the same as a model memory-release leak." + "reason": "Processor input-size mismatch vs model memory leak after deletion are not the same bug." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "Auto-docstring annotation crash and mistral_common import failure are separate issues with different code paths." + "reason": "auto_docstring annotation handling and Gemma 4 processor dependency import failure are unrelated." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "A specific missing import for Qwen2_5_VL is not the same as a vague dependency issue with transformers." + "reason": "Specific missing export/import of Qwen2_5_VLForConditionalGeneration is not the same as a generic dependency issue." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS bus error with a community CLIP model and SigLIP2 documentation/example errors do not share the same underlying bug." + "reason": "MacOS CLIP bus error and SigLIP2 doc example errors are different model-specific problems." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error-message clarity and TrainerState token counting are unrelated." + "reason": "LayoutLMv3 error-message clarity and TrainerState token counting are unrelated concerns." }, { - "left": "issue:32090", - "right": "issue:33290", + "left": "issue:45440", + "right": "issue:45593", "accept": false, - "reason": "NoneType broadcast errors in Trainer and DeepSpeed Adafactor OOMs are different problems." + "reason": "DeepseekV3 model divergence and D-FINE auxiliary-loss omission are unrelated implementation bugs." }, { - "left": "issue:45440", - "right": "issue:45593", + "left": "issue:32090", + "right": "issue:33290", "accept": false, - "reason": "DeepSeekV3MoE divergence from remote implementation is unrelated to D-FINE missing auxiliary losses." + "reason": "Trainer broadcast TypeError and Adafactor/DeepSpeed OOM are different failure modes." }, { "left": "issue:45072", "right": "issue:45229", "accept": false, - "reason": "Bfloat16 dtype mismatches in inference and Gemma4 multi-GPU CUDA OOM are not the same code-path bug." + "reason": "Bfloat16 dtype mismatch in inference is unrelated to Gemma4 multi-GPU OOM." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Gradient checkpointing being disallowed in eval mode is separate from resume_from_checkpoint key mismatches." + "reason": "Gradient checkpointing in eval mode and resume_from_checkpoint key mismatch are separate trainer issues." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer special-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer issues." + "reason": "BigBirdTokenizer special-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." }, { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Qwen3-VL pad_token_id loading failure and Qwen3-MoE sliding_window behavior are different failures." + "reason": "Missing pad_token_id on Qwen3-VL and sliding_window behavior on Qwen3-MoE are distinct model/config issues." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels truncation and LayoutLMv2Tokenizer NER padding crashes do not match." + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to LayoutLMv2Tokenizer crashes." }, { "left": "issue:44464", "right": "issue:45491", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is not the same bug as Gemma3 NaN embeddings with mixed-length batches." + "reason": "Compiled-forward chunked generation inconsistency is not the same bug as Gemma3 NaN embeddings from all-padding windows." }, { "left": "issue:29942", "right": "issue:36010", "accept": false, - "reason": "Flash Attention 2 test failures and GenerationMixin import errors are unrelated." + "reason": "Flash Attention 2 test failures and GenerationMixin import errors do not describe the same underlying defect." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Flash-attention generation problems and a category-count runtime limit are unrelated." + "reason": "Flash-attention generation regression and category-cardinality runtime error are unrelated." }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "LayoutLMv3 box-information messaging and a MacOS CLIP bus error are unrelated." + "reason": "LayoutLMv3 error-message clarity and MacOS bus-error crash are clearly different bugs." }, { "left": "issue:29127", "right": "issue:35141", "accept": false, - "reason": "A LayoutLMv3 error-message issue is unrelated to token embedding reinitialization in post_init." + "reason": "Box-information error messaging in LayoutLMv3 is unrelated to post_init reinitializing resized embeddings." }, { - "left": "issue:43262", - "right": "issue:44206", - "accept": false, - "reason": "Audio chat-template sampling-rate defaults and Lasr feature-extractor argument handling are different issues." - }, - { - "left": "issue:42915", - "right": "issue:43931", + "left": "issue:43761", + "right": "issue:44079", "accept": false, - "reason": "Qwen3Moe FP8 failure and Qwen3-VL weight-shape mismatch are different model-loading problems." + "reason": "CLIPVisionModel hidden_states regression and ModelOutput key assignment bug are different implementation issues." }, { - "left": "issue:44610", - "right": "issue:45491", + "left": "issue:43262", + "right": "issue:44206", "accept": false, - "reason": "Processor output-size mismatch and Gemma3 NaN embeddings are unrelated." + "reason": "Audio chat-template sampling-rate default and LASR extractor center-arg crash are unrelated processor bugs." } ] }, @@ -7771,19 +7727,19 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:33290", "issue:34567", "issue:34689", "issue:39692", "issue:41628", + "issue:42915", "issue:43232", "issue:43316", - "issue:43381", "issue:43388", "issue:43441", "issue:43452", @@ -7826,10 +7782,11 @@ "issue:45600" ], "soft_pairs": [ + "issue:42915|issue:43931", "issue:43441|issue:45588", "issue:43232|issue:43388", - "issue:44442|issue:45072", "issue:43526|issue:44857", + "issue:44442|issue:45072", "issue:43723|issue:45479", "issue:44479|issue:45290", "issue:43526|issue:44610", @@ -7838,9 +7795,9 @@ "issue:43452|issue:45588", "issue:43931|issue:45216", "issue:45588|issue:45600", + "issue:43316|issue:43525", "issue:43575|issue:43701", "issue:44977|issue:45584", - "issue:43316|issue:43525", "issue:43646|issue:44077", "issue:43761|issue:44188", "issue:43994|issue:44206", @@ -7848,209 +7805,208 @@ "issue:43575|issue:44387", "issue:43873|issue:45588", "issue:43673|issue:44743", - "issue:33290|issue:34567", - "issue:44442|issue:44898", "issue:34689|issue:43611", + "issue:44442|issue:44898", + "issue:33290|issue:34567", "issue:41628|issue:45372", - "issue:43493|issue:43827", - "issue:43381|issue:43873" + "issue:43493|issue:43827" ], - "prepared_review_unit_hash": "9915e1270a236c4a6cc4fc796445f8c50c134232a644b0af10eac3c5cbab091d", + "prepared_review_unit_hash": "8f53e0b4048b294f190c17fd3fc339215226eda9c492f6cefae3c6ea7e565fb0", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36204, - "estimated_input_tokens": 9051, - "estimated_eval_tokens": 18358 + "serialized_chars": 36195, + "estimated_input_tokens": 9049, + "estimated_eval_tokens": 18354 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No duplicate clusters are strongly supported here: the items span many unrelated Transformers issues (generation, loading, processors, quantization, configs, docs). I rejected all soft edges as they do not look like the same underlying bug/change.", - "confidence": 0.94, - "canonical_issue_reason": null, + "summary": "This cluster is heterogeneous: the soft-similarity links mostly connect unrelated bugs that only share broad subsystems (generation, tokenizers, vision, quantization, etc.), not the same underlying defect. No PRs are present.", + "confidence": 0.89, + "canonical_issue_reason": "No single canonical issue stands out; the items span many unrelated bugs across different models, processors, and runtime paths, so treating any one as the duplicate target would be misleading.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "If forced to pick a representative issue, issue:43873 is the best anchor because it is open, has the most discussion/inbound references, and is a broadly scoped report; however, it is not a true representative of the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Different models and failures: FineGrainedFP8Config vs a Qwen3-VL weight-shape mismatch." + }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both involve FlashAttention, but one is a Ministral-3 v5 RC failure and the other is an `integrations/flash_attention.py` sink-less model crash; different code paths and symptoms." + "reason": "Both involve flash attention, but one is a Ministral-3 generation regression and the other is an AttributeError in flash_attention.py for sink-less models." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation kwargs sync/gpus bug and `gather_for_metrics` tuple-label truncation are unrelated training/evaluation issues." + "reason": "Unrelated code paths: generation kwargs after sync_gpus vs gather_for_metrics dropping tuple labels." }, { - "left": "issue:44442", - "right": "issue:45072", + "left": "issue:43526", + "right": "issue:44857", "accept": false, - "reason": "Tokenizer loading for FastSpeech2Conformer and dtype mismatch in bfloat16 inference are different problems in different subsystems." + "reason": "Different components and symptoms: BeitImageProcessorFast label reduction vs LwDetrImageLoss AMP crash." }, { - "left": "issue:43526", - "right": "issue:44857", + "left": "issue:44442", + "right": "issue:45072", "accept": false, - "reason": "BEiT label reduction bug and LwDetrImageLoss AMP crash are not the same underlying image-processing issue." + "reason": "Tokenizer-loading failure for FastSpeech2Conformer is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:43723", "right": "issue:45479", "accept": false, - "reason": "Tokenizer loading regression and zero-loss sequence-classification behavior are separate bugs." + "reason": "Tokenizer loading in v5 is unrelated to the zero-loss classification bug for num_labels=1." }, { "left": "issue:44479", "right": "issue:45290", "accept": false, - "reason": "Video input regression for Qwen VL models and chat template crashes on tool-call messages are different code paths." + "reason": "Different regressions: video-input handling for Qwen variants vs chat templating with tool-call assistant messages." }, { "left": "issue:43526", "right": "issue:44610", "accept": false, - "reason": "Both touch image preprocessing, but one is label reduction in BEiT and the other is wrong resize output for OmDet-Turbo." + "reason": "Both are processor-related, but one is label reduction and the other is an input-size mismatch for OmDet-Turbo." }, { "left": "issue:45003", "right": "issue:45440", "accept": false, - "reason": "Unsafe `sys.modules` access and DeepseekV3MoE divergence are unrelated implementation concerns." + "reason": "Generic modeling_utils sys.modules access is unrelated to DeepseekV3MoE divergence from remote implementation." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Cache/recurrent-state reset bug and Whisper batch_decode special-token handling are distinct generation vs decoding issues." + "reason": "Different failures in different areas: recurrent cache reset vs Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43452", "right": "issue:45588", "accept": false, - "reason": "GGUF loading failure and FlashAttention sink-less crash are unrelated load/runtime bugs." + "reason": "Tokenizer/model loading with gguf_file is unrelated to a flash_attention AttributeError." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and Qwen3.5 save_pretrained regression are both model-specific loading issues but not the same concrete bug." + "reason": "Same family name, but different bugs: Qwen3-VL loading mismatch vs Qwen3.5 save_pretrained checkpoint corruption." }, { "left": "issue:45588", "right": "issue:45600", "accept": false, - "reason": "FlashAttention crash and removed Sam3LiteText config references are completely different failures." + "reason": "Flash-attention sink handling is unrelated to stale auto_mappings references breaking CI." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Both are config/API discrepancies, but they affect different classes and different missing/changed attributes." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "Tensor-parallel OOM and resume_from_checkpoint key mismatch are unrelated training/runtime issues." + "reason": "OOM during tensor-parallel model load is unrelated to resume_from_checkpoint key mismatches." }, { "left": "issue:44977", "right": "issue:45584", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and Whisper empty-transcription failure are unrelated model-specific generation issues." - }, - { - "left": "issue:43316", - "right": "issue:43525", - "accept": false, - "reason": "Config API discrepancy and missing `pad_token_id` are both config-related, but they are different model families and different missing/changed fields." + "reason": "Different generation bugs: flash-attention on Qwen3.5 vs empty-transcription handling in Whisper." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "General custom model initialization breakage and `patchtsmixer`'s optional `post_init` restriction are related only at a broad API level, not the same bug." + "reason": "Custom model initialization breakage is broader and different from patchtsmixer's optional post_init constraint." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "CLIP hidden_states regression and attention-kernel divergence under `torch.compile` are unrelated model/runtime issues." + "reason": "CLIPVisionModel hidden_states regression is unrelated to attention-kernel divergence under torch.compile." }, { "left": "issue:43994", "right": "issue:44206", "accept": false, - "reason": "SigLIP2 output quality bug and LasrFeatureExtractor unsupported-arg crash are different failures." + "reason": "SigLIP2 bad outputs/pipeline behavior is not the same bug as an unsupported center argument crash in LasrFeatureExtractor." }, { "left": "issue:39692", "right": "issue:45491", "accept": false, - "reason": "SigLIP2 documentation errors and Gemma3 NaN embeddings are unrelated." + "reason": "Unrelated models and failures: SigLIP2 docs errors vs Gemma3 NaN embeddings with sliding-window attention." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "OOM from tensor-parallel loading and increased reserved CUDA memory under int4 quantization are both memory issues but not the same underlying cause." + "reason": "Both mention OOM, but one is tensor-parallel loading and the other is increased reserved memory under int4 quantization." }, { "left": "issue:43873", "right": "issue:45588", "accept": false, - "reason": "Quantization offloading problems and FlashAttention sink-less crash do not share a concrete code path." + "reason": "Quantization/offloading behavior is unrelated to a flash-attention sink-less-model crash." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Both mention cache, but one is missing GenerationMixin cache in v5 chunked prefill while the other resets recurrent states when cache and seq_len>1; different mechanisms." + "reason": "Cache handling bugs in different generation paths and models; not the same concrete defect." }, { - "left": "issue:33290", - "right": "issue:34567", + "left": "issue:34689", + "right": "issue:43611", "accept": false, - "reason": "Adafactor+deepspeed OOM and `num_input_tokens_seen` not updating are unrelated trainer bugs." + "reason": "Both are model-loading regressions, but they affect different versions and different root causes." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "Tokenizer loading for FastSpeech2Conformer and Perceiver resolution/interpolation failure are different model/processor issues." + "reason": "FastSpeech2Conformer tokenizer loading is unrelated to Perceiver interpolation failures at non-default resolution." }, { - "left": "issue:34689", - "right": "issue:43611", + "left": "issue:33290", + "right": "issue:34567", "accept": false, - "reason": "Both are model-loading regressions, but one is Llama 3.2 vision-specific and the other is a `base_model_prefix` compatibility break; not the same bug." + "reason": "OOM with Adafactor/DeepSpeed is unrelated to TrainerState token counting not updating." }, { "left": "issue:41628", "right": "issue:45372", "accept": false, - "reason": "Different import errors for different symbols/packages; no shared underlying change." + "reason": "Different import errors in different modules: AutoImageProcessor vs ReasoningEffort from mistral_common." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation discrepancy and docs still referencing `pipeline()` are unrelated." - }, - { - "left": "issue:43381", - "right": "issue:43873", - "accept": false, - "reason": "Gradient checkpointing in eval mode and quantization offloading are separate training/runtime topics." + "reason": "SigLIP2 implementation parity is unrelated to docs still using removed pipeline() APIs." } ] }, @@ -8068,11 +8024,11 @@ "review_unit_index": 25, "review_unit_count": 43, "status": "reviewed", - "reason": null, - "source": "cache", - "cache_hit": true, + "reason": null, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:16998", "issue:22355", @@ -8082,7 +8038,6 @@ "issue:34689", "issue:35141", "issue:36010", - "issue:36331", "issue:41950", "issue:42175", "issue:43065", @@ -8098,6 +8053,7 @@ "issue:43688", "issue:43716", "issue:43720", + "issue:43742", "issue:43749", "issue:43761", "issue:43825", @@ -8120,26 +8076,29 @@ "issue:45127", "issue:45137", "issue:45200", - "issue:45584" + "issue:45584", + "issue:45656" ], "soft_pairs": [ + "issue:43381|issue:43873", "issue:33290|issue:35141", "issue:33290|issue:34689", "issue:44077|issue:44479", - "issue:43720|issue:44442", + "issue:33290|issue:45656", "issue:43643|issue:44913", - "issue:43749|issue:43931", + "issue:43720|issue:44442", "issue:44190|issue:44496", + "issue:43749|issue:43931", "issue:43873|issue:44871", - "issue:33290|issue:36331", - "issue:43381|issue:43856", - "issue:43540|issue:44514", "issue:44162|issue:44792", + "issue:43540|issue:44514", "issue:43316|issue:44877", + "issue:43381|issue:43856", "issue:34689|issue:36010", "issue:44855|issue:45137", "issue:42175|issue:43316", "issue:43404|issue:43441", + "issue:43381|issue:43742", "issue:43388|issue:43688", "issue:16998|issue:22355", "issue:25251|issue:28282", @@ -8149,198 +8108,217 @@ "issue:45127|issue:45200", "issue:43295|issue:43825", "issue:43761|issue:44479", - "issue:44871|issue:45584" + "issue:44871|issue:45584", + "issue:43295|issue:44871" ], - "prepared_review_unit_hash": "d4684e7b357a233274b9d7c0846683d127e845dd30a6e7563de1cd8bbee487c5", + "prepared_review_unit_hash": "e9eadfb41649ef36c77e05fde7c230f447c74f3fbf43208be18e3543843be3a7", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35211, - "estimated_input_tokens": 8803, - "estimated_eval_tokens": 17862 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36605, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is noisy and mostly heterogeneous: the soft-similarity links do not describe the same underlying bug/change. The closest items are broad loading/training regressions, but they affect different subsystems and code paths, so none of the soft edges should be merged.", - "confidence": 0.89, - "canonical_issue_reason": "issue:43749 is the broadest, highest-impact regression in the set and the best anchor for a cluster that otherwise spans many unrelated failures.", + "summary": "No true duplicate cluster is present here: the soft links mostly connect unrelated Transformer bugs across config, loading, pipelines, training, and multimodal regressions. I would not merge any of the suggested pairs.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43749 has the strongest combination of generality, user impact, and discussion activity, so it is the most suitable representative issue for this noisy cluster.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Different failure modes: gradient checkpointing in eval mode vs quantization/offloading behavior." + }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "Different failures: DeepSpeed/Adafactor OOM vs embedding reinitialization after resize." + "reason": "OOM with Adafactor+DeepSpeed is unrelated to embedding reinit after resize." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "Unrelated symptoms and paths: optimizer memory use vs Llama 3.2 vision model loading." + "reason": "Training OOM and Llama 3.2 vision model-loading regression are different bugs." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "Different model families and bugs: optional post_init contract vs video-input regression." + "reason": "Optional post_init API change is unrelated to the v5 video input regression." }, { - "left": "issue:43720", - "right": "issue:44442", + "left": "issue:33290", + "right": "issue:45656", "accept": false, - "reason": "Different code paths: packed-weight loading for BitNet vs tokenizer loading for FastSpeech2Conformer." + "reason": "OOM in Adafactor+DeepSpeed is not the same as optimizer step being called twice." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "Generic config-loading issue vs GPTNeoX rotary_pct persistence bug; not the same defect." + "reason": "trust_remote_code config field loss is different from GPTNeoX rotary_pct reload behavior." }, { - "left": "issue:43749", - "right": "issue:43931", + "left": "issue:43720", + "right": "issue:44442", "accept": false, - "reason": "Different loading failures: FSDP CPU-RAM-efficient loading regression vs Qwen3-VL weight-shape mismatch." + "reason": "Packed-weight loading in BitNet is unrelated to AutoTokenizer failing on a specific tokenizer class." }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Unrelated: dataset-loading failure in an example script vs missing model_type/configuration error." + "reason": "Local dataset loading in a training script is unrelated to an unrecognized model config." }, { - "left": "issue:43873", - "right": "issue:44871", + "left": "issue:43749", + "right": "issue:43931", "accept": false, - "reason": "Quantization/offloading behavior vs Gemma eos_token configuration mismatch are distinct bugs." + "reason": "FSDP CPU-efficient loading breakage is not the same as Qwen3-VL weight shape mismatch." }, { - "left": "issue:33290", - "right": "issue:36331", + "left": "issue:43873", + "right": "issue:44871", "accept": false, - "reason": "Different training API regression: memory OOM in DeepSpeed vs custom compute_loss signature mismatch." + "reason": "Quantization/offloading issue is unrelated to Gemma eos_token_id config mismatch." }, { - "left": "issue:43381", - "right": "issue:43856", + "left": "issue:44162", + "right": "issue:44792", "accept": false, - "reason": "Gradient-checkpointing eval-mode restriction is not the same as MoE memory inefficiency." + "reason": "ESM2 model breakage does not match a janus test failure for image generation." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Both involve multimodal/video processing, but the failing code paths differ: Qwen3OmniMoe video ingestion vs batched chat-template padding." + "reason": "Qwen3OmniMoe video processing ValueError is different from Qwen2.5-VL batched chat-template crash." }, { - "left": "issue:44162", - "right": "issue:44792", + "left": "issue:43316", + "right": "issue:44877", "accept": false, - "reason": "ESM2 runtime breakage and Janus image-generation test failure are unrelated." + "reason": "Gemma3TextConfig API mismatch is unrelated to granite_speech config loading strictness." }, { - "left": "issue:43316", - "right": "issue:44877", + "left": "issue:43381", + "right": "issue:43856", "accept": false, - "reason": "Different config-schema mismatches: Gemma3TextConfig API discrepancy vs granite_speech strict-config loading." + "reason": "Gradient checkpointing/eval is unrelated to Qwen3 MoE memory usage." }, { "left": "issue:34689", "right": "issue:36010", "accept": false, - "reason": "Model-loading regression for a specific model vs import failure for GenerationMixin; not the same bug." + "reason": "Llama 3.2 vision loading regression is not the same as GenerationMixin import error." }, { "left": "issue:44855", "right": "issue:45137", "accept": false, - "reason": "Python 3.13 TorchScript parsing issue vs DeepSpeed ZeRO3 deque error are unrelated." + "reason": "Python 3.13 IndentationError in DebertaV2 import is unrelated to DeepSpeed ZeRO3 deque error." }, { "left": "issue:42175", "right": "issue:43316", "accept": false, - "reason": "Packaging/backend dependency issue vs configuration API mismatch; no shared underlying defect." + "reason": "Missing TensorFlow in torch-only install is unrelated to Gemma3TextConfig API discrepancy." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Both mention Mistral-family models, but one is a weight-tying bug and the other is a FlashAttention regression." + "reason": "lm_head tying in Mistral3 is not the same bug as FlashAttention failure in Ministral-3." + }, + { + "left": "issue:43381", + "right": "issue:43742", + "accept": false, + "reason": "Gradient checkpointing eval-mode bug is unrelated to MobileLLM-125M loading key error." }, { "left": "issue:43388", "right": "issue:43688", "accept": false, - "reason": "Different training/eval math bugs: metric gathering drops labels vs auxiliary-loss normalization." + "reason": "gather_for_metrics label truncation is unrelated to auxiliary-loss normalization." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "Model_max_length handling is unrelated to the missing transformers.onnx import error." + "reason": "model_max_length question for DeBERTa-V3 is unrelated to missing transformers.onnx import." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Pipeline top_k output-shape regression is unrelated to missing PyTorch backend import failure." + "reason": "top_k pipeline output shape regression is unrelated to PyTorch-missing AutoModel import error." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Config API discrepancy vs image preprocessor/model dtype mismatch are distinct issues." + "reason": "Gemma3TextConfig API discrepancy is unrelated to Mistral-3 image preprocessor dtype mismatch." }, { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "Dummy Conv2d implementation in Sam3PixelDecoder is not the same as Sam3Video CUDA OOM." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to Sam3Video CUDA OOM." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Both are SAM/video-related, but one is pipeline processor selection and the other is a missing encoder output attribute." + "reason": "video-classification pipeline processor selection is unrelated to Sam3VisionEncoderOutput attribute error." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "LoRA/extended-vocabulary collapse with tied embeddings is unrelated to mm_token_type_ids defaults." + "reason": "LoRA merge collapse with extended vocab is unrelated to Gemma 4 mm_token_type_ids defaulting." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Processor/tokenizer regression is unrelated to the pipeline translation-task error message." + "reason": "processor.tokenizer regression is unrelated to pipeline translation-support error messaging." }, { "left": "issue:43761", "right": "issue:44479", "accept": false, - "reason": "CLIPVision hidden_states regression and Qwen video-input regression are different code paths." + "reason": "CLIPVision hidden_states regression is unrelated to Qwen v5.3.0 video-input regression." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Gemma eos_token configuration mismatch is unrelated to Whisper empty-transcription generation failure." + "reason": "Gemma eos_token_id inconsistency is unrelated to Whisper empty-transcription failure." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "processor/tokenizer regression is unrelated to Gemma eos_token_id mismatch." } ] }, @@ -8359,17 +8337,16 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:32090", "issue:35141", "issue:41720", "issue:42617", "issue:43065", - "issue:43295", "issue:43452", "issue:43550", "issue:43575", @@ -8402,8 +8379,10 @@ "issue:44938", "issue:44964", "issue:44977", + "issue:45127", "issue:45245", "issue:45290", + "issue:45357", "issue:45362", "issue:45381", "issue:45406", @@ -8414,9 +8393,8 @@ ], "soft_pairs": [ "issue:44734|issue:45406", - "issue:43295|issue:44871", - "issue:43550|issue:43701", "issue:44077|issue:44534", + "issue:43550|issue:43701", "issue:43606|issue:43701", "issue:43646|issue:44913", "issue:43906|issue:45290", @@ -8424,8 +8402,8 @@ "issue:43452|issue:44496", "issue:43931|issue:44060", "issue:44617|issue:45584", - "issue:43575|issue:43742", "issue:45563|issue:45584", + "issue:43575|issue:43742", "issue:42617|issue:43575", "issue:44521|issue:45381", "issue:44933|issue:45362", @@ -8436,209 +8414,216 @@ "issue:44464|issue:44871", "issue:43065|issue:44610", "issue:44792|issue:44977", - "issue:43827|issue:43994", "issue:44534|issue:44913", + "issue:43827|issue:43994", "issue:41720|issue:45245", "issue:44792|issue:45600", "issue:44190|issue:44964", - "issue:44938|issue:45542" + "issue:44938|issue:45542", + "issue:45127|issue:45357" ], - "prepared_review_unit_hash": "ef5aaecf07fb1f8966d365afb26f28d9d87e6c2bba6adc5490eb2f4f17b3c38c", + "prepared_review_unit_hash": "189105baee3bb645fe2b30979616c1ec1b53dfd6ea2a079be1300c4b5f7a502b", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 35609, - "estimated_input_tokens": 8903, - "estimated_eval_tokens": 18062 + "serialized_chars": 36224, + "estimated_input_tokens": 9056, + "estimated_eval_tokens": 18368 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose grab-bag of Transformers issues that share broad regression/loading/chat/template vocabulary but do not look like the same underlying bug. I would not merge any of the soft pairs as duplicates, and there are no PRs to choose from.", - "confidence": 0.38, - "canonical_issue_reason": "issue:43646 is the broadest representative here: it frames a general v5 model-initialization regression, which is the closest high-level theme to several other init/load-related reports. Still, it is not a true duplicate of the rest of the cluster.", + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs across model loading, generation, quantization, serving, and docs. I do not see a true duplicate group, so there is no good canonical artifact here. All soft-edge pairs look like similarity-only false positives.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best single issue for the cluster because it is the most umbrella-like and representative of the recurring initialization/reload regressions, even though the other items are mostly distinct bugs.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "Different failures: KV-cache continuation tensor indexing in serve vs missing _tokenizer on Gemma4Processor." + "reason": "Both are runtime crashes, but they hit different paths: KV-cache continuation indexing vs missing `_tokenizer` on `Gemma4Processor`." }, { - "left": "issue:43295", - "right": "issue:44871", + "left": "issue:44077", + "right": "issue:44534", "accept": false, - "reason": "Both are regressions, but one is about processor.tokenizer/image handling and the other is an eos_token_id config mismatch." + "reason": "Different regressions: one is about `post_init` on `patchtsmixer`, the other about junk-filled non-persistent buffers in v5." }, { "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "Unrelated bugs: torch.compile SDPA failure for Bamba vs resume_from_checkpoint key mismatch." - }, - { - "left": "issue:44077", - "right": "issue:44534", - "accept": false, - "reason": "Both touch v5 behavior, but one is about optional post_init enforcement and the other about non-persistent buffer initialization." + "reason": "Unrelated failures: SDPA/torch.compile on Bamba vs checkpoint key mismatch on resume." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "Device-mismatch on CPU offload is a different path from checkpoint resume key handling." + "reason": "CPU offload device mismatch for `bark-small` is unrelated to resume checkpoint key mismatch." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "General custom-model init breakage vs a specific GPTNeoX rotary_pct reload bug." + "reason": "Custom model initialization breakage is not the same as `GPTNeoXConfig` reloading `rotary_pct` incorrectly." }, { "left": "issue:43906", "right": "issue:45290", "accept": false, - "reason": "The first is an isolated reproduction of another issue; the second is a specific apply_chat_template tool-call crash. Not enough to call them the same bug." + "reason": "Different chat/template bugs: isolated reproduction of an earlier issue vs crash on tool-call assistant messages with no content." }, { "left": "issue:43761", "right": "issue:44242", "accept": false, - "reason": "Different model behaviors: CLIPVision hidden_states regression vs MoE load-balancing loss omission." + "reason": "CLIP hidden states regression and MoE load-balancing loss logic are separate code paths and symptoms." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "GGUF loading support breakage is distinct from an unrecognized model/config.json model_type error." + "reason": "Tokenizer/model loading with `gguf_file` is unrelated to missing `model_type` in a config for a different model." }, { "left": "issue:43931", "right": "issue:44060", "accept": false, - "reason": "One is a Qwen3-VL shape mismatch during load; the other is a tied-weights warning bug in Qwen3-Next." + "reason": "Weight shape mismatch in Qwen3-VL loading is not the same as the incorrect tied-weights warning in Qwen3-Next." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after token alignment." + "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after `align_special_tokens`." }, { - "left": "issue:43575", - "right": "issue:43742", + "left": "issue:45563", + "right": "issue:45584", "accept": false, - "reason": "Tensor-parallel OOM for a 57B model vs a key error loading a small MobileLLM checkpoint." + "reason": "Stale `num_return_sequences` warning and Whisper empty-transcription crash are different issues." }, { - "left": "issue:45563", - "right": "issue:45584", + "left": "issue:43575", + "right": "issue:43742", "accept": false, - "reason": "A stale warning in paged generate() is unrelated to Whisper generation failing on empty transcription." + "reason": "Tensor-parallel OOM on Qwen2 is unrelated to the key error loading MobileLLM-125M." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py runtime failure is not the same bug as Qwen2-57B tensor-parallel OOM." + "reason": "Failure to run `3d_parallel.py` is not the same defect as tensor-parallel OOM while loading Qwen2." }, { "left": "issue:44521", "right": "issue:45381", "accept": false, - "reason": "Both are multimodal/chat regressions, but one is assistant mask generation and the other is video vision_position_ids." + "reason": "Both are multimodal, but one is all-zero assistant masks and the other is wrong `vision_position_ids`; different bugs." }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "Missing image_utils import is a different issue from Qwen3.5 chat generation crashing." + "reason": "A bad import from `image_utils` is unrelated to Qwen3.5 crashing with `transformers chat`." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Video input regression for Qwen VL models is unrelated to serve KV-cache indexing." + "reason": "Video-input regression for several Qwen VL models is unrelated to serve-side KV-cache continuation indexing." + }, + { + "left": "issue:43575", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/offload behavior, but one is TP OOM during load and the other is quantization offloading not behaving as expected." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer GPU broadcast TypeError and embedding reinitialization after resize are different code paths and symptoms." + "reason": "Trainer GPU broadcast TypeError and embedding reinitialization in `post_init` are different bugs." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is not the same as a stale warning about num_return_sequences." + "reason": "Compiled-forward inconsistency in generation is unrelated to a stale warning about `num_return_sequences`." }, { "left": "issue:44464", "right": "issue:44871", "accept": false, - "reason": "Generation inconsistency under compilation is unrelated to Gemma eos_token_id config mismatch." + "reason": "Different areas: chunked generation with compiled forward vs EOS token config mismatch." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "A dummy Conv2d in Sam3PixelDecoder is a different model/component than OmDet-Turbo processor size mismatch." + "reason": "`Sam3PixelDecoder` dummy Conv2d and OmDet-Turbo processor input-size mismatch are separate model/processor issues." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation failure are different bugs." + "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation issue are unrelated." }, { - "left": "issue:43827", - "right": "issue:43994", + "left": "issue:44534", + "right": "issue:44913", "accept": false, - "reason": "Docs still referencing pipeline() is not the same as a model producing nonsensical outputs with AutoModel/pipeline." + "reason": "Non-persistent buffer junk in v5 is not the same as `rotary_pct` silently resetting on reload." }, { - "left": "issue:44534", - "right": "issue:44913", + "left": "issue:43827", + "right": "issue:43994", "accept": false, - "reason": "Both are v5-related regressions, but one is buffer initialization corruption and the other is config reload losing rotary_pct." + "reason": "Docs still referencing `pipeline()` is unrelated to SigLIP2 producing nonsensical outputs in `AutoModel`/pipeline usage." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Auto device mapping cudaErrorAssert and category-count overflow are unrelated runtime errors." + "reason": "A800 cudaErrorAssert with Qwen3 auto device mapping is unrelated to the 2^24 category limit runtime error." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "Janus image-generation test failure is unrelated to auto_mappings.py referencing removed Sam3LiteText configs." + "reason": "Janus image-generation test failure and removed Sam3LiteText auto-mapping references are unrelated." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Local dataset loading in a no-trainer script is a different failure from loading a specific multimodal model." + "reason": "Local dataset loading in a training script is unrelated to loading `Phi-4-multimodal-instruct`." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Python 3.14 import/load failure is unrelated to the TensorFlow backend warning from tensorboard-only installs." + "reason": "Python 3.14 import/load failure and the tensorboard-without-TF backend error are distinct startup issues." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge collapse with extended vocab and incorrect visual encoder keys on save are different bugs." } ] }, @@ -8657,10 +8642,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36246", "issue:38175", @@ -8686,23 +8671,23 @@ "issue:44077", "issue:44079", "issue:44112", - "issue:44186", "issue:44190", "issue:44220", + "issue:44246", "issue:44279", "issue:44355", "issue:44466", + "issue:44514", "issue:44561", "issue:44683", "issue:44734", "issue:44743", "issue:44855", "issue:44913", - "issue:45127", + "issue:45003", "issue:45216", "issue:45245", "issue:45276", - "issue:45357", "issue:45375", "issue:45406", "issue:45446", @@ -8712,7 +8697,6 @@ "issue:45632" ], "soft_pairs": [ - "issue:45127|issue:45357", "issue:43388|issue:43582", "issue:44855|issue:45588", "issue:43761|issue:44743", @@ -8721,8 +8705,8 @@ "issue:43540|issue:44734", "issue:43644|issue:44913", "issue:45276|issue:45584", - "issue:44077|issue:44079", "issue:43976|issue:44855", + "issue:44077|issue:44079", "issue:43404|issue:45216", "issue:43550|issue:44190", "issue:44112|issue:45600", @@ -8735,229 +8719,230 @@ "issue:43479|issue:45375", "issue:44734|issue:45632", "issue:36246|issue:38617", - "issue:45406|issue:45632", "issue:44561|issue:45632", + "issue:45406|issue:45632", "issue:44016|issue:44220", - "issue:44186|issue:44734", + "issue:44514|issue:44734", "issue:42175|issue:44279", "issue:38175|issue:43065", "issue:44683|issue:45446", - "issue:44466|issue:44913" + "issue:44466|issue:44913", + "issue:44246|issue:45003" ], - "prepared_review_unit_hash": "6ebd4ce4f81c384b5c5d40646fadb65e913e3b7487141f878ad3b632a6d610fe", + "prepared_review_unit_hash": "207398b18306d23c5c219dc73583950b8a6ccef171f1b8df2d881742d5c6f762", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37242, - "estimated_input_tokens": 9311, - "estimated_eval_tokens": 18878 + "serialized_chars": 37128, + "estimated_input_tokens": 9282, + "estimated_eval_tokens": 18820 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: mostly Transformers regressions, import errors, and model-specific bugs that only look similar at a high level. The soft links are mostly superficial matches on phrases like \"regression\", \"import\", \"save_pretrained\", or \"tied weights\", not true duplicates.", - "confidence": 0.89, - "canonical_issue_reason": "issue:44466 is the clearest representative of the tied-weights / save-pretrained regression theme that appears repeatedly in this cluster, and its failure mode is concrete and well-scoped.", + "summary": "Mostly false-positive soft matches across a heterogeneous set of import errors, model-specific regressions, CI failures, and unrelated runtime issues. No soft pair looks like the same underlying bug/change, and there are no PRs in the cluster.", + "confidence": 0.27, + "canonical_issue_reason": "Issue 41628 is the closest broad representative because it is a top-level `transformers` import/export regression, which is the most generic user-facing API break in this set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44466 is the best single canonical issue in this set: it describes a specific reproducible bug, has a clear code-path, and is more representative than the more incidental import/docs/CI-only reports.", + "best_issue_reason": "41628 is the best single anchor only in a weak sense: it is the most general import-surface failure, while the rest of the cluster is too mixed to support a true canonical duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:45127", - "right": "issue:45357", - "accept": false, - "reason": "Both mention tied weights / save behavior, but one is LoRA merge collapse with extended vocabulary and the other is incorrect visual encoder keys in Qwen3.5 save_pretrained; different bug classes." - }, { "left": "issue:43388", "right": "issue:43582", "accept": false, - "reason": "Label dropping in gather_for_metrics is unrelated to an Apple Silicon TypeError in caching_allocator_warmup." + "reason": "Different subsystems and symptoms: metric label trimming vs Apple Silicon allocator warmup TypeError." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 IndentationError on DebertaV2 import and a flash_attention s_aux=None AttributeError are unrelated failure modes." + "reason": "Unrelated failures: Python 3.13 import/indentation parsing vs flash-attention `s_aux=None` handling." }, { "left": "issue:43761", "right": "issue:44743", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and Qwen3.5 recurrent-state reset with cache are different model/code-path bugs." + "reason": "Both are model regressions, but the concrete bugs differ: hidden_states return handling vs recurrent-state reset with cache." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "A vague data-related issue and a torch audio feature extraction bug are not the same underlying problem." + "reason": "No common bug: a placeholder issue title vs an audio feature extraction helper error." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Compiled Python file errors are not the same as the DebertaV2 @torch.jit.script parsing/indentation issue." + "reason": "Both mention import/compile problems, but one is compiled Python file handling and the other is a Python 3.13 parser/indentation issue." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Qwen3OmniMoe video-input processing and transformers serve KV-cache continuation indexing are distinct code paths." + "reason": "Different code paths: video input processing in Qwen3OmniMoe vs serving KV-cache continuation tensor indexing." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffers being filled with junk is not the same as rotary_pct reverting on config reload." + "reason": "Different regressions: non-persistent buffer serialization vs GPTNeoX config reload persistence." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Gemma4 embedding resizing and Whisper empty-transcription generation after align_special_tokens are unrelated." + "reason": "Unrelated model behaviors: token embedding resize propagation vs Whisper empty-transcription generation." }, { - "left": "issue:44077", - "right": "issue:44079", + "left": "issue:43976", + "right": "issue:44855", "accept": false, - "reason": "Optional post_init handling in patchtsmixer and ModelOutput key assignment are different bugs." + "reason": "Both are version/platform related, but one is a package compatibility complaint and the other is a Python 3.13 import parsing failure." }, { - "left": "issue:43976", - "right": "issue:44855", + "left": "issue:44077", + "right": "issue:44079", "accept": false, - "reason": "A packaging/Python-version compatibility report is not the same as the DebertaV2 import parsing error." + "reason": "Different bugs: `post_init` validation policy vs `ModelOutput` key assignment semantics." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "Both touch tied weights, but one is Mistral3 lm_head tying and the other is Qwen3.5 save_pretrained checkpoint corruption; not mergeable as one fix." + "reason": "Both involve tied weights/serialization, but they are different models and different failure modes, so not one concrete bug." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "torch.compile SDPA failure in Bamba is unrelated to loading a local dataset in an image-classification training script." + "reason": "Torch.compile/SDPA model bug vs local dataset loading in an example script; unrelated." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "A stale GraniteSpeech CI test and broken auto_mappings references for removed Sam3LiteText configs are separate CI failures." + "reason": "Both are CI-facing, but one is a stale test and the other is a broken config mapping; different underlying causes." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Multi-label classification returning empty results and local dataset loading in a different example script are not the same bug." + "reason": "Example script behavior bug vs dataset-loading issue; not the same underlying defect." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Processor.tokenizer regression with images passed to tokenizer is unrelated to Gemma4 resize_token_embeddings not updating output structures." + "reason": "Custom processor/tokenizer regression vs Gemma4 embedding resize propagation; different APIs and code paths." }, { "left": "issue:41950", "right": "issue:44016", "accept": false, - "reason": "A video-classification pipeline processor lookup bug and a notebook syntax error do not share the same underlying defect." + "reason": "Video pipeline processor lookup bug vs a notebook syntax error; unrelated." }, { "left": "issue:45245", "right": "issue:45584", "accept": false, - "reason": "Category-count overflow and Whisper empty transcription after token alignment are unrelated." + "reason": "Different user-visible failures: category-cardinality runtime error vs Whisper empty transcription generation." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Different import errors for different symbols/modules: Qwen2_5_VLImageProcessor vs top-level AutoImageProcessor." + "reason": "Both are import errors, but they are different missing symbols on different APIs and likely separate fixes." }, { "left": "issue:39692", "right": "issue:44190", "accept": false, - "reason": "SigLIP2 docs example errors and local dataset loading failure in an image-classification script are separate issues." + "reason": "SigLIP2 docs/example errors vs image-classification dataset loading; not the same bug." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Phi4MultimodalConfig default-init behavior and missing deepstack_visual_indexes in Qwen3_5MoeVisionConfig are different config bugs." + "reason": "Both are config-related, but one is default-init handling for multimodal configs and the other is a missing strict field in Qwen3_5MoeVisionConfig." }, { "left": "issue:44734", "right": "issue:45632", "accept": false, - "reason": "KV-cache continuation tensor indexing in serve and trust_remote_code cache-path collisions are unrelated." + "reason": "Serving KV-cache indexing crash vs trust_remote_code cache-path collision; unrelated." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Both are import errors, but for different missing names in different modules; no shared code-path is evident." + "reason": "Different missing imports from different modules and different root causes." }, { - "left": "issue:45406", + "left": "issue:44561", "right": "issue:45632", "accept": false, - "reason": "Gemma4Processor missing _tokenizer in serve is not the same as a local-model cache path collision." + "reason": "Removal of `is_torch_fx_available` breaking remote-code models vs local cache-path collision; unrelated." }, { - "left": "issue:44561", + "left": "issue:45406", "right": "issue:45632", "accept": false, - "reason": "Removal of is_torch_fx_available breaking trust_remote_code models is unrelated to cache-directory name collisions." + "reason": "Gemma4Processor missing `_tokenizer` attribute vs trust_remote_code cache collision; different failure surfaces." }, { "left": "issue:44016", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax error and _torch_extract_fbank_features() failure are unrelated." + "reason": "Notebook syntax error vs audio feature extraction helper bug; unrelated." }, { - "left": "issue:44186", + "left": "issue:44514", "right": "issue:44734", "accept": false, - "reason": "LayoutLMv2Tokenizer batching/NER crash and serve KV-cache indexing are different bugs." + "reason": "Both involve Qwen-family multimodal processing, but one is batched chat-template padding and the other is KV-cache continuation indexing." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Missing tensorflow backend in a torch extra install is not the same as a generic dependency issue with transformers." + "reason": "Dependency/backend installation issue vs generic dependency complaint; too vague and not the same bug." }, { "left": "issue:38175", "right": "issue:43065", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and a dummy Conv2d in Sam3PixelDecoder are different model-specific problems." + "reason": "SigLIP2 probability output bug vs Sam3PixelDecoder dummy Conv2d; different models and code paths." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both involve flex_attention, but one is a torch>=2.9 compiled failure and the other is a wrong version check for AuxRequest import; distinct fixes." + "reason": "Compiled flex attention on newer torch vs incorrect PyTorch version guard for AuxRequest import; unrelated." }, { "left": "issue:44466", "right": "issue:44913", "accept": false, - "reason": "Inconsistent lm_head.weight serialization and GPTNeoX rotary_pct reset are separate regression paths, not one duplicate bug." + "reason": "Both are state/persistence regressions, but one is tied-weight serialization and the other is config reload of `rotary_pct`; not the same defect." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import-time performance issue vs unsafe `sys.modules` access; related area but different bugs and fixes." } ] }, @@ -8976,10 +8961,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:42907", "issue:43295", @@ -9003,19 +8988,19 @@ "issue:44075", "issue:44112", "issue:44242", - "issue:44246", "issue:44279", "issue:44336", "issue:44351", "issue:44368", + "issue:44462", "issue:44496", "issue:44521", "issue:44683", "issue:44933", "issue:44964", "issue:44987", - "issue:45003", "issue:45042", + "issue:45081", "issue:45200", "issue:45216", "issue:45278", @@ -9031,15 +9016,14 @@ "issue:45632" ], "soft_pairs": [ - "issue:44246|issue:45003", "issue:43299|issue:45216", "issue:43643|issue:44496", "issue:42907|issue:43441", "issue:44016|issue:44062", "issue:43295|issue:45335", "issue:44964|issue:45600", - "issue:43673|issue:44336", "issue:44112|issue:44964", + "issue:43673|issue:44336", "issue:43976|issue:44987", "issue:43531|issue:45600", "issue:45561|issue:45632", @@ -9060,226 +9044,224 @@ "issue:44351|issue:45278", "issue:44683|issue:45542", "issue:43976|issue:45405", - "issue:43479|issue:43525" + "issue:43479|issue:43525", + "issue:44462|issue:45081" ], - "prepared_review_unit_hash": "00ef7a5f1ae07831723f016cfe37a94f95addb305a7da7154d4287f7f15cb857", + "prepared_review_unit_hash": "cef4760c3638eba31387854bc09a2f12a357248c32a0c756099ca1753df41244", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36626, - "estimated_input_tokens": 9157, - "estimated_eval_tokens": 18570 + "serialized_chars": 36768, + "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 set of unrelated issues that only look similar at a high level (imports, loading failures, multimodal regressions, save/load bugs). I found one plausible duplicate pair around the broader v5 import-regression breakage, but the rest should stay separate.", + "summary": "The cluster is a loose mix of unrelated Transformers issues: model loading/saving regressions, import/config errors, multimodal quirks, and a few utility/docs/test bugs. I don\u2019t see strong duplicates among the soft-edge pairs.", "confidence": 0.84, - "canonical_issue_reason": null, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: a version-upgrade regression causing many import failures. It\u2019s the most representative single issue, even though the cluster as a whole is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45278 is the best cluster anchor because it captures the widest surface area of the reported breakage and is more general than the model-specific or one-off bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44246", - "right": "issue:45003", - "accept": false, - "reason": "Both mention import/performance-ish behavior, but one is about slow imports and the other is an unsafe sys.modules access bug; different failures and code paths." - }, { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Both are Qwen3.5/Qwen3-related, but one is a model loading regression and the other is incorrect save_pretrained output; distinct bugs." + "reason": "Both are Qwen-family regressions, but one is a loading failure for Qwen3VL MoE and the other is a save_pretrained serialization bug for Qwen3.5; not the same code-path." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "One is missing fields from AutoConfig with trust_remote_code, the other is an unrecognized-model/config.json problem; not the same underlying issue." + "reason": "Both concern config/model loading, but one is trust_remote_code returning incomplete config fields and the other is model-type recognition from config.json; different failures." }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Same model family, but saving dequantized weights and FlashAttention load failure are different code paths and symptoms." + "reason": "Different models and different bugs: dequantized save failures vs FlashAttention loading regression." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Notebook syntax error vs AddedToken constructor regression; unrelated." + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Processor/tokenizer image handling regression versus resize_token_embeddings not updating decoder embeddings; different components." + "reason": "Different subsystems and symptoms: processor/tokenizer multimodal regression vs resize_token_embeddings not updating decoder weights." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "A Phi-4 multimodal loading issue versus stale auto_mappings references to removed configs; different bugs." + "reason": "Phi-4 multimodal loading failure and removed Sam3LiteText auto-mapping references are separate model-specific issues." }, { - "left": "issue:43673", - "right": "issue:44336", + "left": "issue:44112", + "right": "issue:44964", "accept": false, - "reason": "Chunked prefill cache regression versus ANSI escape codes in loading_report; unrelated." + "reason": "CI test flake in GraniteSpeech is not the same bug as a model-load failure for Phi-4 multimodal." }, { - "left": "issue:44112", - "right": "issue:44964", + "left": "issue:43673", + "right": "issue:44336", "accept": false, - "reason": "CI test flake in GraniteSpeech versus a model loading failure; not the same underlying problem." + "reason": "Chunked-prefill cache regression and ANSI-color output in loading_report are unrelated." }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Python version compatibility issue versus a model load failure on one package; no shared root cause." + "reason": "Python version compatibility regression and a specific model-loading failure are different issues." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior versus auto_mappings CI breakage; unrelated." + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup are unrelated." }, { "left": "issue:45561", "right": "issue:45632", "accept": false, - "reason": "Parallel test-file race versus trust_remote_code cache-path collision; both are filesystem-ish but not the same bug." + "reason": "A pytest-xdist file race and a trust_remote_code cache-path collision are different concurrency problems in different areas." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Different import failures in different modules; no clear shared root cause." + "reason": "Both are import-related, but one is a missing Qwen2.5-VL export and the other is a nonexistent image_utils import; not the same symbol or fix." }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "Embedding resize bug versus Whisper empty-transcription generation failure; unrelated." + "reason": "resize_token_embeddings behavior for t5gemma and Whisper empty-transcription generation are unrelated." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "One is a Python-version breakage, the other is a vague dependency issue; too broad to be the same bug." + "reason": "Python support regression and a generic dependency issue are not the same underlying bug." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Unrecognized model config versus torchvision requirement regression; unrelated." + "reason": "Model config recognition and PIL/torchvision backend requirements are different code paths." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both are model-load failures, but for different models and likely different causes." + "reason": "Different model families and different loading failures." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "dtype propagation bug in BLIP-2 versus incorrect Qwen3.5 saved checkpoint contents; different behavior." + "reason": "dtype propagation for BLIP2 and Qwen3.5 save_pretrained serialization are unrelated." }, { "left": "issue:43526", "right": "issue:43901", "accept": false, - "reason": "Image processor label reduction bug versus docs/behavior mismatch in text classification pipeline; unrelated." + "reason": "BeitImageProcessorFast label reduction bug and TextClassificationPipeline docs mismatch are not duplicates." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Both are Qwen3.5/Qwen3 model regressions, but loading failure and incorrect visual-encoder save keys are different defects." + "reason": "Both involve Qwen3.5/Qwen3 serialization, but one is a loading failure and the other is incorrect visual-encoder keys on save; close but not the same bug." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "SGD argument handling versus a tied-embeddings warning; unrelated." + "reason": "SGD optimizer argument handling and tie_word_embeddings warnings are unrelated." }, { "left": "issue:44521", "right": "issue:45325", "accept": false, - "reason": "Both are multimodal/Qwen issues, but assistant mask generation and rope-index scaling are separate code paths." + "reason": "Both are multimodal, but assistant mask generation and rope-index scaling are different concrete bugs." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Load-balancing loss behavior versus mm_token_type_ids defaults; different model-training bugs." + "reason": "Router loss calculation and mm_token_type_ids defaults are different training-time behaviors." }, { "left": "issue:43381", "right": "issue:45593", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction versus D-FINE auxiliary-loss behavior; unrelated." + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Pipeline docs/behavior mismatch versus model config loading failure; unrelated." + "reason": "Documentation mismatch and model config loading failure are not the same bug." }, { "left": "issue:43927", "right": "issue:45632", "accept": false, - "reason": "Saved custom token IDs being lost versus trust_remote_code cache collision; different persistence/path issues." + "reason": "DiaConfig save/load token-id loss and local trust_remote_code cache collision are different serialization/caching issues." }, { "left": "issue:44351", "right": "issue:45278", - "accept": true, - "reason": "The specific missing HybridCache import looks like one instance of the broader v5.5 import-regression report that causes multiple import errors after the same update." + "accept": false, + "reason": "Both mention imports, but 45278 is a broad report of many failures while 44351 is a specific missing HybridCache export; not enough to treat as one duplicate." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Compiled flex_attention regression versus a TensorFlow backend/install issue; unrelated." + "reason": "Compiled flex_attention on torch>=2.9 and tensorboard/TensorFlow backend detection are unrelated." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Python support breakage versus an unreleased MIN_PEFT_VERSION bump; unrelated." + "reason": "Python version support and PEFT minimum-version bump are different dependency problems." }, { "left": "issue:43479", "right": "issue:43525", "accept": false, - "reason": "Different config/model attribute initialization bugs in different model families." + "reason": "Phi4MultimodalConfig default subconfigs and Llama4Config missing pad_token_id are separate config bugs." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashing are different tokenizer-loading failures." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately cautious. The soft-edge verdicts are conservative overall, and the one accepted pair is a plausible duplicate within the same broader import-regression context." - }, - "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 }, @@ -9291,10 +9273,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:38175", "issue:39692", @@ -9316,6 +9298,7 @@ "issue:43784", "issue:43854", "issue:43901", + "issue:43994", "issue:44016", "issue:44162", "issue:44261", @@ -9337,7 +9320,6 @@ "issue:44898", "issue:44987", "issue:45020", - "issue:45081", "issue:45276", "issue:45290", "issue:45397", @@ -9346,11 +9328,10 @@ "issue:45593" ], "soft_pairs": [ - "issue:44462|issue:45081", "issue:41950|issue:45020", "issue:44297|issue:44462", - "issue:44496|issue:44877", "issue:44279|issue:45405", + "issue:44496|issue:44877", "issue:41950|issue:43650", "issue:43299|issue:43854", "issue:44462|issue:44779", @@ -9359,225 +9340,232 @@ "issue:39692|issue:43901", "issue:43650|issue:45290", "issue:43723|issue:44162", - "issue:44016|issue:44291", "issue:43450|issue:44279", + "issue:44016|issue:44291", "issue:42994|issue:43317", - "issue:44261|issue:44263", "issue:43441|issue:44683", "issue:43262|issue:44521", + "issue:44261|issue:44263", "issue:43065|issue:44898", - "issue:44623|issue:45397", "issue:44466|issue:45276", "issue:43122|issue:43650", + "issue:44623|issue:45397", "issue:44623|issue:44749", "issue:43784|issue:45542", "issue:44671|issue:44779", "issue:43673|issue:43761", "issue:44987|issue:45542", - "issue:43653|issue:45593" + "issue:43653|issue:45593", + "issue:43901|issue:43994" ], - "prepared_review_unit_hash": "0cbba02162bc3b3ac0ecdf1889c69dfbf5ed2106d5873b09886d84182025c38d", + "prepared_review_unit_hash": "6775c67a49a9ea93be93acbb5f00430962cc7aeab7611b22389da0af1753b8be", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36365, - "estimated_input_tokens": 9092, - "estimated_eval_tokens": 18440 + "serialized_chars": 36281, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: it mixes tokenizer, processor, model-loading, quantization, attention-kernel, docs, and unrelated model-specific bugs. There are a few loose tokenizer/loading similarities, but none are strong enough to merge as duplicates. No PRs are present.", - "confidence": 0.77, - "canonical_issue_reason": "No single issue cleanly represents the whole cluster; if one representative must be chosen, issue 43723 is the broadest tokenizer-loading regression and the closest shared theme among the more similar reports.", + "summary": "The candidate pairs are mostly same-subsystem lookalikes rather than true duplicates: tokenizer, processor/chat_template, config-loading, and serialization regressions are mixed together. I would keep them separate.", + "confidence": 0.67, + "canonical_issue_reason": "issue:44521 is the strongest anchor in this set: it is open, well-discussed, has inbound references, and clearly describes a concrete multimodal `apply_chat_template` bug with a focused reproduction path.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43723 is the best representative because it is concrete, broadly scoped, and closest to the recurring tokenizer/AutoTokenizer loading complaints; the other items are mostly separate bugs in different subsystems.", + "best_issue_reason": "issue:43994 is the cleanest standalone bug report for triage: it names the exact model, the observable bad behavior, and the API surfaces involved (`AutoModel` and `pipeline`).", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44462", - "right": "issue:45081", - "accept": false, - "reason": "Different failures: AutoTokenizer ignoring tokenizer.json vs a Mistral regex patch crash." - }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Video-classification processor lookup bug and remote_code loading regressions are unrelated." + "reason": "Different bugs: video-classification image-processor lookup vs broad remote_code loading regressions." }, { "left": "issue:44297", "right": "issue:44462", "accept": false, - "reason": "Both concern tokenizer persistence/loading, but one is save_pretrained metadata and the other is repository tokenizer.json selection." + "reason": "Related tokenizer area, but one is save-time class metadata mismatch and the other is load-time ignoring `tokenizer.json`." }, { - "left": "issue:44496", - "right": "issue:44877", + "left": "issue:44279", + "right": "issue:45405", "accept": false, - "reason": "Both involve config loading, but the concrete bugs are different: missing model_type vs strict config rejection for granite_speech." + "reason": "Unrelated: generic dependency issue vs a PEFT minimum-version bump problem." }, { - "left": "issue:44279", - "right": "issue:45405", + "left": "issue:44496", + "right": "issue:44877", "accept": false, - "reason": "Generic dependency issue and a PEFT minimum-version bump are not the same bug." + "reason": "Both concern config/model loading, but they target different failure modes and different models; not the same underlying bug." }, { "left": "issue:41950", "right": "issue:43650", "accept": false, - "reason": "Video-classification processor lookup is unrelated to the placeholder 'ADD THE DATA' issue." + "reason": "Completely unrelated subjects." }, { "left": "issue:43299", "right": "issue:43854", "accept": false, - "reason": "Different model-loading failures for different models and code paths." + "reason": "Both are model-loading failures, but for different models and likely different code paths." }, { "left": "issue:44462", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer-related, but one is AutoTokenizer file selection and the other is Deepseek tokenization correctness." + "reason": "Different tokenizer issues: repository tokenizer selection vs incorrect tokenization behavior in v5." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Different regressions: processor.tokenizer access/passing images to tokenizer vs Deepseek tokenization output changes." + "reason": "Processor/tokenizer regression vs Deepseek tokenization output bug; not the same concrete defect." }, { "left": "issue:38175", "right": "issue:44617", "accept": false, - "reason": "Unexpected zero probabilities and CUDA OOM are unrelated symptoms." + "reason": "Different model families and symptoms; SigLIP2 probabilities vs Sam3Video OOM." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Both are docs-related, but they affect different examples and different mismatches." + "reason": "Both mention docs, but one is a SigLIP2 example error report and the other is a documentation mismatch for pipeline return behavior." }, { "left": "issue:43650", "right": "issue:45290", "accept": false, - "reason": "Unrelated: placeholder issue versus a chat_template crash on tool-call assistant messages." + "reason": "No overlap beyond `apply_chat_template`-adjacent wording; one is a placeholder issue and the other is a specific tool-call crash." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "Tokenizer-loading regression and an ESM2 model breakage are different underlying problems." - }, - { - "left": "issue:44016", - "right": "issue:44291", - "accept": false, - "reason": "Notebook syntax error and init_empty_weights argument regression are unrelated." + "reason": "Tokenizer loading regression vs ESM2 model breakage; different artifacts." }, { "left": "issue:43450", "right": "issue:44279", "accept": false, - "reason": "Batched video processor shape bug does not match a generic dependency issue." + "reason": "Video processor batching shape bug is unrelated to a dependency issue." }, { - "left": "issue:42994", - "right": "issue:43317", + "left": "issue:44016", + "right": "issue:44291", "accept": false, - "reason": "Both touch quantization/offload, but one is saving quantized models and the other is loading dequantized models with gpu+cpu offload." + "reason": "Notebook syntax error and `init_empty_weights` argument incompatibility are separate failures." }, { - "left": "issue:44261", - "right": "issue:44263", + "left": "issue:42994", + "right": "issue:43317", "accept": false, - "reason": "Different bugs: missing rms_norm_eps precision issue vs torch.split return values in a different indexer." + "reason": "Both involve quantization/offload, but one is save failure and the other is dequantized load with device_map offload; different bug paths." }, { "left": "issue:43441", "right": "issue:44683", "accept": false, - "reason": "Both involve attention acceleration, but they are different kernels/code paths and different failure modes." + "reason": "Attention backend failures, but one is FlashAttention on a specific model and the other is compiled flex_attention on newer torch." }, { "left": "issue:43262", "right": "issue:44521", "accept": false, - "reason": "Both mention chat_template behavior, but one is audio sampling-rate defaulting and the other is multimodal assistant_masks being all zero." + "reason": "Both involve processors and `apply_chat_template`, but one is a default sampling-rate issue and the other is assistant mask generation for multimodal inputs." }, { - "left": "issue:43065", - "right": "issue:44898", + "left": "issue:44261", + "right": "issue:44263", "accept": false, - "reason": "Different vision-model bugs in different components; only a broad subsystem overlap." + "reason": "Different model internals and different symptoms." }, { - "left": "issue:44623", - "right": "issue:45397", + "left": "issue:43065", + "right": "issue:44898", "accept": false, - "reason": "Processor save_pretrained missing files and gemma-4 zero3 from_pretrained are different save/load failures." + "reason": "Dummy convolution implementation vs interpolation-related image classification failure; not the same bug." }, { "left": "issue:44466", "right": "issue:45276", "accept": false, - "reason": "Different serialization/runtime behaviors: lm_head.weight save inconsistency vs resize_token_embeddings not propagating." + "reason": "Serialization of tied weights vs resize-token-embeddings propagation; related area, but different concrete defects." }, { "left": "issue:43122", "right": "issue:43650", "accept": false, - "reason": "Different issues; one is a tokenizer regression and the other is an unrelated placeholder title." + "reason": "Tokenizer behavior regression is unrelated to the placeholder issue." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Processor save-pretrained missing files vs zero3 from_pretrained loading failure; separate code paths." }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "Processor serialization bug and a performance regression after upgrade are unrelated." + "reason": "Unrelated: processor saving bug vs a data-filtering performance complaint." }, { "left": "issue:43784", "right": "issue:45542", "accept": false, - "reason": "Both are environment/dependency flavored, but one is a missing nn import and the other is a TensorFlow backend packaging problem." + "reason": "Different dependency/import failures affecting different backends." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Different model/tokenizer correctness bugs for different models." + "reason": "CamemBERT masked-LM prediction regression and Deepseek tokenization regression are distinct model/tokenizer issues." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Generation cache regression and CLIPVision hidden_states regression are separate model-path failures." + "reason": "Generation cache/chunked-prefill bug vs CLIPVision hidden-states regression; different code paths." }, { "left": "issue:44987", "right": "issue:45542", "accept": false, - "reason": "Model-loading failure for a specific repo and a tf-backend dependency issue are unrelated." + "reason": "Loading a specific model under v5.1.0 vs optional TensorFlow backend import issue." + }, + { + "left": "issue:45020", + "right": "issue:45276", + "accept": false, + "reason": "Remote-code loading regressions and embedding-resize propagation are unrelated." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "No shared underlying bug; one is a precision/config issue, the other is a torch.split indexing issue." }, { - "left": "issue:43653", - "right": "issue:45593", + "left": "issue:44297", + "right": "issue:44779", "accept": false, - "reason": "BigBirdTokenizer special-token decode bug and D-FINE auxiliary-loss behavior are unrelated." + "reason": "One is save_pretrained metadata, the other is tokenization output correctness; not duplicates." } ] }, @@ -9596,10 +9584,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36246", "issue:41628", @@ -9619,7 +9607,6 @@ "issue:43881", "issue:43901", "issue:43976", - "issue:43994", "issue:44016", "issue:44162", "issue:44188", @@ -9639,6 +9626,7 @@ "issue:44589", "issue:44617", "issue:44661", + "issue:44857", "issue:45003", "issue:45200", "issue:45250", @@ -9651,24 +9639,23 @@ "issue:45593" ], "soft_pairs": [ - "issue:43901|issue:43994", - "issue:44297|issue:45397", "issue:42175|issue:43976", "issue:44361|issue:45593", + "issue:44297|issue:45397", "issue:43475|issue:44016", "issue:45561|issue:45563", "issue:43824|issue:43976", - "issue:43352|issue:45250", "issue:43295|issue:44617", + "issue:43352|issue:45250", "issue:43525|issue:43881", "issue:44589|issue:45362", "issue:44188|issue:44336", "issue:41628|issue:43976", - "issue:44295|issue:45397", "issue:43901|issue:44246", "issue:42491|issue:43299", "issue:41950|issue:44220", "issue:44336|issue:44373", + "issue:44295|issue:45397", "issue:44242|issue:45593", "issue:44462|issue:45356", "issue:44661|issue:45542", @@ -9680,216 +9667,217 @@ "issue:43526|issue:45003", "issue:44336|issue:44485", "issue:43673|issue:44561", - "issue:44162|issue:45200" + "issue:44162|issue:45200", + "issue:44242|issue:44857" ], - "prepared_review_unit_hash": "70dfeb08b2d5286599d58cf93fa8bf9e4b712feafb235bb4b10e01795d4ef228", + "prepared_review_unit_hash": "478478f23f51660d6591cb4e3235c76de4de77e65cb37715452cc810a9509f6c", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36330, - "estimated_input_tokens": 9083, - "estimated_eval_tokens": 18422 + "serialized_chars": 36304, + "estimated_input_tokens": 9076, + "estimated_eval_tokens": 18408 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is very heterogeneous: most soft pairs are only loosely related by broad themes like import errors, tokenizer regressions, or model-loading failures, but they do not describe the same concrete bug/change. I would not merge any of the suggested soft edges.", - "confidence": 0.73, - "canonical_issue_reason": "No strong canonical issue exists for the whole set because the items span unrelated failures. If a representative must be chosen, issue #43824 is the strongest broad regression report in the group (high discussion and inbound refs) and sits near the general import/error theme.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers bug reports, with many model-specific regressions and import/load failures but no clear duplicate group. The closest common theme is v5.x breakage around model loading and processor/tokenizer imports.", + "confidence": 0.23, + "canonical_issue_reason": "Issue 43299 is the best representative of the repeated theme here: a user-facing v5.0.0.dev0 regression that breaks model loading, which is broader than the many single-model errors in the set.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43824 is the best available representative only by signal/impact; it is not truly a duplicate of the others, but it is the clearest high-traffic regression among the import-related reports.", + "best_issue_reason": "43299 is the strongest canonical choice because it is a clear, current-version regression in a core code path (model loading) and reads like the most central problem among the heterogeneous reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43901", - "right": "issue:43994", - "accept": false, - "reason": "Both involve pipeline/model behavior, but one is a docs/behavior mismatch for `return_all_scores` and the other is a SigLIP2 output quality regression. Different bugs and code paths." - }, - { - "left": "issue:44297", - "right": "issue:45397", - "accept": false, - "reason": "Tokenizer class metadata mismatch is unrelated to the Gemma-4 Zero3 `from_pretrained` failure; different failure modes and subsystems." - }, { "left": "issue:42175", "right": "issue:43976", "accept": false, - "reason": "One is about missing TensorFlow in a `.[torch]` install, the other about Python version compatibility in 5.1.0. Same ecosystem, but not the same bug." + "reason": "Different problems: backend extra packaging vs Python version compatibility/install breakage." }, { "left": "issue:44361", "right": "issue:45593", "accept": false, - "reason": "`MLukeTokenizer` task failure and D-FINE auxiliary-loss behavior are unrelated model-specific issues." + "reason": "Unrelated model/task bugs: MLukeTokenizer task failure vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer config metadata mismatch is unrelated to gemma-4 zero3/from_pretrained loading." }, { "left": "issue:43475", "right": "issue:44016", "accept": false, - "reason": "SAM3 video output attribute error and a notebook syntax error are not the same underlying problem." + "reason": "SAM3Video attribute error vs a notebook syntax error; no shared underlying bug." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "A pytest-xdist file race and a stale generation warning are distinct test/runtime issues." + "reason": "pytest-xdist file race is unrelated to a stale generation warning." }, { "left": "issue:43824", "right": "issue:43976", "accept": false, - "reason": "Qwen2_5_VL import failure and Python 3.9/3.10 compatibility are different problems." + "reason": "Importing a specific Qwen2.5 VL class is unrelated to Python 3.9+/3.10+ support." }, { - "left": "issue:43352", - "right": "issue:45250", + "left": "issue:43295", + "right": "issue:44617", "accept": false, - "reason": "A model not supporting Flash Attention 2 yet is not the same as a generic Flash Attention 2 issue." + "reason": "Processor/tokenizer regression and CUDA OOM are different failure modes and code paths." }, { - "left": "issue:43295", - "right": "issue:44617", + "left": "issue:43352", + "right": "issue:45250", "accept": false, - "reason": "Processor/tokenizer regression and SAM3 CUDA OOM are unrelated." + "reason": "Both mention Flash Attention, but one is an unsupported-model error and the other is a generic FA2 issue; not the same bug." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Both are model-loading failures, but they involve different models and likely different missing/config attributes." + "reason": "Missing Llama4Config pad_token_id is unrelated to glm-4v-9b loading failure." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage deserialization failure is unrelated to a Qwen3.5 chat crash." + "reason": "Float8 storage type error and Qwen3.5 chat crash are different runtime failures." }, { "left": "issue:44188", "right": "issue:44336", "accept": false, - "reason": "Attention kernel divergence under `torch.compile` and ANSI codes in loading_report are unrelated." + "reason": "torch.compile attention-kernel divergence is unrelated to ANSI codes in loading_report." }, { "left": "issue:41628", "right": "issue:43976", "accept": false, - "reason": "`AutoImageProcessor` import failure is not the same as the Python-version compatibility report." - }, - { - "left": "issue:44295", - "right": "issue:45397", - "accept": false, - "reason": "Buffer access on `position_ids` and Gemma-4 Zero3 loading are different bugs." + "reason": "AutoImageProcessor import failure is a separate issue from package Python-version compatibility." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "A pipeline docs/behavior mismatch and slow `import transformers` do not share the same concrete code path." + "reason": "Docs mismatch for return_all_scores has nothing to do with slow imports." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both mention Qwen3-MoE-family models, but one is a LoRA compatibility issue across branches and the other is a Qwen3VL-MoE loading regression." + "reason": "Both involve Qwen3 MoE, but one is a LoRA compatibility issue on hf4.x\u2192hf5.x and the other is a direct loading regression in dev0; not clearly the same concrete bug." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video-classification looking for image processors and `_torch_extract_fbank_features()` are different pipeline bugs." + "reason": "Video-classification using image processors is unrelated to _torch_extract_fbank_features()." }, { "left": "issue:44336", - "right": "issue:44373", + "right": "issue:44485", "accept": false, - "reason": "ANSI output in `loading_report` and a wrong docstring for `position_ids` are unrelated." + "reason": "Terminal ANSI escape handling and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer access error is not the same as gemma-4 zero3 from_pretrained loading failure." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "Both concern missing auxiliary-style losses, but they affect different models and different control flags; not the same bug." + "reason": "Missing load-balancing loss and D-FINE auxiliary losses are different model-training bugs." }, { "left": "issue:44462", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they target different models and different tokenizer logic." + "reason": "Tokenizer.json being ignored is a different tokenizer regression than Kimi-K2.5 codec handling." }, { "left": "issue:44661", "right": "issue:45542", "accept": false, - "reason": "`add-new-model-like` mapping logic and a TensorBoard/TensorFlow backend install issue are unrelated." + "reason": "Model-mapping code generation bug and missing TensorFlow backend installation are unrelated." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Slow import and unrecognized model/config loading are not the same failure." + "reason": "Slow import performance is unrelated to an unrecognized-model/config.json error." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "Unsafe `sys.modules` access in `modeling_utils` and a test-worker file race are unrelated." + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to a pytest-xdist capture-file race." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "These are both import errors, but they concern different symbols and different packages/modules." + "reason": "Qwen2.5 VL image-processor import error and huggingface_hub is_offline_mode import error are different missing symbols." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Different missing imports from different parts of Transformers; not a single underlying defect." + "reason": "Different missing exports: Qwen2.5 VLImageProcessor vs HybridCache." }, { "left": "issue:43531", "right": "issue:45362", "accept": false, - "reason": "Both involve Qwen-family models, but one is a sliding-window issue and the other is a chat crash; too different to merge." + "reason": "Qwen3-MoE sliding_window behavior and Qwen3.5 chat crashes are not the same concrete bug." }, { "left": "issue:43526", "right": "issue:45003", "accept": false, - "reason": "`reduce_labels` returning one label and unsafe `sys.modules` access are unrelated implementation bugs." + "reason": "BEiT reduce_labels bug is unrelated to sys.modules access in modeling_utils." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "Terminal-color handling in loading_report and GLM-5 RoPE implementation are unrelated." + "reason": "ANSI-code output issue is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:43673", "right": "issue:44561", "accept": false, - "reason": "Chunked-prefill cache handling and removal of `is_torch_fx_available` affect different code paths and problems." + "reason": "Chunked-prefill cache missing and removal of is_torch_fx_available are separate regressions." }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 breakage and Gemma-4 `mm_token_type_ids` defaults are unrelated model-specific regressions." + "reason": "ESM2 breakage is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Load-balancing loss and LwDetr float16 AMP crashes are different issues." } ] }, @@ -9908,10 +9896,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36246", "issue:41628", @@ -9929,24 +9917,21 @@ "issue:43540", "issue:43596", "issue:43653", - "issue:43673", "issue:43901", "issue:44060", "issue:44077", "issue:44162", - "issue:44242", "issue:44246", "issue:44295", "issue:44336", "issue:44368", "issue:44466", "issue:44496", - "issue:44734", "issue:44749", "issue:44779", - "issue:44857", "issue:44871", "issue:44877", + "issue:44933", "issue:44977", "issue:44987", "issue:44995", @@ -9957,16 +9942,17 @@ "issue:45292", "issue:45335", "issue:45375", + "issue:45397", "issue:45459", "issue:45563", "issue:45584", - "issue:45588" + "issue:45588", + "issue:45656" ], "soft_pairs": [ - "issue:44295|issue:44749", - "issue:44242|issue:44857", "issue:44246|issue:44871", "issue:44877|issue:45375", + "issue:44295|issue:44749", "issue:36246|issue:43596", "issue:43066|issue:44779", "issue:43441|issue:43540", @@ -9987,207 +9973,208 @@ "issue:43262|issue:44246", "issue:43653|issue:44162", "issue:43475|issue:43540", - "issue:44466|issue:45335", "issue:44368|issue:45563", + "issue:44466|issue:45335", "issue:44368|issue:44977", - "issue:43673|issue:44734" + "issue:45397|issue:45656", + "issue:44933|issue:45459" ], - "prepared_review_unit_hash": "b1afb76ba21b8b63d237b7b6e63c48c624cbf674c4d2f9ad07a463ac1540ff8e", + "prepared_review_unit_hash": "433a7361c79fe07a5ffe4217778c40995cd6b1955d83400d011812e9f6ae7808", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 36283, - "estimated_input_tokens": 9071, - "estimated_eval_tokens": 18398 + "serialized_chars": 35622, + "estimated_input_tokens": 8906, + "estimated_eval_tokens": 18068 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The items form a loose collection of Transformers v5 regressions and model-specific bugs, but none of the soft pairs look like true duplicates of the same underlying issue. No PRs are present.", - "confidence": 0.27, - "canonical_issue_reason": "issue:44987 is the broadest, user-facing loading failure among the set and the best weak anchor for this otherwise heterogeneous cluster.", + "summary": "This cluster is a loose mix of Transformers v5-era regressions across tokenization, config strictness, embedding resizing/tied weights, generation, DeepSpeed, and flash-attn. I don\u2019t see any pair that clearly describes the same concrete bug, so I rejected all soft edges.", + "confidence": 0.69, + "canonical_issue_reason": "Issue 44466 is the broadest representative of the tied-weights / serialization regression subgroup, which is the most connected theme in this otherwise heterogeneous cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44987 is the most central open issue and the closest thing to a cluster anchor, but the surrounding issues are too diverse to treat as one duplicate family.", + "best_issue_reason": "44466 is the best overall anchor issue because it is concrete, regression-oriented, and closest to several nearby weight-handling reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44295", - "right": "issue:44749", - "accept": false, - "reason": "One is a position_ids buffer access bug; the other is a general slowdown after upgrading Transformers. Different failure modes and code paths." - }, - { - "left": "issue:44242", - "right": "issue:44857", - "accept": false, - "reason": "Load-balancing loss logic and an AMP/CUDA crash in LwDetrImageLoss are unrelated bugs." - }, { "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Import-time slowness and Gemma eos_token_id inconsistency do not describe the same underlying defect." + "reason": "Different bugs: intermittent import slowdown vs Gemma eos_token_id config mismatch." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Both involve config handling, but one is strict config loading and the other is a missing Qwen vision field; not the same bug." + "reason": "Both involve strict config handling, but they are different model/config failures with different symptoms." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated: position_ids buffer access error vs performance regression after upgrade." }, { "left": "issue:36246", "right": "issue:43596", "accept": false, - "reason": "An import error for Qwen2.5 VL image processor is unrelated to a DeepSpeed ZeRO3/BertModel index error." + "reason": "Different code paths: missing Qwen2_5_VLImageProcessor import vs DeepSpeed ZeRO3/BertModel index error." }, { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Tokenizer decoder-type warnings and incorrect DeepSeek tokenizer output are both tokenizer-related but are not the same concrete issue." + "reason": "Both tokenizer-adjacent, but one is a decoder-type warning and the other is incorrect DeepSeek tokenization behavior." }, { "left": "issue:43441", "right": "issue:43540", "accept": false, - "reason": "FlashAttention failure for Ministral-3 and a video-input ValueError in Qwen3OmniMoe affect different models and code paths." + "reason": "Different models and failures: flash-attn issue for Ministral-3 vs video input validation in Qwen3OmniMoe." }, { "left": "issue:45292", "right": "issue:45584", "accept": false, - "reason": "Embedding resize propagation and Whisper empty-transcription generation after align_special_tokens are different bugs." + "reason": "Resize-token-embeddings bug is unrelated to Whisper empty-transcription behavior after align_special_tokens." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "DeepSeek tokenizer correctness and resize_token_embeddings behavior are distinct problems." + "reason": "Tokenizer correctness bug vs embedding-resize propagation bug; not the same underlying issue." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Model loading on transformers>=5.1.0 and protobuf decode-error handling are unrelated issues." + "reason": "Model loading failure vs protobuf-specific error masking in tokenizer decoding." }, { "left": "issue:41628", "right": "issue:41950", "accept": false, - "reason": "An AutoImageProcessor import error and video-classification pipeline choosing image processors are adjacent but not the same failure." + "reason": "Related around processors, but one is a missing import and the other is a pipeline selecting the wrong processor type." }, { "left": "issue:42886", "right": "issue:45229", "accept": false, - "reason": "Offline tokenizer cache loading and Gemma4 multi-GPU OOM are unrelated." + "reason": "Offline cache loading bug vs Gemma4 multi-GPU OOM; completely different problems." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "A stale indexer cache on second forward and missing default mm_token_type_ids are different bugs." + "reason": "Stale indexer cache on second forward pass is unrelated to missing mm_token_type_ids defaults." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "Docs mismatch for return_all_scores and unsafe sys.modules access in modeling_utils are unrelated." + "reason": "Docs mismatch vs unsafe sys.modules access in modeling_utils; different layers and symptoms." }, { "left": "issue:43065", "right": "issue:43408", "accept": false, - "reason": "A dummy Conv2d in Sam3PixelDecoder and a sam3_video vs sam3_tracker config warning are not the same defect." + "reason": "Sam3 pixel decoder dummy conv issue vs sam3_video/sam3_tracker config warning; not the same bug." }, { "left": "issue:43232", "right": "issue:44336", "accept": false, - "reason": "Generation kwargs update after sync_gpus and ANSI codes in loading_report are unrelated." + "reason": "Generation kwargs update bug vs ANSI code emission in loading_report; unrelated." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MOE weight conversion with accelerate+deepspeed and a ZeRO3 deque pop error are different DeepSpeed failures." + "reason": "Qwen3 MoE weight conversion problem vs DeepSpeed ZeRO3 deque underflow; different failure modes." }, { "left": "issue:44496", "right": "issue:45003", "accept": false, - "reason": "Config model-type recognition and sys.modules access safety are unrelated." + "reason": "Unrecognized model config vs unsafe sys.modules access; not the same underlying cause." }, { "left": "issue:44060", "right": "issue:44077", "accept": false, - "reason": "Incorrect tied-weights warnings in Qwen3-Next and patchtsmixer post_init policy are not the same change." + "reason": "Tied-weights warning bug vs patchtsmixer post_init policy change; unrelated." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "GlmMoeDsa stale cache and flash_attention s_aux=None crashes are different runtime errors." + "reason": "Cache/indexer crash vs flash_attention s_aux=None AttributeError; different subsystems." }, { "left": "issue:43295", "right": "issue:45292", "accept": false, - "reason": "A processor.tokenizer regression in v4.57.5 and embedding resize propagation are not the same bug." + "reason": "Processor.tokenizer regression vs resize_token_embeddings not updating output embeddings; separate code paths." }, { "left": "issue:43262", "right": "issue:44246", "accept": false, - "reason": "Audio chat-template sampling-rate defaults and import slowness are unrelated." + "reason": "Audio sampling-rate default bug vs occasional import slowness; no duplication." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "BigBirdTokenizer special-token registration and ESM2 breakage are distinct model/tokenizer issues." + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 breakage; different models and symptoms." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "SAM3 video attribute loss and Qwen3OmniMoe video-input processing fail in different parts of the stack." + "reason": "Missing Sam3VisionEncoderOutput attribute vs Qwen3OmniMoe video processing ValueError; unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "Tied-weights warning vs stale paged generate warning; different generation/config issues." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Device-dependent tied-weight serialization and t5gemma decoder embedding resizing are related only at a broad embedding level." + "reason": "Both touch embedding resizing / weight tying, but one is serialization behavior and the other is decoder.embed_tokens not updating; not the same bug." }, { "left": "issue:44368", - "right": "issue:45563", + "right": "issue:44977", "accept": false, - "reason": "A tied-word-embeddings warning and a stale num_return_sequences warning are different warning logic bugs." + "reason": "Weight-tying warning in LoRA fine-tuning vs Qwen3.5 flash-attention generation failure." }, { - "left": "issue:44368", - "right": "issue:44977", + "left": "issue:45397", + "right": "issue:45656", "accept": false, - "reason": "Qwen3.5 flash-attention generation failure is not the same as a tie_word_embeddings warning." + "reason": "DeepSpeed zero3 from_pretrained issue vs optimizer.step being called twice; distinct DeepSpeed bugs." }, { - "left": "issue:43673", - "right": "issue:44734", + "left": "issue:44933", + "right": "issue:45459", "accept": false, - "reason": "Chunked-prefill cache missing and KV-cache continuation indexing errors are both generation-related but not the same underlying bug." + "reason": "Missing import from image_utils vs tokenizer decode-error masking without protobuf; unrelated." } ] }, @@ -10206,10 +10193,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:28282", "issue:30990", @@ -10226,7 +10213,6 @@ "issue:43643", "issue:43646", "issue:43673", - "issue:43701", "issue:43704", "issue:43873", "issue:43931", @@ -10239,6 +10225,7 @@ "issue:44368", "issue:44623", "issue:44661", + "issue:44734", "issue:44749", "issue:44869", "issue:44933", @@ -10256,27 +10243,28 @@ "issue:45375", "issue:45381", "issue:45397", - "issue:45459", "issue:45478", - "issue:45600" + "issue:45600", + "issue:45656" ], "soft_pairs": [ - "issue:44933|issue:45459", + "issue:43673|issue:44734", "issue:44869|issue:45356", "issue:45137|issue:45600", "issue:43931|issue:44368", - "issue:44297|issue:44623", "issue:43232|issue:43673", - "issue:43673|issue:45278", + "issue:44297|issue:44623", "issue:43646|issue:44162", - "issue:44361|issue:44749", + "issue:43673|issue:45278", "issue:42673|issue:43704", - "issue:44749|issue:45310", + "issue:44361|issue:44749", + "issue:44623|issue:45656", "issue:28282|issue:30990", - "issue:45230|issue:45397", - "issue:44038|issue:44623", + "issue:44749|issue:45310", "issue:42994|issue:43482", - "issue:43701|issue:44749", + "issue:44038|issue:44623", + "issue:45230|issue:45397", + "issue:44749|issue:45656", "issue:42371|issue:45092", "issue:43475|issue:45381", "issue:45375|issue:45478", @@ -10289,233 +10277,232 @@ "issue:43475|issue:44661", "issue:43873|issue:44945", "issue:42831|issue:44222", - "issue:45310|issue:45397", "issue:45362|issue:45375", "issue:43232|issue:44188" ], - "prepared_review_unit_hash": "3753b2526078ac8678968a86cad8be22ed81ef90384e1965b57f3b8d9d4d58e7", + "prepared_review_unit_hash": "1c9c20279671b4c20ede0526c99aca945387ebf6c33651a39c31ec244368e7a8", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36696, - "estimated_input_tokens": 9174, - "estimated_eval_tokens": 18604 + "serialized_chars": 36727, + "estimated_input_tokens": 9182, + "estimated_eval_tokens": 18620 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous. Only issue:42673 and issue:43704 look like a true duplicate pair (the same Qwen3ForCausalLM VRAM leak in multiple dataloader threads). The other soft pairs are better treated as non-matches, including the Qwen3.5-Moe from_pretrained pair.", - "confidence": 0.79, - "canonical_issue_reason": "issue:45278 is the broadest umbrella issue in the set, covering the largest current upgrade-regression surface ('many import errors after update from 4.57.0 to 5.5.0'), even though the overall cluster is not tightly unified.", + "summary": "Mostly a mixed bag of unrelated transformer/model issues; only a few pairs look like true duplicates, especially the Qwen3.5 MoE from_pretrained regressions and the Qwen3ForCausalLM VRAM leak reports.", + "confidence": 0.64, + "canonical_issue_reason": "issue:45375 is the most concrete Qwen3.5 MoE report: it names the missing `deepstack_visual_indexes` field and explains the `@strict`-related config loss that likely drives the broader from_pretrained failures in the sibling reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best global issue candidate because it is the most general and current regression report; the rest of the cluster is too fragmented to serve as a single canonical bug.", + "best_issue_reason": "issue:45375 is the best representative issue in this set because it pinpoints the root cause rather than only describing the symptom; the related generic reports appear to be the same bug family.