bart-base

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

• Loss: 1.7146
• Accuracy: 0.8180
• Precision: 0.8189
• Recall: 0.8180
• Precision Macro: 0.7608
• Recall Macro: 0.7799
• Macro Fpr: 0.0157
• Weighted Fpr: 0.0151
• Weighted Specificity: 0.9781
• Macro Specificity: 0.9868
• Weighted Sensitivity: 0.8234
• Macro Sensitivity: 0.7799
• F1 Micro: 0.8234
• F1 Macro: 0.7642
• F1 Weighted: 0.8237

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: 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: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	Precision Macro	Recall Macro	Macro Fpr	Weighted Fpr	Weighted Specificity	Macro Specificity	Weighted Sensitivity	Macro Sensitivity	F1 Micro	F1 Macro	F1 Weighted
1.2028	1.0	643	0.8430	0.7599	0.7601	0.7599	0.6004	0.6367	0.0232	0.0221	0.9655	0.9817	0.7599	0.6367	0.7599	0.6064	0.7489
0.715	2.0	1286	0.7332	0.7932	0.8020	0.7932	0.7386	0.7321	0.0190	0.0183	0.9745	0.9845	0.7932	0.7321	0.7932	0.7214	0.7853
0.578	3.0	1929	0.8045	0.7940	0.8075	0.7940	0.7231	0.7069	0.0185	0.0182	0.9775	0.9848	0.7940	0.7069	0.7940	0.6998	0.7901
0.3938	4.0	2572	0.8291	0.8156	0.8171	0.8156	0.7937	0.7218	0.0169	0.0159	0.9711	0.9858	0.8156	0.7218	0.8156	0.7369	0.8105
0.3238	5.0	3215	0.8889	0.7940	0.8146	0.7940	0.7464	0.7515	0.0188	0.0182	0.9762	0.9847	0.7940	0.7515	0.7940	0.7361	0.7995
0.246	6.0	3858	1.1629	0.7955	0.8067	0.7955	0.7483	0.7600	0.0186	0.0180	0.9749	0.9847	0.7955	0.7600	0.7955	0.7362	0.7946
0.1791	7.0	4501	1.1354	0.8180	0.8151	0.8180	0.7832	0.7697	0.0165	0.0156	0.9747	0.9862	0.8180	0.7697	0.8180	0.7736	0.8147
0.1305	8.0	5144	1.2825	0.8110	0.8148	0.8110	0.7422	0.7489	0.0169	0.0164	0.9765	0.9858	0.8110	0.7489	0.8110	0.7369	0.8088
0.0924	9.0	5787	1.4217	0.8040	0.8114	0.8040	0.7465	0.7809	0.0178	0.0171	0.9762	0.9853	0.8040	0.7809	0.8040	0.7560	0.8015
0.0953	10.0	6430	1.5552	0.8025	0.8056	0.8025	0.7702	0.7822	0.0183	0.0173	0.9712	0.9849	0.8025	0.7822	0.8025	0.7661	0.8001
0.0617	11.0	7073	1.5224	0.8040	0.8144	0.8040	0.7457	0.7512	0.0176	0.0171	0.9762	0.9853	0.8040	0.7512	0.8040	0.7422	0.8070
0.0582	12.0	7716	1.6428	0.7971	0.8148	0.7971	0.7470	0.7655	0.0183	0.0179	0.9771	0.9849	0.7971	0.7655	0.7971	0.7465	0.8022
0.0511	13.0	8359	1.4952	0.8195	0.8208	0.8195	0.7645	0.7580	0.0162	0.0155	0.9759	0.9864	0.8195	0.7580	0.8195	0.7586	0.8187
0.0476	14.0	9002	1.7132	0.7971	0.7958	0.7971	0.7637	0.7328	0.0189	0.0179	0.9708	0.9845	0.7971	0.7328	0.7971	0.7417	0.7913
0.0375	15.0	9645	1.7058	0.8002	0.8110	0.8002	0.7369	0.7696	0.0182	0.0175	0.9757	0.9851	0.8002	0.7696	0.8002	0.7437	0.8017
0.0241	16.0	10288	1.7146	0.8180	0.8189	0.8180	0.7852	0.7787	0.0162	0.0156	0.9761	0.9863	0.8180	0.7787	0.8180	0.7780	0.8174
0.0226	17.0	10931	1.7035	0.8203	0.8238	0.8203	0.7732	0.7781	0.0160	0.0154	0.9774	0.9865	0.8203	0.7781	0.8203	0.7714	0.8206
0.0189	18.0	11574	1.8079	0.8164	0.8160	0.8164	0.7583	0.7677	0.0166	0.0158	0.9749	0.9861	0.8164	0.7677	0.8164	0.7578	0.8149
0.026	19.0	12217	1.8187	0.8125	0.8170	0.8125	0.7675	0.7833	0.0169	0.0162	0.9748	0.9858	0.8125	0.7833	0.8125	0.7719	0.8138
0.0101	20.0	12860	1.8354	0.8187	0.8220	0.8187	0.7748	0.7818	0.0161	0.0156	0.9772	0.9864	0.8187	0.7818	0.8187	0.7710	0.8180
0.0216	21.0	13503	1.8372	0.8156	0.8219	0.8156	0.7502	0.7858	0.0163	0.0159	0.9789	0.9863	0.8156	0.7858	0.8156	0.7618	0.8164
0.0138	22.0	14146	1.8472	0.8203	0.8263	0.8203	0.7613	0.7796	0.0159	0.0154	0.9786	0.9866	0.8203	0.7796	0.8203	0.7662	0.8222
0.0169	23.0	14789	1.8104	0.8218	0.8252	0.8218	0.7719	0.7595	0.0160	0.0152	0.9749	0.9865	0.8218	0.7595	0.8218	0.7607	0.8209
0.0079	24.0	15432	1.9253	0.8110	0.8202	0.8110	0.7622	0.7576	0.0171	0.0164	0.9759	0.9858	0.8110	0.7576	0.8110	0.7524	0.8123
0.0017	25.0	16075	1.9111	0.8156	0.8193	0.8156	0.7554	0.7742	0.0164	0.0159	0.9775	0.9862	0.8156	0.7742	0.8156	0.7594	0.8155
0.0071	26.0	16718	1.8809	0.8187	0.8244	0.8187	0.7595	0.7749	0.0161	0.0156	0.9783	0.9865	0.8187	0.7749	0.8187	0.7601	0.8199
0.0032	27.0	17361	1.8246	0.8273	0.8333	0.8273	0.7727	0.7807	0.0152	0.0147	0.9786	0.9871	0.8273	0.7807	0.8273	0.7718	0.8289
0.0014	28.0	18004	1.8354	0.8265	0.8337	0.8265	0.7624	0.7806	0.0154	0.0148	0.9784	0.9870	0.8265	0.7806	0.8265	0.7648	0.8282
0.0004	29.0	18647	1.8558	0.8234	0.8277	0.8234	0.7616	0.7801	0.0157	0.0151	0.9778	0.9867	0.8234	0.7801	0.8234	0.7646	0.8234
0.0012	30.0	19290	1.8392	0.8234	0.8281	0.8234	0.7608	0.7799	0.0157	0.0151	0.9781	0.9868	0.8234	0.7799	0.8234	0.7642	0.8237

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.1