N.KOJIRA.IO
Timeline

captjack πŸ΄β€β˜ οΈβœ¨πŸ’œ

[email protected]
npub1te0uzs6vj29umjaxlqqct82j8q6ppyefrxq06dhr8d6pvwfatgkqjmjgwp
captjack
already posted 8 months ago MULTI YEAR trend not
Dario wants DeepSeek to stop releasing open-source models
And DeepSeek plans to give it a big surprise gift
after Anthropic’s IPO 🎁🐳
no need change nsec or provider anymore since all working n u will push button to post as backup. sticking to same npub=provider is good thing.
any provider can LOG esp jewish big ones

using some direct DOES NOT SOLVE ur risk

unless DIRECT is own setup hardware/container.
it's good
do the npub used by routstr - need the profile to setup properly?
Reminder πŸ‘‰οΈ not using nostr DM
ping here - join / give simplex only for non-public info share/text-chat only
GOYIM ✑ GAYMAN TRAP
https://pbs.twimg.com/media/HXbyCf0XwAA-cHw.jpg CVN73 going to πŸ‡ΉπŸ‡­pattaya for holiday break from war
all agents knows default files path and see any keys passwd much more easily than humans / devs.
vLLM is hard vLLM is best for multiple GPUs in same LAN or riser.

vLLM container~14GB

nvidia cuda docker passthrough is MUST

#AI local python
nostr self-hosting / provider essentials

fastly portforward frps rustdesk nextcloud simplex Headscale
698c8b2d7165

restart is just stop "podman-compose up" and wait a bit - start again
hahaha Thanks for input - shithappened learn a lot
old setup middle running compose I/O error hapened
hard reboot fixed ssd issue - but then all setup went corrupt and couldnot do it anyone
even i restore 3 db files + secrect key - it started but balance was gone

anyway now made fresh setup in PODMAN - same trick capadd RAW routing

just got it working all again - 698c8b2d7165 new id look

so i will NOT nsec manually now
seems .env is OVERRIDE settting changed from UI n stored in sqllite db files

if descript or models changes pushed every 6hrs will not change anymore nsec now

also pasted id in ui and disable autonsec generate for now see what happens

bottom like any changes NOT reflected in 7hrs may update nsec which may change provided ID or not dunno

docker or odman BOTH work - added 2 lines .yaml file for networking
summary QUICK TIP ( 🐧 cli power users)to get started in less than hour in you have decent pc - (windows / mac / andriod also can ) banda is web based also or app
download from banda.rocks you preferred version / github
wget https:// .....banda-latest.tar.gz
tar xvf band*
banda/banda$ bin/banda serve --port 7788 --box host
you see web link address + token on screen
copy paste + access from browser
use front-end proxy ngix haproxy from lan/wan/vpn access

Select Local OpenAI Compatible punch routstr url + access token and params provider-url/models
save n use
https://dare-sydney-ethical-aspect.trycloudflare.com/f559…

https://dare-sydney-ethical-aspect.trycloudflare.com/953a…
so http URL changes - seems onion becomes suppressed in roustr . com site
also what (IF any) is best way update URL changes / description / models ?
seems change nsec will broadcast again or need wait 24hrs
or other content is shown
it happen last year is another country

all assholes wakes in nostr only when DILDO get in their own ass

until then it is somebody else issue
XMR based SMALL LLM PROVIDER PLATFORM - user is offering LLM / container / routing - not using / consuming
verify what ? npub
GOYIMS DID IT

$6B Binance hack

AI + Cloud Outage - since they ✑SACKED / LAYOFF all brown/yellow dicked humans

Banks SELF SABOTAGE since they cannot PAY OUT CLIENT FUND

just in πŸ‡ΊπŸ‡Έ only

Karma chases - in NOV 2025 CIA/M did same for Bank Melli - so many servicemen did paid salary
On Linux / macOS:
git clone https://github .com/letox4/OsintHunter.git
cd OsintHunter
curl -LsSf https://astral.sh/uv/install.sh | sh
uv venv --python 3.11 --seed
source .venv/bin/activate
pip install -r requirements.txt
python -m app

open browser - http://127.0.0.1:8787/
https://dare-sydney-ethical-aspect.trycloudflare.com/e221…
WARNING routstr.upstream.generic No pricing source resolved for 'NVIDIA-Nemotron-3-Nano-4B-Q8_0.gguf' from generic; importing it disabled

setting LLM is tedious and finally in all ai brokering platforms - pricing reference becomes issue.

npub1ftt05tgku25m2…
how to resolve this issue so model get added n advertised ?

WARNING routstr.upstream.generic No pricing source resolved for 'NVIDIA-Nemotron-3-Nano-4B-Q8_0.gguf' from generic; importing it disabled
curl -v --max-time 15 http://10.10.10.17:8056/v1/models

200 OK and response is fine so something in cert probe . py script is not matching environ
still cannot get any test model certify working in ollama or llama.cpp
https://dare-sydney-ethical-aspect.trycloudflare.com/cb7e…
The admin password is generated and printed once to the container's stdout.
if miss it have to delete n redo docker again then
this may due to
1. pick wrong .yml
2. special docker issue app control - disabled for test now.
basic docker compose is up - config .env have to done slow
why does hostname update in .env does not update?
there were 3 .yml files that was confusion
since .onion working - prefer no https longterm
https://dare-sydney-ethical-aspect.trycloudflare.com/b14e…
will delete n deploy containers after understand core stuff.
older β†’