state: generation 13223 (cycle)

.modelhub_state/architecture_compatibility_blacklist.json

@@ -3023,7 +3023,7 @@
       "taskType": "text-generation"
     }
   },
-  "generatedAt": "2026-09-23T00:49:38.161191+00:00",
+  "generatedAt": "2026-09-23T00:53:50.940821+00:00",
   "summary": {
     "activeBlockCount": 152,
     "byGpuFramework": {

.modelhub_state/market_intelligence.json

@@ -396,7 +396,7 @@
     }
   },
   "frameworkUpdatedAt": null,
-  "generatedAt": "2026-09-23T00:52:49.207477+00:00",
+  "generatedAt": "2026-09-23T00:54:00.373153+00:00",
   "gpuStats": {
     "Ascend_910-b4": {
       "available": true,

.modelhub_state/official_capabilities.json

@@ -1,5 +1,5 @@
 {
-  "catalogUpdatedAt": "2026-09-23T00:52:49.207477+00:00",
+  "catalogUpdatedAt": "2026-09-23T00:54:00.373153+00:00",
   "configuredTaskTypes": [
     "text-generation"
   ],
@@ -56,7 +56,7 @@
     "time-series-forecasting"
   ],
   "errors": [],
-  "generatedAt": "2026-09-23T00:52:49.207477+00:00",
+  "generatedAt": "2026-09-23T00:54:00.373153+00:00",
   "gpuCatalog": {
     "Ascend_910-b3": {
       "canVerify": true,
@@ -6669,6 +6669,6 @@
       "updateTime": "2025-12-22 08:59:53"
     }
   ],
-  "taskTreeUpdatedAt": "2026-09-23T00:52:49.207477+00:00",
+  "taskTreeUpdatedAt": "2026-09-23T00:54:00.373153+00:00",
   "version": 1
 }

