yy-fighting
1482601ae5
Add num_prompts = 5×concurrency convention and update all experiments
...
- Document the rule in docs/EXPERIMENT_GUIDE.md section 7
- Update all config.env SCENARIOS to follow the 5x rule
- Migrate dsv4_h200_sglang, dsv4_h200_vllm, dsv4_p800_sglang
from global NUM_PROMPTS to per-scenario num_prompts
- Keep long-context phase2 as exception (low counts documented)
- Add legacy 3-field fallback in run_bench.sh loops
2026-07-09 09:26:41 +00:00
yy-fighting
7e614fb702
Add DSpark backend to long-context matrix experiment
...
- Add start_dspark.sh with --spec-method dspark --spec-tokens 5
- Add BACKENDS env var to run only selected backends
- Fix client/backend mapping for DSpark (bench client uses vllm, parser uses vllm)
- Add three-way compare.py for sglang/vllm/dspark
- Reuse existing sglang/vllm results from 20260708-100016
- Add DSpark long-context results and comparison
2026-07-09 02:36:57 +00:00
yy-fighting
27ceaac47f
results(long-context-matrix): complete 64k/128k/256k matrix run 20260708-100016; use common scripts
2026-07-08 11:21:15 +00:00
yy-fighting
9b135aa4ff
revert: remove intermediate matrix results committed by mistake (again)
2026-07-08 11:06:54 +00:00
yy-fighting
80959cf765
docs: move SLO_STANDARDS.md from scripts/ to docs/; update references
2026-07-08 11:06:49 +00:00
yy-fighting
ccf65e745b
revert: remove intermediate matrix results committed by mistake
2026-07-08 11:04:23 +00:00
yy-fighting
9b485c05f7
cleanup(scripts): remove unused legacy scripts; update README and workflow docs
2026-07-08 11:04:16 +00:00
yy-fighting
868ebc6cdf
feat(long-context-matrix): add SGLang vs vLLM long-context matrix experiment
2026-07-08 10:09:32 +00:00
