|
|
--- |
|
|
library_name: transformers |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- accuracy |
|
|
model-index: |
|
|
- name: trainer |
|
|
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. --> |
|
|
|
|
|
# trainer |
|
|
|
|
|
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.3546 |
|
|
- Accuracy: 0.8807 |
|
|
|
|
|
## 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: 0.032227 |
|
|
- train_batch_size: 512 |
|
|
- eval_batch_size: 512 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 8 |
|
|
- total_train_batch_size: 4096 |
|
|
- optimizer: Use OptimizerNames.SCHEDULE_FREE_ADAMW with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
|
- lr_scheduler_type: constant |
|
|
- lr_scheduler_warmup_steps: 1000 |
|
|
- training_steps: 1000000 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|
|:-------------:|:------:|:----:|:---------------:|:--------:| |
|
|
| No log | 0 | 0 | 4.3903 | 0.0137 | |
|
|
| No log | 0.0044 | 122 | 1.1251 | 0.6574 | |
|
|
| No log | 0.0087 | 244 | 0.8266 | 0.7365 | |
|
|
| No log | 0.0131 | 366 | 0.7493 | 0.7590 | |
|
|
| No log | 0.0175 | 488 | 0.6913 | 0.7755 | |
|
|
| 9.1782 | 0.0218 | 610 | 0.6348 | 0.7927 | |
|
|
| 9.1782 | 0.0262 | 732 | 0.5897 | 0.8064 | |
|
|
| 9.1782 | 0.0306 | 854 | 0.5569 | 0.8170 | |
|
|
| 9.1782 | 0.0349 | 976 | 0.5262 | 0.8266 | |
|
|
| 5.0917 | 0.0393 | 1098 | 0.4957 | 0.8360 | |
|
|
| 5.0917 | 0.0437 | 1220 | 0.4761 | 0.8424 | |
|
|
| 5.0917 | 0.0480 | 1342 | 0.4616 | 0.8464 | |
|
|
| 5.0917 | 0.0524 | 1464 | 0.4479 | 0.8510 | |
|
|
| 4.0398 | 0.0568 | 1586 | 0.4397 | 0.8536 | |
|
|
| 4.0398 | 0.0611 | 1708 | 0.4293 | 0.8564 | |
|
|
| 4.0398 | 0.0655 | 1830 | 0.4231 | 0.8592 | |
|
|
| 4.0398 | 0.0699 | 1952 | 0.4139 | 0.8614 | |
|
|
| 3.5268 | 0.0743 | 2074 | 0.4088 | 0.8635 | |
|
|
| 3.5268 | 0.0786 | 2196 | 0.4035 | 0.8649 | |
|
|
| 3.5268 | 0.0830 | 2318 | 0.4000 | 0.8666 | |
|
|
| 3.5268 | 0.0874 | 2440 | 0.3950 | 0.8678 | |
|
|
| 3.3084 | 0.0917 | 2562 | 0.3915 | 0.8688 | |
|
|
| 3.3084 | 0.0961 | 2684 | 0.3866 | 0.8705 | |
|
|
| 3.3084 | 0.1005 | 2806 | 0.3843 | 0.8712 | |
|
|
| 3.3084 | 0.1048 | 2928 | 0.3804 | 0.8726 | |
|
|
| 3.1769 | 0.1092 | 3050 | 0.3776 | 0.8733 | |
|
|
| 3.1769 | 0.1136 | 3172 | 0.3729 | 0.8749 | |
|
|
| 3.1769 | 0.1179 | 3294 | 0.3723 | 0.8751 | |
|
|
| 3.1769 | 0.1223 | 3416 | 0.3698 | 0.8759 | |
|
|
| 3.0785 | 0.1267 | 3538 | 0.3659 | 0.8772 | |
|
|
| 3.0785 | 0.1310 | 3660 | 0.3644 | 0.8775 | |
|
|
| 3.0785 | 0.1354 | 3782 | 0.3599 | 0.8788 | |
|
|
| 3.0785 | 0.1398 | 3904 | 0.3584 | 0.8794 | |
|
|
| 2.9831 | 0.1441 | 4026 | 0.3567 | 0.8800 | |
|
|
| 2.9831 | 0.1485 | 4148 | 0.3528 | 0.8817 | |
|
|
| 2.9831 | 0.1529 | 4270 | 0.3535 | 0.8811 | |
|
|
| 2.9831 | 0.1572 | 4392 | 0.3541 | 0.8809 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.52.2 |
|
|
- Pytorch 2.8.0.dev20250521+cu128 |
|
|
- Datasets 3.6.0 |
|
|
- Tokenizers 0.21.1 |
|
|
|
