onnxruntime
Split KV on MHA and Attention ops
#18007
Merged

Commits
  • split kv on mha and att ops
    aciddelgado committed 2 years ago
  • update num_splits heuristic
    aciddelgado committed 2 years ago
  • Merge branch 'main' into aciddelgado/mha_splitkv
    aciddelgado committed 2 years ago
  • helper function for split buffers
    aciddelgado committed 2 years ago
  • Merge branch 'main' into aciddelgado/mha_splitkv
    aciddelgado committed 2 years ago
  • lint
    aciddelgado committed 2 years ago
Loading