Arabic T5v1.1 for question paraphrasing
This is a fine-tuned arabic-t5-small on the task of question paraphrasing.
A demo of the trained model using HF Spaces can be found here
Training data
The model was fine-tuned using the Semantic Question Similarity in Arabic data on kaggle.
Only the rows of the dataset where the label is True (the two questions have the same meaning) were taken.
The training data was then also mirrored; so if q1 and q2 were two questions with the same meaning, then (q1, q2) and (q2, q1) were both present in the training set. The evaluation set was kept unmirrored of course.
Training config
batch size |
128 |
dropout rate |
0.1 |
learning rate |
0.001 |
lr schedule |
constant |
weight decay |
1e-7 |
epochs |
3 |
Results
training loss |
0.7086 |
evaluation loss |
0.9819 |
meteor |
49.277 |
sacreBLEU-1 |
57.088 |
sacreBLEU-2 |
39.846 |
sacreBLEU-3 |
29.444 |
sacreBLEU-4 |
22.601 |
Rouge F1 max |
1.299 |