t5-small-mse-summarization

This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1108
• Rouge1: 43.1145
• Rouge2: 23.2262
• Rougel: 37.218
• Rougelsum: 41.0897
• Bleurt: -0.8051
• Gen Len: 18.549

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: 5e-05
• train_batch_size: 64
• eval_batch_size: 256
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bleurt	Gen Len
1.5207	1.0	267	1.2922	38.8738	19.1958	32.8458	36.9993	-0.9061	18.668
1.363	2.0	534	1.2340	39.8466	20.0452	33.9101	37.7708	-0.8925	18.657
1.3062	3.0	801	1.2057	40.5536	20.8249	34.5221	38.4648	-0.8625	18.602
1.272	4.0	1068	1.1782	41.0078	21.2186	35.0101	38.9186	-0.8595	18.602
1.2312	5.0	1335	1.1688	41.521	21.7934	35.704	39.4718	-0.842	18.486
1.2052	6.0	1602	1.1557	42.1037	22.4291	36.3554	40.1124	-0.8432	18.533
1.1842	7.0	1869	1.1440	42.4438	22.6456	36.5729	40.3134	-0.8288	18.553
1.1643	8.0	2136	1.1408	42.245	22.4859	36.3637	40.2193	-0.8284	18.622
1.1495	9.0	2403	1.1320	42.5362	22.5034	36.5092	40.4552	-0.8211	18.57
1.1368	10.0	2670	1.1301	42.5159	22.462	36.4646	40.3968	-0.819	18.538
1.1203	11.0	2937	1.1243	42.2803	22.5963	36.3454	40.2987	-0.8242	18.522
1.1116	12.0	3204	1.1197	42.8078	22.8409	36.7344	40.8186	-0.821	18.565
1.099	13.0	3471	1.1193	42.7423	22.9397	36.7894	40.7298	-0.8125	18.552
1.0976	14.0	3738	1.1176	42.9002	23.2394	37.0215	40.9211	-0.8156	18.568
1.0816	15.0	4005	1.1133	43.0007	23.3093	37.2037	40.9719	-0.8059	18.519
1.084	16.0	4272	1.1146	42.9053	23.2391	37.0542	40.8826	-0.8104	18.533
1.0755	17.0	4539	1.1124	43.0429	23.2773	37.1389	41.0755	-0.8086	18.544
1.0748	18.0	4806	1.1121	43.2243	23.4179	37.2039	41.143	-0.8048	18.548
1.072	19.0	5073	1.1106	43.1776	23.3061	37.3105	41.1392	-0.8039	18.549
1.0671	20.0	5340	1.1108	43.1145	23.2262	37.218	41.0897	-0.8051	18.549

Framework versions

• Transformers 4.21.2
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1