# Latest **URL:** https://discuss.tvm.apache.org/latest.md?page=1 [Latest](https://discuss.tvm.apache.org/latest.md) · [Categories](https://discuss.tvm.apache.org/categories.md) **Page:** 2 --- ## [How to reproduce the results of autotune](https://discuss.tvm.apache.org/t/how-to-reproduce-the-results-of-autotune/18857)
**Author:** [@alen](https://discuss.tvm.apache.org/u/alen)\ **Replies:** 0\ **Last updated:** [December 18, 2025, 2:09am UTC](https://discuss.tvm.apache.org/t/how-to-reproduce-the-results-of-autotune/18857 "2025-12-18T02:09:16Z")
Hi, I’m using the latest TVM branch and running autotuning with MetaSchedule. I’ve noticed that the tuning results are different every time, even though I’ve already passed the seed parameter. Is there something else I m… --- ## [Customize Optimization Tutorial error](https://discuss.tvm.apache.org/t/customize-optimization-tutorial-error/18849)
**Author:** [@puggykk](https://discuss.tvm.apache.org/u/puggykk)\ **Replies:** 1\ **Last updated:** [December 16, 2025, 1:52pm UTC](https://discuss.tvm.apache.org/t/customize-optimization-tutorial-error/18849 "2025-12-16T13:52:58Z")
When I run Customize Optimization example on colab (not gpu, but cpu) # Import cublas pattern import tvm.relax.backend.cuda.cublas as \_cublas # Define a new pass for CUBLAS dispatch @tvm.transform.module\_pass(opt\_leve… --- ## [Did you forget to bind?](https://discuss.tvm.apache.org/t/did-you-forget-to-bind/17553)
**Author:** [@pedro](https://discuss.tvm.apache.org/u/pedro)\ **Replies:** 0\ **Last updated:** [August 5, 2024, 12:56am UTC](https://discuss.tvm.apache.org/t/did-you-forget-to-bind/17553 "2024-08-05T00:56:53Z")
I am trying to load one of my “whisper” models this way model = model.jit(spec=mod\_spec, device="cuda", out\_format="torch", debug=True) To this I get Traceback (most recent call last): File "/algo/users/ppp/GoodsDet… --- ## [\[Release\] v0.23.0 Release Schedule](https://discuss.tvm.apache.org/t/release-v0-23-0-release-schedule/18850)
**Author:** [@ysh329](https://discuss.tvm.apache.org/u/ysh329)\ **Replies:** 0\ **Last updated:** [December 15, 2025, 9:35am UTC](https://discuss.tvm.apache.org/t/release-v0-23-0-release-schedule/18850 "2025-12-15T09:35:24Z")
In order to continue with our agreed release cycle as per discussion in RFC#67 , I’d like to propose a schedule for our next TVM release: v0.23.0. I suggest we keep the discussion in the GitHub issue I opened for this … --- ## [bash: cd: /home/tvm/apps/uma: No such file or directory](https://discuss.tvm.apache.org/t/bash-cd-home-tvm-apps-uma-no-such-file-or-directory/18846)
**Author:** [@wangyangyang](https://discuss.tvm.apache.org/u/wangyangyang)\ **Replies:** 0\ **Last updated:** [December 15, 2025, 9:33am UTC](https://discuss.tvm.apache.org/t/bash-cd-home-tvm-apps-uma-no-such-file-or-directory/18846 "2025-12-15T09:33:46Z")
Could someone please explain what’s going on? Additionally, are there any tutorials for adapting hardware accelerators to TVM using UMA? I’m a beginner, thank you! Urgent! --- ## [Install error on Colab](https://discuss.tvm.apache.org/t/install-error-on-colab/18745)
**Author:** [@puggykk](https://discuss.tvm.apache.org/u/puggykk)\ **Replies:** 6\ **Last updated:** [December 11, 2025, 8:34pm UTC](https://discuss.tvm.apache.org/t/install-error-on-colab/18745 "2025-12-11T20:34:02Z")
I am following quick start But it failed to install library. How could i solve this? --- ## [where can I find autotune logs](https://discuss.tvm.apache.org/t/where-can-i-find-autotune-logs/18834)
