feat: added dolphin models and changed logging to warning

docker/compose/docker-compose.deepseek-14b-gpu.yml

@@ -23,8 +23,9 @@ services:
     command: >
       --model deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
       --gpu-memory-utilization 0.4
-      --max-model-len 10000 
+      --max-model-len 10000
       --tensor-parallel-size 1
+      --uvicorn-log-level WARNING
     environment:
       - SVC_HOST=deepseek_14b_gpu
       - SVC_PORT=8000
@@ -36,7 +37,7 @@ services:
     networks:
       - backend_net
     healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:8000/health"] 
+      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
       interval: 30s
       retries: 3
       start_period: 60s

docker/compose/docker-compose.dolphin-8b-gpu.yml

@@ -0,0 +1,51 @@
+services:
+  dolphin_8b_gpu:
+    build:
+      context: .
+      dockerfile: docker/vllm.Dockerfile
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: all
+              capabilities: [gpu]
+    ipc: host
+    ulimits:
+      memlock: -1
+      stack: 67108864
+    env_file:
+      - .env
+    restart: unless-stopped
+    depends_on:
+      etcd:
+        condition: service_healthy
+    command: >
+      --model cognitivecomputations/Dolphin3.0-Llama3.1-8B
+      --gpu-memory-utilization 0.5
+      --max-model-len 10000
+      --tensor-parallel-size 1
+      --enable-auto-tool-choice
+      --tool-call-parser llama3_json
+      --uvicorn-log-level WARNING
+    environment:
+      - SVC_HOST=dolphin_8b_gpu
+      - SVC_PORT=8000
+      - ETCD_HOST=etcd
+      - ETCD_PORT=2379
+      - TOOL_SUPPORT=true
+    volumes:
+      - hugging_face_models:/root/.cache/huggingface  # cache models
+    networks:
+      - backend_net
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
+      interval: 30s
+      retries: 3
+      start_period: 60s
+      timeout: 10s
+volumes:
+  hugging_face_models:
+
+networks:
+  backend_net:

docker/compose/docker-compose.llama-3b-gpu.yml

@@ -23,10 +23,11 @@ services:
     command: >
       --model meta-llama/Llama-3.2-3B-Instruct
       --gpu-memory-utilization 0.3
-      --max-model-len 10000 
+      --max-model-len 10000
       --tensor-parallel-size 1
       --enable-auto-tool-choice
       --tool-call-parser llama3_json
+      --uvicorn-log-level WARNING
     environment:
       - SVC_HOST=llama_3b_gpu
       - SVC_PORT=8000
@@ -47,4 +48,4 @@ volumes:
   hugging_face_models:
 
 networks:
-  backend_net:
+  backend_net:

docker/compose/docker-compose.llama-8b-gpu.yml

@@ -22,11 +22,12 @@ services:
         condition: service_healthy
     command: >
       --model meta-llama/Llama-3.1-8B-Instruct
-      --gpu-memory-utilization 0.5 
-      --max-model-len 10000 
+      --gpu-memory-utilization 0.5
+      --max-model-len 10000
       --tensor-parallel-size 1
       --enable-auto-tool-choice
       --tool-call-parser llama3_json
+      --uvicorn-log-level WARNING
     environment:
       - SVC_HOST=llama_8b_gpu
       - SVC_PORT=8000
@@ -47,4 +48,4 @@ volumes:
   hugging_face_models:
 
 networks:
-  backend_net:
+  backend_net: