Spaces:
Runtime error
Runtime error
File size: 4,587 Bytes
9dc5b4d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 |
import os
import openai
import gradio as gr
from langchain.chat_models import ChatOpenAI
from langchain.prompts import ChatPromptTemplate
from langchain.prompts import ChatPromptTemplate
from langchain.output_parsers import ResponseSchema
from langchain.output_parsers import StructuredOutputParser
os.environ['OPENAI_API_KEY'] = ''
openai.api_key = os.environ['OPENAI_API_KEY']
# To control the randomness and creativity of the generated
# text by an LLM, use temperature = 0.0
chat = ChatOpenAI(temperature=0.0)
chat
def get_format_instructions():
gift_schema = ResponseSchema(name="gift",
description="Was the item purchased\
as a gift for someone else? \
Answer True if yes,\
False if not or unknown.")
delivery_days_schema = ResponseSchema(name="delivery_days",
description="How many days\
did it take for the product\
to arrive? If this \
information is not found,\
output -1.")
price_value_schema = ResponseSchema(name="price_value",
description="Extract any\
sentences about the value or \
price, and output them as a \
comma separated Python list.")
response_schemas = [gift_schema,
delivery_days_schema,
price_value_schema]
output_parser = StructuredOutputParser.from_response_schemas(response_schemas)
format_instructions = output_parser.get_format_instructions()
return format_instructions
# To retrieve ChatGPT response in the reqiuired style
def response(template_string, user_input, translate_style_parsing_inst):
prompt_template = ChatPromptTemplate.from_template(template_string)
customer_messages = prompt_template.format_messages(
style=translate_style_parsing_inst,
text=user_input)
# Call the LLM to translate the style or parse the customer message
customer_response = chat(customer_messages)
return customer_response.content, prompt_template.messages[0].prompt.input_variables
# To parse the ChatGPT response into a python dictionary
def parser_response(template_string_parse, user_input_parse):
format_instructions = get_format_instructions()
prompt = ChatPromptTemplate.from_template(template=template_string_parse)
messages = prompt.format_messages(text=user_input_parse,
format_instructions=format_instructions)
response = chat(messages)
output_dict = output_parser.parse(response.content)
return output_dict, type(output_dict)
demo = gr.Blocks()
title = """<h1 align="center">Gradio x Langchain - Models, Prompts, and Parsers</h1>"""
with demo:
gr.HTML(title)
with gr.Tab("Translate"):
with gr.Row():
user_input = gr.Textbox(label="Enter user input for translation or parsing", lines=5, max_lines=5)
template_string = gr.Textbox(label="Enter your prompt here", lines=5, max_lines=5)
translate_style_parsing_inst = gr.Textbox(label="Enter the translation style of choice", lines=5, max_lines=5)
btn_response = gr.Button("ChatGPT Response").style(full_width=True)
with gr.Row():
chat_response = gr.Textbox(label="Response from ChatGPT", lines=5, max_lines=5)
with gr.Column():
template_variables = gr.Textbox(label="Input variables for your prompt")
with gr.Tab("Parse"):
with gr.Row():
user_input_parse = gr.Textbox(label="Enter user input for translation or parsing", lines=5, max_lines=5)
template_string_parse = gr.Textbox(label="Enter your prompt here", lines=5, max_lines=5)
btn_response_parse = gr.Button("Parsed ChatGPT Response").style(full_width=True)
with gr.Row():
with gr.Column(scale=5):
chat_response_parse = gr.Textbox(label="Get your ChatGPT response parsed as a dictionary (json)", lines=5, max_lines=5)
with gr.Column(scale=5):
type_parse_output = gr.Textbox(label="Datatype of this parsed output")
btn_response.click(response, [template_string, user_input, translate_style_parsing_inst], [chat_response, template_variables])
btn_response_parse.click(parser_response, [template_string_parse, user_input_parse], [chat_response_parse, type_parse_output])
demo.launch() #(debug=True)
|