fix args for ensure_shard

exo/api/chatgpt_api.py

@@ -568,7 +568,7 @@ class ChatGPTAPI:
       if model_name not in model_cards: return web.json_response({"error": f"Invalid model: {model_name}. Supported models: {list(model_cards.keys())}"}, status=400)
       shard = build_base_shard(model_name, self.inference_engine_classname)
       if not shard: return web.json_response({"error": f"Could not build shard for model {model_name}"}, status=400)
-      asyncio.create_task(self.node.inference_engine.shard_downloader.ensure_shard(shard))
+      asyncio.create_task(self.node.inference_engine.shard_downloader.ensure_shard(shard, self.inference_engine_classname))
 
       return web.json_response({
         "status": "success",