<!-- 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. -->
t5-base_cola_dense_mare_mlp_einsum
This model is a fine-tuned version of lukeleeai/t5-base_cola_densedense_baseline on the glue dataset. It achieves the following results on the evaluation set:
- Loss: 0.7682
- Accuracy: 0.7517
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 32
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 20
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5856 | 0.19 | 50 | 0.6260 | 0.6913 |
| 0.5836 | 0.37 | 100 | 0.6029 | 0.6913 |
| 0.5724 | 0.56 | 150 | 0.6055 | 0.6932 |
| 0.6635 | 0.75 | 200 | 0.6171 | 0.6922 |
| 0.5634 | 0.93 | 250 | 0.6162 | 0.6999 |
| 0.5361 | 1.12 | 300 | 0.6142 | 0.6932 |
| 0.5426 | 1.31 | 350 | 0.5920 | 0.7057 |
| 0.6255 | 1.5 | 400 | 0.5884 | 0.7095 |
| 0.6312 | 1.68 | 450 | 0.5723 | 0.7095 |
| 0.5686 | 1.87 | 500 | 0.5894 | 0.7057 |
| 0.5486 | 2.06 | 550 | 0.5590 | 0.7124 |
| 0.4436 | 2.24 | 600 | 0.5838 | 0.7220 |
| 0.4405 | 2.43 | 650 | 0.6176 | 0.7315 |
| 0.4785 | 2.62 | 700 | 0.6236 | 0.7296 |
| 0.5759 | 2.8 | 750 | 0.6233 | 0.7191 |
| 0.6156 | 2.99 | 800 | 0.6807 | 0.7392 |
| 0.4843 | 3.18 | 850 | 0.6337 | 0.7373 |
| 0.5408 | 3.36 | 900 | 0.7107 | 0.7392 |
| 0.4327 | 3.55 | 950 | 0.6256 | 0.7239 |
| 0.4318 | 3.74 | 1000 | 0.6951 | 0.7478 |
| 0.4047 | 3.93 | 1050 | 0.6566 | 0.7430 |
| 0.423 | 4.11 | 1100 | 0.6731 | 0.7440 |
| 0.3919 | 4.3 | 1150 | 0.6750 | 0.7392 |
| 0.4041 | 4.49 | 1200 | 0.6464 | 0.7421 |
| 0.3941 | 4.67 | 1250 | 0.6580 | 0.7517 |
| 0.3834 | 4.86 | 1300 | 0.6257 | 0.7459 |
| 0.2678 | 5.05 | 1350 | 0.6464 | 0.7555 |
| 0.3202 | 5.23 | 1400 | 0.7048 | 0.7507 |
| 0.2869 | 5.42 | 1450 | 0.7405 | 0.7565 |
| 0.3359 | 5.61 | 1500 | 0.6393 | 0.7593 |
| 0.3528 | 5.79 | 1550 | 0.6249 | 0.7555 |
| 0.3304 | 5.98 | 1600 | 0.6349 | 0.7565 |
| 0.2862 | 6.17 | 1650 | 0.7497 | 0.7670 |
| 0.2315 | 6.36 | 1700 | 0.7787 | 0.7622 |
| 0.3251 | 6.54 | 1750 | 0.7038 | 0.7555 |
| 0.3584 | 6.73 | 1800 | 0.7732 | 0.7603 |
| 0.1804 | 6.92 | 1850 | 0.8226 | 0.7584 |
| 0.2264 | 7.1 | 1900 | 0.7420 | 0.7613 |
| 0.2374 | 7.29 | 1950 | 0.7825 | 0.7507 |
| 0.203 | 7.48 | 2000 | 0.7575 | 0.7641 |
| 0.238 | 7.66 | 2050 | 1.9945 | 0.7603 |
| 0.2328 | 7.85 | 2100 | 0.7682 | 0.7517 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.9.0
- Tokenizers 0.11.6
