flan-t5-large-extraction-cnndm_8000-all-loss-ep10

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: 1.6575
• Hint Hit: 1.7631
• Hint Hit Num: 2.4865
• Hint Precision: 0.453
• Hint Bleu: 4.5231
• Num: 5.3805
• Gen Len: 18.9892

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: 32
• 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	Hint Hit Num	Hint Precision	Hint Bleu	Num	Gen Len
2.0921	0.8	200	1.7095	1.6641	2.376	0.4415	4.2976	5.2234	18.9901
1.876	1.6	400	1.6834	1.7129	2.4296	0.4501	4.4331	5.2965	18.9881
1.8142	2.4	600	1.6666	1.7411	2.4613	0.4519	4.4792	5.3419	18.9868
1.7704	3.2	800	1.6575	1.7631	2.4865	0.453	4.5231	5.3805	18.9892
1.7315	4.0	1000	1.6601	1.7121	2.4268	0.4469	4.4013	5.2852	18.9858
1.6928	4.8	1200	1.6577	1.6925	2.4148	0.4468	4.3806	5.3039	18.9817
1.6707	5.6	1400	1.6605	1.6666	2.3931	0.444	4.3596	5.2993	18.9846

Framework versions

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