checkpoints_10_1_microsoft_deberta_V1.1_384
This model is a fine-tuned version of VuongQuoc/checkpoints_30_9_microsoft_deberta_V1.0_384 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.7688
- Map@3: 0.8458
- Accuracy: 0.75
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-06
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 1200
Training results
| Training Loss | Epoch | Step | Validation Loss | Map@3 | Accuracy |
|---|---|---|---|---|---|
| 1.5583 | 0.05 | 100 | 1.4269 | 0.7675 | 0.65 |
| 1.1541 | 0.11 | 200 | 1.0838 | 0.7692 | 0.67 |
| 1.0124 | 0.16 | 300 | 0.9475 | 0.8108 | 0.715 |
| 0.9627 | 0.21 | 400 | 0.8969 | 0.8233 | 0.73 |
| 0.9241 | 0.27 | 500 | 0.8473 | 0.8392 | 0.755 |
| 0.885 | 0.32 | 600 | 0.8336 | 0.8333 | 0.745 |
| 0.8606 | 0.37 | 700 | 0.7937 | 0.8508 | 0.76 |
| 0.8495 | 0.43 | 800 | 0.7755 | 0.8517 | 0.76 |
| 0.8787 | 0.48 | 900 | 0.7706 | 0.8475 | 0.75 |
| 0.8535 | 0.53 | 1000 | 0.7714 | 0.8458 | 0.75 |
| 0.8499 | 0.59 | 1100 | 0.7694 | 0.8458 | 0.75 |
| 0.8353 | 0.64 | 1200 | 0.7688 | 0.8458 | 0.75 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.0.0
- Datasets 2.9.0
- Tokenizers 0.13.3