STS-conventional-Fine-Tuning-Capstone-roberta-base-filtered-270

This model is a fine-tuned version of FacebookAI/roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0848
• Accuracy: 0.7172

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: 3e-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
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	113	0.7531	0.6854
No log	2.0	226	0.7443	0.7060
No log	3.0	339	0.9619	0.6779
No log	4.0	452	0.8387	0.7022
0.4999	5.0	565	1.2001	0.6966
0.4999	6.0	678	1.2661	0.7060
0.4999	7.0	791	1.3723	0.7172
0.4999	8.0	904	1.6172	0.7303
0.1394	9.0	1017	1.7880	0.7116
0.1394	10.0	1130	1.8037	0.7228
0.1394	11.0	1243	1.8644	0.7303
0.1394	12.0	1356	1.9682	0.7210
0.1394	13.0	1469	2.0287	0.7266
0.0446	14.0	1582	2.0842	0.7247
0.0446	15.0	1695	2.0848	0.7172

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2