@DreamMixGames People wining about using AI as a search/let alone a tool, it's quite sad for you fully releasing a game, you put work and countless hours into something that AI couldn't. just because you had assistance from something that can search the internet faster then a human can. <3
@sudoingX any suggestions on what I could use with 16gb to get similar output, I'm aware the quality won't be the same, but I'm I'm trying to get a good model within my 16gb limit. rx 9070 non xt for the GPU. originally used Qwen 3.6 35B-A3B, but I was getting issues with hermes.
@LottoLabs bro I typed that shit in randomly before your tweet and noticed it live and was like damn it's a ghost town, I was wondering why lmfao. I was the only one there for a minute.
@loktar00 you'd be better off not letting it throttle after pasting. hitting 5ghz doesn't give as much as a benefit as most people lead on. I have mine locked to 4.5 max and it does better depending on the task. with cooler temps on Air.
@xNeeyah@TGravyz Had a group of friends 6 years going, 2 years ago, 4 guys and 2 girls. All gamers all would hang about in VC. One day one of the guys hit on one of the girls. The friend group shrinks and drama happened. Since that day falling out slowly every year. None of them stand left. o7
@TheAhmadOsman Na fr Ollama doesn't give me nearly the same token output, LM Studio has be running 3.6 at 148TPS on my single rx 9070. with nearly full precision. Q8. it's genuinely impressive. (Vulkan Runtime FTW)
@LottoLabs@stevibe@makulas1913 They just ainβt configuring it to its full potential prolly. I only got 16 gbs VRAM and fit this hoe in with 14.5 with nearly full precision π
@EliLikesRice Anxiety has a weird way of fucking with our heads, I doubt you've wronged people eli, I've watched you for years and have seen nothing but a thoughtful and funny guy. I'm sure they're aren't things you've done wrong. just remember everyone is human, and we're all in it together<3