md_mt5_base_boun_split_second_v1_retrain_on_boun_1612

This model is a fine-tuned version of Buseak/md_mt5_base_boun_split_second_v1_retrain_on_second_imst on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1146
• Bleu: 0.7971
• Gen Len: 18.8069

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.3542	1.0	975	0.1679	0.7453	18.7959
0.3354	2.0	1950	0.1553	0.7492	18.8033
0.3183	3.0	2925	0.1500	0.7576	18.7972
0.3013	4.0	3900	0.1421	0.7618	18.7969
0.2912	5.0	4875	0.1364	0.7673	18.8005
0.2816	6.0	5850	0.1321	0.7732	18.8003
0.2753	7.0	6825	0.1287	0.7825	18.8067
0.2718	8.0	7800	0.1252	0.7866	18.8051
0.2631	9.0	8775	0.1222	0.7885	18.799
0.2576	10.0	9750	0.1197	0.793	18.8041
0.2541	11.0	10725	0.1182	0.7935	18.8056
0.2505	12.0	11700	0.1165	0.7956	18.8056
0.2492	13.0	12675	0.1156	0.7971	18.8056
0.2534	14.0	13650	0.1146	0.7964	18.8069
0.2455	15.0	14625	0.1146	0.7971	18.8069

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0