DeepSeek-R1-Distill-Qwen-1.5B-2-contract-sections-classification-v4-50

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9474
• Accuracy Evaluate: 0.7023
• Precision Evaluate: 0.7283
• Recall Evaluate: 0.7032
• F1 Evaluate: 0.7003
• Accuracy Sklearn: 0.7023
• Precision Sklearn: 0.7252
• Recall Sklearn: 0.7023
• F1 Sklearn: 0.6989
• Acuracia Rotulo Objeto: 0.8554
• Acuracia Rotulo Obrigacoes: 0.7795
• Acuracia Rotulo Valor: 0.6017
• Acuracia Rotulo Vigencia: 0.5932
• Acuracia Rotulo Rescisao: 0.6676
• Acuracia Rotulo Foro: 0.9654
• Acuracia Rotulo Reajuste: 0.4306
• Acuracia Rotulo Fiscalizacao: 0.6435
• Acuracia Rotulo Publicacao: 0.8177
• Acuracia Rotulo Pagamento: 0.4457
• Acuracia Rotulo Casos Omissos: 0.8276
• Acuracia Rotulo Sancoes: 0.7339
• Acuracia Rotulo Dotacao Orcamentaria: 0.7802

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: 16
• eval_batch_size: 16
• 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: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy Evaluate	Precision Evaluate	Recall Evaluate	F1 Evaluate	Accuracy Sklearn	Precision Sklearn	Recall Sklearn	F1 Sklearn	Acuracia Rotulo Objeto	Acuracia Rotulo Obrigacoes	Acuracia Rotulo Valor	Acuracia Rotulo Vigencia	Acuracia Rotulo Rescisao	Acuracia Rotulo Foro	Acuracia Rotulo Reajuste	Acuracia Rotulo Fiscalizacao	Acuracia Rotulo Publicacao	Acuracia Rotulo Pagamento	Acuracia Rotulo Casos Omissos	Acuracia Rotulo Sancoes	Acuracia Rotulo Dotacao Orcamentaria
3.9512	1.0	1000	3.6400	0.0653	0.1038	0.0796	0.0507	0.0653	0.1226	0.0653	0.0461	0.0021	0.0067	0.0143	0.1129	0.1219	0.3077	0.0036	0.0	0.0739	0.0072	0.0	0.0550	0.3297
3.3304	2.0	2000	3.1552	0.0877	0.1109	0.1027	0.0684	0.0877	0.1335	0.0877	0.0645	0.0145	0.0202	0.0659	0.1627	0.1219	0.3077	0.0036	0.0	0.4236	0.0072	0.0148	0.0550	0.1374
2.7872	3.0	3000	2.7514	0.1517	0.1497	0.1604	0.1408	0.1517	0.1648	0.1517	0.1429	0.0785	0.1414	0.2321	0.1680	0.1330	0.2923	0.0071	0.0946	0.5123	0.0181	0.2660	0.0642	0.0769
2.4261	4.0	4000	2.5576	0.2005	0.1879	0.1996	0.1902	0.2005	0.1960	0.2005	0.1943	0.2273	0.2677	0.2350	0.1890	0.0997	0.3192	0.0214	0.1104	0.4187	0.0181	0.5320	0.1009	0.0549
2.2045	5.0	5000	2.4390	0.2325	0.2250	0.2249	0.2204	0.2325	0.2313	0.2325	0.2267	0.3037	0.3367	0.2350	0.2310	0.1330	0.3231	0.0427	0.1104	0.4286	0.0254	0.5714	0.1284	0.0549
2.0064	6.0	6000	2.3316	0.2695	0.2628	0.2597	0.2554	0.2695	0.2714	0.2695	0.2637	0.4008	0.3468	0.2407	0.2336	0.1690	0.3577	0.2206	0.1167	0.4877	0.0507	0.5862	0.1376	0.0275
