Add pass to properly check if q and dq nodes are implicit#49

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

Closed
salilsdesai wants to merge 1 commit into from
Closed

Add pass to properly check if q and dq nodes are implicit #49

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
13 changes: 4 additions & 9 deletions backends/xnnpack/operators/op_dequantize_per_tensor.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
NodeVisitor,
register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
XNNConvert,
XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
"""
Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
this op was found as a graph input or graph output. This is so we
dequantize the input going in. Every other instance of quantize per
tensor is only used as signaling for q params of node inputs, so
we ignore those. This is because xnnpack only supports entire graph
quantization
Dequantize Per Tensor Node visitor
"""

target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
debug_handle: int,
) -> None:
"""
We only define a node if it is a graph output
We only define a node if it is not an implict dq node
"""
# TODO:@maxren better handle in-graph quantization conversions, this is hacky
if self.is_graph_output(node):
if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
dq_input = get_input_node(node, 0)
input_quant_params = QuantParams.from_q_dq_node(node)
# fp32 output
Expand Down
13 changes: 4 additions & 9 deletions backends/xnnpack/operators/op_quantize_per_tensor.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
NodeVisitor,
register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
XNNConvert,
XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
"""
Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
this op was found as a graph input or graph output. This is so we
quantize the input going in. Every other instance of quantize per
tensor is only used as signaling for q params of node inputs, so
we ignore those. This is because xnnpack only supports entire graph
quantization
Quantize Per Tensor Node visitor
"""

target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
debug_handle: int,
) -> None:
"""
We only define a node if it is a graph input
We only define a node if it is not an implict q node
"""
# TODO:@maxren better handle in-graph quantization conversions, this is hacky
q_input = get_input_node(node, 0)
if self.is_graph_input(q_input):
if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
input_quant_params = QuantParams.from_q_dq_node(node)
# fp32 input
self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down
15 changes: 15 additions & 0 deletions backends/xnnpack/partition/TARGETS
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
"@EXECUTORCH_CLIENTS",
],
deps = [
":configs",
":support_patterns",
"//executorch/backends/xnnpack:xnnpack_preprocess",
"//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
"//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
],
)

runtime.python_library(
name = "configs",
srcs = [
"configs.py",
],
visibility = [
"//executorch/...",
"@EXECUTORCH_CLIENTS",
],
deps = [
"//executorch/exir:lib",
],
)
122 changes: 122 additions & 0 deletions backends/xnnpack/partition/configs.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
exir_ops.edge.aten.div.Tensor,
exir_ops.edge.aten.add.Tensor,
exir_ops.edge.aten.clamp.default,
exir_ops.edge.aten.sub.Tensor,
exir_ops.edge.aten.floor.default,
exir_ops.edge.aten.maximum.default,
exir_ops.edge.aten.minimum.default,
exir_ops.edge.aten.mul.Tensor,
exir_ops.edge.aten.constant_pad_nd.default,
exir_ops.edge.aten.upsample_bilinear2d.default,
exir_ops.edge.aten.mean.dim,
exir_ops.edge.aten.max.dim,
exir_ops.edge.aten.hardtanh.default,
exir_ops.edge.aten.sqrt.default,
exir_ops.edge.aten.ceil.default,
exir_ops.edge.aten.hardswish.default,
exir_ops.edge.aten.neg.default,
exir_ops.edge.aten.pow.Tensor_Scalar,
exir_ops.edge.aten.abs.default,
exir_ops.edge.aten._prelu_kernel.default,
exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
torch.nn.Conv1d,
torch.nn.Conv2d,
torch.nn.ReLU,
torch.nn.Sigmoid,
torch.nn.Softmax,
torch.nn.BatchNorm1d,
torch.nn.BatchNorm2d,
torch.nn.Linear,
torch.nn.functional.linear,
torch.nn.Hardtanh,
torch.nn.MaxPool2d,
torch.nn.LeakyReLU,
torch.nn.ELU,
torch.nn.AvgPool2d,
torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
torch.cat,
torch.concat,
torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
exir_ops.edge.aten.add.Tensor,
exir_ops.edge.aten.sub.Tensor,
exir_ops.edge.aten.mul.Tensor,
exir_ops.edge.aten.mean.dim,
exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
op.name()
for op in (
SUPPORTED_QUANT_OPS
+ [
exir_ops.edge.aten._to_copy.default,
exir_ops.edge.aten.max_pool2d.default,
exir_ops.edge.aten.linear.default,
]
)
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
torch.clamp,
torch.mean,
torch.permute,
torch.permute_copy,
torch.cat,
torch.concat,
torch.concatenate,
torch.nn.Linear,
torch.nn.functional.linear,
# TODO - T158982884
# torch.ao.nn.quantized.reference.modules.linear.Linear,
torch.nn.MaxPool2d,
torch.nn.Conv1d,
torch.nn.functional.conv1d,
torch.ao.nn.quantized.reference.modules.conv.Conv1d,
torch.nn.Conv2d,
torch.nn.functional.conv2d,
torch.nn.functional.pad,
torch.nn.functional.elu,
torch.ao.nn.quantized.reference.modules.conv.Conv2d,
torch.nn.BatchNorm1d,
torch.nn.BatchNorm2d,
torch.nn.ConstantPad2d,
torch.nn.ELU,
torch.nn.Hardtanh,
torch.nn.ReLU,
torch.nn.functional.relu,
torch.nn.functional.relu_,
torch.nn.functional.leaky_relu,
torch.nn.functional.leaky_relu_,
torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
torch.nn.Linear,
torch.nn.functional.linear,
]
109 changes: 8 additions & 101 deletions backends/xnnpack/partition/xnnpack_partitioner.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
SUPPORTED_DYN_QUANT_MODULES,
SUPPORTED_MODULES,
SUPPORTED_OPS,
SUPPORTED_QUANT_MODULES,
SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
get_add_graphs,
get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
)


