ac0hik's picture
Update README.md
999d750
metadata
license: mit
base_model: camembert-base
tags:
  - generated_from_trainer
datasets:
  - tweet_sentiment_multilingual
metrics:
  - accuracy
model-index:
  - name: camembert_model
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: tweet_sentiment_multilingual
          type: tweet_sentiment_multilingual
          config: french
          split: validation
          args: french
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.7654320987654321

camembert_model

This model is a fine-tuned version of camembert-base on the tweet_sentiment_multilingual dataset (French portion of it) . It achieves the following results on the evaluation set:

  • Loss: 0.7877
  • Accuracy: 0.7654

Model description

A sentiment Classifier for the french language classifies french text to positive, negative or neutral.

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: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 115 0.8510 0.6265
No log 2.0 230 0.7627 0.7130
No log 3.0 345 0.6966 0.7160
No log 4.0 460 0.6862 0.7438
0.7126 5.0 575 0.6637 0.75
0.7126 6.0 690 0.7121 0.7654
0.7126 7.0 805 0.7641 0.7438
0.7126 8.0 920 0.7662 0.7654
0.2932 9.0 1035 0.7765 0.7747
0.2932 10.0 1150 0.7877 0.7654

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.15.0
  • Tokenizers 0.15.0