Create app.py

app.py

@@ -0,0 +1,49 @@
+import requests
+import gradio as gr
+from key import HF_ACCESS_TOKEN
+import numpy as np
+import io
+import soundfile as sf
+
+API_URL = "https://api-inference.huggingface.co/models/openai/whisper-large-v3"
+headers = {"Authorization": f"Bearer {HF_ACCESS_TOKEN}"}
+
+def query(audio_data):
+    with io.BytesIO() as f:
+        sf.write(f, audio_data[1], audio_data[0], format='wav')
+        data = f.getvalue()
+    
+    response = requests.post(API_URL, headers=headers, data=data)
+    return response.json()['text']
+
+def transcribe(audio_data):
+    global output_text
+    print("Received audio data:", audio_data)
+    
+    if audio_data is None:
+        print("Audio data is None. Check the microphone and input configuration.")
+        return None
+    
+    sr, y = audio_data
+    y = y.astype(np.float32)
+    y /= np.max(np.abs(y))
+    
+    # Add your transcription logic here if needed
+    transcription = query(audio_data)    
+    return transcription
+
+dark_minimalist = gr.Theme.from_hub("Taithrah/Minimal")
+
+iface = gr.Interface(theme=dark_minimalist,
+    fn=transcribe,
+    inputs=gr.Microphone(label="Speak into the microphone",),
+    outputs="text",
+    allow_flagging="never",
+    css="""
+    footer {
+        visibility: hidden;
+    }
+    """
+)
+
+iface.launch()