"developers" by @pkgsrc
Twitter accounts of pkgsrc developers
Look! Just loads of modern device/fs drivers for old NT versions. Oh, they are vibe-coded of course. "What if they don't work perfectly?!?" But does it matter? You'll have retro-fun anyway. github.com/ages2001?tab=repo…
Screenshot of the quoted page showing the first top 3 repos.

ALT Screenshot of the quoted page showing the first top 3 repos.

1
13
385
spoiler.
2
3
13
1,291
iMil 🇪🇸🦇 retweeted
Sherwood
@shcallawayOct 22
Today we're launching cve.now 🚨 Yesterday was a rough day for security. php2shell, LiteLLM, Railway... How are we supposed to stay on top of all of this? cve.now aggregates 4 vulnerability sources into one free feed and uses AI to explain each one in plain English. No login. cve.now
123
264
78
1,829
356,362
Signal galère chez vous aussi ? la journée n'est pas finie...
2
2
1,128
iMil 🇪🇸🦇 retweeted
Vlad Rico@RicoVladOct 22
Achieved RCE POC in a PHP 8.5.4 lab (but cheating a little bit on the php.ini with debug mode to leak address .. for now ) Thanks @BlWasp_ for the help CVE-2026-109618 CVE-2026-109619
1
10
1
46
6,427
My now-traditional(?) BazelCon 2026 recap is here! jmmv.dev/2026/10/bazelcon-20… . I'll warn you: it's _long_ and it talks about AI but... fear not, it's not AI-generated. Enjoy!

BazelCon 2026 recap - Julio Merino (jmmv.dev)

We are right in the middle of fall, and that means BazelCon 2026 just took place. This time around it was hosted in Amsterdam and spanned three days: one day of tutorials and an overlapping Buildbarn...

jmmv.dev
2
1
11
825
Give this project more visibility tweeps! Possibly the best personal assistant around, not overly complex, simple to install and run, works on anything, RT @afjcjsbx !
afjcjsbx
@afjcjsbxOct 21
Do you want a personal AI assistant to integrate on WeChat, Feishu, Telegram? you can take a look at PicoClaw A super lightweight open source agent with long-term memory, planned tasks, MCP tools, plugins and web interface Installable with a single command, just enter the API key of the model and use it github.com/afjcjsbx/picoclaw

GitHub - afjcjsbx/picoclaw: Ultra-lightweight, open-source, self-hosted personal AI assistant in Go...

Ultra-lightweight, open-source, self-hosted personal AI assistant in Go for RISC-V, SBCs, and embedded systems. Active PicoClaw fork with WebUI, MCP, memory, multi-agent workflows, plugins, and cha...

github.com
3
3
1
6
940
iMil 🇪🇸🦇 retweeted
afjcjsbx
@afjcjsbxOct 21
Do you want a personal AI assistant to integrate on WeChat, Feishu, Telegram? you can take a look at PicoClaw A super lightweight open source agent with long-term memory, planned tasks, MCP tools, plugins and web interface Installable with a single command, just enter the API key of the model and use it github.com/afjcjsbx/picoclaw

GitHub - afjcjsbx/picoclaw: Ultra-lightweight, open-source, self-hosted personal AI assistant in Go...

Ultra-lightweight, open-source, self-hosted personal AI assistant in Go for RISC-V, SBCs, and embedded systems. Active PicoClaw fork with WebUI, MCP, memory, multi-agent workflows, plugins, and cha...

github.com
3
2
12
56,786
The year is 3026 and I'm working on various projets in // with local agents. For those interested, harness == maki, model on the left is QFN iq3_xxs (RX 7600 8GB / 64GB DDR4), model on the right is QFN q5_k_xl (RTX 5080 + RTX 3090 / 40GB total VRAM + 96 DDR4) #AI #LocalLLM
1
2
3
694
Debian? is that you?
solst/ICE of Astarte
@IceSolstOct 20
I don’t understand? How
4
6
28
3,723
iMil 🇪🇸🦇 retweeted
JVMag🇨🇭
@JVMag_chOct 20
Si vous vous ennuyez au travail, sachez que vous pouvez désormais jouer à Need for Speed Underground 2 depuis votre navigateur 😍 il y a même du multijoueur ! Lien : nfsu2.xyz/
75
513
52
4,446
365,685
Hear me out: @support_huihui abliterated Qwen Flash Next 3.8 Swift iq2_xs on Strata running on a *8GB RX 7600*, 64GB DDR4 and a cheap NVMe drive at FREAKING 30-40 TOKENS/S @coldniko is the goat.
1
3
8
812
Now trying Gyro-M
116
Some neat & useful(!) swag I brought back from BazelCon 😎 I have something even cooler to show but that'll have to wait until I write an article about it. h/t @engflow
1
7
250
Replying to @_adriend_
@_adriend_ toi qui est déjà versé dans le monde des LLMs, tu as essayé Strata github.com/Niko1221/Strata ? Les résultats sont incroyables...

