整理 pytorch 单机多 GPU 训练方法与原理

ALBERT model Pretraining and Fine Tuning using TF2.0

Efficient and Scalable Physics-Informed Deep Learning and Scientific Machine Learning on top of Tensorflow for multi-worker distributed computing

Distributed Reinforcement Learning for LLM Fine-Tuning with multi-GPU utilization

jupyter notebooks to fine tune whisper models on Vietnamese using Colab and/or Kaggle and/or AWS EC2

Tensorflow2 training code with jit compiling on multi-GPU.

Deep learning using TensorFlow low-level APIs

A lightweight Python template for deep learning project or research with PyTorch.

A pytorch project template for intensive AI research. Separate datamodule and models and thus support for multiple data-loaders and multiple models in same project

In depth tutorial for conducting distributed training at NSM Clusters for custom workloads

Context-Aware Image Captioning with BLIP-2

使用TensorFlow训练自己的图片，基于多GPU

Production-ready multi-GPU distributed training framework with DDP/FSDP, gradient compression, and 89% scaling efficiency at 16 GPUs. Includes TensorBoard monitoring, auto-checkpointing, and Kubernetes deployment.

PyTorch/Lightning implementation of https://github.com/kang205/SASRec

SHUKUN Technology Co.,Ltd Algorithm intern (2020/12-2021/5). Multi-GPU, Multi-node training for deep learning models. Horovod, NVIDIA clara train sdk, configuration tutorial,performance testing.

Production-scale video style transfer (AdaIN + RAFT Optical Flow) achieving 6.45 FPS and trained via DDP on 118K images.

Code for various probabilistic deep learning models

DPython is a custom Python launcher for automatic single-GPU or multi-machine distributed training over LAN using Hugging Face Accelerate. It supports Windows, auto-detects local and remote GPUs, falls back safely to single-GPU when needed, and simplifies real-world multi-GPU training on low-VRAM systems.

End-to-End Neural Diarization with python