Quantong Qiu
90eecc4ede
Add P800 long-context matrix experiment (64k/128k) for DeepSeek-V4-Flash-INT8
- New experiment: experiments/dsv4_p800_long_context_matrix/
- Single backend: sglang-xpu in Docker on Kunlun P800
- Context groups: 64k (max_context_len=80000, max_running=10) and
128k (max_context_len=140000, max_running=5)
- Scenarios: concurrency sweeps x output lengths 256/1024/4096
- Parameterized start_sglang.sh enforces max_running >= 2 to avoid
P800 INT8 req_to_token_pool allocation failures
- Reuses proven P800 INT8 environment, NSA/EAGLE settings, and
bench_serving invocations from dsv4_p800_max_context_length
2026-07-09 04:45:04 +00:00
..
2026-07-08 11:06:49 +00:00
2026-07-08 11:06:49 +00:00
2026-07-08 06:56:45 +00:00
2026-07-09 02:36:57 +00:00
2026-07-08 15:33:48 +00:00
2026-07-08 06:56:45 +00:00
2026-07-08 11:06:49 +00:00
2026-07-08 06:31:03 +00:00
2026-07-08 16:13:01 +00:00
2026-07-09 01:42:46 +00:00
2026-07-09 04:45:04 +00:00
2026-07-09 04:08:40 +00:00
2026-07-09 03:30:58 +00:00
2026-07-08 06:40:39 +00:00
2026-07-08 10:48:43 +00:00