PiePline

Neural networks training pipeline based on PyTorch. Designed to standardize training process and accelerate experiments.

• Core is about 2K lines, covered by tests, that you don't need to write again
• Flexible and customizable training process
• Checkpoints management and train process resuming (source and target device independent)
• Metrics processing and visualization by builtin (tensorboard, Matplotlib) or custom monitors
• Training best practices (e.g. learning rate decaying and hard negative mining)
• Metrics logging and comparison (DVC compatible)

Getting started:

Documentation

| Stable: | Latest: |

• See the full documentation there
• Read getting started guide

See the examples

• MNIST classification - notebook, file, Kaggle kernel
• Segmentation - notebook, file
• Resume training process - file

PiePline short overview:

import torch

from neural_pipeline.builtin.monitors.tensorboard import TensorboardMonitor
from neural_pipeline.monitoring import LogMonitor
from neural_pipeline import DataProducer, TrainConfig, TrainStage,\
    ValidationStage, Trainer, FileStructManager

from somethig import MyNet, MyDataset

fsm = FileStructManager(base_dir='data', is_continue=False)
model = MyNet().cuda()

train_dataset = DataProducer([MyDataset()], batch_size=4, num_workers=2)
validation_dataset = DataProducer([MyDataset()], batch_size=4, num_workers=2)

train_config = TrainConfig(model, [TrainStage(train_dataset),
                                   ValidationStage(validation_dataset)], torch.nn.NLLLoss(),
                           torch.optim.SGD(model.parameters(), lr=1e-4, momentum=0.5))

trainer = Trainer(train_config, fsm, torch.device('cuda:0')).set_epoch_num(50)
trainer.monitor_hub.add_monitor(TensorboardMonitor(fsm, is_continue=False))\
                   .add_monitor(LogMonitor(fsm))
trainer.train()

This example of training MyNet on MyDataset with vizualisation in Tensorflow and with metrics logging for further experiments comparison.

Installation:

pip install piepline

For builtin module using install:

pip install tensorboardX matplotlib

Install latest version before it's published on PyPi

pip install -U git+https://github.com/PiePline/piepline