xnli_xlm_r_only_sw / README.md
Dan Semin
update model card README.md
1a828d5
metadata
license: mit
tags:
  - text-classification
  - generated_from_trainer
datasets:
  - xnli
metrics:
  - accuracy
model-index:
  - name: xnli_xlm_r_only_sw
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: xnli
          type: xnli
          config: sw
          split: train
          args: sw
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.6903614457831325

xnli_xlm_r_only_sw

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

  • Loss: 0.9651
  • Accuracy: 0.6904

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: 128
  • eval_batch_size: 128
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.8628 1.0 3068 0.7719 0.6659
0.7407 2.0 6136 0.7147 0.6944
0.6791 3.0 9204 0.7591 0.6940
0.6293 4.0 12272 0.7538 0.6968
0.5833 5.0 15340 0.7716 0.6988
0.5425 6.0 18408 0.8323 0.6956
0.5029 7.0 21476 0.8407 0.6948
0.4707 8.0 24544 0.8840 0.6908
0.4437 9.0 27612 0.9506 0.6880
0.4234 10.0 30680 0.9651 0.6904

Framework versions

  • Transformers 4.24.0
  • Pytorch 1.13.0
  • Datasets 2.6.1
  • Tokenizers 0.13.1