learning_rate: 0.0001
train_batch_size: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 256
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7791	0.0483	1000	0.7469	0.8340
0.7006	0.0965	2000	0.7024	0.8423
0.6624	0.1448	3000	0.6724	0.8477
0.6255	0.1930	4000	0.6500	0.8523
0.6027	0.2413	5000	0.6289	0.8560
0.5836	0.2895	6000	0.6169	0.8583
0.5743	0.3378	7000	0.6033	0.8609
0.5613	0.3860	8000	0.5947	0.8626
0.5486	0.4343	9000	0.5874	0.8642
0.5447	0.4825	10000	0.5778	0.8657
0.5389	0.5308	11000	0.5710	0.8668
0.5295	0.5791	12000	0.5650	0.8688
0.5188	0.6273	13000	0.5551	0.8704
0.5102	0.6756	14000	0.5491	0.8715
0.5096	0.7238	15000	0.5462	0.8723
0.5052	0.7721	16000	0.5386	0.8736
0.4981	0.8203	17000	0.5339	0.8747
0.491	0.8686	18000	0.5272	0.8757
0.4894	0.9168	19000	0.5243	0.8764
0.4853	0.9651	20000	0.5232	0.8768
0.4812	1.0133	21000	0.5152	0.8779
0.4732	1.0616	22000	0.5143	0.8789
0.474	1.1098	23000	0.5101	0.8791
0.4701	1.1581	24000	0.5060	0.8803
0.4678	1.2063	25000	0.5025	0.8806
0.4661	1.2546	26000	0.5003	0.8811
0.464	1.3028	27000	0.4949	0.8822
0.461	1.3511	28000	0.4929	0.8825
0.4574	1.3994	29000	0.4916	0.8829
0.4598	1.4476	30000	0.4896	0.8834
0.4549	1.4959	31000	0.4878	0.8839
0.4523	1.5441	32000	0.4849	0.8846
0.4482	1.5924	33000	0.4820	0.8850
0.4477	1.6406	34000	0.4802	0.8854
0.4467	1.6889	35000	0.4789	0.8853
0.4434	1.7371	36000	0.4765	0.8862
0.443	1.7854	37000	0.4752	0.8865
0.4417	1.8336	38000	0.4741	0.8865
0.4393	1.8819	39000	0.4713	0.8870
0.4362	1.9302	40000	0.4708	0.8874
0.4356	1.9784	41000	0.4687	0.8877
0.4343	2.0266	42000	0.4677	0.8880
0.4333	2.0749	43000	0.4638	0.8888
0.4319	2.1231	44000	0.4645	0.8889
0.4363	2.1714	45000	0.4633	0.8886
0.4281	2.2197	46000	0.4602	0.8892
0.4242	2.2679	47000	0.4609	0.8895
0.4262	2.3162	48000	0.4576	0.8898
0.4231	2.3644	49000	0.4554	0.8904
0.4197	2.4127	50000	0.4562	0.8903
0.4231	2.4609	51000	0.4556	0.8902
0.422	2.5092	52000	0.4522	0.8909
0.4222	2.5574	53000	0.4526	0.8906
0.4208	2.6057	54000	0.4497	0.8916
0.42	2.6539	55000	0.4510	0.8914
0.4218	2.7022	56000	0.4492	0.8920
0.4162	2.7505	57000	0.4479	0.8922
0.4168	2.7987	58000	0.4466	0.8922
0.418	2.8470	59000	0.4466	0.8921
0.4164	2.8952	60000	0.4447	0.8928
0.4133	2.9435	61000	0.4437	0.8929
0.4103	2.9917	62000	0.4418	0.8932
0.4106	3.0400	63000	0.4397	0.8939
0.4122	3.0882	64000	0.4392	0.8938
0.4082	3.1365	65000	0.4380	0.8942
0.4069	3.1847	66000	0.4379	0.8942
0.4076	3.2330	67000	0.4369	0.8944
0.4079	3.2812	68000	0.4355	0.8946
0.4045	3.3295	69000	0.4351	0.8946
0.4032	3.3777	70000	0.4350	0.8950
0.4043	3.4260	71000	0.4329	0.8952
0.4018	3.4742	72000	0.4319	0.8952
0.4028	3.5225	73000	0.4324	0.8954
0.4017	3.5708	74000	0.4303	0.8956
0.4013	3.6190	75000	0.4312	0.8957
0.4003	3.6673	76000	0.4290	0.8959
0.397	3.7155	77000	0.4289	0.8958
0.3978	3.7638	78000	0.4280	0.8962
0.3985	3.8120	79000	0.4266	0.8967
0.3972	3.8603	80000	0.4257	0.8966
0.3926	3.9085	81000	0.4241	0.8969
0.397	3.9568	82000	0.4244	0.8968
0.3963	4.0050	83000	0.4239	0.8971
0.3952	4.0533	84000	0.4234	0.8970
0.3916	4.1015	85000	0.4217	0.8973
0.3938	4.1498	86000	0.4190	0.8979
0.3926	4.1980	87000	0.4188	0.8981
0.3924	4.2463	88000	0.4198	0.8978
0.3895	4.2945	89000	0.4195	0.8978
0.3918	4.3428	90000	0.4186	0.8982
0.3906	4.3911	91000	0.4182	0.8983
0.3927	4.4393	92000	0.4178	0.8983
0.3902	4.4876	93000	0.4170	0.8986
0.3924	4.5358	94000	0.4179	0.8982
0.3879	4.5841	95000	0.4144	0.8989
0.3897	4.6323	96000	0.4150	0.8989
0.3903	4.6806	97000	0.4158	0.8989
0.3893	4.7288	98000	0.4190	0.8983
0.5424	4.7771	99000	0.6255	0.8573
0.4768	4.8253	100000	0.4770	0.8850
0.4418	4.8736	101000	0.4506	0.8913
0.4562	4.9219	102000	0.4713	0.8871
0.4494	4.9701	103000	0.4575	0.8898

Framework versions

Transformers 4.56.0.dev0
Pytorch 2.8.0+cu128
Datasets 4.0.0
Tokenizers 0.21.4

htdung167/ViLegalBERT-v0

ViLegalBERT-v0

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

htdung167/
ViLegalBERT-v0