Keren Zhou

On cpu backend, I only find l2 is support in the forward pooling phase but not backward pooling phase. So it is not supported in neon?

Some misleading syntax

1

In `Dot_engine-inl.h`, the default type of a template argument is set by `template` While in other places like `Tensor.h`, I notice that `DType` is set by `#define MSHADOW_DEFAULT_DTYPE = default_real_t`....

Some implementations in Table 3 are missing

2

RGCN, HetSANN, and HGT code are not provided in this repo. Do you have any plan to upload your implementation in the future?

Read hpctoolkit database without metric-db

10

Hi all, When we were at the scalable tool workshop, we found that hatchet was not able to read hpctoolkit database without metric-db files. I would like to query if...

Hatchet multi-process problem when handling large graphs

8

### Problem The following code will be hanging there forever with the default number of procs for `g_range.filter`. However, if `num_procs=1`, `g_range.filter` could return reasonable results. ### Code > import...

[MemoryProfiler] Don't declare non built-in type variables in the header file

7

The problem appears for every new analysis header file

https://docs.nvidia.com/cuda/cuda-driver-api/group__CUDA__MALLOC__ASYNC.html#group__CUDA__MALLOC__ASYNC On GVProf side, if memory is recycled through a memory pool, we have to update its allocation context to better identify a memory object.