Update app.py

app.py

@@ -1,32 +1,23 @@
 import gradio as gr
-import whisper
-import numpy as np
-import wave
-import io
+from transformers import pipeline
 
-# Load Whisper model
-model = whisper.load_model("base")  # You can change to any other model like "small", "medium", etc.
+# Load Whisper model from Hugging Face
+# This uses the `transformers` library's pipeline to load the model
+transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-large-v3")
 
 def transcribe(audio):
-    # Convert the uploaded audio file to a numpy array
-    with wave.open(io.BytesIO(audio), "rb") as wav_reader:
-        samples = wav_reader.getnframes()
-        audio_data = wav_reader.readframes(samples)
-        audio_as_np_int16 = np.frombuffer(audio_data, dtype=np.int16)
-        audio_as_np_float32 = audio_as_np_int16.astype(np.float32) / np.iinfo(np.int16).max
-    
     # Transcribe the audio using the Whisper model
-    result = model.transcribe(audio_as_np_float32)
-    return result["text"]
+    result = transcriber(audio)["text"]
+    return result
 
 # Create a Gradio Interface
 interface = gr.Interface(
     fn=transcribe, 
-    inputs=gr.Audio(source="upload", type="bytes"), 
+    inputs=gr.Audio(source="upload", type="filepath"), 
     outputs="text",
     title="Whisper Speech-to-Text API",
-    description="Upload an audio file and get a transcription using OpenAI's Whisper model."
+    description="Upload an audio file and get a transcription using OpenAI's Whisper model from Hugging Face."
 )
 
 # Launch the interface as an API
-interface.launch()
+interface.launch()