scripting_documentation/RosettaScripts/xsd/to_EnableSmartAnnealer

Autogenerated Tag Syntax Documentation:

The smart annealer uses tensorflow to decrease the sample space of a typical packing run

<EnableSmartAnnealer name="(&string;)" model="(generation2 &string;)"
        cutoff="(0.25 &real;)" pick_again="(true &bool;)"
        disable_during_quench="(true &bool;)" />

model: Choose which neural network to use for the smart annealer. Look at database/protocol_data/tensorflow_graphs/smart_annealer/ to see the options.
cutoff: Choose a number from 0 to 1 to tune how aggressive the smart annealer is. Higher numbers are more agressive (risky) but have a potentially greater speedup (speedup requires pick_again=false)
pick_again: f disabled, the smart annealer just skips unfruitful amino acids. Enabling this option tells the annealer to pick a fruitful rotamer to sample this round instead of skipping the round. Will not give you a speedup but may give you a better final outcome.
disable_during_quench: Run the final quenching stage as normal, regardless of how bad an amino acid may be.

scripting_documentation/RosettaScripts/xsd/to_EnableSmartAnnealer_type