anjleeg's picture
update model card README.md
043102b
metadata
license: mit
tags:
  - generated_from_trainer
datasets:
  - glue
metrics:
  - matthews_correlation
model-index:
  - name: roberta-base-finetuned-cola
    results: []

roberta-base-finetuned-cola

This model is a fine-tuned version of roberta-base on the glue dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4497
  • Matthews Correlation: 0.6272

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: 1
  • eval_batch_size: 1
  • seed: 42
  • distributed_type: IPU
  • gradient_accumulation_steps: 16
  • total_train_batch_size: 64
  • total_eval_batch_size: 20
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 5
  • training precision: Mixed Precision

Training results

Training Loss Epoch Step Validation Loss Matthews Correlation
0.4453 1.0 133 0.4348 0.5391
0.3121 2.0 266 0.3938 0.5827
0.1149 3.0 399 0.4497 0.6272
0.1194 4.0 532 0.5005 0.6076
0.1639 5.0 665 0.5645 0.5943

Framework versions

  • Transformers 4.18.0
  • Pytorch 1.10.0+cpu
  • Datasets 2.4.0
  • Tokenizers 0.12.1