GitHub - Niko1221/Strata: Qwen3.8-Flash-Next on any consumer hardware: one-click install for...

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input. - Niko1221/Strata

github.com
1
2
15
1,402
more replies
30 tok/s en q2_0 avec une RX 7600 / 8GB et 64GB de DDR4
1
62
Probablement un peu plus en iq2_xxs mais j'ai pas encore eu le temps de test
1
64
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 19
Strata 0.2.0 is now live in beta! It includes its own in-app harness and terminal-serving interface. github.com/Niko1221/Strata A lot of performance improvements and support for low-end devices 👇
141
239
73
1,295
179,921
iMil 🇪🇸🦇 retweeted
afjcjsbx
@afjcjsbxOct 18
PicoClaw v0.5.1 just dropped! ⚡️ ​Featuring OpenCode Go provider support, full Plugin CLI, smart artifact spilling for oversized outputs, and reduced memory usage on embedded boards. ​Grab the release: github.com/afjcjsbx/picoclaw ​#OpenSource #AI #Golang #RISCV

GitHub - afjcjsbx/picoclaw: Ultra-lightweight, open-source, self-hosted personal AI assistant in Go...

Ultra-lightweight, open-source, self-hosted personal AI assistant in Go for RISC-V, SBCs, and embedded systems. Active PicoClaw fork with WebUI, MCP, memory, multi-agent workflows, plugins, and cha...

github.com
1
1
6
221
Real iOS 16.5 booting on Android hardware, Not a theme, Not an emulator. HackiOS shows a Poco X5 Pro/ Note 12 Turbo booting into what looks like iOS 16.5 (Settings About as iPhone X, home screen, widgets). They say custom UEFI + XNU kernel not a skin. But no public source code, no independent verification, and no installable build. They claim custom UEFI + XNU adaptation and say it’ll be free and open.
Md Ismail Šojal 🕷️
@0x0SojalSecOct 16
WTf, Running real iOS on Android hardware. HackiOS is attempting a full code rewrite to run actual iOS on Android hardware. They’re in the final polishing stage. If it works, the project will be completely free and open no paywalls, no premium tiers.
9
49
544
39,690
La semaine dernière on a fait connaissance avec #FreeBSD. Alors demain, ça va devenir un aquatique et piquant. Il est temps qu'on parle tu gros posson tout zaune ! #OpenBSD install / packages / ports / build... Et comme toujours, ça se passe ici à 8h: twitch.tv/lefinnois_

Lefinnois_ - Twitch

Bienvenue sur ma chaîne.Je suis Denis Bodor, alias Lefinnois (oui, c'est une référence à Neuromancer de W. Gibson).Ici je parle tech, dev, hardware et bas niveau, sur mon temps libre.Côté pro, je...

