init v0.23.0

Signed-off-by: Sun Ruoxi <sunruoxi@4paradigm.com>
2026-08-27 15:11:51 +08:00
parent b582a8e7d1
commit 7f8a1b1f7a
2849 changed files with 712887 additions and 22001 deletions
--- a/tests/e2e/weekly/single_node/configs/Qwen3-32B.yaml
+++ b/tests/e2e/weekly/single_node/configs/Qwen3-32B.yaml
@@ -0,0 +1,55 @@
+# ==========================================
+# ACTUAL TEST CASES
+# ==========================================
+#workflow暂缺A2yaml
+#          - name: qwen3-32b
+#            os: linux-aarch64-a2b3-4
+#            config_file_path: Qwen3-32B.yaml
+test_cases:
+  - name: "Qwen3-32B-TP4"
+    model: "Qwen/Qwen3-32B"
+    envs:
+      TASK_QUEUE_ENABLE: "1"
+      OMP_PROC_BIND: "false"
+      HCCL_OP_EXPANSION_MODE: "AIV"
+      PAGED_ATTENTION_MASK_LEN: "5500"
+      SERVER_PORT: "DEFAULT_PORT"
+    server_cmd:
+      - "--no-enable-prefix-caching"
+      - "--tensor-parallel-size"
+      - "4"
+      - "--port"
+      - "$SERVER_PORT"
+      - "--max-model-len"
+      - "36864"
+      - "--max-num-batched-tokens"
+      - "36864"
+      - "--block-size"
+      - "128"
+      - "--trust-remote-code"
+      - "--gpu-memory-utilization"
+      - "0.9"
+      - "--additional-config"
+      - '{"enable_weight_nz_layout":true}'
+    benchmarks:
+      acc:
+        case_type: accuracy
+        dataset_path: vllm-ascend/gsm8k
+        request_conf: vllm_api_general_chat
+        dataset_conf: gsm8k/gsm8k_gen_0_shot_cot_chat_prompt
+        max_out_len: 32768
+        num_prompts: 32
+        batch_size: 32
+        baseline: 95
+        threshold: 5
+      perf:
+        case_type: performance
+        dataset_path: vllm-ascend/GSM8K-in3500-bs400
+        request_conf: vllm_api_stream_chat
+        dataset_conf: gsm8k/gsm8k_gen_0_shot_cot_str_perf
+        num_prompts: 80
+        max_out_len: 1500
+        batch_size: 20
+        request_rate: 0
+        baseline: 1
+        threshold: 0.97