fix: update code

app.py

@@ -1,37 +1,27 @@
-from transformers import HubertForCTC, Wav2Vec2Processor
+from transformers import pipeline
 import gradio as gr
-import time
-import torch
-import soundfile as sf
 import requests
 import os
 
-API_URL = "https://api-inference.huggingface.co/models/omarxadel/hubert-large-arabic-egyptian"
-token = os.environ['apikey']
-headers = {"Authorization": token}
+transcriber = pipeline("automatic-speech-recognition", model="omarxadel/hubert-large-arabic-egyptian")
 
+def transcribe(stream, new_chunk):
+    sr, y = new_chunk
+    y = y.astype(np.float32)
+    y /= np.max(np.abs(y))
 
-def transcribe(audio, state=""):
-    time.sleep(2)
+    if stream is not None:
+        stream = np.concatenate([stream, y])
+    else:
+        stream = y
+    return stream, transcriber({"sampling_rate": sr, "raw": stream})["text"]
 
-    # Load model from HuggingFace Hub
-    with open(audio, "rb") as f:
-        data = f.read()
-    response = requests.post(API_URL, headers=headers, data=data)
-    output = response.json()["text"]
-    state += output + " "
-    return state, state
 
+demo = gr.Interface(
+    transcribe,
+    ["state", gr.Audio(source="microphone", streaming=True)],
+    ["state", "text"],
+    live=True,
+)
 
-gr.Interface(
-    fn=transcribe,
-    inputs=[
-        gr.Audio(source="microphone", type="filepath", streaming=True),
-        "state"
-    ],
-    outputs=[
-        "textbox",
-        "state"
-    ],
-    live=True).launch(share=True)
-
+demo.launch()