<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
xlnet-base-cased_fold_2_binary_v1
This model is a fine-tuned version of xlnet-base-cased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.8748
- F1: 0.8066
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: 2e-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: 25
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| No log | 1.0 | 290 | 0.4803 | 0.7433 |
| 0.434 | 2.0 | 580 | 0.4385 | 0.8099 |
| 0.434 | 3.0 | 870 | 0.5382 | 0.8078 |
| 0.254 | 4.0 | 1160 | 0.6944 | 0.7982 |
| 0.254 | 5.0 | 1450 | 0.9908 | 0.8058 |
| 0.1479 | 6.0 | 1740 | 1.1090 | 0.8062 |
| 0.0874 | 7.0 | 2030 | 1.2405 | 0.8042 |
| 0.0874 | 8.0 | 2320 | 1.3174 | 0.8012 |
| 0.0505 | 9.0 | 2610 | 1.5211 | 0.7909 |
| 0.0505 | 10.0 | 2900 | 1.4014 | 0.8126 |
| 0.0301 | 11.0 | 3190 | 1.4798 | 0.8047 |
| 0.0301 | 12.0 | 3480 | 1.4668 | 0.8091 |
| 0.0279 | 13.0 | 3770 | 1.5286 | 0.8075 |
| 0.0233 | 14.0 | 4060 | 1.6752 | 0.8006 |
| 0.0233 | 15.0 | 4350 | 1.5265 | 0.8132 |
| 0.019 | 16.0 | 4640 | 1.6440 | 0.7949 |
| 0.019 | 17.0 | 4930 | 1.7471 | 0.8097 |
| 0.0096 | 18.0 | 5220 | 1.7329 | 0.8121 |
| 0.0075 | 19.0 | 5510 | 1.7472 | 0.8191 |
| 0.0075 | 20.0 | 5800 | 1.8043 | 0.8161 |
| 0.0052 | 21.0 | 6090 | 1.8102 | 0.8141 |
| 0.0052 | 22.0 | 6380 | 1.7944 | 0.8116 |
| 0.0044 | 23.0 | 6670 | 1.8211 | 0.8141 |
| 0.0044 | 24.0 | 6960 | 1.8741 | 0.8066 |
| 0.0046 | 25.0 | 7250 | 1.8748 | 0.8066 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
