Natural Language Inference

Bases: TextFineTuner

A bolt for fine-tuning Hugging Face models for text classification tasks.

This class extends the TextFineTuner and specializes in fine-tuning models for text classification. It provides additional functionalities for loading and preprocessing text classification datasets in various formats.

Parameters:

Name	Type	Description	Default
input	BatchInput	The batch input data.	required
output	OutputConfig	The output data.	required
state	State	The state manager.	required

CLI Usage:

    genius NLIFineTuner rise \
        batch \
            --input_folder ./input \
        batch \
            --output_folder ./output \
        none \
        --id MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7-lol
        fine_tune \
            --args \
                model_name=my_model \
                tokenizer_name=my_tokenizer \
                num_train_epochs=3 \
                per_device_train_batch_size=8

data_collator(examples)

Customize the data collator.

Parameters:

Name	Type	Description	Default
examples	Dict	The examples to collate.	required

Returns:

Name	Type	Description
dict	Dict	The collated data.

load_dataset(dataset_path, **kwargs)

Load a commonsense reasoning dataset from a directory.

Parameters:

Name	Type	Description	Default
dataset_path	str	The path to the dataset directory.	required
**kwargs	Any	Additional keyword arguments.	{}

Returns:

Name	Type	Description
Dataset	Union[Dataset, DatasetDict, None]	The loaded dataset.

Raises:

Type	Description
Exception	If there was an error loading the dataset.

Supported Data Formats and Structures:

Hugging Face Dataset

Dataset files saved by the Hugging Face datasets library.

JSONL

Each line is a JSON object representing an example.

{"premise": "The premise text", "hypothesis": "The hypothesis text", "label": 0 or 1 or 2}

CSV

Should contain 'premise', 'hypothesis', and 'label' columns.

premise,hypothesis,label
"The premise text","The hypothesis text",0

Parquet

Should contain 'premise', 'hypothesis', and 'label' columns.

JSON

An array of dictionaries with 'premise', 'hypothesis', and 'label' keys.

[{"premise": "The premise text", "hypothesis": "The hypothesis text", "label": 0}]

XML

Each 'record' element should contain 'premise', 'hypothesis', and 'label' child elements.

<record>
    <premise>The premise text</premise>
    <hypothesis>The hypothesis text</hypothesis>
    <label>0</label>
</record>

YAML

Each document should be a dictionary with 'premise', 'hypothesis', and 'label' keys.

- premise: "The premise text"
  hypothesis: "The hypothesis text"
  label: 0

TSV

Should contain 'premise', 'hypothesis', and 'label' columns separated by tabs.

Excel (.xls, .xlsx)

Should contain 'premise', 'hypothesis', and 'label' columns.

SQLite (.db)

Should contain a table with 'premise', 'hypothesis', and 'label' columns.

Feather

Should contain 'premise', 'hypothesis', and 'label' columns.

prepare_train_features(examples)

Tokenize the examples and prepare the features for training.

Parameters:

Name	Type	Description	Default
examples	dict	A dictionary of examples.	required

Returns:

Name	Type	Description
dict	Dict	The processed features.