Weekly GitHub Report for Tensorflow: December 07, 2024 - December 14, 2024
Weekly GitHub Report for Tensorflow
Thank you for subscribing to our weekly newsletter! Each week, we deliver a comprehensive summary of your GitHub project's latest activity right to your inbox, including an overview of your project's issues, pull requests, contributors, and commit activity.
Table of Contents
I. News
1.1 Recent Version Releases:
The current version of this repository is v2.18.0
1.2 Other Noteworthy Updates:
II. Issues
2.1 Top 5 Active Issues:
We consider active issues to be issues that that have been commented on most frequently within the last week.
-
Division by zero error at random places if GPU is used: This issue involves a division by zero error occurring at random locations when a program using TensorFlow's C API is executed on a GPU, specifically a Quadro RTX 6000, while it runs without issues on a CPU. The problem is challenging to diagnose as it appears to be deeply rooted in TensorFlow or CUDA, and the user is uncertain whether the bug is in the TensorFlow binary or one of the CUDA libraries.
- The comments discuss potential collaboration to resolve the issue, with one user offering to contribute and suggesting a screen-sharing session to better understand the problem. Another user suggests that the issue might be due to incompatible software versions and advises updating them according to the documentation. The original poster agrees to try updating the libraries and proposes scheduling a meeting to further investigate the issue, while also reporting a new problem with CUDA diagnostics after attempting an update.
- Number of comments this week: None
-
InvalidArgumentError when using MirroredStrategy but not with tf.distribute.get_strategy(): This issue involves an InvalidArgumentError encountered when using TensorFlow's MirroredStrategy, which does not occur when using tf.distribute.get_strategy(). The error arises specifically when deploying the code on multiple GPUs, and the user is seeking a solution to ensure the code runs without errors under MirroredStrategy, similar to its behavior with tf.distribute.get_strategy().
- The comments reveal that the issue is reproducible with specific versions of TensorFlow and Keras, and different users have attempted to replicate the problem with varying success. A temporary workaround involving downgrading to older versions of TensorFlow and Keras was found to resolve the issue, and there is a plan to investigate further for a permanent fix in future releases. The user confirmed the workaround's effectiveness and inquired about a long-term solution.
- Number of comments this week: None
-
error: defining a type within 'offsetof' is a Clang extension [-Werror,-Wgnu-offsetof-extensions]: This issue involves a user encountering a build error while attempting to compile the Selective Framework for iOS using TensorFlow Lite, specifically related to a Clang extension warning about defining a type within 'offsetof'. The error occurs during the compilation of the 'upb.c' file, and the user is seeking assistance to resolve this problem as they are new to the process.
- The comments section includes a request for more detailed information about the steps and environment used by the user, followed by the user providing some additional details about their setup. Another commenter suggests a potential solution by adding a specific flag to the bazel command to bypass the warning, and requests further information if the issue persists.
- Number of comments this week: None
-
The warning "The structure of
inputsdoesn't match the expected structure" when training a functional model: This issue involves a warning message encountered when training a functional model in TensorFlow, indicating a mismatch between the structure of the inputs and the expected structure, which raises concerns about the model's functionality despite the training proceeding normally. The user has identified that the warning arises from a comparison failure in the code and has provided a standalone code snippet to reproduce the issue, noting that the problem persists regardless of the data format used.- The comments discuss the cause of the warning, which is due to a mismatch between the input structure and TensorFlow's expectations, and suggest code modifications to resolve it. A user points out that using tuples instead of lists for inputs can avoid the warning, but saving and reloading the model reintroduces the issue. Another user confirms that the suggested fix resolves the warning for them, while a different user reports not encountering the warning in a different environment, prompting a discussion about environmental differences.
- Number of comments this week: None
-
Tflite x86 lib and dll for windows: This issue is about a user who is attempting to build an x86 library and DLL for TensorFlow Lite on Windows but is encountering difficulties in doing so. The user is seeking assistance from the community to either provide the necessary files or guidance on how to successfully build them.
- The comments section involves a request for more information about the user's setup and steps taken, followed by suggestions to use CMake with specific commands. The user reports an error during the build process, and further advice is given to try building from a specific commit, as there is no official x86 build available. The discussion highlights the challenges of building TensorFlow Lite for x86 on Windows and suggests building from source as the only viable option.
- Number of comments this week: None
2.2 Top 5 Stale Issues:
We consider stale issues to be issues that has had no activity within the last 30 days. The team should work together to get these issues resolved and closed as soon as possible.
