[Apache TVM Discuss] [Questions] [conda] [tvm-py] `target.build.llvm` is not enabled

I don't think install from source needs any system privileges, even PYTHON_PATH, you can use export method to make it visible. --- [Visit Topic](https://discuss.tvm.apache.org/t/conda-tvm-py-target-build-llvm-is-not-enabled/9346/7) to respond. You are receiving this because you enabled m

[Apache TVM Discuss] [Questions] [conda] [tvm-py] `target.build.llvm` is not enabled

no, I mean in cmake.config, set llvm config and recompile libtvm.so from tvm source code. see https://tvm.apache.org/docs/install/from_source.html#developers-get-source-from-github. Install tvm from source, and export to PYTHON_PATH, and it will be visible to conda env. --- [Visit Topic]

[Apache TVM Discuss] [Questions] Does TVM support dynamic range sum?

Hi, have you solved it ? I would like to use some dynamic computed loop counts inside te.compute and don't know how to do it, I guess it is the same issue as yours. Looking forward to your feedback, thanks. --- [Visit Topic](https://discuss.tvm.apache.org/t/does-tvm-support-dynamic-range-

[Apache TVM Discuss] [Questions] [conda] [tvm-py] `target.build.llvm` is not enabled

Have you set llvm toolchain during the compilation of libtvm.so ? --- [Visit Topic](https://discuss.tvm.apache.org/t/conda-tvm-py-target-build-llvm-is-not-enabled/9346/2) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click h

[Apache TVM Discuss] [Questions] Tiled tensors for external (python) functions and TIR == te.extern?

[quote="cron, post:1, topic:9083"] side of `te.extern` is blocking any kind of optimization which leads from introducing this stage into an [/quote] Hi, I am also trying to apply some schedule primitive methods on tensor.ExternOp, but it seems not supported according to this post (3 years' ag

[Apache TVM Discuss] [Questions] Optimizing a loop body expressed with tvm.extern

May I ask, in this case, the only way to optimize the extern op is to do it by hand (with tir intrinsics) right ? For example like topi.sparse.csrmm implementation, no schedule configs could be associated with it. --- [Visit Topic](https://discuss.tvm.apache.org/t/optimizing-a-loop-body-e

[Apache TVM Discuss] [Questions] [Tutorial Fix] TVM 2019 FCRC Code

Actually I came to FCRC to understand Relay and TIR, I would be very appreciated if you can help me with the questions in my other post https://discuss.tvm.apache.org/t/tir-relay-why-tvm-tir-api-generates-relay-ir/9132 --- [Visit Topic](https://discuss.tvm.apache.org/t/tutorial-fix-tvm-20

[Apache TVM Discuss] [Questions] [Tutorial Fix] TVM 2019 FCRC Code

Hi I just started to learn TVM and noted that some code in FCRC workshop are too old to run with lastest tvm (0.8_dev) FCRC code lab link [https://sampl.cs.washington.edu/tvmfcrc/] So I put the fix here for any starters. 1. relay.Module not exist anymore mod = tvm.IRModule({}) 2. get p

[Apache TVM Discuss] [Questions] [TIR/Relay] why tvm.tir api generates relay IR

The reason might be IRModule.astext() method will parse the Module in Relay IR for better understanding, I guess. Which means astext() can parse IRModule back to Relay IR. Correct me if I misunderstood, thanks. --- [Visit Topic](https://discuss.tvm.apache.org/t/tir-relay-why-tvm-tir-api-g

[Apache TVM Discuss] [Questions] [TIR/Relay] why tvm.tir api generates relay IR

Hello, I am also struggling the tvm.transform.Sequential API usage，could you please help me to understand the following code piece ? # driver/build_module.py下面的_build_for_device函数 opt_mixed += [ tvm.tir.transform.ThreadSync("shared"), tvm.tir.transform.ThreadSync

[Apache TVM Discuss] [Questions] How to dump IRs for each pass when building a model

Hello, where is this tvm.build_config come from ? I am using tvm 0.8 dev but this does not exist --- [Visit Topic](https://discuss.tvm.apache.org/t/how-to-dump-irs-for-each-pass-when-building-a-model/454/7) to respond. You are receiving this because you enabled mailing list mode. To uns