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44933", - "right": "issue:45459", + "left": "issue:43673", + "right": "issue:44734", "accept": false, - "reason": "Both are import/tokenizer-related, but they describe different failures and code paths: a nonexistent image_utils import versus hidden tokenizer decode errors when protobuf is missing." + "reason": "Both are generation/cache-related, but they describe different failures in different code paths (`chunked_prefill` cache missing vs KV cache continuation indexing)." }, { "left": "issue:44869", "right": "issue:45356", "accept": false, - "reason": "Different tokenizer regressions in different model families; Whisper timestamp decoding is unrelated to Kimi-K2.5 codec/fix_mistral_regex behavior." + "reason": "Both involve tokenizer/decoding behavior, but one is a Whisper word-timestamp crash and the other is a Kimi tokenizer codec regression; not the same bug." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "DeepSpeed ZeRO3 deque underflow and stale auto_mappings references to removed Sam3LiteText configs are unrelated bugs." + "reason": "DeepSpeed ZeRO3 deque handling and Sam3LiteText auto-mapping CI breakage are unrelated." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch on load is unrelated to the tie_word_embeddings warning during Qwen3.5 LoRA fine-tuning." - }, - { - "left": "issue:44297", - "right": "issue:44623", - "accept": false, - "reason": "Both touch save_pretrained, but one is tokenizer class metadata and the other is missing processor files; not the same bug." + "reason": "A weight-shape mismatch loading error and a tie_word_embeddings warning are different issues." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Different generation issues: post-sync_gpus kwargs update versus missing GenerationMixin cache during chunked_prefill." + "reason": "Both touch generation internals, but one is sync_gpus/update_kwargs ordering and the other is chunked_prefill cache missing; not the same concrete bug." }, { - "left": "issue:43673", - "right": "issue:45278", + "left": "issue:44297", + "right": "issue:44623", "accept": false, - "reason": "A generation cache regression and a broad import-error regression are different failure classes and code paths." + "reason": "`tokenizer.save_pretrained` metadata mismatch and `processor.save_pretrained` missing files are related broadly, but they are distinct serialization bugs." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "These are separate upgrade regressions affecting different subsystems; custom model initialization is not the same as the ESM2 breakage." + "reason": "`trust_remote_code`/missing config fields and ESM2 being broken are too different in scope and failure mode." }, { - "left": "issue:44361", - "right": "issue:44749", + "left": "issue:43673", + "right": "issue:45278", "accept": false, - "reason": "MLukeTokenizer AttributeError and a large performance slowdown after upgrade are different symptoms and different subsystems." + "reason": "One is a specific v5.0 chunked_prefill cache issue; the other is a broad import-error complaint after upgrading. Too vague to treat as the same bug." }, { "left": "issue:42673", "right": "issue:43704", "accept": true, - "reason": "These appear to be exact duplicates: same Qwen3ForCausalLM VRAM leak when used in multiple dataloader threads." + "reason": "Same title and same specific symptom: Qwen3ForCausalLM VRAM leaking when used in multiple dataloader threads." }, { - "left": "issue:44749", - "right": "issue:45310", + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Tokenizer AttributeError on tasks vs a slowdown after upgrading Transformers; no shared code-path." + }, + { + "left": "issue:44623", + "right": "issue:45656", "accept": false, - "reason": "Data-filtering slowdown and Qwen3.5-Moe from_pretrained loading failure are unrelated bugs." + "reason": "Processor save/export file generation and deepspeed optimizer stepping are unrelated." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "Both are loading-related, but missing PyTorch for AutoModel and Sentence Transformers getting stuck loading are different issues." + "reason": "PyTorch ImportError and Sentence Transformers stuck loading are different environment/runtime problems." }, { - "left": "issue:45230", - "right": "issue:45397", + "left": "issue:44749", + "right": "issue:45310", + "accept": false, + "reason": "Upgrade slowdown during filtering and Qwen3.5 MoE from_pretrained failure are unrelated." + }, + { + "left": "issue:42994", + "right": "issue:43482", "accept": false, - "reason": "issue:45230 is too generic to establish sameness, while issue:45397 is a specific gemma-4 zero3 from_pretrained problem." + "reason": "Saving quantized models and loading Qwen2.5-GGUF under Transformers v5 are opposite operations and different failure modes." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL MoE loading and processor.save_pretrained missing files are different code paths and failure modes." + "reason": "Qwen3-VL-Moe v5 bug and processor.save_pretrained file loss are distinct issues." }, { - "left": "issue:42994", - "right": "issue:43482", + "left": "issue:45230", + "right": "issue:45397", "accept": false, - "reason": "Quantized saving failure and Qwen2.5-GGUF loading failure are both quantization-adjacent but not the same bug." + "reason": "The first is a generic placeholder report; the second is a gemma-4 DeepSpeed/ZeRO3 loading bug. No evidence they are the same." }, { - "left": "issue:43701", - "right": "issue:44749", + "left": "issue:44749", + "right": "issue:45656", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to the data-filtering slowdown regression." + "reason": "A data-filtering slowdown after upgrade and a double optimizer step under deepspeed are unrelated." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 API deprecation guidance is unrelated to the InternVL2 remote-code/meta-initialization incompatibility." + "reason": "TF32 API-settings guidance and InternVL2 meta-initialization incompatibility are not the same bug." }, { "left": "issue:43475", "right": "issue:45381", "accept": false, - "reason": "Different vision/video bugs: missing fpn_position_embeddings versus wrong vision_position_ids." + "reason": "Different model families and different attribute/position-id failures; both are vision issues but not the same code-path." }, { "left": "issue:45375", "right": "issue:45478", - "accept": false, - "reason": "Both mention Qwen3.5 Moe, but one is a missing vision config field and the other is a from_pretrained loading error; not the same underlying issue." + "accept": true, + "reason": "Both are the same Qwen3.5 MoE from_pretrained regression family; 45375 gives the concrete missing-field root cause and 45478 reports the same failure under Transformers>=5.4.0." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "SAM 3 Video encoder attribute errors and trust_remote_code AutoConfig missing fields are unrelated." + "reason": "SAM 3 video encoder attribute error and AutoConfig `trust_remote_code` field loss are unrelated." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both are import errors, but missing AutoImageProcessor and a nonexistent image_utils import are different failures." + "reason": "Missing AutoImageProcessor import and a generic image_utils import issue are too vague to merge as duplicates." }, { "left": "issue:45081", "right": "issue:45103", "accept": false, - "reason": "Tokenizer regex patch crash and auto-docstring annotation handling are unrelated subsystems." + "reason": "Tokenizer regex patch crash and auto_docstring annotation handling crash are unrelated subsystems." }, { "left": "issue:45310", "right": "issue:45375", - "accept": false, - "reason": "Same broad model family, but different components and failure modes: text MoE from_pretrained error versus vision config field loss." + "accept": true, + "reason": "Same Qwen3.5 MoE from_pretrained regression; 45375 identifies the missing config field that explains the generic 45310 failure." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized saving failure and missing pad_token_id in Llama4Config are unrelated." + "reason": "Quantized save failure and Llama4Config missing `pad_token_id` are unrelated." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "Vitpose model breakage and Gemma 4 mm_token_type_ids behavior are unrelated." + "reason": "SAM3 vitpose breakage and Gemma 4 mm_token_type_ids handling are different model-specific bugs." }, { "left": "issue:43475", "right": "issue:44661", "accept": false, - "reason": "SAM 3 Video encoder bug is unrelated to add-new-model-like failing on TOKENIZER_MAPPING_NAMES membership." + "reason": "A SAM 3 video encoder attribute problem and an `add-new-model-like` tokenizer-mapping failure are unrelated." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization/offloading behavior and pipeline-parallel output correctness are different problems." + "reason": "Quantization offloading behavior and pipeline-parallel output correctness are different issues." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FP8 accuracy degradation and FP8 save_pretrained/moe issues are different failure modes." - }, - { - "left": "issue:45310", - "right": "issue:45397", - "accept": false, - "reason": "Both involve Qwen/Gemma model loading regressions, but they affect different model families and code paths." + "reason": "FineGrainedFP8 accuracy regression and FP8 save_pretrained failure are different bugs, even though both mention FP8." }, { "left": "issue:45362", "right": "issue:45375", "accept": false, - "reason": "Qwen3.5 chat crash and Qwen3_5MoeVisionConfig field loss are separate bugs." + "reason": "Qwen3.5-35B chat crash and Qwen3.5 MoE missing config field are separate failures." }, { "left": "issue:43232", "right": "issue:44188", "accept": false, - "reason": "Generation kwargs update after sync_gpus is unrelated to torch.compile attention-kernel branching divergence." + "reason": "Generation kwargs/sync_gpus behavior and attention-kernel divergence under torch.compile are different code paths." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall framing is mostly conservative, but the summary overstates the duplicate structure: the packet only supports one accepted duplicate pair (42673/43704). The claim that there are two true duplicates, including an \u2018identical Qwen3.5-Moe from_pretrained\u2019 pair, is not grounded in the provided soft pairs. Please revise the summary to reflect only the supported duplicate and keep the rest as tentative non-matches." + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on ambiguous pairs. The accepted duplicates are mergeable as single bugfixes (especially the identical Qwen3ForCausalLM VRAM leak reports and the Qwen3.5 MoE from_pretrained regressions). The summary does not overstate confidence beyond the evidence." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -10528,10 +10515,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30333", "issue:30990", @@ -10590,8 +10577,8 @@ "issue:44496|issue:45561", "issue:43526|issue:45561", "issue:43232|issue:44792", - "issue:43504|issue:45561", "issue:43377|issue:44242", + "issue:43504|issue:45561", "issue:36010|issue:41628", "issue:43673|issue:44188", "issue:30990|issue:35141", @@ -10601,13 +10588,13 @@ "issue:43976|issue:44933", "issue:42175|issue:44933", "issue:44964|issue:45137", - "issue:44623|issue:45310", "issue:43784|issue:43824", "issue:44188|issue:44373", + "issue:44623|issue:45310", "issue:44060|issue:45125", "issue:43278|issue:45137", - "issue:44734|issue:45103", "issue:43723|issue:44568", + "issue:44734|issue:45103", "issue:42898|issue:45356", "issue:42898|issue:44779", "issue:30333|issue:33290", @@ -10615,220 +10602,220 @@ "issue:42898|issue:44448", "issue:41628|issue:45459" ], - "prepared_review_unit_hash": "26ddd03dcc17abcdb55c0e83281ea8ab7733baa824e3ca8774d2795f71a65276", + "prepared_review_unit_hash": "b97603aa2b89fc85bbd78da1a3711cf9f5db846ddb2749d87ab977cf08672675", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36969, - "estimated_input_tokens": 9243, - "estimated_eval_tokens": 18742 + "serialized_chars": 36960, + "estimated_input_tokens": 9240, + "estimated_eval_tokens": 18736 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft-edge pair looks like a true duplicate. The set mixes unrelated issues across import/export regressions, tokenizer behavior changes, multimodal model loading, training/runtime bugs, and serialization problems. The only repeated theme is general v5 tokenizer regressions, but the specific bugs are model- or behavior-specific rather than one underlying defect.", - "confidence": 0.93, - "canonical_issue_reason": "No single canonical issue cleanly represents this set because the items are heterogeneous. If forced to pick the broadest hub, issue 42898 is the most general tokenizer-regression report, but it is not a true umbrella for the others.", + "summary": "Loose cluster of mostly unrelated Transformers v5 regressions, import failures, tokenizer/model bugs, and a few testing/runtime issues. The titles are semantically similar in places, but none of the soft pairs look like the same underlying bug or fix.", + "confidence": 0.91, + "canonical_issue_reason": "issue:42898 is the broadest and most central representative of the cluster\u2019s recurring v5 tokenizer-regression theme; it has the strongest overlap with other tokenizer-behavior reports, even though it is not a true duplicate of them.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 42898 is the best representative among the listed issues only in the limited sense that it is the broadest and most central tokenizer-regression report, with some inbound attention. It still does not unify the other bugs into one duplicate cluster.", + "best_issue_reason": "issue:42898 is the best single issue to anchor this cluster because it captures the common v5 regression/tokenization breakage pattern more generally than the other candidates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41628", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols (`AutoImageProcessor` vs `HybridCache`) and likely different fixes." + "reason": "Both are import errors, but they involve different symbols (`AutoImageProcessor` vs `HybridCache`) and different failure paths." }, { "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Both mention Qwen3.5 MoE, but one is FP8 `save_pretrained` and the other is `from_pretrained` loading; different code paths and symptoms." + "reason": "Both mention MoE/FP8 save-load problems, but one is about `save_pretrained` for FP8 and the other about a `from_pretrained` error for Qwen3.5 MoE." }, { "left": "issue:42994", "right": "issue:44164", "accept": false, - "reason": "Both involve saving/loading, but one is quantized model saving and the other is `extra_state` handling; not the same underlying bug." + "reason": "Both concern saving/loading, but one is a quantized model save failure and the other is an `extra_state` handling bug; not the same code path." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Same model family, but one is a broad Qwen3-VL-Moe issue and the other is a specific Qwen3.5 MoE load error." + "reason": "Different model families and symptoms; Qwen3-VL-Moe incompatibility is not the same bug as Qwen3.5 Moe `from_pretrained` failure." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Unrecognized model loading and a pytest-xdist file race are unrelated." + "reason": "Unrecognized model config is unrelated to the xdist `captured_info.txt` race in patched testing utilities." }, { "left": "issue:43526", "right": "issue:45561", "accept": false, - "reason": "BEiT label reduction and test-worker file capture race are unrelated." + "reason": "Image processor label reduction bug and pytest-xdist file race are unrelated." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation kwargs update after `sync_gpus` and a Janus image-generation test failure do not point to the same bug." + "reason": "Generation cache update issue and a janus image-generation test failure do not appear to be the same defect." }, { - "left": "issue:43504", - "right": "issue:45561", + "left": "issue:43377", + "right": "issue:44242", "accept": false, - "reason": "BEiT legacy-field loading and a testing race condition are unrelated." + "reason": "MIMI padding-mask inconsistency and MoE load-balancing loss gating are distinct model behaviors." }, { - "left": "issue:43377", - "right": "issue:44242", + "left": "issue:43504", + "right": "issue:45561", "accept": false, - "reason": "One is a MIMI padding-mask correctness bug; the other is MoE load-balancing loss logic." + "reason": "Legacy-field model-loading failure is unrelated to the xdist captured-info race." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "Both are import regressions, but for different APIs and likely different package exports." + "reason": "Both are import errors, but they concern different API symbols and likely different fixes." }, { "left": "issue:43673", "right": "issue:44188", "accept": false, - "reason": "`GenerationMixin` cache handling and attention-kernel divergence under `torch.compile` are distinct generation-path problems." + "reason": "Generation cache missing in v5 and attention-kernel divergence under `torch.compile` are different bugs." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Sentence-Transformers loading hang and embedding reinitialization after resizing are unrelated." + "reason": "Sentence-transformers loading hang and embedding reinitialization after resize are unrelated." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Multimodal config initialization and `ModelOutput` key assignment are different subsystems with different failure modes." + "reason": "Multimodal config defaulting bug and `ModelOutput` key assignment bug are different layers of the stack." }, { "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 insertion for a specific tokenizer." + "reason": "Both are tokenizer regressions in v5, but one is about `clean_up_tokenization_spaces` and the other about `add_special_tokens` not adding BOS/EOS; different failures." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Config defaults for Phi-4 multimodal and Whisper timestamp decoding are unrelated." + "reason": "Config initialization bug and Whisper timestamp decode crash are unrelated." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version compatibility and a missing image_utils import are unrelated." + "reason": "Python version support issue and image_utils import failure are not the same bug." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Package dependency omission and a missing import are different packaging issues." + "reason": "Missing TensorFlow backend and nonexistent image_utils import are different packaging/API problems." }, { "left": "issue:44964", "right": "issue:45137", "accept": false, - "reason": "Phi-4 multimodal loading and DeepSpeed ZeRO3 deque exhaustion are unrelated." - }, - { - "left": "issue:44623", - "right": "issue:45310", - "accept": false, - "reason": "Processor save file omission and Qwen3.5 MoE load error are different issues." + "reason": "Phi-4 multimodal loading failure and DeepSpeed ZeRO3 deque error are unrelated runtime issues." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Different import failures (`nn` missing vs `Qwen2_5_VLForConditionalGeneration` missing) with no evidence of one root cause." + "reason": "Both are import errors, but they affect different missing names and different modules/models." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "Attention-kernel divergence and a wrong docstring are not duplicates." + "reason": "Attention-kernel branching bug and wrong docstring are not the same underlying issue." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor save-pretrained file omission and Qwen3.5 MoE load failure are different save/load paths." }, { "left": "issue:44060", "right": "issue:45125", "accept": false, - "reason": "Tied-weights warning for Qwen3-Next and missing tensor-parallel plan for Qwen3.5-Moe are different model bugs." + "reason": "Tied-weights warning bug and missing `_tp_plan` for tensor parallelism are separate model-integration issues." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Embedding dtype drift and DeepSpeed deque errors are unrelated." + "reason": "Embedding dtype drift in eval and a ZeRO3 deque underflow are unrelated." }, { - "left": "issue:44734", - "right": "issue:45103", + "left": "issue:43723", + "right": "issue:44568", "accept": false, - "reason": "KV-cache continuation crash and auto-docstring crash under future annotations are different code paths." + "reason": "Tokenizer loading error and `add_special_tokens` BOS/EOS regression are different tokenizer problems." }, { - "left": "issue:43723", - "right": "issue:44568", + "left": "issue:44734", + "right": "issue:45103", "accept": false, - "reason": "Tokenizer loading in v5 and special-token insertion behavior are separate tokenizer issues." + "reason": "KV-cache continuation crash in serving and auto-docstring annotation handling are unrelated." }, { "left": "issue:42898", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different failure mechanisms." + "reason": "Both are tokenizer regressions, but they target different models and different broken behaviors (cleanup-space behavior vs codec handling/warning)." }, { "left": "issue:42898", "right": "issue:44779", "accept": false, - "reason": "Different tokenizer regressions for different models; not the same underlying defect." + "reason": "Deepseek tokenizer regression and `clean_up_tokenization_spaces` change are different tokenizer bugs." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLflow reporting after failed jobs and DeepSpeed Adafactor OOM are unrelated." + "reason": "MLflow reporting after training failure and DeepSpeed Adafactor OOM are unrelated." }, { "left": "issue:43874", "right": "issue:45092", "accept": false, - "reason": "Missing multimodal-image-patch API and remote-code checkpoint meta-init incompatibility are different problems." + "reason": "Missing image-patch helper and remote-code meta-init incompatibility are different multimodal model issues." }, { "left": "issue:42898", "right": "issue:44448", "accept": false, - "reason": "A tokenizer cleanup default change and Pegasus output divergence between v4 and v5 are broader but distinct regressions." + "reason": "Tokenizer cleanup behavior and Pegasus output differences across v4/v5 are related only at a high level, not the same bug." }, { "left": "issue:41628", "right": "issue:45459", "accept": false, - "reason": "Missing exported import and hidden tokenizer decode errors when protobuf is absent are unrelated." + "reason": "Importing `AutoImageProcessor` and tokenizer protobuf error masking are unrelated." } ] }, @@ -10847,10 +10834,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:22355", "issue:28282", @@ -10898,19 +10885,22 @@ "issue:45310", "issue:45357", "issue:45406", - "issue:45459" + "issue:45459", + "issue:45656" ], "soft_pairs": [ + "issue:44295|issue:45656", "issue:44222|issue:44297", "issue:42898|issue:43295", "issue:43479|issue:43644", + "issue:44297|issue:45656", "issue:44442|issue:45459", "issue:44222|issue:44623", "issue:42831|issue:44623", "issue:44206|issue:45103", - "issue:44297|issue:45310", "issue:41628|issue:42757", "issue:22355|issue:28282", + "issue:44297|issue:45310", "issue:44222|issue:44484", "issue:44483|issue:45406", "issue:43723|issue:44977", @@ -10920,226 +10910,238 @@ "issue:30333|issue:34689", "issue:30333|issue:35141", "issue:28282|issue:30333", - "issue:44295|issue:44623", "issue:44514|issue:45103", + "issue:44295|issue:44623", "issue:43479|issue:44913", - "issue:44261|issue:45092", "issue:43723|issue:44871", + "issue:44261|issue:45092", "issue:44373|issue:44485", "issue:43723|issue:44991", "issue:43723|issue:44442", - "issue:45092|issue:45357", "issue:30990|issue:38175", + "issue:45092|issue:45357", "issue:42886|issue:42994", - "issue:42831|issue:45310", "issue:41628|issue:43824", + "issue:42831|issue:45310", "issue:42831|issue:44038", "issue:42757|issue:45278", "issue:43576|issue:43784", "issue:44263|issue:44373" ], - "prepared_review_unit_hash": "ba3ec67ecfba363ef1198c79241a487abcb2ac1d9e101b050d4b683d6ca89ea9", + "prepared_review_unit_hash": "0e1319c4a3e411eee8e22e454f428779249a0162d0ecb665de5bc1ae0caa2f21", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 36, - "serialized_chars": 37646, - "estimated_input_tokens": 9412, - "estimated_eval_tokens": 19080 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38809, + "estimated_input_tokens": 9703, + "estimated_eval_tokens": 19662 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is too heterogeneous to treat as one duplicate set. Most soft pairs share only a broad subsystem label (tokenizer loading, save_pretrained, import errors, v5 regressions, multimodal configs) but describe different concrete failures, so all soft edges should be rejected.", - "confidence": 0.95, - "canonical_issue_reason": "No safe canonical issue: the cluster spans unrelated bugs and regressions, so there is no single underlying issue to merge around.", + "summary": "This cluster is mostly a mix of unrelated Transformers issues that happen to share broad themes like import failures, tokenizer/processor loading, and save_pretrained regressions. None of the soft-edge pairs look like true duplicates or mergeable PR-equivalents.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: it explicitly describes multiple import regressions after a version upgrade, matching the recurring import/loading failure theme better than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43723 is the broadest representative of the tokenizer/loading regression subset, but the set is still too mixed for it to serve as a true cluster canonical.", + "best_issue_reason": "issue:45278 is the best representative of the cluster because it captures the most general recurring breakage pattern here\u2014package import failures after upgrading\u2014while the other issues are more specific symptom reports.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different code paths: position_ids buffer handling vs Deepspeed optimizer stepping." + }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "Both involve save_pretrained, but one is FP8 MoE model saving and the other is tokenizer_class metadata mismatch; different code paths and failures." + "reason": "Both involve save_pretrained, but one is FP8/MoE saving and the other is tokenizer metadata mismatch." }, { "left": "issue:42898", "right": "issue:43295", "accept": false, - "reason": "Both are v5 regressions, but one is clean_up_tokenization_spaces behavior and the other is processor.tokenizer/image handling; not the same bug." + "reason": "Tokenizer cleanup behavior change is unrelated to processor.tokenizer access and image-to-tokenizer regression." }, { "left": "issue:43479", "right": "issue:43644", "accept": false, - "reason": "Unrelated areas: Phi4MultimodalConfig default config initialization versus non-persistent buffer corruption." + "reason": "Config None-default initialization is not the same bug as non-persistent buffer junk values." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer config export issue is unrelated to optimizer step duplication." }, { "left": "issue:44442", "right": "issue:45459", "accept": false, - "reason": "Both mention tokenizer loading/errors, but one is a missing tokenizer class and the other is protobuf error masking; different root causes." + "reason": "One is a tokenizer class loading failure; the other is error masking when protobuf is absent." }, { "left": "issue:44222", "right": "issue:44623", "accept": false, - "reason": "Both concern saving, but one is FP8 model save_pretrained and the other is processor.save_pretrained missing files; not the same artifact or fix." + "reason": "Different save_pretrained failures affecting different object types and outputs." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "Accuracy issues in FineGrainedFP8 are unrelated to processor serialization missing files." + "reason": "Accuracy regression in FineGrainedFP8 is unrelated to missing processor save files." }, { "left": "issue:44206", "right": "issue:45103", "accept": false, - "reason": "Feature-specific feature extractor crash versus auto_docstring crash from future annotations; no shared underlying bug." - }, - { - "left": "issue:44297", - "right": "issue:45310", - "accept": false, - "reason": "Tokenizer metadata mismatch and Qwen3.5 MoE from_pretrained error are different loading problems with different failing paths." + "reason": "Unsupported feature arg in a feature extractor is not the same as docstring parsing crashing on future annotations." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols from different packages; not the same defect." + "reason": "Both are import errors, but they concern different symbols, packages, and root causes." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Both are import errors, but one is missing transformers.onnx and the other is missing PyTorch; different dependency problems." + "reason": "Missing transformers.onnx and missing PyTorch are separate environment/import problems." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to Qwen3.5 from_pretrained failure." }, { "left": "issue:44222", "right": "issue:44484", "accept": false, - "reason": "FP8 model saving bug and a max_shard_size question are unrelated save_pretrained behaviors." + "reason": "MoE FP8 save_pretrained and max_shard_size default are different save_pretrained behaviors." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "A chat-completions API regression and a Gemma4Processor attribute error are different service/component failures." + "reason": "Chat completions request validation is unrelated to Gemma4Processor missing _tokenizer." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "Both involve model loading/generation in v5, but one is tokenizer loading and the other is flash-attention generation behavior; different failures." + "reason": "Tokenizer loading regression and flash-attention generation issue are different failures." }, { "left": "issue:44263", "right": "issue:44360", "accept": false, - "reason": "Both reference the DSA indexer, but one is a torch.split return-value issue and the other says the indexer lacks a ReLU; not the same bug." + "reason": "Different GLM-DSA bugs: torch.split outputs vs missing ReLU." }, { "left": "issue:30333", "right": "issue:36331", "accept": false, - "reason": "MLflow reporting after failed training and CustomTrainer.compute_loss signature breakage are unrelated training issues." + "reason": "MLflow reporting on failed training and CustomTrainer signature mismatch are unrelated trainer issues." }, { "left": "issue:44263", "right": "issue:45092", "accept": false, - "reason": "DSA indexer behavior and old InternVL2 remote-code/meta-init incompatibility are unrelated." + "reason": "GlmMoeDsaIndexer split behavior is unrelated to InternVL2 meta-init compatibility." }, { "left": "issue:30333", "right": "issue:34689", "accept": false, - "reason": "Failed-job MLflow reporting and Llama 3.2 Vision model loading regression are different problems." + "reason": "MLflow failure reporting and Llama 3.2 vision model loading are different problems." }, { "left": "issue:30333", "right": "issue:35141", "accept": false, - "reason": "Training-job reporting and embedding reinitialization during post_init are unrelated." + "reason": "Failed job reporting is unrelated to embedding resize/post_init weight reinitialization." }, { "left": "issue:28282", "right": "issue:30333", "accept": false, - "reason": "AutoModel missing PyTorch and MLflow not reporting failed jobs are different issues with no shared code path." + "reason": "AutoModel missing PyTorch and MLflow job reporting are unrelated issues." }, { - "left": "issue:44295", - "right": "issue:44623", + "left": "issue:44514", + "right": "issue:45103", "accept": false, - "reason": "Reading position_ids after buffer registration and processor save_pretrained missing files are unrelated serialization/access bugs." + "reason": "Batched chat template crash and auto_docstring attribute error do not share the same bug path." }, { - "left": "issue:44514", - "right": "issue:45103", + "left": "issue:44295", + "right": "issue:44623", "accept": false, - "reason": "Qwen2_5_VLProcessor batched chat-template crash and docstring processing under future annotations are unrelated." + "reason": "Buffer registration read error is unrelated to processor.save_pretrained missing files." }, { "left": "issue:43479", "right": "issue:44913", "accept": false, - "reason": "Phi4MultimodalConfig initialization and GPTNeoXConfig rotary_pct reload behavior are different config bugs." + "reason": "Default config initialization and GPTNeoX rotary_pct reload are different config bugs." }, { - "left": "issue:44261", - "right": "issue:45092", + "left": "issue:43723", + "right": "issue:44871", "accept": false, - "reason": "MLA q_a_layernorm epsilon precision issue is unrelated to InternVL2 meta-initialization incompatibility." + "reason": "Tokenizer loading failure and eos_token_id inconsistency are separate symptoms." }, { - "left": "issue:43723", - "right": "issue:44871", + "left": "issue:44261", + "right": "issue:45092", "accept": false, - "reason": "Tokenizer loading failure and Gemma-3 eos_token_id configuration mismatch are different model/tokenizer issues." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is unrelated to InternVL2 meta initialization." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "Wrong docstring for position_ids is documentation-only; GLM-5 RoPE implementation is a code behavior issue." + "reason": "Wrong docstring for position_ids is not the same as GLM-5 RoPE implementation behavior." }, { "left": "issue:43723", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer loading regressions, but they affect different checkpoints and failure modes." + "reason": "Both are tokenizer-load complaints, but they involve different models and failure modes." }, { "left": "issue:43723", "right": "issue:44442", "accept": false, - "reason": "AutoTokenizer failing for a custom tokenizer and FastSpeech2ConformerTokenizer loading failure are different concrete cases." + "reason": "FastSpeech2ConformerTokenizer load failure is a model-specific tokenizer bug, not the same as generic v5 tokenizer-loading regressions." }, { - "left": "issue:45092", - "right": "issue:45357", + "left": "issue:30990", + "right": "issue:38175", "accept": false, - "reason": "InternVL2 meta-init incompatibility and Qwen3.5 save_pretrained key corruption are unrelated multimodal issues." + "reason": "Sentence Transformers loading hang and SigLIP2 zero-probabilities are unrelated." }, { - "left": "issue:30990", - "right": "issue:38175", + "left": "issue:45092", + "right": "issue:45357", "accept": false, - "reason": "Sentence Transformers loading stall and SigLIP2 zero-probability outputs are different symptom classes." + "reason": "Meta-initialization incompatibility and incorrect visual encoder keys in save_pretrained are different bugs." }, { "left": "issue:42886", @@ -11148,40 +11150,40 @@ "reason": "Offline tokenizer cache loading and quantized model saving failure are unrelated." }, { - "left": "issue:42831", - "right": "issue:45310", + "left": "issue:41628", + "right": "issue:43824", "accept": false, - "reason": "FP8 accuracy concerns and Qwen3.5 MoE from_pretrained errors are different bug reports." + "reason": "Both are import errors, but they concern different exported symbols and model families." }, { - "left": "issue:41628", - "right": "issue:43824", + "left": "issue:42831", + "right": "issue:45310", "accept": false, - "reason": "Missing AutoImageProcessor and missing Qwen2_5_VLForConditionalGeneration imports are different symbol/export issues." + "reason": "FP8 accuracy regression is not the same as Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe bug are unrelated." + "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe bug reports do not describe the same defect." }, { "left": "issue:42757", "right": "issue:45278", "accept": false, - "reason": "Missing is_offline_mode from huggingface_hub and a broad set of import errors after upgrade are not the same defect." + "reason": "Missing is_offline_mode is a specific symbol regression, while 45278 is a broader post-upgrade import-error umbrella." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "Broken transformers env command and sentence-transformers NameError are unrelated." + "reason": "Broken transformers env CLI and sentence-transformers import NameError are unrelated." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "DSA indexer behavior and a wrong docstring for position_ids are not the same bug." + "reason": "DSA indexer torch.split behavior and a position_ids docstring issue are not the same bug." } ] }, @@ -11200,10 +11202,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:29127", "issue:30990", @@ -11261,14 +11263,14 @@ "issue:43122|issue:44514", "issue:42371|issue:44263", "issue:30990|issue:33357", - "issue:44295|issue:44297", "issue:42907|issue:44945", "issue:44704|issue:45290", "issue:43976|issue:44945", + "issue:44295|issue:44297", "issue:43232|issue:44743", "issue:39401|issue:44488", - "issue:43232|issue:45245", "issue:42907|issue:43937", + "issue:43232|issue:45245", "issue:43784|issue:45600", "issue:41628|issue:45278", "issue:43408|issue:43475", @@ -11276,244 +11278,244 @@ "issue:44360|issue:44485", "issue:43116|issue:45200", "issue:39401|issue:44521", - "issue:44038|issue:44297", "issue:44261|issue:44373", - "issue:43122|issue:44779", "issue:29127|issue:30990", - "issue:44038|issue:44295", + "issue:43122|issue:44779", + "issue:44038|issue:44297", "issue:42371|issue:42831", "issue:42907|issue:44964", - "issue:44038|issue:44222", + "issue:44038|issue:44295", "issue:36296|issue:43116", "issue:42907|issue:43452", + "issue:44038|issue:44222", "issue:43992|issue:44704" ], - "prepared_review_unit_hash": "e8f793b4998c525b8020670f52bff410d66b6f47d84abfd0d170e2ba05669c88", + "prepared_review_unit_hash": "f127afb17721693d1657e908cb09d392c6af449c59144e1fed8dedcb187218ce", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 33, - "serialized_chars": 36560, - "estimated_input_tokens": 9140, - "estimated_eval_tokens": 18536 + "serialized_chars": 36562, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is very noisy and mostly contains unrelated transformer issues; the soft pairs generally share broad subsystem vocabulary but not the same underlying bug. I\u2019d anchor the cluster on the most active open issue, but I would not merge any of the soft pairs.", - "confidence": 0.58, - "canonical_issue_reason": "Issue #44521 is the strongest anchor: it is open, has the highest inbound reference count in this set, and describes a concrete core API bug rather than a niche or purely diagnostic problem.", + "summary": "This cluster is largely heterogeneous: the soft pairs share broad themes (tokenization, generation, save/load, multimodal), but they do not describe the same concrete bug or change, so none should be merged as duplicates. No PRs are present.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44704 is the clearest representative if one must be chosen: it names a specific, actionable code-path bug in `AutoProcessor.from_pretrained` forwarding kwargs to `cached_file`.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #44521 is the best representative issue overall because it is recent, active, externally referenced, and about a user-visible multimodal API failure with clear reproduction value.", + "best_issue_reason": "Issue 44704 is the best standalone issue here because it is narrowly scoped, reproducible, and points to a specific fix path in a common API.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Both are tokenizer-related, but one is a version-to-version tokenization regression and the other is an all-zero assistant_mask bug in multimodal chat templating; different code paths." + "reason": "Both are tokenizer-related, but one is a v4\u2192v5 tokenization regression and the other is an assistant-mask bug in multimodal chat templating." }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Generation kwargs after sync_gpus is unrelated to D-FINE auxiliary-loss handling." + "reason": "Different subsystems: generation kwargs after `sync_gpus` vs D-FINE auxiliary losses with denoising off." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and AutoProcessor kwargs passthrough are different API-layer bugs." + "reason": "Related to loading paths, but one is tokenizer mapping fallback and the other is `AutoProcessor` kwarg forwarding; different bugs." }, { "left": "issue:42886", "right": "issue:43066", "accept": false, - "reason": "Offline tokenizer cache loading and wrong tokenizer decoder type in v5 are distinct issues." + "reason": "Offline cache loading failure and wrong tokenizer decoder type are distinct problems with different failure modes." }, { "left": "issue:43122", "right": "issue:44514", "accept": false, - "reason": "Both involve tokenization/chat-template behavior, but the first is a tokenizer regression while the second is a batched apply_chat_template crash with padding=False." + "reason": "Generic tokenization drift is not the same as a batched `apply_chat_template` crash with `padding=False`." }, { "left": "issue:42371", "right": "issue:44263", "accept": false, - "reason": "TF32 configuration warnings and a torch.split return-value bug are unrelated." + "reason": "TF32 API guidance and a GLM MoE `torch.split()` issue are unrelated." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence Transformers loading hang and a MacOS CLIP bus error are different failures." - }, - { - "left": "issue:44295", - "right": "issue:44297", - "accept": false, - "reason": "Reading position_ids after buffer registration is unrelated to tokenizer_class mismatch in tokenizer_config.json." + "reason": "Both concern loading, but one is a hang in Sentence Transformers and the other is a MacOS bus error on a CLIP model." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Saving dequantized models and incorrect output under pipeline parallelism are different problems." + "reason": "Saving dequantized models and incorrect pipeline-parallel LLM output are different code paths and fixes." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "AutoProcessor cached_file argument forwarding is not the same bug as apply_chat_template crashing on tool-call assistant messages." + "reason": "`cached_file` kwarg forwarding is unrelated to chat template crashes with tool-call assistant messages." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python version compatibility and pipeline-parallel output corruption are unrelated." + "reason": "Python version compatibility and pipeline-parallel inference correctness are not the same bug." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Reading a registered buffer and saving mismatched `tokenizer_class` are separate serialization/attribute issues." }, { "left": "issue:43232", "right": "issue:44743", "accept": false, - "reason": "Both concern generation behavior, but one is sync_gpus bookkeeping and the other is recurrent-state reset in modular_qwen3_5.py; not the same fix." + "reason": "Generation state handling after `sync_gpus` is distinct from recurrent state reset in `modular_qwen3_5.py`." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Qwen3 tokenizer offset_mapping and loading cjvt/sleng-bert are unrelated tokenizer/model issues." + "reason": "Wrong `offset_mapping` and inability to load a specific tokenizer are different tokenizer bugs." }, { - "left": "issue:43232", - "right": "issue:45245", + "left": "issue:42907", + "right": "issue:43937", "accept": false, - "reason": "Generation state handling and a category-count runtime limit are unrelated." + "reason": "Dequantized model save failure is unrelated to invalid `GenerationConfig` validation for GLM-5." }, { - "left": "issue:42907", - "right": "issue:43937", + "left": "issue:43232", + "right": "issue:45245", "accept": false, - "reason": "Dequantized save failure and GLM-5 GenerationConfig validation are different code paths." + "reason": "These involve unrelated failures: generation kwargs vs a category-count runtime limit." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Sentence-transformers import NameError and stale auto_mappings references are separate import/config breakages." + "reason": "Import-time `nn` NameError in sentence-transformers is unrelated to stale auto-mapping references in CI." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "A missing AutoImageProcessor import and broad post-upgrade import errors are not the same concrete bug." + "reason": "Missing `AutoImageProcessor` export is not the same as the broader set of import errors after upgrade." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "The model-type warning and the missing fpn_position_embeddings attribute are distinct SAM 3 problems." + "reason": "Model-type warning on load and missing `fpn_position_embeddings` are different SAM3 issues." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "num_labels propagation in Qwen3.5 and flash_attention_3 degenerate training are unrelated." + "reason": "`num_labels` propagation and degenerate training under flash_attention_3 are unrelated." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "DSA indexer ReLU discussion and GLM-5 RoPE implementation are different model internals." + "reason": "A missing ReLU in the DSA indexer is not the same as the GLM-5 RoPE implementation discussion." }, { "left": "issue:43116", "right": "issue:45200", "accept": false, - "reason": "Example-script multi-label output bug and Gemma 4 mm_token_type_ids defaulting are not the same issue." + "reason": "Example-script multi-label output bug and Gemma 4 multimodal token-type defaults are distinct problems." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Offset_mapping wrongness and assistant_mask all-zero behavior are both chat/tokenizer adjacent, but they are separate defects." - }, - { - "left": "issue:44038", - "right": "issue:44297", - "accept": false, - "reason": "A Qwen3-VL-Moe bug and tokenizer_class mismatch on save_pretrained are unrelated." + "reason": "Wrong tokenizer offsets and zeroed assistant masks for multimodal chat are different behaviors in separate paths." }, { "left": "issue:44261", "right": "issue:44373", "accept": false, - "reason": "A runtime precision bug and a wrong docstring are not the same underlying change." + "reason": "A precision/config issue in MLA q_a_layernorm is unrelated to a wrong docstring for `position_ids`." }, { - "left": "issue:43122", - "right": "issue:44779", + "left": "issue:29127", + "right": "issue:30990", "accept": false, - "reason": "Both mention tokenizer regressions, but they affect different models and symptoms; not enough evidence for one merged fix." + "reason": "An unclear layoutlmv3 error message and a Sentence Transformers loading hang are not the same bug." }, { - "left": "issue:29127", - "right": "issue:30990", + "left": "issue:43122", + "right": "issue:44779", "accept": false, - "reason": "LayoutLMv3 error-message clarity and a sentence-transformers loading hang are unrelated." + "reason": "Both mention tokenizer behavior, but one is a version-wide tokenization change and the other is a DeepSeek-specific regression." }, { "left": "issue:44038", - "right": "issue:44295", + "right": "issue:44297", "accept": false, - "reason": "Qwen3-VL-Moe and position_ids buffer reading are unrelated bugs." + "reason": "Qwen3-VL-Moe/v5 model behavior and tokenizer save metadata mismatch are different issues." }, { "left": "issue:42371", "right": "issue:42831", "accept": false, - "reason": "TF32 settings guidance and FineGrainedFP8 accuracy are both precision-adjacent but not the same issue." + "reason": "TF32 settings guidance is unrelated to FineGrainedFP8 accuracy problems." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Dequantized save failure and inability to load Phi-4 multimodal instruct are different model-handling problems." + "reason": "Saving dequantized models and loading Phi-4 multimodal with latest transformers are different failures." }, { "left": "issue:44038", - "right": "issue:44222", + "right": "issue:44295", "accept": false, - "reason": "Both mention MOE/FP8, but one is a Qwen3-VL-Moe bug and the other is FP8 save_pretrained behavior; different concrete fixes." + "reason": "Qwen3-VL-Moe compatibility and `position_ids` buffer access are unrelated." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training and a multi-label example script bug are unrelated." + "reason": "Tensor-parallel training bugs and multi-label example-script behavior do not share the same underlying cause." }, { "left": "issue:42907", "right": "issue:43452", "accept": false, - "reason": "Saving dequantized models and gguf_file breaking from_pretrained are different serialization/load paths." + "reason": "Dequantized model save failure is unrelated to `gguf_file` breaking tokenizer/model loading." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "A Qwen3-VL-Moe version bug and FP8 `save_pretrained` are different model/tooling issues." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "UMT5Encoder missing embed_tokens.weight and AutoProcessor cached_file kwargs forwarding are unrelated." + "reason": "Missing `embed_tokens.weight` in UMT5 loading is unrelated to `AutoProcessor` kwargs being dropped." } ] }, @@ -11532,10 +11534,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:22355", "issue:29127", @@ -11589,8 +11591,8 @@ "issue:43824|issue:44351", "issue:45216|issue:45478", "issue:43122|issue:44977", - "issue:42907|issue:44222", "issue:42757|issue:44351", + "issue:42907|issue:44222", "issue:43278|issue:44945", "issue:43784|issue:45446", "issue:42222|issue:43441", @@ -11600,8 +11602,8 @@ "issue:43937|issue:44485", "issue:22355|issue:29127", "issue:42371|issue:43874", - "issue:44075|issue:44936", "issue:43116|issue:43493", + "issue:44075|issue:44936", "issue:44222|issue:45478", "issue:44038|issue:45478", "issue:43976|issue:45446", @@ -11618,14 +11620,14 @@ "issue:42673|issue:43906", "issue:45405|issue:45446" ], - "prepared_review_unit_hash": "229a26532f5694a6f7da1fe516833d26ce206244868c92ef4c783038f76bcbc9", + "prepared_review_unit_hash": "1d40c377029041e73f9e8ad079b8911d5603d2de14603bbfc9c71bd3dc7706d0", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 47, @@ -11640,208 +11642,205 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous and should not be broadly collapsed. The only strong duplicate-like edge is the Qwen3.5 MoE save/load regression pair; the rest are distinct bugs across unrelated subsystems.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45216 is the clearest root report for the shared Qwen3.5 MoE serialization regression: it names the 5.4.0 save_pretrained corruption that plausibly explains the paired from_pretrained failure in 45478.", + "summary": "These items are mostly unrelated regression and bug reports across import errors, tokenizer behavior, model saving/loading, and training/runtime issues. None of the soft pairs is a safe duplicate match from titles alone; the only plausible umbrella candidate is the broad import-regression report.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45278 is the most umbrella-like report: it explicitly covers many import errors after upgrading to 5.5.0, so it is the best representative among these issues for a broad regression cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45216 is the best representative issue because it is the most concrete, version-pinned report in the only accepted duplicate-like pair.", + "best_issue_reason": "issue:45278 is the best global issue candidate because it is broad, regression-focused, and could naturally collect related import-breakage reports; the others are narrower, model-specific, or isolated to different code paths.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different code paths (Qwen2.5-VL vs HybridCache)." + "reason": "Both are import errors, but for different missing symbols and likely different code paths; not the same bug." }, { "left": "issue:45216", "right": "issue:45478", - "accept": true, - "reason": "Same Qwen3.5 MoE save/load regression around transformers 5.4.0; a broken save_pretrained checkpoint can plausibly surface as a from_pretrained failure." + "accept": false, + "reason": "Both mention Qwen3.5 and v5.4+, but one is save_pretrained corruption and the other is from_pretrained failure; different operations." }, { "left": "issue:43122", "right": "issue:44977", "accept": false, - "reason": "Tokenizer regression vs flash-attention generation failure; different models and different failure modes." + "reason": "Tokenizer regression vs flash-attention generation failure; different subsystems and symptoms." }, { - "left": "issue:42907", - "right": "issue:44222", + "left": "issue:42757", + "right": "issue:44351", "accept": false, - "reason": "Both mention saving/quantization, but they are different model families and different bugs." + "reason": "Different missing imports from different dependencies/modules; not a single underlying defect." }, { - "left": "issue:42757", - "right": "issue:44351", + "left": "issue:42907", + "right": "issue:44222", "accept": false, - "reason": "Different import targets from different packages; same symptom class, not the same bug." + "reason": "Both concern saving quantized/dequantized models, but the model families and failure modes differ." }, { "left": "issue:43278", "right": "issue:44945", "accept": false, - "reason": "One is an embedding dtype regression, the other is a pipeline-parallel output correctness issue." + "reason": "Embedding dtype mismatch during eval is unrelated to pipeline-parallel incorrect output." }, { "left": "issue:43784", "right": "issue:45446", "accept": false, - "reason": "Unrelated import/name error in sentence-transformers vs a PyTorch version-guard bug in flex_attention." + "reason": "Sentence-transformers import NameError is unrelated to the PyTorch version check in flex_attention." }, { "left": "issue:42222", "right": "issue:43441", "accept": false, - "reason": "Different affected models and different breakages; no shared underlying code-path is evident." + "reason": "VitPose import breakage and Ministral FlashAttention failure are distinct model-specific issues." }, { "left": "issue:42673", "right": "issue:45072", "accept": false, - "reason": "VRAM leak under threaded dataloaders is unrelated to bfloat16 dtype mismatches in CI." + "reason": "VRAM leak in Qwen3 threads is unrelated to dtype mismatches in bfloat16 inference." }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "Both are import-regression reports, but one is a specific missing symbol and the other is a broad upgrade complaint." + "reason": "The latter is a broad import-regression report, but these are not the same concrete import failure." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Training-state corruption with FSDP/PEFT is unrelated to cache handling in chunked prefill." + "reason": "Wrong optimizer states under FSDP/PEFT is unrelated to missing GenerationMixin cache in chunked prefill." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "GenerationConfig validation on GLM-5 is not the same bug as a RoPE implementation discussion." + "reason": "GenerationConfig validation error and RoPE implementation discussion are different bugs." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "A missing module import and a layoutlmv3 error-message clarity issue are unrelated." + "reason": "Transformers.onnx import failure and LayoutLMv3 error-message clarity are unrelated." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 settings guidance and a missing multimodal method are unrelated issues." + "reason": "TF32 API guidance and a missing image-patch method are not the same issue." }, { - "left": "issue:44075", - "right": "issue:44936", + "left": "issue:43116", + "right": "issue:43493", "accept": false, - "reason": "Optimizer arguments being ignored is separate from evaluate() failing after train()." + "reason": "Example-script multi-label output bug and SigLIP2 implementation discrepancy are different model/workflow problems." }, { - "left": "issue:43116", - "right": "issue:43493", + "left": "issue:44075", + "right": "issue:44936", "accept": false, - "reason": "Multi-label example-script failure and SigLIP2 implementation discrepancy are different bugs." + "reason": "Unused SGD arguments and trainer.evaluate() after train() are separate training issues." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "FP8 save_pretrained for MoE is not the same as the Qwen3.5 from_pretrained regression." + "reason": "One is FP8 save_pretrained, the other is from_pretrained on Qwen3.5 Moe; same family but different code paths." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Different Qwen model families and different failure surfaces; not the same underlying bug." + "reason": "Qwen3-VL-Moe and Qwen3.5 Moe are different model variants with different failures." }, { "left": "issue:43976", "right": "issue:45446", "accept": false, - "reason": "Python version compatibility issue vs an import version-check bug in flex_attention." + "reason": "Python support regression and a PyTorch import/version-check bug are not the same defect." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "Example-script multi-label output bug is unrelated to the pipeline translation-task messaging issue." + "reason": "Example script empty results and pipeline() error messaging are distinct bugs." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "Both involve generation APIs, but one is invalid config handling and the other is a stale warning." + "reason": "GenerationConfig invalidity and a stale num_return_sequences warning do not share the same code-path failure." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and different observable failures." + "reason": "Both are tokenizer regressions, but they affect different models and report different incorrect behavior." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "General tokenizer-loading regression vs a Qwen3 offset-mapping bug; too broad to merge." + "reason": "Tokenizer loading regression and offset_mapping regression are not the same concrete bug." }, { "left": "issue:43931", "right": "issue:44945", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch is unrelated to pipeline-parallel generation correctness." + "reason": "Weight-shape mismatch on model loading is unrelated to pipeline-parallel wrong outputs." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "A qwen3_omni_moe helper bug and a flex_attention import guard bug are unrelated." + "reason": "Feature-extraction length helper behavior and a flex_attention import/version check are unrelated." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "Tokenizer mapping returning None and flash_attention crashing on s_aux=None are different code paths." + "reason": "Tokenizer mapping None handling and flash_attention s_aux handling are separate runtime bugs." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs lag after pipeline removals is unrelated to a runtime version-check import bug." + "reason": "Docs still referencing pipeline() and a PyTorch version-check bug are unrelated." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "Both mention model generation, but GLM-5 config validation and Qwen3.5 flash-attention output issues are distinct." + "reason": "GLM-5 generation config validation and Qwen3.5 flash-attention generation failure are different issues." }, { "left": "issue:45341", "right": "issue:45561", "accept": false, - "reason": "A testing_utils bug and an xdist race in captured_info.txt are separate test-infrastructure problems." + "reason": "A small testing_utils bug and a xdist race on captured_info.txt are not the same defect." }, { "left": "issue:42673", "right": "issue:43906", "accept": false, - "reason": "An active VRAM leak report is unrelated to an isolated reproduction of another issue." + "reason": "A VRAM leak report and an isolated reproduction of another issue are not the same underlying bug." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "A released-version dependency bump problem is unrelated to the flex_attention version-check import bug." + "reason": "PEFT version pinning to an unreleased release and a PyTorch import/version-check bug are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The only plausible duplicate-like pair is the Qwen3.5 MoE save/load regression pair (45216/45478), and the rest of the soft-edge rejections are appropriately cautious given the titles and limited evidence." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -11853,10 +11852,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:38617", "issue:39401", @@ -11879,9 +11878,9 @@ "issue:43673", "issue:43742", "issue:43874", - "issue:43931", "issue:43937", "issue:43986", + "issue:44075", "issue:44117", "issue:44220", "issue:44261", @@ -11895,7 +11894,6 @@ "issue:44496", "issue:44829", "issue:44933", - "issue:45003", "issue:45071", "issue:45081", "issue:45216", @@ -11905,7 +11903,8 @@ "issue:45459", "issue:45464", "issue:45478", - "issue:45542" + "issue:45542", + "issue:45656" ], "soft_pairs": [ "issue:44291|issue:45071", @@ -11914,8 +11913,8 @@ "issue:44484|issue:45310", "issue:44829|issue:45216", "issue:38617|issue:41628", - "issue:44410|issue:45310", "issue:44117|issue:45081", + "issue:44410|issue:45310", "issue:42673|issue:43582", "issue:42222|issue:45542", "issue:42371|issue:44261", @@ -11926,9 +11925,10 @@ "issue:42175|issue:42222", "issue:44933|issue:45278", "issue:43012|issue:43577", + "issue:44075|issue:45656", "issue:44483|issue:45464", - "issue:44484|issue:45478", "issue:42617|issue:45341", + "issue:44484|issue:45478", "issue:43066|issue:43122", "issue:43673|issue:45341", "issue:44410|issue:45478", @@ -11940,249 +11940,240 @@ "issue:43012|issue:43408", "issue:44263|issue:44485", "issue:41628|issue:45071", - "issue:45071|issue:45542", - "issue:43931|issue:44410", - "issue:45003|issue:45341", - "issue:43986|issue:45341" + "issue:45071|issue:45542" ], - "prepared_review_unit_hash": "c3675029f1aba2fcfc785a6f32b0da05067e6f369c827cec3c3fa2829d09417f", + "prepared_review_unit_hash": "4e5fa95f5a5c333e1b3519b7ad53293ce9627066366d4e5592b0237bf34fdd46", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 36, - "serialized_chars": 38279, - "estimated_input_tokens": 9570, - "estimated_eval_tokens": 19396 + "soft_pair_count": 34, + "serialized_chars": 37647, + "estimated_input_tokens": 9412, + "estimated_eval_tokens": 19080 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is very heterogeneous: most items are separate regressions, model-specific bugs, or unrelated API/import issues. I found no soft-edge pair that looks like the same underlying bug/change, so all candidate links are rejected. If a single issue must anchor the cluster, the broad open umbrella is the import-error report after the 5.5.0 upgrade.", - "confidence": 0.88, - "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set, describing multiple import errors after upgrading to 5.5.0; it best fits as a cluster anchor even though the surrounding items are not true duplicates.", + "summary": "This cluster is mostly a noisy mix of unrelated Transformers regressions (imports, tokenizers, model loading, backend/config issues). I would not merge any of the soft-edge pairs as duplicates; the closest umbrella-style issue is the broad import-regression report.", + "confidence": 0.19, + "canonical_issue_reason": "issue:45278 is the broadest, most representative issue in the set: it describes multiple import errors after upgrading to 5.5.0, which can plausibly subsume some of the more specific import complaints here.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most representative issue here because it is broad and version-level, whereas the others are narrower model- or subsystem-specific reports.", + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it is the widest-scope, user-facing regression report among otherwise diverse and mostly unrelated bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "Both are upgrade/regression reports, but they describe different failures: unexpected `_is_hf_initialized` handling during weight loading vs `PreTrainedConfig` type-checking breakage." + "reason": "Different failure modes: unexpected model-init arg handling vs PreTrainedConfig type-checking regression." }, { "left": "issue:43450", "right": "issue:43986", "accept": false, - "reason": "Both involve video processing, but one is an incorrect batched output shape and the other is a crash when `torchvision` is missing. Different bugs, not one change." + "reason": "One is a batched video-shape bug, the other is an AutoProcessor crash due to missing torchvision." }, { "left": "issue:39401", "right": "issue:43421", "accept": false, - "reason": "Tokenizer offset mapping and runtime post-processor updates are different tokenization behaviors with different symptoms and likely different fixes." + "reason": "Both are tokenizer-related, but one is offset mapping and the other is runtime post-processor updates for special tokens." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "`max_shard_size` behavior and a Qwen3.5 MoE `from_pretrained` error are unrelated issues." + "reason": "Unrelated: save_pretrained shard-size question vs Qwen3.5 MoE from_pretrained error." }, { "left": "issue:44829", "right": "issue:45216", "accept": false, - "reason": "FlashAttention training degeneration and a bad saved checkpoint for Qwen3.5 are distinct model-path problems." + "reason": "Different code paths: flash_attention_3 training degeneration vs Qwen3.5 checkpoint save regression." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols (`layer_type_validation` vs `AutoImageProcessor`) and do not look like the same underlying regression." + "reason": "Both are import errors, but they involve different missing symbols/modules and different root causes." }, { - "left": "issue:44410", - "right": "issue:45310", + "left": "issue:44117", + "right": "issue:45081", "accept": false, - "reason": "These affect different model families and different concrete failures: missing projections in qwen3next versus a Qwen3.5 MoE loading error." + "reason": "Tokenizer mapping None handling is not the same bug as the Mistral regex/backend_tokenizer crash." }, { - "left": "issue:44117", - "right": "issue:45081", + "left": "issue:44410", + "right": "issue:45310", "accept": false, - "reason": "A tokenizer mapping `None` assumption bug is not the same as Mistral regex patching crashing on a tokenizer backend attribute." + "reason": "Different model-specific problems: missing Qwen3Next projections vs Qwen3.5 MoE load failure." }, { "left": "issue:42673", "right": "issue:43582", "accept": false, - "reason": "A VRAM leak in multi-threaded dataloader use and an Apple Silicon `TypeError` in `caching_allocator_warmup` are unrelated runtime problems." + "reason": "VRAM leak in Qwen3ForCausalLM threads is unrelated to an Apple Silicon TypeError in allocator warmup." }, { "left": "issue:42222", "right": "issue:45542", "accept": false, - "reason": "VitPose model breakage and a TensorBoard/TensorFlow backend install issue are entirely different areas." + "reason": "Different subsystems: broken vitpose models vs TensorFlow/backend installation mismatch." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 API guidance and a missing `rms_norm_eps` in MLA layernorm are unrelated." + "reason": "TF32 settings API change is unrelated to MLA layernorm epsilon/config precision issues." }, { "left": "issue:42673", "right": "issue:43577", "accept": false, - "reason": "VRAM leakage and incorrect dtype propagation in BLIP2 loading are different defects." + "reason": "A memory leak and a dtype-loading issue are different bugs." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "A model-type warning for sam3 tracker/video and a key error loading MobileLLM are unrelated." + "reason": "Model-type warning for sam3_tracker does not match the MobileLLM key error." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "An invalid `GenerationConfig` for GLM-5 and an unrecognized model/config issue for Olmo-Hybrid are different loading/configuration problems." + "reason": "GenerationConfig validation and unrecognized model/config.json missing model_type are different loading errors." }, { "left": "issue:38617", "right": "issue:44220", "accept": false, - "reason": "Different import/symbol areas: configuration utility import failure versus audio feature extraction issue." + "reason": "ImportError for layer_type_validation is unrelated to the fbank feature extraction issue." }, { "left": "issue:42175", "right": "issue:42222", "accept": false, - "reason": "A packaging/backend dependency issue for TensorFlow is not the same as broken VitPose models." + "reason": "Backend packaging/TF install problem is not the same as broken vitpose model behavior." }, { "left": "issue:44933", "right": "issue:45278", "accept": false, - "reason": "`image_utils` missing import is a specific symbol problem; `45278` is a broad report about many import errors. Similar theme, but not the same bug." + "reason": "45278 is broader, but 44933 is a specific image_utils import issue; not enough evidence they are the same underlying bug." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "A PyTorch precision warning during compilation and BLIP2 dtype staying float32 are different dtype-related symptoms with different code paths." + "reason": "A PyTorch precision warning during compilation is not the same as a model staying float32 after dtype loading." }, { - "left": "issue:44483", - "right": "issue:45464", + "left": "issue:44075", + "right": "issue:45656", "accept": false, - "reason": "Both mention chat/completions, but one is a general request rejection and the other is a streaming failure on a specific model." + "reason": "SGD args ignored and optimizer.step called twice are distinct optimizer bugs." }, { - "left": "issue:44484", - "right": "issue:45478", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "One concerns shard size defaults, the other a Qwen3.5 MoE loading bug; no shared underlying change." + "reason": "Both mention chat/completions, but one is a broad v5.3 request acceptance problem and the other is a Qwen3.5 streaming failure." }, { "left": "issue:42617", "right": "issue:45341", "accept": false, - "reason": "`3d_parallel.py` execution failure and a small `testing_utils.py` bug are unrelated." + "reason": "Not able to run 3d_parallel.py is unrelated to a testing_utils.py bug." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "max_shard_size default discussion is unrelated to the Qwen3.5 MoE from_pretrained error." }, { "left": "issue:43066", "right": "issue:43122", "accept": false, - "reason": "Wrong tokenizer decoder type and different tokenization output between versions are related only loosely by tokenization, not the same concrete bug." + "reason": "Both are tokenizer regression reports, but one is decoder type and the other is changed tokenization behavior; not the same concrete bug." }, { "left": "issue:43673", "right": "issue:45341", "accept": false, - "reason": "A GenerationMixin cache regression and a testing utility bug do not share a code path." + "reason": "Generation cache/chunked_prefill is unrelated to a testing_utils bug." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Different Qwen-related model-loading failures with different symptoms and likely different fixes." + "reason": "Different Qwen-related failures: missing layer projections vs from_pretrained error." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removal of `tie_embeddings_and_encoder_decoder` and unexpected key warnings for position id are unrelated v5 regressions." + "reason": "Removal of a config option and unexpected position-id keys are different migration/regression issues." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Missing image-patch counting support and missing `rms_norm_eps` are different multimodal/model-architecture issues." + "reason": "Missing image-patch count helper is unrelated to the MLA rms_norm_eps/config precision bug." }, { "left": "issue:42222", "right": "issue:45459", "accept": false, - "reason": "VitPose model breakage and protobuf-related tokenizer error masking are unrelated." + "reason": "Broken vitpose models and protobuf masking tokenizer errors are unrelated." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "Both mention FlashAttention, but they target different models and different failure modes; not mergeable as one concrete fix." + "reason": "Both mention flash attention, but the symptoms and affected paths are different enough to be separate bugs." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "A missing `torchvision` dependency crash and a shard-size question are unrelated." + "reason": "Missing torchvision crash is unrelated to max_shard_size behavior." }, { "left": "issue:43012", "right": "issue:43408", "accept": false, - "reason": "PyTorch precision warning and sam3 model-type warning are unrelated." + "reason": "Different areas: bfloat16 compile warning vs sam3 tracker/model-type warning." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "A `torch.split` return-value issue in one GLM component is not the same as a RoPE implementation discussion." + "reason": "GLM Moe torch.split indexing and GLM-5 RoPE implementation are different concrete bugs." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "Different regressions: missing import symbol versus config type-checking breakage." + "reason": "AutoImageProcessor import failure and PreTrainedConfig type checking break are separate regressions." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "PreTrainedConfig type checking and TensorBoard/TensorFlow backend installation are unrelated." - }, - { - "left": "issue:43931", - "right": "issue:44410", - "accept": false, - "reason": "Weight-shape mismatch in a Qwen3-VL checkpoint and missing projections in qwen3next are different model-loading problems." - }, - { - "left": "issue:45003", - "right": "issue:45341", - "accept": false, - "reason": "Unsafe `sys.modules[]` access in modeling utilities and a testing utility bug are separate issues." + "reason": "PreTrainedConfig typing and TensorFlow backend installation issues do not match." } ] }, @@ -12201,15 +12192,14 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:38617", "issue:42757", "issue:42886", - "issue:42907", "issue:43012", "issue:43064", "issue:43066", @@ -12247,6 +12237,7 @@ "issue:44704", "issue:44829", "issue:44987", + "issue:45003", "issue:45071", "issue:45103", "issue:45245", @@ -12256,7 +12247,10 @@ "issue:45588" ], "soft_pairs": [ + "issue:43931|issue:44410", + "issue:45003|issue:45341", "issue:44117|issue:44987", + "issue:43986|issue:45341", "issue:43097|issue:43452", "issue:43931|issue:44038", "issue:43335|issue:45071", @@ -12265,8 +12259,8 @@ "issue:43976|issue:45071", "issue:44829|issue:45245", "issue:43986|issue:45372", - "issue:43064|issue:44315", "issue:43531|issue:45341", + "issue:43064|issue:44315", "issue:38617|issue:42757", "issue:44261|issue:44485", "issue:44410|issue:45375", @@ -12286,223 +12280,240 @@ "issue:38617|issue:43097", "issue:43012|issue:44368", "issue:43012|issue:43550", - "issue:42886|issue:43381", - "issue:42907|issue:44655" + "issue:42886|issue:43381" ], - "prepared_review_unit_hash": "432ba1e2c6439c68f48fac2ebc2194a640354c16e7252428cae68ed566bc55c0", + "prepared_review_unit_hash": "0d9430a51f35ebaa3cc5c79f5ec2314fdaa96d7f009da4ad10966dcdffb3f601", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37287, - "estimated_input_tokens": 9322, - "estimated_eval_tokens": 18900 + "soft_pair_count": 34, + "serialized_chars": 37877, + "estimated_input_tokens": 9470, + "estimated_eval_tokens": 19196 }, "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, model-specific loading/config bugs, and unrelated training/runtime issues. None of the proposed soft pairs look like the same underlying bug or change, so I would not merge any of them.", - "confidence": 0.48, - "canonical_issue_reason": "Issue 43097 is the broadest representative here: a clear Transformers v5 API/removal regression, which best matches the cluster\u2019s general \u201cv5 breakage\u201d theme better than the narrower model-specific reports.", + "summary": "This cluster is a loose mix of unrelated Transformers bug reports; the soft pairs are mostly superficial title matches and do not look like the same underlying fix or bug.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated import errors, loading failures, config regressions, warnings, and training-path bugs rather than one duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": "43097 is the best single issue to anchor the cluster because it is a concrete, user-facing v5 regression and more representative than the many highly specific model or backend bugs.", + "best_issue_reason": "If a representative issue is needed, issue 43824 is the strongest user-facing regression with multiple inbound references and a clear reproducible import failure, but it does not subsume the rest of the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen-family, but one is a weight-shape loading mismatch and the other is missing layer projections in a different model path; not the same bug." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access and a testing_utils bug are unrelated code paths." + }, { "left": "issue:44117", "right": "issue:44987", "accept": false, - "reason": "Both involve model loading, but one is a tokenizer-mapping None bug and the other is a specific failure loading a model repo; different root causes." + "reason": "Tokenizer mapping returning None is a different failure from loading physical-intelligence/fast; not the same defect." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Video-model loading without torchvision and a testing utility bug are unrelated." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "One is a v5 API removal/regression, the other is gguf handling in AutoTokenizer/AutoModel loading; not the same bug." + "reason": "Removed tie_embeddings flag and gguf_file loading breakage are different APIs and failure modes." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Different model families and failures: Qwen3-VL weight-shape mismatch vs Qwen3-VL-Moe compatibility issue." + "reason": "Similar model family, but these describe different Qwen3-VL-related regressions with different symptoms and code paths." }, { "left": "issue:43335", "right": "issue:45071", "accept": false, - "reason": "Unrelated config bugs: SwitchTransformers sparse-layer handling vs PretrainedConfig type checking." + "reason": "SwitchTransformers sparse-layer config bug and PretrainedConfig type-checking regression are unrelated." }, { "left": "issue:43408", "right": "issue:44368", "accept": false, - "reason": "Both are warning-related, but they concern different models and different config fields; not a shared code defect." + "reason": "These are different warnings for different models/configs; not the same underlying issue." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different GLM components and failures: missing image-processor method vs a torch.split return-value issue." + "reason": "One is a missing image-processor method, the other is a torch.split indexing bug; different subsystems." }, { "left": "issue:43976", "right": "issue:45071", "accept": false, - "reason": "A Python version compatibility report is not the same as a PretrainedConfig type-checking regression." + "reason": "Python-version compatibility breakage is not the same as PretrainedConfig type-checking." }, { "left": "issue:44829", "right": "issue:45245", "accept": false, - "reason": "Completely different areas: FlashAttention training degeneration vs category-count runtime limits." + "reason": "Degenerate training with flash_attention_3 and category-count runtime limits are unrelated." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Both crash during loading, but one is a missing torchvision dependency path and the other is a mistral_common import breakage." + "reason": "Missing torchvision during video processor loading and a mistral_common import error are distinct dependency failures." }, { - "left": "issue:43064", - "right": "issue:44315", + "left": "issue:43531", + "right": "issue:45341", "accept": false, - "reason": "Training/optimization problems, but one is FSDP2+PEFT optimizer state corruption and the other is Liger Kernel application with model_init." + "reason": "sliding_window behavior in Qwen3-MoE and a testing_utils bug are unrelated." }, { - "left": "issue:43531", - "right": "issue:45341", + "left": "issue:43064", + "right": "issue:44315", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior and a testing_utils bug are unrelated." + "reason": "FSDP2+PEFT optimizer-state corruption and Liger Kernel not applying with model_init affect different training mechanisms." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they are different missing symbols from different modules and likely different API removals." + "reason": "Import errors for different symbols from different modules are not the same bug." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "Different model-specific numerical/config issues; the titles do not point to the same code path." + "reason": "MLA layernorm epsilon precision and GLM-5 RoPE implementation are different model-internal issues." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Both are model-structure/config complaints, but they affect different model families and different missing fields/projections." + "reason": "Missing qkv/gate projections and a missing config field are different model/config regressions." }, { "left": "issue:38617", "right": "issue:44351", "accept": false, - "reason": "Separate import failures for different symbols; no evidence they share the same underlying breakage." + "reason": "Both are import errors, but for different symbols and modules; not a duplicate." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Tokenizer construction bug vs tokenizer-mapping None handling; related area, but distinct defects." + "reason": "AddedToken special-arg collision and TOKENIZER_MAPPING_NAMES returning None are separate tokenizer bugs." }, { "left": "issue:44038", "right": "issue:44410", "accept": false, - "reason": "Different Qwen variants and different missing-layer/projection problems; too broad to merge." + "reason": "Different Qwen-family architecture issues with different missing components and symptoms." }, { "left": "issue:43097", "right": "issue:45588", "accept": false, - "reason": "A v5 API removal and a flash-attention crash on sink-less models are unrelated issues." + "reason": "tie_embeddings removal and s_aux=None crash in flash_attention are unrelated." }, { "left": "issue:43986", "right": "issue:44117", "accept": false, - "reason": "Video-model processor loading without torchvision is unrelated to tokenizer mapping lookup returning None." + "reason": "A video-model processor dependency crash and a tokenizer-mapping bug do not share the same code path." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "Both mention kwargs handling, but one is AutoProcessor/cached_file forwarding and the other is auto_docstring annotation parsing." + "reason": "Passing kwargs to cached_file and docstring processing under future annotations are unrelated." }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "Different failure modes: offline/network access leakage vs missing encoder weights during load." + "reason": "Both are loading-related, but one is offline/local_files_only network leakage and the other is a missing embedding weight during from_pretrained; different fixes." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "A deprecated API removal and quantization offloading behavior are not the same underlying bug." + "reason": "Removed tie_embeddings and offloading with quantization are unrelated issues." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "Weight-loading failure and a question about max_shard_size default are unrelated." + "reason": "Missing model weight on load and a save_pretrained shard-size question are different operations." }, { "left": "issue:43824", "right": "issue:43986", "accept": false, - "reason": "Both are multimodal loading issues, but one is a missing class import and the other is a torchvision dependency crash." + "reason": "Different model families and failure causes: import error for Qwen2.5-VL vs missing torchvision for video models." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "The latter is only an isolated reproduction of another issue; the titles do not indicate the same concrete tokenizer-decoder bug." + "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue are not the same concrete bug." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Both concern attention implementations, but they are different models and different backend problems." + "reason": "FlashAttention support rejection and GPT2 attention-scaling being ignored are distinct attention-path problems." }, { "left": "issue:43825", "right": "issue:44655", "accept": false, - "reason": "Pipeline error-message wording and Pipeline save_pretrained support are different code paths." + "reason": "Pipeline translation-message regression and saving Pipeline objects are different features." }, { "left": "issue:38617", "right": "issue:43097", "accept": false, - "reason": "Importing a missing symbol and removing a v5 helper are different kinds of breakage." + "reason": "ImportError for layer_type_validation and tie_embeddings removal are unrelated." }, { "left": "issue:43012", "right": "issue:44368", "accept": false, - "reason": "Both are warnings, but they arise from different model/config settings and do not share a specific bug." + "reason": "A bfloat16 precision warning during compilation and a tie_word_embeddings warning are different warnings with different triggers." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "A bfloat16 precision warning and a torch.compile/SDPA failure are unrelated." + "reason": "Precision warning and a torch.compile SDPA failure are unrelated." }, { "left": "issue:42886", "right": "issue:43381", "accept": false, - "reason": "Offline cache loading and gradient checkpointing in eval mode are unrelated runtime issues." + "reason": "Offline cache loading and gradient checkpointing in eval mode are separate bugs." } ] }, @@ -12521,15 +12532,16 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:36010", "issue:36683", "issue:38617", "issue:42757", + "issue:42907", "issue:42994", "issue:43012", "issue:43064", @@ -12556,7 +12568,6 @@ "issue:44077", "issue:44188", "issue:44230", - "issue:44242", "issue:44263", "issue:44380", "issue:44483", @@ -12576,6 +12587,7 @@ "issue:45593" ], "soft_pairs": [ + "issue:42907|issue:44655", "issue:43208|issue:43550", "issue:44380|issue:45479", "issue:43352|issue:43976", @@ -12585,10 +12597,10 @@ "issue:43618|issue:44380", "issue:36683|issue:43582", "issue:43618|issue:43761", - "issue:44483|issue:45290", "issue:43012|issue:44683", - "issue:44038|issue:44230", + "issue:44483|issue:45290", "issue:42757|issue:43824", + "issue:44038|issue:44230", "issue:44483|issue:45381", "issue:43299|issue:43867", "issue:38617|issue:43502", @@ -12603,270 +12615,266 @@ "issue:44077|issue:44380", "issue:44263|issue:45440", "issue:43064|issue:44492", - "issue:43976|issue:44938", "issue:36683|issue:44077", + "issue:43976|issue:44938", "issue:36683|issue:43525", "issue:36010|issue:42757", "issue:44188|issue:44380", "issue:43618|issue:45216", "issue:43352|issue:45593", - "issue:43992|issue:45310", - "issue:36683|issue:44242" + "issue:43992|issue:45310" ], - "prepared_review_unit_hash": "071412328c1d477766c491187d28806f9963e2a4015824a7ed9d67992ada4164", + "prepared_review_unit_hash": "dc2e88a88cadc3ae328d838b1b02afb69e1e7839e46541eac314ef5d96bdf232", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 36, - "serialized_chars": 38130, - "estimated_input_tokens": 9533, - "estimated_eval_tokens": 19322 + "serialized_chars": 38114, + "estimated_input_tokens": 9529, + "estimated_eval_tokens": 19314 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous. I only see a few true duplicates: two Qwen3-VL-Moe loading reports, two Qwen3.5-Moe from_pretrained reports, and a likely shared CLIP v5 output regression. The rest share broad subsystems or buzzwords but are different bugs/changes.", - "confidence": 0.81, - "canonical_issue_reason": "issue:45310 is the cleanest anchor: it states the Qwen3.5 MoE from_pretrained regression clearly and has an exact duplicate (issue:45478).", + "summary": "This cluster is mostly a set of unrelated Transformers bugs across different models, import paths, and save/load flows. The only clear duplicate-looking case in the provided candidates is the repeated Qwen3.5 Moe from_pretrained regression (45310 vs 45478).", + "confidence": 0.93, + "canonical_issue_reason": "issue:45310 is the better canonical for the duplicated Qwen3.5 Moe from_pretrained regression: it is earlier, more established, and matches the same symptom reported again in 45478.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the best representative issue because it is specific, reproducible, and directly duplicated by issue:45478.", + "best_issue_reason": "issue:45310 is the strongest representative issue because it is a concrete, model-specific regression report with clear reproduction context and it anchors the only near-exact duplicate in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Both concern saving, but one is a dequantized Minstral/Devstral model bug and the other is generic Pipeline save_pretrained serialization." + }, { "left": "issue:43208", "right": "issue:43550", "accept": false, - "reason": "Different model families and different failure modes; both are training/compile bugs, not the same bug." + "reason": "xLSTM training bugs and Bamba torch.compile+SDPA failure are different models and different failure modes." }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Unrelated bugs: attention scaling backend behavior vs a zero-loss classification issue." + "reason": "GPT2 backend scaling regression is unrelated to the zero-loss classification bug with num_labels=1." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "Shared version context only; one is FlashAttention support for a model, the other is a Python compatibility problem." + "reason": "Nemotron FlashAttention support and Python version compatibility are separate issues." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "Different code paths: distributed optimizer state handling vs attention backend scaling." + "reason": "FSDP2/PEFT optimizer-state corruption is a different code path from GPT2 attention scaling under SDPA/FlashAttention." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "Unrelated xLSTM training bugs vs an image processor label reduction bug." + "reason": "xLSTM training regressions and BeitImageProcessorFast label reduction are unrelated." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "Both involve chat-related APIs, but one is request validation and the other is batched chat template crashing." + "reason": "One is the chat/completions endpoint rejecting requests; the other is a chat-template crash on batched input." }, { "left": "issue:43618", "right": "issue:44380", "accept": false, - "reason": "Different model/output bug vs GPT2 attention scaling backend behavior." + "reason": "CLIPOutput attentions missing is a distinct regression from GPT2 attention scaling behavior." }, { "left": "issue:36683", "right": "issue:43582", "accept": false, - "reason": "Completely different subsystems: config attribute missing vs Apple Silicon allocator warmup TypeError." + "reason": "Gemma3Config missing vocab_size and Apple Silicon TypeError in caching_allocator_warmup are unrelated." }, { "left": "issue:43618", "right": "issue:43761", - "accept": true, - "reason": "Both describe the same CLIP v5 output regression where requested outputs are no longer populated." - }, - { - "left": "issue:44483", - "right": "issue:45290", "accept": false, - "reason": "Different chat-stack failures: API request rejection vs assistant/tool-call template crash." + "reason": "Both are CLIP-related, but one drops attentions while the other returns hidden_states=None; different symptoms and fixes." }, { "left": "issue:43012", "right": "issue:44683", "accept": false, - "reason": "Different compiler/runtime issues: a warning during bfloat16 compilation vs a flex_attention failure on newer torch." + "reason": "A PyTorch warning during bfloat16 compile is not the same bug as compiled flex_attention failing on torch>=2.9." }, { - "left": "issue:44038", - "right": "issue:44230", + "left": "issue:44483", + "right": "issue:45290", "accept": false, - "reason": "Same model family, but different requests: loading regression vs fp8 inference support." + "reason": "The v5 chat/completions request rejection and the assistant tool-call chat_template crash are different failures." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but they concern different symbols and packages." + "reason": "These are separate import errors from different modules: huggingface_hub versus transformers." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Qwen3-VL-Moe loading regression is not the same as fp8 inference support for Qwen3-VL/Qwen3.5 MoE." }, { "left": "issue:44483", "right": "issue:45381", "accept": false, - "reason": "Different bugs: chat completions request handling vs Qwen2.5-VL video position ids." + "reason": "Chat/completions API rejection is unrelated to the qwen2.5-vl video vision_position_ids bug." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Both are load errors, but one is Qwen3-VL-Moe regression and the other is a sorted-state_dict loading failure." + "reason": "Qwen3VL MoE loading failure and state_dict-sorted load error are different model-loading bugs." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "Different import/local-files-only behaviors; not the same underlying bug." + "reason": "layer_type_validation import failure is unrelated to local_files_only still making network calls." }, { "left": "issue:43756", "right": "issue:44485", "accept": false, - "reason": "Different RoPE/model bugs in different models; only broad architectural similarity." + "reason": "Smollm3 RoPE layer dropping and GLM-5 RoPE implementation are different model-specific problems." }, { "left": "issue:43901", "right": "issue:44655", "accept": false, - "reason": "Docs mismatch vs actual save_pretrained failure are not the same underlying issue." + "reason": "A docs mismatch about return_all_scores is not the same as Pipeline save_pretrained failure." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Both concern saving, but quantized model saving and pipeline saving are different problems." + "reason": "Quantized model saving failure and generic Pipeline saving failure are separate save-path bugs." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Both involve checkpoint loading/saving, but they target different failure modes and models." + "reason": "State_dict sorting load error and Qwen3.5 save_pretrained regression are different checkpoint issues." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Different missing-attribute bugs in unrelated model configs/outputs." + "reason": "Gemma3Config vocab_size and Sam3VisionEncoderOutput missing fpn_position_embeddings are unrelated missing-attribute bugs." }, { "left": "issue:44655", "right": "issue:45478", "accept": false, - "reason": "Pipeline save_pretrained and Qwen3.5 MoE loading are unrelated." + "reason": "Pipeline save_pretrained failure is unrelated to the Qwen3.5 Moe from_pretrained regression." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Different model support vs generation-config validation bugs." + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig invalid are different model/setup bugs." }, { "left": "issue:43066", "right": "issue:45381", "accept": false, - "reason": "Tokenizer decoder typing and video vision_position_ids are unrelated issues." + "reason": "Tokenizer decoder type mismatch is unrelated to qwen2.5-vl video position-id errors." }, { "left": "issue:44077", "right": "issue:44380", "accept": false, - "reason": "Different bugs in different model internals; not the same change." + "reason": "PatchTSMixer post_init validation and GPT2 attention scaling are different issues." }, { "left": "issue:44263", "right": "issue:45440", "accept": false, - "reason": "Different MoE implementations and different failure modes." + "reason": "torch.split return-value handling in GlmMoeDsaIndexer is not the same as DeepseekV3MoE divergence from remote implementation." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "Optimizer-state corruption and a typo in cache strategy are unrelated." + "reason": "Optimizer-state corruption under FSDP2/PEFT is unrelated to a cache-strategy typo." }, { - "left": "issue:43976", - "right": "issue:44938", + "left": "issue:36683", + "right": "issue:44077", "accept": false, - "reason": "Both are version-compatibility reports, but for different Python versions and different breakages." + "reason": "Gemma3Config missing vocab_size and PatchTSMixer post_init gating do not share the same bug." }, { - "left": "issue:36683", - "right": "issue:44077", + "left": "issue:43976", + "right": "issue:44938", "accept": false, - "reason": "Different model-specific attribute bug vs patchtsmixer post_init policy change." + "reason": "Python-version compatibility problems here are different: one is Transformers 5.1.0 vs Python 3.9/3.10 constraints, the other is Python 3.14 load failure." }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Different missing config attributes in different models." + "reason": "Gemma3Config lacking vocab_size and Llama4Config lacking pad_token_id are separate missing-field regressions." }, { "left": "issue:36010", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they are for different modules and symbols." + "reason": "These are different import errors from different symbols and packages." }, { "left": "issue:44188", "right": "issue:44380", "accept": false, - "reason": "Different attention issues: branching divergence under compile vs backend-specific scaling behavior." + "reason": "Diverging attention kernels under torch.compile is a different backend-branching bug from GPT2 attention scaling." