t5-end2end-question-generation

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

• Loss: 1.3875
• Rouge1: 29.8409
• Rouge2: 15.2583
• Rougel: 25.4802
• Rougelsum: 28.8023
• Gen Len: 18.9971
• Bleu: 1.8149
• Bleu 0: 71.9158
• Bleu 1: 46.3975
• Bleu 2: 31.3479
• Bleu 3: 20.236

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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Bleu	Bleu 0	Bleu 1	Bleu 2	Bleu 3
1.4252	0.21	500	1.4638	29.5937	14.6438	25.1309	28.5076	18.9990	1.7595	70.9726	44.8789	29.8013	18.9402
1.3591	0.42	1000	1.4619	29.4017	14.7271	25.1139	28.3406	19.0	1.7286	70.9671	45.415	30.2413	19.1132
1.426	0.64	1500	1.4313	29.9163	15.0542	25.5098	28.852	19.0	1.8109	71.924	46.0312	30.8421	19.6842
1.5525	0.85	2000	1.4177	30.0353	15.2661	25.6495	28.9867	19.0	1.8387	72.1696	46.3888	31.1768	20.1203
1.5035	1.06	2500	1.4185	29.7649	15.1864	25.4353	28.738	19.0	1.7868	71.9618	46.6091	31.4797	20.209
1.4294	1.27	3000	1.4138	29.5473	14.877	25.1373	28.5195	18.9990	1.7516	71.3163	45.6707	30.7404	19.6335
1.4336	1.49	3500	1.4058	29.9003	15.213	25.4924	28.8375	19.0	1.799	71.8573	46.2609	31.2086	20.0675
1.4434	1.7	4000	1.3978	30.0046	15.2722	25.6091	28.9496	18.9990	1.839	72.2448	46.6283	31.463	20.2921
1.4285	1.91	4500	1.3984	30.0478	15.1083	25.4469	28.9337	18.9990	1.8247	71.6695	45.7508	30.7813	19.7828
1.3926	2.12	5000	1.3982	30.0837	15.4009	25.6203	29.0334	18.9990	1.8237	72.2626	46.662	31.5043	20.2789
1.369	2.33	5500	1.3980	29.9042	15.1828	25.4962	28.8323	18.9990	1.8064	71.8783	46.1411	31.0047	19.9691
1.3577	2.55	6000	1.3936	29.9335	15.2821	25.5855	28.9161	19.0	1.8099	71.8881	46.3101	31.3396	20.3185
1.3636	2.76	6500	1.3908	29.9512	15.2434	25.5476	28.9224	18.9995	1.8242	71.9772	46.3212	31.2688	20.1704
1.3799	2.97	7000	1.3900	29.9393	15.1658	25.4702	28.8729	18.9971	1.8055	71.9431	46.1286	30.9969	19.9389
1.3318	3.18	7500	1.3934	29.7982	15.132	25.3908	28.7333	18.9995	1.7908	71.7081	46.1832	31.1416	20.1409
1.3208	3.4	8000	1.3928	29.9378	15.1421	25.4586	28.8793	19.0	1.8258	71.7795	45.969	30.9173	19.9664
1.3135	3.61	8500	1.3888	29.9264	15.2179	25.5529	28.875	19.0	1.8363	71.9537	46.2706	31.2245	20.2624
1.323	3.82	9000	1.3868	29.8749	15.2251	25.4639	28.7949	18.9971	1.812	71.6918	46.1503	31.0437	19.9965
1.3325	4.03	9500	1.3868	29.8804	15.2658	25.4848	28.8238	18.9971	1.8105	71.9146	46.3617	31.2842	20.1447
1.296	4.24	10000	1.3882	29.941	15.28	25.5209	28.9109	18.9971	1.817	71.994	46.3801	31.216	20.0596
1.3027	4.46	10500	1.3883	29.8492	15.2017	25.4398	28.7911	18.9971	1.7994	71.8366	46.0939	30.9953	19.9115
1.3046	4.67	11000	1.3880	29.8538	15.2605	25.4897	28.8236	18.9971	1.8136	71.9285	46.3689	31.2969	20.1728
1.294	4.88	11500	1.3875	29.8409	15.2583	25.4802	28.8023	18.9971	1.8149	71.9158	46.3975	31.3479	20.236

Framework versions

• Transformers 4.27.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2