mt5-small-finetuned-emails-gpt-summaries-batchs8-epochs20

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: 3.0969
• Rouge1: 0.1523
• Rouge2: 0.0732
• Rougel: 0.1381
• Rougelsum: 0.1512

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: 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	Rouge1	Rouge2	Rougel	Rougelsum
20.8315	1.0	18	12.1572	0.0110	0.0	0.0110	0.0110
16.2522	2.0	36	9.4668	0.0172	0.0025	0.0172	0.0168
13.8556	3.0	54	8.5239	0.0149	0.0025	0.0152	0.0146
11.5721	4.0	72	7.3135	0.0128	0.0013	0.0128	0.0125
10.2536	5.0	90	7.3664	0.0118	0.0013	0.0120	0.0115
8.8152	6.0	108	5.2895	0.0288	0.0070	0.0290	0.0286
7.909	7.0	126	3.9341	0.0285	0.0013	0.0284	0.0281
6.7655	8.0	144	3.6549	0.0427	0.0069	0.0406	0.0394
6.243	9.0	162	3.5561	0.0474	0.0133	0.0437	0.0462
5.5584	10.0	180	3.4047	0.0823	0.0253	0.0786	0.0763
5.1966	11.0	198	3.3160	0.1221	0.0484	0.1186	0.1209
4.9761	12.0	216	3.2536	0.1469	0.0751	0.1364	0.1481
4.7006	13.0	234	3.2081	0.1353	0.0637	0.1162	0.1318
4.4686	14.0	252	3.1687	0.1453	0.0707	0.1347	0.1399
4.3912	15.0	270	3.1480	0.1574	0.0782	0.1449	0.1498
4.2511	16.0	288	3.1414	0.1541	0.0734	0.1412	0.1509
4.1233	17.0	306	3.1239	0.1572	0.0734	0.1439	0.1536
4.0948	18.0	324	3.1061	0.1451	0.0689	0.1316	0.1438
4.0434	19.0	342	3.0981	0.1517	0.0732	0.1377	0.1507
4.1557	20.0	360	3.0969	0.1523	0.0732	0.1381	0.1512

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0