Update run time limit based on DMRG and SHCI

[jtcantin]

To address issue #50 "validate run time limit for all problem instances/tasks"

All files pass json validation.

Run times updated based on the following:

Coarse set:

• Cr dimer: 24 hrs (significantly based on source paper calculation times)
• Benzene: 24 hrs (significantly based on source paper calculation times)
• Ozone: 24 hrs (significantly based on source paper calculation times)
• Hubbard instances: 2 hrs (these run rather fast for both SHCI and DMRG)
• Mn mono: see catalysts below

Small molecules and transition metals:

• With reference energy: larger of DMRG and SHCI runtimes
• Without reference energy: 16 hrs (longest DMRG or SHCI time spent on one of these so far)

Catalysts:

• With reference energy: larger of DMRG and SHCI runtimes
• Without reference energy: 24 hrs

Planted solutions: same as source Hamiltonian

Note: The runtime is not allowed to be less than two hours and all times are rounded up to the nearest hour.

@jp7745 @bellonzi @0tt3r: What are your thoughts on this?

[jtcantin]

Note: The runtime is not allowed to be less than two hours and all times are rounded up to the nearest hour.

[0tt3r]

I think having everything be consistently one number - 24 hours - seems more reasonable to me.

[jtcantin]

Having one number for everything was my initial thought as well.
My concern with that, though, would be the smaller systems (such those with 6 to 12 orbitals). For example, would it make sense to have CO2, defined in a six orbital basis, have a runtime limit of 24hrs?

[jp7745]

I think the run time limit should basically be set as what can be achieved in about (state_of_the_art)*(1+margin) ... and probably rounded to some multiple of 12 hours.

I think some systems are large and a practitioner would reasonably wait more than 24 hours and some systems are small and a practitioner would NOT wait more than 24 hours.

[jp7745]

I converted this PR to "DRAFT" to stabilize results through the upcoming presentations. We will revisit this later next week. Thanks all.