Weekly GitHub Report for Tensorflow: July 21, 2025 - July 28, 2025 (12:05:49)
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 v2.19.0
1.2 Version Information:
Released on March 5, 2025, TensorFlow version 2.19.0 introduces breaking changes to the LiteRT (tf.lite) API, including the deprecation of tf.lite.Interpreter
in Python with a migration path to ai_edge_litert.interpreter
, and updates to C++ constants for better API compatibility. Key improvements include support for the bfloat16
data type in the tfl.Cast
operation runtime kernel, alongside the discontinuation of standalone libtensorflow
package publishing.
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. Bot comments are omitted.
-
TensorFlow was not built with CUDA kernel binaries compatible with compute capability 12.0 CUDA_ERROR_INVALID_HANDLE: This issue reports a bug where TensorFlow was not built with CUDA kernel binaries compatible with the NVIDIA RTX 5080 GPU's compute capability 12.0, causing CUDA kernels to be just-in-time compiled from PTX and resulting in a CUDA_ERROR_INVALID_HANDLE during kernel launch. The user struggles to build TensorFlow from source for CUDA 12.8.1/9.8 due to unsupported GPU architecture errors and seeks advice on resolving these build and runtime compatibility problems.
- The comments reveal that the issue does not occur with PyTorch and that the problem may stem from incompatible TensorFlow and CUDA versions; attempts to build TensorFlow from source fail due to unsupported architecture errors, but a workaround involving a specific NVIDIA container image was suggested and reported to work for a similar GPU model, though the CUDA_ERROR_INVALID_HANDLE persists for the original user.
- Number of comments this week: 5
-
tf.keras.Model have not attribute : submodules but the document still use the error attribute: This issue reports a discrepancy between the TensorFlow documentation and the actual API behavior, where the
tf.keras.Model
class is said to have asubmodules
attribute, but users encounter anAttributeError
indicating that this attribute does not exist. The user demonstrates this problem with a custom model subclass and points out that the guide example usestf.Module
instead oftf.keras.Model
, leading to confusion and errors when accessingsubmodules
.- The comment clarifies that the documentation example incorrectly uses
tf.Module
rather than subclassing fromtf.keras.Model
, which is necessary to have thesubmodules
attribute available. A corrected example subclassingtf.keras.Model
is provided, which runs without error and properly exposes thesubmodules
attribute, resolving the confusion. - Number of comments this week: 1
- The comment clarifies that the documentation example incorrectly uses
-
The model does not have any trainable weights.: This issue reports a documentation bug where a custom TensorFlow Keras model defined with
tf.Variable
for weights triggers a warning stating that the model does not have any trainable weights. The user demonstrates that replacingtf.Variable
withself.add_weight
in the model definition resolves the warning and allows training to proceed correctly.- The single comment expresses interest in contributing to the issue resolution, indicating a willingness to work on fixing or improving the documentation or example code.
- Number of comments this week: 1
Since there were fewer than 5 open issues, all of the open issues have been listed above.
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.
- TF-TRT Warning: Could not find TensorRT: This issue describes a problem where TensorFlow on Ubuntu 22.04 cannot detect TensorRT despite having the appropriate NVIDIA driver (535) and CUDA version (12.4) installed for an RTX 3050 Ti GPU. The user reports persistent errors after multiple reinstallations and driver adjustments, seeking assistance to resolve the TF-TRT warning about the missing TensorRT component.
SystemError
intf.ensure_shape
andtf.compat.v1.ensure_shape
whendtype
ofshape
istf.uint64
and its value is too large.: This issue reports a bug in TensorFlow where usingtf.ensure_shape
ortf.compat.v1.ensure_shape
with ashape
tensor of typetf.uint64
containing very large values close to 2^64 causes aSystemError
andOverflowError
. Specifically, when such large values are passed, the functions fail in eager execution mode, triggering an internal error related to type checking, which is unexpected behavior.- Feature Request: Integrate different Digital Signal Processing into tf.signal: This issue is a feature request proposing the integration of advanced digital signal processing (DSP) functionalities, similar to those found in the julius library, into TensorFlow's tf.signal module. The requester highlights the current lack of sophisticated audio data augmentation tools within TensorFlow compared to PyTorch and suggests that adding these capabilities would enhance audio model training by enabling native, efficient preprocessing and augmentation workflows.
