CheatGPT

This project demonstrates the integration of multiple NLP models and databases in a streamlit application. It uses OpenAI's GPT-3 to answer questions based on provided contexts. The contexts are in the form of text extracted from PDF files uploaded by the user. The application utilizes the following libraries:

• SentenceTransformer: This library is used to encode text input into 384 dimensional embeddings.
• Streamlit: This is a web framework for creating and sharing data-driven applications.
• IPython.display: This library is used to display audio files in the streamlit application.
• gTTS: This library is used to convert text to speech.
• audio_recorder_streamlit: This library is used to record audio in the streamlit application.
• speech_recognition: This library is used to transcribe speech.
• wave: This library is used to save audio files in .wav format.
• io: This library is used to load audio files into memory.
• Pinecone: This library is used to create and query a vector index for text embeddings.
• requests: This library is used to send HTTP requests.
• PyPDF2: This library is used to extract text from PDF files.
• pytesseract: This library is used to OCR (Optical Character Recognition) image files.
• Pillow: This library is used to load images.

Usage

Before running the application, you will need to create a .env file in the root directory of the project with the following keys:

• PINECONE_KEY: API key for Pinecone.
• GPT_KEY: API key for OpenAI GPT-3.

Once you have created the .env file, you can run the application by running the following command in your terminal:

1. Clone the repository:

git clone https://github.com/adi611/The-CheatGPT.git

2. Navigate to the project directory:

cd The-CheatGPT

3. Install the required dependencies:

pip install -r requirements.txt

4. Run the Streamlit app:

streamlit run app.py

Follow the instructions in the Streamlit app to interact with the different features.

Features

Text to Speech

This section allows you to enter text and convert it to speech. The text is converted to an audio file using the gTTS library and played back using the IPython.display library.

Speech to Text

This section allows you to record audio and transcribe it to text using the speech_recognition library. The audio is recorded using the audio_recorder_streamlit library and saved to a .wav file using the wave library. The text is transcribed using the Google Speech Recognition API.

OCR

This section allows you to upload an image file and extract text using OCR. The image file is loaded using the Pillow library and the text is extracted using the pytesseract library.

Question Answering

This section allows you to ask a question and receive an answer from a pre-built corpus of text. The corpus is indexed using the Pinecone library and the embeddings are generated using the SentenceTransformer library. The question is then answered using the OpenAI GPT-3 API.

Database Update

This section allows you to upload a PDF file and update the pre-built corpus of text. The PDF file is loaded using the PyPDF2 library and the text is extracted from each page. The text is then split into chunks and indexed using the Pinecone library.

Libraries used

• SentenceTransformers
• Streamlit
• IPython
• gTTS
• audio-recorder-streamlit
• SpeechRecognition
• PyPDF2
• Pillow
• dotenv
• Pinecone
• requests
• pytesseract

License

This project is licensed under the MIT License - see the LICENSE.md file for details.