bert-base-Massive-intent_24

This model is a fine-tuned version of gokuls/bert_base_24 on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 0.8019
• Accuracy: 0.8559

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
1.6309	1.0	180	0.9071	0.7727
0.7005	2.0	360	0.6913	0.8185
0.4328	3.0	540	0.6321	0.8455
0.2875	4.0	720	0.6583	0.8333
0.2036	5.0	900	0.6765	0.8426
0.1437	6.0	1080	0.7043	0.8446
0.1088	7.0	1260	0.7193	0.8510
0.0812	8.0	1440	0.7489	0.8426
0.0622	9.0	1620	0.7450	0.8495
0.0453	10.0	1800	0.7722	0.8500
0.0346	11.0	1980	0.7849	0.8470
0.0227	12.0	2160	0.8088	0.8515
0.0166	13.0	2340	0.8019	0.8559
0.0114	14.0	2520	0.7968	0.8549
0.0078	15.0	2700	0.7949	0.8549

Framework versions

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