File size: 1,780 Bytes
4367c4a |
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 |
---
language:
- ar
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: ardj5.2
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. -->
# ardj5.2
This model is a fine-tuned version of [faissalb/whisper-small-ardj5.1](https://huggingface.co/faissalb/whisper-small-ardj5.1) on the test dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2645
- Wer: 10.7393
## 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-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 6000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0157 | 2.15 | 1000 | 0.2628 | 10.7443 |
| 0.0079 | 4.3 | 2000 | 0.2637 | 10.9840 |
| 0.0092 | 6.44 | 3000 | 0.2635 | 10.7842 |
| 0.0094 | 8.59 | 4000 | 0.2634 | 10.7892 |
| 0.0039 | 10.74 | 5000 | 0.2639 | 10.7692 |
| 0.0051 | 12.89 | 6000 | 0.2645 | 10.7393 |
### Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.3
- Tokenizers 0.13.3
|