license: apache-2.0 | |
tags: | |
- generated_from_trainer | |
model-index: | |
- name: wav2vec2-korean-v3 | |
results: [] | |
<!-- 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. --> | |
# wav2vec2-korean-v3 | |
This model is a fine-tuned version of [Hyuk/wav2vec2-korean-v3](https://huggingface.co/Hyuk/wav2vec2-korean-v3) on the None dataset. | |
It achieves the following results on the evaluation set: | |
- eval_loss: 0.9860 | |
- eval_wer: 0.3797 | |
- eval_runtime: 56.5559 | |
- eval_samples_per_second: 10.167 | |
- eval_steps_per_second: 1.273 | |
- epoch: 9.73 | |
- step: 12300 | |
## 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: 0.0003 | |
- train_batch_size: 4 | |
- eval_batch_size: 8 | |
- seed: 42 | |
- gradient_accumulation_steps: 2 | |
- total_train_batch_size: 8 | |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 | |
- lr_scheduler_type: linear | |
- lr_scheduler_warmup_steps: 400 | |
- num_epochs: 10 | |
- mixed_precision_training: Native AMP | |
### Framework versions | |
- Transformers 4.11.3 | |
- Pytorch 1.10.0+cu113 | |
- Datasets 1.18.3 | |
- Tokenizers 0.10.3 | |