Edit model card

robbert0410_lrate7.5b16

This model is a fine-tuned version of pdelobelle/robbert-v2-dutch-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5355
  • Precisions: 0.8523
  • Recall: 0.8173
  • F-measure: 0.8307
  • Accuracy: 0.9209

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: 7.5e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 16

Training results

Training Loss Epoch Step Validation Loss Precisions Recall F-measure Accuracy
0.6127 1.0 236 0.3656 0.8687 0.6888 0.7011 0.8817
0.3078 2.0 472 0.3390 0.8253 0.7452 0.7612 0.8947
0.1742 3.0 708 0.3899 0.7602 0.7560 0.7469 0.8957
0.1242 4.0 944 0.4402 0.8560 0.7678 0.7861 0.9055
0.0749 5.0 1180 0.4206 0.8163 0.8139 0.8127 0.9121
0.0533 6.0 1416 0.4824 0.8257 0.7936 0.8060 0.9124
0.0366 7.0 1652 0.4927 0.8506 0.7956 0.8158 0.9176
0.0273 8.0 1888 0.5638 0.8631 0.7855 0.8093 0.9202
0.0206 9.0 2124 0.5507 0.8322 0.7957 0.8096 0.9141
0.0154 10.0 2360 0.5355 0.8523 0.8173 0.8307 0.9209
0.0105 11.0 2596 0.5812 0.8301 0.7961 0.8088 0.9162
0.0086 12.0 2832 0.6084 0.8357 0.8065 0.8192 0.9130
0.0046 13.0 3068 0.6035 0.8310 0.7948 0.8104 0.9137
0.0036 14.0 3304 0.6034 0.8223 0.7980 0.8074 0.9134
0.0043 15.0 3540 0.6146 0.8198 0.7869 0.7999 0.9120
0.0018 16.0 3776 0.6070 0.8244 0.7894 0.8029 0.9134

Framework versions

  • Transformers 4.34.0
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.5
  • Tokenizers 0.14.0
Downloads last month
3
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 Tommert25/RobBERTBestModelOct11

Finetuned
(40)
this model
Finetunes
1 model