├── LICENSE
├── README.md
├── generate_embedding.py
├── gif.gif
├── requirements.txt
├── run.py
├── static
├── css
│ └── styles.css
├── js
│ └── script.js
└── send-icon.png
└── templates
└── index.html
/LICENSE:
--------------------------------------------------------------------------------
1 | Apache License
2 | Version 2.0, January 2004
3 | http://www.apache.org/licenses/
4 |
5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6 |
7 | 1. Definitions.
8 |
9 | "License" shall mean the terms and conditions for use, reproduction,
10 | and distribution as defined by Sections 1 through 9 of this document.
11 |
12 | "Licensor" shall mean the copyright owner or entity authorized by
13 | the copyright owner that is granting the License.
14 |
15 | "Legal Entity" shall mean the union of the acting entity and all
16 | other entities that control, are controlled by, or are under common
17 | control with that entity. For the purposes of this definition,
18 | "control" means (i) the power, direct or indirect, to cause the
19 | direction or management of such entity, whether by contract or
20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the
21 | outstanding shares, or (iii) beneficial ownership of such entity.
22 |
23 | "You" (or "Your") shall mean an individual or Legal Entity
24 | exercising permissions granted by this License.
25 |
26 | "Source" form shall mean the preferred form for making modifications,
27 | including but not limited to software source code, documentation
28 | source, and configuration files.
29 |
30 | "Object" form shall mean any form resulting from mechanical
31 | transformation or translation of a Source form, including but
32 | not limited to compiled object code, generated documentation,
33 | and conversions to other media types.
34 |
35 | "Work" shall mean the work of authorship, whether in Source or
36 | Object form, made available under the License, as indicated by a
37 | copyright notice that is included in or attached to the work
38 | (an example is provided in the Appendix below).
39 |
40 | "Derivative Works" shall mean any work, whether in Source or Object
41 | form, that is based on (or derived from) the Work and for which the
42 | editorial revisions, annotations, elaborations, or other modifications
43 | represent, as a whole, an original work of authorship. For the purposes
44 | of this License, Derivative Works shall not include works that remain
45 | separable from, or merely link (or bind by name) to the interfaces of,
46 | the Work and Derivative Works thereof.
47 |
48 | "Contribution" shall mean any work of authorship, including
49 | the original version of the Work and any modifications or additions
50 | to that Work or Derivative Works thereof, that is intentionally
51 | submitted to Licensor for inclusion in the Work by the copyright owner
52 | or by an individual or Legal Entity authorized to submit on behalf of
53 | the copyright owner. For the purposes of this definition, "submitted"
54 | means any form of electronic, verbal, or written communication sent
55 | to the Licensor or its representatives, including but not limited to
56 | communication on electronic mailing lists, source code control systems,
57 | and issue tracking systems that are managed by, or on behalf of, the
58 | Licensor for the purpose of discussing and improving the Work, but
59 | excluding communication that is conspicuously marked or otherwise
60 | designated in writing by the copyright owner as "Not a Contribution."
61 |
62 | "Contributor" shall mean Licensor and any individual or Legal Entity
63 | on behalf of whom a Contribution has been received by Licensor and
64 | subsequently incorporated within the Work.
65 |
66 | 2. Grant of Copyright License. Subject to the terms and conditions of
67 | this License, each Contributor hereby grants to You a perpetual,
68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69 | copyright license to reproduce, prepare Derivative Works of,
70 | publicly display, publicly perform, sublicense, and distribute the
71 | Work and such Derivative Works in Source or Object form.
72 |
73 | 3. Grant of Patent License. Subject to the terms and conditions of
74 | this License, each Contributor hereby grants to You a perpetual,
75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76 | (except as stated in this section) patent license to make, have made,
77 | use, offer to sell, sell, import, and otherwise transfer the Work,
78 | where such license applies only to those patent claims licensable
79 | by such Contributor that are necessarily infringed by their
80 | Contribution(s) alone or by combination of their Contribution(s)
81 | with the Work to which such Contribution(s) was submitted. If You
82 | institute patent litigation against any entity (including a
83 | cross-claim or counterclaim in a lawsuit) alleging that the Work
84 | or a Contribution incorporated within the Work constitutes direct
85 | or contributory patent infringement, then any patent licenses
86 | granted to You under this License for that Work shall terminate
87 | as of the date such litigation is filed.