.modelhub_state/outcome_checkpoint.json

@@ -1,6 +1,6 @@
 {
-  "generatedAt": "2026-09-23T00:07:02.522062+00:00",
+  "generatedAt": "2026-09-23T00:53:50.855166+00:00",
-  "lastSyncTime": "2026-09-23T00:07:02.456935+00:00",
+  "lastSyncTime": "2026-09-23T00:53:50.560096+00:00",
   "recentLimit": 300,
   "report": {
     "architectureCompatibilityBlocks": {
@@ -5544,7 +5544,7 @@
         "decisionSuccessRate": 0.0075,
         "decisionTotal": 668,
         "failureBreakdown": {
-          "ambiguous_runtime": 160,
+          "ambiguous_runtime": 161,
           "architecture_compatibility": 30,
           "attention_backend": 1,
           "context_length": 42,
@@ -5557,7 +5557,7 @@
           "tokenizer_compatibility": 79,
           "参数/模板问题": 3
         },
-        "failureCount": 829,
+        "failureCount": 830,
         "failureRate": 0.994,
         "framework": "vllm",
         "pendingCount": 0,
@@ -5567,8 +5567,8 @@
         "successRate": 0.006,
         "targetGpu": "hygon_k100-ai",
         "taskType": "text-generation",
-        "total": 834,
+        "total": 835,
-        "unresolvedFailureCount": 163
+        "unresolvedFailureCount": 164
       },
       "hygon_k100-ai|vllm|unknown": {
         "attributableFailureCount": 1,
@@ -5696,7 +5696,7 @@
         "decisionSuccessRate": 0.0259,
         "decisionTotal": 3666,
         "failureBreakdown": {
-          "ambiguous_runtime": 1533,
+          "ambiguous_runtime": 1534,
           "architecture_compatibility": 112,
           "attention_backend": 3,
           "backend_operator": 86,
@@ -5710,15 +5710,15 @@
           "tokenizer_compatibility": 413,
           "参数/模板问题": 48
         },
-        "failureCount": 6018,
+        "failureCount": 6019,
         "failureRate": 0.9845,
         "pendingCount": 0,
         "pendingRate": 0.0,
         "platformFailureCount": 866,
         "successCount": 95,
         "successRate": 0.0155,
-        "total": 6113,
+        "total": 6114,
-        "unresolvedFailureCount": 1581
+        "unresolvedFailureCount": 1582
       },
       "vllm-customized": {
         "attributableFailureCount": 12,
@@ -5867,7 +5867,7 @@
         "unresolvedFailureCount": 173
       }
     },
-    "generatedAt": "2026-09-23T00:07:02.509722+00:00",
+    "generatedAt": "2026-09-23T00:53:50.842972+00:00",
     "gpuSummaries": {
       "Ascend_910-b3": {
         "attributableFailureCount": 102,
@@ -6263,7 +6263,7 @@
         "decisionSuccessRate": 0.0464,
         "decisionTotal": 755,
         "failureBreakdown": {
-          "ambiguous_runtime": 319,
+          "ambiguous_runtime": 320,
           "architecture_compatibility": 32,
           "attention_backend": 1,
           "context_length": 46,
@@ -6278,15 +6278,15 @@
           "日志缺失": 66,
           "验证失败": 27
         },
-        "failureCount": 1527,
+        "failureCount": 1528,
         "failureRate": 0.9776,
         "pendingCount": 0,
         "pendingRate": 0.0,
         "platformFailureCount": 4,
         "successCount": 35,
         "successRate": 0.0224,
-        "total": 1562,
+        "total": 1563,
-        "unresolvedFailureCount": 803
+        "unresolvedFailureCount": 804
       }
     },
     "observedGpuMemoryGiB": {
@@ -20228,6 +20228,29 @@
