cd2-ai / app.py
brurei's picture
Update app.py
aebc2c9
raw
history blame
2.26 kB
# coding=utf8
from gpt_index import SimpleDirectoryReader, GPTListIndex, GPTSimpleVectorIndex, LLMPredictor, PromptHelper
from langchain import OpenAI
import gradio as gr
import random
import time
import sys
import os
from transformers import pipeline
p = pipeline("automatic-speech-recognition")
os.environ["OPENAI_API_KEY"] = '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'
md = """This is some code:
hello
```py
def fn(x, y, z):
print(x, y, z)
"""
def transcribe(audio):
text = p(audio)["text"]
return text
def construct_index(directory_path):
max_input_size = 10000
num_outputs = 10000
max_chunk_overlap = 20000
chunk_size_limit = 600000
prompt_helper = PromptHelper(max_input_size, num_outputs, max_chunk_overlap, chunk_size_limit=chunk_size_limit)
llm_predictor = LLMPredictor(llm=OpenAI(temperature=0.0, model_name="text-davinci-003", max_tokens=num_outputs))
documents = SimpleDirectoryReader(directory_path).load_data()
index = GPTSimpleVectorIndex.from_documents(documents)
index.save_to_disk('index.json')
return index
def chatbot(input_text):
index = GPTSimpleVectorIndex.load_from_disk('index.json')
response = index.query(input_text)
return str(response.response)
with gr.Blocks() as demo:
gpt = gr.Chatbot(label="GPT SUPEr", elem_id="chatbot").style(height=800)
msg = gr.Textbox( show_label=False,
placeholder="Bem vindo ao ExpoSuper, Qual sua pergunta?",
).style(container=False)
clear = gr.Button("Limpar Conversa")
gr.Audio(source="microphone", type="filepath",label="ESTÁ COM DIFICULDADES EM ESCREVER? CLIQUE E ME DIGA O QUE DESEJA")
def respond(message, chat_history):
chat_history.append((message, chatbot(message)))
time.sleep(1)
vetor = []
realPath = str(os.path.dirname(os.path.realpath(__file__)))
if str(message).upper()=="OLA" or str(message).upper()=="OLÁ" or str(message).upper()=="OI":
vetor = vetor + [((realPath + "\\images\\logo.png",), "")]
return "", chat_history+vetor
clear.click(lambda:None, None, gpt, queue=False,)
msg.submit(respond, [msg, gpt], [msg,gpt])
index = construct_index("docs")
demo.launch()