Amaan39/Roberta-Webis-CPC

results

This model is a fine-tuned version of FacebookAI/roberta-base on an Webi-CPC-11 dataset. It achieves the following results on the evaluation set:

• Loss: 1.5703
• Accuracy: 0.8432

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	197	0.4964	0.6875
No log	2.0	394	0.4895	0.8326
0.4856	3.0	591	0.3711	0.8422
0.4856	4.0	788	0.3289	0.8464
0.4856	5.0	985	0.4112	0.8559
0.2928	6.0	1182	0.5872	0.8453
0.2928	7.0	1379	0.5353	0.8284
0.1493	8.0	1576	0.6069	0.8273
0.1493	9.0	1773	0.9225	0.8464
0.1493	10.0	1970	1.3133	0.8422
0.0641	11.0	2167	1.2524	0.8369
0.0641	12.0	2364	1.1893	0.8347
0.0394	13.0	2561	1.3631	0.8358
0.0394	14.0	2758	1.1922	0.8273
0.0394	15.0	2955	1.2648	0.8316
0.0205	16.0	3152	1.0889	0.8422
0.0205	17.0	3349	1.2235	0.8422
0.0094	18.0	3546	1.4707	0.8358
0.0094	19.0	3743	1.3305	0.8475
0.0094	20.0	3940	1.4021	0.8263
0.0151	21.0	4137	1.2689	0.8358
0.0151	22.0	4334	1.4997	0.8273
0.0061	23.0	4531	1.4872	0.8358
0.0061	24.0	4728	1.5773	0.8347
0.0061	25.0	4925	1.6127	0.8358
0.0037	26.0	5122	1.5534	0.8326
0.0037	27.0	5319	1.5532	0.8453
0.0036	28.0	5516	1.4986	0.8432
0.0036	29.0	5713	1.5698	0.8422
0.0036	30.0	5910	1.5703	0.8432

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Tokenizers 0.21.0