Reorganize experiments into hardware-specific subdirectories

Browse Source

Move all experiments under hardware-specific folders:
- experiments/h200/     : H200 GPU experiments (15 dirs)
- experiments/h20/      : H20 GPU experiments (2 dirs)
- experiments/p800/     : Kunlun P800 experiments (3 dirs)
- experiments/pro6000/    : RTX 6000D experiments (2 dirs)

This improves discoverability and keeps hardware-specific configs
isolated from each other.

...

This commit is contained in:

SSKJ Dev 2026-07-16 04:11:07 +00:00

parent 90fc378d2b

commit a4e38b9e33

211 changed files with 0 additions and 0 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Show Stats Download Patch File Download Diff File Expand all files Collapse all files

0

experiments/dsv4_h200_sglang_tp_dp_matrix/adaptive_config.env → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/adaptive_config.env

0

experiments/dsv4_h20_sglang_tp_dp_matrix/compare.py → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/compare.py

0

experiments/dsv4_h20_sglang_tp_dp_matrix/config.env → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/config.env

0

experiments/dsv4_h20_sglang_tp_dp_matrix/generate_scenarios.py → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/generate_scenarios.py

0

experiments/dsv4_h20_sglang_tp_dp_matrix/matrix.json → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/matrix.json

0

experiments/dsv4_h200_sglang_tp_dp_matrix/run_adaptive_concurrency.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/run_adaptive_concurrency.sh

0

experiments/dsv4_h20_sglang_tp_dp_matrix/run_adaptive_concurrency_add16.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/run_adaptive_concurrency_add16.sh

0

experiments/dsv4_h20_sglang_tp_dp_matrix/run_bench.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/run_bench.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/run_sglang_in_container.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/run_sglang_in_container.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/start_sglang_container.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/start_sglang_container.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/start_sglang_docker.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/start_sglang_docker.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/start_sglang_dp.sh → experiments/h20/dsv4_h20_sglang_tp_dp_matrix/start_sglang_dp.sh

0

experiments/dsv4_h200_vllm_tp_dp_matrix/adaptive_config.env → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/adaptive_config.env

0

experiments/dsv4_h20_vllm_tp_dp_matrix/compare.py → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/compare.py

0

experiments/dsv4_h20_vllm_tp_dp_matrix/config.env → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/config.env

0

experiments/dsv4_h20_vllm_tp_dp_matrix/generate_scenarios.py → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/generate_scenarios.py

0

experiments/dsv4_h20_vllm_tp_dp_matrix/matrix.json → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/matrix.json

0

experiments/dsv4_h200_vllm_tp_dp_matrix/run_adaptive_concurrency.sh → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/run_adaptive_concurrency.sh

0

experiments/dsv4_h20_vllm_tp_dp_matrix/run_adaptive_concurrency_add16.sh → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/run_adaptive_concurrency_add16.sh

0

experiments/dsv4_h20_vllm_tp_dp_matrix/run_bench.sh → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/run_bench.sh

0

experiments/dsv4_h20_vllm_tp_dp_matrix/start_vllm_docker.sh → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/start_vllm_docker.sh

0

experiments/dsv4_h200_vllm_tp_dp_matrix/start_vllm_dp.sh → experiments/h20/dsv4_h20_vllm_tp_dp_matrix/start_vllm_dp.sh

0

experiments/dsv4_h200_256k_4k_probe/config.env → experiments/h200/dsv4_h200_256k_4k_probe/config.env

0

experiments/dsv4_h200_256k_4k_probe/results/20260708-094838/comparison.md → experiments/h200/dsv4_h200_256k_4k_probe/results/20260708-094838/comparison.md

0

experiments/dsv4_h200_256k_4k_probe/results/20260708-094838/sglang/report.md → experiments/h200/dsv4_h200_256k_4k_probe/results/20260708-094838/sglang/report.md

0

experiments/dsv4_h200_256k_4k_probe/results/20260708-094838/sglang/results.json → experiments/h200/dsv4_h200_256k_4k_probe/results/20260708-094838/sglang/results.json

0

experiments/dsv4_h200_256k_4k_probe/results/20260708-094838/vllm/report.md → experiments/h200/dsv4_h200_256k_4k_probe/results/20260708-094838/vllm/report.md

0

experiments/dsv4_h200_256k_4k_probe/results/20260708-094838/vllm/results.json → experiments/h200/dsv4_h200_256k_4k_probe/results/20260708-094838/vllm/results.json

0

experiments/dsv4_h200_256k_4k_probe/run_bench.sh → experiments/h200/dsv4_h200_256k_4k_probe/run_bench.sh

0

experiments/dsv4_h200_256k_4k_probe/start_sglang.sh → experiments/h200/dsv4_h200_256k_4k_probe/start_sglang.sh

0

experiments/dsv4_h200_256k_4k_probe/start_vllm.sh → experiments/h200/dsv4_h200_256k_4k_probe/start_vllm.sh

