biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Frontend] Add max_tokens prometheus metric (#9881)

Browse Source 
Signed-off-by: Tomer Asida <tomera@ai21.com>
This commit is contained in:
tomeras91
2024-11-05 00:53:24 +02:00
committed by GitHub
parent 9a5664d4a4
commit ac04a97a9f
5 changed files with 23 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
tests/entrypoints/openai/test_metrics.py
View File

@@ -70,10 +70,14 @@ EXPECTED_VALUES = {
[("_sum", _NUM_REQUESTS * _NUM_GENERATION_TOKENS_PER_REQUEST),
("_count", _NUM_REQUESTS)],
"vllm:request_params_n": [("_count", _NUM_REQUESTS)],
"vllm:request_params_max_tokens":
[("_sum", _NUM_REQUESTS * _NUM_GENERATION_TOKENS_PER_REQUEST),
("_count", _NUM_REQUESTS)],
"vllm:prompt_tokens": [("_total",
_NUM_REQUESTS * _NUM_PROMPT_TOKENS_PER_REQUEST)],
"vllm:generation_tokens":
[("_total", _NUM_REQUESTS * _NUM_PROMPT_TOKENS_PER_REQUEST)],
"vllm:generation_tokens": [
("_total", _NUM_REQUESTS * _NUM_PROMPT_TOKENS_PER_REQUEST)
],
"vllm:request_success": [("_total", _NUM_REQUESTS)],
}
@@ -149,6 +153,9 @@ EXPECTED_METRICS = [
"vllm:request_params_n_sum",
"vllm:request_params_n_bucket",
"vllm:request_params_n_count",
"vllm:request_params_max_tokens_sum",
"vllm:request_params_max_tokens_bucket",
"vllm:request_params_max_tokens_count",
"vllm:num_preemptions_total",
"vllm:prompt_tokens_total",
"vllm:generation_tokens_total",