GitHub - Panlichen/models: Models and examples built with OneFlow

Please refer to the official repository for detailed documentation.

Running experiments in the OCCL paper

cd Vision/classification/image/resnet50/examples
bash train_ofccl_graph_distributed_fp32.sh <NUM_LOCAL_GPUS>

Notes:

Prepare the ImageNet dataset in advance.
If the environment virable ONEFLOW_ENABLE_OFCCL in train_ofccl_graph_distributed_fp32.sh is set to 1, OCCL will be used during training; otherwise, NCCL will be employed.

Name		Name	Last commit message	Last commit date
Latest commit History 274 Commits
.github/workflows		.github/workflows
Audio		Audio
DeepReinforcementLearning/FlappyBird		DeepReinforcementLearning/FlappyBird
Demo		Demo
GPT		GPT
NLP		NLP
Quantization		Quantization
RecommenderSystems		RecommenderSystems
Vision		Vision
ci/check		ci/check
ops		ops
scripts		scripts
.clang-format		.clang-format
.gitignore		.gitignore
CMakeLists.txt		CMakeLists.txt
LICENSE		LICENSE
README.md		README.md
README_zh-CN.md		README_zh-CN.md
dev-requirements.txt		dev-requirements.txt