File size: 1,620 Bytes
5fb7f8c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 |
---
language:
- hre
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- ntviet/Hre-audio-dataset6
model-index:
- name: Whisper Small Hre 6 - female voice - 500 steps, metric CER
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. -->
# Whisper Small Hre 6 - female voice - 500 steps, metric CER
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Hre audio dataset 6 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0693
- Cer Ortho: 9.3458
- Cer: 0.0
## 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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 50
- training_steps: 500
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Cer Ortho | Cer |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:---:|
| 0.1016 | 4.13 | 500 | 0.0693 | 9.3458 | 0.0 |
### Framework versions
- Transformers 4.39.3
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
|