Loading Datasets

You can load a Unitxt dataset, using the HuggingFace Dataset API without installing the Unitxt package by using the following code:

from datasets import load_dataset

dataset = load_dataset('unitxt/data', 'card=cards.wnli,template=templates.classification.multi_class.relation.default',trust_remote_code=True)

print(dataset)
DatasetDict({
  train: Dataset({
      features: ['metrics', 'source', 'target', 'references', 'group', 'postprocessors'],
      num_rows: 599
  })
  validation: Dataset({
      features: ['metrics', 'source', 'target', 'references', 'group', 'postprocessors'],
      num_rows: 36
  })
  test: Dataset({
      features: ['metrics', 'source', 'target', 'references', 'group', 'postprocessors'],
      num_rows: 71
  })
})

The following prints the source text (input to the model) of the first sample in the training set:

print(dataset['train'][0]['source'])

Given a premise and hypothesis classify the entailment of the hypothesis to one of entailment, not entailment.
premise: Grace was happy to trade me her sweater for my jacket. She thinks it looks dowdy on her.
hypothesis: The sweater looks dowdy on her.
The entailment class is

This prints the reference text (expected output of the model) of the first sample in the training set:

print(dataset['train'][0]['references'][0])

'not entailment'

Loading a customized dataset

Unitxt enables formatting the dataset in different ways.

As example, here we load wnli in 3 shots format:

dataset = load_dataset('unitxt/data', 'card=cards.wnli,template=templates.classification.multi_class.relation.default,num_demos=3,demos_pool_size=100',trust_remote_code=True)

Now the source text (input to the model) of the first sample in the training set has in-context examples:

print(dataset['train'][0]['source'])

Given a premise and hypothesis classify the entailment of the hypothesis to one of entailment, not entailment.
premise: The journalists interviewed the stars of the new movie. They were very cooperative, so the interview lasted for a long time.
hypothesis: The journalists were very cooperative, so the interview lasted for a long time.
The entailment class is entailment

premise: The table won't fit through the doorway because it is too narrow.
hypothesis: The table is too narrow.
The entailment class is entailment

premise: Sam pulled up a chair to the piano, but it was broken, so he had to stand instead.
hypothesis: The chair was broken, so he had to stand instead.
The entailment class is not entailment

premise: Grace was happy to trade me her sweater for my jacket. She thinks it looks dowdy on her.
hypothesis: The sweater looks dowdy on her.
The entailment class is

Loading a Dataset with Multiple Templates or Number of Demonstrations

You can sample a template for each instance from a pool of templates by assigning the template argument a list of templates. Similarly, you can sample the number of demonstrations by assigning num_demos a list of integers.

Here is an example of using random templates and a varying number of demonstrations for each instance of the data:

dataset = load_dataset('unitxt/data', 'card=cards.wnli,template=[templates.classification.multi_class.relation.default,templates.key_val],num_demos=[0,1,3],demos_pool_size=100',trust_remote_code=True)