Add quantizer support for quantized_conv1d_ncl#17938
Add quantizer support for quantized_conv1d_ncl#17938mcremon-meta wants to merge 2 commits intomainfrom
Conversation
Summary: Remove typed quantized_conv1d operator variants and their implementations to start fresh. This removes: - Type dispatch logic from type_dispatch.py - Type dispatch tests from test_type_dispatch_passes.py - Kernel registrations from functions.yaml and functions_hifi.yaml - Operator definitions from ops_registrations.py - Reference implementations from ref_implementations.py - C++ implementations from generic and HiFi operators - Jarvis custom op definitions and argument checks Differential Revision: D95277804
Summary: Update quantizer patterns to route conv1d operations to quantized_conv1d_ncl instead of quantized_conv2d_nchw: - Conv1dPattern.replacement_op() now returns quantized_conv1d_ncl - Conv1dReluPattern0 and Conv1dReluPattern1 now return quantized_conv1d_ncl - Added fused_conv1d_relu test and _build_conv1d_relu_graph helper Differential Revision: D95279330
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/17938
Note: Links to docs will display an error until the docs builds have been completed. ❌ 9 New Failures, 5 Unrelated FailuresAs of commit 8588fb4 with merge base 2cb1ef5 ( NEW FAILURES - The following jobs have failed:
FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@mcremon-meta has exported this pull request. If you are a Meta employee, you can view the originating Diff in D95279330. |
This PR needs a
|
Summary:
Update quantizer patterns to route conv1d operations to quantized_conv1d_ncl instead of quantized_conv2d_nchw:
Differential Revision: D95279330