Hugging Face's logo

haryoaw
/
scenario-non-kd-scr-ner-half-mdeberta_data-univner_full44

Token Classification
Transformers
Safetensors
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Community
scenario-non-kd-scr-ner-half-mdeberta_data-univner_full44 / README.md
haryoaw's picture
haryoaw
Upload tokenizer
3f000b0 verified
preview code
|
raw
history blame
7.74 kB
metadata
base_model: microsoft/mdeberta-v3-base
library_name: transformers
license: mit
metrics:
  - precision
  - recall
  - f1
  - accuracy
tags:
  - generated_from_trainer
model-index:
  - name: scenario-non-kd-scr-ner-half-mdeberta_data-univner_full44
    results: []

scenario-non-kd-scr-ner-half-mdeberta_data-univner_full44

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3227
  • Precision: 0.6152
  • Recall: 0.5804
  • F1: 0.5973
  • Accuracy: 0.9617

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: 3e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 44
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.357 0.2910 500 0.2880 0.3006 0.1147 0.1660 0.9285
0.2425 0.5821 1000 0.2106 0.3468 0.2782 0.3087 0.9383
0.1767 0.8731 1500 0.1785 0.4322 0.3857 0.4076 0.9469
0.1352 1.1641 2000 0.1621 0.4749 0.4745 0.4747 0.9520
0.1107 1.4552 2500 0.1556 0.5238 0.4991 0.5111 0.9553
0.1031 1.7462 3000 0.1480 0.5536 0.5207 0.5367 0.9576
0.0912 2.0373 3500 0.1435 0.5286 0.5578 0.5428 0.9579
0.0661 2.3283 4000 0.1496 0.5510 0.5698 0.5602 0.9589
0.066 2.6193 4500 0.1502 0.5587 0.5742 0.5663 0.9594
0.0646 2.9104 5000 0.1440 0.5779 0.5803 0.5791 0.9609
0.0492 3.2014 5500 0.1590 0.5898 0.5656 0.5774 0.9608
0.0428 3.4924 6000 0.1613 0.5819 0.5634 0.5725 0.9603
0.0447 3.7835 6500 0.1602 0.5970 0.5742 0.5854 0.9615
0.0407 4.0745 7000 0.1667 0.5744 0.5995 0.5867 0.9611
0.0311 4.3655 7500 0.1762 0.5897 0.5754 0.5824 0.9610
0.0308 4.6566 8000 0.1707 0.5928 0.5862 0.5895 0.9609
0.0303 4.9476 8500 0.1717 0.5882 0.5915 0.5899 0.9610
0.0217 5.2386 9000 0.1826 0.5808 0.6025 0.5915 0.9611
0.0212 5.5297 9500 0.1827 0.5949 0.6006 0.5977 0.9613
0.0228 5.8207 10000 0.1942 0.5760 0.5809 0.5784 0.9601
0.02 6.1118 10500 0.1973 0.5982 0.5913 0.5947 0.9611
0.0146 6.4028 11000 0.2058 0.5938 0.5871 0.5904 0.9608
0.0161 6.6938 11500 0.2025 0.5973 0.5878 0.5925 0.9612
0.0166 6.9849 12000 0.2053 0.5972 0.5921 0.5947 0.9613
0.0115 7.2759 12500 0.2259 0.6083 0.5601 0.5832 0.9609
0.0116 7.5669 13000 0.2133 0.5944 0.6029 0.5986 0.9608
0.0114 7.8580 13500 0.2208 0.5883 0.5973 0.5928 0.9608
0.0098 8.1490 14000 0.2363 0.6118 0.5745 0.5926 0.9611
0.0084 8.4400 14500 0.2387 0.6094 0.5748 0.5916 0.9611
0.0097 8.7311 15000 0.2285 0.5819 0.5998 0.5907 0.9602
0.0083 9.0221 15500 0.2402 0.5992 0.5806 0.5897 0.9610
0.0064 9.3132 16000 0.2456 0.6297 0.5679 0.5972 0.9617
0.0068 9.6042 16500 0.2487 0.6035 0.5752 0.5890 0.9607
0.0072 9.8952 17000 0.2403 0.5910 0.6009 0.5959 0.9610
0.0062 10.1863 17500 0.2465 0.5981 0.5972 0.5976 0.9615
0.0045 10.4773 18000 0.2562 0.6062 0.5776 0.5915 0.9611
0.0055 10.7683 18500 0.2542 0.6139 0.5826 0.5978 0.9615
0.0054 11.0594 19000 0.2596 0.6128 0.5807 0.5963 0.9616
0.0037 11.3504 19500 0.2631 0.5872 0.6015 0.5943 0.9607
0.0048 11.6414 20000 0.2613 0.5998 0.6012 0.6005 0.9615
0.004 11.9325 20500 0.2576 0.6108 0.5892 0.5998 0.9616
0.0042 12.2235 21000 0.2647 0.5943 0.6027 0.5984 0.9611
0.0029 12.5146 21500 0.2773 0.6058 0.5819 0.5936 0.9613
0.0037 12.8056 22000 0.2785 0.6111 0.5874 0.5990 0.9612
0.0031 13.0966 22500 0.2819 0.6281 0.5790 0.6026 0.9618
0.0029 13.3877 23000 0.2794 0.6002 0.5915 0.5958 0.9609
0.0024 13.6787 23500 0.2842 0.6017 0.6019 0.6018 0.9615
0.0034 13.9697 24000 0.2889 0.6133 0.5806 0.5965 0.9616
0.0021 14.2608 24500 0.2876 0.6133 0.5803 0.5963 0.9616
0.0025 14.5518 25000 0.2871 0.6130 0.5845 0.5984 0.9614
0.0027 14.8428 25500 0.2921 0.6087 0.5835 0.5958 0.9613
0.0021 15.1339 26000 0.2888 0.5822 0.5998 0.5909 0.9607
0.0017 15.4249 26500 0.2899 0.6095 0.5911 0.6002 0.9613
0.0026 15.7159 27000 0.2968 0.6065 0.5839 0.5950 0.9613
0.002 16.0070 27500 0.3023 0.6158 0.5752 0.5949 0.9614
0.0015 16.2980 28000 0.2988 0.6006 0.5954 0.5980 0.9614
0.002 16.5891 28500 0.2983 0.5905 0.6045 0.5974 0.9611
0.0017 16.8801 29000 0.3006 0.6080 0.5838 0.5956 0.9614
0.0016 17.1711 29500 0.3078 0.5986 0.5921 0.5953 0.9612
0.0016 17.4622 30000 0.3066 0.6084 0.5892 0.5987 0.9617
0.0015 17.7532 30500 0.3153 0.6110 0.5786 0.5943 0.9617
0.0015 18.0442 31000 0.3134 0.5952 0.5954 0.5953 0.9611
0.0009 18.3353 31500 0.3201 0.6045 0.5904 0.5974 0.9615
0.0017 18.6263 32000 0.3149 0.6095 0.5875 0.5983 0.9614
0.0014 18.9173 32500 0.3227 0.6152 0.5804 0.5973 0.9617

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.1.1+cu121
  • Datasets 2.14.5
  • Tokenizers 0.19.1