text stringlengths 36 4k | label stringclasses 4
values | repo stringclasses 27
values | source stringclasses 2
values |
|---|---|---|---|
implement llama 2 using Tensorflow
### Issue type
Feature Request
### Have you reproduced the bug with TensorFlow Nightly?
No
### Source
source
### TensorFlow version
tf 2.8
### Custom code
No
### OS platform and distribution
_No response_
### Mobile device
_No response_
### Python version
_No response_... | feature | tensorflow/tensorflow | nlbse2024 |
Will there a MLP model in the future version?
### Issue type
Feature Request
### Have you reproduced the bug with TensorFlow Nightly?
No
### Source
source
### TensorFlow version
2.13.0
### Custom code
No
### OS platform and distribution
_No response_
### Mobile device
_No response_
### Python version
_N... | feature | tensorflow/tensorflow | nlbse2024 |
pybind11_proto from python to C++
@BlaziusMaximus thanks for the explanation.
I've been exploring how to update the [`import_graph_def()`](https://github.com/tensorflow/tensorflow/blob/v2.13.0/tensorflow/python/framework/importer.py#L353-L411) code-path to use pybind11_protobuf and I could use your help with the follo... | feature | tensorflow/tensorflow | nlbse2024 |
[Feature] The Heaviside step function as a activation function
Some of the implementations like Single Layer Perceptron needs discrete outputs like 0 or 1. Adding this could make the model building ease. | feature | tensorflow/tensorflow | nlbse2024 |
float8 support for array ops
### Issue type
Feature Request
### Have you reproduced the bug with TensorFlow Nightly?
No
### Source
binary
### TensorFlow version
2.12.0
### Custom code
No
### OS platform and distribution
macOS-13.2.1-arm64-arm-64bit
### Mobile device
_No response_
### Python version
3.9.... | feature | tensorflow/tensorflow | nlbse2024 |
Issue still tittle
### Issue type
Documentation Feature Request
### Have you reproduced the bug with TensorFlow Nightly?
Yes
### Source
source
### TensorFlow version
Tf2.8
### Custom code
Yes
### OS platform and distribution
_No response_
### Mobile device
_No response_
### Python version
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
the tf keras models load_model() used for loading ml model is not able to load model
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
tf 2.8
### Custom Code
Yes
### OS Platform and Dis... | feature | tensorflow/tensorflow | nlbse2024 |
tf.keras.metrics.Precision treats label as binary?
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
binary
### Tensorflow Version
2.14.0-dev20230611
### Custom Code
Yes
### OS Platform and Distributio... | feature | tensorflow/tensorflow | nlbse2024 |
Could not interpret loss function "NDCG Lambda Weight V2"
A erro happened!
Could not interpret loss function NDCG Lambda Weight V2
-------------------------------------------------------------------
model.compile (
optimizer=optimizer,
loss=tfr.keras.losses.NDCGLambdaWeightV2(topn=60),
metrics=[tfr.keras.metrics... | feature | tensorflow/tensorflow | nlbse2024 |
GPU Delegate dynamic tensor input shape (Feature Request)
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
tf 2.12.0
### Custom Code
Yes
### OS Platform and Distribution
Android
### ... | feature | tensorflow/tensorflow | nlbse2024 |
Add docs reference to latest numpy version for `tf.experimental.numpy` functions
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Have you reproduced the bug with TF nightly?
