metadata
license: mit
base_model: facebook/w2v-bert-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec-reptiles
results: []
wav2vec-reptiles
This model is a fine-tuned version of facebook/w2v-bert-2.0 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 483.1183
- Pcc Accuracy: -0.0911
- Pcc Fluency: -0.0702
- Pcc Total Score: -0.0823
- Pcc Content: -0.0799
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: 5e-05
- train_batch_size: 4
- eval_batch_size: 6
- 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: cosine
- lr_scheduler_warmup_ratio: 0.4
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Pcc Accuracy | Pcc Fluency | Pcc Total Score | Pcc Content |
|---|---|---|---|---|---|---|---|
| 3259.2658 | 1.07 | 500 | 2804.9829 | -0.2619 | -0.2376 | -0.2620 | -0.2464 |
| 1980.8576 | 2.13 | 1000 | 2444.1970 | -0.2591 | -0.2359 | -0.2573 | -0.2394 |
| 1728.8568 | 3.2 | 1500 | 1480.2877 | -0.2539 | -0.2310 | -0.2519 | -0.2337 |
| 490.517 | 4.27 | 2000 | 577.7591 | -0.2334 | -0.2102 | -0.2303 | -0.2141 |
| 761.9771 | 5.34 | 2500 | 503.8331 | -0.1885 | -0.1646 | -0.1829 | -0.1714 |
| 286.5463 | 6.4 | 3000 | 496.2837 | -0.1460 | -0.1234 | -0.1390 | -0.1315 |
| 341.2555 | 7.47 | 3500 | 488.9790 | -0.1150 | -0.0937 | -0.1072 | -0.1026 |
| 544.4691 | 8.54 | 4000 | 484.3144 | -0.0955 | -0.0745 | -0.0870 | -0.0841 |
| 377.6802 | 9.61 | 4500 | 483.1183 | -0.0911 | -0.0702 | -0.0823 | -0.0799 |
Framework versions
- Transformers 4.37.0
- Pytorch 2.1.2
- Datasets 2.17.1
- Tokenizers 0.15.1