[BE] Actually suppress vmap warning from gradcheck (pytorch#144287)

This is the much safer change compared to pytorch#144283

Before:
```
PYTORCH_TEST_CUDA_MEM_LEAK_CHECK=1 PYTORCH_TEST_WITH_SLOW_GRADCHECK=1 python test/test_optim.py -k TestDifferentiableOptimizer.test_sgd
/data/users/janeyx/pytorch/torch/autograd/gradcheck.py:1156: FutureWarning: Please use torch.vmap instead of torch._vmap_internals.vmap.
  result = vmap(vjp)(torch.stack(grad_outputs))
/data/users/janeyx/pytorch/torch/autograd/gradcheck.py:1156: FutureWarning: Please use torch.vmap instead of torch._vmap_internals.vmap.
  result = vmap(vjp)(torch.stack(grad_outputs))
.
----------------------------------------------------------------------
Ran 1 test in 0.028s
```

(the env vars aren't necessary)

After:
```
PYTORCH_TEST_CUDA_MEM_LEAK_CHECK=1 PYTORCH_TEST_WITH_SLOW_GRADCHECK=1 python test/test_optim.py -k TestDifferentiableOptimizer.test_sgd
.
----------------------------------------------------------------------
Ran 1 test in 0.028s
```

Pull Request resolved: pytorch#144287
Approved by: https://github.com/cyyever, https://github.com/soulitzer

Author: janeyx99

torch/autograd/gradcheck.py

@@ -1151,7 +1151,7 @@ def vjp(v):
     # NB: this doesn't work for CUDA tests: https://github.com/pytorch/pytorch/issues/50209
     with warnings.catch_warnings():
         warnings.filterwarnings("ignore", message="There is a performance drop")
-        warnings.filterwarnings("ignore", message="Please use torch.vmap")
+        warnings.filterwarnings("ignore", message="Please use `torch.vmap`")
         try:
             result = vmap(vjp)(torch.stack(grad_outputs))
         except RuntimeError as ex: