File size: 2,019 Bytes
f38bc15 |
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 |
---
language:
- fy
base_model: distil-small.en
tags:
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_6_1
metrics:
- wer
model-index:
- name: DistilFT-Frisian-1h
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_6_fy_NL
type: mozilla-foundation/common_voice_6_1
args: 'config: fy-NL, split: train-1h'
metrics:
- name: Wer
type: wer
value: 77.86134378898592
---
<!-- 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. -->
# DistilFT-Frisian-1h
This model is a fine-tuned version of [distil-small.en](https://huggingface.co/distil-small.en) on the mozilla-foundation/common_voice_6_fy_NL dataset.
It achieves the following results on the evaluation set:
- Loss: 2.0756
- Wer: 77.8613
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 2000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-------:|:----:|:---------------:|:-------:|
| 2.2753 | 5.6180 | 500 | 2.6854 | 87.9023 |
| 1.5355 | 11.2360 | 1000 | 2.2463 | 81.0515 |
| 1.201 | 16.8539 | 1500 | 2.1075 | 78.7239 |
| 1.0818 | 22.4719 | 2000 | 2.0756 | 77.8613 |
### Framework versions
- Transformers 4.41.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1
|