pos_tagger_3112

This model is a fine-tuned version of dbmdz/bert-base-turkish-cased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5781
  • Precision: 0.8885
  • Recall: 0.8926
  • F1: 0.8906
  • Accuracy: 0.9222

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: 20

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 244 0.2966 0.8889 0.8935 0.8912 0.9215
No log 2.0 488 0.2884 0.8958 0.8969 0.8963 0.9258
0.3778 3.0 732 0.3138 0.8919 0.8913 0.8916 0.9224
0.3778 4.0 976 0.3086 0.8935 0.8960 0.8948 0.9252
0.1601 5.0 1220 0.3354 0.8897 0.8940 0.8919 0.9232
0.1601 6.0 1464 0.3486 0.8863 0.8911 0.8887 0.9212
0.1077 7.0 1708 0.3700 0.8899 0.8945 0.8922 0.9236
0.1077 8.0 1952 0.3920 0.8846 0.8905 0.8875 0.9207
0.0709 9.0 2196 0.4220 0.8860 0.8911 0.8885 0.9211
0.0709 10.0 2440 0.4467 0.8889 0.8934 0.8912 0.9226
0.0447 11.0 2684 0.4730 0.8834 0.8891 0.8862 0.9183
0.0447 12.0 2928 0.5008 0.8869 0.8921 0.8895 0.9217
0.0289 13.0 3172 0.5190 0.8866 0.8919 0.8892 0.9213
0.0289 14.0 3416 0.5384 0.8865 0.8904 0.8884 0.9204
0.0204 15.0 3660 0.5499 0.8852 0.8907 0.8879 0.9205
0.0204 16.0 3904 0.5564 0.8864 0.8923 0.8893 0.9213
0.015 17.0 4148 0.5688 0.8883 0.8920 0.8901 0.9220
0.015 18.0 4392 0.5740 0.8874 0.8927 0.8901 0.9213
0.0116 19.0 4636 0.5742 0.8879 0.8925 0.8902 0.9219
0.0116 20.0 4880 0.5781 0.8885 0.8926 0.8906 0.9222

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
19
Safetensors
Model size
110M 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 Buseak/pos_tagger_3112

Finetuned
(97)
this model