kubectl -n kubeflow get pods List pods in namespace kubeflow
kubectl -n kubeflow logs <podname> Show log of in namespace kubeflow
kubectl -n kubeflow delete <podname> Delete pod in namespace kubeflow
kubectl -n kubeflow delete pod <podname> Delete in namespace kubeflow
docker system prune -a delete all unused and dangling containers, etc.
docker image prune -a delete all unused and dangling images
tar -c --use-compress-program=pigz -f tar.file dir_to_zip Use pigz parallel implementation of gzip
apt-cache show <package name> Get information about a package
git checkout -b myfeature develop Branch off from develop branch, switch to branch myfeature
Merge work from myfeature into develop branch
git checkout develop
git merge --no-ff myfeature
git branch -d myfeature
git push origin develop
git add -A stage all changes
git add -u stage modifications and deletions, without new files
GIT_SSH_COMMAND="ssh -i ~/.ssh/id_rsa_example" <git command> Specify identity file
git commit -F <changelog> Commit and use file as commit message
The Git index is where you put changes that you want to be included in the next commit.
<leader>ga Stage the current file
<leader>gs Bring up a summary window vaguely akin to git-status
<leader>gc A wrapper around git-commit. A new tab will be used to obtain
a commit message. Write and close that window (:wq or |:Gwrite|)
to finish the commit.
<leader>gt ???
<leader>gd Perform a |vimdiff| against the index version of the current file.
<leader>ge |:edit| a |fugitive-object|
<leader>gr From the working copy, reset all changes. OR from the index file,
add all changes from the current file to the index file.
<leader>gw From the working copy, add all changes from the current file to
the index file. OR from the index file, reset all changes to
the working copy.
<leader>gl
<leader>gp
<leader>gm Rename/relocate file, update buffer
<leader>gb
<leader>gb
<leader>go
<leader>gps
<leader>gpl
[c In vimdiff, jump to previous changeset
]c In vimdiff, jump to next changeset
ls List active sessions
a -t n -d Attach to session number n and disconnect any other screens
, Name window
$ Name session
%norm cs'" Run the same command on every line (or selected lines)
Python compatible tabs
filetype plugin indent on
" show existing tab with 4 spaces width
set tabstop=4
" when indenting with '>', use 4 spaces width
set shiftwidth=4
" On pressing tab, insert 4 spaces
set expandtab
zz Center current line on screen
Ctrl + e Move screen up without moving cursor
Ctrl + y Move screen down without moving cursor
Ctrl + o Jump to last edit location
Ctrl + i Jump to next edit location
Ctrl + w{h,j,k,l} move focus to {left, lower, upper, right} split
:vsplit <another file> Open another file in side-by-side split
Ctrl + wx Swap split
Ctrl + w= Equalize split
Ctrl + wo Close other windows
:tabm i Move tab to position i
:tabm +i Move tab right i positions
:tabm -i Move tab left i positions
<leader>h Switch to left buffer
<leader>l Switch to right buffer
ma Set position for mark a
`a Jump to location at mark a
:bd Unload the buffer, delete it from the buffer list, losing changes
:bw Unload the buffer, delete it from the buffer list, writing changes
gqq Format entire document
da[ Delete around left bracket
yi" yank in quotation marks
<C-R> " Paste ring buffer into command line
y( yank from beginning of sentence
y) yank to end of sentence
- Press
<F5>to purge the cache for the current directory to get new files, remove deleted files and apply new ignore options. - Press
<c-f>and<c-b>to cycle between modes. - Press
<c-d>to switch to filename only search instead of full path. - Press
<c-r>to switch to regexp mode. - Use
<c-j>,<c-k>or the arrow keys to navigate the result list. - Use
<c-t>or<c-v>,<c-x>to open the selected entry in a new tab or in a new split. - Use
<c-n>,<c-p>to select the next/previous string in the prompt's history. - Use
<c-y>to create a new file and its parent directories. - Use
<c-z>to mark/unmark multiple files and<c-o>to open them.
conda info --envs View a list of environments
conda create --name myclone --clone myenv Clone an environment
conda remove --name myenv --all Remove an environment
conda list --explicit > spec-file.txt List packages installed in environment, write to spec-file.txt
conda create --name myenv --file spec-file.txt Create new environment from spec-file.txt
Accidentally pressed Ctrl+s: Ctrl+q to unlock
fuser -v /dev/nvidia* What process(es) are using the GPU?
rsync -avzP –include ‘*.txt’ –exclude ‘*’ /src/dir [email protected]:/dst/dir Sync a local directory to a remote directory, with progress bar
rsync -e "ssh -i ~/.ssh/abc_rsa" /src/dir [email protected]:/dst/dir Use an identity file
Ref: https://linuxtechlab.com/start-with-rsync-command-8-rsync-examples/
scp -i ~/.ssh/abc_rsa Use identity file abc_rsa
scp -r Copy a directory
tensor.detach() creates a tensor that shares storage with tensor that does not require grad
tensor.clone() creates a copy of tensor that imitates the original tensor's requires_grad field
You should use detach() when attempting to remove a tensor from a computation graph, and clone as a way to copy the tensor while still keeping the copy as a part of the computation graph it came from.
tensor.data returns a new tensor that shares storage with tensor. However, it always has requires_grad=False (even if the original tensor had requires_grad=True
You should try not to call tensor.data in 0.4.0. What are your use cases for tensor.data?
tensor.clone() makes a copy of tensor. variable.clone() and variable.detach() in 0.3.1 act the same as tensor.clone() and tensor.detach() in 0.4.0.
Ref: https://discuss.pytorch.org/t/clone-and-detach-in-v0-4-0/16861/2
mysql --host=<ip> --port=<port> -u <username> -p Connect to server, prompt for password
Ref: https://gist.github.com/hofmannsven/9164408
Ref: https://jhui.github.io/2018/02/09/PyTorch-neural-networks/
The difference between torch.nn and torch.nn.functional is very subtle. In fact, many torch.nn.functional have a corresponding equivalent in torch.nn. For layers with trainable parameters, we use torch.nn to create the layer.
In many code samples, it uses torch.nn.functional for simpler operations that have no trainable parameters or configurable parameters.
ssh -D 8080 -q -C -N <ip> Open socks proxy to on port 8080
find /wbdata/Datasets/output1/ -type f \( -name "*.jpg" \) -printf "%f\n" |sed 's/\.jpg$//1' > all.txt
ffmpeg -r 2 -i '%*.jpg' fullstitch.mp4 Stitch a bunch of jpegs together into a video at 2 fps
ssh user@server "echo \"cat ~/.ssh/id_rsa.pub\" >> .ssh/authorized_keys" Copy public key to authorized_keys on remote machine
ps afx Get a listing of all processes and child processes with full commands
usermod -a -G sudo geek Add existing user to existing group