no need to download model

Dockerfile

@@ -6,15 +6,12 @@ RUN apt-get update && apt-get install -y \
     curl \
     && rm -rf /var/lib/apt/lists/*
 
-# Install Python dependencies
 WORKDIR /app
+
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
-# Copy app and download model
 COPY app ./app
-COPY download-model.sh .
-RUN chmod +x download-model.sh && ./download-model.sh
 
 EXPOSE 4002
 CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "4002"]

app/main.py

@@ -1,12 +1,12 @@
 from fastapi import FastAPI, File, UploadFile
-import os
 import shutil
+import os
 from whispercpp import Whisper
 
 app = FastAPI()
 
-# Load model using the updated API
+# Load the model (auto-download and cache it)
-whisper = Whisper.from_pretrained("./app/model/ggml-base.en.bin")
+whisper = Whisper.from_pretrained("base.en")  # Options: tiny, base, small, etc.
 
 @app.post("/transcribe")
 async def transcribe_audio(audio: UploadFile = File(...)):
@@ -14,6 +14,6 @@ async def transcribe_audio(audio: UploadFile = File(...)):
     with open(temp_file, "wb") as buffer:
         shutil.copyfileobj(audio.file, buffer)
 
-    text = whisper.transcribe(temp_file)
+    result = whisper.transcribe(temp_file)
     os.remove(temp_file)
-    return {"text": text}
+    return {"text": result}