README.md

metadata

license: mit
tags:
  - generated_from_trainer
model-index:
  - name: bert_base_tcm_teste
    results: []

bert_base_tcm_teste

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

• Loss: 0.0155
• Criterio Julgamento Precision: 0.7965
• Criterio Julgamento Recall: 0.8654
• Criterio Julgamento F1: 0.8295
• Criterio Julgamento Number: 104
• Data Sessao Precision: 0.7162
• Data Sessao Recall: 0.9636
• Data Sessao F1: 0.8217
• Data Sessao Number: 55
• Modalidade Licitacao Precision: 0.9554
• Modalidade Licitacao Recall: 0.9667
• Modalidade Licitacao F1: 0.9610
• Modalidade Licitacao Number: 421
• Numero Exercicio Precision: 0.9323
• Numero Exercicio Recall: 0.9676
• Numero Exercicio F1: 0.9496
• Numero Exercicio Number: 185
• Objeto Licitacao Precision: 0.5270
• Objeto Licitacao Recall: 0.6610
• Objeto Licitacao F1: 0.5865
• Objeto Licitacao Number: 59
• Valor Objeto Precision: 0.8444
• Valor Objeto Recall: 0.9268
• Valor Objeto F1: 0.8837
• Valor Objeto Number: 41
• Overall Precision: 0.8723
• Overall Recall: 0.9318
• Overall F1: 0.9011
• Overall Accuracy: 0.9966

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-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: 50.0

Training results

Training Loss	Epoch	Step	Validation Loss	Criterio Julgamento Precision	Criterio Julgamento Recall	Criterio Julgamento F1	Criterio Julgamento Number	Data Sessao Precision	Data Sessao Recall	Data Sessao F1	Data Sessao Number	Modalidade Licitacao Precision	Modalidade Licitacao Recall	Modalidade Licitacao F1	Modalidade Licitacao Number	Numero Exercicio Precision	Numero Exercicio Recall	Numero Exercicio F1	Numero Exercicio Number	Objeto Licitacao Precision	Objeto Licitacao Recall	Objeto Licitacao F1	Objeto Licitacao Number	Valor Objeto Precision	Valor Objeto Recall	Valor Objeto F1	Valor Objeto Number	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
0.0193	0.96	2750	0.0190	0.7016	0.8365	0.7632	104	0.6585	0.9818	0.7883	55	0.9446	0.9715	0.9578	421	0.9036	0.9622	0.9319	185	0.2261	0.4407	0.2989	59	0.7	0.8537	0.7692	41	0.7882	0.9121	0.8457	0.9946
0.0165	1.92	5500	0.0133	0.7203	0.8173	0.7658	104	0.675	0.9818	0.8	55	0.9447	0.9739	0.9591	421	0.9430	0.9838	0.9630	185	0.4691	0.6441	0.5429	59	0.8043	0.9024	0.8506	41	0.8466	0.9318	0.8872	0.9964
0.0089	2.88	8250	0.0150	0.7636	0.8077	0.7850	104	0.7895	0.8182	0.8036	55	0.9491	0.9739	0.9613	421	0.9282	0.9784	0.9526	185	0.4444	0.6102	0.5143	59	0.8636	0.9268	0.8941	41	0.8640	0.9179	0.8901	0.9965
0.0066	3.84	11000	0.0150	0.7692	0.8654	0.8145	104	0.7333	0.8	0.7652	55	0.9464	0.9644	0.9553	421	0.9278	0.9730	0.9499	185	0.5	0.6780	0.5755	59	0.7708	0.9024	0.8315	41	0.8588	0.9214	0.8890	0.9966
0.0055	4.8	13750	0.0176	0.75	0.8654	0.8036	104	0.7903	0.8909	0.8376	55	0.9490	0.9715	0.9601	421	0.9326	0.9730	0.9524	185	0.4568	0.6271	0.5286	59	0.7872	0.9024	0.8409	41	0.8587	0.9272	0.8916	0.9963
0.0066	5.76	16500	0.0155	0.7965	0.8654	0.8295	104	0.7162	0.9636	0.8217	55	0.9554	0.9667	0.9610	421	0.9323	0.9676	0.9496	185	0.5270	0.6610	0.5865	59	0.8444	0.9268	0.8837	41	0.8723	0.9318	0.9011	0.9966
0.0031	6.72	19250	0.0181	0.775	0.8942	0.8304	104	0.7879	0.9455	0.8595	55	0.9533	0.9691	0.9611	421	0.9326	0.9730	0.9524	185	0.4875	0.6610	0.5612	59	0.8261	0.9268	0.8736	41	0.8682	0.9364	0.9010	0.9965
0.0066	7.68	22000	0.0192	0.7798	0.8173	0.7981	104	0.6986	0.9273	0.7969	55	0.9353	0.9620	0.9485	421	0.8995	0.9676	0.9323	185	0.4	0.5763	0.4722	59	0.7551	0.9024	0.8222	41	0.8344	0.9145	0.8726	0.9961
0.0052	8.64	24750	0.0201	0.8036	0.8654	0.8333	104	0.7869	0.8727	0.8276	55	0.9465	0.9667	0.9565	421	0.9326	0.9730	0.9524	185	0.5060	0.7119	0.5915	59	0.8043	0.9024	0.8506	41	0.8692	0.9295	0.8983	0.9966
0.0015	9.61	27500	0.0202	0.7838	0.8365	0.8093	104	0.7313	0.8909	0.8033	55	0.9482	0.9572	0.9527	421	0.9326	0.9730	0.9524	185	0.4865	0.6102	0.5414	59	0.8043	0.9024	0.8506	41	0.8646	0.9156	0.8894	0.9966
0.0015	10.57	30250	0.0225	0.7798	0.8173	0.7981	104	0.6912	0.8545	0.7642	55	0.9508	0.9644	0.9575	421	0.9375	0.9730	0.9549	185	0.5395	0.6949	0.6074	59	0.8478	0.9512	0.8966	41	0.8693	0.9225	0.8951	0.9964

Framework versions

• Transformers 4.21.0.dev0
• Pytorch 1.11.0+cu113
• Datasets 2.3.2
• Tokenizers 0.12.1