Reinerzhou/DeepLearningSystem
Deep Learning System core principles introduction.
Deep Learning System core principles introduction.
本项目旨在分享大模型相关技术原理以及实战经验(大模型工程化、大模型应用落地)
📚200+ Tensor/CUDA Cores Kernels, ⚡️flash-attn-mma, ⚡️hgemm with WMMA, MMA and CuTe (98%~100% TFLOPS of cuBLAS/FA2 🎉🎉).
📖A curated list of Awesome LLM Inference Paper with codes, TensorRT-LLM, vLLM, streaming-llm, AWQ, SmoothQuant, WINT8/4, Continuous Batching, FlashAttention, PagedAttention etc.
欢迎来到 LLM-Dojo,这里是一个开源大模型学习场所,使用简洁且易阅读的代码构建模型训练框架(支持各种主流模型如Qwen、Llama、GLM等等)、RLHF框架(DPO/CPO/KTO/PPO)等各种功能。👩🎓👨🎓
LightSeq: A High Performance Library for Sequence Processing and Generation
LMDeploy is a toolkit for compressing, deploying, and serving LLMs.
InteroEvo integration with huggingface models
MLA的设计可以保证效果与传统的MHA效果相同的情况下,实现更低的kv-cache开销。但是官方并没有给出矩阵融合后的推理代码,这对于对齐论文中的效果是必要的一步。本仓库的代码用来实现MLA的参数融合,以及融合后的pytorch推理代码。
8-bit CUDA functions for PyTorch
[CVPR 2024] MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model
Curated tutorials and resources for Large Language Models, AI Painting, and more.
AI绘画资料合集(包含国内外可使用平台、使用教程、参数教程、部署教程、业界新闻等等) Stable diffusion、AnimateDiff、Stable Cascade 、Stable SDXL Turbo
Bringing Old Photo Back to Life (CVPR 2020 oral)
PULSE: Pretrained and Unified Language Service Engine
LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalability, and high-speed performance.
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
🚀 A simple way to train and use PyTorch models with multi-GPU, TPU, mixed-precision
GoogleTest - Google Testing and Mocking Framework
Easy Docker setup for Stable Diffusion with user-friendly UI
Config files for my GitHub profile.