No
### Source
binary
### Tensorflow Version
2.12.0
### Custom Code
No
### OS Platform... | feature | tensorflow/tensorflow | nlbse2024 |
Network-Level Control of Precision (fp32 in training, fp16 in inference)
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
binary
### Tensorflow Version
tf 2.12
### Custom Code
No
### OS Platform and Distribution
_... | feature | tensorflow/tensorflow | nlbse2024 |
How to write custom XLA op for tensorflow
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
binary
### Tensorflow Version
2.12.0
### Custom Code
Yes
### OS Platform and Distribution
Linux
### Mobile ... | feature | tensorflow/tensorflow | nlbse2024 |
Need Add Perl to Supported API Languages
This page lists all the languages with support for the TensorFlow API:
https://www.tensorflow.org/api_docs
Thanks to an official grant from The Perl Foundation, we have recently released a Perl API:
https://github.com/EntropyOrg/perl-AI-TensorFlow-Libtensorflow
https://metacpa... | feature | tensorflow/tensorflow | nlbse2024 |
Make time series_from_array() more intuitive to use
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
No
### Source
source
### Tensorflow Version
2.10.0 (Tensorflow-macos)
### Custom Code
Yes
### OS Platform and Distribution
MacO... | feature | tensorflow/tensorflow | nlbse2024 |
How to build with xnnpack
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
tf 2.10.0
### Custom Code
Yes
### OS Platform and Distribution
ubuntu 20.04
### Mobile device
Qualcomm
##... | feature | tensorflow/tensorflow | nlbse2024 |
Freeze zero weights during training
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
2.11
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No re... | feature | tensorflow/tensorflow | nlbse2024 |
Cast int32 to bfloat16 does not run on A100 GPU
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
TensorFlow version 2.13.0-dev20230215
### Custom Code
Yes
### OS Platform and Distribut... | feature | tensorflow/tensorflow | nlbse2024 |
Feature Request: More descriptive get_weights() method
### Issue Type
Feature Request
### Tensorflow Version
tf 2.11.0
### OS Platform and Distribution
Windows 11
### Python version
3.10.9
Hello!
Currently, I can get the weights of each layer of a Keras model by calling TensorFlow's `get_weights()` method. Add... | feature | tensorflow/tensorflow | nlbse2024 |
Proto file missing in nightly wheels
### Have you reproduced the bug with TF nightly?
Yes
### Tensorflow Version
tf 2.12
### Current Behaviour?
I work on the TensorFlow-DirectML plugin, and we use the .proto files included in the TF wheel to generate pb.cc/pb.h files needed by the plugin. One of the files needed ... | feature | tensorflow/tensorflow | nlbse2024 |
How to change specific body line colour eg. Hip-Knee-Ankle line in swift
<details><summary>Click to expand!</summary>
### Issue Type
Bug
### Have you reproduced the bug with TF nightly?
Yes
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal... | feature | tensorflow/tensorflow | nlbse2024 |
keras.models.clone_model not working
<details><summary>Click to expand!</summary>
### Issue Type
Bug
### Have you reproduced the bug with TF nightly?
No
### Source
source
### Tensorflow Version
tf 2.10
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
How to train with mobilenet models using model_spec.get() ??
Hello. I am using this tutorial to train my dataset for the object detection and efficientdet_lite (0-4) models are working fine.
https://www.tensorflow.org/lite/models/modify/model_maker/object_detection
I found a documentation for **model_spec.get** and i... | feature | tensorflow/tensorflow | nlbse2024 |
Request: Some way to initialise a dynamically-sized state variable in TFLite
**System information**
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04): **All**
- TensorFlow installed from (source or binary): **Binary**
- TensorFlow version (or github SHA if from source): **2.9.1**
**Provide the text output fro... | feature | tensorflow/tensorflow | nlbse2024 |
Support for keepdims and padding in tf.boolean_mask or tf.ragged.boolean_mask
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
tf 2.8
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
##... | feature | tensorflow/tensorflow | nlbse2024 |
TensorFlow Lite Image Classification - add a new model
Hello. I tested your TensorFlow Lite Image Classification android app and it works perfect with these models (MobileNet V1, EfficientNet Lite0, EfficientNet Lite1, EfficientNet Lite2) which are downloading automatically by the download.gradle file.
https://github... | feature | tensorflow/tensorflow | nlbse2024 |
Expecting the same names when converting to tflite model
Hello,
I have created very simple model, and I want to export it to tflite model, to benchmark it on android device, and run simple inferences. This is code to create model:
```python
import tensorflow as tf
import numpy as np
from tensorflow import keras
from ... | feature | tensorflow/tensorflow | nlbse2024 |
Tensorflow Lite iOS Interpreter production logs
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
2.10.0
### Custom Code
No
### OS Platform and Distribution
iOS
### Mobile device
iOS
### Python version
_No response_
### Bazel version
... | feature | tensorflow/tensorflow | nlbse2024 |
Feature request (TensorFlow Lite): Implement mechanism to load Flex ops in C for Android
### System information
- **Have I written custom code**: no
- **OS Platform and Distribution**: macOS Monterey 12.6
- **TensorFlow installed from**: source
- **TensorFlow version**: 2.10.0
- **Python version**: 3.10.8
- **Bazel v... | feature | tensorflow/tensorflow | nlbse2024 |
`tf.identity` docs missing instructions, GPU to CPU
### Issue Type
Documentation Feature Request
### Source
binary
### Tensorflow Version
2.10.0
### Custom Code
No
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python version
_No response_
### Bazel version
_No respon... | feature | tensorflow/tensorflow | nlbse2024 |
Document behavior of tf.keras.layers.Bidirectional with return_state=True
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Source
source
### Tensorflow Version
2.10
### Custom Code
No
</details>
It is unclear from the documentation how a `tf.keras.layers.Bidirec... | feature | tensorflow/tensorflow | nlbse2024 |
How to enable GSPMD?
