update some test_tensor.py cases with 0 in shape (#2368)

test/test_tensor.py

@@ -1,6 +1,5 @@
 import numpy as np
 import torch
-import struct
 import unittest, copy
 import mmap
 from tinygrad.tensor import Tensor, Device
@@ -179,6 +178,7 @@ class TestTinygrad(unittest.TestCase):
     assert a.shape == b.shape, f"shape mismatch (Tensor.ones_like){a.shape} != (torch){b.shape}"
 
   def test_ndim(self):
+    assert Tensor(1).ndim == 0
     assert Tensor.randn(1).ndim == 1
     assert Tensor.randn(2,2,2).ndim == 3
     assert Tensor.randn(1,1,1,1,1,1).ndim == 6
@@ -216,7 +216,7 @@ class TestTinygrad(unittest.TestCase):
     assert Tensor.randn(1,2,5).numel() == 10
     assert Tensor.randn(1,1,1,1,1,1).numel() == 1
     assert Tensor([]).numel() == 0
-    # assert Tensor.randn(1,0,2,5) == 0 # TODO: fix empty tensors
+    assert Tensor.randn(1,0,2,5).numel() == 0
 
   def test_element_size(self):
     for _, dtype in dtypes.fields().items():
@@ -230,8 +230,8 @@ class TestTinygrad(unittest.TestCase):
     x.dot(layer).mean().backward()
 
   def test_zerosized_tensors(self):
-    Tensor([]).realize()
-    Tensor([]).numpy()
+    np.testing.assert_equal(Tensor([]).numpy(), np.array([]))
+    np.testing.assert_equal(Tensor(None).numpy(), np.array([]))
 
   def test_tensor_ndarray_dtype(self):
     arr = np.array([1]) # where dtype is implicitly int64
@@ -269,10 +269,6 @@ class TestTinygrad(unittest.TestCase):
     np.testing.assert_allclose(ua_arr, (Tensor(ua_arr)/Tensor(1)).numpy())
 
 class TestZeroShapeTensor(unittest.TestCase):
-  def test_from_empty(self):
-    np.testing.assert_equal(Tensor([]).numpy(), np.array([]))
-    np.testing.assert_equal(Tensor(None).numpy(), np.array([]))
-
   def test_shape_stride(self):
     t = Tensor.rand(3, 2, 0)
     assert t.shape == (3, 2, 0)