salt_language_Classification
This model is a fine-tuned version of google/t5-efficient-tiny on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.0615
- Accuracy: 0.9782
- Precision: 0.9787
- Recall: 0.9782
- F1: 0.9782
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: 0.001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- training_steps: 20000
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
---|---|---|---|---|---|---|---|
0.2011 | 0.025 | 500 | 0.4979 | 0.8733 | 0.9001 | 0.8733 | 0.8714 |
0.234 | 0.05 | 1000 | 0.1886 | 0.9345 | 0.9354 | 0.9345 | 0.9345 |
0.2083 | 0.075 | 1500 | 0.1833 | 0.9328 | 0.9391 | 0.9328 | 0.9328 |
0.1838 | 0.1 | 2000 | 0.1457 | 0.9476 | 0.9479 | 0.9476 | 0.9475 |
0.1737 | 0.125 | 2500 | 0.1659 | 0.9409 | 0.9438 | 0.9409 | 0.9411 |
0.1591 | 0.15 | 3000 | 0.1450 | 0.9516 | 0.9524 | 0.9516 | 0.9517 |
0.1571 | 0.175 | 3500 | 0.1351 | 0.9459 | 0.9485 | 0.9459 | 0.9461 |
0.1513 | 0.2 | 4000 | 0.1510 | 0.9456 | 0.9515 | 0.9456 | 0.9460 |
0.1439 | 0.225 | 4500 | 0.1339 | 0.9546 | 0.9578 | 0.9546 | 0.9547 |
0.1394 | 0.25 | 5000 | 0.1052 | 0.9657 | 0.9658 | 0.9657 | 0.9656 |
0.1472 | 0.275 | 5500 | 0.1088 | 0.9610 | 0.9629 | 0.9610 | 0.9609 |
0.1385 | 0.3 | 6000 | 0.0792 | 0.9694 | 0.9696 | 0.9694 | 0.9694 |
0.1349 | 0.325 | 6500 | 0.1063 | 0.9610 | 0.9632 | 0.9610 | 0.9613 |
0.1215 | 0.35 | 7000 | 0.0855 | 0.9688 | 0.9694 | 0.9688 | 0.9687 |
0.133 | 0.375 | 7500 | 0.1049 | 0.9630 | 0.9640 | 0.9630 | 0.9630 |
0.1226 | 0.4 | 8000 | 0.0938 | 0.9667 | 0.9675 | 0.9667 | 0.9667 |
0.1222 | 0.425 | 8500 | 0.1134 | 0.9570 | 0.9604 | 0.9570 | 0.9573 |
0.1165 | 0.45 | 9000 | 0.0997 | 0.9688 | 0.9697 | 0.9688 | 0.9687 |
0.1174 | 0.475 | 9500 | 0.1002 | 0.9661 | 0.9680 | 0.9661 | 0.9659 |
0.1165 | 0.5 | 10000 | 0.0807 | 0.9728 | 0.9728 | 0.9728 | 0.9728 |
0.1065 | 0.525 | 10500 | 0.0750 | 0.9745 | 0.9754 | 0.9745 | 0.9746 |
0.1089 | 0.55 | 11000 | 0.0896 | 0.9688 | 0.9703 | 0.9688 | 0.9689 |
0.1125 | 0.575 | 11500 | 0.0632 | 0.9782 | 0.9787 | 0.9782 | 0.9782 |
0.11 | 0.6 | 12000 | 0.0775 | 0.9691 | 0.9708 | 0.9691 | 0.9692 |
0.1028 | 0.625 | 12500 | 0.0833 | 0.9698 | 0.9708 | 0.9698 | 0.9698 |
0.1052 | 0.65 | 13000 | 0.0663 | 0.9751 | 0.9755 | 0.9751 | 0.9751 |
0.1068 | 0.675 | 13500 | 0.0648 | 0.9772 | 0.9774 | 0.9772 | 0.9772 |
0.1029 | 0.7 | 14000 | 0.0962 | 0.9688 | 0.9706 | 0.9688 | 0.9689 |
0.1014 | 0.725 | 14500 | 0.0686 | 0.9772 | 0.9775 | 0.9772 | 0.9771 |
0.0978 | 0.75 | 15000 | 0.0802 | 0.9745 | 0.9752 | 0.9745 | 0.9745 |
0.095 | 0.775 | 15500 | 0.0646 | 0.9758 | 0.9763 | 0.9758 | 0.9758 |
0.0996 | 0.8 | 16000 | 0.0711 | 0.9758 | 0.9761 | 0.9758 | 0.9758 |
0.0967 | 0.825 | 16500 | 0.0683 | 0.9761 | 0.9768 | 0.9761 | 0.9761 |
0.0939 | 0.85 | 17000 | 0.0572 | 0.9792 | 0.9795 | 0.9792 | 0.9791 |
0.0966 | 0.875 | 17500 | 0.0527 | 0.9792 | 0.9794 | 0.9792 | 0.9791 |
0.0925 | 0.9 | 18000 | 0.0581 | 0.9798 | 0.9802 | 0.9798 | 0.9799 |
0.0945 | 0.925 | 18500 | 0.0693 | 0.9768 | 0.9776 | 0.9768 | 0.9768 |
0.0923 | 0.95 | 19000 | 0.0615 | 0.9785 | 0.9790 | 0.9785 | 0.9785 |
0.0896 | 0.975 | 19500 | 0.0643 | 0.9758 | 0.9766 | 0.9758 | 0.9758 |
0.0979 | 1.0 | 20000 | 0.0619 | 0.9765 | 0.9770 | 0.9765 | 0.9765 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 426
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for yigagilbert/salt_language_Classification
Base model
google/t5-efficient-tiny
Finetuned
this model
Evaluation results
- Accuracy on generatorself-reported0.978
- Precision on generatorself-reported0.979
- Recall on generatorself-reported0.978
- F1 on generatorself-reported0.978