metadata
license: apache-2.0
tags:
- automatic-speech-recognition
- multilingual_librispeech
- generated_from_trainer
datasets:
- multilingual_librispeech
model-index:
- name: wav2vec2-xlsr-53-300m-mls-german-ft
results: []
wav2vec2-xlsr-53-300m-mls-german-ft
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the MULTILINGUAL_LIBRISPEECH - GERMAN dataset. It achieves the following results on the evaluation set:
- Loss: 0.1936
- Wer: 0.1316
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: 0.0001
- train_batch_size: 32
- 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: 1000
- num_epochs: 100.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.9888 | 7.25 | 500 | 2.9192 | 1.0 |
| 2.9313 | 14.49 | 1000 | 2.8698 | 1.0 |
| 1.068 | 21.74 | 1500 | 0.2647 | 0.2565 |
| 0.8151 | 28.99 | 2000 | 0.2067 | 0.1719 |
| 0.764 | 36.23 | 2500 | 0.1975 | 0.1568 |
| 0.7332 | 43.48 | 3000 | 0.1812 | 0.1463 |
| 0.5952 | 50.72 | 3500 | 0.1923 | 0.1428 |
| 0.6655 | 57.97 | 4000 | 0.1900 | 0.1404 |
| 0.574 | 65.22 | 4500 | 0.1822 | 0.1370 |
| 0.6211 | 72.46 | 5000 | 0.1937 | 0.1355 |
| 0.5883 | 79.71 | 5500 | 0.1872 | 0.1335 |
| 0.5666 | 86.96 | 6000 | 0.1874 | 0.1324 |
| 0.5436 | 94.2 | 6500 | 0.1924 | 0.1311 |
Framework versions
- Transformers 4.13.0.dev0
- Pytorch 1.10.0
- Datasets 1.15.2.dev0
- Tokenizers 0.10.3