t5-base-rouge-squad-qg

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

• Loss: 0.3358
• Rouge1: 0.3098
• Rouge2: 0.0914
• Rougel: 0.2967
• Rougelsum: 0.3043

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: 24
• eval_batch_size: 24
• 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
4.2041	1.0	3	2.1811	0.1282	0.0401	0.1198	0.1266
1.7635	2.0	6	0.6616	0.0	0.0	0.0	0.0
0.8464	3.0	9	0.5626	0.0	0.0	0.0	0.0
0.6561	4.0	12	0.4082	0.1282	0.0401	0.1198	0.1266
0.723	5.0	15	0.3290	0.1282	0.0401	0.1198	0.1266
0.2988	6.0	18	0.2900	0.3815	0.1296	0.3636	0.3763
0.1786	7.0	21	0.2800	0.3815	0.1296	0.3636	0.3763
0.2887	8.0	24	0.2849	0.4952	0.1959	0.4743	0.4872
0.3224	9.0	27	0.2869	0.3017	0.0980	0.2813	0.2986
0.5636	10.0	30	0.2889	0.3017	0.0980	0.2813	0.2986
0.271	11.0	33	0.2968	0.2384	0.0879	0.2257	0.2335
0.1144	12.0	36	0.3020	0.2473	0.0854	0.2343	0.2357
0.1005	13.0	39	0.3084	0.2517	0.0914	0.2388	0.2492
0.3569	14.0	42	0.3118	0.3098	0.0914	0.2967	0.3043
0.1051	15.0	45	0.3117	0.3098	0.0914	0.2967	0.3043
0.2862	16.0	48	0.3139	0.2950	0.1310	0.2818	0.2903
0.2077	17.0	51	0.3178	0.2950	0.1310	0.2818	0.2903
0.1055	18.0	54	0.3239	0.4408	0.1350	0.4223	0.4368
0.1761	19.0	57	0.3325	0.4408	0.1350	0.4223	0.4368
0.0704	20.0	60	0.3416	0.3098	0.0914	0.2967	0.3043
0.3277	21.0	63	0.3445	0.3098	0.0914	0.2967	0.3043
0.0859	22.0	66	0.3435	0.3098	0.0914	0.2967	0.3043
0.268	23.0	69	0.3412	0.3098	0.0914	0.2967	0.3043
0.1323	24.0	72	0.3378	0.3098	0.0914	0.2967	0.3043
0.0744	25.0	75	0.3351	0.3098	0.0914	0.2967	0.3043
0.1864	26.0	78	0.3343	0.3098	0.0914	0.2967	0.3043
0.1473	27.0	81	0.3341	0.3098	0.0914	0.2967	0.3043
0.0461	28.0	84	0.3346	0.3098	0.0914	0.2967	0.3043
0.0614	29.0	87	0.3354	0.3098	0.0914	0.2967	0.3043
0.0766	30.0	90	0.3358	0.3098	0.0914	0.2967	0.3043

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.21.0