Back
Similar todos
Load previous page…
profit from 
#crisp self-built ai development rig, total cost = 2k€, that will safe a ton of money monthly for our staging infrastructure
render locally, use remote as bridge server to neural nets 
#mused
start to build in-house AI inference API & job manager #crisp
setup metashape headless on gcp vm & use python bindings to automate processing 360 video #mused
Updating Automatic 11, Installing a new video gpu, today its a day to learn how to train local models :B 
#dailywork
plan prod deployment strategy 
#wandlimb
fail w/medium dataset from 360 video (~6,000 images) like small, but full requires too much memory, next strategy is render each input separately #mused
finish migrating 
#mirage kubernetes intel and nvidia gpu instances to scaleway, getting last-generation NVIDIA L40S + L4 GPUs, running much smoother now! (previously: old A40 and A16)
create a neural network 
#ciphey
Setup JupyterLab for AI development
Get the deep neural network working #ciphey
deploy new #crisp dynamic billing system for ai tokens usage in production w/ some scalability optimization
deploy new #crisp ai features in production which have been cooking for 3 months
Finished home AI server build: 2 x 2697v2, 128 GB ram, 2 x Nvidia Tesla P40 24 GB gpus. 
#2markdown
