mt5-small-finetuned-mt5-small-v2

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.3614
• Rouge1: 0.1151
• Rouge2: 0.0251
• Rougel: 0.1143
• Rougelsum: 0.1144

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: 5.6e-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: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
3.5787	1.0	12500	2.7685	0.0863	0.0192	0.0858	0.0857
3.036	2.0	25000	2.6270	0.0911	0.0203	0.0905	0.0905
2.8761	3.0	37500	2.5564	0.101	0.0233	0.1004	0.1004
2.7709	4.0	50000	2.5080	0.1034	0.0231	0.1028	0.1028
2.6959	5.0	62500	2.4671	0.1068	0.0235	0.1061	0.1062
2.6328	6.0	75000	2.4539	0.11	0.026	0.1093	0.1093
2.5839	7.0	87500	2.4302	0.1101	0.0261	0.1092	0.1093
2.5418	8.0	100000	2.4083	0.1113	0.0252	0.1106	0.1108
2.5067	9.0	112500	2.3999	0.1115	0.0257	0.1107	0.1106
2.4762	10.0	125000	2.3857	0.1161	0.0264	0.1153	0.1153
2.4505	11.0	137500	2.3741	0.1141	0.0262	0.1133	0.1134
2.4281	12.0	150000	2.3737	0.1153	0.0259	0.1146	0.1147
2.4103	13.0	162500	2.3648	0.1156	0.0255	0.1148	0.1147
2.3961	14.0	175000	2.3652	0.1131	0.0246	0.1123	0.1123
2.3837	15.0	187500	2.3636	0.1141	0.0255	0.1133	0.1134
2.3772	16.0	200000	2.3614	0.1151	0.0251	0.1143	0.1144

Framework versions

• Transformers 4.40.0
• Pytorch 2.1.0+cu118
• Datasets 2.19.0
• Tokenizers 0.19.1