Merge branch 'develop' into SNOW-172-feat#51/Swagger

.github/workflows/ai.yml

@@ -0,0 +1,37 @@
+name: ai
+
+on:
+  workflow_dispatch:
+
+  push:
+    branches:
+      - develop
+    paths:
+      - "AI/**"
+
+permissions:
+  contents: read
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        working-directory: ./AI/roop
+
+    steps:
+      - uses: actions/checkout@v3
+
+      - name: Log in to Docker Hub
+        uses: docker/login-action@f054a8b539a109f9f41c372932f1ae047eff08c9
+        with:
+          username: ${{ secrets.DOCKER_ID }}
+          password: ${{ secrets.DOCKER_PASSWORD }}
+
+      - name: Run scripts in server
+        uses: appleboy/ssh-action@master
+        with:
+          key: ${{ secrets.KEY }}
+          host: ${{ secrets.AI_HOST }}
+          username: ${{ secrets.USER_NAME }}
+          script: ${{ secrets.AI_SCRIPT }}

.github/workflows/server.yml

@@ -28,7 +28,7 @@ jobs:
           distribution: temurin
           java-version: 17
 
-      - run : echo "${{secrets.DEV_YML}}" > ./src/main/resources/application-dev.yml
+      - run : echo "${{ secrets.DEV_YML }}" > ./src/main/resources/application-dev.yml
 
       - name: Log in to Docker Hub
         uses: docker/login-action@f054a8b539a109f9f41c372932f1ae047eff08c9

AI/roop/.gitignore

@@ -7,3 +7,6 @@ d_id/conf.yaml
 shotstack/conf.yaml
 s3/conf.yaml
 elevenlabs/conf.yaml
+
+# 환경변수 파일
+*.env

AI/roop/Dockerfile

@@ -1,16 +1,20 @@
-FROM python:3.10
+FROM python:3.10 AS builder
 
 ENV PYTHONUNBUFFERED 1
+ARG CACHEBUST=1
 WORKDIR /app
-COPY . ./
-
-VOLUME ./image:./image
+COPY requirements.* ./
 
 RUN pip install --upgrade pip
+RUN pip install --no-cache-dir -r ./requirements.txt
+
+COPY . ./
+
+FROM python:3.10-slim AS deployer
 
-RUN apt-get update
-RUN apt-get -y install ffmpeg
+COPY --from=builder /usr/local/lib/python3.10/site-packages /usr/local/lib/python3.10/site-packages
+COPY --from=builder /app /app
 
-RUN pip install -r ./requirements.txt
+WORKDIR /app
 
-CMD ["python3", "app.py"]
+CMD ["python3", "app.py"]

AI/roop/Dockerfile-celery

@@ -0,0 +1,25 @@
+FROM python:3.10 AS builder
+
+ENV PYTHONUNBUFFERED 1
+ARG CACHEBUST=1
+WORKDIR /app
+COPY requirements* ./
+
+RUN apt-get update && apt-get -y install ffmpeg
+
+RUN pip install --upgrade pip
+RUN pip install --no-cache-dir -r ./requirements-celery.txt
+
+COPY . ./
+
+FROM python:3.10-slim AS deployer
+
+COPY --from=builder /usr/local/lib/python3.10/site-packages /usr/local/lib/python3.10/site-packages
+COPY --from=builder /app /app
+
+RUN apt-get update && apt-get -y install ffmpeg
+RUN pip install --no-cache-dir celery
+
+WORKDIR /app
+
+CMD ["celery", "-A", "tasks", "worker", "--loglevel=info"]

AI/roop/app.py

@@ -1,172 +1,28 @@
-import subprocess
-from urllib.parse import urlparse
+from flask import Flask, request, Response
 
-from flask import Flask, request, jsonify
-from d_id.did_reqeust import DIdAPI
-import hashlib
-import time
-
-from elevenlabs import elevenlabs_request
-from shotstack.shot_stack import ShotStackAPI
-from s3 import s3_request
+from tasks import generate_meme
 
 app = Flask(__name__)
-s3 = s3_request.s3_connection()
 
