fuel
fuel copied to clipboard
mila-iqia
A data pipeline framework for machine learning
I added a transformer that performs a fixed size crop. It is similar to the already existing RandomFixedSizeCrop transformer. One key difference is that the user can specify the crop...
This is a converter and a loader for caltech101, which could be useful for some other people. It might lack a bit of the polish that other datasets have, but...
``` from fuel.datasets import TextFile from blocks.serialization import dump, load dictionary = {'': 0, '': 1, 'this': 2, 'a': 3, 'one': 4} dataset = TextFile(['example_data.gz'], dictionary, None, level='word', encoding='utf8', preprocess=None).open()...
I am trying to fit my data to Fuel and found some example code snippets on the Internet. They were using classes such as Padding, Mapping, etc in fuel.transformers. but...
Fuel is supposed to be Theano-independent. This seemed to be [causing somebody trouble on the mailing list](https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/fuel-users/lXfXvrkA8zA/VcrouU3tCQAJ) so maybe we should just provide a short circuit.
We added a OneHotEncoding class that transforms a given source (in int) to one-hot-encoding (also known as 1-of-K encoding) as well as a corresponding test. One-hot-encoding is useful if (1)...
It doesn't mention that Fuel tries to import Theano and use `theano.config.floatX`.
The flag introduced in this commit could be used to split a stream into multiple ones and use each branch independently. `AbstractDataStream` now accepts an optional flag `copy` which is...
