ft_kor_test_2 / README.md
molto's picture
Upload 10 files
4d3eeb6
---
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: ft_kor_test_1
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. -->
# ft_kor_test_1
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0205
- Cer: 0.0037
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 20
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 5.8814 | 0.1 | 500 | 3.3282 | 1.0 |
| 2.922 | 0.2 | 1000 | 1.5452 | 0.4197 |
| 1.0454 | 0.3 | 1500 | 0.5135 | 0.1411 |
| 0.6881 | 0.4 | 2000 | 0.3212 | 0.0964 |
| 0.5735 | 0.51 | 2500 | 0.2526 | 0.0805 |
| 0.5236 | 0.61 | 3000 | 0.2255 | 0.0691 |
| 0.4813 | 0.71 | 3500 | 0.2167 | 0.0662 |
| 0.4442 | 0.81 | 4000 | 0.1816 | 0.0575 |
| 0.4244 | 0.91 | 4500 | 0.1717 | 0.0542 |
| 0.4026 | 1.01 | 5000 | 0.1573 | 0.0525 |
| 0.3691 | 1.11 | 5500 | 0.1423 | 0.0455 |
| 0.3606 | 1.21 | 6000 | 0.1340 | 0.0429 |
| 0.3451 | 1.32 | 6500 | 0.1305 | 0.0417 |
| 0.3421 | 1.42 | 7000 | 0.1231 | 0.0389 |
| 0.3319 | 1.52 | 7500 | 0.1167 | 0.0379 |
| 0.3265 | 1.62 | 8000 | 0.1158 | 0.0373 |
| 0.3114 | 1.72 | 8500 | 0.1105 | 0.0343 |
| 0.299 | 1.82 | 9000 | 0.1015 | 0.0322 |
| 0.3023 | 1.92 | 9500 | 0.0968 | 0.0309 |
| 0.2952 | 2.02 | 10000 | 0.0926 | 0.0301 |
| 0.2719 | 2.13 | 10500 | 0.0937 | 0.0297 |
| 0.2726 | 2.23 | 11000 | 0.0902 | 0.0285 |
| 0.2615 | 2.33 | 11500 | 0.0876 | 0.0284 |
| 0.2611 | 2.43 | 12000 | 0.0839 | 0.0264 |
| 0.2505 | 2.53 | 12500 | 0.0848 | 0.0269 |
| 0.2494 | 2.63 | 13000 | 0.0788 | 0.0246 |
| 0.2442 | 2.73 | 13500 | 0.0798 | 0.0249 |
| 0.2448 | 2.83 | 14000 | 0.0769 | 0.0243 |
| 0.2365 | 2.93 | 14500 | 0.0755 | 0.0240 |
| 0.234 | 3.04 | 15000 | 0.0750 | 0.0221 |
| 0.2282 | 3.14 | 15500 | 0.0717 | 0.0219 |
| 0.2173 | 3.24 | 16000 | 0.0673 | 0.0210 |
| 0.2124 | 3.34 | 16500 | 0.0680 | 0.0211 |
| 0.2161 | 3.44 | 17000 | 0.0656 | 0.0206 |
| 0.2089 | 3.54 | 17500 | 0.0664 | 0.0204 |
| 0.213 | 3.64 | 18000 | 0.0623 | 0.0190 |
| 0.2094 | 3.74 | 18500 | 0.0635 | 0.0184 |
| 0.1998 | 3.85 | 19000 | 0.0635 | 0.0184 |
| 0.2024 | 3.95 | 19500 | 0.0620 | 0.0183 |
| 0.1935 | 4.05 | 20000 | 0.0572 | 0.0174 |
| 0.1873 | 4.15 | 20500 | 0.0607 | 0.0180 |
| 0.1789 | 4.25 | 21000 | 0.0583 | 0.0163 |
| 0.1842 | 4.35 | 21500 | 0.0663 | 0.0187 |
| 0.1773 | 4.45 | 22000 | 0.0532 | 0.0156 |
