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

Update README.md

#2
by booth-algo - opened
base: refs/heads/main
←
from: refs/pr/2
Discussion Files changed
+1
-1
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -31,7 +31,7 @@ configs:
31
 
32
  # AgentPerfBench
33
 
34
- LLM inference benchmark dataset measuring serving performance (TTFT, TPOT, ITL, throughput) across 9 models, 14 GPU configurations, 2 serving engines, and 20+ workload profiles spanning single-turn chat, multi-turn agent sessions, and synthetic stress tests. Includes per-kernel CUDA profiling data for roofline analysis.
35
 
36
  ## Dataset Configurations
37
 
 
31
 
32
  # AgentPerfBench
33
 
34
+ LLM inference benchmark dataset measuring serving performance (TTFT, TPOT, ITL, throughput) across 9 models, 14 GPU configurations, 2 serving engines, and 20+ workload profiles spanning single-turn chat, multi-turn agent sessions, and synthetic stress tests. Includes per-kernel NVIDIA Nsight Compute profiling data for roofline analysis.
35
 
36
  ## Dataset Configurations
37