<!-- 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. -->
lbri500
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3739
- Wer: 0.2225
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: 0.0003
- train_batch_size: 16
- eval_batch_size: 16
- 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: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0.02 | 200 | 3.4253 | 1.0000 |
| No log | 0.04 | 400 | 1.4537 | 0.6604 |
| 2.74 | 0.06 | 600 | 1.1070 | 0.5025 |
| 2.74 | 0.09 | 800 | 0.8496 | 0.4207 |
| 0.3705 | 0.11 | 1000 | 0.8767 | 0.4470 |
| 0.3705 | 0.13 | 1200 | 0.7220 | 0.3894 |
| 0.3705 | 0.15 | 1400 | 0.7478 | 0.4116 |
| 0.3273 | 0.17 | 1600 | 0.6434 | 0.3624 |
| 0.3273 | 0.19 | 1800 | 2.7013 | 0.5531 |
| 0.2925 | 0.22 | 2000 | 1.1436 | 0.3535 |
| 0.2925 | 0.24 | 2200 | 1.1855 | 0.3640 |
| 0.2925 | 0.26 | 2400 | 0.7101 | 0.3149 |
| 0.2651 | 0.28 | 2600 | 0.5961 | 0.3177 |
| 0.2651 | 0.3 | 2800 | 0.8519 | 0.3193 |
| 0.2491 | 0.32 | 3000 | 0.5905 | 0.2986 |
| 0.2491 | 0.34 | 3200 | 0.5106 | 0.2890 |
| 0.2491 | 0.37 | 3400 | 0.5417 | 0.2984 |
| 0.2403 | 0.39 | 3600 | 0.6281 | 0.3248 |
| 0.2403 | 0.41 | 3800 | 0.5387 | 0.2924 |
| 0.2245 | 0.43 | 4000 | 0.7894 | 0.3626 |
| 0.2245 | 0.45 | 4200 | 0.5342 | 0.2928 |
| 0.2245 | 0.47 | 4400 | 0.3809 | 0.2605 |
| 0.2111 | 0.49 | 4600 | 0.3956 | 0.2551 |
| 0.2111 | 0.52 | 4800 | 0.6096 | 0.2653 |
| 0.1974 | 0.54 | 5000 | 0.4152 | 0.2546 |
| 0.1974 | 0.56 | 5200 | 0.4424 | 0.2573 |
| 0.1974 | 0.58 | 5400 | 0.5658 | 0.2741 |
| 0.193 | 0.6 | 5600 | 0.4838 | 0.2555 |
| 0.193 | 0.62 | 5800 | 0.4207 | 0.2506 |
| 0.1807 | 0.65 | 6000 | 0.4004 | 0.2472 |
| 0.1807 | 0.67 | 6200 | 0.3907 | 0.2524 |
| 0.1807 | 0.69 | 6400 | 0.4393 | 0.2738 |
| 0.1708 | 0.71 | 6600 | 0.5922 | 0.2680 |
| 0.1708 | 0.73 | 6800 | 0.5787 | 0.2751 |
| 0.1644 | 0.75 | 7000 | 0.4308 | 0.2452 |
| 0.1644 | 0.77 | 7200 | 0.3662 | 0.2294 |
| 0.1644 | 0.8 | 7400 | 0.6298 | 0.2600 |
| 0.1532 | 0.82 | 7600 | 0.4775 | 0.2383 |
| 0.1532 | 0.84 | 7800 | 0.3913 | 0.2196 |
| 0.148 | 0.86 | 8000 | 0.3752 | 0.2098 |
| 0.148 | 0.88 | 8200 | 0.3315 | 0.2115 |
| 0.148 | 0.9 | 8400 | 0.3518 | 0.2157 |
| 0.1401 | 0.93 | 8600 | 0.3517 | 0.2159 |
| 0.1401 | 0.95 | 8800 | 0.3515 | 0.2180 |
| 0.1351 | 0.97 | 9000 | 0.3792 | 0.2203 |
| 0.1351 | 0.99 | 9200 | 0.3739 | 0.2225 |
Framework versions
- Transformers 4.27.2
- Pytorch 1.10.0
- Datasets 2.10.1
- Tokenizers 0.13.2
