tico 0.1.0.dev250604__py3-none-any.whl → 0.1.0.dev250608__py3-none-any.whl

tico/__init__.py

@@ -21,7 +21,7 @@ from tico.config import CompileConfigV1, get_default_config
 from tico.utils.convert import convert, convert_from_exported_program, convert_from_pt2
 
 # THIS LINE IS AUTOMATICALLY GENERATED BY setup.py
-__version__ = "0.1.0.dev250604"
+__version__ = "0.1.0.dev250608"
 
 MINIMUM_SUPPORTED_VERSION = "2.5.0"
 SECURE_TORCH_VERSION = "2.6.0"

tico/passes/remove_redundant_reshape.py

@@ -25,7 +25,12 @@ from tico.utils import logging
 from tico.utils.passes import PassBase, PassResult
 from tico.utils.trace_decorators import trace_graph_diff_on_pass
 from tico.utils.utils import set_new_meta_val
-from tico.utils.validate_args_kwargs import ReshapeArgs
+from tico.utils.validate_args_kwargs import (
+    AddTensorArgs,
+    PermuteArgs,
+    ReshapeArgs,
+    SoftmaxArgs,
+)
 
 
 def passes():
@@ -70,8 +75,8 @@ class RemoveRedundantReshapePattern1(PassBase):
             # Assumes that other node do not use ops in the pattern for simplisity.
             if len(reshape1.users) != 1:
                 continue
-            assert len(reshape1.args) == 2, len(reshape1.args)
-            reshape1_input, reshape1_size = reshape1.args
+            reshape1_args = ReshapeArgs(*reshape1.args, **reshape1.kwargs)  # type: ignore[arg-type]
+            reshape1_input, reshape1_size = reshape1_args.input, reshape1_args.size
             # `(AxBxC) - aten.reshape` - (1xAxBxC)
             if [1] + list(extract_shape(reshape1_input)) != list(
                 extract_shape(reshape1)
@@ -84,8 +89,8 @@ class RemoveRedundantReshapePattern1(PassBase):
                 continue
             if len(permute.users) != 1:
                 continue
-            assert len(permute.args) == 2, len(permute.args)
-            permute_input, permute_dims = permute.args
+            permute_args = PermuteArgs(*permute.args, **permute.kwargs)  # type: ignore[arg-type]
+            permute_input, permute_dims = permute_args.input, permute_args.dims
             # (1xAxBxC) - `aten.permute` - (1xAxCxB)
             if permute_dims != [0, 1, 3, 2]:
                 continue
@@ -103,7 +108,8 @@ class RemoveRedundantReshapePattern1(PassBase):
                 continue
             if len(reshape2.users) != 1:
                 continue
-            reshape2_input, reshape2_size = reshape2.args
+            reshape2_args = ReshapeArgs(*reshape2.args, **reshape2.kwargs)  # type: ignore[arg-type]
+            reshape2_input, reshape2_size = reshape2_args.input, reshape2_args.size
             # (1xAxCxB) - `aten.reshape - (AxCxB)
             if list(extract_shape(reshape2_input)) != [1] + list(
                 extract_shape(reshape2)
@@ -154,8 +160,8 @@ class RemoveRedundantReshapePattern2(PassBase):
                 continue
             if len(reshape1.users) != 1:
                 continue
-            assert len(reshape1.args) == 2, len(reshape1.args)
-            reshape1_input, reshape1_size = reshape1.args
+            reshape1_args = ReshapeArgs(*reshape1.args, **reshape1.kwargs)  # type: ignore[arg-type]
+            reshape1_input, reshape1_size = reshape1_args.input, reshape1_args.size
             # `(AxBxC) - aten.reshape` - (1xAxBxC)
             if [1] + list(extract_shape(reshape1_input)) != list(
                 extract_shape(reshape1)
@@ -168,8 +174,8 @@ class RemoveRedundantReshapePattern2(PassBase):
                 continue
             if len(permute.users) != 1:
                 continue
-            assert len(permute.args) == 2, len(permute.args)
-            permute_input, permute_dims = permute.args
+            permute_args = PermuteArgs(*permute.args, **permute.kwargs)  # type: ignore[arg-type]
+            permute_input, permute_dims = permute_args.input, permute_args.dims
             # (1xAxBxC) - `aten.permute` - (Bx1xAxC)
             if permute_dims != [2, 0, 1, 3]:
                 continue
@@ -180,7 +186,8 @@ class RemoveRedundantReshapePattern2(PassBase):
                 continue
             if len(reshape2.users) != 1:
                 continue
-            reshape2_input, reshape2_size = reshape2.args
+            reshape2_args = ReshapeArgs(*reshape2.args, **reshape2.kwargs)  # type: ignore[arg-type]
+            reshape2_input, reshape2_size = reshape2_args.input, reshape2_args.size
             # (Bx1xAxC) - `aten.reshape - (Bx(A*C))
             reshape2_input_shape = list(extract_shape(reshape2_input))
             assert len(reshape2_input_shape) == 4
@@ -237,8 +244,8 @@ class RemoveRedundantReshapePattern3(PassBase):
                 continue
             if not reshape_1.target in ops.aten.reshape:
                 continue
-            assert len(reshape_1.args) == 2, len(reshape_1.args)
-            softmax, reshape_1_size = reshape_1.args
+            reshape_1_args = ReshapeArgs(*reshape_1.args, **reshape_1.kwargs)  # type: ignore[arg-type]
+            softmax, reshape_1_size = reshape_1_args.input, reshape_1_args.size
 
