beplee

242 posts

beplee

beplee

@bepleecanton

No matter your origin, great food knows no borders.

Katılım Mart 2023
158 Takip Edilen7 Takipçiler
beplee
beplee@bepleecanton·
@IBuzovskyi Thanks, i was searching how to increase turn budget😍
English
1
0
1
89
beplee
beplee@bepleecanton·
Hermes - @grok 4.3 asked for command approval a lot less than any local model I have used so far.
English
1
0
0
9
beplee
beplee@bepleecanton·
Thanks @OpenRouter ! Was putting together my own wiki using Grok 4.3 and just hit my credit limit 😂
beplee tweet media
English
0
0
0
11
beplee
beplee@bepleecanton·
Turned out --no-mmap is not a good idea on my setup (RTX 5080 + 32GB RAM), even with swap + zram. It caused the whole system to freeze hard twice when RAM maxed out. Working stable config for now ./llama-server \ -m Qwen3.6-35B-A3B-MXFP4_MOE_MTP.gguf \ -c 200000 \ -ngl 99 \ -ncmoe 30 \ -np 1 \ -ctk q8_0 \ -ctv q8_0 \ -fa 1 \ --spec-type draft-mtp \ --spec-draft-n-max 2 \ --spec-draft-p-min 0.75 \ --temp 1.0 \ --top_p 0.95 \ --top_k 20 \ --min_p 0.0 \ --presence_penalty 1.5 \ --repeat_penalty 1.0 \ --jinja \ --chat-template-file chat_template.jinja \ --cont-batching
English
0
0
0
14
beplee
beplee@bepleecanton·
@morganlinton hermes - super grok - using chrome cdp, /browser connect. Probably can do that quickly.
English
0
0
0
40
Morgan
Morgan@morganlinton·
I have a feature request for Grok Build that I think would be pretty neat. Getting a lot of feedback from people on features they'd like to see added to PasteLocal. Tried asking Grok Build to spec out a feature someone suggested, and passed in the tweet, but it couldn't access the contents of the tweet. Since I'm logged in, and subscribing to SuperHeavy, feels like it would be able to access the tweet and spec out this feature request right? This would be such a great way for people building with Grok Build to take feature requests from X and easily start building them with Grok Build.
Morgan tweet media
English
10
1
42
3K
beplee
beplee@bepleecanton·
@mr_r0b0t @NousResearch This is awesome. Save a lot of time ngl. And I just finished setup gbrain the other day, was having some issue with local embedding. Thank you so much. Gonna dig in right away.
English
2
0
1
9
mr-r0b0t
mr-r0b0t@mr_r0b0t·
My Grok usage has been exceptionally efficient in @NousResearch Hermes Agent! If you're wanting to stretch your allocation, this is a good place to start!
mr-r0b0t tweet media
English
5
0
23
1.3K
beplee
beplee@bepleecanton·
@mr_r0b0t @NousResearch I'm just starting for like 2 weeks, the whole AI journey😅 I recall I read abt it from a guy, Kapathy or Karpathy? i can't recall the name. Anyways, I'm gonna let Hermes deep dive your posts or articles to learn more about building wiki.😁
English
1
0
1
22
mr-r0b0t
mr-r0b0t@mr_r0b0t·
I certainly think so 🤩 Agent memory is a space where there are so very many solutions, it kind of becomes a preference question! I like to keep my host data separate between my devices, some people share brains across all their agents tho! It’s easy to go either way if you go with obsidian tho!
English
3
0
2
17
beplee
beplee@bepleecanton·
@sudoingX That's probably me 😂 Third clean install. Hermes kept telling me what to do instead of doing it. Old backup was the culprit. Good thing I'm a newbie, reinstalls are fast. Now using Timeshift and being careful. Today Hermes + Super Grok cut my token waste by 65%
English
1
0
3
388
Sudo su
Sudo su@sudoingX·
if you are working with agentic systems and you quietly feel like you are faking it, like everyone else got the memo and you are just barely keeping up, i want to tell you what is actually happening. it is not a talent gap. it is a setup gap. the people who look fluent are not smarter than you. they set up the boring foundation early, so the friction you fight all day, they simply do not have anymore. every hour you lose to context dying in a closed window, to a long run dropping on a disconnect, to a machine you cannot reach, that hour feels like proof you are not good enough. it is not. it is proof you are under tooled. i felt like an imposter for a long time too. the feeling did not go away when i got smarter, it went away when i set up the five and the friction stopped. you are not behind. you are one weekend of setup away from feeling like you belong.
Sudo su@sudoingX

anyone thinking about, learning, or already working with agentic systems, you should know this. the first few steps of your setup matter more than any model or framework you pick later. get them right and you never lose your flow. the foundation nobody posts about: > 1. tailscale. a private mesh network across every machine you own. laptop, desktop, rented node, all on one secure tailnet, reachable from anywhere. nothing else works well until this does. > 2. termius, over that tailnet. one SSH client that reaches every node, phone included. you are never away from your stack. > 3. tmux. persistent sessions. disconnect, close the laptop, come back, every session exactly where you left it. agentic work runs long, your terminal has to survive that. > 4. a private git repo. the one i am most glad i found. it is the memory layer across all my agents, they pull, they work, they merge back, the codebase stays alive between sessions. context that would die in a chat window lives in the repo instead. > 5. script everything from day one. ssh aliases for every node, setup scripts, the boring boilerplate automated. if you will do a thing more than twice, it is a script. everything past these five is decorative. know these cold. and the habit that ties it together: ask the AI itself. for the config, for the error, for any of it, let the agent do the lifting, then double check what it hands you. lock the five, build the habit, and you make it. skip it, anon, and you ngmi.

