diff --git a/.modelhub_state/account_capacity.json b/.modelhub_state/account_capacity.json
index 7cb3a4ed53..7dff30668d 100644
--- a/.modelhub_state/account_capacity.json
+++ b/.modelhub_state/account_capacity.json
@@ -42,13 +42,13 @@
     },
     "efca47eed826bada0502": {
       "accountIndex": 1,
-      "knownCap": 29
+      "knownCap": 31
     },
     "f0b5423ecb0841256040": {
       "accountIndex": 8,
       "knownCap": 36
     }
   },
-  "updatedAt": "2026-10-22T10:21:54.549041+00:00",
+  "updatedAt": "2026-10-22T10:30:26.332233+00:00",
   "version": 1
 }
diff --git a/.modelhub_state/recovery_intents.jsonl b/.modelhub_state/recovery_intents.jsonl
index 11eca8d131..5f6c925ae8 100644
--- a/.modelhub_state/recovery_intents.jsonl
+++ b/.modelhub_state/recovery_intents.jsonl
@@ -3719,18 +3719,18 @@
 {"batchId": "cd328024a836450bad5bca0f9deb0dc5", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T08:32:09.107626+00:00", "framework": "vllm_tokenizer_patch", "intentId": "a6351cfe68ae4069bba877c8e4ed68f4", "lastModified": "2026-10-21T17:00:21+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/LFM2-8B-A1B-AWQ-4bit", "repoId": "cyankiwi/LFM2-8B-A1B-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
 {"batchId": "fcb4c513ee164e18881abb7e0a95b872", "completedAt": "2026-10-22T08:52:04.649002+00:00", "configFingerprint": "de86b185053df1a87f4f6a1e0579898f12e492a36aefb07c55e30e7bd8393778", "configSource": "modelhub_live", "createdAt": "2026-10-22T08:51:49.378548+00:00", "framework": "llamacpp", "intentId": "723f3e088afe4f28ab3f0afa2a9ce170", "lastModified": "2026-10-21T03:15:59+00:00", "modelAddress": "https://modelscope.cn/models/XHToken/Spark-X2.6-4B-GGUF", "reason": null, "reconciledAt": "2026-10-22T10:28:48.879581+00:00", "repoId": "XHToken/Spark-X2.6-4B-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "recovered_active", "targetGpu": "Ascend_910-b4", "taskId": "5714121", "taskType": "text-generation"}
 {"batchId": "dc5a0a21871e47ea8525a7dbf0a587d1", "completedAt": "2026-10-22T09:51:23.797519+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:51:20.782801+00:00", "framework": "vllm_tokenizer_patch", "intentId": "e8bbd59a29384e9e8f51bef2c46629cc", "lastModified": "2026-10-21T16:31:43+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-8bit", "reason": null, "reconciledAt": "2026-10-22T10:28:48.877649+00:00", "repoId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "recovered_active", "targetGpu": "Ascend_910-b4", "taskId": "5715097", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "705b8b3614c86a1ac722c97f43e177e84ca20282e9799fffb1b890510d641728", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362367+00:00", "framework": "llamacpp", "intentId": "a5a983283abe48258b5c47785063dbb9", "lastModified": "2026-10-21T18:16:46+00:00", "modelAddress": "https://modelscope.cn/models/AtomicChat/Qwen3.5-9B-DFlash-GGUF", "repoId": "AtomicChat/Qwen3.5-9B-DFlash-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362485+00:00", "framework": "vllm_tokenizer_patch", "intentId": "fe0e6c3d72334482b515c892addd9563", "lastModified": "2026-10-21T17:03:19+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/MiniCPM-SALA-AWQ-4bit", "repoId": "cyankiwi/MiniCPM-SALA-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362542+00:00", "framework": "vllm_tokenizer_patch", "intentId": "608393511b9649e1a1a51be4f97c89c4", "lastModified": "2026-10-22T01:37:07+00:00", "modelAddress": "https://modelscope.cn/models/RecursiveSI/flint-int4", "repoId": "RecursiveSI/flint-int4", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362594+00:00", "framework": "vllm_tokenizer_patch", "intentId": "42c59f3460334991bb61ea53fd08030a", "lastModified": "2026-10-21T16:21:36+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/AgentCPM-Report-AWQ-8bit", "repoId": "cyankiwi/AgentCPM-Report-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362644+00:00", "framework": "vllm_tokenizer_patch", "intentId": "ce264e5166be4073b8b2fb7b2578a2cb", "lastModified": "2026-10-21T16:19:53+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/DASD-30B-A3B-Thinking-Preview-AWQ-4bit", "repoId": "cyankiwi/DASD-30B-A3B-Thinking-Preview-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362694+00:00", "framework": "vllm_tokenizer_patch", "intentId": "30eeb610bca04f0083bac59819313ede", "lastModified": "2026-10-21T16:56:31+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/NVIDIA-Nemotron-Nano-9B-v2-AWQ-4bit", "repoId": "cyankiwi/NVIDIA-Nemotron-Nano-9B-v2-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362743+00:00", "framework": "vllm_tokenizer_patch", "intentId": "220c2f4474844268956c20275198a187", "lastModified": "2026-10-21T16:38:25+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "repoId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362792+00:00", "framework": "vllm_tokenizer_patch", "intentId": "20df7779003e4192bec11807d5d53493", "lastModified": "2026-10-21T16:49:53+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/Olmo-3.1-32B-Think-AWQ-4bit", "repoId": "cyankiwi/Olmo-3.1-32B-Think-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362842+00:00", "framework": "vllm_tokenizer_patch", "intentId": "5413904011b34ee982e10cfd88c8ec32", "lastModified": "2026-10-21T17:00:21+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/LFM2-8B-A1B-AWQ-4bit", "repoId": "cyankiwi/LFM2-8B-A1B-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362891+00:00", "framework": "vllm_tokenizer_patch", "intentId": "310606ad1a084956ade2802cb520bdc5", "lastModified": "2026-10-21T16:58:14+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/KAT-Dev-AWQ-4bit", "repoId": "cyankiwi/KAT-Dev-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362940+00:00", "framework": "vllm_tokenizer_patch", "intentId": "ed257fd2110845158e2f6837fb956221", "lastModified": "2026-10-21T17:09:55+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/Nemotron-Cascade-14B-Thinking-AWQ-8bit", "repoId": "cyankiwi/Nemotron-Cascade-14B-Thinking-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
-{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362989+00:00", "framework": "vllm_tokenizer_patch", "intentId": "df2113df44fc45978914a4dd6816961c", "lastModified": "2026-10-21T16:41:10+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/Nemotron-Cascade-8B-Thinking-AWQ-4bit", "repoId": "cyankiwi/Nemotron-Cascade-8B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "pending", "targetGpu": "Ascend_910-b4", "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650766+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362743+00:00", "framework": "vllm_tokenizer_patch", "intentId": "220c2f4474844268956c20275198a187", "lastModified": "2026-10-21T16:38:25+00:00", "modelAddress": "https://modelscope.cn/models/cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "reason": null, "repoId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "submitted", "targetGpu": "Ascend_910-b4", "taskId": "5715909", "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650780+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362989+00:00", "framework": "vllm_tokenizer_patch", "intentId": "df2113df44fc45978914a4dd6816961c", "repoId": "cyankiwi/Nemotron-Cascade-8B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650777+