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "CLIP output regression and Qwen3.5 checkpoint save regression are unrelated." + "reason": "CLIP attentions missing and Qwen3.5 saved checkpoint corruption are unrelated regressions." }, { "left": "issue:43352", "right": "issue:45593", "accept": false, - "reason": "Different model bugs and different affected functionality." + "reason": "Nemotron FlashAttention support and D-FINE auxiliary-loss handling are unrelated model bugs." }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "Different load-path bugs: missing embed tokens vs Qwen3.5 MoE from_pretrained regression." - }, - { - "left": "issue:36683", - "right": "issue:44242", - "accept": false, - "reason": "Unrelated model/config attribute issue vs MoE auxiliary-loss bookkeeping." + "reason": "UMT5Encoder missing embed_tokens.weight and Qwen3.5 Moe from_pretrained error affect different models and load paths." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The cluster summary is grounded in the packet and stays mostly conservative. The exact Qwen3.5 MoE duplicate pair is well supported, the Qwen3-VL-Moe loading pair is plausible from the issue titles, and the rejected soft edges are generally justified by distinct titles/subsystems. The CLIP pair is the only mildly inferential call, but it remains reasonably conservative." - }, - "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 }, @@ -12878,10 +12886,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:28282", "issue:33453", @@ -12907,13 +12915,13 @@ "issue:44077", "issue:44112", "issue:44222", + "issue:44242", "issue:44246", "issue:44351", "issue:44355", "issue:44360", "issue:44387", "issue:44484", - "issue:44485", "issue:44556", "issue:44749", "issue:44857", @@ -12933,8 +12941,9 @@ "issue:45600" ], "soft_pairs": [ - "issue:43010|issue:45278", + "issue:36683|issue:44242", "issue:36683|issue:43352", + "issue:43010|issue:45278", "issue:41669|issue:44246", "issue:43867|issue:45375", "issue:36683|issue:43064", @@ -12954,252 +12963,233 @@ "issue:42890|issue:44964", "issue:42890|issue:45600", "issue:42890|issue:44112", - "issue:43010|issue:44075", "issue:43502|issue:44062", "issue:43723|issue:44556", + "issue:43010|issue:44075", "issue:44222|issue:45125", "issue:42890|issue:45341", "issue:44387|issue:44556", - "issue:43010|issue:43065", "issue:42548|issue:44351", - "issue:43010|issue:43232", + "issue:43010|issue:43065", "issue:45125|issue:45478", + "issue:43010|issue:43232", "issue:44355|issue:44749", - "issue:45125|issue:45375", - "issue:43502|issue:44485" + "issue:45125|issue:45375" ], - "prepared_review_unit_hash": "f633bd5b996c7dd0e5e707363c810291dfd0a509926ff311f5dc40f01e3ab4ae", + "prepared_review_unit_hash": "ed52206f93f072cdd4ac73d55f3d7e148339e3e682b2bc57be4cb15ee7018d6f", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37341, - "estimated_input_tokens": 9336, - "estimated_eval_tokens": 18928 + "serialized_chars": 37357, + "estimated_input_tokens": 9340, + "estimated_eval_tokens": 18936 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of unrelated import/load/regression reports rather than one duplicate family. Most soft pairs share only broad subsystem or version-regression similarity, not the same concrete bug.", - "confidence": 0.68, - "canonical_issue_reason": "issue:45278 is the broadest umbrella-style report in the set ('Many import errors after update from 4.57.0 to 5.5.0'), so it is the best canonical anchor for this mostly import-regression cluster.", + "summary": "This cluster is mostly a collection of distinct regressions and feature gaps, not true duplicates. There is a loose theme around import/load regressions and some Qwen/Sam model-family issues, but the soft pairs do not describe the same concrete bug or change.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like report in the set: a general post-upgrade import regression with multiple failing imports, which best fits the cluster\u2019s main import-related overlap.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most general and centrally phrased issue, making it the best single issue to represent or route this cluster despite the weak overall cohesion.", + "best_issue_reason": "issue:45278 is the best representative issue because it is the most general symptom report and could plausibly subsume nearby import/export complaints better than the narrower, model-specific tickets.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43010", - "right": "issue:45278", + "left": "issue:36683", + "right": "issue:44242", "accept": false, - "reason": "Both mention v5 regressions, but one is about no_grad on cache/layer updates and the other is broad import failures; different bugs and code paths." + "reason": "Gemma3Config missing vocab_size and load-balancing loss are unrelated code paths." }, { "left": "issue:36683", "right": "issue:43352", "accept": false, - "reason": "Gemma3 config vocab_size and Nemotron Flash Attention support are unrelated model-specific problems." + "reason": "Config attribute error is unrelated to Flash Attention support gating." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Cache update no_grad semantics and import errors after upgrade are different bugs." }, { "left": "issue:41669", "right": "issue:44246", "accept": false, - "reason": "Both concern import performance, but one is about removing import * from models and the other is an intermittent slow import report; not clearly the same fix." + "reason": "Import-star slowdown and intermittent import slowness may be related in theme, but not the same concrete issue." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "One is a state_dict loading failure, the other is a missing config field dropped by strict parsing; different underlying issue." + "reason": "State-dict sorting load failure is unrelated to a missing Qwen3.5 vision config field." }, { "left": "issue:36683", "right": "issue:43064", "accept": false, - "reason": "A missing config attribute in Gemma3 is not the same as an FSDP2/PEFT optimizer-state mismatch." + "reason": "Gemma3Config vocab_size error and FSDP2/PEFT optimizer-state corruption are unrelated." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Both involve from_pretrained, but UMT5Encoder weight loading and Qwen3.5 MoE loading errors are model-specific and distinct." + "reason": "Both are from_pretrained/load problems, but they affect different models and different failure causes." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "A missing set_seed in a model integration test is unrelated to xdist races on captured_info.txt." + "reason": "Missing set_seed in a test and xdist file races are separate testing issues." }, { "left": "issue:44938", "right": "issue:45071", "accept": false, - "reason": "Python 3.14 loading failure and PretrainedConfig type-checking regression are different compatibility issues." + "reason": "Python 3.14 import failure and PreTrainedConfig type-checking breakage are distinct regressions." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "Both are save/load regressions, but one is sorted state_dict loading and the other is incorrect visual encoder keys on save_pretrained." + "reason": "State-dict ordering load error and incorrect visual encoder keys during save are different sides of model serialization." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "Torch no_grad decoration on update methods is unrelated to the Python version support issue." + "reason": "Generation cache updates and Python-version compatibility are unrelated." }, { "left": "issue:44938", "right": "issue:45381", "accept": false, - "reason": "Python 3.14 import/load failure and Qwen2.5-VL vision_position_ids bug are unrelated." + "reason": "Python 3.14 load failure and Qwen2.5-VL video position-id bug are unrelated." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "A missing ReLU in the DSA indexer and a DeepSeekV3MoE implementation divergence are not the same bug." + "reason": "DSA indexer missing ReLU and DeepSeekV3 implementation divergence are not the same bug." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIPOutput attentions assignment and patchtsmixer post_init validity are separate model/API issues." + "reason": "CLIPOutput attentions regression and patchtsmixer post_init policy are unrelated." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "Both mention tensor parallelism, but one is a generic training bug and the other is an OOM when loading a specific Qwen model; not the same change." + "reason": "Tensor parallel training bug and model load OOM under TP are different failure modes." }, { "left": "issue:43867", "right": "issue:44484", "accept": false, - "reason": "Loading failure from sorted state_dict is unrelated to the 50GB default shard size question." + "reason": "Load failure on sorted state_dict and a question about save_pretrained shard size are unrelated." }, { "left": "issue:43618", "right": "issue:45588", "accept": false, - "reason": "CLIPOutput attentions and flash_attention.py crashing on s_aux=None are different code paths." + "reason": "Missing CLIPOutput attentions and flash_attention s_aux=None crash are separate issues." }, { "left": "issue:43618", "right": "issue:44857", "accept": false, - "reason": "An output-field assignment regression is unrelated to LwDetrImageLoss AMP/CUDA crashes." + "reason": "CLIPOutput attentions regression is unrelated to LwDetrImageLoss AMP crash." }, { "left": "issue:28282", "right": "issue:33453", "accept": false, - "reason": "PyTorch-missing AutoModel import error and tokenizer loading regression are distinct failures." + "reason": "PyTorch-missing ImportError and tokenizer-loading regression are different import/load problems." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "A flaky test seed issue is unrelated to loading Phi-4 multimodal models." + "reason": "A flaky test and a Phi-4 multimodal load failure are unrelated." }, { "left": "issue:42890", "right": "issue:45600", "accept": false, - "reason": "Captured-info test race and stale auto_mappings config references are different CI problems." + "reason": "Test seed flakiness and removed Sam3LiteText config references are different CI issues." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "Both are test failures, but they involve different tests and causes." - }, - { - "left": "issue:43010", - "right": "issue:44075", - "accept": false, - "reason": "No_grad decoration for cache/layer updates is unrelated to SGD arguments being ignored." + "reason": "Missing set_seed in SamHQ tests and GraniteSpeech stale device override are unrelated test failures." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "Local-files-only network leakage and AddedToken keyword duplication are unrelated tokenizer/runtime bugs." + "reason": "Local-files-only network leakage and AddedToken keyword collision are unrelated." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Tokenizer loading regression in v5 and checkpoint reload failure after upgrade are both load-related but not the same concrete bug." - }, - { - "left": "issue:44222", - "right": "issue:45125", - "accept": false, - "reason": "FP8 save_pretrained for MoE and missing _tp_plan for Qwen3_5Moe tensor parallelism are different features and failures." - }, - { - "left": "issue:42890", - "right": "issue:45341", - "accept": false, - "reason": "A seed-related integration test issue is not the same as the testing_utils bug." + "reason": "Tokenizer loading regression in v5 and checkpoint reload failure after upgrade are separate issues." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Int4 quantization OOM and checkpoint reload incompatibility are different regressions." - }, - { - "left": "issue:43010", - "right": "issue:43065", - "accept": false, - "reason": "Torch no_grad on update methods and a dummy Conv2d in Sam3PixelDecoder are unrelated." + "reason": "Int4 CUDA memory growth and checkpoint reload compatibility are unrelated." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import-name failures, but for different symbols and likely different root causes." + "reason": "Missing PreTrainedModel export and missing HybridCache export are different symbol-export regressions." }, { - "left": "issue:43010", - "right": "issue:43232", + "left": "issue:44222", + "right": "issue:45125", "accept": false, - "reason": "Cache/layer update decoration and generation kwargs synchronization are different generation bugs." + "reason": "FP8 save_pretrained for MoE and missing _tp_plan for Qwen3_5Moe are different issues." }, { "left": "issue:45125", "right": "issue:45478", "accept": false, - "reason": "Both involve Qwen3.5 MoE, but one is a missing tensor-parallel plan and the other is a broader from_pretrained error; not clearly the same defect." + "reason": "Missing tensor-parallel plan and from_pretrained error for Qwen3.5 MoE are not the same code-path bug." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Cache/Layer no_grad decoration and generation kwargs sync_gpus handling are unrelated." }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled Python file errors and preprocessing slowdown after upgrade are unrelated." + "reason": "Compiled Python file errors and post-upgrade filtering slowdown are different regressions." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Tensor-parallel plan missing and missing deepstack_visual_indexes are different Qwen3.5 MoE config issues." - }, - { - "left": "issue:43502", - "right": "issue:44485", - "accept": false, - "reason": "Local-files-only network leak is unrelated to GLM-5 RoPE implementation." + "reason": "Missing _tp_plan and missing deepstack_visual_indexes are both Qwen3.5-related, but they are different config/serialization bugs." } ] }, @@ -13218,10 +13208,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30333", "issue:30990", @@ -13232,6 +13222,7 @@ "issue:38617", "issue:41628", "issue:41669", + "issue:42371", "issue:42548", "issue:42757", "issue:42831", @@ -13246,7 +13237,6 @@ "issue:43874", "issue:43901", "issue:44038", - "issue:44075", "issue:44222", "issue:44230", "issue:44261", @@ -13262,27 +13252,27 @@ "issue:44556", "issue:44623", "issue:44655", - "issue:44871", "issue:44908", "issue:44961", "issue:45003", "issue:45071", + "issue:45092", "issue:45125", "issue:45230", "issue:45468", "issue:45542" ], "soft_pairs": [ + "issue:43502|issue:44485", "issue:30990|issue:31795", - "issue:44075|issue:44871", "issue:44315|issue:44360", "issue:30333|issue:33453", "issue:43010|issue:44368", "issue:43502|issue:44222", - "issue:43299|issue:44556", "issue:43502|issue:44961", - "issue:44222|issue:44230", + "issue:43299|issue:44556", "issue:44355|issue:44655", + "issue:44222|issue:44230", "issue:30333|issue:36296", "issue:44355|issue:45071", "issue:42831|issue:43874", @@ -13301,273 +13291,281 @@ "issue:44908|issue:44961", "issue:44485|issue:44961", "issue:36296|issue:43901", - "issue:41628|issue:44355", "issue:36296|issue:43526", + "issue:41628|issue:44355", "issue:44273|issue:45003", "issue:44261|issue:44360", - "issue:44623|issue:45230", "issue:44273|issue:45542", + "issue:44623|issue:45230", "issue:43576|issue:44556", "issue:44038|issue:45125", + "issue:44393|issue:45071", "issue:41669|issue:43673", - "issue:44393|issue:45071" + "issue:44556|issue:45092", + "issue:42371|issue:44492" ], - "prepared_review_unit_hash": "5ed5a3429d4394102d7d2041bbc737baf1b3868115b2cafc2363cd3d7af80e04", + "prepared_review_unit_hash": "1789773476e0aae6ebd5601fef9749378a4947df56ed7aeba4ef6980d1708413", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 38, - "serialized_chars": 38351, - "estimated_input_tokens": 9588, - "estimated_eval_tokens": 19432 + "soft_pair_count": 40, + "serialized_chars": 38963, + "estimated_input_tokens": 9741, + "estimated_eval_tokens": 19738 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most items are unrelated bug reports, model-specific regressions, or docs/import issues. I did not find any soft pair that is clearly the same underlying bug/change, so all soft edges are rejected.", - "confidence": 0.91, - "canonical_issue_reason": "Issue 43576 is the best cluster anchor because it is the broadest v5 compatibility/regression report in the set; the others are more specific model, import, or docs problems.", + "summary": "Mostly a loose cluster of Transformers v5-era regressions, import/load failures, and a few unrelated bug reports; there is no single obvious duplicate chain, so this looks more like theme-based similarity than one underlying issue.", + "confidence": 0.36, + "canonical_issue_reason": "issue:43576 is the broadest representative of the set: it concerns a v5 breakage and sits closest to several other loading/import regressions, even though the cluster is too heterogeneous to treat as one duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": "43576 is the most representative issue to keep as canonical, since it best matches the cluster\u2019s recurring theme of v5 breakage and broad user-facing impact.", + "best_issue_reason": "issue:43576 is the best umbrella issue here because it captures the broadest v5-related breakage theme among the candidates, while the other issues are narrower or clearly unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:30990", - "right": "issue:31795", + "left": "issue:43502", + "right": "issue:44485", "accept": false, - "reason": "A loading hang for Sentence Transformers and a docs confusion issue about forward arguments are unrelated bugs." + "reason": "Different problems: offline/local-files-only network behavior versus GLM-5 RoPE implementation." }, { - "left": "issue:44075", - "right": "issue:44871", + "left": "issue:30990", + "right": "issue:31795", "accept": false, - "reason": "Optimizer argument handling and Gemma eos_token_id inconsistency affect different code paths and symptoms." + "reason": "One is a loading hang report; the other is documentation confusion about forward arguments." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger Kernel application with model_init is unrelated to the DSA ReLU discussion." + "reason": "Different subsystems and symptoms: model_init/Liger Kernel versus a DSA indexer activation issue." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLFlow reporting on failed training jobs and tokenizer loading regression are different failures." + "reason": "Training-job MLflow reporting failure is unrelated to tokenizer loading regression." }, { "left": "issue:43010", "right": "issue:44368", "accept": false, - "reason": "Cache update no_grad decoration and a tie_word_embeddings warning are not the same bug." + "reason": "Cache update no_grad semantics and a tie_word_embeddings warning are unrelated." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "local_files_only network leakage and FP8 moe save_pretrained behavior are separate issues." - }, - { - "left": "issue:43299", - "right": "issue:44556", - "accept": false, - "reason": "Both are model-loading complaints, but they target different models, versions, and likely different root causes." + "reason": "Both mention model behavior, but one is local file loading and the other is FP8 save_pretrained for MoE." }, { "left": "issue:43502", "right": "issue:44961", "accept": false, - "reason": "A real bug about local_files_only does not match the placeholder 'racoon' issue." + "reason": "No shared bug path; API request suppression versus an unrelated placeholder title." }, { - "left": "issue:44222", - "right": "issue:44230", + "left": "issue:43299", + "right": "issue:44556", "accept": false, - "reason": "FP8 save_pretrained and FP8 inference support are different code paths and not mergeable as one fix." + "reason": "Both are loading-related version regressions, but they affect different model families and likely different code paths." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "Compiled Python file errors and Pipeline save_pretrained failures are unrelated." + "reason": "Compiled Python file errors are unrelated to saving Pipeline objects." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both involve FP8/MoE, but one is save_pretrained and the other is inference support; not the same fix." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLFlow job status reporting and tensor parallel training bugs do not describe the same failure." + "reason": "MLflow failure on training job vs tensor-parallel training bug are distinct issues." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled-file errors are unrelated to PretrainedConfig type-checking regressions." + "reason": "Compiled-file runtime errors and PretrainedConfig type checking are different failure modes." }, { "left": "issue:42831", "right": "issue:43874", "accept": false, - "reason": "FP8 accuracy concerns and a missing image-processor method are different problems." + "reason": "FineGrainedFP8 accuracy and a missing image-patch method in GLM46V are unrelated." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer loading regression and Llama 3.2 vision model loading failure are not the same bug." + "reason": "Both are loading regressions, but one is tokenizer loading and the other is a specific Llama 3.2 vision model loading breakage." }, { "left": "issue:42757", "right": "issue:43576", "accept": false, - "reason": "A missing huggingface_hub import and a broken transformers env command are separate issues." + "reason": "Importing from huggingface_hub and a broken transformers env command are not the same underlying bug." }, { "left": "issue:33453", "right": "issue:36296", "accept": false, - "reason": "Tokenizer loading and tensor-parallel training problems are unrelated." + "reason": "Tokenizer loading regression and tensor-parallel training bug are unrelated." }, { "left": "issue:43827", "right": "issue:44961", "accept": false, - "reason": "Docs still referencing pipeline() and a placeholder issue have nothing in common." + "reason": "Docs still referencing pipeline() is unrelated to the placeholder racoon issue." }, { "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "A broken env command and a type-checking regression are distinct v5 issues." + "reason": "A broken env command and a PretrainedConfig type-checking regression do not share a concrete fix." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "The env command issue is not the same as a specific missing Qwen2.5-VL export." + "reason": "The env command issue is separate from a missing Qwen2_5_VLForConditionalGeneration import." }, { "left": "issue:38617", "right": "issue:43576", "accept": false, - "reason": "Missing layer_type_validation export and the env command regression affect different APIs." + "reason": "Missing layer_type_validation import and a broken env command are different export/CLI problems." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Generation cache missing in chunked_prefill and a cache-strategy typo are not the same underlying bug." + "reason": "Chunked prefill cache regression and a typo in cache strategies are not the same bug." }, { "left": "issue:44961", "right": "issue:45468", "accept": false, - "reason": "The placeholder issue 'racoon' is unrelated to Gemma-4 audio positional encoding." + "reason": "Placeholder racoon issue does not match Gemma4 audio positional encoding." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Docs confusion in model.forward and an ImportError for layer_type_validation are unrelated." + "reason": "Documentation confusion about model.forward is unrelated to an ImportError." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "AutoImageProcessor and PreTrainedModel import failures are similar in surface form but different missing symbols." + "reason": "Both are top-level import errors, but they concern different symbols and are not clearly the same export bug." }, { "left": "issue:43576", "right": "issue:44351", "accept": false, - "reason": "A broken env command is not the same as a missing HybridCache export." + "reason": "Broken env command and missing HybridCache import are separate issues." }, { "left": "issue:44908", "right": "issue:44961", "accept": false, - "reason": "A scheduler kwarg bug and a placeholder issue are unrelated." + "reason": "Scheduler kwargs handling has nothing to do with the placeholder racoon issue." }, { "left": "issue:44485", "right": "issue:44961", "accept": false, - "reason": "GLM-5 RoPE implementation discussion is unrelated to the placeholder issue." + "reason": "GLM-5 RoPE implementation and the placeholder racoon issue are unrelated." }, { - "left": "issue:36296", - "right": "issue:43901", + "left": "issue:44273", + "right": "issue:45003", "accept": false, - "reason": "Tensor parallel training bug and docs mentioning return_all_scores are different concerns." + "reason": "Lazy loading failure and unsafe sys.modules access may be related, but the titles are not specific enough to treat as the same bug with confidence." }, { - "left": "issue:41628", - "right": "issue:44355", + "left": "issue:36296", + "right": "issue:43901", "accept": false, - "reason": "ImportError for AutoImageProcessor and compiled-file runtime errors are unrelated." + "reason": "Tensor parallel training bug and TextClassificationPipeline docs mismatch are unrelated." }, { "left": "issue:36296", "right": "issue:43526", "accept": false, - "reason": "Tensor parallel training and BeitImageProcessorFast reduce_labels behavior are unrelated bugs." + "reason": "Tensor parallel training bug is unrelated to BeitImageProcessorFast reduce_labels behavior." }, { - "left": "issue:44273", - "right": "issue:45003", + "left": "issue:41628", + "right": "issue:44355", "accept": false, - "reason": "Generic lazy loading failure and unsafe sys.modules access may be related in theme, but there is no clear same code-path match here." + "reason": "Importing AutoImageProcessor and errors from compiled Python files are different failure classes." }, { - "left": "issue:44261", - "right": "issue:44360", + "left": "issue:44273", + "right": "issue:45542", "accept": false, - "reason": "MLA q_a_layernorm precision/config issue and DSA indexer ReLU discussion are unrelated." + "reason": "Lazy loading problems and a TensorBoard/TensorFlow backend error are not the same issue." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "Missing processor.save_pretrained files and a generic bug report are not the same issue." - }, - { - "left": "issue:44273", - "right": "issue:45542", - "accept": false, - "reason": "Lazy loading malfunction and a TensorFlow/tensorboard backend error are unrelated." + "reason": "Missing files on processor.save_pretrained and a generic bug report are not the same concrete bug." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "Broken env command and checkpoint reload compatibility after upgrading to v5.2/5.3 are different regression reports." + "reason": "A broken env command in v5 and checkpoint reload incompatibility across versions are related only at a broad version-breakage level." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "A Qwen3-VL-Moe bug report and missing _tp_plan for Qwen3.5Moe are different model-specific issues." + "reason": "Qwen3-VL-Moe bug report and missing _tp_plan for Qwen3_5Moe are different model-specific issues." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL bounding-box output error and PretrainedConfig type checking are unrelated." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Removing import * for performance and a missing GenerationMixin cache are unrelated changes." + "reason": "Removing import * usage and a GenerationMixin cache regression are not the same bug/change." }, { - "left": "issue:44393", - "right": "issue:45071", + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both concern version incompatibilities, but they target different model-loading paths and aren\u2019t a single concrete fix." + }, + { + "left": "issue:42371", + "right": "issue:44492", "accept": false, - "reason": "Qwen3-VL 2D bounding box errors and PretrainedConfig type-checking regressions are unrelated." + "reason": "TF32 settings guidance and a cache-strategy typo are unrelated." } ] }, @@ -13586,10 +13584,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:31515", "issue:31795", @@ -13623,7 +13621,6 @@ "issue:44393", "issue:44485", "issue:44492", - "issue:44556", "issue:44623", "issue:44655", "issue:44704", @@ -13635,30 +13632,29 @@ "issue:45071", "issue:45092", "issue:45230", + "issue:45290", "issue:45310", "issue:45341", "issue:45468", "issue:45478" ], "soft_pairs": [ - "issue:44556|issue:45092", - "issue:42371|issue:44492", "issue:42890|issue:44908", "issue:44704|issue:44908", "issue:44246|issue:44273", "issue:43408|issue:44315", - "issue:44297|issue:45230", "issue:43352|issue:44393", "issue:43576|issue:44908", "issue:43976|issue:44393", + "issue:44297|issue:45230", "issue:44908|issue:45071", "issue:44295|issue:45230", "issue:41669|issue:44351", "issue:41669|issue:45341", "issue:44492|issue:44945", "issue:41669|issue:44492", - "issue:44315|issue:44829", "issue:44393|issue:45478", + "issue:44315|issue:44829", "issue:31795|issue:33453", "issue:43827|issue:44908", "issue:44077|issue:44861", @@ -13686,322 +13682,328 @@ "issue:43606|issue:45468", "issue:42890|issue:43704", "issue:42913|issue:43066", - "issue:44908|issue:44998" + "issue:44908|issue:44998", + "issue:44861|issue:45290", + "issue:44998|issue:45341", + "issue:44492|issue:44998" ], - "prepared_review_unit_hash": "aab66a8d61556db53ea0992991f80ef79097ee011f5e5d3a74a6011f9484bf74", + "prepared_review_unit_hash": "8cc1a61a8796759623853fd4a44409bcce5464a1d6ecc362934b4042a841965e", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 46, - "serialized_chars": 40777, - "estimated_input_tokens": 10195, - "estimated_eval_tokens": 20646 + "soft_pair_count": 47, + "serialized_chars": 41097, + "estimated_input_tokens": 10275, + "estimated_eval_tokens": 20806 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous set of v5-era regression reports, with only a couple of clear duplicate-style matches. The strongest true duplicate pair is the tokenizer v4\u2192v5 behavior change report, plus a likely lazy-loading pair; most other soft links are only thematic similarities.", - "confidence": 0.53, - "canonical_issue_reason": "42913 is the broadest, cleanest tokenizer-v5 regression report in this set, and it can absorb the more specific wrong-decoder-type symptom better than the narrower variants.", + "summary": "This is a loose cluster of mostly Transformers v5 loading/tokenizer/save_pretrained regressions, mixed with unrelated performance, test, and model-specific issues. Very few pairs look like true duplicates.", + "confidence": 0.64, + "canonical_issue_reason": "issue:43723 is the most central and broadly phrased loading regression here: AutoTokenizer.from_pretrained failing in v5. It best anchors the tokenizer/loading subset of the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "It is the best single anchor for triage because it describes the broader tokenizer behavior change across v4\u2192v5, making it the most representative issue among the tokenizer-related duplicates.", + "best_issue_reason": "issue:43723 is the best representative issue for the cluster because it describes the core v5 tokenizer-loading problem without being overly narrow to one model or edge case.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44556", - "right": "issue:45092", - "accept": false, - "reason": "Both are about loading older checkpoints in v5, but one is a generic reload regression and the other is a specific remote-code/meta-init incompatibility; too broad to treat as the same bug." - }, - { - "left": "issue:42371", - "right": "issue:44492", - "accept": false, - "reason": "TF32 behavior settings and a cache-strategy typo are unrelated." - }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "SAM HQ test seeding flakiness and an inverse_sqrt scheduler kwargs bug are unrelated." + "reason": "Different subsystems: test seeding in SAM HQ vs inverse_sqrt scheduler kwargs." }, { "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "AutoProcessor kwargs passthrough and scheduler kwargs handling are unrelated code paths." + "reason": "Processor file-resolution kwargs vs scheduler configuration are unrelated." }, { "left": "issue:44246", "right": "issue:44273", - "accept": true, - "reason": "Both point to lazy-loading / import-time behavior being broken; 44273 is the broader root-cause report and 44246 is the symptom report." + "accept": false, + "reason": "Both mention loading/slow behavior, but one is import time and the other is lazy loading; not the same bug." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "A model-type mismatch warning and Liger Kernel application during model_init are unrelated." - }, - { - "left": "issue:44297", - "right": "issue:45230", - "accept": false, - "reason": "Tokenizer save_pretrained mismatch vs a generic bug report do not establish the same concrete failure." + "reason": "Model-type mismatch warning vs model_init Liger kernel application are different issues." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash-Attn2 unsupported error and Qwen3-VL 2D bbox hallucination are different problems." + "reason": "Flash Attention support error vs Qwen3-VL bbox hallucination/error are unrelated model-specific bugs." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "Broken `transformers env` output and scheduler kwargs handling are unrelated." + "reason": "Broken CLI env command vs scheduler kwarg handling are unrelated." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version compatibility and Qwen3-VL bbox output are unrelated." + "reason": "Python version compatibility vs Qwen3-VL bounding-box output are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained mismatch vs generic bug report; no concrete shared failure mode." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "A scheduler kwargs bug is unrelated to PretrainedConfig type checking." + "reason": "Scheduler kwargs bug vs PretrainedConfig type-checking regression are different code paths." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "A position_ids buffer read error and a generic bug report are not the same underlying issue." + "reason": "Buffer/position_ids access bug cannot be tied to the generic report." }, { "left": "issue:41669", "right": "issue:44351", "accept": false, - "reason": "Import-* performance work and a HybridCache import error are unrelated." + "reason": "Import-star performance regression vs missing HybridCache import are not the same issue." }, { "left": "issue:41669", "right": "issue:45341", "accept": false, - "reason": "Import-* slowdown and a testing_utils bug are unrelated." + "reason": "Import performance regression vs a testing_utils bug are unrelated." }, { "left": "issue:44492", "right": "issue:44945", "accept": false, - "reason": "Cache-strategy typo and incorrect pipeline-parallelism output are unrelated." + "reason": "Cache-strategy typo vs pipeline-parallelism output bug are unrelated." }, { "left": "issue:41669", "right": "issue:44492", "accept": false, - "reason": "They only share a broad import/runtime theme; the concrete bugs differ." + "reason": "Import-star performance regression vs a cache-strategy typo are not duplicates." }, { - "left": "issue:44315", - "right": "issue:44829", + "left": "issue:44393", + "right": "issue:45478", "accept": false, - "reason": "Liger Kernel application and flash_attention_3 training degeneration are different failures." + "reason": "Different model families and different from_pretrained failure modes." }, { - "left": "issue:44393", - "right": "issue:45478", + "left": "issue:44315", + "right": "issue:44829", "accept": false, - "reason": "Qwen3-VL 2D bbox issues and Qwen3.5 MoE from_pretrained errors are unrelated." + "reason": "model_init/Liger application bug vs flash_attention_3 training degeneration are distinct concrete problems." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion for forward arguments and tokenizer loading regression are not the same bug." + "reason": "Documentation confusion in forward args vs tokenizer loading regression are different." }, { "left": "issue:43827", "right": "issue:44908", "accept": false, - "reason": "Pipeline-removal docs and scheduler kwargs handling are unrelated." + "reason": "Docs still referencing pipeline vs scheduler kwargs bug are unrelated." }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Post-init policy for patchtsmixer and tied-weights key handling are unrelated." + "reason": "patchtsmixer post_init policy vs tied-weight-key AttributeError are unrelated." }, { "left": "issue:41669", "right": "issue:44908", "accept": false, - "reason": "Import-* slowdown and scheduler kwargs are unrelated." + "reason": "Import performance regression vs scheduler kwargs handling are unrelated." }, { "left": "issue:43723", "right": "issue:44393", "accept": false, - "reason": "Tokenizer loading via AutoTokenizer and Qwen3-VL bbox output are unrelated." + "reason": "Tokenizer loading regression vs Qwen3-VL bbox output bug are unrelated." }, { "left": "issue:44655", "right": "issue:44861", "accept": false, - "reason": "Pipeline save_pretrained failure and tied-weights AttributeError are unrelated." + "reason": "Pipeline save_pretrained issue vs _get_tied_weight_keys AttributeError are different save/load paths." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "`transformers env` CLI breakage and lazy loading are only loosely related at best, not the same concrete bug." + "reason": "CLI env command bug vs lazy loading issue are unrelated." }, { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "DSA indexer math and Gemma audio positional encoding are unrelated." + "reason": "DSA indexer ReLU bug vs Gemma audio positional encoding bug are unrelated." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "Generic bug report and Qwen3.5 MoE from_pretrained error lack enough evidence to be the same underlying issue." + "reason": "Generic bug report lacks enough detail to establish the same from_pretrained regression." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy loading and inverse_sqrt scheduler kwargs are unrelated." + "reason": "Lazy loading problem vs scheduler kwargs problem are unrelated." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-* performance and VRAM leakage in dataloader threads are unrelated." + "reason": "Import-time performance issue vs VRAM leak in Qwen3ForCausalLM are different bugs." }, { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are unrelated." + "reason": "Timestamp calculation in Qwen3VL processor vs GLM-5 RoPE implementation are different model-specific bugs." }, { - "left": "issue:43519", - "right": "issue:44908", + "left": "issue:42890", + "right": "issue:43704", "accept": false, - "reason": "Timestamp calculation and scheduler kwargs handling are unrelated." + "reason": "Test seeding flakiness vs VRAM leak are unrelated." }, { - "left": "issue:31795", - "right": "issue:34689", + "left": "issue:42913", + "right": "issue:43066", "accept": false, - "reason": "Model.forward docs confusion and Llama 3.2 model loading breakage are unrelated." + "reason": "Both are tokenizer-v5 related, but one is general behavior drift and the other is a specific decoder-type mismatch; not enough to call them the same bug." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Slow checkpoint loading and forward-doc confusion are unrelated." + "reason": "Slow checkpoint loading vs documentation confusion are different issues." }, { "left": "issue:44315", "right": "issue:45092", "accept": false, - "reason": "Liger Kernel application and old checkpoint meta-init incompatibility are unrelated." + "reason": "Both involve model creation/loading, but the concrete failures differ: Liger kernel not applied vs remote-code meta-init incompatibility." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leakage and GLM-5 RoPE implementation are unrelated." + "reason": "VRAM leak in dataloader threads vs RoPE implementation bug are unrelated." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "GLM-5 RoPE implementation and Gemma audio positional encoding are unrelated." + "reason": "Different model implementations and different failure modes." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 settings and VRAM leakage are unrelated." + "reason": "TF32 settings guidance vs VRAM leak are unrelated." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "local_files_only network access and timestamp calculation are unrelated." + "reason": "local_files_only network leakage vs Qwen3VL timestamp math are different bugs." }, { - "left": "issue:43519", - "right": "issue:44623", + "left": "issue:31795", + "right": "issue:34689", "accept": false, - "reason": "Timestamp calculation and processor.save_pretrained missing files are unrelated." + "reason": "Forward-API docs confusion vs Llama model-loading regression are unrelated." }, { - "left": "issue:43519", - "right": "issue:43704", + "left": "issue:31515", + "right": "issue:31795", "accept": false, - "reason": "Timestamp calculation and VRAM leakage are unrelated." + "reason": "Slow from_pretrained checkpoint loading vs forward-API documentation confusion are different." }, { - "left": "issue:44263", - "right": "issue:45468", + "left": "issue:44315", + "right": "issue:45092", "accept": false, - "reason": "torch.split return-value handling and Gemma audio positional encoding are unrelated." + "reason": "Not the same concrete code-path problem; one is kernel application, the other remote-code/meta-init handling." }, { - "left": "issue:44261", - "right": "issue:45468", + "left": "issue:43704", + "right": "issue:44485", "accept": false, - "reason": "MLA q_a_layernorm epsilon precision and Gemma audio positional encoding are unrelated." + "reason": "Different model bugs with no shared underlying change." }, { - "left": "issue:43606", - "right": "issue:45468", + "left": "issue:42371", + "right": "issue:43704", "accept": false, - "reason": "CPU offload device mismatch and Gemma audio positional encoding are unrelated." + "reason": "TF32 config note vs VRAM leak are unrelated." }, { - "left": "issue:42890", - "right": "issue:43704", + "left": "issue:43502", + "right": "issue:43519", "accept": false, - "reason": "SAM HQ test seeding and VRAM leakage are unrelated." + "reason": "Network access despite local_files_only vs timestamp miscalculation are unrelated." }, { "left": "issue:42913", "right": "issue:43066", - "accept": true, - "reason": "Both are tokenizer behavior regressions in v5; 43066 is the specific wrong-decoder-type manifestation of the broader behavior-change report in 42913." + "accept": false, + "reason": "Related topic, but not the same concrete tokenizer defect." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "A scheduler bug and an unrelated placeholder issue title are not duplicates." + "reason": "Scheduler kwargs bug vs unrelated placeholder issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight-key AttributeError vs chat-template tool-call crash are different code paths." + }, + { + "left": "issue:44998", + "right": "issue:45341", + "accept": false, + "reason": "Unrelated placeholder issue vs testing_utils bug." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Typo in cache strategies vs unrelated placeholder issue." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the two accepted soft links are conservative: 42913\u219443066 is a plausible tokenizer-v5 duplicate pair, and 44246\u219444273 is a reasonable lazy-loading/import-time symptom-root pairing. The remaining rejections stay appropriately narrow and do not overclaim duplicate status." - }, - "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 }, @@ -14013,10 +14015,10 @@ "review_unit_count": 43, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-15354-374", + "cluster_id": "cluster-15354-375", "nodes": [ "issue:30333", "issue:31515", @@ -14028,19 +14030,13 @@ "issue:43572", "issue:43618", "issue:44230", - "issue:44492", "issue:44556", "issue:44623", "issue:44861", "issue:44998", - "issue:45290", - "issue:45310", - "issue:45341" + "issue:45310" ], "soft_pairs": [ - "issue:44861|issue:45290", - "issue:44998|issue:45341", - "issue:44492|issue:44998", "issue:43618|issue:44998", "issue:44556|issue:44998", "issue:43525|issue:44861", @@ -14053,118 +14049,100 @@ "issue:31515|issue:33453", "issue:43298|issue:44230" ], - "prepared_review_unit_hash": "e246d8fe1c5c61c73346920283c41fc4ad81617ab315caa983d79526f103264a", + "prepared_review_unit_hash": "568e9da5ac7036f2042b93d9219b1919aead051afe305b4c78b848ee444cc123", "component_budget": { - "node_count": 374, - "item_count": 374, - "soft_pair_count": 1395, - "serialized_chars": 720165, - "estimated_input_tokens": 180042, - "estimated_eval_tokens": 360340 + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 }, "budget": { - "node_count": 18, - "item_count": 18, - "soft_pair_count": 14, - "serialized_chars": 14254, - "estimated_input_tokens": 3564, - "estimated_eval_tokens": 7384 + "node_count": 15, + "item_count": 15, + "soft_pair_count": 11, + "serialized_chars": 11708, + "estimated_input_tokens": 2927, + "estimated_eval_tokens": 6110 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All soft-pair candidates look like false positives: they share broad wording (loading, config, AttributeError, fp8, etc.) but point to different features, models, or code paths. No clear canonical duplicate emerges from this set.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This cluster is mostly driven by generic similarity around loading/import/config errors, but the pairs do not look like the same underlying bug. I would not merge any of the soft edges.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44556 is the clearest concrete regression report in the set: checkpoint reload failures after upgrading versions. It is more actionable and broadly representative than the narrower model-specific or unrelated items.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:44556 is the best single representative issue: it states a specific from_pretrained/reload regression with version context and is more directly triageable than the other titles.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44861", - "right": "issue:45290", - "accept": false, - "reason": "Different failures: tied-weights key handling crash vs chat template crash on tool-call assistant messages." - }, - { - "left": "issue:44998", - "right": "issue:45341", - "accept": false, - "reason": "Unrelated topics; one is a placeholder-like issue title, the other is a testing_utils bug." - }, - { - "left": "issue:44492", - "right": "issue:44998", - "accept": false, - "reason": "A cache-strategy typo is unrelated to the unemployment issue." - }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attentions regression is unrelated to the unrelated placeholder issue." + "reason": "CLIPOutput attention assignment regression vs an unrelated nonsense issue title; no shared bug." }, { "left": "issue:44556", "right": "issue:44998", "accept": false, - "reason": "Checkpoint reload regression is a concrete model-loading bug; the other issue is unrelated." + "reason": "Checkpoint reload regression vs unrelated issue; the similarity is only superficial text noise." }, { "left": "issue:43525", "right": "issue:44861", "accept": false, - "reason": "Both involve attribute errors, but on different configs and different code paths; not the same bug." + "reason": "Missing pad_token_id on a config class and an AttributeError in tied-weight key handling are different code paths." }, { "left": "issue:43572", "right": "issue:44861", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig is distinct from _tied_weight_keys list handling." + "reason": "StableLmConfig pad_token_idx regression and _get_tied_weight_keys list-handling crash are unrelated failures." }, { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Slow from_pretrained checkpoint loading and an ImportError for layer_type_validation are different problems." + "reason": "Slow from_pretrained checkpoint loading is not the same as an ImportError for layer_type_validation." }, { "left": "issue:44230", "right": "issue:44623", "accept": false, - "reason": "FP8 Qwen support for inference is unrelated to processor.save_pretrained missing files." + "reason": "FP8 Qwen3-VL/MoE support request and processor.save_pretrained missing files are different issues." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Pretrained loading performance and tensor-parallel training bugs are different code paths." + "reason": "Loading slowdown and tensor parallel training bug are unrelated subsystems and symptoms." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow reporting on training failure is unrelated to checkpoint loading speed." + "reason": "MLflow job status reporting on failure is unrelated to checkpoint loading speed." }, { "left": "issue:44230", "right": "issue:45310", "accept": false, - "reason": "Both mention Qwen3.5 MOE, but one is FP8 inference support and the other is a from_pretrained error; not the same concrete bug." + "reason": "FP8 inference support request and a from_pretrained error in Qwen3.5 MoE are adjacent only at the model family level, not the same bug." }, { "left": "issue:31515", "right": "issue:33453", "accept": false, - "reason": "Tokenizer-loading regression is not the same as slow checkpoint loading." + "reason": "Checkpoint loading slowness is different from a tokenizer loading regression." }, { "left": "issue:43298", "right": "issue:44230", "accept": false, - "reason": "Continuous batching audio-model support and FP8 Qwen support are separate feature gaps." + "reason": "Continuous batching audio-model support and FP8 Qwen3-VL/MoE support are different feature gaps." } ] }, @@ -14183,8 +14161,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-26350-2", "nodes": [ @@ -14194,7 +14172,7 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "623006636a96386421082465c5cb50daeec3679cfec9da9816efbb6ca8491a46", + "prepared_review_unit_hash": "b2e28509e7f45ff200e1213557c3efa19a970bca947a336aec1a85f193d954a3", "component_budget": { "node_count": 2, "item_count": 2, @@ -14216,18 +14194,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two distinct community contribution requests, not duplicates: one is for Flash Attention 2 support and the other for GGUF support, even though both mention adding architecture support.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue: issue 26350 and issue 33260 request different features and do not describe the same underlying work.", + "summary": "These are two separate feature requests for different model-loading/attention backends: Flash Attention 2 support versus GGUF support. They share a broad \u2018more architectures\u2019 framing, but not the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 26350 is a clearer representative only in the sense that it is the older and more discussed request, but it is not a true canonical for the other issue because it targets a different feature area.", "canonical_pr_reason": null, - "best_issue_reason": "If a representative is needed, issue 26350 is slightly better only because it is older and has more discussion activity, but it is not a true duplicate target for 33260.", + "best_issue_reason": "Issue 26350 has more discussion activity and is the older community contribution request, so it is the better cluster representative if one must be chosen.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Different feature requests in different areas (Flash Attention 2 vs GGUF); shared wording about architecture support is too broad to treat as the same issue." + "reason": "Reject: Flash Attention 2 support and GGUF support are different capabilities with different code paths and implementation goals; they are not duplicates of the same underlying request." } ] }, @@ -14246,8 +14224,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-27088-4", "nodes": [ @@ -14264,12 +14242,12 @@ "issue:43834|issue:44365", "issue:44365|issue:44370" ], - "prepared_review_unit_hash": "052890fc48eabc94310a0eb090528ec3080b4b22c178a55da2229a5f08bbe7fd", + "prepared_review_unit_hash": "f8263581d7ec624aa8ba217d4a0d6c3b86667d7f48eed0197cbe25a458fe7538", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4454, + "serialized_chars": 4455, "estimated_input_tokens": 1114, "estimated_eval_tokens": 2484 }, @@ -14277,7 +14255,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4454, + "serialized_chars": 4455, "estimated_input_tokens": 1114, "estimated_eval_tokens": 2484 }, @@ -14286,48 +14264,48 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are all translation-request issues with highly templated titles, but only #27088 is explicitly about Turkish; the others look like generic placeholder instances rather than the same exact underlying request. I would not merge any pair as duplicates on the available evidence.", - "confidence": 0.83, - "canonical_issue_reason": "#27088 is the most concrete and specific issue in the set: it names Turkish explicitly, while the others are generic template-style titles with no visible language details here.", + "summary": "These are all i18n documentation-translation issues, but they are not the same underlying task: #27088 is specifically for Turkish, while #43834, #44365, and #44370 are generic template-style issues for unspecified languages. They look similar textually, but represent different translations rather than duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "#27088 is the most concrete and actionable issue because it names a specific target language (Turkish) instead of the generic placeholder wording used by the others.", "canonical_pr_reason": null, - "best_issue_reason": "#27088 is the best anchor for this cluster because it is the only issue that clearly identifies the target language and therefore has the strongest standalone signal.", + "best_issue_reason": "#27088 is the best representative issue since it is specific, non-placeholder, and clearly scoped to one translation effort.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Both are translation-related, but the evidence only shows a generic template on one side and a concrete Turkish request on the other; that is not enough to say they are the same underlying issue." + "reason": "Same translation template family, but Turkish vs. an unspecified language target means they are different tasks, not duplicates." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "The titles are similar at a template level, but there is no concrete match on the target language or request details beyond docs translation in general." + "reason": "These are structurally similar issue templates, but they refer to different language targets, so they should not be merged." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Same broad translation theme, but not the same specific request: one is explicitly Turkish, the other is only a placeholder-style language template." + "reason": "The titles are similar, but one is concrete Turkish and the other is a generic placeholder issue; not the same underlying request." }, { "left": "issue:43834", "right": "issue:44370", "accept": false, - "reason": "Both appear to be generic translation-template issues, but there is no evidence they refer to the same language or the same exact change/request." + "reason": "Both are generic i18n template issues, but there is no evidence they target the same exact language or translation effort." }, { "left": "issue:43834", "right": "issue:44365", "accept": false, - "reason": "High textual similarity alone is not enough here; these look like separate template instances rather than one duplicate report." + "reason": "Highly similar wording, but placeholder language-code/language-name entries indicate separate translation requests rather than one duplicate bug." }, { "left": "issue:44365", "right": "issue:44370", "accept": false, - "reason": "The available metadata does not show a shared concrete target beyond a generic docs-translation template, so they should not be treated as the same issue." + "reason": "They look like cloned templates, but without the same explicit language target they are not safely duplicate-identical." } ] }, @@ -14346,8 +14324,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-36817-14", "nodes": [ @@ -14396,12 +14374,12 @@ "issue:43837|issue:45027", "issue:43837|issue:45177" ], - "prepared_review_unit_hash": "6f6fffab25a25d3979a0af8362116356cd1286ca221634ef1b23283525d68557", + "prepared_review_unit_hash": "ed9c3d5f92306e8f7d99385858fdba72b519907f63d89e4eb1eee77785cae922", "component_budget": { "node_count": 14, "item_count": 14, "soft_pair_count": 28, - "serialized_chars": 17648, + "serialized_chars": 17645, "estimated_input_tokens": 4412, "estimated_eval_tokens": 9080 }, @@ -14409,7 +14387,7 @@ "node_count": 14, "item_count": 14, "soft_pair_count": 28, - "serialized_chars": 17648, + "serialized_chars": 17645, "estimated_input_tokens": 4412, "estimated_eval_tokens": 9080 }, @@ -14418,78 +14396,78 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The items are mostly independent model-support requests, not duplicates: they span different model families and even different modalities (text, audio, OCR, perception). The apparent overlaps are only broad thematic similarity, so none should be merged as the same underlying issue.", + "summary": "This cluster is a loose semantic group of unrelated model-request issues across vision, speech, OCR, and BERT-like architectures. No soft pairs look like the same underlying bug/change, so none should be merged as duplicates.", "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue fits this set: the requests target different models and capabilities, so they are not duplicates of one another.", + "canonical_issue_reason": "issue:42738 is the broadest umbrella in the set ('BERT-like models with RoPE'), but the cluster is still too heterogeneous for a true duplicate canonical.", "canonical_pr_reason": null, - "best_issue_reason": "If a representative is needed, issue 43671 is the most active and clearly described among the audio-related requests, but it still does not subsume the others.", + "best_issue_reason": "issue:42738 is the best representative only by breadth, not because it matches the other requests; it is the least specific of the issues listed.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "Both are model-add requests, but EuroBert and DEIMv2 are different architectures with separate implementation work." + "reason": "EuroBert config and DEIMv2 are different model families with no shared concrete fix." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert config support is unrelated to adding Qwen3-TTS support." + "reason": "EuroBert config and Qwen3-TTS are unrelated modalities (text model config vs speech TTS support)." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are different models and different modalities." + "reason": "NeoBERT and Qwen3-ASR are different concrete models; this is not the same request." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and the weight-sparse transformer request are distinct model families and changes." + "reason": "NeoBERT and a weight-sparse transformer request target different architecture work." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT text-model support is unrelated to LFM2.5 Audio support." + "reason": "NeoBERT and LFM2.5 Audio are unrelated model requests in different domains." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are separate model additions, not the same request." + "reason": "NeoBERT and ModernVBERT are distinct model families, not the same underlying change." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral TTS support are unrelated model integrations." + "reason": "NeoBERT and Voxtral TTS are different speech/model support requests." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture support and Qwen3-ASR support are different code paths." + "reason": "MiniCPM3 and Qwen3-ASR are different model additions, not duplicates." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 and the sparse-transformer request are distinct model changes." + "reason": "MiniCPM3 support and a weight-sparse transformer request are different concrete changes." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are different model families." + "reason": "MiniCPM3 and ModernVBERT are separate model families." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 support is unrelated to Voxtral TTS support." + "reason": "MiniCPM3 and Voxtral TTS are unrelated across model family and modality." }, { "left": "issue:41211", @@ -14501,97 +14479,97 @@ "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT and Qwen3-ASR are different models and modalities." + "reason": "ModernVBERT and Qwen3-ASR are different concrete model support requests." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are separate additions." + "reason": "ModernVBERT and Qwen3-TTS are different model families/modalities." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT and BERT-like RoPE models are related only at a high level, not the same request." + "reason": "ModernVBERT is a specific model request, while BERT-like models with RoPE is a broad architectural umbrella; not the same issue." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like RoPE support and Voxtral TTS support are unrelated." + "reason": "BERT-like RoPE models and Voxtral TTS are unrelated requests." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Both mention Qwen3, but TTS and ASR are different modalities and likely different implementations." + "reason": "Qwen3-TTS and Qwen3-ASR are related speech area requests but different concrete model support changes." }, { "left": "issue:43671", - "right": "issue:45027", + "right": "issue:43909", "accept": false, - "reason": "Qwen3-TTS and Voxtral-4B-TTS are different TTS models with separate support work." + "reason": "Qwen3-TTS and LFM2.5 Audio are different speech/audio model requests." }, { "left": "issue:43671", - "right": "issue:45177", + "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS support and DeepSeek-OCR2 support are unrelated." + "reason": "Qwen3-TTS and weight-sparse transformer are unrelated work items." }, { "left": "issue:43671", - "right": "issue:44121", + "right": "issue:45027", "accept": false, - "reason": "A TTS model request is unrelated to the sparse-transformer request." + "reason": "Both are TTS-related, but they target different models and are not the same concrete change." }, { "left": "issue:43671", - "right": "issue:43909", + "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and LFM2.5 Audio are different audio-model integrations." + "reason": "Speech TTS support and DeepSeek-OCR2 are different modalities and code paths." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are different models and tasks." + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio model requests." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "ASR support and EUPE perception-encoder support are unrelated." + "reason": "Qwen3-ASR and EUPE target different model families and tasks." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR and the weight-sparse transformer request are different changes." + "reason": "ASR support and a sparse-transformer architecture request are not the same underlying change." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio support and SAM3-LiteText are unrelated model requests." + "reason": "LFM2.5 Audio and SAM3-LiteText are unrelated modalities." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Sparse-transformer support and EUPE are different model architectures." + "reason": "Weight-sparse transformer and EUPE are different architecture/model requests." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral TTS are different modalities and models." + "reason": "ASR and TTS are different speech tasks with separate implementations." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR and DeepSeek-OCR2 are unrelated model additions." + "reason": "ASR support and OCR model support are unrelated." } ] }, @@ -14610,8 +14588,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-37934-6", "nodes": [ @@ -14632,7 +14610,7 @@ "issue:43668|issue:44031", "issue:43668|issue:43813" ], - "prepared_review_unit_hash": "cbb40701e334e357ececad06a42e98e6256a988d685d99894e3bc251c84264fb", + "prepared_review_unit_hash": "f03f7869256ab5f3c3eaae8eb5cce1256d3dab1b0accfdb14d1b9214cf225b9d", "component_budget": { "node_count": 6, "item_count": 6, @@ -14654,60 +14632,60 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These soft-linked issues are not duplicates: they span unrelated bugs in different areas (model normalization, Mask2Former loss handling, config typing, a typo, and tokenizer warnings). No soft pair looks like the same underlying fix.", - "confidence": 0.95, - "canonical_issue_reason": "issue:44031 is the broadest, most discussion-heavy bug report in the set, so it is the closest thing to a cluster anchor even though the items are actually heterogeneous.", + "summary": "These are not duplicates as a cluster; the titles describe separate bugs spanning different models, configs, and warning paths. The soft similarity is mostly superficial wording overlap.", + "confidence": 0.93, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43240 is the most concrete, code-path-oriented bug report and is the strongest standalone issue here, but it still does not unify the rest of the set.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Different concerns: model normalization implementation vs. a config type-hint typo." + "reason": "Different components and bug types: Llama4 norm implementation question vs ModernBERT config type hint. No shared code path." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "One is about Mask2Former's ignored loss value; the other is about kwargs not reaching a loss function\u2014related only loosely by training code." + "reason": "Both involve loss-related model code, but one is Mask2Former's unused ignore_value and the other is kwargs not being forwarded to a loss function. Different concrete defects." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Different subsystems and symptoms: Mask2Former loss handling vs tokenizer regex warnings." + "reason": "Mask2Former loss handling vs tokenizer regex warnings after a version change; unrelated subsystems and symptoms." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "A runtime model bug and a documentation/attribute typo are not the same underlying issue." + "reason": "Mask2Former model bug vs a typo in a quantization attribute name. Not the same underlying issue." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former loss behavior does not match a ModernBERT config type annotation issue." + "reason": "Unrelated model-loss bug vs ModernBERT config annotation issue." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwarg forwarding bug vs incorrect config type hint; no shared concrete code-path." + "reason": "One is about forwarding kwargs into loss calculation; the other is an incorrect type hint. Different bug classes and code paths." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "Unrelated: ModernBERT `norm_eps` typing vs tokenizer regex warning after 4.57.3." + "reason": "Config type hint error vs tokenizer warning regression. No evidence they share a fix." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Both are small fixes, but one is a type-hint error and the other is a typo in a quantization attribute name." + "reason": "ModernBERT config annotation issue vs a typo in a quantization symbol. Different files and changes." } ] }, @@ -14726,8 +14704,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41084-2", "nodes": [ @@ -14737,7 +14715,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "91c1774746b59900bfb7d0d382719d9768d814ddfa7e0d55e6aaf06c0eac58bf", + "prepared_review_unit_hash": "693908d4afb82957a1dbf0e251754131d5f0928c1ee2034a1bcc72df8163724b", "component_budget": { "node_count": 2, "item_count": 2, @@ -14759,18 +14737,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are about different requests: one is a generic feature request for block decoding, the other is specific support for DeepSeek v3.2. There is no strong evidence they describe the same underlying bug or change.", - "confidence": 0.94, - "canonical_issue_reason": "No clear canonical issue: the titles and likely scopes differ substantially, so neither issue is a duplicate of the other.", + "summary": "The two issues are not duplicates: one is about Set Block Decoding and the other is a DeepSeek v3.2 support request.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 43037 is the more specific and actionable issue; 41084 is vague and appears to describe a different topic.", "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good global canonical for the pair because they target different functionality and do not appear to overlap in concrete implementation scope.", + "best_issue_reason": "43037 is the best issue to represent this set because its title states a concrete feature request, while 41084 is too ambiguous.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Reject: 'Set Block Decoding' is a broad decoding feature request, while 'DeepSeek v3.2 support' is a model-specific compatibility request. They do not look like the same underlying bug or change." + "reason": "Reject: the titles point to unrelated concerns, with no shared targets, filenames, or explicit linkage indicating the same underlying bug or change." } ] }, @@ -14789,8 +14767,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43598-2", "nodes": [ @@ -14800,7 +14778,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "ecf2a1eee88737bc84261a7778e4ae86a436f83ad6646b2edd52ac142c5a3d3c", + "prepared_review_unit_hash": "961c486aa4857ce960c12c71c606d39047858f6b10168b94629e939dbcfa6b89", "component_budget": { "node_count": 2, "item_count": 2, @@ -14822,18 +14800,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are not duplicates: one is about making Trainer\u2019s `remove_unused_column` behavior more customizable, while the other is about FLOPS estimation for custom trainers. They touch the same general Trainer area but describe different problems and fixes.", - "confidence": 0.93, - "canonical_issue_reason": null, + "summary": "The two issues both mention Trainer customization, but they address different behaviors: one is about `remove_unused_column`, the other about FLOPS estimation with custom trainers. They do not look like the same underlying bug or change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #43598 is the clearer, more specific feature/behavior request around `remove_unused_column` customizability in Trainer.", "canonical_pr_reason": null, - "best_issue_reason": "No single canonical issue emerges because the issues are unrelated; if forced to pick a representative, #43598 is a clearer, more concrete Trainer customization request.", + "best_issue_reason": "#43598 is the best representative issue because it is narrowly scoped and describes a concrete Trainer behavior to revisit.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Different underlying concerns: `remove_unused_column` customization vs FLOPS estimation for custom trainers. Shared Trainer context is too broad to treat as the same bug." + "reason": "These are related only at a broad Trainer/customization level; one concerns column removal logic, the other FLOPS estimation for custom trainers. Different code paths and fixes." } ] }, @@ -14852,8 +14830,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43684-2", "nodes": [ @@ -14863,7 +14841,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "35c0a273185b383ecb8b759903ad08e0b53bbac811f47bd292a329177ab47b8e", + "prepared_review_unit_hash": "d11b7211d10638fd8e7c726ac91edd7fcd9d50882ce3aeb5aa7a8223796d52c8", "component_budget": { "node_count": 2, "item_count": 2, @@ -14885,18 +14863,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one requests core model registration for Qwen3-Omni, while the other requests sequence-classification support for Qwen3.5. They touch different model variants and different functionality layers.", + "summary": "These are related only at a very broad level (both request new Qwen model support), but they target different model families and different AutoModel classes, so they are not duplicates.", "confidence": 0.96, - "canonical_issue_reason": "No canonical duplicate issue: the requests target different capabilities (AutoModel/AutoModelForConditionalGeneration registration vs AutoModelForSequenceClassification support) and different Qwen model families.", + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the broader, more foundational request and is the closest thing to a representative issue, but it still should not be treated as a duplicate of 44405.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying changes: Qwen3-Omni model registration for base/generative auto classes vs Qwen3.5 sequence-classification head support. Not the same bug or mergeable change." + "reason": "Different underlying requests: one asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, the other asks for AutoModelForSequenceClassification support for Qwen3.5. They do not describe the same bug or change." } ] }, @@ -14915,8 +14893,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43835-4", "nodes": [ @@ -14933,12 +14911,12 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "db474929e338b9669489c06f815c46120c6b25f1f41f95afd8094b2742732575", + "prepared_review_unit_hash": "aad5f9af27c90c100a61759884863d662215881ebdd249e333ac00ddd2a1dfcd", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4266, + "serialized_chars": 4265, "estimated_input_tokens": 1067, "estimated_eval_tokens": 2390 }, @@ -14946,7 +14924,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4266, + "serialized_chars": 4265, "estimated_input_tokens": 1067, "estimated_eval_tokens": 2390 }, @@ -14955,48 +14933,48 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four items are extremely sparse issue reports with only generic or near-empty titles. The three \"huggingface\" issues look superficially related, but there is not enough body detail to confirm they describe the same underlying problem. No PRs are present.", + "summary": "The three \"huggingface \" issues look like a likely duplicate cluster, with 43845 as the best representative; 43835 ('Fett') appears unrelated. No PRs are present.", "confidence": 0.41, - "canonical_issue_reason": "issue:43859 is the only report with any discussion activity, so it is the best available anchor among otherwise low-information issues.", + "canonical_issue_reason": "43845 is the earliest of the near-identical \"huggingface \" reports and is the best representative of that cluster; 43835 has a different title and does not fit.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43859 is the best global issue candidate for triage because it has at least minimal activity and matches the repeated title pattern, but the reports are too under-specified to be strongly meaningful.", + "best_issue_reason": "43845 is the cleanest canonical candidate because it is earliest and shares the same title pattern as 43846 and 43859.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Both are too sparse and the titles are not enough to confirm the same bug or request." + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Only weak textual similarity is available; there is no body evidence that they describe the same underlying issue." + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "The reports are under-specified and do not provide enough shared detail to justify duplicate merging." + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." }, { "left": "issue:43845", "right": "issue:43846", "accept": false, - "reason": "These titles match exactly, but the bodies are too thin to confirm they are truly the same issue." + "reason": "Titles match, but there is no description or target detail to prove they are the same underlying issue." }, { "left": "issue:43845", "right": "issue:43859", "accept": false, - "reason": "There is not enough substantive content beyond a generic title to treat them as the same report." + "reason": "Same broad title, but the available metadata is too thin to confirm a single underlying bug/change." }, { "left": "issue:43846", "right": "issue:43859", "accept": false, - "reason": "The similarity is only superficial; the available text does not establish a shared underlying bug." + "reason": "Same broad title, but the available metadata is too thin to confirm a single underlying bug/change." } ] }, @@ -15015,8 +14993,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-44371-5", "nodes": [ @@ -15037,7 +15015,7 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "ebc191fa19a7d622c923a6869092f3584dbc52106d12019420a73aa669b968bc", + "prepared_review_unit_hash": "5efaaf7e352bd932c952e40cd7fafea3360f8771eec3fbe0c9369baf4d7bb191", "component_budget": { "node_count": 5, "item_count": 5, @@ -15059,66 +15037,66 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are low-information issues with placeholder/spam-like titles and no evidence of a shared concrete bug or change; they do not form a duplicate cluster.", - "confidence": 0.94, - "canonical_issue_reason": "No strong canonical issue: the set is dominated by spam/placeholder titles and lacks any substantive description or shared bug details to anchor deduplication.", + "summary": "This cluster is mostly low-signal spam/generic issues with no concrete shared bug or change, so there is no clear duplicate set or canonical representative.", + "confidence": 0.92, + "canonical_issue_reason": "No canonical issue: the items are either spam placeholders or unrelated generic titles, with no substantive body text, targets, or shared technical context.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #44418 is the least placeholder-like title and is the only one that reads like a normal issue, but it still has no concrete enough content to serve as a reliable canonical.", + "best_issue_reason": "If forced to pick a representative, issue 44418 is the only non-spam-looking item, but it still does not relate to the others and is not a strong cluster anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are placeholder spam titles; no shared underlying bug or change is described." + "reason": "Both are just '' with no descriptive content; identical titles alone are not enough to establish the same underlying issue." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are placeholder spam titles; nothing indicates the same issue." + "reason": "Same as above: no body details or technical context, so there is no basis to treat them as duplicates of one bug." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is spam-like and the other is a generic documentation suggestion, but there is no concrete overlap." + "reason": "One is spam and the other is a generic documentation request; these do not indicate the same underlying problem." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Unrelated titles with no shared bug, feature, or code path." + "reason": "'' and 'Check out \"Google LLC\"' are clearly unrelated on their face, with no shared issue description." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "Generic documentation suggestion versus placeholder spam; not the same underlying request." + "reason": "A documentation enhancement suggestion is not the same as a spam placeholder; no shared concrete topic is visible." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "Generic documentation suggestion versus placeholder spam; no evidence of duplication." + "reason": "The titles suggest unrelated content, and there is no body evidence tying them to the same bug or change." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are spam placeholders, but there is no specific issue/change content to confirm they are duplicates." + "reason": "Both are spam placeholders, but there is no substantive evidence they refer to the same underlying report." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Placeholder spam title and an unrelated named topic; not the same issue." + "reason": "These titles are unrelated and provide no common technical context or explicit link." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Placeholder spam title and an unrelated named topic; no duplicate evidence." + "reason": "No shared bug/change is evident; one is spam and the other is an unrelated promotional-looking title." } ] }, @@ -15137,8 +15115,8 @@ "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45145-2", "nodes": [ @@ -15148,7 +15126,7 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "b15563227e049b26fb20bc6684e70f0210f8d8da6a2719650aa288369fc25ff6", + "prepared_review_unit_hash": "aed16a21dd49a1b185838ebfa549975ae96756705e354436ee70da2041d94afa", "component_budget": { "node_count": 2, "item_count": 2, @@ -15170,28 +15148,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues are very likely duplicates: they have the exact same title and soft-similarity evidence, but the packet does not prove their bodies are identical. #45145 is the best canonical issue because it was created first and already has an inbound reference.", - "confidence": 0.89, - "canonical_issue_reason": "#45145 is the better canonical issue because it was created earlier and already has an inbound reference, making it the most established representative for this likely-duplicate pair.", + "summary": "These two closed issues appear highly similar and likely refer to the same energy-efficiency proposal, but the evidence is soft-only, so the match should remain tentative rather than treated as a confirmed duplicate.", + "confidence": 0.87, + "canonical_issue_reason": "Issue #45145 is the better canonical thread because it was created first and has slightly more discussion signal (an inbound reference).", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best issue to keep as the primary record because it is earlier and already has more thread connectivity than #45151.", + "best_issue_reason": "#45145 is the strongest representative issue: earliest created, same title/topic, and marginally more established in the graph.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", - "accept": true, - "reason": "Likely duplicate based on identical title and soft-similarity evidence; the packet does not prove the bodies are identical, but it is reasonable to treat them as the same request." + "accept": false, + "reason": "Reject: the pair is supported only by soft similarity, and deterministic_accept is false; identical titles alone are not enough to confirm a duplicate." } ] }, "evaluator_result": { "accept": false, - "feedback": "The packet supports that the issues share the exact same title and have soft-similarity evidence, but it does not directly show their bodies are \"near-identical\" or that they are definitively the same request. The duplicate verdict should be more conservative (e.g., \"likely duplicate based on identical title and soft similarity\")." + "feedback": "Reject: the pair is only supported by soft similarity, and the packet\u2019s deterministic_accept is false. The analyst\u2019s soft-edge verdict is too strong for the evidence. The summary also overstates certainty by calling them duplicates; it should stay more tentative (e.g. \u201cappear highly similar\u201d)." }, "evaluator_used": true, "retried": true, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -15252,13 +15230,13 @@ { "label": "pull_request", "component_index": 1, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-39895-13", "nodes": [ @@ -15274,7 +15252,7 @@ "pull_request:45170", "pull_request:45213", "pull_request:45587", - "pull_request:45622" + "pull_request:45655" ], "soft_pairs": [ "pull_request:39895|pull_request:43424", @@ -15285,153 +15263,165 @@ "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:43823|pull_request:45622", "pull_request:43995|pull_request:45213", "pull_request:43995|pull_request:44827", - "pull_request:44827|pull_request:45622", "pull_request:44827|pull_request:45587", "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:45655", "pull_request:44827|pull_request:45213", "pull_request:43823|pull_request:45213", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "156f95bf4d19d1d8dc1429b29f2b42d210047a30155c553c1e9bbd86cc7c1c6b", + "prepared_review_unit_hash": "9beab45c0cdf32ca446e10094003609c32cf583151a269e0617e184210320c2b", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 35170, - "estimated_input_tokens": 8793, - "estimated_eval_tokens": 17842 + "serialized_chars": 35210, + "estimated_input_tokens": 8803, + "estimated_eval_tokens": 17862 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 18, - "serialized_chars": 35170, - "estimated_input_tokens": 8793, - "estimated_eval_tokens": 17842 + "serialized_chars": 35210, + "estimated_input_tokens": 8803, + "estimated_eval_tokens": 17862 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly false-positive similarity across unrelated PRs. The only true duplicate pair is the two `fixed_cross_entropy` fixes (43251 and 43254), which both forward `weight`/`label_smoothing` to `nn.functional.cross_entropy` in `src/transformers/loss/loss_utils.py` for issue 43240.", - "confidence": 0.94, + "summary": "This cluster is mostly a set of unrelated PRs across different models, docs, and utilities. The only true duplicate is the pair of `fixed_cross_entropy` PRs (`43251` and `43254`), which address the same loss-function kwargs passthrough bug. Everything else should be kept separate.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43251 is the best representative of the real duplicate pair: it has the explicit issue link, the same concrete `fixed_cross_entropy` fix, and the more complete/traceable description.", + "canonical_pr_reason": "`43251` is the best canonical PR for the duplicate loss-function fix: it explicitly links issue `43240`, has the fuller kwargs handling (`weight`, `label_smoothing`), and the broader review context confirms it targets the same concrete bug as `43254`.", "best_issue_reason": null, - "best_pr_reason": "PR 43251 is the strongest canonical PR overall because it directly addresses the underlying bug, has the issue target attached, and matches the exact code-path change duplicated by 43254.", + "best_pr_reason": "`43251` is the strongest representative PR in the only duplicate pair because it is explicitly issue-backed and directly fixes the cross-entropy kwargs forwarding bug with the clearest implementation context.", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "Unrelated changes: VideoPrism model/docs addition vs an ExecuTorch export test for dynamic shapes." + "reason": "Unrelated PRs: one adds VideoPrism model support, the other adds an Executorch dynamic-shape test. Same broad framework but not the same change or bug." }, { "left": "pull_request:41855", "right": "pull_request:43823", "accept": false, - "reason": "Tokenizer helper methods for Mistral Common are unrelated to adding MobileLLM." + "reason": "Different tokenizer maintenance vs adding a new MobileLLM model. No shared concrete bug or code path." }, { "left": "pull_request:41855", "right": "pull_request:43424", "accept": false, - "reason": "Missing tokenizer methods vs dynamic-shape export tests; different subsystems and bugs." + "reason": "Tokenizer method additions and Executorch export tests are unrelated changes." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Processor auto-loading robustness work is not the same change as introducing MobileLLM." + "reason": "Processor robustness refactor and new MobileLLM model integration are different PRs in different areas." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss utility kwargs fix and a new model implementation are unrelated." + "reason": "Loss utility kwargs fix vs new model addition; no same underlying bug or change." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "`cross_entropy` kwargs forwarding is unrelated to Mistral4 test/integration changes." + "reason": "Cross-entropy kwargs fix and Mistral4 test/model work are unrelated." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "ExecuTorch export testing and Mistral4 test fixes do not target the same underlying bug." + "reason": "Executorch dynamic-shape test and Mistral4 fixes are different code paths." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "Dynamic-shape export testing is unrelated to adding MobileLLM." + "reason": "Executorch export test versus MobileLLM model addition; not a duplicate." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New MobileLLM model vs a class-annotation resolution fix; unrelated." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "MobileLLM model scaffolding and Falcon output-interface refactoring are different changes." + "reason": "New model integration and Falcon output-interface refactor are different changes." }, { "left": "pull_request:43823", - "right": "pull_request:45622", + "right": "pull_request:45213", "accept": false, - "reason": "New model addition vs PEFT constructor fix; no shared underlying bug or change." + "reason": "MobileLLM model PR and a model-creation skill/meta PR are not the same underlying work." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactoring and the model-creation skill are unrelated." + "reason": "Falcon refactor and model-creation skill are unrelated." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon output-collection refactor is not the same as Mistral4 test fixes." + "reason": "Falcon output-capture refactor and Mistral4 test fix are different models and code paths." }, { "left": "pull_request:44827", - "right": "pull_request:45622", + "right": "pull_request:45587", "accept": false, - "reason": "Mistral4 tests and PEFT constructor internals only superficially overlap; they fix different code paths." + "reason": "Mistral4 test/model work and docs-only continuous batching memory management are unrelated." }, { "left": "pull_request:44827", - "right": "pull_request:45587", + "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test fixes are unrelated to docs for continuous batching memory management." + "reason": "Mistral4 fixes and layrnorm-to-layernorm renaming are different changes; same area of Transformers does not make them duplicates." }, { "left": "pull_request:44827", - "right": "pull_request:45170", + "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 test/integration changes are not the same as the `layrnorm` to `layernorm` rename across CLIP-like models." + "reason": "Mistral4 test/model work and `cls.config` resolution ordering fix are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "A Mistral4 test fix is unrelated to the model-creation skill PR." + "reason": "Mistral4 test/model work and the model-creation skill are not the same concrete bug/change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "New MobileLLM model PR and meta skill PR are unrelated." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Both PRs patch the same `fixed_cross_entropy` function in `loss_utils.py` to forward supported kwargs (`weight` and `label_smoothing`) for the same issue 43240, so they are the same underlying fix." + "reason": "These are the same underlying `fixed_cross_entropy` bug fix: both add support for passing cross-entropy kwargs (`weight`/`label_smoothing`) to `nn.functional.cross_entropy` for issue `43240`. The diffs are near-identical and could be merged as one PR." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet. The only accepted pair, 43251/43254, clearly patch the same `fixed_cross_entropy` function in the same file for the same issue, so that mergeable-duplicate verdict is justified. The remaining soft-edge rejections are conservative and consistent with the visible titles/diffs; no evidence is overstated." + "feedback": "The summary is well grounded in the packet and stays conservative. The only accepted duplicate pair is `43251`/`43254`, and the shared issue link plus nearly identical `fixed_cross_entropy` diffs support that. The other pairwise rejections are also appropriately cautious and do not overclaim duplication from broad topical similarity alone." }, "evaluator_used": true, "retried": false, @@ -15442,13 +15432,13 @@ { "label": "pull_request", "component_index": 2, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-40546-2", "nodes": [ @@ -15458,12 +15448,12 @@ "soft_pairs": [ "pull_request:40546|pull_request:42781" ], - "prepared_review_unit_hash": "dd556f46d7e4e8ce6fe09a5efa0f34970929730aa5aac8cc8ee5de80c4a4f3d7", + "prepared_review_unit_hash": "056a0e683381452301093e34d282e3062a6ead88d93df8c8dc2e8b752e6fd914", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7226, + "serialized_chars": 7225, "estimated_input_tokens": 1807, "estimated_eval_tokens": 3870 }, @@ -15471,7 +15461,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7226, + "serialized_chars": 7225, "estimated_input_tokens": 1807, "estimated_eval_tokens": 3870 }, @@ -15480,18 +15470,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related VibeVoice PRs, but they are not the same change: #40546 adds the base VibeVoice model, while #42781 adds VibeVoice Realtime plus an acoustic tokenizer and generation changes. They overlap on shared model plumbing, but the underlying scope is different.", - "confidence": 0.89, + "summary": "These are related VibeVoice workstreams, but not duplicates: one PR adds the base VibeVoice model integration, while the other adds a realtime-oriented extension plus acoustic tokenizer and generation changes.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "#40546 is the better canonical representative of the core VibeVoice addition: it introduces the base model/docs and has substantial review activity. #42781 is a narrower, draft realtime follow-up with extra tokenizer/generation work.", + "canonical_pr_reason": "PR #40546 is the better canonical representative because it is the original, more complete base VibeVoice implementation and is farther along in review, whereas #42781 is a separate draft extension for realtime support.", "best_issue_reason": null, - "best_pr_reason": "#40546 is the best overall PR representative because it is the primary base implementation of VibeVoice, whereas #42781 is a draft extension focused on realtime support and a new tokenizer.", + "best_pr_reason": "PR #40546 best captures the core VibeVoice integration work in the cluster; #42781 builds on that model family but targets a different feature set and code path.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Shared VibeVoice files indicate overlap, but #40546 implements the base model while #42781 adds realtime-specific functionality and an acoustic tokenizer; they are related follow-ups, not the same concrete PR." + "reason": "They overlap on the VibeVoice family and shared registry/docs files, but the concrete changes differ: #40546 implements the base model, while #42781 adds realtime-specific support with an acoustic tokenizer and related generation updates. That is related work, not the same underlying change." } ] }, @@ -15505,13 +15495,13 @@ { "label": "pull_request", "component_index": 3, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-40755-2", "nodes": [ @@ -15521,7 +15511,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "40b36191e1af9c572de66ba2a08ca7e589e1ddaa2a7ab35a06a4f36a8558ae56", + "prepared_review_unit_hash": "c31f57c4eace8cfaa3614dd6e1d646f2aa5dc0cc01c85d76b69e66b6aa42aeea", "component_budget": { "node_count": 2, "item_count": 2, @@ -15543,18 +15533,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two open PRs are unrelated feature additions for different models: TimesFM covariate forecasting versus DINOv3 image classification support. They do not look like the same underlying change and should not be clustered as duplicates.", + "summary": "These are unrelated PRs: one adds TimesFM covariate forecasting support, the other adds DINOv3 image classification support. They touch different models, code paths, and documentation, so they should not be clustered together as duplicates.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #41224 is the better canonical PR for this cluster only as a representative artifact: it has a narrower, clearly defined feature addition and more review activity. However, it is not a duplicate of #40755.", + "canonical_pr_reason": "No canonical PR across the pair; they implement different features for different model families and are not mergeable into one change.", "best_issue_reason": null, - "best_pr_reason": "PR #41224 is the stronger representative PR because its scope is more concrete and self-contained. PR #40755 is a separate TimesFM enhancement with different code paths and should remain distinct.", + "best_pr_reason": "PR #41224 is the cleaner standalone representative because it has a narrower, well-scoped feature addition with focused model/auto/tests changes.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different models and different feature areas: TimesFM forecasting with covariates versus DINOv3 image classification support. No shared concrete bug/change path, so they are not duplicates." + "reason": "Different underlying changes: TimesFM covariate forecasting vs. DINOv3 image classification support. They affect separate model implementations and cannot plausibly be merged into one PR." } ] }, @@ -15568,13 +15558,13 @@ { "label": "pull_request", "component_index": 4, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41356-2", "nodes": [ @@ -15584,7 +15574,7 @@ "soft_pairs": [ "pull_request:41356|pull_request:44339" ], - "prepared_review_unit_hash": "5ca5622c9633bbe7eef9d70a2eb0e3cf7982d42dc10a27ffb1990f021d804327", + "prepared_review_unit_hash": "aead1308c167b0bae6ff11bf9358340baa5e76f816618383418b505a10a9cd01", "component_budget": { "node_count": 2, "item_count": 2, @@ -15606,41 +15596,38 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are PRs for adding DEIMv2 support to Transformers. PR #44339 is the more complete later submission and appears to subsume the earlier #41356 work, so it is the better canonical PR candidate. There is no issue in this cluster.", - "confidence": 0.86, + "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": "#44339 is the broader, later DEIMv2 implementation and includes the core model integration plus additional framework wiring (auto-mappings, conversion, loss/modular updates), making it the best representative PR.", + "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": "#44339 is the most complete and maintainable PR for the DEIMv2 addition; it covers the same feature area as #41356 but with substantially more finished integration work.", + "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": true, - "reason": "Both PRs target the same DEIMv2 model addition, and #44339 looks like a superset that could plausibly replace the earlier implementation in one merged change." + "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": { - "accept": true, - "feedback": "The summary is grounded in the packet: both items are PRs tied to the same issue target (41211) and the later PR #44339 clearly adds broader DEIMv2 integration while overlapping core files with #41356. The soft-edge verdict is conservative, phrased as a plausible supersession rather than a certainty. No issue is present in this cluster." - }, - "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 }, { "label": "pull_request", "component_index": 5, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-41797-2", "nodes": [ @@ -15650,7 +15637,7 @@ "soft_pairs": [ "pull_request:41797|pull_request:45186" ], - "prepared_review_unit_hash": "f1d7c5ed988d770f7a01eb956b28dcb7cdb08b3bf52a7c2e22052331bb08a42c", + "prepared_review_unit_hash": "b8316d46d769ca2ca3baad4fe9701cf71ca86151a807743105f2249ca450ade0", "component_budget": { "node_count": 2, "item_count": 2, @@ -15672,18 +15659,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate model-addition PRs: one adds DeepSeek OCR, the other adds Isaac. They overlap in Transformers scaffolding and docs, but they implement different models and code paths, so they are not duplicates.", - "confidence": 0.97, + "summary": "These are two separate model-addition PRs for different models (DeepSeek OCR vs Isaac), not the same underlying bug or change.