t5-end2end-questions-generation-cvqualtrics-squad-V1

Model description

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

• Loss: 1.2337

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
2.6162	0.34	100	1.8890
1.9995	0.67	200	1.6871
1.8697	1.01	300	1.6146
1.7682	1.34	400	1.5530
1.7323	1.68	500	1.5232
1.7256	2.01	600	1.4921
1.6506	2.35	700	1.4640
1.6438	2.68	800	1.4406
1.6399	3.02	900	1.4137
1.5786	3.36	1000	1.3924
1.5805	3.69	1100	1.3788
1.5824	4.03	1200	1.3626
1.5295	4.36	1300	1.3454
1.5333	4.7	1400	1.3356
1.537	5.03	1500	1.3230
1.5002	5.37	1600	1.3157
1.4936	5.7	1700	1.3046
1.4937	6.04	1800	1.2958
1.4649	6.38	1900	1.2826
1.4742	6.71	2000	1.2744
1.4641	7.05	2100	1.2603
1.4472	7.38	2200	1.2595
1.4403	7.72	2300	1.2526
1.4508	8.05	2400	1.2475
1.4191	8.39	2500	1.2412
1.4367	8.72	2600	1.2354
1.4272	9.06	2700	1.2386
1.4104	9.4	2800	1.2323
1.4179	9.73	2900	1.2337