up

main.py

@@ -1,6 +1,7 @@
 import os
 import logging
 import uvicorn
+import json
 
 import gradio as gr
 import numpy as np
@@ -43,7 +44,7 @@ logger.info(f"CUDA Version: {cuda_version}, GPU Device: {device_name}")
 attention = "flash_attention_2" if is_flash_attn_2_available() else "sdpa"
 logger.info(f"Using attention: {attention}")
 
-model_id = "openai/whisper-large-v3-turbo"
+model_id = os.getenv("MODEL_ID", "openai/whisper-large-v3-turbo")
 logger.info(f"Loading Whisper model: {model_id}")
 model = AutoModelForSpeechSeq2Seq.from_pretrained(
     model_id,