fix CI

Author: zucchini-nlp

tests/models/video_llama_3/test_modeling_video_llama_3.py

@@ -883,8 +883,8 @@ def test_small_model_integration_test_batch_wo_image(self):
         EXPECTED_DECODED_TEXT = Expectations(
             {
                 ("cuda", None): [
-                    "user\n\nDescribe the image.\nassistant\nThe image captures a vibrant nighttime scene on a bustling city street. A woman in a striking red dress",
-                    "user\nWhat is relativity?\nassistant\nRelativity is a scientific theory that describes the relationship between space and time. It was first proposed by",
+                    'user\n\nDescribe the image.\nassistant\nThe image captures a vibrant night scene in a bustling Japanese city. A woman in a striking red dress',
+                    'user\nWhat is relativity?\nassistant\nRelativity is a scientific theory that describes the relationship between space and time. It was first proposed by'
                 ],
                 ("xpu", None): [
                     "user\n\nDescribe the image.\nassistant\nThe image captures a vibrant nighttime scene on a bustling city street. A woman in a striking red dress",
@@ -898,6 +898,7 @@ def test_small_model_integration_test_batch_wo_image(self):
             self.processor.batch_decode(output, skip_special_tokens=True),
             EXPECTED_DECODED_TEXT,
         )
+
     def test_small_model_integration_test_batch_different_resolutions(self):
         model = VideoLlama3ForConditionalGeneration.from_pretrained(
             "lkhl/VideoLLaMA3-2B-Image-HF", dtype=torch.bfloat16, device_map=torch_device
@@ -917,8 +918,8 @@ def test_small_model_integration_test_batch_different_resolutions(self):
         EXPECTED_DECODED_TEXT = Expectations(
             {
                 ("cuda", None): [
-                    "user\n\nDescribe the image.\nassistant\nThe image captures a vibrant nighttime scene on a bustling city street. A woman in a striking red dress",
-                    "user\n\nDescribe the image.\nassistant\nThe image depicts a striking urban scene at night. A person is standing in the center of a wet",
+                    'user\n\nDescribe the image.\nassistant\nThe image captures a vibrant night scene in a bustling Japanese city. A woman in a striking red dress',
+                    'user\n\nDescribe the image.\nassistant\nThe image depicts a striking urban scene at night. A person is standing in the center of a wet'
                 ],
                 ("xpu", None): [
                     "user\n\nDescribe the image.\nassistant\nThe image captures a vibrant night scene in a bustling Japanese city. A woman in a striking red dress",