<!-- 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. -->
wavlm-2phone-jdrt-fw07-clean-param
This model is a fine-tuned version of jonatasgrosman/exp_w2v2t_ja_wavlm_s729 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1329
- Wer: 0.6848
- Cer: 0.5329
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: 4
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
---|---|---|---|---|---|
0.9132 | 1.0 | 939 | 1.7012 | 0.8806 | 0.6618 |
0.4641 | 2.0 | 1879 | 1.4535 | 0.8315 | 0.6077 |
0.3852 | 3.0 | 2819 | 1.2937 | 0.7774 | 0.5898 |
0.3486 | 4.0 | 3759 | 1.1727 | 0.7119 | 0.5512 |
0.3343 | 5.0 | 4695 | 1.1329 | 0.6848 | 0.5329 |
Framework versions
- Transformers 4.34.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3