phobert-base-v2-finetuned

This model is a fine-tuned version of vinai/phobert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2000
• Accuracy: 0.9593
• F1: 0.9594

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.14	50	0.2415	0.9190	0.9192
No log	0.28	100	0.1917	0.9378	0.9379
No log	0.42	150	0.1861	0.9434	0.9434
No log	0.56	200	0.1760	0.9493	0.9493
No log	0.69	250	0.1706	0.9484	0.9484
No log	0.83	300	0.1710	0.9467	0.9467
No log	0.97	350	0.1609	0.9507	0.9507
0.2152	1.11	400	0.1678	0.9445	0.9446
0.2152	1.25	450	0.1626	0.9515	0.9515
0.2152	1.39	500	0.2076	0.9341	0.9343
0.2152	1.53	550	0.1559	0.9537	0.9538
0.2152	1.67	600	0.1562	0.9526	0.9526
0.2152	1.81	650	0.1377	0.9591	0.9591
0.2152	1.94	700	0.1396	0.9579	0.9580
0.1375	2.08	750	0.1526	0.9504	0.9505
0.1375	2.22	800	0.1507	0.9577	0.9577
0.1375	2.36	850	0.1485	0.9568	0.9568
0.1375	2.5	900	0.1419	0.9571	0.9572
0.1375	2.64	950	0.1552	0.9526	0.9527
0.1375	2.78	1000	0.1419	0.9588	0.9588
0.1375	2.92	1050	0.1338	0.9602	0.9602
0.1105	3.06	1100	0.1414	0.9599	0.9600
0.1105	3.19	1150	0.1420	0.9608	0.9608
0.1105	3.33	1200	0.1498	0.9574	0.9575
0.1105	3.47	1250	0.1402	0.9596	0.9596
0.1105	3.61	1300	0.1477	0.9596	0.9597
0.1105	3.75	1350	0.1362	0.9599	0.9599
0.1105	3.89	1400	0.1322	0.9563	0.9563
0.0916	4.03	1450	0.1384	0.9568	0.9569
0.0916	4.17	1500	0.1613	0.9596	0.9597
0.0916	4.31	1550	0.1509	0.9602	0.9602
0.0916	4.44	1600	0.1342	0.9591	0.9591
0.0916	4.58	1650	0.1479	0.9602	0.9602
0.0916	4.72	1700	0.1518	0.9588	0.9588
0.0916	4.86	1750	0.1474	0.9605	0.9605
0.0796	5.0	1800	0.1558	0.9543	0.9544
0.0796	5.14	1850	0.1645	0.9582	0.9582
0.0796	5.28	1900	0.1674	0.9577	0.9577
0.0796	5.42	1950	0.1669	0.9602	0.9602
0.0796	5.56	2000	0.1699	0.9588	0.9587
0.0796	5.69	2050	0.1514	0.9593	0.9594
0.0796	5.83	2100	0.1533	0.9568	0.9569
0.0796	5.97	2150	0.1577	0.9588	0.9588
0.0666	6.11	2200	0.1636	0.9585	0.9585
0.0666	6.25	2250	0.1717	0.9554	0.9555
0.0666	6.39	2300	0.1606	0.9563	0.9563
0.0666	6.53	2350	0.1649	0.9588	0.9588
0.0666	6.67	2400	0.1660	0.9579	0.9580
0.0666	6.81	2450	0.1593	0.9557	0.9558
0.0666	6.94	2500	0.1615	0.9577	0.9577
0.0563	7.08	2550	0.1848	0.9602	0.9602
0.0563	7.22	2600	0.1679	0.9596	0.9597
0.0563	7.36	2650	0.1716	0.9596	0.9596
0.0563	7.5	2700	0.1716	0.9585	0.9585
0.0563	7.64	2750	0.1888	0.9613	0.9613
0.0563	7.78	2800	0.1938	0.9596	0.9596
0.0563	7.92	2850	0.1897	0.9588	0.9588
0.0455	8.06	2900	0.1913	0.9554	0.9555
0.0455	8.19	2950	0.1874	0.9563	0.9563
0.0455	8.33	3000	0.1913	0.9588	0.9588
0.0455	8.47	3050	0.1925	0.9596	0.9596
0.0455	8.61	3100	0.1961	0.9577	0.9577
0.0455	8.75	3150	0.1904	0.9577	0.9577
0.0455	8.89	3200	0.1940	0.9610	0.9610
0.0389	9.03	3250	0.1894	0.9588	0.9588
0.0389	9.17	3300	0.1926	0.9596	0.9596
0.0389	9.31	3350	0.1977	0.9596	0.9596
0.0389	9.44	3400	0.1932	0.9571	0.9571
0.0389	9.58	3450	0.1972	0.9579	0.9580
0.0389	9.72	3500	0.1965	0.9577	0.9577
0.0389	9.86	3550	0.1996	0.9588	0.9588
0.0338	10.0	3600	0.2000	0.9593	0.9594

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2