-
Notifications
You must be signed in to change notification settings - Fork 280
[WC] WC/Mixed Precision/AWQ transpose_a support #3794
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
[WC] WC/Mixed Precision/AWQ transpose_a support #3794
Conversation
|
@andrey-churkin, please review updated AWQ ONNX backend |
| return transformed_model | ||
|
|
||
| def _get_activation_node_and_port(self, node: NNCFNode, nncf_graph: NNCFGraph) -> tuple[NNCFNode, int]: | ||
| def _get_activation_node_port_and_channel(self, node: NNCFNode, nncf_graph: NNCFGraph) -> tuple[NNCFNode, int]: |
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.
tuple[NNCFNode, int, int]
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.
Done
295211e to
7753663
Compare
|
|
||
| if any(weight_dim > 3 for weight_dim in all_weight_dims): | ||
| max_val = max(weight_dim for weight_dim in all_weight_dims) | ||
| msg = f"Compression with {max_val} dimentional weight is not supported" |
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.
we always keep convolutions with 4D weights in backup mode
@daniil-lyakhov is it valid condition for error?
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.
Removed
9017b66 to
1312d77
Compare
|
@andreyanufr @andrey-churkin, please take a look |
7ee7578 to
f829a24
Compare
|
@daniil-lyakhov Could you please start the |
andrey-churkin
left a comment
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.
LGTM
Sure |
Co-authored-by: andreyanufr <andrey.anufriev@intel.com>
6dee0b0
into
openvinotoolkit:develop
Changes
transpose_asupporttranspose_asupportprocess_statisticstranspose_aparam supportReason for changes
Related tickets
173277
Tests
transpose_aparam to check base WC/ mixed precision algotranspose_aandnon_mergable_patternto check non mergeable AWQ branch & activation transpose supportact_ch_axissupport inprocess_statisticsfntrahspose_amodel supplied to an algorithm