Skip to content

Remove type dispatch for quantized_conv1d ops #17939

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

Open
mcremon-meta wants to merge 1 commit into
base: main
Choose a base branch
from
+1,655 −912
Open

Remove type dispatch for quantized_conv1d ops #17939

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion backends/cadence/aot/BUCK
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -154,7 +154,8 @@ fbcode_target(_kind = executorch_generated_lib,
"//executorch/backends/cadence/generic/operators:op_quantize_per_tensor",
"//executorch/backends/cadence/generic/operators:op_quantized_add",
"//executorch/backends/cadence/generic/operators:op_quantized_conv2d",
"//executorch/backends/cadence/generic/operators:op_quantized_conv1d",
"//executorch/backends/cadence/generic/operators:op_quantized_conv1d_ncl",
"//executorch/backends/cadence/generic/operators:op_quantized_conv1d_nlc",
"//executorch/backends/cadence/generic/operators:op_quantized_fully_connected",
"//executorch/backends/cadence/generic/operators:op_quantized_layer_norm",
"//executorch/backends/cadence/generic/operators:op_quantized_linear",
Expand Down
30 changes: 10 additions & 20 deletions backends/cadence/aot/functions.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -359,6 +359,16 @@
- arg_meta: null
kernel_name: impl::generic::quantized_conv2d_nhwc_per_tensor_out

- func: cadence::quantized_conv1d_ncl.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_ncl_per_tensor_out

- func: cadence::quantized_conv1d_nlc.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_nlc_per_tensor_out

- func: cadence::quantized_conv2d_nchw_asym8sxsym8s_asym8s.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
Expand Down Expand Up @@ -419,26 +429,6 @@
- arg_meta: null
kernel_name: impl::generic::quantized_conv2d_nhwc_depthwise_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_conv1d_ncl_asym8sxsym8s_asym8s.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_ncl_asym8sxsym8s_asym8s_per_tensor_out

- func: cadence::quantized_conv1d_ncl_asym8uxsym8u_asym8u.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_ncl_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_conv1d_nlc_asym8sxsym8s_asym8s.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_nlc_asym8sxsym8s_asym8s_per_tensor_out

- func: cadence::quantized_conv1d_nlc_asym8uxsym8u_asym8u.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::generic::quantized_conv1d_nlc_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_fully_connected.out(Tensor src, Tensor weight, Tensor bias, int src_zero_point, Tensor weight_zero_point, Tensor out_multiplier, Tensor out_shift, int out_zero_point, Tensor? offset, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
Expand Down
30 changes: 10 additions & 20 deletions backends/cadence/aot/functions_hifi.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -445,26 +445,6 @@
- arg_meta: null
kernel_name: impl::HiFi::quantized_conv2d_nhwc_depthwise_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_conv1d_ncl_asym8sxsym8s_asym8s.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::quantized_conv1d_ncl_asym8sxsym8s_asym8s_per_tensor_out

- func: cadence::quantized_conv1d_ncl_asym8uxsym8u_asym8u.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::quantized_conv1d_ncl_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_conv1d_nlc_asym8sxsym8s_asym8s.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::quantized_conv1d_nlc_asym8sxsym8s_asym8s_per_tensor_out

- func: cadence::quantized_conv1d_nlc_asym8uxsym8u_asym8u.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::quantized_conv1d_nlc_asym8uxsym8u_asym8u_per_tensor_out

- func: cadence::quantized_layer_norm.out(Tensor input, Tensor in_scale, Tensor in_zero_point, int[] normalized_shape, Tensor weight, Tensor bias, float eps, float output_scale, int output_zero_point, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
Expand Down Expand Up @@ -563,3 +543,13 @@
kernels:
- arg_meta: null
kernel_name: impl::HiFi::quantized_fully_connected_asym8uxasym8u_asym8u_per_tensor_out

- func: cadence::quantized_conv1d_ncl.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::native::quantized_conv1d_ncl_per_tensor_out

- func: cadence::quantized_conv1d_nlc.per_tensor_out(Tensor input, Tensor weight, Tensor bias, int[] stride, SymInt[] padding, int[] dilation, int groups, int input_zero_point, int weight_zero_point, float bias_scale, float out_scale, int out_zero_point, int out_multiplier, int out_shift, *, Tensor(a!) out) -> Tensor(a!)
kernels:
- arg_meta: null
kernel_name: impl::HiFi::native::quantized_conv1d_nlc_per_tensor_out
Loading
Loading