NVIDIA / apex

A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch
BSD 3-Clause "New" or "Revised" License
8.43k stars 1.41k forks source link

Issue installing APEX on ubuntu #830

Open made-by-chris opened 4 years ago

made-by-chris commented 4 years ago

Usage:

I'm trying to install and run https://github.com/microsoft/DialoGPT which uses Apex

environment:

Ubuntu 16.04 conda 4.8.2 Python 3.6.9 :: Anaconda, Inc. processor: Intel® Core™ i7-9700F CPU @ 3.00GHz × 8 graphics: GeForce RTX 2060/PCIe/SSE2

Problem:

After running pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./ this occurs:

ERROR: Command errored out with exit status 1: /home/bl/anaconda3/envs/LSP/bin/python -u -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/home/bl/DialoGPT/apex/setup.py'"'"'; __file__='"'"'/home/bl/DialoGPT/apex/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' --cpp_ext --cuda_ext install --record /tmp/pip-record-j92zt9oj/install-record.txt --single-version-externally-managed --compile --install-headers /home/bl/anaconda3/envs/LSP/include/python3.6m/apex Check the logs for full command output.

( which is causes another exception Given no hashes to check 137 links for project 'pip': discarding no candidates )

AlphaPav commented 4 years ago

I have the same problem as you. Did you solve it?

made-by-chris commented 4 years ago

no not yet, im still stuck at the same step

sepal commented 4 years ago

The suggestion here worked for me:

https://github.com/NVIDIA/apex/issues/690#issuecomment-587996985

Mandule commented 4 years ago

I meet this error too

monologuesmw commented 4 years ago

I meet this error too