| 0.1877 | 4.55 | 22500 | 0.0583 | 0.0163 |
| 0.1844 | 4.65 | 23000 | 0.0543 | 0.0155 |
| 0.1711 | 4.76 | 23500 | 0.0522 | 0.0150 |
| 0.1703 | 4.86 | 24000 | 0.0503 | 0.0148 |
| 0.1712 | 4.96 | 24500 | 0.0524 | 0.0153 |
| 0.1642 | 5.06 | 25000 | 0.0505 | 0.0148 |
| 0.1622 | 5.16 | 25500 | 0.0476 | 0.0138 |
| 0.1544 | 5.26 | 26000 | 0.0500 | 0.0143 |
| 0.157 | 5.36 | 26500 | 0.0505 | 0.0139 |
| 0.1632 | 5.46 | 27000 | 0.0487 | 0.0138 |
| 0.1516 | 5.57 | 27500 | 0.0440 | 0.0126 |
| 0.1532 | 5.67 | 28000 | 0.0467 | 0.0127 |
| 0.1523 | 5.77 | 28500 | 0.0486 | 0.0135 |
| 0.1471 | 5.87 | 29000 | 0.0489 | 0.0129 |
| 0.1498 | 5.97 | 29500 | 0.0458 | 0.0123 |
| 0.1511 | 6.07 | 30000 | 0.0424 | 0.0123 |
| 0.1422 | 6.17 | 30500 | 0.0444 | 0.0118 |
| 0.1394 | 6.27 | 31000 | 0.0519 | 0.0148 |
| 0.1483 | 6.38 | 31500 | 0.0436 | 0.0120 |
| 0.1394 | 6.48 | 32000 | 0.0465 | 0.0126 |
| 0.1363 | 6.58 | 32500 | 0.0397 | 0.0110 |
| 0.1372 | 6.68 | 33000 | 0.0418 | 0.0110 |
| 0.1353 | 6.78 | 33500 | 0.0412 | 0.0110 |
| 0.1356 | 6.88 | 34000 | 0.0397 | 0.0109 |
| 0.1321 | 6.98 | 34500 | 0.0380 | 0.0100 |
| 0.1323 | 7.08 | 35000 | 0.0373 | 0.0101 |
| 0.1251 | 7.18 | 35500 | 0.0365 | 0.0099 |
| 0.1238 | 7.29 | 36000 | 0.0381 | 0.0100 |
| 0.1247 | 7.39 | 36500 | 0.0394 | 0.0103 |
| 0.128 | 7.49 | 37000 | 0.0389 | 0.0102 |
| 0.1245 | 7.59 | 37500 | 0.0382 | 0.0096 |
| 0.1224 | 7.69 | 38000 | 0.0358 | 0.0090 |
| 0.12 | 7.79 | 38500 | 0.0495 | 0.0113 |
| 0.1217 | 7.89 | 39000 | 0.0476 | 0.0108 |
| 0.1198 | 7.99 | 39500 | 0.0512 | 0.0130 |
| 0.1125 | 8.1 | 40000 | 0.0431 | 0.0109 |
| 0.1107 | 8.2 | 40500 | 0.0456 | 0.0111 |
| 0.1101 | 8.3 | 41000 | 0.0889 | 0.0176 |
| 0.1136 | 8.4 | 41500 | 0.0449 | 0.0103 |
| 0.1131 | 8.5 | 42000 | 0.0320 | 0.0082 |
| 0.1145 | 8.6 | 42500 | 0.0311 | 0.0083 |
| 0.1039 | 8.7 | 43000 | 0.0317 | 0.0086 |
| 0.1115 | 8.8 | 43500 | 0.0384 | 0.0086 |
| 0.1098 | 8.91 | 44000 | 0.0328 | 0.0085 |
| 0.1114 | 9.01 | 44500 | 0.0331 | 0.0083 |
| 0.0982 | 9.11 | 45000 | 0.0305 | 0.0079 |
| 0.1041 | 9.21 | 45500 | 0.0359 | 0.0084 |
| 0.1033 | 9.31 | 46000 | 0.0298 | 0.0076 |
| 0.1024 | 9.41 | 46500 | 0.0310 | 0.0076 |
| 0.0981 | 9.51 | 47000 | 0.0309 | 0.0075 |
| 0.1033 | 9.61 | 47500 | 0.0311 | 0.0076 |
| 0.0995 | 9.71 | 48000 | 0.0309 | 0.0079 |