88 |
89 | 4. Redistribution. You may reproduce and distribute copies of the
90 | Work or Derivative Works thereof in any medium, with or without
91 | modifications, and in Source or Object form, provided that You
92 | meet the following conditions:
93 |
94 | (a) You must give any other recipients of the Work or
95 | Derivative Works a copy of this License; and
96 |
97 | (b) You must cause any modified files to carry prominent notices
98 | stating that You changed the files; and
99 |
100 | (c) You must retain, in the Source form of any Derivative Works
101 | that You distribute, all copyright, patent, trademark, and
102 | attribution notices from the Source form of the Work,
103 | excluding those notices that do not pertain to any part of
104 | the Derivative Works; and
105 |
106 | (d) If the Work includes a "NOTICE" text file as part of its
107 | distribution, then any Derivative Works that You distribute must
108 | include a readable copy of the attribution notices contained
109 | within such NOTICE file, excluding those notices that do not
110 | pertain to any part of the Derivative Works, in at least one
111 | of the following places: within a NOTICE text file distributed
112 | as part of the Derivative Works; within the Source form or
113 | documentation, if provided along with the Derivative Works; or,
114 | within a display generated by the Derivative Works, if and
115 | wherever such third-party notices normally appear. The contents
116 | of the NOTICE file are for informational purposes only and
117 | do not modify the License. You may add Your own attribution
118 | notices within Derivative Works that You distribute, alongside
119 | or as an addendum to the NOTICE text from the Work, provided
120 | that such additional attribution notices cannot be construed
121 | as modifying the License.
122 |
123 | You may add Your own copyright statement to Your modifications and
124 | may provide additional or different license terms and conditions
125 | for use, reproduction, or distribution of Your modifications, or
126 | for any such Derivative Works as a whole, provided Your use,
127 | reproduction, and distribution of the Work otherwise complies with
128 | the conditions stated in this License.
129 |
130 | 5. Submission of Contributions. Unless You explicitly state otherwise,
131 | any Contribution intentionally submitted for inclusion in the Work
132 | by You to the Licensor shall be under the terms and conditions of
133 | this License, without any additional terms or conditions.
134 | Notwithstanding the above, nothing herein shall supersede or modify
135 | the terms of any separate license agreement you may have executed
136 | with Licensor regarding such Contributions.
137 |
138 | 6. Trademarks. This License does not grant permission to use the trade
139 | names, trademarks, service marks, or product names of the Licensor,
140 | except as required for reasonable and customary use in describing the
141 | origin of the Work and reproducing the content of the NOTICE file.
142 |
143 | 7. Disclaimer of Warranty. Unless required by applicable law or
144 | agreed to in writing, Licensor provides the Work (and each
145 | Contributor provides its Contributions) on an "AS IS" BASIS,
146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147 | implied, including, without limitation, any warranties or conditions
148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149 | PARTICULAR PURPOSE. You are solely responsible for determining the
150 | appropriateness of using or redistributing the Work and assume any
151 | risks associated with Your exercise of permissions under this License.
152 |
153 | 8. Limitation of Liability. In no event and under no legal theory,
154 | whether in tort (including negligence), contract, or otherwise,
155 | unless required by applicable law (such as deliberate and grossly
156 | negligent acts) or agreed to in writing, shall any Contributor be
157 | liable to You for damages, including any direct, indirect, special,
158 | incidental, or consequential damages of any character arising as a
159 | result of this License or out of the use or inability to use the
160 | Work (including but not limited to damages for loss of goodwill,
161 | work stoppage, computer failure or malfunction, or any and all
162 | other commercial damages or losses), even if such Contributor
163 | has been advised of the possibility of such damages.
164 |
165 | 9. Accepting Warranty or Additional Liability. While redistributing
166 | the Work or Derivative Works thereof, You may choose to offer,
167 | and charge a fee for, acceptance of support, warranty, indemnity,
168 | or other liability obligations and/or rights consistent with this
169 | License. However, in accepting such obligations, You may act only
170 | on Your own behalf and on Your sole responsibility, not on behalf
171 | of any other Contributor, and only if You agree to indemnify,
172 | defend, and hold each Contributor harmless for any liability
173 | incurred by, or claims asserted against, such Contributor by reason
174 | of your accepting any such warranty or additional liability.
175 |
176 | END OF TERMS AND CONDITIONS
177 |
178 | APPENDIX: How to apply the Apache License to your work.
179 |
180 | To apply the Apache License to your work, attach the following
181 | boilerplate notice, with the fields enclosed by brackets "[]"
182 | replaced with your own identifying information. (Don't include
183 | the brackets!) The text should be enclosed in the appropriate
184 | comment syntax for the file format. We also recommend that a
185 | file or class name and description of purpose be included on the
186 | same "printed page" as the copyright notice for easier
187 | identification within third-party archives.
188 |
189 | Copyright [yyyy] [name of copyright owner]
190 |
191 | Licensed under the Apache License, Version 2.0 (the "License");
192 | you may not use this file except in compliance with the License.
193 | You may obtain a copy of the License at
194 |
195 | http://www.apache.org/licenses/LICENSE-2.0
196 |
197 | Unless required by applicable law or agreed to in writing, software
198 | distributed under the License is distributed on an "AS IS" BASIS,
199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200 | See the License for the specific language governing permissions and
201 | limitations under the License.
202 |
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # ChatPDFLike
2 |
3 | An end-to-end document question-answering application using large language model APIs.
4 |
5 | **Note**: This project is not affiliated with or endorsed by [ChatPDF](https://www.chatpdf.com/). This is an independent project attempting to replicate similar functionality.
6 |
7 | ## Overview
8 |
9 | ChatPDF-Like is a web application that allows users to upload PDF documents and interact with them using natural language queries. The application leverages large language models (LLMs) like OpenAI's GPT-3.5 Turbo to understand the content of the PDF and provide concise and accurate answers to user questions.
