flan-t5-rouge-durga-q5-clean-2

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: 0.0035
• Rouge1: 0.7335
• Rouge2: 0.7000
• Rougel: 0.7304
• Rougelsum: 0.7329

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: 3
• eval_batch_size: 3
• 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
1.3408	1.0	34	1.4618	0.2667	0.0665	0.2596	0.2602
0.8265	2.0	68	0.9113	0.3595	0.1570	0.3484	0.3491
0.6005	3.0	102	0.6349	0.3902	0.1965	0.3804	0.3836
0.7385	4.0	136	0.4371	0.4156	0.2511	0.4098	0.4121
0.4385	5.0	170	0.3191	0.4665	0.2901	0.4620	0.4636
0.5228	6.0	204	0.2193	0.4971	0.3658	0.4899	0.4932
0.7345	7.0	238	0.1589	0.5328	0.4149	0.5317	0.5341
0.6454	8.0	272	0.1229	0.5480	0.4471	0.5447	0.5496
0.1834	9.0	306	0.0956	0.5834	0.4832	0.5813	0.5832
0.1491	10.0	340	0.0741	0.5920	0.4950	0.5871	0.5894
0.224	11.0	374	0.0631	0.5901	0.5017	0.5887	0.5924
0.1733	12.0	408	0.0417	0.6601	0.5955	0.6572	0.6587
0.0927	13.0	442	0.0462	0.6558	0.5896	0.6527	0.6553
0.106	14.0	476	0.0296	0.6695	0.6164	0.6681	0.6715
0.1517	15.0	510	0.0223	0.7063	0.6618	0.7032	0.7036
0.0332	16.0	544	0.0249	0.7100	0.6666	0.7079	0.7073
0.1027	17.0	578	0.0179	0.7063	0.6558	0.7023	0.7026
0.0683	18.0	612	0.0131	0.7212	0.6782	0.7186	0.7202
0.003	19.0	646	0.0113	0.7287	0.6950	0.7256	0.7282
0.0083	20.0	680	0.0145	0.7123	0.6573	0.7105	0.7118
0.0005	21.0	714	0.0060	0.7197	0.6780	0.7169	0.7187
0.002	22.0	748	0.0132	0.7249	0.6897	0.7217	0.7246
0.0282	23.0	782	0.0034	0.7335	0.7000	0.7304	0.7329
0.0011	24.0	816	0.0042	0.7337	0.7008	0.7305	0.7321
0.012	25.0	850	0.0044	0.7335	0.7000	0.7304	0.7329
0.0083	26.0	884	0.0041	0.7334	0.6999	0.7302	0.7322
0.0172	27.0	918	0.0034	0.7280	0.6906	0.7250	0.7259
0.0125	28.0	952	0.0039	0.7279	0.6909	0.7253	0.7276
0.0162	29.0	986	0.0038	0.7335	0.7000	0.7304	0.7329
0.0172	30.0	1020	0.0035	0.7335	0.7000	0.7304	0.7329

Framework versions

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