Skip to content

Load external ONNX tensor data before transforming models #2227

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
happyme531 wants to merge 1 commit into from
Closed

Load external ONNX tensor data before transforming models #2227

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
1 change: 1 addition & 0 deletions olive/passes/onnx/conversion.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -655,6 +655,7 @@ def _run_for_config(
) -> ONNXModelHandler:
output_model_path = resolve_onnx_path(output_model_path)
model_ir = model.load_ir_model()
ir.external_data.load_to_model(model_ir)
version_converter.convert_version(model_ir, config.target_opset, fallback=True)
return ir_model_to_olive_model(model_ir, output_model_path, config)

Expand Down
1 change: 1 addition & 0 deletions olive/passes/onnx/hqq_quantization.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,7 @@ def _run_for_config(
return model
output_model_path = resolve_onnx_path(output_model_path, Path(model.model_path).name)
ir_model = model.load_ir_model()
ir.external_data.load_to_model(ir_model)
ir_model.graph.opset_imports[MSFT_DOMAIN] = 1
self._quantize_model(
ir_model,
Expand Down
4 changes: 3 additions & 1 deletion olive/passes/onnx/onnxscript_fusion.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
from pathlib import Path

from onnxscript.rewriter import ort_fusions
import onnx_ir as ir

from olive.hardware.accelerator import AcceleratorSpec
from olive.model import ONNXModelHandler
Expand All @@ -30,7 +31,8 @@ def _run_for_config(
output_model_path = resolve_onnx_path(output_model_path, Path(model.model_path).name)

model_ir = model.load_ir_model()

ir.external_data.load_to_model(model_ir)

# TODO(exporter team): Different fusions support different devices
model_ir, function_stats = ort_fusions.optimize_for_ort(model_ir)
logger.debug("Function stats: %s", function_stats)
Expand Down
1 change: 1 addition & 0 deletions olive/passes/onnx/rtn_quantization.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,6 +77,7 @@ def _run_for_config(
) -> ONNXModelHandler:
output_model_path = resolve_onnx_path(output_model_path, Path(model.model_path).name)
ir_model = model.load_ir_model()
ir.external_data.load_to_model(ir_model)
ir_model.graph.opset_imports[MSFT_DOMAIN] = 1
self._quantize_model(
ir_model,
Expand Down