# ==========================================
# Shared Configurations
# ==========================================

_envs: &envs
  TASK_QUEUE_ENABLE: "1"
  VLLM_USE_V1: "1"
  HCCL_OP_EXPANSION_MODE: "AIV"
  VLLM_ASCEND_ENABLE_FLASHCOMM: "1"
  SERVER_PORT: "DEFAULT_PORT"

_server_cmd: &server_cmd
  - "--tensor-parallel-size"
  - "4"
  - "--port"
  - "$SERVER_PORT"
  - "--distributed_executor_backend"
  - "mp"
  - "--trust-remote-code"
  - "--reasoning-parser"
  - "deepseek_r1"
  - "--gpu-memory-utilization"
  - "0.9"

_benchmarks_gsm8k_lite: &benchmarks_gsm8k_lite
  acc:
    case_type: accuracy
    dataset_path: vllm-ascend/gsm8k_lite
    request_conf: vllm_api_general_chat
    dataset_conf: gsm8k/gsm8k_gen_0_shot_cot_chat_prompt
    max_out_len: 10240
    num_prompts: 32
    batch_size: 32
    baseline: 100
    threshold: 10

_benchmarks_perf1: &benchmarks_perf1
  perf:
    case_type: performance
    dataset_path: vllm-ascend/GSM8K-in3500-bs8000-qwen3
    request_conf: vllm_api_stream_chat
    dataset_conf: gsm8k/gsm8k_gen_0_shot_cot_str_perf
    num_prompts: 340
    max_out_len: 1500
    batch_size: 60
    request_rate: 0
    baseline: 724.59
    threshold: 0.97

# ==========================================
# ACTUAL TEST CASES
# ==========================================

test_cases:
  - name: "Qwq_32b_Feature_A3"
    model: "Qwen/QwQ-32B"
    envs:
      <<: *envs
    server_cmd: *server_cmd
    server_cmd_extra:
      - "--max-num-seqs"
      - "256"
      - "--max-model-len"
      - "32768"
      - "--max-num-batched-tokens"
      - "32768"
      - "--enforce-eager"
      - "--additional-config"
      - '{"ascend_scheduler_config":{"enabled":false}}'
  - name: "Qwq_32B_gsm8k_Accuracy_A3"
    model: "Qwen/QwQ-32B"
    envs:
      <<: *envs
      LD_PRELOAD: "/usr/lib/aarch64-linux-gnu/libjemalloc.so.2:$LD_PRELOAD"
      VLLM_ASCEND_ENABLE_DENSE_OPTIMIZE: "1"
      VLLM_ASCEND_ENABLE_PREFETCH_MLP: "1"
    server_cmd: *server_cmd
    server_cmd_extra:
      - "--max-model-len"
      - "35840"
      - "--max-num-batched-tokens"
      - "40960"
      - "--block-size"
      - "128"
      - "--async-scheduling"
      - "--no-enable-prefix-caching"
      - "--compilation-config"
      - '{"cudagraph_mode": "FULL_DECODE_ONLY"}'
    benchmarks:
      <<: *benchmarks_gsm8k_lite
  - name: "Qwq_Performance_A3_3500_1500_A3"
    model: "Qwen/QwQ-32B"
    envs:
      <<: *envs
      OMP_PROC_BIND: "false"
      ASCEND_RT_VISIBLE_DEVICES: "0,1,2,3"
      TASK_QUEUE_ENABLE: "1"
    server_cmd: *server_cmd
    server_cmd_extra:
      - "--max-model-len"
      - "5500"
      - "--max-num-batched-tokens"
      - "40960"
      - "--block-size"
      - "128"
      - "--async-scheduling"
      - "--compilation-config"
      - '{"cudagraph_mode": "FULL_DECODE_ONLY", "cudagraph_capture_sizes": [ 1,5,8,72,76,80,100,120,140,144,160,192,216,240,252,288,320,336,360,384,400,408,416,420,432,480,540,576,600 ]}'
      - "--additional-config"
      - '{"pa_shape_list": [32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223,224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239,240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255,256 ]}'
    benchmarks:
      <<: *benchmarks_perf1