-
Notifications
You must be signed in to change notification settings - Fork 97
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.
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
Change create_tt_tensor_from_py_data to use from_vector #16999
Conversation
…b borrow handling to helper method and clarify it
f8e0eca
to
4352603
Compare
…/github.com/tenstorrent/tt-metal into jjiang/16837-tensor_creation_and_conversion
…b borrow handling to helper method and clarify it
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The files seem to have a bunch of un-related changes, such as adding ttnn/cpp/ttnn/operations/data_movement/permute/device/kernels/compute/transpose_xw_tiled.cpp which I'm pretty sure is in main.
Please update the PR appropriately to only have relevant changes.
models/bringup_testing/README.md
Outdated
|
||
## Model Bring-Up | ||
|
||
- **Adding a Model** - |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Will these sections be filled in as part of this PR?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry, it seems like I messed up a few rebases somewhere along the way by accidentally merging which added a bunch of files to my changelog. What's the best way to fix this? Git complains when I try to simply revert the relevant commits.
I guess I could just cherry pick all of my actual commits to main and then delete this branch when the time comes to merge?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Cherry-pick this on top of main would work.
…and oddshape testing, move shape testing into roundtrip
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nice! Just a few nit comments, the code is almost ready!
I do see your post commit is failing for 2 test cases though (the rest of failures seem to be due to timeouts):
- https://github.com/tenstorrent/tt-metal/actions/runs/12918830342/job/36028847154 (just a single test case in test_tensor_creation_with_memory_config)
- https://github.com/tenstorrent/tt-metal/actions/runs/12918830342/job/36028840916 (I think this is the test you added here?)
…tal type error, improve bfloat16 conversions
Ticket
#16837
Problem description
Duplicate bespoke code in pytensor.cpp for bfloat handling, should probably use the tensor.cpp version for better maintainability.
What's changed
Call from_vector within pytensor.cpp instead of directly doing bfloat packing in pytensor.cpp, clean up from_span and create_owned_tensor_from_row_major_data helper method in tensor.cpp, add test cases to cover the new path.
Checklist