|
--- |
|
license: mit |
|
base_model: prajjwal1/bert-tiny |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: tinybert_29_med_intents |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# tinybert_29_med_intents |
|
|
|
This model is a fine-tuned version of [prajjwal1/bert-tiny](https://huggingface.co/prajjwal1/bert-tiny) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3344 |
|
- Accuracy: 0.9199 |
|
|
|
## 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: 16 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 50 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:-----:|:---------------:|:--------:| |
|
| No log | 1.0 | 430 | 2.8232 | 0.4144 | |
|
| 3.1061 | 2.0 | 860 | 2.4220 | 0.4890 | |
|
| 2.6532 | 3.0 | 1290 | 2.0921 | 0.5967 | |
|
| 2.28 | 4.0 | 1720 | 1.8178 | 0.6878 | |
|
| 1.9726 | 5.0 | 2150 | 1.5987 | 0.7431 | |
|
| 1.7268 | 6.0 | 2580 | 1.4221 | 0.7569 | |
|
| 1.5454 | 7.0 | 3010 | 1.2797 | 0.7762 | |
|
| 1.5454 | 8.0 | 3440 | 1.1608 | 0.7818 | |
|
| 1.3826 | 9.0 | 3870 | 1.0589 | 0.8039 | |
|
| 1.2445 | 10.0 | 4300 | 0.9737 | 0.8177 | |
|
| 1.1266 | 11.0 | 4730 | 0.8920 | 0.8343 | |
|
| 1.0328 | 12.0 | 5160 | 0.8279 | 0.8398 | |
|
| 0.9528 | 13.0 | 5590 | 0.7646 | 0.8453 | |
|
| 0.8538 | 14.0 | 6020 | 0.7186 | 0.8564 | |
|
| 0.8538 | 15.0 | 6450 | 0.6733 | 0.8619 | |
|
| 0.7987 | 16.0 | 6880 | 0.6347 | 0.8812 | |
|
| 0.7367 | 17.0 | 7310 | 0.5945 | 0.8840 | |
|
| 0.6931 | 18.0 | 7740 | 0.5674 | 0.8950 | |
|
| 0.6339 | 19.0 | 8170 | 0.5429 | 0.9061 | |
|
| 0.606 | 20.0 | 8600 | 0.5132 | 0.9033 | |
|
| 0.5647 | 21.0 | 9030 | 0.4991 | 0.9061 | |
|
| 0.5647 | 22.0 | 9460 | 0.4709 | 0.9033 | |
|
| 0.5375 | 23.0 | 9890 | 0.4642 | 0.9116 | |
|
| 0.4961 | 24.0 | 10320 | 0.4421 | 0.9116 | |
|
| 0.4695 | 25.0 | 10750 | 0.4390 | 0.9088 | |
|
| 0.4499 | 26.0 | 11180 | 0.4126 | 0.9088 | |
|
| 0.4315 | 27.0 | 11610 | 0.4149 | 0.9088 | |
|
| 0.4005 | 28.0 | 12040 | 0.4036 | 0.9116 | |
|
| 0.4005 | 29.0 | 12470 | 0.3938 | 0.9033 | |
|
| 0.3929 | 30.0 | 12900 | 0.3846 | 0.9061 | |
|
| 0.3707 | 31.0 | 13330 | 0.3856 | 0.9116 | |
|
| 0.369 | 32.0 | 13760 | 0.3727 | 0.9088 | |
|
| 0.3517 | 33.0 | 14190 | 0.3739 | 0.9088 | |
|
| 0.3355 | 34.0 | 14620 | 0.3604 | 0.9088 | |
|
| 0.3226 | 35.0 | 15050 | 0.3518 | 0.9144 | |
|
| 0.3226 | 36.0 | 15480 | 0.3570 | 0.9116 | |
|
| 0.3197 | 37.0 | 15910 | 0.3502 | 0.9144 | |
|
| 0.3038 | 38.0 | 16340 | 0.3463 | 0.9144 | |
|
| 0.3038 | 39.0 | 16770 | 0.3448 | 0.9116 | |
|
| 0.2918 | 40.0 | 17200 | 0.3448 | 0.9144 | |
|
| 0.2937 | 41.0 | 17630 | 0.3460 | 0.9144 | |
|
| 0.2845 | 42.0 | 18060 | 0.3414 | 0.9199 | |
|
| 0.2845 | 43.0 | 18490 | 0.3412 | 0.9199 | |
|
| 0.2785 | 44.0 | 18920 | 0.3401 | 0.9227 | |
|
| 0.2781 | 45.0 | 19350 | 0.3372 | 0.9199 | |
|
| 0.2665 | 46.0 | 19780 | 0.3364 | 0.9199 | |
|
| 0.2722 | 47.0 | 20210 | 0.3352 | 0.9199 | |
|
| 0.2683 | 48.0 | 20640 | 0.3359 | 0.9199 | |
|
| 0.267 | 49.0 | 21070 | 0.3345 | 0.9199 | |
|
| 0.2641 | 50.0 | 21500 | 0.3344 | 0.9199 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.1.0+cu118 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|