Bartosz Nogalski

still figuring it out ¯\_(ツ)_/¯

[Bartosz Nogalski @B_Nogalski]

@MiaAI_lab Everything looks better than model that doesnt exist. At this point i could train some 20M model and it would be better than something that was never released 🥲

[Bartosz Nogalski @B_Nogalski]

@MiaAI_lab @niklaslenz_ai I live in Swiss and the best i can get is 5G... no fiber in my area 🫠

[Bartosz Nogalski @B_Nogalski]

@thsottiaux Just make it an opt in and let user decide what they prefer.

[Bartosz Nogalski @B_Nogalski]

@QwenDevs Something in the middle 😅

[Bartosz Nogalski @B_Nogalski]

@bot @bot you know what to do

[Bartosz Nogalski @B_Nogalski]

@sudoingX Try shipping something with so many regulations on your back 😅

[Bartosz Nogalski @B_Nogalski]

@KrishnaBuildsAI

[Bartosz Nogalski @B_Nogalski]

@choblin29 No wonder Nous team doesnt sleep and keep delivering. The real money is on the table 😅

[Bartosz Nogalski @B_Nogalski]

@0xkydo I can totally relate 😂

[Bartosz Nogalski @B_Nogalski]

@elonmusk Only issue is that the cost is out of scale 🥲

[Bartosz Nogalski @B_Nogalski]

@thsottiaux Are you consulting your declarations with your boss? Or anyone?

[Bartosz Nogalski @B_Nogalski]

@jocarrasqueira @GoogleDeepMind

[Bartosz Nogalski @B_Nogalski]

@Blackwellboy @ViC305 @Blackfrost_AI

[Bartosz Nogalski @B_Nogalski]

@HermesAgentTips Glm 5.3 via api + glm 5.3 flash local for private stuff

[Bartosz Nogalski @B_Nogalski]

@VictorTaelin Does it really matter? OS models are already good enough for 99% of daily agent work and thats where they are actually usefull and worth the hardware price

[Bartosz Nogalski @B_Nogalski]

@Priyannkaaaa Easy. Glm 5.5 flash 😅

[Bartosz Nogalski @B_Nogalski]

@plotarmordev As long as i will be able to put it in my stack along with my sparks, I dont care who will deliver the hardware :)

[Bartosz Nogalski @B_Nogalski]

@MiaAI_lab If you are not hardcore developer/ chasing beatiful frontends/ doing pentesting etc. and require an AI agent only for daily work like scripting, visualizations, organising and managing things then 2x dgx with glm 5.3 is already enough. If you are in the first group then well 🥲

[Bartosz Nogalski @B_Nogalski]

@ishuagra02 Simply set a weekly + 1 minute cron to send a message. It will start the reset timer. Set it once and dont worry about it ever again (unless async reset appears :) )

[Bartosz Nogalski @B_Nogalski]

@antigravity 😲