roberta-large-pp-500000-1e-06-128-negcommonsensebalanced-1e-06-64

This model is a fine-tuned version of mhr2004/roberta-large-pp-500000-1e-06-128 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3431

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: 1e-06
• train_batch_size: 256
• eval_batch_size: 1024
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss
0.5214	1.0	795	0.4808
0.4627	2.0	1590	0.4284
0.4322	3.0	2385	0.4102
0.4075	4.0	3180	0.3885
0.3858	5.0	3975	0.3778
0.3741	6.0	4770	0.3692
0.3632	7.0	5565	0.3628
0.3502	8.0	6360	0.3584
0.3412	9.0	7155	0.3551
0.3293	10.0	7950	0.3609
0.3234	11.0	8745	0.3475
0.3112	12.0	9540	0.3456
0.3067	13.0	10335	0.3451
0.3011	14.0	11130	0.3437
0.296	15.0	11925	0.3457
0.2856	16.0	12720	0.3424
0.2837	17.0	13515	0.3402
0.2789	18.0	14310	0.3404
0.2758	19.0	15105	0.3401
0.2702	20.0	15900	0.3393
0.2708	21.0	16695	0.3439
0.2628	22.0	17490	0.3378
0.2625	23.0	18285	0.3432
0.259	24.0	19080	0.3409
0.2572	25.0	19875	0.3431

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0