t5-small-finetuned-rahul-summariza

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

• Loss: 0.7002
• Rouge1: 29.5043
• Rouge2: 23.832
• Rougel: 27.5786
• Rougelsum: 28.404
• Gen Len: 19.0

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: 5.6e-05
• train_batch_size: 8
• eval_batch_size: 2
• 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	Gen Len
1.123	1.0	16	0.8258	27.2788	21.3634	25.7114	26.7324	19.0
0.9067	2.0	32	0.7539	28.873	23.5401	27.2337	27.939	19.0
0.8137	3.0	48	0.7280	29.1767	23.6599	27.7065	28.3569	19.0
0.7872	4.0	64	0.7230	29.0451	23.4597	27.2762	28.1324	19.0
0.7338	5.0	80	0.7133	29.4821	23.8113	27.4912	28.326	19.0
0.6913	6.0	96	0.7101	29.4237	23.8523	27.4109	28.2418	19.0
0.6679	7.0	112	0.7097	29.4237	23.8523	27.4109	28.2418	19.0
0.6963	8.0	128	0.7046	29.4237	23.8523	27.4109	28.2418	19.0
0.6223	9.0	144	0.7052	29.4237	23.7633	27.493	28.3362	19.0
0.6494	10.0	160	0.7019	29.4237	23.7633	27.493	28.3362	19.0
0.616	11.0	176	0.7010	29.4237	23.7633	27.493	28.3362	19.0
0.6058	12.0	192	0.7028	29.4237	23.7633	27.493	28.3362	19.0
0.5964	13.0	208	0.6996	29.4237	23.7633	27.493	28.3362	19.0
0.5958	14.0	224	0.6997	29.4237	23.7633	27.493	28.3362	19.0
0.57	15.0	240	0.6996	29.5043	23.832	27.5786	28.404	19.0
0.5714	16.0	256	0.6998	29.5043	23.832	27.5786	28.404	19.0
0.5648	17.0	272	0.6999	29.5043	23.832	27.5786	28.404	19.0
0.5258	18.0	288	0.7005	29.5043	23.832	27.5786	28.404	19.0
0.5692	19.0	304	0.7001	29.5043	23.832	27.5786	28.404	19.0
0.5708	20.0	320	0.7002	29.5043	23.832	27.5786	28.404	19.0

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu113
• Datasets 2.7.1
• Tokenizers 0.13.2