3694614dd9
Add experiments/dsv4_p800_max_context_length/ to find the longest input context that P800 SGLang INT8 can serve for DeepSeek-V4-Flash-INT8. - Docker-based server start with dynamic --context-length. - Single-request, single-output-token probes for each candidate length. - Records success/failure, server args, and metrics per length. - Generates results.json and report.md summarizing the max supported length. - Smoke-tested at 4096 tokens successfully.
21 lines
591 B
JSON
21 lines
591 B
JSON
{
|
|
"metadata": {
|
|
"experiment": "dsv4_p800_sglang",
|
|
"run_id": "20260708-054332",
|
|
"timestamp": "2026-07-08T05:43:32+00:00",
|
|
"model": "/data1/models/DeepSeek-V4-Flash-INT8",
|
|
"backend": "sglang",
|
|
"engine": "sglang-xpu",
|
|
"hardware": "8x Kunlun P800 XPU",
|
|
"accelerator": "Kunlun P800 XPU",
|
|
"chip": "kunlun_p800",
|
|
"script": "experiments/dsv4_p800_sglang/run_bench.sh",
|
|
"env": "",
|
|
"git_commit": "4ef6543",
|
|
"git_dirty": "dirty",
|
|
"description": "Kunlun P800 SGLang benchmark for DeepSeek-V4-Flash-INT8"
|
|
},
|
|
"config": {},
|
|
"scenarios": []
|
|
}
|