<!-- 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. -->
Speech7
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: nan
- Wer: 1
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: 0.01
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 3000
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 4.2042 | 1.44 | 100 | 4.2839 | 1 |
| 4.1119 | 2.88 | 200 | 4.2368 | 1 |
| 4.1696 | 4.32 | 300 | 4.2242 | 1 |
| 4.1838 | 5.76 | 400 | 4.2262 | 1 |
| 4.2368 | 7.19 | 500 | 4.2247 | 1 |
| 4.1376 | 8.63 | 600 | 4.2179 | 1 |
| 4.1417 | 10.07 | 700 | 4.2209 | 1 |
| 4.2254 | 11.51 | 800 | 4.2471 | 1 |
| 4.2302 | 12.95 | 900 | 4.2145 | 1 |
| 4.1778 | 14.39 | 1000 | 4.3393 | 1 |
| 4.1574 | 15.83 | 1100 | 4.2917 | 1 |
| 4.2026 | 17.27 | 1200 | 4.2731 | 1 |
| 4.141 | 18.71 | 1300 | 4.2302 | 1 |
| 4.2525 | 20.14 | 1400 | 4.2104 | 1 |
| 4.2325 | 21.58 | 1500 | 4.2543 | 1 |
| 4.1789 | 23.02 | 1600 | 4.4020 | 1 |
| 4.1456 | 24.46 | 1700 | 4.2143 | 1 |
| 4.1754 | 25.9 | 1800 | 4.2123 | 1 |
| 12.3485 | 27.34 | 1900 | 50.3232 | 1 |
| 4.2031 | 28.78 | 2000 | 4.2259 | 1 |
| 4.1497 | 30.22 | 2100 | 4.3216 | 1 |
| 4.2171 | 31.65 | 2200 | 4.2108 | 1 |
| 4.1981 | 33.09 | 2300 | 4.3025 | 1 |
| 4.2091 | 34.53 | 2400 | 4.2173 | 1 |
| 4.2005 | 35.97 | 2500 | 4.2747 | 1 |
| 4.2386 | 37.41 | 2600 | 4.2027 | 1 |
| 4.2343 | 38.85 | 2700 | 4.2137 | 1 |
| 4.0967 | 40.29 | 2800 | 4.2804 | 1 |
| 4.1737 | 41.73 | 2900 | 4.2072 | 1 |
| 4.171 | 43.17 | 3000 | 4.2186 | 1 |
| 4.2117 | 44.6 | 3100 | 4.2161 | 1 |
| 4.1021 | 46.04 | 3200 | 4.2389 | 1 |
| 4.2572 | 47.48 | 3300 | 4.2126 | 1 |
| 3.4461 | 48.92 | 3400 | 4.2700 | 1 |
| 0.7289 | 50.36 | 3500 | 4.2700 | 1 |
| 0.4496 | 51.8 | 3600 | 4.2700 | 1 |
| 0.1189 | 53.24 | 3700 | 4.2700 | 1 |
| 8.233 | 54.68 | 3800 | 4.2700 | 1 |
| 3.8072 | 56.12 | 3900 | 4.2700 | 1 |
| 0.0 | 57.55 | 4000 | nan | 1 |
| 0.0 | 58.99 | 4100 | nan | 1 |
| 0.0 | 60.43 | 4200 | nan | 1 |
| 0.0 | 61.87 | 4300 | nan | 1 |
| 0.0 | 63.31 | 4400 | nan | 1 |
| 0.0 | 64.75 | 4500 | nan | 1 |
| 0.0 | 66.19 | 4600 | nan | 1 |
| 0.0 | 67.63 | 4700 | nan | 1 |
| 0.0 | 69.06 | 4800 | nan | 1 |
| 0.0 | 70.5 | 4900 | nan | 1 |
| 0.0 | 71.94 | 5000 | nan | 1 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3
