maximuspowers commited on
Commit
6e6750f
·
verified ·
1 Parent(s): 08ae843

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +9 -6
app.py CHANGED
@@ -146,16 +146,19 @@ with iface:
146
  with gr.Row():
147
  gr.Markdown(
148
  """
149
- # Social Bias Named Entity Recognition (with BERT) 🕵
150
- Enter a sentence to predict biased parts of speech tags. This model uses multi-label `BertForTokenClassification` to label the entities:
 
 
 
 
 
 
151
  - **Generalizations (GEN)**
152
  - **Unfairness (UNFAIR)**
153
  - **Stereotypes (STEREO)**
154
 
155
- Labels follow the BIO format. Try it out!
156
-
157
- - **[Blog Post](https://huggingface.co/blog/maximuspowers/bias-entity-recognition)**
158
- - **[Model Page](https://huggingface.co/maximuspowers/bias-detection-ner)**
159
  """
160
  )
161
  with gr.Row():
 
146
  with gr.Row():
147
  gr.Markdown(
148
  """
149
+ # GUS-Net 🕵
150
+ [GUS-Net](https://huggingface.co/ethical-spectacle/social-bias-ner) is a `BertForTokenClassification` based model, trained on the [GUS dataset](https://huggingface.co/datasets/ethical-spectacle/gus-dataset-v1). It preforms multi-label named-entity recognition of socially biased entities, intended to reveal the underlying structure of bias rather than a one-size fits all definition.
151
+
152
+ You can find the full collection of resources introduced in our paper [here](https://huggingface.co/collections/ethical-spectacle/gus-net-66edfe93801ea45d7a26a10f).
153
+
154
+ This [blog post](https://huggingface.co/blog/maximuspowers/bias-entity-recognition) walks through the training and architecture of the model.
155
+
156
+ Enter a sentence for named-entity recognition of biased entities:
157
  - **Generalizations (GEN)**
158
  - **Unfairness (UNFAIR)**
159
  - **Stereotypes (STEREO)**
160
 
161
+ Labels follow the BIO format. Try it out:
 
 
 
162
  """
163
  )
164
  with gr.Row():