Edit model card

distilbert-base-uncased__hate_speech_offensive__train-16-4

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0903
  • Accuracy: 0.4805

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: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.0974 1.0 10 1.1139 0.1
1.0637 2.0 20 1.0988 0.1
0.9758 3.0 30 1.1013 0.1
0.9012 4.0 40 1.0769 0.3
0.6993 5.0 50 1.0484 0.6
0.5676 6.0 60 1.0223 0.6
0.4069 7.0 70 0.9190 0.6
0.3192 8.0 80 1.1370 0.6
0.1112 9.0 90 1.1728 0.6
0.07 10.0 100 1.1998 0.6
0.0397 11.0 110 1.3700 0.6
0.027 12.0 120 1.3329 0.6
0.021 13.0 130 1.2697 0.6
0.0177 14.0 140 1.4195 0.6
0.0142 15.0 150 1.5342 0.6
0.0118 16.0 160 1.5999 0.6
0.0108 17.0 170 1.6327 0.6

Framework versions

  • Transformers 4.15.0
  • Pytorch 1.10.2+cu102
  • Datasets 1.18.2
  • Tokenizers 0.10.3
Downloads last month
13
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.