metadata
license: mit
base_model: microsoft/deberta-v3-large
tags:
- generated_from_trainer
metrics:
- f1
model-index:
- name: opus-em-deberta-3-large-v2
results: []
opus-em-deberta-3-large-v2
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: 4.4267
- F1: 0.1942
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: 2e-05
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| 1.2929 | 1.0 | 179 | 13.4522 | 0.1942 |
| 0.1541 | 2.0 | 359 | 8.4684 | 0.1942 |
| 0.1257 | 3.0 | 538 | 7.6370 | 0.1942 |
| 0.1684 | 4.0 | 718 | 0.7054 | 0.6376 |
| 0.0911 | 5.0 | 897 | 5.1195 | 0.1942 |
| 0.145 | 6.0 | 1077 | 0.2694 | 0.7984 |
| 0.1191 | 7.0 | 1256 | 2.9415 | 0.2027 |
| 0.1008 | 8.0 | 1436 | 0.1785 | 0.9023 |
| 0.0231 | 9.0 | 1615 | 8.5722 | 0.1942 |
| 0.0521 | 9.97 | 1790 | 4.4267 | 0.1942 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0