<details><summary>Click to expand!</summary>
### Issue Type
Support
### Source
binary
### Tensorflow Version
tf 2.7
### Custom Code
No
### OS Platform and Distribution
Linux
### Mobile device
_No response_
### Python version
_No response_
### Bazel version
_No response_
### GC... | feature | tensorflow/tensorflow | nlbse2024 |
Support for C++ Builder 11.2
<details><summary>Click to expand!</summary>
### Issue Type
Others
### Source
binary
### Tensorflow Version
TF 2.8
### Custom Code
No
### OS Platform and Distribution
Windows 11 x64
### Mobile device
_No response_
### Python version
_No response_
### Bazel version
_No re... | feature | tensorflow/tensorflow | nlbse2024 |
LeakyRelu in Tensorflow lite with the Hexagon Delegate not supported
When using a tflite model (8-bits quantized via TensorFlow lite conversion framework) that includes the activation function "LeakyRelu", the Hexagon delegate from tensorflow framework cannot perform the DNN inference on the whole graph, but rather it... | feature | tensorflow/tensorflow | nlbse2024 |
Making tensorflow.tile similar to numpy.tile and torch.tile
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
2.9.1
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python version
_... | feature | tensorflow/tensorflow | nlbse2024 |
Supporting int32 value type in TextFileInitializer.
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
tf 2.9
### Custom Code
No
### OS Platform and Distribution
Linux
### Mobile device
_No response_
### Python version
_No response_
##... | feature | tensorflow/tensorflow | nlbse2024 |
How to build TensorFlow from source with Clang?
<details><summary>Click to expand!</summary>
### Issue Type
Build/Install
### Source
source
### Tensorflow Version
v2.10.0
### Custom Code
No
### OS Platform and Distribution
Ubuntu 20.04
### Mobile device
_No response_
### Python version
3.8
### Bazel ... | feature | tensorflow/tensorflow | nlbse2024 |
Adan optimizer
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
tf 2.9.1
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python version
_No response_
### Bazel version
_No respo... | feature | tensorflow/tensorflow | nlbse2024 |
GPU detection src location
<details><summary>Click to expand!</summary>
### Issue Type
Others
### Source
source
### Tensorflow Version
2
### Custom Code
No
### OS Platform and Distribution
Ubuntu 20.04
### Mobile device
_No response_
### Python version
_No response_
### Bazel version
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
tf.image.combined_non_max_suppression crash with segmentation fault
<details><summary>Click to expand!</summary>
### Issue Type
Bug
### Source
binary
### Tensorflow Version
2.11.0-dev20220916
### Custom Code
No
### OS Platform and Distribution
Ubuntu 18.04.4 LTS (x86_64)
### Mobile device
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
Include "name" parameter in tf.keras.models.clone_model() function to give a new name to the clone.
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
2.8.2
### Custom Code
No
### OS Platform and Distribution
_No response_
### Mobile devic... | feature | tensorflow/tensorflow | nlbse2024 |
Feature: Balanced Accuracy
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
tf 2.10.0-rc3
### Custom Code
No
### OS Platform and Distribution
Debian 11
### Mobile device
_No response_
### Python version
3.9
### Bazel version
_No res... | feature | tensorflow/tensorflow | nlbse2024 |
tf.image.rot90 should add a note for the case k<0
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Source
source
### Tensorflow Version
TF2.8
### Custom Code
Yes
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python versio... | feature | tensorflow/tensorflow | nlbse2024 |
`tf.pad` fails when executing in `tf.autodiff.ForwardAccumulator`
<details><summary>Click to expand!</summary>
### Issue Type
Bug
### Source
binary
### Tensorflow Version
tf 2.9.1
### Custom Code
Yes
### OS Platform and Distribution
Linux Ubuntu 20.04
### Mobile device
_No response_
### Python version
... | feature | tensorflow/tensorflow | nlbse2024 |
when will tensorflow support FP8?
