N_roberta_imdb_padding10model

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

• Loss: 0.5407
• Accuracy: 0.9524

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: 16
• eval_batch_size: 16
• 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
0.2113	1.0	1563	0.2381	0.9337
0.1641	2.0	3126	0.1671	0.9498
0.1084	3.0	4689	0.2624	0.9476
0.0731	4.0	6252	0.2613	0.9496
0.0488	5.0	7815	0.3130	0.9481
0.0398	6.0	9378	0.3571	0.9473
0.0254	7.0	10941	0.3278	0.9494
0.0282	8.0	12504	0.4027	0.9466
0.0214	9.0	14067	0.3643	0.9493
0.0151	10.0	15630	0.4171	0.9495
0.0146	11.0	17193	0.4385	0.9505
0.0075	12.0	18756	0.4331	0.9523
0.0133	13.0	20319	0.3997	0.952
0.0053	14.0	21882	0.4604	0.9530
0.0044	15.0	23445	0.4686	0.952
0.0003	16.0	25008	0.5177	0.9508
0.0028	17.0	26571	0.5353	0.9514
0.0012	18.0	28134	0.5197	0.9525
0.0001	19.0	29697	0.5354	0.9519
0.0006	20.0	31260	0.5407	0.9524

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3