jeju-ko-nmt-v6-word-v1
This model is a fine-tuned version of leadawon/jeju-ko-nmt-v6 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6390
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: 5e-06
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
0.8194 | 0.03 | 500 | 0.7499 |
0.7814 | 0.06 | 1000 | 0.7211 |
0.7559 | 0.09 | 1500 | 0.7036 |
0.7441 | 0.12 | 2000 | 0.7017 |
0.741 | 0.15 | 2500 | 0.6865 |
0.7329 | 0.18 | 3000 | 0.6825 |
0.7341 | 0.21 | 3500 | 0.6754 |
0.7358 | 0.24 | 4000 | 0.6671 |
0.723 | 0.28 | 4500 | 0.6641 |
0.7229 | 0.31 | 5000 | 0.6639 |
0.726 | 0.34 | 5500 | 0.6642 |
0.7219 | 0.37 | 6000 | 0.6577 |
0.7213 | 0.4 | 6500 | 0.6552 |
0.7192 | 0.43 | 7000 | 0.6627 |
0.7154 | 0.46 | 7500 | 0.6582 |
0.7108 | 0.49 | 8000 | 0.6570 |
0.7154 | 0.52 | 8500 | 0.6530 |
0.7178 | 0.55 | 9000 | 0.6601 |
0.7046 | 0.58 | 9500 | 0.6480 |
0.7127 | 0.61 | 10000 | 0.6507 |
0.6985 | 0.64 | 10500 | 0.6493 |
0.7093 | 0.67 | 11000 | 0.6500 |
0.7055 | 0.7 | 11500 | 0.6469 |
0.7061 | 0.73 | 12000 | 0.6534 |
0.7018 | 0.76 | 12500 | 0.6458 |
0.7027 | 0.8 | 13000 | 0.6424 |
0.7068 | 0.83 | 13500 | 0.6387 |
0.7044 | 0.86 | 14000 | 0.6441 |
0.7028 | 0.89 | 14500 | 0.6413 |
0.6992 | 0.92 | 15000 | 0.6359 |
0.7007 | 0.95 | 15500 | 0.6479 |
0.6887 | 0.98 | 16000 | 0.6416 |
0.7053 | 1.01 | 16500 | 0.6381 |
0.6795 | 1.04 | 17000 | 0.6391 |
0.6812 | 1.07 | 17500 | 0.6392 |
0.6826 | 1.1 | 18000 | 0.6405 |
0.6775 | 1.13 | 18500 | 0.6391 |
0.6798 | 1.16 | 19000 | 0.6378 |
0.6895 | 1.19 | 19500 | 0.6359 |
0.687 | 1.22 | 20000 | 0.6364 |
0.6887 | 1.25 | 20500 | 0.6357 |
0.6758 | 1.28 | 21000 | 0.6356 |
0.6732 | 1.32 | 21500 | 0.6368 |
0.6832 | 1.35 | 22000 | 0.6381 |
0.6805 | 1.38 | 22500 | 0.6347 |
0.6821 | 1.41 | 23000 | 0.6373 |
0.6828 | 1.44 | 23500 | 0.6401 |
0.678 | 1.47 | 24000 | 0.6402 |
0.6892 | 1.5 | 24500 | 0.6358 |
0.6855 | 1.53 | 25000 | 0.6339 |
0.6748 | 1.56 | 25500 | 0.6363 |
0.6734 | 1.59 | 26000 | 0.6361 |
0.6748 | 1.62 | 26500 | 0.6348 |
0.6812 | 1.65 | 27000 | 0.6355 |
0.6844 | 1.68 | 27500 | 0.6380 |
0.6794 | 1.71 | 28000 | 0.6393 |
0.6834 | 1.74 | 28500 | 0.6390 |
0.6843 | 1.77 | 29000 | 0.6411 |
0.6732 | 1.8 | 29500 | 0.6414 |
0.6758 | 1.84 | 30000 | 0.6395 |
0.6782 | 1.87 | 30500 | 0.6407 |
0.6787 | 1.9 | 31000 | 0.6402 |
0.6752 | 1.93 | 31500 | 0.6393 |
0.6769 | 1.96 | 32000 | 0.6392 |
0.6768 | 1.99 | 32500 | 0.6390 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu116
- Tokenizers 0.13.2
- Downloads last month
- 3
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.