Skip to content

fix: inter-token latency metric calculation #99

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

fix: inter-token latency metric calculation #99

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b4600ac
fix: exclude ttft from itl in openai clients
Amirresm Nov 8, 2025
b4e1f8a
fix: use correct output token count in itl
Amirresm Nov 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/llmperf/ray_clients/openai_chat_completions_client.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,7 +109,7 @@ def llm_request(self, request_config: RequestConfig) -> Dict[str, Any]:
print(f"Warning Or Error: {e}")
print(error_response_code)

metrics[common_metrics.INTER_TOKEN_LAT] = sum(time_to_next_token) #This should be same as metrics[common_metrics.E2E_LAT]. Leave it here for now
metrics[common_metrics.INTER_TOKEN_LAT] = sum(time_to_next_token) - ttft
metrics[common_metrics.TTFT] = ttft
metrics[common_metrics.E2E_LAT] = total_request_time
metrics[common_metrics.REQ_OUTPUT_THROUGHPUT] = output_throughput
Expand Down
8 changes: 4 additions & 4 deletions token_benchmark_ray.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -117,8 +117,8 @@ def launch_request(thread_index):
num_output_tokens = get_token_length(gen_text)
with completed_requests_lock:
if num_completed_requests < max_num_completed_requests:
if num_output_tokens:
request_metrics[common_metrics.INTER_TOKEN_LAT] /= request_metrics[common_metrics.NUM_OUTPUT_TOKENS]
if num_output_tokens > 1:
request_metrics[common_metrics.INTER_TOKEN_LAT] /= (num_output_tokens - 1)
else:
request_metrics[common_metrics.INTER_TOKEN_LAT] = 0
request_metrics[common_metrics.NUM_OUTPUT_TOKENS] = num_output_tokens
Expand Down Expand Up @@ -154,8 +154,8 @@ def launch_request(thread_index):
num_output_tokens = get_token_length(gen_text)
with completed_requests_lock:
if num_completed_requests < max_num_completed_requests:
if num_output_tokens:
request_metrics[common_metrics.INTER_TOKEN_LAT] /= num_output_tokens
if num_output_tokens > 1:
request_metrics[common_metrics.INTER_TOKEN_LAT] /= (num_output_tokens - 1)
else:
request_metrics[common_metrics.INTER_TOKEN_LAT] = 0
request_metrics[common_metrics.NUM_OUTPUT_TOKENS] = num_output_tokens
Expand Down