**Author:** [@itslenz](https://discuss.tvm.apache.org/u/itslenz)\ **Replies:** 1\ **Last updated:** [December 11, 2025, 4:12pm UTC](https://discuss.tvm.apache.org/t/where-can-i-find-autotune-logs/18834 "2025-12-11T16:12:49Z")
I’m working on fine-tuning a model and need autotune logs (CUDA/ARM) for kernel configurations, performance data, and tuning results. Unfortunately, I don’t have access to the necessary hardware for autotuning. Could an… --- ## [Alternative for the bundle\_deploy](https://discuss.tvm.apache.org/t/alternative-for-the-bundle-deploy/18832)
**Author:** [@Xunflash](https://discuss.tvm.apache.org/u/Xunflash)\ **Replies:** 0\ **Last updated:** [December 11, 2025, 5:04am UTC](https://discuss.tvm.apache.org/t/alternative-for-the-bundle-deploy/18832 "2025-12-11T05:04:40Z")
After version 0.20.0, what is the recommended alternative for the bundle\_deploy functionality that was removed? I am referring to the code that was previously located here: --- ## [Hexagon Backend Compile Issues](https://discuss.tvm.apache.org/t/hexagon-backend-compile-issues/18831)
**Author:** [@linlw](https://discuss.tvm.apache.org/u/linlw)\ **Replies:** 0\ **Last updated:** [December 11, 2025, 5:04am UTC](https://discuss.tvm.apache.org/t/hexagon-backend-compile-issues/18831 "2025-12-11T05:04:24Z")
TVM 0.22 Version #compile code target\_hexagon = tvm.target.hexagon(“v68”) hexagon\_target = tvm.target.Target(target\_hexagon, host=target\_hexagon) mod = tvm.compile(ir\_mod, target=hexagon\_target) #error remote\_compile\_l… --- ## [where can I find autotune logs](https://discuss.tvm.apache.org/t/where-can-i-find-autotune-logs/18830)
**Author:** [@itslenz](https://discuss.tvm.apache.org/u/itslenz)\ **Replies:** 0\ **Last updated:** [December 11, 2025, 5:03am UTC](https://discuss.tvm.apache.org/t/where-can-i-find-autotune-logs/18830 "2025-12-11T05:03:55Z")
I’m working on fine-tuning a model and need autotune logs (CUDA/ARM) for kernel configurations, performance data, and tuning results. Unfortunately, I don’t have access to the necessary hardware for autotuning. Could an… --- ## [How to tile buffers to fit into on-chip memory?](https://discuss.tvm.apache.org/t/how-to-tile-buffers-to-fit-into-on-chip-memory/15119)
**Author:** [@Necrotos](https://discuss.tvm.apache.org/u/Necrotos)\ **Replies:** 4\ **Last updated:** [December 11, 2025, 5:03am UTC](https://discuss.tvm.apache.org/t/how-to-tile-buffers-to-fit-into-on-chip-memory/15119 "2025-12-11T05:03:43Z")
Hello, I am trying to understand how to use TVM for conv2d on a custom accelerator. For this I have created a toy example of a conv2d operation. So far I understand how to tile the computation, reorder loops and how to … --- ## [Semantic Preserving/ Type Preserving Transformations](https://discuss.tvm.apache.org/t/semantic-preserving-type-preserving-transformations/18829)
**Author:** [@Abhiroop](https://discuss.tvm.apache.org/u/Abhiroop)\ **Replies:** 0\ **Last updated:** [December 11, 2025, 5:03am UTC](https://discuss.tvm.apache.org/t/semantic-preserving-type-preserving-transformations/18829 "2025-12-11T05:03:33Z")
This is a more research-oriented question on TVM. But to the community, is there any value in having type-preserving (or something stronger like semantic-preserving) transformations in TVM? For context, in PL literature,… --- ## [tvm.register\_global\_func not exist](https://discuss.tvm.apache.org/t/tvm-register-global-func-not-exist/18828)
