Commit
·
78af37e
1
Parent(s):
615f1ca
Added readme link to vaccine model
Browse files
README.md
CHANGED
|
@@ -6,6 +6,8 @@ Base model: [gpt2-large](https://huggingface.co/gpt2-large)
|
|
| 6 |
|
| 7 |
Fine-tuned to generate responses on a dataset of [COVID-19 public health tweets](https://github.com/TheRensselaerIDEA/generative-response-modeling). For more information about the dataset, task and training, see [our paper](https://arxiv.org/abs/2204.04353). This checkpoint corresponds to the lowest validation perplexity (3.36 at 2 epochs) seen during training. See Training metrics for Tensorboard logs.
|
| 8 |
|
|
|
|
|
|
|
| 9 |
**Data input format:** <span style="color:red"><|message|></span>public health message<span style="color:red"><|author|></span>public health Twitter handle<span style="color:red"><|response|></span>
|
| 10 |
|
| 11 |
Example:
|
|
|
|
| 6 |
|
| 7 |
Fine-tuned to generate responses on a dataset of [COVID-19 public health tweets](https://github.com/TheRensselaerIDEA/generative-response-modeling). For more information about the dataset, task and training, see [our paper](https://arxiv.org/abs/2204.04353). This checkpoint corresponds to the lowest validation perplexity (3.36 at 2 epochs) seen during training. See Training metrics for Tensorboard logs.
|
| 8 |
|
| 9 |
+
Also see: our [Vaccine public health tweet response model](https://huggingface.co/TheRensselaerIDEA/gpt2-large-vaccine-tweet-response).
|
| 10 |
+
|
| 11 |
**Data input format:** <span style="color:red"><|message|></span>public health message<span style="color:red"><|author|></span>public health Twitter handle<span style="color:red"><|response|></span>
|
| 12 |
|
| 13 |
Example:
|