soumith / inferno

Inferno

A scikit-learn compatible neural network library that wraps pytorch.

Example

import numpy as np
from sklearn.datasets import make_classification
import torch
from torch import nn
import torch.nn.functional as F

from inferno.net import NeuralNetClassifier


X, y = make_classification(1000, 20, n_informative=10, random_state=0)
X = X.astype(np.float32)


class MyModule(nn.Module):
    def __init__(self, num_units=10, nonlin=F.relu):
        super(MyModule, self).__init__()

        self.dense0 = nn.Linear(20, num_units)
        self.nonlin = nonlin
        self.dropout = nn.Dropout(0.5)
        self.dense1 = nn.Linear(num_units, 10)
        self.output = nn.Linear(10, 2)

    def forward(self, X, **kwargs):
        X = self.nonlin(self.dense0(X))
        X = self.dropout(X)
        X = F.relu(self.dense1(X))
        X = F.softmax(self.output(X))
        return X


net = NeuralNetClassifier(
    MyModule,
    max_epochs=10,
    lr=0.1,
)

net.fit(X, y)
y_proba = net.predict_proba(X)

In an sklearn Pipeline:

from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler


pipe = Pipeline([
    ('scale', StandardScaler()),
    ('net', net),
])

pipe.fit(X, y)
y_proba = pipe.predict_proba(X)

With grid search

from sklearn.model_selection import GridSearchCV


params = {
    'lr': [0.01, 0.02],
    'max_epochs': [10, 20],
    'module__num_units': [10, 20],
}
gs = GridSearchCV(net, params, refit=False, cv=3, scoring='accuracy')

gs.fit(X, y)
print(gs.best_score_, gs.best_params_)

Installation

conda

You need a working conda installation. Get the correct miniconda for your system from here.

For users

conda env create
source activate inferno
python setup.py install

For developers

conda env create
source activate inferno
conda install --file requirements-dev.txt
python setup.py develop

py.test  # unit tests
pylint inferno  # static code checks

pip

Same as for conda, but to install main requirements, run:

pip install -r requirements.txt

For installation instructions for pytorch, visit the pytorch website.

In general, this should work:

pip install http://download.pytorch.org/whl/cu80/torch-0.2.0.post1-cp36-cp36m-manylinux1_x86_64.whl