Fixed litellm

templates/compose/litellm.yaml

@@ -4,13 +4,15 @@
 # logo: svgs/litellm.svg
 # port: 4000
 
+version: "3"
+
 volumes:
-  pg-data: null
+  pg_data: null
-  redis-data: null
+  redis_data: null
 
 services:
   litellm:
-    image: "ghcr.io/berriai/litellm-database:main-latest"
+    image: "ghcr.io/berriai/litellm-database:main-stable"
     restart: always
     depends_on:
       postgres:
@@ -18,9 +20,10 @@ services:
       redis:
         condition: service_healthy
     environment:
+      - SERVICE_FQDN_LITELLM_4000
       - LITELLM_LOG=ERROR
       - LITELLM_MODE=PRODUCTION
-      - "LITELLM_MASTER_KEY=${SERVICE_PASSWORD_32_LITELLM_MASTER_KEY}"
+      - "LITELLM_MASTER_KEY=${SERVICE_PASSWORD_MASTERKEY}"
       - "UI_USERNAME=${SERVICE_USER_UI}"
       - "UI_PASSWORD=${SERVICE_PASSWORD_UI}"
       - "DATABASE_URL=postgresql://${SERVICE_USER_POSTGRES}:${SERVICE_PASSWORD_POSTGRES}@postgres:5432/${POSTGRES_DB:-litellm}"
@@ -36,16 +39,14 @@ services:
       - "VOYAGE_API_KEY=${VOYAGE_API_KEY}"
       - "VOYAGE_API_BASE=${VOYAGE_API_BASE}"
     volumes:
-      - "./litellm-config.yaml:/app/config.yaml"
       - type: bind
-        source: ./docker/app/config.yaml
+        source: ./litellm-config.yaml
         target: /app/config.yaml
         content: |
           general_settings:
             proxy_batch_write_at: 60
 
           router_settings:
-            routing_strategy: usage-based-routing-v2
             redis_host: os.environ/REDIS_HOST
             redis_port: os.environ/REDIS_PORT
             redis_password: os.environ/REDIS_PASSWORD
@@ -57,6 +58,7 @@ services:
             log_raw_request_response: true
             # turn_off_message_logging: false
             # redact_user_api_key_info: false
+            service_callback: ["prometheus_system"]
             drop_params: true
             # max_budget: 100
             # budget_duration: 30d
@@ -71,19 +73,14 @@ services:
               password: os.environ/REDIS_PASSWORD
               namespace: "litellm_cache"
               ttl: 600
-            # success_callback:
+            success_callback:
               # - "langfuse"
-            #   - "prometheus"
+              - "prometheus"
-            # failure_callback:
+            failure_callback:
               # - "langfuse"
-            #   - "prometheus"
+              - "prometheus"
           model_list:
             # OpenAI
-            - model_name: gpt-3.5-turbo
-              litellm_params:
-                model: openai/gpt-3.5-turbo
-                api_key: os.environ/OPENAI_API_KEY
-                api_base: os.environ/OPENAI_API_BASE
             - model_name: gpt-4
               litellm_params:
                 model: openai/gpt-4
@@ -118,8 +115,8 @@ services:
                 model: voyage/voyage-law-2
                 api_key: "os.environ/VOYAGE_API_KEY"
                 api_base: "os.environ/VOYAGE_API_BASE"
-                rpm: 300
+                # rpm: 300
-                tpm: 1000000
+                # tpm: 1000000
             - model_name: voyage-multilingual-2
               model_info:
                 mode: embedding
@@ -132,8 +129,8 @@ services:
                 api_base: "os.environ/VOYAGE_API_BASE"
                 input_cost_per_token: 0.00000012
                 output_cost_per_token: 0
-                rpm: 300
+                # rpm: 300
-                tpm: 1000000
+                # tpm: 1000000
     healthcheck:
       test:
         - CMD
@@ -157,7 +154,7 @@ services:
       - POSTGRES_PASSWORD=$SERVICE_PASSWORD_POSTGRES
       - POSTGRES_USER=$SERVICE_USER_POSTGRES
     volumes:
-      - "pg-data:/var/lib/postgresql/data"
+      - "pg_data:/var/lib/postgresql/data"
     healthcheck:
       test:
         - CMD-SHELL
@@ -169,7 +166,7 @@ services:
     image: redis:7-alpine
     command: redis-server --appendonly yes
     volumes:
-      - redis-data:/data
+      - redis_data:/data
     healthcheck:
       test: ["CMD", "redis-cli", "ping"]
       interval: 5s