wuyii8941 opened a new issue, #19574:
URL: https://github.com/apache/tvm/issues/19574
## Description
`LegalizeOps` lowering of `relax.nn.batch_norm` computes mean and variance
from the input data (training mode) instead of using the provided
`running_mean` and `running_var` parameters (inference mode). This produces
incorrect results when `batch_norm` is used for inference.
When `DecomposeOpsForInference` is applied before `LegalizeOps`, the
batch_norm is correctly decomposed using the provided running statistics.
## Reproducer
```python
import numpy as np
import tvm
from tvm import relax
import tvm.relax.op as R
N, C, HW = 1, 4, 4
bb = relax.BlockBuilder()
x = relax.Var('x', relax.TensorStructInfo((N, C, HW, HW), 'float32'))
gamma = relax.Var('gamma', relax.TensorStructInfo((C,), 'float32'))
beta = relax.Var('beta', relax.TensorStructInfo((C,), 'float32'))
mean = relax.Var('mean', relax.TensorStructInfo((C,), 'float32'))
var = relax.Var('var', relax.TensorStructInfo((C,), 'float32'))
with bb.function('main', [x, gamma, beta, mean, var]):
with bb.dataflow():
bn = bb.emit(R.nn.batch_norm(x, gamma, beta, mean, var, axis=1,
epsilon=1e-5))
out = bb.emit_output(relax.TupleGetItem(bn, 0))
bb.emit_func_output(out)
mod = bb.finalize()
x_np = np.ones((N, C, HW, HW), dtype=np.float32)
gamma_np = np.array([1.0, 2.0, 0.5, 3.0], dtype=np.float32)
beta_np = np.array([0.0, 1.0, -1.0, 0.5], dtype=np.float32)
mean_np = np.array([0.0, 0.5, 1.0, -1.0], dtype=np.float32)
var_np = np.array([1.0, 0.25, 4.0, 0.01], dtype=np.float32)
inputs = [x_np, gamma_np, beta_np, mean_np, var_np]
def run(mod, passes, inputs):
pipeline = tvm.ir.transform.Sequential(passes)
mod_l = pipeline(mod)
exe = tvm.relax.build(mod_l, target='llvm')
vm = tvm.relax.VirtualMachine(exe, device=tvm.cpu())
tvm_inputs = [tvm.runtime.tensor(x, device=tvm.cpu()) for x in inputs]
return vm['main'](*tvm_inputs).numpy()
out_legalize = run(mod, [relax.transform.LegalizeOps()], inputs)
out_correct = run(mod, [relax.transform.DecomposeOpsForInference(),
relax.transform.LegalizeOps()], inputs)
# Expected: (1.0 - (-1.0)) / sqrt(0.01 + 1e-5) * 3.0 + 0.5 = 60.47
# Channel 3: legalize=0.5 (wrong), correct=60.47
print(f"LegalizeOps: {out_legalize[0, 3, 0, 0]:.4f}") # 0.5000 (WRONG)
print(f"Correct: {out_correct[0, 3, 0, 0]:.4f}") # 60.4700
```
## Expected behavior
`LegalizeOps` batch_norm should use the provided `running_mean` and
`running_var` for inference, producing the same result as
`DecomposeOpsForInference`.
## Actual behavior
`LegalizeOps` computes mean/var from the input tensor (training-mode batch
normalization), producing incorrect results. The max error vs ground truth is
59.97.
## Environment
- TVM version: 0.24.dev0 (commit 0b0afd8dd, 2026-04-24)
- Target: llvm (CPU)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