0

experiments/dsv4_h200_64k_sglang_vs_vllm/config.env → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/config.env

0

experiments/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/comparison.md → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/comparison.md

0

experiments/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/sglang/report.md → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/sglang/report.md

0

experiments/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/sglang/results.json → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/sglang/results.json

0

experiments/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/vllm/report.md → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/vllm/report.md

0

experiments/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/vllm/results.json → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/results/20260708-092556/vllm/results.json

0

experiments/dsv4_h200_64k_sglang_vs_vllm/run_bench.sh → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/run_bench.sh

0

experiments/dsv4_h200_64k_sglang_vs_vllm/start_sglang.sh → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/start_sglang.sh

0

experiments/dsv4_h200_64k_sglang_vs_vllm/start_vllm.sh → experiments/h200/dsv4_h200_64k_sglang_vs_vllm/start_vllm.sh

0

experiments/dsv4_h200_dspark/README.md → experiments/h200/dsv4_h200_dspark/README.md

0

experiments/dsv4_h200_dspark/config.env → experiments/h200/dsv4_h200_dspark/config.env

0

experiments/dsv4_h200_dspark/parse_results.py → experiments/h200/dsv4_h200_dspark/parse_results.py

0

experiments/dsv4_h200_dspark/run_bench.sh → experiments/h200/dsv4_h200_dspark/run_bench.sh

0

experiments/dsv4_h200_long_context_matrix/compare.py → experiments/h200/dsv4_h200_long_context_matrix/compare.py

0

experiments/dsv4_h200_long_context_matrix/config.env → experiments/h200/dsv4_h200_long_context_matrix/config.env

0

experiments/dsv4_h200_long_context_matrix/results/20260708-100016/comparison.md → experiments/h200/dsv4_h200_long_context_matrix/results/20260708-100016/comparison.md

0

experiments/dsv4_h200_long_context_matrix/results/20260708-100016/sglang/report.md → experiments/h200/dsv4_h200_long_context_matrix/results/20260708-100016/sglang/report.md

0

experiments/dsv4_h200_long_context_matrix/results/20260708-100016/sglang/results.json → experiments/h200/dsv4_h200_long_context_matrix/results/20260708-100016/sglang/results.json

0

experiments/dsv4_h200_long_context_matrix/results/20260708-100016/vllm/report.md → experiments/h200/dsv4_h200_long_context_matrix/results/20260708-100016/vllm/report.md

0

experiments/dsv4_h200_long_context_matrix/results/20260708-100016/vllm/results.json → experiments/h200/dsv4_h200_long_context_matrix/results/20260708-100016/vllm/results.json

0

experiments/dsv4_h200_long_context_matrix/results/20260709-020227/comparison.md → experiments/h200/dsv4_h200_long_context_matrix/results/20260709-020227/comparison.md

0

experiments/dsv4_h200_long_context_matrix/results/20260709-020227/dspark/report.md → experiments/h200/dsv4_h200_long_context_matrix/results/20260709-020227/dspark/report.md

0

experiments/dsv4_h200_long_context_matrix/results/20260709-020227/dspark/results.json → experiments/h200/dsv4_h200_long_context_matrix/results/20260709-020227/dspark/results.json

0

experiments/dsv4_h200_long_context_matrix/run_bench.sh → experiments/h200/dsv4_h200_long_context_matrix/run_bench.sh

0

experiments/dsv4_h200_long_context_matrix/start_dspark.sh → experiments/h200/dsv4_h200_long_context_matrix/start_dspark.sh

0

experiments/dsv4_h200_long_context_matrix/start_sglang.sh → experiments/h200/dsv4_h200_long_context_matrix/start_sglang.sh

0

experiments/dsv4_h200_long_context_matrix/start_vllm.sh → experiments/h200/dsv4_h200_long_context_matrix/start_vllm.sh

0

experiments/dsv4_h200_max_context_length/README.md → experiments/h200/dsv4_h200_max_context_length/README.md

0

experiments/dsv4_h200_max_context_length/config.env → experiments/h200/dsv4_h200_max_context_length/config.env

0

experiments/dsv4_h200_max_context_length/extract_metrics.py → experiments/h200/dsv4_h200_max_context_length/extract_metrics.py

0

experiments/dsv4_h200_max_context_length/kv_cache_analysis.md → experiments/h200/dsv4_h200_max_context_length/kv_cache_analysis.md

0

experiments/dsv4_h200_max_context_length/parse_results.py → experiments/h200/dsv4_h200_max_context_length/parse_results.py

0

experiments/dsv4_h200_max_context_length/results/20260708-143534/report.md → experiments/h200/dsv4_h200_max_context_length/results/20260708-143534/report.md

0

experiments/dsv4_h200_max_context_length/results/20260708-143534/results.json → experiments/h200/dsv4_h200_max_context_length/results/20260708-143534/results.json

