Added support for bias in optimized linear operation #9527
Conversation
DavitGrigoryan132
requested review from
manuelcandales and
swolchok
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/9527
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit b5b3702 with merge base 20abf34 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
Thank you for your pull request and welcome to our community. Action RequiredIn order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you. ProcessIn order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA. Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with If you have received this in error or have any questions, please contact us at [email protected]. Thanks! |
Signed-off-by: David Grigoryan <[email protected]>
DavitGrigoryan132
force-pushed
the
optimized_linear_op_bias
branch
from
9c392a9 to
b5b3702
Compare
facebook-github-bot
added
the
CLA Signed
|
@pytorchbot label "topic: not user facing" |
| for (size_t row = 0; row < m; ++row) { | ||
| out_ptr[col * m + row] = bias_ptr[row]; | ||
| } |
There was a problem hiding this comment.
this inner loop should probably be memcpy
| auto bias_ptr = bias_value.const_data_ptr<CTYPE>(); | ||
| CTYPE* out_ptr = out.mutable_data_ptr<CTYPE>(); | ||
| // Broadcast the bias to every column of the output. | ||
| for (size_t col = 0; col < n; ++col) { |
There was a problem hiding this comment.
use c10::irange. you'll need to include the header and add a buck dep. -- for (const auto col : c10::irange(n)) {
| op_linear_out(x, y, out); | ||
| op_linear_out(x, y, bias, out); | ||
|
|
||
| Tensor expected = tf.full({3, 5}, 192); | ||
| Tensor expected = tf.full({3, 5}, 193); | ||
|
|
||
| EXPECT_TENSOR_EQ(out, expected); |
There was a problem hiding this comment.
please add bias checking as a second op_linear_out call in this function rather than replacing the existing one; as written, the empty-bias path is now untested.
Summary
#8234
This PR is adding a functionality to use bias in optimized linear operation