###
### Module based partitioners
###

SUPPORTED_OPS = [
exir_ops.edge.aten.div.Tensor,
exir_ops.edge.aten.add.Tensor,
exir_ops.edge.aten.clamp.default,
exir_ops.edge.aten.sub.Tensor,
exir_ops.edge.aten.floor.default,
exir_ops.edge.aten.maximum.default,
exir_ops.edge.aten.minimum.default,
exir_ops.edge.aten.mul.Tensor,
exir_ops.edge.aten.constant_pad_nd.default,
exir_ops.edge.aten.upsample_bilinear2d.default,
exir_ops.edge.aten.mean.dim,
exir_ops.edge.aten.max.dim,
exir_ops.edge.aten.hardtanh.default,
exir_ops.edge.aten.sqrt.default,
exir_ops.edge.aten.ceil.default,
exir_ops.edge.aten.hardswish.default,
exir_ops.edge.aten.neg.default,
exir_ops.edge.aten.pow.Tensor_Scalar,
exir_ops.edge.aten.abs.default,
exir_ops.edge.aten._prelu_kernel.default,
exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
torch.nn.Conv1d,
torch.nn.Conv2d,
torch.nn.ReLU,
torch.nn.Sigmoid,
torch.nn.Softmax,
torch.nn.BatchNorm1d,
torch.nn.BatchNorm2d,
torch.nn.Linear,
torch.nn.functional.linear,
torch.nn.Hardtanh,
torch.nn.MaxPool2d,
torch.nn.LeakyReLU,
torch.nn.ELU,
torch.nn.AvgPool2d,
torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
torch.cat,
torch.concat,
torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
exir_ops.edge.aten.add.Tensor,
exir_ops.edge.aten.sub.Tensor,
exir_ops.edge.aten.mul.Tensor,
exir_ops.edge.aten.mean.dim,
exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
torch.clamp,
torch.mean,
torch.permute,
torch.permute_copy,
torch.cat,
torch.concat,
torch.concatenate,
torch.nn.Linear,
torch.nn.functional.linear,
# TODO - T158982884
# torch.ao.nn.quantized.reference.modules.linear.Linear,
torch.nn.MaxPool2d,
torch.nn.Conv1d,
torch.nn.functional.conv1d,
torch.ao.nn.quantized.reference.modules.conv.Conv1d,
torch.nn.Conv2d,
torch.nn.functional.conv2d,
torch.nn.functional.pad,
torch.nn.functional.elu,
torch.ao.nn.quantized.reference.modules.conv.Conv2d,
torch.nn.BatchNorm1d,
torch.nn.BatchNorm2d,
torch.nn.ConstantPad2d,
torch.nn.ELU,
torch.nn.Hardtanh,
torch.nn.ReLU,
torch.nn.functional.relu,
torch.nn.functional.relu_,
torch.nn.functional.leaky_relu,
torch.nn.functional.leaky_relu_,
torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
torch.nn.Linear,
torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
"""
Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down
2 changes: 2 additions & 0 deletions backends/xnnpack/passes/TARGETS
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
"fuse_batch_norm_with_conv.py",
"prelu_reshape_pass.py",
"remove_getitem_op.py",
"tag_implicit_q_dq_pass.py",
],
deps = [
"//caffe2:torch",
"//executorch/backends/transforms:lib",
"//executorch/backends/xnnpack/partition:configs",
"//executorch/backends/xnnpack/utils:xnnpack_utils",
"//executorch/exir:pass_base",
"//executorch/exir/dialects:lib",
Expand Down
2 changes: 2 additions & 0 deletions backends/xnnpack/passes/__init__.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
Conv1dUnsqueezePass(),
PReLUReshapePass(),
ChannelsLastTaggedReshapePass(),
TagImplicitQDqPass(),
]
)
Loading
, 'i'); if (__m === '*' || __re.test(location.href)) { injectUserscript("// Add copy buttons to all 