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362940+00:00", "framework": "vllm_tokenizer_patch", "intentId": "ed257fd2110845158e2f6837fb956221", "repoId": "cyankiwi/Nemotron-Cascade-14B-Thinking-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650774+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362891+00:00", "framework": "vllm_tokenizer_patch", "intentId": "310606ad1a084956ade2802cb520bdc5", "repoId": "cyankiwi/KAT-Dev-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650772+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362842+00:00", "framework": "vllm_tokenizer_patch", "intentId": "5413904011b34ee982e10cfd88c8ec32", "repoId": "cyankiwi/LFM2-8B-A1B-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650769+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362792+00:00", "framework": "vllm_tokenizer_patch", "intentId": "20df7779003e4192bec11807d5d53493", "repoId": "cyankiwi/Olmo-3.1-32B-Think-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650763+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362694+00:00", "framework": "vllm_tokenizer_patch", "intentId": "30eeb610bca04f0083bac59819313ede", "repoId": "cyankiwi/NVIDIA-Nemotron-Nano-9B-v2-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650760+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362644+00:00", "framework": "vllm_tokenizer_patch", "intentId": "ce264e5166be4073b8b2fb7b2578a2cb", "repoId": "cyankiwi/DASD-30B-A3B-Thinking-Preview-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650757+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362594+00:00", "framework": "vllm_tokenizer_patch", "intentId": "42c59f3460334991bb61ea53fd08030a", "repoId": "cyankiwi/AgentCPM-Report-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650754+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362542+00:00", "framework": "vllm_tokenizer_patch", "intentId": "608393511b9649e1a1a51be4f97c89c4", "repoId": "RecursiveSI/flint-int4", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650750+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362485+00:00", "framework": "vllm_tokenizer_patch", "intentId": "fe0e6c3d72334482b515c892addd9563", "repoId": "cyankiwi/MiniCPM-SALA-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
+{"batchId": "22f829e7938f436b9dc44a19d4ca2c1c", "completedAt": "2026-10-22T10:30:27.650731+00:00", "configFingerprint": "705b8b3614c86a1ac722c97f43e177e84ca20282e9799fffb1b890510d641728", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:30:25.362367+00:00", "framework": "llamacpp", "intentId": "a5a983283abe48258b5c47785063dbb9", "repoId": "AtomicChat/Qwen3.5-9B-DFlash-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
 {"batchId": "45ea95c21a0b4d1aa4821b5572df10e4", "completedAt": "2026-10-22T10:28:25.680038+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:28:22.993917+00:00", "framework": "vllm_tokenizer_patch", "intentId": "d8a7639ee2ef4169ae54db90657b114e", "repoId": "cyankiwi/Nemotron-Cascade-8B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
 {"batchId": "45ea95c21a0b4d1aa4821b5572df10e4", "completedAt": "2026-10-22T10:28:25.680036+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:28:22.993867+00:00", "framework": "vllm_tokenizer_patch", "intentId": "46cd8f444d4e4787b58fe01970bcc303", "repoId": "cyankiwi/Nemotron-Cascade-14B-Thinking-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
 {"batchId": "45ea95c21a0b4d1aa4821b5572df10e4", "completedAt": "2026-10-22T10:28:25.680033+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T10:28:22.993814+00:00", "framework": "vllm_tokenizer_patch", "intentId": "e916632f941a4e4b94e261dde550a318", "repoId": "cyankiwi/KAT-Dev-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