twitch.tv
2
8
12
845
Writing articles in Markdown comes with friction. One thing that has always irritated me is inserting cross-links to other articles. I don't know elisp and would have never automated this... but AI does so... why not ask it? 5 minutes, and done. "SPC #" for my own blog helpers!
7
499
i don't get the moe obsession, every model release this week is moe the last good dense model that fits on a decent gpu was qwen 3.8 27b in august, before that gemma 4 31b in april. that 20 to 30b dense class is where small gpus live, and it's gone quiet and on small gpus dense is the one that pays off, give us dense models that run well on small gpus.
51
8
1
178
19,652
Maybe try QFN for real? It literally destroys 27B on every task. It runs on a 8GB GPU w/ 64GB DDR{3,4,5}
1
1
301
...with Strata of course :)
1
1
137
Qwen Flash Next UD-IQ4_XS is good, but note the level of detail UD-Q4_K_XL brings in this small X-Wing simulator, created in 10 minutes with Strata * xs imil.net/stuff/xwing.html * k_xl imil.net/stuff/xwing-q4_k_xl…
1
570
A principle that’s hard for me to not apply, for better or worse: Faced with a problem, don’t just work to solve it. Instead, figure out what would have made it so easy that someone would have already solved it by now. Make it so. Then be the lucky someone.
1
6
581
cc @coldniko
🤖 Made with AI
2
2
16
1,856
Same rig, unsloth-UD-Q4_K_XL
1
124
I've been very quiet around here for unclear reasons, but with BazelCon 2026 behind our backs... I need to write a couple of articles so let's pick this up again!
2
9
470
Por ahí vamos otra vez eh @CorreosAtiende ?
2
424
Replying to @Rapid_API
@Rapid_API @nokia One month since your September 15 reply, still no response or provider payout despite several follow-ups. Could you please provide an update?
91
iMil 🇪🇸🦇 retweeted
VecSzn
@Vec_SznOct 14
Replying to @iMilnb
It’s back now!
1
1
1
144
.@Vec_Szn did you take down the PolyStrata repo? 404 this morning :(
1
3
916
more replies
Oh ok, stay strong! I'm working on MiniCPM-V-4.7 support in PolyStrata ;)
2
1
227
While at it I merged latest Strata 0.1.43.1
109
iMil 🇪🇸🦇 retweeted
Jun Song
@jun_songOct 12
Running AI locally is not about cutting API costs. It is about owning your digital sovereignty again.
44
31
7
474
13,905
I've just PR'd AMD GPU support to @Vec_Szn's fork of Strata, unleashing 60 tok/s on Qwen 3.6 35B A3B with an RX 7600 / 8GB! github.com/VecSzn/PolyStrata…

poly: run strata-poly on AMD cards through the HIP backend by iMilnb · Pull Request #11 · VecSzn/...

What strata-poly (the Qwen3.6 / Ornith / Gemma 4 / Tiel engine) was CUDA-only; this builds and runs it on AMD GPUs through the HIP backend the main engine already uses, and lets setup.py install it...

github.com
1
9
573
Hopla! Le replay du stream de dimanche, "Découverte de #FreeBSD" est en ligne sur Youtube -> youtube.com/watch?v=yhq4ga5t…

Stream Twitch du 11/10/2026. Découvrons FreeBSD: installation, commandes, paquets, ports, etc.

On installe FreeBSD 15.1 dans une VM Bhyve instanciée sur un FreeBS...

youtube.com
6
11
485
Current: [strata] done: 6719 tokens in 66 s (102.4 tok/s) (disconnect, cancel=True), expert cache 97.5% hit (+0.4% of the routed experts over PCIe) The model is qwen3.8-flash-next-unsloth-ud-iq4_xs on 1xRTX5080 + 1xRTX3090 + 64GB DDR4 insane. kudos to @coldniko
1
13
768
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 1
People still dont fully grasp Strata You can now buy old cheap DDR3 machine with 128GB or 64GB RAM connect 8GB+ GPU and run Qwen3.8-Flash-Next at 70tps+ Remember guys: More cores/threads in CPU - faster it will go :) github.com/Niko1221/Strata

GitHub - Niko1221/Strata: Qwen3.8-Flash-Next on any consumer hardware: one-click install for...

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input. - Niko1221/Strata

github.com
194
272
91
2,882
229,565
smolClaw is now updated with @afjcjsbx's picoclaw fork release 0.5.0! smolbsd.org/smolclaw/ Great job Mauro!
1
4
3
596
iMil 🇪🇸🦇 retweeted
afjcjsbx
@afjcjsbxOct 11
PicoClaw v0.5.0 is here! ⚡️ ​Huge update with Agent Plugins 1.0, MCP OAuth, Voice Mode, Image Generation, WebUI conversation forks, and a new low-power profile for embedded hardware! ​Full changelog & release: github.com/afjcjsbx/picoclaw ​#OpenSource #AI #Golang #RISCV

GitHub - afjcjsbx/picoclaw: Ultra-lightweight, open-source, self-hosted personal AI assistant in Go...