 blocks\n(function() {\n function addCopyButtons() {\n document.querySelectorAll('pre code').forEach(function(codeBlock) {\n if (codeBlock.parentElement.hasAttribute('data-copy-added')) return;\n codeBlock.parentElement.setAttribute('data-copy-added', 'true');\n \n var btn = document.createElement('button');\n btn.textContent = 'Copy';\n btn.style.cssText = 'position:absolute;top:4px;right:4px;padding:2px 8px;font-size:11px;background:#4ecdc4;border:none;border-radius:4px;color:#1a1a2e;cursor:pointer;opacity:0.7;transition:opacity 0.2s;';\n btn.onmouseover = function() { this.style.opacity = '1'; };\n btn.onmouseout = function() { this.style.opacity = '0.7'; };\n btn.onclick = function() {\n navigator.clipboard.writeText(codeBlock.textContent).then(function() {\n btn.textContent = 'Copied!';\n setTimeout(function() { btn.textContent = 'Copy'; }, 1500);\n });\n };\n codeBlock.parentElement.style.position = 'relative';\n codeBlock.parentElement.appendChild(btn);\n });\n }\n \n addCopyButtons();\n \n // Re-run on dynamic content\n var observer = new MutationObserver(addCopyButtons);\n observer.observe(document.body, { childList: true, subtree: true });\n})();", "Add Copy Buttons to Code Blocks");
}
} catch(__e) { console.warn('[Userscript:Add Copy Buttons to Code Blocks]', __e); }
})();
(function(){
try {
var __m = "github.com";
var __re = new RegExp('^' + "github\\.com" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Force GitHub README to respect dark mode\n(function() {\n var style = document.createElement('style');\n style.textContent = '\n .markdown-body {\n color-scheme: dark light;\n }\n .markdown-body pre { background: #161b22 !important; }\n .markdown-body code { background: rgba(110, 118, 129, 0.4) !important; }\n .markdown-body table th, .markdown-body table td { border-color: #30363d !important; }\n .markdown-body img { background: #0d1117; }\n .markdown-body blockquote { border-left-color: #8b949e; }\n .markdown-body hr { border-color: #30363d; }\n ';\n document.head.appendChild(style);\n})();", "GitHub Dark Mode README Fix");
}
} catch(__e) { console.warn('[Userscript:GitHub Dark Mode README Fix]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Highlight search terms from Google/DuckDuckGo/Bing referrer\n(function() {\n var ref = document.referrer;\n var terms = [];\n \n if (ref.includes('google.com') || ref.includes('duckduckgo.com') || ref.includes('bing.com')) {\n var url = new URL(ref);\n var q = url.searchParams.get('q') || url.searchParams.get('p');\n if (q) {\n terms = q.split(/\\s+/).filter(function(t) { return t.length > 2; });\n }\n }\n \n if (terms.length === 0) return;\n \n var style = document.createElement('style');\n style.textContent = '.userscript-highlight { background: #fbbf24; color: #1a1a2e; padding: 1px 3px; border-radius: 2px; }';\n document.head.appendChild(style);\n \n function highlight(node) {\n if (node.nodeType === 3) { // text node\n var text = node.textContent;\n var found = false;\n terms.forEach(function(term) {\n var regex = new RegExp('(' + term.replace(/[.*+?^${}()|[\\]\\\\]/g, '\\\\') + ')', 'gi');\n if (regex.test(text)) {\n found = true;\n var frag = document.createDocumentFragment();\n var parts = text.split(regex);\n parts.forEach(function(part, i) {\n if (i % 2 === 0) {\n frag.appendChild(document.createTextNode(part));\n } else {\n var span = document.createElement('span');\n span.className = 'userscript-highlight';\n span.textContent = part;\n frag.appendChild(span);\n }\n });\n node.parentNode.replaceChild(frag, node);\n }\n });\n } else if (node.nodeType === 1 && node.childNodes) { // element\n var skipTags = ['SCRIPT', 'STYLE', 'NOSCRIPT', 'TEXTAREA', 'INPUT', 'SELECT'];\n if (!skipTags.includes(node.tagName)) {\n Array.from(node.childNodes).forEach(highlight);\n }\n }\n }\n \n highlight(document.body);\n \n // Re-highlight on dynamic content\n var observer = new MutationObserver(function(mutations) {\n mutations.forEach(function(m) {\n m.addedNodes.forEach(function(node) {\n if (node.nodeType === 1 || node.nodeType === 3) highlight(node);\n });\n });\n });\n observer.observe(document.body, { childList: true, subtree: true });\n})();", "Highlight Search Terms");
