Skip to content

Commit 351bb53

Browse files
committed
Add Reinforcement Learning integration
1 parent 527e5cd commit 351bb53

4 files changed

Lines changed: 46 additions & 0 deletions

File tree

dashboards/reinforcement-learning/metadata.yaml

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -5,3 +5,6 @@ sample_dashboards:
55
display_name: "Reinforcement Learning Workload Performance Dashboard"
66
description: |-
77
This dashboard provides metrics for generic Reinforcement Learning training workloads, including loop duration, reward, loss, and throughput.
8+
related_integrations:
9+
- id: rl
10+
platform: GKE

integrations/rl/documentation.yaml

Lines changed: 11 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
exporter_type: included
2+
app_name_short: RL
3+
app_name: Reinforcement Learning
4+
app_site_name: Reinforcement Learning
5+
app_site_url: https://opentelemetry.io/docs/specs/semconv/
6+
exporter_name: the OpenTelemetry SDK
7+
exporter_repo_url: https://opentelemetry.io/docs/specs/semconv/
8+
dashboard_available: true
9+
multiple_dashboards: false
10+
dashboard_display_name: Reinforcement Learning Workload Performance Dashboard
11+
sample_promql_query: up{job="rl"}

integrations/rl/metadata.yaml

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
id: rl
2+
short_name: Reinforcement Learning
3+
display_name: Reinforcement Learning
4+
description: |
5+
This integration collects metrics for Reinforcement Learning (RL) training workloads using OpenTelemetry (OTLP) with semantic conventions.
Lines changed: 27 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,27 @@
1+
platforms:
2+
- type: GKE
3+
launch_stage: GA
4+
version: 1
5+
detections:
6+
- characteristic_metric:
7+
metric_type: prometheus.googleapis.com/rl.loop.duration/histogram
8+
exporter_metadata:
9+
name: OpenTelemetry
10+
doc_url: https://opentelemetry.io/docs/specs/semconv/
11+
default_metrics:
12+
- name: prometheus.googleapis.com/rl.loop.duration/histogram
13+
prometheus_name: rl.loop.duration
14+
kind: CUMULATIVE
15+
value_type: DISTRIBUTION
16+
- name: prometheus.googleapis.com/rl.environment.reward.mean/gauge
17+
prometheus_name: rl.environment.reward.mean
18+
kind: GAUGE
19+
value_type: DOUBLE
20+
- name: prometheus.googleapis.com/rl.train.loss/gauge
21+
prometheus_name: rl.train.loss
22+
kind: GAUGE
23+
value_type: DOUBLE
24+
- name: prometheus.googleapis.com/rl.train.mfu/histogram
25+
prometheus_name: rl.train.mfu
26+
kind: CUMULATIVE
27+
value_type: DISTRIBUTION

0 commit comments

Comments
 (0)