Ultra-lightweight, open-source, self-hosted personal AI assistant in Go for RISC-V, SBCs, and embedded systems. Active PicoClaw fork with WebUI, MCP, memory, multi-agent workflows, plugins, and cha...

github.com
1
2
3
264
1. reduce source tree to strictly needed files 2. remove unneeded code paths 3. while at it, make it faster
3
1
18
1,405
Local inference BTW, Qwen 3.8 Flash Next iq4_xs on Strata github.com/Niko1221/Strata, all thanks to @coldniko

GitHub - Niko1221/Strata: Qwen3.8-Flash-Next on any consumer hardware: one-click install for...

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input. - Niko1221/Strata

github.com
1
7
248
iMil 🇪🇸🦇 retweeted
田都 アリエル@maruru_289Oct 11
Replying to @iMilnb
- GPU Kernels and Type-Specific Computations: Completely rewritten in SYCL (cannot be merged). - Multi-GPU Workload Distribution: I also plan to implement distributed execution across multiple computers over a network in the future (cannot be merged).
2
1
1
345
You're not ready for supasmolBSD
1
12
586
🚨🇫🇷 La taxe GAFAM passe de 3% à 15%. L'Assemblée présente ça comme une "taxation des géants américains" qui ne touchera pas les français. Sauf que... C'est faux. 😞 Google, Meta, Amazon n'ont aucune raison d'absorber une hausse de taxe de 12% sur leurs marges. → Ils vont la répercuter sur les annonceurs qui achètent de la publicité Google ou Meta (PME françaises, artisans, e-commerçants). → Sur les vendeurs Amazon qui paient leurs commissions. → Sur les abonnés aux services premium. Un impôt sur une entreprise dont le pricing est libre de s'ajuster est toujours, à terme, un impôt sur ses clients.
BFM Business
@bfmbusinessOct 9
Entre 3,5 et 4 milliards d'euros de recettes : les députés votent en commission pour porter la taxe Gafam de 3% à 15% l.bfmtv.com/li5g

Entre 3,5 et 4 milliards d'euros de recettes: les députés votent en commission pour porter la taxe...

Les députés ont adopté en commission des finances un amendement visant à porter de 3% à 15% la taxe sur les géants du numérique. Une mesure qui rapporterait 4 milliards d'euros de recettes, mais dont...

bfmtv.com
48
126
10
705
43,775
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 10
Was always a big fan of @gospaceport … now I’m seeing my project in his video 🥲 youtu.be/x7s6Mf4xF9A?is=dIDB…

Running Qwen 3.8 Flash Next on a $1000 Local AI Build with Strata

We put Strata to the test on a $1,000 dual RTX 3060 build to see ho...

youtube.com
14
28
1
210
11,679
Incredible. On an RX 7600 / 8GB + 32GB DDR4 ROCm 10.1.0, qwen3.6 35b a3b q4_xs with llama.cpp is *slower* (~15 t/s) than qwen 3.8 QFN iq1_m with Strata (20+ t/s) cc @coldniko
1
1
10
807
iMil 🇪🇸🦇 retweeted
afjcjsbx
@afjcjsbxOct 9
People buying Mac Minis to run OpenClaw while the #PicoClaw gateway takes just ~38 MB of RAM (with only ~18 MB of private memory!). ⚡️ And this isn't even a clean install, it's driving my daily setup with 6 MCP servers and 1 plugin. Peak efficiency! 🔥 #OpenClaw #AI #Assistant
2
1
2
153
My bank account is not friend with @coldniko
306
iMil 🇪🇸🦇 retweeted
afjcjsbx
@afjcjsbxOct 8
Since PicoClaw by @SipeedLab seems unmaintained, I’ve forked it to keep it alive. I’ve already shipped several new features! ​We’re looking for users (feedback/bug reports) and contributors to join in. ​github.com/afjcjsbx/picoclaw… ​#OpenSource #AIAssistant #AI

GitHub - afjcjsbx/picoclaw: Ultra-lightweight, open-source, self-hosted personal AI assistant in Go...

