cannot init scalar tensor
- Dominant language
- C++
- Stars
- 3.6k
- Forks
- 1.3k
- PR merge metrics
- No merged PRs in 30d
Description
Hi, @dcslin , since some onnx models need scalar-tensor but we can't support.
The error is:
> Traceback (most recent call last):
File "../../test/python/test_operation.py", line 4015, in test_tmp
x = tensor.from_numpy(x)
File "/usr/local/lib/python3.5/dist-packages/singa/tensor.py", line 766, in from_numpy
ret.copy_from_numpy(np_array)
File "/usr/local/lib/python3.5/dist-packages/singa/tensor.py", line 307, in copy_from_numpy
assert np_array.size == self.size(), 'tensor shape should be the same'
AssertionError: tensor shape should be the same
please use this test case:
```python3
x = np.array(256, ndmin=0).astype(np.int32)
x = tensor.from_numpy(x)
```
Contributor guide
Assessment
This issue has not been assessed yet.