Simple dataset to dataloader library for pytorch
Project description
This is a simple library for creating readable dataset pipelines and reusing best practices for issues such as imbalanced datasets. There are just two components to keep track of: Dataset and Datastream.
Dataset is a simple mapping between an index and an example. It provides pipelining of functions in a readable syntax originally adapted from tensorflow 2’s tf.data.Dataset.
Datastream combines a Dataset and a sampler into a stream of examples. It provides a simple solution to oversampling / stratification, weighted sampling, and finally converting to a torch.utils.data.DataLoader.
Install
pip install pytorch-datastream
Usage
The list below is meant to showcase functions that are useful in most standard and non-standard cases. It is not meant to be an exhaustive list. See the documentation for a more extensive list on API and usage.
Dataset.from_subscriptable
Dataset.from_dataframe
Dataset
.map
.subset
.split
.cache
.with_columns
Datastream.merge
Datastream.zip
Datastream
.map
.data_loader
.zip_index
.update_weights_
.update_example_weight_
.weight
.state_dict
.load_state_dict
Merge / stratify / oversample datastreams
The fruit datastreams given below repeatedly yields the string of its fruit type.
>>> datastream = Datastream.merge([
... (apple_datastream, 2),
... (pear_datastream, 1),
... (banana_datastream, 1),
... ])
>>> next(iter(datastream.data_loader(batch_size=8)))
['apple', 'apple', 'pear', 'banana', 'apple', 'apple', 'pear', 'banana']
Zip independently sampled datastreams
The fruit datastreams given below repeatedly yields the string of its fruit type.
>>> datastream = Datastream.zip([
... apple_datastream,
... Datastream.merge([pear_datastream, banana_datastream]),
... ])
>>> next(iter(datastream.data_loader(batch_size=4)))
[('apple', 'pear'), ('apple', 'banana'), ('apple', 'pear'), ('apple', 'banana')]
More usage examples
See the documentation for more usage examples.
Install from source
To patch the code locally for Python 3.6 run patch-python3.6.sh.
$ ./patch-python3.6.sh
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
Built Distributions
Hashes for pytorch_datastream-0.4.3-py39-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 704c7d0551f3c6ed17f654164b19713bdd5c7f923bd3daa063a5ba248547e779 |
|
MD5 | 6d350941eb6fc37c4f61c68b7e42a30e |
|
BLAKE2b-256 | 57fb5c5a962aee0ce863addd8344cc4ea0d9f81e507e60f1a9c8bac95c602efa |
Hashes for pytorch_datastream-0.4.3-py38-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 5aa7bb3b07257ed923d6fe14f340707fc1a520c77d37df44e12de7de6ad020e3 |
|
MD5 | aae13c62bd193f647eaf90921aa5a81f |
|
BLAKE2b-256 | b32c05a30977223bf05ede5409b893159c96f430e4567e7190afd64ed6f117ce |
Hashes for pytorch_datastream-0.4.3-py37-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 8a8d810146d7dbb7d6fafe43a6a532083c12a17a5735c756f15502d38de8101f |
|
MD5 | 43e8e543ec70752e3598fa505069c63a |
|
BLAKE2b-256 | c70bccce9fccb93449d0e2d686f09131e88fda4677febadcab664948b48280fb |
Hashes for pytorch_datastream-0.4.3-py36-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 29fa31b58135942ae2094eed85c36df281118bff3abca8fe2c76bef382eb6134 |
|
MD5 | 061ed9084eaa0b790d2024108d7d766a |
|
BLAKE2b-256 | 795740988edb5756845f161bc6235f16cf248c5fb0ad97978c96f91f1b311782 |