entailed_after_rte-bert-base-uncased

This model is a fine-tuned version of kennethge123/superglue_rte-bert-base-uncased on the bigbench dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7322
  • Accuracy: 0.5714

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: 5e-05
  • train_batch_size: 4
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 30 0.6876 0.5714
No log 2.0 60 0.8029 0.5714
No log 3.0 90 0.7246 0.5714
No log 4.0 120 0.7152 0.5714
No log 5.0 150 0.7887 0.5714
No log 6.0 180 0.7498 0.5714
No log 7.0 210 0.8149 0.4286
No log 8.0 240 0.7055 0.5714
No log 9.0 270 0.7209 0.5714
No log 10.0 300 0.6922 0.5714
No log 11.0 330 0.7186 0.5714
No log 12.0 360 0.6916 0.5714
No log 13.0 390 0.7233 0.5714
No log 14.0 420 0.7109 0.5714
No log 15.0 450 0.7051 0.5714
No log 16.0 480 0.6968 0.5714
0.7046 17.0 510 0.7068 0.5714
0.7046 18.0 540 0.7319 0.5714
0.7046 19.0 570 0.7301 0.5714
0.7046 20.0 600 0.7322 0.5714

Framework versions

  • Transformers 4.37.0
  • Pytorch 1.13.1+cu117
  • Datasets 2.15.0
  • Tokenizers 0.15.2
Downloads last month
17
Safetensors
Model size
109M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for kennethge123/entailed_after_rte-bert-base-uncased

Dataset used to train kennethge123/entailed_after_rte-bert-base-uncased

Evaluation results