Edit model card

bert-squad-covidqa

This model is a fine-tuned version of hung200504/bert-squadv2 on the covid_qa_deepset dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5141

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: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss
4.1636 0.09 5 1.4553
0.8433 0.18 10 0.6359
0.8245 0.26 15 0.5610
0.5916 0.35 20 0.5416
0.5899 0.44 25 0.5148
0.4838 0.53 30 0.4996
0.4501 0.61 35 0.4929
0.7377 0.7 40 0.4610
0.455 0.79 45 0.4645
0.478 0.88 50 0.4745
0.3672 0.96 55 0.4803
0.6509 1.05 60 0.4875
0.3094 1.14 65 0.5089
0.3203 1.23 70 0.5751
0.3955 1.32 75 0.5416
0.6197 1.4 80 0.4848
0.455 1.49 85 0.4716
0.4086 1.58 90 0.4738
0.5028 1.67 95 0.4818
0.4953 1.75 100 0.4867
0.557 1.84 105 0.4826
0.3139 1.93 110 0.4832
0.3217 2.02 115 0.4921
0.4175 2.11 120 0.5056
0.3471 2.19 125 0.5204
0.209 2.28 130 0.5321
0.5151 2.37 135 0.5285
0.441 2.46 140 0.5141
0.3022 2.54 145 0.5031
0.3789 2.63 150 0.5002
0.2917 2.72 155 0.5041
0.372 2.81 160 0.5097
0.4001 2.89 165 0.5105
0.1803 2.98 170 0.5141

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.5
  • Tokenizers 0.14.1
Downloads last month
4
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 hung200504/bert-squad-covidqa

Dataset used to train hung200504/bert-squad-covidqa