             # softmax
             assert isinstance(softmax, torch.fx.Node), type(softmax)
@@ -246,8 +253,12 @@ class RemoveRedundantReshapePattern3(PassBase):
                 continue
             if not softmax.target in ops.aten.softmax:
                 continue
-            assert len(softmax.args) == 3, len(softmax.args)
-            add, softmax_dim, softmax_half_to_float = softmax.args
+            softmax_args = SoftmaxArgs(*softmax.args, **softmax.kwargs)  # type: ignore[arg-type]
+            add, softmax_dim, softmax_half_to_float = (
+                softmax_args.input,
+                softmax_args.dim,
+                softmax_args.half_to_float,
+            )
             assert isinstance(add, torch.fx.Node), type(add)
             assert isinstance(softmax_dim, int), type(softmax_dim)
             assert isinstance(softmax_half_to_float, bool), type(softmax_half_to_float)
@@ -259,8 +270,8 @@ class RemoveRedundantReshapePattern3(PassBase):
             # add
             if not add.target in ops.aten.add:
                 continue
-            assert len(add.args) == 2, len(add.args)
-            reshape_2, reshape_3 = add.args
+            add_args = AddTensorArgs(*add.args, **add.kwargs)  # type: ignore[arg-type]
+            reshape_2, reshape_3 = add_args.input, add_args.other
             assert isinstance(reshape_2, torch.fx.Node), type(reshape_2)
             assert isinstance(reshape_3, torch.fx.Node), type(reshape_3)
 
@@ -269,16 +280,16 @@ class RemoveRedundantReshapePattern3(PassBase):
                 continue
             if not reshape_2.target in ops.aten.reshape:
                 continue
-            assert len(reshape_2.args) == 2, len(reshape_2.args)
-            reshape_2_input, reshape_2_size = reshape_2.args
+            reshape_2_args = ReshapeArgs(*reshape_2.args, **reshape_2.kwargs)  # type: ignore[arg-type]
+            reshape_2_input, reshape_2_size = reshape_2_args.input, reshape_2_args.size
             assert isinstance(reshape_2_input, torch.fx.Node), type(reshape_2_input)
             # reshape_3
             if not reshape_3.op == "call_function":
                 continue
             if not reshape_3.target in ops.aten.reshape:
                 continue
-            assert len(reshape_3.args) == 2, len(reshape_3.args)
-            reshape_3_input, reshape_3_size = reshape_3.args
+            reshape_3_args = ReshapeArgs(*reshape_3.args, **reshape_3.kwargs)  # type: ignore[arg-type]
+            reshape_3_input, reshape_3_size = reshape_3_args.input, reshape_3_args.size
             assert isinstance(reshape_3_input, torch.fx.Node), type(reshape_3_input)
 
             # Check condition
@@ -342,9 +353,8 @@ class RemoveRedundantReshapePattern4(PassBase):
                 continue
             if not reshape1.target in ops.aten.reshape:
                 continue
-            assert len(reshape1.args) == 2, len(reshape1.args)
-
-            reshape1_input, size = list(reshape1.args)
+            reshape1_args = ReshapeArgs(*reshape1.args, **reshape1.kwargs)  # type: ignore[arg-type]
+            reshape1_input, size = reshape1_args.input, reshape1_args.size
             assert isinstance(reshape1_input, torch.fx.Node), type(reshape1_input)
             assert isinstance(size, list), type(size)
             for s in size:
@@ -359,9 +369,9 @@ class RemoveRedundantReshapePattern4(PassBase):
                 continue
             if not reshape2.target in ops.aten.reshape:
                 continue
-            assert len(reshape2.args) == 2, len(reshape2.args)
+            reshape2_args = ReshapeArgs(*reshape2.args, **reshape2.kwargs)  # type: ignore[arg-type]
 
-            reshape2_input, reshape2_size = list(reshape2.args)
+            reshape2_input, reshape2_size = reshape2_args.input, reshape2_args.size
             assert isinstance(reshape2_input, torch.fx.Node), type(reshape2_input)
             assert isinstance(reshape2_size, list), type(reshape2_size)
             for s in reshape2_size:
@@ -411,7 +421,7 @@ class RemoveRedundantReshapePattern5(PassBase):
             if not node.target in ops.aten.reshape:
                 continue
 
-            args = ReshapeArgs(*node.args)
+            args = ReshapeArgs(*node.args, **node.kwargs)  # type: ignore[arg-type]
             output_shape = args.size
             input_shape = list(extract_shape(args.input))
 

tico/serialize/circle_graph.py

@@ -91,6 +91,10 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
         self.tensors: List[circle.Tensor.TensorT] = []
         self.operators: List[circle.Operator.OperatorT] = []
         self.name_to_tid: Dict[str, int] = {}
+        # Mapping from Circle tensor names to their originating FX nodes.
+        # Used to trace back tensor definitions to their source and finalize
+        # human-readable tensor names after serialization.
+        self.name_to_node: Dict[str, torch.fx.Node] = {}
         self.counter: defaultdict = defaultdict(int)
 
     # Generate a unique name with prefix.
@@ -111,6 +115,7 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
 
     def _add_tensor(self, tensor: circle.Tensor.TensorT) -> None:
         self.tensors.append(tensor)
+        assert tensor.name not in self.name_to_tid
         self.name_to_tid[tensor.name] = len(self.tensors) - 1
 
     def add_operator(self, op: circle.Operator.OperatorT) -> None:
@@ -138,10 +143,12 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
         return name in self.name_to_tid
 
     def add_tensor_from_node(
-        self, node: torch.fx.node.Node, data: Optional[np.ndarray] = None
+        self, node: torch.fx.Node, data: Optional[np.ndarray] = None
     ) -> None:
         tensor = circle.Tensor.TensorT()
         tensor.name = self._gen_unique_name_with_prefix(node.name)
+        assert tensor.name not in self.name_to_node
+        self.name_to_node[tensor.name] = node
         assert node.meta.get("val") is not None
         tensor.type = extract_circle_dtype(node)
         tensor.shape = list(extract_shape(node))
@@ -165,10 +172,15 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
         tensor.buffer = bid
         self._add_tensor(tensor)
 
