MCPcopy Create free account
hub / github.com/FlashSampling/FlashSampling / print_per_run

Function print_per_run

benchmarking/vllm/collect_results.py:102–111  ·  view source on GitHub ↗
(results_dir: Path, variant_key: str)

Source from the content-addressed store, hash-verified

100
101
102def print_per_run(results_dir: Path, variant_key: str):
103 display_name = dict(VARIANTS)[variant_key]
104 df, run_name = read_summary(results_dir, variant_key)
105 print(f"{variant_key}: {run_name}")
106
107 pivoted = df.pivot(index="max_concurrency", columns="run_number", values="median_tpot_ms")
108 pivoted.columns = [f"Run {c}" for c in pivoted.columns]
109 pivoted.index.name = "Concurrency"
110 print(f"\n{display_name} — Median TPOT (ms) per run\n")
111 print(tabulate(pivoted, headers="keys", tablefmt="grid", floatfmt=".2f"))
112
113
114def main():

Callers 1

mainFunction · 0.85

Calls 1

read_summaryFunction · 0.85

Tested by

no test coverage detected