|
--- |
|
base_model: yihongLiu/furina |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: furina_seed42_eng_amh_esp |
|
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. --> |
|
|
|
# furina_seed42_eng_amh_esp |
|
|
|
This model is a fine-tuned version of [yihongLiu/furina](https://huggingface.co/yihongLiu/furina) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0131 |
|
- Spearman Corr: 0.8481 |
|
|
|
## 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: 32 |
|
- eval_batch_size: 128 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 64 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 30 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Spearman Corr | |
|
|:-------------:|:-----:|:-----:|:---------------:|:-------------:| |
|
| No log | 0.59 | 200 | 0.0247 | 0.6933 | |
|
| No log | 1.18 | 400 | 0.0227 | 0.7395 | |
|
| No log | 1.76 | 600 | 0.0217 | 0.7604 | |
|
| 0.0471 | 2.35 | 800 | 0.0192 | 0.7685 | |
|
| 0.0471 | 2.94 | 1000 | 0.0186 | 0.7805 | |
|
| 0.0471 | 3.53 | 1200 | 0.0178 | 0.7884 | |
|
| 0.0196 | 4.12 | 1400 | 0.0179 | 0.7961 | |
|
| 0.0196 | 4.71 | 1600 | 0.0177 | 0.7990 | |
|
| 0.0196 | 5.29 | 1800 | 0.0178 | 0.8035 | |
|
| 0.0196 | 5.88 | 2000 | 0.0164 | 0.8057 | |
|
| 0.0145 | 6.47 | 2200 | 0.0167 | 0.8070 | |
|
| 0.0145 | 7.06 | 2400 | 0.0159 | 0.8110 | |
|
| 0.0145 | 7.65 | 2600 | 0.0165 | 0.8121 | |
|
| 0.0106 | 8.24 | 2800 | 0.0161 | 0.8110 | |
|
| 0.0106 | 8.82 | 3000 | 0.0159 | 0.8148 | |
|
| 0.0106 | 9.41 | 3200 | 0.0155 | 0.8195 | |
|
| 0.008 | 10.0 | 3400 | 0.0151 | 0.8227 | |
|
| 0.008 | 10.59 | 3600 | 0.0149 | 0.8253 | |
|
| 0.008 | 11.18 | 3800 | 0.0154 | 0.8244 | |
|
| 0.008 | 11.76 | 4000 | 0.0147 | 0.8251 | |
|
| 0.0064 | 12.35 | 4200 | 0.0148 | 0.8249 | |
|
| 0.0064 | 12.94 | 4400 | 0.0149 | 0.8287 | |
|
| 0.0064 | 13.53 | 4600 | 0.0147 | 0.8297 | |
|
| 0.0052 | 14.12 | 4800 | 0.0142 | 0.8347 | |
|
| 0.0052 | 14.71 | 5000 | 0.0148 | 0.8314 | |
|
| 0.0052 | 15.29 | 5200 | 0.0141 | 0.8341 | |
|
| 0.0052 | 15.88 | 5400 | 0.0139 | 0.8386 | |
|
| 0.0045 | 16.47 | 5600 | 0.0139 | 0.8350 | |
|
| 0.0045 | 17.06 | 5800 | 0.0137 | 0.8389 | |
|
| 0.0045 | 17.65 | 6000 | 0.0136 | 0.8402 | |
|
| 0.004 | 18.24 | 6200 | 0.0139 | 0.8400 | |
|
| 0.004 | 18.82 | 6400 | 0.0138 | 0.8414 | |
|
| 0.004 | 19.41 | 6600 | 0.0136 | 0.8433 | |
|
| 0.0036 | 20.0 | 6800 | 0.0140 | 0.8420 | |
|
| 0.0036 | 20.59 | 7000 | 0.0136 | 0.8434 | |
|
| 0.0036 | 21.18 | 7200 | 0.0137 | 0.8451 | |
|
| 0.0036 | 21.76 | 7400 | 0.0133 | 0.8445 | |
|
| 0.0032 | 22.35 | 7600 | 0.0135 | 0.8451 | |
|
| 0.0032 | 22.94 | 7800 | 0.0136 | 0.8447 | |
|
| 0.0032 | 23.53 | 8000 | 0.0136 | 0.8449 | |
|
| 0.003 | 24.12 | 8200 | 0.0132 | 0.8463 | |
|
| 0.003 | 24.71 | 8400 | 0.0131 | 0.8472 | |
|
| 0.003 | 25.29 | 8600 | 0.0133 | 0.8477 | |
|
| 0.003 | 25.88 | 8800 | 0.0135 | 0.8472 | |
|
| 0.0028 | 26.47 | 9000 | 0.0134 | 0.8478 | |
|
| 0.0028 | 27.06 | 9200 | 0.0131 | 0.8477 | |
|
| 0.0028 | 27.65 | 9400 | 0.0131 | 0.8478 | |
|
| 0.0026 | 28.24 | 9600 | 0.0130 | 0.8482 | |
|
| 0.0026 | 28.82 | 9800 | 0.0131 | 0.8479 | |
|
| 0.0026 | 29.41 | 10000 | 0.0131 | 0.8481 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.1 |
|
|
