textToSQL's picture
Update app.py
81033d4
import whisper
import gradio as gr
import openai
import os
openai.api_key = os.environ["OPENAI_API_KEY"]
model = whisper.load_model("small")
#option 1
def transcribe(audio):
model = whisper.load_model("base")
result = model.transcribe(audio)
return result["text"]
#option 2
# def transcribe(audio):
# #time.sleep(3)
# # load audio and pad/trim it to fit 30 seconds
# audio = whisper.load_audio(audio)
# audio = whisper.pad_or_trim(audio)
# # make log-Mel spectrogram and move to the same device as the model
# mel = whisper.log_mel_spectrogram(audio).to(model.device)
# # detect the spoken language
# _, probs = model.detect_language(mel)
# print(f"Detected language: {max(probs, key=probs.get)}")
# # decode the audio
# options = whisper.DecodingOptions(fp16 = False)
# result = whisper.decode(model, mel, options)
# return result.text
def process_text(input_text):
# Apply your function here to process the input text
output_text = input_text.upper()
return output_text
def get_completion(prompt, model='gpt-3.5-turbo'):
messages = [
{"role": "system", "content": """You are a .... You are provided with the transcription of a ... . \
Extract the following information from the transcription, replace curly brackets {} with relevant extracted information ... \
...the rest of your prompt...
"""
},
{"role": "user", "content": prompt}
]
response = openai.ChatCompletion.create(
model = model,
messages = messages,
temperature = 0,
)
return response.choices[0].message['content']
with gr.Blocks() as demo:
gr.Markdown("""
# Title <br>
Description
""")
title = "title"
audio = gr.Audio(type="filepath")
b1 = gr.Button("Transcribe audio")
b2 = gr.Button("<Placeholder for the prompted action>")
# b3 = gr.Button("Email report to your doctor")
text1 = gr.Textbox(lines=5)
text2 = gr.Textbox(lines=5)
prompt = text1
b1.click(transcribe, inputs=audio, outputs=text1)
b2.click(get_completion, inputs=text1, outputs=text2)
# b1.click(transcribe, inputs=audio, outputs=text1)
# b2.click(get_completion, inputs=prompt, outputs=text2)
demo.launch()
#demo.launch(share=True, auth=("username", "password"))
# In this example, the process_text function just converts the input text to uppercase, but you can replace it with your desired function. The Gradio Blocks interface will have two buttons: "Transcribe audio" and "Process text". The first button transcribes the audio and fills the first textbox, and the second button processes the text from the first textbox and fills the second textbox.
# gr.Interface(
# title = 'OpenAI Whisper ASR Gradio Web UI',
# fn=transcribe,
# inputs=[
# gr.inputs.Audio(source="microphone", type="filepath")
# ],
# outputs=[
# "textbox"
# ],
# live=True).launch()