add-bert-Massive-intent_24

This model is a fine-tuned version of gokuls/add_bert_12_layer_model_complete_training_new on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 0.9139
• Accuracy: 0.8549

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7362	1.0	180	0.9126	0.7590
0.7607	2.0	360	0.7920	0.7870
0.5345	3.0	540	0.7643	0.8062
0.3975	4.0	720	0.7447	0.8131
0.2823	5.0	900	0.7352	0.8269
0.214	6.0	1080	0.7413	0.8308
0.1642	7.0	1260	0.7857	0.8357
0.1215	8.0	1440	0.8389	0.8337
0.0896	9.0	1620	0.8059	0.8515
0.0636	10.0	1800	0.8186	0.8519
0.0439	11.0	1980	0.8643	0.8510
0.0297	12.0	2160	0.8882	0.8485
0.0168	13.0	2340	0.9139	0.8549
0.0094	14.0	2520	0.9200	0.8529
0.0048	15.0	2700	0.9243	0.8544

Framework versions

• Transformers 4.30.2
• Pytorch 1.14.0a0+410ce96
• Datasets 2.13.0
• Tokenizers 0.13.3