Spaces:
Running
Running
| import gradio as gr | |
| import requests | |
| import os | |
| import json | |
| import google.generativeai as genai | |
| from bs4 import BeautifulSoup | |
| from groq import Groq | |
| # Load environment variables | |
| genai.configure(api_key=os.environ["geminiapikey"]) | |
| read_key = os.environ.get('HF_TOKEN', None) | |
| custom_css = """ | |
| #md { | |
| height: 400px; | |
| font-size: 30px; | |
| background: #202020; | |
| padding: 20px; | |
| color: white; | |
| border: 1 px solid white; | |
| } | |
| """ | |
| api_key = os.getenv('groq') | |
| if api_key is None: | |
| raise ValueError("groq_whisper environment variable is not set") | |
| # Initialize the Groq client | |
| client = Groq(api_key=api_key) | |
| def predict(prompt): | |
| generation_config = { | |
| "temperature": 0.4, | |
| "top_p": 0.95, | |
| "top_k": 40, | |
| "max_output_tokens": 8192, | |
| "response_mime_type": "text/plain", | |
| } | |
| model = genai.GenerativeModel( | |
| model_name="gemini-2.0-flash-exp", | |
| generation_config=generation_config, | |
| ) | |
| chat_session = model.start_chat( | |
| history=[] | |
| ) | |
| response = chat_session.send_message(f"{prompt}\n antworte immer auf deutsch") | |
| response_value = response.candidates[0].content.parts[0].text | |
| return response_value | |
| #very simple (and extremly fast) websearch | |
| def websearch(search_term): | |
| headers = { | |
| "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36" | |
| } | |
| url = f"https://www.google.com/search?q={search_term}" | |
| response = requests.get(url, headers=headers) | |
| soup = BeautifulSoup(response.content, 'html.parser') | |
| response_text = soup.find('body') | |
| prompt = f"{search_term}\n use this result from a google search to answer the question \n {response_text.text}" | |
| result = predict(prompt) | |
| return result | |
| def process_audio(file_path): | |
| try: | |
| # Open the audio file | |
| with open(file_path, "rb") as file: | |
| # Create a translation of the audio file | |
| translation = client.audio.transcriptions.create( | |
| file=(os.path.basename(file_path), file.read()), # Correct passing of filename | |
| model="whisper-large-v3-turbo", # Required model to use for translation | |
| prompt="transcribe", # Optional | |
| language="de", # Optional | |
| response_format="json", # Optional | |
| temperature=0.0 # Optional | |
| ) | |
| # Return the translation text | |
| suche = websearch(translation.text) | |
| result = predict(suche) | |
| return result | |
| return translation.text | |
| except Exception as e: | |
| return f"An error occurred: {str(e)}" | |
| # Create the Gradio interface | |
| with gr.Blocks(css=custom_css) as demo: | |
| with gr.Row(): | |
| details_output = gr.Markdown(label="answer", elem_id="md") | |
| #details_output = gr.Textbox(label="Ausgabe", value = f"\n\n\n\n") | |
| with gr.Row(): | |
| ort_input = gr.Textbox(label="prompt", placeholder="ask anything...") | |
| audio_input=gr.Microphone(type="filepath") | |
| with gr.Row(): | |
| button = gr.Button("Senden") | |
| # Connect the button to the function | |
| button.click(fn=websearch, inputs={ort_input, audio_input], outputs=details_output) | |
| # Launch the Gradio application | |
| demo.launch() | |
| """ | |
| with gr.Blocks() as speech: | |
| with gr.Row(): | |
| sr_outputs = gr.Textbox(label="Antwort") | |
| with gr.Row(): | |
| sr_inputs = gr.Microphone(type="filepath") | |
| sr_inputs.change(process_audio, inputs=sr_inputs, outputs=sr_outputs) | |
| speech.launch() | |
| """ | |