finetuned-bert-base-uncased-on-HOPE

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: 1.3515
• Accuracy: 0.5345

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7368	1.0	289	1.6685	0.4526
1.3051	2.0	578	1.4303	0.5176
1.0563	3.0	867	1.3849	0.5438
1.1101	4.0	1156	1.4233	0.5185
0.9118	5.0	1445	1.5438	0.5023
0.7889	6.0	1734	1.6832	0.5014
0.4892	7.0	2023	1.8469	0.4824
0.3739	8.0	2312	2.0680	0.4734
0.3813	9.0	2601	2.1392	0.4706
0.3459	10.0	2890	2.2772	0.4761
0.2323	11.0	3179	2.3445	0.4688
0.1977	12.0	3468	2.4754	0.4761
0.2351	13.0	3757	2.5912	0.4661
0.1991	14.0	4046	2.6713	0.4743
0.2239	15.0	4335	2.7262	0.4706
0.155	16.0	4624	2.7958	0.4697
0.1675	17.0	4913	2.8367	0.4724
0.1471	18.0	5202	2.8619	0.4715
0.1973	19.0	5491	2.8744	0.4770
0.1902	20.0	5780	2.8865	0.4752

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1