<!-- 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. -->
my_awesome_asr_mind_model_g
This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5919
- Wer: 0.8075
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- training_steps: 40000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 4.5631 | 200.0 | 1000 | 11.4025 | 0.9608 |
| 3.0542 | 400.0 | 2000 | 12.4850 | 0.9657 |
| 2.304 | 600.0 | 3000 | 13.9588 | 0.9755 |
| 1.6633 | 800.0 | 4000 | 14.9246 | 1.0049 |
| 1.1861 | 1000.0 | 5000 | 15.8909 | 1.0392 |
| 0.8167 | 1200.0 | 6000 | 16.8013 | 1.1127 |
| 0.5285 | 1400.0 | 7000 | 17.8491 | 1.1667 |
| 0.3458 | 1600.0 | 8000 | 18.9104 | 1.2304 |
| 0.2389 | 1800.0 | 9000 | 19.8274 | 1.2549 |
| 0.1797 | 2000.0 | 10000 | 20.9407 | 1.3137 |
| 0.1399 | 2200.0 | 11000 | 21.9922 | 1.2990 |
| 0.1247 | 2400.0 | 12000 | 22.2851 | 1.2892 |
| 0.1057 | 2600.0 | 13000 | 23.1505 | 1.2843 |
| 0.0941 | 2800.0 | 14000 | 23.6575 | 1.3088 |
| 0.0806 | 3000.0 | 15000 | 24.0379 | 1.3137 |
| 0.0786 | 3200.0 | 16000 | 24.4104 | 1.3186 |
| 0.0756 | 3400.0 | 17000 | 24.6755 | 1.3039 |
| 0.0726 | 3600.0 | 18000 | 24.9217 | 1.3284 |
| 0.0703 | 3800.0 | 19000 | 25.0183 | 1.3676 |
| 0.0691 | 4000.0 | 20000 | 25.0077 | 1.3578 |
| 0.1578 | 4200.0 | 21000 | 1.6879 | 0.9057 |
| 0.1069 | 4400.0 | 22000 | 1.7429 | 0.9198 |
| 0.0904 | 4600.0 | 23000 | 1.7689 | 0.9104 |
| 0.0789 | 4800.0 | 24000 | 1.7784 | 0.8821 |
| 0.0722 | 5000.0 | 25000 | 1.8151 | 0.9104 |
| 0.0683 | 5200.0 | 26000 | 1.9367 | 0.9528 |
| 0.0605 | 5400.0 | 27000 | 1.8784 | 0.9198 |
| 0.0827 | 5600.0 | 28000 | 0.5633 | 0.7746 |
| 0.0684 | 5800.0 | 29000 | 0.5884 | 0.7981 |
| 0.0625 | 6000.0 | 30000 | 0.5694 | 0.7981 |
| 0.0589 | 6200.0 | 31000 | 0.5863 | 0.7934 |
| 0.0552 | 6400.0 | 32000 | 0.5806 | 0.7840 |
| 0.0524 | 6600.0 | 33000 | 0.5765 | 0.7981 |
| 0.0513 | 6800.0 | 34000 | 0.5865 | 0.7840 |
| 0.0483 | 7000.0 | 35000 | 0.5980 | 0.7934 |
| 0.0471 | 7200.0 | 36000 | 0.5889 | 0.7981 |
| 0.0461 | 7400.0 | 37000 | 0.5821 | 0.8028 |
| 0.0444 | 7600.0 | 38000 | 0.5915 | 0.7981 |
| 0.0455 | 7800.0 | 39000 | 0.5960 | 0.8028 |
| 0.0451 | 8000.0 | 40000 | 0.5919 | 0.8075 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1
- Datasets 2.14.5
- Tokenizers 0.13.3
