qqtang/sskj
1
2
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
88 Commits 1 Branch 0 Tags
Commit Graph

4 Commits

Author SHA1 Message Date
iiGray
60826dda09 add run_adaptive_concurrency_add16 2026-07-13 17:26:33 +08:00
Root
a7f453b997 feat: add run_bench.sh and fix config for dsv4_pro6000_sglang_tp_dp_matrix 
- Migrate run_bench.sh from dsv4_h200_sglang_tp_dp_matrix
- Add missing generate_scenarios.py, compare.py, matrix.json
- Fix MODEL_PATH to /root/data/models/ds_flash_v4
- Fix DATASET_PATH to local dummy_sharegpt.json
- Tune CONTEXT_LENGTH, MAX_RUNNING_REQUESTS, MEM_FRACTION_STATIC for RTX 6000D 83GB
- Exclude GPU 2 due to DRAM Uncorrectable ECC hardware error
 2026-07-13 05:22:31 +00:00
Quantong Qiu
302d1a9822 Rename DSL to OSL (Output Sequence Length) in dsv4_pro6000 experiment 
DSL -> OSL to better reflect the meaning: Output Sequence Length.
 2026-07-13 02:28:11 +00:00
Quantong Qiu
8ebe963269 Add dsv4_pro6000_sglang_tp_dp_matrix experiment for RTX 6000D 
- New experiment directory with adaptive concurrency benchmark for SGLang
- Platform config for 8x NVIDIA RTX 6000D 84GB
- Adapted from dsv4_h200_sglang_tp_dp_matrix with updated paths:
  - Model path: /data/hf_models/DeepSeek-V4-Flash
  - Venv path: /root/.miniconda3/envs/sglang
  - Dataset path: /data/yy/sskj/datasets/
 2026-07-13 02:18:34 +00:00