Nov 15 2022 11:44 PM
Hi, I am a newbie, experimenting project bonsai using the cookbook and tutorials.
I am currently following the bonsai cookbook by microsoft, and amid doing data driven modelling. I am there here....
Train a brain using the data driven simulator. Considering all the necessary information is in the conf\simulator\ file to know what the inputs and outputs of the simulator are, ddm_predictor.py can use your trained model directly to run in a sequential manner. When registered with the Bonsai platform, it will Idle until you change the purpose of the simulator to Train using the Bonsai UI or bonsai-cli.
bonsai brain create -n <brainname>bonsai brain version update-inkling -n <brainname> --version <version number> -f <inklingfilename.in...bonsai brain version start-training -n <brainname>
In the second terminal start the simulator e.g.
python ddm_predictor simulator=$YOUR_SIM_CONFIG.yaml
Should i add dataset to bonsai first or do i create brain using cli, what is the best practice?
Nov 16 2022 06:09 AM
@Ashfaq_Ahamed You can create the dataset or brain in any order, but the dataset needs to exist before you can start training your brain with your data-driven sim.