Update main.py

main.py

@@ -1,7 +1,4 @@
 import torch
-import torchaudio
-import wave
-import numpy as np
 from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
 from datasets import load_dataset
 from googletrans import Translator
@@ -45,22 +42,15 @@ async def process_audio(file: UploadFile = File(...)):
         file_location = save_directory / file.filename
 
         with open(file_location, "wb") as saved_file:
-            content = await file.read()
-            saved_file.write(content)
+            saved_file.write(file.file.read())saved_file.write(content)
 
-        # Load audio file using torchaudio
-        waveform, sample_rate = torchaudio.load(file_location)
-
-        # Ensure the waveform is a single-channel array
-        waveform = waveform[0] if waveform.size(0) > 1 else waveform
-
-        # ASR
-        original = pipe(waveform.numpy())
-        original_version = original[0]["text"]
+        # JP
+        original = pipe(file_location)
+        original_version = original["text"]
 
         # EN
-        result = pipe(waveform.numpy(), generate_kwargs={"task": "translate"})
-        hasil = result[0]["text"]
+        result = pipe(file_location, generate_kwargs={"task": "translate"})
+        hasil = result["text"]
 
         # ID
         detect = detect_google(hasil)