1. Data

• Choose a dataset from HuggingFace (HF), the Local Filesystem (LF), an Annotation Tool (AT) server, or a Built-in (BI) dataset

dataset = Dataset("conll2003",  source="HF")  # HuggingFace
dataset = Dataset("my_dataset", source="LF")  # Local Filesystem
dataset = Dataset("swe_nerc",   source="BI")  # Built-in

• Set up the dataset

dataset.set_up()

2. Training

• Define the training by choosing a pretrained model and a dataset

training = Training("my_training", model="bert-base-cased", dataset="conll2003")

• Run the training and get the performance of the fine-tuned model

training.run()
training.get_result(metric="f1", level="entity", phase="test")
# 0.9045

3. Evaluation

• Load the model

model = Model.from_training("my_training")

• Evaluate the model

results = model.evaluate_on_dataset("ehealth_kd", phase="test")
results["micro"]["entity"]["f1"]
# 0.9045

4. Inference

• Load the model

model = Model.from_training("my_training")

• Let the model predict

model.predict("The United Nations has never recognised Jakarta's move.")
# [[
#  {'char_start': '4', 'char_end': '18', 'token': 'United Nations', 'tag': 'ORG'},
#  {'char_start': '40', 'char_end': '47', 'token': 'Jakarta', 'tag': 'LOC'}
# ]]

There is much more to it than that! See the documentation to get started.