0

experiments/dsv4_h200_max_context_length/run_bench.sh → experiments/h200/dsv4_h200_max_context_length/run_bench.sh

0

experiments/dsv4_h200_max_context_length/start_sglang.sh → experiments/h200/dsv4_h200_max_context_length/start_sglang.sh

0

experiments/dsv4_h200_max_context_length/start_vllm.sh → experiments/h200/dsv4_h200_max_context_length/start_vllm.sh

0

experiments/dsv4_h200_sglang/README.md → experiments/h200/dsv4_h200_sglang/README.md

0

experiments/dsv4_h200_sglang/config.env → experiments/h200/dsv4_h200_sglang/config.env

0

experiments/dsv4_h200_sglang/parse_results.py → experiments/h200/dsv4_h200_sglang/parse_results.py

0

experiments/dsv4_h200_sglang/results/20260708-065101/report.md → experiments/h200/dsv4_h200_sglang/results/20260708-065101/report.md

0

experiments/dsv4_h200_sglang/results/20260708-065101/results.json → experiments/h200/dsv4_h200_sglang/results/20260708-065101/results.json

0

experiments/dsv4_h200_sglang/run_bench.sh → experiments/h200/dsv4_h200_sglang/run_bench.sh

0

experiments/dsv4_h200_sglang/start_server.sh → experiments/h200/dsv4_h200_sglang/start_server.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/README.md → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/README.md

0

experiments/dsv4_h200_sglang_tp_dp_matrix/README_adaptive_concurrency.md → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/README_adaptive_concurrency.md

0

experiments/dsv4_h20_sglang_tp_dp_matrix/adaptive_config.env → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/adaptive_config.env

0

experiments/dsv4_h200_sglang_tp_dp_matrix/compare.py → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/compare.py

0

experiments/dsv4_h200_sglang_tp_dp_matrix/config.env → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/config.env

0

experiments/dsv4_h200_sglang_tp_dp_matrix/generate_scenarios.py → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/generate_scenarios.py

0

experiments/dsv4_h200_sglang_tp_dp_matrix/matrix.json → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/matrix.json

0

experiments/dsv4_h20_sglang_tp_dp_matrix/run_adaptive_concurrency.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/run_adaptive_concurrency.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/run_adaptive_concurrency_add16.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/run_adaptive_concurrency_add16.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/run_bench.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/run_bench.sh

0

experiments/dsv4_h20_sglang_tp_dp_matrix/run_sglang_in_container.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/run_sglang_in_container.sh

0

experiments/dsv4_h200_sglang_tp_dp_matrix/smoke_matrix.json → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/smoke_matrix.json

0

experiments/dsv4_h20_sglang_tp_dp_matrix/start_sglang_container.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/start_sglang_container.sh

0

experiments/dsv4_h20_sglang_tp_dp_matrix/start_sglang_docker.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/start_sglang_docker.sh

0

experiments/dsv4_h20_sglang_tp_dp_matrix/start_sglang_dp.sh → experiments/h200/dsv4_h200_sglang_tp_dp_matrix/start_sglang_dp.sh

0

experiments/dsv4_h200_sglang_vs_vllm/README.md → experiments/h200/dsv4_h200_sglang_vs_vllm/README.md

0

experiments/dsv4_h200_sglang_vs_vllm/config.env → experiments/h200/dsv4_h200_sglang_vs_vllm/config.env

0

experiments/dsv4_h200_sglang_vs_vllm/results/20260708-083346/comparison.md → experiments/h200/dsv4_h200_sglang_vs_vllm/results/20260708-083346/comparison.md

0

experiments/dsv4_h200_sglang_vs_vllm/results/20260708-083346/sglang/report.md → experiments/h200/dsv4_h200_sglang_vs_vllm/results/20260708-083346/sglang/report.md

0

experiments/dsv4_h200_sglang_vs_vllm/results/20260708-083346/sglang/results.json → experiments/h200/dsv4_h200_sglang_vs_vllm/results/20260708-083346/sglang/results.json

0

experiments/dsv4_h200_sglang_vs_vllm/results/20260708-083346/vllm/report.md → experiments/h200/dsv4_h200_sglang_vs_vllm/results/20260708-083346/vllm/report.md

0

experiments/dsv4_h200_sglang_vs_vllm/results/20260708-083346/vllm/results.json → experiments/h200/dsv4_h200_sglang_vs_vllm/results/20260708-083346/vllm/results.json

0

experiments/dsv4_h200_sglang_vs_vllm/run_bench.sh → experiments/h200/dsv4_h200_sglang_vs_vllm/run_bench.sh

0

experiments/dsv4_h200_sglang_vs_vllm/start_sglang.sh → experiments/h200/dsv4_h200_sglang_vs_vllm/start_sglang.sh

0

experiments/dsv4_h200_sglang_vs_vllm/start_vllm.sh → experiments/h200/dsv4_h200_sglang_vs_vllm/start_vllm.sh

Some files were not shown because too many files have changed in this diff Show More