flan-t5-large-extraction-all-dm_4000-ep10-nonstop

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

• Loss: 1.5811
• Hint Hit Num: 2.366
• Hint Precision: 0.4373
• Num: 5.3574
• Gen Len: 18.7764

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: 2e-05
• train_batch_size: 12
• eval_batch_size: 96
• seed: 1799
• 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	Hint Hit Num	Hint Precision	Num	Gen Len
2.2334	0.6	200	1.5968	2.7154	0.4867	5.6214	18.96
2.0104	1.2	400	1.5582	2.6402	0.4732	5.5972	18.9258
1.8919	1.8	600	1.5531	2.5878	0.4689	5.5198	18.8178
1.8495	2.4	800	1.5572	2.4834	0.4556	5.4372	18.8354
1.7984	2.99	1000	1.5569	2.426	0.4514	5.352	18.791
1.7446	3.59	1200	1.5443	2.4482	0.4502	5.4084	18.8278
1.7167	4.19	1400	1.5532	2.4502	0.4524	5.3864	18.8066
1.6838	4.79	1600	1.5603	2.3258	0.4387	5.2672	18.7164
1.6713	5.39	1800	1.5602	2.3874	0.4431	5.3434	18.7818
1.6403	5.99	2000	1.5636	2.313	0.4315	5.3092	18.7958
1.5963	6.59	2200	1.5745	2.353	0.4389	5.3184	18.7622
1.6203	7.19	2400	1.5757	2.3634	0.4348	5.3772	18.784
1.5808	7.78	2600	1.5822	2.33	0.4318	5.3452	18.8062
1.5436	8.38	2800	1.5791	2.3764	0.4394	5.3548	18.7792
1.5707	8.98	3000	1.5786	2.3592	0.4349	5.37	18.807
1.5512	9.58	3200	1.5820	2.362	0.4378	5.3418	18.773

Framework versions

• Transformers 4.18.0
• Pytorch 1.10.0+cu111
• Datasets 2.5.1
• Tokenizers 0.12.1