dsr1-fp4-1k1k-mid-curve

best_of

1.00

burstiness

1.00

completed

5,120

duration

162

max_concurrency

512

mean_e2el_ms

15,579ms

mean_itl_ms

774ms

mean_tpot_ms

15.80ms

mean_ttft_ms

1,014ms

median_e2el_ms

15,540ms

median_itl_ms

771ms

median_tpot_ms

16.04ms

median_ttft_ms

766ms

num_prompts

5,120

output_throughput

29,136tok/s

p99_e2el_ms

18,848ms

p99_itl_ms

1,087ms

p99_tpot_ms

16.50ms

p99_ttft_ms

3,700ms

peak_output_tokens_per_s

35,997s

request_throughput

31.59tok/s

std_e2el_ms

1,219ms

std_itl_ms

154ms

std_tpot_ms

0.75ms

std_ttft_ms

890ms

total_input_tokens

4,717,859

total_output_tokens

4,722,209

total_token_throughput

58,246tok/s