<table border="1" cellspacing="0" cellpadding="8">
    <tr>
        <th>Issue</th>
        <td>
            <a href=https://github.com/llvm/llvm-project/issues/56432>56432</a>
        </td>
    </tr>

    <tr>
        <th>Summary</th>
        <td>
            ML pretrained models are insufficiently opensource.
        </td>
    </tr>

    <tr>
      <th>Labels</th>
      <td>
            new issue
      </td>
    </tr>

    <tr>
      <th>Assignees</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Reporter</th>
      <td>
          Hello1024
      </td>
    </tr>
</table>

<pre>
    A build of LLVM with the ML optimizations enabled will by default include pretrained machine learning models made by Google:

https://github.com/llvm/llvm-project/blob/main/llvm/lib/Analysis/CMakeLists.txt#L8

That URL will be downloaded and included in the binary.

For a FOSS project, it's important that compiled-in files also meet the letter and spirit of the license, and it appears that unfortunately (and perhaps unintentionally) this file doesn't meet that test.

Specifically, it's a binary file built with an unpublished method, using closed source code operating on a closed source dataset.  There is available an open source tool, but this file is not produced with it, and it does not produce files with comparable performance.

Details here:
https://github.com/google/ml-compiler-opt/releases/tag/inlining-Oz-v1.1


Calling on LLVM maintainers (and the code author, @jacob-hegna ) to decide the correct course of action to take next.
</pre>
<img width="1px" height="1px" alt="" src="http://email.email.llvm.org/o/eJyVVMFu4zYQ_Rr7QliQZTuODjqkCdIeHCzQbHsfiSOJW5oUyFGy3q_vI-UEzqVAAYGkyOHMmzdv2Hp9aR5UOxurle_V6fT3i3o3MioZWb2clJ_EnM0vEuNdVOyotaxhYa1qL0pzT7MVZVxnZ81qCiyBjIPJmboRC2WZgjNuUGev2Ubsww5Xf_d-sLzaPazKp1V5HUeRKaa96hnfABxzW3T-jB9r3z6mzRT8D-4Ev631LaYzYt7YmLT34MheoolYPr7QP3wyUWIhP3Ftd7q_jfp9JFF__Xm6psVK-3dnPYBqRU5_ZJcWmZbWOAqX4tbFsw-K1PO311f1Ce5RGYzHqMx58kHICW4jEvKZDFjcwF2PRVRko1dnZsnuLYtwyJHjZIKRVJh8YDp2kZPnDEsUTRPojYvf2fUIMzsSthe1qu6T0cRhpCni0Dhhl8pI1l5WVY1LJmYAyJcj-DvKBwh4E47yJcXXiTvTm265_pkcXelYPCUhySIgcgg6za01cUx6YBm9ThfnmOTQWR-xHf0cOgYnUIUHWAgNh97B71cLTUKRpVDq-8iBFbDTGxmbBJli4bL7sBXvbYrUznKTJWbnJdVHz10WMVAauaEz8XBrcy1PNkxVo5CjASaYPpPr-AtDTywAFFXC9yns_5D0sLQA5Gs3V1GEDRoOO4HRN5GTeIUGjMZZk9po8-3X5m1bbG_jLuMjCnPlLndx6glJvQh9XMWQRJSZphnFCCnz1b78QZ1vNyMPjlTWhUdfdwZmi30IkDPmOUROUqQuqSiZCdpKOf4pxVo3O13valqLEcsNXo7bt2BpfUplc3HuoSIDLUKlqWpL0Yr1HGzzvx8AE-OcaTrc7XfVemz4qLWu-0N93Pfc9wfed-X2uKW6qnVFx-0aigGYZnX4bVVVjt9VdoH16vC0Nk1VVlV5LI9VudvuQPR9uT3UVV_y_eHujmvwxWDWFglH4cOwDk2G1M5DxKHNr8znIcVoBsecw8H_QnzzB1vrt2W1X-fgTQb_L8uY4SQ">