Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Implement constraints API for MeanOp #2041

Closed
odjuricicTT opened this issue Jan 30, 2025 · 0 comments · Fixed by #2273
Closed

Implement constraints API for MeanOp #2041

odjuricicTT opened this issue Jan 30, 2025 · 0 comments · Fixed by #2273
Assignees
@arminaleTT
Milestone
[Performance - 2]...

Comments

@odjuricicTT
Copy link
Contributor

This is needed to run llama sharded.

Reference op constraints implementation: #1554
@odjuricicTT odjuricicTT added this to the [Performance - 2] March milestone Jan 30, 2025
@odjuricicTT odjuricicTT mentioned this issue Feb 4, 2025
Open
17 tasks
@arminaleTT arminaleTT mentioned this issue Feb 24, 2025
Merged
1 task
arminaleTT added a commit that referenced this issue Feb 26, 2025
@arminaleTT
Enable constraints and runtime measurement for Reshape and Mean Ops (#…
Loading
Loading status checks…
9ff6ac1 
…2273)

### Ticket
#2041 
#2042 

### Problem description
To enable llama to go through the optimizer, we need constraints and
runtime APIs for all ops (tracked in #2084). This PR enables these APIs
for `mean` and `reshape`.

### What's changed
- Added `getOpRuntime()` and `getOpConstraints()` interface methods to
`ReshapeOp` and `MeanOp`
- Added unit tests for both APIs on both Ops
- Closes #2041 
- Closes #2042

Note: This PR seems long for a "simple" integration change because both
APIs require a manual translation of mlir types down to ttnn types. As a
result, opportunities for sharing code across Ops is limited. About
1/3rd of the diff is in unit tests

### Checklist
- [x] New/Existing tests provide coverage for changes
jserbedzijaTT pushed a commit that referenced this issue Mar 1, 2025
@arminaleTT @jserbedzijaTT
Enable constraints and runtime measurement for Reshape and Mean Ops (#…
09ea275 
…2273)

### Ticket
#2041 
#2042 

### Problem description
To enable llama to go through the optimizer, we need constraints and
runtime APIs for all ops (tracked in #2084). This PR enables these APIs
for `mean` and `reshape`.

### What's changed
- Added `getOpRuntime()` and `getOpConstraints()` interface methods to
`ReshapeOp` and `MeanOp`
- Added unit tests for both APIs on both Ops
- Closes #2041 
- Closes #2042

Note: This PR seems long for a "simple" integration change because both
APIs require a manual translation of mlir types down to ttnn types. As a
result, opportunities for sharing code across Ops is limited. About
1/3rd of the diff is in unit tests

### Checklist
- [x] New/Existing tests provide coverage for changes
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@arminaleTT arminaleTT

Labels
None yet
Projects
None yet
Milestone
[Performance - 2] March
Development

Successfully merging a pull request may close this issue.

Enable constraints and runtime measurement for Reshape and Mean Ops tenstorrent/tt-mlir
2 participants
@odjuricicTT @arminaleTT