As of our latest update, there are no stale issues for the project this week.
2.3 Open Issues
This section lists, groups, and then summarizes issues that were created within the last week in the repository.
Issues Opened This Week: 14
Summarized Issues:
- TensorFlow Runtime Errors: Users have reported various runtime errors when using TensorFlow across different environments. One issue involves an unexpected failure during tensor allocation in an Android Studio application due to a dimension mismatch in a TensorFlow Lite model. Another user experiences an error in PyCharm on Ubuntu 22.04, where certain TensorFlow functions are reported as missing despite the code being correct. Additionally, a division by zero error occurs unpredictably when using a GPU with TensorFlow's C API, which does not happen on a CPU, suggesting potential incompatibilities with CUDA libraries.
- Compilation and Linking Issues: Several users have encountered compilation and linking issues when building TensorFlow or its components. A compile error on Windows 11 arises from a conflict in the
TrieRawHashMap.cppfile within the LLVM project. Another issue involves undefined symbol errors when linking an Android static library with TensorFlow Lite's GPU delegate using CMake on Ubuntu. Additionally, a compilation error occurs with ndk-build for a custom C++ file due to an undefined symbol related to the NNAPI delegate in TensorFlow Lite.
- TensorFlow Version and Compatibility: Users have raised concerns about version compatibility and updates in TensorFlow. There is a request to update the Python version in TensorFlow's Docker images to ensure compatibility and improved functionality. Another issue highlights a compatibility problem with the cross-compile toolchain for building TensorFlow Lite on a Raspberry Pi Zero, which requires an armv6 architecture. Additionally, there is a request for the publication of TensorFlowLite version 2.18.0 to Maven Central and CocoaPods Specs for Android and iOS platforms.
- TensorFlow Functionality Bugs: Several bugs have been reported affecting TensorFlow's functionality. A
ValueErroroccurs in TensorFlow 2.18.0 when using a dataset with an unknown shape in themodel.evaluatefunction, which did not happen in TensorFlow 2.13. Another bug involves the 'height_shift_range' and 'width_shift_range' parameters in the ImageDataGenerator causing images to shift in the opposite directions than intended. Additionally, a documentation bug in thetf.raw_ops.MaxPoolGradWithArgmaxfunction incorrectly states that theargmaxparameter can accept bothint32andint64data types.
- TensorFlow Environment and Setup Issues: Users have reported issues related to the setup and environment configuration of TensorFlow. An ImportError occurs when attempting to run TensorFlow on PyCharm due to a DLL initialization error. Another issue describes an annoying warning message, "Ignoring Assert operator," that appears during the
model.fit()operation with Keras 3.x and TensorFlow 2.17+, disrupting the progress bar display. These issues highlight the challenges users face in configuring their environments for optimal TensorFlow performance.
2.4 Closed Issues
This section lists, groups, and then summarizes issues that were closed within the last week in the repository. This section also links the associated pull requests if applicable.
Issues Closed This Week: 25
Summarized Issues:
- TensorFlow 2.17 Bugs: Several issues have been reported in TensorFlow version 2.17, including crashes with operations like
tf.raw_ops.Cholesky,tf.raw_ops.MatrixDeterminant, andtf.raw_ops.RaggedTensorToVariantGradientwhen executed with specific inputs on a GPU. These bugs result in "Aborted (core dumped)" errors, and they have been confirmed to occur in TensorFlow Nightly as well. Users have suggested workarounds and fixes, but the issues highlight the need for further investigation and resolution by the TensorFlow team.
- TensorFlow Lite and Android Issues: Developers have encountered various issues with TensorFlow Lite on Android, including build failures and missing versions on Maven. These problems have led to difficulties in implementing TensorFlow Lite in Android projects, prompting requests for solutions and updates from the community. The issues highlight the challenges of maintaining compatibility and availability across different platforms and versions.
- Numerical Precision and Performance: Discussions have arisen regarding the numerical precision of TensorFlow operations compared to other frameworks like PyTorch and NumPy. Specifically, the
tf.math.cumsumoperation shows varying precision across different data types and hardware, with TensorFlow matching NumPy on CPU but performing better on GPU. These findings have sparked interest in understanding the precision trade-offs and performance characteristics of TensorFlow.
- Build and Compilation Issues: Several issues have been reported related to build failures and compilation warnings in TensorFlow projects. These include problems with Gradle annotations, compiler warnings in TensorFlow Lite, and compatibility issues with Microsoft Visual C++. Such issues often require patches or updates to ensure successful builds and maintain code quality.
- TensorFlow 2.17.0 and 2.16.1 Crashes: Users have reported crashes in TensorFlow versions 2.17.0 and 2.16.1 on Ubuntu 20.04, particularly with operations like
SparseTensorDenseMatMulandUnBatch. These crashes are often due to invalid input shapes or specific input parameters, resulting in "Aborted (core dumped)" errors. The issues have been addressed in later versions, but they underscore the importance of robust error handling in TensorFlow.
- TensorFlow 2.18 GPU Utilization: A bug in TensorFlow 2.18 has been identified where the software fails to utilize the GPU for certain operations on a Linux system, despite the GPU being recognized. This issue was resolved by updating the LD_LIBRARY_PATH, but it highlights the challenges of ensuring proper GPU utilization and compatibility in TensorFlow.
- Spam and Misleading Content: A spam entry was reported on a GitHub project, falsely advertising the availability of a movie for download and streaming. This issue underscores the need for vigilance and moderation to maintain the integrity of open-source projects and prevent the spread of misleading information.
2.5 Issue Discussion Insights
This section will analyze the tone and sentiment of discussions within this project's open and closed issues that occurred within the past week. It aims to identify potentially heated exchanges and to maintain a constructive project environment.
Based on our analysis, there are no instances of toxic discussions in the project's open issues from the past week.
III. Pull Requests
3.1 Open Pull Requests
This section lists and summarizes pull requests that were created within the last week in the repository.
Pull Requests Opened This Week: 5
Pull Requests:
This pull request addresses a bug related to the incorrect export of symbols when building shared libraries on Windows platforms without using Bazel or MSYS, by modifying the tensorflow/lite/CMakeLists.txt file to prevent the inheritance of definitions to other libraries, thereby ensuring proper symbol exports in the tensorflowlite_c.dll.
3.2 Closed Pull Requests
This section lists and summarizes pull requests that were closed within the last week in the repository. Similar pull requests are grouped, and associated commits are linked if applicable.
Pull Requests Closed This Week: 13
Summarized Pull Requests:
This pull request updates the .clang-format file in the TensorFlow project to allow customization of the code formatting style based on Google's style guide and to disable automatic pointer alignment derivation, providing instructions for further customization.
3.3 Pull Request Discussion Insights
This section will analyze the tone and sentiment of discussions within this project's open and closed pull requests that occurred within the past week. It aims to identify potentially heated exchanges and to maintain a constructive project environment.
- spam
- Toxicity Score: 0.55 (Defensive responses, Lack of consensus, Growing impatience.)
- This GitHub conversation involves several users, with username1 expressing frustration over username2's proposed solution not working as expected. Username2 responds defensively, attempting to clarify their approach, which leads to a tense exchange. Username3 intervenes, trying to mediate and offer alternative solutions, but the tone remains strained. The conversation is marked by a lack of consensus and growing impatience among the participants.
IV. Contributors
4.1 Contributors
Active Contributors:
We consider an active contributor in this project to be any contributor who has made at least 1 commit, opened at least 1 issue, created at least 1 pull request, or made more than 2 comments in the last month.
| Contributor | Commits | Pull Requests | Issues | Comments |
|---|---|---|---|---|
| gaikwadrahul8 | 13 | 11 | 1 | 88 |
| tilakrayal | 0 | 0 | 0 | 51 |
| Venkat6871 | 3 | 3 | 0 | 43 |
| mihaimaruseac | 0 | 1 | 0 | 21 |
| LongZE666 | 0 | 0 | 9 | 5 |
| LakshmiKalaKadali | 4 | 1 | 0 | 7 |
| x0w3n | 0 | 0 | 7 | 5 |
| pkgoogle | 0 | 0 | 0 | 12 |
| yuvashrikarunakaran | 0 | 3 | 0 | 5 |
| phpYj | 0 | 0 | 2 | 5 |
ReadMe Summary: TensorFlow is an open-source platform for machine learning, offering a flexible ecosystem of tools and libraries for researchers and developers to build and deploy ML applications. It supports Python and C++ APIs, with installation options for GPU and CPU, and provides nightly binaries for testing. Stay updated with release announcements and contribute to the project by following the guidelines and engaging with the community through various forums and resources.