metadata
license: mit
base_model: microsoft/deberta-v2-xxlarge
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: output
results: []
output
This model is a fine-tuned version of microsoft/deberta-v2-xxlarge on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.7550
- Accuracy: 0.6786
- Macro F1: 0.6773
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: 3e-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 |
|---|---|---|---|---|---|
| 1.6193 | 0.2286 | 100 | 1.6018 | 0.2184 | 0.1357 |
| 1.305 | 0.4571 | 200 | 0.9285 | 0.591 | 0.5953 |
| 0.8772 | 0.6857 | 300 | 0.8561 | 0.6256 | 0.6250 |
| 0.8552 | 0.9143 | 400 | 0.8332 | 0.6511 | 0.6473 |
| 0.798 | 1.1429 | 500 | 0.8210 | 0.6641 | 0.6579 |
| 0.7713 | 1.3714 | 600 | 0.7759 | 0.666 | 0.6669 |
| 0.7758 | 1.6 | 700 | 0.7634 | 0.6667 | 0.6615 |
| 0.7442 | 1.8286 | 800 | 0.7960 | 0.6613 | 0.6590 |
| 0.752 | 2.0571 | 900 | 0.7715 | 0.667 | 0.6690 |
| 0.7123 | 2.2857 | 1000 | 0.7600 | 0.6696 | 0.6698 |
| 0.7066 | 2.5143 | 1100 | 0.7599 | 0.6701 | 0.6684 |
| 0.7024 | 2.7429 | 1200 | 0.7551 | 0.6757 | 0.6763 |
| 0.7117 | 2.9714 | 1300 | 0.7550 | 0.6786 | 0.6773 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.2
- Datasets 2.19.0
- Tokenizers 0.19.1