apairon
3bf8055539
continuous-integration/drone/push Build is passing
Damit übt npm test den (...)+-Zweig des duration-Musters aus statt ihn nur zu kommentieren (Review-Nachbefund, low).
50 lines
1.3 KiB
YAML
50 lines
1.3 KiB
YAML
# Beispiel für den embedding-Abschnitt der Server-config.yml
|
|
# (tibi-server, server.embedding) — validiert gegen schemas/config/embedding.schema.json.
|
|
providers:
|
|
- name: bge-m3-external
|
|
type: flagembedding
|
|
model: BAAI/bge-m3
|
|
dimensions: 1024
|
|
baseURL: https://embed.intern/embed
|
|
timeoutPerAttempt: 8s
|
|
slowThreshold: 1500ms
|
|
- name: bge-m3-local
|
|
type: native
|
|
model: BAAI/bge-m3
|
|
dimensions: 1024
|
|
modelPath: /models/bge-m3
|
|
timeoutPerAttempt: 60s
|
|
groups:
|
|
- name: bge-m3-main
|
|
description: Bevorzugt extern, sonst lokal, kehrt automatisch zurück.
|
|
members:
|
|
- bge-m3-external
|
|
- bge-m3-local
|
|
onSingleQuery:
|
|
policy: prefer-in-order
|
|
preferredOrder:
|
|
- bge-m3-external
|
|
- bge-m3-local
|
|
timeoutPerAttempt: 8s
|
|
timeoutTotal: 1m30s # Mehrfach-Einheiten sind erlaubt (Go time.ParseDuration)
|
|
|
|
autoFailback: true
|
|
retryAttempts: 1
|
|
retryBackoff: 200ms
|
|
retryBackoffMax: 2s
|
|
latencyWindow: "100"
|
|
|
|
onBulkWrite:
|
|
distribution: lowest-latency
|
|
maxConcurrentTotal: 6
|
|
maxConcurrentPerEndpoint: 2
|
|
timeoutPerAttempt: 15s
|
|
timeoutTotal: 120s
|
|
health:
|
|
checkInterval: 30s
|
|
checkTimeout: 3s
|
|
unhealthyAfterFailures: 3
|
|
healthyAfterSuccesses: 2
|
|
cooldownAfterUnhealthy: 60s
|
|
goldenPath: testdata/embedding-golden
|