| 0.1012 | 9.82 | 48500 | 0.0283 | 0.0071 |
| 0.1039 | 9.92 | 49000 | 0.0276 | 0.0070 |
| 0.0957 | 10.02 | 49500 | 0.0298 | 0.0071 |
| 0.0933 | 10.12 | 50000 | 0.0297 | 0.0073 |
| 0.0961 | 10.22 | 50500 | 0.0278 | 0.0069 |
| 0.0939 | 10.32 | 51000 | 0.0278 | 0.0071 |
| 0.0928 | 10.42 | 51500 | 0.0279 | 0.0071 |
| 0.0915 | 10.52 | 52000 | 0.0271 | 0.0065 |
| 0.0907 | 10.63 | 52500 | 0.0385 | 0.0099 |
| 0.0951 | 10.73 | 53000 | 0.0556 | 0.0127 |
| 0.0949 | 10.83 | 53500 | 0.0767 | 0.0189 |
| 0.0923 | 10.93 | 54000 | 0.0317 | 0.0074 |
| 0.0852 | 11.03 | 54500 | 0.0474 | 0.0114 |
| 0.0863 | 11.13 | 55000 | 0.0304 | 0.0067 |
| 0.0858 | 11.23 | 55500 | 0.0289 | 0.0063 |
| 0.0852 | 11.33 | 56000 | 0.0399 | 0.0117 |
| 0.0821 | 11.43 | 56500 | 0.0498 | 0.0111 |
| 0.0822 | 11.54 | 57000 | 0.0452 | 0.0113 |
| 0.0838 | 11.64 | 57500 | 0.0397 | 0.0079 |
| 0.0771 | 11.74 | 58000 | 0.0568 | 0.0120 |
| 0.0813 | 11.84 | 58500 | 0.0465 | 0.0087 |
| 0.078 | 11.94 | 59000 | 0.0524 | 0.0092 |
| 0.0809 | 12.04 | 59500 | 0.0545 | 0.0100 |
| 0.0755 | 12.14 | 60000 | 0.0273 | 0.0057 |
| 0.077 | 12.24 | 60500 | 0.0277 | 0.0060 |
| 0.0772 | 12.35 | 61000 | 0.0265 | 0.0057 |
| 0.0728 | 12.45 | 61500 | 0.0311 | 0.0057 |
| 0.0766 | 12.55 | 62000 | 0.0301 | 0.0066 |
| 0.0805 | 12.65 | 62500 | 0.0323 | 0.0067 |
| 0.0732 | 12.75 | 63000 | 0.0298 | 0.0061 |
| 0.0735 | 12.85 | 63500 | 0.0229 | 0.0052 |
| 0.0738 | 12.95 | 64000 | 0.0242 | 0.0054 |
| 0.0709 | 13.05 | 64500 | 0.0237 | 0.0053 |
| 0.0702 | 13.16 | 65000 | 0.0236 | 0.0050 |
| 0.0702 | 13.26 | 65500 | 0.0255 | 0.0053 |
| 0.0676 | 13.36 | 66000 | 0.0236 | 0.0052 |
| 0.0704 | 13.46 | 66500 | 0.0224 | 0.0053 |
| 0.07 | 13.56 | 67000 | 0.0238 | 0.0054 |
| 0.0671 | 13.66 | 67500 | 0.0232 | 0.0054 |
| 0.0709 | 13.76 | 68000 | 0.0228 | 0.0051 |
| 0.0636 | 13.86 | 68500 | 0.0227 | 0.0052 |
| 0.0661 | 13.96 | 69000 | 0.0223 | 0.0049 |
| 0.0645 | 14.07 | 69500 | 0.0222 | 0.0048 |
| 0.0639 | 14.17 | 70000 | 0.0243 | 0.0051 |
| 0.0608 | 14.27 | 70500 | 0.0250 | 0.0050 |
| 0.0631 | 14.37 | 71000 | 0.0234 | 0.0048 |
| 0.0656 | 14.47 | 71500 | 0.0228 | 0.0048 |
| 0.0616 | 14.57 | 72000 | 0.0239 | 0.0050 |
| 0.0631 | 14.67 | 72500 | 0.0237 | 0.0049 |
| 0.0662 | 14.77 | 73000 | 0.0234 | 0.0047 |
| 0.0622 | 14.88 | 73500 | 0.0289 | 0.0056 |
