| --- |
| library_name: transformers |
| tags: |
| - generated_from_trainer |
| model-index: |
| - name: patchtst-tsmixup |
| 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. --> |
|
|
| # patchtst-tsmixup |
|
|
| 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: 0.1553 |
| - Mse: 280.0361 |
| - Mae: 0.6489 |
| - Rmse: 16.7343 |
| - Smape: 100.3318 |
|
|
| ## 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: 0.0001 |
| - train_batch_size: 256 |
| - eval_batch_size: 512 |
| - seed: 42 |
| - gradient_accumulation_steps: 2 |
| - total_train_batch_size: 512 |
| - 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: 1000 |
| - num_epochs: 100 |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | Mse | Mae | Rmse | Smape | |
| |:-------------:|:------:|:-----:|:---------------:|:--------:|:------:|:-------:|:--------:| |
| | 0.1797 | 0.0952 | 1000 | 0.1756 | 447.3596 | 0.7397 | 21.1509 | 90.8971 | |
| | 0.1709 | 0.1904 | 2000 | 0.1691 | 425.0924 | 0.7153 | 20.6178 | 112.3049 | |
| | 0.1722 | 0.2857 | 3000 | 0.1662 | 516.2153 | 0.7009 | 22.7204 | 89.5236 | |
| | 0.1694 | 0.3809 | 4000 | 0.1643 | 321.2047 | 0.6708 | 17.9222 | 93.0515 | |
| | 0.1648 | 0.4761 | 5000 | 0.1626 | 350.6870 | 0.6731 | 18.7266 | 94.0748 | |
| | 0.1672 | 0.5713 | 6000 | 0.1612 | 370.8825 | 0.6797 | 19.2583 | 84.6619 | |
| | 0.1623 | 0.6666 | 7000 | 0.1605 | 400.0790 | 0.6715 | 20.0020 | 89.7598 | |
| | 0.1638 | 0.7618 | 8000 | 0.1613 | 387.6971 | 0.6771 | 19.6900 | 122.3799 | |
| | 0.1609 | 0.8570 | 9000 | 0.1602 | 335.3427 | 0.6603 | 18.3124 | 109.3877 | |
| | 0.1618 | 0.9522 | 10000 | 0.1592 | 318.1492 | 0.6688 | 17.8367 | 76.3322 | |
| | 0.1588 | 1.0474 | 11000 | 0.1586 | 345.3675 | 0.6628 | 18.5841 | 94.5032 | |
| | 0.1601 | 1.1426 | 12000 | 0.1580 | 326.8865 | 0.6540 | 18.0800 | 81.2504 | |
| | 0.1585 | 1.2379 | 13000 | 0.1575 | 279.7964 | 0.6532 | 16.7271 | 107.6181 | |
| | 0.1567 | 1.3331 | 14000 | 0.1575 | 328.3490 | 0.6622 | 18.1204 | 91.9899 | |
| | 0.1592 | 1.4283 | 15000 | 0.1567 | 376.8973 | 0.6523 | 19.4138 | 89.7952 | |
| | 0.16 | 1.5235 | 16000 | 0.1576 | 327.5271 | 0.6580 | 18.0977 | 105.7316 | |
| | 0.1586 | 1.6188 | 17000 | 0.1568 | 399.5775 | 0.6602 | 19.9894 | 88.6057 | |
| | 0.1593 | 1.7140 | 18000 | 0.1565 | 359.5630 | 0.6604 | 18.9621 | 325.5064 | |
| | 0.1562 | 1.8092 | 19000 | 0.1566 | 281.2739 | 0.6545 | 16.7712 | 80.4528 | |
| | 0.1601 | 1.9044 | 20000 | 0.1570 | 287.3577 | 0.6543 | 16.9516 | 79.5544 | |
| | 0.1551 | 1.9997 | 21000 | 0.1561 | 279.2150 | 0.6444 | 16.7097 | 102.6016 | |
| | 0.1532 | 2.0948 | 22000 | 0.1554 | 282.9574 | 0.6454 | 16.8213 | 85.0121 | |
| | 0.1564 | 2.1901 | 23000 | 0.1554 | 332.3758 | 0.6485 | 18.2312 | 76.0350 | |
| | 0.1568 | 2.2853 | 24000 | 0.1551 | 356.0441 | 0.6528 | 18.8691 | 92.2597 | |
| | 0.1569 | 2.3805 | 25000 | 0.1562 | 333.3135 | 0.6536 | 18.2569 | 180.8556 | |
| | 0.1569 | 2.4757 | 26000 | 0.1551 | 291.0384 | 0.6491 | 17.0598 | 80.7309 | |
| | 0.1532 | 2.5710 | 27000 | 0.1553 | 280.0361 | 0.6489 | 16.7343 | 100.3318 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.51.3 |
| - Pytorch 2.7.1+cu126 |
| - Datasets 2.17.1 |
| - Tokenizers 0.21.1 |
|
|