**Author:** [@Haleski](https://discuss.tvm.apache.org/u/Haleski)\ **Replies:** 0\ **Last updated:** [December 11, 2025, 5:03am UTC](https://discuss.tvm.apache.org/t/tvm-register-global-func-not-exist/18828 "2025-12-11T05:03:07Z")
hi, when i test test\_tir\_transform\_storage\_rewrite.py. the following register\_mem will cause the bug: def register\_mem(scope\_tb, max\_bits): # Register mem @tvm.register\_global\_func("tvm.info.mem.%s" % scope\_tb) … --- ## [How to bring an ASIC Accelerator to TVM](https://discuss.tvm.apache.org/t/how-to-bring-an-asic-accelerator-to-tvm/18808)
**Author:** [@sexjun](https://discuss.tvm.apache.org/u/sexjun)\ **Replies:** 2\ **Last updated:** [December 9, 2025, 1:38am UTC](https://discuss.tvm.apache.org/t/how-to-bring-an-asic-accelerator-to-tvm/18808 "2025-12-09T01:38:33Z")
At present, the AI ​​compiler we wrote ourselves uses the front end of relax, and then we have re-implemented the optimization, memory allocation, scheduling and code generation ourselves. This is because we did not hav… --- ## [Does tvm Bring Tensorflow support to tvm.relax.frontend from v0.21.0](https://discuss.tvm.apache.org/t/does-tvm-bring-tensorflow-support-to-tvm-relax-frontend-from-v0-21-0/18804)
**Author:** [@MankalaSrujan](https://discuss.tvm.apache.org/u/MankalaSrujan)\ **Replies:** 0\ **Last updated:** [December 1, 2025, 2:16am UTC](https://discuss.tvm.apache.org/t/does-tvm-bring-tensorflow-support-to-tvm-relax-frontend-from-v0-21-0/18804 "2025-12-01T02:16:12Z")
As of TVM version v0.22.0 we don’t have support in tvm.relax.frontend to convert any tensorflow model to tvm relax ir module. Does TVM has any plans in future to add support for tensorflow model in tvm.relax.frontend? --- ## [Install error on Local](https://discuss.tvm.apache.org/t/install-error-on-local/18803)
**Author:** [@puggykk](https://discuss.tvm.apache.org/u/puggykk)\ **Replies:** 0\ **Last updated:** [December 1, 2025, 2:15am UTC](https://discuss.tvm.apache.org/t/install-error-on-local/18803 "2025-12-01T02:15:54Z")
I followed this But cmake .. && cmake --build . --parallel $(nproc) This error occured Also this error occured. Can you solve this? --- ## [tvm.relax.transform.ConvertLayout pass is working only for relax.nn.conv2d operator](https://discuss.tvm.apache.org/t/tvm-relax-transform-convertlayout-pass-is-working-only-for-relax-nn-conv2d-operator/18802)
**Author:** [@MankalaSrujan](https://discuss.tvm.apache.org/u/MankalaSrujan)\ **Replies:** 0\ **Last updated:** [December 1, 2025, 2:15am UTC](https://discuss.tvm.apache.org/t/tvm-relax-transform-convertlayout-pass-is-working-only-for-relax-nn-conv2d-operator/18802 "2025-12-01T02:15:48Z")
When I am running following piece of code: def test\_simple\_network\_1(): def create\_ir\_module(): @I.ir\_module class SimpleNetwork1: @R.function def main(inp\_0: R.Tensor((1, 32,… --- ## [Error with tune\_tir](https://discuss.tvm.apache.org/t/error-with-tune-tir/18801)
**Author:** [@a333222](https://discuss.tvm.apache.org/u/a333222)\ **Replies:** 0\ **Last updated:** [December 1, 2025, 2:15am UTC](https://discuss.tvm.apache.org/t/error-with-tune-tir/18801 "2025-12-01T02:15:46Z")
mod, params = relax.frontend.detach\_params(mod) mod = relax.get\_pipeline(“zero”)(mod) db = ms.tir\_integration.tune\_tir( mod=mod, target=target, work\_dir=“./tuning\_spr\_logs”, max\_trials\_global=8000, # SPR은 코어 많아서… --- ## [Error on IRModule example](https://discuss.tvm.apache.org/t/error-on-irmodule-example/18800)
**Author:** [@puggykk](https://discuss.tvm.apache.org/u/puggykk)\ **Replies:** 0\ **Last updated:** [December 1, 2025, 2:15am UTC](https://discuss.tvm.apache.org/t/error-on-irmodule-example/18800 "2025-12-01T02:15:44Z")
I run example code on colab ; installation with !python -m pip install --pre -U -f https://mlc.ai/wheels mlc-ai-nightly-cpu. By the way, It occurred error on Deploy on GPU --- ## [New releases on PyPI?](https://discuss.tvm.apache.org/t/new-releases-on-pypi/18448)
**Author:** [@mattpitkin](https://discuss.tvm.apache.org/u/mattpitkin)\ **Replies:** 3\ **Last updated:** [November 30, 2025, 12:04pm UTC](https://discuss.tvm.apache.org/t/new-releases-on-pypi/18448 "2025-11-30T12:04:34Z")
Hi, I wanted to ask whether new releases of apache-tvm will ever by published on PyPI? The last stable release was v0.12.0 back in June 2023, with dev releases up to 0.14dev273 in November 2023. The version in the GitHub… --- ## [MetaScheduleTuneTIR does not work well](https://discuss.tvm.apache.org/t/metascheduletunetir-does-not-work-well/18726)
**Author:** [@alen](https://discuss.tvm.apache.org/u/alen)\ **Replies:** 3\ **Last updated:** [November 29, 2025, 1:46pm UTC](https://discuss.tvm.apache.org/t/metascheduletunetir-does-not-work-well/18726 "2025-11-29T13:46:55Z")
code version: 0.21.0(latest release) torch version: 2.4.0+cu121 import tvm from tvm import IRModule, relax from tvm.relax.frontend import nn class RelaxModel(nn.Module): def \_\_init\_\_(self): super(RelaxMode… --- ## [TVM building issue on AArch64 architecture](https://discuss.tvm.apache.org/t/tvm-building-issue-on-aarch64-architecture/18779)
**Author:** [@andy](https://discuss.tvm.apache.org/u/andy)\ **Replies:** 2\ **Last updated:** [November 28, 2025, 6:47pm UTC](https://discuss.tvm.apache.org/t/tvm-building-issue-on-aarch64-architecture/18779 "2025-11-28T18:47:31Z")
Ubuntu22, AArch64. gcc version is 12.3.0 Could anyone help me? The error info when building: This warning is for project developers. Use -Wno-dev to suppress it. – Found CUDA\_TOOLKIT\_ROOT\_DIR=/usr/local/cuda-12.2 – F… --- ## [How to keep data in local buffer between matrix and vector ops (avoid extra global memory copies)?](https://discuss.tvm.apache.org/t/how-to-keep-data-in-local-buffer-between-matrix-and-vector-ops-avoid-extra-global-memory-copies/18797)
**Author:** [@Haleski](https://discuss.tvm.apache.org/u/Haleski)\ **Replies:** 0\ **Last updated:** [November 28, 2025, 1:10pm UTC](https://discuss.tvm.apache.org/t/how-to-keep-data-in-local-buffer-between-matrix-and-vector-ops-avoid-extra-global-memory-copies/18797 "2025-11-28T13:10:03Z")
Hi everyone, I’m working on a custom accelerator backend for TVM and ran into a scheduling / memory issue that I’m not sure how to model correctly. Hardware background On our hardware we have: A matrix compute unit A… --- ## [Reproducing old VTA schedule](https://discuss.tvm.apache.org/t/reproducing-old-vta-schedule/18750)
**Author:** [@l0r3m](https://discuss.tvm.apache.org/u/l0r3m)\ **Replies:** 1\ **Last updated:** [November 28, 2025, 1:09pm UTC](https://discuss.tvm.apache.org/t/reproducing-old-vta-schedule/18750 "2025-11-28T13:09:56Z")
Hello, TL;DR: I cannot make the tensorization part of the old Simple Matrix Multiply tutorial for VTA work in modern TVM. I am trying to bring back support for the VTA by porting the old VTA specific code to the modern… --- ## [Quantization in Post Relay TVM](https://discuss.tvm.apache.org/t/quantization-in-post-relay-tvm/18770)
