Is it not possible to gradient checkpoint PatchEmbed in VisionTransformer?

I tried to add gradient checkpoint to vision transformer through the following code

class PatchEmbed(nn.Module):
    Image to Patch Embedding.

    def __init__(
        kernel_size: Tuple[int, int] = (16, 16),
        stride: Tuple[int, int] = (16, 16),
        padding: Tuple[int, int] = (0, 0),
        in_chans: int = 3,
        embed_dim: int = 768,
        use_checkpoint: bool = False
    ) -> None:
            kernel_size (Tuple): kernel size of the projection layer.
            stride (Tuple): stride of the projection layer.
            padding (Tuple): padding size of the projection layer.
            in_chans (int): Number of input image channels.
            embed_dim (int): Patch embedding dimension.

        self.use_checkpoint = use_checkpoint
        self.proj = nn.Conv2d(
            in_chans, embed_dim, kernel_size=kernel_size, stride=stride, padding=padding

    def forward(self, x):
        if self.use_checkpoint:
            return torch.utils.checkpoint.checkpoint(self._forward, x)
            return self._forward(x)

    def _forward(self, x: torch.Tensor) -> torch.Tensor:
        x = self.proj(x)
        # B C H W -> B H W C
        x = x.permute(0, 2, 3, 1)
        return x

However, it gives the following error when use_checkpoint is True.

RuntimeError: Expected to have finished reduction in the prior iteration before starting a new one. This error indicates that your module has parameters that were not used in producing loss. You can enable unused parameter detection by passing the keyword argument `find_unused_parameters=True` to `torch.nn.parallel.DistributedDataParallel`, and by 
making sure all `forward` function outputs participate in calculating loss. 
If you already have done the above, then the distributed data parallel module wasn't able to locate the output tensors in the return value of your module's `forward` function. Please include the loss function and the structure of the return value of `forward` of your module when reporting this issue (e.g. list, dict, iterable).
Parameter indices which did not receive grad for rank 6: 1 2
 In addition, you can set the environment variable TORCH_DISTRIBUTED_DEBUG to either INFO or DETAIL to print out information about which particular parameters did not receive gradient on this rank as part of this error

I have already checkpointed the rest of the transformer blocks but still need a bit extra memory and thought maybe I could get that by checkpointing the PatchEmbed too. However, it gives error when I try to do so.