Hugging Face's logo

quancute
/
videberta-base-ner-ghtk-hirach_NER-first_1000_data-3090-15Nov

Text Classification
Transformers
Safetensors
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Community
videberta-base-ner-ghtk-hirach_NER-first_1000_data-3090-15Nov / README.md
quancute's picture
quancute
End of training
62e2dc2 verified
preview code
|
raw
history blame contribute delete
5.27 kB
metadata
library_name: transformers
base_model: Fsoft-AIC/videberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: videberta-base-ner-ghtk-hirach_NER-first_1000_data-3090-15Nov
    results: []

videberta-base-ner-ghtk-hirach_NER-first_1000_data-3090-15Nov

This model is a fine-tuned version of Fsoft-AIC/videberta-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0921
  • Accuracy: 0.9816
  • F1: 0.0426
  • Precision: 0.25
  • Recall: 0.0233

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: 2.5e-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: 40

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Precision Recall
No log 1.0 250 0.0910 0.9825 0.0 0.0 0.0
0.1567 2.0 500 0.0949 0.9825 0.0 0.0 0.0
0.1567 3.0 750 0.0959 0.9825 0.0 0.0 0.0
0.0772 4.0 1000 0.0962 0.9825 0.0 0.0 0.0
0.0772 5.0 1250 0.0975 0.9825 0.0 0.0 0.0
0.0767 6.0 1500 0.0969 0.9825 0.0 0.0 0.0
0.0767 7.0 1750 0.0984 0.9825 0.0 0.0 0.0
0.0758 8.0 2000 0.0966 0.9825 0.0 0.0 0.0
0.0758 9.0 2250 0.0960 0.9825 0.0 0.0 0.0
0.0739 10.0 2500 0.0955 0.9825 0.0 0.0 0.0
0.0739 11.0 2750 0.0958 0.9825 0.0 0.0 0.0
0.0711 12.0 3000 0.0940 0.9825 0.0 0.0 0.0
0.0711 13.0 3250 0.0942 0.9820 0.0 0.0 0.0
0.0672 14.0 3500 0.0958 0.9825 0.0 0.0 0.0
0.0672 15.0 3750 0.0943 0.9825 0.0851 0.5 0.0465
0.0639 16.0 4000 0.0926 0.9829 0.0455 1.0 0.0233
0.0639 17.0 4250 0.0964 0.9820 0.0435 0.3333 0.0233
0.0611 18.0 4500 0.0970 0.9820 0.0435 0.3333 0.0233
0.0611 19.0 4750 0.0969 0.9825 0.0444 0.5 0.0233
0.058 20.0 5000 0.0952 0.9820 0.0435 0.3333 0.0233
0.058 21.0 5250 0.0950 0.9820 0.0435 0.3333 0.0233
0.0547 22.0 5500 0.0954 0.9816 0.0426 0.25 0.0233
0.0547 23.0 5750 0.0963 0.9816 0.0426 0.25 0.0233
0.0525 24.0 6000 0.0946 0.9820 0.0435 0.3333 0.0233
0.0525 25.0 6250 0.0942 0.9820 0.0435 0.3333 0.0233
0.0502 26.0 6500 0.0909 0.9825 0.0444 0.5 0.0233
0.0502 27.0 6750 0.0958 0.9816 0.0426 0.25 0.0233
0.048 28.0 7000 0.0934 0.9808 0.0408 0.1667 0.0233
0.048 29.0 7250 0.0946 0.9804 0.04 0.1429 0.0233
0.0458 30.0 7500 0.0938 0.9808 0.0408 0.1667 0.0233
0.0458 31.0 7750 0.0913 0.9829 0.0870 0.6667 0.0465
0.044 32.0 8000 0.0913 0.9816 0.0426 0.25 0.0233
0.044 33.0 8250 0.0915 0.9808 0.0408 0.1667 0.0233
0.0427 34.0 8500 0.0924 0.9808 0.0408 0.1667 0.0233
0.0427 35.0 8750 0.0912 0.9812 0.0417 0.2 0.0233
0.041 36.0 9000 0.0922 0.9808 0.0408 0.1667 0.0233
0.041 37.0 9250 0.0933 0.9812 0.0417 0.2 0.0233
0.0404 38.0 9500 0.0929 0.9812 0.0417 0.2 0.0233
0.0404 39.0 9750 0.0922 0.9816 0.0426 0.25 0.0233
0.0401 40.0 10000 0.0921 0.9816 0.0426 0.25 0.0233

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.1.0
  • Tokenizers 0.19.1