File size: 5,589 Bytes
091c9ec 3346c04 091c9ec | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 | ---
library_name: transformers
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: base_sami_22k_ftpseudo
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. -->
# base_sami_22k_ftpseudo
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 272.0352
- Wer: 0.4500
- Cer: 0.1412
## 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: 0.0005
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.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_ratio: 0.25
- num_epochs: 60.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|
| 3988.287 | 1.0 | 1080 | 303.8312 | 0.5742 | 0.1722 |
| 674.1274 | 2.0 | 2160 | 270.2952 | 0.4506 | 0.1415 |
| 517.9397 | 3.0 | 3240 | 339.8132 | 0.4931 | 0.1630 |
| 478.4345 | 4.0 | 4320 | 292.6837 | 0.4896 | 0.1691 |
| 476.3436 | 5.0 | 5400 | 331.5307 | 0.5095 | 0.1799 |
| 507.3865 | 6.0 | 6480 | 359.2405 | 0.5620 | 0.2006 |
| 511.676 | 7.0 | 7560 | 406.4503 | 0.5881 | 0.2164 |
| 546.8947 | 8.0 | 8640 | 367.2822 | 0.5835 | 0.2145 |
| 578.7039 | 9.0 | 9720 | 441.9370 | 0.6559 | 0.2586 |
| 610.5788 | 10.0 | 10800 | 454.6888 | 0.6856 | 0.2609 |
| 658.402 | 11.0 | 11880 | 472.7102 | 0.7457 | 0.3024 |
| 692.8005 | 12.0 | 12960 | 474.2548 | 0.6993 | 0.2891 |
| 726.6378 | 13.0 | 14040 | 476.5723 | 0.7154 | 0.2942 |
| 775.3879 | 14.0 | 15120 | 469.1171 | 0.7360 | 0.2998 |
| 803.9686 | 15.0 | 16200 | 503.4136 | 0.7707 | 0.3031 |
| 818.0579 | 16.0 | 17280 | 544.9781 | 0.7587 | 0.3132 |
| 808.2149 | 17.0 | 18360 | 493.0830 | 0.7396 | 0.3015 |
| 767.5317 | 18.0 | 19440 | 527.8341 | 0.7296 | 0.3046 |
| 739.8194 | 19.0 | 20520 | 500.3179 | 0.7558 | 0.3085 |
| 716.691 | 20.0 | 21600 | 545.5074 | 0.7235 | 0.2984 |
| 682.661 | 21.0 | 22680 | 516.1239 | 0.7511 | 0.2900 |
| 657.0491 | 22.0 | 23760 | 549.7004 | 0.6968 | 0.2776 |
| 629.1355 | 23.0 | 24840 | 500.3793 | 0.6974 | 0.2808 |
| 607.2812 | 24.0 | 25920 | 528.1496 | 0.6959 | 0.2700 |
| 595.4605 | 25.0 | 27000 | 495.3539 | 0.7015 | 0.2834 |
| 555.9978 | 26.0 | 28080 | 500.2841 | 0.7071 | 0.2782 |
| 544.9409 | 27.0 | 29160 | 476.8067 | 0.7075 | 0.2840 |
| 517.4491 | 28.0 | 30240 | 513.6489 | 0.6824 | 0.2703 |
| 502.3091 | 29.0 | 31320 | 450.8210 | 0.6880 | 0.2624 |
| 477.324 | 30.0 | 32400 | 469.6162 | 0.6562 | 0.2616 |
| 461.2854 | 31.0 | 33480 | 480.2810 | 0.6640 | 0.2484 |
| 452.682 | 32.0 | 34560 | 477.9762 | 0.6638 | 0.2652 |
| 424.353 | 33.0 | 35640 | 444.6511 | 0.6533 | 0.2520 |
| 417.6179 | 34.0 | 36720 | 412.5329 | 0.6504 | 0.2526 |
| 389.705 | 35.0 | 37800 | 485.3770 | 0.6744 | 0.2633 |
| 375.7767 | 36.0 | 38880 | 467.3829 | 0.6474 | 0.2664 |
| 361.8829 | 37.0 | 39960 | 469.9674 | 0.6312 | 0.2517 |
| 352.311 | 38.0 | 41040 | 457.0285 | 0.6495 | 0.2545 |
| 340.1846 | 39.0 | 42120 | 463.1925 | 0.6345 | 0.2462 |
| 323.3272 | 40.0 | 43200 | 421.0725 | 0.6171 | 0.2394 |
| 312.6201 | 41.0 | 44280 | 443.3647 | 0.6201 | 0.2384 |
| 301.6251 | 42.0 | 45360 | 429.3776 | 0.6105 | 0.2350 |
| 284.7902 | 43.0 | 46440 | 466.1553 | 0.6021 | 0.2321 |
| 279.8459 | 44.0 | 47520 | 487.2148 | 0.6162 | 0.2319 |
| 260.5616 | 45.0 | 48600 | 445.4757 | 0.6023 | 0.2306 |
| 254.3347 | 46.0 | 49680 | 439.6965 | 0.6054 | 0.2392 |
| 244.043 | 47.0 | 50760 | 459.5868 | 0.5885 | 0.2317 |
| 227.4755 | 48.0 | 51840 | 492.8037 | 0.6002 | 0.2308 |
| 216.7 | 49.0 | 52920 | 452.6693 | 0.5934 | 0.2283 |
| 211.8976 | 50.0 | 54000 | 482.3886 | 0.5947 | 0.2288 |
| 202.0287 | 51.0 | 55080 | 475.8258 | 0.6053 | 0.2353 |
| 186.2731 | 52.0 | 56160 | 465.3925 | 0.5908 | 0.2311 |
| 187.1888 | 53.0 | 57240 | 459.6522 | 0.5890 | 0.2247 |
| 179.0453 | 54.0 | 58320 | 473.7304 | 0.5789 | 0.2243 |
| 165.2614 | 55.0 | 59400 | 453.9692 | 0.5788 | 0.2238 |
| 160.4416 | 56.0 | 60480 | 474.8051 | 0.5732 | 0.2212 |
| 153.8781 | 57.0 | 61560 | 478.4581 | 0.5729 | 0.2202 |
| 151.1706 | 58.0 | 62640 | 467.0158 | 0.5688 | 0.2196 |
| 147.0876 | 59.0 | 63720 | 474.2252 | 0.5603 | 0.2160 |
| 143.0797 | 60.0 | 64800 | 469.5599 | 0.5641 | 0.2168 |
### Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1
- Datasets 3.2.0
- Tokenizers 0.21.0
|