@@ -4020,14 +4020,3 @@
 {"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392500+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955596+00:00", "framework": "vllm_tokenizer_patch", "intentId": "6d2c460bcd744ff7a243bfd9ed10d2bc", "repoId": "cyankiwi/NVIDIA-Nemotron-Nano-9B-v2-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
 {"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392497+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955545+00:00", "framework": "vllm_tokenizer_patch", "intentId": "d3b8e9c5dcd3464b99440ca1873e3d63", "repoId": "cyankiwi/DASD-30B-A3B-Thinking-Preview-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
 {"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392495+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955494+00:00", "framework": "vllm_tokenizer_patch", "intentId": "043e3f0e927044dc88be413106c0c3ef", "repoId": "cyankiwi/AgentCPM-Report-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392491+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955441+00:00", "framework": "vllm_tokenizer_patch", "intentId": "f0e5fa0b54b2490d8c96867c06748d37", "repoId": "RecursiveSI/flint-int4", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392486+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955383+00:00", "framework": "vllm_tokenizer_patch", "intentId": "2aea87643d5a4079915c34de151a5a6c", "repoId": "cyankiwi/MiniCPM-SALA-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "f7d9fb6f458f4e2f9e86dbf58a76ca55", "completedAt": "2026-10-22T09:55:48.392466+00:00", "configFingerprint": "705b8b3614c86a1ac722c97f43e177e84ca20282e9799fffb1b890510d641728", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:55:45.955268+00:00", "framework": "llamacpp", "intentId": "9d408fdbeb4e4eb3b93f867276e357a2", "repoId": "AtomicChat/Qwen3.5-9B-DFlash-GGUF", "safeConfigVector": {"gpuNum": 1, "maxModelLen": 4096}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323266+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.713219+00:00", "framework": "vllm_tokenizer_patch", "intentId": "9ca2c61132be4d8f9950d76b63bcc04e", "repoId": "cyankiwi/Nemotron-Cascade-8B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323264+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.713153+00:00", "framework": "vllm_tokenizer_patch", "intentId": "a3a5361fc28e40e69ea81b1bdcb3a9e7", "repoId": "cyankiwi/Nemotron-Cascade-14B-Thinking-AWQ-8bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323261+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.713099+00:00", "framework": "vllm_tokenizer_patch", "intentId": "8706ef5ea7bb486387f8c6d590c9f315", "repoId": "cyankiwi/KAT-Dev-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323258+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.713044+00:00", "framework": "vllm_tokenizer_patch", "intentId": "d9278681e51c4b6d8db5e41d5a659e54", "repoId": "cyankiwi/LFM2-8B-A1B-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323254+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.712994+00:00", "framework": "vllm_tokenizer_patch", "intentId": "c3069ce075734cd197808d51050e71db", "repoId": "cyankiwi/Olmo-3.1-32B-Think-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323251+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.712945+00:00", "framework": "vllm_tokenizer_patch", "intentId": "c4a26ba23ff9474081ce21cb35e4a4cf", "repoId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323248+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.712894+00:00", "framework": "vllm_tokenizer_patch", "intentId": "902f1e7ff6374c20b977ee28fc61864b", "repoId": "cyankiwi/NVIDIA-Nemotron-Nano-9B-v2-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
