<!-- 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-visquad2-3-4-2023-1
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: 75.5555
- Loss: 1.9410
- Exact: 38.9950
- F1: 56.6891
- Total: 3821
- Hasans Exact: 55.8236
- Hasans F1: 81.3075
- Hasans Total: 2653
- Noans Exact: 0.7705
- Noans F1: 0.7705
- Noans Total: 1168
- Best Exact: 59.6179
- Best Exact Thresh: 0.6305
- Best F1 Thresh: 0.9555
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: 8
- total_train_batch_size: 64
- 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 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1.7058 | 1.0 | 1055 | 63.5823 | 1.3396 | 34.9385 | 53.3844 | 3821 | 50.3204 | 76.8873 | 2653 | 0.0 | 0.0 | 1168 | 50.1963 | 0.8713 | 0.9414 |
0.5411 | 2.0 | 2110 | 72.7630 | 1.0530 | 37.5818 | 55.6036 | 3821 | 54.1274 | 80.0834 | 2653 | 0.0 | 0.0 | 1168 | 57.3672 | 0.8339 | 0.9185 |
0.3807 | 3.0 | 3166 | 75.2055 | 1.0871 | 39.8063 | 56.8927 | 3821 | 57.2936 | 81.9024 | 2653 | 0.0856 | 0.0856 | 1168 | 60.2460 | 0.9277 | 0.9277 |
0.2835 | 4.0 | 4221 | 74.8165 | 1.1504 | 39.2044 | 56.6028 | 3821 | 56.3890 | 81.4471 | 2653 | 0.1712 | 0.1712 | 1168 | 59.1992 | 0.8443 | 0.8989 |
0.218 | 5.0 | 5276 | 75.1668 | 1.2219 | 39.4138 | 57.1114 | 3821 | 56.4267 | 81.9158 | 2653 | 0.7705 | 0.7705 | 1168 | 59.6179 | 0.7805 | 0.9053 |
0.1721 | 6.0 | 6332 | 75.9461 | 1.4275 | 39.5185 | 57.1022 | 3821 | 56.3513 | 81.6765 | 2653 | 1.2842 | 1.2842 | 1168 | 60.1675 | 0.7748 | 0.9764 |
0.1325 | 7.0 | 7387 | 75.7665 | 1.4863 | 39.5708 | 57.0923 | 3821 | 56.3890 | 81.6245 | 2653 | 1.3699 | 1.3699 | 1168 | 60.2722 | 0.7701 | 0.9771 |
0.1054 | 8.0 | 8443 | 75.6723 | 1.7071 | 38.4978 | 56.5420 | 3821 | 55.2582 | 81.2466 | 2653 | 0.4281 | 0.4281 | 1168 | 59.2515 | 0.6401 | 0.9894 |
0.0858 | 9.0 | 9498 | 75.4763 | 1.8596 | 38.6025 | 56.5256 | 3821 | 55.4090 | 81.2229 | 2653 | 0.4281 | 0.4281 | 1168 | 59.2515 | 0.6451 | 0.9918 |
0.0704 | 10.0 | 10550 | 75.5555 | 1.9410 | 38.9950 | 56.6891 | 3821 | 55.8236 | 81.3075 | 2653 | 0.7705 | 0.7705 | 1168 | 59.6179 | 0.6305 | 0.9555 |
Framework versions
- Transformers 4.27.3
- Pytorch 1.13.1+cu117
- Datasets 2.10.1
- Tokenizers 0.13.2