         "total": 1,
         "unresolvedFailureCount": 0
       },
+      "hygon_k100-ai|vllm|text-generation|gemma3|compressed-tensors": {
+        "attributableFailureCount": 0,
+        "decisionFailureRate": 0.0,
+        "decisionSuccessRate": 0.0,
+        "decisionTotal": 0,
+        "failureBreakdown": {
+          "ambiguous_runtime": 1
+        },
+        "failureCount": 1,
+        "failureRate": 1.0,
+        "framework": "vllm",
+        "modelType": "gemma3",
+        "pendingCount": 0,
+        "pendingRate": 0.0,
+        "platformFailureCount": 0,
+        "quantizationMethod": "compressed-tensors",
+        "successCount": 0,
+        "successRate": 0.0,
+        "targetGpu": "hygon_k100-ai",
+        "taskType": "text-generation",
+        "total": 1,
+        "unresolvedFailureCount": 1
+      },
       "hygon_k100-ai|vllm|text-generation|llama|awq": {
         "attributableFailureCount": 1,
         "decisionFailureRate": 0.2,
@@ -45161,6 +45184,30 @@
         "total": 1,
         "unresolvedFailureCount": 0
       },
+      "hygon_k100-ai|vllm|text-generation|gemma3|compressed-tensors|34": {
+        "attributableFailureCount": 0,
+        "decisionFailureRate": 0.0,
+        "decisionSuccessRate": 0.0,
+        "decisionTotal": 0,
+        "failureBreakdown": {
+          "ambiguous_runtime": 1
+        },
+        "failureCount": 1,
+        "failureRate": 1.0,
+        "framework": "vllm",
+        "loadSizeLog2Bucket": 34,
+        "modelType": "gemma3",
+        "pendingCount": 0,
+        "pendingRate": 0.0,
+        "platformFailureCount": 0,
+        "quantizationMethod": "compressed-tensors",
+        "successCount": 0,
+        "successRate": 0.0,
+        "targetGpu": "hygon_k100-ai",
+        "taskType": "text-generation",
+        "total": 1,
+        "unresolvedFailureCount": 1
+      },
       "hygon_k100-ai|vllm|text-generation|llama|awq|30": {
         "attributableFailureCount": 1,
         "decisionFailureRate": 1.0,
@@ -45494,15 +45541,15 @@
         "unresolvedFailureCount": 0
       }
     },
-    "terminalRecords": 16781,
+    "terminalRecords": 16782,
-    "totalRecords": 16989,
+    "totalRecords": 16990,
     "totals": {
       "attributableFailureCount": 6006,
       "decisionFailureRate": 0.8621,
       "decisionSuccessRate": 0.1379,
       "decisionTotal": 6967,
       "failureBreakdown": {
-        "ambiguous_runtime": 4277,
+        "ambiguous_runtime": 4278,
         "architecture_compatibility": 212,
         "attention_backend": 3,
         "backend_operator": 113,
@@ -45518,15 +45565,15 @@
         "日志缺失": 719,
         "验证失败": 676
       },
-      "failureCount": 15820,
+      "failureCount": 15821,
       "failureRate": 0.9427,
       "pendingCount": 0,
       "pendingRate": 0.0,
       "platformFailureCount": 929,
       "successCount": 961,
       "successRate": 0.0573,
-      "total": 16781,
+      "total": 16782,
-      "unresolvedFailureCount": 8885
+      "unresolvedFailureCount": 8886
     },
     "warnings": [
       "GPU MetaX_c-500 本地统计失败率偏高（≥50%），建议重点关注。",
@@ -45583,6 +45630,6 @@
     ]
   },
   "storageMode": "decision_state_only",
-  "summarizedRecords": 16989,
+  "summarizedRecords": 16990,
   "version": 1
 }

