fix(wasm): eliminate UI hang during prompt prefill

quant.h

@@ -15461,9 +15461,15 @@ int tq_generate(tq_model_t* model, tq_tokenizer_t* tokenizer,
         fprintf(stderr, "\n");
     }
 
-    /* Prefill: process all prompt tokens */
+    /* Prefill: process all prompt tokens.
+     * On Emscripten with ASYNCIFY, yield every 2 tokens so the browser
+     * can repaint (shows "Thinking..." and avoids "page unresponsive"). */
     for (int i = 0; i < n_prompt; i++) {
         tq_forward(model, state, prompt_tokens[i], i);
+#ifdef __EMSCRIPTEN__
+        extern void emscripten_sleep(unsigned int ms);
+        if (i % 2 == 1) emscripten_sleep(0);
+#endif
     }
 
     /* Repetition penalty setup */

wasm/index.html

@@ -448,6 +448,11 @@ <h2>LLM in Your Browser</h2>
     };
 
     const chatPrompt = getChatPrompt(text);
+
+    // Ensure "Thinking..." paints before WASM blocks the thread.
+    // Double-rAF guarantees the browser completes one paint cycle.
+    await new Promise(r => requestAnimationFrame(() => requestAnimationFrame(r)));
+
     const promptPtr = Module.allocateUTF8(chatPrompt);
     try {
         await Module._wasm_generate_async(promptPtr, 0.7, 256);