10 |
11 | ## Features
12 |
13 | - **PDF Document Upload**: Upload local PDF files or provide a URL to a PDF document.
14 | - **Natural Language Interaction**: Ask questions about the content of the PDF in natural language.
15 | - **Relevant Answers**: Receive concise answers based on the content of the document.
16 | - **Source References**: View sources (sections of the PDF) that were used to generate the answer.
17 | - **Multiple LLM Providers**: Support for both OpenAI and Ollama models.
18 | - **Web Interface**: Simple and intuitive web interface built with Flask and JavaScript.
19 |
20 | ## How It Works
21 |
22 | The application follows these main steps:
23 |
24 | 1. **Text Extraction and Processing**:
25 | - The PDF is parsed using `PyPDF2`.
26 | - Text is extracted from each page, and large pieces of text are split into manageable chunks.
27 |
28 | 2. **Embedding Generation**:
29 | - For each text chunk, an embedding vector is generated using the selected embedding model (e.g., OpenAI's `text-embedding-ada-002`).
30 | - These embeddings represent the semantic meaning of the text chunks and are stored for similarity calculations.
31 |
32 | 3. **User Query Handling**:
33 | - When a user asks a question, an embedding vector for the query is generated using the same embedding model.
34 |
35 | 4. **Similarity Search**:
36 | - The application computes the cosine similarity between the query embedding and the text chunk embeddings.
37 | - The most relevant text chunks are selected based on the highest similarity scores.
38 |
39 | 5. **Prompt Construction**:
40 | - A prompt is created for the language model, incorporating the user's question and the most relevant text chunks.
41 |
42 | 6. **Answer Generation**:
43 | - The prompt is sent to the language model (e.g., OpenAI's GPT-3.5 Turbo).
44 | - The model generates an answer to the user's question based on the provided context.
45 |
46 | 7. **Response Display**:
47 | - The answer is displayed to the user in the web interface.
48 | - References to the source text chunks are also provided for transparency.
49 |
50 | ## Getting Started
51 |
52 | ### Prerequisites
53 |
54 | - **Python**: Version 3.6 or higher is required.
55 | - **API Keys**:
56 | - **OpenAI API Key**: Required to use OpenAI's models for embeddings and answer generation.
57 | - **Ollama API Key**: Optional. Required if you want to use Ollama models.
58 |
59 | ### Installation
60 |
61 | 1. **Clone the Repository**
62 |
63 | ```bash
64 | git clone https://github.com/Ulov888/chatpdflike.git
65 | cd chatpdflike
66 | ```
67 |
68 | 2. **Install Dependencies**
69 |
70 | Using `pip`, install the required packages:
71 |
72 | ```bash
73 | pip install -r requirements.txt
74 | ```
75 |
76 | ### API Keys
77 |
78 | To use OpenAI's API:
79 |
80 | 1. Sign up for an API key at [OpenAI](https://platform.openai.com/account/api-keys).
81 | 2. Set the `OPENAI_API_KEY` environment variable:
82 |
83 | ```bash
84 | export OPENAI_API_KEY="your_openai_api_key"
85 | ```
86 |
87 | To use Ollama's API (if desired):
88 |
89 | 1. Obtain an API key from Ollama.
90 | 2. Set the `OLLAMA_API_KEY` environment variable:
91 |
92 | ```bash
93 | export OLLAMA_API_KEY="your_ollama_api_key"
94 | ```
95 |
96 | ## Usage
97 |
98 | 1. **Start the Application**
99 |
100 | Run the Flask application:
101 |
102 | ```bash
103 | python run.py
104 | ```
105 |
106 | By default, the server runs on `http://0.0.0.0:8080`.
107 |
108 | 2. **Access the Web Interface**
109 |
110 | Open a web browser and navigate to `http://localhost:8080`.
111 |
112 | 3. **Upload a PDF Document**
113 |
114 | You can either:
115 |
116 | - Click on "Upload PDF" to select and upload a PDF file from your computer.
117 | - Enter a URL to a PDF document and click "Submit".
118 |
119 | 4. **Interact with the PDF**
120 |
121 | - Once the PDF is processed, you can ask questions about its content using the chat interface on the right side of the screen.
122 | - Type your question in the input box and press "Send".
123 |
124 | 5. **View Answers**
125 |
126 | - The application's response will appear below your question.
127 | - Source references (e.g., page numbers and excerpts) are provided for context.
128 |
129 | 
130 |
131 | ## Customization
132 |
133 | ### Prompt Strategies
134 |
135 | The behavior of the language model can be customized by modifying the prompt strategies in `generate_embedding.py`, specifically in the `create_prompt` method of the `Chatbot` class.
136 |
137 | Strategies include:
138 |
139 | - **Paper**: For summarizing scientific papers.
140 | - **Handbook**: For summarizing financial handbooks (answers in Chinese).
141 | - **Contract**: For understanding contracts (answers in Chinese).
142 | - **Default**: General-purpose strategy (answers in Chinese).
