search

iLampard / lamp

PyTorch implementation of the paper "Language Models Can Improve Event Prediction by Few-Shot Abductive Reasoning", NeurIPS 2023
Apache License 2.0
47 stars 6 forks source link

Questions about the third phase of training #8

Closed lshy687 closed 1 month ago

lshy687 commented 2 months ago

Hello, I have a question regarding Step 3. Is the training data in this step derived from the results of the initial test set after inference by the LLM? Thank you for your clarification.

iLampard commented 1 month ago

Hi,

the training data at step 3 is still derived from the train set in step 1.