Hugging Face's logo

admko
/
sembr2023-distilbert-base-uncased

Token Classification
Transformers
PyTorch
distilbert
Generated from Trainer
Inference Endpoints
Model card Files Community
sembr2023-distilbert-base-uncased / README.md
admko's picture
admko
End of training
d6a4d88
preview code
|
raw
history blame
No virus
16.2 kB
metadata
license: apache-2.0
base_model: distilbert-base-uncased
tags:
  - generated_from_trainer
datasets:
  - sembr2023
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: sembr2023-distilbert-base-uncased
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: sembr2023
          type: sembr2023
          config: sembr2023
          split: test
          args: sembr2023
        metrics:
          - name: Precision
            type: precision
            value: 0.7588342440801458
          - name: Recall
            type: recall
            value: 0.8253863426231673
          - name: F1
            type: f1
            value: 0.790712387700873
          - name: Accuracy
            type: accuracy
            value: 0.9613411398987951

sembr2023-distilbert-base-uncased

This model is a fine-tuned version of distilbert-base-uncased on the sembr2023 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2562
  • Precision: 0.7588
  • Recall: 0.8254
  • F1: 0.7907
  • Iou: 0.6539
  • Accuracy: 0.9613
  • Balanced Accuracy: 0.9000
  • Overall Accuracy: 0.9468

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: 0.0001
  • train_batch_size: 64
  • eval_batch_size: 128
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • training_steps: 1000

