Amazing dataset.
Q1. Does anyone have an estimate of how many $ it costs to run ~265M DFT computations?
Q2. How about 1B, 10B, 100B and 1000B?
This might allow us to train GPT3 sized molecular models without overfitting.
Amazing dataset.
Q1. Does anyone have an estimate of how many $ it costs to run ~265M DFT computations?
Q2. How about 1B, 10B, 100B and 1000B?
This might allow us to train GPT3 sized molecular models without overfitting.