ctbert_climate_step2_three_category

This model is a fine-tuned version of Teapotboy/ctbert_climate_step1 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6038
  • Accuracy: 0.805
  • Precision: 0.8034
  • Recall: 0.805
  • F1-weighted: 0.8008
  • F1: 0.7798

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: 1e-05
  • train_batch_size: 128
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.05
  • num_epochs: 16

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1-weighted F1
1.0552 1.0 13 1.0037 0.5925 0.4696 0.5925 0.5014 0.4092
0.9542 2.0 26 0.7901 0.6475 0.7247 0.6475 0.5755 0.5004
0.8064 3.0 39 0.6620 0.7125 0.7298 0.7125 0.6807 0.6323
0.5555 4.0 52 0.6044 0.7425 0.7416 0.7425 0.7298 0.6985
0.4958 5.0 65 0.6352 0.7425 0.7548 0.7425 0.7238 0.6840
0.4388 6.0 78 0.6029 0.7675 0.7701 0.7675 0.7563 0.7249
0.3266 7.0 91 0.6036 0.775 0.7782 0.775 0.7623 0.7326
0.2613 8.0 104 0.6038 0.805 0.8034 0.805 0.8008 0.7798

Framework versions

  • Transformers 4.28.1
  • Pytorch 2.0.0+cu118
  • Datasets 2.12.0
  • Tokenizers 0.13.3
Downloads last month
8
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.