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Iou Accuracy Balanced Accuracy Overall Accuracy
0.4048 0.07 10 0.4102 0 0.0 0.0 0.0 0.9115 0.5 0.9115
0.3834 0.14 20 0.3950 0 0.0 0.0 0.0 0.9115 0.5 0.9115
0.3817 0.21 30 0.3616 0 0.0 0.0 0.0 0.9115 0.5 0.9115
0.2982 0.28 40 0.3127 0 0.0 0.0 0.0 0.9115 0.5 0.9115
0.2505 0.35 50 0.2716 0.6064 0.7118 0.6549 0.4869 0.9336 0.8335 0.9149
0.2176 0.42 60 0.2119 0.8427 0.5653 0.6767 0.5114 0.9522 0.7775 0.9411
0.2018 0.49 70 0.2031 0.7785 0.7056 0.7402 0.5876 0.9562 0.8430 0.9391
0.1609 0.56 80 0.1825 0.8199 0.6992 0.7548 0.6061 0.9598 0.8422 0.9437
0.153 0.62 90 0.1820 0.7830 0.7362 0.7589 0.6115 0.9586 0.8582 0.9421
0.1336 0.69 100 0.1760 0.7955 0.7323 0.7626 0.6163 0.9597 0.8570 0.9447
0.1351 0.76 110 0.1691 0.8314 0.7243 0.7742 0.6316 0.9626 0.8550 0.9460
0.1393 0.83 120 0.1742 0.8028 0.7506 0.7758 0.6338 0.9616 0.8664 0.9454
0.1013 0.9 130 0.1657 0.7991 0.7608 0.7795 0.6387 0.9619 0.8711 0.9454
0.0969 0.97 140 0.1681 0.7926 0.7702 0.7812 0.6410 0.9618 0.8753 0.9461
0.1009 1.04 150 0.1624 0.8247 0.7475 0.7842 0.6450 0.9636 0.8660 0.9485
0.086 1.11 160 0.1532 0.8062 0.7793 0.7925 0.6564 0.9639 0.8806 0.9496
0.0732 1.18 170 0.1710 0.7633 0.8185 0.7899 0.6528 0.9615 0.8969 0.9465
0.0804 1.25 180 0.1731 0.8098 0.7735 0.7912 0.6546 0.9639 0.8779 0.9472
0.0865 1.32 190 0.1659 0.7905 0.7863 0.7884 0.6507 0.9627 0.8830 0.9486
0.0867 1.39 200 0.1716 0.7704 0.8082 0.7888 0.6513 0.9617 0.8924 0.9452
0.0769 1.46 210 0.1644 0.8122 0.7678 0.7894 0.6520 0.9637 0.8753 0.9486
0.0639 1.53 220 0.1668 0.7927 0.7786 0.7856 0.6469 0.9624 0.8794 0.9474
0.0642 1.6 230 0.1669 0.7960 0.7830 0.7895 0.6521 0.9630 0.8818 0.9482
0.0598 1.67 240 0.1776 0.7959 0.7802 0.7880 0.6501 0.9628 0.8804 0.9464
0.0679 1.74 250 0.1938 0.7436 0.8338 0.7861 0.6476 0.9599 0.9030 0.9423
0.0578 1.81 260 0.1669 0.7965 0.7855 0.7910 0.6542 0.9633 0.8830 0.9494
0.0587 1.88 270 0.1750 0.7982 0.7833 0.7907 0.6538 0.9633 0.8820 0.9468
0.0624 1.94 280 0.1759 0.7684 0.8249 0.7956 0.6606 0.9625 0.9004 0.9482
0.0494 2.01 290 0.1868 0.7606 0.8194 0.7889 0.6514 0.9612 0.8972 0.9447
0.0418 2.08 300 0.1746 0.8089 0.7769 0.7926 0.6564 0.9640 0.8795 0.9501
0.0406 2.15 310 0.1900 0.7778 0.8157 0.7963 0.6616 0.9631 0.8966 0.9477
0.0417 2.22 320 0.2066 0.7754 0.8165 0.7954 0.6603 0.9628 0.8968 0.9467
0.0438 2.29 330 0.2037 0.7623 0.8282 0.7939 0.6582 0.9619 0.9015 0.9461
0.0454 2.36 340 0.1882 0.7671 0.8221 0.7936 0.6579 0.9622 0.8989 0.9487
0.0372 2.43 350 0.1896 0.7937 0.8082 0.8009 0.6679 0.9644 0.8939 0.9489
0.0301 2.5 360 0.1846 0.8058 0.7913 0.7985 0.6646 0.9647 0.8864 0.9503
0.0334 2.57 370 0.1867 0.7723 0.8212 0.7960 0.6611 0.9628 0.8988 0.9498
0.038 2.64 380 0.2099 0.7538 0.8276 0.7890 0.6516 0.9608 0.9007 0.9463
0.0427 2.71 390 0.2050 0.7625 0.8272 0.7935 0.6577 0.9619 0.9011 0.9460
0.0373 2.78 400 0.2002 0.7779 0.8086 0.7930 0.6569 0.9626 0.8931 0.9487
0.0308 2.85 410 0.2002 0.7501 0.8260 0.7863 0.6478 0.9603 0.8997 0.9449
0.035 2.92 420 0.2034 0.7656 0.8173 0.7906 0.6537 0.9617 0.8965 0.9470
0.0265 2.99 430 0.1877 0.7766 0.8202 0.7978 0.6637 0.9632 0.8987 0.9487
0.0358 3.06 440 0.2140 0.7555 0.8300 0.7910 0.6542 0.9612 0.9020 0.9470
0.0286 3.12 450 0.2084 0.7775 0.8052 0.7911 0.6544 0.9624 0.8914 0.9487
0.0335 3.19 460 0.2084 0.7706 0.8144 0.7919 0.6555 0.9621 0.8954 0.9468
0.0304 3.26 470 0.2231 0.7506 0.8280 0.7874 0.6494 0.9604 0.9007 0.9458
0.0255 3.33 480 0.2194 0.7501 0.8348 0.7902 0.6532 0.9608 0.9039 0.9459
0.0209 3.4 490 0.2186 0.7602 0.8300 0.7936 0.6578 0.9618 0.9023 0.9469
0.0237 3.47 500 0.2138 0.7731 0.8196 0.7957 0.6607 0.9628 0.8981 0.9488
0.0263 3.54 510 0.2268 0.7565 0.8313 0.7921 0.6558 0.9614 0.9027 0.9472
