add more qwen2.5 models: mlx-community/Qwen2.5-7B-Instruct-4bit mlx-community/Qwen2.5-Math-7B-Instruct-4bit mlx-community/Qwen2.5-72B-Instruct-4bit mlx-community/Qwen2.5-Math-72B-Instruct-4bit

exo/models.py

@@ -31,7 +31,19 @@ model_base_shards = {
   ### llava
   "llava-1.5-7b-hf": {"MLXDynamicShardInferenceEngine": Shard(model_id="llava-hf/llava-1.5-7b-hf", start_layer=0, end_layer=0, n_layers=32),},
   ### qwen
+  "qwen-2.5-7b": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-7B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),
+  },
+  "qwen-2.5-math-7b": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Math-7B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),
+  },
   "qwen-2.5-14b": {
     "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-14B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=48),
   },
+  "qwen-2.5-72b": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-72B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=80),
+  },
+  "qwen-2.5-math-72b": {
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Math-72B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=80),
+  },
 }

tinychat/examples/tinychat/index.html

@@ -37,7 +37,11 @@
 <option value="mistral-large">Mistral Large</option>
 <option value="deepseek-coder-v2-lite">Deepseek Coder V2 Lite</option>
 <option value="llava-1.5-7b-hf">LLaVa 1.5 7B (Vision Model)</option>
+<option value="qwen-2.5-7b">Qwen 2.5 7B</option>
+<option value="qwen-2.5-math-7b">Qwen 2.5 7B (Math)</option>
 <option value="qwen-2.5-14b">Qwen 2.5 14B</option>
+<option value="qwen-2.5-72b">Qwen 2.5 72B</option>
+<option value="qwen-2.5-math-72b">Qwen 2.5 72B (Math)</option>
 </select>
 </div>
 <div @popstate.window="