Welcome to BatchFlow documentation!

BatchFlow helps you conveniently work with random or sequential batches of your data and define data processing and machine learning workflows even for datasets that do not fit into memory.

Main features:

• flexible batch generaton

• deterministic and stochastic pipelines

• datasets and pipelines joins and merges

• data processing actions

• flexible model configuration

• within batch parallelism

• batch prefetching

• ready to use ML models and proven NN architectures

• convenient layers and helper functions to build custom models

• a powerful research engine with parallel model training and extended experiment logging.

Contents

• A short introduction
• Classes and capabilities
  • Index
  • Dataset
  • Batch class
  • Batch class for handling images
  • Pipeline
  • Named expressions
  • Within batch parallelism
  • Inter-batch parallelism
  • Working with models
  • Torch models
  • Research
  • Model zoo
• API
  • batchflow
  • batchflow.opensets
  • batchflow.models
  • batchflow.research

Basic usage

my_workflow = my_dataset.pipeline()
                .load('/some/path')
                .do_something()
                .do_something_else()
                .some_additional_action()
                .save('/to/other/path')

The trick here is that all the processing actions are lazy. They are not executed until their results are needed, e.g. when you request a preprocessed batch:

my_workflow.run(BATCH_SIZE, shuffle=True, n_epochs=5)

or

for batch in my_workflow.gen_batch(BATCH_SIZE, shuffle=True, n_epochs=5):
    # only now the actions are fired and data is being changed with the workflow defined earlier
    # actions are executed one by one and here you get a fully processed batch

or

NUM_ITERS = 1000
for i in range(NUM_ITERS):
    processed_batch = my_workflow.next_batch(BATCH_SIZE, shuffle=True, n_epochs=None)
    # only now the actions are fired and data is changed with the workflow defined earlier

Train a neural network

BatchFlow includes ready-to-use proven architectures like VGG, Inception, ResNet and many others. To apply them to your data just choose a model, specify the inputs (like the number of classes or images shape) and call train_model. Of course, you can also choose a loss function, an optimizer and many other parameters, if you want.:

from batchflow.models.torch import ResNet34

my_workflow = my_dataset.pipeline()
              .init_model('dynamic', ResNet34, config={
                          'inputs/images/shape': B('image_shape'),
                          'labels/classes': 10,
                          'initial_block/inputs': 'images'})
              .load('/some/path')
              .some_transform()
              .another_transform()
              .train_model('ResNet34', images=B('images'), labels=B('labels'))
              .run(BATCH_SIZE, shuffle=True)

Installation

Python package

With modern pipenv:

pipenv install git+https://github.com/analysiscenter/batchflow.git#egg=batchflow

With old-fashioned pip:

pip3 install git+https://github.com/analysiscenter/batchflow.git

After that just import batchflow:

import batchflow as bf

Git submodule

Note

BatchFlow module is in the beta stage. Your suggestions and improvements are very welcome.

Note

BatchFlow supports python 3.5 or higher.

In many cases it is much more convenient to install batchflow as a submodule in your project repository than as a python package:

git submodule add https://github.com/analysiscenter/batchflow.git
git submodule init
git submodule update

If your python file is located in another directory, you might need to add a path to batchflow submodule location:

import sys
sys.path.insert(0, "/path/to/batchflow")
import batchflow as bf

What is great about using a submodule is that every commit in your project can be linked to its own commit of a submodule. This is extremely convenient in a fast paced research environment.

Relative import is also possible:

from .batchflow import Dataset

Requirements

BatchFlow has the following strict requirments:

• numpy

• dill

• tqdm

• matplotlib

Some BatchFlow modules also depend on other mandatory packages:

• pandas and multiprocess for batchflow.research

• torch for batchflow.models.torch

Citing BatchFlow

Please cite BatchFlow in your publications if it helps your research.

Roman Khudorozhkov et al. BatchFlow library for fast ML workflows. 2017. doi:10.5281/zenodo.1041203

@misc{roman_kh_2017_1041203,
  author       = {Khudorozhkov, Roman and others},
  title        = {BatchFlow library for fast ML workflows},
  year         = 2017,
  doi          = {10.5281/zenodo.1041203},
  url          = {https://doi.org/10.5281/zenodo.1041203}
}