bert-base-uncased-nsp-10000-1e-06-16

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

• Loss: 0.4019

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: 64
• 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
No log	1.0	157	0.6964
0.7027	2.0	314	0.6912
0.6942	3.0	471	0.6809
0.6806	4.0	628	0.6548
0.6806	5.0	785	0.6324
0.6538	6.0	942	0.6146
0.6265	7.0	1099	0.6019
0.6071	8.0	1256	0.5906
0.5947	9.0	1413	0.5812
0.5947	10.0	1570	0.5710
0.5781	11.0	1727	0.5594
0.5666	12.0	1884	0.5410
0.5422	13.0	2041	0.5192
0.5422	14.0	2198	0.4981
0.5154	15.0	2355	0.4777
0.4849	16.0	2512	0.4636
0.4559	17.0	2669	0.4514
0.4421	18.0	2826	0.4442
0.4421	19.0	2983	0.4328
0.4223	20.0	3140	0.4251
0.4069	21.0	3297	0.4197
0.3988	22.0	3454	0.4165
0.3912	23.0	3611	0.4120
0.3912	24.0	3768	0.4086
0.3803	25.0	3925	0.4060
0.3847	26.0	4082	0.4050
0.3749	27.0	4239	0.4032
0.3749	28.0	4396	0.4032
0.37	29.0	4553	0.4020
0.3694	30.0	4710	0.4019

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1