fosh-detector-v2-tmp

This model is a fine-tuned version of google-bert/bert-base-multilingual-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1027
• Accuracy: 0.9684
• Precision: 0.8378
• Recall: 0.8692
• F1: 0.8532

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: 5e-05
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Use OptimizerNames.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: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.3176	0.1305	50	0.2282	0.9150	0.6981	0.3458	0.4625
0.1735	0.2611	100	0.1324	0.9526	0.8242	0.7009	0.7576
0.1358	0.3916	150	0.1283	0.9615	0.8542	0.7664	0.8079
0.1332	0.5222	200	0.1142	0.9664	0.8411	0.8411	0.8411
0.1125	0.6527	250	0.1268	0.9634	0.85	0.7944	0.8213
0.1022	0.7833	300	0.1251	0.9644	0.7983	0.8879	0.8407
0.1107	0.9138	350	0.0972	0.9644	0.8381	0.8224	0.8302
0.1017	1.0444	400	0.1070	0.9654	0.8396	0.8318	0.8357
0.0925	1.1749	450	0.0977	0.9664	0.8288	0.8598	0.8440
0.0834	1.3055	500	0.0867	0.9674	0.8936	0.7850	0.8358
0.0748	1.4360	550	0.0951	0.9704	0.8348	0.8972	0.8649
0.0697	1.5666	600	0.0950	0.9713	0.8611	0.8692	0.8651
0.0818	1.6971	650	0.0871	0.9674	0.8136	0.8972	0.8533
0.073	1.8277	700	0.0813	0.9684	0.8319	0.8785	0.8545
0.0742	1.9582	750	0.0841	0.9713	0.875	0.8505	0.8626
0.0599	2.0888	800	0.0926	0.9713	0.8824	0.8411	0.8612
0.0522	2.2193	850	0.1015	0.9674	0.8426	0.8505	0.8465
0.0581	2.3499	900	0.1000	0.9694	0.88	0.8224	0.8502
0.0562	2.4804	950	0.1066	0.9674	0.8364	0.8598	0.8479
0.0553	2.6110	1000	0.0989	0.9674	0.8190	0.8879	0.8520
0.0546	2.7415	1050	0.0921	0.9694	0.8725	0.8318	0.8517
0.0541	2.8721	1100	0.0920	0.9644	0.8034	0.8785	0.8393
0.0494	3.0026	1150	0.0981	0.9713	0.8611	0.8692	0.8651
0.0358	3.1332	1200	0.1033	0.9713	0.8679	0.8598	0.8638
0.038	3.2637	1250	0.1109	0.9684	0.8378	0.8692	0.8532
0.0408	3.3943	1300	0.0996	0.9684	0.8205	0.8972	0.8571
0.0456	3.5248	1350	0.0976	0.9684	0.8440	0.8598	0.8519
0.0367	3.6554	1400	0.1075	0.9694	0.8455	0.8692	0.8571
0.0332	3.7859	1450	0.1081	0.9684	0.8440	0.8598	0.8519
0.0376	3.9164	1500	0.1027	0.9684	0.8378	0.8692	0.8532

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.3.2
• Tokenizers 0.21.1