0.0196 3.61 520 0.2157 0.7626 0.8242 0.7922 0.6559 0.9617 0.8996 0.9473
0.0225 3.68 530 0.2265 0.7664 0.8185 0.7916 0.6551 0.9619 0.8971 0.9469
0.0226 3.75 540 0.2316 0.7628 0.8139 0.7875 0.6495 0.9611 0.8947 0.9461
0.0226 3.82 550 0.2100 0.7812 0.8130 0.7968 0.6622 0.9633 0.8954 0.9493
0.024 3.89 560 0.2213 0.7670 0.8196 0.7924 0.6562 0.9620 0.8977 0.9487
0.0194 3.96 570 0.2227 0.7673 0.8164 0.7911 0.6544 0.9618 0.8962 0.9483
0.0153 4.03 580 0.2296 0.7621 0.8245 0.7921 0.6557 0.9617 0.8997 0.9462
0.0168 4.1 590 0.2362 0.7522 0.8237 0.7863 0.6479 0.9604 0.8987 0.9463
0.0215 4.17 600 0.2413 0.7629 0.8183 0.7896 0.6524 0.9614 0.8968 0.9470
0.0173 4.24 610 0.2411 0.7550 0.8196 0.7859 0.6474 0.9605 0.8969 0.9461
0.0196 4.31 620 0.2315 0.7625 0.8239 0.7920 0.6557 0.9617 0.8995 0.9471
0.0186 4.38 630 0.2593 0.7370 0.8377 0.7841 0.6449 0.9592 0.9043 0.9432
0.0183 4.44 640 0.2368 0.7660 0.8255 0.7947 0.6593 0.9623 0.9005 0.9477
0.0178 4.51 650 0.2326 0.7705 0.8237 0.7962 0.6614 0.9627 0.8999 0.9486
0.017 4.58 660 0.2457 0.7448 0.8352 0.7874 0.6493 0.9601 0.9037 0.9454
0.0175 4.65 670 0.2311 0.7694 0.8209 0.7943 0.6588 0.9624 0.8985 0.9485
0.0163 4.72 680 0.2426 0.7696 0.8208 0.7944 0.6589 0.9624 0.8985 0.9484
0.0163 4.79 690 0.2400 0.7527 0.8305 0.7897 0.6525 0.9609 0.9020 0.9462
0.0122 4.86 700 0.2328 0.7656 0.8208 0.7922 0.6559 0.9619 0.8982 0.9476
0.0205 4.93 710 0.2483 0.7474 0.8305 0.7868 0.6485 0.9602 0.9016 0.9452
0.0154 5.0 720 0.2350 0.7612 0.8206 0.7898 0.6526 0.9614 0.8978 0.9472
0.0155 5.07 730 0.2581 0.7570 0.8263 0.7901 0.6531 0.9612 0.9003 0.9463
0.0137 5.14 740 0.2482 0.7592 0.8242 0.7904 0.6534 0.9613 0.8994 0.9472
0.0124 5.21 750 0.2507 0.7639 0.8185 0.7903 0.6533 0.9616 0.8970 0.9466
0.0121 5.28 760 0.2509 0.7586 0.8229 0.7894 0.6521 0.9612 0.8987 0.9470
0.016 5.35 770 0.2552 0.7625 0.8173 0.7890 0.6515 0.9613 0.8963 0.9466
0.0131 5.42 780 0.2500 0.7573 0.8231 0.7889 0.6513 0.9610 0.8988 0.9467
0.0111 5.49 790 0.2647 0.7496 0.8312 0.7883 0.6506 0.9605 0.9021 0.9453
0.0113 5.56 800 0.2529 0.7646 0.8226 0.7926 0.6564 0.9619 0.8990 0.9475
0.0141 5.62 810 0.2521 0.7623 0.8230 0.7915 0.6549 0.9616 0.8990 0.9474
0.0123 5.69 820 0.2558 0.7523 0.8282 0.7884 0.6508 0.9607 0.9008 0.9458
0.0113 5.76 830 0.2560 0.7585 0.8208 0.7884 0.6507 0.9610 0.8977 0.9465
0.0176 5.83 840 0.2560 0.7553 0.8263 0.7892 0.6518 0.9609 0.9002 0.9466
0.0137 5.9 850 0.2539 0.7628 0.8233 0.7919 0.6555 0.9617 0.8992 0.9471
0.0159 5.97 860 0.2574 0.7574 0.8246 0.7896 0.6523 0.9611 0.8995 0.9464
0.0107 6.04 870 0.2573 0.7561 0.8257 0.7894 0.6520 0.9610 0.8999 0.9464
0.01 6.11 880 0.2562 0.7633 0.8204 0.7908 0.6540 0.9616 0.8978 0.9471
0.0112 6.18 890 0.2574 0.7618 0.8218 0.7907 0.6538 0.9615 0.8984 0.9469
0.0111 6.25 900 0.2563 0.7598 0.8239 0.7906 0.6537 0.9614 0.8993 0.9469
0.0135 6.32 910 0.2558 0.7578 0.8263 0.7905 0.6536 0.9613 0.9003 0.9468
0.0171 6.39 920 0.2554 0.7601 0.8251 0.7913 0.6546 0.9615 0.8999 0.9470
0.0115 6.46 930 0.2563 0.7602 0.8254 0.7915 0.6549 0.9615 0.9001 0.9469
0.0127 6.53 940 0.2561 0.7591 0.8258 0.7910 0.6543 0.9614 0.9002 0.9468
0.0107 6.6 950 0.2563 0.7584 0.8258 0.7906 0.6538 0.9613 0.9001 0.9467
0.0129 6.67 960 0.2567 0.7579 0.8259 0.7905 0.6535 0.9613 0.9002 0.9467
0.012 6.74 970 0.2567 0.7580 0.8257 0.7904 0.6534 0.9612 0.9000 0.9467
0.0141 6.81 980 0.2563 0.7587 0.8254 0.7907 0.6538 0.9613 0.9000 0.9468
0.0116 6.88 990 0.2562 0.7588 0.8254 0.7907 0.6539 0.9613 0.9000 0.9468
0.0097 6.94 1000 0.2562 0.7588 0.8254 0.7907 0.6539 0.9613 0.9000 0.9468

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.0.1
  • Datasets 2.14.6
  • Tokenizers 0.14.1