Dlopen backend 5 (#779)

Major change to the backend that allows for pluggable versions of llama.cpp/ggml. This was squashed merged from dlopen_backend_5 where the history is preserved.

gpt4all-backend/llamamodel_impl.h

@@ -0,0 +1,39 @@
+#ifndef LLAMAMODEL_H_I_KNOW_WHAT_I_AM_DOING_WHEN_INCLUDING_THIS_FILE
+#error This file is NOT meant to be included outside of llamamodel.cpp. Doing so is DANGEROUS. Be sure to know what you are doing before proceeding to #define LLAMAMODEL_H_I_KNOW_WHAT_I_AM_DOING_WHEN_INCLUDING_THIS_FILE
+#endif
+#ifndef LLAMAMODEL_H
+#define LLAMAMODEL_H
+
+#include <string>
+#include <functional>
+#include <vector>
+#include "llmodel.h"
+
+struct LLamaPrivate;
+class LLamaModel : public LLModel {
+public:
+    LLamaModel();
+    ~LLamaModel();
+
+    bool loadModel(const std::string &modelPath) override;
+    bool isModelLoaded() const override;
+    size_t stateSize() const override;
+    size_t saveState(uint8_t *dest) const override;
+    size_t restoreState(const uint8_t *src) override;
+    void prompt(const std::string &prompt,
+        std::function<bool(int32_t)> promptCallback,
+        std::function<bool(int32_t, const std::string&)> responseCallback,
+        std::function<bool(bool)> recalculateCallback,
+        PromptContext &ctx) override;
+    void setThreadCount(int32_t n_threads) override;
+    int32_t threadCount() const override;
+
+protected:
+    void recalculateContext(PromptContext &promptCtx,
+        std::function<bool(bool)> recalculate) override;
+
+private:
+    LLamaPrivate *d_ptr;
+};
+
+#endif // LLAMAMODEL_H