pretrain_Law_model_bartpho_version1

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

• Loss: 0.2879
• Rouge1: 0.4482
• Rouge2: 0.3252
• Rougel: 0.3897
• Rougelsum: 0.4055

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: 8
• eval_batch_size: 8
• 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	Rouge1	Rouge2	Rougel	Rougelsum
No log	1.0	245	0.3270	0.4333	0.2996	0.3650	0.3815
No log	2.0	490	0.3024	0.4400	0.3072	0.3731	0.3938
0.6695	3.0	735	0.2883	0.4411	0.3186	0.3829	0.3996
0.6695	4.0	980	0.2820	0.4484	0.3255	0.3883	0.4046
0.2387	5.0	1225	0.2822	0.4500	0.3221	0.3876	0.4050
0.2387	6.0	1470	0.2806	0.4483	0.3224	0.3875	0.4049
0.1843	7.0	1715	0.2852	0.4471	0.3235	0.3870	0.4037
0.1843	8.0	1960	0.2844	0.4466	0.3216	0.3850	0.4020
0.1515	9.0	2205	0.2866	0.4504	0.3265	0.3900	0.4066
0.1515	10.0	2450	0.2879	0.4482	0.3252	0.3897	0.4055

Framework versions

• Transformers 4.32.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3