flan-t5-bleu-squad-qg-120b

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

• Loss: 19.0104
• Bleu: 0.0272
• Precisions: [0.32, 0.15428571428571428, 0.04666666666666667, 0.016]
• Brevity Penalty: 0.3491
• Length Ratio: 0.4872
• Translation Length: 400
• Reference Length: 821

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: 24
• eval_batch_size: 24
• seed: 42
• optimizer: Use OptimizerNames.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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Precisions	Brevity Penalty	Length Ratio	Translation Length	Reference Length
28.5845	1.0	5	32.7207	0.0224	[0.3176470588235294, 0.18620689655172415, 0.058333333333333334, 0.021052631578947368]	0.2430	0.4141	340	821
24.0236	2.0	10	26.1977	0.0252	[0.29210526315789476, 0.16363636363636364, 0.05, 0.017391304347826087]	0.3133	0.4629	380	821
19.6925	3.0	15	22.4190	0.0268	[0.3, 0.15428571428571428, 0.04666666666666667, 0.016]	0.3491	0.4872	400	821
18.2172	4.0	20	19.9877	0.0272	[0.32, 0.15428571428571428, 0.04666666666666667, 0.016]	0.3491	0.4872	400	821
17.5741	5.0	25	19.0104	0.0272	[0.32, 0.15428571428571428, 0.04666666666666667, 0.016]	0.3491	0.4872	400	821

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0