**Author:** [@pedrot123](https://discuss.tvm.apache.org/u/pedrot123)\ **Replies:** 0\ **Last updated:** [November 17, 2025, 8:34pm UTC](https://discuss.tvm.apache.org/t/quantization-in-post-relay-tvm/18770 "2025-11-17T20:34:47Z")
Hello, I was wondering if there are options for quantization in post Relay TVM. To be able to build support and research for custom accelerators, integer support through quantization is needed. In the old VTA flow, relay… --- ## [Announcing apache-tvm-ffi v0.1.2](https://discuss.tvm.apache.org/t/announcing-apache-tvm-ffi-v0-1-2/18756)
**Author:** [@tqchen](https://discuss.tvm.apache.org/u/tqchen)\ **Replies:** 0\ **Last updated:** [November 11, 2025, 3:28pm UTC](https://discuss.tvm.apache.org/t/announcing-apache-tvm-ffi-v0-1-2/18756 "2025-11-11T15:28:20Z")
Dear Community: We are happy to announce apache-tvm-ffi v0.1.2 This is a minor release patch adding compatibilities to cuda-python streams, and older version of ml\_dtypes, see release note here Release v0.1.2 · apache/… --- ## [Are there modern codegen/BYOC examples?](https://discuss.tvm.apache.org/t/are-there-modern-codegen-byoc-examples/18749)
**Author:** [@pedrot123](https://discuss.tvm.apache.org/u/pedrot123)\ **Replies:** 1\ **Last updated:** [November 10, 2025, 4:19pm UTC](https://discuss.tvm.apache.org/t/are-there-modern-codegen-byoc-examples/18749 "2025-11-10T16:19:30Z")
I was wondering if there are any modern examples for using your own codegen or BYOC. The only tutorial online is from 2020 and still uses relay and not relax. Any info on this would be appreciated. Thank you! --- ## [The issues encountered with the take op during tensorize](https://discuss.tvm.apache.org/t/the-issues-encountered-with-the-take-op-during-tensorize/18746)
**Author:** [@liubin](https://discuss.tvm.apache.org/u/liubin)\ **Replies:** 0\ **Last updated:** [November 8, 2025, 6:13pm UTC](https://discuss.tvm.apache.org/t/the-issues-encountered-with-the-take-op-during-tensorize/18746 "2025-11-08T18:13:44Z")
def get\_take\_2d\_intrin(M: T.int64, D: T.int64, L: T.int64, dtype=“float32”, idx\_dtype=“int32”): @T.prim\_func def desc(a: T.handle, indices: T.handle, out: T.handle): A = T.match\_buffer(a, (L, D), dtype=dtype) xx = T.matc… --- ## [Linker Problems while building TVM](https://discuss.tvm.apache.org/t/linker-problems-while-building-tvm/18744)
**Author:** [@liamsemeria](https://discuss.tvm.apache.org/u/liamsemeria)\ **Replies:** 0\ **Last updated:** [November 8, 2025, 6:13pm UTC](https://discuss.tvm.apache.org/t/linker-problems-while-building-tvm/18744 "2025-11-08T18:13:31Z")
I’ve been following the Install from Source guide and by following it exactly I get this at the end of my cmake: \[100%\] Built target tvm\_runtime \[100%\] Built target tvm\_ffi\_testing ld.lld: error: undefined symbol: std::… --- ## [Advice on Learning TVM for Embedded Development](https://discuss.tvm.apache.org/t/advice-on-learning-tvm-for-embedded-development/18705)
**Author:** [@nielskoe](https://discuss.tvm.apache.org/u/nielskoe)\ **Replies:** 3\ **Last updated:** [November 7, 2025, 12:36am UTC](https://discuss.tvm.apache.org/t/advice-on-learning-tvm-for-embedded-development/18705 "2025-11-07T00:36:32Z")
Hi all, sorry to reach out to you with such a general question. I having been reading and watching things about TVM and I am very interested! Both conceptually as I think the underlying tech is really amazing, and for u… [Previous page](https://discuss.tvm.apache.org/latest.md) [Next page](https://discuss.tvm.apache.org/latest.md?page=2)