File size: 4,177 Bytes
6b55d88 |
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 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 |
---
library_name: peft
language:
- sw
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- fsicoli/common_voice_19_0
model-index:
- name: eolang/sw-peft-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. -->
# eolang/sw-peft-2
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the Common Voice 19 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3697
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 1
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.5084 | 0.0191 | 100 | 1.0225 |
| 0.9199 | 0.0382 | 200 | 0.9041 |
| 0.8295 | 0.0573 | 300 | 0.8428 |
| 0.8404 | 0.0764 | 400 | 0.8142 |
| 0.8215 | 0.0955 | 500 | 0.8141 |
| 0.7527 | 0.1147 | 600 | 0.7640 |
| 0.7494 | 0.1338 | 700 | 0.7652 |
| 0.7693 | 0.1529 | 800 | 0.7081 |
| 0.7104 | 0.1720 | 900 | 0.7238 |
| 0.6975 | 0.1911 | 1000 | 0.6990 |
| 0.7164 | 0.2102 | 1100 | 0.7002 |
| 0.6693 | 0.2293 | 1200 | 0.6842 |
| 0.7043 | 0.2484 | 1300 | 0.6831 |
| 0.6521 | 0.2675 | 1400 | 0.6527 |
| 0.6468 | 0.2866 | 1500 | 0.6563 |
| 0.6376 | 0.3058 | 1600 | 0.6180 |
| 0.6008 | 0.3249 | 1700 | 0.6223 |
| 0.6353 | 0.3440 | 1800 | 0.6113 |
| 0.5885 | 0.3631 | 1900 | 0.6033 |
| 0.598 | 0.3822 | 2000 | 0.5987 |
| 0.5749 | 0.4013 | 2100 | 0.5792 |
| 0.5714 | 0.4204 | 2200 | 0.5772 |
| 0.5438 | 0.4395 | 2300 | 0.5688 |
| 0.5442 | 0.4586 | 2400 | 0.5711 |
| 0.5165 | 0.4777 | 2500 | 0.5588 |
| 0.4971 | 0.4968 | 2600 | 0.5408 |
| 0.5026 | 0.5160 | 2700 | 0.5365 |
| 0.5278 | 0.5351 | 2800 | 0.5112 |
| 0.5371 | 0.5542 | 2900 | 0.5160 |
| 0.5013 | 0.5733 | 3000 | 0.5041 |
| 0.4867 | 0.5924 | 3100 | 0.4978 |
| 0.4938 | 0.6115 | 3200 | 0.4830 |
| 0.4522 | 0.6306 | 3300 | 0.4798 |
| 0.4515 | 0.6497 | 3400 | 0.4751 |
| 0.4593 | 0.6688 | 3500 | 0.4631 |
| 0.4539 | 0.6879 | 3600 | 0.4561 |
| 0.4557 | 0.7071 | 3700 | 0.4467 |
| 0.417 | 0.7262 | 3800 | 0.4419 |
| 0.4251 | 0.7453 | 3900 | 0.4368 |
| 0.4062 | 0.7644 | 4000 | 0.4277 |
| 0.3815 | 0.7835 | 4100 | 0.4271 |
| 0.3832 | 0.8026 | 4200 | 0.4155 |
| 0.3818 | 0.8217 | 4300 | 0.4098 |
| 0.3988 | 0.8408 | 4400 | 0.4005 |
| 0.4073 | 0.8599 | 4500 | 0.3964 |
| 0.3894 | 0.8790 | 4600 | 0.3898 |
| 0.3464 | 0.8981 | 4700 | 0.3858 |
| 0.3626 | 0.9173 | 4800 | 0.3800 |
| 0.3753 | 0.9364 | 4900 | 0.3771 |
| 0.3734 | 0.9555 | 5000 | 0.3733 |
| 0.3362 | 0.9746 | 5100 | 0.3718 |
| 0.3607 | 0.9937 | 5200 | 0.3697 |
### Framework versions
- PEFT 0.14.0
- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.3.1
- Tokenizers 0.21.0 |