flan-t5-rouge-durga-q5-clean-1

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

• Loss: 0.0005
• Rouge1: 0.7445
• Rouge2: 0.7185
• Rougel: 0.7434
• Rougelsum: 0.7443

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: 0.0003
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.9251	1.0	25	1.6278	0.2967	0.0873	0.2883	0.2882
1.5255	2.0	50	1.1981	0.3701	0.1385	0.3567	0.3556
2.3769	3.0	75	0.9291	0.3827	0.1966	0.3803	0.3791
1.2209	4.0	100	0.7037	0.4314	0.2566	0.4236	0.4236
0.7555	5.0	125	0.5043	0.4341	0.2524	0.4342	0.4309
1.4086	6.0	150	0.4352	0.4525	0.2964	0.4495	0.4505
1.3036	7.0	175	0.3015	0.5994	0.4670	0.5977	0.5982
0.592	8.0	200	0.2114	0.5736	0.4563	0.5723	0.5720
0.6039	9.0	225	0.1706	0.5588	0.4524	0.5569	0.5584
0.4569	10.0	250	0.1314	0.6186	0.5224	0.6172	0.6172
0.5411	11.0	275	0.0986	0.6435	0.5631	0.6427	0.6456
0.8071	12.0	300	0.0894	0.6291	0.5462	0.6307	0.6286
0.5719	13.0	325	0.0629	0.6876	0.6185	0.6832	0.6838
0.0508	14.0	350	0.0486	0.6538	0.5946	0.6525	0.6478
0.0632	15.0	375	0.0350	0.6864	0.6361	0.6850	0.6853
0.1612	16.0	400	0.0256	0.6867	0.6456	0.6855	0.6851
0.0411	17.0	425	0.0175	0.7375	0.7103	0.7368	0.7375
0.2348	18.0	450	0.0112	0.7256	0.6949	0.7255	0.7260
0.0872	19.0	475	0.0109	0.7247	0.6968	0.7260	0.7267
0.1036	20.0	500	0.0054	0.7421	0.7164	0.7422	0.7422
0.0815	21.0	525	0.0066	0.7432	0.7173	0.7420	0.7427
0.1068	22.0	550	0.0034	0.7445	0.7185	0.7434	0.7443
0.0786	23.0	575	0.0024	0.7421	0.7164	0.7422	0.7422
0.0566	24.0	600	0.0015	0.7421	0.7164	0.7422	0.7422
0.0103	25.0	625	0.0016	0.7445	0.7185	0.7434	0.7443
0.0003	26.0	650	0.0007	0.7445	0.7185	0.7434	0.7443
0.0569	27.0	675	0.0004	0.7445	0.7185	0.7434	0.7443
0.0389	28.0	700	0.0005	0.7445	0.7185	0.7434	0.7443
0.0089	29.0	725	0.0005	0.7445	0.7185	0.7434	0.7443
0.0022	30.0	750	0.0005	0.7445	0.7185	0.7434	0.7443

Framework versions

• Transformers 4.46.0
• Pytorch 2.5.0+cu121
• Datasets 3.0.2
• Tokenizers 0.20.1