-    def add_const_tensor(self, data: ConstData) -> circle.Tensor.TensorT:
+    def add_const_tensor(
+        self, data: ConstData, source_node: Optional[torch.fx.Node] = None
+    ) -> circle.Tensor.TensorT:
         assert is_const(data)
         tensor = circle.Tensor.TensorT()
         tensor.name = self._gen_unique_name_with_prefix("const_tensor")
+        assert tensor.name not in self.name_to_node
+        if source_node is not None:
+            self.name_to_node[tensor.name] = source_node
         assert not self.has_tensor(tensor.name)
         torch_t = torch.as_tensor(data=data)
         torch_t_shape = list(torch_t.size())
@@ -189,10 +201,45 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
         shape: List[int],
         dtype: int,
         qparam: Optional[QuantParam] = None,
+        source_node: Optional[torch.fx.Node] = None,
     ) -> circle.Tensor.TensorT:
+        """
+        Create a new tensor and register it into the Circle subgraph from scratch.
+
+        This function is used to allocate tensors that are not directly derived from
+        values in the FX graph, such as those created by padding or shape-generating
+         operators.
+
+        If a `source_node` is provided, it is used to enrich the tensor's metadata
+        (e.g., by associating the tensor with the module hierarchy path stored in
+        the node's `nn_module_stack`). This enables better traceability and more
+        informative tensor names in the final Circle model.
+
+        Parameters
+        ----------
+        prefix : str
+            A name prefix used to generate a unique tensor name.
+        shape : List[int]
+            The shape of the tensor.
+        dtype : int
+            The Circle-compatible dtype of the tensor. Use `to_circle_dtype()` to convert.
+        qparam : Optional[QuantParam]
+            Optional quantization parameters to apply to the tensor.
+        source_node : Optional[torch.fx.Node]
+            If provided, the FX node from which this tensor originates. Used to generate
+            a richer name and track module origin.
+
+        Returns
+        -------
+        circle.Tensor.TensorT
+            The newly created and registered tensor.
+        """
         assert isinstance(dtype, int), f"{dtype} must be integer. Use to_circle_dtype."
         tensor = circle.Tensor.TensorT()
         tensor.name = self._gen_unique_name_with_prefix(prefix)
+        assert tensor.name not in self.name_to_node
+        if source_node is not None:
+            self.name_to_node[tensor.name] = source_node
         tensor.shape = shape
         if qparam is not None:
             tensor.quantization = to_circle_qparam(qparam)
@@ -255,7 +302,7 @@ class CircleSubgraph(circle.SubGraph.SubGraphT):
 
     # TODO Rename, it doesn't only get_tid but also possibly add a new const tensor
     def get_tid(
-        self, node: Union[torch.fx.node.Node, circle.Tensor.TensorT, ConstData]
+        self, node: Union[torch.fx.Node, circle.Tensor.TensorT, ConstData]
     ) -> int:
         # return -1 if node is None. This is for generating CircleOutputExclude
         if node == None:

tico/serialize/circle_serializer.py

@@ -31,6 +31,7 @@ from tico.serialize.circle_graph import CircleModel, CircleSubgraph
 from tico.serialize.operators.hashable_opcode import OpCode
 from tico.serialize.operators.node_visitor import get_node_visitors
 from tico.utils import logging
+from tico.utils.serialize import finalise_tensor_names
 
 
 multiple_output_ops = [
@@ -150,6 +151,7 @@ def build_circle(edge_program: ExportedProgram) -> bytes:
                 prefix=node.name,
                 shape=list(attr_tensor.shape),
                 dtype=to_circle_dtype(attr_tensor.dtype),
+                source_node=node,
             )
 
             logger.debug(f"get_attr: {node.name} tensor exported.")
@@ -217,6 +219,7 @@ def build_circle(edge_program: ExportedProgram) -> bytes:
             logger.debug(f"call_function: {node.name} ({opcode}) Op exported.")
 
     # Register subgraph
+    finalise_tensor_names(graph)
     model.subgraphs.append(graph)
 
     # Encode operator codes

tico/serialize/operators/op_any.py

@@ -122,7 +122,10 @@ class AnyVisitor(NodeVisitor):
             dst_dtype_circle = circle.TensorType.TensorType.BOOL
             dst_dtype_torch = torch.bool
             ne_tensor: circle.Tensor.TensorT = self.graph.add_tensor_from_scratch(
-                prefix=f"{input.name}_ne", shape=input_shape, dtype=dst_dtype_circle
+                prefix=f"{input.name}_ne",
+                shape=input_shape,
+                dtype=dst_dtype_circle,
+                source_node=input,
             )
             ne_node = self.define_ne_node(
                 [input_tensor, torch.Tensor([0]).to(dtype_torch)], [ne_tensor]

tico/serialize/operators/op_avg_pool2d.py

@@ -151,6 +151,7 @@ class AvgPool2DVisitor(NodeVisitor):
                 prefix=f"{input.name}_pad_output",
                 shape=padded_input_shape,
                 dtype=input_dtype,
+                source_node=node,
             )
             pad_operator = define_pad_node(
                 self.graph, self._op_codes, [input, padding_vec], [padded_input_tensor]

tico/serialize/operators/op_clamp.py

@@ -107,7 +107,10 @@ class ClampVisitor(NodeVisitor):
             input_shape = extract_shape(input)
             input_dtype = extract_circle_dtype(input)
             minimum_tensor = self.graph.add_tensor_from_scratch(
-                prefix=f"{input.name}_min", dtype=input_dtype, shape=list(input_shape)
+                prefix=f"{input.name}_min",
+                dtype=input_dtype,
+                shape=list(input_shape),
+                source_node=node,
             )
             minimum_opertor = self.define_minimum_node(
                 [input, max_val], [minimum_tensor]

tico/serialize/operators/op_conv2d.py

@@ -159,6 +159,7 @@ class Conv2dVisitor(NodeVisitor):
                 shape=pad_output_shape,
                 dtype=input_dtype,
                 qparam=input_qparam,
+                source_node=node,
             )
             # CirclePad
             pad_operator = define_pad_node(

tico/serialize/operators/op_copy.py

@@ -111,6 +111,7 @@ class CopyVisitor(NodeVisitor):
             prefix=f"{dst_name}_shape_output",
             shape=dst_shape_shape,
             dtype=circle.TensorType.TensorType.INT32,
+            source_node=node,
         )
 
