bertweet-base / README.md
dtorber's picture
Model save
f54ee5c verified
|
raw
history blame
1.89 kB
metadata
base_model: vinai/bertweet-base
tags:
  - generated_from_trainer
metrics:
  - f1
  - recall
model-index:
  - name: bertweet-base
    results: []

bertweet-base

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

  • Loss: 0.4555
  • F1 Macro: 0.8329
  • F1: 0.8864
  • F1 Neg: 0.7794
  • Acc: 0.85
  • Prec: 0.9035
  • Recall: 0.8699
  • Mcc: 0.6670

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: 8
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss F1 Macro F1 F1 Neg Acc Prec Recall Mcc
No log 1.0 497 0.4926 0.7671 0.8270 0.7071 0.7825 0.8421 0.8125 0.5350
0.6271 2.0 994 0.4256 0.7808 0.8608 0.7008 0.81 0.8103 0.9180 0.5762
0.4307 3.0 1491 0.4918 0.8193 0.8726 0.7660 0.835 0.8626 0.8828 0.6390

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2