TechMD

2.4K posts

TechMD banner
TechMD

TechMD

@TechMDAI

🔬 MD fueled by a deep passion for Medicine & Tech. 🌐 Exploring the frontiers of VR/AR/MR & BCI. 🤖 AI enthusiast.

Katılım Haziran 2023
2.8K Takip Edilen598 Takipçiler
Sabitlenmiş Tweet
TechMD
TechMD@TechMDAI·
Built a local agentic stack where Hermes orchestrates, Codex 5.3 designs workflows, and DGX Spark runs the model team. Telegram is the front door. Hermes turns messages into tasks. Codex helps build and refine the pipelines. Local models execute natively on Spark. Current local lineup: • gemma-4-26B-A4B • qwen3.5-9b • qwen3.5-4b So the flow is: Telegram → gateway channel → Hermes → Codex-assisted workflow logic → local harness → DGX Spark inference → validated response Not just chatbot plumbing. More like building an operating system for coordinated intelligence.
TechMD tweet media
English
0
1
10
412
TechMD retweetledi
mr-r0b0t
mr-r0b0t@mr_r0b0t·
It’s official! @NVIDIAAI / MiniMax-M2.7-NVFP4 Optimized specifically for your SM120/121 DGX Spark (GB10) and RTX 6000/5090 Blackwell tensor cores! Full native FlashInfer/CUTLASS Finalizing the benchmarks and documentation now 😁
mr-r0b0t tweet media
English
10
3
59
1.4K
TechMD retweetledi
Tibo
Tibo@thsottiaux·
Some of you noticed limits drained faster in Codex, we root caused it to an optimization that we rolled back that had an impact on cache hit rates when compacting across long running sessions. We fixed this and have now reset usage limits for all accounts. Enjoy the weekend.
Kappaemme@Kappaemme1926

CODEX LIMITS ARE FIXED!

English
603
227
5.3K
224K
Ivan Fioravanti ᯅ
Ivan Fioravanti ᯅ@ivanfioravanti·
Cursor + Composer 2.5 finally solved a bug in my fasterliveportrait-mlx webui! GPT 5.5 and Opus 4.7 were trying with no luck 🤷🏻‍♂️ I think I will have to review my coding subscriptions now! Too many actors on the stage now!
Ivan Fioravanti ᯅ tweet media
English
4
3
41
2.9K
TechMD
TechMD@TechMDAI·
The f1 coverage was very horrible . So many poorly called incidents.
English
0
0
2
71
Joey
Joey@aijoey·
termius/dgxspark/hermes/vllmstudio running my inference from bed because I just worked 8hrs and I dont want to be sitting in the office. Theres really no excuses to not be building.
Joey tweet media
Joey@aijoey

got vllm studio running from my mac mini against the dgx spark today added hermes as a selectable agent runtime, wired it through an openai compatible bridge, fixed lan access, and imported my local model zoo into launchable recipes 21 models detected 12 vllm recipes 9 llama.cpp gguf recipes this is the kind of workflow i want locally mac mini as the control surface dgx spark as the inference box hermes as the operator layer vllm studio as the model dashboard no cs degree, just building the stack piece by piece with ai as the teacher

English
2
0
15
988
stevibe
stevibe@stevibe·
My setup: > Mac Studio & Mac mini as permanent code machines > MacBook Air for the road, never half-closes for a dev task > Codex handles all the coding work > Linear for issue handling > Tailscale glues it all > Screens on iOS for remote anytime > Obsidian for managing knowledge > BYOK chat: Chatwise (Mac), Flowdown (iOS) > Remarkable Paper Pro Move for note taking
English
3
2
19
1.9K
Steve💙🇨🇦
Steve💙🇨🇦@xyster·
The new home shuffle is complete. It's a downgrade to an old PCIe 4.0 Threadripper board, but the other new Epyc build is for my RTX 6000 Pros. This fits them better mind you, and PCIe 4.0 should still be fine for my needs. Cable management still needs some work...
Steve💙🇨🇦 tweet media
English
18
6
128
7.2K
Whisperer
Whisperer@cyberwhisperr·
I am coming to a dark realization, which is that NVIDIA DGX spark has the capability to run models, but we are so spoiled by the quick responses of cloud models that just feels laborious I might return this and buy RTX Pro 6000. I need speed cc @TheAhmadOsman
English
8
0
9
868
TechMD retweetledi
0xSero
0xSero@0xSero·
DeepSeek-V4-Pro & Kimi-K2.6 running in Codex app. Cheapest way to taste the frontier. Works w local models and they can do all the computer use stuff (:
0xSero tweet media
English
92
79
1.4K
186.4K
TechMD
TechMD@TechMDAI·
@thsottiaux is listening. Codex app is improving with user input
English
0
0
1
24
TonoKen3🌏文明航海士©｜とのけん3
TonoKen3🌏文明航海士©｜とのけん3@Tono_Ken3·
8x RTXPro6000と8x MacStudioUltraと両方のユーザーになって見えた地平がありますな その計算資源は間違いなくベストセラーになる 140万で買えた時代が懐かしく思えるに違いない。私は今あの日観た夢を売っている 手元に残るは1276GBのDDR5 RDIMMとRTX4060だ母さん 人生は誠に面白いじゃァないか😇
日本語
1
2
29
1.9K
TechMD
TechMD@TechMDAI·
Trying Composer in cursor. It’s been a while cursor. Cursor is where my journey began.
English
0
0
1
19
Charlie Marsh
Charlie Marsh@charliermarsh·
What would you do with unlimited tokens
English
372
8
508
79.7K
TechMD retweetledi
Google Gemma
Google Gemma@googlegemma·
We are entering a new era of on-device automation. ✨ Watch Gemma 4 E4B navigate and drive an iOS simulator directly using Argent. Local models can handle complex interactions and software navigation autonomously.
English
112
543
6.1K
457.4K

Keşfet

@NVIDIAAI @thsottiaux @ivanfioravanti @aijoey @NousResearch @Teknium @0xSero @Dimillian