Fast Byte Latent Transformer in pure C and CUDA. A tokenizer-free, byte-level language model for code completion and small LLMs, based on Meta's BLT and Fast BLT papers.
-
Updated
Oct 23, 2026 - C
#
byte-latent-transformer
Here are 3 public repositories matching this topic...
An implementation of Meta's Byte Latent Transformer architecture
-
Updated
Oct 2, 2025 - Python
Minimal, clean PyTorch implementation of Byte Latent Transformer (BLT), inspired by Karpathy's minbpe.
-
Updated
Oct 16, 2026 - Python
Add this topic to your repo
To associate your repository with the byte-latent-transformer topic, visit your repo's landing page and select "manage topics."