Benchmark for Tuning Accuracy and Efficiency

Overview

The benchmark includes our efforts in using Colossal-AI to train different tasks to achieve SOTA results. We are interested in both validataion accuracy and training speed, and prefer larger batch size to take advantage of more GPU devices. For example, we trained vision transformer with batch size 512 on CIFAR10 and 4096 on ImageNet1k, which are basically not used in existing works. Some of the results in the benchmark trained with 8x A100 are shown below.

Task	Model	Training Time	Top-1 Accuracy
CIFAR10	ViT-Lite-7/4	~ 16 min	~ 90.5%
ImageNet1k	ViT-S/16	~ 16.5 h	~ 74.5%

The train.py script in each task runs training with the specific configuration script in configs/ for different parallelisms. Supported parallelisms include data parallel only (ends with vanilla), 1D (ends with 1d), 2D (ends with 2d), 2.5D (ends with 2p5d), 3D (ends with 3d).

Each configuration scripts basically includes the following elements, taking ImageNet1k task as example:

TOTAL_BATCH_SIZE = 4096
LEARNING_RATE = 3e-3
WEIGHT_DECAY = 0.3

NUM_EPOCHS = 300
WARMUP_EPOCHS = 32

# data parallel only
TENSOR_PARALLEL_SIZE = 1    
TENSOR_PARALLEL_MODE = None

# parallelism setting
parallel = dict(
    pipeline=1,
    tensor=dict(mode=TENSOR_PARALLEL_MODE, size=TENSOR_PARALLEL_SIZE),
)

fp16 = dict(mode=AMP_TYPE.TORCH, ) # amp setting

gradient_accumulation = 2 # accumulate 2 steps for gradient update

BATCH_SIZE = TOTAL_BATCH_SIZE // gradient_accumulation # actual batch size for dataloader

clip_grad_norm = 1.0 # clip gradient with norm 1.0

Upper case elements are basically what train.py needs, and lower case elements are what Colossal-AI needs to initialize the training.

Usage

To start training, use the following command to run each worker:

$ DATA=/path/to/dataset python train.py --world_size=WORLD_SIZE \
                                        --rank=RANK \
                                        --local_rank=LOCAL_RANK \
                                        --host=MASTER_IP_ADDRESS \
                                        --port=MASTER_PORT \
                                        --config=CONFIG_FILE

It is also recommended to start training with torchrun as:

$ DATA=/path/to/dataset torchrun --nproc_per_node=NUM_GPUS_PER_NODE \
                                 --nnodes=NUM_NODES \
                                 --node_rank=NODE_RANK \
                                 --master_addr=MASTER_IP_ADDRESS \
                                 --master_port=MASTER_PORT \
                                 train.py --config=CONFIG_FILE

ColossalAI-Benchmark - Performance benchmarking with ColossalAI

Related tags

Overview

Benchmark for Tuning Accuracy and Efficiency

Overview

Usage

Owner

HPC-AI Tech

Official PaddlePaddle implementation of Paint Transformer

A small demonstration of using WebDataset with ImageNet and PyTorch Lightning

GrailQA: Strongly Generalizable Question Answering

AdaNet is a lightweight TensorFlow-based framework for automatically learning high-quality models with minimal expert intervention

Scikit-event-correlation - Event Correlation and Forecasting over High Dimensional Streaming Sensor Data algorithms

Repository for the Bias Benchmark for QA dataset.

CausalNLP is a practical toolkit for causal inference with text as treatment, outcome, or "controlled-for" variable.

Sequence to Sequence (seq2seq) Recurrent Neural Network (RNN) for Time Series Forecasting

Tensorflow python implementation of "Learning High Fidelity Depths of Dressed Humans by Watching Social Media Dance Videos"

PyTorch implementation of the paper Ultra Fast Structure-aware Deep Lane Detection

Creating predictive checklists from data using integer programming.

Research code for the paper "Variational Gibbs inference for statistical estimation from incomplete data".

labelpix is a graphical image labeling interface for drawing bounding boxes

Time Series Cross-Validation -- an extension for scikit-learn

[ICML 2021] A fast algorithm for fitting robust decision trees.

A Semantic Segmentation Network for Urban-Scale Building Footprint Extraction Using RGB Satellite Imagery

A Lightweight Experiment & Resource Monitoring Tool 📺

Official implementation of "Intrinsic Dimension, Persistent Homology and Generalization in Neural Networks", NeurIPS 2021.

Pytorch implementation of Rosca, Mihaela, et al. "Variational Approaches for Auto-Encoding Generative Adversarial Networks."

Simple Dynamic Batching Inference