143 |
144 | To select a strategy, you can modify the `strategy` parameter when calling `create_prompt`.
145 |
146 | ### Language and Output
147 |
148 | The application is currently configured to provide answers in Chinese for some strategies. You can modify the prompts to change the language or adjust the behavior of the model.
149 |
150 | ## Limitations
151 |
152 | - **OpenAI API Costs**: Using OpenAI's API will incur costs based on usage. Make sure to monitor your API usage to avoid unexpected charges.
153 | - **PDF Parsing**: The application uses `PyPDF2`, which may not handle all PDFs perfectly. Complex PDFs with unusual formatting may not parse correctly.
154 | - **Embedding Limits**: The maximum token limit for embeddings may restrict the size of text chunks or the maximum length of the prompt.
155 | - **Model Responses**: The quality and accuracy of the answers depend on the performance of the language model and the relevance of the retrieved text chunks.
156 |
157 | ## Contributing
158 |
159 | Contributions are welcome! If you have any suggestions or improvements, feel free to submit an issue or pull request.
160 |
161 | ## License
162 |
163 | This project is licensed under the [Apache License](LICENSE).
164 |
--------------------------------------------------------------------------------
/generate_embedding.py:
--------------------------------------------------------------------------------
1 | import logging as logger
2 | import ollama
3 | import openai
4 | import os
5 | import os
6 | import pandas as pd
7 | from flask_cors import CORS
8 | from openai.embeddings_utils import get_embedding, cosine_similarity
9 |
10 | openai.api_key = os.getenv('OPENAI_API_KEY')
11 |
12 | class Chatbot():
13 | def parse_paper(self, pdf):
14 | logger.info("Parsing paper")
15 | number_of_pages = len(pdf.pages)
16 | logger.info(f"Total number of pages: {number_of_pages}")
17 | paper_text = []
18 | for i in range(number_of_pages):
19 | page = pdf.pages[i]
20 | page_text = []
21 |
22 | def visitor_body(text, cm, tm, fontDict, fontSize):
23 | x = tm[4]
24 | y = tm[5]
25 | # ignore header/footer
26 | if (y > 50 and y < 720) and (len(text.strip()) > 1):
27 | page_text.append({
28 | 'fontsize': fontSize,
29 | 'text': text.strip().replace('\x03', ''),
30 | 'x': x,
31 | 'y': y
32 | })
33 |
34 | _ = page.extract_text(visitor_text=visitor_body)
35 |
36 | blob_font_size = None
37 | blob_text = ''
38 | processed_text = []
39 |
40 | for t in page_text:
41 | if t['fontsize'] == blob_font_size:
42 | blob_text += f" {t['text']}"
43 | if len(blob_text) >= 2000:
44 | processed_text.append({
45 | 'fontsize': blob_font_size,
46 | 'text': blob_text,
47 | 'page': i
48 | })
49 | blob_font_size = None
50 | blob_text = ''
51 | else:
52 | if blob_font_size is not None and len(blob_text) >= 1:
53 | processed_text.append({
54 | 'fontsize': blob_font_size,
55 | 'text': blob_text,
56 | 'page': i
57 | })
58 | blob_font_size = t['fontsize']
59 | blob_text = t['text']
60 | paper_text += processed_text
61 | logger.info("Done parsing paper")
62 | return paper_text
63 |
64 | def paper_df(self, pdf):
65 | logger.info('Creating dataframe')
66 | filtered_pdf= []
67 | for row in pdf:
68 | if len(row['text']) < 30:
69 | continue
70 | if len(row['text']) > 8000:
71 | row['text'] = row['text'][:8000]
72 | filtered_pdf.append(row)
73 | df = pd.DataFrame(filtered_pdf)
74 | # remove elements with identical df[text] and df[page] values
75 | df = df.drop_duplicates(subset=['text', 'page'], keep='first')
76 | df['length'] = df['text'].apply(lambda x: len(x))
77 | logger.info('Done creating df')
78 | return df
79 |
80 | def calculate_embeddings(self, df):
81 | logger.info('Calculating embeddings')
82 | embedding_model = "text-embedding-ada-002"
83 | embeddings = df.text.apply([lambda x: get_embedding(x, engine=embedding_model)])
84 | df["embeddings"] = embeddings
85 | logger.info('Done calculating embeddings')
86 | return df
87 |
88 | def search_embeddings(self, df, query, n=2, pprint=True):
89 | query_embedding = get_embedding(
90 | query,
91 | engine="text-embedding-ada-002"
92 | )
93 | df["similarity"] = df.embeddings.apply(lambda x: cosine_similarity(x, query_embedding))
94 |
95 | results = df.sort_values("similarity", ascending=False, ignore_index=True)
96 | results = results.head(n)
97 | global sources
98 | sources = []
99 | for i in range(n):
100 | # append the page number and the text as a dict to the sources list
101 | sources.append({'Page '+str(results.iloc[i]['page']): results.iloc[i]['text'][:150]+'...'})
102 | print(sources)
103 | return results.head(n)
104 |
105 | def create_prompt(self, df, user_input, strategy=None):
106 | result = self.search_embeddings(df, user_input)
107 | if strategy == "paper":
108 | prompt = """You are a large language model whose expertise is reading and summarizing scientific papers.
