Auto mixed precision conversion of GPT-2 onnx model #10711
add auto mixed precision
d0640dc4
Add float_to_float16_max_diff, update constant
680f301c
remove cascaded Cast nodes
43ca36aa
format
7e63ab73
add -a option
72392e1f
viboga
commented
on 2022-03-02
viboga
commented
on 2022-03-02
viboga
commented
on 2022-03-02
viboga
approved these changes
on 2022-03-02
tianleiwu
merged
47ab0c20
into master 4 years ago
tianleiwu
deleted the tlwu/gpt2_mixed_precision branch 4 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub