Fix Customize Optimization tutorial import error #18584 #18697

AshwiniBokka · 2026-01-29T14:30:54Z

Problem

The tutorial docs/how_to/tutorials/customize_opt.py fails with "No module named 'tvm_ffi'" when TVM is not built. This happens because the tutorial imports tvm.relax.backend.cuda.cublas which requires TVM to be built with CUDA support.

Solution

Add clear prerequisite note about building TVM with CUDA
Add helpful error handling with guidance when import fails

Changes

Added note after imports explaining build requirements
Wrapped problematic import with try/except
Provides clear error message pointing to build instructions
Follows Python best practices (stderr, proper indentation)

Testing

File compiles without syntax errors
Error message clearly guides users to solution

Before

import tvm.relax.backend.cuda.cublas as _cublas
# Users get: No module named 'tvm_ffi'

…error

- Add prerequisite note about building TVM with CUDA support - Add error handling for 'tvm_ffi' import failure - Print helpful error messages to stderr - Provide link to build instructions Fixes apache#18584

gemini-code-assist · 2026-01-29T14:31:10Z

Summary of Changes

Hello @AshwiniBokka, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request addresses a common ImportError encountered by users of the customize_opt.py tutorial when TVM is not built with CUDA. The changes improve the user experience by adding clear documentation about the CUDA build requirement and implementing a try-except block to gracefully handle the missing module, providing actionable guidance. Additionally, it updates license-related URLs to use HTTPS.

Highlights

Enhanced Tutorial Robustness: The customize_opt.py tutorial now includes explicit prerequisites and robust error handling for the tvm.relax.backend.cuda.cublas import, guiding users if TVM is not built with CUDA support.
Improved Error Messaging: A try-except block has been added to catch ImportError related to tvm_ffi, providing a clear error message to stderr with instructions on how to build TVM with CUDA.
URL Protocol Update: Several http links to Apache licenses in LICENSE and README.md files have been updated to https for improved security and consistency.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request addresses an import error in the 'Customize Optimization' tutorial that occurs when TVM is not built with CUDA support. The changes include adding a note about the CUDA dependency and implementing a try-except block to catch the ImportError and provide a user-friendly error message. The changes are well-aligned with the problem description. My review includes a critical fix for a Byte Order Mark (BOM) character that was introduced, and a suggestion to make the error handling more robust by checking for CUDA feature availability directly instead of parsing the exception message.

gemini-code-assist · 2026-01-29T14:32:44Z

docs/how_to/tutorials/customize_opt.py

@@ -1,4 +1,4 @@
-# Licensed to the Apache Software Foundation (ASF) under one
+# Licensed to the Apache Software Foundation (ASF) under one


The file starts with a UTF-8 Byte Order Mark (BOM), which can cause issues with script execution in some environments. It should be removed.

Suggested change

# Licensed to the Apache Software Foundation (ASF) under one

# Licensed to the Apache Software Foundation (ASF) under one

gemini-code-assist · 2026-01-29T14:32:45Z

docs/how_to/tutorials/customize_opt.py

+try:
+    import tvm.relax.backend.cuda.cublas as _cublas
+except ImportError as e:
+    if "tvm_ffi" in str(e):
+        import sys
+        print("Error: TVM needs to be built with CUDA support.", file=sys.stderr)
+        print("Please build TVM from source with CUDA enabled.", file=sys.stderr)
+        print("See: https://tvm.apache.org/docs/install/from_source.html", file=sys.stderr)
+        sys.exit(1)
+    else:
+        raise


Checking the exception string for "tvm_ffi" is a bit brittle, as the error message could change in future TVM versions. A more robust approach is to use tvm.runtime.enabled("cuda") to check if CUDA support is compiled in. This directly checks for the feature availability rather than relying on a specific error message string.

Suggested change

try:

import tvm.relax.backend.cuda.cublas as _cublas

except ImportError as e:

if "tvm_ffi" in str(e):

import sys

print("Error: TVM needs to be built with CUDA support.", file=sys.stderr)

print("Please build TVM from source with CUDA enabled.", file=sys.stderr)

print("See: https://tvm.apache.org/docs/install/from_source.html", file=sys.stderr)

sys.exit(1)

else:

raise

try:

import tvm.relax.backend.cuda.cublas as _cublas

except ImportError:

import tvm.runtime

if not tvm.runtime.enabled("cuda"):

import sys

print("Error: This tutorial requires TVM to be built with CUDA support.", file=sys.stderr)

print("Please build TVM from source with CUDA enabled.", file=sys.stderr)

print("See: https://tvm.apache.org/docs/install/from_source.html", file=sys.stderr)

sys.exit(1)

else:

raise

yongwww · 2026-01-29T17:06:37Z

LICENSE

   You may obtain a copy of the License at

-       http://www.apache.org/licenses/LICENSE-2.0
+       https://www.apache.org/licenses/LICENSE-2.0


why do we need to update license here?

yongwww · 2026-01-29T17:09:13Z

docs/how_to/tutorials/customize_opt.py

 from tvm.relax.frontend import nn

+# Note: This tutorial requires TVM to be built with CUDA support.
+# If you encounter 'No module named ''tvm_ffi''' error,


The statement might not be correct, tvm_ffi is now a independent repo (and 3rdparty in tvm), I think we just need to pip install tvm-ffi if we find the No module named tvm_ffi error.

AshwiniBokka added 6 commits January 25, 2026 17:16

docs: update http to https in license link

a65fb44

docs: update http to https in LICENSE file (lines 3 and 195)

8d74f25

docs: add contributor entry

fceb7da

fix: add missing trailing newline to CONTRIBUTORS.md to resolve lint …

2201b99

…error

Merge branch 'main' of https://github.com/apache/tvm

9a599d5

Fix Customize Optimization tutorial import error

1084817

- Add prerequisite note about building TVM with CUDA support - Add error handling for 'tvm_ffi' import failure - Print helpful error messages to stderr - Provide link to build instructions Fixes apache#18584

gemini-code-assist bot reviewed Jan 29, 2026

View reviewed changes

AshwiniBokka added 2 commits January 29, 2026 20:38

[lint] Apply black formatting

4f8fbac

[lint] Fix ASF license header in README.md

f6b4209

yongwww requested changes Jan 29, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix Customize Optimization tutorial import error #18584 #18697

Fix Customize Optimization tutorial import error #18584 #18697

AshwiniBokka commented Jan 29, 2026

Uh oh!

gemini-code-assist bot commented Jan 29, 2026

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

gemini-code-assist bot Jan 29, 2026

Uh oh!

gemini-code-assist bot Jan 29, 2026

Uh oh!

yongwww Jan 29, 2026

Uh oh!

yongwww Jan 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		@@ -1,4 +1,4 @@
		# Licensed to the Apache Software Foundation (ASF) under one
		# Licensed to the Apache Software Foundation (ASF) under one

Fix Customize Optimization tutorial import error #18584 #18697

Are you sure you want to change the base?

Fix Customize Optimization tutorial import error #18584 #18697

Conversation

AshwiniBokka commented Jan 29, 2026

Problem

Solution

Changes

Testing

Before

Uh oh!

gemini-code-assist bot commented Jan 29, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist bot Jan 29, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist bot Jan 29, 2026

Choose a reason for hiding this comment

Uh oh!

yongwww Jan 29, 2026

Choose a reason for hiding this comment

Uh oh!

yongwww Jan 29, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants