life2lang-small-pt

This model is a fine-tuned version of google/flan-t5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0421

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: 0.0001
• train_batch_size: 24
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 96
• 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: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
20.1414	0.0235	100	12.4348
8.0101	0.0470	200	3.5418
3.3624	0.0705	300	2.7387
2.6774	0.0940	400	2.2845
2.2274	0.1175	500	1.9608
1.873	0.1410	600	1.6025
1.5043	0.1645	700	1.4491
1.3942	0.1880	800	1.4141
1.3569	0.2115	900	1.3769
1.325	0.2350	1000	1.3504
1.2896	0.2585	1100	1.3211
1.2544	0.2820	1200	1.2955
1.2245	0.3055	1300	1.2460
1.1967	0.3290	1400	1.2218
1.1717	0.3525	1500	1.2105
1.1531	0.3760	1600	1.1831
1.1396	0.3995	1700	1.1938
1.1276	0.4230	1800	1.1658
1.1156	0.4465	1900	1.1518
1.1089	0.4700	2000	1.1556
1.1013	0.4935	2100	1.1558
1.0942	0.5170	2200	1.1376
1.0905	0.5405	2300	1.1338
1.0833	0.5640	2400	1.1275
1.075	0.5875	2500	1.1251
1.0733	0.6110	2600	1.1191
1.0691	0.6345	2700	1.1147
1.0624	0.6580	2800	1.1123
1.0573	0.6815	2900	1.1064
1.0526	0.7050	3000	1.0976
1.0515	0.7285	3100	1.0947
1.0465	0.7520	3200	1.0926
1.045	0.7755	3300	1.0881
1.0432	0.7990	3400	1.0867
1.0389	0.8225	3500	1.0867
1.0391	0.8460	3600	1.0825
1.032	0.8695	3700	1.0798
1.0302	0.8930	3800	1.0793
1.0321	0.9165	3900	1.0780
1.0298	0.9400	4000	1.0732
1.0254	0.9635	4100	1.0694
1.0241	0.9870	4200	1.0679
1.0174	1.0103	4300	1.0673
1.0214	1.0338	4400	1.0644
1.0208	1.0573	4500	1.0639
1.0196	1.0808	4600	1.0621
1.0168	1.1043	4700	1.0588
1.0158	1.1278	4800	1.0596
1.0147	1.1513	4900	1.0581
1.0144	1.1748	5000	1.0584
1.0109	1.1983	5100	1.0573
1.0093	1.2218	5200	1.0544
1.0086	1.2453	5300	1.0559
1.0084	1.2688	5400	1.0533
1.0092	1.2923	5500	1.0545
1.0079	1.3158	5600	1.0538
1.0075	1.3393	5700	1.0527
1.007	1.3628	5800	1.0522
1.0067	1.3863	5900	1.0510
1.007	1.4098	6000	1.0514
1.0074	1.4333	6100	1.0508
1.0023	1.4568	6200	1.0496
1.0002	1.4803	6300	1.0504
1.0018	1.5038	6400	1.0496
1.0012	1.5273	6500	1.0487
1.0043	1.5508	6600	1.0492
1.0024	1.5743	6700	1.0469
1.0008	1.5978	6800	1.0475
0.9965	1.6213	6900	1.0470
1.0014	1.6448	7000	1.0475
0.9988	1.6683	7100	1.0465
0.9983	1.6918	7200	1.0452
0.9984	1.7153	7300	1.0455
0.9983	1.7388	7400	1.0452
0.9972	1.7623	7500	1.0457
0.9968	1.7858	7600	1.0459
0.9979	1.8093	7700	1.0456
0.9966	1.8328	7800	1.0453
0.9983	1.8563	7900	1.0449
0.9931	1.8798	8000	1.0443
0.9959	1.9033	8100	1.0445
0.996	1.9268	8200	1.0448
0.9964	1.9503	8300	1.0441
0.9956	1.9738	8400	1.0447
0.9955	1.9973	8500	1.0447
0.9879	2.0207	8600	1.0451
0.9943	2.0442	8700	1.0436
0.9956	2.0677	8800	1.0438
0.9934	2.0912	8900	1.0439
0.9955	2.1147	9000	1.0429
0.9937	2.1382	9100	1.0438
0.9954	2.1617	9200	1.0432
0.9951	2.1852	9300	1.0436
0.9934	2.2087	9400	1.0431
0.9945	2.2322	9500	1.0432
0.9926	2.2557	9600	1.0420
0.9911	2.2792	9700	1.0425
0.9931	2.3027	9800	1.0431
0.9941	2.3262	9900	1.0432
0.9921	2.3497	10000	1.0425
0.9909	2.3732	10100	1.0421
0.9923	2.3967	10200	1.0425
0.9924	2.4202	10300	1.0417
0.994	2.4437	10400	1.0425
0.9908	2.4672	10500	1.0422
0.9947	2.4907	10600	1.0423
0.9913	2.5142	10700	1.0421
0.9925	2.5377	10800	1.0424
0.9906	2.5612	10900	1.0420
0.9929	2.5847	11000	1.0421
0.9922	2.6082	11100	1.0423
0.9942	2.6317	11200	1.0426
0.9905	2.6552	11300	1.0419
0.99	2.6787	11400	1.0418
0.9914	2.7022	11500	1.0422
0.9924	2.7256	11600	1.0422
0.9924	2.7491	11700	1.0421
0.993	2.7726	11800	1.0418
0.9931	2.7961	11900	1.0420
0.9914	2.8196	12000	1.0422
0.993	2.8431	12100	1.0422
0.9931	2.8666	12200	1.0421
0.9928	2.8901	12300	1.0420
0.9908	2.9136	12400	1.0420
0.9928	2.9371	12500	1.0421
0.9902	2.9606	12600	1.0421
0.994	2.9841	12700	1.0421

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2