TEST: update turbomind qwen3.5 config

autotest/configs/Qwen/Qwen3.5-397B-A17B-FP8.yml

@@ -11,6 +11,9 @@ h:
       max-batch-size: 256
       cache-max-entry-count: 0.9
   backends:
+  - name: turbomind
+    communicators:
+    - nccl
   - name: pytorch
     communicators:
     - nccl
@@ -81,3 +84,30 @@ h:
     top-p: 0.95
     chat-template-kwargs:
       enable_thinking: true
+- model_type:
+  - chat
+  - vl
+  engine_config:
+    tp: 8
+    cp: 2
+    extra:
+      max-batch-size: 256
+      cache-max-entry-count: 0.9
+  backends:
+  - name: turbomind
+    communicators:
+    - nccl
+  test_coverage:
+  - benchmark
+  - evaluate
+  - func
+  - longtext_benchmark
+  - longtext_evaluate
+  - mllm_evaluate
+  gen_config:
+    temperature: 1.0
+    top-k: 20
+    repetition-penalty: 1.0
+    top-p: 0.95
+    chat-template-kwargs:
+      enable_thinking: true

autotest/configs/Qwen/Qwen3.5-397B-A17B.yml

@@ -11,6 +11,9 @@ h:
       max-batch-size: 256
       cache-max-entry-count: 0.7
   backends:
+  - name: turbomind
+    communicators:
+    - nccl
   - name: pytorch
     communicators:
     - nccl
@@ -88,6 +91,33 @@ h:
     top-p: 0.95
     chat-template-kwargs:
       enable_thinking: true
+- model_type:
+  - chat
+  - vl
+  engine_config:
+    tp: 8
+    cp: 2
+    extra:
+      max-batch-size: 256
+      cache-max-entry-count: 0.7
+  backends:
+  - name: turbomind
+    communicators:
+    - nccl
+  test_coverage:
+  - benchmark
+  - evaluate
+  - func
+  - longtext_benchmark
+  - longtext_evaluate
+  - mllm_evaluate
+  gen_config:
+    temperature: 1.0
+    top-k: 20
+    repetition-penalty: 1.0
+    top-p: 0.95
+    chat-template-kwargs:
+      enable_thinking: true
 ascend:
 - model_type:
   - chat