@NickADobos@trq212@AnthropicAI@alexalbert__ Valid take
So, from what you just said I'm assuming (I haven't read the article, yet) that the original generation of the next token is likely to no longer be via total randomness (with regard to the most probable answer by 0 temperature), but by controlled "randomness" instead?
My week in 4 frames 🥹
• Graduated from 200 level at the Ark Ministry Academy (My Church) 🥳🎉
• I was an Arduino tutor throughout this week
(I'm stressed 🫠)
• Shawn (@itzjustshiro) ended his internship, so I continue my lonely days at the office 🫠
• @yfe_embedded's workshop kicks off today 🥹
P.S. Registration closes on the 29th
All you people that can juggle multiple attention-demanding tasks at the same time are beasts
I'm juggling just 3 (sometimes 4) and I've got times I just sit and hold down the insanity... because does life really have to be so demanding
How do you guys do it?... I need help 😮💨
All you people that can juggle multiple attention-demanding tasks at the same time are beasts
I'm juggling just 3 (sometimes 4) and I've got times I just sit and hold down the insanity... because does life really have to be so demanding
How do you guys do it?... I need help 😮💨
Hi, I'm Nwokolo victor oluebubchukwu AKA NEVO
I'm a 19 y/o
I am the Senior Associate of @aurora_robots and Chief ambassador of @yfe_embedded
between i am into CAD utilizing Solidworks and Onshape.
Now an Author!!!
I recently presented my papers at NAICE and now they're published🔥
One covers a hybrid CFD-ML framework for pipeline leak detection in oil and gas, the other an adaptive ML approach to concept drift in EV battery predictive maintenance.
@Chioma__Amadi This tip would benefit young startup founders even more... I hope they see it
By reserving MEETINGS strictly for urgencies you solidify it's status as IMPORTANT. Now, in a situation where it is called, your people will know to drop everything and hop on board without question
Working on Project Impersonator, an AI project to map social media posts to personality
I ran into a hallucination problem. What I suspect:
max_new_tokens (30%)
Complex system prompt (20%)
LangChain memory (50%)
I will run a few tests to narrow it down, but what do you think?