Spaces:

fudii0921
/

savers

Runtime error

savers / app.py

Update app.py

3019d6a verified over 1 year ago

5.21 kB

import os
import fitz  # PyMuPDF
import gradio as gr
import autogen
from autogen.agentchat.contrib.capabilities import transform_messages, transforms
from autogen.agentchat.contrib.capabilities.text_compressors import LLMLingua
from autogen.agentchat.contrib.capabilities.transforms import TextMessageCompressor
import mysql.connector

import copy
import pprint
import re
from typing import Dict, List, Tuple

# MySQLに接続
conn = mysql.connector.connect(
    host="www.ryhintl.com",
    user="smairuser",
    password="smairuser",
    port=36000,
    database="smair"
)

# カーソルを取得
cursor = conn.cursor(dictionary=True)

# List API Keys
select_one_data_query = "SELECT * FROM agentic_apis"
cursor.execute(select_one_data_query)
result = cursor.fetchall()
# JSONをパースしてkeyを抽出
keys = [item['key'] for item in result]

os.environ["GROQ_API_KEY"] = keys[2]

def extract_text_from_pdf(file):
    text = ""
    with fitz.open("./"+file) as doc:
        for page in doc:
            text += page.get_text()

    return text

def summarize_pdf(file):
    pdf_text = extract_text_from_pdf(file)
    #pdf_text = extract_text_from_pdf(text)
    
    #system_message = "You are a world class researcher."
    system_message = "貴方は世界的なレベルのリサーチャーです。"
    config_list = [{
            "model": "llama-3.3-70b-versatile",
            #"model": "llama-guard-3-8b",
            "api_key": os.environ.get("GROQ_API_KEY"),
            "api_type": "groq"
        }]
    
    researcher = autogen.ConversableAgent(
        "assistant",
        llm_config={"config_list": config_list},
        max_consecutive_auto_reply=1,
        system_message=system_message,
        human_input_mode="NEVER",
    )
    user_proxy = autogen.UserProxyAgent(
        "user_proxy",
        human_input_mode="NEVER",
        is_termination_msg=lambda x: "TERMINATE" in x.get("content", ""),
        max_consecutive_auto_reply=1,
        code_execution_config={"work_dir":"coding", "use_docker":False}
    )
    
    llm_lingua = LLMLingua()
    #text_compressor = TextMessageCompressor(text_compressor=llm_lingua,compression_params={"target_token": 13000},cache=None)
    text_compressor = TextMessageCompressor(text_compressor=llm_lingua,compression_params={"target_token": 5999},cache=None)
    compressed_text = text_compressor.apply_transform([{"content": pdf_text}])
    
    
    #content_list = [item['content'] for item in compressed_text]
    # Limit the message history to the 3 most recent messages
    max_msg_transfrom = transforms.MessageHistoryLimiter(max_messages=3)

    # Limit the token limit per message to 10 tokens
    token_limit_transform = transforms.MessageTokenLimiter(max_tokens_per_message=3, min_tokens=10)
    
    '''messages = [
        {"role": "user", "content": "hello"},
        {"role": "assistant", "content": [{"type": "text", "text": "there"}]},
        {"role": "user", "content": "how"},
        {"role": "assistant", "content": [{"type": "text", "text": "are you doing?"}]},
        {"role": "user", "content": "very very very very very very long string"},
    ]
    processed_messages = max_msg_transfrom.apply_transform(copy.deepcopy(messages))
    pprint.pprint(processed_messages)
    processed_messages = token_limit_transform.apply_transform(copy.deepcopy(messages))
    pprint.pprint(processed_messages)'''
    

    context_handling = transform_messages.TransformMessages(
        transforms=[
            transforms.MessageHistoryLimiter(max_messages=10),
            transforms.MessageTokenLimiter(max_tokens=6000, max_tokens_per_message=2000, min_tokens=500),
            #transforms.MessageTokenLimiter(max_tokens=1000, max_tokens_per_message=50, min_tokens=500),
        ]
    )

    context_handling.add_to_agent(researcher)
    
    
    #context_handling = transform_messages.TransformMessages(transforms=[text_compressor])
    #context_handling.add_to_agent(researcher)
    
    #message = "Summarize this research paper for me in Japanese, include the important information" + pdf_text
    message = "この資料を日本語で要約し、重要な情報を含めてください。節約されたトークン数も表示してください。" + pdf_text
    result = user_proxy.initiate_chat(recipient=researcher, clear_history=True, message=message, silent=True)
    
    #print(text_compressor.get_logs([], []))
    # タプルの例
    tresult = text_compressor.get_logs([], [])
    #print(tresult)

    # 文字列から数字を抽出
    #saved_tokens = str(int(tresult[0].split()[0]))
    #print(saved_tokens)
    
    return result.chat_history[1]["content"]
    #return result.chat_history[1]["content"]+"\n\n"+saved_tokens+"トークンが節約できました。"

iface = gr.Interface(
    fn=summarize_pdf,
    #inputs=gr.inputs.File(label="Upload PDF"),
    inputs=gr.Dropdown(
        choices=["yoin.pdf", "spo_revenue.pdf", "lings.pdf", "korea-ai.pdf"],  # ドロップダウンの選択肢を指定
        label="PDFを選択"  # ラベルを指定
    ),
    outputs="text",
    title="Research Paper Summarizer",
    description="Select a PDF and get a summary in Japanese."
)

iface.launch()

	import os
	import fitz # PyMuPDF
	import gradio as gr
	import autogen
	from autogen.agentchat.contrib.capabilities import transform_messages, transforms
	from autogen.agentchat.contrib.capabilities.text_compressors import LLMLingua
	from autogen.agentchat.contrib.capabilities.transforms import TextMessageCompressor
	import mysql.connector

	import copy
	import pprint
	import re
	from typing import Dict, List, Tuple

	# MySQLに接続
	conn = mysql.connector.connect(
	host="www.ryhintl.com",
	user="smairuser",
	password="smairuser",
	port=36000,
	database="smair"
	)