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "41797 is the more established and broader PR, with substantial implementation and review activity, but it is not the same change as 45186.", + "canonical_pr_reason": "No canonical PR: the PRs add different model integrations with different files, configs, tests, and code paths.", "best_issue_reason": null, - "best_pr_reason": "41797 is the better representative PR for the cluster because it is the earlier, more reviewed, and more complete model addition, though it should not be merged with 45186.", + "best_pr_reason": "PR #45186 is the better representative only because it is the more complete/newer integration, but it is still not a duplicate of #41797.", "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "They add different models (DeepSeek OCR vs Isaac) with different configurations, model files, processors, tests, and docs. Shared auto-registration/docs scaffolding is not enough to treat them as the same underlying change." + "reason": "Different model implementations and integration surfaces; they share generic boilerplate paths but do not fix the same concrete problem or change the same code path." } ] }, @@ -15697,13 +15684,13 @@ { "label": "pull_request", "component_index": 6, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-42166-2", "nodes": [ @@ -15713,7 +15700,7 @@ "soft_pairs": [ "pull_request:42166|pull_request:45426" ], - "prepared_review_unit_hash": "24d82348f0c461f8db8b70c3f527755390623f2ceb2ac7e9425e7d911981ef5c", + "prepared_review_unit_hash": "33bc07fa001bd8a3da2c492ed62e4470777347a05a993c0d468d91f455b88e2a", "component_budget": { "node_count": 2, "item_count": 2, @@ -15735,18 +15722,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate model-integration PRs: one adds InternVLFlash, the other adds AXK1. They overlap only on shared Auto/registry plumbing, not on the same underlying bug or feature.", - "confidence": 0.97, + "summary": "These are two unrelated model-addition PRs that only overlap on shared Transformers registration/auto-mapping files. They do not fix the same bug or implement the same change, so they should not be clustered as duplicates.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR for deduping here: PR 42166 introduces InternVLFlash, while PR 45426 introduces AXK1, so they are different feature additions sharing some common registration files.", + "canonical_pr_reason": "No true canonical PR here: PR #42166 adds InternVLFlash, while PR #45426 adds AXK1. They are separate model integrations and cannot be merged into one duplicate change.", "best_issue_reason": null, - "best_pr_reason": "If a representative PR is needed, PR 42166 is the more complete integration (docs + tests + model code), but it is still not a duplicate of PR 45426.", + "best_pr_reason": "If one PR must represent the cluster, #42166 is the better candidate because it is a fuller, more mature standalone model integration with docs and tests. It is still not a duplicate of #45426.", "soft_edge_verdicts": [ { - "left": "add internvl_flash model", - "right": "Feature/add axk1", + "left": "pull_request:42166", + "right": "pull_request:45426", "accept": false, - "reason": "Different model additions with different codepaths; shared auto-registry files are incidental and not enough to treat them as the same change." + "reason": "Reject: they add different models (InternVLFlash vs AXK1) and only share generic registration/auto-file edits. This is not the same underlying feature or code-path fix." } ] }, @@ -15760,113 +15747,13 @@ { "label": "pull_request", "component_index": 7, - "component_count": 25, - "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-42227-4", - "nodes": [ - "pull_request:42227", - "pull_request:43291", - "pull_request:43665", - "pull_request:44815" - ], - "soft_pairs": [ - "pull_request:42227|pull_request:43291", - "pull_request:42227|pull_request:43665", - "pull_request:42227|pull_request:44815", - "pull_request:43291|pull_request:43665", - "pull_request:43291|pull_request:44815", - "pull_request:43665|pull_request:44815" - ], - "prepared_review_unit_hash": "58545cda1a2c00e4fdb3beb8d023ee40979a0c91a817d33b6ad75ffb04275f27", - "component_budget": { - "node_count": 4, - "item_count": 4, - "soft_pair_count": 6, - "serialized_chars": 10170, - "estimated_input_tokens": 2543, - "estimated_eval_tokens": 5342 - }, - "budget": { - "node_count": 4, - "item_count": 4, - "soft_pair_count": 6, - "serialized_chars": 10170, - "estimated_input_tokens": 2543, - "estimated_eval_tokens": 5342 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": false, - "analyst_result": { - "summary": "These PRs are not duplicates of one another. 42227 is a focused Whisper pipeline bug fix, 43291 is a Whisper tokenizer/test change, 43665 is unrelated test-file churn, and 44815 is a broad FP8/dequantization fix.", - "confidence": 0.97, - "canonical_issue_reason": null, - "canonical_pr_reason": "42227 is the clearest representative PR in this set: it has a specific Whisper bug, a focused code change, and an explicit issue target. The others are either test-only, unrelated, or much broader in scope.", - "best_issue_reason": null, - "best_pr_reason": "42227 is the best standalone PR because it addresses a concrete user-facing bug in the ASR pipeline and has the strongest traceability to a reported issue.", - "soft_edge_verdicts": [ - { - "left": "pull_request:42227", - "right": "pull_request:43291", - "accept": false, - "reason": "Both touch Whisper, but they fix different code paths: 42227 passes return_language through the ASR pipeline, while 43291 changes tokenizer decode behavior/tests. Not the same underlying bug." - }, - { - "left": "pull_request:42227", - "right": "pull_request:43665", - "accept": false, - "reason": "Unrelated changes. 42227 is a Whisper pipeline fix; 43665 only adds bogus test assertions in CLIP/ViT test files and does not address the same bug." - }, - { - "left": "pull_request:42227", - "right": "pull_request:44815", - "accept": false, - "reason": "Completely different area and purpose: Whisper ASR return-language handling versus FP8/dequant/model-loading changes." - }, - { - "left": "pull_request:43291", - "right": "pull_request:43665", - "accept": false, - "reason": "43291 is a Whisper tokenizer fix, while 43665 is unrelated test churn in CLIP and ViT. No shared concrete change or bug." - }, - { - "left": "pull_request:43291", - "right": "pull_request:44815", - "accept": false, - "reason": "Different subsystems and problems: Whisper decoding versus FP8/dequantization and model loading." - }, - { - "left": "pull_request:43665", - "right": "pull_request:44815", - "accept": false, - "reason": "No overlap in underlying bug or change; 43665 is test-only noise, while 44815 is a substantive quantization/dequant 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": 8, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-42975-2", "nodes": [ @@ -15876,12 +15763,12 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "00cce191563cf78862acd7b7f129232fa445b495cedb4ec580326a457d948c6f", + "prepared_review_unit_hash": "0324948a0ea0e48eb4334c7015ca2ae64844c60e4ee6e2d5a4c346f30aac5c7d", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5935, + "serialized_chars": 5934, "estimated_input_tokens": 1484, "estimated_eval_tokens": 3224 }, @@ -15889,7 +15776,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5935, + "serialized_chars": 5934, "estimated_input_tokens": 1484, "estimated_eval_tokens": 3224 }, @@ -15898,18 +15785,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are broad GitHub Actions maintenance changes, but they address different version bumps and different goals: 42975 is specifically about Node 24 compatibility for workflow actions, while 42976 is a separate upgrade of several actions including docker/build-push-action. The overlap is superficial rather than the same underlying change.", - "confidence": 0.94, + "summary": "Both items are broad GitHub Actions dependency bump PRs, but they target different action families and motivations: one is specifically for Node 24 compatibility, the other upgrades docker/build-push-action and related workflow actions. They overlap in workflow files, but they are not the same concrete change.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the better canonical representative because it has the clearest, more specific intent (Node 24 compatibility) and its diff is centered on that workflow-action compatibility update.", + "canonical_pr_reason": "PR 42975 is the more specific, clearly scoped change: it explicitly targets Node 24 compatibility and updates the core workflow actions that are most likely to break there.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the best representative of the cluster\u2019s theme because its title and diff clearly identify the maintenance goal, whereas 42976 is a broader/latest-versions bump with a different concrete scope.", + "best_pr_reason": "PR 42975 is the best representative of the cluster because its title and diff align around a single maintenance goal across multiple workflows.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Reject: these are related maintenance PRs, but they do not fix the same concrete bug or implement the same change. 42975 updates actions for Node 24 compatibility; 42976 bumps different actions, notably docker/build-push-action, to newer versions. Shared workflow files alone is not enough to treat them as duplicates." + "reason": "Do not merge: these are related workflow dependency upgrades, but they fix different problems and update different actions. One is a Node 24 compatibility pass; the other is a docker/build-push-action version bump plus other latest-action updates. They are not the same underlying bug/change." } ] }, @@ -15922,14 +15809,14 @@ }, { "label": "pull_request", - "component_index": 9, - "component_count": 25, + "component_index": 8, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-42978-9", "nodes": [ @@ -15957,12 +15844,12 @@ "pull_request:45490|pull_request:45630", "pull_request:45626|pull_request:45630" ], - "prepared_review_unit_hash": "7849dfa95f3843516ea5bd439f458f66fef73f85344bca5d78ab6e08847f78d8", + "prepared_review_unit_hash": "726f37b33213407ece90f5cf7072b0283ecae6f8b8dbe12835b99103e5847f49", "component_budget": { "node_count": 9, "item_count": 9, "soft_pair_count": 12, - "serialized_chars": 29349, + "serialized_chars": 29352, "estimated_input_tokens": 7338, "estimated_eval_tokens": 14932 }, @@ -15970,7 +15857,7 @@ "node_count": 9, "item_count": 9, "soft_pair_count": 12, - "serialized_chars": 29349, + "serialized_chars": 29352, "estimated_input_tokens": 7338, "estimated_eval_tokens": 14932 }, @@ -15979,84 +15866,162 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a set of unrelated model-addition and infrastructure PRs that only resemble each other because they follow the same Transformers template. None of the soft-similarity pairs look like the same underlying change.", - "confidence": 0.96, + "summary": "This cluster is mostly a mix of unrelated model-addition PRs and two unrelated infrastructure/support changes. They share broad patterns like docs/Auto registration, but they do not describe the same underlying bug or change, so no duplicate merges are justified.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single PR is a valid canonical duplicate target: the PRs cover different models (ViT NEPA, Molmo, Molmo2, LFM2 Audio, CTSM, PP-FormulaNet, Kimi2_6) plus unrelated PP/GGUF support work.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "No global best PR for deduplication: the cluster is heterogeneous, so any one PR would be a poor representative of the rest.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model integrations: ViT NEPA vs Molmo. Shared docs/auto-registration files are template overlap, not the same fix." + "reason": "Different model families (ViT NEPA vs Molmo); similar scaffolding files are just model onboarding boilerplate." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "ViT NEPA and Molmo2 are distinct model additions with different codepaths and files; similarity is only from model onboarding boilerplate." + "reason": "ViT NEPA and Molmo2 are distinct additions with different code paths and target models." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "ViT NEPA vs CTSM are separate model PRs; shared auto/doc files do not indicate one underlying bug or change." + "reason": "ViT NEPA and CTSM are unrelated model additions; shared docs/auto files are not enough." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo and Molmo2 are related names but still separate model implementations, docs, and tests; not mergeable as one concrete change." + "reason": "Molmo and Molmo2 are separate model integrations, not the same concrete change." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo vs CTSM are unrelated model additions; overlap is only in standard registration files." + "reason": "Molmo and CTSM are different model additions with no shared underlying bug or feature." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM touch different model packages and have different purposes; no shared bug or code-path fix." + "reason": "Molmo2 and CTSM are unrelated model support PRs." }, { "left": "pull_request:43973", "right": "pull_request:45283", "accept": false, - "reason": "Lfm2.5 audio model support and Qwen3.5 GGUF loading are different features in different subsystems." + "reason": "LFM2 Audio model support and Qwen3.5 GGUF loading support address different code paths and problems." }, { "left": "pull_request:44601", "right": "pull_request:45626", "accept": false, - "reason": "Distributed pipeline-parallel loading support is infrastructure; PP-FormulaNet is a new model PR. They do not fix the same problem." + "reason": "Native pipeline-parallel loading support is infrastructure work; PP-FormulaNet is a specific model addition." }, { "left": "pull_request:45283", "right": "pull_request:45626", "accept": false, - "reason": "Qwen3.5 GGUF loading support and PP-FormulaNet model support are unrelated changes with no shared concrete code-path issue." + "reason": "GGUF loading support for Qwen3.5 is unrelated to adding PP-FormulaNet model support." }, { "left": "pull_request:45490", "right": "pull_request:45626", "accept": false, - "reason": "CTSM model addition and PP-FormulaNet model addition are distinct model PRs; shared auto/doc churn is not enough to treat them as duplicates." + "reason": "CTSM model addition and PP-FormulaNet model support are separate models, not duplicate changes." }, { "left": "pull_request:45490", "right": "pull_request:45630", "accept": false, - "reason": "CTSM and Kimi2-6 are separate model onboarding PRs; the overlap is generic template code, not the same change." + "reason": "CTSM and Kimi2-6 are different model integrations." }, { "left": "pull_request:45626", "right": "pull_request:45630", "accept": false, - "reason": "PP-FormulaNet and Kimi2-6 are different model additions with different architecture and file sets; not the same underlying change." + "reason": "PP-FormulaNet and Kimi2-6 are distinct model additions with no shared concrete bug or 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": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-3", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43665|pull_request:44815" + ], + "prepared_review_unit_hash": "8e12ce4b9fbaf83b0fa661523c693b52dbd743f9c6a2272e3310f512f194fe5b", + "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 pull requests appear unrelated despite generic titles and soft similarity signals: one fixes Whisper decoding/tests, one injects placeholder test failures in CLIP/ViT, and one is a substantive FP8 dequantization fix.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44815 is the strongest canonical candidate because it has the clearest, most substantive bug-fix scope and the most descriptive title; the other PRs are either narrowly Whisper test/code adjustments or low-information test-only changes.", + "best_issue_reason": null, + "best_pr_reason": "PR 44815 is the best representative of the cluster only in terms of being the most concrete and actionable change set. It is still not actually related to the other PRs, so this is a weak global representative choice.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Whisper tokenizer/timestamp decoding changes are unrelated to CLIP/ViT test placeholder assertions; they do not fix the same code-path problem." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decode logic and FP8 dequant/model-loading work address different subsystems and different bugs." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "A test-only placeholder-failure PR is not the same underlying change as a core FP8 dequantization fix." } ] }, @@ -16070,13 +16035,13 @@ { "label": "pull_request", "component_index": 10, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43488-3", "nodes": [ @@ -16089,7 +16054,7 @@ "pull_request:43488|pull_request:45476", "pull_request:43532|pull_request:45476" ], - "prepared_review_unit_hash": "a4ab09b31b96c3653260eda26715a6a76b4d0ef4a44ca92d77d7eeca4f5d8c66", + "prepared_review_unit_hash": "73001e92f9cb0c988f758824a176d8899c21cacb25674f3bbcf0d3319c06aa52", "component_budget": { "node_count": 3, "item_count": 3, @@ -16111,30 +16076,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three open PRs are not duplicates of the same change; they are separate test/dummy edits touching different workflows and files. The shared Vit test file is incidental and does not indicate a common underlying bug or fix.", - "confidence": 0.97, + "summary": "These are three unrelated open \"[don't merge]\" PRs that only overlap incidentally on a shared vit test file. One is a setup/version check, one is CI/workflow plumbing, and one is a CI-call workflow plus CLIP/chinese-CLIP test/tokenizer edits. They do not look like the same underlying bug or change.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "No clear canonical PR: the three PRs address unrelated areas (packaging/versioning, self-comment CI, and PR CI caller wiring) and are all artificial '[don't merge]' changes rather than one shared fix.", + "canonical_pr_reason": "No canonical PR: the overlapping filenames are superficial and the PRs target different code paths and purposes.", "best_issue_reason": null, - "best_pr_reason": "If a representative PR must be chosen, #43532 is the broadest and most structurally coherent change, but it is still not a superset of the others and should not be treated as the same fix.", + "best_pr_reason": "If one must be picked, #43532 is the broadest/most central around the shared test-file overlap, but it still isn't a true duplicate representative because its main changes are CI/workflow-related.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "Different dummy changes: #43488 tweaks setup/dependency/version metadata, while #43532 rewires self-comment CI and test utilities. Shared Vit test filename is incidental." + "reason": "Shared vit test filename is incidental; #43488 is a setup/version/dependency check, while #43532 is workflow and test-infra plumbing." }, { "left": "pull_request:43488", "right": "pull_request:45476", "accept": false, - "reason": "#43488 is packaging/version churn; #45476 adds a PR CI caller workflow and unrelated placeholder code in CLIP/chinese-CLIP files. Not the same underlying bug or change." + "reason": "These edit different areas: #43488 is repo packaging/versioning, while #45476 adds a CI caller workflow and unrelated CLIP/chinese-CLIP test/tokenizer changes." }, { "left": "pull_request:43532", "right": "pull_request:45476", "accept": false, - "reason": "Both are artificial '[don't merge]' PRs, but they modify different CI paths and unrelated files. They would not plausibly be merged as one PR." + "reason": "Both mention test files, but the concrete changes are different: #43532 focuses on self-comment CI/slack/testing utils, while #45476 focuses on calling an external CI workflow and CLIP/chinese-CLIP code." } ] }, @@ -16148,13 +16113,13 @@ { "label": "pull_request", "component_index": 11, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43612-2", "nodes": [ @@ -16164,7 +16129,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "c9b12ea72345d7c47de5b53c0e9b0077ef10d01716f7c6d7e93491cecc99de10", + "prepared_review_unit_hash": "7dc1daf25738868a8fe46504d4f4e8366ac6bb7fb6190ddba39bbbe00de74477", "component_budget": { "node_count": 2, "item_count": 2, @@ -16186,18 +16151,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two separate open PRs add different pipelines: Promptable Concept Segmentation and Promptable Visual Segmentation. They overlap in registration/docs scaffolding, but they target different tasks, different pipeline modules, and different model/test surfaces, so they are not duplicates.", - "confidence": 0.96, + "summary": "These are two separate pipeline-addition PRs for different segmentation features: promptable concept segmentation vs promptable visual segmentation. They share some infra and docs plumbing, but they implement different pipelines, model integrations, and tests, so they should not be deduplicated.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: these are adjacent but distinct feature additions, each implementing a different pipeline and model integration path.", + "canonical_pr_reason": "PR 43613 is the better standalone representative because it is the broader promptable visual segmentation addition with more model and test coverage; however, it is not a duplicate of 43612.", "best_issue_reason": null, - "best_pr_reason": "PR 43613 is the broader and more complete feature set, but it still does not subsume 43612 because it implements a different segmentation pipeline.", + "best_pr_reason": "PR 43613 is the stronger overall PR in this cluster: it adds the visual segmentation pipeline with wider integration coverage, while 43612 targets a different concept-segmentation feature.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different underlying change: concept segmentation vs visual segmentation. They share some boilerplate files, but the pipeline names, task docs, model integrations, and tests point to separate features, not one mergeable PR." + "reason": "Different concrete features and code paths: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. Shared docs/registration files are just framework plumbing, not evidence of the same underlying change." } ] }, @@ -16211,13 +16176,13 @@ { "label": "pull_request", "component_index": 12, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", - "reason": null, - "source": "cache", - "cache_hit": true, + "reason": null, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43651-2", "nodes": [ @@ -16227,40 +16192,40 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "a9a9806e32e4c7866b48300debebd0a249261b2f4356537f609bdde6bf6081f8", + "prepared_review_unit_hash": "db66e871d06d38205487959f85958aadd9e691d08b82983eb00e10a4fcd4b144", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4293, - "estimated_input_tokens": 1074, - "estimated_eval_tokens": 2404 + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4293, - "estimated_input_tokens": 1074, - "estimated_eval_tokens": 2404 + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open Trainer PRs that both add small extensibility hooks, but they address different code paths: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for PEFT/custom trainers. They are related in theme but not the same change.", - "confidence": 0.95, + "summary": "Two open PRs in the same file, but they address different trainer extensibility points: one changes gradient-accumulation loss scaling, the other adds an overridable signature-column helper for column filtering.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue exists in the cluster; if one PR must stand in as the representative, 43663 is slightly more canonical because it introduces a reusable Trainer helper that centralizes an existing code path, whereas 43651 is a narrower behavioral tweak.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "43663 is the better representative PR because it is a cleaner internal refactor with a clear extension point and broader applicability within Trainer. 43651 is valid but more specialized and behavior-specific.", + "best_pr_reason": "PR 43651 is the more self-contained bug-fix style change, with a concrete behavior correction in training loss scaling; PR 43663 is a separate refactor/extensibility change around signature inspection.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Reject: these PRs change different mechanisms in Trainer. 43651 adjusts loss normalization during gradient accumulation; 43663 extracts signature-column logic for model input filtering. Same file and similar extensibility theme, but not the same underlying bug or mergeable as one fix." + "reason": "Different code paths and user-facing behaviors: 43651 adjusts gradient accumulation loss normalization, while 43663 adds a helper for inspecting model forward signatures to filter columns. They share a file but not the same underlying bug or mergeable fix." } ] }, @@ -16274,13 +16239,13 @@ { "label": "pull_request", "component_index": 13, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43656-3", "nodes": [ @@ -16292,55 +16257,55 @@ "pull_request:43656|pull_request:43842", "pull_request:43836|pull_request:43842" ], - "prepared_review_unit_hash": "7b790654607e5ff8383e0fa7d6220c2a30e32782182f2e872831003575c68c20", + "prepared_review_unit_hash": "2093d26d368c88c30ffe68fc98577b7f0283380e453bfb419208877ddde508a8", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 5993, - "estimated_input_tokens": 1499, - "estimated_eval_tokens": 3254 + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 }, "budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 5993, - "estimated_input_tokens": 1499, - "estimated_eval_tokens": 3254 + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "PR 43842 is the clearest, narrowest fix for the `TypeAdapter` NameError in `src/transformers/cli/serve.py`. PR 43836 looks like a broader variant of the same bugfix, but PR 43656 is too weakly supported to merge into this cluster because it targets a different issue and appears to be a more general postponed-annotations change.", - "confidence": 0.72, + "summary": "All three PRs are trying to fix the same CLI `TypeAdapter` annotation/runtime `NameError` in `src/transformers/cli/serve.py`. PR 43842 is the cleanest targeted fix; the others look like overlapping variants of the same bug rather than distinct changes.", + "confidence": 0.86, "canonical_issue_reason": null, - "canonical_pr_reason": "43842 is the best canonical PR because it directly and minimally fixes the runtime `TypeAdapter` NameError, with a small diff and a clear issue target.", + "canonical_pr_reason": "PR 43842 is the most focused and explicit fix: it addresses the `TypeAdapter` NameError directly, scopes the import to `TYPE_CHECKING`, and avoids broader unrelated annotation churn.", "best_issue_reason": null, - "best_pr_reason": "43842 is the strongest representative of the cluster: it addresses the concrete failure directly, with the least extra typing/refactor churn.", + "best_pr_reason": "PR 43842 best represents the cluster because it is the narrowest, clearest change for the exact runtime bug described by the title and diff.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", - "accept": false, - "reason": "This is not strong enough to treat as the same fix: 43656 targets issue 43576 and makes a broader postponed-annotations change, while 43842 is a focused NameError fix. Shared file/symptom alone is insufficient." + "accept": true, + "reason": "Both modify the same file to fix the same `TypeAdapter` annotation NameError in the CLI serve path; they are alternative implementations of the same bug fix." }, { "left": "pull_request:43836", "right": "pull_request:43842", "accept": true, - "reason": "Both PRs address the same `TypeAdapter` NameError in `src/transformers/cli/serve.py`, and 43836 appears to be a broader implementation of the same underlying bugfix. They look plausibly mergeable into one PR." + "reason": "Both target the same `TypeAdapter` NameError in `serve.py` and appear to be overlapping fixes on the same code path, despite 43836 being broader/more refactor-heavy." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The summary is a bit too strong for the packet. PR 43842 is clearly the most focused TypeAdapter fix, but the claim that 43656 is the same TypeAdapter NameError is not well supported by the provided diff preview: 43656 shows a broader postponed-annotations change and targets a different issue (43576), with only soft similarity evidence against 43842. That is not enough to confidently merge it into the same bugfix cluster. The 43836/43842 pairing is more plausible, but the overall cluster summary should stay more conservative." + "accept": true, + "feedback": "The summary is grounded in the packet: all three PRs touch `src/transformers/cli/serve.py` and their titles/diffs point to the same `TypeAdapter`/annotation runtime-name issue. The soft-edge pairings are conservative enough, since each pair appears to be an overlapping implementation of the same bugfix rather than a distinct change." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -16348,7 +16313,7 @@ { "label": "pull_request", "component_index": 14, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -16414,13 +16379,13 @@ { "label": "pull_request", "component_index": 15, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 2, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43996-28", "nodes": [ @@ -16498,7 +16463,7 @@ "pull_request:44056|pull_request:44101", "pull_request:44073|pull_request:44161", "pull_request:44101|pull_request:44161", - "pull_request:43997|pull_request:44017", + "pull_request:44017|pull_request:44018", "pull_request:43997|pull_request:44056", "pull_request:44071|pull_request:44076", "pull_request:44074|pull_request:44101", @@ -16527,340 +16492,424 @@ "pull_request:44044|pull_request:44154", "pull_request:44024|pull_request:44076" ], - "prepared_review_unit_hash": "1b3e5d56a718451a482744ef4b9c7675dab838e890ca2e6baf3ea925226498ee", + "prepared_review_unit_hash": "f5a10d022afa46230b04913144ca64337d835aede37bc8383a6be6b119d241bb", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84498, - "estimated_input_tokens": 21125, - "estimated_eval_tokens": 42506 + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 }, "budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 72, - "serialized_chars": 76931, - "estimated_input_tokens": 19233, - "estimated_eval_tokens": 38722 + "serialized_chars": 76934, + "estimated_input_tokens": 19234, + "estimated_eval_tokens": 38724 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of model-by-model output-tracing refactors for the same framework-level initiative, but most PRs are not duplicates of each other because they touch different model code paths. The only clear duplicate subgroups are the GPT-J/CodeGen variants and the GPT-Neo variants.", - "confidence": 0.92, + "summary": "This cluster is mostly a batch of model-specific output-tracing refactors under the same umbrella theme, but they are not duplicates across architectures. The only clear duplicates are the GPT-J and GPT-Neo follow-up PRs that rewrite the same code paths.", + "confidence": 0.84, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44722 is the best canonical anchor for the GPT-J/CodeGen duplicate subgroup: it targets the same concrete code paths as the earlier GPT-J variants, has the shared output-tracing refactor intent, and looks like the later consolidated version of that change.", + "canonical_pr_reason": "PR 44722 is the strongest representative: it targets the core GPT-J/CodeGen output-tracing refactor, overlaps directly with earlier GPT-J PRs, and looks like the most complete version of that change.", "best_issue_reason": null, - "best_pr_reason": "PR 44722 is the strongest single representative of the same underlying change because it is focused, explicit about the output-tracing refactor, and most plausibly subsumes the earlier GPT-J-specific iterations.", + "best_pr_reason": "PR 44722 best captures the underlying duplicate set for the GPT-J/CodeGen output-tracing rewrite; the rest are mostly separate model ports, not one shared fix.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Both are output-tracing refactors, but they apply to different model families (CVT/FNet vs GPT-J/CodeGen), so they are not the same underlying change." + "reason": "Different model families (CVT/FNet vs GPT-J); only the refactor theme and issue target match." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Different model code paths (CVT/FNet vs DeBERTa-v2); same initiative, but not a duplicate PR." + "reason": "CVT/FNet vs DeBERTa are unrelated code paths; same tracking umbrella is not enough." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Both target GPT-J output tracing in the same model file and appear to be the same concrete refactor path, with only minor iteration differences." + "reason": "Same GPT-J output-tracing refactor on the same file; 44085 reads like a narrow follow-up of the same change." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "These are separate model implementations (ResNet/RT-DETR ResNet vs EfficientNet), so they are not the same change." + "reason": "ResNet/RegNet/RT-DETR vs EfficientNet are different model implementations; not the same bug/change." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet and GPT-J/CodeGen are unrelated code paths; both are standardized-output refactors, but not duplicates." + "reason": "EfficientNet and GPT-J are unrelated models; same refactor pattern only." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen and EfficientNet touch different model implementations, so this is only thematic similarity." + "reason": "GPT-J/CodeGen vs EfficientNet are different code paths." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTa-v2 are different models and fix different forward/output plumbing." + "reason": "MobileNetV2 vs DeBERTa are unrelated model families." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen vs MGP-STR: same refactor pattern, different underlying code paths." + "reason": "GPT-J/CodeGen and MGP-STR are different models; not mergeable as one fix." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen and MPT are distinct model implementations; do not merge based on shared decorator work." + "reason": "GPT-J/CodeGen vs MPT are separate refactors, not the same concrete change." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Both are GPT-Neo output-tracing refactors in the same model file, and the diff previews point to the same underlying code-path migration." + "reason": "Both are GPT-Neo output-tracing refactors on the same model/file; this looks like the same underlying change revised later." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen and GPT-Neo are separate model families; same migration theme, not the same bug/change." + "reason": "GPT-J/CodeGen vs GPT-Neo are different code paths despite similar wording." }, { "left": "pull_request:44018", - "right": "pull_request:44044", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo vs GPT-J/CodeGen are separate model-specific refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and DeBERTa-v2 are unrelated implementations, so these are not duplicates." + "reason": "ResNet is unrelated to GPT-J; they only share the standardized-tracing motif." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J/CodeGen are unrelated model code paths." + "reason": "MGP-STR and GPT-J are unrelated implementations." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet and GPT-J/CodeGen are different model refactors; same broad initiative only." + "reason": "ResNet vs GPT-J are different model families." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo vs ResNet: different model families and different forward implementations." + "reason": "GPT-Neo and ResNet are unrelated changes." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet and MPT are unrelated code paths." + "reason": "ResNet vs MPT are separate model-specific refactors." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are different model implementations; not the same change." + "reason": "GPT-Neo vs Flaubert are unrelated models." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are distinct refactors despite similar output-tracing patterns." + "reason": "GPT-Neo vs GPT-J are different code paths." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT and GPT-J/CodeGen are separate model-specific changes." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44116", - "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are different models with different output-handling code." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-Neo and MGP-STR are unrelated implementations." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44085", - "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are different model families." - }, - { - "left": "pull_request:44068", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-Neo and MGP-STR are not the same underlying fix." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44086", - "accept": false, - "reason": "MPT and MGP-STR are distinct code paths." + "reason": "MPT vs GPT-J are unrelated implementations." }, { "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-2 and MGP-STR are separate model refactors." + "reason": "GPT2 and MGP-STR are different models; same refactor theme only." }, { "left": "pull_request:44085", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen and MGP-STR are different model-specific changes." + "reason": "GPT-J and MGP-STR are unrelated model code paths." }, { "left": "pull_request:44059", "right": "pull_request:44722", "accept": false, - "reason": "GPT-2 and GPT-J/CodeGen are unrelated model implementations." + "reason": "GPT2 vs GPT-J are different model families." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert/XLM are different model files and output paths." + "reason": "ViLT vs Flaubert/XLM are unrelated model changes." }, { "left": "pull_request:44017", "right": "pull_request:44722", "accept": false, - "reason": "Segformer and GPT-J/CodeGen are different model families." + "reason": "SegFormer vs GPT-J are different models; only the tracing pattern is shared." }, { "left": "pull_request:44017", "right": "pull_request:44161", "accept": false, - "reason": "Segformer and LongT5 are unrelated code paths." + "reason": "SegFormer vs LongT5 are unrelated refactors." }, { "left": "pull_request:44017", "right": "pull_request:44073", "accept": false, - "reason": "Segformer and VisualBert are different model implementations; only the refactor pattern matches." + "reason": "SegFormer and VisualBERT are separate model-specific changes." }, { "left": "pull_request:44068", "right": "pull_request:44071", "accept": false, - "reason": "GPT-Neo and MPT are distinct model-specific output refactors." + "reason": "GPT-Neo vs MPT are different code paths." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT vs MGP-STR are unrelated models." }, { "left": "pull_request:43997", "right": "pull_request:44073", "accept": false, - "reason": "RegNet and VisualBert are unrelated model code paths." + "reason": "RegNet and VisualBERT are different model families." }, { "left": "pull_request:44056", "right": "pull_request:44101", "accept": false, - "reason": "MPNet and Flaubert/XLM are different model implementations." + "reason": "MPNet vs Flaubert/XLM are unrelated changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44018", + "accept": false, + "reason": "SegFormer vs GPT-Neo are different model implementations." }, { "left": "pull_request:43997", - "right": "pull_request:44017", + "right": "pull_request:44056", "accept": false, - "reason": "RegNet and Segformer are different models; do not treat as duplicates." + "reason": "RegNet vs MPNet are separate code paths." }, { - "left": "pull_request:44056", + "left": "pull_request:43997", "right": "pull_request:44161", "accept": false, - "reason": "MPNet and LongT5 are distinct code paths." + "reason": "RegNet vs LongT5 are unrelated model-specific refactors." }, { - "left": "pull_request:43996", - "right": "pull_request:44161", + "left": "pull_request:44074", + "right": "pull_request:44101", "accept": false, - "reason": "CVT/FNet and LongT5 are unrelated model-specific refactors." + "reason": "TextNet vs Flaubert/XLM are different model families." }, { - "left": "pull_request:43996", - "right": "pull_request:44017", + "left": "pull_request:44074", + "right": "pull_request:44161", "accept": false, - "reason": "CVT/FNet and Segformer are different model families." + "reason": "TextNet vs LongT5 are unrelated changes." }, { "left": "pull_request:44010", "right": "pull_request:44071", "accept": false, - "reason": "SqueezeBert and MPT are unrelated code paths." + "reason": "SqueezeBERT vs MPT are different model paths." }, { "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "SqueezeBert and VisualBert are different model implementations." + "reason": "SqueezeBERT vs VisualBERT are unrelated refactors." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT vs DeBERTa are different code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa vs TextNet are unrelated model-specific changes." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet vs ImageGPT are different models; same standardized-output pattern only." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT2 vs GPT-Neo are separate refactors." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT2 vs MPT are unrelated code paths." }, { "left": "pull_request:44018", "right": "pull_request:44129", "accept": false, - "reason": "GPT-Neo and SpeechT5 are separate model refactors." + "reason": "GPT-Neo vs SpeechT5 are different model families." }, { - "left": "pull_request:43997", - "right": "pull_request:44056", + "left": "pull_request:44018", + "right": "pull_request:44086", "accept": false, - "reason": "RegNet and MPNet are unrelated model code paths." + "reason": "GPT-Neo vs MGP-STR are unrelated changes." }, { - "left": "pull_request:43997", + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo vs GPT-J are different code paths." + }, + { + "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "RegNet and LongT5 are different models and different forward paths." + "reason": "MPNet vs LongT5 are unrelated models." }, { - "left": "pull_request:44074", - "right": "pull_request:44101", + "left": "pull_request:44056", + "right": "pull_request:44017", "accept": false, - "reason": "TextNet and Flaubert/XLM are unrelated implementations." + "reason": "MPNet vs SegFormer are different code paths." }, { - "left": "pull_request:43997", - "right": "pull_request:44074", + "left": "pull_request:44010", + "right": "pull_request:44018", "accept": false, - "reason": "RegNet and TextNet are different model families." + "reason": "SqueezeBERT vs GPT-Neo are unrelated model-specific refactors." }, { "left": "pull_request:44071", "right": "pull_request:44129", "accept": false, - "reason": "MPT and SpeechT5 are distinct code paths." + "reason": "MPT vs SpeechT5 are different model families." }, { - "left": "pull_request:44017", - "right": "pull_request:44074", + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet vs LongT5 are unrelated changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", "accept": false, - "reason": "Segformer and TextNet are different model-specific changes." + "reason": "CVT/FNet vs SegFormer are different model code paths." }, { "left": "pull_request:44010", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBert and DeBERTa-v2 are unrelated model implementations." + "reason": "SqueezeBERT vs DeBERTa are unrelated model refactors." }, { "left": "pull_request:44044", - "right": "pull_request:44074", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa vs SpeechT5 are different code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing change set; 44722 is the broader later version of the same refactor." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both touch GPT-J output tracing on the same file; they look like the same fix in different iterations." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated models." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "Different model families; only the output-tracing theme overlaps." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet vs SpeechT5 are unrelated changes." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "ResNet/RegNet/RT-DETR vs EfficientNet are different model-specific refactors." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44072", + "accept": false, + "reason": "MobileNetV2 vs EfficientNet are separate model implementations." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa vs SpeechT5 are unrelated code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa-v2 and TextNet are different code paths." + "reason": "DeBERTa vs VITS are different models." }, { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "FocalNet and ImageGPT are different models; similar standardized output handling only." + "reason": "FocalNet vs ImageGPT are unrelated model refactors." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is conservative and grounded in the packet. The accepted duplicate pairs are supported by overlapping GPT-J/CodeGen and GPT-Neo file changes, while the false verdicts appropriately separate different model families/code paths. The canonical PR choice for the GPT-J/CodeGen subgroup is also reasonable given the later consolidated diff." + "feedback": "Grounded overall. The summary stays conservative by treating cross-model output-tracing rewrites as theme similarity rather than duplicates, and the accepted soft pairs are the ones with clear shared code paths/issue targets (GPT-J/CodeGen and GPT-Neo iterations). No obvious overstatement of evidence." }, "evaluator_used": true, "retried": false, @@ -16871,13 +16920,13 @@ { "label": "pull_request", "component_index": 15, - "component_count": 25, + "component_count": 24, "review_unit_index": 2, "review_unit_count": 2, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43996-28", "nodes": [ @@ -16895,14 +16944,14 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "1b0eb1546754af6fb76c3546a0bdac8ace894306fe83f2ffc8e2d8cbb4e9727b", + "prepared_review_unit_hash": "09dffba8a9af89e6dab017f2d24cf82f66a75633cc9236c1ce1ebaa2844b6719", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84498, - "estimated_input_tokens": 21125, - "estimated_eval_tokens": 42506 + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 }, "budget": { "node_count": 6, @@ -16917,42 +16966,42 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they touch different architectures and code paths. They are related by technique, not by the same underlying bug or mergeable change, so the soft links should be rejected.", - "confidence": 0.89, + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they target different architectures and code paths. They look thematically similar, not duplicate changes to the same underlying bug or implementation.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "44129 is the strongest representative of the cluster because it has an explicit issue target and a concrete SpeechT5 layer change, making it the best anchor for this refactor theme.", + "canonical_pr_reason": "If one representative PR is needed, 44073 is the clearest example of the pattern: it adds the standardized output-capture plumbing directly in the model class and shows the intended refactor most explicitly.", "best_issue_reason": null, - "best_pr_reason": "44129 is the best standalone PR in the set: it has the clearest scope, an explicit linked issue, and a nontrivial model-specific implementation rather than just a generic refactor.", + "best_pr_reason": "44073 is the best representative because it is a straightforward, explicit migration to the new output-capture mechanism in a single model file, making the cluster\u2019s theme easiest to understand.", "soft_edge_verdicts": [ { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "Both are output-tracing refactors, but one is for VisualBert and the other for TextNet. Different model implementations and code paths, so not the same underlying change." + "reason": "Both are output-tracing refactors, but they affect different models (VisualBert vs TextNet) and different forward/output implementations, so they are not the same change." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "MPNet and VisualBert are unrelated model files; the shared theme is only the standardized tracing refactor pattern, not a single bug or mergeable code change." + "reason": "MPNet and VisualBert are separate model implementations with distinct output-recording plumbing; this is a shared refactor theme, not a duplicate fix." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "These are separate model migrations (MPNet vs TextNet). They may use similar decorator plumbing, but they do not fix the same concrete code path." + "reason": "These touch different model families and different tracing internals, so they do not appear to be one mergeable code change." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert are distinct architectures with independent output handling; similarity in refactor style is insufficient for deduplication." + "reason": "RegNet and SqueezeBert are unrelated model code paths; both migrate output handling, but they do not fix the same concrete bug or implementation." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 changes are in different model implementations and are not plausibly one combined PR for the same bug/change." + "reason": "SqueezeBert output tracing and SpeechT5 output tracing are separate model-specific refactors; SpeechT5 also has cache/attention behavior that makes it a different change." } ] }, @@ -16966,13 +17015,13 @@ { "label": "pull_request", "component_index": 16, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-43998-13", "nodes": [ @@ -17035,292 +17084,298 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "b3aca3b33fc546ba1c85e8cb116a7e6251bd04049022e033ce5adda412fccde9", + "prepared_review_unit_hash": "9d5612e997f54af5d1cc2cda75ab0e697129ebd8d2620efc5205add3ddf4c7fe", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41325, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41319, + "estimated_input_tokens": 10330, + "estimated_eval_tokens": 20916 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 43, - "serialized_chars": 41325, - "estimated_input_tokens": 10332, - "estimated_eval_tokens": 20920 + "serialized_chars": 41319, + "estimated_input_tokens": 10330, + "estimated_eval_tokens": 20916 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are all members of the same broad output-tracing refactor theme, but they target different model implementations and code paths. I did not find any duplicate-worthy pair: they are related stylistically, not the same concrete change.", - "confidence": 0.91, + "summary": "A loose cluster of model-specific output-tracing refactors across unrelated Transformers implementations. The PRs share a naming/theme pattern and helper adoption, but they target different code paths and model families, so they are not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44003 is the most representative single example of the cluster: it applies the output-tracing refactor to two closely related Mamba-family models and includes explicit recorder wiring, making it a clearer exemplar than the smaller one-model cleanups.", + "canonical_pr_reason": "PR 44003 is the strongest representative: it has the largest diff, spans two closely related Mamba-family files, and adds the most explicit output-capturing plumbing. It is still just a representative, not a duplicate of the others.", "best_issue_reason": null, - "best_pr_reason": "PR 44003 is the strongest standalone representative because it is more substantive than the other refactors and shows the full tracing/capture integration across a closely related pair of files.", + "best_pr_reason": "If one PR must stand in for this cluster, 44003 is the most central concrete example of the shared refactor theme, with broader scope than the single-model PRs.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different models and files; same refactor pattern, but not the same underlying bug or change." + "reason": "Different model implementations (`timm_backbone` vs `upernet`); same refactor style, not the same code-path bug/change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Different model code paths; this is only a similar output-tracing cleanup, not a mergeable duplicate." + "reason": "Unrelated architectures (`timm_backbone` vs `vision_text_dual_encoder`); only the output-tracing theme is shared." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Separate architectures with similar API cleanup; not one concrete change." + "reason": "Different model families (`timm_backbone` vs `univnet`); no common concrete fix." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different model families and distinct implementation details; too broad to treat as one PR." + "reason": "Both are refactors, but they touch different Mamba/timm code paths and do not appear to fix the same underlying bug." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Both are tracing refactors, but they touch unrelated models and code paths." + "reason": "Different models (`timm_backbone` vs `rwkv`); shared helper usage is not enough for duplication." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "Different model implementations; same theme only, not the same fix." + "reason": "Different domains (`mobilenet_v1` vision classifier vs `univnet` audio model); not the same change." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Separate model files and distinct forward/output changes." + "reason": "`mobilenet_v1` and `upernet` are separate model paths; this is a parallel refactor, not a duplicate fix." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "Both are output-tracing refactors, but they are for different models and not one underlying change." + "reason": "Different model families (`mobilenet_v1` vs `depth_anything`) and separate forward signatures; not mergeable as one PR." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Different code paths in different model wrappers; not duplicates." + "reason": "`mobilenet_v1` and `vision_encoder_decoder` are unrelated code paths despite similar output-tracing cleanup." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Same refactor style, but the encoder-decoder wrapper is unrelated to MobileNetV1." + "reason": "`mobilenet_v1` vs `speech_encoder_decoder`; same helper refactor pattern, different concrete models." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Unrelated models; similar cleanup pattern is not enough to merge them." + "reason": "Multimodal encoder/decoder model vs audio generation model; different forward logic and outputs." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different subsystems and model implementations; not the same concrete bug/change." + "reason": "`vision_text_dual_encoder` and `upernet` are separate architectures; no shared bug beyond the refactor theme." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and different affected files; not the same concrete code-path problem." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "These touch different model families and different output plumbing." + "reason": "`vision_text_dual_encoder` and `depth_anything` are unrelated models; this is only a similar API cleanup." }, { "left": "pull_request:44000", "right": "pull_request:44026", "accept": false, - "reason": "Same broad tracing refactor, but distinct model wrappers and forward signatures." + "reason": "Both are wrapper-style models, but they are different wrappers with different signatures and targets." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "Separate model code paths; only the refactor pattern is shared." + "reason": "`univnet` and `upernet` have unrelated forward paths; same refactor motif only." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "Different models and different tracing implementations; not mergeable as one PR." + "reason": "Audio generation vs depth estimation; different model code and outputs." }, { "left": "pull_request:44001", "right": "pull_request:44026", "accept": false, - "reason": "Distinct architectures; same modernization theme only." + "reason": "`univnet` and `vision_encoder_decoder` are separate implementations; not one fix split across PRs." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "Not the same underlying code-path problem; just another model adopting the same refactor pattern." + "reason": "`upernet` and `rwkv` are unrelated architectures; same helper adoption does not make them duplicates." }, { "left": "pull_request:44002", "right": "pull_request:44025", "accept": false, - "reason": "Different model families and distinct forward/output semantics." + "reason": "`upernet` segmentation wrapper vs `depth_anything` depth model; distinct code paths and behavior." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Unrelated models; too broad to count as a duplicate." + "reason": "Audio model vs Mamba-family language model; no common concrete bug." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and output plumbing; similar theme only." + "reason": "Segmentation wrapper vs Mamba-family model; different outputs and implementation details." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "These are separate model refactors, not one shared defect or change." + "reason": "Different model families and file sets; only the refactor pattern is shared." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "Different model families and different implementation targets." + "reason": "Mamba-family output tracing vs SuperPoint feature extraction; unrelated concrete changes." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Both involve output-related refactoring, but they are different code paths in different models." + "reason": "`mamba`/`falcon_mamba` refactor vs `codegen` output tracing; distinct model code paths." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "Different model code and different forward signatures; not a duplicate pair." + "reason": "`timm_backbone` and `codegen` are unrelated subsystems; not the same bug or change." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "Same general cleanup theme, but unrelated models and concrete changes." + "reason": "`univnet` audio model vs `codegen` attention/cache path; separate changes." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "Separate model implementations; not the same underlying fix." + "reason": "`upernet` segmentation vs `codegen` transformer internals; no duplicate relationship." }, { "left": "pull_request:44000", "right": "pull_request:44004", "accept": false, - "reason": "Distinct model code paths and refactor details." + "reason": "`vision_text_dual_encoder` and `codegen` have different forward contracts and behavior." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Different model families; cannot be merged as one PR without becoming a multi-model sweep." + "reason": "Both are refactors, but `depth_anything` and `vision_encoder_decoder` are different model families and wrappers." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Related refactor style, but different encoder/decoder wrapper code paths." + "reason": "`depth_anything` vs `speech_encoder_decoder`; separate implementations with different output shapes and targets." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Different models and different output-tracing implementations." + "reason": "Depth estimation model vs SuperPoint; unrelated code paths." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "Separate model files and distinct refactor scopes." + "reason": "`depth_anything` and `rwkv` are unrelated model families; not mergeable as one fix." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Not the same concrete bug/change; just another model adopting the pattern." + "reason": "Depth model vs DPR encoder; different forward behavior and outputs." }, { "left": "pull_request:44026", - "right": "pull_request:44029", + "right": "pull_request:44027", "accept": false, - "reason": "Different model wrappers and code paths; similar tracing refactor only." + "reason": "Encoder-decoder wrappers for different modalities; similar cleanup, not the same change." }, { "left": "pull_request:44026", - "right": "pull_request:44027", + "right": "pull_request:44028", "accept": false, - "reason": "Different architectures and implementation details." + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated models." }, { "left": "pull_request:44026", - "right": "pull_request:44028", + "right": "pull_request:44029", "accept": false, - "reason": "Separate model code paths; not duplicates." + "reason": "Different architectures (`vision_encoder_decoder` vs `rwkv`); no shared concrete bug." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "Different model families and output contracts." + "reason": "Wrapper model vs DPR retriever; distinct code paths and outputs." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "Different model implementations; same theme only." + "reason": "Speech encoder-decoder wrapper vs SuperPoint; unrelated implementations." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Unrelated model code paths despite similar refactoring style." + "reason": "Speech encoder-decoder vs RWKV; different subsystems and fixes." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Distinct models and forward/output changes; not mergeable as one PR." + "reason": "Speech encoder-decoder wrapper vs DPR encoder; not the same bug/change." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "Different subsystems; only the output-tracing pattern is shared." + "reason": "SuperPoint and RWKV are unrelated models; the similarity is only in refactor style." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "Separate model implementations with different concrete changes." + "reason": "Feature-point model vs DPR retrieval model; separate code paths." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "Different models and output paths; similar refactor theme is insufficient for duplication." + "reason": "RWKV and DPR are distinct model families with different output semantics; not duplicates." } ] }, @@ -17334,13 +17389,13 @@ { "label": "pull_request", "component_index": 17, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-44662-3", "nodes": [ @@ -17353,52 +17408,52 @@ "pull_request:44662|pull_request:45075", "pull_request:45075|pull_request:45401" ], - "prepared_review_unit_hash": "0477964a07c6baee8c710646c6adf3639d01bb7fa747f5c0b69efbe896877e62", + "prepared_review_unit_hash": "021776661ded15d64ebd90a932bd2c1753b9bf0a9f1399f63e00a0b5db02f234", "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": 10914, + "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": 10914, + "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 independent model-integration PRs for different architectures (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). They share boilerplate auto-mapping/docs patterns, but not the same underlying bug or change, so none should be clustered as duplicates.", + "summary": "All three pull requests are separate model-addition implementations (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). Their overlap is limited to standard Transformers boilerplate such as docs, auto-mapping, and package registration, so they are not duplicates of the same underlying change.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: each pull request introduces a distinct model implementation and associated docs/tests, not the same fix or feature.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 45075 is the strongest representative only because it is the most mature of the set (highest review activity, non-draft), but it is still a separate model addition rather than a duplicate of the others.", + "best_pr_reason": "No single PR is a better duplicate canonical here; each PR targets a different model and code path, so they should remain separate.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "PenguinVL and Voxtral TTS are unrelated model additions for different code paths and model families; shared auto-registry/docs edits are generic boilerplate, not evidence of the same change." + "reason": "Different models and distinct implementations; shared files are only generic docs/auto-registration boilerplate." }, { "left": "pull_request:44662", "right": "pull_request:45075", "accept": false, - "reason": "PenguinVL vs DeepSeek-OCR-2 are separate model implementations with different files and functionality; overlap is limited to standard Transformers scaffolding." + "reason": "PenguinVL and DeepSeek-OCR-2 are unrelated model additions with no shared concrete bug or change." }, { "left": "pull_request:45075", "right": "pull_request:45401", "accept": false, - "reason": "DeepSeek-OCR-2 and Voxtral TTS target different models and different processing/modeling stacks, so they are not mergeable as one PR for the same underlying fix or feature." + "reason": "Different model integrations (DeepSeek-OCR-2 vs Voxtral TTS); overlap is limited to common Transformers scaffolding." } ] }, @@ -17412,13 +17467,13 @@ { "label": "pull_request", "component_index": 18, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-44979-2", "nodes": [ @@ -17428,12 +17483,12 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "86fc2b03cc1172bd52938ea751667b81abf5cc04940494b34905aefa056ed19b", + "prepared_review_unit_hash": "8fefa35855c3771492ad6b4e9149686ad25b96d32a4760d45850b1a26a332fb4", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4408, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -17441,7 +17496,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4408, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -17450,18 +17505,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two PRs are related to fusion, but they address different layers of the feature: one introduces a new generic Module Fusion API, while the other wires up n-to-1 kernel fusion through KernelConfig and integration plumbing. They do not look like duplicate implementations of the same concrete change.", - "confidence": 0.89, + "summary": "These PRs are related around module/kernel fusion, but they are not duplicates: #44979 adds a new generic Module Fusion API and tests, while #45363 wires up n-to-1 kernel fusion through KernelConfig and touches different integration points. They address adjacent layers of the feature, not the same concrete change.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45363 is the better representative of the cluster because it shows the more complete end-user feature path: KernelConfig wiring, integration changes, and a usage example. PR #44979 is a lower-level API scaffold for module fusion rather than the same change.", + "canonical_pr_reason": "#45363 is the more complete, user-facing implementation of the fusion feature, spanning KernelConfig, integrations, and modeling hooks; #44979 is the lower-level API foundation.", "best_issue_reason": null, - "best_pr_reason": "PR #45363 is the most substantive and feature-complete PR in the set, with model/config plumbing and an example demonstrating the fusion workflow. It is more suitable as the cluster anchor than the standalone API proposal in #44979.", + "best_pr_reason": "#45363 best represents the cluster because it implements the end-to-end KernelConfig-based fusion path and has the wider surface area of the feature.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Different scope and code path: #44979 adds a generic module fusion API and tests, while #45363 implements kernel fusion via KernelConfig across modeling/integration code. They are related conceptually but not the same fix/change." + "reason": "Related feature work, but not the same change: #44979 introduces the generic module fusion API, while #45363 implements n-to-1 kernel fusion via KernelConfig in different code paths. They would not be merged as one PR." } ] }, @@ -17475,13 +17530,13 @@ { "label": "pull_request", "component_index": 19, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45086-2", "nodes": [ @@ -17491,7 +17546,7 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "d14de3bcfe2fdfc447a1cfa690dbd864ff539603c6575c3c4a397ad77f7e66ba", + "prepared_review_unit_hash": "a9942be8259f16d437fe7acea1cf847428493ee1ccc8e07a5bf4d9abab58095d", "component_budget": { "node_count": 2, "item_count": 2, @@ -17513,18 +17568,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests address the same concrete bug in `_patch_mistral_regex`: using `tokenizer.backend_tokenizer.pre_tokenizer` caused an `AttributeError`, and both switch to `tokenizer.pre_tokenizer` with the same underlying fix. PR 45317 is the stronger canonical choice because it includes the fix plus test coverage.", + "summary": "Both pull requests target the same Mistral regex AttributeError in the same function and make the same code change; one adds tests, so this is a strong duplicate pair.", "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "45317 is the better canonical PR: it fixes the same AttributeError as 45086 and also adds a test file, making it the more complete and merge-ready version of the same change.", + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same core fix plus test coverage and has more discussion/review activity, making it the more complete representative of this change.", "best_issue_reason": null, - "best_pr_reason": "45317 is the best PR representative because it targets the same issue, applies the same code-path fix, and includes validation via tests.", + "best_pr_reason": "PR 45317 is the best representative of the underlying change: same fix as 45086, with added tests and stronger validation of the bug fix.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying bug and same fix in the same file; 45317 is a near-direct superset of 45086 with added tests, so they are duplicates/versions of the same change." + "reason": "Same concrete bug and same code-path fix in _patch_mistral_regex, same target issue, and same file diff; 45317 is effectively a superset with tests." } ] }, @@ -17538,13 +17593,13 @@ { "label": "pull_request", "component_index": 20, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45115-2", "nodes": [ @@ -17554,7 +17609,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "eac41f007bdb1fb80d3140400e7840d04f0a9b6e6804b7d6a024327e017ce14a", + "prepared_review_unit_hash": "91f123a6a533a1fe0a4ae85e2313900cc5c21fa356623589f75889a2d337bcd3", "component_budget": { "node_count": 2, "item_count": 2, @@ -17576,18 +17631,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch NemotronH internals, but they are not duplicates: one is a narrow conversion-mapping/refactor pass, while the other is a broader split into NemotronH dense/sparse packages with docs and tests. The overlap is mostly in shared base model files, not the same concrete change.", - "confidence": 0.91, + "summary": "Two open PRs around NemotronH share some model files, but they target different levels of the stack: one is a narrower refactor/conversion-mapping update, while the other is a broader model split/renaming and docs update. They are related, but not the same underlying change.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45569 is the broader, more central NemotronH restructuring change: it introduces the dense/sparse model split, updates auto-mapping, docs, and tests, and wraps the old NemotronH entrypoint for backward compatibility. PR #45115 is a narrower converter/inheritance refactor.", + "canonical_pr_reason": "PR #45569 is the better cluster representative because it contains the broader, user-visible model restructuring (NemotronH dispatching to dense/sparse variants, auto-mapping, docs, and tests), whereas #45115 is a narrower inheritance/conversion-mapping refactor.", "best_issue_reason": null, - "best_pr_reason": "PR #45569 best represents the cluster because it captures the main architectural transition for NemotronH and its new variants; the other PR is a smaller related refactor.", + "best_pr_reason": "PR #45569 is the strongest single PR in the cluster: it addresses the concrete architecture packaging change end-to-end and is more complete than the narrower #45115 refactor.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Shared NemotronH files are not enough here: #45115 is about GraniteMoeHybrid-compatible conversion/key renaming and inheritance refactor, while #45569 adds separate nemotron_h_dense/nemotron_h_sparse modules, docs, and tests. Different concrete changes, not one mergeable fix." + "reason": "Shared NemotronH internals do not make these duplicates. #45115 focuses on conversion key remapping and GraniteMoeHybrid-style inheritance, while #45569 splits NemotronH into dense/sparse variants, updates auto/docs, and adds tests; they are different changes that would not plausibly merge into one PR." } ] }, @@ -17601,13 +17656,13 @@ { "label": "pull_request", "component_index": 21, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45332-2", "nodes": [ @@ -17617,7 +17672,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "97af8317c894556305aeb36f7b95bc9ef2ff945cd44fae30412e91d8d6fa6633", + "prepared_review_unit_hash": "be24af849aeeb265ac9a9d804ac70f0135f23bdf05edf7f439f3667678be0ecb", "component_budget": { "node_count": 2, "item_count": 2, @@ -17639,18 +17694,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related but not duplicates: PR #45333 adds heterogeneous per-layer configuration support, while PR #45332 builds on that to add broader heterogeneous model/masking/cache integration. They share infrastructure files, but they are different changes and not mergeable as a single duplicate fix.", - "confidence": 0.92, + "summary": "Two related but distinct PRs: one adds heterogeneous per-layer configuration plumbing, the other adds broader heterogeneous model/runtime support. They share config utilities and tests, but they are not the same concrete change and would not be merged as one duplicate PR.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45332 is the better canonical representative because it is the broader end-to-end heterogeneous model support change and subsumes the config groundwork it depends on.", + "canonical_pr_reason": "PR 45332 is the more representative change for this cluster because it covers the end-to-end heterogeneous model/runtime integration, while 45333 is narrower config-only plumbing.", "best_issue_reason": null, - "best_pr_reason": "PR #45332 is the strongest cluster anchor: it covers the concrete modeling/cache/masking behavior, not just the config plumbing.", + "best_pr_reason": "PR 45332 is the best standalone representative since it exercises the actual heterogeneous code paths beyond configuration serialization.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related prerequisite/follow-up work, but not the same underlying change. #45333 is config-only; #45332 adds broader model/runtime support and cannot be treated as a duplicate PR." + "reason": "Related feature work, but not the same underlying fix/change: 45333 adds per-layer config support, while 45332 wires heterogeneous behavior into cache/modeling/integration paths. They share files and a theme, not a single mergeable PR scope." } ] }, @@ -17664,13 +17719,13 @@ { "label": "pull_request", "component_index": 22, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45394-5", "nodes": [ @@ -17689,76 +17744,76 @@ "pull_request:45592|pull_request:45615", "pull_request:45596|pull_request:45615" ], - "prepared_review_unit_hash": "dd61d5a6c57e73f578e65abeb1b8113581a5b875c2fd43a2f89882c2019cb2db", + "prepared_review_unit_hash": "320fdd3021344fe0050204df1c4ce6e9de4fb30fed5894810055e9e7ec9f3c48", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 7, - "serialized_chars": 13273, - "estimated_input_tokens": 3319, - "estimated_eval_tokens": 6894 + "serialized_chars": 13269, + "estimated_input_tokens": 3318, + "estimated_eval_tokens": 6892 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 7, - "serialized_chars": 13273, - "estimated_input_tokens": 3319, - "estimated_eval_tokens": 6894 + "serialized_chars": 13269, + "estimated_input_tokens": 3318, + "estimated_eval_tokens": 6892 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are not duplicates of one another. They span different models and different fix types: a real model-parallel code fix, a separate XCLIP split-module fix, and several test-only expectation updates for XPU/padding behavior.", - "confidence": 0.94, + "summary": "This cluster is a mix of unrelated model-specific fixes and test expectation updates. There isn\u2019t a single underlying bug or change spanning the PRs, so none of the soft edges should be merged as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue artifacts are present. Among the PRs, #45487 is the strongest representative because it contains a concrete runtime bug fix in model code (device placement / model-parallel handling) rather than only test expectation updates.", + "canonical_pr_reason": "PR 45487 is the strongest representative because it contains an actual code-path fix for model parallel behavior, not just test expectation updates, and it affects multiple related models.", "best_issue_reason": null, - "best_pr_reason": "#45487 is the best standalone PR in the set: it addresses an actual code-path problem, touches model implementation files, and fixes a broader class of model-parallel failures than the test-only PRs.", + "best_pr_reason": "PR 45487 is the best overall PR candidate: it addresses a concrete runtime/model-parallel bug in model code and has broader functional impact than the test-only PRs.", "soft_edge_verdicts": [ { "left": "pull_request:45394", "right": "pull_request:45596", "accept": false, - "reason": "Different models and different problems: XCLIP split-module handling vs BLT XPU generation expectations. Same broad theme of failed tests is not enough." + "reason": "Different models and different fixes: XCLIP model-parallel split-module metadata vs BLT XPU test expectations." }, { "left": "pull_request:45487", "right": "pull_request:45592", "accept": false, - "reason": "#45487 fixes model-parallel device placement in model code; #45592 only changes FastVLM test padding-side behavior. Not the same underlying bug/change." + "reason": "AltCLIP/ChineseCLIP model-parallel/device handling fix is unrelated to a FastVLM test padding-side adjustment." }, { "left": "pull_request:45487", "right": "pull_request:45596", "accept": false, - "reason": "Both mention XPU/model issues, but one is a substantive AltCLIP/ChineseCLIP code fix and the other is BLT test expectation updates. Different concrete changes." + "reason": "One fixes model parallel behavior in model code; the other only updates BLT XPU test expectations." }, { "left": "pull_request:45487", "right": "pull_request:45615", "accept": false, - "reason": "#45487 fixes runtime model-parallel behavior; #45615 only adds XPU expectations in qianfan_ocr tests. Different artifacts and no shared code-path bug." + "reason": "Different models and different change types; no shared concrete bug or code path." }, { "left": "pull_request:45592", "right": "pull_request:45596", "accept": false, - "reason": "FastVLM test padding-side issue and BLT XPU expectation issue are unrelated test fixes for different models." + "reason": "Both are test-related, but they target different models and different failure modes." }, { "left": "pull_request:45592", "right": "pull_request:45615", "accept": false, - "reason": "#45592 is about FastVLM padding/test setup; #45615 is about qianfan_ocr XPU expected outputs. Not the same change." + "reason": "FastVLM padding-side test fix is unrelated to Qianfan OCR XPU expectation updates." }, { "left": "pull_request:45596", "right": "pull_request:45615", "accept": false, - "reason": "Both are test expectation tweaks for different models, but they do not fix the same concrete bug or code-path problem." + "reason": "Both add XPU expectations, but for different model test suites and different outputs, so they are not the same underlying change." } ] }, @@ -17772,13 +17827,13 @@ { "label": "pull_request", "component_index": 23, - "component_count": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45415-2", "nodes": [ @@ -17788,12 +17843,12 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "8451c32533e2c451e4f0e06b012fba13d350698fddb3b0cbe92ab707c6f62423", + "prepared_review_unit_hash": "2c66e933be0d2eb5f890a40250a622430bab7d383479d16ceb2104eeea872895", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5513, + "serialized_chars": 5514, "estimated_input_tokens": 1379, "estimated_eval_tokens": 3014 }, @@ -17801,7 +17856,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5513, + "serialized_chars": 5514, "estimated_input_tokens": 1379, "estimated_eval_tokens": 3014 }, @@ -17810,18 +17865,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two different typing-related pull requests. PR 45415 is a broad type-checking sweep across many Transformers files, while PR 45425 adds specific typing helpers/protocols and related fixes. They overlap in a few files but do not look like the same underlying change.", - "confidence": 0.84, + "summary": "Both PRs are typing-related cleanups in transformers, but they target different scopes and fixes: one is a broad type-checking sweep across many files, the other adds specific typing helpers and targeted annotations in a smaller set of modules.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45425 is the better canonical representative because it is a focused typing-support change with a clear purpose, whereas PR 45415 is a much broader repo-wide type-checking sweep.", + "canonical_pr_reason": "PR #45425 is the better canonical representative because it is a more focused, concrete typing infrastructure change, while #45415 is a much broader repo-wide type-checking sweep.", "best_issue_reason": null, - "best_pr_reason": "PR 45425 is the strongest standalone PR in the cluster because it makes concrete typing infrastructure additions and is more self-contained than the large type-checking sweep in PR 45415.", + "best_pr_reason": "PR #45425 is the strongest standalone candidate in this cluster due to its narrower, more coherent scope and clearer code-level purpose.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Overlap in typing-related files, but the changes are not the same concrete fix: 45415 is a broad type-checking pass across many modules, while 45425 adds specific typing aliases/protocols and targeted support code." + "reason": "Both are typing-related, but they are not the same underlying change: #45415 is a broad type-checking pass across many files plus config changes, while #45425 adds specific typing protocols/helpers and targeted fixes. They would not plausibly be merged as one PR." } ] }, @@ -17835,76 +17890,13 @@ { "label": "pull_request", "component_index": 24, - "component_count": 25, - "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-45524-2", - "nodes": [ - "pull_request:45524", - "pull_request:45650" - ], - "soft_pairs": [ - "pull_request:45524|pull_request:45650" - ], - "prepared_review_unit_hash": "a4b3fca1cdc691cd23bb743cc758fb03f46d9f2a4b5e5abf16b9814674b4362b", - "component_budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 4378, - "estimated_input_tokens": 1095, - "estimated_eval_tokens": 2446 - }, - "budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 4378, - "estimated_input_tokens": 1095, - "estimated_eval_tokens": 2446 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": false, - "analyst_result": { - "summary": "Both PRs target flash_attn-related availability checks in `import_utils.py` for Python 3.13, but they are not the same patch: 45524 is the narrower KeyError/missing-mapping fix, while 45650 also changes the FA3 package lookup behavior.", - "confidence": 0.79, - "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45524 is the cleaner canonical fix: it directly guards `PACKAGE_DISTRIBUTION_MAPPING` lookups with `.get(..., [])` for the flash_attn checks without changing the underlying package-selection logic.", - "best_issue_reason": null, - "best_pr_reason": "PR 45524 is the best representative of the underlying bug because it matches the issue description most closely and avoids the extra behavior change introduced by 45650.", - "soft_edge_verdicts": [ - { - "left": "pull_request:45524", - "right": "pull_request:45650", - "accept": false, - "reason": "Reject: they overlap on the same file and issue target, but 45650 is not just a duplicate of the KeyError fix; it also changes `is_flash_attn_3_available()` to check `flash_attn` instead of `flash_attn_interface`, so the concrete code-path change is different." - } - ] - }, - "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": 25, + "component_count": 24, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", "reason": null, - "source": "cache", - "cache_hit": true, + "source": "llm", + "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-45639-2", "nodes": [ @@ -17914,7 +17906,7 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "35953697445f117e60c25c0cdab94e3f444d5f83a9849552c0f9b89d95da8e1e", + "prepared_review_unit_hash": "583ba3da5b7a107c6b9ff217eb0a4ec45c274a74359cec1d346a1b461d15c4c2", "component_budget": { "node_count": 2, "item_count": 2, @@ -17936,18 +17928,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two PRs address the same xdist collision problem around captured_info debug artifacts; 45645 is the broader, more complete fix and 45639 is a narrower subset of it.", - "confidence": 0.82, + "summary": "Both PRs address the same captured_info xdist collision problem; #45645 is the broader, more complete fix, while #45639 is a narrower subset focused on the testing_utils path.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the best canonical representative because it includes the core xdist-safe output-file change in testing_utils and also updates CI artifact collection to handle multiple worker files.", + "canonical_pr_reason": "PR #45645 is the better canonical PR because it includes the core xdist-safe file handling change plus the CI workflow update needed to consume multiple captured_info artifacts, making it the more complete fix for the same bug.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 is the strongest overall fix: it covers the same root collision bug as 45639 and also preserves/log-collects the resulting debug artifacts correctly in CI.", + "best_pr_reason": "PR #45645 best represents the cluster since it covers the underlying captured_info artifact collision end-to-end, not just the library-side write path.", "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 appears to subsume 45639 by adding the same fix plus CI-side aggregation for the new per-worker files." + "reason": "Accept: both PRs fix the same concrete xdist collision around captured_info output; #45639 is a narrower implementation of the same bug that #45645 also addresses, and they could plausibly be merged into one PR." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index f534d7adb60a027a6be0bc44c0de878da5c92306..5ab25b7e31a23d33c050f97a8acfb05776f3a1ea 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260427t000017z", + "analysis_id": "hybrid-model-20260427t120026z", "archived_artifacts": { - "hybrid": "snapshots/20260427T000017Z/analysis-runs/hybrid-model-20260427t000017z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260427T000017Z/analysis-runs/hybrid-model-20260427t000017z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-04-27T00:01:34Z", + "published_at": "2026-04-27T12:09:03Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260427T000017Z", + "snapshot_id": "20260427T120026Z", "variant": "hybrid" } diff --git a/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..2c931c276370566b14a30ab66c73e22c68c078ab --- /dev/null +++ b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json @@ -0,0 +1,1963 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260427T120026Z", + "generated_at": "2026-04-27T12:09:02Z", + "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-43656-4", + "summary": "Cluster of 1 issues and 3 PRs centered on issue #43824.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 43824, + "canonical_pr_number": 43656, + "issue_numbers": [ + 43824 + ], + "pr_numbers": [ + 43656, + 43836, + 43842 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43656, + "right_pr_number": 43836, + "code_similarity": 0.36, + "size_similarity": 0.176, + "file_overlap": 1.0, + "area_overlap": 0.051, + "patch_similarity": 0.048, + "shared_filenames": [ + "src/transformers/cli/serve.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/serve.py", + "left_ranges": [ + [ + 11, + 18 + ], + [ + 30, + 36 + ], + [ + 315, + 323 + ], + [ + 665, + 671 + ], + [ + 931, + 937 + ], + [ + 1843, + 1849 + ], + [ + 1868, + 1874 + ] + ], + "right_ranges": [ + [ + 11, + 18 + ], + [ + 359, + 429 + ], + [ + 584, + 590 + ], + [ + 1892, + 1910 + ], + [ + 1917, + 1923 + ] + ] + } + ] + }, + { + "left_pr_number": 43656, + "right_pr_number": 43842, + "code_similarity": 0.405, + "size_similarity": 0.5, + "file_overlap": 1.0, + "area_overlap": 0.0, + "patch_similarity": 0.036, + "shared_filenames": [ + "src/transformers/cli/serve.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/serve.py", + "left_ranges": [ + [ + 11, + 18 + ], + [ + 30, + 36 + ], + [ + 315, + 323 + ], + [ + 665, + 671 + ], + [ + 931, + 937 + ], + [ + 1843, + 1849 + ], + [ + 1868, + 1874 + ] + ], + "right_ranges": [ + [ + 54, + 61 + ], + [ + 587, + 593 + ] + ] + } + ] + }, + { + "left_pr_number": 43836, + "right_pr_number": 43842, + "code_similarity": 0.332, + "size_similarity": 0.088, + "file_overlap": 1.0, + "area_overlap": 0.033, + "patch_similarity": 0.017, + "shared_filenames": [ + "src/transformers/cli/serve.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/serve.py", + "left_ranges": [ + [ + 11, + 18 + ], + [ + 359, + 429 + ], + [ + 584, + 590 + ], + [ + 1892, + 1910 + ], + [ + 1917, + 1923 + ] + ], + "right_ranges": [ + [ + 54, + 61 + ], + [ + 587, + 593 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-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-45081-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45081, + "canonical_pr_number": 45317, + "issue_numbers": [ + 45081 + ], + "pr_numbers": [ + 45086, + 45317 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45086, + "right_pr_number": 45317, + "code_similarity": 0.589, + "size_similarity": 0.34, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.136, + "shared_filenames": [ + "src/transformers/tokenization_utils_tokenizers.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/tokenization_utils_tokenizers.py", + "left_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ], + "right_ranges": [ + [ + 1360, + 1370 + ], + [ + 1374, + 1380 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-43698-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #43698.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 43698, + "canonical_pr_number": 43779, + "issue_numbers": [ + 43698 + ], + "pr_numbers": [ + 43779, + 43816 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 43779, + "right_pr_number": 43816, + "code_similarity": 0.418, + "size_similarity": 0.538, + "file_overlap": 1.0, + "area_overlap": 0.02, + "patch_similarity": 0.02, + "shared_filenames": [ + "src/transformers/integrations/integration_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/integration_utils.py", + "left_ranges": [ + [ + 2227, + 2241 + ], + [ + 2303, + 2309 + ] + ], + "right_ranges": [ + [ + 2278, + 2291 + ], + [ + 2309, + 2322 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45561-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45561, + "canonical_pr_number": 45639, + "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": 0.532, + "size_similarity": 0.619, + "file_overlap": 0.5, + "area_overlap": 0.543, + "patch_similarity": 0.451, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3529, + 3536 + ], + [ + 3753, + 3787 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 86 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [], + "duplicate_prs": [ + { + "cluster_id": "cluster-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-43656-4", + "canonical_pr_number": 43656, + "duplicate_pr_numbers": [ + 43836, + 43842 + ], + "target_issue_number": 43824, + "reason": "PRs in cluster-43656-4 are treated as duplicates because they converge on issue #43824 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-43698-3", + "canonical_pr_number": 43779, + "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": 45639, + "duplicate_pr_numbers": [ + 45645 + ], + "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." + } + ], + "best_issue": { + "cluster_id": "cluster-41211-3", + "issue_number": 41211, + "reason": "Issue #41211 is the strongest global issue candidate because it is open, belongs to a cluster with 3 artifacts, and carries 6 discussion comments plus 6 inbound references.", + "score": 148.0 + }, + "best_pr": { + "cluster_id": "cluster-43979-11", + "pr_number": 44007, + "reason": "PR #44007 is the strongest global PR candidate because it is open, belongs to a cluster with 11 artifacts, links to 1 issues, and carries 2 review/discussion events.", + "score": 238.886 + } +} diff --git a/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..4ee9598350f0caec03ca2629999ea587b8125993 --- /dev/null +++ b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,17954 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260427T120026Z", + "generated_at": "2026-04-27T12:09:02Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 13, + "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": "9f08dd43c33773c61325575f36c0aa14b48447224fc818d2a894eb9e6d42ec60", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues do not appear to be duplicates. #16618 is the broadest model-loading failure report, while #15980 is specifically about the error message shown when downloading a private model without authentication, and #13244 is about Tapas tokenization diverging from TensorFlow behavior.", + "confidence": 0.81, + "canonical_issue_reason": "#16618 is the best canonical issue because it is the most general user-facing failure report in this set and centers on model loading, which is the closest match to a core bug rather than a narrow UX or preprocessing discrepancy.", + "canonical_pr_reason": null, + "best_issue_reason": "#16618 is the best overall issue candidate because it describes a concrete model-load failure and is broader than the other two, which are clearly narrower in scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization mismatch and failed loading of bert-base-uncased are different problems affecting different code paths." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Bad auth-related download messaging for private models is not the same underlying bug as an inability to load bert-base-uncased." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 1, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41093", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44661", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45070", + "issue:45081", + "issue:45084", + "issue:45161", + "issue:45237", + "issue:45362", + "issue:45464", + "issue:45507" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:45507", + "issue:44292|issue:45237", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:44509|issue:44512", + "issue:35545|issue:36010", + "issue:43638|issue:44805", + "issue:43596|issue:44805", + "issue:44805|issue:45161", + "issue:44805|issue:44918", + "issue:44661|issue:44805", + "issue:30277|issue:43638", + "issue:34634|issue:35545", + "issue:43596|issue:43638", + "issue:43975|issue:44457", + "issue:43039|issue:45161", + "issue:44292|issue:44457", + "issue:34928|issue:44805", + "issue:24643|issue:30277", + "issue:43975|issue:45237", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:39290|issue:45081", + "issue:43828|issue:45161", + "issue:43425|issue:44292", + "issue:44918|issue:45161", + "issue:34928|issue:41720", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:44292", + "issue:44292|issue:44918", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43716|issue:44560", + "issue:43716|issue:44918", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:43039|issue:44805", + "issue:34928|issue:36010", + "issue:41762|issue:43596", + "issue:43901|issue:44512", + "issue:39290|issue:41720", + "issue:43039|issue:44560", + "issue:43866|issue:44863", + "issue:43854|issue:43866", + "issue:43425|issue:45070", + "issue:39290|issue:43531", + "issue:41093|issue:43638", + "issue:43404|issue:43866", + "issue:45362|issue:45464", + "issue:43792|issue:45070", + "issue:43541|issue:43828", + "issue:36010|issue:39290", + "issue:44841|issue:45084", + "issue:41093|issue:41762", + "issue:43296|issue:43366" + ], + "prepared_review_unit_hash": "1209ca2d249a0324b4088139e576eaa76d140574aed3ef7dbf500677feb27d16", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 46483, + "estimated_input_tokens": 11621, + "estimated_eval_tokens": 23498 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a mix of unrelated model/runtime failures, but there is a small coherent subcluster around zero-3/indexing errors and a docs-cleanup subcluster around v5 pipeline removals.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44805 is the broadest representative of the repeated mask/index-shape failure family and best matches the closest duplicate-like reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44805 is the most central and generic issue here, with the clearest overlap to other index/shape-mismatch reports.", + "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 failure." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different models and different bugs; dtype mismatch in preprocessing is unrelated to AMD GPU execution failure." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Both are runtime dtype-related, but they affect different models and code paths." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different model families and different failures; same broad MoE area is not enough." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Different inference stacks and error modes; not the same concrete bug." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collective mismatch is unrelated to ONNX export errors." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "Export-to-ONNX failure and FP8 config failure are different problems." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Different models and runtime environments; no shared code-path bug." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are v5 docs cleanups, but they target different removed features and are not the same doc change." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ONNX export issue vs import-path regression; unrelated." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Both are index errors, but one is ZeRO-3 loading on Bert while the other is a mask/tensor shape mismatch; not the same bug." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Different error signatures and likely different root causes." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Shape-mismatch IndexError vs GPT-OSS tensor-parallel issue; unrelated." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Model-mapping generation failure is not the same as the mask/tensor indexing error." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Collective mismatch and ZeRO-3 index error are different failure classes." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "Processor voice preset bug is unrelated to ONNX export." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Near-identical ZeRO-3 index-0-out-of-bounds failures on Bert models with the same underlying loading path." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Both affect output correctness, but one is detokenization in v5 and the other is LoRA merge/reload inconsistency." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger kernel cross-entropy dispatch and GPT-OSS tensor parallel issues are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "NVFP4 runtime error and LoRA save/reload mismatch are different bugs." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "FSDP/activation-checkpoint recomputation mismatch is a different training bug from the mask/tensor index error." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "DeepSpeed weight-dimension error and collective mismatch are different issues." + }, + { + "left": "issue:43975", + "right": "issue:45237", + "accept": false, + "reason": "Different models and different failure mechanisms." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "ZeRO-3 Bert loading error is unrelated to GPT-OSS tensor-parallel behavior." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "Activation-checkpoint/FSDP mismatch is unrelated to the custom-trainer argument regression." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different model loading path and different failure mode." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Image-preprocessor dtype mismatch is unrelated to invalid-method calls in GraniteMoEHybrid." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Gemma3/vLLM config attribute issue and tokenizer regex patch crash are unrelated." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Different models and different runtime pathways; same MoE label is too broad." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Torch version incompatibility and NVFP4 runtime error are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRL embedding unpacking issue is not the same as GPT-OSS tensor-parallel failure." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "FSDP checkpointing mismatch and Qwen3 auto-device-map assert are different bugs." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs regressions in v5, but they concern different removed commands and are separate cleanup tasks." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different model-specific dtype/video-processing failures." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Preprocessor dtype mismatch and TRL embedding unpacking are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Different model/runtime bugs despite both touching numeric kernels." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "Collective mismatch and trainer-signature regression are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Liger-kernel cross-entropy dispatch is unrelated to indexing/shape mismatch errors." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "Training/checkpointing bug and import regression are unrelated." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Both are ZeRO-3 index errors, but they hit different model families and likely different root causes." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Docs mentioning return_all_scores is unrelated to the removed run command docs issue." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Different models and different runtime errors." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger kernel cross-entropy issue and Qwen3-vl video StopIteration are unrelated." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Checkpoint corruption and NemotronH loading failure are different issues." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Different model-loading failures with no clear shared root cause." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch compatibility and pydantic PretrainedConfig-field breakage are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Gemma3/vLLM attribute issue and Qwen3-MoE sliding_window bug are different model regressions." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Same broad index-error family, but different concrete symptoms and code paths." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "lm_head tying bug is unrelated to checkpoint corruption." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Related model family, but the failure modes differ: chat generation crash vs streaming inference error." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper model runtime failure and pydantic field regression are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Different MoE/runtime failures with distinct stack traces." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin is unrelated to Gemma3/vLLM attribute issues." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Processor failure and template-node compilation error are unrelated." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are index errors, but the reported causes and affected paths differ." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL load failure and GGUF gpt-oss support request are different issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: it correctly identifies a loose zero-3/indexing subcluster and a separate v5 docs-removal subcluster, while treating the rest as mostly unrelated runtime/model failures. The soft-edge verdicts are generally conservative, and the one accepted pair (43596/43638) is well-supported by the titles as a closely related ZeRO-3 Bert index-out-of-bounds family." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 2, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36331", + "issue:39186", + "issue:39290", + "issue:39692", + "issue:40990", + "issue:41093", + "issue:41553", + "issue:41762", + "issue:42915", + "issue:43054", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43644", + "issue:43645", + "issue:43742", + "issue:43828", + "issue:43866", + "issue:43873", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:43994", + "issue:44291", + "issue:44292", + "issue:44360", + "issue:44387", + "issue:44403", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44863", + "issue:44918", + "issue:44928", + "issue:44960", + "issue:45005", + "issue:45070", + "issue:45161", + "issue:45399", + "issue:45464", + "issue:45507" + ], + "soft_pairs": [ + "issue:43572|issue:45070", + "issue:44387|issue:45005", + "issue:43541|issue:45161", + "issue:43828|issue:45464", + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:43531|issue:45070", + "issue:44661|issue:45464", + "issue:39290|issue:43742", + "issue:43541|issue:44589", + "issue:43054|issue:43994", + "issue:41093|issue:43596", + "issue:42915|issue:43541", + "issue:40990|issue:44960", + "issue:39186|issue:44918", + "issue:43404|issue:45005", + "issue:43645|issue:45399", + "issue:44530|issue:45005", + "issue:44863|issue:45005", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:42915|issue:45399", + "issue:44918|issue:45507", + "issue:44492|issue:44512", + "issue:43975|issue:44292", + "issue:43950|issue:44534", + "issue:43644|issue:43950", + "issue:44292|issue:45507", + "issue:39692|issue:43054", + "issue:43828|issue:44928", + "issue:41553|issue:43883", + "issue:43866|issue:44403", + "issue:44560|issue:45507", + "issue:36331|issue:41762" + ], + "prepared_review_unit_hash": "3d3cfe81a9eaae410c2b37143b572c1a4a6fa3f93905260e1f0a076ba86fe67e", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37566, + "estimated_input_tokens": 9392, + "estimated_eval_tokens": 19040 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous and does not form one large duplicate cluster. The only clear duplicate family is the non-persistent buffer regression in Transformers v5 (`43644`, `43950`, `44534`).", + "confidence": 0.93, + "canonical_issue_reason": "issue:43950 is the clearest canonical representative of the duplicate family because it most precisely identifies the root bug: `from_pretrained()` corrupting `register_buffer(persistent=False)` buffers in Transformers 5.x.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43950 is the best issue to anchor the duplicate set: it is the most specific, technically precise, and directly describes the underlying regression rather than just the symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Different bugs: one is a missing `pad_token_idx` config field in `StableLmConfig`, the other is a Pydantic `PretrainedConfig` field regression in v5." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "Both involve v5 behavior changes, but one is an int4 memory/OOM issue and the other is tied-weight handling for translation models; not the same bug." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Different failure modes: torch dynamo grouped_mm tracing vs tensor-parallel support for GPT-OSS MoE." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Unrelated model/runtime problems: dtype mismatch under autocast for Phi-tiny-MoE vs streaming inference failure for Qwen3.5-0.8B." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "`init_empty_weights` unexpected `_is_hf_initialized` argument is a distinct constructor/API mismatch, not the Pydantic field regression." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "One is GGUF support for GPT-OSS architecture, the other is a streaming chat/completions failure for Qwen3.5; no shared code-path bug." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding-window handling is unrelated to the Pydantic `PretrainedConfig` field issue." + }, + { + "left": "issue:44661", + "right": "issue:45464", + "accept": false, + "reason": "Tokenizer mapping/add-new-model-like failure is unrelated to Qwen3.5 streaming inference." + }, + { + "left": "issue:39290", + "right": "issue:43742", + "accept": false, + "reason": "Gemma3 sliding-window attribute error and MobileLLM key error are different model-loading issues." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Grouped_mm tracing failure and missing Float8 storage are distinct runtime errors." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Both concern SigLIP2, but one is degraded text embeddings while the other is bad AutoModel/pipeline outputs; not enough evidence of the same underlying bug." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Different shape/index errors in separate loading paths; no clear duplicate relation." + }, + { + "left": "issue:42915", + "right": "issue:43541", + "accept": false, + "reason": "FineGrainedFP8Config failure for Qwen3 MoE is unrelated to Mixtral dynamo tracing." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "Perplexity regression on GPT-OSS and a generic GLM5 issue are not the same underlying defect." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP `'weight' must be 2-D` and Qwen3.5 SFT unpacking failure are separate training-time bugs." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a Mistral3 model-specific missing tie and the other is a broader v5 translation-model issue; not the same concrete bug." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Jupyter custom-model initialization regression is unrelated to flash-attn fallback checks." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash is unrelated to tied-weight handling." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "NemotronH checkpoint loading and translation-model tied weights are different problems." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is not the same as tied-weight regressions." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "DSA indexer/ReLU discussion and docs command removal are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "FineGrainedFP8 Qwen3 MoE failure is unrelated to flash-attn fallback gating." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "TRL embedding unpacking failure and GraniteMoEHybrid invalid method are separate model/training bugs." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "A typo in cache strategy docs is unrelated to docs still mentioning a removed command." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "DeepSeek detokenization regression and Qwen-3-8B-NVFP4 runtime failure are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same underlying v5 regression: non-persistent buffers registered with `persistent=False` are getting corrupted/junk values during loading." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Clear duplicate wording and same bug: Transformers 5.x fills/corrupts non-persistent buffers during `from_pretrained()`." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific failures with no shared bug or code path." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "SigLIP2 docs example errors and SigLIP2 embedding quality regression are related only at model family level, not the same concrete bug." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Different Qwen3.5-related failures: autocast dtype mismatch vs RLHF NaN explosion from position IDs/SDPA fallback." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad AutoTokenizer error messaging for Voxtral is unrelated to missing `all_tied_weights_keys` in Molmo." + }, + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "Corrupted Ovis2 checkpoint and noisy load logging are not the same bug." + }, + { + "left": "issue:44560", + "right": "issue:45507", + "accept": false, + "reason": "Qwen3-vl-embedding video `StopIteration` and GraniteMoEHybrid invalid method are unrelated." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Trainer `compute_loss` signature mismatch and Gemma3/ZeRO-3 index error are separate issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: the only clearly supported duplicate family is the non-persistent buffer regression (`43644` / `43950` / `44534`), and the accepted links are consistent with the issue titles. The soft-edge rejections are generally conservative and do not overclaim duplicate status from weak similarity alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 3, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:38175", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43257", + "issue:43278", + "issue:43404", + "issue:43493", + "issue:43643", + "issue:43742", + "issue:43782", + "issue:43792", + "issue:43824", + "issue:43827", + "issue:43854", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43940", + "issue:43950", + "issue:43975", + "issue:44220", + "issue:44291", + "issue:44368", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44509", + "issue:44589", + "issue:44661", + "issue:44863", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45305", + "issue:45313", + "issue:45356", + "issue:45399", + "issue:45520", + "issue:45588" + ], + "soft_pairs": [ + "issue:41553|issue:43881", + "issue:43872|issue:44589", + "issue:38175|issue:43054", + "issue:42947|issue:44387", + "issue:43054|issue:43493", + "issue:43881|issue:43883", + "issue:44403|issue:44451", + "issue:43975|issue:45520", + "issue:43643|issue:45020", + "issue:42947|issue:43856", + "issue:44403|issue:44488", + "issue:43782|issue:44960", + "issue:42491|issue:44368", + "issue:41762|issue:42915", + "issue:43975|issue:44451", + "issue:45020|issue:45356", + "issue:44661|issue:45020", + "issue:43824|issue:43872", + "issue:43054|issue:43278", + "issue:44991|issue:45399", + "issue:43827|issue:44509", + "issue:43792|issue:44220", + "issue:45081|issue:45399", + "issue:43404|issue:43950", + "issue:43257|issue:43866", + "issue:42915|issue:45020", + "issue:43742|issue:45399", + "issue:41553|issue:44991", + "issue:43742|issue:45020", + "issue:43940|issue:45313", + "issue:43872|issue:44291", + "issue:43854|issue:44863", + "issue:44928|issue:45588", + "issue:41553|issue:44488", + "issue:42915|issue:43872", + "issue:43257|issue:45305" + ], + "prepared_review_unit_hash": "e38c6858ba3425ef79e36fe70ca05af31300dee0d9cfb35e5e54ee2f89d9a4c8", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38600, + "estimated_input_tokens": 9650, + "estimated_eval_tokens": 19556 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous loading/regression reports; only two soft pairs look like true duplicates. The strongest duplicate family is the `_is_hf_initialized`/`init_empty_weights` TypeError, and the docs-removal reports are also the same issue.", + "confidence": 0.71, + "canonical_issue_reason": "issue:44291 is the cleanest representative bug report: it states the exact `_is_hf_initialized` root cause, the `init_empty_weights` context, and generalizes beyond one model-specific symptom.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44291 is the best single issue to keep because it has the clearest repro and root-cause description; the other items are either narrower variants or unrelated regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different failures: bad AutoTokenizer error-message behavior vs a model-specific loading crash." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Same broad incompatibility area, but not the same concrete bug or code path." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Both mention SigLIP2, but one is zero probabilities and the other is degraded text embeddings; different symptoms." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Different problems: LoRA gradient checkpointing vs int4 memory reservation/OOM." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Same model family, but one is quality discrepancy and the other is HF-vs-JAX implementation mismatch; not enough to treat as the same bug." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Unrelated model-loading crash vs missing attribute on a different model class." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "One is loading noise/logging, the other is a specific model load failure." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Different subsystems: tokenizer detokenization vs flash_attn import/key handling on Python 3.13." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Both involve remote code, but one is missing fields from AutoConfig and the other is a broader recent-version breakage report." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Different training-performance issues: gradient checkpointing vs Qwen3 MoE memory usage." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Generic load-noise complaint vs a specific tokenizer/model loading failure." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Different model families and different failure modes." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA/Qwen3-MoE incompatibility vs a warning about tie_word_embeddings." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Both are Qwen/Gemma loading issues, but they affect different models and different error paths." + }, + { + "left": "issue:43975", + "right": "issue:44451", + "accept": false, + "reason": "Different symptoms and model targets." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Both are version regressions, but one is broad remote_code breakage and the other is a Kimi tokenizer codec/regex warning bug." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Different areas: tokenizer mapping helper failure vs remote_code loading regressions." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "ImportError for a VL class is unrelated to the bitsandbytes/_is_hf_initialized TypeError." + }, + { + "left": "issue:43054", + "right": "issue:43278", + "accept": false, + "reason": "Different embedding issues: SigLIP2 text quality vs dtype drift between train and eval." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading failure vs flash-attn fallback gating logic." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same docs bug: stale pipeline-task references after v5 removals." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Likely related audio path, but the reports are not specific enough to prove the same concrete bug." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer regex crash vs flash-attn2 fallback gating are different issues." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "Different state-management bugs: untied lm_head weights vs non-persistent buffer corruption." + }, + { + "left": "issue:43257", + "right": "issue:43866", + "accept": false, + "reason": "Different model/domain and different failure mode." + }, + { + "left": "issue:42915", + "right": "issue:45020", + "accept": false, + "reason": "Different Qwen3-MoE FP8 training bug vs a broad remote_code regression report." + }, + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "KeyError while loading a model vs flash-attn fallback selection logic." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Tokenizer loading problems, but different models and different error behavior." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "Different bug classes; both are loading issues only at a very broad level." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both mention ZeRO-3 and Qwen, but they are different model versions with different load failures." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same exact `_is_hf_initialized` / `init_empty_weights` TypeError family." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Model-loading test failure for GLM-4.7-Flash vs NemotronH checkpoint compatibility issue." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Different training/runtime failures: RLHF gradient explosion vs flash_attention auxiliary-state crash." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Different tokenizer/model load reports." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Different Qwen3 MoE training/loading bugs; no shared concrete code path is evident." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering is grounded in the packet and the verdicts stay conservative. The two accepted pairs are well-supported: 43872/44291 share the same `_is_hf_initialized` / `init_empty_weights` TypeError family, and 43827/44509 are both the v5 docs stale-pipeline-task issue. The rest are reasonably treated as heterogeneous or only loosely related." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 4, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:41093", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43317", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43596", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43720", + "issue:43824", + "issue:43828", + "issue:43872", + "issue:43950", + "issue:43957", + "issue:43994", + "issue:44155", + "issue:44479", + "issue:44530", + "issue:44534", + "issue:44560", + "issue:44589", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44918", + "issue:44960", + "issue:44977", + "issue:45030", + "issue:45313", + "issue:45325", + "issue:45357", + "issue:45538", + "issue:45584" + ], + "soft_pairs": [ + "issue:36331|issue:41093", + "issue:41093|issue:44805", + "issue:43872|issue:45538", + "issue:44479|issue:44560", + "issue:44530|issue:44863", + "issue:43366|issue:43828", + "issue:24643|issue:30064", + "issue:44155|issue:45584", + "issue:43366|issue:43531", + "issue:43550|issue:44910", + "issue:43645|issue:43646", + "issue:43950|issue:45325", + "issue:44560|issue:44918", + "issue:43366|issue:44589", + "issue:43824|issue:43957", + "issue:42915|issue:43957", + "issue:42617|issue:43366", + "issue:38175|issue:43994", + "issue:43646|issue:43950", + "issue:44560|issue:44805", + "issue:36010|issue:42915", + "issue:43638|issue:44661", + "issue:44877|issue:45030", + "issue:43596|issue:44661", + "issue:43317|issue:44756", + "issue:43644|issue:44534", + "issue:45313|issue:45357", + "issue:38175|issue:39692", + "issue:42491|issue:43720", + "issue:44960|issue:44977" + ], + "prepared_review_unit_hash": "b898a8eee5d636182263cdf90a11ba17a159618c37a89709713edbad6149c79e", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36154, + "estimated_input_tokens": 9039, + "estimated_eval_tokens": 18334 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous transformer bug reports. Only two soft duplicate pairs look strong enough to accept: the non-persistent-buffer regression and the SigLIP2 wrong-output reports. The rest are different failure modes or different model/code paths.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 43644 is the best cluster anchor: it is the clearest report of the non-persistent-buffer corruption regression and is earlier than the near-exact duplicate 44534.", + "canonical_pr_reason": null, + "best_issue_reason": "43644 is the strongest issue to keep as representative because it is specific, early, and directly describes the underlying regression rather than a symptom variant.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Different bugs: custom trainer signature mismatch vs a mask/tensor shape IndexError." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": false, + "reason": "Both are IndexError-shaped reports, but they appear to involve different examples/models and are not clearly the same failure." + }, + { + "left": "issue:43872", + "right": "issue:45538", + "accept": false, + "reason": "Unrelated: bitsandbytes constructor incompatibility vs CLIPTokenizer max-length behavior." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both concern Qwen video paths, but one is a broad v5.3.0 regression and the other is a StopIteration in the embedding model; not the same concrete bug." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "Different model families and failures: Qwen3.5 PagedAttentionCache crash vs NemotronH checkpoint loading." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF/gpt-oss support is unrelated to the Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "DeepSpeed training weight-shape error vs void segmentation map processing are unrelated." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "Different audio/text workflows and different bugs: embedding/token leak across tracks vs empty-transcription generation failure." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Different Qwen-related issues: GGUF architecture support vs sliding_window behavior." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different attention/compile failures on different models and code paths." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": false, + "reason": "Close in theme, but one is notebook-specific and the other is a broader custom-init report; not enough to confirm the same bug." + }, + { + "left": "issue:43950", + "right": "issue:45325", + "accept": false, + "reason": "Non-persistent buffer corruption is unrelated to the Qwen2.5-VL rope-index scaling bug." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Different Qwen3.5 video/training failures: StopIteration in video embedding vs TRL SFT embedding unpacking." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "GGUF model support is unrelated to the Float8 storage TypeError." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Import error for a missing model class vs meta-device loading regression." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Qwen3Moe FP8 failure and meta-device loading are different bugs." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py runtime issue vs GGUF gpt-oss support are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": true, + "reason": "Same underlying SigLIP2 wrong-output bug for the same checkpoint; one report says zero probabilities, the other says nonsensical outputs." + }, + { + "left": "issue:43646", + "right": "issue:43950", + "accept": false, + "reason": "Custom model initialization regression vs non-persistent buffer corruption are different failures." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different errors and likely different models/code paths; not the same bug." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "GenerationMixin import failure is unrelated to Qwen3Moe FineGrainedFP8Config errors." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "DeepSpeed zero3 Bert index error vs tokenizer-mapping add-new-model-like failure are different bugs." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Both are config-validation problems, but they affect different models and are not clearly the same regression." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different load-time failures: zero-sized tensor index error vs TOKENIZER_MAPPING_NAMES edge case." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "device_map/offload loading failure is unrelated to mmap-on-Strix-Halo OOM avoidance." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Exact same regression: Transformers v5 fills non-persistent buffers with junk." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "One is ZeRO-3 weight loading for language_model; the other is save_pretrained writing incorrect visual encoder keys." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Same model family, but one is an output correctness bug and the other is a documentation example/config/quantization issue." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Different loading/backward-compatibility problems: Qwen3_moe LoRA compatibility vs BitNet packed-weight unpacking." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and symptoms: GLM5 vs Qwen3.5 flash-attention generation bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The representative issue choice for 43644 over 44534 is well supported by the titles/timestamps, and the two accepted soft pairs are plausible same-bug duplicates. The rejected pairs are consistently framed as different bugs or different code paths without overclaiming." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 5, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44488", + "issue:44661", + "issue:44683", + "issue:44743", + "issue:44756", + "issue:44910", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45362", + "issue:45406", + "issue:45588" + ], + "soft_pairs": [ + "issue:44451|issue:44488", + "issue:43708|issue:44075", + "issue:43854|issue:43931", + "issue:43856|issue:43873", + "issue:42491|issue:45276", + "issue:41720|issue:42915", + "issue:43257|issue:43854", + "issue:39692|issue:43994", + "issue:44488|issue:44991", + "issue:43856|issue:44387", + "issue:45362|issue:45406", + "issue:42915|issue:44756", + "issue:43644|issue:43645", + "issue:44661|issue:44683", + "issue:42491|issue:43931", + "issue:44910|issue:45588", + "issue:43404|issue:43454", + "issue:42491|issue:43856", + "issue:45276|issue:45335", + "issue:36010|issue:39404", + "issue:44164|issue:44291", + "issue:44402|issue:44488", + "issue:41720|issue:44155", + "issue:43873|issue:44387", + "issue:39692|issue:44960", + "issue:43425|issue:43957", + "issue:44387|issue:44756", + "issue:43854|issue:45216", + "issue:44291|issue:44756", + "issue:44292|issue:44912", + "issue:42947|issue:43381", + "issue:45276|issue:45292", + "issue:44743|issue:44960", + "issue:43381|issue:44928", + "issue:43957|issue:44756", + "issue:44164|issue:44756", + "issue:43646|issue:43708" + ], + "prepared_review_unit_hash": "bf7f871294ff767ca3a302ab9e9626ce5155d2422a3611d7b3e1beb0628cab13", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 37, + "serialized_chars": 38104, + "estimated_input_tokens": 9526, + "estimated_eval_tokens": 19308 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab-bag of unrelated Transformers bug reports, with a few small duplicate families. The clearest duplicate set is around `resize_token_embeddings` not updating all tied embeddings, and there is also a likely shared lm_head weight-tying regression in multimodal generation models. No PRs are present.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45276 is the best canonical issue for the strongest duplicate family: it states the broader `resize_token_embeddings` propagation bug and subsumes the more specific follow-ups.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45276 is the most representative issue in the cluster because it is concrete, well-scoped, and has clear duplicate support from closely related reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are tokenizer/model load failures, but they target different checkpoints and describe separate report details; not enough evidence of the same underlying bug." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer resume/max_steps logic and SGD optimizer args are unrelated subsystems and failure modes." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different models and different loading errors; no shared concrete bug path is evident." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/efficiency, but one is MoE training memory usage and the other is quantization offloading behavior." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "These concern unrelated problems: Qwen3 MoE version compatibility vs Gemma4 token embedding resizing." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "One is an auto device-mapping CUDA assert, the other is a Qwen3 MoE LoRA version-compatibility issue." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Different code paths and symptoms; no duplicate relationship is supported." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but one is a docs/example mismatch and quantization issue while the other is bad inference output from AutoModel/pipeline." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Different models and different tokenizer-loading failures; too broad to treat as the same bug." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both are memory-related, but the causes differ: training inefficiency vs increased reserved memory under int4 quantization." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Different model families and different crash points (chat flow vs serve processor attribute error)." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility and Strix Halo mmap OOM are unrelated issues." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers 5.0 regressions, but one is junk non-persistent buffers and the other is notebook custom-model initialization." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Tokenizer-mapping registration failure and torch>=2.9 flex_attention compilation are unrelated." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "Version incompatibility for a Qwen3 MoE LoRA checkpoint is not the same bug as a Qwen3-VL shape mismatch." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both touch flash attention, but the root causes differ: misread 3D position_ids vs `s_aux=None` attribute handling." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": true, + "reason": "Both report missing lm_head weight tying causing broken generation in multimodal model classes; they appear to share the same fix path." + }, + { + "left": "issue:42491", + "right": "issue:43856", + "accept": false, + "reason": "These are unrelated Qwen3 MoE issues with different symptoms and code paths." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": true, + "reason": "Both describe `resize_token_embeddings` failing to propagate to nested/tied embeddings in related model families, which looks like the same underlying regression." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "Importing `GenerationMixin` and Whisper `return_language` pipeline behavior are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Save/load extra_state handling and `init_empty_weights` constructor argument handling are different loading-path bugs." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "One is tokenizer vocab-size mismatch, the other is a specific model loading failure; not the same bug." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "CUDA assert in Qwen3 auto device mapping and AudioFlamingo3 token/embedding leakage are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both involve quantization and memory, but one is offloading behavior and the other is int4 reserved-memory OOM." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "SigLIP2 example/model issues and GLM5 are entirely different reports." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 incompatibility and meta-device model loading failures are different compatibility problems." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention OOM, but one is int4 quantization reserved memory and the other is mmap on Strix Halo." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different model-loading regressions: GLM-4.7 unit tests vs Qwen3.5 save_pretrained checkpoint correctness." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "Unexpected `_is_hf_initialized` during meta init and Strix Halo mmap OOM are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Both are quantization-related, but they affect different models and different quantization paths, so they are not clearly the same bug." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Gradient checkpointing in PEFT/LoRA and gradient checkpointing in eval mode are different issues." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "These are effectively the same `resize_token_embeddings` regression, with 45276 phrased more broadly." + }, + { + "left": "issue:44743", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3.5 recurrent-state reset and GLM5 are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Gradient checkpointing eval-mode behavior and 3D position_ids causing SDPA fallback are unrelated." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading breakage and disabling mmap on Strix Halo are different problems." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "Save/from_pretrained extra_state handling is unrelated to mmap-related OOM." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization regressions and trainer checkpoint step math are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The summary stays within the packet, the resize_token_embeddings family is well supported by the strongest soft pairs, and the other verdicts avoid over-claiming duplicates where evidence is weak. No PRs are present in the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 6, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:32090", + "issue:36032", + "issue:39404", + "issue:40990", + "issue:42947", + "issue:43257", + "issue:43299", + "issue:43421", + "issue:43425", + "issue:43452", + "issue:43582", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43856", + "issue:43883", + "issue:43901", + "issue:43906", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44671", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44898", + "issue:44912", + "issue:44918", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45292", + "issue:45335" + ], + "soft_pairs": [ + "issue:44479|issue:44877", + "issue:45292|issue:45335", + "issue:44402|issue:44451", + "issue:43645|issue:44164", + "issue:43257|issue:44863", + "issue:43425|issue:43828", + "issue:43828|issue:43957", + "issue:44589|issue:44756", + "issue:43931|issue:44863", + "issue:43906|issue:44530", + "issue:43611|issue:43645", + "issue:43425|issue:43716", + "issue:39404|issue:44912", + "issue:44530|issue:45084", + "issue:43901|issue:44509", + "issue:43299|issue:43931", + "issue:44265|issue:44898", + "issue:44479|issue:44805", + "issue:43994|issue:44079", + "issue:44671|issue:44991", + "issue:43425|issue:44918", + "issue:43856|issue:44756", + "issue:43596|issue:44683", + "issue:44589|issue:45084", + "issue:39404|issue:44671", + "issue:40990|issue:42947", + "issue:36032|issue:43452", + "issue:44991|issue:45081", + "issue:43582|issue:43632", + "issue:32090|issue:39404", + "issue:43957|issue:44164", + "issue:40990|issue:43421", + "issue:44164|issue:44479", + "issue:43883|issue:44488", + "issue:44488|issue:45081" + ], + "prepared_review_unit_hash": "f48db81fd8dfa6cf288828f0a5c95cc80c3976bcf58b4726d4c67c6a6ae5dbc7", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 38227, + "estimated_input_tokens": 9557, + "estimated_eval_tokens": 19370 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated Transformers regressions; only one pair looks like a true duplicate/same bug. Most soft matches are too broad or target different code paths/models.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45292 is the broader, earlier report of resize_token_embeddings not propagating to output embeddings; issue:45335 looks like a specific t5gemma manifestation of the same bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45292 is the best representative issue because it states the general failure mode without narrowing to one model subclass.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Different regressions: video-input handling for Qwen VL models vs strict config blocking granite_speech loading." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same underlying bug: resize_token_embeddings fails to update downstream embeddings; 45335 is a t5gemma-specific instance." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "Both concern tokenizer loading, but the model symptoms and likely root causes differ; not clearly the same bug." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Custom model init in notebooks vs save/from_pretrained extra_state handling are different code paths." + }, + { + "left": "issue:43257", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures for different architectures; only broad subsystem overlap." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch version incompatibility is not the same as a Phi autocast dtype mismatch." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "One is autocast dtype mismatch; the other is meta-device loading breakage. Different failures." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup error vs Apple Silicon mmap OOM are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44863", + "accept": false, + "reason": "Different checkpoint-loading problems for different models; no clear shared bug." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Isolated reproduction of an older issue vs PagedAttentionCache group-type crash are different issues." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Base_model_prefix loading regression and Jupyter custom-model initialization regression are separate bugs." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "Torch compatibility issue vs Mistral-3 image dtype mismatch; not the same defect." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline language return regression vs quantization fallback issue are unrelated." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and template-node compilation error are different code paths." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs-related, but they cover different API/task deprecations." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3VL MoE loading regression and Qwen3-VL weight-shape mismatch are not the same concrete bug." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to Perceiver interpolate_pos_encoding behavior." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video input regression and mask-shape IndexError are different problems." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Incorrect SigLIP2 outputs vs ModelOutput key assignment bug are distinct." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction issue and tokenizer-loading failure are unrelated." + }, + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch version compatibility and Qwen3.5 embedding unpacking/trainer issue are different." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Training memory efficiency and mmap OOM avoidance are not the same bug." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Deepspeed zero3/BertModel index error is unrelated to compiled flex_attention on newer torch." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Different failure surfaces: storage type lookup vs template compilation." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper pipeline regression and CamemBERT prediction regression are unrelated." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "High perplexity on a model is not the same as ineffective gradient checkpointing with LoRA." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "T5 tokenizer special-token conflict and gguf_file loading breakage are different tokenizer/model-loading bugs." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer load failures, but they involve different models and different failure mechanisms." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "Apple Silicon allocator TypeError and _is_hf_initialized flag regression are unrelated." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper return_language pipeline regression." + }, + { + "left": "issue:43957", + "right": "issue:44164", + "accept": false, + "reason": "Meta-device loading breakage and extra_state save/from_pretrained handling are different issues." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity quality issue and runtime special-token post-processor update feature are unrelated." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "extra_state serialization and Qwen video-input regression are different bugs." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Missing tied-weights metadata in Molmo and tokenizer loading failure for sleng-bert are unrelated." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer-loading failures affecting different models and code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well-grounded in the packet and stays conservative. The one accepted duplicate pair (45292/45335) is plausibly the same resize_token_embeddings issue with a model-specific manifestation, and the other soft pairs are appropriately treated as distinct or only superficially related." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 7, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:42915", + "issue:42947", + "issue:43262", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43425", + "issue:43452", + "issue:43482", + "issue:43540", + "issue:43541", + "issue:43611", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43825", + "issue:43883", + "issue:43994", + "issue:44206", + "issue:44220", + "issue:44403", + "issue:44451", + "issue:44466", + "issue:44488", + "issue:44530", + "issue:44534", + "issue:44610", + "issue:44756", + "issue:44811", + "issue:44849", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45161", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45507", + "issue:45584" + ], + "soft_pairs": [ + "issue:43425|issue:45161", + "issue:43611|issue:43644", + "issue:44811|issue:45584", + "issue:44488|issue:44987", + "issue:44206|issue:44220", + "issue:43716|issue:45084", + "issue:42947|issue:44756", + "issue:34567|issue:40444", + "issue:43883|issue:44991", + "issue:43644|issue:43646", + "issue:43643|issue:44530", + "issue:44451|issue:44987", + "issue:43452|issue:43482", + "issue:43541|issue:45084", + "issue:36032|issue:42915", + "issue:43317|issue:43482", + "issue:42915|issue:44530", + "issue:39404|issue:43825", + "issue:43404|issue:44466", + "issue:44849|issue:45325", + "issue:36032|issue:43645", + "issue:44987|issue:44991", + "issue:43262|issue:43377", + "issue:45084|issue:45507", + "issue:43646|issue:44534", + "issue:43645|issue:44534", + "issue:43994|issue:44610", + "issue:45216|issue:45357", + "issue:43322|issue:43540", + "issue:44403|issue:45081", + "issue:44451|issue:45081" + ], + "prepared_review_unit_hash": "c205bbb24062553b876015592e5dca510175526c0c1b7562ba861f4cf91a1b7c", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36325, + "estimated_input_tokens": 9082, + "estimated_eval_tokens": 18420 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly unrelated bug reports with a few near-duplicates. The Qwen3.5 `save_pretrained` reports (45216/45357) are the strongest duplicate pair; the GGUF reports (43452/43482) are too weakly connected from titles alone to call duplicates confidently.", + "confidence": 0.66, + "canonical_issue_reason": "Issue 45216 is the best canonical issue because it describes the broader Qwen3.5 `save_pretrained` regression, while 45357 is a narrower follow-up about the same broken serialization path.", + "canonical_pr_reason": null, + "best_issue_reason": "45216 is the strongest anchor for the cluster: it states the general failure mode more broadly than the follow-up report and best represents the shared regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43425", + "right": "issue:45161", + "accept": false, + "reason": "Different problems: Torch version incompatibility vs MoE tensor-parallel loading failure." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Both are v5 regressions, but one is model loading with `base_model_prefix` and the other is non-persistent buffer serialization; not the same bug." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both involve Whisper special tokens, but one is `batch_decode()` behavior and the other is empty-transcription generation after `align_special_tokens`; too different to treat as duplicates." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Separate model-loading failures for different checkpoints; same broad subsystem only." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Both are audio-feature related, but the titles point to different failure points (`LasrFeatureExtractor` center arg vs `_torch_extract_fbank_features()`), so the duplication is not clear enough." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Unrelated: image preprocessor/model dtype mismatch vs template compilation error." + }, + { + "left": "issue:42947", + "right": "issue:44756", + "accept": false, + "reason": "Different bugs: ineffective gradient checkpointing vs mmap-induced OOM on Strix Halo." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "Trainer token-count tracking and multi-image iterable finetuning are unrelated." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Different models and different loading failures; no concrete shared code-path indicated." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 regressions around model handling, but one is junk buffers and the other is custom model initialization; not the same defect." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "Completely different areas: remote-code config fields vs Qwen3.5 paged-attention group parsing." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Both are model load failures, but for different models and likely different root causes." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": false, + "reason": "The titles both mention GGUF loading, but the packet does not establish the same regression or concrete code-path failure, so this is too weak to merge as duplicates." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Different failures: torch dynamo grouped_mm tracing vs template-node compilation." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "Tokenizer method-name conflict vs Qwen3Moe FP8 config failure; unrelated." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Different loading paths: device-map/offload dequantized model loading vs GGUF model loading." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Both mention Qwen MoE models, but the actual bugs are different (FP8 config vs invalid linear_attention group type)." + }, + { + "left": "issue:39404", + "right": "issue:43825", + "accept": false, + "reason": "Whisper `return_language` regression and generic pipeline translation-task messaging are not the same concrete bug." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both involve tied weights, but one is a missing tie in Mistral3 and the other is device-dependent serialization behavior." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen-family bugs on different models/code paths." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer method conflict vs custom model initialization breakage; unrelated." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Separate model-loading failures; insufficient evidence of a shared underlying defect." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Audio chat-template sampling-rate default vs MIMI batching/padding-mask outputs; different bugs." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Template compilation error vs GraniteMoEHybrid invalid method call; unrelated." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 issues, but one is custom model initialization and the other is non-persistent buffer corruption; not clearly the same bug." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Both are vision-related mismatches, but one is nonsensical model outputs and the other is a concrete processor size mismatch." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same Qwen3.5 `save_pretrained` serialization regression; 45357 is a narrower report of the same broken checkpoint save behavior." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Different models and failures: Llava Next segfault vs Qwen3OmniMoe video-processing ValueError." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Generic loading noise vs Mistral regex patch crash; not the same bug." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "Different model-loading failure and tokenizer-patch crash; no shared concrete issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary is mostly grounded, but one soft-edge accept looks too strong for the evidence available. The Qwen3.5 `save_pretrained` pair (45216/45357) is well supported, but the GGUF pair (43452/43482) is only linked by broad title similarity and a shared subsystem; the packet does not establish the same regression, so accepting it overstates the evidence. The rest of the rejects are conservatively phrased, and there are no PRs." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 8, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36032", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43296", + "issue:43317", + "issue:43377", + "issue:43404", + "issue:43421", + "issue:43425", + "issue:43482", + "issue:43531", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43742", + "issue:43818", + "issue:43819", + "issue:43856", + "issue:44062", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44479", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44829", + "issue:44877", + "issue:44987", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45084", + "issue:45127", + "issue:45325", + "issue:45357", + "issue:45479", + "issue:45538", + "issue:45542" + ], + "soft_pairs": [ + "issue:41553|issue:43582", + "issue:43377|issue:43819", + "issue:44220|issue:44683", + "issue:43632|issue:44756", + "issue:43404|issue:45127", + "issue:44265|issue:44610", + "issue:42915|issue:45084", + "issue:44987|issue:45081", + "issue:43296|issue:43531", + "issue:44479|issue:45325", + "issue:44062|issue:44987", + "issue:42915|issue:44291", + "issue:40990|issue:44811", + "issue:43611|issue:44534", + "issue:36032|issue:43742", + "issue:43317|issue:43856", + "issue:42175|issue:45542", + "issue:43632|issue:44534", + "issue:40444|issue:45325", + "issue:43531|issue:44877", + "issue:42915|issue:44589", + "issue:43421|issue:44568", + "issue:43818|issue:45357", + "issue:36032|issue:41720", + "issue:43425|issue:44403", + "issue:44683|issue:45020", + "issue:43482|issue:45538", + "issue:44829|issue:45479", + "issue:44403|issue:44991" + ], + "prepared_review_unit_hash": "0dfae2517b3208f1f971e1b4fb26f2c7085ba959a420ec3aafa9bc6d2c1b857d", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36489, + "estimated_input_tokens": 9123, + "estimated_eval_tokens": 18502 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated issues, with a few loosely similar transformer/tokenizer regressions but no clear duplicate pair. I would not merge any of the soft-edge candidates into one bug/change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue cleanly represents the cluster; the reports span many distinct code paths and model families.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Different failures in different subsystems: tokenizer loading error vs Apple Silicon allocator TypeError." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Both are model correctness bugs, but they affect unrelated components and code paths (MIMI padding mask vs DAC STE behavior)." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Audio feature extraction bug vs compiled flex_attention regression; no shared underlying fix." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Separate regressions: `_is_hf_initialized`/v5 init behavior vs mmap OOM on Strix Halo." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both involve weight tying behavior, but one is missing tying in Mistral3 while the other is LoRA merge collapse with extended vocab; not the same bug." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to the OmDet-Turbo processor image-size mismatch." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FP8 config issue and a template-node compilation error are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer/model loading problems, but they target different models and distinct code paths." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL load failure in vLLM is unrelated to Qwen3-MoE sliding-window behavior." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Both concern Qwen VL temporal handling, but one is a broad video-input regression and the other is a still-image position-id scaling bug." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken keyword conflict is unrelated to loading a specific model repo with newer transformers." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "FP8 model config failure and `_is_hf_initialized` init_empty_weights error are different regression classes." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity evaluation issue vs Whisper batch_decode special-token handling; no shared code path." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Base-model-prefix loading regression is unrelated to non-persistent buffer initialization." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Tokenizer method-name conflict vs MobileLLM key error; no same underlying bug." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "Dequantized model offload failure and Qwen3 MoE training memory usage are separate problems." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both mention TensorFlow packaging, but one is missing TF in the torch extra and the other is an undefined backend when only tensorboard is installed; not the same concrete bug." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Different transformer v5 regressions affecting different initialization/storage behaviors." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "IterableDataset multi-image prompt failure is unrelated to Qwen2.5-VL still-image temporal position scaling." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window bug and granite_speech config strictness are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "FP8 config failure vs missing Float8 storage type; related to float8 broadly but not the same bug." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both involve special tokens, but one is a backend update feature request and the other is a tokenizer regression; too different to be the same change." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA tower/temporal-attention problem is unrelated to Qwen3.5 visual-encoder save_pretrained key regression." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer name conflict vs CUDA assert during device-mapped Qwen3 loading; unrelated." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch version incompatibility and loading noise are not the same bug." + }, + { + "left": "issue:44683", + "right": "issue:45020", + "accept": false, + "reason": "Flex attention compilation regression is unrelated to remote_code loading breakage." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Qwen2.5-GGUF/v5 loading issue and CLIPTokenizer model_max_length are unrelated." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both mention degenerate classification behavior, but one is tied to flash_attention_3 while the other is a num_labels=1/problem_type loss issue." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Generic loading noise is unrelated to the est-roberta tokenizer loading failure." + } + ] + }, + "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": 13, + "review_unit_index": 9, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43262", + "issue:43296", + "issue:43377", + "issue:43454", + "issue:43482", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43701", + "issue:43746", + "issue:43749", + "issue:43792", + "issue:43828", + "issue:43881", + "issue:43883", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44265", + "issue:44291", + "issue:44387", + "issue:44466", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44617", + "issue:44683", + "issue:44898", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45381", + "issue:45491", + "issue:45563", + "issue:45584" + ], + "soft_pairs": [ + "issue:33666|issue:40444", + "issue:43792|issue:44589", + "issue:44079|issue:44534", + "issue:35707|issue:40444", + "issue:43377|issue:45491", + "issue:43482|issue:43525", + "issue:44291|issue:44589", + "issue:43749|issue:44079", + "issue:43644|issue:44079", + "issue:44079|issue:45072", + "issue:40444|issue:45381", + "issue:43632|issue:45072", + "issue:43881|issue:44488", + "issue:44265|issue:45072", + "issue:43531|issue:44291", + "issue:41720|issue:45081", + "issue:45325|issue:45381", + "issue:43611|issue:43646", + "issue:42175|issue:43296", + "issue:43454|issue:44466", + "issue:43994|issue:44898", + "issue:43828|issue:44683", + "issue:44387|issue:44617", + "issue:42915|issue:43482", + "issue:43526|issue:43883", + "issue:43792|issue:43957", + "issue:43701|issue:45325", + "issue:43262|issue:43746", + "issue:43632|issue:45563", + "issue:43881|issue:44991", + "issue:40990|issue:45584" + ], + "prepared_review_unit_hash": "cfea4018848479f06d5a136282f5ab5742ef8e6ff1a2d13a933374cda3a11ede", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36409, + "estimated_input_tokens": 9103, + "estimated_eval_tokens": 18462 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous set of unrelated issue reports, with only one clear duplicate-style pair around the v5 non-persistent-buffer regression. The rest share broad Transformer/v5 themes at best, but not the same underlying bug.", + "confidence": 0.57, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the strongest representative of the set: it states a concrete, version-specific regression and also matches the near-identical follow-up issue:44534.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Different models and failure modes; both are training/finetuning issues, not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Both are loading/runtime failures, but the errors and model contexts are different." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "Both are v5 regressions, but one is ModelOutput key handling and the other is buffer initialization." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Progressive generation with caches is unrelated to multi-image iterable finetuning." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both involve batching/padding, but they are distinct model-specific bugs with different symptoms." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "Separate model-loading failures with different root causes." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Init-empty-weights argument error is unrelated to the Float8 storage lookup failure." + }, + { + "left": "issue:43749", + "right": "issue:44079", + "accept": false, + "reason": "FSDP CPU RAM-efficient loading is a different path than ModelOutput key assignment." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer junk and ModelOutput key handling are unrelated regressions." + }, + { + "left": "issue:44079", + "right": "issue:45072", + "accept": false, + "reason": "Different subsystems: output container semantics vs bfloat16 dtype mismatch." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen2.5-VL multimodal bugs, one training data handling and one video position ids." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "_is_hf_initialized/init-loading regression is unrelated to dtype mismatches in inference." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Two model-loading failures, but no evidence they share the same underlying cause." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export/torch_compilable_check is unrelated to bfloat16 inference dtype issues." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior is unrelated to init_empty_weights argument handling." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to Mistral tokenizer patching." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL position-id bugs, but one is still-image temporal scaling and the other is video vision positions; not clearly the same defect." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both concern v5 custom-model compatibility, but the titles indicate different failure surfaces and not the same concrete bug." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Packaging/backend selection problem versus a model-load failure." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Tied-weight generation failure and serialization inconsistency are related thematically, but not the same underlying bug." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Different vision models and different positional/interpolation issues." + }, + { + "left": "issue:43828", + "right": "issue:44683", + "accept": false, + "reason": "Autocast dtype mismatch and compiled flex_attention failure are separate issues." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both mention OOM, but one is an int4 quantization memory regression and the other is a model-specific runtime OOM." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Different model families and unrelated loading failures." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "Image-processor label reduction bug versus missing tied-weights attribute." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "Generic run/load failure versus meta-device loading regression; not the same bug." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint-resume key mismatch is unrelated to Qwen2.5-VL position-id scaling." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting and GraniteSpeech PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:43632", + "right": "issue:45563", + "accept": false, + "reason": "_is_hf_initialized regression is unrelated to a stale warning in paged generate()." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Different model-loading issues for different models and tokenizers." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "Perplexity anomaly and Whisper empty-transcription generation failure are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 10, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:36331", + "issue:39692", + "issue:40444", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43296", + "issue:43316", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43454", + "issue:43526", + "issue:43550", + "issue:43577", + "issue:43606", + "issue:43708", + "issue:43720", + "issue:43749", + "issue:43792", + "issue:43819", + "issue:43873", + "issue:43881", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44423", + "issue:44466", + "issue:44479", + "issue:44534", + "issue:44610", + "issue:44671", + "issue:44683", + "issue:44743", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44918", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45325", + "issue:45357", + "issue:45406" + ], + "soft_pairs": [ + "issue:43317|issue:43873", + "issue:34634|issue:35707", + "issue:44291|issue:44877", + "issue:43606|issue:44265", + "issue:44186|issue:44265", + "issue:43606|issue:43819", + "issue:35707|issue:41720", + "issue:44610|issue:45325", + "issue:43454|issue:43550", + "issue:43377|issue:44898", + "issue:43454|issue:45072", + "issue:43377|issue:44610", + "issue:39692|issue:40444", + "issue:44466|issue:45357", + "issue:43749|issue:44534", + "issue:43792|issue:44683", + "issue:43296|issue:44991", + "issue:43296|issue:44877", + "issue:43454|issue:43720", + "issue:41720|issue:43322", + "issue:43404|issue:45357", + "issue:41720|issue:44977", + "issue:43526|issue:43881", + "issue:44423|issue:45406", + "issue:44186|issue:45072", + "issue:43873|issue:44756", + "issue:43577|issue:43708", + "issue:42617|issue:42915", + "issue:44877|issue:44991", + "issue:44479|issue:44918", + "issue:44610|issue:44898", + "issue:43550|issue:43720", + "issue:36331|issue:42915", + "issue:44186|issue:44610", + "issue:36331|issue:42617", + "issue:43316|issue:44913", + "issue:44671|issue:44743", + "issue:42915|issue:43317", + "issue:43550|issue:43606" + ], + "prepared_review_unit_hash": "d0542c4053d39ed37e52d81f28902ed72d8757045509a2b332feda2af72eb6f1", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 39476, + "estimated_input_tokens": 9869, + "estimated_eval_tokens": 19994 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is very heterogeneous: most items are distinct model-specific loading, training, or inference bugs, and none of the soft pairs look like true duplicates. The best available anchor is a broad loading/config issue, but only as a loose representative, not a merged duplicate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:44877 is the broadest loading/configuration bug in the set and has the most nearby soft links, so it is the least specific representative of this heterogeneous cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44877 is the best cluster anchor for triage because it is broad and central among the loading/configuration failures, though it is not a strong duplicate canon.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both involve offloading/quantization, but one is a specific dequantized-model load failure and the other is a general offloading regression; not the same bug." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice preset and progressive generation with inputs_embeds/past_key_values are unrelated code paths." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "One is init_empty_weights/_is_hf_initialized argument handling; the other is strict config loading for granite_speech. Different failures." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch in Bark-small is unrelated to torch.export failing on torch_compilable_check." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Tokenizer crashes on NER/padding are unrelated to export-time compilation failures." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Device mismatch during offload and DAC.from_latents mismatch with missing STE are different bugs." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with embeddings/cache and Qwen3 auto device-map cudaErrorAssert are different problems." + }, + { + "left": "issue:44610", + "right": "issue:45325", + "accept": false, + "reason": "Processor output size mismatch and Qwen2.5-VL rope index temporal scaling are distinct multimodal preprocessing bugs." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "lm_head weight tying in AyaVision and Bamba torch.compile SDPA failure are not the same issue." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "MIMI batching/padding-mask inconsistency and Perceiver non-default resolution failure are different model behaviors." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Weight tying bug and bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "MIMI batched-vs-single output mismatch and OmDet-Turbo processor/model size mismatch are different subsystems." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "Documentation example errors for SigLIP2 and Qwen2.5-VL multi-image IterableDataset failure are separate issues." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Serialization of tied lm_head weights depends on device, while the other is incorrect visual encoder key saving for Qwen3.5." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM-efficient loading and non-persistent buffer initialization junk are different failures." + }, + { + "left": "issue:43792", + "right": "issue:44683", + "accept": false, + "reason": "Whisper-large-v2 runtime failure and compiled flex_attention on torch>=2.9 are unrelated." + }, + { + "left": "issue:43296", + "right": "issue:44991", + "accept": false, + "reason": "PaddleOCR-VL loading in vLLM and tokenizer loading for EMBEDDIA/est-roberta are different model/load paths." + }, + { + "left": "issue:43296", + "right": "issue:44877", + "accept": false, + "reason": "vLLM PaddleOCR-VL load failure and granite_speech strict config loading are both load-related but not the same concrete bug." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "AyaVision lm_head tying and BitNet packed-weight unpacking during accelerate loading are different code paths." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 auto device mapping cudaErrorAssert and Llava Next segmentation fault are separate model-loading failures." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Mistral3 lm_head tying and Qwen3.5 visual encoder save_pretrained regression are unrelated serialization/tie bugs." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3 auto device mapping crash and Qwen3.5 flash-attention generation failure are distinct regressions." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BEiT Fast label reduction bug and glm-4v-9b loading failure are unrelated." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Continuous-batching multimodal crash due to .to() on strings and Gemma4Processor missing _tokenizer are different serve bugs." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "NER/padding tokenizer crash and dtype mismatch in inference are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "General offloading/quantization problem and Strix Halo mmap OOM mitigation are not the same bug." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Blip2 dtype propagation and Trainer resume_from_checkpoint max_steps calculation are unrelated." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel.py startup failure and Qwen3Moe FineGrainedFP8Config failure are different issues." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config blocking granite_speech and tokenizer loading failure for est-roberta are different load-time problems." + }, + { + "left": "issue:44479", + "right": "issue:44918", + "accept": false, + "reason": "Video-input regression in Qwen models and TRL unpacking of Qwen3.5 input embeddings are different failures." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "OmDet-Turbo processor/model resolution mismatch and Perceiver interpolation failure are separate vision preprocessing bugs." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba and packed-weight unpacking failure in BitNet are not the same concrete bug." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "CustomTrainer compute_loss signature change and Qwen3Moe FP8 loading failure are unrelated." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Tokenizer padding/truncation crash and processor output-size mismatch are different model/preprocessor bugs." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer compute_loss keyword regression and 3d_parallel.py execution failure are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and GPTNeoXConfig rotary_pct reload regression are different config issues." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM regression and Qwen3.5 recurrent-state cache reset bug are unrelated model behaviors." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Qwen3Moe FP8 loading failure and dequantized model offload failure are different problems." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Bamba torch.compile SDPA failure and Bark-small CPU-offload device mismatch are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 11, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30064", + "issue:33357", + "issue:34567", + "issue:34634", + "issue:35707", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42617", + "issue:42886", + "issue:42915", + "issue:43322", + "issue:43388", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43873", + "issue:43927", + "issue:44062", + "issue:44186", + "issue:44292", + "issue:44361", + "issue:44368", + "issue:44451", + "issue:44479", + "issue:44492", + "issue:44509", + "issue:44514", + "issue:44545", + "issue:44625", + "issue:44843", + "issue:44898", + "issue:44991", + "issue:45059", + "issue:45072", + "issue:45161", + "issue:45357", + "issue:45381", + "issue:45406", + "issue:45464", + "issue:45479" + ], + "soft_pairs": [ + "issue:43531|issue:43572", + "issue:41720|issue:42617", + "issue:44292|issue:44479", + "issue:41720|issue:45161", + "issue:43540|issue:43782", + "issue:43388|issue:45059", + "issue:44186|issue:44361", + "issue:41720|issue:43873", + "issue:42915|issue:43322", + "issue:36010|issue:42617", + "issue:45406|issue:45464", + "issue:36331|issue:41720", + "issue:44514|issue:44545", + "issue:33357|issue:35707", + "issue:41720|issue:43746", + "issue:30064|issue:34634", + "issue:44492|issue:44509", + "issue:43606|issue:43720", + "issue:43550|issue:43653", + "issue:44368|issue:45357", + "issue:44625|issue:45479", + "issue:43421|issue:43927", + "issue:43606|issue:43653", + "issue:43782|issue:45381", + "issue:34567|issue:35707", + "issue:34634|issue:41720", + "issue:42886|issue:44843", + "issue:44898|issue:45072", + "issue:43421|issue:43653", + "issue:44451|issue:44991", + "issue:44062|issue:44451" + ], + "prepared_review_unit_hash": "45fb029a16afca766af9c5707ea2328dc3e33fe137caedf1988a31d4facd0a5d", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37035, + "estimated_input_tokens": 9259, + "estimated_eval_tokens": 18774 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue reports. The only clear duplicate pairs here are the batched `Qwen2_5_VLProcessor.apply_chat_template(..., padding=False)` crash and the HF_HUB_OFFLINE tokenizer-load regression; the rest look like different bugs in different code paths or model families.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 42886 is the cleanest canonical report for the offline-tokenizer regression: it states the user-facing failure, while 44843 provides the root-cause detail (`model_info()` being called in offline mode).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 42886 is the best representative issue in this set because it is a clear, broadly scoped bug report with a concrete failure mode and version/environment context.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Different models and different failures: Qwen3-MoE `sliding_window` vs StableLmConfig `pad_token_idx`." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Unrelated symptoms and subsystems; one is Qwen3 auto-device-map CUDA assert, the other is a 3d_parallel.py runtime failure." + }, + { + "left": "issue:44292", + "right": "issue:44479", + "accept": false, + "reason": "Different bug classes: Qwen-3-8B-NVFP4 runtime error vs Qwen VL video-input regression." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Both involve model execution, but they are different concrete failures and code paths (CUDA assert vs TP-only MoE issue)." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Both are Qwen multimodal reports, but one is a video-processing ValueError and the other is a `from_pretrained` weight-loading error." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Different pipelines: Trainer metric gathering/label dropping vs SAM3 text+box prompt behavior." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer bugs in different models; no shared failing code path is evident." + }, + { + "left": "issue:41720", + "right": "issue:43873", + "accept": false, + "reason": "Qwen3 device-mapping CUDA assert is unrelated to quantization offloading behavior." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Different failures on different models: FineGrainedFP8Config vs Llava Next loading segfault." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "ImportError for `GenerationMixin` is unrelated to a 3d_parallel.py execution problem." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Different components and symptoms: Gemma4Processor server crash vs Qwen3.5 streaming inference API failure." + }, + { + "left": "issue:36331", + "right": "issue:41720", + "accept": false, + "reason": "Trainer API signature change vs model-device-mapping CUDA assert; not the same bug." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same `Qwen2_5_VLProcessor.apply_chat_template(..., padding=False)` crash with effectively identical title and symptom." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS CLIP bus error and progressive generation with `inputs_embeds`/`past_key_values` are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Different loading/runtime failures: Qwen3 CUDA assert vs GraniteSpeech local checkpoint/PEFT loading." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Image segmentation-map processing and BarkProcessor voice preset handling are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategies is not the same as outdated docs for removed pipeline tasks." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Different model-specific loading/offload failures with different root causes." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "torch.compile/SDPA failure for Bamba is unrelated to BigBirdTokenizer special-token registration." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "Warning about `tie_word_embeddings` during LoRA fine-tuning is not the same as incorrect visual-encoder keys in `save_pretrained`." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Both mention classification config, but one is Qwen3.5 config propagation and the other is a broader zero-loss training bug with `num_labels=1`." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Runtime special-token/post-processor update issue is different from config token-ID loss after save/load." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Different bug families: CPU offload device mismatch vs tokenizer decode/special-token registration." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen VL issues: `weight_only=True` loading error vs video `vision_position_ids` regression." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "Trainer state token-count tracking is unrelated to progressive generation cache behavior." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice preset bug is unrelated to Qwen3 auto device mapping." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same HF_HUB_OFFLINE tokenizer-load regression; 44843 pinpoints the unconditional `model_info()` call in `_patch_mistral_regex`." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Different failures in different models: Perceiver positional interpolation vs dtype mismatch in bfloat16 inference." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading complaints, but they involve different model repos and likely different compatibility issues." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "`AddedToken(..., special=...)` argument duplication is unrelated to loading ScandiBERT." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative. The two accepted pairs are supported by near-identical or directly complementary titles, and the rejected pairs are phrased cautiously without overclaiming shared root cause. No clear evidence of unsupported duplication claims." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 12, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:39692", + "issue:40444", + "issue:43335", + "issue:43381", + "issue:43421", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43618", + "issue:43653", + "issue:43697", + "issue:43716", + "issue:43720", + "issue:43746", + "issue:43756", + "issue:43819", + "issue:43844", + "issue:43873", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44077", + "issue:44186", + "issue:44220", + "issue:44488", + "issue:44589", + "issue:44610", + "issue:44625", + "issue:44661", + "issue:44671", + "issue:44811", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45183", + "issue:45356", + "issue:45381", + "issue:45479" + ], + "soft_pairs": [ + "issue:44077|issue:44625", + "issue:44987|issue:45042", + "issue:44918|issue:44936", + "issue:43531|issue:44589", + "issue:43381|issue:43844", + "issue:43994|issue:44671", + "issue:43421|issue:44186", + "issue:39692|issue:44610", + "issue:43525|issue:43531", + "issue:45081|issue:45356", + "issue:43596|issue:44220", + "issue:45183|issue:45479", + "issue:43746|issue:45381", + "issue:39692|issue:43873", + "issue:43540|issue:44008", + "issue:43618|issue:43697", + "issue:40444|issue:43746", + "issue:43746|issue:43873", + "issue:44220|issue:44661", + "issue:43716|issue:44008", + "issue:43335|issue:43756", + "issue:43720|issue:43819", + "issue:43653|issue:43720", + "issue:30064|issue:36010", + "issue:43421|issue:44811", + "issue:44062|issue:44488", + "issue:44843|issue:45081", + "issue:44186|issue:44898", + "issue:43540|issue:45381", + "issue:43550|issue:45072" + ], + "prepared_review_unit_hash": "c8f89df0e08377584f068d2578d779bd1b23025eaa8b680682c3d6a7cfe5ccff", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36406, + "estimated_input_tokens": 9102, + "estimated_eval_tokens": 18460 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a mixed bag of Transformers bug reports across tokenizers, multimodal processors, model configs, and training/runtime regressions; the soft pairs mostly share superficial subsystem overlap rather than the same underlying defect.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 43421 is the cleanest central representative of the tokenizer/runtime-regression subset, but the overall cluster is too heterogeneous to treat as one true duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43421 is the strongest representative because it is concrete and reproducible, and several nearby items are tokenizer-adjacent even though they are not duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "Different models and failures: patchtsmixer post-init policy versus Qwen3.5 label propagation." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Unrelated problems: loading a specific model checkpoint versus PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Different trainer issues: unpacking Qwen3.5 embeddings with TRL versus evaluate() failing after train()." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Separate bugs in different paths: Qwen3-MoE sliding_window handling versus float8 storage lookup." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Both are training-related, but one is eval-mode gradient checkpointing and the other is ZeRO-3 gradient growth." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different model-specific inference regressions: SigLIP2 nonsense outputs versus CamemBERT MLM predictions." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Tokenizer-related, but not the same bug: runtime special-token post-processing versus LayoutLMv2 NER/padding crashes." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "Both concern vision docs/processors, but one is a SigLIP2 example and the other is OmDet-Turbo input sizing." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Config attribute error in Llama4 versus sliding-window behavior in Qwen3-MoE." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Related tokenizer area, but different defects: backend_tokenizer crash versus Kimi-K2.5 codec/regression warning issues." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Distinct initialization paths: ZeRO-3/BertModel indexing versus audio fbank feature extraction." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Validation bug versus loss formulation bug; they affect different code paths and symptoms." + }, + { + "left": "issue:43746", + "right": "issue:45381", + "accept": false, + "reason": "Different domains: PEFT checkpoint loading for GraniteSpeech versus Qwen2.5-VL video position IDs." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "Doc/example issue versus a general offloading-with-quantization runtime issue." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Separate multimodal bugs: Qwen3OmniMoe video processing versus Gemma 3n audio_mel_mask name collision." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Different model regressions with different outputs and surfaces; not the same underlying defect." + }, + { + "left": "issue:40444", + "right": "issue:43746", + "accept": false, + "reason": "Different loading/training problems affecting unrelated models and workflows." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "PEFT checkpoint loading is unrelated to quantization/offloading behavior." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction issue versus tokenizer-mapping/model registration failure." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Different multimodal mismatch bugs: image preprocessor dtype versus audio tensor attribute collision." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "Different model config/content bugs: sparse-layer creation versus Smollm3 RoPE-layer omission." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "Different quantization issues: accelerate unpacking of packed weights versus DAC STE mismatch." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "Tokenizer special-token registration versus BitNet packed-weight loading are unrelated failures." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Image segmentation-map processing has nothing to do with the GenerationMixin import error." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Both touch tokenization, but one is runtime special-token updates and the other is Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken construction error versus model loading failure for cjvt/sleng-bert." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Same patch family, but one is offline hub access in model_info() and the other is a backend_tokenizer attribute crash." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "Different vision/tokenizer bugs: LayoutLMv2 NER padding versus Perceiver image-size handling." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal, but video-processing failures in different Qwen variants are not the same bug." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Different low-level issues: torch.compile with SDPA versus bfloat16 dtype mismatches." + } + ] + }, + "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": 13, + "review_unit_index": 13, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30064", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:41720", + "issue:42175", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43643", + "issue:43644", + "issue:43650", + "issue:43653", + "issue:43720", + "issue:43749", + "issue:43782", + "issue:43818", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:43927", + "issue:43994", + "issue:44060", + "issue:44186", + "issue:44291", + "issue:44368", + "issue:44442", + "issue:44448", + "issue:44485", + "issue:44509", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44821", + "issue:44877", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45325", + "issue:45406" + ], + "soft_pairs": [ + "issue:43577|issue:43818", + "issue:43643|issue:44877", + "issue:43550|issue:43927", + "issue:43531|issue:43643", + "issue:43782|issue:44977", + "issue:43644|issue:43749", + "issue:44448|issue:44671", + "issue:43577|issue:43582", + "issue:44561|issue:45020", + "issue:43653|issue:43819", + "issue:43825|issue:43994", + "issue:43550|issue:43819", + "issue:44821|issue:45084", + "issue:43825|issue:43827", + "issue:44186|issue:44442", + "issue:43650|issue:43906", + "issue:41720|issue:44991", + "issue:44805|issue:44936", + "issue:44485|issue:44509", + "issue:30064|issue:36331", + "issue:43606|issue:45072", + "issue:42175|issue:45081", + "issue:43782|issue:45325", + "issue:44610|issue:45072", + "issue:34634|issue:36010", + "issue:44060|issue:44368", + "issue:45081|issue:45406", + "issue:43720|issue:45072", + "issue:43653|issue:44671", + "issue:43749|issue:44291" + ], + "prepared_review_unit_hash": "cc8f72d4fb4f7e60c29ff24d8495872414bb925574834cdc5e2f2b1e854f49c4", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36878, + "estimated_input_tokens": 9220, + "estimated_eval_tokens": 18696 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is highly heterogeneous: it spans unrelated bugs in processors, tokenizers, model loading, dtype/device mismatches, docs, and training behavior. None of the soft pairs look like the same concrete underlying bug or change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue 45020 is the broadest umbrella-like report here (recent remote_code regressions), but the overall cluster is too diverse to collapse into a single duplicate issue.", + "canonical_pr_reason": null, + "best_issue_reason": "45020 is the best representative issue by scope, but it is still only a loose umbrella and not a true duplicate target for most of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "Different failures: BLIP2 dtype casting vs Video-LLaVA missing temporal attention/weight sharing." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both involve config loading, but one is trust_remote_code missing fields and the other is strict config blocking granite_speech; different bugs." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to DiaConfig save/load token ID loss." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to trust_remote_code config field loss." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Different Qwen model/runtime issues: weight_only load error vs flash-attention generation failure." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer junk in v5 is a different problem from FSDP CPU RAM efficient loading." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Both are v5 regressions, but Pegasus output drift and CamemBERT masked-LM predictions are separate model-specific bugs." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "BLIP2 dtype mismatch is not the same as Apple Silicon caching_allocator_warmup TypeError." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Removal of is_torch_fx_available is one specific remote_code breakage; 45020 is a broader umbrella report." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "BigBirdTokenizer special-token decode bug is unrelated to DAC.from_latents/STE mismatch." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message text in v5 is unrelated to SigLIP2 nonsensical outputs." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Bamba torch.compile with SDPA and DAC latent reconstruction are different code paths and symptoms." + }, + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "AutoImageProcessor URL loading and non-template node compilation are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Both touch v5 pipeline removals, but one is runtime error messaging and the other is stale docs; not the same concrete bug." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 NER/padding crash and FastSpeech2ConformerTokenizer loading failure are separate tokenizer issues." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "The placeholder 'ADD THE DATA' issue is unrelated to the isolated reproduction of issue 38071." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "Qwen3 auto device mapping cuda assert and EMBEDDIA tokenizer loading failure are unrelated." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask/index shape mismatch is not the same as trainer.evaluate() failing after train()." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE discussion is unrelated to docs still mentioning removed pipeline tasks." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "ImageProcessor void segmentation maps and CustomTrainer.compute_loss signature mismatch are different bugs." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch in bark-small is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "Packaging/backend dependency issue is unrelated to Mistral tokenizer regex patch crash." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different VL bugs: Qwen3VL weight_only load error versus Qwen2.5-VL RoPE position scaling." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "OmDet-Turbo processor size mismatch and bfloat16 dtype mismatches are unrelated." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset and GenerationMixin import error are unrelated issues." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-Next tied-weight warning and Qwen3.5 tie_word_embeddings warning are similar only at a high level, but not the same bug." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Mistral tokenizer regex crash and Gemma4Processor _tokenizer missing are different processor/tokenizer failures." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "BitNet packed-weight unpacking during accelerate loading is unrelated to dtype mismatches in inference." + }, + { + "left": "issue:43653", + "right": "issue:44671", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and CamemBERT masked-LM output issues are different tokenizer/model bugs." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and init_empty_weights unexpected _is_hf_initialized argument are different loading-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": 2, + "component_count": 13, + "review_unit_index": 14, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:34567", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:39692", + "issue:40990", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43452", + "issue:43454", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43638", + "issue:43645", + "issue:43653", + "issue:43696", + "issue:43697", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43756", + "issue:43819", + "issue:43854", + "issue:43927", + "issue:43931", + "issue:44479", + "issue:44554", + "issue:44568", + "issue:44589", + "issue:44610", + "issue:44756", + "issue:44811", + "issue:44857", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45084", + "issue:45356", + "issue:45488", + "issue:45538", + "issue:45600" + ], + "soft_pairs": [ + "issue:43526|issue:43697", + "issue:43582|issue:44857", + "issue:43262|issue:43819", + "issue:43526|issue:44554", + "issue:43329|issue:45600", + "issue:45084|issue:45538", + "issue:43653|issue:44912", + "issue:34634|issue:36331", + "issue:44568|issue:45488", + "issue:43454|issue:43577", + "issue:43452|issue:44991", + "issue:43638|issue:44936", + "issue:43257|issue:43329", + "issue:43746|issue:44811", + "issue:43854|issue:44977", + "issue:44589|issue:45042", + "issue:38175|issue:44610", + "issue:39692|issue:44912", + "issue:43257|issue:43696", + "issue:36010|issue:43645", + "issue:44568|issue:45356", + "issue:43819|issue:43927", + "issue:43550|issue:44912", + "issue:43720|issue:44610", + "issue:43749|issue:44756", + "issue:43262|issue:44811", + "issue:43531|issue:45042", + "issue:43329|issue:44479", + "issue:43257|issue:43931", + "issue:36010|issue:36331", + "issue:43257|issue:43756", + "issue:34567|issue:40990" + ], + "prepared_review_unit_hash": "14f853e63b62b864b1a931307eff15bbb05e40759ae9d8c8e59b886a96891cd0", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37477, + "estimated_input_tokens": 9370, + "estimated_eval_tokens": 18996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated closed/open issue reports across different Transformers subsystems; no PRs are present, and none of the suggested soft pairs look like true duplicates of the same underlying bug.", + "confidence": 0.24, + "canonical_issue_reason": "There is no strong single canonical because the set is heterogeneous. If a fallback anchor is needed, the broad tokenizer regression in issue 45488 is the clearest, most reusable bug report.