Hugging Face's logo

haryoaw
/
scenario-non-kd-pre-ner-full-mdeberta_data-univner_en44

Token Classification
Transformers
Safetensors
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Community
scenario-non-kd-pre-ner-full-mdeberta_data-univner_en44 / README.md
haryoaw's picture
haryoaw
Initial Commit
a731de1 verified
preview code
|
raw
history blame
3.79 kB
metadata
library_name: transformers
license: mit
base_model: haryoaw/scenario-TCR-NER_data-univner_en
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: scenario-non-kd-pre-ner-full-mdeberta_data-univner_en44
    results: []

scenario-non-kd-pre-ner-full-mdeberta_data-univner_en44

This model is a fine-tuned version of haryoaw/scenario-TCR-NER_data-univner_en on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1490
  • Precision: 0.8349
  • Recall: 0.8271
  • F1: 0.8310
  • Accuracy: 0.9846

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
0.0023 1.2755 500 0.1120 0.8158 0.8344 0.8250 0.9851
0.0019 2.5510 1000 0.1055 0.7790 0.8468 0.8115 0.9840
0.0019 3.8265 1500 0.1115 0.8134 0.8395 0.8263 0.9847
0.0013 5.1020 2000 0.1248 0.7938 0.8251 0.8091 0.9835
0.0016 6.3776 2500 0.1281 0.8302 0.8147 0.8224 0.9845
0.0012 7.6531 3000 0.1241 0.8179 0.8137 0.8158 0.9841
0.001 8.9286 3500 0.1191 0.8184 0.8302 0.8243 0.9842
0.0005 10.2041 4000 0.1335 0.7951 0.8395 0.8167 0.9837
0.001 11.4796 4500 0.1376 0.8125 0.8344 0.8233 0.9843
0.0005 12.7551 5000 0.1282 0.8244 0.8313 0.8278 0.9847
0.0007 14.0306 5500 0.1247 0.8294 0.8354 0.8324 0.9848
0.0003 15.3061 6000 0.1440 0.8143 0.8354 0.8247 0.9841
0.0009 16.5816 6500 0.1456 0.8384 0.8219 0.8301 0.9840
0.0003 17.8571 7000 0.1401 0.8154 0.8416 0.8283 0.9849
0.0002 19.1327 7500 0.1478 0.8306 0.8323 0.8314 0.9850
0.0003 20.4082 8000 0.1448 0.8306 0.8271 0.8288 0.9850
0.0003 21.6837 8500 0.1416 0.8240 0.8385 0.8312 0.9847
0.0002 22.9592 9000 0.1501 0.8159 0.8395 0.8276 0.9843
0.0001 24.2347 9500 0.1457 0.8243 0.8354 0.8298 0.9845
0.0001 25.5102 10000 0.1474 0.8258 0.8344 0.8301 0.9846
0.0001 26.7857 10500 0.1493 0.8244 0.8406 0.8324 0.9847
0.0001 28.0612 11000 0.1497 0.8418 0.8209 0.8312 0.9847
0.0001 29.3367 11500 0.1490 0.8349 0.8271 0.8310 0.9846

Framework versions

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