flan-t5-base-YT-transcript-sum

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

• Loss: 1.4111
• Rouge1: 25.4013
• Rouge2: 12.4728
• Rougel: 21.5206
• Rougelsum: 23.6322
• 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: 5e-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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	216	1.5817	23.8566	11.0314	20.1664	22.2953	18.9954
No log	2.0	432	1.4907	24.2446	11.6603	20.6712	22.4196	18.9861
1.7643	3.0	648	1.4510	25.4355	12.9236	21.584	23.7272	19.0
1.7643	4.0	864	1.4312	24.8929	12.5927	21.3295	23.3504	19.0
1.4359	5.0	1080	1.4145	25.242	12.9269	21.6351	23.6509	19.0
1.4359	6.0	1296	1.4111	25.4013	12.4728	21.5206	23.6322	19.0
1.2819	7.0	1512	1.4135	25.6542	13.103	22.2059	23.9474	19.0
1.2819	8.0	1728	1.4145	26.0783	13.7584	22.343	24.3255	19.0
1.2819	9.0	1944	1.4163	25.4385	13.1278	21.7173	23.8295	18.9861
1.1688	10.0	2160	1.4208	25.7625	13.5586	22.2246	24.2042	19.0
1.1688	11.0	2376	1.4165	25.5482	13.1163	21.9475	23.8181	18.9907
1.0951	12.0	2592	1.4215	25.7614	13.5565	22.1965	24.0657	19.0
1.0951	13.0	2808	1.4285	26.3345	14.2027	22.7422	24.6261	18.9907
1.0549	14.0	3024	1.4277	25.8835	13.8044	22.3845	24.269	19.0
1.0549	15.0	3240	1.4321	25.8292	13.7231	22.3506	24.3188	19.0

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3