- [DOCS] Missing complex input for Round op: This issue reports a documentation bug in TensorFlow where the
Round
operation is described as supporting complex tensor inputs, but in practice, attempting to use a complex tensor with this operation results in an error, requiring users to round the real and imaginary parts separately. The user provides a reproducible example and error logs showing that the operation fails due to a missing device kernel for complex inputs, highlighting a discrepancy between the official documentation and actual functionality. - tf.raw_ops.Unbatch aborts with "Check failed: d < dims()": This issue reports a bug in TensorFlow version 2.17 where the
tf.raw_ops.Unbatch
operation aborts with a fatal error message "Check failed: d < dims()" when invoked with certain tensor inputs. The problem occurs on Linux Ubuntu 20.04.3 LTS using Python 3.11.8, and the user has provided a minimal reproducible example demonstrating that the operation crashes due to a shape-check failure in the tensor shape code.
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: 5
Summarized Issues:
- CUDA compatibility and runtime errors: TensorFlow fails to run CUDA kernel binaries on NVIDIA RTX 5080 GPUs with compute capability 12.0 due to being built without compatible CUDA kernel binaries, causing long JIT compilation times and CUDA errors like CUDA_ERROR_INVALID_HANDLE during kernel launches. This issue highlights the incompatibility between TensorFlow builds and the latest CUDA versions on specific hardware.
- issues/97387
- TensorFlow Keras API attribute errors: The TensorFlow documentation incorrectly references a non-existent
submodules
attribute ontf.keras.Model
instances, leading to AttributeErrors in example code. Additionally, usingtf.Variable
directly in custom Keras models triggers warnings about missing trainable weights, with the recommended fix being to useself.add_weight
instead to properly register parameters. - issues/97403, issues/97599
- Control flow and operator fusion bugs in Keras: Using
tf.no_op
as a sequential control mechanism in Keras causes a DAG cycle after operator fusion when splitting matrix multiplications across multipletf.functions
, resulting in cyclic logical graphs and dependency confusion. This bug affects the correct execution order and graph construction in TensorFlow 2.10.0. - issues/97512
- Module import errors after TensorFlow version downgrade: Downgrading TensorFlow from version 2.19.0 to earlier versions like 2.18.0, 2.17.0, and 2.16.2 fixes some malfunctions but causes a
ModuleNotFoundError
because thekeras
module is missing in these older versions, despite expectations that it should be included. This issue complicates version management and backward compatibility. - issues/97620
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: 9
Summarized Issues:
- Build Failures and Compilation Issues: Multiple issues report build failures when compiling TensorFlow or TensorFlow Lite on Ubuntu 22.04 and other environments. These failures are caused by missing build targets, undefined version macros, or incorrect build configurations requiring manual workarounds or updates to build files.
- issues/96569, issues/96654
- Dependency and Version Conflicts: There are problems upgrading TensorFlow due to conflicting dependencies with protobuf and numpy, causing installation failures and errors during downgrades. Users seek advice on pinning dependencies or other workarounds to resolve these version conflicts.
- issues/96694
- DLL Load Failures on Windows with Python 3.x: Several issues describe ImportErrors caused by DLL load failures of the native TensorFlow runtime (_pywrap_tensorflow_internal) on Windows systems with Python versions 3.10, 3.11, and 3.12. These failures are linked to missing DLLs, incompatible Python versions, or missing redistributables, preventing TensorFlow from functioning properly.
- issues/97345, issues/97432, issues/97500, issues/97514
- Python Version Support Requests: Users request official TensorFlow support for newer Python versions such as Python 3.13, citing compatibility problems encountered on Linux distributions. This indicates a need for TensorFlow to update compatibility with the latest Python releases.
- issues/97361
- Code Corrections in TensorFlow: A specific issue addresses a correction in TensorFlow’s internal code, particularly in the errors_impl.py file, to fix bugs or improve code quality.
- issues/97474
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 or closed issues from the past week.
III. Pull Requests
3.1 Open Pull Requests
This section provides a summary of pull requests that were opened in the repository over the past week. The top three pull requests with the highest number of commits are highlighted as 'key' pull requests. Other pull requests are grouped based on similar characteristics for easier analysis. Up to 25 pull requests are displayed in this section, while any remaining pull requests beyond this limit are omitted for brevity.
Pull Requests Opened This Week: 6
Key Open Pull Requests
1. Attribute exporter absl: This pull request enables explicit conversion from llvm::StringRef to absl::string_view by updating the attribute_exporter.cc file and its BUILD dependencies to support this compatibility, which was previously failing due to the lack of direct conversion between these types.
- URL: pull/97525
- Merged: No
2. Fix: Remove dangling reference to ExperimentalTFLiteToTosaBytecode: This pull request addresses the NameError caused by a dangling reference to the removed C++ symbol ExperimentalTFLiteToTosaBytecode in TensorFlow 2.19 by replacing the obsolete Python function with a clearer notification to users that the function is no longer available.
