t5_small_SA_abbr_replaced

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

• Loss: 0.5633
• Rouge1: 0.2325
• Rouge2: 0.1112
• Rougel: 0.2228
• Rougelsum: 0.2228
• Gen Len: 13.115
• Bert Score F1: 0.8407
• Bert Score Precision: 0.8551
• Bert Score Recall: 0.8282

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Bert Score F1	Bert Score Precision	Bert Score Recall
0.6839	0.95	500	0.6228	0.1495	0.0654	0.1433	0.1421	10.3894	0.6689	0.6835	0.656
0.5918	1.9	1000	0.6084	0.1965	0.0775	0.1893	0.1904	13.6991	0.8138	0.827	0.8024
0.5917	2.85	1500	0.5943	0.2075	0.0804	0.1975	0.1982	12.5841	0.8293	0.8452	0.8153
0.5524	3.8	2000	0.5859	0.2104	0.0856	0.1994	0.2	13.4602	0.8373	0.8523	0.8242
0.5635	4.74	2500	0.5830	0.2069	0.0905	0.1974	0.1984	12.1327	0.8379	0.8549	0.8231
0.5455	5.69	3000	0.5756	0.2113	0.0963	0.2038	0.2032	12.6903	0.8398	0.8545	0.8269
0.5156	6.64	3500	0.5710	0.2147	0.0993	0.2093	0.2096	13.115	0.8393	0.8527	0.8278
0.5134	7.59	4000	0.5683	0.2252	0.105	0.216	0.2168	12.4513	0.8418	0.8572	0.8285
0.5381	8.54	4500	0.5661	0.2228	0.1051	0.215	0.2151	13.4602	0.8401	0.8534	0.8287
0.5092	9.49	5000	0.5643	0.2186	0.1027	0.2135	0.2132	13.1593	0.8385	0.8524	0.8265
0.5181	10.44	5500	0.5643	0.2299	0.1087	0.2196	0.2202	13.0	0.8398	0.8543	0.8273
0.4824	11.39	6000	0.5633	0.2325	0.1112	0.2228	0.2228	13.115	0.8407	0.8551	0.8282

Framework versions

• Transformers 4.27.2
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2