Toby Roseman

Hi @andrusza2 - I have discussed this issue with my team. We think the best fix is to change: ```python @precondition(allow=VALUE | SYMBOL) def value_inference(self): ``` To: ```python @precondition(allow=VALUE) def...

Regarding this potentially also being an issue with `flatten2d`: I think it should be fine. I think we should always know the `shape` and `axis` values. In which case, it...

Looking at your code sample, I'm not able to reproduce this issue. Could you attach your torch model (i.e. `model` from your code)? Or tell us how to reproduce it...

@MatthieuTPHR - Thank you for the detailed description. With the following imports, I can reproduce your results. ```python import torch import coremltools as ct from coremltools.models.neural_network import flexible_shape_utils import numpy...

Is there are reason you need to use a PyTorch script rather than a PyTorch traced model? Replacing this line: `s = torch.jit.script(m)` with the following line: `s = torch.jit.trace(m,...

@sandeepmistry - can you share a simple TensorFlow model that reproduces this issue?

[Loading an untrusted TensorFlow model is not secure](https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md). We really need a minimal, stand alone example, i.e. a short amount of code that can be copied and pasted to reproduce...

This is related to #1359.

@dzhelonkin - Your first code segment works for me using the most recent version of coremltools. Please try again with the latest version of coremltools. If you still get the...

Is this still an issue with our latest beta release? To install our latest beta, run: `pip install -U --pre coremltools`