Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.
You must be logged in to block users.
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Open Source AI Accelerator Research Platform Standard / 开源推理研究平台
Python 1.9k 345
Implemented the DeepSeek v3 model from scratch which includes Multi-Head Latent Attention and MoE architecture
Python
Deep Matrix Capsules Implementation
Jupyter Notebook 16 6
CUDAGrad: A CUDA Native Autograd Engine
Jupyter Notebook
Flash Attention 2 from Scratch in Triton
A triton kernel for NF4 dequantization faster than HF and a gradient checkpointing implementation
There was an error while loading. Please reload this page.