torch.fx.experimental.proxy_tensor.maybe_disable_thunkify

torch.fx.experimental.proxy_tensor.maybe_disable_thunkify()

Within a context, disable thunkification. See maybe_enable_thunkify() for more details. This is helpful if you have a wrapper function which you want to enable thunkification on, but in some segment on the inside (say, the original user function), you want to disable thunkification as you know it is not needed there.

Return type

Generator[None, None, None]