Added support for named attrs in DecomposingContext. #213
Conversation
vladimirjovanovicTT
requested review from
nvukobratTT,
nsmithtt,
pilkicTT and
sdjordjevicTT
vladimirjovanovicTT
force-pushed
the
vladimirjovanovic/named_attrs_in_decomposing_context
branch
from
58d5795 to
7e4fbb2
Compare
| @@ -12,7 +12,7 @@ jobs: | |||
| fail-fast: false | |||
| matrix: | |||
| build: | |||
| - runs-on: self-hosted | |||
| - runs-on: runner | |||
There was a problem hiding this comment.
Is this intentional?
There was a problem hiding this comment.
Yup, this is required now that runners have been re-configured. However, I think there is separate PR to tackle this change
There was a problem hiding this comment.
Yes, I needed to set this up here in order to run CI.
| @@ -12,7 +12,7 @@ jobs: | |||
| fail-fast: false | |||
| matrix: | |||
| build: | |||
| - runs-on: self-hosted | |||
| - runs-on: runner | |||
There was a problem hiding this comment.
Yup, this is required now that runners have been re-configured. However, I think there is separate PR to tackle this change
There was a problem hiding this comment.
Looks good, just sync with @vmilosevic on who will push the runner update change :))
|
The runner is already updated on main, so it is okay. If there is any difference, it will show as a conflict; otherwise, it should merge without problems. |
nvukobratTT
deleted the
vladimirjovanovic/named_attrs_in_decomposing_context
branch
Added support for named attrs in DecomposingContext, since they are needed for lowering to MLIR.
Currently using it only in softmax_bw, as an example. As we find more uses for it (as training matures), we will add it at other locations.
Solves #183