metadata
license: mit
base_model: microsoft/deberta-v3-large
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: taskA-DeBERTa-large-conf-1.0.0
results: []
taskA-DeBERTa-large-conf-1.0.0
This model is a fine-tuned version of microsoft/deberta-v3-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.0557
- Accuracy: 0.8003
- Precision: 0.6080
- Recall: 0.5677
- F1: 0.5872
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: 4e-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.5219 | 0.39 | 500 | 0.5422 | 0.8147 | 0.725 | 0.4179 | 0.5302 |
| 0.5372 | 0.78 | 1000 | 0.6654 | 0.7873 | 0.9194 | 0.1643 | 0.2787 |
| 0.4255 | 1.16 | 1500 | 0.7670 | 0.8133 | 0.6774 | 0.4841 | 0.5647 |
| 0.3873 | 1.55 | 2000 | 1.2863 | 0.7592 | 0.5132 | 0.7262 | 0.6014 |
| 0.4026 | 1.94 | 2500 | 0.9782 | 0.7967 | 0.5964 | 0.5793 | 0.5877 |
| 0.2893 | 2.33 | 3000 | 1.0557 | 0.8003 | 0.6080 | 0.5677 | 0.5872 |
| 0.2513 | 2.71 | 3500 | 1.2664 | 0.7779 | 0.5470 | 0.6542 | 0.5958 |
| 0.231 | 3.1 | 4000 | 1.2628 | 0.7895 | 0.5745 | 0.6110 | 0.5922 |
| 0.1684 | 3.49 | 4500 | 1.1848 | 0.8061 | 0.6211 | 0.5764 | 0.5979 |
| 0.1647 | 3.88 | 5000 | 1.2085 | 0.8068 | 0.6262 | 0.5648 | 0.5939 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2