[Installation] Pypi server unavailable

mit-han-lab / torchsparse

[MICRO'23, MLSys'22] TorchSparse: Efficient Training and Inference Framework for Sparse Convolution on GPUs.

https://torchsparse.mit.edu

MIT License

1.22k stars 143 forks source link

[Installation] Pypi server unavailable #300

Closed mella30 closed 2 months ago

mella30 commented 7 months ago

Is there an existing issue for this?

[X] I have searched the existing issues

Have you followed all the steps in the FAQ?

[X] I have tried the steps in the FAQ.

Current Behavior

I click on the link to download the torchsparse wheel I need but the page is not reachable.

Error Line

Environment

- GCC: -
- NVCC: -
- PyTorch: -
- PyTorch CUDA: -

Full Error Log

Error Log

This Page Cannot Be Displayed The system cannot communicate with the external server ( pypi.hanlab.ai ). The Internet server may be busy, may be permanently down, or may be unreachable because of network problems. Please check the spelling of the Internet address entered. If it is correct, try this request later. If you have questions, please contact your local helpdesk and provide the codes shown below. Date: Mon, 08 Apr 2024 08:44:00 CEST ... URL: GET http://pypi.hanlab.ai/simple/torchsparse Category: URL Filtering Bypassed Reason: UNKNOWN Notification: GATEWAY_TIMEOUT

Jovendish commented 7 months ago

same issue, The website is still unavailable, on April 10, 2024

mbhenaff commented 7 months ago

Same here.

duguqiubai-wp commented 7 months ago

both Pypi server and alternate website are unavailable

mbendjilali commented 6 months ago

Still unreachable here.

ColsonZhang commented 6 months ago

Still unreachable here. Please fix it.

kentang-mit commented 6 months ago

Hi all,

We've relaunched the pypi server. Sorry for the inconvenience.

Best, Haotian

wjz51 commented 5 months ago

Still unreachable here.

LarsPh commented 5 months ago

Still unreachable on May 26.

ybc-ybc commented 5 months ago

Hi all,

We've relaunched the pypi server. Sorry for the inconvenience.

Best, Haotian

Still unreachable.

ys-2020 commented 5 months ago

Hi All, the pypi server has been restarted and can now be accessed. Sorry for the inconvenience. Many thanks to @kentang-mit .

zhongqiu1245 commented 5 months ago

Still broken

JBKnights commented 5 months ago

Still broken for me too

Tortoise0Knight commented 5 months ago

Still unreachable

Cheburum commented 5 months ago

Still broken for me

daniil-chesakov commented 5 months ago

Still broken

RuiyuM commented 4 months ago

still borken

mella30 commented 4 months ago

Hi All, the pypi server has been restarted and can now be accessed. Sorry for the inconvenience. Many thanks to @kentang-mit .

Can you restart again, maybe on a regular basis? Does not seem to run stable..

@kentang-mit

vettelx commented 4 months ago

Still broken @kentang-mit

AoxiangFan commented 4 months ago

still broken

weweng1999 commented 4 months ago

It is still broken.

brycekl commented 4 months ago

It is still broken.

jacksonsunny29 commented 3 months ago

Build it from source. Worked for me.

jacksonsunny29 commented 3 months ago

pip install --upgrade git+https://github.com/mit-han-lab/torchsparse.git@v1.1.0 Change the versions accordingly. Check the github repo for the versions. The current one is v2.0.0

mella30 commented 3 months ago

Build it from source. Worked for me.

There has to be a GPU available at build time for that to work (at least for the GPU-accelerated version). For me, this is not the case during cloud deployment.

jacksonsunny29 commented 3 months ago

I see your point. Indeed a GPU is required at build time for the usage of cuda.

zhijian-liu commented 2 months ago

Could you try to install TorchSparse from source using

pip install git+https://github.com/mit-han-lab/torchsparse.git