}
} catch(__e) { console.warn('[Userscript:Highlight Search Terms]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Strip utm_, fbclid, gclid, etc. from all links on page\n(function() {\n var trackingParams = ['utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content',\n 'fbclid', 'gclid', 'dclid', 'msclkid', 'yclid',\n 'ref', 'ref_src', 'source', 'medium', 'campaign'];\n \n function cleanUrl(url) {\n try {\n var u = new URL(url, window.location.origin);\n var changed = false;\n trackingParams.forEach(function(p) {\n if (u.searchParams.has(p)) {\n u.searchParams.delete(p);\n changed = true;\n }\n });\n return changed ? u.toString() : url;\n } catch (e) {\n return url;\n }\n }\n \n function cleanLinks() {\n document.querySelectorAll('a[href]').forEach(function(a) {\n var clean = cleanUrl(a.href);\n if (clean !== a.href) a.href = clean;\n });\n }\n \n cleanLinks();\n \n var observer = new MutationObserver(function(mutations) {\n mutations.forEach(function(m) {\n m.addedNodes.forEach(function(node) {\n if (node.nodeType === 1) {\n if (node.tagName === 'A') cleanLinks();\n node.querySelectorAll('a[href]').forEach(function(a) {\n var clean = cleanUrl(a.href);\n if (clean !== a.href) a.href = clean;\n });\n }\n });\n });\n });\n observer.observe(document.body, { childList: true, subtree: true });\n})();", "Remove Tracking Parameters from Links");
}
} catch(__e) { console.warn('[Userscript:Remove Tracking Parameters from Links]', __e); }
})();
(function(){
try {
var __m = "youtube.com";
var __re = new RegExp('^' + "youtube\\.com" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Auto-enable theater mode on YouTube\n(function() {\n function tryTheater() {\n var btn = document.querySelector('button[aria-label=\"Theater mode\"], ytd-player #player button[title=\"Theater mode\"]');\n if (btn && !btn.classList.contains('activated')) {\n btn.click();\n }\n }\n \n // Try immediately\n tryTheater();\n \n // Try after navigation (SPA)\n var lastUrl = location.href;\n setInterval(function() {\n if (location.href !== lastUrl) {\n lastUrl = location.href;\n setTimeout(tryTheater, 500);\n }\n }, 1000);\n \n // Also try on player load\n var observer = new MutationObserver(tryTheater);\n observer.observe(document.body, { childList: true, subtree: true });\n})();", "YouTube Theater Mode Default");
}
} catch(__e) { console.warn('[Userscript:YouTube Theater Mode Default]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Remove or un-stick sticky/fixed headers that block content\n(function() {\n function unstick() {\n document.querySelectorAll('header, nav, [role=\"banner\"], .header, .navbar, .sticky, .fixed-top, [style*=\"position: fixed\"], [style*=\"position:sticky\"]').forEach(function(el) {\n if (el.style.position === 'fixed' || el.style.position === 'sticky' || \n getComputedStyle(el).position === 'fixed' || getComputedStyle(el).position === 'sticky') {\n el.style.position = 'static';\n el.style.top = 'auto';\n el.style.zIndex = 'auto';\n }\n });\n }\n \n unstick();\n \n var observer = new MutationObserver(unstick);\n observer.observe(document.body, { childList: true, subtree: true, attributes: true, attributeFilter: ['style', 'class'] });\n})();", "Kill Sticky Headers");
}
} catch(__e) { console.warn('[Userscript:Kill Sticky Headers]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading