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45488 is the most triage-useful fallback: it describes a concrete, user-visible regression with a clear failure mode and broad impact on a major tokenizer family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "Both are vision-model regressions, but one is a BeitImageProcessor label-reduction bug and the other is an RTDetrV2 output-drift report; different code paths." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Unrelated failures: Apple Silicon allocator warmup TypeError vs. CUDA AMP loss crash in LwDetrImageLoss." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio processor chat-template sampling-rate bug vs. DAC latent/STE mismatch; different components and symptoms." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "BeitImageProcessorFast label handling and an MPS attention correctness issue are not the same underlying bug." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Video multimodal token counting bug vs. auto_mappings CI breakage; different files and failure modes." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Compile/template-node error is unrelated to CLIPTokenizer's model_max_length behavior." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "BigBird special-token registration bug vs. gpt-oss MXFP4 load fallback; no shared code path." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "BarkProcessor voice_preset issue and CustomTrainer.compute_loss signature breakage are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both are tokenizer regressions, but one is missing BOS/EOS insertion and the other overrides tokenizer.json pre-tokenizer settings; not the same concrete bug." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "AyaVision lm_head tying bug and BLIP2 dtype propagation bug affect different model-loading paths." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "GGUF/from_pretrained breakage and est-roberta tokenizer loading failure are distinct issues." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO3 index error for a non-pretrained Bert model is unrelated to evaluate() failing after train()." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate+deepspeed is unrelated to the multimodal video token-counting bug." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT/local-checkpoint loading problem and Whisper batch_decode skip_special_tokens behavior are different subsystems." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "GLM-4.7-Flash load failure in tests vs. Qwen3.5 flash-attention generation regression; separate model-specific bugs." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup error and PIL image-processor torchvision dependency regression are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "SigLIP2 zero-probabilities and OmDet-Turbo input-size mismatch are both vision issues but not the same bug." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Documentation example/model mismatch and gpt-oss MXFP4 loading are different failures." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3 MoE weight conversion and CUDA OOM on GPT-oss-20b are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "ImportError for GenerationMixin and Jupyter custom-model initialization breakage are not the same issue." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but the failure modes are different: missing special tokens vs. codec/regex handling regression." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC forward/from_latents mismatch and DiaConfig token-id persistence loss are separate model/config bugs." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "SDPA/torch.compile failure in Bamba is unrelated to quantization fallback in gpt-oss-20b." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight loading bug in BitNet and OmDet-Turbo processor size mismatch are different code paths." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "FSDP CPU RAM efficient loading breakage and Strix Halo mmap OOM are both memory/loading themed but not the same concrete bug." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Both touch audio processors, but one is chat-template sampling rate and the other is batch_decode special-token handling." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding-window issue and PIL backend torchvision dependency regression are unrelated." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "Both mention video/multimodal regressions, but one is undefined helper usage in token counting and the other is a broader model-specific video-input regression; not enough to treat as the same bug." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3 MoE weight conversion bug and Qwen3-VL-30B-A3B weight-shape mismatch are distinct model-loading failures." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "GenerationMixin import failure and CustomTrainer compute_loss signature incompatibility are unrelated API breaks." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MoE loading/conversion bug and Smollm3 RoPE-layer mismatch concern different models and mechanisms." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "TrainerState token-count tracking and GPT-oss perplexity are not the same underlying problem." + } + ] + }, + "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": 13, + "review_unit_index": 15, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:28282", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:36010", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43317", + "issue:43329", + "issue:43450", + "issue:43452", + "issue:43526", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43643", + "issue:43646", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43824", + "issue:43906", + "issue:43927", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44560", + "issue:44737", + "issue:44821", + "issue:44898", + "issue:44912", + "issue:44938", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45072", + "issue:45137", + "issue:45357", + "issue:45375", + "issue:45381", + "issue:45405", + "issue:45479", + "issue:45538", + "issue:45588" + ], + "soft_pairs": [ + "issue:43329|issue:43450", + "issue:43697|issue:45357", + "issue:43577|issue:45072", + "issue:41720|issue:43540", + "issue:44479|issue:45381", + "issue:44987|issue:45538", + "issue:43452|issue:43582", + "issue:42175|issue:44938", + "issue:44291|issue:45042", + "issue:43317|issue:45538", + "issue:43927|issue:44912", + "issue:33357|issue:34567", + "issue:42915|issue:43742", + "issue:43540|issue:44560", + "issue:43540|issue:44479", + "issue:43906|issue:45381", + "issue:42915|issue:45538", + "issue:42175|issue:43824", + "issue:42915|issue:44821", + "issue:43526|issue:45479", + "issue:44062|issue:44991", + "issue:42175|issue:44488", + "issue:44737|issue:45588", + "issue:45137|issue:45405", + "issue:38175|issue:44898", + "issue:28282|issue:34689", + "issue:43646|issue:43749", + "issue:36010|issue:42175", + "issue:43643|issue:45375", + "issue:43582|issue:43782", + "issue:43742|issue:44291" + ], + "prepared_review_unit_hash": "f0302386fd268a03a070db91126c50832351480e62d8bd79021dbb840e7f3b06", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37136, + "estimated_input_tokens": 9284, + "estimated_eval_tokens": 18824 + }, + "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 around the Qwen video-input regression. Most other soft pairs share only broad subsystem similarity and should stay separate.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 44479 is the broadest and most central report in the nearest video-regression subcluster, covering the 5.3.0 Qwen video-input breakage across multiple models.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44479 best represents the cluster\u2019s strongest duplicate candidate because it states the shared regression at a higher level than the narrower follow-up reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Different bugs: undefined video token-count helpers vs batched video processor shape handling." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "Different failures: RTDetrV2 output drift vs save/load visual-encoder key serialization." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype, but they affect different models and code paths." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Qwen3 auto device-map CUDA assert is not the same as Qwen3OmniMoe video-input ValueError." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Both describe the same Qwen video-input regression in v5.3.0; 45381 looks like a specific symptom of the broader bug." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Generic model-loading regression vs a tokenizer model_max_length bug." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "GGUF loading failure and Apple Silicon warmup TypeError are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Packaging/backend inclusion issue vs Python 3.14 load failure." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Unexpected init_empty_weights argument and torchvision requirement regression are different bugs." + }, + { + "left": "issue:43317", + "right": "issue:45538", + "accept": false, + "reason": "Dequantized model offload failure is unrelated to CLIPTokenizer max_length." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "Config field loss on save/load and MXFP4 quantization fallback are different issues." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS bus error on CLIP loading is unrelated to TrainerState token counter updates." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure and MobileLLM key error are distinct loading problems." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-related, but the models and failure modes differ (ValueError vs StopIteration)." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError is not the same concrete regression as the broader Qwen 5.3.0 video-input bug." + }, + { + "left": "issue:43906", + "right": "issue:45381", + "accept": false, + "reason": "A reproduction of another issue is not the same bug as the Qwen video regression." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "TensorFlow backend packaging and missing Qwen2_5_VL export are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "FP8 config failure is unrelated to loading AutoImageProcessor from URL." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and sequence-classification zero-loss bug are different code paths." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "AddedToken special-arg conflict is unrelated to tokenizer loading for est-roberta." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "Backend install issue vs model loading failure for cjvt/sleng-bert." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "CPU arange placement bug and flash_attention s_aux=None crash are unrelated." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error is unrelated to a PEFT version bump." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Unexpected zero probabilities and Perceiver non-default-resolution failure are different vision bugs." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Missing PyTorch ImportError and Llama 3.2 Vision model loading break are not the same issue." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model init breakage and FSDP CPU RAM efficient loading are distinct failures." + }, + { + "left": "issue:36010", + "right": "issue:42175", + "accept": false, + "reason": "GenerationMixin import failure is unrelated to TensorFlow backend packaging." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Missing fields after trust_remote_code load and dropped deepstack_visual_indexes are different config-serialization bugs." + }, + { + "left": "issue:43582", + "right": "issue:43782", + "accept": false, + "reason": "Apple Silicon warmup TypeError and Qwen3VL weight_only load error are unrelated." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "MobileLLM key error and init_empty_weights argument error are different loading issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: most pairs are clearly cross-subsystem or cross-code-path matches, and the one accepted pair (44479/45381) is a conservative duplicate call within the same Qwen video-input regression family. The canonical/best-issue rationale is also appropriately cautious given the titles and cluster structure." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 16, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:34567", + "issue:36331", + "issue:41762", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43531", + "issue:43577", + "issue:43643", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43819", + "issue:43994", + "issue:44062", + "issue:44291", + "issue:44484", + "issue:44488", + "issue:44514", + "issue:44561", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44704", + "issue:44792", + "issue:44811", + "issue:44821", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:44936", + "issue:44964", + "issue:45042", + "issue:45081", + "issue:45216", + "issue:45290", + "issue:45325", + "issue:45356", + "issue:45381", + "issue:45405", + "issue:45440", + "issue:45600" + ], + "soft_pairs": [ + "issue:43756|issue:45325", + "issue:28282|issue:29127", + "issue:44484|issue:44704", + "issue:43493|issue:43994", + "issue:34567|issue:36331", + "issue:44291|issue:44821", + "issue:44062|issue:45081", + "issue:43452|issue:44488", + "issue:44792|issue:44964", + "issue:44514|issue:45290", + "issue:45405|issue:45600", + "issue:43643|issue:44561", + "issue:43688|issue:43749", + "issue:44514|issue:45381", + "issue:43697|issue:45216", + "issue:43742|issue:44589", + "issue:45042|issue:45081", + "issue:43531|issue:44821", + "issue:43452|issue:43577", + "issue:44625|issue:44877", + "issue:43452|issue:44843", + "issue:43688|issue:45440", + "issue:43653|issue:44568", + "issue:43720|issue:44898", + "issue:41762|issue:44936", + "issue:43295|issue:45356", + "issue:43819|issue:44811" + ], + "prepared_review_unit_hash": "00d8efedb29f76e158422f41b3ed228efe7c09dd2cd2e67981a2bddb8e9666c9", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36184, + "estimated_input_tokens": 9046, + "estimated_eval_tokens": 18348 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Only one pair looks like a true duplicate: the two SigLIP2 issues describing the same model-implementation mismatch. The rest are either different models, different code paths, or only loosely related symptoms.", + "confidence": 0.89, + "canonical_issue_reason": "issue:43493 is the broader root-cause report about SigLIP2 diverging from the original JAX implementation; issue:43994 is a downstream symptom on a specific checkpoint/pipeline path.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43493 is the best representative issue because it states the underlying bug more generally and is more useful as the canonical tracker.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Both mention RoPE/position handling, but they concern different model families and different mechanisms." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "One is a missing PyTorch dependency error; the other is a LayoutLMv3 validation/error-message clarity issue." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "Unrelated: save_pretrained shard sizing vs AutoProcessor kwargs forwarding." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Same underlying SigLIP2 bug: HF outputs diverge from the reference implementation and manifest as bad results on the named checkpoint/pipeline." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token counting and CustomTrainer.compute_loss signature are separate trainer issues." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Different loading failures: init_empty_weights/_is_hf_initialized versus AutoImageProcessor URL loading." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer bugs with different failure modes and code paths." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but one is gguf_file/AutoTokenizer-AutoModelForCausalLM behavior and the other is a specific model that fails to load for a different reason." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal models and different failure points; one is a failed test, the other is model loading." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both are chat-template crashes, but under different inputs and triggers, so not the same concrete bug." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "Release/versioning issue vs CI break from removed config references; unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Different trust_remote_code regressions affecting different APIs and failure modes." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are unrelated." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen2.5-VL problems: chat templating crash vs incorrect vision_position_ids." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "Different models and different regressions: RTDetrV2 output drift vs Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "KeyError when loading MobileLLM-125M is unrelated to missing Float8 storage type." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "PIL backend torchvision dependency and Mistral regex backend_tokenizer crash are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE has nothing to do with loading AutoImageProcessor from URL." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading breakage and BLIP2 dtype propagation are different bugs." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "num_labels propagation and strict config loading are not the same underlying issue." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Different tokenizer/processor loading paths; offline model_info bug is unrelated to gguf_file breakage." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both are model correctness issues, but they affect different architectures and subsystems." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Different tokenizer special-token problems with different tokenizers and symptoms." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "BitNet accelerate-loading unpacking bug and Perceiver image classification failure are unrelated." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO-3 loading failure and trainer.evaluate() after train() failure are separate bugs." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor/tokenizer API regression and Kimi-K2.5 codec/mistral-regex regression are different issues." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC latent decoding mismatch and Whisper batch_decode skip_special_tokens behavior are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted pair, issue:43493 vs issue:43994, is plausibly the same SigLIP2 bug at different levels of specificity, and the other verdicts do not overclaim beyond the titles/summaries provided." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 17, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:30064", + "issue:32090", + "issue:33357", + "issue:33666", + "issue:34567", + "issue:35141", + "issue:38175", + "issue:41720", + "issue:43122", + "issue:43262", + "issue:43295", + "issue:43329", + "issue:43334", + "issue:43452", + "issue:43504", + "issue:43525", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43906", + "issue:43927", + "issue:44361", + "issue:44442", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44554", + "issue:44625", + "issue:44743", + "issue:44843", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45005", + "issue:45084", + "issue:45325", + "issue:45356", + "issue:45588" + ], + "soft_pairs": [ + "issue:44466|issue:45005", + "issue:44464|issue:44743", + "issue:24643|issue:29127", + "issue:43525|issue:43643", + "issue:41720|issue:43906", + "issue:43577|issue:43697", + "issue:38175|issue:43295", + "issue:44361|issue:44442", + "issue:44843|issue:44991", + "issue:43653|issue:44361", + "issue:44479|issue:44936", + "issue:43122|issue:43906", + "issue:44977|issue:45356", + "issue:28282|issue:32090", + "issue:28282|issue:33357", + "issue:28282|issue:35141", + "issue:30064|issue:33666", + "issue:43577|issue:45588", + "issue:43531|issue:43742", + "issue:43577|issue:44554", + "issue:44464|issue:44977", + "issue:43653|issue:43927", + "issue:43329|issue:45325", + "issue:44479|issue:44625", + "issue:43550|issue:44361", + "issue:43452|issue:44451", + "issue:41720|issue:44464", + "issue:43504|issue:43720", + "issue:43504|issue:43550", + "issue:32090|issue:33666", + "issue:32090|issue:34567", + "issue:43334|issue:43572", + "issue:43262|issue:45084" + ], + "prepared_review_unit_hash": "e63e0114647d63a010600a30b7caaf2c92f304c113def96e5f22a6fc518c378e", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37914, + "estimated_input_tokens": 9479, + "estimated_eval_tokens": 19214 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a heterogeneous set of unrelated bug reports across tokenizer loading, config propagation, vision/multimodal preprocessing, trainer, and generation. The soft edges mostly share only broad subsystem similarity, not the same concrete underlying bug, so none should be merged.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out: the set spans many distinct failure modes and architectures, and the apparent overlaps are mostly broad-topic false positives rather than duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue for deduplication; any one choice would misrepresent the rest because the reported bugs are materially different.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a serialization/device-dependent lm_head.weight bug and the other is a translation-model loading issue in v5; different failure modes." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Both concern generation/caching, but one is inconsistent chunked generation under compiled forward and the other is recurrent states resetting with cache; not the same bug." + }, + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "DeepSpeed training weight-dimension RuntimeError and LayoutLMv3 box-validation messaging are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Both involve missing config fields, but one is Llama4 pad_token_id access and the other is trust_remote_code returning incomplete AutoConfig; different code paths and symptoms." + }, + { + "left": "issue:41720", + "right": "issue:43906", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert and an isolated reproduction of another issue are not enough to establish the same bug." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "Blip2 dtype remaining float32 and RTDetrV2 output drift in v5 are separate model behavior bugs." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Siglip2 zero probabilities and processor.tokenizer regression are unrelated multimodal model issues." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Both are tokenizer-related load failures, but for different tokenizer classes and different error shapes." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Both affect tokenizer loading, but one is an offline model_info call regression and the other is a specific tokenizer load failure for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and MLukeTokenizer AttributeError are unrelated tokenizer defects." + }, + { + "left": "issue:44479", + "right": "issue:44936", + "accept": false, + "reason": "Video-input regression in Qwen VL models and trainer.evaluate() failing after train are different subsystems and bugs." + }, + { + "left": "issue:43122", + "right": "issue:43906", + "accept": false, + "reason": "Different tokenization across versions and an isolated reproduction of another issue do not indicate the same concrete defect." + }, + { + "left": "issue:44977", + "right": "issue:45356", + "accept": false, + "reason": "Qwen3.5 flash-attention generation failure and Kimi-K2.5 tokenizer regression are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "Missing PyTorch import and a NoneType broadcast error in Trainer are distinct errors." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "AutoModel import failure without PyTorch and MacOS bus error in CLIP are different problems." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "ImportError for missing torch and output embedding reinitialization after resize are unrelated." + }, + { + "left": "issue:30064", + "right": "issue:33666", + "accept": false, + "reason": "Void segmentation map processing and Qwen2-VL multi-GPU training are different areas and failure modes." + }, + { + "left": "issue:43577", + "right": "issue:45588", + "accept": false, + "reason": "Blip2 dtype handling and flash_attention crash on s_aux=None are not the same bug." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM load key error are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Blip2 precision loading and an MPS attention correctness issue with mismatched head dims are separate defects." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Chunked generation under compiled forward and Qwen3.5 flash-attention generation are different generation bugs." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBirdTokenizer decode output issue and DiaConfig custom token ID loss after save/load are different tokenizer/config bugs." + }, + { + "left": "issue:43329", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal token-indexing issues, but one is an undefined video-branch implementation bug and the other is still-image temporal scaling in rope index; not the same root cause." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video regression in several Qwen VL models and Qwen3.5 num_labels propagation are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba-9B-v2 and tokenizer AttributeError in MLukeTokenizer are different subsystems." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and inability to load a specific model are both loading problems, but not the same concrete bug." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping and compiled chunked generation inconsistency are unrelated runtime failures." + }, + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Legacy field handling in Beit semantic segmentation and BitNet packed-weight unpacking during accelerate loading are distinct model-loading issues." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer broadcast NoneType error and Qwen2-VL multi-GPU training request are not the same bug." + }, + { + "left": "issue:32090", + "right": "issue:34567", + "accept": false, + "reason": "Trainer GPU broadcast error and TrainerState.num_input_tokens_seen not updating are different trainer issues." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both mention missing pad-token-related fields, but they affect different configs and models; no shared code-path evidence." + }, + { + "left": "issue:43262", + "right": "issue:45084", + "accept": false, + "reason": "Audio processor chat-template sample-rate default and template compilation TypeError are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 18, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:33666", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:43257", + "issue:43278", + "issue:43299", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43540", + "issue:43572", + "issue:43606", + "issue:43646", + "issue:43688", + "issue:43716", + "issue:43720", + "issue:43742", + "issue:43746", + "issue:43756", + "issue:43824", + "issue:43825", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:44060", + "issue:44361", + "issue:44423", + "issue:44466", + "issue:44488", + "issue:44521", + "issue:44589", + "issue:44610", + "issue:44625", + "issue:44734", + "issue:44821", + "issue:44843", + "issue:44849", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45200", + "issue:45229", + "issue:45290", + "issue:45357", + "issue:45405", + "issue:45520", + "issue:45584" + ], + "soft_pairs": [ + "issue:43504|issue:44610", + "issue:43278|issue:43688", + "issue:44849|issue:45357", + "issue:43756|issue:44913", + "issue:43525|issue:43572", + "issue:43540|issue:43716", + "issue:43742|issue:44821", + "issue:43493|issue:43825", + "issue:33666|issue:36010", + "issue:45200|issue:45229", + "issue:43688|issue:43873", + "issue:43720|issue:43927", + "issue:43606|issue:43927", + "issue:43257|issue:43299", + "issue:43278|issue:43844", + "issue:39692|issue:43493", + "issue:44488|issue:44843", + "issue:43540|issue:44977", + "issue:43720|issue:44361", + "issue:44964|issue:45405", + "issue:33666|issue:36331", + "issue:43493|issue:43844", + "issue:43756|issue:44060", + "issue:44625|issue:45584", + "issue:44589|issue:45520", + "issue:43646|issue:44466", + "issue:44521|issue:45290", + "issue:44423|issue:44734", + "issue:43824|issue:44589", + "issue:43688|issue:43746" + ], + "prepared_review_unit_hash": "bb1b6cd3da4e2bede23b035b6c1d01b64aae75136ee2e23a10346537d78235cb", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36766, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are a loose similarity cluster, not a duplicate set: they cover many unrelated models, configs, and failure modes. No soft pair looks like the same underlying bug/change, so all candidate edges are rejected.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single canonical issue is appropriate because the cluster is heterogeneous; the closest-to-broad regression report is 43646, but it still does not subsume the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different models and bugs: Beit legacy-field loading vs OmDet processor size mismatch." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Unrelated problems: dtype drift between train/eval vs auxiliary loss normalization." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Same family (Qwen3.5) but different failures: hidden-states output vs save_pretrained visual-encoder keys." + }, + { + "left": "issue:43756", + "right": "issue:44913", + "accept": false, + "reason": "Both involve rotary/RoPE config, but one is a layer-drop mismatch and the other is reload/default persistence." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Similar symptom class, but different configs/fields and models: pad_token_id vs pad_token_idx." + }, + { + "left": "issue:43540", + "right": "issue:43716", + "accept": false, + "reason": "Different multimodal code paths and modalities: Qwen3OmniMoe video processing vs Mistral-3 image dtype mismatch." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "Different loading failures: MobileLLM key error vs AutoImageProcessor URL loading." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy is unrelated to a pipeline translation error-message regression." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Unrelated: multi-GPU training for Qwen2-VL vs GenerationMixin import error." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Different issues in Gemma 4: text-only mm_token_type_ids defaults vs multi-GPU inference CUDA OOM." + }, + { + "left": "issue:43688", + "right": "issue:43873", + "accept": false, + "reason": "Different subsystems: auxiliary-loss normalization vs quantization/offloading behavior." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "Different models and failure modes: packed-weight accelerate loading vs config token-ID persistence." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to DiaConfig save/load token-ID loss." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Both are Qwen3 MoE-related, but one is deepspeed conversion/loading and the other is a v5 model-load break." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Different bugs: BF16/FP32 dtype drift vs DeepSpeed gradient growth." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Documentation example errors are not the same bug as SigLIP2 implementation mismatch." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Different causes: a specific model failing to load vs offline-mode break from unconditional hub metadata lookup." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen3.5-related, but one is video-input processing and the other is flash-attention generation behavior." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Different components and failures: BitNet packed weights vs MLukeTokenizer AttributeError." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Model loading regression vs dependency version bump; not the same underlying issue." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Unrelated training and trainer API issues." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 discrepancy is unrelated to DeepSpeed/gradient magnitude behavior." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Both mention model internals, but one is RoPE-layer dropping and the other is tied-weights warning corruption." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "Different models and symptoms: Qwen3.5 num_labels propagation vs Whisper empty-transcription generation failure." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Different incompatibilities: missing Float8 storage type vs flash_attn import-utils KeyError on Python 3.13." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Both are v5 regressions, but one is custom model initialization and the other is lm_head serialization depending on device." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Same API area, but different bugs: multimodal assistant masks vs tool-call/no-content crash." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both are serving crashes, but the code paths differ: multimodal continuous batching vs KV-cache continuation indexing." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Import error for a specific model class is unrelated to the Float8 storage lookup failure." + }, + { + "left": "issue:43688", + "right": "issue:43746", + "accept": false, + "reason": "Different model families and issues: loss normalization vs PEFT/local-checkpoint loading." + } + ] + }, + "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": 13, + "review_unit_index": 19, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:15354", + "issue:29942", + "issue:30333", + "issue:33357", + "issue:33666", + "issue:37428", + "issue:38175", + "issue:39401", + "issue:39692", + "issue:43232", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43504", + "issue:43525", + "issue:43540", + "issue:43572", + "issue:43575", + "issue:43582", + "issue:43606", + "issue:43653", + "issue:43701", + "issue:43844", + "issue:43927", + "issue:43957", + "issue:43994", + "issue:44112", + "issue:44291", + "issue:44315", + "issue:44361", + "issue:44464", + "issue:44493", + "issue:44554", + "issue:44568", + "issue:44610", + "issue:44792", + "issue:44849", + "issue:44964", + "issue:44977", + "issue:45072", + "issue:45084", + "issue:45198", + "issue:45200", + "issue:45356", + "issue:45405", + "issue:45446", + "issue:45520", + "issue:45584" + ], + "soft_pairs": [ + "issue:39692|issue:43295", + "issue:33357|issue:33666", + "issue:43572|issue:44291", + "issue:43994|issue:44493", + "issue:43504|issue:43606", + "issue:44464|issue:45584", + "issue:43575|issue:45405", + "issue:43927|issue:44361", + "issue:43504|issue:43653", + "issue:43278|issue:43701", + "issue:44792|issue:45405", + "issue:43606|issue:44112", + "issue:29942|issue:30333", + "issue:15354|issue:38175", + "issue:44554|issue:45446", + "issue:43540|issue:45200", + "issue:45084|issue:45520", + "issue:43844|issue:44964", + "issue:39401|issue:44568", + "issue:43582|issue:44977", + "issue:43232|issue:44849", + "issue:45072|issue:45198", + "issue:37428|issue:43957", + "issue:44315|issue:44849", + "issue:44554|issue:45356", + "issue:43334|issue:43525", + "issue:43295|issue:44610" + ], + "prepared_review_unit_hash": "5bac5679595785c57eb405b2946199a2f8dcab5f010d59530ed496d861ed24a0", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35856, + "estimated_input_tokens": 8964, + "estimated_eval_tokens": 18184 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous: most items are distinct bug reports across unrelated models/features, and none of the soft pairs look like the same underlying issue. No clear canonical issue/PR emerges.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Both involve processor/tokenizer behavior, but one is a SigLIP2 docs-example failure and the other is a v4.57.5 regression in custom code; different root causes." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error vs Qwen2-VL multi-GPU training; unrelated models and failure modes." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Both are 5.x compatibility/loading problems, but they hit different APIs and errors (`pad_token_idx` vs `_is_hf_initialized`); not the same bug." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 wrong outputs and a generic position-id key warning are different symptoms with no shared concrete code path." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "BEiT legacy-field loading failure and Bark CPU-offload device mismatch are unrelated." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation with compiled forward vs Whisper empty-transcription handling after token alignment; different generation paths." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Tensor-parallel OOM for a large model vs an unreleased dependency version bump; no overlap in bug/change." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig save/load token-ID loss and MLukeTokenizer task-time AttributeError are separate tokenizer/config problems." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Different models and failures: BEiT legacy-field load issue vs BigBird mask-token registration/decoding bug." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "Embedding dtype drift in evaluate vs checkpoint resume key mismatch are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "Janus image generation test failure vs PEFT version pin issue; no common underlying defect." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "CPU offload device mismatch in Bark vs a stale GraniteSpeech CI device-override test; not the same bug." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures and MLFlow-run state reporting are completely different subsystems." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript export limitation for GeneratorExp vs SigLIP2 zero probabilities; different models and issues." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness bug vs a PyTorch version-check/import bug in flex_attention; unrelated code paths." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Video input processing in Qwen3OmniMoe vs Gemma 4 token-type-id defaulting; different features and failures." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template-node compilation error vs Python 3.13 flash_attn import KeyError; unrelated runtime problems." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "DeepSpeed ZeRO-3 gradient growth vs Phi-4 multimodal load failure; different models and symptoms." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Qwen3 offset_mapping bug vs mdeberta BOS/EOS insertion regression; both tokenizer issues but not the same concrete defect." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "Apple Silicon allocator warmup TypeError and Qwen3.5 flash-attention generation failure are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwarg update after sync_gpus vs output_hidden_states bug in Qwen3.5; different generation code paths." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "bfloat16 dtype mismatches in inference vs Wav2Vec2 save/tokenization failure; no shared underlying bug." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash-attention import error vs meta-device model loading breakage; distinct failures." + }, + { + "left": "issue:44315", + "right": "issue:44849", + "accept": false, + "reason": "Liger Kernel not applied during model_init and Qwen3.5 hidden-states generation bug are unrelated." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "MPS attention correctness issue vs Kimi-K2.5 tokenizer codec regression; different components and root causes." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention missing pad_token_id, but they affect different configs/models and appear to be separate loading bugs." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Processor/tokenizer API regression vs OmDet-Turbo image-size mismatch; unrelated symptoms and fixes." + } + ] + }, + "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": 13, + "review_unit_index": 20, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:37428", + "issue:38175", + "issue:42915", + "issue:43278", + "issue:43295", + "issue:43388", + "issue:43450", + "issue:43493", + "issue:43526", + "issue:43572", + "issue:43653", + "issue:43701", + "issue:43824", + "issue:43827", + "issue:43844", + "issue:43856", + "issue:43873", + "issue:43901", + "issue:43975", + "issue:43994", + "issue:44112", + "issue:44206", + "issue:44360", + "issue:44442", + "issue:44464", + "issue:44466", + "issue:44514", + "issue:44521", + "issue:44561", + "issue:44568", + "issue:44589", + "issue:44792", + "issue:44849", + "issue:44898", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45003", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45245", + "issue:45290", + "issue:45325", + "issue:45381", + "issue:45412", + "issue:45563", + "issue:45584" + ], + "soft_pairs": [ + "issue:45200|issue:45325", + "issue:44466|issue:45127", + "issue:43493|issue:43701", + "issue:44898|issue:45412", + "issue:43873|issue:44464", + "issue:44792|issue:45563", + "issue:37428|issue:43824", + "issue:43388|issue:43526", + "issue:44514|issue:44521", + "issue:43572|issue:44987", + "issue:43844|issue:44936", + "issue:43572|issue:44589", + "issue:43824|issue:43975", + "issue:43295|issue:43450", + "issue:43295|issue:43873", + "issue:44360|issue:44561", + "issue:44442|issue:45198", + "issue:38175|issue:45412", + "issue:43295|issue:43994", + "issue:42915|issue:43824", + "issue:38175|issue:43493", + "issue:43278|issue:43856", + "issue:43827|issue:43901", + "issue:44849|issue:45003", + "issue:44568|issue:44991", + "issue:44206|issue:45584", + "issue:43653|issue:44112", + "issue:43450|issue:43493", + "issue:45290|issue:45381", + "issue:45127|issue:45245" + ], + "prepared_review_unit_hash": "879606e3e90ea6998c436e4063c4feca00ef48c27909f404f50a0e9731b56873", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36805, + "estimated_input_tokens": 9202, + "estimated_eval_tokens": 18660 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most soft pairs only share a subsystem or wording, not the same underlying defect. I would not merge any of the candidate pairs, and there is no safe single canonical issue for the whole set.", + "confidence": 0.95, + "canonical_issue_reason": "No single canonical issue fits: the items span unrelated bugs across multimodal processing, tokenizers, training, generation, and import/runtime failures.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a safe representative because the soft pairs do not describe one shared bug/change; they are separate regressions in different code paths.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Different defects: Gemma 4 token-type defaults vs Qwen2.5-VL RoPE/temporal position scaling." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "One is lm_head.weight serialization varying by device; the other is LoRA merge collapse with extended vocab and tied weights." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 implementation mismatch and resume-from-checkpoint key mismatch are unrelated failure modes." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver interpolation failure and RT-DETR memory not being released are distinct bugs." + }, + { + "left": "issue:43873", + "right": "issue:44464", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to chunked generation inconsistencies under compiled forward." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "Janus model test failure and a stale paged-generate warning are different issues." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but for different symbols and different code paths." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both involve labels, but one is gather_for_metrics dropping tuple elements and the other is BeitImageProcessorFast reduce_labels collapsing arrays." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both mention apply_chat_template, but one is a batching crash with padding=False and the other is incorrect assistant masks for multimodal inputs." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "StableLm config pad_token_idx regression is unrelated to loading physical-intelligence/fast." + }, + { + "left": "issue:43844", + "right": "issue:44936", + "accept": false, + "reason": "ZeRO-3 gradient growth and trainer.evaluate() failing after train() are different training bugs." + }, + { + "left": "issue:43572", + "right": "issue:44589", + "accept": false, + "reason": "Missing pad_token_idx in config and Float8 storage lookup errors are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "Qwen2_5_VL import failure and DeepSeek detokenization bug are different model-specific issues." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor/tokenizer regression and batched video shape error are separate multimodal bugs." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression is unrelated to quantization/offloading behavior." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to removal of is_torch_fx_available." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "FastSpeech2Conformer tokenizer loading and Wav2Vec2 save/tokenization failures are not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 zero-probabilities and RT-DETR GC/memory release issues are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43994", + "accept": false, + "reason": "Custom processor/tokenizer regression does not match the SigLIP2 nonsensical output issue." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Qwen2.5-VL import failure are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": false, + "reason": "Both involve SigLIP2 but one is zero probabilities while the other is HF vs JAX discrepancy; not the same concrete bug." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "Embedding dtype mismatch on evaluate and Qwen3 MoE memory usage are different training issues." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs cleanup around pipeline changes, but they cover different docs and different API mentions." + }, + { + "left": "issue:44849", + "right": "issue:45003", + "accept": false, + "reason": "Qwen3.5 hidden_states bug and unsafe sys.modules access are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer regressions for different models; not one shared bug." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Unsupported center arg in LasrFeatureExtractor and Whisper empty-transcription failure are separate regressions." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and GraniteSpeech stale CI test are unrelated." + }, + { + "left": "issue:43450", + "right": "issue:43493", + "accept": false, + "reason": "Batched video shape bug and SigLIP2 model discrepancy are different issues." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat-template crash with tool-call messages and Qwen2.5-VL video vision_position_ids are distinct bugs." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merge/collapsed weights and the 2^24 categories runtime error are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 21, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:32090", + "issue:37428", + "issue:38175", + "issue:39692", + "issue:43262", + "issue:43278", + "issue:43335", + "issue:43425", + "issue:43450", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43723", + "issue:43819", + "issue:43825", + "issue:43931", + "issue:44112", + "issue:44292", + "issue:44351", + "issue:44448", + "issue:44462", + "issue:44493", + "issue:44554", + "issue:44561", + "issue:44568", + "issue:44610", + "issue:44743", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45200", + "issue:45229", + "issue:45310", + "issue:45381", + "issue:45478", + "issue:45491", + "issue:45520", + "issue:45588", + "issue:45593" + ], + "soft_pairs": [ + "issue:44462|issue:44561", + "issue:43819|issue:45083", + "issue:44561|issue:44779", + "issue:43450|issue:43575", + "issue:45200|issue:45588", + "issue:37428|issue:43425", + "issue:44610|issue:44857", + "issue:44857|issue:45593", + "issue:43262|issue:43450", + "issue:43819|issue:44112", + "issue:44448|issue:44743", + "issue:45310|issue:45478", + "issue:43606|issue:45083", + "issue:43278|issue:43825", + "issue:29942|issue:32090", + "issue:37428|issue:44351", + "issue:39692|issue:43526", + "issue:43611|issue:43931", + "issue:43335|issue:43550", + "issue:43819|issue:44493", + "issue:45200|issue:45381", + "issue:43450|issue:43526", + "issue:39692|issue:44493", + "issue:29127|issue:32090", + "issue:43335|issue:43653", + "issue:44855|issue:45083", + "issue:44448|issue:44779", + "issue:38175|issue:44493", + "issue:44292|issue:45520", + "issue:44977|issue:45229", + "issue:43335|issue:45083", + "issue:43335|issue:43504", + "issue:43550|issue:44112", + "issue:44568|issue:45491", + "issue:43335|issue:45072", + "issue:43723|issue:44554" + ], + "prepared_review_unit_hash": "426c12fcf93507ac587fb1a1074b1ed3a39d9643245429443df2de339371d397", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38521, + "estimated_input_tokens": 9631, + "estimated_eval_tokens": 19518 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab-bag of unrelated bug reports. The only clear duplicate pair is the Qwen3.5 MoE from_pretrained error (issues 45310 and 45478); everything else looks like separate regressions or different subsystems.", + "confidence": 0.89, + "canonical_issue_reason": "Issue 45310 is the earlier and slightly better-documented report for the same Qwen3.5 MoE from_pretrained error also filed as 45478.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45310 is the best representative because it is the earliest clear duplicate target in the set and has the same concrete failure as 45478.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44462", + "right": "issue:44561", + "accept": false, + "reason": "Both are v5-era loading regressions, but one is AutoTokenizer/tokenizer.json and the other is removal of is_torch_fx_available for trust_remote_code; different bugs." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "Different code paths: DAC.from_latents parity with forward pass versus a Qwen3 Omni MoE helper length calculation issue." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "Both mention tokenizer breakage in v5, but one is about a removed utility and remote-code loading; the other is DeepSeek tokenization behavior." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Video processor shape mismatch and Qwen2 model loading OOM are unrelated." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 text-only mm_token_type_ids defaulting and flash_attention crashing on s_aux=None are different failures." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "ImportError for a flash-attention helper and a Torch 2.10 compatibility issue are not the same bug." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Processor output size mismatch versus AMP/CUDA loss crash are unrelated model-specific issues." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Both involve training losses, but one is a float16 CUDA crash and the other is missing auxiliary losses when denoising is off." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate default and batched video processor shape are different subsystems." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC latent/forward mismatch is unrelated to a stale CI device override test." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Pegasus v4/v5 output differences and Qwen3.5 recurrent state resetting are distinct model behaviors." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Same title and same Qwen3.5 MoE from_pretrained error; 45478 is a later duplicate of 45310." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch for bark-small versus a Qwen3 Omni MoE feature-length helper issue." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Embedding dtype drift in train/eval is unrelated to pipeline task support error messaging." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention test failures and Trainer _gpu_broadcast_one NoneType errors are different problems." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols in different modules." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "SigLIP2 doc example errors and BEiT fast processor label reduction bug are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43931", + "accept": false, + "reason": "base_model_prefix loading breakage in v5 and Qwen3-VL weight-shape mismatch are different model-loading failures." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer configuration bug and Bamba torch.compile SDPA failure are unrelated." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC latent mismatch and unexpected position-id key warnings are different issues." + }, + { + "left": "issue:45200", + "right": "issue:45381", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaults and Qwen2.5-VL video position ids are separate input-prep bugs." + }, + { + "left": "issue:43450", + "right": "issue:43526", + "accept": false, + "reason": "Batched video output shape bug versus BEiT fast reduce_labels behavior are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "Documentation/quantization issues for SigLIP2 do not match the position-id warning regression." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and Trainer GPU broadcast type error are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation bug and BigBirdTokenizer special-token registration bug are not the same defect." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 IndentationError in DebertaV2 import is unrelated to Qwen3 Omni MoE helper behavior." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Pegasus v4/v5 output change and DeepSeek tokenizer regression are separate tokenizer/model issues." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Unexpected zero probabilities in SigLIP2 and unexpected position-id keys are different symptoms with different causes." + }, + { + "left": "issue:44292", + "right": "issue:45520", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime failure and flash_attn import KeyError under Python 3.13 are not the same bug." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug and Gemma4 multi-GPU CUDA OOM are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "SwitchTransformers sparse-layer misconfiguration and Qwen3 Omni MoE helper length behavior are different code paths." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "Sparse-layer creation bug and BEiT semantic segmentation legacy-field loading bug are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "torch.compile SDPA failure and stale CI device override test failure are not the same underlying bug." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "mDeBERTa add_special_tokens behavior and Gemma3 NaN embeddings from sliding-window attention are different tokenizer/model issues." + }, + { + "left": "issue:43335", + "right": "issue:45072", + "accept": false, + "reason": "SwitchTransformers sparse layer creation and bfloat16 dtype mismatch in inference are distinct." + }, + { + "left": "issue:43723", + "right": "issue:44554", + "accept": false, + "reason": "AutoTokenizer loading regression in v5 and MPS attention correctness on differing head dims are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet, and the lone accepted duplicate pair (45310/45478) is well-supported by identical titles and chronology. The other pairwise rejections are conservative and do not overclaim duplication. The only minor caution is that 'slightly better-documented' is a bit stronger than the packet directly shows, but it does not materially affect the verdict." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 22, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:25251", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:40444", + "issue:41720", + "issue:43065", + "issue:43066", + "issue:43278", + "issue:43329", + "issue:43335", + "issue:43412", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43701", + "issue:43720", + "issue:43873", + "issue:43976", + "issue:43994", + "issue:44112", + "issue:44448", + "issue:44451", + "issue:44462", + "issue:44466", + "issue:44493", + "issue:44743", + "issue:44857", + "issue:44977", + "issue:44987", + "issue:45072", + "issue:45083", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45520", + "issue:45563", + "issue:45588", + "issue:45593" + ], + "soft_pairs": [ + "issue:40444|issue:43526", + "issue:44448|issue:44977", + "issue:43335|issue:45593", + "issue:43720|issue:44112", + "issue:43066|issue:44462", + "issue:45372|issue:45588", + "issue:34689|issue:35141", + "issue:44977|issue:45563", + "issue:44857|issue:45072", + "issue:43606|issue:44857", + "issue:43412|issue:45588", + "issue:37428|issue:44987", + "issue:43873|issue:44493", + "issue:43873|issue:45229", + "issue:45372|issue:45406", + "issue:29942|issue:35141", + "issue:43701|issue:45083", + "issue:43976|issue:45440", + "issue:45245|issue:45563", + "issue:43550|issue:45083", + "issue:43278|issue:43329", + "issue:32090|issue:33357", + "issue:41720|issue:45229", + "issue:45372|issue:45459", + "issue:44448|issue:44466", + "issue:44451|issue:45520", + "issue:43065|issue:43994", + "issue:25251|issue:32090", + "issue:43065|issue:45412", + "issue:43653|issue:44743", + "issue:44857|issue:45588" + ], + "prepared_review_unit_hash": "c27b3e69374e7ce3587541174fdf24a16d5043249d5db43cc8b96801fb2a99eb", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36493, + "estimated_input_tokens": 9124, + "estimated_eval_tokens": 18504 + }, + "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; these are mostly unrelated model-, tokenizer-, device-, or CI-specific issues. There is no clear canonical issue or PR for this mixed set.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Different bugs: Qwen2.5-VL multi-image IterableDataset failure vs BeitImageProcessorFast label reduction." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failure modes: Pegasus v4/v5 output drift vs Qwen3.5 flash-attention generation issue." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model logic: SwitchTransformers sparse-layer config bug vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "Different subsystems: BitNet accelerate loading failure vs GraniteSpeech CI stale device-override test." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Both tokenizer-related, but one is wrong decoder type in v5 and the other ignores tokenizer.json; not the same bug." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 processor import issue vs flash_attention.py AttributeError on s_aux=None are distinct code paths." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Different failure surfaces: Llama 3.2 Vision load regression vs embedding reinit after resize/post_init." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug is unrelated to the stale paged-generate warning." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Different loss/model paths: LwDetrImageLoss AMP crash vs dtype mismatches in SwitchTransformers/TimmWrapperModel inference." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "Both runtime/device related, but CPU offload mismatch for bark-small is not the same as LwDetr AMP/CUDA crash." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export/training-guard and erfinv support issue is unrelated to flash_attention.py s_aux=None crash." + }, + { + "left": "issue:37428", + "right": "issue:44987", + "accept": false, + "reason": "Both mention loading/flash-attention, but one is a missing internal import and the other is a model-loading failure for a specific repo." + }, + { + "left": "issue:43873", + "right": "issue:44493", + "accept": false, + "reason": "Quantization offload behavior is unrelated to unexpected position-id keys in model outputs." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Quantized offloading bug is not the same as Gemma4 multi-GPU CUDA OOM." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma4 processor import error and Gemma4Processor missing _tokenizer in serve are different bugs." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to token-embedding resize reinitialization." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint resume key mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python-version compatibility issue is unrelated to DeepseekV3MoE divergence from remote implementation." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-count limit error is unrelated to the stale num_return_sequences warning." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure is unrelated to qwen3_omni_moe length helper behavior." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Embedding dtype drift in eval is unrelated to the multimodal token-count video-branch bug." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Trainer broadcast TypeError and MacOS CLIP bus error are different failures." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3 auto device-map cudaErrorAssert is distinct from Gemma4 multi-GPU OOM." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma4 processor import breakage is unrelated to tokenizer-error masking when protobuf is missing." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Pegasus output differences and lm_head.weight serialization inconsistencies are different regressions." + }, + { + "left": "issue:44451", + "right": "issue:45520", + "accept": false, + "reason": "ScandiBERT loading failure is unrelated to flash_attn import key errors on Python 3.13." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Sam3PixelDecoder dummy Conv2d issue is unrelated to SigLIP2 producing nonsensical outputs." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as RT-DETR memory-not-released behavior." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug is unrelated to Qwen3 recurrent-state reset with cache." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "Different areas: LwDetr float16 AMP crash vs flash_attention.py AttributeError on sink-less models." + } + ] + }, + "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": 13, + "review_unit_index": 23, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:29127", + "issue:29942", + "issue:30064", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34567", + "issue:35141", + "issue:36010", + "issue:39692", + "issue:42915", + "issue:42994", + "issue:43257", + "issue:43262", + "issue:43278", + "issue:43334", + "issue:43381", + "issue:43421", + "issue:43526", + "issue:43531", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44279", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44479", + "issue:44610", + "issue:44871", + "issue:44977", + "issue:45072", + "issue:45103", + "issue:45198", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45412", + "issue:45440", + "issue:45491", + "issue:45593" + ], + "soft_pairs": [ + "issue:44186|issue:45198", + "issue:43278|issue:43381", + "issue:43421|issue:45491", + "issue:42915|issue:42994", + "issue:44361|issue:45198", + "issue:43994|issue:45412", + "issue:44265|issue:45198", + "issue:44206|issue:44479", + "issue:44871|issue:44977", + "issue:43257|issue:43278", + "issue:43756|issue:45440", + "issue:29942|issue:30064", + "issue:44610|issue:45412", + "issue:45103|issue:45372", + "issue:43824|issue:44279", + "issue:33357|issue:39692", + "issue:29127|issue:34567", + "issue:45440|issue:45593", + "issue:32090|issue:33290", + "issue:45072|issue:45229", + "issue:43381|issue:43701", + "issue:43653|issue:44442", + "issue:43334|issue:43531", + "issue:43526|issue:44186", + "issue:44464|issue:45491", + "issue:29942|issue:36010", + "issue:44977|issue:45245", + "issue:29127|issue:33357", + "issue:29127|issue:35141", + "issue:43761|issue:44079", + "issue:43262|issue:44206" + ], + "prepared_review_unit_hash": "385a1812659d0d9ee762cb3117e2b3ac62b993b9f33c9c73222eb12aa2e9b89a", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37066, + "estimated_input_tokens": 9267, + "estimated_eval_tokens": 18790 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are distinct bugs across tokenizers, vision/audio processors, training, generation, and model-loading paths. None of the soft-similarity pairs look like the same underlying fix, so all should be rejected. No PRs are present in the cluster.", + "confidence": 0.82, + "canonical_issue_reason": "issue:45198 is the strongest hub in the soft-pair graph and describes a concrete, user-facing tokenizer/save_pretrained failure. That said, the surrounding items are still too diverse for a true duplicate cluster, so this is only a weak canonical choice.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45198 is the best representative of the small tokenizer-failure subgroup because it is concrete and appears in multiple soft-similarity links. It is still not a good global duplicate anchor for the whole set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Different models and failure modes: LayoutLMv2Tokenizer NER/padding crash vs Wav2Vec2 save_pretrained/tokenization failure." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "Unrelated bugs: BF16->FP32 dtype drift in evaluate vs gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Different subsystems: runtime post-processor/special-token updates vs Gemma3 NaN embeddings from sliding-window padding." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "Both are quantization-related but not the same bug: FineGrainedFP8Config failure vs quantized model saving failure." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer bugs on different code paths: MLuke task AttributeError vs Wav2Vec2 tokenization/save failure." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "No shared underlying defect: SigLIP2 bad outputs vs RT-DETR memory not released on deletion." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export failure from torch_compilable_check is unrelated to Wav2Vec2 tokenization/save_pretrained errors." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Different regressions in different models: unsupported center arg in LASR extractor vs Qwen video input regression." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Distinct issues: Gemma eos_token_id config mismatch vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:43257", + "right": "issue:43278", + "accept": false, + "reason": "Different problem classes: Qwen3 MoE weight conversion with accelerate/deepspeed vs training/eval dtype change." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Separate model-specific correctness issues: Smollm3 RoPE layer count vs DeepseekV3 implementation divergence." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures and void segmentation map processing are unrelated bugs." + }, + { + "left": "issue:44610", + "right": "issue:45412", + "accept": false, + "reason": "Processor input-size mismatch vs model memory leak after deletion are not the same bug." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "auto_docstring annotation handling and Gemma 4 processor dependency import failure are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Specific missing export/import of Qwen2_5_VLForConditionalGeneration is not the same as a generic dependency issue." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS CLIP bus error and SigLIP2 doc example errors are different model-specific problems." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and TrainerState token counting are unrelated concerns." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3 model divergence and D-FINE auxiliary-loss omission are unrelated implementation bugs." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast TypeError and Adafactor/DeepSpeed OOM are different failure modes." + }, + { + "left": "issue:45072", + "right": "issue:45229", + "accept": false, + "reason": "Bfloat16 dtype mismatch in inference is unrelated to Gemma4 multi-GPU OOM." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Gradient checkpointing in eval mode and resume_from_checkpoint key mismatch are separate trainer issues." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and AutoTokenizer loading FastSpeech2ConformerTokenizer are different tokenizer bugs." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Missing pad_token_id on Qwen3-VL and sliding_window behavior on Qwen3-MoE are distinct model/config issues." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to LayoutLMv2Tokenizer crashes." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency is not the same bug as Gemma3 NaN embeddings from all-padding windows." + }, + { + "left": "issue:29942", + "right": "issue:36010", + "accept": false, + "reason": "Flash Attention 2 test failures and GenerationMixin import errors do not describe the same underlying defect." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation regression and category-cardinality runtime error are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 error-message clarity and MacOS bus-error crash are clearly different bugs." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Box-information error messaging in LayoutLMv3 is unrelated to post_init reinitializing resized embeddings." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and ModelOutput key assignment bug are different implementation issues." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sampling-rate default and LASR extractor center-arg crash are unrelated processor bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 24, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:33290", + "issue:34567", + "issue:34689", + "issue:39692", + "issue:41628", + "issue:42915", + "issue:43232", + "issue:43316", + "issue:43388", + "issue:43441", + "issue:43452", + "issue:43493", + "issue:43525", + "issue:43526", + "issue:43575", + "issue:43611", + "issue:43646", + "issue:43673", + "issue:43701", + "issue:43723", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44077", + "issue:44188", + "issue:44206", + "issue:44387", + "issue:44442", + "issue:44479", + "issue:44610", + "issue:44743", + "issue:44811", + "issue:44857", + "issue:44898", + "issue:44977", + "issue:45003", + "issue:45072", + "issue:45216", + "issue:45290", + "issue:45372", + "issue:45440", + "issue:45479", + "issue:45491", + "issue:45584", + "issue:45588", + "issue:45600" + ], + "soft_pairs": [ + "issue:42915|issue:43931", + "issue:43441|issue:45588", + "issue:43232|issue:43388", + "issue:43526|issue:44857", + "issue:44442|issue:45072", + "issue:43723|issue:45479", + "issue:44479|issue:45290", + "issue:43526|issue:44610", + "issue:45003|issue:45440", + "issue:44743|issue:44811", + "issue:43452|issue:45588", + "issue:43931|issue:45216", + "issue:45588|issue:45600", + "issue:43316|issue:43525", + "issue:43575|issue:43701", + "issue:44977|issue:45584", + "issue:43646|issue:44077", + "issue:43761|issue:44188", + "issue:43994|issue:44206", + "issue:39692|issue:45491", + "issue:43575|issue:44387", + "issue:43873|issue:45588", + "issue:43673|issue:44743", + "issue:34689|issue:43611", + "issue:44442|issue:44898", + "issue:33290|issue:34567", + "issue:41628|issue:45372", + "issue:43493|issue:43827" + ], + "prepared_review_unit_hash": "8f53e0b4048b294f190c17fd3fc339215226eda9c492f6cefae3c6ea7e565fb0", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36195, + "estimated_input_tokens": 9049, + "estimated_eval_tokens": 18354 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the soft-similarity links mostly connect unrelated bugs that only share broad subsystems (generation, tokenizers, vision, quantization, etc.), not the same underlying defect. No PRs are present.", + "confidence": 0.89, + "canonical_issue_reason": "No single canonical issue stands out; the items span many unrelated bugs across different models, processors, and runtime paths, so treating any one as the duplicate target would be misleading.", + "canonical_pr_reason": null, + "best_issue_reason": "If forced to pick a representative issue, issue:43873 is the best anchor because it is open, has the most discussion/inbound references, and is a broadly scoped report; however, it is not a true representative of the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Different models and failures: FineGrainedFP8Config vs a Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both involve flash attention, but one is a Ministral-3 generation regression and the other is an AttributeError in flash_attention.py for sink-less models." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Unrelated code paths: generation kwargs after sync_gpus vs gather_for_metrics dropping tuple labels." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "Different components and symptoms: BeitImageProcessorFast label reduction vs LwDetrImageLoss AMP crash." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer-loading failure for FastSpeech2Conformer is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43723", + "right": "issue:45479", + "accept": false, + "reason": "Tokenizer loading in v5 is unrelated to the zero-loss classification bug for num_labels=1." + }, + { + "left": "issue:44479", + "right": "issue:45290", + "accept": false, + "reason": "Different regressions: video-input handling for Qwen variants vs chat templating with tool-call assistant messages." + }, + { + "left": "issue:43526", + "right": "issue:44610", + "accept": false, + "reason": "Both are processor-related, but one is label reduction and the other is an input-size mismatch for OmDet-Turbo." + }, + { + "left": "issue:45003", + "right": "issue:45440", + "accept": false, + "reason": "Generic modeling_utils sys.modules access is unrelated to DeepseekV3MoE divergence from remote implementation." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Different failures in different areas: recurrent cache reset vs Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43452", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer/model loading with gguf_file is unrelated to a flash_attention AttributeError." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Same family name, but different bugs: Qwen3-VL loading mismatch vs Qwen3.5 save_pretrained checkpoint corruption." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "Flash-attention sink handling is unrelated to stale auto_mappings references breaking CI." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Both are config/API discrepancies, but they affect different classes and different missing/changed attributes." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "OOM during tensor-parallel model load is unrelated to resume_from_checkpoint key mismatches." + }, + { + "left": "issue:44977", + "right": "issue:45584", + "accept": false, + "reason": "Different generation bugs: flash-attention on Qwen3.5 vs empty-transcription handling in Whisper." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Custom model initialization breakage is broader and different from patchtsmixer's optional post_init constraint." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression is unrelated to attention-kernel divergence under torch.compile." + }, + { + "left": "issue:43994", + "right": "issue:44206", + "accept": false, + "reason": "SigLIP2 bad outputs/pipeline behavior is not the same bug as an unsupported center argument crash in LasrFeatureExtractor." + }, + { + "left": "issue:39692", + "right": "issue:45491", + "accept": false, + "reason": "Unrelated models and failures: SigLIP2 docs errors vs Gemma3 NaN embeddings with sliding-window attention." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both mention OOM, but one is tensor-parallel loading and the other is increased reserved memory under int4 quantization." + }, + { + "left": "issue:43873", + "right": "issue:45588", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to a flash-attention sink-less-model crash." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Cache handling bugs in different generation paths and models; not the same concrete defect." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Both are model-loading regressions, but they affect different versions and different root causes." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "FastSpeech2Conformer tokenizer loading is unrelated to Perceiver interpolation failures at non-default resolution." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "OOM with Adafactor/DeepSpeed is unrelated to TrainerState token counting not updating." + }, + { + "left": "issue:41628", + "right": "issue:45372", + "accept": false, + "reason": "Different import errors in different modules: AutoImageProcessor vs ReasoningEffort from mistral_common." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation parity is unrelated to docs still using removed pipeline() APIs." + } + ] + }, + "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": 13, + "review_unit_index": 25, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:25251", + "issue:28282", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:36010", + "issue:41950", + "issue:42175", + "issue:43065", + "issue:43295", + "issue:43316", + "issue:43381", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43475", + "issue:43540", + "issue:43643", + "issue:43688", + "issue:43716", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43761", + "issue:43825", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:44077", + "issue:44162", + "issue:44190", + "issue:44442", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44617", + "issue:44792", + "issue:44855", + "issue:44871", + "issue:44877", + "issue:44913", + "issue:45127", + "issue:45137", + "issue:45200", + "issue:45584", + "issue:45656" + ], + "soft_pairs": [ + "issue:43381|issue:43873", + "issue:33290|issue:35141", + "issue:33290|issue:34689", + "issue:44077|issue:44479", + "issue:33290|issue:45656", + "issue:43643|issue:44913", + "issue:43720|issue:44442", + "issue:44190|issue:44496", + "issue:43749|issue:43931", + "issue:43873|issue:44871", + "issue:44162|issue:44792", + "issue:43540|issue:44514", + "issue:43316|issue:44877", + "issue:43381|issue:43856", + "issue:34689|issue:36010", + "issue:44855|issue:45137", + "issue:42175|issue:43316", + "issue:43404|issue:43441", + "issue:43381|issue:43742", + "issue:43388|issue:43688", + "issue:16998|issue:22355", + "issue:25251|issue:28282", + "issue:43316|issue:43716", + "issue:43065|issue:44617", + "issue:41950|issue:43475", + "issue:45127|issue:45200", + "issue:43295|issue:43825", + "issue:43761|issue:44479", + "issue:44871|issue:45584", + "issue:43295|issue:44871" + ], + "prepared_review_unit_hash": "e9eadfb41649ef36c77e05fde7c230f447c74f3fbf43208be18e3543843be3a7", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36605, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No true duplicate cluster is present here: the soft links mostly connect unrelated Transformer bugs across config, loading, pipelines, training, and multimodal regressions. I would not merge any of the suggested pairs.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Different failure modes: gradient checkpointing in eval mode vs quantization/offloading behavior." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "OOM with Adafactor+DeepSpeed is unrelated to embedding reinit after resize." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Training OOM and Llama 3.2 vision model-loading regression are different bugs." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "Optional post_init API change is unrelated to the v5 video input regression." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "OOM in Adafactor+DeepSpeed is not the same as optimizer step being called twice." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "trust_remote_code config field loss is different from GPTNeoX rotary_pct reload behavior." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "Packed-weight loading in BitNet is unrelated to AutoTokenizer failing on a specific tokenizer class." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading in a training script is unrelated to an unrecognized model config." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU-efficient loading breakage is not the same as Qwen3-VL weight shape mismatch." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization/offloading issue is unrelated to Gemma eos_token_id config mismatch." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 model breakage does not match a janus test failure for image generation." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Qwen3OmniMoe video processing ValueError is different from Qwen2.5-VL batched chat-template crash." + }, + { + "left": "issue:43316", + "right": "issue:44877", + "accept": false, + "reason": "Gemma3TextConfig API mismatch is unrelated to granite_speech config loading strictness." + }, + { + "left": "issue:43381", + "right": "issue:43856", + "accept": false, + "reason": "Gradient checkpointing/eval is unrelated to Qwen3 MoE memory usage." + }, + { + "left": "issue:34689", + "right": "issue:36010", + "accept": false, + "reason": "Llama 3.2 vision loading regression is not the same as GenerationMixin import error." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 IndentationError in DebertaV2 import is unrelated to DeepSpeed ZeRO3 deque error." + }, + { + "left": "issue:42175", + "right": "issue:43316", + "accept": false, + "reason": "Missing TensorFlow in torch-only install is unrelated to Gemma3TextConfig API discrepancy." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "lm_head tying in Mistral3 is not the same bug as FlashAttention failure in Ministral-3." + }, + { + "left": "issue:43381", + "right": "issue:43742", + "accept": false, + "reason": "Gradient checkpointing eval-mode bug is unrelated to MobileLLM-125M loading key error." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "gather_for_metrics label truncation is unrelated to auxiliary-loss normalization." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length question for DeBERTa-V3 is unrelated to missing transformers.onnx import." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "top_k pipeline output shape regression is unrelated to PyTorch-missing AutoModel import error." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy is unrelated to Mistral-3 image preprocessor dtype mismatch." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to Sam3Video CUDA OOM." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "video-classification pipeline processor selection is unrelated to Sam3VisionEncoderOutput attribute error." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA merge collapse with extended vocab is unrelated to Gemma 4 mm_token_type_ids defaulting." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "processor.tokenizer regression is unrelated to pipeline translation-support error messaging." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "CLIPVision hidden_states regression is unrelated to Qwen v5.3.0 video-input regression." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency is unrelated to Whisper empty-transcription failure." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "processor/tokenizer regression is unrelated to Gemma eos_token_id mismatch." + } + ] + }, + "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": 13, + "review_unit_index": 26, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:32090", + "issue:35141", + "issue:41720", + "issue:42617", + "issue:43065", + "issue:43452", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43701", + "issue:43742", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43906", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44521", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:44933", + "issue:44938", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45245", + "issue:45290", + "issue:45357", + "issue:45362", + "issue:45381", + "issue:45406", + "issue:45542", + "issue:45563", + "issue:45584", + "issue:45600" + ], + "soft_pairs": [ + "issue:44734|issue:45406", + "issue:44077|issue:44534", + "issue:43550|issue:43701", + "issue:43606|issue:43701", + "issue:43646|issue:44913", + "issue:43906|issue:45290", + "issue:43761|issue:44242", + "issue:43452|issue:44496", + "issue:43931|issue:44060", + "issue:44617|issue:45584", + "issue:45563|issue:45584", + "issue:43575|issue:43742", + "issue:42617|issue:43575", + "issue:44521|issue:45381", + "issue:44933|issue:45362", + "issue:44479|issue:44734", + "issue:43575|issue:43873", + "issue:32090|issue:35141", + "issue:44464|issue:45563", + "issue:44464|issue:44871", + "issue:43065|issue:44610", + "issue:44792|issue:44977", + "issue:44534|issue:44913", + "issue:43827|issue:43994", + "issue:41720|issue:45245", + "issue:44792|issue:45600", + "issue:44190|issue:44964", + "issue:44938|issue:45542", + "issue:45127|issue:45357" + ], + "prepared_review_unit_hash": "189105baee3bb645fe2b30979616c1ec1b53dfd6ea2a079be1300c4b5f7a502b", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36224, + "estimated_input_tokens": 9056, + "estimated_eval_tokens": 18368 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the items span unrelated bugs across model loading, generation, quantization, serving, and docs. I do not see a true duplicate group, so there is no good canonical artifact here. All soft-edge pairs look like similarity-only false positives.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "Both are runtime crashes, but they hit different paths: KV-cache continuation indexing vs missing `_tokenizer` on `Gemma4Processor`." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Different regressions: one is about `post_init` on `patchtsmixer`, the other about junk-filled non-persistent buffers in v5." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "Unrelated failures: SDPA/torch.compile on Bamba vs checkpoint key mismatch on resume." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch for `bark-small` is unrelated to resume checkpoint key mismatch." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Custom model initialization breakage is not the same as `GPTNeoXConfig` reloading `rotary_pct` incorrectly." + }, + { + "left": "issue:43906", + "right": "issue:45290", + "accept": false, + "reason": "Different chat/template bugs: isolated reproduction of an earlier issue vs crash on tool-call assistant messages with no content." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "CLIP hidden states regression and MoE load-balancing loss logic are separate code paths and symptoms." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Tokenizer/model loading with `gguf_file` is unrelated to missing `model_type` in a config for a different model." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Weight shape mismatch in Qwen3-VL loading is not the same as the incorrect tied-weights warning in Qwen3-Next." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription failure after `align_special_tokens`." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "Stale `num_return_sequences` warning and Whisper empty-transcription crash are different issues." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Tensor-parallel OOM on Qwen2 is unrelated to the key error loading MobileLLM-125M." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "Failure to run `3d_parallel.py` is not the same defect as tensor-parallel OOM while loading Qwen2." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal, but one is all-zero assistant masks and the other is wrong `vision_position_ids`; different bugs." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "A bad import from `image_utils` is unrelated to Qwen3.5 crashing with `transformers chat`." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Video-input regression for several Qwen VL models is unrelated to serve-side KV-cache continuation indexing." + }, + { + "left": "issue:43575", + "right": "issue:43873", + "accept": false, + "reason": "Both mention memory/offload behavior, but one is TP OOM during load and the other is quantization offloading not behaving as expected." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer GPU broadcast TypeError and embedding reinitialization in `post_init` are different bugs." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Compiled-forward inconsistency in generation is unrelated to a stale warning about `num_return_sequences`." + }, + { + "left": "issue:44464", + "right": "issue:44871", + "accept": false, + "reason": "Different areas: chunked generation with compiled forward vs EOS token config mismatch." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "`Sam3PixelDecoder` dummy Conv2d and OmDet-Turbo processor input-size mismatch are separate model/processor issues." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation issue are unrelated." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk in v5 is not the same as `rotary_pct` silently resetting on reload." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing `pipeline()` is unrelated to SigLIP2 producing nonsensical outputs in `AutoModel`/pipeline usage." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "A800 cudaErrorAssert with Qwen3 auto device mapping is unrelated to the 2^24 category limit runtime error." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "Janus image-generation test failure and removed Sam3LiteText auto-mapping references are unrelated." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading in a training script is unrelated to loading `Phi-4-multimodal-instruct`." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import/load failure and the tensorboard-without-TF backend error are distinct startup issues." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge collapse with extended vocab and incorrect visual encoder keys on save are different bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 27, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36246", + "issue:38175", + "issue:38617", + "issue:39692", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:43065", + "issue:43116", + "issue:43295", + "issue:43388", + "issue:43404", + "issue:43479", + "issue:43540", + "issue:43550", + "issue:43582", + "issue:43644", + "issue:43650", + "issue:43761", + "issue:43976", + "issue:44016", + "issue:44077", + "issue:44079", + "issue:44112", + "issue:44190", + "issue:44220", + "issue:44246", + "issue:44279", + "issue:44355", + "issue:44466", + "issue:44514", + "issue:44561", + "issue:44683", + "issue:44734", + "issue:44743", + "issue:44855", + "issue:44913", + "issue:45003", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45375", + "issue:45406", + "issue:45446", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45632" + ], + "soft_pairs": [ + "issue:43388|issue:43582", + "issue:44855|issue:45588", + "issue:43761|issue:44743", + "issue:43650|issue:44220", + "issue:44355|issue:44855", + "issue:43540|issue:44734", + "issue:43644|issue:44913", + "issue:45276|issue:45584", + "issue:43976|issue:44855", + "issue:44077|issue:44079", + "issue:43404|issue:45216", + "issue:43550|issue:44190", + "issue:44112|issue:45600", + "issue:43116|issue:44190", + "issue:43295|issue:45276", + "issue:41950|issue:44016", + "issue:45245|issue:45584", + "issue:36246|issue:41628", + "issue:39692|issue:44190", + "issue:43479|issue:45375", + "issue:44734|issue:45632", + "issue:36246|issue:38617", + "issue:44561|issue:45632", + "issue:45406|issue:45632", + "issue:44016|issue:44220", + "issue:44514|issue:44734", + "issue:42175|issue:44279", + "issue:38175|issue:43065", + "issue:44683|issue:45446", + "issue:44466|issue:44913", + "issue:44246|issue:45003" + ], + "prepared_review_unit_hash": "207398b18306d23c5c219dc73583950b8a6ccef171f1b8df2d881742d5c6f762", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37128, + "estimated_input_tokens": 9282, + "estimated_eval_tokens": 18820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly false-positive soft matches across a heterogeneous set of import errors, model-specific regressions, CI failures, and unrelated runtime issues. No soft pair looks like the same underlying bug/change, and there are no PRs in the cluster.", + "confidence": 0.27, + "canonical_issue_reason": "Issue 41628 is the closest broad representative because it is a top-level `transformers` import/export regression, which is the most generic user-facing API break in this set.", + "canonical_pr_reason": null, + "best_issue_reason": "41628 is the best single anchor only in a weak sense: it is the most general import-surface failure, while the rest of the cluster is too mixed to support a true canonical duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43388", + "right": "issue:43582", + "accept": false, + "reason": "Different subsystems and symptoms: metric label trimming vs Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Unrelated failures: Python 3.13 import/indentation parsing vs flash-attention `s_aux=None` handling." + }, + { + "left": "issue:43761", + "right": "issue:44743", + "accept": false, + "reason": "Both are model regressions, but the concrete bugs differ: hidden_states return handling vs recurrent-state reset with cache." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "No common bug: a placeholder issue title vs an audio feature extraction helper error." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Both mention import/compile problems, but one is compiled Python file handling and the other is a Python 3.13 parser/indentation issue." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Different code paths: video input processing in Qwen3OmniMoe vs serving KV-cache continuation tensor indexing." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Different regressions: non-persistent buffer serialization vs GPTNeoX config reload persistence." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Unrelated model behaviors: token embedding resize propagation vs Whisper empty-transcription generation." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "Both are version/platform related, but one is a package compatibility complaint and the other is a Python 3.13 import parsing failure." + }, + { + "left": "issue:44077", + "right": "issue:44079", + "accept": false, + "reason": "Different bugs: `post_init` validation policy vs `ModelOutput` key assignment semantics." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Both involve tied weights/serialization, but they are different models and different failure modes, so not one concrete bug." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "Torch.compile/SDPA model bug vs local dataset loading in an example script; unrelated." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Both are CI-facing, but one is a stale test and the other is a broken config mapping; different underlying causes." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Example script behavior bug vs dataset-loading issue; not the same underlying defect." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Custom processor/tokenizer regression vs Gemma4 embedding resize propagation; different APIs and code paths." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video pipeline processor lookup bug vs a notebook syntax error; unrelated." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Different user-visible failures: category-cardinality runtime error vs Whisper empty transcription generation." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they are different missing symbols on different APIs and likely separate fixes." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "SigLIP2 docs/example errors vs image-classification dataset loading; not the same bug." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Both are config-related, but one is default-init handling for multimodal configs and the other is a missing strict field in Qwen3_5MoeVisionConfig." + }, + { + "left": "issue:44734", + "right": "issue:45632", + "accept": false, + "reason": "Serving KV-cache indexing crash vs trust_remote_code cache-path collision; unrelated." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Different missing imports from different modules and different root causes." + }, + { + "left": "issue:44561", + "right": "issue:45632", + "accept": false, + "reason": "Removal of `is_torch_fx_available` breaking remote-code models vs local cache-path collision; unrelated." + }, + { + "left": "issue:45406", + "right": "issue:45632", + "accept": false, + "reason": "Gemma4Processor missing `_tokenizer` attribute vs trust_remote_code cache collision; different failure surfaces." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error vs audio feature extraction helper bug; unrelated." + }, + { + "left": "issue:44514", + "right": "issue:44734", + "accept": false, + "reason": "Both involve Qwen-family multimodal processing, but one is batched chat-template padding and the other is KV-cache continuation indexing." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "Dependency/backend installation issue vs generic dependency complaint; too vague and not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:43065", + "accept": false, + "reason": "SigLIP2 probability output bug vs Sam3PixelDecoder dummy Conv2d; different models and code paths." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Compiled flex attention on newer torch vs incorrect PyTorch version guard for AuxRequest import; unrelated." + }, + { + "left": "issue:44466", + "right": "issue:44913", + "accept": false, + "reason": "Both are state/persistence regressions, but one is tied-weight serialization and the other is config reload of `rotary_pct`; not the same defect." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Import-time performance issue vs unsafe `sys.modules` access; related area but different bugs and fixes." + } + ] + }, + "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": 13, + "review_unit_index": 28, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43577", + "issue:43643", + "issue:43673", + "issue:43824", + "issue:43854", + "issue:43901", + "issue:43927", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44112", + "issue:44242", + "issue:44279", + "issue:44336", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44933", + "issue:44964", + "issue:44987", + "issue:45042", + "issue:45081", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45325", + "issue:45335", + "issue:45357", + "issue:45405", + "issue:45542", + "issue:45561", + "issue:45584", + "issue:45593", + "issue:45600", + "issue:45632" + ], + "soft_pairs": [ + "issue:43299|issue:45216", + "issue:43643|issue:44496", + "issue:42907|issue:43441", + "issue:44016|issue:44062", + "issue:43295|issue:45335", + "issue:44964|issue:45600", + "issue:44112|issue:44964", + "issue:43673|issue:44336", + "issue:43976|issue:44987", + "issue:43531|issue:45600", + "issue:45561|issue:45632", + "issue:43824|issue:44933", + "issue:45335|issue:45584", + "issue:43976|issue:44279", + "issue:44496|issue:45042", + "issue:43854|issue:44964", + "issue:43577|issue:45216", + "issue:43526|issue:43901", + "issue:43299|issue:45357", + "issue:44075|issue:44368", + "issue:44521|issue:45325", + "issue:44242|issue:45200", + "issue:43381|issue:45593", + "issue:43901|issue:44496", + "issue:43927|issue:45632", + "issue:44351|issue:45278", + "issue:44683|issue:45542", + "issue:43976|issue:45405", + "issue:43479|issue:43525", + "issue:44462|issue:45081" + ], + "prepared_review_unit_hash": "cef4760c3638eba31387854bc09a2f12a357248c32a0c756099ca1753df41244", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36768, + "estimated_input_tokens": 9192, + "estimated_eval_tokens": 18640 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is a loose mix of unrelated Transformers issues: model loading/saving regressions, import/config errors, multimodal quirks, and a few utility/docs/test bugs. I don\u2019t see strong duplicates among the soft-edge pairs.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: a version-upgrade regression causing many import failures. It\u2019s the most representative single issue, even though the cluster as a whole is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best cluster anchor because it captures the widest surface area of the reported breakage and is more general than the model-specific or one-off bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both are Qwen-family regressions, but one is a loading failure for Qwen3VL MoE and the other is a save_pretrained serialization bug for Qwen3.5; not the same code-path." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Both concern config/model loading, but one is trust_remote_code returning incomplete config fields and the other is model-type recognition from config.json; different failures." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Different models and different bugs: dequantized save failures vs FlashAttention loading regression." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Different subsystems and symptoms: processor/tokenizer multimodal regression vs resize_token_embeddings not updating decoder weights." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Phi-4 multimodal loading failure and removed Sam3LiteText auto-mapping references are separate model-specific issues." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI test flake in GraniteSpeech is not the same bug as a model-load failure for Phi-4 multimodal." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Chunked-prefill cache regression and ANSI-color output in loading_report are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Python version compatibility regression and a specific model-loading failure are different issues." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup are unrelated." + }, + { + "left": "issue:45561", + "right": "issue:45632", + "accept": false, + "reason": "A pytest-xdist file race and a trust_remote_code cache-path collision are different concurrency problems in different areas." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but one is a missing Qwen2.5-VL export and the other is a nonexistent image_utils import; not the same symbol or fix." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "resize_token_embeddings behavior for t5gemma and Whisper empty-transcription generation are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "Python support regression and a generic dependency issue are not the same underlying bug." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Model config recognition and PIL/torchvision backend requirements are different code paths." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Different model families and different loading failures." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "dtype propagation for BLIP2 and Qwen3.5 save_pretrained serialization are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and TextClassificationPipeline docs mismatch are not duplicates." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both involve Qwen3.5/Qwen3 serialization, but one is a loading failure and the other is incorrect visual-encoder keys on save; close but not the same bug." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "SGD optimizer argument handling and tie_word_embeddings warnings are unrelated." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal, but assistant mask generation and rope-index scaling are different concrete bugs." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Router loss calculation and mm_token_type_ids defaults are different training-time behaviors." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are unrelated." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Documentation mismatch and model config loading failure are not the same bug." + }, + { + "left": "issue:43927", + "right": "issue:45632", + "accept": false, + "reason": "DiaConfig save/load token-id loss and local trust_remote_code cache collision are different serialization/caching issues." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "Both mention imports, but 45278 is a broad report of many failures while 44351 is a specific missing HybridCache export; not enough to treat as one duplicate." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and tensorboard/TensorFlow backend detection are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version support and PEFT minimum-version bump are different dependency problems." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Phi4MultimodalConfig default subconfigs and Llama4Config missing pad_token_id are separate config bugs." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json and Mistral regex patch crashing are different tokenizer-loading failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 29, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41950", + "issue:42994", + "issue:43065", + "issue:43122", + "issue:43262", + "issue:43295", + "issue:43299", + "issue:43317", + "issue:43441", + "issue:43450", + "issue:43650", + "issue:43653", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43854", + "issue:43901", + "issue:43994", + "issue:44016", + "issue:44162", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44291", + "issue:44297", + "issue:44462", + "issue:44466", + "issue:44496", + "issue:44521", + "issue:44617", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:44877", + "issue:44898", + "issue:44987", + "issue:45020", + "issue:45276", + "issue:45290", + "issue:45397", + "issue:45405", + "issue:45542", + "issue:45593" + ], + "soft_pairs": [ + "issue:41950|issue:45020", + "issue:44297|issue:44462", + "issue:44279|issue:45405", + "issue:44496|issue:44877", + "issue:41950|issue:43650", + "issue:43299|issue:43854", + "issue:44462|issue:44779", + "issue:43295|issue:44779", + "issue:38175|issue:44617", + "issue:39692|issue:43901", + "issue:43650|issue:45290", + "issue:43723|issue:44162", + "issue:43450|issue:44279", + "issue:44016|issue:44291", + "issue:42994|issue:43317", + "issue:43441|issue:44683", + "issue:43262|issue:44521", + "issue:44261|issue:44263", + "issue:43065|issue:44898", + "issue:44466|issue:45276", + "issue:43122|issue:43650", + "issue:44623|issue:45397", + "issue:44623|issue:44749", + "issue:43784|issue:45542", + "issue:44671|issue:44779", + "issue:43673|issue:43761", + "issue:44987|issue:45542", + "issue:43653|issue:45593", + "issue:43901|issue:43994" + ], + "prepared_review_unit_hash": "6775c67a49a9ea93be93acbb5f00430962cc7aeab7611b22389da0af1753b8be", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36281, + "estimated_input_tokens": 9071, + "estimated_eval_tokens": 18398 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate pairs are mostly same-subsystem lookalikes rather than true duplicates: tokenizer, processor/chat_template, config-loading, and serialization regressions are mixed together. I would keep them separate.", + "confidence": 0.67, + "canonical_issue_reason": "issue:44521 is the strongest anchor in this set: it is open, well-discussed, has inbound references, and clearly describes a concrete multimodal `apply_chat_template` bug with a focused reproduction path.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43994 is the cleanest standalone bug report for triage: it names the exact model, the observable bad behavior, and the API surfaces involved (`AutoModel` and `pipeline`).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Different bugs: video-classification image-processor lookup vs broad remote_code loading regressions." + }, + { + "left": "issue:44297", + "right": "issue:44462", + "accept": false, + "reason": "Related tokenizer area, but one is save-time class metadata mismatch and the other is load-time ignoring `tokenizer.json`." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "Unrelated: generic dependency issue vs a PEFT minimum-version bump problem." + }, + { + "left": "issue:44496", + "right": "issue:44877", + "accept": false, + "reason": "Both concern config/model loading, but they target different failure modes and different models; not the same underlying bug." + }, + { + "left": "issue:41950", + "right": "issue:43650", + "accept": false, + "reason": "Completely unrelated subjects." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Both are model-loading failures, but for different models and likely different code paths." + }, + { + "left": "issue:44462", + "right": "issue:44779", + "accept": false, + "reason": "Different tokenizer issues: repository tokenizer selection vs incorrect tokenization behavior in v5." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Processor/tokenizer regression vs Deepseek tokenization output bug; not the same concrete defect." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "Different model families and symptoms; SigLIP2 probabilities vs Sam3Video OOM." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Both mention docs, but one is a SigLIP2 example error report and the other is a documentation mismatch for pipeline return behavior." + }, + { + "left": "issue:43650", + "right": "issue:45290", + "accept": false, + "reason": "No overlap beyond `apply_chat_template`-adjacent wording; one is a placeholder issue and the other is a specific tool-call crash." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "Tokenizer loading regression vs ESM2 model breakage; different artifacts." + }, + { + "left": "issue:43450", + "right": "issue:44279", + "accept": false, + "reason": "Video processor batching shape bug is unrelated to a dependency issue." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error and `init_empty_weights` argument incompatibility are separate failures." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Both involve quantization/offload, but one is save failure and the other is dequantized load with device_map offload; different bug paths." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "Attention backend failures, but one is FlashAttention on a specific model and the other is compiled flex_attention on newer torch." + }, + { + "left": "issue:43262", + "right": "issue:44521", + "accept": false, + "reason": "Both involve processors and `apply_chat_template`, but one is a default sampling-rate issue and the other is assistant mask generation for multimodal inputs." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Different model internals and different symptoms." + }, + { + "left": "issue:43065", + "right": "issue:44898", + "accept": false, + "reason": "Dummy convolution implementation vs interpolation-related image classification failure; not the same bug." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Serialization of tied weights vs resize-token-embeddings propagation; related area, but different concrete defects." + }, + { + "left": "issue:43122", + "right": "issue:43650", + "accept": false, + "reason": "Tokenizer behavior regression is unrelated to the placeholder issue." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Processor save-pretrained missing files vs zero3 from_pretrained loading failure; separate code paths." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated: processor saving bug vs a data-filtering performance complaint." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Different dependency/import failures affecting different backends." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and Deepseek tokenization regression are distinct model/tokenizer issues." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Generation cache/chunked-prefill bug vs CLIPVision hidden-states regression; different code paths." + }, + { + "left": "issue:44987", + "right": "issue:45542", + "accept": false, + "reason": "Loading a specific model under v5.1.0 vs optional TensorFlow backend import issue." + }, + { + "left": "issue:45020", + "right": "issue:45276", + "accept": false, + "reason": "Remote-code loading regressions and embedding-resize propagation are unrelated." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "No shared underlying bug; one is a precision/config issue, the other is a torch.split indexing issue." + }, + { + "left": "issue:44297", + "right": "issue:44779", + "accept": false, + "reason": "One is save_pretrained metadata, the other is tokenization output correctness; not duplicates." + } + ] + }, + "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": 13, + "review_unit_index": 30, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42175", + "issue:42491", + "issue:42757", + "issue:43295", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43673", + "issue:43824", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44162", + "issue:44188", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44297", + "issue:44336", + "issue:44351", + "issue:44361", + "issue:44373", + "issue:44462", + "issue:44485", + "issue:44496", + "issue:44561", + "issue:44589", + "issue:44617", + "issue:44661", + "issue:44857", + "issue:45003", + "issue:45200", + "issue:45250", + "issue:45356", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45561", + "issue:45563", + "issue:45593" + ], + "soft_pairs": [ + "issue:42175|issue:43976", + "issue:44361|issue:45593", + "issue:44297|issue:45397", + "issue:43475|issue:44016", + "issue:45561|issue:45563", + "issue:43824|issue:43976", + "issue:43295|issue:44617", + "issue:43352|issue:45250", + "issue:43525|issue:43881", + "issue:44589|issue:45362", + "issue:44188|issue:44336", + "issue:41628|issue:43976", + "issue:43901|issue:44246", + "issue:42491|issue:43299", + "issue:41950|issue:44220", + "issue:44336|issue:44373", + "issue:44295|issue:45397", + "issue:44242|issue:45593", + "issue:44462|issue:45356", + "issue:44661|issue:45542", + "issue:44246|issue:44496", + "issue:45003|issue:45561", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:43531|issue:45362", + "issue:43526|issue:45003", + "issue:44336|issue:44485", + "issue:43673|issue:44561", + "issue:44162|issue:45200", + "issue:44242|issue:44857" + ], + "prepared_review_unit_hash": "478478f23f51660d6591cb4e3235c76de4de77e65cb37715452cc810a9509f6c", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36304, + "estimated_input_tokens": 9076, + "estimated_eval_tokens": 18408 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated Transformers bug reports, with many model-specific regressions and import/load failures but no clear duplicate group. The closest common theme is v5.x breakage around model loading and processor/tokenizer imports.", + "confidence": 0.23, + "canonical_issue_reason": "Issue 43299 is the best representative of the repeated theme here: a user-facing v5.0.0.dev0 regression that breaks model loading, which is broader than the many single-model errors in the set.", + "canonical_pr_reason": null, + "best_issue_reason": "43299 is the strongest canonical choice because it is a clear, current-version regression in a core code path (model loading) and reads like the most central problem among the heterogeneous reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Different problems: backend extra packaging vs Python version compatibility/install breakage." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "Unrelated model/task bugs: MLukeTokenizer task failure vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer config metadata mismatch is unrelated to gemma-4 zero3/from_pretrained loading." + }, + { + "left": "issue:43475", + "right": "issue:44016", + "accept": false, + "reason": "SAM3Video attribute error vs a notebook syntax error; no shared underlying bug." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "pytest-xdist file race is unrelated to a stale generation warning." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "Importing a specific Qwen2.5 VL class is unrelated to Python 3.9+/3.10+ support." + }, + { + "left": "issue:43295", + "right": "issue:44617", + "accept": false, + "reason": "Processor/tokenizer regression and CUDA OOM are different failure modes and code paths." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Both mention Flash Attention, but one is an unsupported-model error and the other is a generic FA2 issue; not the same bug." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Missing Llama4Config pad_token_id is unrelated to glm-4v-9b loading failure." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage type error and Qwen3.5 chat crash are different runtime failures." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "torch.compile attention-kernel divergence is unrelated to ANSI codes in loading_report." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "AutoImageProcessor import failure is a separate issue from package Python-version compatibility." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Docs mismatch for return_all_scores has nothing to do with slow imports." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both involve Qwen3 MoE, but one is a LoRA compatibility issue on hf4.x\u2192hf5.x and the other is a direct loading regression in dev0; not clearly the same concrete bug." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification using image processors is unrelated to _torch_extract_fbank_features()." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "Terminal ANSI escape handling and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer access error is not the same as gemma-4 zero3 from_pretrained loading failure." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Missing load-balancing loss and D-FINE auxiliary losses are different model-training bugs." + }, + { + "left": "issue:44462", + "right": "issue:45356", + "accept": false, + "reason": "Tokenizer.json being ignored is a different tokenizer regression than Kimi-K2.5 codec handling." + }, + { + "left": "issue:44661", + "right": "issue:45542", + "accept": false, + "reason": "Model-mapping code generation bug and missing TensorFlow backend installation are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Slow import performance is unrelated to an unrecognized-model/config.json error." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is unrelated to a pytest-xdist capture-file race." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Qwen2.5 VL image-processor import error and huggingface_hub is_offline_mode import error are different missing symbols." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Different missing exports: Qwen2.5 VLImageProcessor vs HybridCache." + }, + { + "left": "issue:43531", + "right": "issue:45362", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Qwen3.5 chat crashes are not the same concrete bug." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BEiT reduce_labels bug is unrelated to sys.modules access in modeling_utils." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "ANSI-code output issue is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Chunked-prefill cache missing and removal of is_torch_fx_available are separate regressions." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Load-balancing loss and LwDetr float16 AMP crashes 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": 2, + "component_count": 13, + "review_unit_index": 31, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36246", + "issue:41628", + "issue:41950", + "issue:42886", + "issue:43065", + "issue:43066", + "issue:43232", + "issue:43257", + "issue:43262", + "issue:43295", + "issue:43408", + "issue:43441", + "issue:43475", + "issue:43540", + "issue:43596", + "issue:43653", + "issue:43901", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44368", + "issue:44466", + "issue:44496", + "issue:44749", + "issue:44779", + "issue:44871", + "issue:44877", + "issue:44933", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45229", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45459", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45656" + ], + "soft_pairs": [ + "issue:44246|issue:44871", + "issue:44877|issue:45375", + "issue:44295|issue:44749", + "issue:36246|issue:43596", + "issue:43066|issue:44779", + "issue:43441|issue:43540", + "issue:45292|issue:45584", + "issue:44779|issue:45292", + "issue:44987|issue:45459", + "issue:41628|issue:41950", + "issue:42886|issue:45229", + "issue:44995|issue:45200", + "issue:43901|issue:45003", + "issue:43065|issue:43408", + "issue:43232|issue:44336", + "issue:43257|issue:45137", + "issue:44496|issue:45003", + "issue:44060|issue:44077", + "issue:44995|issue:45588", + "issue:43295|issue:45292", + "issue:43262|issue:44246", + "issue:43653|issue:44162", + "issue:43475|issue:43540", + "issue:44368|issue:45563", + "issue:44466|issue:45335", + "issue:44368|issue:44977", + "issue:45397|issue:45656", + "issue:44933|issue:45459" + ], + "prepared_review_unit_hash": "433a7361c79fe07a5ffe4217778c40995cd6b1955d83400d011812e9f6ae7808", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35622, + "estimated_input_tokens": 8906, + "estimated_eval_tokens": 18068 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of Transformers v5-era regressions across tokenization, config strictness, embedding resizing/tied weights, generation, DeepSpeed, and flash-attn. I don\u2019t see any pair that clearly describes the same concrete bug, so I rejected all soft edges.", + "confidence": 0.69, + "canonical_issue_reason": "Issue 44466 is the broadest representative of the tied-weights / serialization regression subgroup, which is the most connected theme in this otherwise heterogeneous cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "44466 is the best overall anchor issue because it is concrete, regression-oriented, and closest to several nearby weight-handling reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Different bugs: intermittent import slowdown vs Gemma eos_token_id config mismatch." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both involve strict config handling, but they are different model/config failures with different symptoms." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Unrelated: position_ids buffer access error vs performance regression after upgrade." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Different code paths: missing Qwen2_5_VLImageProcessor import vs DeepSpeed ZeRO3/BertModel index error." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both tokenizer-adjacent, but one is a decoder-type warning and the other is incorrect DeepSeek tokenization behavior." + }, + { + "left": "issue:43441", + "right": "issue:43540", + "accept": false, + "reason": "Different models and failures: flash-attn issue for Ministral-3 vs video input validation in Qwen3OmniMoe." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "Resize-token-embeddings bug is unrelated to Whisper empty-transcription behavior after align_special_tokens." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Tokenizer correctness bug vs embedding-resize propagation bug; not the same underlying issue." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model loading failure vs protobuf-specific error masking in tokenizer decoding." + }, + { + "left": "issue:41628", + "right": "issue:41950", + "accept": false, + "reason": "Related around processors, but one is a missing import and the other is a pipeline selecting the wrong processor type." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline cache loading bug vs Gemma4 multi-GPU OOM; completely different problems." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "Stale indexer cache on second forward pass is unrelated to missing mm_token_type_ids defaults." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "Docs mismatch vs unsafe sys.modules access in modeling_utils; different layers and symptoms." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "Sam3 pixel decoder dummy conv issue vs sam3_video/sam3_tracker config warning; not the same bug." + }, + { + "left": "issue:43232", + "right": "issue:44336", + "accept": false, + "reason": "Generation kwargs update bug vs ANSI code emission in loading_report; unrelated." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Qwen3 MoE weight conversion problem vs DeepSpeed ZeRO3 deque underflow; different failure modes." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized model config vs unsafe sys.modules access; not the same underlying cause." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Tied-weights warning bug vs patchtsmixer post_init policy change; unrelated." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Cache/indexer crash vs flash_attention s_aux=None AttributeError; different subsystems." + }, + { + "left": "issue:43295", + "right": "issue:45292", + "accept": false, + "reason": "Processor.tokenizer regression vs resize_token_embeddings not updating output embeddings; separate code paths." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio sampling-rate default bug vs occasional import slowness; no duplication." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 breakage; different models and symptoms." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "Missing Sam3VisionEncoderOutput attribute vs Qwen3OmniMoe video processing ValueError; unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "Tied-weights warning vs stale paged generate warning; different generation/config issues." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both touch embedding resizing / weight tying, but one is serialization behavior and the other is decoder.embed_tokens not updating; not the same bug." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "Weight-tying warning in LoRA fine-tuning vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "DeepSpeed zero3 from_pretrained issue vs optimizer.step being called twice; distinct DeepSpeed bugs." + }, + { + "left": "issue:44933", + "right": "issue:45459", + "accept": false, + "reason": "Missing import from image_utils vs tokenizer decode-error masking without protobuf; unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 32, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43704", + "issue:43873", + "issue:43931", + "issue:44038", + "issue:44162", + "issue:44188", + "issue:44222", + "issue:44297", + "issue:44361", + "issue:44368", + "issue:44623", + "issue:44661", + "issue:44734", + "issue:44749", + "issue:44869", + "issue:44933", + "issue:44945", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45200", + "issue:45230", + "issue:45278", + "issue:45310", + "issue:45356", + "issue:45362", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45478", + "issue:45600", + "issue:45656" + ], + "soft_pairs": [ + "issue:43673|issue:44734", + "issue:44869|issue:45356", + "issue:45137|issue:45600", + "issue:43931|issue:44368", + "issue:43232|issue:43673", + "issue:44297|issue:44623", + "issue:43646|issue:44162", + "issue:43673|issue:45278", + "issue:42673|issue:43704", + "issue:44361|issue:44749", + "issue:44623|issue:45656", + "issue:28282|issue:30990", + "issue:44749|issue:45310", + "issue:42994|issue:43482", + "issue:44038|issue:44623", + "issue:45230|issue:45397", + "issue:44749|issue:45656", + "issue:42371|issue:45092", + "issue:43475|issue:45381", + "issue:45375|issue:45478", + "issue:43475|issue:43643", + "issue:41628|issue:44933", + "issue:45081|issue:45103", + "issue:45310|issue:45375", + "issue:42994|issue:43525", + "issue:42222|issue:45200", + "issue:43475|issue:44661", + "issue:43873|issue:44945", + "issue:42831|issue:44222", + "issue:45362|issue:45375", + "issue:43232|issue:44188" + ], + "prepared_review_unit_hash": "1c9c20279671b4c20ede0526c99aca945387ebf6c33651a39c31ec244368e7a8", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36727, + "estimated_input_tokens": 9182, + "estimated_eval_tokens": 18620 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a mixed bag of unrelated transformer/model issues; only a few pairs look like true duplicates, especially the Qwen3.5 MoE from_pretrained regressions and the Qwen3ForCausalLM VRAM leak reports.", + "confidence": 0.64, + "canonical_issue_reason": "issue:45375 is the most concrete Qwen3.5 MoE report: it names the missing `deepstack_visual_indexes` field and explains the `@strict`-related config loss that likely drives the broader from_pretrained failures in the sibling reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45375 is the best representative issue in this set because it pinpoints the root cause rather than only describing the symptom; the related generic reports appear to be the same bug family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43673", + "right": "issue:44734", + "accept": false, + "reason": "Both are generation/cache-related, but they describe different failures in different code paths (`chunked_prefill` cache missing vs KV cache continuation indexing)." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenizer/decoding behavior, but one is a Whisper word-timestamp crash and the other is a Kimi tokenizer codec regression; not the same bug." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque handling and Sam3LiteText auto-mapping CI breakage are unrelated." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "A weight-shape mismatch loading error and a tie_word_embeddings warning are different issues." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Both touch generation internals, but one is sync_gpus/update_kwargs ordering and the other is chunked_prefill cache missing; not the same concrete bug." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "`tokenizer.save_pretrained` metadata mismatch and `processor.save_pretrained` missing files are related broadly, but they are distinct serialization bugs." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "`trust_remote_code`/missing config fields and ESM2 being broken are too different in scope and failure mode." + }, + { + "left": "issue:43673", + "right": "issue:45278", + "accept": false, + "reason": "One is a specific v5.0 chunked_prefill cache issue; the other is a broad import-error complaint after upgrading. Too vague to treat as the same bug." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Same title and same specific symptom: Qwen3ForCausalLM VRAM leaking when used in multiple dataloader threads." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Tokenizer AttributeError on tasks vs a slowdown after upgrading Transformers; no shared code-path." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Processor save/export file generation and deepspeed optimizer stepping are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "PyTorch ImportError and Sentence Transformers stuck loading are different environment/runtime problems." + }, + { + "left": "issue:44749", + "right": "issue:45310", + "accept": false, + "reason": "Upgrade slowdown during filtering and Qwen3.5 MoE from_pretrained failure are unrelated." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Saving quantized models and loading Qwen2.5-GGUF under Transformers v5 are opposite operations and different failure modes." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe v5 bug and processor.save_pretrained file loss are distinct issues." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "The first is a generic placeholder report; the second is a gemma-4 DeepSpeed/ZeRO3 loading bug. No evidence they are the same." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "A data-filtering slowdown after upgrade and a double optimizer step under deepspeed are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API-settings guidance and InternVL2 meta-initialization incompatibility are not the same bug." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "Different model families and different attribute/position-id failures; both are vision issues but not the same code-path." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": true, + "reason": "Both are the same Qwen3.5 MoE from_pretrained regression family; 45375 gives the concrete missing-field root cause and 45478 reports the same failure under Transformers>=5.4.0." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "SAM 3 video encoder attribute error and AutoConfig `trust_remote_code` field loss are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Missing AutoImageProcessor import and a generic image_utils import issue are too vague to merge as duplicates." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Tokenizer regex patch crash and auto_docstring annotation handling crash are unrelated subsystems." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": true, + "reason": "Same Qwen3.5 MoE from_pretrained regression; 45375 identifies the missing config field that explains the generic 45310 failure." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized save failure and Llama4Config missing `pad_token_id` are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "SAM3 vitpose breakage and Gemma 4 mm_token_type_ids handling are different model-specific bugs." + }, + { + "left": "issue:43475", + "right": "issue:44661", + "accept": false, + "reason": "A SAM 3 video encoder attribute problem and an `add-new-model-like` tokenizer-mapping failure are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading behavior and pipeline-parallel output correctness are different issues." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "FineGrainedFP8 accuracy regression and FP8 save_pretrained failure are different bugs, even though both mention FP8." + }, + { + "left": "issue:45362", + "right": "issue:45375", + "accept": false, + "reason": "Qwen3.5-35B chat crash and Qwen3.5 MoE missing config field are separate failures." + }, + { + "left": "issue:43232", + "right": "issue:44188", + "accept": false, + "reason": "Generation kwargs/sync_gpus behavior and attention-kernel divergence under torch.compile are different code paths." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative on ambiguous pairs. The accepted duplicates are mergeable as single bugfixes (especially the identical Qwen3ForCausalLM VRAM leak reports and the Qwen3.5 MoE from_pretrained regressions). The summary does not overstate confidence beyond the evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 33, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:36010", + "issue:41628", + "issue:42175", + "issue:42898", + "issue:42994", + "issue:43232", + "issue:43278", + "issue:43377", + "issue:43479", + "issue:43504", + "issue:43526", + "issue:43673", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43976", + "issue:44038", + "issue:44060", + "issue:44079", + "issue:44164", + "issue:44188", + "issue:44222", + "issue:44242", + "issue:44351", + "issue:44373", + "issue:44448", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:44869", + "issue:44933", + "issue:44964", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45137", + "issue:45310", + "issue:45356", + "issue:45459", + "issue:45561" + ], + "soft_pairs": [ + "issue:41628|issue:44351", + "issue:44222|issue:45310", + "issue:42994|issue:44164", + "issue:44038|issue:45310", + "issue:44496|issue:45561", + "issue:43526|issue:45561", + "issue:43232|issue:44792", + "issue:43377|issue:44242", + "issue:43504|issue:45561", + "issue:36010|issue:41628", + "issue:43673|issue:44188", + "issue:30990|issue:35141", + "issue:43479|issue:44079", + "issue:42898|issue:44568", + "issue:43479|issue:44869", + "issue:43976|issue:44933", + "issue:42175|issue:44933", + "issue:44964|issue:45137", + "issue:43784|issue:43824", + "issue:44188|issue:44373", + "issue:44623|issue:45310", + "issue:44060|issue:45125", + "issue:43278|issue:45137", + "issue:43723|issue:44568", + "issue:44734|issue:45103", + "issue:42898|issue:45356", + "issue:42898|issue:44779", + "issue:30333|issue:33290", + "issue:43874|issue:45092", + "issue:42898|issue:44448", + "issue:41628|issue:45459" + ], + "prepared_review_unit_hash": "b97603aa2b89fc85bbd78da1a3711cf9f5db846ddb2749d87ab977cf08672675", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36960, + "estimated_input_tokens": 9240, + "estimated_eval_tokens": 18736 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Loose cluster of mostly unrelated Transformers v5 regressions, import failures, tokenizer/model bugs, and a few testing/runtime issues. The titles are semantically similar in places, but none of the soft pairs look like the same underlying bug or fix.", + "confidence": 0.91, + "canonical_issue_reason": "issue:42898 is the broadest and most central representative of the cluster\u2019s recurring v5 tokenizer-regression theme; it has the strongest overlap with other tokenizer-behavior reports, even though it is not a true duplicate of them.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42898 is the best single issue to anchor this cluster because it captures the common v5 regression/tokenization breakage pattern more generally than the other candidates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different symbols (`AutoImageProcessor` vs `HybridCache`) and different failure paths." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "Both mention MoE/FP8 save-load problems, but one is about `save_pretrained` for FP8 and the other about a `from_pretrained` error for Qwen3.5 MoE." + }, + { + "left": "issue:42994", + "right": "issue:44164", + "accept": false, + "reason": "Both concern saving/loading, but one is a quantized model save failure and the other is an `extra_state` handling bug; not the same code path." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Different model families and symptoms; Qwen3-VL-Moe incompatibility is not the same bug as Qwen3.5 Moe `from_pretrained` failure." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Unrecognized model config is unrelated to the xdist `captured_info.txt` race in patched testing utilities." + }, + { + "left": "issue:43526", + "right": "issue:45561", + "accept": false, + "reason": "Image processor label reduction bug and pytest-xdist file race are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation cache update issue and a janus image-generation test failure do not appear to be the same defect." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "MIMI padding-mask inconsistency and MoE load-balancing loss gating are distinct model behaviors." + }, + { + "left": "issue:43504", + "right": "issue:45561", + "accept": false, + "reason": "Legacy-field model-loading failure is unrelated to the xdist captured-info race." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they concern different API symbols and likely different fixes." + }, + { + "left": "issue:43673", + "right": "issue:44188", + "accept": false, + "reason": "Generation cache missing in v5 and attention-kernel divergence under `torch.compile` are different bugs." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "Sentence-transformers loading hang and embedding reinitialization after resize are unrelated." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Multimodal config defaulting bug and `ModelOutput` key assignment bug are different layers of the stack." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is about `clean_up_tokenization_spaces` and the other about `add_special_tokens` not adding BOS/EOS; different failures." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Config initialization bug and Whisper timestamp decode crash are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version support issue and image_utils import failure are not the same bug." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Missing TensorFlow backend and nonexistent image_utils import are different packaging/API problems." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading failure and DeepSpeed ZeRO3 deque error are unrelated runtime issues." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they affect different missing names and different modules/models." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Attention-kernel branching bug and wrong docstring are not the same underlying issue." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Processor save-pretrained file omission and Qwen3.5 MoE load failure are different save/load paths." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Tied-weights warning bug and missing `_tp_plan` for tensor parallelism are separate model-integration issues." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "Embedding dtype drift in eval and a ZeRO3 deque underflow are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer loading error and `add_special_tokens` BOS/EOS regression are different tokenizer problems." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "KV-cache continuation crash in serving and auto-docstring annotation handling are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but they target different models and different broken behaviors (cleanup-space behavior vs codec handling/warning)." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Deepseek tokenizer regression and `clean_up_tokenization_spaces` change are different tokenizer bugs." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow reporting after training failure and DeepSpeed Adafactor OOM are unrelated." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-patch helper and remote-code meta-init incompatibility are different multimodal model issues." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "Tokenizer cleanup behavior and Pegasus output differences across v4/v5 are related only at a high level, not the same bug." + }, + { + "left": "issue:41628", + "right": "issue:45459", + "accept": false, + "reason": "Importing `AutoImageProcessor` and tokenizer protobuf error masking are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 34, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30333", + "issue:30990", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:38175", + "issue:41628", + "issue:42757", + "issue:42831", + "issue:42886", + "issue:42898", + "issue:42994", + "issue:43295", + "issue:43479", + "issue:43576", + "issue:43644", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:44038", + "issue:44206", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44442", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44514", + "issue:44623", + "issue:44871", + "issue:44913", + "issue:44977", + "issue:44991", + "issue:45092", + "issue:45103", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45459", + "issue:45656" + ], + "soft_pairs": [ + "issue:44295|issue:45656", + "issue:44222|issue:44297", + "issue:42898|issue:43295", + "issue:43479|issue:43644", + "issue:44297|issue:45656", + "issue:44442|issue:45459", + "issue:44222|issue:44623", + "issue:42831|issue:44623", + "issue:44206|issue:45103", + "issue:41628|issue:42757", + "issue:22355|issue:28282", + "issue:44297|issue:45310", + "issue:44222|issue:44484", + "issue:44483|issue:45406", + "issue:43723|issue:44977", + "issue:44263|issue:44360", + "issue:30333|issue:36331", + "issue:44263|issue:45092", + "issue:30333|issue:34689", + "issue:30333|issue:35141", + "issue:28282|issue:30333", + "issue:44514|issue:45103", + "issue:44295|issue:44623", + "issue:43479|issue:44913", + "issue:43723|issue:44871", + "issue:44261|issue:45092", + "issue:44373|issue:44485", + "issue:43723|issue:44991", + "issue:43723|issue:44442", + "issue:30990|issue:38175", + "issue:45092|issue:45357", + "issue:42886|issue:42994", + "issue:41628|issue:43824", + "issue:42831|issue:45310", + "issue:42831|issue:44038", + "issue:42757|issue:45278", + "issue:43576|issue:43784", + "issue:44263|issue:44373" + ], + "prepared_review_unit_hash": "0e1319c4a3e411eee8e22e454f428779249a0162d0ecb665de5bc1ae0caa2f21", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38809, + "estimated_input_tokens": 9703, + "estimated_eval_tokens": 19662 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a mix of unrelated Transformers issues that happen to share broad themes like import failures, tokenizer/processor loading, and save_pretrained regressions. None of the soft-edge pairs look like true duplicates or mergeable PR-equivalents.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set: it explicitly describes multiple import regressions after a version upgrade, matching the recurring import/loading failure theme better than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative of the cluster because it captures the most general recurring breakage pattern here\u2014package import failures after upgrading\u2014while the other issues are more specific symptom reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different code paths: position_ids buffer handling vs Deepspeed optimizer stepping." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8/MoE saving and the other is tokenizer metadata mismatch." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "Tokenizer cleanup behavior change is unrelated to processor.tokenizer access and image-to-tokenizer regression." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Config None-default initialization is not the same bug as non-persistent buffer junk values." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer config export issue is unrelated to optimizer step duplication." + }, + { + "left": "issue:44442", + "right": "issue:45459", + "accept": false, + "reason": "One is a tokenizer class loading failure; the other is error masking when protobuf is absent." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Different save_pretrained failures affecting different object types and outputs." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "Accuracy regression in FineGrainedFP8 is unrelated to missing processor save files." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "Unsupported feature arg in a feature extractor is not the same as docstring parsing crashing on future annotations." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they concern different symbols, packages, and root causes." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Missing transformers.onnx and missing PyTorch are separate environment/import problems." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to Qwen3.5 from_pretrained failure." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "MoE FP8 save_pretrained and max_shard_size default are different save_pretrained behaviors." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat completions request validation is unrelated to Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading regression and flash-attention generation issue are different failures." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Different GLM-DSA bugs: torch.split outputs vs missing ReLU." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting on failed training and CustomTrainer signature mismatch are unrelated trainer issues." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "GlmMoeDsaIndexer split behavior is unrelated to InternVL2 meta-init compatibility." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "MLflow failure reporting and Llama 3.2 vision model loading are different problems." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "Failed job reporting is unrelated to embedding resize/post_init weight reinitialization." + }, + { + "left": "issue:28282", + "right": "issue:30333", + "accept": false, + "reason": "AutoModel missing PyTorch and MLflow job reporting are unrelated issues." + }, + { + "left": "issue:44514", + "right": "issue:45103", + "accept": false, + "reason": "Batched chat template crash and auto_docstring attribute error do not share the same bug path." + }, + { + "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:44913", + "accept": false, + "reason": "Default config initialization and GPTNeoX rotary_pct reload are different config bugs." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer loading failure and eos_token_id inconsistency are separate symptoms." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm is unrelated to InternVL2 meta initialization." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Wrong docstring for position_ids is not the same as GLM-5 RoPE implementation behavior." + }, + { + "left": "issue:43723", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-load complaints, but they involve different models and failure modes." + }, + { + "left": "issue:43723", + "right": "issue:44442", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer load failure is a model-specific tokenizer bug, not the same as generic v5 tokenizer-loading regressions." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading hang and SigLIP2 zero-probabilities are unrelated." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Meta-initialization incompatibility and incorrect visual encoder keys in save_pretrained are different bugs." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Offline tokenizer cache loading and quantized model saving failure are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different exported symbols and model families." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FP8 accuracy regression is not the same as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe bug reports do not describe the same defect." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Missing is_offline_mode is a specific symbol regression, while 45278 is a broader post-upgrade import-error umbrella." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Broken transformers env CLI and sentence-transformers import NameError are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "DSA indexer torch.split behavior and a position_ids docstring issue are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 35, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:29127", + "issue:30990", + "issue:33357", + "issue:36296", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42831", + "issue:42886", + "issue:42907", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43784", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44117", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44373", + "issue:44485", + "issue:44488", + "issue:44514", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44743", + "issue:44779", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45200", + "issue:45245", + "issue:45278", + "issue:45290", + "issue:45593", + "issue:45600" + ], + "soft_pairs": [ + "issue:43122|issue:44521", + "issue:43232|issue:45593", + "issue:44117|issue:44704", + "issue:42886|issue:43066", + "issue:43122|issue:44514", + "issue:42371|issue:44263", + "issue:30990|issue:33357", + "issue:42907|issue:44945", + "issue:44704|issue:45290", + "issue:43976|issue:44945", + "issue:44295|issue:44297", + "issue:43232|issue:44743", + "issue:39401|issue:44488", + "issue:42907|issue:43937", + "issue:43232|issue:45245", + "issue:43784|issue:45600", + "issue:41628|issue:45278", + "issue:43408|issue:43475", + "issue:44625|issue:44829", + "issue:44360|issue:44485", + "issue:43116|issue:45200", + "issue:39401|issue:44521", + "issue:44261|issue:44373", + "issue:29127|issue:30990", + "issue:43122|issue:44779", + "issue:44038|issue:44297", + "issue:42371|issue:42831", + "issue:42907|issue:44964", + "issue:44038|issue:44295", + "issue:36296|issue:43116", + "issue:42907|issue:43452", + "issue:44038|issue:44222", + "issue:43992|issue:44704" + ], + "prepared_review_unit_hash": "f127afb17721693d1657e908cb09d392c6af449c59144e1fed8dedcb187218ce", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 33, + "serialized_chars": 36562, + "estimated_input_tokens": 9141, + "estimated_eval_tokens": 18538 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is largely heterogeneous: the soft pairs share broad themes (tokenization, generation, save/load, multimodal), but they do not describe the same concrete bug or change, so none should be merged as duplicates. No PRs are present.", + "confidence": 0.94, + "canonical_issue_reason": "Issue 44704 is the clearest representative if one must be chosen: it names a specific, actionable code-path bug in `AutoProcessor.from_pretrained` forwarding kwargs to `cached_file`.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44704 is the best standalone issue here because it is narrowly scoped, reproducible, and points to a specific fix path in a common API.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Both are tokenizer-related, but one is a v4\u2192v5 tokenization regression and the other is an assistant-mask bug in multimodal chat templating." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Different subsystems: generation kwargs after `sync_gpus` vs D-FINE auxiliary losses with denoising off." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Related to loading paths, but one is tokenizer mapping fallback and the other is `AutoProcessor` kwarg forwarding; different bugs." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong tokenizer decoder type are distinct problems with different failure modes." + }, + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Generic tokenization drift is not the same as a batched `apply_chat_template` crash with `padding=False`." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API guidance and a GLM MoE `torch.split()` issue are unrelated." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Both concern loading, but one is a hang in Sentence Transformers and the other is a MacOS bus error on a CLIP model." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Saving dequantized models and incorrect pipeline-parallel LLM output are different code paths and fixes." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "`cached_file` kwarg forwarding is unrelated to chat template crashes with tool-call assistant messages." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version compatibility and pipeline-parallel inference correctness are not the same bug." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Reading a registered buffer and saving mismatched `tokenizer_class` are separate serialization/attribute issues." + }, + { + "left": "issue:43232", + "right": "issue:44743", + "accept": false, + "reason": "Generation state handling after `sync_gpus` is distinct from recurrent state reset in `modular_qwen3_5.py`." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Wrong `offset_mapping` and inability to load a specific tokenizer are different tokenizer bugs." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Dequantized model save failure is unrelated to invalid `GenerationConfig` validation for GLM-5." + }, + { + "left": "issue:43232", + "right": "issue:45245", + "accept": false, + "reason": "These involve unrelated failures: generation kwargs vs a category-count runtime limit." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Import-time `nn` NameError in sentence-transformers is unrelated to stale auto-mapping references in CI." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Missing `AutoImageProcessor` export is not the same as the broader set of import errors after upgrade." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Model-type warning on load and missing `fpn_position_embeddings` are different SAM3 issues." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "`num_labels` propagation and degenerate training under flash_attention_3 are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is not the same as the GLM-5 RoPE implementation discussion." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example-script multi-label output bug and Gemma 4 multimodal token-type defaults are distinct problems." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Wrong tokenizer offsets and zeroed assistant masks for multimodal chat are different behaviors in separate paths." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "A precision/config issue in MLA q_a_layernorm is unrelated to a wrong docstring for `position_ids`." + }, + { + "left": "issue:29127", + "right": "issue:30990", + "accept": false, + "reason": "An unclear layoutlmv3 error message and a Sentence Transformers loading hang are not the same bug." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both mention tokenizer behavior, but one is a version-wide tokenization change and the other is a DeepSeek-specific regression." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe/v5 model behavior and tokenizer save metadata mismatch are different issues." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 settings guidance is unrelated to FineGrainedFP8 accuracy problems." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Saving dequantized models and loading Phi-4 multimodal with latest transformers are different failures." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe compatibility and `position_ids` buffer access are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor-parallel training bugs and multi-label example-script behavior do not share the same underlying cause." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Dequantized model save failure is unrelated to `gguf_file` breaking tokenizer/model loading." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "A Qwen3-VL-Moe version bug and FP8 `save_pretrained` are different model/tooling issues." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Missing `embed_tokens.weight` in UMT5 loading is unrelated to `AutoProcessor` kwargs being dropped." + } + ] + }, + "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": 13, + "review_unit_index": 36, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:33453", + "issue:38617", + "issue:39401", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42757", + "issue:42907", + "issue:43064", + "issue:43116", + "issue:43122", + "issue:43278", + "issue:43441", + "issue:43493", + "issue:43673", + "issue:43784", + "issue:43824", + "issue:43825", + "issue:43827", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43937", + "issue:43976", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44222", + "issue:44351", + "issue:44485", + "issue:44779", + "issue:44936", + "issue:44945", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45216", + "issue:45278", + "issue:45341", + "issue:45405", + "issue:45446", + "issue:45478", + "issue:45561", + "issue:45563", + "issue:45588" + ], + "soft_pairs": [ + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:43122|issue:44977", + "issue:42757|issue:44351", + "issue:42907|issue:44222", + "issue:43278|issue:44945", + "issue:43784|issue:45446", + "issue:42222|issue:43441", + "issue:42673|issue:45072", + "issue:38617|issue:45278", + "issue:43064|issue:43673", + "issue:43937|issue:44485", + "issue:22355|issue:29127", + "issue:42371|issue:43874", + "issue:43116|issue:43493", + "issue:44075|issue:44936", + "issue:44222|issue:45478", + "issue:44038|issue:45478", + "issue:43976|issue:45446", + "issue:43116|issue:43825", + "issue:43937|issue:45563", + "issue:39401|issue:44779", + "issue:33453|issue:39401", + "issue:43931|issue:44945", + "issue:45083|issue:45446", + "issue:44117|issue:45588", + "issue:43827|issue:45446", + "issue:43937|issue:44977", + "issue:45341|issue:45561", + "issue:42673|issue:43906", + "issue:45405|issue:45446" + ], + "prepared_review_unit_hash": "1d40c377029041e73f9e8ad079b8911d5603d2de14603bbfc9c71bd3dc7706d0", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 35998, + "estimated_input_tokens": 9000, + "estimated_eval_tokens": 18256 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated regression and bug reports across import errors, tokenizer behavior, model saving/loading, and training/runtime issues. None of the soft pairs is a safe duplicate match from titles alone; the only plausible umbrella candidate is the broad import-regression report.", + "confidence": 0.68, + "canonical_issue_reason": "issue:45278 is the most umbrella-like report: it explicitly covers many import errors after upgrading to 5.5.0, so it is the best representative among these issues for a broad regression cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best global issue candidate because it is broad, regression-focused, and could naturally collect related import-breakage reports; the others are narrower, model-specific, or isolated to different code paths.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and likely different code paths; not the same bug." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5 and v5.4+, but one is save_pretrained corruption and the other is from_pretrained failure; different operations." + }, + { + "left": "issue:43122", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer regression vs flash-attention generation failure; different subsystems and symptoms." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports from different dependencies/modules; not a single underlying defect." + }, + { + "left": "issue:42907", + "right": "issue:44222", + "accept": false, + "reason": "Both concern saving quantized/dequantized models, but the model families and failure modes differ." + }, + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Embedding dtype mismatch during eval is unrelated to pipeline-parallel incorrect output." + }, + { + "left": "issue:43784", + "right": "issue:45446", + "accept": false, + "reason": "Sentence-transformers import NameError is unrelated to the PyTorch version check in flex_attention." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "VitPose import breakage and Ministral FlashAttention failure are distinct model-specific issues." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in Qwen3 threads is unrelated to dtype mismatches in bfloat16 inference." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "The latter is a broad import-regression report, but these are not the same concrete import failure." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Wrong optimizer states under FSDP/PEFT is unrelated to missing GenerationMixin cache in chunked prefill." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GenerationConfig validation error and RoPE implementation discussion are different bugs." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Transformers.onnx import failure and LayoutLMv3 error-message clarity are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance and a missing image-patch method are not the same issue." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script multi-label output bug and SigLIP2 implementation discrepancy are different model/workflow problems." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Unused SGD arguments and trainer.evaluate() after train() are separate training issues." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "One is FP8 save_pretrained, the other is from_pretrained on Qwen3.5 Moe; same family but different code paths." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3-VL-Moe and Qwen3.5 Moe are different model variants with different failures." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python support regression and a PyTorch import/version-check bug are not the same defect." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example script empty results and pipeline() error messaging are distinct bugs." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GenerationConfig invalidity and a stale num_return_sequences warning do not share the same code-path failure." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and report different incorrect behavior." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression and offset_mapping regression are not the same concrete bug." + }, + { + "left": "issue:43931", + "right": "issue:44945", + "accept": false, + "reason": "Weight-shape mismatch on model loading is unrelated to pipeline-parallel wrong outputs." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "Feature-extraction length helper behavior and a flex_attention import/version check are unrelated." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping None handling and flash_attention s_aux handling are separate runtime bugs." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() and a PyTorch version-check bug are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "GLM-5 generation config validation and Qwen3.5 flash-attention generation failure are different issues." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "A small testing_utils bug and a xdist race on captured_info.txt are not the same defect." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "A VRAM leak report and an isolated reproduction of another issue are not the same underlying bug." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "PEFT version pinning to an unreleased release and a PyTorch import/version-check bug are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 37, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42617", + "issue:42673", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43408", + "issue:43421", + "issue:43441", + "issue:43450", + "issue:43577", + "issue:43582", + "issue:43673", + "issue:43742", + "issue:43874", + "issue:43937", + "issue:43986", + "issue:44075", + "issue:44117", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44291", + "issue:44410", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44496", + "issue:44829", + "issue:44933", + "issue:45071", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45459", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45656" + ], + "soft_pairs": [ + "issue:44291|issue:45071", + "issue:43450|issue:43986", + "issue:39401|issue:43421", + "issue:44484|issue:45310", + "issue:44829|issue:45216", + "issue:38617|issue:41628", + "issue:44117|issue:45081", + "issue:44410|issue:45310", + "issue:42673|issue:43582", + "issue:42222|issue:45542", + "issue:42371|issue:44261", + "issue:42673|issue:43577", + "issue:43408|issue:43742", + "issue:43937|issue:44496", + "issue:38617|issue:44220", + "issue:42175|issue:42222", + "issue:44933|issue:45278", + "issue:43012|issue:43577", + "issue:44075|issue:45656", + "issue:44483|issue:45464", + "issue:42617|issue:45341", + "issue:44484|issue:45478", + "issue:43066|issue:43122", + "issue:43673|issue:45341", + "issue:44410|issue:45478", + "issue:43097|issue:44493", + "issue:43874|issue:44261", + "issue:42222|issue:45459", + "issue:43441|issue:44829", + "issue:43986|issue:44484", + "issue:43012|issue:43408", + "issue:44263|issue:44485", + "issue:41628|issue:45071", + "issue:45071|issue:45542" + ], + "prepared_review_unit_hash": "4e5fa95f5a5c333e1b3519b7ad53293ce9627066366d4e5592b0237bf34fdd46", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37647, + "estimated_input_tokens": 9412, + "estimated_eval_tokens": 19080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a noisy mix of unrelated Transformers regressions (imports, tokenizers, model loading, backend/config issues). I would not merge any of the soft-edge pairs as duplicates; the closest umbrella-style issue is the broad import-regression report.", + "confidence": 0.19, + "canonical_issue_reason": "issue:45278 is the broadest, most representative issue in the set: it describes multiple import errors after upgrading to 5.5.0, which can plausibly subsume some of the more specific import complaints here.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it is the widest-scope, user-facing regression report among otherwise diverse and mostly unrelated bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Different failure modes: unexpected model-init arg handling vs PreTrainedConfig type-checking regression." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "One is a batched video-shape bug, the other is an AutoProcessor crash due to missing torchvision." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Both are tokenizer-related, but one is offset mapping and the other is runtime post-processor updates for special tokens." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "Unrelated: save_pretrained shard-size question vs Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "Different code paths: flash_attention_3 training degeneration vs Qwen3.5 checkpoint save regression." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols/modules and different root causes." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer mapping None handling is not the same bug as the Mistral regex/backend_tokenizer crash." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Different model-specific problems: missing Qwen3Next projections vs Qwen3.5 MoE load failure." + }, + { + "left": "issue:42673", + "right": "issue:43582", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM threads is unrelated to an Apple Silicon TypeError in allocator warmup." + }, + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "Different subsystems: broken vitpose models vs TensorFlow/backend installation mismatch." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 settings API change is unrelated to MLA layernorm epsilon/config precision issues." + }, + { + "left": "issue:42673", + "right": "issue:43577", + "accept": false, + "reason": "A memory leak and a dtype-loading issue are different bugs." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Model-type warning for sam3_tracker does not match the MobileLLM key error." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "GenerationConfig validation and unrecognized model/config.json missing model_type are different loading errors." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for layer_type_validation is unrelated to the fbank feature extraction issue." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "Backend packaging/TF install problem is not the same as broken vitpose model behavior." + }, + { + "left": "issue:44933", + "right": "issue:45278", + "accept": false, + "reason": "45278 is broader, but 44933 is a specific image_utils import issue; not enough evidence they are the same underlying bug." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "A PyTorch precision warning during compilation is not the same as a model staying float32 after dtype loading." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "SGD args ignored and optimizer.step called twice are distinct optimizer bugs." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both mention chat/completions, but one is a broad v5.3 request acceptance problem and the other is a Qwen3.5 streaming failure." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "Not able to run 3d_parallel.py is unrelated to a testing_utils.py bug." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "max_shard_size default discussion is unrelated to the Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both are tokenizer regression reports, but one is decoder type and the other is changed tokenization behavior; not the same concrete bug." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache/chunked_prefill is unrelated to a testing_utils bug." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen-related failures: missing layer projections vs from_pretrained error." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removal of a config option and unexpected position-id keys are different migration/regression issues." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image-patch count helper is unrelated to the MLA rms_norm_eps/config precision bug." + }, + { + "left": "issue:42222", + "right": "issue:45459", + "accept": false, + "reason": "Broken vitpose models and protobuf masking tokenizer errors are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both mention flash attention, but the symptoms and affected paths are different enough to be separate bugs." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "Missing torchvision crash is unrelated to max_shard_size behavior." + }, + { + "left": "issue:43012", + "right": "issue:43408", + "accept": false, + "reason": "Different areas: bfloat16 compile warning vs sam3 tracker/model-type warning." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "GLM Moe torch.split indexing and GLM-5 RoPE implementation are different concrete bugs." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor import failure and PreTrainedConfig type checking break are separate regressions." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "PreTrainedConfig typing and TensorFlow backend installation issues do not match." + } + ] + }, + "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": 13, + "review_unit_index": 38, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:38617", + "issue:42757", + "issue:42886", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43335", + "issue:43352", + "issue:43381", + "issue:43408", + "issue:43452", + "issue:43502", + "issue:43531", + "issue:43550", + "issue:43824", + "issue:43825", + "issue:43873", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44351", + "issue:44368", + "issue:44380", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44704", + "issue:44829", + "issue:44987", + "issue:45003", + "issue:45071", + "issue:45103", + "issue:45245", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45588" + ], + "soft_pairs": [ + "issue:43931|issue:44410", + "issue:45003|issue:45341", + "issue:44117|issue:44987", + "issue:43986|issue:45341", + "issue:43097|issue:43452", + "issue:43931|issue:44038", + "issue:43335|issue:45071", + "issue:43408|issue:44368", + "issue:43874|issue:44263", + "issue:43976|issue:45071", + "issue:44829|issue:45245", + "issue:43986|issue:45372", + "issue:43531|issue:45341", + "issue:43064|issue:44315", + "issue:38617|issue:42757", + "issue:44261|issue:44485", + "issue:44410|issue:45375", + "issue:38617|issue:44351", + "issue:44062|issue:44117", + "issue:44038|issue:44410", + "issue:43097|issue:45588", + "issue:43986|issue:44117", + "issue:44704|issue:45103", + "issue:43502|issue:43992", + "issue:43097|issue:43873", + "issue:43992|issue:44484", + "issue:43824|issue:43986", + "issue:43066|issue:43906", + "issue:43352|issue:44380", + "issue:43825|issue:44655", + "issue:38617|issue:43097", + "issue:43012|issue:44368", + "issue:43012|issue:43550", + "issue:42886|issue:43381" + ], + "prepared_review_unit_hash": "0d9430a51f35ebaa3cc5c79f5ec2314fdaa96d7f009da4ad10966dcdffb3f601", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37877, + "estimated_input_tokens": 9470, + "estimated_eval_tokens": 19196 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of unrelated Transformers bug reports; the soft pairs are mostly superficial title matches and do not look like the same underlying fix or bug.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated import errors, loading failures, config regressions, warnings, and training-path bugs rather than one duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "If a representative issue is needed, issue 43824 is the strongest user-facing regression with multiple inbound references and a clear reproducible import failure, but it does not subsume the rest of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Both are Qwen-family, but one is a weight-shape loading mismatch and the other is missing layer projections in a different model path; not the same bug." + }, + { + "left": "issue:45003", + "right": "issue:45341", + "accept": false, + "reason": "Unsafe sys.modules access and a testing_utils bug are unrelated code paths." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "Tokenizer mapping returning None is a different failure from loading physical-intelligence/fast; not the same defect." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Video-model loading without torchvision and a testing utility bug are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Removed tie_embeddings flag and gguf_file loading breakage are different APIs and failure modes." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Similar model family, but these describe different Qwen3-VL-related regressions with different symptoms and code paths." + }, + { + "left": "issue:43335", + "right": "issue:45071", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug and PretrainedConfig type-checking regression are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:44368", + "accept": false, + "reason": "These are different warnings for different models/configs; not the same underlying issue." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "One is a missing image-processor method, the other is a torch.split indexing bug; different subsystems." + }, + { + "left": "issue:43976", + "right": "issue:45071", + "accept": false, + "reason": "Python-version compatibility breakage is not the same as PretrainedConfig type-checking." + }, + { + "left": "issue:44829", + "right": "issue:45245", + "accept": false, + "reason": "Degenerate training with flash_attention_3 and category-count runtime limits are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Missing torchvision during video processor loading and a mistral_common import error are distinct dependency failures." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "sliding_window behavior in Qwen3-MoE and a testing_utils bug are unrelated." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "FSDP2+PEFT optimizer-state corruption and Liger Kernel not applying with model_init affect different training mechanisms." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Import errors for different symbols from different modules are not the same bug." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "MLA layernorm epsilon precision and GLM-5 RoPE implementation are different model-internal issues." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Missing qkv/gate projections and a missing config field are different model/config regressions." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols and modules; not a duplicate." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken special-arg collision and TOKENIZER_MAPPING_NAMES returning None are separate tokenizer bugs." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Different Qwen-family architecture issues with different missing components and symptoms." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "tie_embeddings removal and s_aux=None crash in flash_attention are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "A video-model processor dependency crash and a tokenizer-mapping bug do not share the same code path." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "Passing kwargs to cached_file and docstring processing under future annotations are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Both are loading-related, but one is offline/local_files_only network leakage and the other is a missing embedding weight during from_pretrained; different fixes." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Removed tie_embeddings and offloading with quantization are unrelated issues." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing model weight on load and a save_pretrained shard-size question are different operations." + }, + { + "left": "issue:43824", + "right": "issue:43986", + "accept": false, + "reason": "Different model families and failure causes: import error for Qwen2.5-VL vs missing torchvision for video models." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue are not the same concrete bug." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "FlashAttention support rejection and GPT2 attention-scaling being ignored are distinct attention-path problems." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Pipeline translation-message regression and saving Pipeline objects are different features." + }, + { + "left": "issue:38617", + "right": "issue:43097", + "accept": false, + "reason": "ImportError for layer_type_validation and tie_embeddings removal are unrelated." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "A bfloat16 precision warning during compilation and a tie_word_embeddings warning are different warnings with different triggers." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Precision warning and a torch.compile SDPA failure are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:43381", + "accept": false, + "reason": "Offline cache loading and gradient checkpointing in eval mode are separate bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 39, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:36010", + "issue:36683", + "issue:38617", + "issue:42757", + "issue:42907", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43525", + "issue:43526", + "issue:43550", + "issue:43582", + "issue:43618", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43867", + "issue:43901", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44077", + "issue:44188", + "issue:44230", + "issue:44263", + "issue:44380", + "issue:44483", + "issue:44485", + "issue:44492", + "issue:44514", + "issue:44655", + "issue:44683", + "issue:44938", + "issue:45216", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45440", + "issue:45478", + "issue:45479", + "issue:45593" + ], + "soft_pairs": [ + "issue:42907|issue:44655", + "issue:43208|issue:43550", + "issue:44380|issue:45479", + "issue:43352|issue:43976", + "issue:43064|issue:44380", + "issue:43208|issue:43526", + "issue:44483|issue:44514", + "issue:43618|issue:44380", + "issue:36683|issue:43582", + "issue:43618|issue:43761", + "issue:43012|issue:44683", + "issue:44483|issue:45290", + "issue:42757|issue:43824", + "issue:44038|issue:44230", + "issue:44483|issue:45381", + "issue:43299|issue:43867", + "issue:38617|issue:43502", + "issue:43756|issue:44485", + "issue:43901|issue:44655", + "issue:42994|issue:44655", + "issue:43867|issue:45216", + "issue:36683|issue:43475", + "issue:44655|issue:45478", + "issue:43352|issue:43937", + "issue:43066|issue:45381", + "issue:44077|issue:44380", + "issue:44263|issue:45440", + "issue:43064|issue:44492", + "issue:36683|issue:44077", + "issue:43976|issue:44938", + "issue:36683|issue:43525", + "issue:36010|issue:42757", + "issue:44188|issue:44380", + "issue:43618|issue:45216", + "issue:43352|issue:45593", + "issue:43992|issue:45310" + ], + "prepared_review_unit_hash": "dc2e88a88cadc3ae328d838b1b02afb69e1e7839e46541eac314ef5d96bdf232", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38114, + "estimated_input_tokens": 9529, + "estimated_eval_tokens": 19314 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated Transformers bugs across different models, import paths, and save/load flows. The only clear duplicate-looking case in the provided candidates is the repeated Qwen3.5 Moe from_pretrained regression (45310 vs 45478).", + "confidence": 0.93, + "canonical_issue_reason": "issue:45310 is the better canonical for the duplicated Qwen3.5 Moe from_pretrained regression: it is earlier, more established, and matches the same symptom reported again in 45478.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the strongest representative issue because it is a concrete, model-specific regression report with clear reproduction context and it anchors the only near-exact duplicate in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Both concern saving, but one is a dequantized Minstral/Devstral model bug and the other is generic Pipeline save_pretrained serialization." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba torch.compile+SDPA failure are different models and different failure modes." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "GPT2 backend scaling regression is unrelated to the zero-loss classification bug with num_labels=1." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "Nemotron FlashAttention support and Python version compatibility are separate issues." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption is a different code path from GPT2 attention scaling under SDPA/FlashAttention." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training regressions and BeitImageProcessorFast label reduction are unrelated." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "One is the chat/completions endpoint rejecting requests; the other is a chat-template crash on batched input." + }, + { + "left": "issue:43618", + "right": "issue:44380", + "accept": false, + "reason": "CLIPOutput attentions missing is a distinct regression from GPT2 attention scaling behavior." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config missing vocab_size and Apple Silicon TypeError in caching_allocator_warmup are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Both are CLIP-related, but one drops attentions while the other returns hidden_states=None; different symptoms and fixes." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "A PyTorch warning during bfloat16 compile is not the same bug as compiled flex_attention failing on torch>=2.9." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "The v5 chat/completions request rejection and the assistant tool-call chat_template crash are different failures." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "These are separate import errors from different modules: huggingface_hub versus transformers." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Qwen3-VL-Moe loading regression is not the same as fp8 inference support for Qwen3-VL/Qwen3.5 MoE." + }, + { + "left": "issue:44483", + "right": "issue:45381", + "accept": false, + "reason": "Chat/completions API rejection is unrelated to the qwen2.5-vl video vision_position_ids bug." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Qwen3VL MoE loading failure and state_dict-sorted load error are different model-loading bugs." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "layer_type_validation import failure is unrelated to local_files_only still making network calls." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Smollm3 RoPE layer dropping and GLM-5 RoPE implementation are different model-specific problems." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "A docs mismatch about return_all_scores is not the same as Pipeline save_pretrained failure." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model saving failure and generic Pipeline saving failure are separate save-path bugs." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "State_dict sorting load error and Qwen3.5 save_pretrained regression are different checkpoint issues." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3Config vocab_size and Sam3VisionEncoderOutput missing fpn_position_embeddings are unrelated missing-attribute bugs." + }, + { + "left": "issue:44655", + "right": "issue:45478", + "accept": false, + "reason": "Pipeline save_pretrained failure is unrelated to the Qwen3.5 Moe from_pretrained regression." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig invalid are different model/setup bugs." + }, + { + "left": "issue:43066", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer decoder type mismatch is unrelated to qwen2.5-vl video position-id errors." + }, + { + "left": "issue:44077", + "right": "issue:44380", + "accept": false, + "reason": "PatchTSMixer post_init validation and GPT2 attention scaling are different issues." + }, + { + "left": "issue:44263", + "right": "issue:45440", + "accept": false, + "reason": "torch.split return-value handling in GlmMoeDsaIndexer is not the same as DeepseekV3MoE divergence from remote implementation." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "Optimizer-state corruption under FSDP2/PEFT is unrelated to a cache-strategy typo." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3Config missing vocab_size and PatchTSMixer post_init gating do not share the same bug." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Python-version compatibility problems here are different: one is Transformers 5.1.0 vs Python 3.9/3.10 constraints, the other is Python 3.14 load failure." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3Config lacking vocab_size and Llama4Config lacking pad_token_id are separate missing-field regressions." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "These are different import errors from different symbols and packages." + }, + { + "left": "issue:44188", + "right": "issue:44380", + "accept": false, + "reason": "Diverging attention kernels under torch.compile is a different backend-branching bug from GPT2 attention scaling." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIP attentions missing and Qwen3.5 saved checkpoint corruption are unrelated regressions." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Nemotron FlashAttention support and D-FINE auxiliary-loss handling are unrelated model bugs." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "UMT5Encoder missing embed_tokens.weight and Qwen3.5 Moe from_pretrained error affect different models and load paths." + } + ] + }, + "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": 13, + "review_unit_index": 40, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42548", + "issue:42890", + "issue:43010", + "issue:43064", + "issue:43065", + "issue:43232", + "issue:43352", + "issue:43502", + "issue:43575", + "issue:43618", + "issue:43723", + "issue:43867", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44075", + "issue:44077", + "issue:44112", + "issue:44222", + "issue:44242", + "issue:44246", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44387", + "issue:44484", + "issue:44556", + "issue:44749", + "issue:44857", + "issue:44938", + "issue:44964", + "issue:45071", + "issue:45125", + "issue:45278", + "issue:45341", + "issue:45357", + "issue:45375", + "issue:45381", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45588", + "issue:45600" + ], + "soft_pairs": [ + "issue:36683|issue:44242", + "issue:36683|issue:43352", + "issue:43010|issue:45278", + "issue:41669|issue:44246", + "issue:43867|issue:45375", + "issue:36683|issue:43064", + "issue:43992|issue:45478", + "issue:42890|issue:45561", + "issue:44938|issue:45071", + "issue:43867|issue:45357", + "issue:43010|issue:43976", + "issue:44938|issue:45381", + "issue:44360|issue:45440", + "issue:43618|issue:44077", + "issue:36296|issue:43575", + "issue:43867|issue:44484", + "issue:43618|issue:45588", + "issue:43618|issue:44857", + "issue:28282|issue:33453", + "issue:42890|issue:44964", + "issue:42890|issue:45600", + "issue:42890|issue:44112", + "issue:43502|issue:44062", + "issue:43723|issue:44556", + "issue:43010|issue:44075", + "issue:44222|issue:45125", + "issue:42890|issue:45341", + "issue:44387|issue:44556", + "issue:42548|issue:44351", + "issue:43010|issue:43065", + "issue:45125|issue:45478", + "issue:43010|issue:43232", + "issue:44355|issue:44749", + "issue:45125|issue:45375" + ], + "prepared_review_unit_hash": "ed52206f93f072cdd4ac73d55f3d7e148339e3e682b2bc57be4cb15ee7018d6f", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37357, + "estimated_input_tokens": 9340, + "estimated_eval_tokens": 18936 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a collection of distinct regressions and feature gaps, not true duplicates. There is a loose theme around import/load regressions and some Qwen/Sam model-family issues, but the soft pairs do not describe the same concrete bug or change.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45278 is the broadest, most umbrella-like report in the set: a general post-upgrade import regression with multiple failing imports, which best fits the cluster\u2019s main import-related overlap.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best representative issue because it is the most general symptom report and could plausibly subsume nearby import/export complaints better than the narrower, model-specific tickets.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "Gemma3Config missing vocab_size and load-balancing loss are unrelated code paths." + }, + { + "left": "issue:36683", + "right": "issue:43352", + "accept": false, + "reason": "Config attribute error is unrelated to Flash Attention support gating." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Cache update no_grad semantics and import errors after upgrade are different bugs." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Import-star slowdown and intermittent import slowness may be related in theme, but not the same concrete issue." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "State-dict sorting load failure is unrelated to a missing Qwen3.5 vision config field." + }, + { + "left": "issue:36683", + "right": "issue:43064", + "accept": false, + "reason": "Gemma3Config vocab_size error and FSDP2/PEFT optimizer-state corruption are unrelated." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "Both are from_pretrained/load problems, but they affect different models and different failure causes." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Missing set_seed in a test and xdist file races are separate testing issues." + }, + { + "left": "issue:44938", + "right": "issue:45071", + "accept": false, + "reason": "Python 3.14 import failure and PreTrainedConfig type-checking breakage are distinct regressions." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "State-dict ordering load error and incorrect visual encoder keys during save are different sides of model serialization." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Generation cache updates and Python-version compatibility are unrelated." + }, + { + "left": "issue:44938", + "right": "issue:45381", + "accept": false, + "reason": "Python 3.14 load failure and Qwen2.5-VL video position-id bug are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "DSA indexer missing ReLU and DeepSeekV3 implementation divergence are not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "CLIPOutput attentions regression and patchtsmixer post_init policy are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor parallel training bug and model load OOM under TP are different failure modes." + }, + { + "left": "issue:43867", + "right": "issue:44484", + "accept": false, + "reason": "Load failure on sorted state_dict and a question about save_pretrained shard size are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:45588", + "accept": false, + "reason": "Missing CLIPOutput attentions and flash_attention s_aux=None crash are separate issues." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "CLIPOutput attentions regression is unrelated to LwDetrImageLoss AMP crash." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "PyTorch-missing ImportError and tokenizer-loading regression are different import/load problems." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "A flaky test and a Phi-4 multimodal load failure are unrelated." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Test seed flakiness and removed Sam3LiteText config references are different CI issues." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "Missing set_seed in SamHQ tests and GraniteSpeech stale device override are unrelated test failures." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Local-files-only network leakage and AddedToken keyword collision are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading regression in v5 and checkpoint reload failure after upgrade are separate issues." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Int4 CUDA memory growth and checkpoint reload compatibility are unrelated." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Missing PreTrainedModel export and missing HybridCache export are different symbol-export regressions." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained for MoE and missing _tp_plan for Qwen3_5Moe are different issues." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Missing tensor-parallel plan and from_pretrained error for Qwen3.5 MoE are not the same code-path bug." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Cache/Layer no_grad decoration and generation kwargs sync_gpus handling are unrelated." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled Python file errors and post-upgrade filtering slowdown are different regressions." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Missing _tp_plan and missing deepstack_visual_indexes are both Qwen3.5-related, but they are different config/serialization bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 41, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41628", + "issue:41669", + "issue:42371", + "issue:42548", + "issue:42757", + "issue:42831", + "issue:43010", + "issue:43299", + "issue:43502", + "issue:43526", + "issue:43576", + "issue:43673", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43901", + "issue:44038", + "issue:44222", + "issue:44230", + "issue:44261", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44368", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44655", + "issue:44908", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45468", + "issue:45542" + ], + "soft_pairs": [ + "issue:43502|issue:44485", + "issue:30990|issue:31795", + "issue:44315|issue:44360", + "issue:30333|issue:33453", + "issue:43010|issue:44368", + "issue:43502|issue:44222", + "issue:43502|issue:44961", + "issue:43299|issue:44556", + "issue:44355|issue:44655", + "issue:44222|issue:44230", + "issue:30333|issue:36296", + "issue:44355|issue:45071", + "issue:42831|issue:43874", + "issue:33453|issue:34689", + "issue:42757|issue:43576", + "issue:33453|issue:36296", + "issue:43827|issue:44961", + "issue:43576|issue:45071", + "issue:43576|issue:43824", + "issue:38617|issue:43576", + "issue:43673|issue:44492", + "issue:44961|issue:45468", + "issue:31795|issue:38617", + "issue:41628|issue:42548", + "issue:43576|issue:44351", + "issue:44908|issue:44961", + "issue:44485|issue:44961", + "issue:36296|issue:43901", + "issue:36296|issue:43526", + "issue:41628|issue:44355", + "issue:44273|issue:45003", + "issue:44261|issue:44360", + "issue:44273|issue:45542", + "issue:44623|issue:45230", + "issue:43576|issue:44556", + "issue:44038|issue:45125", + "issue:44393|issue:45071", + "issue:41669|issue:43673", + "issue:44556|issue:45092", + "issue:42371|issue:44492" + ], + "prepared_review_unit_hash": "1789773476e0aae6ebd5601fef9749378a4947df56ed7aeba4ef6980d1708413", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 40, + "serialized_chars": 38963, + "estimated_input_tokens": 9741, + "estimated_eval_tokens": 19738 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a loose cluster of Transformers v5-era regressions, import/load failures, and a few unrelated bug reports; there is no single obvious duplicate chain, so this looks more like theme-based similarity than one underlying issue.", + "confidence": 0.36, + "canonical_issue_reason": "issue:43576 is the broadest representative of the set: it concerns a v5 breakage and sits closest to several other loading/import regressions, even though the cluster is too heterogeneous to treat as one duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43576 is the best umbrella issue here because it captures the broadest v5-related breakage theme among the candidates, while the other issues are narrower or clearly unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "Different problems: offline/local-files-only network behavior versus GLM-5 RoPE implementation." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "One is a loading hang report; the other is documentation confusion about forward arguments." