CoolFace
Modelpublic

gokuls/hBERTv2_new_pretrain_mnli

sourceHugging Faceupdated 3y agoView on Hugging Face
0likes17downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

hBERTv2newpretrain_mnli

This model is a fine-tuned version of gokuls/bert_12_layer_model_v2_complete_training_new on the GLUE MNLI dataset. It achieves the following results on the evaluation set:

  • —Loss: 1.0964
  • —Accuracy: 0.3182

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
  • —trainbatchsize: 128
  • —evalbatchsize: 128
  • —seed: 10
  • —distributed_type: multi-GPU
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —num_epochs: 50

Training results

Training LossEpochStepValidation LossAccuracy
1.10311.030681.10080.3274
1.09892.061361.09640.3182
1.09873.092041.09800.3265
1.09874.0122721.09950.3182
1.09865.0153401.09860.3274
1.09916.0184081.10160.3545
1.09917.0214761.10000.3274

Framework versions

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