<!-- 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. -->
libri-alpha-0.75-Temp-1-attention-3-layers-distil-with-6-layers-att-take-2
This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:
- Loss: 42.0471
- Wer: 0.2540
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: 0.002
- train_batch_size: 64
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
860.105 | 3.6 | 400 | 37.1616 | 0.4129 |
640.8073 | 7.21 | 800 | 38.0915 | 0.3892 |
578.6465 | 10.81 | 1200 | 41.3839 | 0.3648 |
478.3375 | 14.41 | 1600 | 43.8448 | 0.3231 |
400.7667 | 18.02 | 2000 | 35.5516 | 0.3103 |
348.2905 | 21.62 | 2400 | 41.2895 | 0.2954 |
297.0109 | 25.22 | 2800 | 42.0566 | 0.2889 |
262.856 | 28.83 | 3200 | 38.5730 | 0.2779 |
227.4767 | 32.43 | 3600 | 42.4243 | 0.2657 |
200.5691 | 36.04 | 4000 | 42.0213 | 0.2675 |
181.1116 | 39.64 | 4400 | 42.0471 | 0.2540 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1
- Datasets 2.8.0
- Tokenizers 0.13.2