PROJECT.01 // ATTENTION.MATRIX
TirZ - M1 (460M)
TirZ M1 is a GPT-style decoder-only transformer language model comprising 417,887,232 trainable parameters, designed and trained from scratch by Tirth Patel (@Tirth9978) under the OpenTirZ organization. Built upon the foundational principles established in the landmark Attention Is All You Need paper .
- Large Language Model
- Python
- PyTorch
- Transformer
- LLM