-{"batchId": "cfd6fe03fe0142b59aacabe7f4cfd5b8", "completedAt": "2026-10-22T09:54:10.323245+00:00", "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configSource": "modelhub_live", "createdAt": "2026-10-22T09:54:07.712844+00:00", "framework": "vllm_tokenizer_patch", "intentId": "c5574e0bf83c4939b9e2fd27ebdd8a6e", "repoId": "cyankiwi/DASD-30B-A3B-Thinking-Preview-AWQ-4bit", "safeConfigVector": {"gpuNum": 1}, "status": "capacity_deferred", "targetGpu": "Ascend_910-b4", "taskId": null, "taskType": "text-generation"}
diff --git a/.modelhub_state/routing_intelligence.json b/.modelhub_state/routing_intelligence.json
index 834627af0b..8927432c92 100644
--- a/.modelhub_state/routing_intelligence.json
+++ b/.modelhub_state/routing_intelligence.json
@@ -1,6 +1,6 @@
 {
   "acceptedByCategory": {
-    "unified_success_first": 3845
+    "unified_success_first": 3846
   },
   "acceptedByRoute": {
     "text-generation|Ascend_910-b3|llamacpp": 36,
@@ -8,7 +8,7 @@
     "text-generation|Ascend_910-b3|vllm_tokenizer_patch": 265,
     "text-generation|Ascend_910-b4|llamacpp": 25,
     "text-generation|Ascend_910-b4|vllm": 36,
-    "text-generation|Ascend_910-b4|vllm_tokenizer_patch": 272,
+    "text-generation|Ascend_910-b4|vllm_tokenizer_patch": 273,
     "text-generation|Biren_166m|vllm": 81,
     "text-generation|Biren_166m|vllm_fix_tokenizer": 252,
     "text-generation|Cambricon_mlu-370-x4|vllm": 73,
@@ -39,8 +39,8 @@
     "text-generation|hygon_k100-ai|vllm": 35,
     "text-generation|hygon_k100-ai|vllm-patch-tokenizer": 238
   },
-  "acceptedSinceRefresh": 3845,
-  "acceptedTotal": 3845,
-  "generatedAt": "2026-10-22T10:28:25.568925+00:00",
+  "acceptedSinceRefresh": 3846,
+  "acceptedTotal": 3846,
+  "generatedAt": "2026-10-22T10:30:27.470229+00:00",
   "version": 1
 }
diff --git a/ledger/submissions.jsonl b/ledger/submissions.jsonl
index 1db2b0d06d..b6fd2fd181 100644
--- a/ledger/submissions.jsonl
+++ b/ledger/submissions.jsonl
@@ -1248,3 +1248,4 @@
 {"framework": "vllm_tokenizer_patch", "modelAddress": "https://modelscope.cn/models/XingChen-AGI/Xing4.0-29B-A4B", "modelId": "XingChen-AGI/Xing4.0-29B-A4B", "submitTime": "2026-10-13T12:52:58.552750+00:00", "targetGpu": "Ascend_910-b3", "taskId": "5512698", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-tokenizer-patch-ascend-910-b3"}
 {"framework": "vllm-patch-tokenizer", "modelAddress": "https://modelscope.cn/models/XingChen-AGI/Xing4.0-29B-A4B", "modelId": "XingChen-AGI/Xing4.0-29B-A4B", "submitTime": "2026-10-13T13:12:18.147765+00:00", "targetGpu": "hygon_k100-ai", "taskId": "5513114", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-patch-tokenizer-hygon-k100-ai"}
 {"framework": "llamacpp", "modelAddress": "https://modelscope.cn/models/DavidAU/DarkSapling-V2-Ultra-Quality-7B-GGUF", "modelId": "DavidAU/DarkSapling-V2-Ultra-Quality-7B-GGUF", "submitTime": "2026-10-15T20:46:16.830124+00:00", "targetGpu": "hygon_k100-ai", "taskId": "5569369", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-llamacpp-hygon-k100-ai"}
+{"framework": "vllm_tokenizer_patch", "modelAddress": "https://modelscope.cn/models/cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "modelId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "submitTime": "2026-10-22T10:30:26.211847+00:00", "targetGpu": "Ascend_910-b4", "taskId": "5715909", "taskType": "text-generation", "templateId": "modelhub-live-text-generation-vllm-tokenizer-patch-ascend-910-b4"}
diff --git a/manifest.json b/manifest.json
index 3aacb2f322..5a78a55d10 100644
