File size: 1,954 Bytes
3fbf29e |
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 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 |
---
language:
- th
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- fruk19/C_SMALL
metrics:
- wer
model-index:
- name: South_asri
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: aicookcook
type: fruk19/C_SMALL
config: default
split: None
args: 'config: th'
metrics:
- name: Wer
type: wer
value: 3.7677461386031106
---
<!-- 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. -->
# South_asri
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the aicookcook dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0347
- Wer: 3.7677
## 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: 4
- 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: 6
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 0.0304 | 2.0 | 6000 | 0.0440 | 5.5648 |
| 0.0061 | 4.0 | 12000 | 0.0358 | 4.1532 |
| 0.0007 | 6.0 | 18000 | 0.0347 | 3.7677 |
### Framework versions
- Transformers 4.41.2
- Pytorch 2.0.1+cu117
- Datasets 2.20.0
- Tokenizers 0.19.1
|