mdeberta-squad-qasina-LR-5e-06-BS-16-ML-384

This model is a fine-tuned version of mhdafifan/mdeberta-squad-id-LR-2e-05-BS-16-ML-384 on the QASiNa dataset. It achieves the following results on the evaluation set:

• Loss: 1.0449
• F1: 79.2234
• Exact Match: 67.9245

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-06
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Exact Match
1.0264	1.0	46	1.0156	78.8963	68.8679
0.7833	2.0	92	0.9699	79.2803	69.8113
0.5752	3.0	138	0.9659	79.2234	67.9245
0.4626	4.0	184	1.0120	78.9875	67.9245
0.3366	5.0	230	1.0449	79.2234	67.9245

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.1+cu128
• Datasets 4.4.1
• Tokenizers 0.22.1