English
19
14
241
14.5K
NVIDIA GeForce
NVIDIA GeForce@NVIDIAGeForce·
Recruits, your first prize is here... A custom GeForce RTX 5080 Founders Edition + PC copy of the game. Comment #007FirstLightRTX to win 👇
English
27.8K
2K
14.7K
2.2M
Tony Simons
Tony Simons@tonysimons_·
Try this GPT Images 2.0 prompt to instantly level up your product photography images: Use the attached product photo as the reference subject. Turn it into a premium commercial product photography shot. Do NOT redesign or change the actual product packaging, branding, colors, or shape. Preserve the real product identity exactly. The goal is to elevate the presentation, not invent a new product. Create a realistic studio-quality scene with: - cinematic commercial lighting - soft shadows - subtle reflections - shallow depth of field - premium modern aesthetic - realistic materials and textures - professional composition Place the product in an environment that matches its brand and category naturally. Add tasteful supporting props only if they improve the composition. The image should feel like: - a high-end tech advertisement - Apple-style product photography - clean luxury ecommerce branding - magazine-quality commercial work Avoid: - fake-looking CGI - warped text - excessive clutter - random floating objects - over-stylized cyberpunk aesthetics - changing logos or typography Prioritize realism, lighting quality, composition, and premium presentation. Camera style: 85mm lens, f/2.8, soft studio lighting, ultra detailed, photorealistic, commercial photography, natural reflections, realistic shadows. Make the final image social-media-ready and visually striking.
Tony Simons tweet mediaTony Simons tweet media
English
2
0
9
571
beplee
beplee@bepleecanton·
@Layemie001 Lost me at pulling out. Never did never will.
English
0
0
0
98
LMD (Arc.)
LMD (Arc.)@Layemie001·
How to get a car out of trouble when both rear wheels are stuck without asking for help. Every driver must see this!
English
2.4K
9.7K
209.4K
7.7M
JUMPERZ
JUMPERZ@jumperz·
love seeing my discord stay in sync with Hermes Kanban..everything here was done in plain English. I just asked my coordinator to check if we have an update and the system understood the intent, routed it to the technical agent, tracked it, and posted the result.. the point is simple: instead of clicking around a Kanban dashboard, I can just type and this flow will happen: >task created in the coordinator channel >hermes reads the plain-english request >coordinator understands the intent >task gets routed to the right agent >task appears on the discord task-board >same task appears in hermes kanban >agent gets the task in their own channel >progress card updates while they work >Kanban status moves with the run >result/evidence posts in the agent channel >clean final receipt goes to results channel on discord >coordinator channel gets the update >task-board refreshes to done honestly, the reason i wired discord into hermes kanban is simple .. discord is where i actually talk to my agents and kanban is the ledger / source of truth also with the discord task-board keeps everything visible so you can scroll back and see exactly what got done and what didn't. never felt this organized. if you're running hermes / discord as an orchestration layer .. set this up asap.. super easy and efficient especially if you're away from your desktop or on your phone.
JUMPERZ tweet mediaJUMPERZ tweet mediaJUMPERZ tweet mediaJUMPERZ tweet media
English
13
10
156
44.5K
Dailymeow
Dailymeow@Dailymeoww1·
In Japan, a woman noticed that while her little daughter was playing, she would constantly end up inside the cat’s cage, trapped and crying for help. But the strange part was this: the cage was normally closed, and the woman could never figure out how her daughter was getting inside or how the door was closing afterward 🤔 When it kept happening every single day, she finally decided to place a small camera in the area where her daughter played with the cats. And when she watched the footage, she couldn’t believe her eyes… 😳 Turns out, her orange cat was opening the cage door, letting the little girl go inside, then walking back out and closing the door on her. 😂 Ah, orange cats… they really are on another level 😹❤️
Dailymeow tweet mediaDailymeow tweet mediaDailymeow tweet media
English
200
1.4K
9.4K
1.6M
beplee
beplee@bepleecanton·
My setup currently, Ubuntu 24.04 LTS, 5080, 32GB RAM, turboquant plus, Qwopus-GLM-18B-Healed-Q4_K_M.gguf from Kyle ( forgot his last name) , ctx 131k. I get around 60 t/s kv q4, turbo4. k q8 v q4 cause weird stuff that only cpu compute even tho ngl 999, turbo4 works fine no issue. I need one tool call, thanks.
English
0
0
0
56
left curve dev
left curve dev@leftcurvedev_·
You know what day it is 🔥 🔧 Setup Sunday Tell me your hardware and i’ll recommend the best models & setups you can run Did something cool on your rig this week? Found a new github repo? Drop it in the comments! Let’s help each other 🫡
English
49
4
40
4.3K
beplee
beplee@bepleecanton·
@garrytan I asked Hermes to install GBrain and everything is working great except the embedding part. How do I configure GBrain to use Ollama’s embedding model please? I tried dummy key for openai and point to localhost but it's not working.
English
1
0
1
620
Garry Tan
Garry Tan@garrytan·
GBrain v0.31.1 just shipped real MCP thin client support. So basically you can run ONE "home GBrain server" and everything else can just connect to it via MCP and it'll work pretty close to as well as running it locally. GBrain just went client-server.
Garry Tan tweet media
English
49
40
490
52.2K
Kris Kashtanova
Kris Kashtanova@icreatelife·
Prove you work with AI with just one phrase
English
1.3K
10
417
82.9K
ZOTAC
ZOTAC@ZOTAC·
Week 1 LIVE (May 4–10)! ZOTAC 20th drops 🎉 Loot: ALLOY case, RTX 5060 LP, RTX 5050 OC White. Like+Follow+Tag2 #ZOTAC20th: gleam.io/QkvMH/zotac-20… 🔥 Bonus: Share ZOTAC card = extra entry + boost chances
ZOTAC tweet media
English
322
309
602
16.2K