Loading…
Loading…
This model is a fine-tuned version of google/mobilebert-uncased on the GLUE STSB dataset. It achieves the following results on the evaluation set:
More information needed
More information needed
More information needed
The following hyperparameters were used during training:
| Training Loss | Epoch | Step | Validation Loss | Pearson | Spearmanr | Combined Score |
|---|---|---|---|---|---|---|
| 3.5023 | 1.0 | 45 | 2.4151 | 0.0290 | 0.0235 | 0.0262 |
| 2.148 | 2.0 | 90 | 2.4051 | 0.0483 | 0.0436 | 0.0460 |
| 2.1403 | 3.0 | 135 | 2.7281 | 0.0523 | 0.0502 | 0.0513 |
| 2.0908 | 4.0 | 180 | 2.5970 | 0.0511 | 0.0509 | 0.0510 |
| 2.0699 | 5.0 | 225 | 2.5728 | 0.0601 | 0.0621 | 0.0611 |
| 2.0307 | 6.0 | 270 | 2.4059 | 0.0700 | 0.0781 | 0.0740 |
| 1.9823 | 7.0 | 315 | 2.3720 | 0.0872 | 0.0832 | 0.0852 |
| 1.9283 | 8.0 | 360 | 2.4539 | 0.0836 | 0.0814 | 0.0825 |
| 1.9014 | 9.0 | 405 | 2.3975 | 0.0950 | 0.0915 | 0.0932 |
| 1.883 | 10.0 | 450 | 2.6096 | 0.0894 | 0.0876 | 0.0885 |
| 1.8453 | 11.0 | 495 | 2.4121 | 0.0970 | 0.0912 | 0.0941 |
| 1.8275 | 12.0 | 540 | 2.6028 | 0.1038 | 0.0920 | 0.0979 |