mangoes package

Subpackages

• mangoes.utils package
  • Submodules
    • mangoes.utils.arrays module
    • mangoes.utils.decorators module
    • mangoes.utils.exceptions module
    • mangoes.utils.io module
    • mangoes.utils.metrics module
    • mangoes.utils.multiproc module
    • mangoes.utils.options module
    • mangoes.utils.persist module
    • mangoes.utils.reader module
  • Module contents
• mangoes.modeling package
  • Submodules
    • mangoes.modeling.transformer_base module
    • mangoes.modeling.finetuning module
    • mangoes.modeling.coref module
    • mangoes.modeling.training_utils module
  • Module contents

Submodules

• mangoes.base module
• mangoes.constants module
• mangoes.context module
  • Examples
• mangoes.corpus module
• mangoes.counting module
• mangoes.dataset module
  • Warnings
  • References
• mangoes.evaluate module
• mangoes.reduction module
  • Examples
  • See Also
• mangoes.visualize module
• mangoes.vocabulary module
• mangoes.weighting module
  • Examples
  • See Also

Module contents

Mangoes is a toolbox to produce and evaluate word embeddings.

Mangoes can be used to :

• Construct a words co-occurrence matrix from a Corpus

• Create Word Embeddings from this matrix

• Analyze a corpus (counting words, sentences, …) and create vocabularies from it

• Evaluate word embeddings

• Generate context dependent embeddings from a pretrained language model.

• Get task outputs from a pretrained language model.

• Fine-tune a pretrained language model.