add requiredMem method to llmodel impls

most of these can just shortcut out of the model loading logic llama is a bit worse to deal with because we submodule it so I have to at least parse the hparams, and then I just use the size on disk as an estimate for the mem size (which seems reasonable since we mmap() the llama files anyway)

gpt4all-backend/llmodel.h

@@ -59,6 +59,7 @@ public:
 
     virtual bool loadModel(const std::string &modelPath) = 0;
     virtual bool isModelLoaded() const = 0;
+    virtual size_t requiredMem(const std::string &modelPath) = 0;
     virtual size_t stateSize() const { return 0; }
     virtual size_t saveState(uint8_t */*dest*/) const { return 0; }
     virtual size_t restoreState(const uint8_t */*src*/) { return 0; }