typo(vllm): gpu_memory_utilisation typo (#553)

* Update vllm_model.py

* Update use-vllm-as-backend.mdx

* Update quicktour.mdx

docs/source/quicktour.mdx

@@ -148,7 +148,7 @@ accelerate).
 ### VLLM
 
 - **pretrained** (str): HuggingFace Hub model ID name or the path to a pre-trained model to load.
-- **gpu_memory_utilisation** (float): The fraction of GPU memory to use.
+- **gpu_memory_utilization** (float): The fraction of GPU memory to use.
 - **batch_size** (int): The batch size for model training.
 - **revision** (str): The revision of the model.
 - **dtype** (str, None): The data type to use for the model.

docs/source/use-vllm-as-backend.mdx

@@ -57,4 +57,4 @@ model: # Model specific parameters
 
 > [!WARNING]
 > In the case of OOM issues, you might need to reduce the context size of the
-> model as well as reduce the `gpu_memory_utilisation` parameter.
+> model as well as reduce the `gpu_memory_utilization` parameter.

src/lighteval/models/vllm/vllm_model.py

@@ -76,7 +76,7 @@
 @dataclass
 class VLLMModelConfig:
     pretrained: str
-    gpu_memory_utilisation: float = 0.9  # lower this if you are running out of memory
+    gpu_memory_utilization: float = 0.9  # lower this if you are running out of memory
     revision: str = "main"  # revision of the model
     dtype: str | None = None
     tensor_parallel_size: int = 1  # how many GPUs to use for tensor parallelism
@@ -174,7 +174,7 @@ def _create_auto_model(self, config: VLLMModelConfig, env_config: EnvConfig) ->
         """
         self.model_args = {
             "model": config.pretrained,
-            "gpu_memory_utilization": float(config.gpu_memory_utilisation),
+            "gpu_memory_utilization": float(config.gpu_memory_utilization),
             "revision": config.revision + (f"/{config.subfolder}" if config.subfolder is not None else ""),
             "dtype": config.dtype,
             "trust_remote_code": config.trust_remote_code,