sskj/experiments at 98cdb67b660b574ef5d9f92c21df71c05fc11b03 - sskj

History

shishi 98cdb67b66 fix(910c/glm52): 修复sglang0.5.2解析兼容性+TP=16设备挂载+health超时

sglang 0.5.2 bench_serving 输出与解析器不兼容,致 add16 c=32 崩溃:
- 缺 total_throughput -> total_tps 恒0,无法检测吞吐增益
- 缺 p95_*(仅p99) -> TTFT SLO 失效
- gain previous<=0 返回 inf -> json.loads(inf) 崩溃

修复:
1. adaptive_concurrency.py: 缺失时从 ttfts/itls 数组补算 p95/p50;
   total_tps 回退 input+output throughput; gain 返回 Infinity
2. parse_backend.py: 同上补算逻辑; 补 from __future__ import annotations
   (py3.9 下 dict|None 语法无法 import)
3. start_vllm_docker.sh: --device davinci0~15 支持 TP=16;
   health 超时可配(默认480x5s=40min,TP=16编译16 graph约60min);
   补驱动挂载+/mnt; 修容器名双后缀
4. run_adaptive_concurrency_add16.sh: --tokenizer 替代 --model;
   TORCH_DEVICE_BACKEND_AUTOLOAD=0; CONTAINER_PYTHON 路径; 导出 ENGINE_TP/DP
5. config.env: 固定 CONTAINER_NAME/DOCKER_IMAGE/GPU_MEM_UTIL
6. TP8_vs_TP16_report.md: TP=8 vs TP=16 手动测速对比报告

验证: TP=8 add16 c=16->c=32 不再崩溃; TP=16 编译完成变 healthy 推理正常

2026-07-28 16:43:05 +08:00

910c

fix(910c/glm52): 修复sglang0.5.2解析兼容性+TP=16设备挂载+health超时

2026-07-28 16:43:05 +08:00

h20

Add vLLM TP×DP matrix benchmarking scripts and Docker support

2026-07-23 02:06:47 +00:00

h200

update 8k and 32k

2026-07-17 16:47:25 +08:00

p800

fix: replace 5 buggy parse_results.py (raw_requests bloat) with shared parse_backend.py wrapper