evaluating the policy in single agent mode

[MohammadSajadMarvi]

Hi, First of all thank you so much for the great work

I am currently training a policy which is almost the same as the baseline but with minor changes. I am a bit confused regarding the evaluation of the model, in case we want to evaluate how model is performing shouldn't we control only the ego and not all the agents in the scene?
In the case I want to change this from multiple controlled agent to single agent I saw there is this line in the config file
max_controlled_agents: 1 # Maximum number of agents controlled by the model. Make sure this aligns with the variable kMaxAgentCount in src/consts.hpp based on this I set the inline constexpr madrona::CountT kMaxAgentCount = 1;
in consts.hpp
and rebuild the code, now when I run the evaluation code I'm facing this error

 /gpudrive/src/types.hpp(259): error: the size of an array must be greater than zero
          PartnerObservation obs[consts::kMaxAgentCount - 1];
                                 ^

/gpudrive/src/types.hpp(264): error: static assertion failed
      static_assert(sizeof(PartnerObservations) == sizeof(float) *
      ^

/gpudrive/src/types.hpp(335): error: the size of an array must be greater than zero
          madrona::Entity e[consts::kMaxAgentCount - 1];
                            ^

3 errors detected in the compilation of "/gpudrive/src/sim.cpp".


Error at /gpudrive/external/madrona/src/mw/cpp_compile.cpp:100 in CompileOutput madrona::cu::jitCompileCPPSrc(const char *, const char *, const char **, uint32_t, const char **, uint32_t, bool)
NVRTC_ERROR_COMPILATION
Aborted

I would appreciate if you have any tips on how to solve this problem