|
|
--- |
|
|
base_model: jcblaise/gpt2-tagalog |
|
|
datasets: |
|
|
- jcblaise/hatespeech_filipino |
|
|
language: |
|
|
- tl |
|
|
- en |
|
|
--- |
|
|
|
|
|
# Model Card for Model ID |
|
|
|
|
|
<!-- Provide a quick summary of what the model is/does. --> |
|
|
|
|
|
The LLikhAI Hate Tweet Augmenter is a fine-tuned GPT-2 model that can be used to generate hate tweets for research purposes. |
|
|
|
|
|
## Model Details |
|
|
|
|
|
### Model Description |
|
|
|
|
|
<!-- Provide a longer summary of what this model is. --> |
|
|
|
|
|
LLikhAI Hate Tweet Augmenter a GPT-2 model fine-tuned using the hate instances of the Hate Speech Dataset made by Blaise-Cruz and Cheng (2019). The model is made to augment low-resource Filipino hate speech datasets in order for these to have more instances in the hopes of developing more robust models. |
|
|
|
|
|
- **Language(s) (NLP):** All languages from original GPT-2 model. Fine-tuned for Tagalog. |
|
|
- **Finetuned from model:** jcblaise/gpt2-tagalog |
|
|
|
|
|
## 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. --> |
|
|
|
|
|
### Direct Use |
|
|
|
|
|
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> |
|
|
|
|
|
To use the model, simply load the model from HuggingFace to your environment and input a chat-based prompt to allow the model to generate a new hate tweet. |
|
|
|
|
|
### Out-of-Scope Use |
|
|
|
|
|
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. --> |
|
|
|
|
|
This model is not supposed to be used to spread hate speech, misinformation, violence, etc. on social media. This model is only supposed to be used for research purposes and for augmenting hate speech datasets. |
|
|
|
|
|
## Bias, Risks, and Limitations |
|
|
|
|
|
<!-- This section is meant to convey both technical and sociotechnical limitations. --> |
|
|
|
|
|
This is GPT-2 model finetuned on the Hate Speech Dataset made by Blaise-Cruz and Cheng (2019). As such, this model carries the scope and limitations of both the original model and the original dataset. Given that the Hate Speech Dataset contains mostly political tweets, the generated tweets will have political tones. |
|
|
|
