Lasq

[Lasq @lasq88]

there's always local :)

[Lasq @lasq88]

Oops, might need a bigger plan. 😉 GLM 5.2 is an absolute powerhouse, but I miss how incredibly efficient GLM 5.1 was on the wallet—felt like I could push 4x the volume and never even blink at a rate limit. Back to Composer for now. #AI #LLM #GLM5

[Lasq @lasq88]

@MainzOnX Food in Google offices in Asia is way better than in the US

[Lasq @lasq88]

@MarcinAdamAlek1 @Piegziu W centrach dużych miastach jak Tokio tak, chociaż wszystkie zabezpieczenia działają tylko do pewnej magnitudy. Jak walnie 9+ to żaden cud inżynierii nie pomoże. Na peryferiach, w małych miasteczkach i wiochach nie ma takich zabezpieczeń i budynki normalnie się walą

[Lasq @lasq88]

@KrzysztofKakol @Piegziu Byłem akurat 1 stycznia 2024. W Tokio tylko lekkie tąpnięcie było czuć ale na zachodnim wybrzeżu niestety duże straty i sporo ofiar śmiertelnych. Duże miasta są lepiej przygotowane chociaż nie ma rozwiązań idealnych, w małych miasteczkach jest gorzej.

[Lasq @lasq88]

@sudoingX Hey @Teknium - would really love to see this merged into the main release. I'm using Composer in Hermes-Agent and it's really great but running a forked instance in a long term is not ideal. Thanks!

[Lasq @lasq88]

@GeorgeXenos2 @Teknium @NousResearch For now just .md files but the goal is to build a system to automatically analyse those and model the findings in a graph database. I also send notifications to my Discord. This is the system for now, but I will be expanding it.

[Lasq @lasq88]

Hermes Agent is great and I'm pretty impressed with it. The only issue I have it keeps blocking my simple daily cron job to scrape CTI blogs. Is there a way to turn off / loosen security checks for these cron jobs? @Teknium @NousResearch

[Lasq @lasq88]

@ethanolivertroy @Teknium @NousResearch Not much yet, as I just started experimenting with it. I will definitely give it a go for Malware Analysis

[Lasq @lasq88]

@CoinMonero @Teknium @NousResearch Yeah that's exactly what I did. It "fixed it", then documented these pitfalls in the SKILL which... caused the same issue again 😅

[Lasq @lasq88]

@Teknium @NousResearch Just scraping some Threat Intel vendors blogs on latest cyber threats. Anyway, seems the issue was in the SKILL.md, the agent tried to hardcode what not to include to trigger the security scan, and... triggering the security scan by this, lol.