thusken's picture
End of training
d1b5066
|
raw
history blame
1.95 kB
metadata
license: cc-by-4.0
tags:
  - generated_from_trainer
model-index:
  - name: nb-bert-base-ctr-regression
    results: []

nb-bert-base-ctr-regression

This model is a fine-tuned version of NbAiLab/nb-bert-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0073
  • Mse: 0.0073

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: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Mse
0.0106 1.0 1103 0.0069 0.0069
0.0073 2.0 2206 0.0072 0.0072
0.0058 3.0 3309 0.0063 0.0063
0.0038 4.0 4412 0.0073 0.0073
0.0025 5.0 5515 0.0064 0.0064
0.0019 6.0 6618 0.0065 0.0065
0.0014 7.0 7721 0.0066 0.0066
0.0011 8.0 8824 0.0067 0.0067
0.0008 9.0 9927 0.0066 0.0066
0.0007 10.0 11030 0.0066 0.0066

Framework versions

  • Transformers 4.17.0
  • Pytorch 1.10.2+cu113
  • Datasets 1.18.4
  • Tokenizers 0.12.1