update

app.py

@@ -39,7 +39,7 @@ logger = configure_logger()
 
 def parse_args():
     parser = argparse.ArgumentParser()
-    parser.add_argument("--model", type=str, default="Eagle2.5-VL-8B-Preview")
+    parser.add_argument("--model", type=str, default="Eagle-2.5-8B")
     parser.add_argument(
         "--local-path",
         type=str,
@@ -57,7 +57,7 @@ def fetch_model(model_name: str):
     if args.local_path:
         model_path = args.local_path
     else:
-        model_path = f"NVEagle/{args.model}"
+        model_path = f"nvidia/{args.model}"
 
     if model_name in DEPLOY_MODELS:
         model_info = DEPLOY_MODELS[model_name]

eagle_vl/serve/inference.py

@@ -18,7 +18,7 @@ from .chat_utils import Conversation, get_conv_template
 logger = logging.getLogger(__name__)
 
 
-def load_model(model_path: str = "NVEagle/Eagle2.5-VL-8B-Preview"):
+def load_model(model_path: str = "nvidia/Eagle-2.5-8B"):
 
     token = os.environ.get("HF_TOKEN")
     # hotfix the model to use flash attention 2