<!-- 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. -->
hbertv2-emotion_w_in
This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new_wt_init on the emotion dataset. It achieves the following results on the evaluation set:
- Loss: 0.2778
- Accuracy: 0.9305
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 |
|---|---|---|---|---|
| 0.647 | 1.0 | 250 | 0.3054 | 0.889 |
| 0.2593 | 2.0 | 500 | 0.2384 | 0.913 |
| 0.1844 | 3.0 | 750 | 0.2164 | 0.923 |
| 0.1493 | 4.0 | 1000 | 0.1990 | 0.9235 |
| 0.1244 | 5.0 | 1250 | 0.1908 | 0.9265 |
| 0.1006 | 6.0 | 1500 | 0.2328 | 0.923 |
| 0.0857 | 7.0 | 1750 | 0.2379 | 0.93 |
| 0.0724 | 8.0 | 2000 | 0.2638 | 0.9235 |
| 0.0591 | 9.0 | 2250 | 0.2778 | 0.9305 |
| 0.0407 | 10.0 | 2500 | 0.3189 | 0.927 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3
