phoenix/litellm
1
0
Fork 0
forked from phoenix/litellm-mirror
Code Pull requests Activity

feat(prometheus.py): track team based metrics on prometheus

Browse source
This commit is contained in:
Krrish Dholakia 2024-04-03 13:43:21 -07:00
parent cbe4aa386b
commit fcaa452ccd
2 changed files with 18 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

25
litellm/integrations/prometheus.py
View file

@ -25,21 +25,21 @@ class PrometheusLogger:
self.litellm_requests_metric = Counter( self.litellm_requests_metric = Counter(
name="litellm_requests_metric", name="litellm_requests_metric",
documentation="Total number of LLM calls to litellm", documentation="Total number of LLM calls to litellm",
labelnames=["user", "key", "model"], labelnames=["end_user", "key", "model", "team"],
) )
# Counter for spend # Counter for spend
self.litellm_spend_metric = Counter( self.litellm_spend_metric = Counter(
"litellm_spend_metric", "litellm_spend_metric",
"Total spend on LLM requests", "Total spend on LLM requests",
labelnames=["user", "key", "model"], labelnames=["end_user", "key", "model", "team"],
) )
# Counter for total_output_tokens # Counter for total_output_tokens
self.litellm_tokens_metric = Counter( self.litellm_tokens_metric = Counter(
"litellm_total_tokens", "litellm_total_tokens",
"Total number of input + output tokens from LLM requests", "Total number of input + output tokens from LLM requests",
labelnames=["user", "key", "model"], labelnames=["end_user", "key", "model", "team"],
) )
except Exception as e: except Exception as e:
print_verbose(f"Got exception on init prometheus client {str(e)}") print_verbose(f"Got exception on init prometheus client {str(e)}")
@ -66,19 +66,24 @@ class PrometheusLogger:
proxy_server_request = litellm_params.get("proxy_server_request") or {} proxy_server_request = litellm_params.get("proxy_server_request") or {}
end_user_id = proxy_server_request.get("body", {}).get("user", None) end_user_id = proxy_server_request.get("body", {}).get("user", None)
user_api_key = litellm_params.get("metadata", {}).get("user_api_key", None) user_api_key = litellm_params.get("metadata", {}).get("user_api_key", None)
user_api_team = litellm_params.get("metadata", {}).get(
"user_api_key_team_id", None
)
tokens_used = response_obj.get("usage", {}).get("total_tokens", 0) tokens_used = response_obj.get("usage", {}).get("total_tokens", 0)
print_verbose( print_verbose(
f"inside track_prometheus_metrics, model {model}, response_cost {response_cost}, tokens_used {tokens_used}, end_user_id {end_user_id}, user_api_key {user_api_key}" f"inside track_prometheus_metrics, model {model}, response_cost {response_cost}, tokens_used {tokens_used}, end_user_id {end_user_id}, user_api_key {user_api_key}"
) )
self.litellm_requests_metric.labels(end_user_id, user_api_key, model).inc() self.litellm_requests_metric.labels(
self.litellm_spend_metric.labels(end_user_id, user_api_key, model).inc( end_user_id, user_api_key, model, user_api_team
response_cost ).inc()
) self.litellm_spend_metric.labels(
self.litellm_tokens_metric.labels(end_user_id, user_api_key, model).inc( end_user_id, user_api_key, model, user_api_team
tokens_used ).inc(response_cost)
) self.litellm_tokens_metric.labels(
end_user_id, user_api_key, model, user_api_team
).inc(tokens_used)
except Exception as e: except Exception as e:
traceback.print_exc() traceback.print_exc()
verbose_logger.debug( verbose_logger.debug(

3
litellm/proxy/_new_secret_config.yaml
View file

@ -10,6 +10,9 @@ model_list:
# api_key: my-fake-key # api_key: my-fake-key
# api_base: https://exampleopenaiendpoint-production.up.railway.app/ # api_base: https://exampleopenaiendpoint-production.up.railway.app/
litellm_settings:
success_callback: ["prometheus"]
# litellm_settings: # litellm_settings:
# drop_params: True # drop_params: True
# max_budget: 800021 # max_budget: 800021