|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: dys_asr_10min_rub |
|
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. --> |
|
|
|
# dys_asr_10min_rub |
|
|
|
This model is a fine-tuned version of [Splend1dchan/wav2vec2-large-10min-lv60-self](https://huggingface.co/Splend1dchan/wav2vec2-large-10min-lv60-self) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0359 |
|
- Wer: 0.9969 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 1000 |
|
- num_epochs: 7 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:| |
|
| 12.8427 | 0.79 | 500 | 2.4894 | 1.0 | |
|
| 1.4059 | 1.58 | 1000 | 0.2969 | 1.1495 | |
|
| 0.425 | 2.37 | 1500 | 0.1267 | 1.0623 | |
|
| 0.2377 | 3.16 | 2000 | 0.0781 | 1.0374 | |
|
| 0.1148 | 3.95 | 2500 | 0.0393 | 1.0062 | |
|
| 0.1136 | 4.74 | 3000 | 0.0432 | 1.0031 | |
|
| 0.0987 | 5.53 | 3500 | 0.0411 | 0.9969 | |
|
| 0.0746 | 6.32 | 4000 | 0.0359 | 0.9969 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.17.0 |
|
- Pytorch 1.13.1+cu116 |
|
- Datasets 1.18.3 |
|
- Tokenizers 0.13.2 |
|
|
