Llama 4 From Scratch

This repository is an implementation for Llama 4 model with BPE tokenizer, a Feed forward network and a Rotation Positional Encoder.

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
.idea		.idea
.gitignore		.gitignore
BPE Tokenizer.py		BPE Tokenizer.py
Feedforward.py		Feedforward.py
Llama Attention.py		Llama Attention.py
README.md		README.md
requirements.txt		requirements.txt

Provide feedback