TweetEval_XLNET_5E / README.md
pig4431's picture
update model card README.md
4c08b2f
|
raw
history blame
9.19 kB
metadata
license: mit
tags:
  - generated_from_trainer
datasets:
  - tweet_eval
metrics:
  - accuracy
model-index:
  - name: TweetEval_XLNET_5E
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: tweet_eval
          type: tweet_eval
          config: sentiment
          split: train
          args: sentiment
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9333333333333333

TweetEval_XLNET_5E

This model is a fine-tuned version of xlnet-base-cased on the tweet_eval dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4591
  • Accuracy: 0.9333

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: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 5

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.5575 0.04 50 0.2675 0.9
0.4177 0.08 100 0.2193 0.9067
0.2911 0.12 150 0.2482 0.9
0.3503 0.16 200 0.2424 0.9
0.3412 0.2 250 0.1913 0.9267
0.2747 0.24 300 0.1783 0.92
0.2999 0.28 350 0.2495 0.9133
0.3141 0.32 400 0.2460 0.9
0.2935 0.37 450 0.2034 0.92
0.2619 0.41 500 0.2600 0.9067
0.2454 0.45 550 0.2178 0.92
0.2809 0.49 600 0.2254 0.9133
0.288 0.53 650 0.1849 0.92
0.2769 0.57 700 0.1896 0.9267
0.3079 0.61 750 0.2153 0.9133
0.2598 0.65 800 0.3279 0.9067
0.3149 0.69 850 0.1985 0.92
0.2872 0.73 900 0.1801 0.9333
0.2554 0.77 950 0.2023 0.9267
0.2645 0.81 1000 0.2208 0.9067
0.2509 0.85 1050 0.2012 0.9333
0.2404 0.89 1100 0.1995 0.9067
0.2361 0.93 1150 0.1808 0.9133
0.2298 0.97 1200 0.2226 0.9333
0.193 1.01 1250 0.2535 0.9267
0.1603 1.06 1300 0.2163 0.9467
0.1916 1.1 1350 0.2479 0.92
0.1963 1.14 1400 0.1964 0.94
0.1667 1.18 1450 0.3139 0.9133
0.1668 1.22 1500 0.2204 0.9267
0.1677 1.26 1550 0.2468 0.9333
0.1601 1.3 1600 0.2394 0.94
0.1714 1.34 1650 0.2326 0.94
0.197 1.38 1700 0.1861 0.94
0.1777 1.42 1750 0.2518 0.94
0.1925 1.46 1800 0.1806 0.94
0.2068 1.5 1850 0.1319 0.9467
0.1716 1.54 1900 0.1199 0.9667
0.1442 1.58 1950 0.1694 0.96
0.1929 1.62 2000 0.1990 0.9467
0.1654 1.66 2050 0.2972 0.9333
0.1759 1.7 2100 0.1584 0.9467
0.1788 1.75 2150 0.2266 0.94
0.1796 1.79 2200 0.2746 0.9333
0.172 1.83 2250 0.2313 0.9333
0.1637 1.87 2300 0.2918 0.9267
0.2359 1.91 2350 0.2121 0.9267
0.1778 1.95 2400 0.2022 0.9333
0.1581 1.99 2450 0.2936 0.9067
0.1312 2.03 2500 0.2531 0.9333
0.1178 2.07 2550 0.2525 0.9267
0.0924 2.11 2600 0.2715 0.9333
0.0774 2.15 2650 0.2123 0.9533
0.091 2.19 2700 0.2128 0.9467
0.0948 2.23 2750 0.2187 0.9533
0.1121 2.27 2800 0.2438 0.9467
0.1259 2.31 2850 0.2197 0.9467
0.0747 2.35 2900 0.2727 0.9333
0.114 2.39 2950 0.3197 0.9333
0.086 2.44 3000 0.3643 0.9333
0.1326 2.48 3050 0.2791 0.94
0.1017 2.52 3100 0.2661 0.9333
0.0719 2.56 3150 0.2797 0.94
0.1424 2.6 3200 0.1819 0.96
0.106 2.64 3250 0.2770 0.94
0.0996 2.68 3300 0.2213 0.94
0.0835 2.72 3350 0.2894 0.9333
0.0808 2.76 3400 0.3424 0.9333
0.1406 2.8 3450 0.2166 0.94
0.0345 2.84 3500 0.3146 0.9333
0.1247 2.88 3550 0.2824 0.9467
0.076 2.92 3600 0.2650 0.9467
0.134 2.96 3650 0.2758 0.9267
0.0521 3.0 3700 0.2693 0.9467
0.0366 3.04 3750 0.3428 0.9333
0.0682 3.08 3800 0.2779 0.9533
0.0624 3.12 3850 0.2563 0.9467
0.0402 3.17 3900 0.3086 0.94
0.052 3.21 3950 0.3324 0.94
0.0579 3.25 4000 0.3165 0.9467
0.0411 3.29 4050 0.3507 0.9467
0.0507 3.33 4100 0.3108 0.9533
0.0326 3.37 4150 0.3645 0.94
0.085 3.41 4200 0.3390 0.94
0.022 3.45 4250 0.3367 0.94
0.0689 3.49 4300 0.3433 0.94
0.0458 3.53 4350 0.3359 0.9533
0.0384 3.57 4400 0.3642 0.9467
0.0415 3.61 4450 0.3429 0.9467
0.0362 3.65 4500 0.3727 0.9467
0.0351 3.69 4550 0.3293 0.9467
0.06 3.73 4600 0.4717 0.92
0.0344 3.77 4650 0.3668 0.94
0.0518 3.81 4700 0.3461 0.94
0.046 3.86 4750 0.4020 0.9267
0.0735 3.9 4800 0.2660 0.9467
0.0453 3.94 4850 0.3364 0.9333
0.039 3.98 4900 0.4398 0.92
0.0497 4.02 4950 0.3476 0.94
0.0183 4.06 5000 0.3871 0.94
0.0558 4.1 5050 0.4066 0.9267
0.0358 4.14 5100 0.3926 0.92
0.0507 4.18 5150 0.3312 0.9467
0.0111 4.22 5200 0.3976 0.9267
0.0363 4.26 5250 0.4753 0.92
0.0283 4.3 5300 0.4234 0.9267
0.0097 4.34 5350 0.4547 0.9333
0.0018 4.38 5400 0.4687 0.9267
0.0344 4.42 5450 0.4274 0.9333
0.021 4.46 5500 0.4448 0.9333
0.0092 4.5 5550 0.4672 0.9333
0.0354 4.55 5600 0.4666 0.9333
0.029 4.59 5650 0.4614 0.9333
0.0182 4.63 5700 0.4840 0.9333
0.043 4.67 5750 0.4327 0.9333
0.0259 4.71 5800 0.4639 0.9333
0.0224 4.75 5850 0.4607 0.9333
0.0302 4.79 5900 0.4606 0.9333
0.0224 4.83 5950 0.4654 0.9333
0.0431 4.87 6000 0.4681 0.9333
0.0284 4.91 6050 0.4622 0.9333
0.0326 4.95 6100 0.4602 0.9333
0.018 4.99 6150 0.4591 0.9333

Framework versions

  • Transformers 4.24.0
  • Pytorch 1.13.0
  • Datasets 2.7.1
  • Tokenizers 0.13.2