sushreebarsa comments

Results 267 comments of


                                            sushreebarsa

loss and accuracy different on Custom fit function using train_step function compared to normal training

@arpit196 Could you please share the notebook link or colab gist in order to reproduce the issue correctly. I am getting different error while replicating the above code. Thank you!

loss and accuracy different on Custom fit function using train_step function compared to normal training

@gadagashwini I was able replicate the issue on colab, please find the gist [here](https://colab.research.google.com/gist/sushreebarsa/ae373ac8a2d68e507c943ffaecd80502/cifar10.ipynb ). Thank you!

Check-fail in MaxPoolGrad

@shijy16 I tried to replicate the issue on colab, could you please find the gist [here](https://colab.research.google.com/gist/sushreebarsa/8b364141eef923b353c0061b2cc8d2b4/57502.ipynb#scrollTo=UPASBHLgjuT-) and confirm the same? Thank you!

"No common supertype of ..." while loading model

@ahasselbring I tried to replicate the issue on colab and faced following error; ``` ValueError: Types are not compatible: TensorSpec(shape=(None,), dtype=tf.float32, name=None) with type of vs VariableSpec(shape=(1,), dtype=tf.float32, name='x') with...

when value is not float, tf.keras.backend.set_epsilon doesnt throw exception

@cheyennee I tried to replicate the issue reported, please find the gist [here](https://colab.research.google.com/gist/sushreebarsa/d2fbb19e7af28f4c1de8d65faa31d801/untitled113.ipynb) and confirm the same? Thank you!

Early stopping halts training after second epoch with default patience

@mogeid In order to expedite the trouble-shooting process, please provide a code snippet to reproduce the issue reported here. Thanks!

Early stopping halts training after second epoch with default patience

@gowthamkpr I was able to replicate the issue on colab using TF v2.8.0 ,please find the gist [here](https://colab.research.google.com/gist/sushreebarsa/7ccaaeaca9ebb2928fb10386540c885d/16393.ipynb) for reference.Thanks!

TypeSpec `tf.keras.layers.Input` fails to propagate the name property

@nmatare I tried to replicate the issue on colab, please find the [gist](https://colab.research.google.com/gist/sushreebarsa/edbad638fd6791af852c7426417fd81a/untitled719.ipynb#scrollTo=P89jqIbqgzyk) here and confirm the same? Thank you!

`tf.keras.Model()` behaves incorrectly if `inputs` is a dictionary view

@gowthamkpr I was able to replicate the issue on colab, please find the [gist](https://colab.research.google.com/gist/sushreebarsa/50ab03223579fc9b0dc637b29e7f88d6/17169.ipynb) here. Thank you!

Keras load LSTM/GRU model with constant mask/initial_state will raise error

@RunnerZhong Could you have a look at the gist [here](https://colab.research.google.com/gist/sushreebarsa/8d47eb86fb9347e955c108c866f57841/17173.ipynb#scrollTo=VMXJNT8_XgeT) and confirm the issue? Thank you!