Giters

mojivalipour / symbolicgpt

Symbolic regression is the task of identifying a mathematical expression that best fits a provided dataset of input and output values. In this work, we present SymbolicGPT, a novel transformer-based language model for symbolic regression.

Home Page:https://git.uwaterloo.ca/data-analytics-lab/symbolicgpt2

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Reproduce the results

Pengwei-Jin opened this issue · comments

commented

Hi, I tried to reproduce the results for 2 variables model but got a worse result shown in the image
XYE_2Var_200-201Points_512EmbeddingSize_SymbolicGPT_GPT_PT_EMB_SUM_Skeleton_Padding_NOT_VAR_MINIMIZE
. I just downloaded the datasets and retrained the model using parameters provided in README. So, is there any detail I haven't noticed?

commented

Please let me know more about the config of the computer that you trained with such as GPU etc. Did you train the model for 20 epochs?

commented

Hi, I retrained the model for 20 epochs twice and both got the same results with you. Maybe something unexpected occurred for the first time. Thank you for your reply. : )