José Ángel González
commited on
Commit
•
dcfdda4
1
Parent(s):
b0ec8dc
Update README.md
Browse files
README.md
CHANGED
|
@@ -22,6 +22,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 22 |
# barthez-deft-sciences_de_l_information
|
| 23 |
|
| 24 |
This model is a fine-tuned version of [moussaKam/barthez](https://huggingface.co/moussaKam/barthez) on an unknown dataset.
|
|
|
|
|
|
|
|
|
|
| 25 |
It achieves the following results on the evaluation set:
|
| 26 |
- Loss: 2.0258
|
| 27 |
- Rouge1: 34.5672
|
|
|
|
| 22 |
# barthez-deft-sciences_de_l_information
|
| 23 |
|
| 24 |
This model is a fine-tuned version of [moussaKam/barthez](https://huggingface.co/moussaKam/barthez) on an unknown dataset.
|
| 25 |
+
|
| 26 |
+
**Note**: this model is one of the preliminary experiments and it slightly underperforms the models published in the paper (using [MBartHez](https://huggingface.co/moussaKam/mbarthez) and HAL/Wiki pre-training + copy mechanisms)
|
| 27 |
+
|
| 28 |
It achieves the following results on the evaluation set:
|
| 29 |
- Loss: 2.0258
|
| 30 |
- Rouge1: 34.5672
|