Downloads · 30 days
5
14% of all-time downloads
mdeberta-xlwa-en-it
pgajo/mdeberta-xlwa-en-it
mdeberta-xlwa-en-it is a question answering model from pgajo. Use it when the input is a question plus a passage. It is set up for transformers.
All-time downloads
35
Public
Parameters
278M
1.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.1 GB · 98%
At a glance
- Task
- Question Answering
- Library
- transformers
- Model type
- deberta-v2
- Access
- Public
- Created
- Feb 11, 2024
- Updated
- Feb 11, 2024
- SHA
- dfac6b62
Try a prompt
Model card
From the Hugging Face model README
Model description:
Model: mdeberta
Dataset: TASTEset
Unshuffled ratio: []
Shuffled ratio: []
Best exact match epoch: 3
Best exact match: 82.32
Best epoch: 3
Drop duplicates: []
Max epochs = 10
Optimizer lr = 3e-05
Optimizer eps = 1e-08
Batch size = 32
Dataset path = pgajo/mdeberta_xlwa_en-it
Results
| epoch | train_loss | train_f1 | train_exact | dev_loss | dev_f1 | dev_exact | test_loss | test_f1 | test_exact |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 0.32 | 91.77 | 90.75 | 1.26 | 81.86 | 80.66 | 0 | 0 | 0 |
| 2 | 0.04 | 98.92 | 98.74 | 1.67 | 82.05 | 81.26 | 0 | 0 | 0 |
| 3 | 0.02 | 99.4 | 99.29 | 1.46 | 82.58 | 82.32 | 0 | 0 | 0 |
| 4 | 0.02 | 99.53 | 99.45 | 1.73 | 82.6 | 81.97 | 0 | 0 | 0 |
| 5 | 0.01 | 99.71 | 99.65 | 1.63 | 82.05 | 81.52 | 0 | 0 | 0 |
| 6 | 0.01 | 99.66 | 99.59 | 1.86 | 82.16 | 82.02 | 0 | 0 | 0 |
- Task
- Question Answering
- Library
- transformers
- Type
- deberta-v2
- Created
- Feb 11, 2024
- Updated
- Feb 11, 2024