         shape_operator = self.define_shape_node([dst], [shape_output])
@@ -131,6 +132,7 @@ class CopyVisitor(NodeVisitor):
                     prefix=f"{src_name}_broadcast_to_output",
                     shape=dst_shape,
                     dtype=src_type,
+                    source_node=node,
                 )
             )
 

tico/serialize/operators/op_cumsum.py

@@ -63,7 +63,10 @@ class CumsumVisitor(NodeVisitor):
             cast_name = f"{input.name}_cast"
             cast_dtype = circle.TensorType.TensorType.INT64
             cast_tensor = self.graph.add_tensor_from_scratch(
-                prefix=cast_name, dtype=cast_dtype, shape=input_shape
+                prefix=cast_name,
+                dtype=cast_dtype,
+                shape=input_shape,
+                source_node=node,
             )
             cast_operator = create_builtin_operator(
                 self.graph, cast_op_index, [input], [cast_tensor]

tico/serialize/operators/op_depthwise_conv2d.py

@@ -171,6 +171,7 @@ class DepthwiseConv2dVisitor(NodeVisitor):
                 prefix=f"{node.name}_input_pad_output",
                 shape=pad_output_shape,
                 dtype=input_dtype,
+                source_node=node,
             )
             # CirclePad
             pad_operator = define_pad_node(

tico/serialize/operators/op_log1p.py

@@ -65,7 +65,10 @@ class Log1pVisitor(NodeVisitor):
         input_shape = list(extract_shape(input))
         dst_dtype_circle = extract_circle_dtype(input)
         add_tensor: circle.Tensor.TensorT = self.graph.add_tensor_from_scratch(
-            prefix=f"{input.name}_add", shape=input_shape, dtype=dst_dtype_circle
+            prefix=f"{input.name}_add",
+            shape=input_shape,
+            dtype=dst_dtype_circle,
+            source_node=node,
         )
         const_one = torch.tensor([1]).to(extract_torch_dtype(input))
 

tico/serialize/operators/op_max_pool2d_with_indices.py

@@ -106,6 +106,7 @@ class MaxPool2DWithIndicesVisitor(NodeVisitor):
                 shape=padded_input_shape,
                 dtype=input_dtype,
                 qparam=input_qparam,
+                source_node=node,
             )
             if input_qparam is not None:
                 padding_value = get_integer_dtype_min(input_qparam.dtype)

tico/serialize/operators/op_mm.py

@@ -111,7 +111,9 @@ class MatmulDefaultVisitor(NodeVisitor):
     fullyconnected( lhs[H, K], trs_output[W', K] ) -> output(H, W')
     """
 
-    def define_fc_with_transpose(self, inputs, outputs) -> circle.Operator.OperatorT:
+    def define_fc_with_transpose(
+        self, node, inputs, outputs
+    ) -> circle.Operator.OperatorT:
         lhs, rhs = inputs
 
         # get transpose shape
@@ -128,8 +130,9 @@ class MatmulDefaultVisitor(NodeVisitor):
             prefix=f"{rhs_name}_transposed_output",
             shape=rhs_shape_transpose,
             dtype=rhs_type,
+            source_node=node,
         )
-        trs_perm = self.graph.add_const_tensor(data=[1, 0])
+        trs_perm = self.graph.add_const_tensor(data=[1, 0], source_node=node)
         trs_operator = self.define_transpose_node([rhs, trs_perm], [trs_output])
         self.graph.add_operator(trs_operator)
 
@@ -138,7 +141,7 @@ class MatmulDefaultVisitor(NodeVisitor):
         fc_weight = trs_output
         fc_shape = [fc_weight.shape[0]]
         fc_bias = self.graph.add_const_tensor(
-            data=[0.0] * fc_shape[0],
+            data=[0.0] * fc_shape[0], source_node=node
         )
 
         operator = self.define_fc_node([fc_input, fc_weight, fc_bias], outputs)
@@ -169,6 +172,6 @@ class MatmulDefaultVisitor(NodeVisitor):
         if not is_const(other) and prior_latency:
             operator = self.define_bmm_node(inputs, outputs)
         else:
-            operator = self.define_fc_with_transpose(inputs, outputs)
+            operator = self.define_fc_with_transpose(node, inputs, outputs)
 
         return operator

tico/serialize/operators/op_pow.py

@@ -42,7 +42,10 @@ class BasePowVisitor(NodeVisitor):
         cast_name = f"{node.name}_cast"
         cast_dtype = circle.TensorType.TensorType.FLOAT32
         cast_tensor = self.graph.add_tensor_from_scratch(
-            prefix=cast_name, dtype=cast_dtype, shape=node_shape
+            prefix=cast_name,
+            dtype=cast_dtype,
+            shape=node_shape,
+            source_node=node,
         )
         cast_operator = create_builtin_operator(
             self.graph, op_index, [node], [cast_tensor]

tico/serialize/operators/op_reciprocal.py

@@ -47,7 +47,7 @@ class ReciprocalVisitor(NodeVisitor):
         input = args.input
 
         input_tensor = torch.tensor(1, dtype=extract_torch_dtype(input))
-        x = self.graph.add_const_tensor(input_tensor)
+        x = self.graph.add_const_tensor(input_tensor, source_node=node)
         inputs = [x, input]
         outputs = [node]
 

tico/serialize/operators/op_repeat.py

@@ -76,6 +76,7 @@ class RepeatVisitor(NodeVisitor):
                         prefix=f"{node.name}_concat_{idx}",
                         shape=repeated_shape,
                         dtype=tensor_dtype,
+                        source_node=node,
                     )
                 inputs = [concat_input] * r
                 if repeat_dim_cnt == 1:

tico/serialize/operators/op_split_with_sizes.py

@@ -82,7 +82,10 @@ class SplitWithSizesVisitor(NodeVisitor):
                 shape = list(fake_tensor.size())
                 dtype = to_circle_dtype(fake_tensor.dtype)
                 tensor = self.graph.add_tensor_from_scratch(
-                    f"{node.name}_unused_{idx}", shape, dtype
+                    f"{node.name}_unused_{idx}",
+                    shape,
+                    dtype,
+                    source_node=node,
                 )
                 outputs.append(tensor)
 

