neural_compressor.data
Built-in dataloaders, datasets, transforms, filters for multiple framework backends.
Subpackages
neural_compressor.data.datasets
neural_compressor.data.datasets.bert_dataset
neural_compressor.data.datasets.coco_dataset
neural_compressor.data.datasets.dataset
neural_compressor.data.datasets.dummy_dataset
neural_compressor.data.datasets.dummy_dataset_v2
neural_compressor.data.datasets.imagenet_dataset
neural_compressor.data.datasets.style_transfer_dataset
neural_compressor.data.filters
neural_compressor.data.transforms
Package Contents
Classes
|
A base class for all framework datasets. |
|
The base class of dataset. |
|
An iterable Dataset. |
|
Tensorflow imageNet database in tf record format. |
|
Tensorflow COCO dataset in tf record format. |
|
Entrance of all configured DataLoaders. |
|
DefaultDataLoader for multiple framework backends. |
|
Transforms collection class. |
|
The base class for transform. |
|
Composes several transforms together. |
|
Just collect the infos to construct a Postprocess. |
|
Convert label to label - label_shift. |
|
Combination of a series of transforms which is applicable to images in Imagenet. |
|
Combination of a series of transforms which is applicable to images in Imagenet. |
|
Postprocess the predictions of bert on SQuAD. |
|
Postprocess the predictions of bert on SQuADV1.1. |
|
Resize image with aspect ratio and pad it to max shape(optional). |
|
Resize the input image to the given size. |
|
Scale the values of image to [0,1]. |
|
Normalize a image with mean and standard deviation. |
|
Coco decoding will be performed automatically from Neural Compressor v1.4. |
|
Label shift by 1 and rescale. |
|
The filter register for all frameworks. |
|
The base class for transform. |
|
The label balance filter for COCO Record. |
Functions
|
Register dataset subclasses. |
|
Check if the dataloader meets requirement of neural_compressor. |
|
Class decorator used to register all transform subclasses. |
|
Register all transform subclasses. |
Attributes
|