File size: 2,270 Bytes
7b15aef 5e2a607 7b15aef 5e2a607 7b15aef 5e2a607 |
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 |
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- google/fleurs
metrics:
- wer
model-index:
- name: xls-r-fleurs_nl-run2
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: audiofolder
type: audiofolder
config: default
split: validation
args: default
metrics:
- name: Wer
type: wer
value: 0.38678223185265437
---
<!-- 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. -->
# xls-r-fleurs_nl-run2
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the FLEURS (nl) dataset.
It achieves the following results:
- Wer (Validation): 38.05%
- Wer (Test): 37.20%
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 7.4175 | 0.55 | 100 | 3.8140 | 1.0 |
| 3.2001 | 1.1 | 200 | 2.9283 | 1.0 |
| 2.9162 | 1.64 | 300 | 2.8922 | 1.0 |
| 2.8398 | 2.19 | 400 | 2.3875 | 0.9756 |
| 1.2121 | 2.74 | 500 | 0.8873 | 0.7050 |
| 0.5833 | 3.29 | 600 | 0.6149 | 0.5276 |
| 0.4103 | 3.84 | 700 | 0.5275 | 0.4575 |
| 0.2901 | 4.38 | 800 | 0.5056 | 0.4510 |
| 0.2584 | 4.93 | 900 | 0.5021 | 0.4363 |
| 0.1929 | 5.48 | 1000 | 0.4881 | 0.4087 |
| 0.1747 | 6.03 | 1100 | 0.4961 | 0.4209 |
| 0.1469 | 6.58 | 1200 | 0.4893 | 0.3868 |
### Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3 |