metadata

base_model: dccuchile/bert-base-spanish-wwm-cased
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: ABL_trad_2a
    results: []

ABL_trad_2a

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: 3.5485
Accuracy: 0.6917
F1: 0.6904

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: 6
eval_batch_size: 6
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 78

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.9204	1.0	2000	0.8942	0.5825	0.5814
0.8116	2.0	4000	0.8397	0.615	0.6112
0.7443	3.0	6000	0.8179	0.635	0.6340
0.7074	4.0	8000	0.8130	0.65	0.6475
0.6663	5.0	10000	0.8134	0.6625	0.6590
0.6385	6.0	12000	0.8092	0.6708	0.6694
0.5702	7.0	14000	0.8355	0.6708	0.6699
0.5436	8.0	16000	0.8662	0.6758	0.6728
0.5002	9.0	18000	0.9040	0.6825	0.6804
0.4602	10.0	20000	0.9345	0.6967	0.6948
0.4324	11.0	22000	1.0021	0.6858	0.6842
0.3862	12.0	24000	1.0713	0.6917	0.6899
0.3586	13.0	26000	1.1688	0.7017	0.6989
0.3536	14.0	28000	1.2699	0.695	0.6923
0.3041	15.0	30000	1.4034	0.6917	0.6911
0.2851	16.0	32000	1.5512	0.6917	0.6897
0.3071	17.0	34000	1.6367	0.69	0.6885
0.2586	18.0	36000	1.7689	0.6892	0.6880
0.2192	19.0	38000	1.9568	0.6842	0.6828
0.2418	20.0	40000	2.0536	0.6808	0.6770
0.2066	21.0	42000	2.1762	0.6917	0.6899
0.1449	22.0	44000	2.3044	0.69	0.6879
0.1669	23.0	46000	2.3854	0.6908	0.6891
0.1376	24.0	48000	2.5111	0.6883	0.6849
0.1434	25.0	50000	2.5862	0.685	0.6829
0.1123	26.0	52000	2.6845	0.685	0.6825
0.0949	27.0	54000	2.7896	0.6825	0.6804
0.1264	28.0	56000	2.8471	0.6908	0.6887
0.0774	29.0	58000	2.8967	0.6883	0.6860
0.1046	30.0	60000	2.9571	0.6867	0.6837
0.0967	31.0	62000	2.9687	0.6892	0.6858
0.0689	32.0	64000	3.0554	0.6917	0.6887
0.069	33.0	66000	3.0982	0.6917	0.6888
0.044	34.0	68000	3.1798	0.6917	0.6898
0.0654	35.0	70000	3.2407	0.685	0.6821
0.0438	36.0	72000	3.1972	0.6908	0.6881
0.0553	37.0	74000	3.2033	0.6967	0.6947
0.0534	38.0	76000	3.2928	0.6883	0.6856
0.0568	39.0	78000	3.3607	0.6933	0.6893
0.03	40.0	80000	3.2983	0.6867	0.6836
0.0539	41.0	82000	3.2896	0.6908	0.6870
0.0459	42.0	84000	3.3401	0.6883	0.6855
0.0322	43.0	86000	3.3879	0.6883	0.6847
0.0251	44.0	88000	3.3517	0.6875	0.6855
0.063	45.0	90000	3.3609	0.6875	0.6856
0.0343	46.0	92000	3.4785	0.6825	0.6801
0.052	47.0	94000	3.4663	0.6817	0.6794
0.038	48.0	96000	3.4157	0.6908	0.6889
0.0312	49.0	98000	3.3996	0.6958	0.6933
0.0469	50.0	100000	3.3978	0.6942	0.6918
0.0405	51.0	102000	3.4383	0.6833	0.6801
0.0438	52.0	104000	3.3833	0.7025	0.7004
0.0366	53.0	106000	3.5241	0.6917	0.6879
0.0415	54.0	108000	3.4236	0.6992	0.6972
0.0315	55.0	110000	3.3053	0.7033	0.7009
0.041	56.0	112000	3.4287	0.6975	0.6961
0.024	57.0	114000	3.4783	0.695	0.6919
0.0263	58.0	116000	3.5307	0.6942	0.6917
0.0231	59.0	118000	3.4495	0.6908	0.6883
0.0269	60.0	120000	3.4664	0.6925	0.6907
0.0152	61.0	122000	3.4655	0.6917	0.6893
0.025	62.0	124000	3.4954	0.6967	0.6957
0.0313	63.0	126000	3.4727	0.6967	0.6942
0.0175	64.0	128000	3.5688	0.6908	0.6892
0.0319	65.0	130000	3.4812	0.6975	0.6963
0.0382	66.0	132000	3.5716	0.6975	0.6942
0.0168	67.0	134000	3.5241	0.7008	0.6978
0.0351	68.0	136000	3.5020	0.6917	0.6889
0.0185	69.0	138000	3.4793	0.6908	0.6894
0.0264	70.0	140000	3.5716	0.6883	0.6847
0.0236	71.0	142000	3.5403	0.69	0.6861
0.0238	72.0	144000	3.5622	0.6917	0.6886
0.0163	73.0	146000	3.4998	0.7008	0.6984
0.0229	74.0	148000	3.5659	0.6925	0.6898
0.0168	75.0	150000	3.5080	0.6983	0.6958
0.0149	76.0	152000	3.4678	0.7008	0.6985
0.0175	77.0	154000	3.5733	0.6967	0.6939
0.0118	78.0	156000	3.5485	0.6917	0.6904

Framework versions

Transformers 4.37.2
Pytorch 2.1.0+cu121
Datasets 2.16.1
Tokenizers 0.15.1