| 0.064 | 14.98 | 74000 | 0.0242 | 0.0048 |
| 0.0546 | 15.08 | 74500 | 0.0234 | 0.0049 |
| 0.0573 | 15.18 | 75000 | 0.0254 | 0.0054 |
| 0.0571 | 15.28 | 75500 | 0.0288 | 0.0058 |
| 0.0576 | 15.38 | 76000 | 0.0244 | 0.0053 |
| 0.0562 | 15.48 | 76500 | 0.0299 | 0.0061 |
| 0.0595 | 15.58 | 77000 | 0.0221 | 0.0046 |
| 0.0601 | 15.69 | 77500 | 0.0224 | 0.0046 |
| 0.0575 | 15.79 | 78000 | 0.0216 | 0.0045 |
| 0.059 | 15.89 | 78500 | 0.0222 | 0.0045 |
| 0.0562 | 15.99 | 79000 | 0.0224 | 0.0047 |
| 0.0551 | 16.09 | 79500 | 0.0216 | 0.0044 |
| 0.0539 | 16.19 | 80000 | 0.0223 | 0.0047 |
| 0.0547 | 16.29 | 80500 | 0.0212 | 0.0045 |
| 0.0527 | 16.39 | 81000 | 0.0264 | 0.0049 |
| 0.0527 | 16.49 | 81500 | 0.0247 | 0.0050 |
| 0.0526 | 16.6 | 82000 | 0.0236 | 0.0047 |
| 0.0507 | 16.7 | 82500 | 0.0213 | 0.0042 |
| 0.0522 | 16.8 | 83000 | 0.0221 | 0.0042 |
| 0.0522 | 16.9 | 83500 | 0.0220 | 0.0042 |
| 0.0496 | 17.0 | 84000 | 0.0217 | 0.0043 |
| 0.0495 | 17.1 | 84500 | 0.0214 | 0.0042 |
| 0.0493 | 17.2 | 85000 | 0.0217 | 0.0042 |
| 0.0488 | 17.3 | 85500 | 0.0207 | 0.0040 |
| 0.0492 | 17.41 | 86000 | 0.0210 | 0.0042 |
| 0.0496 | 17.51 | 86500 | 0.0204 | 0.0042 |
| 0.0487 | 17.61 | 87000 | 0.0216 | 0.0041 |
| 0.0466 | 17.71 | 87500 | 0.0199 | 0.0040 |
| 0.0465 | 17.81 | 88000 | 0.0199 | 0.0040 |
| 0.0491 | 17.91 | 88500 | 0.0198 | 0.0040 |
| 0.0469 | 18.01 | 89000 | 0.0204 | 0.0041 |
| 0.0447 | 18.11 | 89500 | 0.0205 | 0.0040 |
| 0.0487 | 18.21 | 90000 | 0.0215 | 0.0040 |
| 0.0455 | 18.32 | 90500 | 0.0207 | 0.0039 |
| 0.047 | 18.42 | 91000 | 0.0207 | 0.0040 |
| 0.0458 | 18.52 | 91500 | 0.0206 | 0.0040 |
| 0.0462 | 18.62 | 92000 | 0.0202 | 0.0039 |
| 0.0473 | 18.72 | 92500 | 0.0212 | 0.0039 |
| 0.043 | 18.82 | 93000 | 0.0208 | 0.0039 |
| 0.0435 | 18.92 | 93500 | 0.0204 | 0.0039 |
| 0.0448 | 19.02 | 94000 | 0.0208 | 0.0038 |
| 0.0435 | 19.13 | 94500 | 0.0205 | 0.0038 |
| 0.0433 | 19.23 | 95000 | 0.0203 | 0.0038 |
| 0.0425 | 19.33 | 95500 | 0.0204 | 0.0037 |
| 0.045 | 19.43 | 96000 | 0.0205 | 0.0038 |
| 0.043 | 19.53 | 96500 | 0.0205 | 0.0037 |
| 0.0435 | 19.63 | 97000 | 0.0206 | 0.0038 |
| 0.0424 | 19.73 | 97500 | 0.0207 | 0.0037 |
| 0.0441 | 19.83 | 98000 | 0.0206 | 0.0037 |
| 0.0452 | 19.94 | 98500 | 0.0205 | 0.0037 |
### Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3