[BUG] DataFrame.from_dict incorrectly sets index values as data values from nested dicts

[mroeschke]

Describe the bug
DataFrame.from_dict incorrectly sets index values as data values from nested dicts

Steps/Code to reproduce bug

In [10]: import cudf

In [11]: import pandas as pd

In [13]: from collections import OrderedDict

In [14]:         a = OrderedDict(
    ...:             [
    ...:                 ("one", OrderedDict([("col_a", "foo1"), ("col_b", "bar1")])),
    ...:                 ("two", OrderedDict([("col_a", "foo2"), ("col_b", "bar2")])),
    ...:                 ("three", OrderedDict([("col_a", "foo3"), ("col_b", "bar3")])),
    ...:             ]
    ...:         )

In [15]: cudf.DataFrame.from_dict(a, orient="columns")
Out[15]: 
     one    two  three
0  col_a  col_a  col_a
1  col_b  col_b  col_b

In [16]: pd.DataFrame.from_dict(a, orient="columns")
Out[16]: 
        one   two three
col_a  foo1  foo2  foo3
col_b  bar1  bar2  bar3

Expected behavior
Out[16]

Environment overview (please complete the following information)

• Environment location: Bare-metal
• Method of cuDF install: conda