[BUG] Grouping by callables doesn't match pandas

[Matt711]

Describe the bug
See the repro

Steps/Code to reproduce bug

import pandas as pd

df = pd.DataFrame(
    {
        "A": ["foo", "bar", "foo", "bar"],
        "B": ["one", "one", "two", "three"],
        "C": [1,3,6,10],
        "D": [2,4,6,8],
    }
)
col1 = df["A"]
col2 = df["B"]

grouped = df.groupby([col1.get, col2.get])
agged = grouped.mean(numeric_only=True)
print(agged.loc[:, ["C", "D"]])

with cudf.pandas

      C    D
0   1.0  2.0
1   3.0  4.0
2   6.0  6.0
3  10.0  8.0

with pandas:

              C    D
bar one     3.0  4.0
    three  10.0  8.0
foo one     1.0  2.0
    two     6.0  6.0

Expected behavior
Should match pandas