import urllib.request import fitz import re import numpy as np import tensorflow_hub as hub import openai import gradio as gr import os from sklearn.neighbors import NearestNeighbors from jina import Document, DocumentArray # Create a new DocumentArray for file storage doc_array = DocumentArray() def download_pdf(url, output_path): urllib.request.urlretrieve(url, output_path) def preprocess(text): text = text.replace('\n', ' ') text = re.sub('\s+', ' ', text) return text # Store a file in the DocumentArray def store_file_in_docarray(file_name, file_content): doc = Document(id=file_name, content=file_content) doc_array.append(doc) # Retrieve a file from the DocumentArray def get_file_from_docarray(file_name): for doc in doc_array: if doc.id == file_name: return doc.content return None def pdf_to_text(path, start_page=1, end_page=None): doc = fitz.open(path) total_pages = doc.page_count if end_page is None: end_page = total_pages text_list = [] for i in range(start_page-1, end_page): text = doc.load_page(i).get_text("text") text = preprocess(text) text_list.append(text) doc.close() return text_list def text_to_chunks(texts, file_names, word_length=150, start_page=1): text_toks = [t.split(' ') for t in texts] page_nums = [] chunks = [] total_pages = len(texts) for idx, words in enumerate(text_toks): current_file_idx = 0 current_page = idx + start_page for i, num_pages in enumerate(file_names.values()): if current_page > num_pages: current_page -= num_pages current_file_idx += 1 else: break for i in range(0, len(words), word_length): chunk = words[i:i+word_length] if (i+word_length) > len(words) and (len(chunk) < word_length) and ( len(text_toks) != (idx+1)): text_toks[idx+1] = chunk + text_toks[idx+1] continue chunk = ' '.join(chunk).strip() chunk = f'[{list(file_names.keys())[current_file_idx]}, Page no. {current_page}]' + ' ' + '"' + chunk + '"' chunks.append(chunk) return chunks class SemanticSearch: def __init__(self): self.use = hub.load('https://tfhub.dev/google/universal-sentence-encoder/4') self.fitted = False def fit(self, data, batch=1000, n_neighbors=5): self.data = data self.embeddings = self.get_text_embedding(data, batch=batch) n_neighbors = min(n_neighbors, len(self.embeddings)) self.nn = NearestNeighbors(n_neighbors=n_neighbors) self.nn.fit(self.embeddings) self.fitted = True def __call__(self, text, return_data=True): inp_emb = self.use([text]) neighbors = self.nn.kneighbors(inp_emb, return_distance=False)[0] if return_data: return [self.data[i] for i in neighbors] else: return neighbors def get_text_embedding(self, texts, batch=1000): embeddings = [] for i in range(0, len(texts), batch): text_batch = texts[i:(i+batch)] emb_batch = self.use(text_batch) embeddings.append(emb_batch) embeddings = np.vstack(embeddings) return embeddings def load_recommender(paths, start_page=1): global recommender all_texts = [] file_names = {} for path in paths: texts = pdf_to_text(path, start_page=start_page) all_texts.extend(texts) file_names[os.path.basename(path)] = len(texts) chunks = text_to_chunks(all_texts, file_names, start_page=start_page) recommender.fit(chunks) return 'Corpus Loaded.' def generate_text(openAI_key, prompt, engine="text-davinci-003"): openai.api_key = openAI_key completions = openai.Completion.create( engine=engine, prompt=prompt, max_tokens=512, n=1, stop=None, temperature=0.7, ) message = completions.choices[0].text return message def generate_answer(question, openAI_key): topn_chunks = recommender(question) prompt = "" prompt += 'search results:\n\n' for c in topn_chunks: prompt += c + '\n\n' prompt += "Instructions: Compose a comprehensive reply to the query using the search results given. "\ "Cite each reference using [PDF FILE NAME, PAGE NUMBER:] notation (every result has this number at the beginning). "\ "Citation should be done at the end of each sentence. If the search results mention multiple subjects "\ "with the same name, create separate answers for each. Only include information found in the results and "\ "don't add any additional information. Make sure the answer is correct and don't output false content. "\ "If the text does not relate to the query, simply state 'Text Not Found in PDF'. Ignore outlier "\ "search results which have nothing to do with the question. Only answer what is asked. The "\ "answer should be short and concise. Answer step-by-step. \n\nQuery: {question}\nAnswer: " prompt += f"Query: {question}\nAnswer:" answer = generate_text(openAI_key, prompt, "text-davinci-003") return answer def question_answer(url, files, question, openAI_key): if openAI_key.strip() == '': return '[ERROR]: Please enter your Open AI Key. Get your key here : https://platform.openai.com/account/api-keys' if url.strip() == '' and not files: return '[ERROR]: Both URL and PDF are empty. Provide at least one.' if url.strip() != '' and files: return '[ERROR]: Both URL and PDF are provided. Please provide only one (either URL or PDF).' pdf_paths = [] if url.strip() != '': glob_url = url output_path = 'corpus.pdf' download_pdf(glob_url, output_path) pdf_paths.append(output_path) else: for file in files: old_file_name = file.name file_name = file.name file_name = file_name[:-12] + file_name[-4:] os.rename(old_file_name, file_name) pdf_paths.append(file_name) if question.strip() == '': return '[ERROR]: Question field is empty' # Store the PDF content in the DocumentArray for pdf_path in pdf_paths: with open(pdf_path, "rb") as f: content = f.read() store_file_in_docarray(pdf_path, content) # Load the recommender load_recommender(pdf_paths) # Generate an answer return generate_answer(question, openAI_key) recommender = SemanticSearch() title = 'PDF GPT (Sandbox)' description = """ PDF GPT allows you to chat with your PDF file using Universal Sentence Encoder and Open AI. It gives hallucination free response than other tools as the embeddings are better than OpenAI. The returned response can even cite the page number and file name in square brackets([]) where the information is located, adding credibility to the responses and helping to locate pertinent information quickly.""" with gr.Blocks() as demo: gr.Markdown(f'

{title}

') gr.Markdown(description) with gr.Row(): with gr.Group(): gr.Markdown(f'

Get your Open AI API key here

') openAI_key = gr.Textbox(label='Enter your OpenAI API key here') url = gr.Textbox(label='Enter PDF URL here') gr.Markdown("

OR

") files = gr.File(label='Upload your PDF/ Research Paper / Book here', file_types=['.pdf'], file_count='multiple') question = gr.Textbox(label='Enter your question here') btn = gr.Button(value='Submit') btn.style(full_width=True) with gr.Group(): answer = gr.Textbox(label='The answer to your question is :') btn.click(question_answer, inputs=[url, files, question, openAI_key], outputs=[answer]) demo.launch()