Hugging Face's logo

haryoaw
/
scenario-MDBT-TCR-MSV-EN

PyTorch
deberta-v2
Generated from Trainer
Eval Results
Model card Files Community
scenario-MDBT-TCR-MSV-EN / README.md
haryoaw's picture
haryoaw
Initial Commit
a6319fb verified
preview code
|
raw
history blame
9.54 kB
metadata
license: mit
base_model: microsoft/mdeberta-v3-base
tags:
  - generated_from_trainer
datasets:
  - massive
metrics:
  - accuracy
  - f1
model-index:
  - name: scenario-MDBT-TCR_data-en-massive_all_1_1
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: massive
          type: massive
          config: all_1.1
          split: validation
          args: all_1.1
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.7227317882391521
          - name: F1
            type: f1
            value: 0.6670992426180887

scenario-MDBT-TCR_data-en-massive_all_1_1

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the massive dataset. It achieves the following results on the evaluation set:

  • Loss: 2.6914
  • Accuracy: 0.7227
  • F1: 0.6671

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: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 64
  • seed: 66
  • 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 Accuracy F1
No log 0.28 100 3.1171 0.2852 0.0691
No log 0.56 200 2.3001 0.4341 0.1961
No log 0.83 300 1.7494 0.5860 0.3648
No log 1.11 400 1.5526 0.6387 0.4610
1.995 1.39 500 1.5531 0.6500 0.4780
1.995 1.67 600 1.4151 0.6671 0.5333
1.995 1.94 700 1.2962 0.6946 0.5785
1.995 2.22 800 1.3865 0.6875 0.5773
1.995 2.5 900 1.2868 0.7121 0.6082
0.6196 2.78 1000 1.3864 0.6981 0.6033
0.6196 3.06 1100 1.4551 0.6925 0.6229
0.6196 3.33 1200 1.4319 0.7092 0.6216
0.6196 3.61 1300 1.4668 0.7035 0.6309
0.6196 3.89 1400 1.4418 0.7056 0.6303
0.347 4.17 1500 1.4875 0.7108 0.6562
0.347 4.44 1600 1.4943 0.7144 0.6564
0.347 4.72 1700 1.5156 0.7122 0.6407
0.347 5.0 1800 1.5642 0.7013 0.6506
0.347 5.28 1900 1.5904 0.7112 0.6440
0.2195 5.56 2000 1.5237 0.7239 0.6596
0.2195 5.83 2100 1.6728 0.7064 0.6285
0.2195 6.11 2200 1.6606 0.7026 0.6457
0.2195 6.39 2300 1.6961 0.7117 0.6461
0.2195 6.67 2400 1.7144 0.7088 0.6451
0.1729 6.94 2500 1.6841 0.7148 0.6585
0.1729 7.22 2600 1.8309 0.7057 0.6420
0.1729 7.5 2700 1.7698 0.7197 0.6580
0.1729 7.78 2800 1.9600 0.7069 0.6430
0.1729 8.06 2900 2.0215 0.6836 0.6281
0.113 8.33 3000 1.8546 0.7191 0.6600
0.113 8.61 3100 1.9063 0.7190 0.6593
0.113 8.89 3200 1.7990 0.7263 0.6578
0.113 9.17 3300 1.8465 0.7215 0.6613
0.113 9.44 3400 1.9787 0.7133 0.6522
0.0826 9.72 3500 1.9424 0.7168 0.6593
0.0826 10.0 3600 2.1079 0.6973 0.6399
0.0826 10.28 3700 2.0101 0.7081 0.6510
0.0826 10.56 3800 2.1830 0.6990 0.6307
0.0826 10.83 3900 2.1300 0.7112 0.6541
0.066 11.11 4000 2.0432 0.7118 0.6480
