flan-t5-small-lamp-4u-finetuned-3

This model is a fine-tuned version of google/flan-t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.4088
• Rouge1: 0.1634
• Rouge2: 0.0510
• Rougel: 0.1494
• Rougelsum: 0.1500

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.6388	1.0	1566	2.5235	0.1413	0.0452	0.1311	0.1316
2.5039	2.0	3132	2.4539	0.1469	0.0474	0.1354	0.1359
2.401	3.0	4698	2.4320	0.1525	0.0486	0.1409	0.1414
2.3748	4.0	6264	2.4193	0.1528	0.0495	0.1414	0.1417
2.2997	5.0	7830	2.4120	0.1559	0.0490	0.1427	0.1430
2.2742	6.0	9396	2.4042	0.1562	0.0508	0.1436	0.1438
2.2404	7.0	10962	2.4039	0.1584	0.0515	0.1457	0.1461
2.2249	8.0	12528	2.4010	0.1624	0.0509	0.1491	0.1495
2.1985	9.0	14094	2.3993	0.1622	0.0520	0.1493	0.1501
2.1509	10.0	15660	2.3993	0.1599	0.0505	0.1454	0.1462
2.1226	11.0	17226	2.4026	0.1631	0.0519	0.1498	0.1503
2.107	12.0	18792	2.4040	0.1623	0.0513	0.1487	0.1491
2.0855	13.0	20358	2.4049	0.1634	0.0517	0.1493	0.1498
2.0678	14.0	21924	2.4028	0.1631	0.0515	0.1489	0.1495
2.0899	15.0	23490	2.4052	0.1628	0.0510	0.1489	0.1496
2.0777	16.0	25056	2.4050	0.1628	0.0503	0.1493	0.1498
2.0572	17.0	26622	2.4076	0.1620	0.0511	0.1481	0.1488
2.0408	18.0	28188	2.4066	0.1625	0.0510	0.1487	0.1495
2.0538	19.0	29754	2.4076	0.1635	0.0510	0.1496	0.1503
2.0283	20.0	31320	2.4088	0.1634	0.0510	0.1494	0.1500

Framework versions

• Transformers 4.35.2
• Pytorch 2.0.1+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0