|
--- |
|
license: apache-2.0 |
|
datasets: |
|
- fgs218ok/WikiEditBias |
|
language: |
|
- en |
|
--- |
|
|
|
# Model Card for distilRoberta model finetuned on WikiEditBias dataset |
|
|
|
<!-- Provide a quick summary of what the model is/does. --> |
|
|
|
The [distilRoberta-base](distilbert/distilroberta-base) model finetuned on WikiEditBias dataset on the format of edit diff data and 200k samples. |
|
|
|
For training data, please ref: https://huggingface.co/datasets/fgs218ok/WikiEditBias/viewer/train200k_val2k_test2k_edit_diff |
|
|
|
## Uses |
|
|
|
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. --> |
|
|
|
### Input format |
|
The input should be in diff format: |
|
|
|
``` |
|
<old_text>{old_sentence1} ... {old_sentenceN}<new_text>{new_sentence1} ... {new_sentenceN} |
|
``` |
|
|
|
For example: |
|
``` |
|
<old_text>He is a boy.<new_text>He is a nice boy |
|
``` |
|
|
|
### Direct Use |
|
|
|
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> |
|
``` |
|
model = AutoModel.from_pretrained("fgs218ok/WikiEditBias200k") |
|
tokenizer = AutoTokenizer.from_pretrained("fgs218ok/WikiEditBias200k") |
|
``` |
|
|
|
## Evaluation |
|
|
|
<!-- This section describes the evaluation protocols and provides the results. --> |
|
This model achieves the 83.45% accuracy on fgs218ok/WikiEditBias200k dataset, outperforms GPT-3.5-turbo by large margin while competitive to GPT4o: |
|
|
|
 |
|
|
|
|
