Kyle1668's picture
Model save
e78d326
metadata
license: apache-2.0
base_model: bert-base-uncased
tags:
  - generated_from_trainer
datasets:
  - ag_news
metrics:
  - f1
model-index:
  - name: ag-news-twitter-38400-bert-base-uncased
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: ag_news
          type: ag_news
          config: default
          split: test
          args: default
        metrics:
          - name: F1
            type: f1
            value: 0.9306013073265345

ag-news-twitter-38400-bert-base-uncased

This model is a fine-tuned version of bert-base-uncased on the ag_news dataset. It achieves the following results on the evaluation set:

  • F1: 0.9306
  • Acc: 0.9308
  • Loss: 0.4724

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: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 20

Training results

Training Loss Epoch Step F1 Acc Validation Loss
0.2511 1.0 2400 0.9219 0.9222 0.2514
0.2172 2.0 4800 0.9259 0.9261 0.2275
0.1631 3.0 7200 0.9375 0.9374 0.2335
0.1118 4.0 9600 0.9276 0.9275 0.3648
0.0661 5.0 12000 0.9280 0.9280 0.3757
0.0548 6.0 14400 0.9306 0.9308 0.4724

Framework versions

  • Transformers 4.35.0
  • Pytorch 2.1.0+cu121
  • Datasets 2.14.6
  • Tokenizers 0.14.1