metadata
license: apache-2.0
base_model: nutella-toast/wav2vec2-large-xls-r-ssw
tags:
- generated_from_trainer
datasets:
- ml-superb-subset
metrics:
- wer
model-index:
- name: wav2vec2-large-xls-r-ssw
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: ml-superb-subset
type: ml-superb-subset
config: ssw
split: dev
args: ssw
metrics:
- name: Wer
type: wer
value: 0.940809968847352
wav2vec2-large-xls-r-ssw
This model is a fine-tuned version of nutella-toast/wav2vec2-large-xls-r-ssw on the ml-superb-subset dataset. It achieves the following results on the evaluation set:
- Loss: 1.0000
- Wer: 0.9408
Model description
Finetuned version of vanilla Wav2Vec2 for CS224S.
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: 0.0003
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
1.3831 | 1.0471 | 100 | 1.3053 | 1.0 |
1.2606 | 2.0942 | 200 | 1.1802 | 0.9720 |
1.0789 | 3.1414 | 300 | 1.0889 | 1.0405 |
0.9249 | 4.1885 | 400 | 1.0000 | 0.9408 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1