jonasknobloch's picture
Update README.md
9d5d427 verified
metadata
tags:
  - generated_from_trainer
datasets:
  - uonlp/CulturaX
metrics:
  - accuracy
model-index:
  - name: gpt2_cx-cs_00000-00019_50k
    results:
      - task:
          name: Causal Language Modeling
          type: text-generation
        dataset:
          name: uonlp/CulturaX cs
          type: uonlp/CulturaX
          args: cs
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.38830943632679016
license: mit
language:
  - cs

gpt2_cx-cs_00000-00019_50k

This model is a fine-tuned version of on the uonlp/CulturaX cs dataset. It achieves the following results on the evaluation set:

  • Loss: 3.5060
  • Accuracy: 0.3883

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: 64
  • eval_batch_size: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 1.0

Training results

Training Loss Epoch Step Validation Loss Accuracy
4.729 0.04 10000 4.6077 0.2836
4.3383 0.07 20000 4.2318 0.3162
4.1706 0.11 30000 4.0651 0.3316
4.0594 0.15 40000 3.9599 0.3416
3.9842 0.19 50000 3.8825 0.3487
3.9298 0.22 60000 3.8244 0.3545
3.8777 0.26 70000 3.7791 0.3592
3.8455 0.3 80000 3.7436 0.3629
3.8104 0.33 90000 3.7120 0.3660
3.7908 0.37 100000 3.6862 0.3687
3.7613 0.41 110000 3.6628 0.3712
3.7492 0.45 120000 3.6434 0.3731
3.7228 0.48 130000 3.6246 0.3751
3.7127 0.52 140000 3.6090 0.3767
3.694 0.56 150000 3.5962 0.3783
3.6871 0.59 160000 3.5831 0.3797
3.6784 0.63 170000 3.5708 0.3810
3.6606 0.67 180000 3.5593 0.3823
3.646 0.71 190000 3.5491 0.3835
3.6453 0.74 200000 3.5410 0.3843
3.6393 0.78 210000 3.5342 0.3851
3.6207 0.82 220000 3.5280 0.3857
3.6288 0.86 230000 3.5218 0.3865
3.6176 0.89 240000 3.5151 0.3872
3.6099 0.93 250000 3.5108 0.3878
3.6093 0.97 260000 3.5079 0.3881

Framework versions

  • Transformers 4.37.1
  • Pytorch 2.1.2+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.1