	# カーソルを取得
	cursor = conn.cursor(dictionary=True)

	# List API Keys
	select_one_data_query = "SELECT * FROM agentic_apis"
	cursor.execute(select_one_data_query)
	result = cursor.fetchall()
	# JSONをパースしてkeyを抽出
	keys = [item['key'] for item in result]

	os.environ["GROQ_API_KEY"] = keys[2]

	def extract_text_from_pdf(file):
	text = ""
	with fitz.open("./"+file) as doc:
	for page in doc:
	text += page.get_text()

	return text

	def summarize_pdf(file):
	pdf_text = extract_text_from_pdf(file)
	#pdf_text = extract_text_from_pdf(text)

	#system_message = "You are a world class researcher."
	system_message = "貴方は世界的なレベルのリサーチャーです。"
	config_list = [{
	"model": "llama-3.3-70b-versatile",
	#"model": "llama-guard-3-8b",
	"api_key": os.environ.get("GROQ_API_KEY"),
	"api_type": "groq"
	}]

	researcher = autogen.ConversableAgent(
	"assistant",
	llm_config={"config_list": config_list},
	max_consecutive_auto_reply=1,
	system_message=system_message,
	human_input_mode="NEVER",
	)
	user_proxy = autogen.UserProxyAgent(
	"user_proxy",
	human_input_mode="NEVER",
	is_termination_msg=lambda x: "TERMINATE" in x.get("content", ""),
	max_consecutive_auto_reply=1,
	code_execution_config={"work_dir":"coding", "use_docker":False}
	)

	llm_lingua = LLMLingua()
	#text_compressor = TextMessageCompressor(text_compressor=llm_lingua,compression_params={"target_token": 13000},cache=None)
	text_compressor = TextMessageCompressor(text_compressor=llm_lingua,compression_params={"target_token": 5999},cache=None)
	compressed_text = text_compressor.apply_transform([{"content": pdf_text}])


	#content_list = [item['content'] for item in compressed_text]
	# Limit the message history to the 3 most recent messages
	max_msg_transfrom = transforms.MessageHistoryLimiter(max_messages=3)

	# Limit the token limit per message to 10 tokens
	token_limit_transform = transforms.MessageTokenLimiter(max_tokens_per_message=3, min_tokens=10)

	'''messages = [
	{"role": "user", "content": "hello"},
	{"role": "assistant", "content": [{"type": "text", "text": "there"}]},
	{"role": "user", "content": "how"},
	{"role": "assistant", "content": [{"type": "text", "text": "are you doing?"}]},
	{"role": "user", "content": "very very very very very very long string"},
	]

	processed_messages = max_msg_transfrom.apply_transform(copy.deepcopy(messages))
	pprint.pprint(processed_messages)

	processed_messages = token_limit_transform.apply_transform(copy.deepcopy(messages))

	pprint.pprint(processed_messages)'''


	context_handling = transform_messages.TransformMessages(
	transforms=[
	transforms.MessageHistoryLimiter(max_messages=10),
	transforms.MessageTokenLimiter(max_tokens=6000, max_tokens_per_message=2000, min_tokens=500),
	#transforms.MessageTokenLimiter(max_tokens=1000, max_tokens_per_message=50, min_tokens=500),
	]
	)

	context_handling.add_to_agent(researcher)


	#context_handling = transform_messages.TransformMessages(transforms=[text_compressor])
	#context_handling.add_to_agent(researcher)

	#message = "Summarize this research paper for me in Japanese, include the important information" + pdf_text
	message = "この資料を日本語で要約し、重要な情報を含めてください。節約されたトークン数も表示してください。" + pdf_text
	result = user_proxy.initiate_chat(recipient=researcher, clear_history=True, message=message, silent=True)

	#print(text_compressor.get_logs([], []))
	# タプルの例
	tresult = text_compressor.get_logs([], [])
	#print(tresult)

	# 文字列から数字を抽出
	#saved_tokens = str(int(tresult[0].split()[0]))
	#print(saved_tokens)

	return result.chat_history[1]["content"]
	#return result.chat_history[1]["content"]+"\n\n"+saved_tokens+"トークンが節約できました。"

	iface = gr.Interface(
	fn=summarize_pdf,
	#inputs=gr.inputs.File(label="Upload PDF"),
	inputs=gr.Dropdown(
	choices=["yoin.pdf", "spo_revenue.pdf", "lings.pdf", "korea-ai.pdf"], # ドロップダウンの選択肢を指定
	label="PDFを選択" # ラベルを指定
	),
	outputs="text",
	title="Research Paper Summarizer",
	description="Select a PDF and get a summary in Japanese."
	)

	iface.launch()