0.066 11.39 4100 2.2643 0.7005 0.6312
0.066 11.67 4200 2.3124 0.7056 0.6504
0.066 11.94 4300 2.1704 0.7169 0.6606
0.066 12.22 4400 2.1669 0.7244 0.6668
0.0465 12.5 4500 2.0924 0.7187 0.6566
0.0465 12.78 4600 2.1401 0.7192 0.6520
0.0465 13.06 4700 2.1376 0.7233 0.6552
0.0465 13.33 4800 2.1814 0.7246 0.6625
0.0465 13.61 4900 2.1595 0.7232 0.6618
0.0321 13.89 5000 2.2037 0.7299 0.6757
0.0321 14.17 5100 2.2631 0.7220 0.6736
0.0321 14.44 5200 2.3036 0.7178 0.6608
0.0321 14.72 5300 2.4098 0.7164 0.6625
0.0321 15.0 5400 2.3241 0.7177 0.6615
0.0238 15.28 5500 2.4564 0.7105 0.6606
0.0238 15.56 5600 2.3782 0.7208 0.6666
0.0238 15.83 5700 2.3832 0.7189 0.6591
0.0238 16.11 5800 2.5115 0.7075 0.6452
0.0238 16.39 5900 2.4870 0.7112 0.6640
0.0208 16.67 6000 2.5268 0.7145 0.6636
0.0208 16.94 6100 2.5253 0.7134 0.6641
0.0208 17.22 6200 2.4308 0.7233 0.6696
0.0208 17.5 6300 2.4632 0.7177 0.6668
0.0208 17.78 6400 2.3885 0.7253 0.6665
0.0169 18.06 6500 2.4380 0.7187 0.6631
0.0169 18.33 6600 2.4620 0.7163 0.6681
0.0169 18.61 6700 2.4921 0.7195 0.6646
0.0169 18.89 6800 2.5746 0.7087 0.6474
0.0169 19.17 6900 2.5031 0.7201 0.6645
0.0139 19.44 7000 2.5396 0.7183 0.6579
0.0139 19.72 7100 2.5645 0.7191 0.6635
0.0139 20.0 7200 2.5458 0.7184 0.6614
0.0139 20.28 7300 2.5119 0.7210 0.6663
0.0139 20.56 7400 2.5254 0.7257 0.6752
0.0079 20.83 7500 2.5765 0.7198 0.6709
0.0079 21.11 7600 2.5612 0.7203 0.6703
0.0079 21.39 7700 2.5182 0.7278 0.6719
0.0079 21.67 7800 2.5369 0.7247 0.6711
0.0079 21.94 7900 2.6488 0.7208 0.6681
0.0045 22.22 8000 2.6237 0.7245 0.6726
0.0045 22.5 8100 2.5783 0.7243 0.6722
0.0045 22.78 8200 2.6651 0.7209 0.6738
0.0045 23.06 8300 2.5498 0.7253 0.6717
0.0045 23.33 8400 2.6436 0.7233 0.6687
0.0056 23.61 8500 2.6572 0.7245 0.6710
0.0056 23.89 8600 2.8399 0.7147 0.6647
0.0056 24.17 8700 2.7875 0.7161 0.6682
0.0056 24.44 8800 2.7095 0.7195 0.6669
0.0056 24.72 8900 2.6328 0.7248 0.6688
0.0056 25.0 9000 2.6524 0.7246 0.6693
0.0056 25.28 9100 2.6860 0.7219 0.6685
0.0056 25.56 9200 2.7291 0.7194 0.6671
0.0056 25.83 9300 2.7558 0.7164 0.6625
0.0056 26.11 9400 2.7021 0.7185 0.6636
0.0023 26.39 9500 2.7087 0.7200 0.6650
0.0023 26.67 9600 2.7187 0.7199 0.6688
0.0023 26.94 9700 2.6568 0.7241 0.6720
0.0023 27.22 9800 2.6873 0.7213 0.6675
0.0023 27.5 9900 2.7043 0.7205 0.6667
0.0024 27.78 10000 2.7342 0.7178 0.6662
0.0024 28.06 10100 2.7089 0.7202 0.6673
0.0024 28.33 10200 2.7063 0.7207 0.6674
0.0024 28.61 10300 2.7048 0.7208 0.6671
0.0024 28.89 10400 2.7010 0.7214 0.6674
0.0015 29.17 10500 2.6951 0.7226 0.6670
0.0015 29.44 10600 2.6964 0.7223 0.6669
0.0015 29.72 10700 2.6925 0.7225 0.6671
0.0015 30.0 10800 2.6914 0.7227 0.6671

Framework versions

  • Transformers 4.33.3
  • Pytorch 2.1.1+cu121
  • Datasets 2.14.5
  • Tokenizers 0.13.3