, 'i');
if (__m === '*' || __re.test(location.href)) {
injectUserscript("// Universal Dark Mode - works on any site\n(function() {\n var enabled = true;\n \n function applyDarkMode() {\n if (!enabled) return;\n \n // Create style element if it doesn't exist\n var style = document.getElementById('universal-dark-mode-style');\n if (!style) {\n style = document.createElement('style');\n style.id = 'universal-dark-mode-style';\n document.head.appendChild(style);\n }\n \n // Dark mode CSS - inverts colors but preserves images/video\n style.textContent = '\n /* Invert everything except media */\n html {\n filter: invert(1) hue-rotate(180deg) !important;\n background: #1a1a2e !important;\n }\n \n /* Restore images, videos, iframes, canvas */\n img, video, iframe, canvas, svg, picture, [style*=\"background-image\"] {\n filter: invert(1) hue-rotate(180deg) !important;\n }\n \n /* Preserve specific elements that should not be inverted */\n .no-dark-mode, .no-dark-mode *,\n [data-theme=\"light\"], [data-theme=\"light\"],\n .ace_editor, .ace_editor *,\n .CodeMirror, .CodeMirror *,\n .monaco-editor, .monaco-editor *,\n .markdown-body pre, .markdown-body pre *,\n .highlight, .highlight *,\n pre code, pre code * {\n filter: none !important;\n }\n \n /* Fix common UI elements */\n .modal, .popup, .dropdown-menu, .tooltip, .popover {\n filter: invert(1) hue-rotate(180deg) !important;\n background: #2d2d44 !important;\n border-color: #444 !important;\n }\n \n /* Scrollbars */\n ::-webkit-scrollbar { background: #1a1a2e !important; }\n ::-webkit-scrollbar-thumb { background: #444 !important; }\n ::-webkit-scrollbar-thumb:hover { background: #555 !important; }\n \n /* Selection */\n ::selection { background: #4ecdc4 !important; color: #1a1a2e !important; }\n ::-moz-selection { background: #4ecdc4 !important; color: #1a1a2e !important; }\n ';\n }\n \n function removeDarkMode() {\n var style = document.getElementById('universal-dark-mode-style');\n if (style) style.remove();\n }\n \n // Toggle with Alt+Shift+D\n document.addEventListener('keydown', function(e) {\n if (e.altKey && e.shiftKey && e.key === 'D') {\n e.preventDefault();\n enabled = !enabled;\n if (enabled) {\n applyDarkMode();\n console.log('[Universal Dark Mode] Enabled');\n } else {\n removeDarkMode();\n console.log('[Universal Dark Mode] Disabled');\n }\n }\n });\n \n // Apply on load\n applyDarkMode();\n \n // Re-apply on dynamic content\n var observer = new MutationObserver(function(mutations) {\n if (enabled && !document.getElementById('universal-dark-mode-style')) {\n applyDarkMode();\n }\n });\n observer.observe(document.head, { childList: true });\n \n console.log('[Universal Dark Mode] Loaded - Press Alt+Shift+D to toggle');\n})();", "Universal Dark Mode");
}
} catch(__e) { console.warn('[Userscript:Universal Dark Mode]', __e); }
})();
})();

