Skip to content
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

Jump to bottom

[AutoTVM] Minor bug fixes in AutoTVM for QNN graphs #4797

Merged
merged 3 commits into from
Feb 4, 2020
Merged

[AutoTVM] Minor bug fixes in AutoTVM for QNN graphs #4797

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5e0c672
[AutoTVM] Minor bug fixes in AutoTVM for QNN graphs.
anijain2305 Feb 1, 2020
4541518
Bring back strided_slice.
anijain2305 Feb 4, 2020
5d22a56
Replace tvm.nd change.
anijain2305 Feb 4, 2020
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
7 changes: 4 additions & 3 deletions python/tvm/autotvm/graph_tuner/utils/traverse_graph.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,10 +126,10 @@ def _traverse_expr(node):
for i, input_idx in enumerate(node_entry["inputs"]):
input_node_entry = node_list[input_idx[0]]
input_type = input_node_entry["types"][input_idx[1]]
if not isinstance(input_node_entry["node"], (Var, Call)):
if not isinstance(input_node_entry["node"], (Var, Constant, Call)):
raise RuntimeError("Graph tuner can only tune target "
"operators with input node of type "
"relay.expr.Var or relay.expr.Call. Now "
"relay.expr.Var/Constant/Call. Now "
"find a target op %s with input type %s"
% (op_name, str(type(input_node_entry["node"]))))
free_var = relay.Var("var_%d" % i, input_type)
Expand Down Expand Up @@ -167,7 +167,8 @@ def _traverse_expr(node):
else:
node_entry["inputs"].append([in_node_idx, 0, 0])
elif isinstance(node, Constant):
pass
node_entry["name"] = "Constant_" + str(node_index)
node_entry["types"] = [node.checked_type]
elif isinstance(node, relay.op.op.Op):
return
else:
Expand Down
2 changes: 1 addition & 1 deletion python/tvm/autotvm/graph_tuner/utils/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,7 +73,7 @@ def is_boundary_node(node_entry, input_names):
# Operators dependent on original layouts.
_LAYOUT_FIXED_OP = ["batch_flatten", "transpose", "reshape",
"multibox_prior", "multibox_transform_loc", "where",
"non_max_suppression", "strided_slice"]
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do we want to change this?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This was causing problem in quantized mobilenet conv2d. QNN inserts strided slice for quantized conv2d. This caused many strided slice, and somehow graph tuning dependency map generation was incorrect. We already have a stridedSliceInferLayout attritbute, so I don't think this is required. But, let me know, I dont fully understand the graph tuner dependency collection code yet. If you know, why this is here, I can also test the changes.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

After talking to Yao, we decided to solve it in a different manner. It might work for the case that I experienced with QNN, but might fail in other cases. We will revisit this.

"non_max_suppression"]

out = node_entry["op"] in _LAYOUT_FIXED_OP or \
("name" in node_entry and node_entry["name"] in input_names)
Expand Down
7 changes: 5 additions & 2 deletions python/tvm/autotvm/task/relay_integration.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,8 @@ def _lower(mod,
grc.codegen(mod["main"])
# default case
compiler = relay.vm.VMCompiler()
compiler.set_params(params)
if params:
compiler.set_params(params)
compiler.lower(mod, target=target)


Expand Down Expand Up @@ -123,7 +124,9 @@ def extract_from_multiple_program(mods, params, ops, target, target_host=None,
# relay op -> topi compute
OP2TOPI = {
tvm.relay.op.nn.conv2d: [topi.nn.conv2d, topi.nn.depthwise_conv2d_nchw,
topi.nn.group_conv2d_nchw, topi.nn.conv2d_NCHWc],
topi.nn.group_conv2d_nchw,
topi.nn.conv2d_NCHWc,
topi.nn.conv2d_NCHWc_int8],
tvm.relay.op.nn.conv2d_transpose: [topi.nn.conv2d_transpose_nchw],
tvm.relay.op.nn.dense: [topi.nn.dense],
tvm.relay.op.nn.batch_matmul: [topi.nn.batch_matmul],
Expand Down
2 changes: 1 addition & 1 deletion python/tvm/relay/backend/vm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -422,7 +422,7 @@ def set_params(self, params):
inputs = {}
for name, param in params.items():
if isinstance(param, np.ndarray):
param = _nd.array(param)
param = tvm.nd.array(param)
inputs[name] = _expr.const(param)
self._set_params_func(inputs)

Expand Down