1.8509	7.0	7000	2.2336	0.2933	0.2896	0.2810	0.2782	0.2933	0.2989	0.2933	0.2876	0.4690	0.3586	0.2436	0.2598	0.1745	0.3962	0.2384	0.1451	0.4828	0.0797	0.6404	0.1376	0.0275
1.7072	8.0	8000	2.1432	0.321	0.3267	0.3075	0.3084	0.321	0.3354	0.321	0.3181	0.5186	0.3653	0.2493	0.2625	0.1911	0.4192	0.2847	0.2587	0.4926	0.1196	0.6650	0.1376	0.0330
1.6043	9.0	9000	2.0637	0.3443	0.3542	0.3288	0.3306	0.3443	0.3637	0.3443	0.3414	0.5702	0.3822	0.2521	0.2808	0.2022	0.4615	0.3132	0.3123	0.5172	0.1341	0.6601	0.1560	0.0330
1.4611	10.0	10000	1.9872	0.3822	0.3950	0.3639	0.3670	0.3822	0.4059	0.3822	0.3803	0.5971	0.4242	0.2751	0.3386	0.2133	0.5154	0.3203	0.4322	0.5369	0.1993	0.6601	0.1743	0.0440
1.3565	11.0	11000	1.9195	0.4037	0.4203	0.3862	0.3889	0.4037	0.4298	0.4037	0.4018	0.6116	0.4360	0.2951	0.3491	0.2188	0.5615	0.3310	0.5047	0.5567	0.2428	0.6601	0.1927	0.0604
1.2776	12.0	12000	1.8541	0.4235	0.4395	0.4056	0.4077	0.4235	0.4499	0.4235	0.4220	0.6281	0.4663	0.3209	0.3596	0.2410	0.5962	0.3416	0.5205	0.5665	0.2681	0.6601	0.2385	0.0659
1.2015	13.0	13000	1.7973	0.4415	0.4623	0.4297	0.4304	0.4415	0.4723	0.4415	0.4417	0.6384	0.4848	0.3524	0.3885	0.2604	0.6192	0.3452	0.4353	0.5714	0.2862	0.7931	0.3119	0.0989
1.0948	14.0	14000	1.7435	0.465	0.4844	0.4533	0.4516	0.465	0.4969	0.465	0.4648	0.6653	0.5253	0.3954	0.3963	0.2825	0.6731	0.3488	0.4479	0.5862	0.2645	0.7980	0.3670	0.1429
1.0335	15.0	15000	1.6933	0.4788	0.4981	0.4701	0.4663	0.4788	0.5114	0.4788	0.4789	0.6880	0.5269	0.4069	0.4042	0.2936	0.6923	0.3488	0.4511	0.5911	0.3007	0.8030	0.4128	0.1923
0.9816	16.0	16000	1.6456	0.4968	0.5181	0.4918	0.4848	0.4968	0.5318	0.4968	0.4966	0.7004	0.5455	0.4298	0.4094	0.3102	0.7115	0.3488	0.4511	0.7044	0.3043	0.8030	0.4495	0.2253
0.9221	17.0	17000	1.5969	0.5142	0.5474	0.5091	0.5038	0.5142	0.5601	0.5142	0.5150	0.7149	0.5875	0.4470	0.4173	0.3324	0.7538	0.3523	0.4637	0.7143	0.2717	0.7980	0.4954	0.2692
0.8315	18.0	18000	1.5564	0.5325	0.5665	0.5267	0.5212	0.5325	0.5794	0.5325	0.5332	0.7335	0.6145	0.4527	0.4199	0.4017	0.7692	0.3523	0.4763	0.7192	0.2681	0.7980	0.5229	0.3187
0.8019	19.0	19000	1.5088	0.551	0.5805	0.5418	0.5361	0.551	0.5949	0.551	0.5508	0.7335	0.6768	0.4728	0.4278	0.4266	0.7885	0.3523	0.4858	0.7291	0.2790	0.7980	0.5321	0.3407
0.7627	20.0	20000	1.4605	0.5627	0.5898	0.5555	0.5485	0.5627	0.6039	0.5627	0.5621	0.7355	0.6869	0.4814	0.4331	0.4460	0.8269	0.3523	0.4858	0.7389	0.2971	0.8030	0.5505	0.3846
0.6963	21.0	21000	1.4205	0.5835	0.6192	0.5789	0.5722	0.5835	0.6276	0.5835	0.5826	0.75	0.6953	0.4842	0.4357	0.4709	0.8654	0.3523	0.5804	0.7389	0.3261	0.8030	0.6055	0.4176
