bert-uncased-massive-intent-classification

This model is a fine-tuned version of bert-base-uncased on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 0.8396
• Accuracy: 0.8854

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: 16
• eval_batch_size: 16
• seed: 33
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.4984	1.0	720	0.6402	0.8495
0.4376	2.0	1440	0.5394	0.8731
0.2318	3.0	2160	0.5903	0.8760
0.1414	4.0	2880	0.6221	0.8805
0.087	5.0	3600	0.7072	0.8819
0.0622	6.0	4320	0.7121	0.8819
0.036	7.0	5040	0.7750	0.8805
0.0234	8.0	5760	0.7767	0.8834
0.0157	9.0	6480	0.8243	0.8805
0.0122	10.0	7200	0.8198	0.8839
0.0092	11.0	7920	0.8105	0.8849
0.0047	12.0	8640	0.8561	0.8844
0.0038	13.0	9360	0.8367	0.8815
0.0029	14.0	10080	0.8396	0.8854
0.0014	15.0	10800	0.8410	0.8849

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1