- Open the site in your browser.
- Add training examples:
- In Example text, type a short sentence (e.g.,
This is great!).
- In Label, type the category (e.g.,
positive or negative).
- Click Add example. Repeat—aim for 5–10 examples per label.
- Train the model:
- Go to “2) Train the classifier” and click Train model.
- Wait until it says Done.
- Predict:
- Type a new sentence in Text to classify.
- Click Predict label to see the result and confidence.
- Save / Load:
- Download CSV saves your examples (headers:
text,label).
- Upload CSV loads examples back in.
Quick Tips
- If Add example is gray, make sure both text and label have content.
- Use consistent labels (e.g., always
positive and negative).
- More balanced examples → better model.
Troubleshooting
- Can’t train? You need at least 2 labels and 2+ examples total.
- Dropdown not showing? You can type labels manually.
- Weird predictions? Add more training examples and retrain.
🌎 Translate This Page
If you want to see the page in Spanish or another language, you can use Google Translate:
- Go to translate.google.com.
- In the left box, type or paste your website address (for example:
https://KJ-sv.github.io/Promptable-Machine/).
- Choose your language — for example, pick Spanish (Español).
- Click the translated link on the right — it will open your site automatically translated.
To go back to English, click the small “Show original” button at the top of the page.