ankan-ban / llama_cu_awq
llama INT4 cuda inference with AWQ
☆46Updated 2 months ago
Related projects: ⓘ
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆82Updated 6 months ago
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆81Updated 2 months ago
- Standalone Flash Attention v2 kernel without libtorch dependency☆93Updated last week
- ☆138Updated 2 months ago
- ☆67Updated last week
- Benchmark code for the "Online normalizer calculation for softmax" paper☆52Updated 6 years ago
- ☆102Updated 3 months ago
- Several optimization methods of half-precision general matrix vector multiplication (HGEMV) using CUDA core.☆40Updated last week
- 使用 cutlass 仓库在 ada 架构上实现 fp8 的 flash attention☆46Updated last month
- PyTorch bindings for CUTLASS grouped GEMM.☆41Updated 3 weeks ago
- Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.☆20Updated last week
- TiledCUDA is a highly efficient kernel template library designed to elevate CUDA C’s level of abstraction for processing tiles.☆114Updated last week
- Penn CIS 5650 (GPU Programming and Architecture) Final Project☆21Updated 9 months ago
- play gemm with tvm☆81Updated last year
- ☆33Updated 5 months ago
- ☆151Updated last year
- Benchmark scripts for TVM☆73Updated 2 years ago
- SparseTIR: Sparse Tensor Compiler for Deep Learning☆129Updated last year
- ☆23Updated this week
- Applied AI experiments and examples for PyTorch☆123Updated last month
- An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).☆173Updated 3 months ago
- PyTorch extension for emulating FP8 data formats on standard FP32 Xeon/GPU hardware.☆98Updated 9 months ago
- System for automated integration of deep learning backends.☆48Updated 2 years ago
- CUDA Matrix Multiplication Optimization☆118Updated 2 months ago
- An extention of TVMScript to write simple and high performance GPU kernels with tensorcore.☆49Updated last month
- [ACL 2024] A novel QAT with Self-Distillation framework to enhance ultra low-bit LLMs.☆69Updated 4 months ago
- QQQ is an innovative and hardware-optimized W4A8 quantization solution.☆59Updated last week
- ☆34Updated 2 years ago
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆166Updated 11 months ago
- ☆27Updated 3 weeks ago