t5_summarize

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

• Loss: 2.6492
• Evaluation Runtime: 28.4792
• Rounded Rouge Scores: {'rouge1': 0.174, 'rouge2': 0.0607, 'rougeL': 0.1367, 'rougeLsum': 0.1369}

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Evaluation Runtime	Rounded Rouge Scores
2.7245	1.0	500	2.6814	29.2864	{'rouge1': 0.1697, 'rouge2': 0.0584, 'rougeL': 0.1344, 'rougeLsum': 0.1345}
2.7318	2.0	1000	2.6707	27.6464	{'rouge1': 0.1735, 'rouge2': 0.0597, 'rougeL': 0.1372, 'rougeLsum': 0.1373}
2.7164	3.0	1500	2.6646	27.3926	{'rouge1': 0.1734, 'rouge2': 0.06, 'rougeL': 0.1371, 'rougeLsum': 0.1372}
2.7054	4.0	2000	2.6600	27.3819	{'rouge1': 0.1739, 'rouge2': 0.0599, 'rougeL': 0.1367, 'rougeLsum': 0.1368}
2.6955	5.0	2500	2.6581	27.9933	{'rouge1': 0.1731, 'rouge2': 0.0601, 'rougeL': 0.1361, 'rougeLsum': 0.1361}
2.6865	6.0	3000	2.6535	28.2157	{'rouge1': 0.1733, 'rouge2': 0.0603, 'rougeL': 0.1363, 'rougeLsum': 0.1364}
2.6821	7.0	3500	2.6521	29.0758	{'rouge1': 0.174, 'rouge2': 0.0606, 'rougeL': 0.1366, 'rougeLsum': 0.1369}
2.681	8.0	4000	2.6508	31.2621	{'rouge1': 0.1743, 'rouge2': 0.0609, 'rougeL': 0.1367, 'rougeLsum': 0.1369}
2.6771	9.0	4500	2.6499	30.4251	{'rouge1': 0.1735, 'rouge2': 0.0605, 'rougeL': 0.1364, 'rougeLsum': 0.1365}
2.6751	10.0	5000	2.6492	28.4792	{'rouge1': 0.174, 'rouge2': 0.0607, 'rougeL': 0.1367, 'rougeLsum': 0.1369}

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2