metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - ecommerce_reviews_with_language_drift
metrics:
  - accuracy
  - f1
model-index:
  - name: distilbert_reviews_with_language_drift
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: ecommerce_reviews_with_language_drift
          type: ecommerce_reviews_with_language_drift
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.818
          - name: F1
            type: f1
            value: 0.8167126877417763
widget:
  - text: >-
      Poor quality of fabric and ridiculously tight at chest. It's way too
      short.
    example_title: Negative
  - text: >-
      One worked perfectly, but the other one has a slight leak and we end up
      with water underneath the filter.
    example_title: Neutral
  - text: I liked the price most! Nothing to dislike here!
    example_title: Positive

distilbert_reviews_with_language_drift

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

Loss: 0.4970
Accuracy: 0.818
F1: 0.8167

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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.593	1.0	500	0.4723	0.799	0.7976
0.3714	2.0	1000	0.4679	0.818	0.8177
0.2652	3.0	1500	0.4970	0.818	0.8167

Framework versions

Transformers 4.19.2
Pytorch 1.11.0+cu113
Datasets 2.2.2
Tokenizers 0.12.1