aetin's picture
Model save
4dd3db3 verified
metadata
base_model: klue/roberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: roberta-interview-intent
    results: []

roberta-interview-intent

This model is a fine-tuned version of klue/roberta-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.8618
  • Accuracy: 0.6771
  • Macro F1: 0.4469
  • Weighted F1: 0.6798

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: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.06
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy Macro F1 Weighted F1
1.8294 1.0 859 1.4177 0.6630 0.3062 0.6381
0.7746 2.0 1718 1.3270 0.6777 0.3558 0.6590
0.5419 3.0 2577 1.3263 0.6806 0.4224 0.6715
0.3855 4.0 3436 1.4520 0.6775 0.4342 0.6726
0.2805 5.0 4295 1.5418 0.6775 0.4364 0.6767
0.2026 6.0 5154 1.5926 0.6734 0.4439 0.6771
0.1448 7.0 6013 1.7215 0.6775 0.4451 0.6802
0.106 8.0 6872 1.8030 0.6728 0.4492 0.6781
0.0778 9.0 7731 1.8198 0.6806 0.4518 0.6828
0.0611 10.0 8590 1.8618 0.6771 0.4469 0.6798

Framework versions

  • Transformers 4.40.2
  • Pytorch 2.8.0+cu128
  • Datasets 2.19.0
  • Tokenizers 0.19.1