Open rajkaramchedu opened 4 months ago
cc: @gus-opentensor
Posting these questions here for future reference when documenting this feature.
The main purpose of btcli profile create is so that the user can simply pass the config file to a command like python neurons/validator.py --netuid 1 --subtensor.network test --wallet.name validator --wallet.hotkey default --logging.debug
instead of typing all these options and their values, correct?
How would I pass the created config file? Like, python neurons/validator.py --config ~/.bittensor/profiles/myfirstprofile
?
Are these below the only key-value pairs created in the profile file? -- name (is this the name of the output file name, like "myfirstprofile" above? -- path (path to the location of the "myfirstprofile" I think?) -- coldkey: does it accept both "0x" prefixed-format and also the SS58 format? -- hotkey: same q as above -- subtensor_network: allowed values I think are "test, finney, local" correct? -- active_netuid: I think the netuid of interest
In the screenshot in the PR, I see that provided inputs are "1" and "2" and "3" for the prompted fields. Does it mean it is not doing any typecheck?
Two terms are used in the PR: "profile" and "config" so a single profile can contain multiple configs, correct?
Lastly, are these config files in plain text format or yaml or ? Asking because can the user simply follow a sample file example and manually create these profiles and configs for each profile? A good option to document if this is true.
See this PR and document what's necessary. https://github.com/opentensor/bittensor/pull/1784