Merge pull request #2 from e-Learning-by-SSE/feat/docker_image

Feat/docker image

Author: Odrec

Dockerfile

@@ -0,0 +1,18 @@
+FROM nvidia/cuda:12.2.2-cudnn8-runtime-ubuntu22.04
+
+# Install required packages
+ENV DEBIAN_FRONTEND=noninteractive
+RUN apt update && apt install -y python3 python3-pip python3-venv python3-dev ffmpeg sudo python-is-python3 rabbitmq-server
+
+# Copy project into image
+WORKDIR /app
+COPY . .
+
+# Install python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Add entrypoint script
+COPY docker-entrypoint.sh /entrypoint.sh
+RUN chmod +x /entrypoint.sh
+
+ENTRYPOINT ["/entrypoint.sh"]

docker-entrypoint.sh

@@ -0,0 +1,6 @@
+#!/bin/bash
+# Starte RabbitMQ im Hintergrund
+sudo service rabbitmq-server start
+
+# Starte die Python-Anwendung
+exec python start.py

src/api/tasks.py

@@ -4,11 +4,16 @@
 from celery import Celery
 from src.utils.file_utils import convert_to_mp3, read_output_files
 from src.utils.transcription_utils import run_whisperx
+from celery.signals import setup_logging
 
 celery_app = Celery(
     "whisperx-tasks", backend="db+sqlite:///celery.db", broker=BROKER_URL
 )
 
+@setup_logging.connect
+def configure_celery_logging(**kwargs):
+    # Suppress task success logging
+    logging.getLogger("celery.app.trace").setLevel(logging.WARNING)
 
 @celery_app.task(name="transcribe_file")
 def transcribe_file(temp_video_path, lang, model, min_speakers, max_speakers, prompt):

src/utils/transcription_utils.py

@@ -15,7 +15,8 @@ def run_whisperx(temp_mp3_path, lang, model, min_speakers, max_speakers, prompt)
         "--language", lang,
         "--output_format", "all",
         "--output_dir", output_dir,
-        "--align_model", "WAV2VEC2_ASR_LARGE_LV60K_960H"
+        "--align_model", "WAV2VEC2_ASR_LARGE_LV60K_960H",
+        "--verbose", "False"
     ]
 
     # Include the prompt if provided