README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - yelp_review_full
metrics:
  - accuracy
model-index:
  - name: 22s-dl-sentiment-1
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: yelp_review_full
          type: yelp_review_full
          args: yelp_review_full
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9542333333333334

22s-dl-sentiment-1

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

• Loss: 0.2574
• Accuracy: 0.9542

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: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Framework versions

• Transformers 4.19.4
• Pytorch 1.11.0+cu113
• Datasets 2.3.0
• Tokenizers 0.12.1