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
tf2.9
### Custom Code
No
### OS Platform and Distribution
Linux ubuntu 20.04
### Mobile device
_No response_
### Python version
_No response_
### Baze... | feature | tensorflow/tensorflow | nlbse2024 |
Please publish official wheels for Apple Silicon
<details><summary>Click to expand!</summary>
### Issue Type
Build/Install
### Source
binary
### Tensorflow Version
2.9.1
### Custom Code
No
### OS Platform and Distribution
macOS 12
### Mobile device
_No response_
### Python version
_No response_
### B... | feature | tensorflow/tensorflow | nlbse2024 |
How to fix the "undefined reference" problem for the TF Lite? A suggestion.
When [CMakeLists.txt](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/CMakeLists.txt) is used to build a shared library of TF Lite, it is impossible to link with it because of the "undefined reference" problem.
Those "und... | feature | tensorflow/tensorflow | nlbse2024 |
Is possible to measure height of person using tensorFlow pose Estimation
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal']
### Custom Code
Yes
### OS Platform a... | feature | tensorflow/tensorflow | nlbse2024 |
Rsqrt doesn't support for tf.lite.OpsSet.EXPERIMENTAL_TFLITE_BUILTINS_ACTIVATIONS_INT16_WEIGHTS_INT8
**System information**
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04):
- TensorFlow installed from (source or binary):
- TensorFlow version (or github SHA if from source):
**Provide the text output from tfl... | feature | tensorflow/tensorflow | nlbse2024 |
Are there on-device training with Tensorflow lite in Arduino IDEs???
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
tf2.8
### Custom Code
No
### OS Platform and Distribution
Windows
### Mobile device
_No response_
### Python version
... | feature | tensorflow/tensorflow | nlbse2024 |
[Feature Request] GELU activation with the Hexagon delegate
**System information**
- OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 20.04
- TensorFlow installed from (source or binary): binary
- TensorFlow version (or github SHA if from source): 2.9.1
I think I'd be able to implement this myself, but... | feature | tensorflow/tensorflow | nlbse2024 |
[TF.io] Incompatible with AWS S3 filepaths
<details><summary>Click to expand!</summary>
### Issue Type
Bug
### Source
binary
### Tensorflow Version
2.9.1
### Custom Code
No
### OS Platform and Distribution
RHEL
### Mobile device
_No response_
### Python version
3.7.12
### Bazel version
_No response_... | feature | tensorflow/tensorflow | nlbse2024 |
CMAKE has no key to disable the build of python wrapper when building TF lite from source using Xcode or Android Studio
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
TF2.9
### Custom Code
No
### OS Platform and Distribution
Linux Ubunt... | feature | tensorflow/tensorflow | nlbse2024 |
Custom Op written in C API errors when requesting GPU_Device stream
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Source
binary
### Tensorflow Version
2.8
### Custom Code
Yes
### OS Platform and Distribution
Linux
### Mobile device
07010841551
### Python... | feature | tensorflow/tensorflow | nlbse2024 |
Reproducible sparse-dense matrix multiplication
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
v2.9.0-18-gd8ce9f9c301 2.9.1
### Custom Code
No
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python ... | feature | tensorflow/tensorflow | nlbse2024 |
Tensorflow GPU Delegate Limit resources (Feature Request)
Hello, sometimes when I'm running the tflite model on gpu delegate it use almost all gpu resources while execution and this behaviour freeze UI because of lack resources. If it possible, will be good to have some initial option with percentage of GPU use.
OS ... | feature | tensorflow/tensorflow | nlbse2024 |
Is it possible to sync tensorflow stream_executor's cuda stream with cuda stream outside?
