FNST_trad_k

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: 3.0507
• Accuracy: 0.6275
• F1: 0.6275

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-06
• 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: 36

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.1468	1.0	1500	1.0961	0.52	0.4256
1.0016	2.0	3000	0.9821	0.5742	0.5705
0.9255	3.0	4500	0.9492	0.5833	0.5865
0.8837	4.0	6000	0.9318	0.5825	0.5837
0.8417	5.0	7500	0.9306	0.5992	0.5978
0.8013	6.0	9000	0.9318	0.6017	0.6045
0.7621	7.0	10500	0.9359	0.605	0.6052
0.7504	8.0	12000	0.9394	0.61	0.6121
0.6951	9.0	13500	0.9744	0.6092	0.6063
0.6763	10.0	15000	0.9820	0.61	0.6099
0.612	11.0	16500	1.0162	0.6158	0.6116
0.593	12.0	18000	1.0400	0.6158	0.6193
0.5561	13.0	19500	1.0735	0.6158	0.6167
0.5342	14.0	21000	1.0789	0.6158	0.6132
0.4931	15.0	22500	1.1443	0.6167	0.6136
0.4758	16.0	24000	1.1832	0.6192	0.6195
0.4346	17.0	25500	1.2587	0.62	0.6196
0.3959	18.0	27000	1.3334	0.6167	0.6178
0.3848	19.0	28500	1.3624	0.6258	0.6245
0.35	20.0	30000	1.4552	0.6233	0.6227
0.3094	21.0	31500	1.5021	0.6208	0.6206
0.3221	22.0	33000	1.6168	0.6242	0.6228
0.2803	23.0	34500	1.6995	0.6225	0.6201
0.2722	24.0	36000	1.8134	0.625	0.6232
0.2355	25.0	37500	1.9296	0.6167	0.6137
0.2285	26.0	39000	2.0198	0.6283	0.6268
0.211	27.0	40500	2.1630	0.6208	0.6220
0.1857	28.0	42000	2.2532	0.6275	0.6244
0.188	29.0	43500	2.4117	0.625	0.6228
0.1787	30.0	45000	2.4971	0.6275	0.6257
0.1687	31.0	46500	2.6493	0.6217	0.6191
0.1534	32.0	48000	2.7295	0.6217	0.6169
0.1606	33.0	49500	2.9021	0.6208	0.6198
0.1537	34.0	51000	2.9315	0.6167	0.6162
0.1284	35.0	52500	3.0047	0.6208	0.6217
0.1359	36.0	54000	3.0507	0.6275	0.6275

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1