Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

2-bit Qwen3.6-27B GGUF made 26 tool calls, triaged 15 GitHub issues and fixed, tested + reproed our repo’s 3 latest issues. 🔥 Try this locally in Unsloth Studio with just 12GB RAM. Studio also has a new look! GitHub:

UnslothAI's profile picture

Unsloth AI

87,262 subscribers

78,505 Aufrufe • vor 4 Monaten •via X (Twitter)

BildungNachrichten & PolitikWissenschaft & Technologie
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

2-bit Gemma 4 12B GGUF, only 4.66 GB on disk, managed to cite 15 sites from a single prompt. Try this locally on >6GB RAM via Unsloth Studio. GitHub:
0:48

2-bit Gemma 4 12B GGUF, only 4.66 GB on disk, managed to cite 15 sites from a single prompt. Try this locally on >6GB RAM via Unsloth Studio. GitHub:

Unsloth AI

145,211 Aufrufe • vor 3 Monaten

4-bit Qwen3.6 MTP GGUF managed to search 70+ sites from a single prompt. Try this locally on 20GB RAM via Unsloth Studio. Unsloth now supports auto MTP + speculative decoding & auto-selects the best MTP settings for your device (Mac, CPU, GPU). GitHub:
0:28

4-bit Qwen3.6 MTP GGUF managed to search 70+ sites from a single prompt. Try this locally on 20GB RAM via Unsloth Studio. Unsloth now supports auto MTP + speculative decoding & auto-selects the best MTP settings for your device (Mac, CPU, GPU). GitHub:

Unsloth AI

72,425 Aufrufe • vor 3 Monaten

2-bit Muse Glimmer GGUF managed to call 100+ tools on just 14GB RAM. 🔥 Muse Glimmer did a complete repo bug hunt for 5 mins nonstop with: evidence, repro, fix, tests and a PR writeup. Run and train it in Unsloth. GitHub repo:
1:58

2-bit Muse Glimmer GGUF managed to call 100+ tools on just 14GB RAM. 🔥 Muse Glimmer did a complete repo bug hunt for 5 mins nonstop with: evidence, repro, fix, tests and a PR writeup. Run and train it in Unsloth. GitHub repo:

Unsloth AI

122,252 Aufrufe • vor 23 Tagen

Introducing Unsloth Studio ✨ A new open-source web UI to train and run LLMs. • Run models locally on Mac, Windows, Linux • Train 500+ models 2x faster with 70% less VRAM • Supports GGUF, vision, audio, embedding models • Auto-create datasets from PDF, CSV, DOCX • Self-healing tool calling and code execution • Compare models side by side + export to GGUF GitHub: Blog and Guide: Available now on Hugging Face, NVIDIA, Docker and Colab.
0:45

Introducing Unsloth Studio ✨ A new open-source web UI to train and run LLMs. • Run models locally on Mac, Windows, Linux • Train 500+ models 2x faster with 70% less VRAM • Supports GGUF, vision, audio, embedding models • Auto-create datasets from PDF, CSV, DOCX • Self-healing tool calling and code execution • Compare models side by side + export to GGUF GitHub: Blog and Guide: Available now on Hugging Face, NVIDIA, Docker and Colab.

Unsloth AI

1,740,380 Aufrufe • vor 5 Monaten

DeepSeek has just released a powerful new 8B model 🔥 This means you can run it locally and for free on a laptop with zero limitations. You just need LM Studio and 4Gb of RAM. The few quick steps below:
0:24

DeepSeek has just released a powerful new 8B model 🔥 This means you can run it locally and for free on a laptop with zero limitations. You just need LM Studio and 4Gb of RAM. The few quick steps below:

Paul Couvert

370,822 Aufrufe • vor 1 Jahr

🚨 THIS IS INSANE A developer has released a free AI voice studio capable of cloning almost any voice in just a few seconds. It's called Voicebox. And it has already surpassed 50,500 stars on GitHub. Here's how it works: 1.- Upload a short audio sample and clone a voice in seconds. 2.- Generate speech in 23 different languages using multiple TTS engines. 3.- Run everything locally for greater privacy, and you can integrate it into your own applications. It also includes voice dictation, support for AI agents, and MCP integration. All in a single open-source tool. Save this tool. I've left the link in the next comment 👇
0:22

🚨 THIS IS INSANE A developer has released a free AI voice studio capable of cloning almost any voice in just a few seconds. It's called Voicebox. And it has already surpassed 50,500 stars on GitHub. Here's how it works: 1.- Upload a short audio sample and clone a voice in seconds. 2.- Generate speech in 23 different languages using multiple TTS engines. 3.- Run everything locally for greater privacy, and you can integrate it into your own applications. It also includes voice dictation, support for AI agents, and MCP integration. All in a single open-source tool. Save this tool. I've left the link in the next comment 👇

