<!-- 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. -->
EElayoutlmv3_jordyvl_rvl_cdip_100_examples_per_class_2023-10-29_two_txt_vis_conc_9_10_11_12_ramp
This model is a fine-tuned version of jordyvl/LayoutLMv3_RVL-CDIP_NK100 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2198
- Accuracy: 0.7975
- Exit 0 Accuracy: 0.065
- Exit 1 Accuracy: 0.0625
- Exit 2 Accuracy: 0.0625
- Exit 3 Accuracy: 0.0625
- Exit 4 Accuracy: 0.175
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: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 24
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 60
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Exit 0 Accuracy | Exit 1 Accuracy | Exit 2 Accuracy | Exit 3 Accuracy | Exit 4 Accuracy |
---|---|---|---|---|---|---|---|---|---|
No log | 0.96 | 16 | 0.9759 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 1.98 | 33 | 0.9941 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 3.0 | 50 | 1.0106 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 3.96 | 66 | 1.0257 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 4.98 | 83 | 1.0394 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 6.0 | 100 | 1.0508 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 6.96 | 116 | 1.0600 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 7.98 | 133 | 1.0698 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 9.0 | 150 | 1.0790 | 0.8 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 9.96 | 166 | 1.0860 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 10.98 | 183 | 1.0935 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 12.0 | 200 | 1.1010 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 12.96 | 216 | 1.1067 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 13.98 | 233 | 1.1129 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 15.0 | 250 | 1.1187 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 15.96 | 266 | 1.1248 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 16.98 | 283 | 1.1310 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 18.0 | 300 | 1.1368 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 18.96 | 316 | 1.1411 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 19.98 | 333 | 1.1460 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 21.0 | 350 | 1.1514 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 21.96 | 366 | 1.1557 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 22.98 | 383 | 1.1606 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 24.0 | 400 | 1.1636 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 24.96 | 416 | 1.1672 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 25.98 | 433 | 1.1715 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 27.0 | 450 | 1.1747 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 27.96 | 466 | 1.1783 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
No log | 28.98 | 483 | 1.1819 | 0.795 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 30.0 | 500 | 1.1847 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 30.96 | 516 | 1.1871 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 31.98 | 533 | 1.1900 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 33.0 | 550 | 1.1930 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 33.96 | 566 | 1.1952 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 34.98 | 583 | 1.1966 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 36.0 | 600 | 1.1980 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 36.96 | 616 | 1.1996 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 37.98 | 633 | 1.2012 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 39.0 | 650 | 1.2028 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 39.96 | 666 | 1.2044 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 40.98 | 683 | 1.2063 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 42.0 | 700 | 1.2077 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 42.96 | 716 | 1.2091 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 43.98 | 733 | 1.2106 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 45.0 | 750 | 1.2121 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 45.96 | 766 | 1.2133 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 46.98 | 783 | 1.2145 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 48.0 | 800 | 1.2157 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 48.96 | 816 | 1.2165 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 49.98 | 833 | 1.2171 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 51.0 | 850 | 1.2176 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 51.96 | 866 | 1.2181 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 52.98 | 883 | 1.2187 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 54.0 | 900 | 1.2191 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 54.96 | 916 | 1.2194 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 55.98 | 933 | 1.2196 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 57.0 | 950 | 1.2197 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
2.1088 | 57.6 | 960 | 1.2198 | 0.7975 | 0.065 | 0.0625 | 0.0625 | 0.0625 | 0.175 |
Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3