oneflow

oneflow copied to clipboard

## Summary This will raise a crash because a 0D tensor cannot be repeated meaningfully. ## Code to reproduce bug ``` import oneflow as flow zero_dim_tensor = flow.tensor(5) result =...

x0w3n

## Summary A crash was triggered after continuous min calculations. ## Code to reproduce bug ``` import oneflow as flow input_tensor = flow.tensor([[1, 2], [3, 4]]) min_result = input_tensor.min(1) print(min_result)...

x0w3n

## Summary In the official example flow.linalg.inv computes normally, but triggers a crash under the specific tensor. ## Code to reproduce bug ``` import oneflow as flow A = flow.tensor([[...

x0w3n

## Summary crash is triggered when dim=0. ## Code to reproduce bug ``` import oneflow as flow x = flow.tensor(5) result = x.cumprod(dim=0) ``` ``` import oneflow as flow x...

x0w3n

## Summary crash is triggered when dim=0, otherwise normal. ## Code to reproduce bug ``` import oneflow as flow input_tensor = flow.tensor([[1, 2, 3], [4, 5, 6]]) invalid_indices = [0,...

x0w3n

## Summary Attempting to split the tensor with an invalid partition size triggers a crash. ## Code to reproduce bug ``` import oneflow as flow a = flow.tensor([1, 2, 3,...

x0w3n

## Summary A crash is triggered when slice_tup_list 's dimension exceeds the boundary. ## Code to reproduce bug ``` import oneflow as flow import numpy as np input = flow.tensor(np.array([1,...

x0w3n

## Summary A crash is triggered when the indices parameter exceeds the dimensions of the input tensor. ## Code to reproduce bug ``` import oneflow as flow tensor = flow.tensor([0,...

x0w3n

## Summary A crash is triggered when processing a negative input. ## Code to reproduce bug ``` import oneflow as flow n = -5 result = flow.randperm(n) ``` output: ```...

x0w3n

## Summary A crash is triggered when processing a nonexistent path. ## Code to reproduce bug ``` import oneflow as flow record_reader = flow.nn.OFRecordReader( "/tmp/test/", batch_size=16, part_name_suffix_length=5, ) val_record =...

x0w3n