LukeLin

Results 109 comments of LukeLin

When doing pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./, shows ModuleNotFoundError: No module named 'packaging'

I also met this problem. https://github.com/NVIDIA/apex/issues/1748#issuecomment-1928910265 solved my problem.

开源四年的一些记录

我记得有个issue 被秒关了， 当时很难过。

usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]

> > Hi, all When I run the `python setup.py` command in cmd, I got these print: > > ``` > > usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]...

pip install -e python failed

You should use `pip install -e .` in the dir where has the `setup.py` or `pyproject.toml`

Can't change rate incoming with tcset

`sudo tc qdisc add dev enp0s1 root fq limit 1000` I try to change the queue limit, but fail `Error: Exclusivity flag on, cannot modify.`

There is an error when quiver's features are applied to the ogb-papers00M dataset on DGL

It may out of memory.

Provide a newer dockerfile

dockerhub [1263810658](https://hub.docker.com/repository/docker/1263810658/torch_quiver/general)

CPU only sample will hang

I find in serving.py line 129 ` result_queue.put((res, time.perf_counter()-start))` will stuck.

quiver sampler has no rand?

What did you mean by the same result? Same subgraph?

Does the input tensor have to be from CPU?

Usually, we assume CPU memory is big enough. You can refer to https://marius-project.org/publications/ to find some papers which store the node in the disk.