bert-large-uncased_finetuning
This model is a fine-tuned version of bert-large-uncased on the massive dataset. It achieves the following results on the evaluation set:
- Loss: 0.6719
- Accuracy: 0.8635
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
1.748 | 1.39 | 500 | 0.6449 | 0.8426 |
0.5674 | 2.78 | 1000 | 0.6501 | 0.8564 |
0.385 | 4.17 | 1500 | 0.6410 | 0.8623 |
0.2833 | 5.56 | 2000 | 0.6784 | 0.8495 |
0.202 | 6.94 | 2500 | 0.7068 | 0.8716 |
0.1405 | 8.33 | 3000 | 0.7838 | 0.8770 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.14.0
- Downloads last month
- 4