ner-fine-tune-roberta-more-data

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

• Loss: 0.3443
• Precision: 0.2432
• Recall: 0.3373
• F1: 0.2826
• Accuracy: 0.9352

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	120	0.2270	0.0	0.0	0.0	0.9541
No log	2.0	240	0.1783	0.1801	0.0684	0.0991	0.9582
No log	3.0	360	0.1810	0.1023	0.0637	0.0785	0.9519
No log	4.0	480	0.1642	0.2227	0.2358	0.2291	0.9518
0.2437	5.0	600	0.1728	0.1880	0.2429	0.2119	0.9401
0.2437	6.0	720	0.1997	0.1536	0.1934	0.1712	0.9344
0.2437	7.0	840	0.2637	0.1827	0.3231	0.2334	0.9237
0.2437	8.0	960	0.2564	0.1730	0.2476	0.2037	0.9320
0.0746	9.0	1080	0.2437	0.2116	0.3184	0.2542	0.9353
0.0746	10.0	1200	0.2524	0.2340	0.3443	0.2786	0.9347
0.0746	11.0	1320	0.2636	0.2071	0.2618	0.2312	0.9373
0.0746	12.0	1440	0.2562	0.2434	0.3255	0.2785	0.9389
0.0309	13.0	1560	0.2793	0.2263	0.3042	0.2596	0.9371
0.0309	14.0	1680	0.2441	0.3455	0.3137	0.3288	0.9586
0.0309	15.0	1800	0.3174	0.2123	0.3090	0.2517	0.9324
0.0309	16.0	1920	0.2784	0.2374	0.2877	0.2601	0.9393
0.0176	17.0	2040	0.2740	0.2758	0.3090	0.2914	0.9461
0.0176	18.0	2160	0.3077	0.2319	0.3467	0.2779	0.9344
0.0176	19.0	2280	0.3088	0.2380	0.3160	0.2715	0.9388
0.0176	20.0	2400	0.2848	0.2613	0.3278	0.2908	0.9414
0.0112	21.0	2520	0.2958	0.2453	0.3420	0.2857	0.9369
0.0112	22.0	2640	0.3089	0.2295	0.3632	0.2813	0.9331
0.0112	23.0	2760	0.3435	0.2359	0.375	0.2896	0.9330
0.0112	24.0	2880	0.3303	0.2434	0.3467	0.2860	0.9366
0.0076	25.0	3000	0.3237	0.2363	0.3160	0.2704	0.9383
0.0076	26.0	3120	0.3235	0.2451	0.3278	0.2805	0.9384
0.0076	27.0	3240	0.3409	0.2491	0.3302	0.2840	0.9361
0.0076	28.0	3360	0.3446	0.2416	0.3373	0.2815	0.9351
0.0076	29.0	3480	0.3470	0.2417	0.3278	0.2783	0.9355
0.0055	30.0	3600	0.3443	0.2432	0.3373	0.2826	0.9352

Framework versions

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