husnu's picture
Update README.md
4dce54e
metadata
tags:
  - generated_from_trainer
datasets:
  - squad
model-index:
  - name: ft_electra-small-turkish-uncased-discriminator_lr-2e-1_epochs-1
    results: []

This model is a fine-tuned version of loodos/electra-small-turkish-uncased-discriminator on the squad dataset. It achieves the following results on the evaluation set:

  • Loss: 5.9506

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: 0.2
  • 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: 1

Training results

Training Loss Epoch Step Validation Loss
5.951 1.0 5818 5.9506

Framework versions

  • Transformers 4.15.0
  • Pytorch 1.10.0+cu111
  • Datasets 1.17.0
  • Tokenizers 0.10.3