import torch
import transformers
import gradio as gr
from ragatouille import RAGPretrainedModel
from huggingface_hub import InferenceClient
import re
from datetime import datetime
import json
import os
import arxiv
from utils import get_md_text_abstract, search_cleaner, get_arxiv_live_search
retrieve_results = 10
show_examples = False
llm_models_to_choose = ['mistralai/Mixtral-8x7B-Instruct-v0.1','mistralai/Mistral-7B-Instruct-v0.2', 'google/gemma-7b-it', 'None']
generate_kwargs = dict(
temperature = None,
max_new_tokens = 512,
top_p = None,
do_sample = False,
)
## RAG Model
RAG = RAGPretrainedModel.from_index("colbert/indexes/arxiv_colbert")
try:
gr.Info("Setting up retriever, please wait...")
rag_initial_output = RAG.search("what is Mistral?", k = 1)
gr.Info("Retriever working successfully!")
except:
gr.Warning("Retriever not working!")
## Header
mark_text = '# π©Ίπ Search Results\n'
header_text = "## Arxiv Paper Summary With QA Retrieval Augmented Generation \n"
try:
with open("README.md", "r") as f:
mdfile = f.read()
date_pattern = r'Index Last Updated : \d{4}-\d{2}-\d{2}'
match = re.search(date_pattern, mdfile)
date = match.group().split(': ')[1]
formatted_date = datetime.strptime(date, '%Y-%m-%d').strftime('%d %b %Y')
header_text += f'Index Last Updated: {formatted_date}\n'
index_info = f"Semantic Search - up to {formatted_date}"
except:
index_info = "Semantic Search"
database_choices = [index_info,'Arxiv Search - Latest - (EXPERIMENTAL)']
## Arxiv API
arx_client = arxiv.Client()
is_arxiv_available = True
check_arxiv_result = get_arxiv_live_search("What is Self Rewarding AI and how can it be used in Multi-Agent Systems?", arx_client, retrieve_results)
if len(check_arxiv_result) == 0:
is_arxiv_available = False
print("Arxiv search not working, switching to default search ...")
database_choices = [index_info]
## Show examples (disabled)
if show_examples:
with open("sample_outputs.json", "r") as f:
sample_outputs = json.load(f)
output_placeholder = sample_outputs['output_placeholder']
md_text_initial = sample_outputs['search_placeholder']
else:
output_placeholder = None
md_text_initial = ''
def rag_cleaner(inp):
rank = inp['rank']
title = inp['document_metadata']['title']
content = inp['content']
date = inp['document_metadata']['_time']
return f"{rank}. <b> {title} </b> \n Date : {date} \n Abstract: {content}"
def get_prompt_text(question, context, formatted = True, llm_model_picked = 'mistralai/Mistral-7B-Instruct-v0.2'):
if formatted:
sys_instruction = f"Context:\n {context} \n Given the following scientific paper abstracts, take a deep breath and lets think step by step to answer the question. Cite the titles of your sources when answering, do not cite links or dates."
message = f"Question: {question}"
if 'mistralai' in llm_model_picked:
return f"<s>" + f"[INST] {sys_instruction}" + f" {message}[/INST]"
elif 'gemma' in llm_model_picked:
return f"<bos><start_of_turn>user\n{sys_instruction}" + f" {message}<end_of_turn>\n"
return f"Context:\n {context} \n Given the following info, take a deep breath and lets think step by step to answer the question: {question}. Cite the titles of your sources when answering.\n\n"
def get_references(question, retriever, k = retrieve_results):
rag_out = retriever.search(query=question, k=k)
return rag_out
def get_rag(message):
return get_references(message, RAG)
def SaveResponseAndRead(result):
documentHTML5='''
<!DOCTYPE html>
<html>
<head>
<title>Read It Aloud</title>
<script type="text/javascript">
function readAloud() {
const text = document.getElementById("textArea").value;
const speech = new SpeechSynthesisUtterance(text);
window.speechSynthesis.speak(speech);
}
</script>
</head>
<body>
<h1>π Read It Aloud</h1>
<textarea id="textArea" rows="10" cols="80">
'''
documentHTML5 = documentHTML5 + result
documentHTML5 = documentHTML5 + '''
</textarea>
<br>
<button onclick="readAloud()">π Read Aloud</button>
</body>
</html>
'''
gr.HTML(documentHTML5)
def save_search_results(prompt, results):
filename = re.sub(r'[^\w\-_\. ]', '_', prompt) + ".txt"
with open(filename, "w") as f:
f.write(f"# {prompt}\n\n")
f.write(results)
def get_past_searches():
txt_files = [f for f in os.listdir(".") if f.endswith(".txt") and f != "requirements.txt"]
return txt_files
with gr.Blocks(theme = gr.themes.Soft()) as demo:
header = gr.Markdown(header_text)
with gr.Row():
with gr.Column():
msg = gr.Textbox(label = 'Search', placeholder = 'What is Mistral?')
