amazon_massive_intent-en-US-distilbert

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

• Loss: 0.7769
• Accuracy: 0.8869

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1629	1.0	720	0.9624	0.7959
0.7179	2.0	1440	0.5693	0.8569
0.3695	3.0	2160	0.5104	0.8692
0.2227	4.0	2880	0.4813	0.8888
0.1302	5.0	3600	0.5207	0.8834
0.0822	6.0	4320	0.5709	0.8829
0.0527	7.0	5040	0.6016	0.8854
0.0345	8.0	5760	0.6373	0.8839
0.0244	9.0	6480	0.6616	0.8908
0.0191	10.0	7200	0.6958	0.8854
0.0131	11.0	7920	0.7212	0.8844
0.0095	12.0	8640	0.7348	0.8864
0.0067	13.0	9360	0.7502	0.8824
0.0055	14.0	10080	0.7469	0.8819
0.0038	15.0	10800	0.7772	0.8854
0.0043	16.0	11520	0.7761	0.8844
0.0025	17.0	12240	0.7625	0.8893
0.002	18.0	12960	0.7782	0.8854
0.0014	19.0	13680	0.7867	0.8859
0.0019	20.0	14400	0.7769	0.8869

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3