109 | You are given a query and a series of text embeddings from a paper in order of their cosine similarity to the query.
110 | You must take the given embeddings and return a very detailed summary of the paper that answers the query.
111 | Given the question: """+ user_input + """
112 |
113 | and the following embeddings as data:
114 |
115 | 1.""" + str(result.iloc[0]['text']) + """
116 | 2.""" + str(result.iloc[1]['text']) + """
117 |
118 | Return a concise and accurate answer:"""
119 | elif strategy == "handbook":
120 | prompt = """You are a large language model whose expertise is reading and summarizing financial handbook.
121 | You are given a query and a series of text embeddings from a handbook in order of their cosine similarity to the query.
122 | You must take the given embeddings and return a very detailed answer in Chinese of the handbook that answers the query.
123 | If not necessary, your answer please use the original text as much as possible.
124 | You should also ensure that your response is written in clear and concise Chinese, using appropriate grammar and vocabulary.
125 | Additionally, your response should focus on answering the specific query provided..
126 | Given the question: """+ user_input + """
127 | and the following embeddings as data:
128 |
129 | 1.""" + str(result.iloc[0]['text']) + """
130 | 2.""" + str(result.iloc[1]['text']) + """
131 |
132 | Return a concise and accurate answer:"""
133 | elif strategy == "contract":
134 | prompt = """As a large language model specializing in reading and summarizing, your task is to read a query and a sequence of text inputs sorted by their cosine similarity to the query.
135 | Your goal is to provide a Chinese answer to the query using the given padding. If possible, please use the original text of your answer.
136 | Please ensure that your response adheres to the terms of the agreement. Your response should focus on addressing the specific query provided,
137 | providing relevant information and details based on the input texts' content. You should also strive for clarity and conciseness in your response,
138 | summarizing key points while maintaining accuracy and relevance. Please note that you should prioritize understanding the context and meaning
139 | behind both the query and input texts before generating a response.
140 | Given the question: """+ user_input + """
141 | and the following embeddings as data:
142 |
143 | 1.""" + str(result.iloc[0]['text']) + """
144 | 2.""" + str(result.iloc[1]['text']) + """
145 |
146 | Return a concise and accurate answer:"""
147 | else:
148 | prompt = """As a language model specialized in reading and summarizing documents, your task is to provide a concise answer in Chinese based on a given query and a series of text embeddings from the document.
149 | The embeddings are provided in order of their cosine similarity to the query. Your response should use as much original text as possible.
150 | Your answer should be highly concise and accurate, providing relevant information that directly answers the query.
151 | You should also ensure that your response is written in clear and concise Chinese, using appropriate grammar and vocabulary.
152 | Please note that you must use the provided text embeddings to generate your response, which means you will need to understand how they relate to the original document.
153 | Additionally, your response should focus on answering the specific query provided..
154 | Given the question: """+ user_input + """
155 |
156 | and the following embeddings as data:
157 |
158 | 1.""" + str(result.iloc[0]['text']) + """
159 | 2.""" + str(result.iloc[1]['text']) + """
160 |
161 | Return a concise and accurate answer:"""
162 | logger.info('Done creating prompt')
163 | return prompt
164 |
165 | def response(self, df, prompt):
166 | logger.info('Sending request to GPT-3')
167 | prompt = self.create_prompt(df, prompt)
168 | r = openai.ChatCompletion.create(model="gpt-3.5-turbo",
169 | messages=[{"role": "user", "content": prompt},
170 | ])
171 | answer = r.choices[0]['message']['content']
172 | logger.info('Done sending request to GPT-3')
173 | response = {'answer': answer, 'sources': sources}
174 | return response
175 |
176 |
177 | class OllamaChatbot(Chatbot):
178 |
179 | def __init__(self):
180 | self.ollama_api_key = os.getenv('OLLAMA_API_KEY')
181 | ollama.api_key = self.ollama_api_key
182 |
183 | def get_ollama_embedding(self, text):
184 | response = ollama.embed(model='llama3.1', input=text)
185 | return response['embedding']
186 |
187 | def calculate_ollama_embeddings(self, df):
188 | logger.info('Calculating embeddings using Ollama')
189 | embeddings = df.text.apply(lambda x: self.get_ollama_embedding(x))
190 | df["embeddings"] = embeddings
191 | logger.info('Done calculating embeddings')
