From 9be971d47c07bf60f8dc0abbc217515a2cb6183e Mon Sep 17 00:00:00 2001
From: Sunny Liu <sunny19981005@outlook.com>
Date: Mon, 21 Apr 2025 11:34:37 -0400
Subject: [PATCH] update test_models.py to conform to new quantization config

---
 tests/utils/test_models.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/tests/utils/test_models.py b/tests/utils/test_models.py
index 83678430a..37267e3f7 100644
--- a/tests/utils/test_models.py
+++ b/tests/utils/test_models.py
@@ -21,8 +21,10 @@ class TestModelsUtils:
                 "base_model": "JackFram/llama-68m",
                 "model_type": "LlamaForCausalLM",
                 "tokenizer_type": "LlamaTokenizer",
-                "load_in_8bit": True,
-                "load_in_4bit": False,
+                "quantization": {
+                    "backend": "bnb",
+                    "bits": 8,
+                },
                 "adapter": "lora",
                 "flash_attention": False,
                 "sample_packing": True,