ner-fine-tune-roberta

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

• Loss: 0.3622
• Precision: 0.2958
• Recall: 0.2972
• F1: 0.2965
• Accuracy: 0.9462

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-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	105	0.2178	0.2884	0.3278	0.3068	0.9468
No log	2.0	210	0.2681	0.2599	0.2642	0.2620	0.9448
No log	3.0	315	0.2576	0.2824	0.2877	0.2850	0.9475
No log	4.0	420	0.2884	0.2608	0.2571	0.2589	0.9467
0.0235	5.0	525	0.2929	0.2565	0.3019	0.2774	0.9432
0.0235	6.0	630	0.2993	0.3127	0.2618	0.2850	0.9528
0.0235	7.0	735	0.3014	0.2792	0.3160	0.2965	0.9449
0.0235	8.0	840	0.3349	0.2671	0.3042	0.2845	0.9426
0.0235	9.0	945	0.3303	0.2930	0.3373	0.3136	0.9455
0.0083	10.0	1050	0.3573	0.3047	0.2759	0.2896	0.9479
0.0083	11.0	1155	0.3500	0.2729	0.2665	0.2697	0.9464
0.0083	12.0	1260	0.3626	0.2947	0.2995	0.2971	0.9462
0.0083	13.0	1365	0.3522	0.2954	0.2877	0.2915	0.9466
0.0083	14.0	1470	0.3610	0.2930	0.2972	0.2951	0.9462
0.0042	15.0	1575	0.3622	0.2958	0.2972	0.2965	0.9462

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1