fix the use of device_map in CP docs (#12902)

up

Author: sayakpaul

docs/source/en/training/distributed_inference.md

@@ -263,8 +263,8 @@ def main():
     world_size = dist.get_world_size()
 
     pipeline = DiffusionPipeline.from_pretrained(
-        "black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16, device_map=device
-    )
+        "black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16
+    ).to(device)
     pipeline.transformer.set_attention_backend("_native_cudnn")
 
     cp_config = ContextParallelConfig(ring_degree=world_size)