Feature: Interfacing ML code with MiMA

[mondus]

Closes #7

@jatkinson1000 - covering #4 and #5

• [x] Adapt cg_drag.f90 to accommodate drag forcing from various different approaches
• [x] Create .f90 files with modules & subroutines for relevant drag calculations
• [x] Adapt PT NN to receive individual tensors of each variable and combine in a prepended layer This way each approach will call the NN twice, once with zonal, once with meridional, but tensors will only be written to shared memory once.

@SimonClifford - covering #3

• [x] Inspect PT model and construct matching TF model
• [x] Migrate weights
• [x] Add TF code into MiMA (requires @jatkinson1000 to write .f90 files above)

We then need to conduct benchmarking:

• [ ] Instrument MiMA for ReBench
• [x] Plan a set of runs
• [x] Perform the set of runs

[SimonClifford]

Closes #13 in the early @SimonClifford commits.

[SimonClifford]

@SimonClifford commits address TensorFlow coupling. Closes #12 Closes #3

[SimonClifford]

@jatkinson1000 commits close #11 and close #12 and therefore close #4 .

[SimonClifford]

@jatkinson1000 closes #5, although we may revisit.

[jatkinson1000]

Also closes #1 in earliest @jatkinson1000 commits (we could merge cmake into this branch to remove unused dependencies that caused failure on Sherlock).

CMake docs are included in docs/GettingStarted.md.