[Doc][Metrics] Fix spec-decode PromQL examples to use the exposed names (#55016)
Signed-off-by: Michele Campi <michele.campi@outlook.com>
This commit is contained in:
@@ -187,12 +187,12 @@ class SpecDecodingProm:
|
|||||||
|
|
||||||
1 + (
|
1 + (
|
||||||
rate(vllm:spec_decode_num_accepted_tokens_total[$interval]) /
|
rate(vllm:spec_decode_num_accepted_tokens_total[$interval]) /
|
||||||
rate(vllm:spec_decode_num_drafts[$interval]))
|
rate(vllm:spec_decode_num_drafts_total[$interval]))
|
||||||
|
|
||||||
A per-position acceptance rate vector can be computed using
|
A per-position acceptance rate vector can be computed using
|
||||||
|
|
||||||
vllm:spec_decode_num_accepted_tokens_per_pos[$interval] /
|
vllm:spec_decode_num_accepted_tokens_per_pos_total[$interval] /
|
||||||
vllm:spec_decode_num_drafts[$interval]
|
vllm:spec_decode_num_drafts_total[$interval]
|
||||||
"""
|
"""
|
||||||
|
|
||||||
_counter_cls = prometheus_client.Counter
|
_counter_cls = prometheus_client.Counter
|
||||||
|
|||||||
Reference in New Issue
Block a user