flan-t5-small-samsum

This model is a fine-tuned version of google/flan-t5-small on the samsum dataset. It achieves the following results on the evaluation set:

• Loss: 1.6629
• Rouge1: 42.6378
• Rouge2: 18.2896
• Rougel: 35.1851
• Rougelsum: 38.8113
• Gen Len: 16.8596

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: 40
• eval_batch_size: 40
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.7932	0.27	100	1.6830	42.7032	18.4803	35.278	38.9331	17.0403
1.8102	0.54	200	1.6701	42.2811	18.2246	35.0893	38.4619	16.7265
1.8279	0.81	300	1.6658	42.6465	18.6939	35.4208	38.9399	16.8120
1.802	1.08	400	1.6633	42.5867	18.3579	35.3253	38.7049	16.6862
1.773	1.36	500	1.6629	42.6378	18.2896	35.1851	38.8113	16.8596
1.7752	1.63	600	1.6598	42.7111	18.3689	35.4218	38.8698	16.9328
1.7688	1.9	700	1.6589	42.6972	18.3536	35.3153	38.7976	17.0073

Framework versions

• Transformers 4.36.0
• Pytorch 2.0.0
• Datasets 2.15.0
• Tokenizers 0.15.0