-
accuracy.py: calculate the pass@k accuracy of n agent runs on one query.
-
avg_accuracy.py: calcuate the average pass@1 accuracy of n agent runs on one dataset.
-
avg_pass_k.py: calcuate the average pass@k accuracy of n agent runs on one dataset.
-
cost.py: calculate the total cost (USD) of n agent runs on one query.
-
llm_calls.py: calculate the average count, latency, and input/output tokens of API calls of n agent runs on one query.
-
tool_calls.py: calculate the average count, execution latency, and success rate of tool execution of n agent runs on one query.
-
traj_stats.py: caculcate the average latency, #iterations, #tool calls, #db queries, #python execs of n agent runs on one query.
-
python_exec_failure_rate.py: calculate the average failure rate of execute_python of n agent runs on one query
-
parellel_rate.py: calculate the average/max parallel rate of API call across all agent API calls on all datatsets.
stats_scripts
Directory actions
More options
Directory actions
More options
stats_scripts
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
parent directory.. | ||||