0.6698	22.0	22000	1.3759	0.5938	0.6262	0.5887	0.5825	0.5938	0.6348	0.5938	0.5931	0.7583	0.7020	0.4986	0.4436	0.5097	0.8692	0.3559	0.5741	0.7389	0.3442	0.8079	0.6055	0.4451
0.6175	23.0	23000	1.3362	0.601	0.6335	0.5976	0.5908	0.601	0.6412	0.601	0.6001	0.7665	0.7088	0.5043	0.4357	0.5263	0.8692	0.3594	0.5804	0.7438	0.3551	0.8079	0.6330	0.4780
0.5775	24.0	24000	1.3042	0.6102	0.6430	0.6066	0.5991	0.6102	0.6513	0.6102	0.6089	0.7810	0.7256	0.5129	0.4357	0.5485	0.8731	0.3594	0.5836	0.7438	0.3587	0.8079	0.6606	0.4945
0.5471	25.0	25000	1.2682	0.619	0.6507	0.6167	0.6089	0.619	0.6586	0.619	0.6177	0.7851	0.7340	0.5186	0.4436	0.5485	0.8769	0.3879	0.5868	0.7635	0.3768	0.8079	0.6881	0.5
0.5339	26.0	26000	1.2367	0.6262	0.6564	0.6243	0.6163	0.6262	0.6637	0.6262	0.6241	0.7996	0.7407	0.5272	0.4462	0.5540	0.8808	0.3879	0.5931	0.7635	0.3804	0.8128	0.6972	0.5330
0.4904	27.0	27000	1.2076	0.6352	0.6632	0.6339	0.6257	0.6352	0.6706	0.6352	0.6329	0.8079	0.7508	0.5301	0.4541	0.5706	0.8846	0.3950	0.5962	0.7783	0.3804	0.8128	0.6972	0.5824
0.4679	28.0	28000	1.1778	0.6405	0.6669	0.6392	0.6313	0.6405	0.6741	0.6405	0.6382	0.8058	0.7525	0.5301	0.4672	0.5762	0.8846	0.4021	0.6088	0.7833	0.4022	0.8177	0.6972	0.5824
0.4567	29.0	29000	1.1544	0.6468	0.6700	0.6461	0.6373	0.6468	0.6772	0.6468	0.6440	0.8140	0.7593	0.5330	0.4698	0.5900	0.8885	0.4057	0.6088	0.7833	0.4022	0.8177	0.7064	0.6209
0.4424	30.0	30000	1.1306	0.652	0.6728	0.6513	0.6422	0.652	0.6806	0.652	0.6490	0.8202	0.7643	0.5330	0.4777	0.5983	0.8962	0.4057	0.6120	0.7833	0.4094	0.8177	0.7064	0.6429
0.409	31.0	31000	1.1112	0.657	0.6785	0.6577	0.6486	0.657	0.6857	0.657	0.6543	0.8244	0.7559	0.5415	0.4751	0.6066	0.9115	0.4093	0.6278	0.7833	0.4203	0.8227	0.7064	0.6648
0.4016	32.0	32000	1.0915	0.664	0.6861	0.6644	0.6564	0.664	0.6922	0.664	0.6614	0.8326	0.7626	0.5559	0.4856	0.6122	0.9115	0.4128	0.6278	0.7833	0.4312	0.8227	0.7064	0.6923
0.3937	33.0	33000	1.0746	0.6685	0.6937	0.6695	0.6620	0.6685	0.6979	0.6685	0.6658	0.8306	0.7694	0.5702	0.4856	0.6122	0.9423	0.4128	0.6246	0.7833	0.4312	0.8227	0.7156	0.7033
0.3765	34.0	34000	1.0577	0.6727	0.6971	0.6744	0.6670	0.6727	0.7010	0.6727	0.6700	0.8388	0.7660	0.5759	0.4934	0.6150	0.9462	0.4164	0.6246	0.7931	0.4348	0.8227	0.7156	0.7253
0.3727	35.0	35000	1.0426	0.6785	0.7022	0.6800	0.6729	0.6785	0.7055	0.6785	0.6756	0.8409	0.7677	0.5903	0.5171	0.6150	0.9538	0.4164	0.6309	0.7931	0.4348	0.8227	0.7156	0.7418
0.3593	36.0	36000	1.0304	0.6803	0.7041	0.6810	0.6743	0.6803	0.7072	0.6803	0.6773	0.8430	0.7694	0.5903	0.5249	0.6177	0.9577	0.4199	0.6309	0.7931	0.4348	0.8227	0.7064	0.7418
