xxxyyyzzz12345

## Summary The following code snippet produces inconsistent and unstable values (output1 ≠ output2). However, the results are always the same when the same code snippet is run on cpu....

Fail to calculate gradient for oneflow.floor_divide when divisor is scalar

1

## Summary The following code snippet produces floor_divide derivative not implemeted error, but the exception only occurs when the divisor is a scalar (variable y in this case). ``` x...

APIs that do not support backward mode

1

## Summary I'm wondering whether oneflow.erfinv and oneflow.nn.Quantization are implemented, and if so, whether they are implemented correctly since they do not support backward mode. Output when requires_grad option is...

Inappropriate gradient computation for oneflow.fmod

2

## Summary Inappropriate gradient computation for oneflow.fmod. Theoretically, the following code snippet should output a 2. However, in reality, the produced result is a tensor [1., 1.]. ``` x =...

Wrong gradient computation for oneflow.nn.init.constant_

1

## Summary Wrong gradient computation for oneflow.nn.init.constant_. Theoretically, the gradient of constant initializer should always be 0. However, the calculated gradient is always 1. ``` x = oneflow.tensor([1.]).requires_grad_() output =...

oneflow.nn.Fold may produce inconsistent results when runned w/ and w/o requires_grad_(True)

2

## Summary oneflow.nn.Fold may produce inconsistent results when runned w/ and w/o requires_grad_(True). The following code snippet produces count != 0 ``` count = 0 input = oneflow.rand(2, 72, 16,dtype=oneflow.float64).cuda()...

Further concern about oneflow.nn.FakeQuantization

1

But when the input is outside the range [quant_min,quant_max], shouldn't the gradient be 0.0 instead of 1.0? The following code snippet sets both quant_min and quant_max to 0 and defines...

oneflow.nn.MovingAverageMinMaxObserver produces "Aborted (core dumped)" bug

1

## Summary The following code snippet produces "Aborted (core dumped)" bug with "dtype=oneflow.float64". However, "dtype=oneflow.float32" won't have this problem. ## Code to reproduce bug input1 = oneflow.rand(9, 40, 20, 10,dtype=oneflow.float64)...

oneflow.nn.conv2D causes "core dumped"

1

## Summary The following code snippet produces "Aborted (core dumped)" bug with "dtype=oneflow.float64". However, "dtype=oneflow.float32" won't have this problem. ## Code to reproduce bug import oneflow input = oneflow.rand(4,10,32,32,dtype=oneflow.float64) m...

Cannot run oneflow.roi_align() properly

1

## Summary Traceback (most recent call last): File "/mnt/sda/yjy/summer/DL-autograd-torch-main/yjy_test.py", line 24, in output = oneflow.roi_align(input0,input1,input2,input3,input4,input5,input6) oneflow._oneflow_internal.exception.OpKernelNotFoundException: File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 139, in Dispatch Dispatch(op_expr, inputs, ctx) File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 131, in...