xls-r-asr_af-run6
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the asr_af dataset. It achieves the following results:
- Wer (Validation): 41.33%
- Wer (Test): 42.49%
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 (Train) |
---|---|---|---|---|
8.6153 | 0.59 | 100 | 4.0306 | 1.0 |
3.2623 | 1.17 | 200 | 3.0277 | 1.0 |
2.9667 | 1.76 | 300 | 2.9436 | 1.0 |
2.4516 | 2.35 | 400 | 1.4571 | 0.9030 |
1.174 | 2.93 | 500 | 0.9461 | 0.7412 |
0.7792 | 3.52 | 600 | 0.6839 | 0.6080 |
0.5749 | 4.11 | 700 | 0.5418 | 0.5068 |
0.4187 | 4.69 | 800 | 0.5341 | 0.4902 |
0.36 | 5.28 | 900 | 0.5231 | 0.4746 |
0.2934 | 5.87 | 1000 | 0.4457 | 0.4133 |
0.2338 | 6.45 | 1100 | 0.4904 | 0.4157 |
0.2245 | 7.04 | 1200 | 0.4952 | 0.4115 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3
- Downloads last month
- 3
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.