File size: 3,779 Bytes
148a9b2 d828b84 148a9b2 d828b84 |
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 |
---
library_name: transformers
license: mit
base_model: facebook/w2v-bert-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: w2v-bert-2.0-Fleurs_AMMI_AFRIVOICE_LRSC-ln-5hrs-v1
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. -->
# w2v-bert-2.0-Fleurs_AMMI_AFRIVOICE_LRSC-ln-5hrs-v1
This model is a fine-tuned version of [facebook/w2v-bert-2.0](https://huggingface.co/facebook/w2v-bert-2.0) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6643
- Wer: 0.2469
- Cer: 0.0788
## 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-05
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 100
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
| 1.8874 | 0.9949 | 98 | 0.6403 | 0.5429 | 0.1657 |
| 0.4899 | 2.0 | 197 | 0.4921 | 0.3300 | 0.1001 |
| 0.3892 | 2.9949 | 295 | 0.4608 | 0.3314 | 0.1019 |
| 0.3259 | 4.0 | 394 | 0.4729 | 0.3080 | 0.0942 |
| 0.2863 | 4.9949 | 492 | 0.4495 | 0.3156 | 0.0951 |
| 0.2333 | 6.0 | 591 | 0.4269 | 0.2624 | 0.0808 |
| 0.2059 | 6.9949 | 689 | 0.4365 | 0.2609 | 0.0839 |
| 0.1722 | 8.0 | 788 | 0.4346 | 0.2552 | 0.0825 |
| 0.1551 | 8.9949 | 886 | 0.4134 | 0.2468 | 0.0766 |
| 0.1318 | 10.0 | 985 | 0.4794 | 0.2631 | 0.0811 |
| 0.1189 | 10.9949 | 1083 | 0.5191 | 0.2530 | 0.0796 |
| 0.1004 | 12.0 | 1182 | 0.5311 | 0.2689 | 0.0794 |
| 0.0959 | 12.9949 | 1280 | 0.5502 | 0.2535 | 0.0778 |
| 0.0831 | 14.0 | 1379 | 0.5060 | 0.2476 | 0.0757 |
| 0.0679 | 14.9949 | 1477 | 0.5023 | 0.2517 | 0.0830 |
| 0.0617 | 16.0 | 1576 | 0.5279 | 0.2403 | 0.0757 |
| 0.0562 | 16.9949 | 1674 | 0.6012 | 0.2411 | 0.0761 |
| 0.0496 | 18.0 | 1773 | 0.6263 | 0.2423 | 0.0755 |
| 0.0442 | 18.9949 | 1871 | 0.5991 | 0.2581 | 0.0794 |
| 0.0401 | 20.0 | 1970 | 0.6323 | 0.2412 | 0.0762 |
| 0.0329 | 20.9949 | 2068 | 0.6417 | 0.2326 | 0.0735 |
| 0.0266 | 22.0 | 2167 | 0.6279 | 0.2381 | 0.0756 |
| 0.0255 | 22.9949 | 2265 | 0.5834 | 0.2470 | 0.0772 |
| 0.0214 | 24.0 | 2364 | 0.6781 | 0.2364 | 0.0735 |
| 0.0217 | 24.9949 | 2462 | 0.6253 | 0.2398 | 0.0752 |
| 0.0163 | 26.0 | 2561 | 0.6940 | 0.2427 | 0.0813 |
| 0.0363 | 26.9949 | 2659 | 0.6632 | 0.2363 | 0.0756 |
| 0.0182 | 28.0 | 2758 | 0.6094 | 0.2363 | 0.0766 |
| 0.014 | 28.9949 | 2856 | 0.6928 | 0.2438 | 0.0770 |
| 0.0157 | 30.0 | 2955 | 0.6863 | 0.2422 | 0.0768 |
| 0.0121 | 30.9949 | 3053 | 0.6643 | 0.2469 | 0.0788 |
### Framework versions
- Transformers 4.46.3
- Pytorch 2.1.0+cu118
- Datasets 3.1.0
- Tokenizers 0.20.3
|