Eric Shi

Would it help if we put the warning at a higher verbosity level, e.g. if you set `wp.config.verbose = True`? I don't want to presume that everyone doing a non-standard...

Can you please clarify the code that produced the output in the original comment? Input: ```python import torch import warp as wp wp.init() a = torch.arange(10) wp1 = wp.from_torch(a, dtype=wp.int32)...

Thanks! I'll take a look.

Here's what I have in a new branch (no need to flip on a verbose flag, this is the default behavior): ``` Warp UserWarning: The input data type int64 does...

This change went in with 5e32b5f0b419834761860a4ef3a7600dbea97bb0 Eventually, we might raise an exception for certain conversions, e.g. requesting a 32-bit type when the Torch array is a 64-bit type.

Due to a lack of activity, I'll close this issue. Please reopen the issue or file a new one if you're still having issues.

Sorry we weren't able to address this issue before removing `warp.sim` from the codebase. As of Warp v1.10, the `warp.sim` module has been removed. It was deprecated back in v1.8...

This issue seems related to #552. After adding some debug commands into `example_cartpole.py`: ``` import faulthandler faulthandler.enable() wp.config.mode = "debug" ``` I can get: ``` Warp 1.6.0 initialized: CUDA not...

@cadop: Can you modify https://github.com/NVIDIA/warp/blob/acfeabcf8f3414e1b131df3b47027dc5789c9113/warp/examples/core/example_dem.py#L178 to use `with wp.ScopedTimer("step", synchronize=True):`? I noticed a similar behavior happening with the timing information printed out to stdout at lower resolutions. Since you turned...

> What causes the change in performance for synchronize? The issue here is that the default output of `wp.ScopedTimer` is not usually the time it takes for a block of...