Commit
•
e016ed8
1
Parent(s):
35f28f8
Update README.md
Browse files
README.md
CHANGED
|
@@ -38,6 +38,8 @@ We have released three models for Scandinavian NLI, of different sizes:
|
|
| 38 |
- alexandrainst/scandi-nli-base (this)
|
| 39 |
- [alexandrainst/scandi-nli-small](https://huggingface.co/alexandrainst/scandi-nli-small)
|
| 40 |
|
|
|
|
|
|
|
| 41 |
The performance and model size of each of them can be found in the Performance section below.
|
| 42 |
|
| 43 |
|
|
@@ -146,6 +148,8 @@ The training split of DanFEVER is generated using [this gist](https://gist.githu
|
|
| 146 |
|
| 147 |
The three languages are sampled equally during training, and they're validated on validation splits of [DanFEVER](https://aclanthology.org/2021.nodalida-main.pdf#page=439) and machine translated versions of [MultiNLI](https://cims.nyu.edu/~sbowman/multinli/) for Swedish and Norwegian Bokmål, sampled equally.
|
| 148 |
|
|
|
|
|
|
|
| 149 |
### Training hyperparameters
|
| 150 |
|
| 151 |
The following hyperparameters were used during training:
|
|
|
|
| 38 |
- alexandrainst/scandi-nli-base (this)
|
| 39 |
- [alexandrainst/scandi-nli-small](https://huggingface.co/alexandrainst/scandi-nli-small)
|
| 40 |
|
| 41 |
+
A demo of the large model can be found in [this Hugging Face Space](https://huggingface.co/spaces/alexandrainst/zero-shot-classification) - check it out!
|
| 42 |
+
|
| 43 |
The performance and model size of each of them can be found in the Performance section below.
|
| 44 |
|
| 45 |
|
|
|
|
| 148 |
|
| 149 |
The three languages are sampled equally during training, and they're validated on validation splits of [DanFEVER](https://aclanthology.org/2021.nodalida-main.pdf#page=439) and machine translated versions of [MultiNLI](https://cims.nyu.edu/~sbowman/multinli/) for Swedish and Norwegian Bokmål, sampled equally.
|
| 150 |
|
| 151 |
+
The full training logs can be found in [this Weights and Biases report](https://wandb.ai/saattrupdan/huggingface/reports/ScandiNLI--VmlldzozMDQyOTk1?accessToken=r9crgxqvvigy2hatdjeobzwipz7f3id5vqg8ooksljhfw6wl0hv1b05asypsfj9v).
|
| 152 |
+
|
| 153 |
### Training hyperparameters
|
| 154 |
|
| 155 |
The following hyperparameters were used during training:
|