flan-t5-base-dialogsum_v2

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9323
• Rouge1: 49.6194
• Rouge2: 23.9441
• Rougel: 47.1784
• Rougelsum: 47.7351
• 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: 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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.8419	1.0	1558	0.9686	49.1717	21.6321	46.5027	46.9099	19.0
0.7931	2.0	3116	0.9393	49.0429	25.016	46.8583	47.1894	19.0
0.7931	3.0	4674	0.9555	49.969	24.7039	47.6287	48.1426	19.0
0.7572	4.0	6232	0.9746	49.1309	23.2537	46.4771	46.9249	19.0
0.7417	5.0	7790	0.9820	49.2093	23.1137	46.6788	47.1939	19.0
0.7491	6.0	9348	0.9592	50.4655	25.5968	48.0925	48.6231	19.0
0.7181	7.0	10906	0.9593	50.6565	24.5502	48.0572	48.5985	19.0
0.7144	8.0	12464	0.9386	50.9351	25.9342	48.4183	48.898	19.0
0.7137	9.0	14022	0.9486	49.5404	23.7464	46.9305	47.4827	19.0
0.6728	10.0	15580	0.9419	49.9498	23.8617	47.3421	47.9081	19.0
0.6783	11.0	17138	0.9459	50.52	25.9799	48.0356	48.596	19.0
0.6708	12.0	18696	0.9552	49.85	23.8992	47.3564	47.8535	19.0
0.6515	13.0	20254	0.9462	49.7239	25.468	47.4924	48.0146	19.0
0.6419	14.0	21812	0.9507	49.8791	25.3529	47.5795	48.0845	19.0
0.6297	15.0	23370	0.9323	49.6194	23.9441	47.1784	47.7351	19.0
0.6354	16.0	24928	0.9408	50.3383	25.6301	48.1241	48.6641	19.0
0.6178	17.0	26486	0.9420	49.5266	24.6134	47.1452	47.6481	19.0
0.608	18.0	28044	0.9549	49.7907	24.9381	47.4622	48.0312	19.0
0.6205	19.0	29602	0.9580	49.687	25.1737	47.4014	47.8466	19.0
0.5946	20.0	31160	0.9519	50.1159	25.3671	47.8232	48.3411	19.0
0.6107	21.0	32718	0.9549	49.9508	24.746	47.5774	48.1047	19.0
0.591	22.0	34276	0.9543	50.1561	25.3451	47.8341	48.3492	19.0
0.6017	23.0	35834	0.9570	49.893	25.1383	47.503	48.059	19.0
0.5942	24.0	37392	0.9584	49.9479	24.9422	47.5917	48.1012	19.0
0.5826	25.0	38950	0.9592	49.9049	25.1102	47.5736	48.1165	19.0

Framework versions

• Transformers 4.35.2
• Pytorch 2.0.1+cu117
• Datasets 2.15.0
• Tokenizers 0.15.0