 
-@app.route('/files', methods=['GET', 'POST'])
+@app.route('/files', methods=['POST'])
 def process_data():
-    try:
-        did = DIdAPI()
-        shotstack = ShotStackAPI()
-        # JSON 데이터를 파싱하여 Python 객체로 변환
-        data = request.get_json()
-        talk_id_queue = []
-        scene = data['scene']
-        member_id = data.get('member_id')
-        voice_id = data.get('voice_id')
-
-        callback_url = 'http://ec2-43-202-91-197.ap-northeast-2.compute.amazonaws.com/meme/create/' + member_id
-
-        # 여기에서 데이터를 원하는 대로 처리
-        for item in scene:
-            text_data = item.get('text')
-            target_image_number = item.get('target_image')
-            source_image_s3_url = item.get('source_image')
-
-            # s3_url에서 파일이름 추출
-            file_name = get_file_name_from_url(source_image_s3_url)
-
-            # s3_url (이미지를 파일로 저장하는 경우)
-            if file_name:
-                source_image = s3_request.s3_get_object(s3, file_name)
-
-                # 현재 시간을 기반으로 한 해시값 생성
-                hash_input = str(time.time())
-                unique_hash = hashlib.sha256(hash_input.encode()).hexdigest()[:4]
-
-                # 파일을 저장할 경로 및 파일 이름 생성
-                file_path = f"./image/{unique_hash}.jpg"
-
-                # BytesIO 객체에 저장된 데이터를 파일로 저장
-                if save_bytes_io_to_file(source_image, file_path):
-                    print(f"파일 저장 성공: {file_path}")
-                else:
-                    print(f"파일 저장 실패: {file_path}")
-
-            face_swap_image_name = face_swap(target_image_number, unique_hash)
-            talk_id = did.run(face_swap_image_name, text_data, voice_id)
-            talk_id_queue.append(talk_id)
-            print(talk_id)
-
-        print(talk_id_queue)
-        result_url_queue = []
-
-        for item in talk_id_queue:
-            result_url_queue.append(did.download_scene(item))
-
-        print(result_url_queue)
-
-        # 초기 타임라인 데이터 객체 -> flask 서버에다 복사해두고, 함수 시작할 때 실행
-        timeline_data = {
-            "timeline": {
-                "tracks": []
-            },
-            "output": {
-                "format": "mp4",
-                "resolution": "sd"
-            },
-            "callback": callback_url
-        }
-
-        start_time = 0
-
-        while result_url_queue:
-            item = result_url_queue.pop(0)
-            duration = int(item[1])
-            print("duration : " + str(duration))
-            shotstack.add_track(timeline_data["timeline"]["tracks"], "video", item[0], start_time, duration)
-            start_time = start_time + duration
-
-        print(timeline_data)
-
-        shotstack_id = shotstack.send_timeline_data(timeline_data)
-
-        # 처리 결과 응답
-        return jsonify({'message': 'Data processed successfully'})
-
-    except Exception as e:
-        # 예외 처리
-        print('Error:', str(e))
-        return jsonify({'error': 'An error occurred during data processing'})
-
-
-@app.route("/voice", methods=['POST'])
-def add_voice():
-    voice_file = request.files['voice_file']
-    name = request.form['name']
-    description = request.form['description']
-    file_name = voice_file.filename
-
-    save_bytes_io_to_file(voice_file, f"elevenlabs/voice/{file_name}")
-
-    voice_id = elevenlabs_request.add_voice(name, description, file_name)
-    return jsonify({"message": "Voice added successfully", "data": voice_id}), 200
-
-
-def face_swap(target_image_number, source_image):
-    command = [
-        "python",
-        "./run.py",
-        "-s",
-        "./image/" + source_image + '.jpg',
-        "-t",
-        './target_image/' + target_image_number + '.jpg',
-        "-o",
-        "./image/faceswap_image/",
-        "--frame-processor",
-        "face_swapper"
-    ]
-
-    subprocess.run(command)
-
-    face_swap_image_name = source_image + "-" + target_image_number + '.jpg'
-
-    return face_swap_image_name
-
-
-def save_bytes_io_to_file(bytes_io, file_path):
-    """
-    BytesIO 객체에 저장된 데이터를 파일로 저장
-
-    :param bytes_io: BytesIO 객체
-    :param file_path: 저장할 파일 경로
-    :return: 저장에 성공하면 True, 실패하면 False를 반환
-    """
-    try:
-        with open(file_path, "wb") as file:
-            file.write(bytes_io.getvalue())
-        return True
-    except Exception as e:
-        print(e)
-        return False
-
-
-def get_file_name_from_url(url):
-    """
-    URL에서 파일 이름을 추출
-
-    :param url: 파일이 위치한 URL
-    :return: 추출된 파일 이름 또는 None (URL에서 파일 이름을 추출할 수 없는 경우)
-    """
-    try:
-        parsed_url = urlparse(url)
-        # URL 경로를 '/'로 분할하여 파일 이름을 추출
-        file_name = parsed_url.path.split('/')[-1]
-        return file_name
-    except Exception as e:
-        print(e)
-        return None
+    data = request.get_json()
+    generate_meme.delay(data)
+    return Response(status=201)
+
+
+# @app.route("/voice", methods=['POST'])
+# def add_voice():
+#     voice_file = request.files['voice_file']
+#     name = request.form['name']
+#     description = request.form['description']
+#     file_name = voice_file.filename
+#
+#     save_bytes_io_to_file(voice_file, f"elevenlabs/voice/{file_name}")
+#
+#     voice_id = elevenlabs.add_voice(name, description, file_name)
+#     return jsonify({"message": "Voice added successfully", "data": voice_id}), 200
 
 
 if __name__ == '__main__':

AI/roop/config/config.py

@@ -0,0 +1,36 @@
+import os
+from dotenv import load_dotenv
+
+BASE_DIR = os.path.dirname(os.path.dirname(__file__))
+
+
+class Config:
+    _instance = None
+
+    DEBUG = False
+    DID_API_KEY = None
+    ELEVENLABS_API_KEY = None
+    SHOT_STACK_API_KEY = None
+    AWS_ACCESS_KEY_ID = None
+    AWS_SECRET_ACCESS_KEY = None
+    AWS_S3_BUCKET_REGION = None
+    AWS_S3_BUCKET_NAME = None
+
+    def __new__(cls):
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+
+            load_dotenv()
+
+            cls._instance.DID_API_KEY = os.environ.get("D-ID_API_KEY")
+            cls._instance.ELEVENLABS_API_KEY = os.environ.get("ELEVENLABS_API_KEY")
+            cls._instance.SHOT_STACK_API_KEY = os.environ.get("SHOT_STACK_API_KEY")
+            cls._instance.AWS_ACCESS_KEY_ID = os.environ.get("AWS_ACCESS_KEY_ID")
+            cls._instance.AWS_SECRET_ACCESS_KEY = os.environ.get("AWS_SECRET_ACCESS_KEY")
+            cls._instance.AWS_S3_BUCKET_REGION = os.environ.get("AWS_S3_BUCKET_REGION")
+            cls._instance.AWS_S3_BUCKET_NAME = os.environ.get("AWS_S3_BUCKET_NAME")
+        return cls._instance
+
+# 사용 예시:
+# config = Config()
+# print(config.ID_API_KEY)