[tests] enable GemmaIntegrationTest on XPU (#33555)

enable GemmaIntegrationTest

tests/models/gemma/test_modeling_gemma.py

@@ -528,7 +528,7 @@ class GemmaModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixi
 
 
 @slow
-@require_torch_gpu
+@require_torch_accelerator
 class GemmaIntegrationTest(unittest.TestCase):
     input_text = ["Hello I am doing", "Hi today"]
     # This variable is used to determine which CUDA device are we using for our runners (A10 or T4)
@@ -748,7 +748,6 @@ class GemmaIntegrationTest(unittest.TestCase):
 
         output = model.generate(**inputs, max_new_tokens=20, do_sample=False)
         output_text = tokenizer.batch_decode(output, skip_special_tokens=True)
-
         self.assertEqual(output_text, EXPECTED_TEXTS[self.cuda_compute_capability_major_version])
 
     @require_read_token
@@ -770,10 +769,8 @@ class GemmaIntegrationTest(unittest.TestCase):
 
         tokenizer = AutoTokenizer.from_pretrained(model_id)
         inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device)
-
         output = model.generate(**inputs, max_new_tokens=20, do_sample=False)
         output_text = tokenizer.batch_decode(output, skip_special_tokens=True)
-
         self.assertEqual(output_text, EXPECTED_TEXTS)
 
     @require_bitsandbytes