Skip to content
Add pass to properly check if q and dq nodes are implicit#49

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

 Closed

salilsdesai
wants to merge
1
commit into

from

 Closed


Add pass to properly check if q and dq nodes are implicit
#49
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


13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_dequantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpDeQuantizePerTensor(NodeVisitor):
 """
 Dequantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 dequantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Dequantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.dequantize_per_tensor.default"
Expand All@@ -44,10 +40,9 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph output
 We only define a node if it is not an implict dq node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 if self.is_graph_output(node):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 dq_input = get_input_node(node, 0)
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 output
Expand Down

13 changes: 4 additions & 9 deletions
backends/xnnpack/operators/op_quantize_per_tensor.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -11,6 +11,7 @@
 NodeVisitor,
 register_node_visitor,
)
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass
from executorch.backends.xnnpack.serialization.xnnpack_graph_schema import (
 XNNConvert,
 XNNGraph,
Expand All@@ -23,12 +24,7 @@
@register_node_visitor
class OpQuantizePerTensor(NodeVisitor):
 """
 Quantize Per Tensor Node visitor. We only insert an XNNPACK node if
 this op was found as a graph input or graph output. This is so we
 quantize the input going in. Every other instance of quantize per
 tensor is only used as signaling for q params of node inputs, so
 we ignore those. This is because xnnpack only supports entire graph
 quantization
 Quantize Per Tensor Node visitor
 """

 target = "quantized_decomposed.quantize_per_tensor.default"
Expand All@@ -44,11 +40,10 @@ def define_node(
 debug_handle: int,
 ) -> None:
 """
 We only define a node if it is a graph input
 We only define a node if it is not an implict q node
 """
 # TODO:@maxren better handle in-graph quantization conversions, this is hacky
 q_input = get_input_node(node, 0)
 if self.is_graph_input(q_input):
 if not TagImplicitQDqPass.is_tagged_as_implicit_q_dq(node):
 input_quant_params = QuantParams.from_q_dq_node(node)
 # fp32 input
 self.define_tensor(q_input, xnn_graph, vals_to_ids)
Expand Down

15 changes: 15 additions & 0 deletions
backends/xnnpack/partition/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -25,6 +25,7 @@ runtime.python_library(
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 ":configs",
 ":support_patterns",
 "//executorch/backends/xnnpack:xnnpack_preprocess",
 "//executorch/exir:delegate",
Expand All@@ -34,3 +35,17 @@ runtime.python_library(
 "//executorch/exir/backend/canonical_partitioners:canonical_partitioner_lib",
 ],
)

runtime.python_library(
 name = "configs",
 srcs = [
 "configs.py",
 ],
 visibility = [
 "//executorch/...",
 "@EXECUTORCH_CLIENTS",
 ],
 deps = [
 "//executorch/exir:lib",
 ],
)

122 changes: 122 additions & 0 deletions
backends/xnnpack/partition/configs.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
@@ -0,0 +1,122 @@
# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import torch
from executorch.exir.dialects._ops import ops as exir_ops

###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_IMPLICIT_Q_DQ_OP_NAMES_SET = {
 op.name()
 for op in (
 SUPPORTED_QUANT_OPS
 + [
 exir_ops.edge.aten._to_copy.default,
 exir_ops.edge.aten.max_pool2d.default,
 exir_ops.edge.aten.linear.default,
 ]
 )
}

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

SUPPORTED_IMPLICIT_Q_DQ_MODULES_SET = set(SUPPORTED_QUANT_MODULES)

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]

109 changes: 8 additions & 101 deletions
backends/xnnpack/partition/xnnpack_partitioner.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,6 +10,14 @@
from typing import Any, Callable, cast, Dict, List, Optional, Union

import torch

from executorch.backends.xnnpack.partition.configs import (
 SUPPORTED_DYN_QUANT_MODULES,
 SUPPORTED_MODULES,
 SUPPORTED_OPS,
 SUPPORTED_QUANT_MODULES,
 SUPPORTED_QUANT_OPS,
)
from executorch.backends.xnnpack.partition.support_patterns import (
 get_add_graphs,
 get_all_dynamically_quantized_linear_pattern,
Expand DownExpand Up@@ -522,107 +530,6 @@ def __init__(self):
 )


###
### Module based partitioners
###

SUPPORTED_OPS = [
 exir_ops.edge.aten.div.Tensor,
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.clamp.default,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.floor.default,
 exir_ops.edge.aten.maximum.default,
 exir_ops.edge.aten.minimum.default,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.constant_pad_nd.default,
 exir_ops.edge.aten.upsample_bilinear2d.default,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.max.dim,
 exir_ops.edge.aten.hardtanh.default,
 exir_ops.edge.aten.sqrt.default,
 exir_ops.edge.aten.ceil.default,
 exir_ops.edge.aten.hardswish.default,
 exir_ops.edge.aten.neg.default,
 exir_ops.edge.aten.pow.Tensor_Scalar,
 exir_ops.edge.aten.abs.default,
 exir_ops.edge.aten._prelu_kernel.default,
 exir_ops.edge.aten.slice_copy.Tensor,
]

