5 ms·

Hm, when I run the example, I get this error: > expected scalar type BFloat16 but found Float Has anyone seen this error? It's pretty hard to google for.

• nathas 4y ago
  Yeah. Try running with PYTORCH_ENABLE_MPS_FALLBACK=1 <script> --full-precision
  • johnfn 4y ago
    This worked!
  • yboris 4y ago
    For me `--full-precision` kept erroring out, but `--precision full` worked correctly.
• bfirsh 4y ago
  Are you running macOS >=12.3?
  • johnfn 4y ago
    Oh, no I'm not, I'm on 12.0. Does this make a difference?
• johnfn 4y ago
  Update: I solved this error more properly by upgrading to the latest version. Thanks bfirsh.
  • wuyishan 4y ago
    I am having the same issue on MacOS 12.2.1 (21D62); Python 3.10.6 What did you upgrade to solve this? Thanks! (I can get it working with `--precision full`)
    • johnfn 4y ago
      I'm now on 12.5.1