floret_ko_ud_demo

🪐 spaCy Project: Demo floret vectors for UD Korean Kaist

Train floret vectors on OSCAR and compare no vectors, standard vectors, and floret vectors on UD Korean Kaist.

📋 project.yml

The project.yml defines the data assets required by the project, as well as the available commands and workflows. For details, see the spaCy projects documentation.

⏯ Commands

The following commands are defined by the project. They can be executed using spacy project run [name]. Commands are only re-run if their inputs have changed.

Command	Description
tokenize-oscar	Download, tokenize, and sentencize data
train-fasttext-standard-vectors	Train standard fasttext vectors
train-floret-vectors	Train floret vectors
init-standard-unpruned-vectors	Create a standard unpruned vectors model
init-standard-vectors	Create a standard vectors model
init-floret-vectors	Create a floret vectors model
convert	Convert the data to spaCy's format
train-no-vectors-model	Train the model without vectors
train-standard-unpruned-model	Train the model with standard, unpruned vectors
train-standard-model	Train the model with standard, pruned vectors
train-floret-model	Train the model with floret vectors
evaluate	Evaluate the models and export metrics

⏭ Workflows

The following workflows are defined by the project. They can be executed using spacy project run [name] and will run the specified commands in order. Commands are only re-run if their inputs have changed.

Workflow	Steps
all	tokenize-oscar → train-fasttext-standard-vectors → train-floret-vectors → init-standard-unpruned-vectors → init-standard-vectors → init-floret-vectors → convert → train-no-vectors-model → train-standard-unpruned-model → train-standard-model → train-floret-model → evaluate

🗂 Assets

The following assets are defined by the project. They can be fetched by running spacy project assets in the project directory.

File	Source	Description
assets/UD_Korean-Kaist	Git