Update app.py

app.py

@@ -1,7 +1,6 @@
 
 import gradio as gr
 import edge_tts
-import asyncio
 import tempfile
 import numpy as np
 import soxr
@@ -10,95 +9,11 @@ import torch
 import sentencepiece as spm
 import onnxruntime as ort
 from huggingface_hub import hf_hub_download, InferenceClient
-import requests
-from bs4 import BeautifulSoup
-import urllib
-import random
 
 theme = gr.themes.Soft(
     primary_hue="blue",
     secondary_hue="orange")
 
-
-# List of user agents to choose from for requests
-_useragent_list = [
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:66.0) Gecko/20100101 Firefox/66.0',
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36',
-    'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36',
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36',
-    'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36',
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36 Edg/111.0.1661.62',
-    'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 Firefox/111.0'
-]
-
-def get_useragent():
-    """Returns a random user agent from the list."""
-    return random.choice(_useragent_list)
-
-def extract_text_from_webpage(html_content):
-    """Extracts visible text from HTML content using BeautifulSoup."""
-    soup = BeautifulSoup(html_content, "html.parser")
-    # Remove unwanted tags
-    for tag in soup(["script", "style", "header", "footer", "nav"]):
-        tag.extract()
-    # Get the remaining visible text
-    visible_text = soup.get_text(strip=True)
-    return visible_text
-
-def search(term, num_results=1, lang="en", advanced=True, sleep_interval=0, timeout=5, safe="active", ssl_verify=None):
-    """Performs a Google search and returns the results."""
-    escaped_term = urllib.parse.quote_plus(term)
-    start = 0
-    all_results = []
-
-    # Fetch results in batches
-    while start < num_results:
-        resp = requests.get(
-            url="https://www.google.com/search",
-            headers={"User-Agent": get_useragent()}, # Set random user agent
-            params={
-                "q": term,
-                "num": num_results - start, # Number of results to fetch in this batch
-                "hl": lang,
-                "start": start,
-                "safe": safe,
-            },
-            timeout=timeout,
-            verify=ssl_verify,
-        )
-        resp.raise_for_status() # Raise an exception if request fails
-
-        soup = BeautifulSoup(resp.text, "html.parser")
-        result_block = soup.find_all("div", attrs={"class": "g"})
-
-        # If no results, continue to the next batch
-        if not result_block:
-            start += 1
-            continue
-
-        # Extract link and text from each result
-        for result in result_block:
-            link = result.find("a", href=True)
-            if link:
-                link = link["href"]
-                try:
-                    # Fetch webpage content
-                    webpage = requests.get(link, headers={"User-Agent": get_useragent()})
-                    webpage.raise_for_status()
-                    # Extract visible text from webpage
-                    visible_text = extract_text_from_webpage(webpage.text)
-                    all_results.append({"link": link, "text": visible_text})
-                except requests.exceptions.RequestException as e:
-                    # Handle errors fetching or processing webpage
-                    print(f"Error fetching or processing {link}: {e}")
-                    all_results.append({"link": link, "text": None})
-            else:
-                all_results.append({"link": None, "text": None})
-
-        start += len(result_block) # Update starting index for next batch
-
-    return all_results
-
 # Speech Recognition Model Configuration
 model_name = "neongeckocom/stt_en_citrinet_512_gamma_0_25"
 sample_rate = 16000
@@ -118,7 +33,7 @@ def resample(audio_fp32, sr):
 def to_float32(audio_buffer):
     return np.divide(audio_buffer, np.iinfo(audio_buffer.dtype).max, dtype=np.float32)
 
-def transcribe(audio_path):
+async def transcribe(audio_path):
     audio_file = AudioSegment.from_file(audio_path)
     sr = audio_file.frame_rate
     audio_buffer = np.array(audio_file.get_array_of_samples())
@@ -138,22 +53,14 @@ def transcribe(audio_path):
 
     return text
 
-def model(text, web_search):
-    if web_search is True:
-        """Performs a web search, feeds the results to a language model, and returns the answer."""
-        web_results = search(text)
-        web2 = ' '.join([f"Link: {res['link']}\nText: {res['text']}\n\n" for res in web_results])
-        formatted_prompt = system_instructions1 + text + "[WEB]" + str(web2) + "[OpenGPT 4o]"
-        stream = client1.text_generation(formatted_prompt, max_new_tokens=512, stream=True, details=True, return_full_text=False)
-        return "".join([response.token.text for response in stream if response.token.text != "</s>"])
-    else:
-        formatted_prompt = system_instructions1 + text + "[OpenGPT 4o]"
-        stream = client1.text_generation(formatted_prompt, max_new_tokens=512, stream=True, details=True, return_full_text=False)
-        return "".join([response.token.text for response in stream if response.token.text != "</s>"])
-
-async def respond(audio, web_search):
-    user = transcribe(audio)
-    reply = model(user, web_search)
+async def model(text):
+    formatted_prompt = system_instructions1 + text + "[OpenGPT 4o]"
+    stream = client1.text_generation(formatted_prompt, max_new_tokens=512, stream=True, details=True, return_full_text=False)
+    return "".join([response.token.text for response in stream if response.token.text != "</s>"])
+
+async def respond(audio):
+    user = await transcribe(audio)
+    reply = await model(user)
     communicate = edge_tts.Communicate(reply)
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
         tmp_path = tmp_file.name
@@ -161,11 +68,9 @@ async def respond(audio, web_search):
     return tmp_path
 
 with gr.Blocks(theme=theme) as demo:    
-    with gr.Row():
-        web_search = gr.Checkbox(label="Web Search", value=False)
-        input = gr.Audio(label="User Input", sources="microphone", type="filepath")
-        output = gr.Audio(label="AI", autoplay=True)
-        gr.Interface(fn=respond, inputs=[input, web_search], outputs=[output], live=True)
+    input = gr.Audio(label="User Input", sources="microphone", type="filepath")
+    output = gr.Audio(label="AI", autoplay=True)
+    gr.Interface(fn=respond, inputs=[input], outputs=[output], live=True)
 
 if __name__ == "__main__":
     demo.queue(max_size=200).launch()