tico/utils/graph.py

@@ -198,3 +198,39 @@ def is_single_value_tensor(t: torch.Tensor):
         return True
 
     return False
+
+
+def get_module_name_chain(node: Optional[torch.fx.Node]) -> str:
+    """
+    Returns a slash-separated string of module names representing the
+    hierarchical path of the FX node within the original model.
+
+    If the node has no `nn_module_stack` metadata, "unknown" is returned.
+
+    Example:
+        "encoder/layer1/linear"
+
+    Parameters
+    ----------
+    node: torch.fx.Node
+        A node from an ExportedProgram graph.
+
+    Returns
+    -------
+    str
+        A human-readable string that describes the full module path.
+    """
+    if node is None:
+        return "unknown"
+    # Let's prefix "tico" for graph inputs
+    if node.op == "placeholder" and "nn_module_stack" not in node.meta:
+        return "tico"
+
+    assert isinstance(node, torch.fx.Node)
+    stack = node.meta.get("nn_module_stack")
+    if stack:
+        assert isinstance(stack, dict)
+        # Retrieving the last element is enough.
+        return next(reversed(stack.values()))[1]
+    else:
+        return "unknown"

tico/utils/serialize.py

@@ -0,0 +1,42 @@
+# Copyright (c) 2025 Samsung Electronics Co., Ltd. All Rights Reserved
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import Optional
+
+import torch
+
+from tico.serialize.circle_graph import CircleSubgraph
+from tico.utils.graph import get_module_name_chain
+
+
+def finalise_tensor_names(
+    graph: CircleSubgraph,
+) -> None:
+    """
+    Replace every `tensor.name` with the *readable* version
+     **after** the graph is fully built.
+
+    Why late?
+    ---------
+    - All intermediate steps (add_input, add_output, get_tid…) rely on the
+       original technical names in ExportedProgram.
+
+    The rewrite is *in-place* and touches **only** the `name` field of
+    each tensor.
+    """
+    assert hasattr(graph, "name_to_node")
+
+    for tensor in graph.tensors:
+        if tensor.name in graph.name_to_node:
+            tensor.name = f"{get_module_name_chain(graph.name_to_node[tensor.name])}::{tensor.name}"

{tico-0.1.0.dev250604.dist-info → tico-0.1.0.dev250608.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: tico
-Version: 0.1.0.dev250604
+Version: 0.1.0.dev250608
 Summary: Convert exported Torch module to circle
 Home-page: UNKNOWN
 License: UNKNOWN

