Update app.py

app.py

@@ -1,22 +1,28 @@
 import torch
 from transformers import pipeline
+import gradio as gr
 
-# Initialize the speech-to-text pipeline from Hugging Face Transformers
-# This uses the "openai/whisper-tiny.en" model for automatic speech recognition (ASR)
-# The `chunk_length_s` parameter specifies the chunk length in seconds for processing
-pipe = pipeline(
-  "automatic-speech-recognition",
-  model="openai/whisper-tiny.en",
-  chunk_length_s=30,
-)
+# Function to transcribe audio using the OpenAI Whisper model
+def transcript_audio(audio_file):
+    # Initialize the speech recognition pipeline
+    pipe = pipeline(
+        "automatic-speech-recognition",
+        model="openai/whisper-tiny.en",
+        chunk_length_s=30,
+    )
+    # Transcribe the audio file and return the result
+    result = pipe(audio_file, batch_size=8)["text"]
+    return result
 
-# Define the path to the audio file that needs to be transcribed
-sample = 'downloaded_audio.mp3'
+# Set up Gradio interface
+audio_input = gr.Audio(sources="upload", type="filepath")  # Audio input
+output_text = gr.Textbox()  # Text output
 
-# Perform speech recognition on the audio file
-# The `batch_size=8` parameter indicates how many chunks are processed at a time
-# The result is stored in `prediction` with the key "text" containing the transcribed text
-prediction = pipe(sample, batch_size=8)["text"]
+# Create the Gradio interface with the function, inputs, and outputs
+iface = gr.Interface(fn=transcript_audio, 
+                     inputs=audio_input, outputs=output_text, 
+                     title="Audio Transcription App",
+                     description="Upload the audio file")
 
-# Print the transcribed text to the console
-print(prediction)
+# Launch the Gradio app
+iface.launch()