|
--- |
|
license: mit |
|
base_model: lemon-mint/LaBSE-EnKo-Nano-Preview-v0.3 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- precision |
|
- recall |
|
- accuracy |
|
model-index: |
|
- name: ko-edu-classifier |
|
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. --> |
|
|
|
It's a training checkpoint. I strongly recommend not to use this model 🤗 |
|
|
|
# ko-edu-classifier |
|
|
|
This model is a fine-tuned version of [lemon-mint/LaBSE-EnKo-Nano-Preview-v0.3](https://huggingface.co/lemon-mint/LaBSE-EnKo-Nano-Preview-v0.3) on the None dataset. |
|
|
|
## 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: 0.0003 |
|
- train_batch_size: 256 |
|
- eval_batch_size: 256 |
|
- seed: 0 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_ratio: 0.2 |
|
- num_epochs: 30 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 Macro | Accuracy | |
|
|:-------------:|:-------:|:----:|:---------------:|:---------:|:------:|:--------:|:--------:| |
|
| 8.2847 | 0.9922 | 128 | 5.7695 | 0.0554 | 0.1667 | 0.0832 | 0.3326 | |
|
| 2.9466 | 1.9845 | 256 | 2.4992 | 0.0297 | 0.1667 | 0.0504 | 0.1783 | |
|
| 2.2442 | 2.9767 | 384 | 2.2770 | 0.0972 | 0.1779 | 0.0789 | 0.1884 | |
|
| 2.11 | 3.9690 | 512 | 2.2539 | 0.1370 | 0.1917 | 0.1233 | 0.1966 | |
|
| 2.0444 | 4.9612 | 640 | 1.9768 | 0.2723 | 0.2069 | 0.1448 | 0.2171 | |
|
| 2.0458 | 5.9535 | 768 | 2.1823 | 0.1460 | 0.2022 | 0.1450 | 0.2021 | |
|
| 2.0249 | 6.9457 | 896 | 2.0237 | 0.2773 | 0.2019 | 0.1478 | 0.2062 | |
|
| 2.0141 | 7.9380 | 1024 | 2.0108 | 0.3220 | 0.2043 | 0.1498 | 0.2081 | |
|
| 2.0178 | 8.9302 | 1152 | 1.9606 | 0.2890 | 0.2066 | 0.1513 | 0.2127 | |
|
| 2.0145 | 9.9225 | 1280 | 2.0984 | 0.3189 | 0.2077 | 0.1561 | 0.2062 | |
|
| 2.0093 | 10.9147 | 1408 | 1.9506 | 0.2829 | 0.2089 | 0.1517 | 0.2157 | |
|
| 2.014 | 11.9070 | 1536 | 1.9494 | 0.3039 | 0.2086 | 0.1538 | 0.2152 | |
|
| 2.0137 | 12.8992 | 1664 | 1.9247 | 0.3109 | 0.2110 | 0.1548 | 0.2190 | |
|
| 2.0055 | 13.8915 | 1792 | 1.8977 | 0.3184 | 0.2121 | 0.1537 | 0.2223 | |
|
| 2.0058 | 14.8837 | 1920 | 1.9747 | 0.3245 | 0.2094 | 0.1539 | 0.2130 | |
|
| 1.9975 | 15.8760 | 2048 | 1.9288 | 0.3084 | 0.2109 | 0.1535 | 0.2187 | |
|
| 1.995 | 16.8682 | 2176 | 1.8964 | 0.3036 | 0.2142 | 0.1590 | 0.2247 | |
|
| 1.9959 | 17.8605 | 2304 | 1.9247 | 0.3164 | 0.2144 | 0.1605 | 0.2209 | |
|
| 2.003 | 18.8527 | 2432 | 1.9297 | 0.3152 | 0.2151 | 0.1595 | 0.2217 | |
|
| 1.9908 | 19.8450 | 2560 | 1.8936 | 0.3065 | 0.2144 | 0.1610 | 0.2256 | |
|
| 1.9843 | 20.8372 | 2688 | 1.9238 | 0.3201 | 0.2168 | 0.1613 | 0.2242 | |
|
| 2.0042 | 21.8295 | 2816 | 1.9712 | 0.3228 | 0.2095 | 0.1577 | 0.2119 | |
|
| 1.9913 | 22.8217 | 2944 | 1.9070 | 0.3134 | 0.2168 | 0.1612 | 0.2250 | |
|
| 1.9855 | 23.8140 | 3072 | 1.9155 | 0.3123 | 0.2166 | 0.1611 | 0.2242 | |
|
| 1.9892 | 24.8062 | 3200 | 1.9338 | 0.3213 | 0.2163 | 0.1619 | 0.2220 | |
|
| 1.9964 | 25.7984 | 3328 | 1.9309 | 0.3125 | 0.2167 | 0.1625 | 0.2226 | |
|
| 1.9704 | 26.7907 | 3456 | 1.9165 | 0.3101 | 0.2187 | 0.1648 | 0.2258 | |
|
| 1.9977 | 27.7829 | 3584 | 1.9165 | 0.3177 | 0.2193 | 0.1653 | 0.2264 | |
|
| 1.9976 | 28.7752 | 3712 | 1.9127 | 0.3099 | 0.2191 | 0.1643 | 0.2269 | |
|
| 1.9728 | 29.7674 | 3840 | 1.9129 | 0.3096 | 0.2186 | 0.1640 | 0.2264 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.43.3 |
|
- Pytorch 2.4.0+cu118 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|
