Vladimir

You need to install the GCC compiler on your system. If this is Ubuntu or Debian Linux, you can do it with `sudo apt install build-essential`. If on CentOS, you...

Your model is too large, so you ran out of memory during synthesis. You need to use `io_stream` instead of `io_parallel` to have any chance of success.

Your Vivado installation is the issue. You're missing some libraries (`multilib` probably).

A question on the approach itself: This will only work for ReLU and will require duplicate overrides for every possible combination of activation, layer and HLS implementation if we want...

It was discussed offline and we converged on the proper approach to this.

Hi @mackncheesiest, thanks for this, we were also thinking about what the best solution to parsing the report would be. One alternative to edalize that could be explored is to...

@jmduarte This will probably not work to help push larger models onto FPGAs and will inevitably cause other problems with scheduling. I don't think we should support this type of...

Currently, `bram_size` (controlled by the `BramFactor` in the config) is the threshold above which weights will be implemented as BRAM ports (basically the `#pragma HLS INTERFACE bram port=weight_name`)

For example: ``` HLSConfig: Model: Precision: ap_fixed ReuseFactor: 42 BramFactor: 1337 ``` Now every layer whose number of weights is larger than 1337 will be implemented as BRAM ports. (The...

`CONFIG::n_scale_bias` approach seems far simpler, but are we sure we won't have to split the functions anyway for optimal QoR?