requirements.txt

# Accelerate provides a simple API allowing you to speedup your machine learning and deep learning pipelines 
# using PyTorch while maintaining an identical interface to your models and data loading methods
accelerate

# Appdirs is a Python module for determining appropriate platform-specific directories
appdirs

# BitsAndBytes is an efficient Pytorch library for quantized neural network training
bitsandbytes

# Black is a code formatter for Python. It reformats your Python code to make it more readable and consistent with the PEP 8 style guide.
black

# This provides Jupyter support for the Black code formatter.
black[jupyter]

# Datasets is a lightweight and extensible library to easily share and access datasets and evaluation metrics for Natural Language Processing (NLP)
datasets

# Fire is a library for automatically generating command line interfaces (CLIs) from absolutely any Python object
fire

# This is a PEFT (Performance Estimation of Fine-Tuning) library that provides an easy-to-use tool for computing the cost of fine-tuning pre-trained models in terms of CO2 emissions, time and money.
git+https://github.com/huggingface/peft.git@e536616888d51b453ed354a6f1e243fecb02ea08

# Transformers provides thousands of pretrained models to perform tasks on texts such as classification, information extraction, answering questions, summarizing texts, translating languages, and more.
git+https://github.com/huggingface/transformers.git

# Gradio allows you to quickly create customizable UI components around your TensorFlow or PyTorch models, or even arbitrary Python functions.
gradio

# SentencePiece is an unsupervised text tokenizer and detokenizer mainly for Neural Network-based text generation systems where the vocabulary size is predetermined prior to the neural model training.
sentencepiece

# Wandb is a tool for helping track the progress of machine learning projects, providing methods for logging metrics from runs of your code, and tools for comparing runs and reproducing experiments.
wandb