wait on download, server ready

Author: amakropoulos

Samples~/AndroidDemo/AndroidDemo.cs

@@ -20,37 +20,29 @@ public class AndroidDemo : MonoBehaviour
         public Text progressText;
         int cores;
 
-        void Awake()
-        {
-            ChatPanel.SetActive(false);
-            DownloadPanel.SetActive(false);
-        }
-
-        void Start()
+        async void Start()
         {
             playerText.onSubmit.AddListener(onInputFieldSubmit);
             playerText.interactable = false;
-            StartCoroutine(Loading());
+            await ShowDownloadScreen();
+            await WarmUp();
         }
 
-        IEnumerator<string> Loading()
+        async Task ShowDownloadScreen()
         {
+            ChatPanel.SetActive(false);
             DownloadPanel.SetActive(true);
-            AIText.text = "Downloading model...";
-            Task downloadTask = llm.DownloadModel(
-                "https://huggingface.co/afrideva/smol_llama-220M-openhermes-GGUF/resolve/main/smol_llama-220m-openhermes.q4_k_m.gguf?download=true",
-                SetProgress
-            );
-            while (!downloadTask.IsCompleted) yield return null;
-            llm.SetTemplate("alpaca");
+            // await llm.WaitUntilModelDownloaded(SetProgress);
             DownloadPanel.SetActive(false);
-
             ChatPanel.SetActive(true);
-            cores = LLMUnitySetup.AndroidGetNumBigCores();
-            AIText.text += $"\nWarming up the model...\nWill use {cores} cores";
-            Task warmup = llmCharacter.Warmup();
-            while (!warmup.IsCompleted) yield return null;
+        }
 
+        async Task WarmUp()
+        {
+            llm.SetTemplate("alpaca");
+            cores = LLMUnitySetup.AndroidGetNumBigCores();
+            AIText.text += $"Warming up the model...\nWill use {cores} cores";
+            await llmCharacter.Warmup();
             AIText.text = $"Ready when you are ({cores} cores)!";
             AIReplyComplete();
         }

Samples~/AndroidDemo/Scene.unity

@@ -1323,22 +1323,24 @@ MonoBehaviour:
   advancedOptions: 0
   remote: 0
   port: 13333
-  numThreads: -1
+  numThreads: 2
   numGPULayers: 0
-  debug: 1
+  debug: 0
   parallelPrompts: -1
   asynchronousStartup: 1
   dontDestroyOnLoad: 1
-  model: smol_llama-220m-openhermes.q4_k_m.gguf
+  model: Phi-3-mini-4k-instruct-q4.gguf
+  downloadOnBuild: 0
+  modelURL: https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-gguf/resolve/main/Phi-3-mini-4k-instruct-q4.gguf?download=true
   lora: 
   contextSize: 0
   batchSize: 512
   basePrompt: 
-  SelectedModel: 0
+  SelectedModel: 3
   modelProgress: 1
   modelCopyProgress: 1
   modelHide: 1
-  chatTemplate: chatml
+  chatTemplate: phi-3
 --- !u!4 &1047848255
 Transform:
   m_ObjectHideFlags: 0