[Apache TVM Discuss] [Questions] Custom datatypes in AutoScheduler?

The modified code looks good to me. Maybe you can double check what's going on inside those registrations, e.g. checking whether the mapping in c++ side is indeed updated when `register` is called --- [Visit Topic](https://discuss.tvm.apache.org/t/custom-datatypes-in-autoscheduler/12377/6

[Apache TVM Discuss] [Questions] Custom datatypes in AutoScheduler?

you can try modifying `reset_global_scope` https://github.com/vinx13/tvm/blob/79a1aaf693feddf1d9835779d7eb587ac0eb21df/python/tvm/autotvm/env.py#L35 and registering your custom data types there, which will be sent to worker processes. If your custom data types are registered in cpp side, it wi

[Apache TVM Discuss] [Questions] Custom datatypes in AutoScheduler?

It is possible that they are not propagated to workers, if they are registered in python side, are you able to try registering them again in workers by setting a custom PopenPool initializer like the one in the PR? --- [Visit Topic](https://discuss.tvm.apache.org/t/custom-datatypes-in-aut

[Apache TVM Discuss] [Questions] Using tvm quantize model is too slower than not quantize

To get better performance, you can try auto tuning or auto scheduling You may find these tutorials helpful https://tvm.apache.org/docs/how_to/tune_with_autotvm/tune_relay_cuda.html https://tvm.apache.org/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html?highlight=tune%20relay --- [V

[Apache TVM Discuss] [Questions] Free(): invalid pointer Aborted

You need to modify `USE_LLVM` and `HIDE_PRIVATE_SYMBOLS` in `build/config.cmake` --- [Visit Topic](https://discuss.tvm.apache.org/t/free-invalid-pointer-aborted/11357/10) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click h

[Apache TVM Discuss] [Questions] How to instantiate and evaluate tvm.tir.expr.Var?

There is a `tvm.arith.Analyzer` to bind variables and simplify https://github.com/apache/tvm/blob/main/python/tvm/arith/analyzer.py#L192 --- [Visit Topic](https://discuss.tvm.apache.org/t/how-to-instantiate-and-evaluate-tvm-tir-expr-var/11368/3) to respond. You are receiving this because

[Apache TVM Discuss] [Questions] Autotuner does not work on Windows

The error happened here https://github.com/apache/tvm/blob/main/python/tvm/contrib/cc.py#L238 can you check llvm/clang are installed correctly? --- [Visit Topic](https://discuss.tvm.apache.org/t/autotuner-does-not-work-on-windows/11332/2) to respond. You are receiving this because you e

[Apache TVM Discuss] [Questions] Errors when obtaining gradients for `nn.batch_norm`

you can add it here https://github.com/apache/tvm/blob/main/python/tvm/relay/op/_tensor_grad.py --- [Visit Topic](https://discuss.tvm.apache.org/t/errors-when-obtaining-gradients-for-nn-batch-norm/11304/2) to respond. You are receiving this because you enabled mailing list mode. To unsu

[Apache TVM Discuss] [Questions] [TensorRT] has no attribute pad_value

https://github.com/apache/tvm/blob/main/python/tvm/relay/op/contrib/tensorrt.py#L745 you need to change it to `pad_value = args[1]` and check `isinstance(pad_value, Constant)` so that you can get the numeric value of pad value --- [Visit Topic](https://discuss.tvm.apache.org/t/tensorrt-cla

[Apache TVM Discuss] [Questions] Cannot build on Mac

There are some C++ ABI issues, you can try building llvm from source or using a different compiler version. (Adding `-lc++abi` is not sufficient as llvm was pre-built in another ABI) --- [Visit Topic](https://discuss.tvm.apache.org/t/cannot-build-on-mac/10923/4) to respond. You are recei