bert-fa-base-uncased-finetuned-PQuAD-3epochs

This model is a fine-tuned version of HooshvareLab/bert-fa-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9011
• {'exact': 74.38140464883779, 'f1': 87.1425541703777, 'total': 8002, 'HasAns_exact': 68.61038107752957, 'HasAns_f1': 85.38349514969813, 'HasAns_total': 6088, 'NoAns_exact': 92.73772204806687, 'NoAns_f1': 92.73772204806687, 'NoAns_total': 1914, 'best_exact': 74.38140464883779, 'best_exact_thresh': 0.0, 'best_f1': 87.14255417037822, 'best_f1_thresh': 0.0}

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: 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
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.7492	1.0	4000	0.7875
0.4947	2.0	8000	0.8019
0.3259	3.0	12000	0.9011

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1