<!-- 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. -->
xlm-roberta-large-finetuned-augument-visquad2-2-4-2023-3
This model is a fine-tuned version of xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:
- Best F1: 76.3263
- Loss: 2.9101
- Exact: 41.0887
- F1: 58.6813
- Total: 3821
- Hasans Exact: 56.0498
- Hasans F1: 81.3876
- Hasans Total: 2653
- Noans Exact: 7.1062
- Noans F1: 7.1062
- Noans Total: 1168
- Best Exact: 60.3769
- Best Exact Thresh: 0.7798
- Best F1 Thresh: 0.9874
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- lr_scheduler_warmup_steps: 1000
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Best F1 | Validation Loss | Exact | F1 | Total | Hasans Exact | Hasans F1 | Hasans Total | Noans Exact | Noans F1 | Noans Total | Best Exact | Best Exact Thresh | Best F1 Thresh |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.9242 | 1.0 | 2807 | 69.6410 | 1.0239 | 37.3201 | 55.1119 | 3821 | 53.7505 | 79.3752 | 2653 | 0.0 | 0.0 | 1168 | 55.0118 | 0.8222 | 0.8968 |
| 0.3756 | 2.0 | 5615 | 73.7526 | 1.0092 | 38.8642 | 55.8953 | 3821 | 55.9744 | 80.5035 | 2653 | 0.0 | 0.0 | 1168 | 59.4085 | 0.9128 | 0.9611 |
| 0.2595 | 3.0 | 8423 | 75.1395 | 1.0121 | 39.7278 | 56.5553 | 3821 | 57.1806 | 81.4165 | 2653 | 0.0856 | 0.0856 | 1168 | 60.6386 | 0.8138 | 0.9174 |
| 0.185 | 4.0 | 11231 | 75.2011 | 1.2309 | 39.2306 | 56.7010 | 3821 | 56.2005 | 81.3625 | 2653 | 0.6849 | 0.6849 | 1168 | 59.7749 | 0.7215 | 0.8729 |
| 0.1336 | 5.0 | 14038 | 75.0330 | 1.4052 | 38.4454 | 56.1488 | 3821 | 55.2582 | 80.7556 | 2653 | 0.2568 | 0.2568 | 1168 | 59.4085 | 0.6660 | 0.8646 |
| 0.0976 | 6.0 | 16846 | 75.4976 | 1.6109 | 38.5763 | 56.1952 | 3821 | 55.4467 | 80.8224 | 2653 | 0.2568 | 0.2568 | 1168 | 59.8534 | 0.6631 | 0.9605 |
| 0.072 | 7.0 | 19654 | 76.0690 | 1.9673 | 39.5970 | 56.9041 | 3821 | 56.0874 | 81.0142 | 2653 | 2.1404 | 2.1404 | 1168 | 60.5862 | 0.7197 | 0.9882 |
| 0.0526 | 8.0 | 22462 | 75.3652 | 2.2945 | 38.8903 | 56.5382 | 3821 | 55.3336 | 80.7511 | 2653 | 1.5411 | 1.5411 | 1168 | 59.8273 | 0.6659 | 0.9573 |
| 0.0389 | 9.0 | 25269 | 76.0674 | 2.6609 | 42.5281 | 59.8494 | 3821 | 56.0121 | 80.9591 | 2653 | 11.9007 | 11.9007 | 1168 | 60.4292 | 0.6494 | 0.9632 |
| 0.0291 | 10.0 | 28070 | 76.3263 | 2.9101 | 41.0887 | 58.6813 | 3821 | 56.0498 | 81.3876 | 2653 | 7.1062 | 7.1062 | 1168 | 60.3769 | 0.7798 | 0.9874 |
Framework versions
- Transformers 4.27.3
- Pytorch 1.13.1+cu117
- Datasets 2.10.1
- Tokenizers 0.13.2
