| GPU available: True (cuda), used: True | |
| TPU available: False, using: 0 TPU cores | |
| IPU available: False, using: 0 IPUs | |
| HPU available: False, using: 0 HPUs | |
| LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0] | |
| | Name | Type | Params | |
| ------------------------------------------------------------------------ | |
| 0 | preprocessor | AudioToMelSpectrogramPreprocessor | 0 | |
| 1 | encoder | ConvASREncoder | 36.3 M | |
| 2 | decoder | ConvASRDecoder | 112 K | |
| 3 | loss | CTCLoss | 0 | |
| 4 | spec_augmentation | SpectrogramAugmentation | 0 | |
| 5 | wer | WER | 0 | |
| ------------------------------------------------------------------------ | |
| 1.8 M Trainable params | |
| 34.7 M Non-trainable params | |
| 36.4 M Total params | |
| 145.798 Total estimated model params size (MB) | |
| Epoch 9, global step 5070: 'val_wer' reached 0.38640 (best 0.38640), saving model to '/content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3864-epoch=9.ckpt' as top 3 | |
| Epoch 19, global step 10140: 'val_wer' reached 0.35093 (best 0.35093), saving model to '/content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt' as top 3 | |
| `Trainer.fit` stopped: `max_epochs=20` reached. | |
| Restoring states from the checkpoint path at /content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt | |
| Restored all states from the checkpoint at /content/experiments/lang-ta/ASR-Model-Language-ta/2024-02-15_22-34-08/checkpoints/ASR-Model-Language-ta--val_wer=0.3509-epoch=19.ckpt | |