Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  1. Select Action:

    Status
    titleCreate

  2. Select your dataset from the dropdown

  3. Choose a model type

    1. symanto_fast (for demonstration purposes, live demo, etc.)

    2. symanto (for actual model training with a purpose)

  4. Select or unselect the multi_label option (select it if the text classification task allows the prediction of multiple labels per text instance. Otherwise, the task is considered as multi-class (or binary if only 2 labels are categorized). 

  5. Select the language of the text

  6. Choose an embedding model

    • Choose symanto_brain_multilingual only if there is no specific language model for your selected langaugelanguage

  7. Select the number of labels for your analysis

...

  1. Provide label names, e.g.

    Status
    titlepositive
    ,
    Status
    titlenegative
    and a label texts, e.g.
    Status
    titlegreat
    ,
    Status
    titlebad
    . You can include your hypothesis in the label text too, such as
    Status
    titleThe text is great
    or
    Status
    titlethe text is bad

  2. Enable “Inspect typical examples” to see how your model performs

    1. Use the slider to select up to 10 examples.

    2. Change the label dropdown to see examples for every each of your labels.

  3. Click create‘Create’

...

If you want to continue using Promptranker

...

  1. Enable “Rank label texts”

  2. Provide label names, e.g.

    Status
    titlepositive
    ,
    Status
    titlenegative

    1. For each label name, provide a list of alternative descriptions. For example, for positive, you can provide

      Status
      titlegreat
      Status
      titlewonderful
      Status
      titlepositive
      and for negative,
      Status
      titlebad
      Status
      titleterrible
      Status
      titlenegative

  3. Provide hypothesis variations, e.g. use an empty one and add several different variations such as

    Status
    titleThis TExt is {}
    ,
    Status
    titleThe author expresses a {} sentiment
    , …

  4. Click “Rank”

  5. The result is a table where each variation is ranked according to a calculated score. The higher the score, the more likely the given labels and hypothesis yield good results in a real scenario.

  6. Enable “Inspect typical examples” to see how your model performs

    1. Use the slider to select up to 10 examples.

    2. Change the label dropdown to see examples for every each of your labels.

  7. Click create. The model is automatically being created with the prompts (labels and hypothesis) that scored the top.

...

Next:

...

K3. Train the model using active learning