metadata
base_model: lnxdx/B4_1000_1e-5_hp-myself-2
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: C2_1000_1e-5_hp-myself-2
results: []
C2_1000_1e-5_hp-myself-2
This model is a fine-tuned version of lnxdx/B4_1000_1e-5_hp-myself-2 on the None dataset. It achieves the following results on the evaluation set:
- Loss on ShEMO train set: 0.7065
- Loss on ShEMO dev set: 0.6634
- WER on ShEMO train set: 26.52
- WER on ShEMO dev set: 30.87
- WER on Common Voice 13 test set: 19.43
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: 1e-05
- 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
- training_steps: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.7564 | 0.62 | 100 | 0.6705 | 0.3131 |
| 0.7761 | 1.25 | 200 | 0.6664 | 0.3140 |
| 0.7722 | 1.88 | 300 | 0.6573 | 0.3137 |
| 0.7035 | 2.5 | 400 | 0.6627 | 0.3157 |
| 0.7026 | 3.12 | 500 | 0.6834 | 0.3107 |
| 0.7213 | 3.75 | 600 | 0.6561 | 0.3169 |
| 0.6996 | 4.38 | 700 | 0.6664 | 0.3096 |
| 0.7146 | 5.0 | 800 | 0.6593 | 0.3148 |
| 0.7071 | 5.62 | 900 | 0.6646 | 0.3125 |
| 0.7065 | 6.25 | 1000 | 0.6634 | 0.3107 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0