manifest.json

@@ -2,24 +2,24 @@
   "agentVersion": "2026.09.22.1",
   "checksums": {
     ".modelhub_state/account_capacity.json": "930f9869793c3d1aa071c53f05b7cf82a4e372f002be88361d6d6189e27c12a1",
-    ".modelhub_state/architecture_compatibility_blacklist.json": "4bf2fad3913eb153e46cd7f4ac8a6cdb74b669512394f6fad485ee310c53b8e0",
+    ".modelhub_state/architecture_compatibility_blacklist.json": "d781fa879b71cbd1b47ebffd8beb3ac9df9d0ead3994f545dffad7194c5f3de7",
     ".modelhub_state/architecture_history_backfill.json": "a92348206605da04f75c9e26e7c818b76f259e0b28983f18b6375ef94802f0ab",
-    ".modelhub_state/market_intelligence.json": "8099d1abb7885af3d93230d8899993314bcc140f474197c9ce53e79ebc511e79",
+    ".modelhub_state/market_intelligence.json": "b47852215bf675a32af45898955dbebfb1b6222ba490b33767ec14014833965c",
-    ".modelhub_state/official_capabilities.json": "ab9f603f0c553fead7631738207fa8197523d90fd988f80c69783d7547592c5c",
+    ".modelhub_state/official_capabilities.json": "32b8dc142e08f3f9a555c1fad59d5fe3866c801599c767663a39c9f2802b5126",
-    ".modelhub_state/outcome_checkpoint.json": "b631b0012293da2ddc9b21e5a692988241af2d3cac82d77a67b89f25decdaa51",
+    ".modelhub_state/outcome_checkpoint.json": "96788325cb4d6f19eda0b672864afdc436a0fab4a53d6788a718bd1fda27610b",
     ".modelhub_state/queue_cleanup_latest.json": "70b12a33050de2b7d18becf5530413055730ba9da8e6ce7b09a99877bc0fd590",
     ".modelhub_state/recent_outcomes.jsonl": "79b6d55284c3f75cc08ad11b8ab20e4edff0f39ec4effad3abbee12ed1b7e175",
-    ".modelhub_state/recovery_active_tasks.jsonl": "fe1ba823eaf0171cc437c3020eec0cf5a60f6587b7154f356c7e279a2d5f25b3",
+    ".modelhub_state/recovery_active_tasks.jsonl": "7741fbdd5f4688490881e83013f5adf79170b440861e6a265444ac3ee2193817",
-    ".modelhub_state/recovery_intents.jsonl": "3518996d70f5f3fe6fe4069f7ae2731b215439c6d353e709276bffb9b0dcc7eb",
+    ".modelhub_state/recovery_intents.jsonl": "16fa19ba7c0bd38b70bd188739487f7365f2043dbbbfaf8ca429d6ce73bdf143",
     ".modelhub_state/routing_intelligence.json": "39ec7b5e77e11b96887ed828d68b094ab47d4d4e68b51e1ea92f4df7a81ab3e3",
     ".modelhub_state/submission_exclusions.jsonl": "80315f370e9cc3cdadaf390e12573ef887794214779379c50b7b37b7631e8989",
     ".modelhub_state/worker_crashes.jsonl": "81a279620f26a6ff30e7cd3dca4b86844ec41a0d30ffee7d8a4b03ab754d2d4c",
     "ledger/submissions.jsonl": "e41ca1b41c2d9fe7c553e3c10665298fcb8b0ac7fb7114270dbe689bfb7c6207",
-    "outcomes/submissions.jsonl": "08d02c149e826ebccbeed3bf9ab749b9748087ec5134fcba0281e2d9376715db"
+    "outcomes/submissions.jsonl": "7b35b3b5a0a9c6d77f01c47c906f4b47280f65ced73b4d4e76cca5bbf3b39909"
   },
-  "generation": 13222,
+  "generation": 13223,
   "phase": "cycle",
   "schemaVersion": 1,
-  "updatedAt": "2026-09-23T00:52:49.839688+00:00",
+  "updatedAt": "2026-09-23T00:54:02.001791+00:00",
   "writerId": "0024d6b5efdf486184421da342f897ee"
 }

