math_question_grade_detection_Bert_databalanced
This model is a fine-tuned version of google-bert/bert-large-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6880
- Accuracy: 0.7603
- Precision: 0.7651
- Recall: 0.7603
- F1: 0.7588
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: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 1100
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
---|---|---|---|---|---|---|---|
No log | 0.2817 | 50 | 2.1003 | 0.2349 | 0.3799 | 0.2349 | 0.2106 |
No log | 0.5634 | 100 | 1.9607 | 0.2762 | 0.3337 | 0.2762 | 0.2498 |
No log | 0.8451 | 150 | 1.5031 | 0.4778 | 0.4633 | 0.4778 | 0.4591 |
No log | 1.1268 | 200 | 1.2546 | 0.5460 | 0.5596 | 0.5460 | 0.5176 |
No log | 1.4085 | 250 | 1.0941 | 0.5746 | 0.5804 | 0.5746 | 0.5675 |
No log | 1.6901 | 300 | 0.9381 | 0.6730 | 0.6943 | 0.6730 | 0.6721 |
No log | 1.9718 | 350 | 0.8974 | 0.6619 | 0.6822 | 0.6619 | 0.6570 |
No log | 2.2535 | 400 | 0.8243 | 0.6889 | 0.6913 | 0.6889 | 0.6856 |
No log | 2.5352 | 450 | 0.8219 | 0.6937 | 0.7131 | 0.6937 | 0.6881 |
1.2537 | 2.8169 | 500 | 0.7642 | 0.7159 | 0.7239 | 0.7159 | 0.7121 |
1.2537 | 3.0986 | 550 | 0.7580 | 0.7175 | 0.7197 | 0.7175 | 0.7068 |
1.2537 | 3.3803 | 600 | 0.7310 | 0.7397 | 0.7523 | 0.7397 | 0.7387 |
1.2537 | 3.6620 | 650 | 0.7562 | 0.7413 | 0.7466 | 0.7413 | 0.7349 |
1.2537 | 3.9437 | 700 | 0.6512 | 0.7730 | 0.7792 | 0.7730 | 0.7726 |
1.2537 | 4.2254 | 750 | 0.6941 | 0.7476 | 0.7484 | 0.7476 | 0.7447 |
1.2537 | 4.5070 | 800 | 0.6866 | 0.7571 | 0.7607 | 0.7571 | 0.7550 |
1.2537 | 4.7887 | 850 | 0.6942 | 0.7603 | 0.7644 | 0.7603 | 0.7588 |
1.2537 | 5.0704 | 900 | 0.7230 | 0.7683 | 0.7821 | 0.7683 | 0.7656 |
1.2537 | 5.3521 | 950 | 0.7123 | 0.7603 | 0.7669 | 0.7603 | 0.7588 |
0.321 | 5.6338 | 1000 | 0.6939 | 0.7667 | 0.7725 | 0.7667 | 0.7652 |
0.321 | 5.9155 | 1050 | 0.6884 | 0.7667 | 0.7723 | 0.7667 | 0.7657 |
0.321 | 6.1972 | 1100 | 0.6880 | 0.7603 | 0.7651 | 0.7603 | 0.7588 |
Framework versions
- Transformers 4.46.3
- Pytorch 2.4.0
- Datasets 3.1.0
- Tokenizers 0.20.3
- Downloads last month
- 10
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for nzm97/math_question_grade_detection_Bert_databalanced
Base model
google-bert/bert-large-uncased