Pierre-Louis Guhur
Pierre-Louis Guhur
I can confirm that one needs to run PyTorch 1.0.0 (tested with python 3.6 + CUDA 10.0.130 + cuDNN 7401) and not PyTorch 1.6.0 (+ CUDA 10.0.130). It might be...
Ok great! I am working usually on three different remote servers, so being able to jump easily from one session to the other one would be a real time saver...
Sorry nope :/ On Sat, Feb 26, 2022, 10:41 AM Siteng ***@***.***> wrote: > @guhur did you manage to find a solution for > this probelm? > > — >...
It works on-the-shelf if you put the credentials in `.env`
I can confirm this bug happens to me on 4.5.1 but not on 4.4.0. My file is https://drive.google.com/drive/folders/1Qx_2_ePIqf_Z6SnpPkocUiPgFeCfePQh?usp=sharing it was tested on a machine located in France.
bug fix here : https://github.com/wkentaro/gdown/pull/210
This error disappeared for me when I used Python 2 instead of Python 3
> In run_bigbatch.sh just use torch.distributed.launch --nprocs_per_node=2(since i had two gpus on one node) --nnodes=$1 --node_rank=$2 --master_addr=$3 --master_port=$4 main.py --my arguments Thanks that was helpful :) Actually, the argument name...
Here how I do to setup jeanzay with `sshfs`: ```bash mkdir -p $PREFIX/scratch/jeanzay sshfs -o reconnect,ServerAliveInterval=15,ServerAliveCountMax=3 jean-zay:/path/to/your/scratch/ $PREFIX/scratch/jeanzay ```
@Stefuniverse's solution is working for me! Thanks