deploy local retrieval models for Hindsight

2026-09-18 00:29:33 -07:00
parent 26d56f7565
commit 2f8bf0e944
4 changed files with 132 additions and 9 deletions
@@ -0,0 +1,97 @@
+services:
+  embedding:
+    image: ghcr.io/huggingface/text-embeddings-inference@sha256:a7d82dfef16c3bf1a95e93f5b226f358312512dbb0d585b48c3cf886f9d470a9
+    pull_policy: never
+    container_name: hindsight-embedding-local
+    restart: unless-stopped
+    command:
+      - --model-id
+      - BAAI/bge-m3
+      - --dtype
+      - float16
+      - --max-input-length
+      - "8192"
+      - --max-client-batch-size
+      - "32"
+      - --max-concurrent-requests
+      - "64"
+      - --max-batch-tokens
+      - "8192"
+      - --tokenization-workers
+      - "4"
+      - --auto-truncate
+      - --served-model-name
+      - BAAI/bge-m3
+    environment:
+      HF_HUB_CACHE: /data
+      TZ: America/Los_Angeles
+    ports:
+      - "192.168.50.100:18083:80"
+    volumes:
+      - /mnt/cache-appdata/ai-models/hindsight/models/embedding:/data
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              device_ids: ["0"]
+              capabilities: [gpu]
+    healthcheck:
+      test: ["CMD", "curl", "-fsS", "http://127.0.0.1:80/health"]
+      interval: 30s
+      timeout: 10s
+      start_period: 120s
+      retries: 6
+    labels:
+      com.getarcaneapp.arcane.icon: huggingface
+      net.unraid.docker.icon: https://github.com/huggingface.png?size=128
+      net.unraid.docker.webui: http://[IP]:[PORT:18083]/docs
+
+  reranker:
+    image: ghcr.io/huggingface/text-embeddings-inference@sha256:a7d82dfef16c3bf1a95e93f5b226f358312512dbb0d585b48c3cf886f9d470a9
+    pull_policy: never
+    container_name: hindsight-reranker-local
+    restart: unless-stopped
+    command:
+      - --model-id
+      - Alibaba-NLP/gte-multilingual-reranker-base
+      - --dtype
+      - float16
+      - --max-input-length
+      - "8192"
+      - --max-client-batch-size
+      - "32"
+      - --max-concurrent-requests
+      - "32"
+      - --max-batch-tokens
+      - "8192"
+      - --tokenization-workers
+      - "4"
+      - --auto-truncate
+      - --served-model-name
+      - Alibaba-NLP/gte-multilingual-reranker-base
+    environment:
+      HF_HUB_CACHE: /data
+      TZ: America/Los_Angeles
+    ports:
+      - "192.168.50.100:18084:80"
+    volumes:
+      - /mnt/cache-appdata/ai-models/hindsight/models/gte-reranker:/data
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              device_ids: ["0"]
+              capabilities: [gpu]
+    healthcheck:
+      test: ["CMD", "curl", "-fsS", "http://127.0.0.1:80/health"]
+      interval: 30s
+      timeout: 10s
+      start_period: 180s
+      retries: 6
+    labels:
+      com.getarcaneapp.arcane.icon: huggingface
+      net.unraid.docker.icon: https://github.com/huggingface.png?size=128
+      net.unraid.docker.webui: http://[IP]:[PORT:18084]/docs
+