roberta-base-nsp-100000-1e-06-32

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1623

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: 1e-06
• train_batch_size: 128
• eval_batch_size: 1024
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss
0.5785	1.0	782	0.3139
0.2847	2.0	1564	0.2379
0.2318	3.0	2346	0.2068
0.2039	4.0	3128	0.1906
0.1897	5.0	3910	0.1776
0.1718	6.0	4692	0.1730
0.1604	7.0	5474	0.1684
0.1495	8.0	6256	0.1638
0.139	9.0	7038	0.1618
0.1361	10.0	7820	0.1625
0.1325	11.0	8602	0.1618
0.1266	12.0	9384	0.1593
0.1211	13.0	10166	0.1591
0.1144	14.0	10948	0.1600
0.1166	15.0	11730	0.1596
0.1093	16.0	12512	0.1581
0.104	17.0	13294	0.1581
0.1025	18.0	14076	0.1576
0.0984	19.0	14858	0.1619
0.0976	20.0	15640	0.1608
0.0937	21.0	16422	0.1623

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1