santipvz/bert_spanish_fine_tuned

bert-base-spanish-wwm-cased-ner

This model is a fine-tuned version of dccuchile/bert-base-spanish-wwm-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0974
• Precision: 0.8664
• Recall: 0.8784
• F1: 0.8724
• Accuracy: 0.9807

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: 32
• eval_batch_size: 8
• 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: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.1848	0.3831	100	0.1168	0.7770	0.8227	0.7992	0.9715
0.0688	0.7663	200	0.0903	0.8460	0.8448	0.8454	0.9769
0.048	1.1494	300	0.0943	0.8374	0.8557	0.8465	0.9776
0.0363	1.5326	400	0.0826	0.8508	0.8718	0.8612	0.9801
0.0373	1.9157	500	0.0906	0.8375	0.8455	0.8415	0.9776
0.0243	2.2989	600	0.0940	0.8523	0.8632	0.8577	0.9788
0.0253	2.6820	700	0.1008	0.8364	0.8541	0.8452	0.9781
0.0204	3.0651	800	0.0974	0.8664	0.8784	0.8724	0.9807
0.0138	3.4483	900	0.1045	0.8322	0.8623	0.8470	0.9778
0.0154	3.8314	1000	0.1020	0.8628	0.8763	0.8695	0.9804
0.0139	4.2146	1100	0.1107	0.8547	0.8681	0.8614	0.9786
0.0114	4.5977	1200	0.1068	0.8579	0.8639	0.8609	0.9795
0.0097	4.9808	1300	0.1167	0.8560	0.8688	0.8624	0.9794
0.0079	5.3640	1400	0.1203	0.8462	0.8630	0.8545	0.9784
0.0069	5.7471	1500	0.1177	0.8511	0.8606	0.8559	0.9785
0.007	6.1303	1600	0.1348	0.8445	0.8529	0.8487	0.9780
0.0056	6.5134	1700	0.1215	0.8519	0.8592	0.8556	0.9788
0.0057	6.8966	1800	0.1212	0.8619	0.8739	0.8679	0.9800
0.0047	7.2797	1900	0.1290	0.8604	0.8702	0.8653	0.9801
0.0037	7.6628	2000	0.1461	0.8448	0.8578	0.8513	0.9784
0.0048	8.0460	2100	0.1371	0.8632	0.8735	0.8683	0.9798
0.0034	8.4291	2200	0.1419	0.8483	0.8655	0.8568	0.9788
0.004	8.8123	2300	0.1397	0.8655	0.8758	0.8706	0.9801
0.0029	9.1954	2400	0.1518	0.8506	0.8597	0.8551	0.9784
0.0034	9.5785	2500	0.1542	0.8602	0.8707	0.8654	0.9795
0.0029	9.9617	2600	0.1465	0.8596	0.8693	0.8644	0.9795
0.0032	10.3448	2700	0.1542	0.8495	0.8574	0.8534	0.9790
0.0031	10.7280	2800	0.1392	0.8665	0.8725	0.8695	0.9800
0.0025	11.1111	2900	0.1378	0.8652	0.8735	0.8693	0.9801
0.0016	11.4943	3000	0.1585	0.8535	0.8690	0.8612	0.9794
0.0026	11.8774	3100	0.1565	0.8472	0.8658	0.8564	0.9793
0.0026	12.2605	3200	0.1533	0.8576	0.8700	0.8637	0.9796
0.0021	12.6437	3300	0.1588	0.8571	0.8653	0.8612	0.9793
0.0024	13.0268	3400	0.1470	0.8613	0.8711	0.8662	0.9804
0.0018	13.4100	3500	0.1584	0.8567	0.8683	0.8625	0.9795
0.002	13.7931	3600	0.1622	0.8521	0.8700	0.8609	0.9792
0.0018	14.1762	3700	0.1624	0.8550	0.8641	0.8595	0.9792
0.0014	14.5594	3800	0.1677	0.8617	0.8611	0.8614	0.9791
0.0022	14.9425	3900	0.1622	0.8525	0.8634	0.8579	0.9789
0.0011	15.3257	4000	0.1666	0.8549	0.8625	0.8587	0.9789
0.0014	15.7088	4100	0.1640	0.8600	0.8702	0.8651	0.9795
0.0012	16.0920	4200	0.1639	0.8598	0.8730	0.8663	0.9798
0.0011	16.4751	4300	0.1645	0.8634	0.8746	0.8690	0.9801
0.0013	16.8582	4400	0.1676	0.8577	0.8681	0.8629	0.9793
0.0017	17.2414	4500	0.1741	0.8554	0.8616	0.8585	0.9788
0.0014	17.6245	4600	0.1683	0.8578	0.8690	0.8634	0.9797
0.001	18.0077	4700	0.1718	0.8556	0.8655	0.8605	0.9792
0.0009	18.3908	4800	0.1675	0.8595	0.8681	0.8638	0.9795
0.0009	18.7739	4900	0.1702	0.8597	0.8683	0.8640	0.9796
0.0011	19.1571	5000	0.1706	0.8616	0.8704	0.8660	0.9798
0.0008	19.5402	5100	0.1706	0.8590	0.8686	0.8637	0.9797
0.0009	19.9234	5200	0.1704	0.8598	0.8690	0.8644	0.9797

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1