In tensorflow, streams including cuda streams are handled by stream_executer. Now I make some modification to tensorflow and create an independent cuda stream outside stream_executer. I was wondering is it possible to sync the in... | feature | tensorflow/tensorflow | nlbse2024 |
Custom Op written in C API compilation Issue
<details><summary>Click to expand!</summary>
### Issue Type
Documentation Feature Request
### Source
binary
### Tensorflow Version
2.8
### Custom Code
Yes
### OS Platform and Distribution
Linux
### Mobile device
07010841551
### Python version
_No response_... | feature | tensorflow/tensorflow | nlbse2024 |
Add tf.Complex to tflite ops. Support padding and slicing of complex tensors
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
2.9.1
### Custom Code
No
### OS Platform and Distribution
Linux Ubuntu 20.04
### Mobile device
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
F1 score metric
### Issue Type
<kbd>Feature Request</kbd>
---
### Tensorflow Version
2.9.1
---
### Current Behaviour?
No F1 score metric. It's a holistic measure for classification.
https://torchmetrics.readthedocs.io/en/stable/classification/f1_score.html
https://scikit-learn.org/stable/modules/generated/skl... | feature | tensorflow/tensorflow | nlbse2024 |
Is it possible to change colour of specific keypoint in iOS
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal']
### Custom Code
Yes
### OS Platform and Distributi... | feature | tensorflow/tensorflow | nlbse2024 |
Building Tflite compiling out pthread
<details><summary>Click to expand!</summary>
### Issue Type
Build/Install
### Source
source
### Tensorflow Version
tf2.9
### Custom Code
No
### OS Platform and Distribution
Linux Ubuntu 16.04.2 LTS
### Mobile device
_No response_
### Python version
_No response_
... | feature | tensorflow/tensorflow | nlbse2024 |
How to display custom label with angles in iOS?
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal']
### Custom Code
Yes
### OS Platform and Distribution
_No resp... | feature | tensorflow/tensorflow | nlbse2024 |
How to calculate 45 degree standing position of body from camera in swift (Pose estimation)
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal']
### Custom Code
Yes... | feature | tensorflow/tensorflow | nlbse2024 |
is it possible to calculate Body degree angle using pose estimation
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
pod 'TensorFlowLiteSwift', '~> 0.0.1-nightly', :subspecs => ['CoreML', 'Metal']
### Custom Code
Yes
### OS Platform and Di... | feature | tensorflow/tensorflow | nlbse2024 |
Restore non max suppression (nms) support for Tensorflow Lite GPU Delegate.
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
binary
### Tensorflow Version
v2.9.0-rc2-42-g8a20d54a3c1 2.9.0
### Custom Code
Yes
### OS Platform and Distribution
Google Colab
### Mobile de... | feature | tensorflow/tensorflow | nlbse2024 |
[TFLite] Dynamic batch size with GPU delegate
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
2.10.0-dev20220427
### Custom Code
Yes
### OS Platform and Distribution
Linux Ubuntu 18.04
### Mobile device
Linux Ubuntu 18.04
### Python v... | feature | tensorflow/tensorflow | nlbse2024 |
Build/release Python 3.10 tflite-runtime wheels to PyPI
<details><summary>Click to expand!</summary>
### Issue Type
Build/Install
### Source
binary
### Tensorflow Version
2.9.0
### Custom Code
No
### OS Platform and Distribution
Linux & macOS
### Mobile device
_No response_
### Python version
3.10
##... | feature | tensorflow/tensorflow | nlbse2024 |
3D SSIM
Hi, is it possible to add 3D SSIM (batched) into TensorFlow? We need this feature for 3D inference. Currently `tf.image.ssim` only accepts batched 2D input.
If not, is there any internal beta version of 3D SSIM for try-out?
Thanks. | feature | tensorflow/tensorflow | nlbse2024 |
tf.keras.optimizers.experimental.AdamW only support constant weight_decay
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
### Source
source
### Tensorflow Version
2.8
### Custom Code
No
### OS Platform and Distribution
_No response_
### Mobile device
_No response_
### Python... | feature | tensorflow/tensorflow | nlbse2024 |
How to resize (downsample) 5D samples in tensorflow?
