metadata
license: apache-2.0
base_model: bert-large-uncased
tags:
- generated_from_trainer
datasets:
- massive
metrics:
- accuracy
model-index:
- name: bert-large-uncased_peft
results: []
bert-large-uncased_peft
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: 0.5312
- Accuracy: 0.8742
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 16
- 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: constant
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.7801 | 1.39 | 500 | 0.6850 | 0.8219 |
| 0.5228 | 2.78 | 1000 | 0.5273 | 0.8736 |
| 0.3262 | 4.17 | 1500 | 0.5020 | 0.8819 |
| 0.2176 | 5.56 | 2000 | 0.5455 | 0.8864 |
| 0.1605 | 6.94 | 2500 | 0.5884 | 0.8834 |
| 0.1105 | 8.33 | 3000 | 0.6838 | 0.8731 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.14.0