roberta-base-riddle-finetuned_new_3choice

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

• Loss: 0.1949
• Accuracy: 0.875

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: 0.0005
• 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	12	0.4873	0.7250
No log	2.0	24	0.3075	0.8000
No log	3.0	36	0.4185	0.8000
No log	4.0	48	0.3031	0.8250
No log	5.0	60	0.3392	0.8250
No log	6.0	72	0.3420	0.8500
No log	7.0	84	0.3509	0.8250
No log	8.0	96	0.3089	0.8500
No log	9.0	108	0.2797	0.8250
No log	10.0	120	0.2378	0.8000
No log	11.0	132	0.2622	0.875
No log	12.0	144	0.2334	0.9000
No log	13.0	156	0.2314	0.9000
No log	14.0	168	0.1987	0.875
No log	15.0	180	0.1949	0.875

Framework versions

• Transformers 4.37.2
• Pytorch 1.13.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.0