skip CI CPU test_data_parallel_resnet_train_step (tinygrad#9685)

flaky

Author: chenyuxyz

test/test_multitensor.py

@@ -362,7 +362,7 @@ def test_data_parallel_resnet(self):
     shard_output_np = shard_output.numpy()
     np.testing.assert_allclose(real_output, shard_output_np, atol=1e-6, rtol=1e-6)
 
-  @unittest.skipIf(CI and Device.DEFAULT in ("CUDA", "NV", "LLVM"), "slow, and flaky on LLVM")
+  @unittest.skipIf(CI and Device.DEFAULT in ("CUDA", "NV", "LLVM", "CPU"), "slow, and flaky on LLVM/CPU")
   @unittest.skipIf(Device.DEFAULT == "WEBGPU" and not OSX, "WEBGPU Vulkan can only run kernels with up to 10 buffers")
   def test_data_parallel_resnet_train_step(self):
     from extra.models.resnet import ResNet18