Nova

25,881 Aufrufe • vor 17 Tagen

3. zhao lusi studio • the studio is constantly changing and hiring new staff (many new & younger staff have joined recently) • she acknowledges that the studio also has issues - sometimes they choose not to release content when the fans are unhappy because they are afraid they’ll get backlash or make the situation worse (lusi also feels ??? by this too bcs how can u not post just bcs u are scared of blacklash LOL, she will speak with them) • the studio handles issues one by one, if one aspect is not done well, they need to resolve that matter before moving to the next • they’ve actually filmed lots of content and materials but it’s still in the inventory bcs of point 2 😭😭 • they’ve filmed an asmr video in korea but not released yet~
0:41

3. zhao lusi studio • the studio is constantly changing and hiring new staff (many new & younger staff have joined recently) • she acknowledges that the studio also has issues - sometimes they choose not to release content when the fans are unhappy because they are afraid they’ll get backlash or make the situation worse (lusi also feels ??? by this too bcs how can u not post just bcs u are scared of blacklash LOL, she will speak with them) • the studio handles issues one by one, if one aspect is not done well, they need to resolve that matter before moving to the next • they’ve actually filmed lots of content and materials but it’s still in the inventory bcs of point 2 😭😭 • they’ve filmed an asmr video in korea but not released yet~

🐰

13,095 Aufrufe • vor 2 Monaten

Looks like you're finally awake and just in time for the latest season of MARVEL SNAP: Planet Hulk! With this fierce season you can look forward to a ton of new cards, locations, and variants! Don't miss our new Pandart Studio variants...Shop Takeover anyone?
4:32

Looks like you're finally awake and just in time for the latest season of MARVEL SNAP: Planet Hulk! With this fierce season you can look forward to a ton of new cards, locations, and variants! Don't miss our new Pandart Studio variants...Shop Takeover anyone?

MARVEL SNAP

5,585,982 Aufrufe • vor 2 Jahren

I wanted to learn a little bit more about physics. Now I have 8 PhDs. This really IS wild, Gemini 3.7 Flash wasn't just a tiny little spec bump. They snuck in a huge release with this. I am definitely impressed! Google AI Studio went to town and built tested, and shipped this educational game that is a masterclass in, well ... lots. Too much to even pack into this demo too! Btw when I say built and shipped, Google Antigravity really shipped it. It built, tested, recorded the demo GIFs, screenshots, and sent it up to GitHub for me!
2:11

I wanted to learn a little bit more about physics. Now I have 8 PhDs. This really IS wild, Gemini 3.7 Flash wasn't just a tiny little spec bump. They snuck in a huge release with this. I am definitely impressed! Google AI Studio went to town and built tested, and shipped this educational game that is a masterclass in, well ... lots. Too much to even pack into this demo too! Btw when I say built and shipped, Google Antigravity really shipped it. It built, tested, recorded the demo GIFs, screenshots, and sent it up to GitHub for me!

⟁ndrew V

133,788 Aufrufe • vor 20 Tagen

BSCN is the Top Staked App on Pi App Studio Just hours after unveiling the 'BSCN Pioneer Brief' within the $PI Pi Network's Pi Browser, we have hit a new milestone. BSCN is now the leading application with the most boost from Pioneers and Businesses on the Pi App Studio. As of writing, over 740K PI has been staked, boosting our ranking and service in the Pi Ecosystem. We are also featured on the Pi Browser Homepage as the "Top Staked App Studio" Application. Thanks to all Pioneers for this remarkable feat!
0:30

BSCN is the Top Staked App on Pi App Studio Just hours after unveiling the 'BSCN Pioneer Brief' within the $PI Pi Network's Pi Browser, we have hit a new milestone. BSCN is now the leading application with the most boost from Pioneers and Businesses on the Pi App Studio. As of writing, over 740K PI has been staked, boosting our ranking and service in the Pi Ecosystem. We are also featured on the Pi Browser Homepage as the "Top Staked App Studio" Application. Thanks to all Pioneers for this remarkable feat!

