Strata: how to configure sampling parameters

The top-level README doesn't mention this, but you can add a "sampling" key to your strata-iq3_s.json like this:

{ "sampling": {"temperature": 1.0, "top_p": 0.95, "top_k": 20}, "exe": "/path/to/Strata/engine/strata", "args": [ ... ], ... } 

From docs/DETAILS.md:

The run config's optional sampling block sets the defaults for requests that leave the fields out ("sampling": {"temperature": 1.0, "top_p": 0.95, "top_k": 20}); a request's own fields always win, and with no block at all a request without sampling keys decodes greedy.