Edit model card

xlnet-finetuned-socialmediatweet

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

  • Loss: 2.6923
  • Accuracy: 0.7130

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: 2e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.0161 1.0 58 2.4538 0.6821
0.0416 2.0 116 2.3751 0.6821
0.0294 3.0 174 2.4929 0.7068
0.031 4.0 232 2.5655 0.7037
0.0422 5.0 290 3.0881 0.6605
0.0751 6.0 348 2.6787 0.6883
0.0264 7.0 406 2.5283 0.7006
0.0123 8.0 464 2.5634 0.7006
0.0277 9.0 522 2.7127 0.6852
0.0448 10.0 580 2.6113 0.6759
0.0261 11.0 638 2.6640 0.6759
0.0111 12.0 696 2.6089 0.6914
0.0239 13.0 754 2.5785 0.6975
0.0255 14.0 812 2.6923 0.7130
0.0242 15.0 870 2.4704 0.7068
0.0131 16.0 928 2.6724 0.6667
0.0059 17.0 986 2.5554 0.7068
0.0066 18.0 1044 2.6696 0.6698
0.001 19.0 1102 2.5653 0.6883
0.0026 20.0 1160 2.5846 0.6883

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0
Downloads last month
4
Safetensors
Model size
117M params
Tensor type
F32
·

Finetuned from

Evaluation results