Ultra-lightweight, open-source, self-hosted personal AI assistant in Go for RISC-V, SBCs, and embedded systems. Active PicoClaw fork with WebUI, MCP, memory, multi-agent workflows, plugins, and cha...

github.com
1
2
4
239
Rejoice picoclaw users: despite @SipeedLab silence, @afjcjsbx is now maintaining an active fork of the cleanest claw around, 0.4.0 available at github.com/afjcjsbx/picoclaw
afjcjsbx
@afjcjsbxOct 7
Replying to @iMilnb
Hey, thanks so much! Really happy to hear that people appreciate PicoClaw being carried forward. Can't wait to see your PRs on the fork whenever you're ready! 🚀
2
3
5
615
Hey @afjcjsbx ! Thanks for taking over picoclaw <3 I've seen your messages, just hadn't the time to re-open the PR on your fork but I'll certainly will, and possibly another one about conversations mixing with previous ones.
1
409
1
1
8
557
Create an animated SVG of a colorful parrot juggling with bottles of gin imil.net/stuff/pinpin_ud_iq4… ~55 tok/s on: RTX5080 + RTX3090, 64GB DDR4 Strata magic cc @coldniko
6
575
iMil 🇪🇸🦇 retweeted
Peasant Smith
@PeasantSmithOct 5
I know Niko's been busy, but I'm kind of impatient This is one day of me with Claude Pro (cheapest plan) access😂 🔥GLM 5.3 Flash ⚖️93.1GB model file (the smallest, for now) 👍About 20t/s decode ⚒️Fully custom version of Strata 😥No MTP (yet) 🐌Xeon E5 2690 V4 🐌32GB ddr4 quad channel 🐌NVME 3.0 ⭐2x Tesla V100 PCIe 32GB I have redated part of the dashboard numbers because it is all new and I don't want to spoil anything until that is ready. I left the experts location open though, in case anyone thinks that is not the model I claim :) I enjoyed working on it, so I plan to carry on with optimisations and eventually release it all, stay tuned! More to come⏳
12
8
3
86
13,938
Hopla ! Le replay du stream "plus avant avec le LLM local" est en ligne sur Youtube (attention, l'intro peut contenir des râleries de vieux con réac) : youtube.com/watch?v=XaK5d-Ca…

Stream Twitch du 04/10/2026. IA locale à peu de frais, suite : llama.cpp, harness, agent, MCP...

Suite de la configuration de la machine d'inférence à moins de 600...

youtube.com
2
6
16
462
wtf
huangserva
@servasyy_aiOct 5
让我震惊的是这个排行榜！ Strata + Qwen3.8-Flash-Next测试出来的效果太好了！ 让我突然觉得我拥有了本地的顶级大模型！ 速度还爆快，还要什么API？
2
364
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 5
You can also donate via: buymeacoffee.com/strataengin… Funds will take two weeks though :(

Niko

Developing Strata engine for Qwen3.8 and Qwen4 fast inference on consumer hardware.

buymeacoffee.com
2
5
21
4,412
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 5
Please share this to support Strata development 👇 Looking for sponsors or rental of at least: - PC with AMD GPU + 64GB RAM - PC with Intel GPU + 64GB RAM - PC with Dual GPU setup + 32GB RAM Location: Slovenia (Europe) I would like to work faster as we’re only 30% in. Please message me Thank you Niko
35
94
8
412
28,158
TLD Zone File Statistics (Glue records, Name servers, DNSSEC, and Domain counts) for October 2026 are online statdns.com
1
4
5
368
Huihui-Qwen3.8-Flash-NEXT + Strata グッジョブ huggingface.co/alesha-pro/Qw…

alesha-pro/Qwen3.8-Flash-Next-abliterated-GSQ-RCO-Strata-GGUF · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co
10
46
1
476
36,567
omw @coldniko 85 tok/s w/ 0.1.39 vs 75 tok/s w/ 0.1.38 BEAAAST (RTX 5080 + RTX 3090 + 64 DDR4)
3
360
iMil 🇪🇸🦇 retweeted
SDF
@sdf_pubnixOct 4
Booting up Version 7 UNIX on Misspiggy. We’re open 10-4 today! icm.museum #vintagecomputer #retrocomputer #unix #pdp11
7
30
2
188
4,316
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 4
- Output t/s: 75.8 -> 80.0 (+8% aggregate) - Prompt t/s: +26% - Parallel requests (4 max) 👀 - Old CPU support: Sandy / Ivy Bridge, Xeon E5 v1/v2, Bulldozer, Nehalem / Westmere - Old GPU support: P40, P100, GTX 10, V100, Titan V, Instinct MI50 / MI60, Radeon VII - Intel Arc support 👇
41
40
24
448
72,289
iMil 🇪🇸🦇 retweeted
Zed
@ZedLLMOct 3
Strata is great! I’m particularly interested in improving multi-GPU performance. I found that the helper cache stays static while the main GPU adapts, allowing duplicate experts and more CPU fallback. A small optimization delivered +62.5% mixed-task and +129.2% code decode speed on dual RTX 4090s. Hope this can make it upstream! @coldniko
3
2
60
3,537
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 4
+20% boost day? 👀
29
11
3
224
44,573
we live in 3026
2
14
1,248
abliterated weights for Strata! huggingface.co/SC117/Qwen3.8…

