From 3e26d7d37e70916f821c4aaff2706c3715af1b1d Mon Sep 17 00:00:00 2001
From: Christian Medina <37550954+cmedinasoriano@users.noreply.github.com>
Date: Thu, 2 Jul 2026 12:09:49 -0400
Subject: [PATCH] fix: use already-quantized 4-bit model for test

---
 test_model_loading.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/test_model_loading.py b/test_model_loading.py
index 122018c..72859e8 100644
--- a/test_model_loading.py
+++ b/test_model_loading.py
@@ -31,9 +31,9 @@ def test_model_loading():
         
         print("   Loading model...")
         model = AutoModelForCausalLM.from_pretrained(
-            "/data/models/Ornith-1.0-35B",
-            quantization_config=bnb_config,
+            "/data/models/Ornith-1.0-35B-4bit",  # Use already-quantized 4-bit model
             device_map="auto",
+            torch_dtype=torch.float16,
             trust_remote_code=True,
             low_cpu_mem_usage=True,
         )