Merge pull request #22 from janhq/inf_sd_s3

Inf sd s3

.gitignore

@@ -3,9 +3,10 @@
 
 # Jan inference
 jan-inference/llm/models/**
-jan-inference/llm/.env
 
 jan-inference/sd/models/**
 jan-inference/sd/output/**
-jan-inference/sd/.env
 jan-inference/sd/sd
+
+# Minio
+minio/**

README.md

@@ -29,7 +29,8 @@ Jan is a free, source-available and [fair code licensed](https://faircode.io/) A
   cd jan
 
   # Pull latest submodule
-  git submodule update --init
+  git submodule update --init --recursive
+
   ```
 
 - **Environment Variables**: You will need to set up several environment variables for services such as Keycloak and Postgres. You can place them in `.env` files in the respective folders as shown in the `docker-compose.yml`.

docker-compose.yml

@@ -12,13 +12,13 @@ services:
       KC_DB_PASSWORD: ${POSTGRES_PASSWORD:-postgres}
       KC_DB_USERNAME: ${POSTGRES_USERNAME:-postgres}
       KC_DB_SCHEMA: ${KC_DB_SCHEMA:-public}
-      KC_HEALTH_ENABLED: 'true'
+      KC_HEALTH_ENABLED: "true"
       KEYCLOAK_ADMIN: ${KEYCLOAK_ADMIN-admin}
       KEYCLOAK_ADMIN_PASSWORD: ${KEYCLOAK_ADMIN_PASSWORD-admin}
     volumes:
       - ./conf/keycloak_conf:/opt/keycloak/data/import
     ports:
-    - "8088:8088"
+      - "8088:8088"
     depends_on:
       keycloak_postgres:
         condition: service_healthy
@@ -58,7 +58,7 @@ services:
   graphql-engine:
     image: hasura/graphql-engine:v2.31.0.cli-migrations-v3
     ports:
-    - "8080:8080"
+      - "8080:8080"
     restart: always
     env_file:
       - conf/sample.env_app-backend
@@ -133,7 +133,6 @@ services:
     build:
       context: ./web-client
       dockerfile: ./dev.Dockerfile
-    container_name: jan_web
     restart: always
     volumes:
       - ./web-client/:/app
@@ -191,14 +190,17 @@ services:
     # Mount the directory that contains the downloaded model.
     volumes:
       - ./jan-inference/sd/models:/models/
-      - ./jan-inference/sd/output/:/output/
-    command:  /bin/bash -c "python -m uvicorn main:app --proxy-headers --host 0.0.0.0 --port 8000"
+    command: /bin/bash -c "python -m uvicorn main:app --proxy-headers --host 0.0.0.0 --port 8000"
     environment:
       # Specify the path to the model for the web application.
-      BASE_URL: http://0.0.0.0:8001
+      S3_ENDPOINT_URL: ${S3_ENDPOINT_URL}
+      S3_PUBLIC_ENDPOINT_URL: ${S3_PUBLIC_ENDPOINT_URL}
+      S3_ACCESS_KEY_ID: ${S3_ACCESS_KEY_ID}
+      S3_SECRET_ACCESS_KEY: ${S3_SECRET_ACCESS_KEY}
+      S3_BUCKET_NAME: ${S3_BUCKET_NAME}
       MODEL_NAME: ${SD_MODEL_FILE}.q4_0.bin
       MODEL_DIR: /models
-      OUTPUT_DIR: /output
+      OUTPUT_DIR: /tmp
       SD_PATH: /sd
       PYTHONUNBUFFERED: 1
     ports:
@@ -213,6 +215,41 @@ services:
       jan_community:
         ipv4_address: 172.20.0.21
 
+  minio:
+    image: minio/minio
+    ports:
+      - 9000:9000
+      - 9001:9001
+    volumes:
+      - ./minio/data:/export
+      - ./minio/config:/root/.minio
+    environment:
+      MINIO_ROOT_USER: ${S3_ACCESS_KEY_ID}
+      MINIO_ROOT_PASSWORD: ${S3_SECRET_ACCESS_KEY}
+    command: server /export --console-address ":9001"
+    networks:
+      jan_community:
+        ipv4_address: 172.20.0.23
+
+  createbuckets:
+    image: minio/mc
+    depends_on:
+      - minio
+    environment:
+      S3_ACCESS_KEY_ID: ${S3_ACCESS_KEY_ID}
+      S3_SECRET_ACCESS_KEY: ${S3_SECRET_ACCESS_KEY}
+      S3_BUCKET_NAME: ${S3_BUCKET_NAME}
+    entrypoint: >
+      /bin/sh -c "
+      /usr/bin/mc config host add myminio http://minio:9000 ${S3_ACCESS_KEY_ID} ${S3_SECRET_ACCESS_KEY};
+      /usr/bin/mc rm -r --force myminio/${S3_BUCKET_NAME};
+      /usr/bin/mc mb myminio/${S3_BUCKET_NAME};
