Fully working version at last.

package.json

@@ -1,6 +1,6 @@
 {
   "name": "localm",
-  "version": "1.1.37",
+  "version": "1.1.38",
   "description": "Chat application",
   "scripts": {
     "build": "esbuild src/index.js --target=es6 --bundle --sourcemap --outfile=./index.js --format=iife --external:fs --external:path --external:child_process --external:ws --external:katex/dist/katex.min.css",

src/app/init-milkdown.js

@@ -105,8 +105,10 @@ export async function initMilkdown({
     availableModels = entries.map(e => ({
       id: e.id || e.modelId || '',
       name: e.name || (e.id || e.modelId || '').split('/').pop(),
-      size: '',
-      requiresAuth: e.classification === 'auth-protected'
+      size:
+        ((e.size_hint || '') + ' ' +
+        (e.info?.params || '')).trim(),
+      requiresAuth: e.classification === 'auth-protected' || e.requiresAuth,
     }));
 
     outputMessage('Models discovered: **' + availableModels.length + '**');

src/app/model-slash.js

@@ -46,7 +46,7 @@ export function createModelSlashPlugin({ getModels, onSlashCommand }) {
   // Function to rebuild menu content
   function rebuildMenu() {
     menu.innerHTML = '';
-    
+
     const availableModels = getModels();
 
     if (availableModels.length === 0) {
@@ -70,7 +70,7 @@ export function createModelSlashPlugin({ getModels, onSlashCommand }) {
       const icon = document.createElement('span');
       icon.className = 'model-icon';
       icon.textContent = model.requiresAuth ? '🔒' : '🤖';
-      
+
       // Create text container
       const textContainer = document.createElement('div');
       textContainer.className = 'model-text-container';
@@ -79,17 +79,17 @@ export function createModelSlashPlugin({ getModels, onSlashCommand }) {
       name.className = 'name';
       name.textContent = model.name;
       textContainer.appendChild(name);
-      
+
       if (model.size) {
         const subtitle = document.createElement('div');
         subtitle.className = 'size';
         subtitle.textContent = `(${model.size})`;
         textContainer.appendChild(subtitle);
       }
-      
+
       item.appendChild(icon);
       item.appendChild(textContainer);
-      
+
       // Add auth indicator if needed
       if (model.requiresAuth) {
         const authSpan = document.createElement('span');
@@ -100,7 +100,7 @@ export function createModelSlashPlugin({ getModels, onSlashCommand }) {
 
       modelList.appendChild(item);
     });
-    
+
     menu.appendChild(modelList);
   }
 

src/worker/boot-worker.js

@@ -63,7 +63,14 @@ export function bootWorker() {
       // run the pipeline
       if (!pipe) throw new Error('pipeline not available');
       self.postMessage({ id, type: 'status', status: 'inference-start', model: modelName });
-      const out = await pipe(prompt, options || {});
+      const out = await pipe(prompt, {
+        max_new_tokens: 250,        // Increase from default
+        temperature: 0.7,
+        do_sample: true,
+        pad_token_id: pipe.tokenizer.eos_token_id,
+        return_full_text: false,     // Only return the generated text
+        ...options
+      });
       const text = extractText(out);
       self.postMessage({ id, type: 'status', status: 'inference-done', model: modelName });
       self.postMessage({ id, type: 'response', result: text });

src/worker/curated-model-list.json

@@ -50,31 +50,6 @@
       "notes": "Parameter count and asset layout should be verified against HF API."
     }
   },
-  {
-    "id": "Xenova/all-MiniLM-L6-v2",
-    "name": "all-MiniLM-L6-v2",
-    "model_type": "sentence-transformers",
-    "architectures": ["MiniLM"],
-    "classification": "encoder",
-    "confidence": "high",
-    "fetchStatus": "ok",
-    "hasTokenizer": true,
-    "hasOnnxModel": true,
-    "isTransformersJsReady": true,
-    "info": {
-      "display_name": "all-MiniLM-L6-v2 (Xenova mirror)",
-      "params": "≈22M",
-      "params_count": 22000000,
-      "architecture": "MiniLM (distilled transformer)",
-      "context_window": 512,
-      "hf_url": "https://huggingface.co/Xenova/all-MiniLM-L6-v2",
-      "recommended_runtime": "transformers.js (wasm)",
-      "is_mobile_capable": true,
-      "verified": false,
-      "assumed": true,
-      "notes": "Encoder-style model (worker prefilter normally excludes sentence-transformers from generation lists)."
-    }
-  },
   {
     "id": "Xenova/distilgpt2",
     "name": "distilgpt2",
@@ -109,6 +84,7 @@
     "fetchStatus": "ok",
     "hasTokenizer": true,
     "hasOnnxModel": true,
+    "size_hint": "500Mb",
     "isTransformersJsReady": true,
     "info": {
       "display_name": "GPT-2 (Xenova mirror)",
@@ -134,6 +110,7 @@
     "hasTokenizer": true,
     "hasOnnxModel": false,
     "isTransformersJsReady": false,
+    "requiresAuth": true,
     "info": {
       "display_name": "Qwen-2.5 Small Instruct (Xenova mirror)",
       "params": "≈2.5B (estimate)",