roberta-mediawatch-claims-el

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

• Loss: 0.5323
• Accuracy: 0.7182
• Precision: 0.5003
• Recall: 0.7806
• F1: 0.6098

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.6019	0.2115	500	0.5692	0.6953	0.4746	0.7222	0.5728
0.5599	0.4230	1000	0.5541	0.6846	0.4655	0.7786	0.5827
0.5501	0.6345	1500	0.5470	0.6894	0.4708	0.7938	0.5910
0.5633	0.8460	2000	0.5412	0.6756	0.4592	0.8283	0.5908
0.4963	1.0575	2500	0.5506	0.7117	0.4939	0.7862	0.6067
0.4945	1.2690	3000	0.5639	0.7501	0.5477	0.6675	0.6017
0.4993	1.4805	3500	0.5400	0.7077	0.4896	0.7929	0.6054
0.4928	1.6920	4000	0.5645	0.6815	0.4647	0.8300	0.5958
0.4929	1.9036	4500	0.5313	0.7220	0.5055	0.7727	0.6112
0.3425	2.1151	5000	0.6383	0.7324	0.5203	0.6911	0.5936
0.3471	2.3266	5500	0.6375	0.7282	0.5136	0.7332	0.6040
0.3616	2.5381	6000	0.6713	0.7255	0.5107	0.7054	0.5924
0.3437	2.7496	6500	0.7096	0.7427	0.5374	0.6465	0.5869
0.3774	2.9611	7000	0.6173	0.7129	0.4949	0.7306	0.5901
0.2218	3.1726	7500	0.9221	0.7232	0.5079	0.6785	0.5809
0.2397	3.3841	8000	0.8595	0.7222	0.5064	0.6995	0.5875
0.2316	3.5956	8500	0.8502	0.7255	0.5111	0.6785	0.5830
0.2359	3.8071	9000	0.8198	0.7248	0.5099	0.6936	0.5877
0.1983	4.0186	9500	0.9416	0.7267	0.5131	0.6574	0.5764
0.1519	4.2301	10000	1.0557	0.7258	0.5116	0.6692	0.5799
0.1607	4.4416	10500	1.1513	0.7360	0.5281	0.6254	0.5726
0.1476	4.6531	11000	1.0914	0.7310	0.5188	0.6726	0.5858
0.1734	4.8646	11500	1.1049	0.7294	0.5168	0.6608	0.5800

Framework versions

• Transformers 4.57.6
• Pytorch 2.10.0+cu130
• Datasets 4.5.0
• Tokenizers 0.22.2