192 | return df
193 |
194 | def ollama_response(self, df, prompt):
195 | logger.info('Sending request to Ollama')
196 | response = ollama.chat(model='llama3.1', messages=[{"role": "user", "content": prompt}])
197 | answer = response['message']['content']
198 | logger.info('Done sending request to Ollama')
199 | response = {'answer': answer, 'sources': sources}
200 | return response
201 |
--------------------------------------------------------------------------------
/gif.gif:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Ulov888/chatpdflike/26dedce74609d48eccc80bede66fe9cc3c871f89/gif.gif
--------------------------------------------------------------------------------
/requirements.txt:
--------------------------------------------------------------------------------
1 | flask
2 | PyPDF2
3 | pandas
4 | openai
5 | requests
6 | flask-cors
7 | matplotlib
8 | scipy
9 | plotly
10 | google-cloud-storage
11 | gunicorn==20.1.0
12 | scikit-learn==0.24.1
13 | ollama
14 |
--------------------------------------------------------------------------------
/run.py:
--------------------------------------------------------------------------------
1 | from flask import Flask, request, render_template
2 | from flask_cors import CORS
3 | from generate_embedding import Chatbot, OllamaChatbot
4 | from io import BytesIO
5 | from PyPDF2 import PdfReader
6 | import requests
7 |
8 |
9 | app = Flask(__name__)
10 | CORS(app)
11 |
12 | @app.route("/", methods=["GET", "POST"])
13 | def index():
14 | return render_template("index.html")
15 |
16 |
17 | @app.route("/process_pdf", methods=['POST'])
18 | def process_pdf():
19 | print("Processing pdf")
20 | file = request.data
21 | pdf = PdfReader(BytesIO(file))
22 | chatbot = Chatbot() # Default to OpenAI
23 | if request.args.get('provider') == 'ollama':
24 | chatbot = OllamaChatbot()
25 | paper_text = chatbot.parse_paper(pdf)
26 | global df
27 | df = chatbot.paper_df(paper_text)
28 | df = chatbot.calculate_embeddings(df)
29 | print("Done processing pdf")
30 | return {'answer': ''}
31 |
32 |
33 | @app.route("/download_pdf", methods=['POST'])
34 | def download_pdf():
35 | chatbot = Chatbot() # Default to OpenAI
36 | if request.args.get('provider') == 'ollama':
37 | chatbot = OllamaChatbot()
38 | url = request.json['url']
39 | r = requests.get(str(url))
40 | print(r.headers)
41 | pdf = PdfReader(BytesIO(r.content))
42 | paper_text = chatbot.parse_paper(pdf)
43 | global df
44 | df = chatbot.paper_df(paper_text)
45 | df = chatbot.calculate_embeddings(df)
46 | print("Done processing pdf")
47 | return {'key': ''}
48 |
49 |
50 | @app.route("/reply", methods=['POST'])
51 | def reply():
52 | chatbot = Chatbot() # Default to OpenAI
53 | if request.args.get('provider') == 'ollama':
54 | chatbot = OllamaChatbot()
55 | query = request.json['query']
56 | query = str(query)
57 | prompt = chatbot.create_prompt(df, query)
58 | response = chatbot.response(df, prompt)
59 | print(response)
60 | return response, 200
61 |
62 |
63 | if __name__ == '__main__':
64 | app.run(host='0.0.0.0', port=8080, debug=True)
65 |
--------------------------------------------------------------------------------
/static/css/styles.css:
--------------------------------------------------------------------------------
1 | html {
2 | height: 100%;
3 | position: fixed;
4 | }
5 |
6 | body {
7 | margin: 0;
8 | position: fixed;
9 | background-color: white;
10 | display: flex;
11 | flex-direction: column;
12 | align-items: center;
13 | height: 100%;
14 | width: 100%;
15 | }
16 |
17 | .upload-btn {
18 | display: inline-block;
19 | padding: 12px 30px;
20 | background-color: #4285F4;
21 | color: #fff;
22 | font-size: 16px;
23 | font-weight: 500;
24 | border-radius: 4px;
25 | cursor: pointer;
26 | transition: background-color 0.3s ease;
27 | }
28 |
29 | input[type="file"] {
30 | display: none;
31 | }
32 |
33 | form {
34 | margin: auto 0;
35 | }
36 |
37 | input[type="text"] {
38 | border: none;
39 | outline: none;
40 | }
41 |
42 | #container {
43 | display: none;
44 | width: 100%;
45 | height: 100vh;
46 | flex-direction: row;
47 | }
48 |
49 | .pdf-viewer {
50 | width: 100%;
51 | height: 100vh;
52 | display: none;
53 | }
54 |
55 | #chat {
56 | width: 100%;
57 | }
58 |
59 | #chat p {
60 | color: white;
61 | margin: 10px;
62 | padding: 10px;
63 | border-radius: 10px;
64 | width: fit-content;
65 | text-align: left;
66 | word-break: break-all;
67 | font-family: Roboto;
68 | font-size: 18px;
69 | font-weight: lighter;
70 | }
71 | .search-container {
72 | display: flex;
73 | align-items: center;
74 | /*margin: 0 auto;*/
75 | max-width: 600px;
76 | height: 50px;
77 | border: 2px solid #ccc;
78 | border-radius: 25px;
79 | overflow: hidden;
80 | padding: 0 10px;
81 | background-color: #fff;
82 | box-shadow: 0px 2px 5px rgba(0, 0, 0, 0.2);
83 | margin-top: 150px;