with gr.Accordion("Advanced Settings", open=False):
with gr.Row(equal_height = True):
llm_model = gr.Dropdown(choices = llm_models_to_choose, value = 'mistralai/Mistral-7B-Instruct-v0.2', label = 'LLM Model')
llm_results = gr.Slider(minimum=4, maximum=10, value=5, step=1, interactive=True, label="Top n results as context")
database_src = gr.Dropdown(choices = database_choices, value = index_info, label = 'Search Source')
stream_results = gr.Checkbox(value = True, label = "Stream output", visible = False)
output_text = gr.Textbox(show_label = True, container = True, label = 'LLM Answer', visible = True, placeholder = output_placeholder)
input = gr.Textbox(show_label = False, visible = False)
gr_md = gr.Markdown(mark_text + md_text_initial)
with gr.Column():
past_searches = gr.Dropdown(choices=get_past_searches(), label="Past Searches")
past_search_content = gr.Textbox(label="Past Search Content", visible=False)
def update_past_search_content(past_search):
if past_search:
with open(past_search, "r") as f:
content = f.read()
return gr.Textbox.update(value=content, visible=True)
else:
return gr.Textbox.update(visible=False)
past_searches.change(update_past_search_content, past_searches, past_search_content)
def update_with_rag_md(message, llm_results_use = 5, database_choice = index_info, llm_model_picked = 'mistralai/Mistral-7B-Instruct-v0.2'):
prompt_text_from_data = ""
database_to_use = database_choice
if database_choice == index_info:
rag_out = get_rag(message)
else:
arxiv_search_success = True
try:
rag_out = get_arxiv_live_search(message, arx_client, retrieve_results)
if len(rag_out) == 0:
arxiv_search_success = False
except:
arxiv_search_success = False
if not arxiv_search_success:
gr.Warning("Arxiv Search not working, switching to semantic search ...")
rag_out = get_rag(message)
database_to_use = index_info
md_text_updated = mark_text
for i in range(retrieve_results):
rag_answer = rag_out[i]
if i < llm_results_use:
md_text_paper, prompt_text = get_md_text_abstract(rag_answer, source = database_to_use, return_prompt_formatting = True)
prompt_text_from_data += f"{i+1}. {prompt_text}"
else:
md_text_paper = get_md_text_abstract(rag_answer, source = database_to_use)
md_text_updated += md_text_paper
prompt = get_prompt_text(message, prompt_text_from_data, llm_model_picked = llm_model_picked)
save_search_results(message, md_text_updated)
return md_text_updated, prompt, get_past_searches()
def ask_llm(prompt, llm_model_picked = 'mistralai/Mistral-7B-Instruct-v0.2', stream_outputs = False):
model_disabled_text = "LLM Model is disabled"
output = ""
if llm_model_picked == 'None':
if stream_outputs:
for out in model_disabled_text:
output += out
yield output
return output
else:
return model_disabled_text
client = InferenceClient(llm_model_picked)
try:
stream = client.text_generation(prompt, **generate_kwargs, stream=stream_outputs, details=False, return_full_text=False)
except:
gr.Warning("LLM Inference rate limit reached, try again later!")
return ""
if stream_outputs:
for response in stream:
output += response
SaveResponseAndRead(response)
yield output
return output
else:
return stream
msg.submit(update_with_rag_md, [msg, llm_results, database_src, llm_model], [gr_md, input, past_searches]).success(ask_llm, [input, llm_model, stream_results], output_text)
demo.queue().launch()