{tico-0.1.0.dev250604.dist-info → tico-0.1.0.dev250608.dist-info}/RECORD

@@ -1,4 +1,4 @@
-tico/__init__.py,sha256=X2eazANf-Y9gV1fcug-QytZyLbNFnJqqW658DamD_wI,1743
+tico/__init__.py,sha256=FjyTxaQjfrgfUCPfMzx5WnfpSY0Vbksu3gx8m6Tc5xY,1743
 tico/pt2_to_circle.py,sha256=PPmFNw20jw2Z2VyM3ln9pX__jTzBOAZiv0gT5a-p-Y8,2666
 tico/config/__init__.py,sha256=xZzCXjZ84qE-CsBi-dfaL05bqpQ3stKKfTXhnrJRyVs,142
 tico/config/base.py,sha256=anwOiJFkUxUi7Cef573JgQcjk6S-FSi6O_TLjYASW-g,1244
@@ -87,15 +87,15 @@ tico/passes/remove_nop.py,sha256=5QE3inFsXgzyPT_t7pKeXNqD1LRf6ed_Mp7YMadA6AI,270
 tico/passes/remove_redundant_assert_nodes.py,sha256=3a2xEQ2iPY7Gqg8jZi8G5bfDDrK2kOO1OHCMv_gJGz0,1592
 tico/passes/remove_redundant_expand.py,sha256=7st92AbWOl7yzM0Y5seaZJQKMFHqkYpH3qYMOlAU5lk,2234
 tico/passes/remove_redundant_permute.py,sha256=sS53eTY4sSnpZWDaaHN8czUmzNwmqh1lF90nYamXzac,3566
-tico/passes/remove_redundant_reshape.py,sha256=eJ2GpYTx7qrXROjmZn4m9xBFIfjSWjlGjYqQrEnU5Qw,15677
+tico/passes/remove_redundant_reshape.py,sha256=aPZcDR0kBExEsWCYfBbLulm_wcjJNnGjn4mgrUIPdpU,16810
 tico/passes/remove_redundant_slice.py,sha256=BAfSkA5jDIEhYx4nMnu6cJadQle3YTw5y39ZLiYfJJ8,2109
 tico/passes/remove_redundant_to_copy.py,sha256=uTIjAn3Eli_RvXC-QOqxBAkV_whDBkkNhu-mvNKAEhs,3136
 tico/passes/restore_linear.py,sha256=UMMHdLmRGq9bfJx_0L9lL2UQBd51PGNP0WywO8KdrDM,4066
 tico/passes/segment_index_select.py,sha256=ifXOIFC12lNwsB-s3k1cJcMHP3UEijPpkMAbwI7lZbQ,5097
 tico/serialize/__init__.py,sha256=IO6FP_xYbGy0dW0HL26GXD3ouxARaxCK7bz9dn4blPQ,26
-tico/serialize/circle_graph.py,sha256=l9fcvV4x3L5Bh2WmEseo--0KFjSVTrlSBaBrqbrmXgg,9498
+tico/serialize/circle_graph.py,sha256=_u0vFDhPdOhEkucmaEhqILo13NKbjyVemPYFfC5YCZg,11619
 tico/serialize/circle_mapping.py,sha256=C9C3ORACQOdvBdnt5KRzlT8zao_TvzQklIxH794OhP0,5719
-tico/serialize/circle_serializer.py,sha256=gJP2QEY7hFrlP_I4JRnnMehThr9ay9SUyoOQ2ppJBl0,8859
+tico/serialize/circle_serializer.py,sha256=KRx_Azx2Je9XNYe-pZuuiSMvbXEddd8M8qDATIt7XXk,8981
 tico/serialize/pack.py,sha256=5HZ9kX3x6C6CyT_FWS6FRmvx_P7Dx21orjUNQxJ2xlo,1297
 tico/serialize/quant_param.py,sha256=s97GJyDOZULnqFUWPakHais31G_qqPuO0awPHCkZDvI,1342
 tico/serialize/operators/__init__.py,sha256=LIvXsNnN4yUCS2CGNQ5XW8p8oXDTV_WHWuOEAw1t6WY,990
@@ -104,20 +104,20 @@ tico/serialize/operators/node_visitor.py,sha256=UYyCwXqSCeRyimThMShstHnt7vKM9tsu
 tico/serialize/operators/op_abs.py,sha256=Y-vy7rcqPT-qD3QS5R8zbApWWTPpjY6xuMMVnbIhYmQ,1827
 tico/serialize/operators/op_add.py,sha256=otm062DMHVAThWmOtSTZdPyP3P5-2cv5VL_UWBJeLms,2346
 tico/serialize/operators/op_alias_copy.py,sha256=Xu9OiILbGf8oddh8yTqovvLfgVs8XYV7Cg4n6CesWcg,2175
-tico/serialize/operators/op_any.py,sha256=WMsHLq7WIcl6rD2G3QqpWRSCR-a6UYX6y5AjB6BDS3U,5049
+tico/serialize/operators/op_any.py,sha256=QqAYby1WPd4l97vX-tBjAsOb_pVb1wPjj7SbWQNToCQ,5117
 tico/serialize/operators/op_arange_start_step.py,sha256=0T5lWwh3TfsFStmVv0v5qG03KENRDBmMix08RXQ4D-U,2132
 tico/serialize/operators/op_argmax.py,sha256=ARyGHlmWVmzwCct93V5x1-VyKqhxMOvV8GuM8yQWXdo,2290
-tico/serialize/operators/op_avg_pool2d.py,sha256=ABxhfowDz7SXlWnW2iQuSA5X52xm0PGLs-N1l9vGXbo,7488
+tico/serialize/operators/op_avg_pool2d.py,sha256=eVd5ngFSImAklvJFfBAaEUvAc-I60e43xCdLjd5zwYA,7522
 tico/serialize/operators/op_bmm.py,sha256=AELjHC9ISFPIzEEl5Kr1s4GSNLZElwZmVZJWkEyCEoA,2189
 tico/serialize/operators/op_cat.py,sha256=XDYOh0XAyrM0TlxVm6Sa0OFFGrKk7aSDcGXC-hYX4gs,2204
-tico/serialize/operators/op_clamp.py,sha256=V3rncHvUAuJ2nXOyywTnOGCvNBeCQGqQIW1_zxKlSsA,4231
+tico/serialize/operators/op_clamp.py,sha256=ZRAsXLGsZqJEh4wXxESEpRJkRtUuJWTDgAem6lr9_5I,4298
 tico/serialize/operators/op_clone.py,sha256=vzDYJ8TS3tc2BAyd_z8nt5VqT1inpymSseMEhd9dva0,2394
 tico/serialize/operators/op_constant_pad_nd.py,sha256=OpP4AP-d1IFcWZolNa-o9ZxzXJQkMdG9WQ66soX3s-E,2675
-tico/serialize/operators/op_conv2d.py,sha256=a8aU0CwPgCYM37rDWmrCpxMoQZQdxzXaPtJAVB7dK0M,7283
-tico/serialize/operators/op_copy.py,sha256=W_ih1yqqMwSCO_l9l_LUn_G_IoTgDJcNUnAbX1ZITZI,6054
