Tracking Large Tensor vector test coverage

[ChaiBapchya]

As part of the ongoing effort to cover certain set of operators for testing large vector support (vector size > 2^32)

Few ops that haven't been covered yet / Need to be fixed

1. ndarray convert types
   e.g. a.tostype('row_sparse')

2. nd.dot

3. nd.where

4. nd.sparse.where

[mxnet-label-bot]

Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it.
Here are my recommended labels: Test

[ChaiBapchya]

@mxnet-label-bot add [Test]

[ChaiBapchya]

From @access2rohit PR #15943
9bff00f

[ChaiBapchya]

From my PR #15941
SequenceLast

======================================================================
FAIL: test_large_vector.test_sequence_last
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/ubuntu/anaconda3/lib/python3.6/site-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/home/ubuntu/workspace/lts/vector/incubator-mxnet/tests/nightly/test_large_vector.py", line 236, in test_sequence_last
    assert_almost_equal(b.asnumpy(), a[-1].asnumpy())
  File "/home/ubuntu/workspace/lts/vector/incubator-mxnet/python/mxnet/test_utils.py", line 522, in assert_almost_equal
    raise AssertionError(msg)
AssertionError:
Items are not equal:
Error 100000.000000 exceeds tolerance rtol=0.000010, atol=0.000000.  Location of maximum error:(1,), a=0.000000, b=10000000000.000000
 a: array([1.1785947e+10, 1.2611686e-44], dtype=float32)
 b: array([1.e+10, 1.e+10], dtype=float32)

----------------------------------------------------------------------