init v0.23.0

Signed-off-by: Sun Ruoxi <sunruoxi@4paradigm.com>

tests/e2e/models/configs/Qwen3-8B-W8A8.yaml

@@ -0,0 +1,25 @@
+model_name: "vllm-ascend/Qwen3-8B-W8A8"
+model_type: "vllm"
+hardware: "Atlas A2 Series"
+
+serve:
+  tensor_parallel_size: 1
+  dtype: auto
+  max_model_len: 4096
+  gpu_memory_utilization: 0.8
+  trust_remote_code: false
+  quantization: ascend
+  enable_thinking: false
+
+tasks:
+- name: "gsm8k"
+  metrics:
+  - name: "exact_match,strict-match"
+    value: 0.80
+  - name: "exact_match,flexible-extract"
+    value: 0.82
+
+num_fewshot: 5
+apply_chat_template: true
+fewshot_as_multiturn: true
+batch_size: "auto"