+      /usr/bin/mc anonymous set public myminio/${S3_BUCKET_NAME};
+      exit 0;
+      "
+    networks:
+      jan_community:
+
   # Service for Traefik, a modern HTTP reverse proxy and load balancer.
   # traefik:
   #   image: traefik:v2.10
@@ -230,10 +267,10 @@ services:
   #     jan_community:
   #       ipv4_address: 172.20.0.22
 
-networks:  
+networks:
   jan_community:
     driver: bridge
     ipam:
       driver: default
       config:
-        - subnet: 172.20.0.0/16
+        - subnet: 172.20.0.0/16

jan-inference/sd/inference.requirements.txt

@@ -1,4 +1,5 @@
 # Inference
 fastapi
 uvicorn
-python-multipart
+python-multipart
+boto3

jan-inference/sd/main.py

@@ -5,6 +5,8 @@ import subprocess
 import os
 from uuid import uuid4
 from pydantic import BaseModel
+import boto3
+from botocore.client import Config
 
 app = FastAPI()
 
@@ -12,8 +14,23 @@ OUTPUT_DIR = os.environ.get("OUTPUT_DIR", "output")
 SD_PATH = os.environ.get("SD_PATH", "./sd")
 MODEL_DIR = os.environ.get("MODEL_DIR", "./models")
 MODEL_NAME = os.environ.get(
-    "MODEL_NAME", "v1-5-pruned-emaonly-ggml-model-q5_0.bin")
-BASE_URL = os.environ.get("BASE_URL", "http://localhost:8000")
+    "MODEL_NAME", "v1-5-pruned-emaonly.safetensors.q4_0.bin")
+
+S3_ENDPOINT_URL = os.environ.get("S3_ENDPOINT_URL", "http://localhost:9000")
+S3_PUBLIC_ENDPOINT_URL = os.environ.get(
+    "S3_PUBLIC_ENDPOINT_URL", "http://localhost:9000")
+S3_ACCESS_KEY_ID = os.environ.get("S3_ACCESS_KEY_ID", "minio")
+S3_SECRET_ACCESS_KEY = os.environ.get("S3_SECRET_ACCESS_KEY", "minio123")
+S3_BUCKET_NAME = os.environ.get("S3_BUCKET_NAME", "jan")
+
+s3 = boto3.resource('s3',
+                    endpoint_url=S3_ENDPOINT_URL,
+                    aws_access_key_id=S3_ACCESS_KEY_ID,
+                    aws_secret_access_key=S3_SECRET_ACCESS_KEY,
+                    config=Config(signature_version='s3v4'),
+                    region_name='us-east-1')
+
+s3_bucket = s3.Bucket(S3_BUCKET_NAME)
 
 
 class Payload(BaseModel):
@@ -33,9 +50,6 @@ if not os.path.exists(OUTPUT_DIR):
 if not os.path.exists(MODEL_DIR):
     os.makedirs(MODEL_DIR)
 
-# Serve files from the "files" directory
-app.mount("/output", StaticFiles(directory=OUTPUT_DIR), name="output")
-
 
 def run_command(payload: Payload, filename: str):
     # Construct the command based on your provided example
@@ -66,21 +80,11 @@ async def run_inference(background_tasks: BackgroundTasks, payload: Payload):
     # We will use background task to run the command so it won't block
     # background_tasks.add_task(run_command, payload, filename)
     run_command(payload, filename)
-
+    s3_bucket.upload_file(f'{os.path.join(OUTPUT_DIR, filename)}', filename)
     # Return the expected path of the output file
-    return {"url": f'{BASE_URL}/serve/{filename}'}
-
-
-@app.get("/serve/{filename}")
-async def serve_file(filename: str):
-    file_path = os.path.join(OUTPUT_DIR, filename)
-
-    if os.path.exists(file_path):
-        return FileResponse(file_path)
-    else:
-        raise HTTPException(status_code=404, detail="File not found")
+    return {"url": f'{S3_PUBLIC_ENDPOINT_URL}/{S3_BUCKET_NAME}/{filename}'}
 
 
 if __name__ == "__main__":
     import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=8000)
+    uvicorn.run(app, host="0.0.0.0", port=8002)

sample.env

@@ -15,3 +15,10 @@ LLM_MODEL_FILE=llama-2-7b-chat.ggmlv3.q4_1.bin
 ## SD
 SD_MODEL_URL=https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors
 SD_MODEL_FILE=v1-5-pruned-emaonly.safetensors
+
+# Minio
+S3_ACCESS_KEY_ID=minio
+S3_SECRET_ACCESS_KEY=minio123
+S3_BUCKET_NAME=jan
+S3_ENDPOINT_URL=http://minio:9000
+S3_PUBLIC_ENDPOINT_URL=http://127.0.0.1:9000