gokuls's picture
End of training
f3b2f7a
metadata
language:
  - en
tags:
  - generated_from_trainer
datasets:
  - glue
metrics:
  - accuracy
model-index:
  - name: hBERTv1_new_pretrain_w_init__mnli
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: GLUE MNLI
          type: glue
          config: mnli
          split: validation_matched
          args: mnli
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.36421887713588286

hBERTv1_new_pretrain_w_init__mnli

This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_wt_init on the GLUE MNLI dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0846
  • Accuracy: 0.3642

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: 4e-05
  • train_batch_size: 128
  • eval_batch_size: 128
  • seed: 10
  • distributed_type: multi-GPU
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.0911 1.0 3068 1.0944 0.3340
1.0984 2.0 6136 1.0967 0.3182
1.0988 3.0 9204 1.0962 0.3545
1.0989 4.0 12272 1.0970 0.3182
1.0986 5.0 15340 1.0896 0.3515
1.0984 6.0 18408 1.0986 0.3545
1.0992 7.0 21476 1.0994 0.3274
1.0993 8.0 24544 1.0999 0.3545
1.099 9.0 27612 1.0965 0.3182
1.0991 10.0 30680 1.1005 0.3182

Framework versions

  • Transformers 4.29.2
  • Pytorch 1.14.0a0+410ce96
  • Datasets 2.12.0
  • Tokenizers 0.13.3