<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
HBERTv1_48_L10_H512_A8_emotion
This model is a fine-tuned version of gokuls/HBERTv1_48_L10_H512_A8 on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.3384
- Accuracy: 0.8935
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 33
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
1.2329 | 1.0 | 250 | 0.9004 | 0.673 |
0.5962 | 2.0 | 500 | 0.4426 | 0.8655 |
0.3688 | 3.0 | 750 | 0.3734 | 0.876 |
0.2824 | 4.0 | 1000 | 0.3868 | 0.8795 |
0.2242 | 5.0 | 1250 | 0.3715 | 0.879 |
0.1805 | 6.0 | 1500 | 0.3384 | 0.8935 |
0.1546 | 7.0 | 1750 | 0.3623 | 0.8815 |
0.1273 | 8.0 | 2000 | 0.3532 | 0.8885 |
0.1029 | 9.0 | 2250 | 0.3974 | 0.886 |
0.0848 | 10.0 | 2500 | 0.4109 | 0.8915 |
Framework versions
- Transformers 4.34.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.14.5
- Tokenizers 0.14.0