crapthings

its working, but slow

Did you bring the dataset with 44k during the training? HQSeg-44K

this work for me, lower the batch_size and nproc_per_node if u have only 1 gpu ``` torchrun --nproc_per_node=2 train.py --checkpoint ./pretrained_checkpoint/sam_vit_h_4b8939.pth --batch_size_train 16 --model-type vit_h --output work_dirs/hq_sam_h torchrun --nproc_per_node=2 train.py...

what about this one? https://swc.rs/ https://deno.land/[email protected]/typescript/overview#how-does-it-work > At a high level, Deno converts TypeScript (as well as TSX and JSX) into JavaScript. It does this via a combination of the...

first clone these packagess in your project/packages - standard-minifier-js # find package.json and change package name to scope:standard-minifier-js - minifier-js # change name to scope:minifier-js ```js esbuild = esbuild ||...

@jankapunkt for large records its slow than mongo distinct

how about ws https://github.com/websockets/ws https://npmcompare.com/compare/derby,socket.io,sockjs,websocket-node,ws

https://github.com/meteor/meteor-feature-requests/issues/268

localStorage package https://github.com/meteor/meteor/blob/982181fff86e09a75f1d2681a78b9d784e666f4f/packages/localstorage/localstorage.js#L51 turn into something like this? https://github.com/crapthings/meteor-multiple-accounts-sandbox/blob/master/mantra-sample-blog-app/packages/localstorage/localstorage.js#L37 ```js // Meteor._localStorage is not an ideal name, but we can change it later. // Let's test to make sure that...

better have this in core method name can be useful