obvirm

268 posts

obvirm banner
obvirm

obvirm

@obvirm

Random Reposting

Los Angeles, CA Katılım Nisan 2021
270 Takip Edilen5 Takipçiler
emptyArray
emptyArray@ArrayEmpty·
I'm finally happy with my @UnslothAI unsloth/Qwen3.6-35B-A3B-MTP-GGUF, @no_stp_on_snek llama.cpp-turboquant, @_HermesAgent setup, @MajorFAFO params suggestions, and my own Rust/SQL/Hermes app. - @ASUS, 32GB RAM, 8GB VRAM 5070 laptop - MoE fully offloaded to CPU - I've fiddled around with --no-mmap and mlock. It pegs my RAM to 95% instead of 45% and after it spins up, I don't see a difference. --cache-type-k q8_0 \ --cache-type-v turbo3 \ -ot "\\.ffn_.*_exps\\..*=CPU" \ --spec-type draft-mtp \ --spec-draft-n-max 3 \ The video shows it has full hermes skills/toolset access easily, knows my system, can be creative, can print out a quick BASIC function, corrected itself when I told it that it was incorrect, and that it doesn't love me....
emptyArray tweet media
English
14
17
180
11K
Jordan Daniel Chesney
Jordan Daniel Chesney@jordandchesney·
You know that super annoying overlay texture pattern you get on AI images, particularly with Nano Banana Pro and GPT 2.0? Here's how to get rid of it. Toss your image into @magnific as a reference, select the Nano Banana Pro model, and prompt: Recreate this image exactly, detail for detail and restore fidelity, making it high resolution, and crisp, and clear. Exactly maintain the original colors. Credit to @DavidLaChanceJr for figuring this out. Hope this helps.
English
13
35
405
29K
奶昔🥤
奶昔🥤@realNyarime·
NVIDIA 赠送一年NVIDIA DSX Air资源,60vCPU+60G RAM 需使用域名邮箱注册并登录,每7天需要手动续期一次,注册链接放评论区 最大配置96C96G 500GB
奶昔🥤 tweet media
中文
158
240
3.1K
347.1K
obvirm retweetledi
Jake
Jake@JakeKAllDay·
A fun finding for the 12GB VRAM + MOE RAM Offload gang: 60 tok/s sustained with Qwen 3.6 35a3b! (MTP +Turboquant and a few other tricks) 🧵
Jake tweet media
English
14
9
39
9.8K
ISM ☀️
ISM ☀️@ism_sol·
gelo berhasil cok bikin pake Codex 👀 gua kasih link video youtube doang di tele langsung nge clip dia
ISM ☀️ tweet media
Indonesia
44
25
453
20.7K
satyaxbt
satyaxbt@satyaXBT·
Kalo lu bikin konten ori walaupun masih sepi, you've earned my respect. Coba kasih tau gue kalian bikin konten tentang apa, gue mau support kreator yang beneran niat
Indonesia
230
29
401
14.5K
obvirm
obvirm@obvirm·
@0xyunss pakek moe yang 35b-A3b bagus bang gw 4070ti dapet 37 t/s
obvirm tweet media
Indonesia
3
0
0
1.4K
yunus
yunus@0xyunss·
menurut kalian gimana self hosting local AI? jujur aja gua pengen beli dgx spark buat hosting qwen3.6-35B atau gemma4, di rumah ada 3090 si cuma 16gb doang tapi kalo dipikir-pikir mending beli credit ga si? kebutuhan gua lebih ke coding sih
yunus tweet media
Indonesia
41
6
170
17.9K
obvirm retweetledi
Jack Appleby
Jack Appleby@jappleby·
Man, I’m really worried about consumers. Not everyone’s tech savvy enough to know this ad is completely AI, and they’re not actually showing the product they’re selling. AI ads are going to really take advantage of people.
English
1.1K
3.5K
44.7K
9.5M
v
v@iavins·
👀
v tweet media
QME
47
41
875
52.2K