distilbert-fa-zwnj-base-MLM-pquad

This model is pretained only on the PQuAD dataset. for educational purposes only.

Tokenizer and base model configs are from HooshvareLab/distilbert-fa-zwnj-base on the generator dataset.

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: 0.0005
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 1000
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

TrainOutput(global_step=31, training_loss=10.31849128969254, metrics={'train_runtime': 42.7618, 'train_samples_per_second': 188.369, 'train_steps_per_second': 0.725, 'total_flos': 263071290359808.0, 'train_loss': 10.31849128969254, 'epoch': 0.98})

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2