84 | margin-bottom: 20px;
85 | }
86 |
87 | .search-container input[type="text"] {
88 | flex: 1;
89 | border: none;
90 | font-size: 18px;
91 | height: 100%;
92 | padding: 0 15px;
93 | background-color: transparent;
94 | color: #666;
95 | font-weight: bold;
96 | outline: none;
97 | }
98 |
99 | .search-container input[type="text"]::placeholder {
100 | color: #ccc;
101 | font-weight: normal;
102 | }
103 |
104 | .search-container button {
105 | border: none;
106 | background-color: #4285F4;
107 | color: #fff;
108 | font-size: 18px;
109 | cursor: pointer;
110 | padding: 10px 15px;
111 | border-radius: 25px;
112 | box-shadow: 0px 2px 5px rgba(0, 0, 0, 0.2);
113 | transition: background-color 0.2s ease-in-out;
114 | }
115 |
116 | .search-container button:hover {
117 | background-color: #2962FF;
118 | }
119 |
120 | @media only screen and (max-width: 600px) {
121 | .search-container {
122 | max-width: 100%;
123 | margin: 0 10px;
124 | height: auto;
125 | border-radius: 0;
126 | box-shadow: none;
127 | }
128 |
129 | .search-container input[type="text"] {
130 | font-size: 16px;
131 | padding: 10px 15px;
132 | }
133 |
134 | .search-container button {
135 | font-size: 16px;
136 | padding: 10px 12px;
137 | border-radius: 20px;
138 | }
139 | }
140 |
141 |
142 |
143 | .file-upload {
144 | display: inline-block;
145 | padding: 12px 30px;
146 | background-color: #4285F4;
147 | color: #fff;
148 | font-size: 16px;
149 | font-weight: 500;
150 | border-radius: 4px;
151 | cursor: pointer;
152 | transition: background-color 0.3s ease;
153 | }
154 |
155 | .file-upload:hover {
156 | background-color: #3367D6;
157 | }
158 |
159 |
--------------------------------------------------------------------------------
/static/js/script.js:
--------------------------------------------------------------------------------
1 | document.addEventListener("DOMContentLoaded", function() {
2 | // This file contains the JavaScript code for the web app
3 |
4 | const input = document.querySelector("input[type='file']");
5 | var uploadBtn = document.querySelector(".upload-btn");
6 | const viewer = document.querySelector("#pdf-viewer");
7 | const container = document.querySelector("#container");
8 | var x = document.querySelector("input[name='pdf-url']");
9 | const form = document.querySelector("form");
10 | const p = document.querySelector("p");
11 | const up = document.querySelector("#up");
12 | const y = document.querySelector("#url");
13 | const send = document.querySelector("#send");
14 |
15 |
16 | send.addEventListener("click", function(event) {
17 | event.preventDefault();
18 | const message = document.querySelector("input[name='chat']").value;
19 | // if the message is empty, do nothing
20 | if (message === "") {
21 | return;
22 | }
23 | const chat = document.querySelector("#chat");
24 | const query = document.createElement("p");
25 | query.innerHTML = message;
26 | chat.appendChild(query);
27 |
28 | const loading = document.createElement("p");
29 | loading.style.color = "lightgray";
30 | loading.style.fontSize = "14px";
31 | loading.innerHTML = "Loading...";
32 | chat.appendChild(loading);
33 |
34 | // call the endpoint /reply with the message and get the reply.
35 | fetch('/reply', {
36 | method: 'POST',
37 | body: JSON.stringify({'query': message, 'key': window.key}),
38 | headers: {
39 | 'Content-Type': 'application/json'
40 | }
41 | })
42 | .then(response => response.json())
43 | // Append the reply to #chat as a simple paragraph without any styling
44 | .then(data => {
45 | console.log(data.answer);
46 | chat.removeChild(loading);
47 |
48 | const reply = document.createElement("p");
49 | reply.style.color = "lightgray";
50 | reply.style.marginBottom = "0px";
51 | reply.style.paddingTop = "0px";
52 | reply.innerHTML = data.answer;
53 | chat.appendChild(reply);
54 | chat.scrollTop = chat.scrollHeight;
55 |
56 | const sources = data.sources;
57 | console.log(sources)
58 | // console.log(typeof JSON.parse(sources))
59 | sources.forEach(function(source) {
60 | for (var page in source) {
61 | var p = document.createElement("p");
62 | p.style.color = "gray";
63 | p.style.fontSize = "12px";
64 | p.style.fontWeight = "bold";
65 | p.style.marginTop = "0px";
66 | p.style.marginBottom = "0px";
67 | p.style.paddingTop = "0px";
68 | p.style.paddingBottom = "5px";
69 | p.innerHTML = page + ": " + "'"+source[page];+"'"
70 | chat.appendChild(p);
71 | }
72 | });
73 | })
74 | .catch(error => {
75 | chat.removeChild(loading);
76 | console.error(error);
77 |
78 | const errorMessage = document.createElement("p");
79 | errorMessage.style.color = "red";
80 | errorMessage.style.marginBottom = "0px";
81 | errorMessage.style.paddingTop = "0px";
82 | errorMessage.innerHTML = "Error: Request to OpenAI failed. Please try again.";
83 | chat.appendChild(errorMessage);
