File size: 1,847 Bytes
f3a9d37 |
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 |
---
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
model-index:
- name: Whisper-medium-Jibbali_lang
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-medium-Jibbali_lang
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0149
## 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: 0.001
- train_batch_size: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.0447 | 1.0 | 300 | 0.0417 |
| 0.0028 | 2.0 | 600 | 0.0259 |
| 0.0084 | 3.0 | 900 | 0.0181 |
| 0.0029 | 4.0 | 1200 | 0.0140 |
| 0.002 | 5.0 | 1500 | 0.0146 |
| 0.0033 | 6.0 | 1800 | 0.0157 |
| 0.0032 | 7.0 | 2100 | 0.0133 |
| 0.0019 | 8.0 | 2400 | 0.0155 |
| 0.0006 | 9.0 | 2700 | 0.0143 |
| 0.0013 | 10.0 | 3000 | 0.0149 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2
|