0.3532	37.0	37000	1.0170	0.6827	0.7052	0.6838	0.6774	0.6827	0.7075	0.6827	0.6795	0.8450	0.7694	0.5903	0.5276	0.6260	0.9615	0.4199	0.6309	0.7980	0.4384	0.8227	0.7064	0.7527
0.3486	38.0	38000	1.0061	0.685	0.7071	0.6857	0.6801	0.685	0.7082	0.685	0.6814	0.8450	0.7727	0.5931	0.5354	0.6316	0.9615	0.4199	0.6309	0.7980	0.4384	0.8227	0.7064	0.7582
0.3383	39.0	39000	0.9962	0.6863	0.7083	0.6869	0.6812	0.6863	0.7099	0.6863	0.6829	0.8471	0.7710	0.5931	0.5354	0.6399	0.9615	0.4235	0.6309	0.7980	0.4420	0.8227	0.7064	0.7582
0.3375	40.0	40000	0.9883	0.6885	0.7114	0.6887	0.6837	0.6885	0.7123	0.6885	0.6851	0.8512	0.7727	0.5931	0.5381	0.6482	0.9615	0.4235	0.6341	0.7980	0.4457	0.8227	0.7064	0.7582
0.3105	41.0	41000	0.9787	0.692	0.7138	0.6925	0.6875	0.692	0.7143	0.692	0.6885	0.8492	0.7744	0.5931	0.5538	0.6565	0.9615	0.4235	0.6372	0.7980	0.4457	0.8227	0.7064	0.7802
0.3194	42.0	42000	0.9722	0.6953	0.7173	0.6957	0.6910	0.6953	0.7178	0.6953	0.6919	0.8512	0.7761	0.5989	0.5669	0.6565	0.9615	0.4306	0.6404	0.7980	0.4457	0.8227	0.7156	0.7802
0.3241	43.0	43000	0.9658	0.6987	0.7218	0.6987	0.6950	0.6987	0.7210	0.6987	0.6953	0.8533	0.7811	0.6017	0.5774	0.6620	0.9615	0.4306	0.6435	0.8030	0.4457	0.8276	0.7156	0.7802
0.3078	44.0	44000	0.9608	0.6993	0.7227	0.6992	0.6957	0.6993	0.7214	0.6993	0.6958	0.8533	0.7811	0.6017	0.5801	0.6620	0.9654	0.4306	0.6435	0.8030	0.4457	0.8276	0.7156	0.7802
0.3084	45.0	45000	0.9572	0.6997	0.7234	0.7002	0.6964	0.6997	0.7222	0.6997	0.6964	0.8512	0.7811	0.6017	0.5853	0.6620	0.9654	0.4306	0.6435	0.8030	0.4457	0.8276	0.7248	0.7802
0.3223	46.0	46000	0.9536	0.702	0.7275	0.7031	0.6997	0.702	0.7249	0.702	0.6987	0.8533	0.7811	0.6017	0.5853	0.6648	0.9654	0.4306	0.6530	0.8177	0.4457	0.8276	0.7339	0.7802
0.2982	47.0	47000	0.9505	0.7017	0.7279	0.7028	0.6997	0.7017	0.7250	0.7017	0.6984	0.8554	0.7795	0.6017	0.5906	0.6648	0.9654	0.4306	0.6435	0.8177	0.4457	0.8276	0.7339	0.7802
0.3047	48.0	48000	0.9490	0.7017	0.7278	0.7029	0.6998	0.7017	0.7249	0.7017	0.6985	0.8533	0.7795	0.6017	0.5906	0.6676	0.9654	0.4306	0.6435	0.8177	0.4457	0.8276	0.7339	0.7802
0.3013	49.0	49000	0.9477	0.7027	0.7284	0.7038	0.7007	0.7027	0.7255	0.7027	0.6994	0.8554	0.7795	0.6017	0.5906	0.6676	0.9654	0.4306	0.6530	0.8177	0.4457	0.8276	0.7339	0.7802
0.3005	50.0	50000	0.9474	0.7023	0.7283	0.7032	0.7003	0.7023	0.7252	0.7023	0.6989	0.8554	0.7795	0.6017	0.5932	0.6676	0.9654	0.4306	0.6435	0.8177	0.4457	0.8276	0.7339	0.7802

Framework versions

• PEFT 0.14.0
• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.3.0
• Tokenizers 0.21.0