Tilak

icon company Google icon location India icon location Google

Results 330 comments of Tilak
trafficstars

Unexpected breaking change: Optimizer.get_weights() removed

@gowthamkpr, I was able to reproduce the issue on tensorflow v2.8, v2.9 and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/edc5936e8c247cf0df328421c1dd685d/untitled568.ipynb).

2022-09-17 23:49:02.600018: I tensorflow/stream_executor/gpu/asm_compiler.cc:323] ptxas warning : Registers are spilled to local memory in function 'fusion_24', 8 bytes spill stores, 16 bytes spill loads ptxas warning : Registers are spilled to local memory in function '__internal_trig_reduction_slowpathd', 4 bytes spill stores, 4 bytes spill loads

@gadagashwini, I was able to reproduce the issue on tensorflow v2.8 and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/4a9d0812b748b3a1c3b9ba40a71131f4/singlerun_neohookean_hyperelastic_fosls_2d_tf2_py.ipynb).

ModelCheckpoint used with `save_best_only` doesn't handle interruptions, even with BackupAndRestore

@gowthamkpr, I was able to reproduce the issue on tensorflow [v2.8](https://colab.research.google.com/gist/tilakrayal/a5ee0fe112e214d5a56dec3998f670bc/2-8-17012.ipynb), [v2.9](https://colab.research.google.com/gist/tilakrayal/362c5b3945c7b16f7e94c9b58b3d909b/2_9_17012.ipynb) and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/e0a1191bc9fd3aa84b194ae9e4f15600/2_11_17012.ipynb).

ConvNeXt not compatible with mixed precision

@gowthamkpr, I was able to reproduce the issue on tensorflow v2.8, [v2.9](https://colab.research.google.com/gist/tilakrayal/a5520f1ef47e04c7d052b1ff75cf7ad2/untitled504.ipynb) and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/30257452eebf9cc45e590f84ed8da4bd/untitled503.ipynb).

Memory leak activated when setting a seed in TensorFlow 2 (2.0 & 2.1) and using eager execution.

@IanRDavies, Could you please check the attached [gist](https://colab.research.google.com/gist/tilakrayal/5b776e3a60e59b513ff5f9588d3c6238/memory_leak_in_tensorflow_2.ipynb) of the latest version and let us know if the issue still persists. Thank you!

Crash due to Use of Uninitialized Value When Initializing TensorShape in MaxPoolingWithArgmaxOp

@sachinprasadhs, I was able to reproduce the issue on tensorflow v2.8, v2.9 and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/608249722e57127ee76f4882bf7e7a69/untitled542.ipynb).

Very slow tf.TensorArray and a possible bug when used with range()

@ghost, Could you please take a look at this [comment](https://github.com/tensorflow/tensorflow/issues/46518#issuecomment-767746510) from the developer and update to the latest tensorflow v2.9 and let us know if the issue still persists. Thank...

Reshape layer drops mask from previous layers

@brian36 , Can you please take a look at this [issue](https://github.com/tensorflow/tensorflow/issues/42403) with similar error.It helps.Thanks

Conv2D with XLA `jit_compile=True` fails to run

@Co1lin, Thanks for opening this issue. This issue is not related to Keras. Could you please post this issue on [Tensorflow/Tensorflow](https://github.com/tensorflow/tensorflow/issues) repo. Thank you!

2.10.0: not ready for pep517 build spec?

@kloczek, Could you please provide the build source you are trying to install. Also please try to install the latest stable v2.9.0 from [here](https://github.com/tensorflow/tensorflow/tree/v2.9.0) and let us know if you...