flan-t5-large-summary

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

• Loss: 0.3277
• Rouge1: 69.554
• Rouge2: 59.2044
• Rougel: 67.6581
• Rougelsum: 67.7174
• Gen Len: 18.7309

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: 8
• eval_batch_size: 8
• 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
No log	1.0	190	0.3832	67.2926	53.7031	64.7974	64.8071	18.8179
No log	2.0	380	0.3090	67.8051	55.3515	65.6548	65.6747	18.8232
0.2632	3.0	570	0.2876	68.3985	56.4896	66.3492	66.3648	18.7889
0.2632	4.0	760	0.2789	68.5676	56.4035	66.101	66.1396	18.8100
0.2632	5.0	950	0.2805	68.2045	56.6864	66.2411	66.3045	18.7784
0.2625	6.0	1140	0.2827	68.4741	57.1103	66.101	66.1991	18.7810
0.2625	7.0	1330	0.2767	68.9503	57.7255	66.8972	66.9236	18.7916
0.2048	8.0	1520	0.2776	69.1733	57.5766	66.7948	66.9347	18.7968
0.2048	9.0	1710	0.2777	69.7098	58.867	67.8471	67.8675	18.7388
0.2048	10.0	1900	0.2855	69.2753	57.8341	67.1037	67.2396	18.7757
0.1653	11.0	2090	0.2819	69.7836	58.5212	67.6087	67.6426	18.7916
0.1653	12.0	2280	0.2918	70.0288	59.2524	68.0356	68.0754	18.7573
0.1653	13.0	2470	0.2940	69.7288	59.2246	67.8426	67.8537	18.7704
0.1421	14.0	2660	0.2995	69.6257	58.8437	67.6509	67.6521	18.7810
0.1421	15.0	2850	0.3042	69.5083	58.6743	67.4254	67.5589	18.7625
0.1201	16.0	3040	0.3114	69.3398	58.9756	67.4566	67.5762	18.7810
0.1201	17.0	3230	0.3158	69.3569	58.8509	67.3013	67.4321	18.7520
0.1201	18.0	3420	0.3215	69.7484	59.1145	67.6617	67.6992	18.7573
0.1076	19.0	3610	0.3254	69.5556	59.2392	67.7187	67.7966	18.7520
0.1076	20.0	3800	0.3277	69.554	59.2044	67.6581	67.7174	18.7309

Framework versions

• Transformers 4.43.2
• Pytorch 2.2.0a0+81ea7a4
• Datasets 2.20.0
• Tokenizers 0.19.1