shishishu

GlassMachine: From Bit to Attention

A high-throughput and memory-efficient inference and serving engine for LLMs

LLM Inference with Deep Learning Accelerator.

Training materials associated with NVIDIA's CUDA Training Series (www.olcf.ornl.gov/cuda-training-series/)

Implementations of various classifiers on MNIST dataset. Both traditional machine learning and deep learning methods are included.

📚200+ Tensor/CUDA Cores Kernels, ⚡️flash-attn-mma, ⚡️hgemm with WMMA, MMA and CuTe (98%~100% TFLOPS of cuBLAS/FA2 🎉🎉).