outcomes/submissions.jsonl

@@ -389,7 +389,6 @@
 {"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.599735+00:00", "modelId": "LiquidAI/LFM2.5-1.2B-Instruct-MLX-bf16", "modelProfile": {"architectures": ["Lfm2ForCausalLM"], "configFingerprint": "ef4c08c913d3f64cae22550f0c4a189795901d4b11e26d8e640e07e857659fb9", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 2340697867, "estimatedRequiredGiB": 2.622, "gpuMemoryEvidence": {"memoryGiB": 24.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 24.0, "maximumRepositorySizeGiB": 20.0, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "lfm2", "modelscopeFileSize": 2345693224, "modelscopeLicense": "other", "modelscopeParams": 1170340608, "modelscopeTags": ["license:other", "model_type:lfm2", "library:mlx", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:liquid", "custom_tag:lfm2.5", "custom_tag:edge", "custom_tag:mlx"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 2345693224}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:25:32.744196+00:00", "targetGpu": "Cambricon_mlu-370-x4", "taskId": "4969674", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.599915+00:00", "modelId": "inclusionAI/Ling-3.0-tiny", "modelProfile": {"architectures": ["BailingMoeV3ForCausalLM"], "configFingerprint": "8b3c0a82b858cf26b1a0f6075074234a66487d5734741dbbfd4743963b967faa", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 15787992416, "estimatedRequiredGiB": 17.659, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "bailing_hybrid", "modelscopeFileSize": 15801179822, "modelscopeLicense": "mit", "modelscopeParams": 7893392800, "modelscopeTags": ["license:mit", "model_type:bailing_hybrid", "library:safetensors", "library:pytorch", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 15801179822}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:25:32.939987+00:00", "targetGpu": "hygon_k100-ai", "taskId": "4969687", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.600157+00:00", "modelId": "solidrust/Llama-3-8B-Instruct-v0.4-AWQ", "modelProfile": {"architectures": ["LlamaForCausalLM"], "configFingerprint": "8b3c0a82b858cf26b1a0f6075074234a66487d5734741dbbfd4743963b967faa", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 5727938576, "estimatedRequiredGiB": 6.412, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "llama", "modelscopeFileSize": 5737300809, "modelscopeLicense": "other", "modelscopeParams": 8030261248, "modelscopeTags": ["license:other", "model_type:llama", "library:transformer", "library:safetensors", "library:pytorch", "task:text-generation", "custom_tag:4-bit", "custom_tag:AWQ", "custom_tag:text-generation", "custom_tag:autotrain_compatible", "custom_tag:endpoints_compatible", "custom_tag:axolotl", "custom_tag:finetune", "custom_tag:facebook", "custom_tag:meta", "custom_tag:pytorch", "custom_tag:llama", "custom_tag:llama-3"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "awq", "repositoryOnDiskBytes": 5737300809}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:25:32.875443+00:00", "targetGpu": "hygon_k100-ai", "taskId": "4969681", "taskType": "text-generation", "verifyResult": null}
-{"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.600203+00:00", "modelId": "aisingapore/Gemma-SEA-LION-v4-27B-IT-FP8-Dynamic", "modelProfile": {"architectures": ["Gemma3ForConditionalGeneration"], "configFingerprint": "8b3c0a82b858cf26b1a0f6075074234a66487d5734741dbbfd4743963b967faa", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 29274355224, "estimatedRequiredGiB": 32.761, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "gemma3", "modelscopeFileSize": 29314389092, "modelscopeLicense": "gemma", "modelscopeParams": 27432406640, "modelscopeTags": ["license:gemma", "model_type:gemma3", "library:safetensors", "library:pytorch", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "compressed-tensors", "repositoryOnDiskBytes": 29314389092}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:25:32.993901+00:00", "targetGpu": "hygon_k100-ai", "taskId": "4969689", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.599837+00:00", "modelId": "aisingapore/Gemma-SEA-LION-v3-9B", "modelProfile": {"architectures": ["Gemma2ForCausalLM"], "configFingerprint": "8b3c0a82b858cf26b1a0f6075074234a66487d5734741dbbfd4743963b967faa", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 18483466448, "estimatedRequiredGiB": 20.702, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "gemma2", "modelscopeFileSize": 18523972484, "modelscopeLicense": "gemma", "modelscopeParams": 9241705984, "modelscopeTags": ["license:gemma", "model_type:gemma2", "library:safetensors", "library:pytorch", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 18523972484}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:25:33.037269+00:00", "targetGpu": "hygon_k100-ai", "taskId": "4969692", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm", "lastSyncTime": "2026-09-20T11:30:54.599773+00:00", "modelId": "aisingapore/Gemma-SEA-LION-v4-27B-IT", "modelProfile": {"architectures": ["Gemma3ForConditionalGeneration"], "configFingerprint": "533968aac3b9c90272dfbe2866791f782e077f0dceee5b6ffffce48ba6674a10", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 54864980000, "estimatedRequiredGiB": 61.361, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "gemma3", "modelscopeFileSize": 54904930780, "modelscopeLicense": "gemma", "modelscopeParams": 27432406640, "modelscopeTags": ["license:gemma", "model_type:gemma3", "library:pytorch", "library:transformer", "library:safetensors", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 54904930780}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:30:23.763321+00:00", "targetGpu": "MetaX_c-500", "taskId": "4969808", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm-mlu", "lastSyncTime": "2026-09-20T12:01:00.671231+00:00", "modelId": "neuralmagic/Meta-Llama-3-8B-Instruct-quantized.w8a8", "modelProfile": {"architectures": ["LlamaForCausalLM"], "configFingerprint": "36ec94e289a8da5fd70b98a2f401fc5159740daf9e2a076da4fd490f34bcbc0b", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 9084013360, "estimatedRequiredGiB": 10.163, "gpuMemoryEvidence": {"memoryGiB": 24.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 24.0, "maximumRepositorySizeGiB": 20.0, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "llama", "modelscopeFileSize": 9093325202, "modelscopeLicense": "llama3", "modelscopeParams": 8030261248, "modelscopeTags": ["license:llama3", "model_type:llama", "library:transformer", "library:safetensors", "library:pytorch", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "compressed-tensors", "repositoryOnDiskBytes": 9093325202}, "outcome": "pending", "status": "waiting", "submitTime": "2026-09-20T03:40:37.351352+00:00", "targetGpu": "Cambricon_mlu-370-x4", "taskId": "4969964", "taskType": "text-generation", "verifyResult": null}