+tico/serialize/operators/op_conv2d.py,sha256=nC_jqzjlrUJ0L_lux_wXBqxDfq67jyroXSgrl5WoNfk,7317
+tico/serialize/operators/op_copy.py,sha256=vaianLQ19-2ZQZ-MdQ07YuOPeFeo_HAx2a0Qfn7I5Kk,6122
 tico/serialize/operators/op_cos.py,sha256=N12bNyuTQIxRnD0eHRPdFVzRQPMy1NFM4iM8oQ4lYzw,2034
-tico/serialize/operators/op_cumsum.py,sha256=HhPF2uKMamk8KIiV0j6P-grU8oZ9AhlcRsRVC28oYVw,3731
-tico/serialize/operators/op_depthwise_conv2d.py,sha256=Ed0aVW0UaRoTDPt6kcF4iDzpx_CqZ0oDjdiMSexVk78,7498
+tico/serialize/operators/op_cumsum.py,sha256=3fmOf1mIeCX1uhTBcSJmRGXejzLtO8UwaI1eEQDC6nA,3798
+tico/serialize/operators/op_depthwise_conv2d.py,sha256=PTos0tQoM8EZoB88s4Tjb7n6pJja5nbNQRDsucVzRwc,7532
 tico/serialize/operators/op_dequantize_per_channel.py,sha256=aPcVxjdgvfSFoLnv9NL-RxO5vZYj8ulqriMP5LHIWs0,3133
 tico/serialize/operators/op_dequantize_per_tensor.py,sha256=u9aK_Xle9rDN0EHLE0YrCTlXY4Q53Ch9Di4qmx7ynps,2304
 tico/serialize/operators/op_div.py,sha256=WjeM2Ux7TyGlSNx2aVC783JvcL0xnY6FBYo1Q_kdb5Q,2201
@@ -135,27 +135,27 @@ tico/serialize/operators/op_index_select.py,sha256=cw7IbvixooikGxzbpUmI9tHS4kjl4
 tico/serialize/operators/op_instance_norm.py,sha256=AhcVm71ChB16BlPNwqBh5tMHCqMShOXHPkE8Ag9jBfQ,3144
 tico/serialize/operators/op_linear.py,sha256=bw_mn2CiJy8CbpPevOV0PMPh0ZMWKAybLZ9cnIKJSsk,2527
 tico/serialize/operators/op_log.py,sha256=1TKvH2lttdAHE0P84vcxmOvGBBRUs6D71Jrei7SdZHE,1827
-tico/serialize/operators/op_log1p.py,sha256=lH0rLxpag7kGeM5UiE5b1Q4JluOE-yiQSsEcQBYv6ts,3066
+tico/serialize/operators/op_log1p.py,sha256=gG7Fs4UDj_Nnp7U60UtPyz0fLv1lBpJVOGGCMm-42pY,3121
 tico/serialize/operators/op_logical_and.py,sha256=WhQ8knuq32BO-WhAqkOgpcUStPkjoPmRWuYNsKveF3w,2163
 tico/serialize/operators/op_logical_not.py,sha256=ugrVcRqR3IvUUaiRVW5cArCYJbzmkcXp88QM846jCww,2129
 tico/serialize/operators/op_lt.py,sha256=_vA7dWpV9wVBxB7JL9pLQT9BsV91NGQBq_0auAtHK5Y,2080
 tico/serialize/operators/op_max_dim.py,sha256=nS_TZl5uq4uv1LwgBD9Wddyac4atKqBiIWKIyeXse2s,2519
-tico/serialize/operators/op_max_pool2d_with_indices.py,sha256=GBeBNEyohVSGMIiY0Z1UUX-9N8OdVGitbGNAq_TJZd4,5631
+tico/serialize/operators/op_max_pool2d_with_indices.py,sha256=SPXIdlrYgJfR982il_ajokNehFYJWOmccUoksS_0qos,5665
 tico/serialize/operators/op_maximum.py,sha256=JjBr6gWEnuakLuk1_feotTHfIIm3s5YqWmqhUMpSPI0,1873
 tico/serialize/operators/op_mean.py,sha256=rVQZOxCJkHFY4kQBAS1HVK0HkcqxgkSy6zvEDLX_WYQ,2267
 tico/serialize/operators/op_minimum.py,sha256=fASjQVcTPCin02umQwFPdq2ss-Ve7S5A33J3QmmQ_wQ,1873
-tico/serialize/operators/op_mm.py,sha256=fHggR9dmlwXw0DAyn__2JbG7e0q1Jhfmi5-2jDlpRDk,6730
+tico/serialize/operators/op_mm.py,sha256=Fgq_HUUKuXOQY_t8lah3SOUqTsGet-KbVttCK4-fjAk,6821
 tico/serialize/operators/op_mul.py,sha256=42Guc0MWBGBCZoj9-4LcLtTMtUPwsmDSVmvkR8tqLhM,3165
 tico/serialize/operators/op_ne.py,sha256=xa2WJL2tYksxw7fIJic_D9ltLEseyCII8HpR32Oq8Do,1900
 tico/serialize/operators/op_neg.py,sha256=fkI3ExyD3QF-qtxBcXqQutPNDbNL8g7lZYE7CyD2wLk,2046
 tico/serialize/operators/op_permute.py,sha256=5DfX3pfZ5FDNmrSqx3-hRwPA7vm36z7BfG-nuyyBTsM,2282
-tico/serialize/operators/op_pow.py,sha256=cvW81yaW2hxsMPwRF_vJu4VXMy2AvdATqJ26eNRrGZk,5379
+tico/serialize/operators/op_pow.py,sha256=z_4G_J1k_keeVE6ZYKSy-kqkdJ_i4p4kHkO0dJZnz-Y,5434
 tico/serialize/operators/op_prelu.py,sha256=0ZybL5pNvBrRvQGy4M6gELrjiEXEsb2wBDdU8x4D75I,1874
 tico/serialize/operators/op_quantize_per_tensor.py,sha256=w-vYxSPnN2gtx-pEkkcMGU0ZjiwaS4y1sxy56pKEq3E,3004
-tico/serialize/operators/op_reciprocal.py,sha256=cK2uddDbV32wPAhD8msUW7AXnLIGIs7v6SxTq9i5Bxc,2376
+tico/serialize/operators/op_reciprocal.py,sha256=6b9_bxjg_0EvgAitSv1MgBi4PJSEgm-21s5qtWI1UR4,2394
 tico/serialize/operators/op_relu.py,sha256=WXCR_chwEUBqjFIQ_4E2avwk-Acy76pmX20rJQCBTQo,1832
 tico/serialize/operators/op_relu6.py,sha256=ZWqEolfAKjOdUC1ZCg0iuu4dBhkJRxVYR2tUzpbvKQM,1829
-tico/serialize/operators/op_repeat.py,sha256=nR-xKZYl4ZzEYj0Tw8N-oKh5wLyv3PxI80vmiNcAm_0,4193
