Update app.py
Browse files
app.py
CHANGED
|
@@ -38,7 +38,8 @@ One of the toughest challenges when building the app and the model was to find p
|
|
| 38 |
<ul>
|
| 39 |
<li> First of all, we would love to engage people and maybe inspire them to work on similar projects, because we believe this kind of projects we can actually make a difference</a>
|
| 40 |
<li> Following that, a broader dataset would help reducing some overfitting and making a more robust model </a>
|
| 41 |
-
<li>
|
|
|
|
| 42 |
<li> Reduce the gender biases existing on some of the data for training Language Models by using the output of this model</a>
|
| 43 |
<li> Share ideas with the community for further improvement </a>
|
| 44 |
</ul>
|
|
@@ -59,7 +60,7 @@ One of the toughest challenges when building the app and the model was to find p
|
|
| 59 |
description = """
|
| 60 |
|
| 61 |
<p style="text-align: justify;">
|
| 62 |
-
Spanish is a beautiful language and it has many ways of referring to people, neutralizing the genders and using some of the resources inside the language. One would say *Todas las personas asistentes* instead of *Todos los asistentes* and it would end in a more inclusive way for talking about people.
|
| 63 |
|
| 64 |
The purpose of this app is to transform Spanish gendered text into a neutral version, suitable for an unbiased environment
|
| 65 |
</p>
|
|
|
|
| 38 |
<ul>
|
| 39 |
<li> First of all, we would love to engage people and maybe inspire them to work on similar projects, because we believe this kind of projects we can actually make a difference</a>
|
| 40 |
<li> Following that, a broader dataset would help reducing some overfitting and making a more robust model </a>
|
| 41 |
+
<li> Increase the model diversity to all forms of Spanish. Data generation was quite time consuming and we needed to focus on some sources, but we must change this in the future</a>
|
| 42 |
+
<li> Further training of even more model backbones and architectures </a>
|
| 43 |
<li> Reduce the gender biases existing on some of the data for training Language Models by using the output of this model</a>
|
| 44 |
<li> Share ideas with the community for further improvement </a>
|
| 45 |
</ul>
|
|
|
|
| 60 |
description = """
|
| 61 |
|
| 62 |
<p style="text-align: justify;">
|
| 63 |
+
Spanish is a beautiful language and it has many ways of referring to people, neutralizing the genders and using some of the resources inside the language. One would say *Todas las personas asistentes* instead of *Todos los asistentes* and it would end in a more inclusive way for talking about people. At the same time, this process must be coherent with the context (i.e: if we consider *las alumnas/los alumnos => el alumando*, we would keep *Los alumnos Juanju y Fernando* instead of *El alumnado Juanju y Fernando*)
|
| 64 |
|
| 65 |
The purpose of this app is to transform Spanish gendered text into a neutral version, suitable for an unbiased environment
|
| 66 |
</p>
|