Simon Says Prompt Optimization and Text Generation

This application allows you to optimize a Simon Says prompt based on your input text using advanced machine learning techniques. You can visualize the optimization process and generate text using the optimized prompt. Additionally, you can upload a pre-optimized tensor for direct inference (if you do, the other parameters will be ignored).

Input Text

Provide the text for which you want to optimize the Simon Says prompt. This text will be used as the target for generating the Simon Says Prompt.

Select Model

Choose a pre-trained language model to use for optimization and text generation. Each model has different capabilities and sizes.

Number of Simon Says Prompt Tokens

Specify the number of tokens to include in the Simon Says prompt. Bigger sizes may make it easier to optimize, but they take up more space.

1 10

Patience

Set the maximum number of steps for the optimization process. It will stop early if the optimization converges before reaching this number, but if it reaches the limit, it will stop without converging.

100 10000

Learning Rate

Adjust the learning rate for the optimization algorithm. This controls how quickly the optimization converges but can also lead to instability if set too high.

0.00001 0.1

Enable this option to append an End-Of-Sequence (EOS) token to the input text. This can help models better understand the input context.

Add EOS Token

Upload Optimized SS Prompt (Optional)

Generated Text

The text generated by the model using the optimized Simon Says prompt.

Download Optimized SS Prompt