Hugging Face's logo

Datasets:
agent-perf-bench
/
AgentPerfBench

Tasks:
Tabular Regression
Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Size:
100K - 1M
Tags:
llm-inference
benchmarking
gpu-profiling
vllm
sglang
agentic-workloads
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
8
AgentPerfBench / mse_validation /notes /mse-validation-debug-process-2026-05-05.md

Commit History

Preserve MSE validation evidence for synthetic replay claims (#3)
4b1b970

booth-algo commited on