metadata
library_name: transformers
tags: []
Fine-tuned roberta-base for detecting paragraphs with eHRAF-assigned two-digit id '900'
Description
This is a fine tuned roberta-base model for detecting whether paragraphs drawn from ethnographic source material classified under the main subject 'Language and Communication' is more specifically about '900'.
Usage
The easiest way to use this model at inference time is with the HF pipelines API.
from transformers import pipeline
classifier = pipeline("text-classification", model="gptmurdock/classifier-900")
classifier("Example text to classify")
Training data
...
Training procedure
...
We use a 60-20-20 train-val-test split, and fine-tuned roberta-base for 5 epochs (lr = 2e-5, batch size = 40).
Evaluation
Evals on the test set are reported below.
| Metric | Value |
|---|---|
| Precision | 98.0 |
| Recall | 97.9 |
| F1 | 97.9 |