Feature/move op past fork transform

Feature/linear collapse repeated

SameResize_batch HLSCustomOp

don't test with bipolar datatype, must support zero
merged rtlsim and cppsim into a single parametrized test

- SameResize_Batch hlslib op always pads with 0, i/odt must support
zero values
- Added assertion to check for PaddingStyle=2, the only tested
value

This reverts commit 2e96d065.

Feature/add past conv check padding

Feature/absorb transform only if linear

Revert "Feature/padded convolution"

Feature/padded convolution

Feature/linear reorder for residual

Feature/dupstreams batch

Implemented standalone Thresholding HLS Op

Merge branch 'feature/dupstreams_batch' of https://github.com/quetric/finn into quetric-feature/dupstreams_batch

Add new transformation to give unique parameter tensors

* use make_new_valueinfo_name instead of own routine
* set dtype on newly created tensor

Merge branch 'dev' of https://github.com/Xilinx/finn into quetric-feature/GiveUniqueParameterTensors

Merge branch 'feature/GiveUniqueParameterTensors' of https://github.com/quetric/finn into quetric-feature/GiveUniqueParameterTensors