+tico/serialize/operators/op_repeat.py,sha256=0wTv1Mg7kg0eHz0CT6atyVAli4T4h5rYXq5opY6op20,4235
 tico/serialize/operators/op_reshape.py,sha256=PdYenXvfQxzYST3yNH6MTxUQ25TulNwiip6N2q76zfQ,2549
 tico/serialize/operators/op_resize_nearest_neighbor.py,sha256=dXaAnZ5M_ko_tH-HolxNpHFXkDUQ8x45myskojP5XZE,2771
 tico/serialize/operators/op_rsqrt.py,sha256=yl2vd8InjhLPbE0vHIrEera6DVXlY9dLgO7yZZCH3RI,1837
@@ -165,7 +165,7 @@ tico/serialize/operators/op_sigmoid.py,sha256=ZubbGG1yU5uvNkEmOmbjj3eq7d9mwEaJdC
 tico/serialize/operators/op_sin.py,sha256=MbttmHTVKhwKK6gH9Vbcbn5aAaxnQ71NdpmQAlTcojU,1827
 tico/serialize/operators/op_slice.py,sha256=g0r8lj5CIxpT6ixOKqUzwKiNhoiuIFwWjbpaiCoOg6w,5259
 tico/serialize/operators/op_softmax.py,sha256=8AwmsAVdSoIMKdfejrw9cy44TbOvvXsA0w3WQDVpI3A,3855
-tico/serialize/operators/op_split_with_sizes.py,sha256=rZ6WR-u_S-jOa6fgN5l4xxVq7ZLLuJLhTbaabOvY6Bs,4195
+tico/serialize/operators/op_split_with_sizes.py,sha256=TgYg1cu-3BSz9SsXfAhoJbo4q5ZzFaoFArkH_obsYlU,4274
 tico/serialize/operators/op_sqrt.py,sha256=9Q5jkuEPrim11XfSQHGDGVTMYk1TQhOfVqMVYD_eIrI,1871
 tico/serialize/operators/op_squeeze.py,sha256=QnNwfAdTC1xBm04C9DkVs8VB5YRN-4fCsIWn189QaPg,2416
 tico/serialize/operators/op_sub.py,sha256=yZskQJF0ylXVk02Uid8djPNIWDJ-0uHJar4UYhlJVkk,2479
@@ -181,12 +181,13 @@ tico/utils/convert.py,sha256=KCllPnvQ8bjEYR1yI72s9aNBp7Py1CzIEEpYSYZcu60,11684
 tico/utils/define.py,sha256=Ypgp7YffM4pgPl4Zh6TmogSn1OxGBMRw_e09qYGflZk,1467
 tico/utils/diff_graph.py,sha256=_eDGGPDPYQD4b--MXX0DLoVgSt_wLfNPt47UlolLLR4,5272
 tico/utils/errors.py,sha256=f3csJjgbXG9W1aHhqEcou008Aor19W57X8oT5Hx8w1M,954
-tico/utils/graph.py,sha256=YTAMbnjxqtNWy8ckOi68WYLXUIIvGwsux4PHmIamVok,6654
+tico/utils/graph.py,sha256=qQIz78XfJK1rrGyq00f-ArqxACItiS-oHbN2nE4xQrY,7653
 tico/utils/logging.py,sha256=IlbBWscsaHidI0dNqro1HEXAbIcbkR3BD5ukLy2m95k,1286
 tico/utils/model.py,sha256=Uqc92AnJXQ2pbvctS2z2F3Ku3yNrwXZ9O33hZVis7is,1250
 tico/utils/padding.py,sha256=GGO27VbaOvtaMYLDrSaKv7uxjeet566aMJD0PyYeMvQ,1484
 tico/utils/passes.py,sha256=kGmDe__5cPaO6i5EDAoXSVe6yXEoX9hAny4ROb3ZEmQ,2409
 tico/utils/register_custom_op.py,sha256=iRQvdqlBqrJxq_pNkvJyDIJD_SYtCUl88wwbbuvSwlk,22952
+tico/utils/serialize.py,sha256=AQXMBOLu-Kg2Rn-qbqsAtHndjZAZIavlKA0QFgJREHM,1420
 tico/utils/trace_decorators.py,sha256=ddLIiKQfSaQrxgF1kNpwjFTQnXENzeSfcr1kuAW4jGI,3221
 tico/utils/utils.py,sha256=NAa3ZX5G-UCQwmz5WnFl0iCEra24PMY5wC0MyX7smUg,11156
 tico/utils/validate_args_kwargs.py,sha256=P4aMnr9EhNCtc_AgJPpuezfQbqFfDn0lhJSWqmumLZ8,25054
@@ -194,9 +195,9 @@ tico/utils/mx/__init__.py,sha256=IO6FP_xYbGy0dW0HL26GXD3ouxARaxCK7bz9dn4blPQ,26
 tico/utils/mx/elemwise_ops.py,sha256=V6glyAHsVR1joqpsgnNytatCD_ew92xNWZ19UFDoMTA,10281
 tico/utils/mx/formats.py,sha256=uzNWyu-1onUlwQfX5cZ6fZSUfHMRqorper7_T1k3jfk,3404
 tico/utils/mx/mx_ops.py,sha256=RcfUTYVi-wilGB2sC35OeARdwDqnixv7dG5iyZ-fQT8,8555
-tico-0.1.0.dev250604.dist-info/LICENSE,sha256=kp4JLII7bzRhPb0CPD5XTDZMh22BQ7h3k3B7t8TiSbw,12644
-tico-0.1.0.dev250604.dist-info/METADATA,sha256=gN7iYzRhOodrWqExhbNBTMQ5S_6iOpb-vte13nECcak,8633
-tico-0.1.0.dev250604.dist-info/WHEEL,sha256=G16H4A3IeoQmnOrYV4ueZGKSjhipXx8zc8nu9FGlvMA,92
-tico-0.1.0.dev250604.dist-info/entry_points.txt,sha256=kBKYSS_IYrSXmUYevmmepqIVPScq5vF8ulQRu3I_Zf0,59
-tico-0.1.0.dev250604.dist-info/top_level.txt,sha256=oqs7UPoNSKZEwqsX8B-KAWdQwfAa7i60pbxW_Jk7P3w,5
-tico-0.1.0.dev250604.dist-info/RECORD,,
+tico-0.1.0.dev250608.dist-info/LICENSE,sha256=kp4JLII7bzRhPb0CPD5XTDZMh22BQ7h3k3B7t8TiSbw,12644
+tico-0.1.0.dev250608.dist-info/METADATA,sha256=LJYAjDzOfuHcUGzJvc-Cgc_QKCnVUHFY2yr9IY8EkgU,8633
+tico-0.1.0.dev250608.dist-info/WHEEL,sha256=G16H4A3IeoQmnOrYV4ueZGKSjhipXx8zc8nu9FGlvMA,92
+tico-0.1.0.dev250608.dist-info/entry_points.txt,sha256=kBKYSS_IYrSXmUYevmmepqIVPScq5vF8ulQRu3I_Zf0,59
+tico-0.1.0.dev250608.dist-info/top_level.txt,sha256=oqs7UPoNSKZEwqsX8B-KAWdQwfAa7i60pbxW_Jk7P3w,5
+tico-0.1.0.dev250608.dist-info/RECORD,,