Bringing Up DeepSeek-V4-Flash on AMD MI300X

6 comments

• maCDzP an hour ago ago

  I train on AMD MI250X and managed to get Gemma 4 31B to work - but it took a lot of work on the software side.

  [-]
  • kkm an hour ago ago

    This is very interesting, planning to write about it?

• kkm 3 hours ago ago

  Also the vllm patch accompanying the blogpost: https://github.com/doublewordai/vllm-amd-blog-doubleword

• mezark 3 hours ago ago

  We at doubleword are bullish for AMD for low-interactivity inference - it does just take a bigger lift on the software side...

  [-]
  • brcmthrowaway an hour ago ago

    Are you long AMD?

• benlm 3 hours ago ago

  Nice work! Would DeepSeek V4 Pro on 8xMI300X work with these patches?