text-to-sparql-Version-3.5

This model is a fine-tuned version of yazdipour/text-to-sparql-t5-small-qald9 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0689
• Gen Len: 19.0
• Bleu-score: 0.3061
• Bleu-precisions: [91.41630901287553, 82.88177339901478, 72.54335260115607, 64.86013986013987]
• Bleu-bp: 0.0040

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 6
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Gen Len	Bleu-score	Bleu-precisions	Bleu-bp
No log	1.0	18	0.9058	19.0	0.3206	[77.68014059753955, 42.92730844793713, 18.040089086859687, 6.298200514138817]	0.0129
No log	2.0	36	0.3255	19.0	0.1908	[85.59046587215602, 65.37982565379825, 42.60614934114202, 29.129662522202487]	0.0037
No log	3.0	54	0.1637	19.0	0.1969	[89.16478555304741, 76.76240208877284, 62.538699690402474, 55.32319391634981]	0.0028
No log	4.0	72	0.1002	19.0	0.2395	[91.40022050716648, 79.41550190597205, 66.11694152923538, 57.22120658135283]	0.0033
No log	5.0	90	0.0761	19.0	0.2712	[91.28540305010893, 81.57894736842105, 70.64896755162242, 62.18637992831541]	0.0036
No log	6.0	108	0.0689	19.0	0.3061	[91.41630901287553, 82.88177339901478, 72.54335260115607, 64.86013986013987]	0.0040

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.2
• Datasets 2.1.0
• Tokenizers 0.15.2