roberta-mc-4

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4442
• Accuracy: 0.5

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: 0.001
• train_batch_size: 8
• eval_batch_size: 8
• 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	Accuracy
1.6061	1.0	24	1.5920	0.7
1.6032	2.0	48	1.5838	0.6
1.6104	3.0	72	1.5750	0.7
1.5851	4.0	96	1.5584	0.6
1.5653	5.0	120	1.5059	0.7
1.5485	6.0	144	1.4743	0.6
1.5175	7.0	168	1.4500	0.7
1.5025	8.0	192	1.4298	0.5
1.466	9.0	216	1.4559	0.5
1.4444	10.0	240	1.4010	0.5
1.4223	11.0	264	1.4699	0.4
1.3804	12.0	288	1.4915	0.4
1.3884	13.0	312	1.4624	0.4
1.3699	14.0	336	1.4798	0.5
1.3705	15.0	360	1.3615	0.5
1.3383	16.0	384	1.3814	0.7
1.3306	17.0	408	1.5099	0.4
1.2886	18.0	432	1.5039	0.4
1.2964	19.0	456	1.4033	0.5
1.285	20.0	480	1.4596	0.4
1.311	21.0	504	1.4100	0.4
1.218	22.0	528	1.3952	0.5
1.2193	23.0	552	1.2449	0.7
1.2618	24.0	576	1.2691	0.7
1.236	25.0	600	1.3427	0.7
1.1773	26.0	624	1.3669	0.5
1.1873	27.0	648	1.5114	0.5
1.1519	28.0	672	1.4285	0.6
1.1172	29.0	696	1.4485	0.5
1.0677	30.0	720	1.4442	0.5

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.4
• Tokenizers 0.13.3