BatchNorm

import set_env
from tvm.relay.frontend.onnx import BatchNorm

BatchNorm??

Init signature: BatchNorm()
Source:        
class BatchNorm(OnnxOpConverter):
    """Operator converter for BatchNorm."""

    @classmethod
    def _impl_v1(cls, inputs, attr, params):
        # TODO(zhreshold): 'spatial' is not properly handled here.
        # TODO(vvchernov): 'training_mode' (onnx tag) is not correctly handled, ignore for now
        out = AttrCvt(
            op_name="batch_norm",
            ignores=["spatial", "is_test", "consumed_inputs", "momentum", "training_mode"],
        )(inputs, attr, params)
        # We only support test mode, so we return data, moving_mean, moving_var,
        # and then moving_mean and moving_var again as placeholders for
        # the expected "saved_mean", "saved_var".
        return _expr.TupleWrapper(_expr.Tuple((*out, out[1], out[2])), 5)
File:           /media/pc/data/lxw/ai/tvm/python/tvm/relay/frontend/onnx.py
Type:           type
Subclasses:

from tvm.relay import expr as _expr
_expr.TupleWrapper?

Init signature: _expr.TupleWrapper(tuple_value, size)
Docstring:     
TupleWrapper.

This class is a Python wrapper for a Relay tuple of known size.
It allows for accessing the fields of the Relay tuple as though
it were a Python tuple.

Parameters
----------
tuple_value: tvm.relay.Expr
    The input tuple

size: int
    The size of the tuple.
File:           /media/pc/data/lxw/ai/tvm/python/tvm/relay/expr.py
Type:           type
Subclasses:

from tvm.relay.frontend.common import AttrCvt

out = AttrCvt(
    op_name="batch_norm",
    ignores=["spatial", "is_test", "consumed_inputs", "momentum", "training_mode"],
)