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Different subsystems and symptoms: model_init/Liger Kernel versus a DSA indexer activation issue." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "Training-job MLflow reporting failure is unrelated to tokenizer loading regression." + }, + { + "left": "issue:43010", + "right": "issue:44368", + "accept": false, + "reason": "Cache update no_grad semantics and a tie_word_embeddings warning are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Both mention model behavior, but one is local file loading and the other is FP8 save_pretrained for MoE." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No shared bug path; API request suppression versus an unrelated placeholder title." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Both are loading-related version regressions, but they affect different model families and likely different code paths." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Compiled Python file errors are unrelated to saving Pipeline objects." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both involve FP8/MoE, but one is save_pretrained and the other is inference support; not the same fix." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow failure on training job vs tensor-parallel training bug are distinct issues." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file runtime errors and PretrainedConfig type checking are different failure modes." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "FineGrainedFP8 accuracy and a missing image-patch method in GLM46V are unrelated." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Both are loading regressions, but one is tokenizer loading and the other is a specific Llama 3.2 vision model loading breakage." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Importing from huggingface_hub and a broken transformers env command are not the same underlying bug." + }, + { + "left": "issue:33453", + "right": "issue:36296", + "accept": false, + "reason": "Tokenizer loading regression and tensor-parallel training bug are unrelated." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline() is unrelated to the placeholder racoon issue." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "A broken env command and a PretrainedConfig type-checking regression do not share a concrete fix." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "The env command issue is separate from a missing Qwen2_5_VLForConditionalGeneration import." + }, + { + "left": "issue:38617", + "right": "issue:43576", + "accept": false, + "reason": "Missing layer_type_validation import and a broken env command are different export/CLI problems." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression and a typo in cache strategies are not the same bug." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "Placeholder racoon issue does not match Gemma4 audio positional encoding." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion about model.forward is unrelated to an ImportError." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Both are top-level import errors, but they concern different symbols and are not clearly the same export bug." + }, + { + "left": "issue:43576", + "right": "issue:44351", + "accept": false, + "reason": "Broken env command and missing HybridCache import are separate issues." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs handling has nothing to do with the placeholder racoon issue." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "GLM-5 RoPE implementation and the placeholder racoon issue are unrelated." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Lazy loading failure and unsafe sys.modules access may be related, but the titles are not specific enough to treat as the same bug with confidence." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor parallel training bug and TextClassificationPipeline docs mismatch are unrelated." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor parallel training bug is unrelated to BeitImageProcessorFast reduce_labels behavior." + }, + { + "left": "issue:41628", + "right": "issue:44355", + "accept": false, + "reason": "Importing AutoImageProcessor and errors from compiled Python files are different failure classes." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading problems and a TensorBoard/TensorFlow backend error are not the same issue." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "Missing files on processor.save_pretrained and a generic bug report are not the same concrete bug." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "A broken env command in v5 and checkpoint reload incompatibility across versions are related only at a broad version-breakage level." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Qwen3-VL-Moe bug report and missing _tp_plan for Qwen3_5Moe are different model-specific issues." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL bounding-box output error and PretrainedConfig type checking are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Removing import * usage and a GenerationMixin cache regression are not the same bug/change." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both concern version incompatibilities, but they target different model-loading paths and aren\u2019t a single concrete fix." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 settings guidance and a cache-strategy typo are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 42, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:42913", + "issue:43066", + "issue:43352", + "issue:43408", + "issue:43502", + "issue:43519", + "issue:43576", + "issue:43606", + "issue:43704", + "issue:43723", + "issue:43827", + "issue:43976", + "issue:44077", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44704", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44945", + "issue:44998", + "issue:45071", + "issue:45092", + "issue:45230", + "issue:45290", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45478" + ], + "soft_pairs": [ + "issue:42890|issue:44908", + "issue:44704|issue:44908", + "issue:44246|issue:44273", + "issue:43408|issue:44315", + "issue:43352|issue:44393", + "issue:43576|issue:44908", + "issue:43976|issue:44393", + "issue:44297|issue:45230", + "issue:44908|issue:45071", + "issue:44295|issue:45230", + "issue:41669|issue:44351", + "issue:41669|issue:45341", + "issue:44492|issue:44945", + "issue:41669|issue:44492", + "issue:44393|issue:45478", + "issue:44315|issue:44829", + "issue:31795|issue:33453", + "issue:43827|issue:44908", + "issue:44077|issue:44861", + "issue:41669|issue:44908", + "issue:43723|issue:44393", + "issue:44655|issue:44861", + "issue:43576|issue:44273", + "issue:44360|issue:45468", + "issue:45230|issue:45310", + "issue:44273|issue:44908", + "issue:41669|issue:43704", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:31795|issue:34689", + "issue:31515|issue:31795", + "issue:44315|issue:45092", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:42371|issue:43704", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:43519|issue:43704", + "issue:44263|issue:45468", + "issue:44261|issue:45468", + "issue:43606|issue:45468", + "issue:42890|issue:43704", + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44861|issue:45290", + "issue:44998|issue:45341", + "issue:44492|issue:44998" + ], + "prepared_review_unit_hash": "8cc1a61a8796759623853fd4a44409bcce5464a1d6ecc362934b4042a841965e", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 47, + "serialized_chars": 41097, + "estimated_input_tokens": 10275, + "estimated_eval_tokens": 20806 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose cluster of mostly Transformers v5 loading/tokenizer/save_pretrained regressions, mixed with unrelated performance, test, and model-specific issues. Very few pairs look like true duplicates.", + "confidence": 0.64, + "canonical_issue_reason": "issue:43723 is the most central and broadly phrased loading regression here: AutoTokenizer.from_pretrained failing in v5. It best anchors the tokenizer/loading subset of the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43723 is the best representative issue for the cluster because it describes the core v5 tokenizer-loading problem without being overly narrow to one model or edge case.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Different subsystems: test seeding in SAM HQ vs inverse_sqrt scheduler kwargs." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "Processor file-resolution kwargs vs scheduler configuration are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Both mention loading/slow behavior, but one is import time and the other is lazy loading; not the same bug." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Model-type mismatch warning vs model_init Liger kernel application are different issues." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention support error vs Qwen3-VL bbox hallucination/error are unrelated model-specific bugs." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Broken CLI env command vs scheduler kwarg handling are unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility vs Qwen3-VL bounding-box output are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained mismatch vs generic bug report; no concrete shared failure mode." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug vs PretrainedConfig type-checking regression are different code paths." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "Buffer/position_ids access bug cannot be tied to the generic report." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "Import-star performance regression vs missing HybridCache import are not the same issue." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import performance regression vs a testing_utils bug are unrelated." + }, + { + "left": "issue:44492", + "right": "issue:44945", + "accept": false, + "reason": "Cache-strategy typo vs pipeline-parallelism output bug are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44492", + "accept": false, + "reason": "Import-star performance regression vs a cache-strategy typo are not duplicates." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Different model families and different from_pretrained failure modes." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "model_init/Liger application bug vs flash_attention_3 training degeneration are distinct concrete problems." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion in forward args vs tokenizer loading regression are different." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Docs still referencing pipeline vs scheduler kwargs bug are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "patchtsmixer post_init policy vs tied-weight-key AttributeError are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44908", + "accept": false, + "reason": "Import performance regression vs scheduler kwargs handling are unrelated." + }, + { + "left": "issue:43723", + "right": "issue:44393", + "accept": false, + "reason": "Tokenizer loading regression vs Qwen3-VL bbox output bug are unrelated." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline save_pretrained issue vs _get_tied_weight_keys AttributeError are different save/load paths." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "CLI env command bug vs lazy loading issue are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer ReLU bug vs Gemma audio positional encoding bug are unrelated." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report lacks enough detail to establish the same from_pretrained regression." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading problem vs scheduler kwargs problem are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-time performance issue vs VRAM leak in Qwen3ForCausalLM are different bugs." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Timestamp calculation in Qwen3VL processor vs GLM-5 RoPE implementation are different model-specific bugs." + }, + { + "left": "issue:42890", + "right": "issue:43704", + "accept": false, + "reason": "Test seeding flakiness vs VRAM leak are unrelated." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both are tokenizer-v5 related, but one is general behavior drift and the other is a specific decoder-type mismatch; not enough to call them the same bug." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Slow checkpoint loading vs documentation confusion are different issues." + }, + { + "left": "issue:44315", + "right": "issue:45092", + "accept": false, + "reason": "Both involve model creation/loading, but the concrete failures differ: Liger kernel not applied vs remote-code meta-init incompatibility." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in dataloader threads vs RoPE implementation bug are unrelated." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Different model implementations and different failure modes." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 settings guidance vs VRAM leak are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "local_files_only network leakage vs Qwen3VL timestamp math are different bugs." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Forward-API docs confusion vs Llama model-loading regression are unrelated." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading vs forward-API documentation confusion are different." + }, + { + "left": "issue:44315", + "right": "issue:45092", + "accept": false, + "reason": "Not the same concrete code-path problem; one is kernel application, the other remote-code/meta-init handling." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "Different model bugs with no shared underlying change." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 config note vs VRAM leak are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Network access despite local_files_only vs timestamp miscalculation are unrelated." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Related topic, but not the same concrete tokenizer defect." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug vs unrelated placeholder issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight-key AttributeError vs chat-template tool-call crash are different code paths." + }, + { + "left": "issue:44998", + "right": "issue:45341", + "accept": false, + "reason": "Unrelated placeholder issue vs testing_utils bug." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Typo in cache strategies vs unrelated placeholder issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 13, + "review_unit_index": 43, + "review_unit_count": 43, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-15354-375", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:33453", + "issue:36296", + "issue:38617", + "issue:43298", + "issue:43525", + "issue:43572", + "issue:43618", + "issue:44230", + "issue:44556", + "issue:44623", + "issue:44861", + "issue:44998", + "issue:45310" + ], + "soft_pairs": [ + "issue:43618|issue:44998", + "issue:44556|issue:44998", + "issue:43525|issue:44861", + "issue:43572|issue:44861", + "issue:31515|issue:38617", + "issue:44230|issue:44623", + "issue:31515|issue:36296", + "issue:30333|issue:31515", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230" + ], + "prepared_review_unit_hash": "568e9da5ac7036f2042b93d9219b1919aead051afe305b4c78b848ee444cc123", + "component_budget": { + "node_count": 375, + "item_count": 375, + "soft_pair_count": 1399, + "serialized_chars": 722206, + "estimated_input_tokens": 180552, + "estimated_eval_tokens": 361360 + }, + "budget": { + "node_count": 15, + "item_count": 15, + "soft_pair_count": 11, + "serialized_chars": 11708, + "estimated_input_tokens": 2927, + "estimated_eval_tokens": 6110 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly driven by generic similarity around loading/import/config errors, but the pairs do not look like the same underlying bug. I would not merge any of the soft edges.", + "confidence": 0.87, + "canonical_issue_reason": "issue:44556 is the clearest concrete regression report in the set: checkpoint reload failures after upgrading versions. It is more actionable and broadly representative than the narrower model-specific or unrelated items.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44556 is the best single representative issue: it states a specific from_pretrained/reload regression with version context and is more directly triageable than the other titles.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIPOutput attention assignment regression vs an unrelated nonsense issue title; no shared bug." + }, + { + "left": "issue:44556", + "right": "issue:44998", + "accept": false, + "reason": "Checkpoint reload regression vs unrelated issue; the similarity is only superficial text noise." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_id on a config class and an AttributeError in tied-weight key handling are different code paths." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "StableLmConfig pad_token_idx regression and _get_tied_weight_keys list-handling crash are unrelated failures." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading is not the same as an ImportError for layer_type_validation." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 Qwen3-VL/MoE support request and processor.save_pretrained missing files are different issues." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Loading slowdown and tensor parallel training bug are unrelated subsystems and symptoms." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLflow job status reporting on failure is unrelated to checkpoint loading speed." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 inference support request and a from_pretrained error in Qwen3.5 MoE are adjacent only at the model family level, not the same bug." + }, + { + "left": "issue:31515", + "right": "issue:33453", + "accept": false, + "reason": "Checkpoint loading slowness is different from a tokenizer loading regression." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching audio-model support and FP8 Qwen3-VL/MoE support are different feature gaps." + } + ] + }, + "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": 13, + "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": "b2e28509e7f45ff200e1213557c3efa19a970bca947a336aec1a85f193d954a3", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1589, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate feature requests for different model-loading/attention backends: Flash Attention 2 support versus GGUF support. They share a broad \u2018more architectures\u2019 framing, but not the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 26350 is a clearer representative only in the sense that it is the older and more discussed request, but it is not a true canonical for the other issue because it targets a different feature area.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 26350 has more discussion activity and is the older community contribution request, so it is the better cluster representative if one must be chosen.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Reject: Flash Attention 2 support and GGUF support are different capabilities with different code paths and implementation goals; they are not duplicates of the same underlying request." + } + ] + }, + "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": 13, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-4", + "nodes": [ + "issue:27088", + "issue:43834", + "issue:44365", + "issue:44370" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:44365|issue:44370" + ], + "prepared_review_unit_hash": "f8263581d7ec624aa8ba217d4a0d6c3b86667d7f48eed0197cbe25a458fe7538", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4455, + "estimated_input_tokens": 1114, + "estimated_eval_tokens": 2484 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4455, + "estimated_input_tokens": 1114, + "estimated_eval_tokens": 2484 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are all i18n documentation-translation issues, but they are not the same underlying task: #27088 is specifically for Turkish, while #43834, #44365, and #44370 are generic template-style issues for unspecified languages. They look similar textually, but represent different translations rather than duplicates.", + "confidence": 0.9, + "canonical_issue_reason": "#27088 is the most concrete and actionable issue because it names a specific target language (Turkish) instead of the generic placeholder wording used by the others.", + "canonical_pr_reason": null, + "best_issue_reason": "#27088 is the best representative issue since it is specific, non-placeholder, and clearly scoped to one translation effort.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Same translation template family, but Turkish vs. an unspecified language target means they are different tasks, not duplicates." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "These are structurally similar issue templates, but they refer to different language targets, so they should not be merged." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "The titles are similar, but one is concrete Turkish and the other is a generic placeholder issue; not the same underlying request." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "Both are generic i18n template issues, but there is no evidence they target the same exact language or translation effort." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "Highly similar wording, but placeholder language-code/language-name entries indicate separate translation requests rather than one duplicate bug." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "They look like cloned templates, but without the same explicit language target they are not safely duplicate-identical." + } + ] + }, + "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": 13, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-14", + "nodes": [ + "issue:36817", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:37015|issue:42503", + "issue:37015|issue:45027", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42738", + "issue:42738|issue:45027", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43671|issue:43909", + "issue:43837|issue:43909", + "issue:43837|issue:45322", + "issue:43837|issue:44121", + "issue:43909|issue:44205", + "issue:44121|issue:45322", + "issue:43837|issue:45027", + "issue:43837|issue:45177" + ], + "prepared_review_unit_hash": "ed9c3d5f92306e8f7d99385858fdba72b519907f63d89e4eb1eee77785cae922", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 28, + "serialized_chars": 17645, + "estimated_input_tokens": 4412, + "estimated_eval_tokens": 9080 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 28, + "serialized_chars": 17645, + "estimated_input_tokens": 4412, + "estimated_eval_tokens": 9080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a loose semantic group of unrelated model-request issues across vision, speech, OCR, and BERT-like architectures. No soft pairs look like the same underlying bug/change, so none should be merged as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "issue:42738 is the broadest umbrella in the set ('BERT-like models with RoPE'), but the cluster is still too heterogeneous for a true duplicate canonical.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42738 is the best representative only by breadth, not because it matches the other requests; it is the least specific of the issues listed.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "EuroBert config and DEIMv2 are different model families with no shared concrete fix." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config and Qwen3-TTS are unrelated modalities (text model config vs speech TTS support)." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different concrete models; this is not the same request." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT and a weight-sparse transformer request target different architecture work." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio are unrelated model requests in different domains." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are distinct model families, not the same underlying change." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS are different speech/model support requests." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 and Qwen3-ASR are different model additions, not duplicates." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 support and a weight-sparse transformer request are different concrete changes." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 and ModernVBERT are separate model families." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 and Voxtral TTS are unrelated across model family and modality." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT and Qwen3-ASR are different concrete model support requests." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "ModernVBERT and Qwen3-TTS are different model families/modalities." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT is a specific model request, while BERT-like models with RoPE is a broad architectural umbrella; not the same issue." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "BERT-like RoPE models and Voxtral TTS are unrelated requests." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are related speech area requests but different concrete model support changes." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-TTS and LFM2.5 Audio are different speech/audio model requests." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS and weight-sparse transformer are unrelated work items." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Both are TTS-related, but they target different models and are not the same concrete change." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Speech TTS support and DeepSeek-OCR2 are different modalities and code paths." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio model requests." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE target different model families and tasks." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "ASR support and a sparse-transformer architecture request are not the same underlying change." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are unrelated modalities." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer and EUPE are different architecture/model requests." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "ASR and TTS are different speech tasks with separate implementations." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "ASR support and OCR model support are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 13, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-37934-6", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43813", + "issue:44031" + ], + "soft_pairs": [ + "issue:37934|issue:43668", + "issue:42936|issue:43240", + "issue:42936|issue:44031", + "issue:42936|issue:43813", + "issue:42936|issue:43668", + "issue:43240|issue:43668", + "issue:43668|issue:44031", + "issue:43668|issue:43813" + ], + "prepared_review_unit_hash": "f03f7869256ab5f3c3eaae8eb5cce1256d3dab1b0accfdb14d1b9214cf225b9d", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 8, + "serialized_chars": 6293, + "estimated_input_tokens": 1574, + "estimated_eval_tokens": 3404 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 8, + "serialized_chars": 6293, + "estimated_input_tokens": 1574, + "estimated_eval_tokens": 3404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are not duplicates as a cluster; the titles describe separate bugs spanning different models, configs, and warning paths. The soft similarity is mostly superficial wording overlap.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Different components and bug types: Llama4 norm implementation question vs ModernBERT config type hint. No shared code path." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Both involve loss-related model code, but one is Mask2Former's unused ignore_value and the other is kwargs not being forwarded to a loss function. Different concrete defects." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former loss handling vs tokenizer regex warnings after a version change; unrelated subsystems and symptoms." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2Former model bug vs a typo in a quantization attribute name. Not the same underlying issue." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Unrelated model-loss bug vs ModernBERT config annotation issue." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "One is about forwarding kwargs into loss calculation; the other is an incorrect type hint. Different bug classes and code paths." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "Config type hint error vs tokenizer warning regression. No evidence they share a fix." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "ModernBERT config annotation issue vs a typo in a quantization symbol. Different files and changes." + } + ] + }, + "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": 13, + "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": "693908d4afb82957a1dbf0e251754131d5f0928c1ee2034a1bcc72df8163724b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1479, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1479, + "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 not duplicates: one is about Set Block Decoding and the other is a DeepSeek v3.2 support request.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 43037 is the more specific and actionable issue; 41084 is vague and appears to describe a different topic.", + "canonical_pr_reason": null, + "best_issue_reason": "43037 is the best issue to represent this set because its title states a concrete feature request, while 41084 is too ambiguous.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Reject: the titles point to unrelated concerns, with no shared targets, filenames, or explicit linkage indicating 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": 13, + "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": "961c486aa4857ce960c12c71c606d39047858f6b10168b94629e939dbcfa6b89", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues both mention Trainer customization, but they address different behaviors: one is about `remove_unused_column`, the other about FLOPS estimation with custom trainers. They do not look like the same underlying bug or change.", + "confidence": 0.92, + "canonical_issue_reason": "Issue #43598 is the clearer, more specific feature/behavior request around `remove_unused_column` customizability in Trainer.", + "canonical_pr_reason": null, + "best_issue_reason": "#43598 is the best representative issue because it is narrowly scoped and describes a concrete Trainer behavior to revisit.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "These are related only at a broad Trainer/customization level; one concerns column removal logic, the other FLOPS estimation for custom trainers. Different code paths and fixes." + } + ] + }, + "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": 13, + "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": "d11b7211d10638fd8e7c726ac91edd7fcd9d50882ce3aeb5aa7a8223796d52c8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related only at a very broad level (both request new Qwen model support), but they target different model families and different AutoModel classes, so they are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying requests: one asks for Qwen3-Omni registration in AutoModel/AutoModelForConditionalGeneration, the other asks for AutoModelForSequenceClassification support for Qwen3.5. They do not describe the same bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 13, + "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": "aad5f9af27c90c100a61759884863d662215881ebdd249e333ac00ddd2a1dfcd", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4265, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The three \"huggingface \" issues look like a likely duplicate cluster, with 43845 as the best representative; 43835 ('Fett') appears unrelated. No PRs are present.", + "confidence": 0.41, + "canonical_issue_reason": "43845 is the earliest of the near-identical \"huggingface \" reports and is the best representative of that cluster; 43835 has a different title and does not fit.", + "canonical_pr_reason": null, + "best_issue_reason": "43845 is the cleanest canonical candidate because it is earliest and shares the same title pattern as 43846 and 43859.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "Different title and no shared targets/files or content evidence; likely unrelated." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": false, + "reason": "Titles match, but there is no description or target detail to prove they are the same underlying issue." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": false, + "reason": "Same broad title, but the available metadata is too thin to confirm a single underlying bug/change." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": false, + "reason": "Same broad title, but the available metadata is too thin to confirm a single underlying bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 11, + "component_count": 13, + "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": "5efaaf7e352bd932c952e40cd7fafea3360f8771eec3fbe0c9369baf4d7bb191", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5794, + "estimated_input_tokens": 1449, + "estimated_eval_tokens": 3154 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly low-signal spam/generic issues with no concrete shared bug or change, so there is no clear duplicate set or canonical representative.", + "confidence": 0.92, + "canonical_issue_reason": "No canonical issue: the items are either spam placeholders or unrelated generic titles, with no substantive body text, targets, or shared technical context.", + "canonical_pr_reason": null, + "best_issue_reason": "If forced to pick a representative, issue 44418 is the only non-spam-looking item, but it still does not relate to the others and is not a strong cluster anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are just '' with no descriptive content; identical titles alone are not enough to establish the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Same as above: no body details or technical context, so there is no basis to treat them as duplicates of one bug." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "One is spam and the other is a generic documentation request; these do not indicate the same underlying problem." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "'' and 'Check out \"Google LLC\"' are clearly unrelated on their face, with no shared issue description." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "A documentation enhancement suggestion is not the same as a spam placeholder; no shared concrete topic is visible." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "The titles suggest unrelated content, and there is no body evidence tying them to the same bug or change." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Both are spam placeholders, but there is no substantive evidence they refer to the same underlying report." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "These titles are unrelated and provide no common technical context or explicit link." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "No shared bug/change is evident; one is spam and the other is an unrelated promotional-looking title." + } + ] + }, + "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": 13, + "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": "aed16a21dd49a1b185838ebfa549975ae96756705e354436ee70da2041d94afa", + "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": "These two closed issues appear highly similar and likely refer to the same energy-efficiency proposal, but the evidence is soft-only, so the match should remain tentative rather than treated as a confirmed duplicate.", + "confidence": 0.87, + "canonical_issue_reason": "Issue #45145 is the better canonical thread because it was created first and has slightly more discussion signal (an inbound reference).", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the strongest representative issue: earliest created, same title/topic, and marginally more established in the graph.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": false, + "reason": "Reject: the pair is supported only by soft similarity, and deterministic_accept is false; identical titles alone are not enough to confirm a duplicate." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "Reject: the pair is only supported by soft similarity, and the packet\u2019s deterministic_accept is false. The analyst\u2019s soft-edge verdict is too strong for the evidence. The summary also overstates certainty by calling them duplicates; it should stay more tentative (e.g. \u201cappear highly similar\u201d)." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 13, + "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": "pull_request", + "component_index": 1, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-39895-13", + "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:45587", + "pull_request:45655" + ], + "soft_pairs": [ + "pull_request:39895|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:41855|pull_request:43424", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45587", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "9beab45c0cdf32ca446e10094003609c32cf583151a269e0617e184210320c2b", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 35210, + "estimated_input_tokens": 8803, + "estimated_eval_tokens": 17862 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 18, + "serialized_chars": 35210, + "estimated_input_tokens": 8803, + "estimated_eval_tokens": 17862 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated PRs across different models, docs, and utilities. The only true duplicate is the pair of `fixed_cross_entropy` PRs (`43251` and `43254`), which address the same loss-function kwargs passthrough bug. Everything else should be kept separate.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "`43251` is the best canonical PR for the duplicate loss-function fix: it explicitly links issue `43240`, has the fuller kwargs handling (`weight`, `label_smoothing`), and the broader review context confirms it targets the same concrete bug as `43254`.", + "best_issue_reason": null, + "best_pr_reason": "`43251` is the strongest representative PR in the only duplicate pair because it is explicitly issue-backed and directly fixes the cross-entropy kwargs forwarding bug with the clearest implementation context.", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "Unrelated PRs: one adds VideoPrism model support, the other adds an Executorch dynamic-shape test. Same broad framework but not the same change or bug." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Different tokenizer maintenance vs adding a new MobileLLM model. No shared concrete bug or code path." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Tokenizer method additions and Executorch export tests are unrelated changes." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor robustness refactor and new MobileLLM model integration are different PRs in different areas." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Loss utility kwargs fix vs new model addition; no same underlying bug or change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Cross-entropy kwargs fix and Mistral4 test/model work are unrelated." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Executorch dynamic-shape test and Mistral4 fixes are different code paths." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Executorch export test versus MobileLLM model addition; not a duplicate." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New MobileLLM model vs a class-annotation resolution fix; unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "New model integration and Falcon output-interface refactor are different changes." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model PR and a model-creation skill/meta PR are not the same underlying work." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor and model-creation skill are unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon output-capture refactor and Mistral4 test fix are different models and code paths." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45587", + "accept": false, + "reason": "Mistral4 test/model work and docs-only continuous batching memory management are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 fixes and layrnorm-to-layernorm renaming are different changes; same area of Transformers does not make them duplicates." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 test/model work and `cls.config` resolution ordering fix are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Mistral4 test/model work and the model-creation skill are not the same concrete bug/change." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "New MobileLLM model PR and meta skill PR are unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "These are the same underlying `fixed_cross_entropy` bug fix: both add support for passing cross-entropy kwargs (`weight`/`label_smoothing`) to `nn.functional.cross_entropy` for issue `43240`. The diffs are near-identical and could be merged as one PR." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative. The only accepted duplicate pair is `43251`/`43254`, and the shared issue link plus nearly identical `fixed_cross_entropy` diffs support that. The other pairwise rejections are also appropriately cautious and do not overclaim duplication from broad topical similarity alone." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-2", + "nodes": [ + "pull_request:40546", + "pull_request:42781" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781" + ], + "prepared_review_unit_hash": "056a0e683381452301093e34d282e3062a6ead88d93df8c8dc2e8b752e6fd914", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7225, + "estimated_input_tokens": 1807, + "estimated_eval_tokens": 3870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related VibeVoice workstreams, but not duplicates: one PR adds the base VibeVoice model integration, while the other adds a realtime-oriented extension plus acoustic tokenizer and generation changes.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #40546 is the better canonical representative because it is the original, more complete base VibeVoice implementation and is farther along in review, whereas #42781 is a separate draft extension for realtime support.", + "best_issue_reason": null, + "best_pr_reason": "PR #40546 best captures the core VibeVoice integration work in the cluster; #42781 builds on that model family but targets a different feature set and code path.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "They overlap on the VibeVoice family and shared registry/docs files, but the concrete changes differ: #40546 implements the base model, while #42781 adds realtime-specific support with an acoustic tokenizer and related generation updates. That is related work, not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 3, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40755-2", + "nodes": [ + "pull_request:40755", + "pull_request:41224" + ], + "soft_pairs": [ + "pull_request:40755|pull_request:41224" + ], + "prepared_review_unit_hash": "c31f57c4eace8cfaa3614dd6e1d646f2aa5dc0cc01c85d76b69e66b6aa42aeea", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are unrelated PRs: one adds TimesFM covariate forecasting support, the other adds DINOv3 image classification support. They touch different models, code paths, and documentation, so they should not be clustered together as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR across the pair; they implement different features for different model families and are not mergeable into one change.", + "best_issue_reason": null, + "best_pr_reason": "PR #41224 is the cleaner standalone representative because it has a narrower, well-scoped feature addition with focused model/auto/tests changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different underlying changes: TimesFM covariate forecasting vs. DINOv3 image classification support. They affect separate model implementations and cannot 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": 4, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41356-2", + "nodes": [ + "pull_request:41356", + "pull_request:44339" + ], + "soft_pairs": [ + "pull_request:41356|pull_request:44339" + ], + "prepared_review_unit_hash": "aead1308c167b0bae6ff11bf9358340baa5e76f816618383418b505a10a9cd01", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two DEIMv2 pull requests target the same model family, but they are not the same change: PR 41356 is an early/basic model + image processor add, while PR 44339 is a much broader follow-up that adds loss code, auto-mappings, conversion tooling, modular files, and more integration work.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44339 is the more complete and representative DEIMv2 implementation; PR 41356 looks like an earlier partial version of the same model bring-up.", + "best_issue_reason": null, + "best_pr_reason": "PR 44339 is the strongest single artifact for the cluster because it subsumes the basic model work and includes the broader integration needed for a full Transformers addition.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41356", + "right": "pull_request:44339", + "accept": false, + "reason": "Same model area, but not the same concrete change: 44339 expands the implementation substantially and adds new code paths beyond the basic PR 41356. They look like related iterations, not mergeable duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41797-2", + "nodes": [ + "pull_request:41797", + "pull_request:45186" + ], + "soft_pairs": [ + "pull_request:41797|pull_request:45186" + ], + "prepared_review_unit_hash": "b8316d46d769ca2ca3baad4fe9701cf71ca86151a807743105f2249ca450ade0", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6722, + "estimated_input_tokens": 1681, + "estimated_eval_tokens": 3618 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6722, + "estimated_input_tokens": 1681, + "estimated_eval_tokens": 3618 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate model-addition PRs for different models (DeepSeek OCR vs Isaac), not the same underlying bug or change.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the PRs add different model integrations with different files, configs, tests, and code paths.", + "best_issue_reason": null, + "best_pr_reason": "PR #45186 is the better representative only because it is the more complete/newer integration, but it is still not a duplicate of #41797.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41797", + "right": "pull_request:45186", + "accept": false, + "reason": "Different model implementations and integration surfaces; they share generic boilerplate paths but do not fix the same concrete problem or change the same code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42166-2", + "nodes": [ + "pull_request:42166", + "pull_request:45426" + ], + "soft_pairs": [ + "pull_request:42166|pull_request:45426" + ], + "prepared_review_unit_hash": "33bc07fa001bd8a3da2c492ed62e4470777347a05a993c0d468d91f455b88e2a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5408, + "estimated_input_tokens": 1352, + "estimated_eval_tokens": 2960 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5408, + "estimated_input_tokens": 1352, + "estimated_eval_tokens": 2960 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated model-addition PRs that only overlap on shared Transformers registration/auto-mapping files. They do not fix the same bug or implement the same change, so they should not be clustered as duplicates.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true canonical PR here: PR #42166 adds InternVLFlash, while PR #45426 adds AXK1. They are separate model integrations and cannot be merged into one duplicate change.", + "best_issue_reason": null, + "best_pr_reason": "If one PR must represent the cluster, #42166 is the better candidate because it is a fuller, more mature standalone model integration with docs and tests. It is still not a duplicate of #45426.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42166", + "right": "pull_request:45426", + "accept": false, + "reason": "Reject: they add different models (InternVLFlash vs AXK1) and only share generic registration/auto-file edits. This is not the same underlying feature or 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": 7, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42975-2", + "nodes": [ + "pull_request:42975", + "pull_request:42976" + ], + "soft_pairs": [ + "pull_request:42975|pull_request:42976" + ], + "prepared_review_unit_hash": "0324948a0ea0e48eb4334c7015ca2ae64844c60e4ee6e2d5a4c346f30aac5c7d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5934, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are broad GitHub Actions dependency bump PRs, but they target different action families and motivations: one is specifically for Node 24 compatibility, the other upgrades docker/build-push-action and related workflow actions. They overlap in workflow files, but they are not the same concrete change.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the more specific, clearly scoped change: it explicitly targets Node 24 compatibility and updates the core workflow actions that are most likely to break there.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the best representative of the cluster because its title and diff align around a single maintenance goal across multiple workflows.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "Do not merge: these are related workflow dependency upgrades, but they fix different problems and update different actions. One is a Node 24 compatibility pass; the other is a docker/build-push-action version bump plus other latest-action updates. They are not the same underlying bug/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 8, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42978-9", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:43451", + "pull_request:43973", + "pull_request:44601", + "pull_request:45283", + "pull_request:45490", + "pull_request:45626", + "pull_request:45630" + ], + "soft_pairs": [ + "pull_request:42978|pull_request:43448", + "pull_request:42978|pull_request:43451", + "pull_request:42978|pull_request:45490", + "pull_request:43448|pull_request:43451", + "pull_request:43448|pull_request:45490", + "pull_request:43451|pull_request:45490", + "pull_request:43973|pull_request:45283", + "pull_request:44601|pull_request:45626", + "pull_request:45283|pull_request:45626", + "pull_request:45490|pull_request:45626", + "pull_request:45490|pull_request:45630", + "pull_request:45626|pull_request:45630" + ], + "prepared_review_unit_hash": "726f37b33213407ece90f5cf7072b0283ecae6f8b8dbe12835b99103e5847f49", + "component_budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 12, + "serialized_chars": 29352, + "estimated_input_tokens": 7338, + "estimated_eval_tokens": 14932 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 12, + "serialized_chars": 29352, + "estimated_input_tokens": 7338, + "estimated_eval_tokens": 14932 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a mix of unrelated model-addition PRs and two unrelated infrastructure/support changes. They share broad patterns like docs/Auto registration, but they do not describe the same underlying bug or change, so no duplicate merges are justified.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single PR is a valid canonical duplicate target: the PRs cover different models (ViT NEPA, Molmo, Molmo2, LFM2 Audio, CTSM, PP-FormulaNet, Kimi2_6) plus unrelated PP/GGUF support work.", + "best_issue_reason": null, + "best_pr_reason": "No global best PR for deduplication: the cluster is heterogeneous, so any one PR would be a poor representative of the rest.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model families (ViT NEPA vs Molmo); similar scaffolding files are just model onboarding boilerplate." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "ViT NEPA and Molmo2 are distinct additions with different code paths and target models." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "ViT NEPA and CTSM are unrelated model additions; shared docs/auto files are not enough." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are separate model integrations, not the same concrete change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM are different model additions with no shared underlying bug or feature." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 and CTSM are unrelated model support PRs." + }, + { + "left": "pull_request:43973", + "right": "pull_request:45283", + "accept": false, + "reason": "LFM2 Audio model support and Qwen3.5 GGUF loading support address different code paths and problems." + }, + { + "left": "pull_request:44601", + "right": "pull_request:45626", + "accept": false, + "reason": "Native pipeline-parallel loading support is infrastructure work; PP-FormulaNet is a specific model addition." + }, + { + "left": "pull_request:45283", + "right": "pull_request:45626", + "accept": false, + "reason": "GGUF loading support for Qwen3.5 is unrelated to adding PP-FormulaNet model support." + }, + { + "left": "pull_request:45490", + "right": "pull_request:45626", + "accept": false, + "reason": "CTSM model addition and PP-FormulaNet model support are separate models, not duplicate changes." + }, + { + "left": "pull_request:45490", + "right": "pull_request:45630", + "accept": false, + "reason": "CTSM and Kimi2-6 are different model integrations." + }, + { + "left": "pull_request:45626", + "right": "pull_request:45630", + "accept": false, + "reason": "PP-FormulaNet and Kimi2-6 are distinct model additions with no shared concrete bug or 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": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-3", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43665|pull_request:44815" + ], + "prepared_review_unit_hash": "8e12ce4b9fbaf83b0fa661523c693b52dbd743f9c6a2272e3310f512f194fe5b", + "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 pull requests appear unrelated despite generic titles and soft similarity signals: one fixes Whisper decoding/tests, one injects placeholder test failures in CLIP/ViT, and one is a substantive FP8 dequantization fix.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44815 is the strongest canonical candidate because it has the clearest, most substantive bug-fix scope and the most descriptive title; the other PRs are either narrowly Whisper test/code adjustments or low-information test-only changes.", + "best_issue_reason": null, + "best_pr_reason": "PR 44815 is the best representative of the cluster only in terms of being the most concrete and actionable change set. It is still not actually related to the other PRs, so this is a weak global representative choice.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Whisper tokenizer/timestamp decoding changes are unrelated to CLIP/ViT test placeholder assertions; they do not fix the same code-path problem." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decode logic and FP8 dequant/model-loading work address different subsystems and different bugs." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "A test-only placeholder-failure PR is not the same underlying change as a core FP8 dequantization 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": 10, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43488-3", + "nodes": [ + "pull_request:43488", + "pull_request:43532", + "pull_request:45476" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:43532", + "pull_request:43488|pull_request:45476", + "pull_request:43532|pull_request:45476" + ], + "prepared_review_unit_hash": "73001e92f9cb0c988f758824a176d8899c21cacb25674f3bbcf0d3319c06aa52", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7469, + "estimated_input_tokens": 1868, + "estimated_eval_tokens": 3992 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7469, + "estimated_input_tokens": 1868, + "estimated_eval_tokens": 3992 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three unrelated open \"[don't merge]\" PRs that only overlap incidentally on a shared vit test file. One is a setup/version check, one is CI/workflow plumbing, and one is a CI-call workflow plus CLIP/chinese-CLIP test/tokenizer edits. They do not look like the same underlying bug or change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the overlapping filenames are superficial and the PRs target different code paths and purposes.", + "best_issue_reason": null, + "best_pr_reason": "If one must be picked, #43532 is the broadest/most central around the shared test-file overlap, but it still isn't a true duplicate representative because its main changes are CI/workflow-related.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Shared vit test filename is incidental; #43488 is a setup/version/dependency check, while #43532 is workflow and test-infra plumbing." + }, + { + "left": "pull_request:43488", + "right": "pull_request:45476", + "accept": false, + "reason": "These edit different areas: #43488 is repo packaging/versioning, while #45476 adds a CI caller workflow and unrelated CLIP/chinese-CLIP test/tokenizer changes." + }, + { + "left": "pull_request:43532", + "right": "pull_request:45476", + "accept": false, + "reason": "Both mention test files, but the concrete changes are different: #43532 focuses on self-comment CI/slack/testing utils, while #45476 focuses on calling an external CI workflow and CLIP/chinese-CLIP code." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 11, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43612-2", + "nodes": [ + "pull_request:43612", + "pull_request:43613" + ], + "soft_pairs": [ + "pull_request:43612|pull_request:43613" + ], + "prepared_review_unit_hash": "7dc1daf25738868a8fe46504d4f4e8366ac6bb7fb6190ddba39bbbe00de74477", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6082, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate pipeline-addition PRs for different segmentation features: promptable concept segmentation vs promptable visual segmentation. They share some infra and docs plumbing, but they implement different pipelines, model integrations, and tests, so they should not be deduplicated.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43613 is the better standalone representative because it is the broader promptable visual segmentation addition with more model and test coverage; however, it is not a duplicate of 43612.", + "best_issue_reason": null, + "best_pr_reason": "PR 43613 is the stronger overall PR in this cluster: it adds the visual segmentation pipeline with wider integration coverage, while 43612 targets a different concept-segmentation feature.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different concrete features and code paths: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. Shared docs/registration files are just framework plumbing, not evidence of 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": 12, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43651-2", + "nodes": [ + "pull_request:43651", + "pull_request:43663" + ], + "soft_pairs": [ + "pull_request:43651|pull_request:43663" + ], + "prepared_review_unit_hash": "db66e871d06d38205487959f85958aadd9e691d08b82983eb00e10a4fcd4b144", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs in the same file, but they address different trainer extensibility points: one changes gradient-accumulation loss scaling, the other adds an overridable signature-column helper for column filtering.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 43651 is the more self-contained bug-fix style change, with a concrete behavior correction in training loss scaling; PR 43663 is a separate refactor/extensibility change around signature inspection.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Different code paths and user-facing behaviors: 43651 adjusts gradient accumulation loss normalization, while 43663 adds a helper for inspecting model forward signatures to filter columns. They share a file but not the same underlying bug or mergeable fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43656-3", + "nodes": [ + "pull_request:43656", + "pull_request:43836", + "pull_request:43842" + ], + "soft_pairs": [ + "pull_request:43656|pull_request:43842", + "pull_request:43836|pull_request:43842" + ], + "prepared_review_unit_hash": "2093d26d368c88c30ffe68fc98577b7f0283380e453bfb419208877ddde508a8", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5992, + "estimated_input_tokens": 1498, + "estimated_eval_tokens": 3252 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three PRs are trying to fix the same CLI `TypeAdapter` annotation/runtime `NameError` in `src/transformers/cli/serve.py`. PR 43842 is the cleanest targeted fix; the others look like overlapping variants of the same bug rather than distinct changes.", + "confidence": 0.86, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43842 is the most focused and explicit fix: it addresses the `TypeAdapter` NameError directly, scopes the import to `TYPE_CHECKING`, and avoids broader unrelated annotation churn.", + "best_issue_reason": null, + "best_pr_reason": "PR 43842 best represents the cluster because it is the narrowest, clearest change for the exact runtime bug described by the title and diff.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": true, + "reason": "Both modify the same file to fix the same `TypeAdapter` annotation NameError in the CLI serve path; they are alternative implementations of the same bug fix." + }, + { + "left": "pull_request:43836", + "right": "pull_request:43842", + "accept": true, + "reason": "Both target the same `TypeAdapter` NameError in `serve.py` and appear to be overlapping fixes on the same code path, despite 43836 being broader/more refactor-heavy." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: all three PRs touch `src/transformers/cli/serve.py` and their titles/diffs point to the same `TypeAdapter`/annotation runtime-name issue. The soft-edge pairings are conservative enough, since each pair appears to be an overlapping implementation of the same bugfix rather than a distinct change." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 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": 24, + "review_unit_index": 1, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43996", + "pull_request:43997", + "pull_request:44007", + "pull_request:44010", + "pull_request:44013", + "pull_request:44017", + "pull_request:44018", + "pull_request:44019", + "pull_request:44024", + "pull_request:44044", + "pull_request:44056", + "pull_request:44059", + "pull_request:44066", + "pull_request:44068", + "pull_request:44071", + "pull_request:44072", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44085", + "pull_request:44086", + "pull_request:44098", + "pull_request:44101", + "pull_request:44116", + "pull_request:44129", + "pull_request:44154", + "pull_request:44161", + "pull_request:44722" + ], + "soft_pairs": [ + "pull_request:43996|pull_request:44085", + "pull_request:43996|pull_request:44044", + "pull_request:44066|pull_request:44085", + "pull_request:44007|pull_request:44072", + "pull_request:44072|pull_request:44722", + "pull_request:44066|pull_request:44072", + "pull_request:44013|pull_request:44044", + "pull_request:44066|pull_request:44086", + "pull_request:44066|pull_request:44071", + "pull_request:44018|pull_request:44068", + "pull_request:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", + "pull_request:44019|pull_request:44722", + "pull_request:44086|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44018|pull_request:44086", + "pull_request:44018|pull_request:44085", + "pull_request:44068|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44059|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44059|pull_request:44722", + "pull_request:44098|pull_request:44101", + "pull_request:44018|pull_request:44071", + "pull_request:44017|pull_request:44722", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:44071|pull_request:44086", + "pull_request:43997|pull_request:44073", + "pull_request:44017|pull_request:44071", + "pull_request:44059|pull_request:44071", + "pull_request:44073|pull_request:44101", + "pull_request:44059|pull_request:44068", + "pull_request:44018|pull_request:44059", + "pull_request:44056|pull_request:44101", + "pull_request:44073|pull_request:44161", + "pull_request:44101|pull_request:44161", + "pull_request: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:43997|pull_request:44071", + "pull_request:44010|pull_request:44071", + "pull_request:44010|pull_request:44073", + "pull_request:44018|pull_request:44129", + "pull_request:43997|pull_request:44074", + "pull_request:43997|pull_request:44129", + "pull_request:44010|pull_request:44018", + "pull_request:44071|pull_request:44129", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154", + "pull_request:44024|pull_request:44076" + ], + "prepared_review_unit_hash": "f5a10d022afa46230b04913144ca64337d835aede37bc8383a6be6b119d241bb", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 + }, + "budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 72, + "serialized_chars": 76934, + "estimated_input_tokens": 19234, + "estimated_eval_tokens": 38724 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a batch of model-specific output-tracing refactors under the same umbrella theme, but they are not duplicates across architectures. The only clear duplicates are the GPT-J and GPT-Neo follow-up PRs that rewrite the same code paths.", + "confidence": 0.84, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44722 is the strongest representative: it targets the core GPT-J/CodeGen output-tracing refactor, overlaps directly with earlier GPT-J PRs, and looks like the most complete version of that change.", + "best_issue_reason": null, + "best_pr_reason": "PR 44722 best captures the underlying duplicate set for the GPT-J/CodeGen output-tracing rewrite; the rest are mostly separate model ports, not one shared fix.", + "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 and issue target match." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "CVT/FNet vs DeBERTa are unrelated code paths; same tracking umbrella is not enough." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Same GPT-J output-tracing refactor on the same file; 44085 reads like a narrow follow-up of the same change." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "ResNet/RegNet/RT-DETR vs EfficientNet are different model implementations; not the same bug/change." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet and GPT-J are unrelated models; same refactor pattern only." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen vs EfficientNet are different code paths." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 vs DeBERTa are unrelated model families." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen and MGP-STR are different models; not mergeable as one fix." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen vs MPT are separate refactors, not the same concrete change." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Both are GPT-Neo output-tracing refactors on the same model/file; this looks like the same underlying change revised later." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen vs GPT-Neo are different code paths despite similar wording." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo vs GPT-J/CodeGen are separate model-specific refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet is unrelated to GPT-J; they only share the standardized-tracing motif." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J are unrelated implementations." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet vs GPT-J are different model families." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are unrelated changes." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet vs MPT are separate model-specific refactors." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo vs Flaubert are unrelated models." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo vs GPT-J are different code paths." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT vs GPT-J are unrelated implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT2 and MGP-STR are different models; same refactor theme only." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J and MGP-STR are unrelated model code paths." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT2 vs GPT-J are different model families." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT vs Flaubert/XLM are unrelated model changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44722", + "accept": false, + "reason": "SegFormer vs GPT-J are different models; only the tracing pattern is shared." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "SegFormer vs LongT5 are unrelated refactors." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "SegFormer and VisualBERT are separate model-specific changes." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-Neo vs MPT are different code paths." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44086", + "accept": false, + "reason": "MPT vs MGP-STR are unrelated models." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44073", + "accept": false, + "reason": "RegNet and VisualBERT are different model families." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44101", + "accept": false, + "reason": "MPNet vs Flaubert/XLM are unrelated changes." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44018", + "accept": false, + "reason": "SegFormer vs GPT-Neo are different model implementations." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet vs MPNet are separate code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet vs LongT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet vs Flaubert/XLM are different model families." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet vs LongT5 are unrelated changes." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBERT vs MPT are different model paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBERT vs VisualBERT are unrelated refactors." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT vs DeBERTa are different code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa vs TextNet are unrelated model-specific changes." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet vs ImageGPT are different models; same standardized-output pattern only." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT2 vs GPT-Neo are separate refactors." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT2 vs MPT are unrelated code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo vs SpeechT5 are different model families." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo vs MGP-STR are unrelated changes." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo vs GPT-J are different code paths." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet vs LongT5 are unrelated models." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44017", + "accept": false, + "reason": "MPNet vs SegFormer are different code paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44018", + "accept": false, + "reason": "SqueezeBERT vs GPT-Neo are unrelated model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT vs SpeechT5 are different model families." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet vs LongT5 are unrelated changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "CVT/FNet vs SegFormer are different model code paths." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBERT vs DeBERTa are unrelated model refactors." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa vs SpeechT5 are different code paths." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing change set; 44722 is the broader later version of the same refactor." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both touch GPT-J output tracing on the same file; they look like the same fix in different iterations." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated models." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "Different model families; only the output-tracing theme overlaps." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet vs SpeechT5 are unrelated changes." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "ResNet/RegNet/RT-DETR vs EfficientNet are different model-specific refactors." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44072", + "accept": false, + "reason": "MobileNetV2 vs EfficientNet are separate model implementations." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa vs SpeechT5 are unrelated code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa vs VITS are different models." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet vs ImageGPT are unrelated model refactors." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded overall. The summary stays conservative by treating cross-model output-tracing rewrites as theme similarity rather than duplicates, and the accepted soft pairs are the ones with clear shared code paths/issue targets (GPT-J/CodeGen and GPT-Neo iterations). No obvious overstatement of evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 24, + "review_unit_index": 2, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43997", + "pull_request:44010", + "pull_request:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44129" + ], + "soft_pairs": [ + "pull_request:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "09dffba8a9af89e6dab017f2d24cf82f66a75633cc9236c1ce1ebaa2844b6719", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84501, + "estimated_input_tokens": 21126, + "estimated_eval_tokens": 42508 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 5, + "serialized_chars": 13067, + "estimated_input_tokens": 3267, + "estimated_eval_tokens": 6790 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These PRs are all model-specific refactors for standardized output tracing/capture, but they target different architectures and code paths. They look thematically similar, not duplicate changes to the same underlying bug or implementation.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "If one representative PR is needed, 44073 is the clearest example of the pattern: it adds the standardized output-capture plumbing directly in the model class and shows the intended refactor most explicitly.", + "best_issue_reason": null, + "best_pr_reason": "44073 is the best representative because it is a straightforward, explicit migration to the new output-capture mechanism in a single model file, making the cluster\u2019s theme easiest to understand.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "Both are output-tracing refactors, but they affect different models (VisualBert vs TextNet) and different forward/output implementations, so they are not the same change." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert are separate model implementations with distinct output-recording plumbing; this is a shared refactor theme, not a duplicate fix." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "These touch different model families and different tracing internals, so they do not appear to be one mergeable code change." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are unrelated model code paths; both migrate output handling, but they do not fix the same concrete bug or implementation." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert output tracing and SpeechT5 output tracing are separate model-specific refactors; SpeechT5 also has cache/attention behavior that makes it a different 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": 16, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-13", + "nodes": [ + "pull_request:43998", + "pull_request:43999", + "pull_request:44000", + "pull_request:44001", + "pull_request:44002", + "pull_request:44003", + "pull_request:44004", + "pull_request:44025", + "pull_request:44026", + "pull_request:44027", + "pull_request:44028", + "pull_request:44029", + "pull_request:44030" + ], + "soft_pairs": [ + "pull_request:43998|pull_request:44002", + "pull_request:43998|pull_request:44000", + "pull_request:43998|pull_request:44001", + "pull_request:43998|pull_request:44003", + "pull_request:43998|pull_request:44029", + "pull_request:43999|pull_request:44001", + "pull_request:43999|pull_request:44002", + "pull_request:43999|pull_request:44025", + "pull_request:43999|pull_request:44026", + "pull_request:43999|pull_request:44027", + "pull_request:44000|pull_request:44001", + "pull_request:44000|pull_request:44002", + "pull_request:44000|pull_request:44025", + "pull_request:44000|pull_request:44026", + "pull_request:44001|pull_request:44002", + "pull_request:44001|pull_request:44025", + "pull_request:44001|pull_request:44026", + "pull_request:44002|pull_request:44029", + "pull_request:44002|pull_request:44025", + "pull_request:44001|pull_request:44003", + "pull_request:44002|pull_request:44003", + "pull_request:44000|pull_request:44003", + "pull_request:44003|pull_request:44028", + "pull_request:44003|pull_request:44004", + "pull_request:43998|pull_request:44004", + "pull_request:44001|pull_request:44004", + "pull_request:44002|pull_request:44004", + "pull_request:44000|pull_request:44004", + "pull_request:44025|pull_request:44026", + "pull_request:44025|pull_request:44027", + "pull_request:44025|pull_request:44028", + "pull_request:44025|pull_request:44029", + "pull_request:44025|pull_request:44030", + "pull_request:44026|pull_request:44029", + "pull_request:44026|pull_request:44027", + "pull_request:44026|pull_request:44028", + "pull_request:44026|pull_request:44030", + "pull_request:44027|pull_request:44028", + "pull_request:44027|pull_request:44029", + "pull_request:44027|pull_request:44030", + "pull_request:44028|pull_request:44029", + "pull_request:44028|pull_request:44030", + "pull_request:44029|pull_request:44030" + ], + "prepared_review_unit_hash": "9d5612e997f54af5d1cc2cda75ab0e697129ebd8d2620efc5205add3ddf4c7fe", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41319, + "estimated_input_tokens": 10330, + "estimated_eval_tokens": 20916 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 43, + "serialized_chars": 41319, + "estimated_input_tokens": 10330, + "estimated_eval_tokens": 20916 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "A loose cluster of model-specific output-tracing refactors across unrelated Transformers implementations. The PRs share a naming/theme pattern and helper adoption, but they target different code paths and model families, so they are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44003 is the strongest representative: it has the largest diff, spans two closely related Mamba-family files, and adds the most explicit output-capturing plumbing. It is still just a representative, not a duplicate of the others.", + "best_issue_reason": null, + "best_pr_reason": "If one PR must stand in for this cluster, 44003 is the most central concrete example of the shared refactor theme, with broader scope than the single-model PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations (`timm_backbone` vs `upernet`); same refactor style, not the same code-path bug/change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Unrelated architectures (`timm_backbone` vs `vision_text_dual_encoder`); only the output-tracing theme is shared." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different model families (`timm_backbone` vs `univnet`); no common concrete fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Both are refactors, but they touch different Mamba/timm code paths and do not appear to fix the same underlying bug." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different models (`timm_backbone` vs `rwkv`); shared helper usage is not enough for duplication." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "Different domains (`mobilenet_v1` vision classifier vs `univnet` audio model); not the same change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "`mobilenet_v1` and `upernet` are separate model paths; this is a parallel refactor, not a duplicate fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families (`mobilenet_v1` vs `depth_anything`) and separate forward signatures; not mergeable as one PR." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "`mobilenet_v1` and `vision_encoder_decoder` are unrelated code paths despite similar output-tracing cleanup." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "`mobilenet_v1` vs `speech_encoder_decoder`; same helper refactor pattern, different concrete models." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Multimodal encoder/decoder model vs audio generation model; different forward logic and outputs." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "`vision_text_dual_encoder` and `upernet` are separate architectures; no shared bug beyond the refactor theme." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and different affected files; not the same concrete code-path problem." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "`vision_text_dual_encoder` and `depth_anything` are unrelated models; this is only a similar API cleanup." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44026", + "accept": false, + "reason": "Both are wrapper-style models, but they are different wrappers with different signatures and targets." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "`univnet` and `upernet` have unrelated forward paths; same refactor motif only." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Audio generation vs depth estimation; different model code and outputs." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44026", + "accept": false, + "reason": "`univnet` and `vision_encoder_decoder` are separate implementations; not one fix split across PRs." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "`upernet` and `rwkv` are unrelated architectures; same helper adoption does not make them duplicates." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44025", + "accept": false, + "reason": "`upernet` segmentation wrapper vs `depth_anything` depth model; distinct code paths and behavior." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Audio model vs Mamba-family language model; no common concrete bug." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Segmentation wrapper vs Mamba-family model; different outputs and implementation details." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and file sets; only the refactor pattern is shared." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Mamba-family output tracing vs SuperPoint feature extraction; unrelated concrete changes." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "`mamba`/`falcon_mamba` refactor vs `codegen` output tracing; distinct model code paths." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "`timm_backbone` and `codegen` are unrelated subsystems; not the same bug or change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "`univnet` audio model vs `codegen` attention/cache path; separate changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "`upernet` segmentation vs `codegen` transformer internals; no duplicate relationship." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "`vision_text_dual_encoder` and `codegen` have different forward contracts and behavior." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Both are refactors, but `depth_anything` and `vision_encoder_decoder` are different model families and wrappers." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "`depth_anything` vs `speech_encoder_decoder`; separate implementations with different output shapes and targets." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Depth estimation model vs SuperPoint; unrelated code paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "`depth_anything` and `rwkv` are unrelated model families; not mergeable as one fix." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Depth model vs DPR encoder; different forward behavior and outputs." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Encoder-decoder wrappers for different modalities; similar cleanup, not the same change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated models." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures (`vision_encoder_decoder` vs `rwkv`); no shared concrete bug." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Wrapper model vs DPR retriever; distinct code paths and outputs." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Speech encoder-decoder wrapper vs SuperPoint; unrelated implementations." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Speech encoder-decoder vs RWKV; different subsystems and fixes." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Speech encoder-decoder wrapper vs DPR encoder; not the same bug/change." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "SuperPoint and RWKV are unrelated models; the similarity is only in refactor style." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "Feature-point model vs DPR retrieval model; separate code paths." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "RWKV and DPR are distinct model families with different output semantics; not duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 17, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-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": "021776661ded15d64ebd90a932bd2c1753b9bf0a9f1399f63e00a0b5db02f234", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10914, + "estimated_input_tokens": 2729, + "estimated_eval_tokens": 5714 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 10914, + "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 pull requests are separate model-addition implementations (PenguinVL, DeepSeek-OCR-2, and Voxtral TTS). Their overlap is limited to standard Transformers boilerplate such as docs, auto-mapping, and package registration, so they are not duplicates of the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "No single PR is a better duplicate canonical here; each PR targets a different model and code path, so they should remain separate.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different models and distinct implementations; shared files are only generic docs/auto-registration boilerplate." + }, + { + "left": "pull_request:44662", + "right": "pull_request:45075", + "accept": false, + "reason": "PenguinVL and DeepSeek-OCR-2 are unrelated model additions with no shared concrete bug or change." + }, + { + "left": "pull_request:45075", + "right": "pull_request:45401", + "accept": false, + "reason": "Different model integrations (DeepSeek-OCR-2 vs Voxtral TTS); overlap is limited to common Transformers scaffolding." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 18, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44979-2", + "nodes": [ + "pull_request:44979", + "pull_request:45363" + ], + "soft_pairs": [ + "pull_request:44979|pull_request:45363" + ], + "prepared_review_unit_hash": "8fefa35855c3771492ad6b4e9149686ad25b96d32a4760d45850b1a26a332fb4", + "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 around module/kernel fusion, but they are not duplicates: #44979 adds a new generic Module Fusion API and tests, while #45363 wires up n-to-1 kernel fusion through KernelConfig and touches different integration points. They address adjacent layers of the feature, not the same concrete change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45363 is the more complete, user-facing implementation of the fusion feature, spanning KernelConfig, integrations, and modeling hooks; #44979 is the lower-level API foundation.", + "best_issue_reason": null, + "best_pr_reason": "#45363 best represents the cluster because it implements the end-to-end KernelConfig-based fusion path and has the wider surface area of the feature.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Related feature work, but not the same change: #44979 introduces the generic module fusion API, while #45363 implements n-to-1 kernel fusion via KernelConfig in different code paths. They would not be merged as one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45086-2", + "nodes": [ + "pull_request:45086", + "pull_request:45317" + ], + "soft_pairs": [ + "pull_request:45086|pull_request:45317" + ], + "prepared_review_unit_hash": "a9942be8259f16d437fe7acea1cf847428493ee1ccc8e07a5bf4d9abab58095d", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests target the same Mistral regex AttributeError in the same function and make the same code change; one adds tests, so this is a strong duplicate pair.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the better canonical PR because it contains the same core fix plus test coverage and has more discussion/review activity, making it the more complete representative of this change.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the best representative of the underlying change: same fix as 45086, with added tests and stronger validation of the bug fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same concrete bug and same code-path fix in _patch_mistral_regex, same target issue, and same file diff; 45317 is effectively a superset with tests." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 20, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45115-2", + "nodes": [ + "pull_request:45115", + "pull_request:45569" + ], + "soft_pairs": [ + "pull_request:45115|pull_request:45569" + ], + "prepared_review_unit_hash": "91f123a6a533a1fe0a4ae85e2313900cc5c21fa356623589f75889a2d337bcd3", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs around NemotronH share some model files, but they target different levels of the stack: one is a narrower refactor/conversion-mapping update, while the other is a broader model split/renaming and docs update. They are related, but not the same underlying change.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45569 is the better cluster representative because it contains the broader, user-visible model restructuring (NemotronH dispatching to dense/sparse variants, auto-mapping, docs, and tests), whereas #45115 is a narrower inheritance/conversion-mapping refactor.", + "best_issue_reason": null, + "best_pr_reason": "PR #45569 is the strongest single PR in the cluster: it addresses the concrete architecture packaging change end-to-end and is more complete than the narrower #45115 refactor.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "Shared NemotronH internals do not make these duplicates. #45115 focuses on conversion key remapping and GraniteMoeHybrid-style inheritance, while #45569 splits NemotronH into dense/sparse variants, updates auto/docs, and adds tests; they are different changes that 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": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45332-2", + "nodes": [ + "pull_request:45332", + "pull_request:45333" + ], + "soft_pairs": [ + "pull_request:45332|pull_request:45333" + ], + "prepared_review_unit_hash": "be24af849aeeb265ac9a9d804ac70f0135f23bdf05edf7f439f3667678be0ecb", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5398, + "estimated_input_tokens": 1350, + "estimated_eval_tokens": 2956 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5398, + "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 adds heterogeneous per-layer configuration plumbing, the other adds broader heterogeneous model/runtime support. They share config utilities and tests, but they are not the same concrete change and would not be merged as one duplicate PR.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45332 is the more representative change for this cluster because it covers the end-to-end heterogeneous model/runtime integration, while 45333 is narrower config-only plumbing.", + "best_issue_reason": null, + "best_pr_reason": "PR 45332 is the best standalone representative since it exercises the actual heterogeneous code paths beyond configuration serialization.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related feature work, but not the same underlying fix/change: 45333 adds per-layer config support, while 45332 wires heterogeneous behavior into cache/modeling/integration paths. They share files and a theme, not a single mergeable PR scope." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 22, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45394-5", + "nodes": [ + "pull_request:45394", + "pull_request:45487", + "pull_request:45592", + "pull_request:45596", + "pull_request:45615" + ], + "soft_pairs": [ + "pull_request:45394|pull_request:45596", + "pull_request:45487|pull_request:45592", + "pull_request:45487|pull_request:45596", + "pull_request:45487|pull_request:45615", + "pull_request:45592|pull_request:45596", + "pull_request:45592|pull_request:45615", + "pull_request:45596|pull_request:45615" + ], + "prepared_review_unit_hash": "320fdd3021344fe0050204df1c4ce6e9de4fb30fed5894810055e9e7ec9f3c48", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 7, + "serialized_chars": 13269, + "estimated_input_tokens": 3318, + "estimated_eval_tokens": 6892 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 7, + "serialized_chars": 13269, + "estimated_input_tokens": 3318, + "estimated_eval_tokens": 6892 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is a mix of unrelated model-specific fixes and test expectation updates. There isn\u2019t a single underlying bug or change spanning the PRs, so none of the soft edges should be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45487 is the strongest representative because it contains an actual code-path fix for model parallel behavior, not just test expectation updates, and it affects multiple related models.", + "best_issue_reason": null, + "best_pr_reason": "PR 45487 is the best overall PR candidate: it addresses a concrete runtime/model-parallel bug in model code and has broader functional impact than the test-only PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45394", + "right": "pull_request:45596", + "accept": false, + "reason": "Different models and different fixes: XCLIP model-parallel split-module metadata vs BLT XPU test expectations." + }, + { + "left": "pull_request:45487", + "right": "pull_request:45592", + "accept": false, + "reason": "AltCLIP/ChineseCLIP model-parallel/device handling fix is unrelated to a FastVLM test padding-side adjustment." + }, + { + "left": "pull_request:45487", + "right": "pull_request:45596", + "accept": false, + "reason": "One fixes model parallel behavior in model code; the other only updates BLT XPU test expectations." + }, + { + "left": "pull_request:45487", + "right": "pull_request:45615", + "accept": false, + "reason": "Different models and different change types; no shared concrete bug or code path." + }, + { + "left": "pull_request:45592", + "right": "pull_request:45596", + "accept": false, + "reason": "Both are test-related, but they target different models and different failure modes." + }, + { + "left": "pull_request:45592", + "right": "pull_request:45615", + "accept": false, + "reason": "FastVLM padding-side test fix is unrelated to Qianfan OCR XPU expectation updates." + }, + { + "left": "pull_request:45596", + "right": "pull_request:45615", + "accept": false, + "reason": "Both add XPU expectations, but for different model test suites and different outputs, so they are not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 23, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45415-2", + "nodes": [ + "pull_request:45415", + "pull_request:45425" + ], + "soft_pairs": [ + "pull_request:45415|pull_request:45425" + ], + "prepared_review_unit_hash": "2c66e933be0d2eb5f890a40250a622430bab7d383479d16ceb2104eeea872895", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are typing-related cleanups in transformers, but they target different scopes and fixes: one is a broad type-checking sweep across many files, the other adds specific typing helpers and targeted annotations in a smaller set of modules.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45425 is the better canonical representative because it is a more focused, concrete typing infrastructure change, while #45415 is a much broader repo-wide type-checking sweep.", + "best_issue_reason": null, + "best_pr_reason": "PR #45425 is the strongest standalone candidate in this cluster due to its narrower, more coherent scope and clearer code-level purpose.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Both are typing-related, but they are not the same underlying change: #45415 is a broad type-checking pass across many files plus config changes, while #45425 adds specific typing protocols/helpers and targeted fixes. They would not plausibly be merged as one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 24, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45639-2", + "nodes": [ + "pull_request:45639", + "pull_request:45645" + ], + "soft_pairs": [ + "pull_request:45639|pull_request:45645" + ], + "prepared_review_unit_hash": "583ba3da5b7a107c6b9ff217eb0a4ec45c274a74359cec1d346a1b461d15c4c2", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same captured_info xdist collision problem; #45645 is the broader, more complete fix, while #45639 is a narrower subset focused on the testing_utils path.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45645 is the better canonical PR because it includes the core xdist-safe file handling change plus the CI workflow update needed to consume multiple captured_info artifacts, making it the more complete fix for the same bug.", + "best_issue_reason": null, + "best_pr_reason": "PR #45645 best represents the cluster since it covers the underlying captured_info artifact collision end-to-end, not just the library-side write path.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Accept: both PRs fix the same concrete xdist collision around captured_info output; #45639 is a narrower implementation of the same bug that #45645 also addresses, and they could plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/manifest.json b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..057442f07fe48a88cb67c0b5a1291299deb3b018 --- /dev/null +++ b/snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260427t120026z", + "artifacts": { + "hybrid": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-04-27T12:09:03Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260427T120026Z", + "variant": "hybrid" +} diff --git a/snapshots/20260427T120026Z/manifest.json b/snapshots/20260427T120026Z/manifest.json index 8119342590201d07f23f02bf306eb5a05b82d597..d9c2a2435f60a9be752a7ce4673da63be31a5d1e 100644 --- a/snapshots/20260427T120026Z/manifest.json +++ b/snapshots/20260427T120026Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 1228 }, "extracted_at": "2026-04-27T12:00:26Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260427t120026z", + "runs": { + "hybrid-model-20260427t120026z": { + "analysis_id": "hybrid-model-20260427t120026z", + "artifacts": { + "hybrid": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260427T120026Z/analysis-runs/hybrid-model-20260427t120026z/manifest.json", + "model": null, + "published_at": "2026-04-27T12:09:03Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260427T120026Z", "watermark": {