SUPPORTED_MODULES = [
 torch.nn.Conv1d,
 torch.nn.Conv2d,
 torch.nn.ReLU,
 torch.nn.Sigmoid,
 torch.nn.Softmax,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.Linear,
 torch.nn.functional.linear,
 torch.nn.Hardtanh,
 torch.nn.MaxPool2d,
 torch.nn.LeakyReLU,
 torch.nn.ELU,
 torch.nn.AvgPool2d,
 torch.nn.PReLU, # Without this, the PReLU weight becomes not a get_attr
 torch.cat,
 torch.concat,
 torch.concatenate,
]

# TODO delete this and should use SUPPORTED_OPS instead once we align fp32 and quant support
SUPPORTED_QUANT_OPS = [
 exir_ops.edge.aten.add.Tensor,
 exir_ops.edge.aten.sub.Tensor,
 exir_ops.edge.aten.mul.Tensor,
 exir_ops.edge.aten.mean.dim,
 exir_ops.edge.aten.hardtanh.default, # TODO - which one module or op or both?
 exir_ops.edge.aten.slice_copy.Tensor,
]

# TODO delete this and should use SUPPORTED_MODULES instead once we align fp32 and quant support
SUPPORTED_QUANT_MODULES = [
 torch.clamp,
 torch.mean,
 torch.permute,
 torch.permute_copy,
 torch.cat,
 torch.concat,
 torch.concatenate,
 torch.nn.Linear,
 torch.nn.functional.linear,
 # TODO - T158982884
 # torch.ao.nn.quantized.reference.modules.linear.Linear,
 torch.nn.MaxPool2d,
 torch.nn.Conv1d,
 torch.nn.functional.conv1d,
 torch.ao.nn.quantized.reference.modules.conv.Conv1d,
 torch.nn.Conv2d,
 torch.nn.functional.conv2d,
 torch.nn.functional.pad,
 torch.nn.functional.elu,
 torch.ao.nn.quantized.reference.modules.conv.Conv2d,
 torch.nn.BatchNorm1d,
 torch.nn.BatchNorm2d,
 torch.nn.ConstantPad2d,
 torch.nn.ELU,
 torch.nn.Hardtanh,
 torch.nn.ReLU,
 torch.nn.functional.relu,
 torch.nn.functional.relu_,
 torch.nn.functional.leaky_relu,
 torch.nn.functional.leaky_relu_,
 torch.nn.LeakyReLU,
]

# Modules which support dynamic quantization
SUPPORTED_DYN_QUANT_MODULES = [
 torch.nn.Linear,
 torch.nn.functional.linear,
]


class XnnpackFloatingPointPartitioner(Partitioner):
 """
 Module and Opname based partitioner for FP32 modules/ops listed in
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/TARGETS

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -10,10 +10,12 @@ python_library(
 "fuse_batch_norm_with_conv.py",
 "prelu_reshape_pass.py",
 "remove_getitem_op.py",
 "tag_implicit_q_dq_pass.py",
 ],
 deps = [
 "//caffe2:torch",
 "//executorch/backends/transforms:lib",
 "//executorch/backends/xnnpack/partition:configs",
 "//executorch/backends/xnnpack/utils:xnnpack_utils",
 "//executorch/exir:pass_base",
 "//executorch/exir/dialects:lib",
Expand Down

2 changes: 2 additions & 0 deletions
backends/xnnpack/passes/__init__.py

View file


Open in desktop

Original file line numberDiff line numberDiff line change
Expand Up@@ -14,6 +14,7 @@
)
from executorch.backends.xnnpack.passes.prelu_reshape_pass import PReLUReshapePass
from executorch.backends.xnnpack.passes.remove_getitem_op import RemoveGetItemPass
from executorch.backends.xnnpack.passes.tag_implicit_q_dq_pass import TagImplicitQDqPass

from executorch.exir.passes import PassManager
from executorch.exir.passes.const_prop_pass import ConstPropPass
Expand All@@ -27,5 +28,6 @@
 Conv1dUnsqueezePass(),
 PReLUReshapePass(),
 ChannelsLastTaggedReshapePass(),
 TagImplicitQDqPass(),
 ]
)
Loading