84 | chat.scrollTop = chat.scrollHeight;
85 | });
86 | document.querySelector("input[name='chat']").value = "";
87 | });
88 |
89 | x.addEventListener("focus", function() {
90 | if (this.value === "Enter URL") {
91 | this.value = "";
92 | this.style.color = "black";
93 | }
94 | });
95 |
96 | y.addEventListener("submit", function(event) {
97 | event.preventDefault();
98 | const url = this.elements["pdf-url"].value;
99 | if (url === "") {
100 | return;
101 | }
102 | // if the url does not end with .pdf, make x.value = "Error: URL does not end with .pdf"
103 | if (!url.endsWith(".pdf")) {
104 | x.value = "Error: URL does not end with .pdf";
105 | return;
106 | }
107 | x.value = "Loading...";
108 | console.log(url);
109 | fetch(url)
110 | .then(response => response.blob())
111 | .then(pdfBlob => {
112 | console.log(pdfBlob);
113 | const pdfUrl = URL.createObjectURL(pdfBlob);
114 | pdfjsLib.getDocument(pdfUrl).promise.then(pdfDoc => {
115 | viewer.src = pdfUrl;
116 | uploadBtn.style.display = "none";
117 | form.style.display = "none";
118 | form.style.marginTop = "0px";
119 | p.style.display = "none";
120 | up.style.display = "none";
121 | container.style.display = "flex";
122 | viewer.style.display = "block";
123 | });
124 | })
125 | .catch(error => {
126 | console.error(error);
127 | });
128 | var loading = document.createElement("p");
129 | loading.style.color = "lightgray";
130 | loading.style.fontSize = "14px";
131 | loading.innerHTML = "Calculating embeddings...";
132 | chat.appendChild(loading);
133 |
134 | // Make a POST request to the server 'myserver/download-pdf' with the URL
135 | fetch('/download_pdf', {
136 | method: 'POST',
137 | body: JSON.stringify({'url': url}),
138 | headers: {
139 | 'Content-Type': 'application/json',
140 | 'Access-Control-Allow-Origin': '*',
141 | 'Access-Control-Allow-Methods': 'GET, POST, PUT, DELETE, PATCH, OPTIONS',
142 | 'Access-Control-Allow-Headers': 'Content-Type, Authorization'
143 | }
144 | })
145 | .then(response => response.json())
146 | // Append the reply to #chat as a simple paragraph without any styling
147 | .then(data => {
148 | chat.removeChild(loading);
149 | window.key = data.key;
150 | })
151 | .catch(error => {
152 | uploadBtn.innerHTML = "Error: Request to server failed. Please try again. Check the URL if there is https:// at the beginning. If not, add it.";
153 | x.innerHTML = "Error: Request to server failed. Please try again. Check the URL if there is https:// at the beginning. If not, add it.";
154 | console.error(error);
155 | });
156 | });
157 |
158 | input.addEventListener("change", async function() {
159 | const file = this.files[0];
160 | const fileArrayBuffer = await file.arrayBuffer();
161 | console.log(fileArrayBuffer);
162 |
163 | var loading = document.createElement("p");
164 | loading.style.color = "lightgray";
165 | loading.style.fontSize = "14px";
166 | loading.innerHTML = "Calculating embeddings...";
167 | chat.appendChild(loading);
168 |
169 | // Make a post request to /process_pdf with the file
170 | fetch('/process_pdf', {
171 | method: 'POST',
172 | body: fileArrayBuffer,
173 | headers: {
174 | 'Content-Type': 'application/pdf',
175 | 'Content-Length': fileArrayBuffer.byteLength,
176 | 'Access-Control-Allow-Origin': '*',
177 | 'Access-Control-Allow-Methods': 'GET, POST, PUT, DELETE, PATCH, OPTIONS',
178 | 'Access-Control-Allow-Headers': 'Content-Type, Authorization'
179 | }
180 | })
181 | .then(response => response.json())
182 | // Append the reply to #chat as a simple paragraph without any styling
183 | .then(data => {
184 | chat.removeChild(loading);
185 | window.key = data.key;
186 | })
187 | .catch(error => {
188 | loading.innerHTML = "Error: Processing the pdf failed due to excess load. Please try again later. Check the URL if there is https:// at the beginning. If not, add it.";
189 | console.error(error);
190 | });
191 |
192 | pdfjsLib.getDocument(fileArrayBuffer).promise.then(pdfDoc => {
193 | viewer.src = URL.createObjectURL(file);
194 | uploadBtn.style.display = "none";
195 | form.style.display = "none";
196 | form.style.marginTop = "0px";
197 | p.style.display = "none";
198 | up.style.display = "none";
199 | container.style.display = "flex";
200 | viewer.style.display = "block";
201 | }).catch(error => {
202 | console.error(error);
203 | });
204 | });
205 | });
206 |
--------------------------------------------------------------------------------
/static/send-icon.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/Ulov888/chatpdflike/26dedce74609d48eccc80bede66fe9cc3c871f89/static/send-icon.png
--------------------------------------------------------------------------------
/templates/index.html:
--------------------------------------------------------------------------------
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
13 |
14 |