Skip to content

[Relax][PyTorch] Support advanced range constraints (addition) #18452

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

Merged
mshr-h merged 1 commit into from
Nov 17, 2025
Merged

[Relax][PyTorch] Support advanced range constraints (addition) #18452

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
46 changes: 41 additions & 5 deletions python/tvm/relax/frontend/torch/exported_program_translator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@
"""PyTorch ExportedProgram of Relax."""
from collections import ChainMap, OrderedDict
from functools import partial
from typing import Callable, Dict, List, Tuple
from typing import Callable, Dict, List, Optional, Tuple

import torch
import tvm
Expand Down Expand Up @@ -1161,6 +1161,40 @@ def create_convert_map(
"_local_scalar_dense.default": self._item,
}

def _process_derived_symbol(
self, symbol, torch_symbol_to_relax_var: Dict[str, tvm.tir.Var]
) -> Tuple[str, Optional[tvm.tir.PrimExpr]]:
"""Process a sympy symbol to generate a descriptive name and TIR expression."""
import sympy

if isinstance(symbol, sympy.Symbol):
return str(symbol), None

if not isinstance(symbol, sympy.Add):
return str(symbol), None

tir_expr = None
for arg in symbol.args:
if isinstance(arg, sympy.Integer):
term = tvm.tir.IntImm("int64", int(arg))
elif isinstance(arg, sympy.Symbol):
term = torch_symbol_to_relax_var.setdefault(
str(arg), tvm.tir.SizeVar(str(arg), "int64")
)
else:
_, term = self._process_derived_symbol(arg, torch_symbol_to_relax_var)

if term is None:
return str(symbol), None
tir_expr = term if tir_expr is None else tir_expr + term

if isinstance(tir_expr, tvm.tir.Add):
for const, var in [(tir_expr.a, tir_expr.b), (tir_expr.b, tir_expr.a)]:
if isinstance(const, tvm.tir.IntImm) and isinstance(var, tvm.tir.Var):
return f"{var.name}___{const.value}", tir_expr

return str(symbol), tir_expr

def create_input_vars(
self, exported_program: torch.export.ExportedProgram
) -> Tuple[Dict[str, relax.Var], Dict[str, relax.Var], Dict[str, Tuple[int, int]]]:
Expand All @@ -1172,12 +1206,16 @@ def create_input_vars(

if hasattr(exported_program, "range_constraints"):
for symbol, value_range in exported_program.range_constraints.items():
symbol_name = str(symbol)
if hasattr(value_range, "lower") and hasattr(value_range, "upper"):
try:
lower = int(value_range.lower)
upper = int(value_range.upper)

symbol_name, _ = self._process_derived_symbol(
symbol, torch_symbol_to_relax_var
)
range_constraints[symbol_name] = (lower, upper)

except (OverflowError, AttributeError, TypeError):
continue

Expand Down Expand Up @@ -1235,10 +1273,8 @@ def from_exported_program(
# Initialize the block builder with a function and a dataflow block.
self.block_builder = relax.BlockBuilder()
func_name = "main"
func_attrs = {"num_input": len(user_input_vars)} if keep_params_as_input else None
func_attrs = {"num_input": len(user_input_vars)} if keep_params_as_input else {}
if range_constraints:
if func_attrs is None:
func_attrs = {}
func_attrs["tir_var_lower_bound"] = {
var_name: lower for var_name, (lower, _) in range_constraints.items()
}
Expand Down
34 changes: 34 additions & 0 deletions tests/python/relax/test_frontend_from_exported_program.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6959,5 +6959,39 @@ def main(
tvm.ir.assert_structural_equal(mod, Expected)


def test_dynamic_shape_with_derived_range_constraints():
class ConcatModel(torch.nn.Module):
def forward(self, x, y):
return torch.cat([x, y], dim=0)

@I.ir_module
class Expected:
@R.function
def main(
x: R.Tensor(("s0", 4), dtype="float32"), y: R.Tensor(("s0___1", 4), dtype="float32")
) -> R.Tuple(R.Tensor(("s0 + s0___1", 4), dtype="float32")):
s0 = T.int64(is_size_var=True)
s0___1 = T.int64(is_size_var=True)
R.func_attr(
{
"tir_var_lower_bound": {"s0": 1, "s0___1": 2},
"tir_var_upper_bound": {"s0": 64, "s0___1": 65},
}
)
with R.dataflow():
lv: R.Tensor((s0 + s0___1, 4), dtype="float32") = R.concat((x, y), axis=0)
gv: R.Tuple(R.Tensor((s0 + s0___1, 4), dtype="float32")) = (lv,)
R.output(gv)
return gv

batch = torch.export.Dim("batch", min=1, max=64)
example_args = (torch.randn(8, 4), torch.randn(9, 4))
dynamic_shapes = {"x": {0: batch}, "y": {0: batch + 1}}
exported_program = export(ConcatModel(), args=example_args, dynamic_shapes=dynamic_shapes)

mod = from_exported_program(exported_program, run_ep_decomposition=True)
tvm.ir.assert_structural_equal(mod, Expected, map_free_vars=True)


if __name__ == "__main__":
tvm.testing.main()
Loading