PhobertLexicalMeta-revision_04_03_2026

This model is a fine-tuned version of phunganhsang/PhoBert_Lexical_Dataset55K on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6988
• Accuracy: 0.8621
• F1: 0.6351

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: Use 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.3497	100	0.2546	0.9085	0.6521
No log	0.6993	200	0.3251	0.8705	0.6354
0.3275	1.0490	300	0.3322	0.8688	0.6384
0.3275	1.3986	400	0.3272	0.8775	0.6448
0.3275	1.7483	500	0.4746	0.8060	0.5917
0.2076	2.0979	600	0.4108	0.8302	0.6037
0.2076	2.4476	700	0.3489	0.8728	0.6453
0.2076	2.7972	800	0.4454	0.8307	0.6104
0.1535	3.1469	900	0.4494	0.8438	0.6173
0.1535	3.4965	1000	0.3678	0.8721	0.6423
0.1535	3.8462	1100	0.4342	0.8479	0.6231
0.1153	4.1958	1200	0.3938	0.8753	0.6506
0.1153	4.5455	1300	0.3813	0.8768	0.6552
0.1153	4.8951	1400	0.4264	0.8660	0.6438
0.0835	5.2448	1500	0.4581	0.8587	0.6318
0.0835	5.5944	1600	0.4207	0.8733	0.6482
0.0835	5.9441	1700	0.4390	0.8728	0.6467
0.0637	6.2937	1800	0.4741	0.8680	0.6444
0.0637	6.6434	1900	0.3859	0.8927	0.6723
0.0637	6.9930	2000	0.4588	0.8725	0.6494
0.0456	7.3427	2100	0.5145	0.8626	0.6395
0.0456	7.6923	2200	0.4199	0.8809	0.6541
0.0336	8.0420	2300	0.5151	0.8671	0.6391
0.0336	8.3916	2400	0.5541	0.8545	0.6225
0.0336	8.7413	2500	0.5328	0.8680	0.6436
0.0262	9.0909	2600	0.6343	0.8459	0.6239
0.0262	9.4406	2700	0.4982	0.8846	0.6620
0.0262	9.7902	2800	0.5181	0.8795	0.6568
0.0208	10.1399	2900	0.5680	0.8624	0.6384
0.0208	10.4895	3000	0.6268	0.8597	0.6387
0.0208	10.8392	3100	0.5802	0.8659	0.6449
0.0171	11.1888	3200	0.5412	0.8743	0.6454
0.0171	11.5385	3300	0.5960	0.8697	0.6399
0.0171	11.8881	3400	0.5734	0.8705	0.6418
0.0132	12.2378	3500	0.6409	0.8580	0.6344
0.0132	12.5874	3600	0.5851	0.8715	0.6486
0.0132	12.9371	3700	0.5801	0.8712	0.6461
0.0111	13.2867	3800	0.5925	0.8711	0.6433
0.0111	13.6364	3900	0.6013	0.8740	0.6473
0.0111	13.9860	4000	0.6728	0.8584	0.6327
0.0085	14.3357	4100	0.6788	0.8570	0.6304
0.0085	14.6853	4200	0.6764	0.8597	0.6350
0.0071	15.0350	4300	0.6437	0.8656	0.6408
0.0071	15.3846	4400	0.7153	0.8566	0.6316
0.0071	15.7343	4500	0.6704	0.8669	0.6406
0.0060	16.0839	4600	0.6561	0.8708	0.6443
0.0060	16.4336	4700	0.6754	0.8639	0.6354
0.0060	16.7832	4800	0.6644	0.8653	0.6383
0.0056	17.1329	4900	0.7199	0.8573	0.6302
0.0056	17.4825	5000	0.7005	0.8586	0.6337
0.0056	17.8322	5100	0.6703	0.8673	0.6402
0.0045	18.1818	5200	0.6351	0.8745	0.6469
0.0045	18.5315	5300	0.7205	0.8570	0.6308
0.0045	18.8811	5400	0.7082	0.8586	0.6333
0.0045	19.2308	5500	0.6728	0.8670	0.6407
0.0045	19.5804	5600	0.6844	0.8643	0.6376
0.0045	19.9301	5700	0.6988	0.8621	0.6351

Framework versions

• Transformers 5.3.0
• Pytorch 2.4.1+cu121
• Datasets 4.6.1
• Tokenizers 0.22.2