--- a/manifest.json
+++ b/manifest.json
@@ -1,7 +1,7 @@
 {
   "agentVersion": "2026.09.04.4",
   "checksums": {
-    ".modelhub_state/account_capacity.json": "29d6ef6a38d178e3250b370bcca69315c52ee5b7842caeb5f83bcae570c1da66",
+    ".modelhub_state/account_capacity.json": "27ad125d4dff7469dc5d2d31300677f08a71ff830f785fbf3019ae621299c75e",
     ".modelhub_state/architecture_compatibility_blacklist.json": "6b2e82148cf8189f6101837f9292f12c87bd0d38e8ca1a537e91c0698777534e",
     ".modelhub_state/architecture_history_backfill.json": "a92348206605da04f75c9e26e7c818b76f259e0b28983f18b6375ef94802f0ab",
     ".modelhub_state/market_intelligence.json": "ee57dd7ec58e1ed537b89f73a17ebcb4bfb077da89b55be643f406037d52386f",
@@ -10,16 +10,16 @@
     ".modelhub_state/queue_cleanup_latest.json": "d1b2c9882d025fc6633c615fec46ac1442177e814b6bed4e9f8da8370b926a2a",
     ".modelhub_state/recent_outcomes.jsonl": "5f39ab623ae6524a7a4e30fa4827e61d28cdbb7bb4cee4ac00310cffb73d3d0b",
     ".modelhub_state/recovery_active_tasks.jsonl": "7dc3bc634feb7ba53917011a6e98b845937b82fc33ad26752637c828e944435a",
-    ".modelhub_state/recovery_intents.jsonl": "8d95e8a4c3166dfb99ef8346fca150ef72aa1312bdb76d916d54afd180519cfb",
-    ".modelhub_state/routing_intelligence.json": "6aa10cf1a8927f98c69dfb6b2bfabd07f6b19d7d87ae4e9a6b73f16d1efb3883",
+    ".modelhub_state/recovery_intents.jsonl": "61d1d8940601c3a35482302f2ef90a9d0874c5655754e692f36445664d5b6432",
+    ".modelhub_state/routing_intelligence.json": "68610a71203491d8b4a641f4a0f950f04a107793311f28746d16a15bc0320b62",
     ".modelhub_state/submission_exclusions.jsonl": "e2677e09d968a0c44c65d7c030b9c8ea2c9d39dc378ce302c2a309ab54d600d9",
     ".modelhub_state/worker_crashes.jsonl": "55ae96a720edf99ab4f90d3da3465977d66990ab194b8b0e7db7d1ff59682921",
-    "ledger/submissions.jsonl": "6e70d026632898da50811334ca97300bafc0d8ef49404ad375c619b869a43d2c",
-    "outcomes/submissions.jsonl": "49da490cd9204a71760158cee91cf85060d13224fc1a67762771061593baff99"
+    "ledger/submissions.jsonl": "ce99420eada0c078ff46d583f81ae20a8339f7e3d7e8a5180aa1a99c6a3646b1",
+    "outcomes/submissions.jsonl": "3d7de488e4750ce82d9ae685ec562d9a9820be8edaf3a5216861f9238ca3778d"
   },
-  "generation": 29115,
-  "phase": "intent",
+  "generation": 29116,
+  "phase": "result",
   "schemaVersion": 1,
-  "updatedAt": "2026-10-22T10:30:25.523243+00:00",
+  "updatedAt": "2026-10-22T10:30:27.830557+00:00",
   "writerId": "f8795cac6afa4a14b81a33808cc7d9c0"
 }
diff --git a/outcomes/submissions.jsonl b/outcomes/submissions.jsonl
index 1ab1501049..8de6ed76a6 100644
--- a/outcomes/submissions.jsonl
+++ b/outcomes/submissions.jsonl
@@ -1092,3 +1092,4 @@
 {"failReason": null, "framework": "vllm_tokenizer_patch", "lastSyncTime": null, "modelId": "cyankiwi/Kimi-Linear-48B-A3B-Instruct-AWQ-4bit", "modelProfile": {"architectures": ["KimiLinearForCausalLM"], "configFingerprint": "8be5b78bdba5c4413c1e4ab91a46e27c34642b25d5dc059ad3a11d50d7e33593", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 30532500512, "estimatedRequiredGiB": 34.133, "gpuMemoryEvidence": {"memoryGiB": 64.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 64.0, "maximumRepositorySizeGiB": 53.333, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "kimi_linear", "modelscopeFileSize": 30542125733, "modelscopeLicense": "mit", "modelscopeParams": 49122681728, "modelscopeTags": ["license:mit", "model_type:kimi_linear", "library:transformer", "library:safetensors", "library:pytorch", "task:text-generation"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "compressed-tensors", "repositoryOnDiskBytes": 30542125733}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-22T07:17:02.348716+00:00", "targetGpu": "MetaX_c-500", "taskId": "5712853", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "llamacpp", "lastSyncTime": null, "modelId": "XHToken/Spark-X2.6-4B-GGUF", "modelProfile": {"architectures": [], "configFingerprint": "de86b185053df1a87f4f6a1e0579898f12e492a36aefb07c55e30e7bd8393778", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 4375021824, "estimatedRequiredGiB": 18.048, "gpuMemoryEvidence": {"memoryGiB": 32.