Issue Type: Feature Request
Source: binary
Tensorflow Version: tf 2.8
Custom Code: Yes
OS Platform and Distribution: Windows 10
Python version: 3.9
---
## Note.
Reposting from [here](https://github.com/keras-team/keras/issues/16260). It's closed in keras and reco... | feature | tensorflow/tensorflow | nlbse2024 |
TFLite Tensor name converted by 2.7.0 is much longer than 1.14.0
### System information
- **OS Platform and Distribution**: Linux Ubuntu 18.04
- **Mobile device**: No
- **TensorFlow installed from (source or binary)**: binary, pip install tensorflow==2.7.0
- **TensorFlow version**: 2.7.0
- **Python version**: 3.8.13
... | feature | tensorflow/tensorflow | nlbse2024 |
Replacing Add Layer with Linear Combination of Learnable Weights
<details><summary>Click to expand!</summary>
### Issue Type
Feature Request
The Add() Layer can be expressed as a linear combination of inputs where each coefficient value is 1.
I was trying to make a custom layer that performs the same as Add, e... | feature | tensorflow/tensorflow | nlbse2024 |
request for feature gather and matmul
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
... | feature | tensorflow/tensorflow | nlbse2024 |
RFE tensorflow-aarch64==2.6.0 build ?
**System information**
TensorFlow version (you are using): 2.6.0
- Are you willing to contribute it (Yes/No): Yes
**Describe the feature and the current behavior/state.**
Brainchip Akida AKD1000 SNN neuromorphic MetaTF SDK support 2.6.0 on x86_64. They claim support for aarch64... | feature | tensorflow/tensorflow | nlbse2024 |
tf.image.extract_patches default value for areas outside the input
Running tf.image.extract_patches by the "Same" padding leads to zeros for areas outside the input. There is no way to change the default value and it is best to have this as another parameter for this function. This background label should be always ze... | feature | tensorflow/tensorflow | nlbse2024 |
`pfor`/`tf.vectorize_map`: Improve feedback/hints messages when it `fallback_to_while_loop`
**System information**
- TensorFlow version (you are using):
- Are you willing to contribute it (Yes/No):
Probably if I can talk with a codeowner of this namepace
**Describe the feature and the current behavior/state.**
As we ... | feature | tensorflow/tensorflow | nlbse2024 |
List of TF ops that tf.vectorized_map supports
Where can I get this information?
Which tf operation is a vectorized operation? and which one is not? | feature | tensorflow/tensorflow | nlbse2024 |
Generalise `xla::Map` to functions over arbitrary shapes
**System information**
- TensorFlow version (you are using): 2.8
- Are you willing to contribute it (Yes/No): No
**Describe the feature and the current behavior/state.**
Currently `xla::Map` only allow functions from shape [] to []. I'd like to use functions w... | feature | tensorflow/tensorflow | nlbse2024 |
please publish TensorFlowLiteObjC 2.8
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
... | feature | tensorflow/tensorflow | nlbse2024 |
capture_tpu_profile on Cloud TPU VM
Hello
It seems that capture_tpu_profile only works for legacy TPU devices, but not for Cloud TPU VM.
I was wondering if and when we can expect to have support for Cloud TPU VM.
At the moment, its is impossible to monitor Jax/Flax TPU workloads, as tensorboard instrumentation does... | feature | tensorflow/tensorflow | nlbse2024 |
XLA tf.bincount support
**System information**
- TensorFlow version (you are using): master
- Are you willing to contribute it (Yes/No): only if we have a clear path and a reviewer on how to contribute this
**Describe the feature and the current behavior/state.**
`tf.bincount` isn't supported by XLA
**Will this change... | feature | tensorflow/tensorflow | nlbse2024 |
gradient_function/gradient_tapes with device annotations
**System information**
- TensorFlow version (you are using): 2.8
- Are you willing to contribute it (Yes/No): Yes
**Describe the feature and the current behavior/state.**
Feature request: adding `colocate_gradients_with_ops` option to `tf.gradients` in 1.x to ... | feature | tensorflow/tensorflow | nlbse2024 |
RuntimeError: Exporting/importing meta graphs is not supported when eager execution is enabled. No graph exists when eager execution is enabled.
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs... | feature | tensorflow/tensorflow | nlbse2024 |
Option to avoid caching with bijectors.
