Popular repositories Loading

1. flash-attention-w-tree-attn flash-attention-w-tree-attnPublic

   Forked from Dao-AILab/flash-attention

   Fast and memory-efficient exact attention

   Python 4

2. nano-patch-sequence-pack nano-patch-sequence-packPublic

   Just a few lines to combine 🤗 Transformers, Flash Attention 2, and torch.compile — simple, clean, fast ⚡

   Python 2

3. nano-dpo nano-dpoPublic

   A minimal implementation of Direct Preference Optimization (DPO) in Chinese

   Jupyter Notebook 1

4. fast-hadamard-transform fast-hadamard-transformPublic

   Forked from Dao-AILab/fast-hadamard-transform

   Fast Hadamard transform in CUDA, with a PyTorch interface

   C 1

5. NeUQI NeUQIPublic

   Python 1

6. Megatron-LM-NEO Megatron-LM-NEOPublic

   Python