Hugging Face's logo

intone
/
BERT-GRADER

Text Classification
Transformers
PyTorch
bert
Inference Endpoints
Model card Files Community
BERT-GRADER / README.md
cappuch's picture
cappuch
Update README.md
60402fe
preview code
|
raw
history blame
1.21 kB
metadata
tags:
  - text-classification
widget:
  - text: >-
      Hello! I'm BERT-Grader! I'm a finetuned BERT model for grading user input
      on values like consistency, grammar and quality!
    example_title: Example 1
library_name: transformers
pipeline_tag: text-classification

BERT-GRADER?

BERT-Grader is a BERT model designed for grading text by mesuring gramar, consistency and quality. Due to an error in the dataset, 'undefined' is 0. BERT-Grader is a finetune of distil-bert, for performance.

Classes

0 (Undefined) - Elementary 1 - Intermidiate 2 - Advanced

Biases and Issues

For BERT-Grader to perform well, the input text should be an actual essay. Some biases may arise due to training data phrases, but there were mesures in place to limit bias. Giberrish is accounted for, and will be rather random. I suggest using a spell-checker with this model, and mesure tokens.

Validation Metrics

loss: 0.166 accuracy: 0.933333..

Citing 

@misc {intone_2023,
    author       = { {intone} },
    title        = { BERT-GRADER (Revision 6bddc72) },
    year         = 2023,
    url          = { https://huggingface.co/intone/BERT-GRADER },
    doi          = { 10.57967/hf/1257 },
    publisher    = { Hugging Face }
}