SC117/Qwen3.8-Flash-Next-GSQ-RCO-abliterated-GGUF · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co
1
378
This is #Qwen 3.8 Flash Next IQ3_S running on an RTX 5080 and an RTX 3090 (40GB total VRAM) + 64GB DDR4 with Strata. Absolutely insane. Much love @coldniko
4
3
26
1,445
iMil 🇪🇸🦇 retweeted
Niko Veit
@coldnikoOct 1
- Prompt processing on multi gpu +20% - AMD performance increased by 12% - Unsloth UD-Q4_K_XL experimental fix - OrcaRouter IQ3_XXS added - RTX 20 series fixes - Many more improvements github.com/Niko1221/Strata/r…

Release Strata v0.1.32 · Niko1221/Strata

Multi-GPU short prompts fixed, AMD decode +12%, Unsloth's Q4 model in setup with 2.8x faster prompts, and 30+ community changes - with the same answers. Multi-GPU: short prompts faster again (#...

github.com
28
32
11
340
22,247
.@coldniko you absolute legend
1
360
iMil 🇪🇸🦇 retweeted
🥭
@MangoSweet78Feb 4
Fucking killed them Lmao.
5
13
3
415
126,910
Qwen3.8-Flash-Next first 32K tokens are 4%-10% faster, LOW RAM mode added (for low hardware). New update here github.com/Niko1221/Strata/r…

Release Strata v0.1.26 · Niko1221/Strata

Prompts up to 32K read 4-10% faster: the draft layer's prompt pass runs in batches. Before answering, Strata runs its draft layer over the prompt, so the drafter (MTP speculative decoding) know...

github.com
21
19
3
179
9,853
Is multi/different GPUs possible? (1xRTX 5080 + 1xRTX 3090)
1
1
155
Ok yes since 0.1.23 :)
1
76
iMil 🇪🇸🦇 retweeted
Hugo Duprez
@HugoDuprezSep 28
Destroy any website with a stickman. Have fun: destroy.spritefusion.com
1,035
7,085
1,076
67,691
5,325,689
Replying to @Iberia
@Iberia buenas! veo que se puede llevar una maleta de 10kg en cabina con billete economy, es verdad? Es un vuelo Valencia -> Ibiza
2
544
iMil 🇪🇸🦇 retweeted
CloudGoat@croudgoatSep 22
Max For AI
@MaxForAISep 22
🚨Qwen4家族首次曝光！！ 刚刚，在2026年云栖大会的开幕式上，新任@Alibaba_Qwen LLM负责人刘大一恒官宣了即将到来的Qwen4家族！ 包含Qwen4-Max Qwen4-Flash&Qwen4-Plus 还有Qwen4-27B！！！ 未来Qwen会训5-10T的模型
18
62
9
1,327
59,267
DO NOT TRASH YOUR OLD GPUS! (or give them to me) This machine is composed of a 2010-era motherboard (Gigabyte Z77-DS3H), a 2010-ish i5-2310 CPU, 16GB of DDR3 (3!!), an NVidia GTX 970 AND an NVidia GTX 1050Ti for a whooping total of 8GB VRAM. Useless? #AI #GPU #llama
iMil 🇪🇸🦇@iMilnbSep 22
Replying to @iMilnb
llama-server running huggingface.co/empero-ai/Qwe…
1
3
17
1,282
smoljose is alive
2
9
950
more replies
now we're talking
1
139
llama-server running huggingface.co/empero-ai/Qwe…
1
1
1,564
Qwen 4 announced, and YES, there WILL be a Qwen 4 27B !! Thank you @Alibaba_Qwen <3 reddit.com/r/LocalLLaMA/comm…

From the LocalLLaMA community on Reddit

Explore this post and more from the LocalLLaMA community

reddit.com
2
16
760
Load more