- URL: pull/97508
- Merged: No
3. Update release notes for TensorFlow 2.19.1: This pull request aims to update the release notes for TensorFlow version 2.19.1 by adding a placeholder for internal release notes that a Release Owner needs to complete before submission.
- URL: pull/97297
- Merged: No
- Associated Commits: 64045
Other Open Pull Requests
- Go API for TensorFlow session configuration: This pull request introduces a type-safe, modular Go API for configuring TensorFlow sessions, including Go protobuf generation and new builder-style structs for session and GPU options. It also adds comprehensive tests and documentation while maintaining full backward compatibility with the existing binary configuration approach.
- pull/97479
- Code cleanup for compiler warnings: This pull request removes an unused variable in the ResizeBilinear888Uint8 function to fix the -Wunused-but-set-variable warning on recent Clang compiler versions. The change improves code quality by addressing compiler feedback without affecting functionality.
- pull/97484
- GPU compatibility check updates in TFLite Model Analyzer: This pull request clarifies the GPU compatibility check by adding a deprecation warning for the related argument and modifying the output message to explicitly state that GPU compatibility checks are skipped when the deprecated flag is enabled. These changes improve user awareness and maintain clarity in the tool's output.
- pull/97513
3.2 Closed Pull Requests
This section provides a summary of pull requests that were closed in the repository over the past week. The top three pull requests with the highest number of commits are highlighted as 'key' pull requests. Other pull requests are grouped based on similar characteristics for easier analysis. Up to 25 pull requests are displayed in this section, while any remaining pull requests beyond this limit are omitted for brevity.
Pull Requests Closed This Week: 8
Key Closed Pull Requests
1. multi_process_runner_test test suite hangs for 2400s: This pull request is a draft aimed at addressing the issue where the multi_process_runner_test test suite hangs for 2400 seconds by adding improved logging to help pinpoint the cause of the timeout, but it was not merged.
- URL: pull/97574
- Merged: No
- Associated Commits: 6fd39, b1cde, cc5a9, a3a2f, 390b7, c1e07, 716b3, 3f49a, 7a0fc, 99339, 9124f, a09bd, 428ff, c32c4, 5b577, 9e380, e0f21, 5556e, e36ba, 1d25e, 577be
2. Change int64 to int64_t in Pow operations for type consistency: This pull request aims to standardize the integer type usage in the Pow operation registrations within TensorFlow by replacing int64
with int64_t
in both CPU and GPU implementations to ensure type consistency.
- URL: pull/97210
- Merged: No
3. Update keras version and refresh lock files: This pull request updates the Keras version specified in the project's requirements and setup files and refreshes the associated lock files to ensure dependency consistency.
- URL: pull/97519
- Merged: 2025-07-24T21:03:14Z
Other Closed Pull Requests
- Protobuf library upgrade and fixes: This pull request upgrades the Protobuf library to version 6.31.1 to address a C++ deserialization crash related to embedded messages. It also includes additional fixes such as implementing a hermetic CUDA build.
- pull/97567
- TensorFlow Lite benchmark_model enablement: This pull request introduces a configuration option to enable the benchmark_model tool in TensorFlow Lite, allowing it to be built and installed using the all target. This facilitates packaging and mirrors a similar enhancement made in a previous pull request.
- pull/96999
- TensorFlow version update to 2.20.0-rc0: This pull request updates version numbers across key files including
core/public/version.h
,tools/pip_package/setup.py
, andtensorflow/tensorflow.bzl
. It also warns about potential lingering old version strings that require manual verification. - pull/97295
- Python dependencies pinning and lock file update: This pull request pins the Python dependencies and updates the corresponding lock files to ensure consistent and reproducible builds.
- pull/97302
- BUILD files update with dependencies: This pull request updates the BUILD files in the TensorFlow project to include necessary dependencies, improving build configuration.
- pull/97305
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.
Based on our analysis, there are no instances of toxic discussions in the project's open or closed pull requests from the past week.
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.
If there are more than 10 active contributors, the list is truncated to the top 10 based on contribution metrics for better clarity.
Contributor | Commits | Pull Requests | Issues | Comments |
---|---|---|---|---|
harshithn31 | 83 | 0 | 0 | 0 |
CloudSmallInsect | 0 | 0 | 8 | 27 |
mihaimaruseac | 0 | 0 | 0 | 27 |
Venkat6871 | 0 | 0 | 0 | 11 |
dhruv-dhiman122 | 6 | 2 | 0 | 2 |
rtg0795 | 6 | 3 | 0 | 0 |
gaikwadrahul8 | 3 | 2 | 0 | 3 |
mynickmynick | 6 | 1 | 0 | 0 |
jiren-the-gray | 0 | 0 | 7 | 0 |
No author found | 6 | 0 | 0 | 0 |