SkyR/roberta-base-ours-run-2

run-2

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

• Loss: 2.1449
• Accuracy: 0.75
• Precision: 0.7115
• Recall: 0.7093
• F1: 0.7103

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: 5e-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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.9838	1.0	50	0.8621	0.645	0.6536	0.6130	0.6124
0.7134	2.0	100	0.8124	0.7	0.6628	0.6421	0.6483
0.4911	3.0	150	0.8571	0.7	0.6726	0.6314	0.6361
0.3104	4.0	200	0.8228	0.76	0.7298	0.7367	0.7294
0.1942	5.0	250	1.1132	0.76	0.7282	0.7031	0.7119
0.1409	6.0	300	1.2218	0.685	0.6516	0.6560	0.6524
0.0976	7.0	350	1.3648	0.715	0.6984	0.7044	0.6946
0.0791	8.0	400	1.5985	0.745	0.7183	0.7113	0.7124
0.0647	9.0	450	1.8884	0.725	0.6818	0.6761	0.6785
0.0275	10.0	500	1.8639	0.725	0.6979	0.7008	0.6958
0.0329	11.0	550	1.8831	0.72	0.6816	0.6869	0.6838
0.0169	12.0	600	2.1426	0.73	0.6864	0.6776	0.6794
0.0072	13.0	650	2.2483	0.725	0.7187	0.7054	0.6968
0.0203	14.0	700	2.2901	0.735	0.6986	0.6885	0.6921
0.0093	15.0	750	2.3134	0.725	0.6830	0.6666	0.6723
0.0089	16.0	800	2.1598	0.73	0.6919	0.6860	0.6885
0.0061	17.0	850	2.0879	0.75	0.7129	0.7132	0.7125
0.0024	18.0	900	2.1285	0.745	0.7062	0.7071	0.7049
0.0043	19.0	950	2.1386	0.74	0.7001	0.7003	0.6985
0.0028	20.0	1000	2.1449	0.75	0.7115	0.7093	0.7103

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.1+cu116
• Tokenizers 0.13.2