GGMLRunner: avoid multiple calls to get_graph

[idostyle]

Not setting up the compute graph twice might result in a minor (/negligible?) performance improvement.

Previously it worked like the following in GGMLRunner#compute:

1. calls alloc_compute_buffer 1.1 calls reset_compute_ctx 1.2 calls get_graph
2. calls reset_compute_ctx
3. calls get_graph

Tested on cpu and vulkan locally.