hmxiong

Transformer-Series

OpenMMLabCamp

paper-reading

深度学习经典、新论文逐段精读

CRATE

Code for CRATE (Coding RAte reduction TransformEr).

CUDA-Learn-Note

🎉CUDA 笔记 / 大模型手撕CUDA / C++笔记，更新随缘: flash_attn、sgemm、sgemv、warp reduce、block reduce、dot product、elementwise、softmax、layernorm、rmsnorm、hist etc.

GaLore

github-slideshow

A robot powered training repository :robot:

hallow

i just wanna learn deep learning

llama2.c

Inference Llama 2 in one file of pure C

ScanNet_Vis

Tarurs

competition files

pytorch-distributed-training

Simple tutorials on Pytorch DDP training

RWKV-LM

RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.

VILA

VILA - a multi-image visual language model with training, inference and evaluation recipe, deployable from cloud to edge (Jetson Orin and laptops)