roberta-base-pp-1000000-1e-06-128-negcommonsensebalanced-1e-06-256

This model is a fine-tuned version of mhr2004/roberta-base-pp-1000000-1e-06-128 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4009

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: 256
• 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
0.5718	1.0	795	0.5297
0.5233	2.0	1590	0.4949
0.4986	3.0	2385	0.4810
0.4843	4.0	3180	0.4702
0.4699	5.0	3975	0.4565
0.4601	6.0	4770	0.4506
0.454	7.0	5565	0.4388
0.4446	8.0	6360	0.4368
0.4353	9.0	7155	0.4323
0.4274	10.0	7950	0.4255
0.4219	11.0	8745	0.4221
0.4178	12.0	9540	0.4173
0.4128	13.0	10335	0.4210
0.4099	14.0	11130	0.4135
0.4071	15.0	11925	0.4088
0.3983	16.0	12720	0.4081
0.3979	17.0	13515	0.4089
0.3961	18.0	14310	0.4083
0.3946	19.0	15105	0.4054
0.3899	20.0	15900	0.4030
0.3907	21.0	16695	0.4034
0.3831	22.0	17490	0.4006
0.3846	23.0	18285	0.4018
0.385	24.0	19080	0.4023
0.3796	25.0	19875	0.4009

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0