roberta-similarity
This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5172
- Accuracy: 0.834
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.6747 | 0.16 | 10 | 0.6562 | 0.672 |
0.5355 | 0.32 | 20 | 0.5163 | 0.772 |
0.5374 | 0.48 | 30 | 0.5901 | 0.74 |
0.5064 | 0.63 | 40 | 0.4904 | 0.782 |
0.4241 | 0.79 | 50 | 0.5793 | 0.73 |
0.5484 | 0.95 | 60 | 0.5381 | 0.776 |
0.5441 | 1.11 | 70 | 0.5375 | 0.764 |
0.445 | 1.27 | 80 | 0.5096 | 0.792 |
0.4436 | 1.43 | 90 | 0.5617 | 0.814 |
0.4677 | 1.59 | 100 | 0.6145 | 0.796 |
0.4306 | 1.75 | 110 | 0.6105 | 0.814 |
0.3197 | 1.9 | 120 | 0.5112 | 0.772 |
0.3373 | 2.06 | 130 | 0.5168 | 0.818 |
0.3128 | 2.22 | 140 | 0.5007 | 0.824 |
0.3286 | 2.38 | 150 | 0.4900 | 0.83 |
0.476 | 2.54 | 160 | 0.4989 | 0.79 |
0.413 | 2.7 | 170 | 0.6129 | 0.748 |
0.3811 | 2.86 | 180 | 0.5137 | 0.818 |
0.3224 | 3.02 | 190 | 0.5178 | 0.806 |
0.2917 | 3.17 | 200 | 0.5382 | 0.802 |
0.3696 | 3.33 | 210 | 0.5610 | 0.822 |
0.3019 | 3.49 | 220 | 0.7040 | 0.792 |
0.3354 | 3.65 | 230 | 0.5342 | 0.826 |
0.2854 | 3.81 | 240 | 0.5047 | 0.832 |
0.3079 | 3.97 | 250 | 0.5124 | 0.83 |
0.3271 | 4.13 | 260 | 0.5876 | 0.808 |
0.276 | 4.29 | 270 | 0.5271 | 0.824 |
0.2519 | 4.44 | 280 | 0.5309 | 0.832 |
0.2107 | 4.6 | 290 | 0.5186 | 0.834 |
0.2471 | 4.76 | 300 | 0.5191 | 0.838 |
0.2751 | 4.92 | 310 | 0.5172 | 0.834 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Tokenizers 0.13.3
- Downloads last month
- 8
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.