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>... | feature | tensorflow/tensorflow | nlbse2024 |
Gradient for tf.sparse.reduce_max
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
**De... | feature | tensorflow/tensorflow | nlbse2024 |
iterating over `tf.Tensor` is not allowed
<em>Please make sure that this is a bug. As per our
[GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md),
we only address code/doc bugs, performance issues, feature requests and
build/installation issues on GitHub. tag:bug_template</em>
**System inf... | feature | tensorflow/tensorflow | nlbse2024 |
Count number of leaves in the ensemble gradient BoostedTree
This is an issue related to the performance of [TensorFlow.BoostedTree](https://www.tensorflow.org/api_docs/python/tf/estimator/BoostedTreesClassifier)
**System information**
- TensorFlow version (you are using): 2.7.0
- Are you willing to contribute it (Yes... | feature | tensorflow/tensorflow | nlbse2024 |
normalization should be done using moving averages?
## URL(s) with the issue:
https://www.tensorflow.org/tutorials/structured_data/time_series#normalize_the_data
## Description of issue (what needs changing):
Hello this is a small feature request. In the Time series tutorial at the normalization section in the 3rd ... | feature | tensorflow/tensorflow | nlbse2024 |
Using scipy fsolve or other solvers (e.g. GEKKO) in combination with backpropagation models
I was wondering if there was any progress or existing methodologies to use existing solvers such as GEKKO to find/update certain values in your network.
E.g: When using PINN networks on a electrical circuit, one might use a var... | feature | tensorflow/tensorflow | nlbse2024 |
[TFLite] Optional Debug Tools can only support log in Desktop Computer
Hi TensorFlowers,
The `tflite::PrintInterpreterState(interpreter.get())` in [minimal.cc](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/examples/minimal/minimal.cc#L62) can only be used on computer system with desktop. Because... | feature | tensorflow/tensorflow | nlbse2024 |
I have made wrapper of Tensor
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
**System... | feature | tensorflow/tensorflow | nlbse2024 |
tf.data.dataset gather element
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
**Syste... | feature | tensorflow/tensorflow | nlbse2024 |
Request for RandomIllumination layer in keras
Something like this already is possible with `ImageDataGenerator` but with corners in it (only applies brightness scale).
What is needed is a full keras layer that randomize illumination differences spatially across the image (or even a single scaling factor to the whole i... | feature | tensorflow/tensorflow | nlbse2024 |
Release tensorflow-macos wheel compatible with Python 3.9 and x86
Currently in version 2.7.0 of `tensorflow-macos`, there are only 3 wheels released in pip: 2 for arm64 supporting Python 3.8-3.9 and 1 for x86 only supporting Python 3.8.
It would be great to have also a wheel for Python 3.9 for the x86 architecture.
Th... | feature | tensorflow/tensorflow | nlbse2024 |
tf.debugging.assert_type throws error when checking type of RaggedTensor
<em>Please make sure that this is a bug. As per our
[GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md),
we only address code/doc bugs, performance issues, feature requests and
build/installation issues on GitHub. tag:... | feature | tensorflow/tensorflow | nlbse2024 |
Support partial parameter warm-start from pretrained checkpoints in tensorflow v2 (non-estimator mode)
<em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and... | feature | tensorflow/tensorflow | nlbse2024 |
New dtype: bcomplex32
**System information**
- TensorFlow version (you are using): 2.7
- Are you willing to contribute it (Yes/No): Probably yes
**Describe the feature and the current behavior/state.**
Add bcomplex32 dtype.
Currently only a real (like in Math, real field) version is supported (bfloat16)
**Will this ... | feature | tensorflow/tensorflow | nlbse2024 |
Can I repeat a 1-D tensor in segment form?
According to the doc, TF has: tf.segment_xxx and tf.repeat. But can I repeat a 1-D tensor in segment form?
#### Example:
segment = [0 0 0 1 1]
value = [0 1 2 3 4]
#### [ 1 ] If repeat_cnt = [2 2 2 1 1]
With tf.repeat, I can easily repeat value to: [0 0 1 1 2 2 3 4]
#### [ ... | feature | tensorflow/tensorflow | nlbse2024 |
LSTM expansion
**System information**
- TensorFlow version (you are using): 1.3
- Are you willing to contribute it (Yes/No): Maybe
**Describe the feature and the current behavior/state.**
Simple LSTMs have limitations. It cannot extract some features, i.e. some mathematical operands, conditional logic, etc. I don't ... | feature | tensorflow/tensorflow | nlbse2024 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.