kobigbird-pure2-93753787

This model is a fine-tuned version of monologg/kobigbird-bert-base on the custom_squad_v2 dataset. It achieves the following results on the evaluation set:

• eval_loss: 3.4619
• eval_runtime: 26.5163
• eval_samples_per_second: 45.406
• eval_steps_per_second: 0.717
• epoch: 0.99
• step: 21

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: 64
• eval_batch_size: 64
• seed: 2
• gradient_accumulation_steps: 8
• total_train_batch_size: 512
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 2
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.24.0
• Pytorch 1.12.1
• Datasets 2.6.1
• Tokenizers 0.13.1