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 32.0, "maximumRepositorySizeGiB": 26.667, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": null, "modelscopeFileSize": 16148689583, "modelscopeLicense": "apache-2.0", "modelscopeParams": 4112079360, "modelscopeTags": ["license:apache-2.0", "library:transformer", "library:gguf", "task:text-generation", "custom_tag:llm", "custom_tag:sparkx2_5", "custom_tag:gguf", "deploy:swingdeploy"], "modelscopeTasks": ["text-generation"], "quantizationMethod": null, "repositoryOnDiskBytes": 16148689583}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-22T08:51:51.071527+00:00", "targetGpu": "Ascend_910-b4", "taskId": "5714121", "taskType": "text-generation", "verifyResult": null}
 {"failReason": null, "framework": "vllm_tokenizer_patch", "lastSyncTime": null, "modelId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-8bit", "modelProfile": {"architectures": ["Ernie4_5_MoeForCausalLM"], "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 24682886272, "estimatedRequiredGiB": 27.602, "gpuMemoryEvidence": {"memoryGiB": 32.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 32.0, "maximumRepositorySizeGiB": 26.667, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "ernie4_5_moe", "modelscopeFileSize": 24697709852, "modelscopeLicense": "apache-2.0", "modelscopeParams": 22090203328, "modelscopeTags": ["license:apache-2.0", "model_type:ernie4_5_moe", "library:pytorch", "library:transformer", "library:safetensors", "task:text-generation", "custom_tag:ERNIE4.5"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "compressed-tensors", "repositoryOnDiskBytes": 24697709852}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-22T09:51:22.357498+00:00", "targetGpu": "Ascend_910-b4", "taskId": "5715097", "taskType": "text-generation", "verifyResult": null}
+{"failReason": null, "framework": "vllm_tokenizer_patch", "lastSyncTime": null, "modelId": "cyankiwi/ERNIE-4.5-21B-A3B-Thinking-AWQ-4bit", "modelProfile": {"architectures": ["Ernie4_5_MoeForCausalLM"], "configFingerprint": "167e80c9a74f5acd91819ac9a661546c8992cc5beb0bdd0a1dc9fc9659a5e778", "configOptimization": {"applied": false, "source": "official"}, "configSource": "modelhub_live", "estimatedLoadBytes": 14278393080, "estimatedRequiredGiB": 15.973, "gpuMemoryEvidence": {"memoryGiB": 32.0, "source": "local_modelhub_preflight_oom"}, "gpuMemoryGiB": 32.0, "maximumRepositorySizeGiB": 26.667, "memorySizingBasis": "recursive_repository_on_disk", "modelCard": {}, "modelType": "ernie4_5_moe", "modelscopeFileSize": 14292651057, "modelscopeLicense": "apache-2.0", "modelscopeParams": 22090203328, "modelscopeTags": ["license:apache-2.0", "model_type:ernie4_5_moe", "library:pytorch", "library:transformer", "library:safetensors", "task:text-generation", "custom_tag:ERNIE4.5"], "modelscopeTasks": ["text-generation"], "quantizationMethod": "compressed-tensors", "repositoryOnDiskBytes": 14292651057}, "outcome": "pending", "status": "pending", "submitTime": "2026-10-22T10:30:26.211847+00:00", "targetGpu": "Ascend_910-b4", "taskId": "5715909", "taskType": "text-generation", "verifyResult": null}