[BUG] The pass "eliminate_nop_dropout" causes output mismatch

[BUG] The pass "eliminate_nop_dropout" causes output mismatch

Issue
Running the single pass eliminate_nop_dropout with onnxoptimizer 0.4.2 changes numerical outputs.

Environment

• Ubuntu 20.04
• Python 3.10
• onnx==1.19.0
• onnxruntime==1.23.2
• onnxoptimizer==0.4.2(latest)

Repro steps (run from this folder)

eliminate_nop_dropout_repro.tar.gz

1. Create a Python environment (Python 3.10) and install dependencies:

python3 -m venv .venv
source .venv/bin/activate
pip install -U pip
pip install -r requirements.txt

2. Optimize a case with only eliminate_nop_dropout (writes model.opt.onnx next to model.onnx):

• python optimize_model.py --case ./case_03140_seed45116679

1. Differential test original and optimized outputs using stored oracle inputs (ORT graph optimization disabled):

• python diff_test.py --case ./case_03140_seed45116679

Expected
eliminate_nop_dropout should be semantics-preserving. Applying this pass alone should not change any output values.

Differential Test Output Details

1. case_03140_seed45116679 — output[14] max_abs=4.022e+04 max_rel=4.022e+12; output[15] max_abs=5.190e+02 max_rel=5.120e+10; overall mismatch.

Case: eliminate_nop_dropout/case_03140_seed45116679 (ORT graph optimizations disabled)
  output[0]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 21)
  output[1]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(2, 1, 24)
  output[2]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(42336,)
  output[3]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 21, 24)
  output[4]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(21, 1)
  output[5]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(21, 1)
  output[6]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(21, 1)
  output[7]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(2, 24)
  output[8]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 24)
  output[9]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 24)
  output[10]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 24)
  output[11]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(1, 24)
  output[12]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(21, 24)
  output[13]: max_abs=0.000e+00, max_rel=0.000e+00, shape=(24, 21, 1)
  output[14]: max_abs=4.022e+04, max_rel=4.022e+12, shape=(1, 21, 24)
  output[15]: max_abs=5.190e+02, max_rel=5.120e+10, shape=(32, 21, 24)
Overall: max_abs=4.022e+04, max_rel=4.022e+12

Attachments

• requirements.txt
• optimize_model.py (runs only eliminate_nop_dropout and saves model.opt.onnx)
• diff_test.py (runs original and optimized with same oracle inputs; ORT graph optimizations disabled)
• run_optimized_model.py (executes optimized model with oracle inputs)
• case_03140_seed45116679/ (original model and oracle inputs)

Run optimize_model.py and diff_test.py from the attached reproduction folder for case_03140_seed45116679, using the listed Python dependencies and disabling ORT graph optimizations as shown. Compare the original and optimized models and investigate the eliminate_nop_dropout pass. Done means the pass preserves all output values and the differential test reports no mismatch.