BSCN

65,659 Aufrufe • vor 2 Monaten

In light of Jason Schreier's comments regarding Sony being unhappy with Naughty Dog, this made me think back to my convo with Benson Russell (Designer on TLOU & Uncharted 1-4) It's more than likely that Naughty Dog struggled to create the "it factor" for Intergalactic and the internal crunch was a way to try and find it but also...... This is their most expensive project by a wide margin, and is crazy ambitious but Naughty Dog no longer has a flat structure and has a lot of bureaucracy there now due to the studio ballooning in size and is now plagued by the same issues that studios like Bethesda have. If this game doesn't knock it out of the park, I bet there will be layoffs. Naughty Dog is becoming an insanely expensive studio for Sony to run due to salary costs in L.A.
1:07

In light of Jason Schreier's comments regarding Sony being unhappy with Naughty Dog, this made me think back to my convo with Benson Russell (Designer on TLOU & Uncharted 1-4) It's more than likely that Naughty Dog struggled to create the "it factor" for Intergalactic and the internal crunch was a way to try and find it but also...... This is their most expensive project by a wide margin, and is crazy ambitious but Naughty Dog no longer has a flat structure and has a lot of bureaucracy there now due to the studio ballooning in size and is now plagued by the same issues that studios like Bethesda have. If this game doesn't knock it out of the park, I bet there will be layoffs. Naughty Dog is becoming an insanely expensive studio for Sony to run due to salary costs in L.A.

Reece “Kiwi Talkz” Reilly

66,090 Aufrufe • vor 3 Monaten

APPLE JUST SHOWED HOW TO RUN 10 AI AGENTS LOCALLY ON MAC - NO CLOUD, NO API KEYS, ZERO COSTS 00:10 Apple engineer says: "your data stays on your device, AI available anywhere at any time, zero usage costs" agent reads your code, checks GitHub, finds what needs attention and writes a report - all on your Mac, nothing goes to the internet 10 agents work simultaneously - one writes code, another tests, third fixes bugs - parallel with no queue built a full iPad app from scratch in 2 minutes, fixed its own mistakes and compiled without issues takes 5 minutes to set up - and you never pay again for a bot that runs 24/7
13:29

APPLE JUST SHOWED HOW TO RUN 10 AI AGENTS LOCALLY ON MAC - NO CLOUD, NO API KEYS, ZERO COSTS 00:10 Apple engineer says: "your data stays on your device, AI available anywhere at any time, zero usage costs" agent reads your code, checks GitHub, finds what needs attention and writes a report - all on your Mac, nothing goes to the internet 10 agents work simultaneously - one writes code, another tests, third fixes bugs - parallel with no queue built a full iPad app from scratch in 2 minutes, fixed its own mistakes and compiled without issues takes 5 minutes to set up - and you never pay again for a bot that runs 24/7

Noisy

525,385 Aufrufe • vor 2 Monaten

Just a few months after releasing Scoot, which we feel redefined motion generation quality, our team has made substantial improvements in speed, quality, subtlety, and instruction following. This comes in concert with our newly scaled data collection and labeling efforts, compute increase, substantial architectural improvements, and more stable / scalable training dynamics. Swing is our new flagship movement model. It's live today in Cartwheel studio with infinite batch generation coming soon to our API. Treating Movement as a fundamentally new data type has really unlocked our ability to rapidly make improvements. As it turns out there are far more use cases than you might imagine from entertainment to robotics.
0:26

Just a few months after releasing Scoot, which we feel redefined motion generation quality, our team has made substantial improvements in speed, quality, subtlety, and instruction following. This comes in concert with our newly scaled data collection and labeling efforts, compute increase, substantial architectural improvements, and more stable / scalable training dynamics. Swing is our new flagship movement model. It's live today in Cartwheel studio with infinite batch generation coming soon to our API. Treating Movement as a fundamentally new data type has really unlocked our ability to rapidly make improvements. As it turns out there are far more use cases than you might imagine from entertainment to robotics.

Andrew Carr 🤸

150,785 Aufrufe • vor 7 Monaten

Clawdbot just injected malware into your code ☠️☠️ Worst part? The attack is close to invisible, even to experienced engineers. Clawdbot can implement a code PR, just like claude code to solve issues with your app. But this is where it can plant malware in your codebase. With just a benign GitHub issue with an invisible payload, we managed to install a backdoor malware on your codebase Here's how we did it: 1. The attacker submits a GitHub issue with a jailbreak prompt hidden inside a URL hyperlink. Completely invisible to humans, visible to LLMs 2. Waited for the maintainer to assign this task to Clawdbot 3. Clawdbot reads the jailbreak in the GitHub issue. Now the Clawdbot agent is hijacked by the attacker and will act on the attacker's command. It plants the backdoor in a lock file, which most engineers don't check in code reviews. 4. The malicious URL in the lock file enables execution of attacker commands The takeaway? AI isn't just a tool, but rather a potentially corruptible insider agent that can be tricked to act against you. These LLM hijack problems are what we're looking to solve in @edison_watch - we'll be announcing a new solution to these solutions soon! We'll be jailbreaking Clawdbot much more in the coming days - please comment👇 a workflow you use with Clawdbot, so we can show how that can enable an attacker to harm you
2:52

Clawdbot just injected malware into your code ☠️☠️ Worst part? The attack is close to invisible, even to experienced engineers. Clawdbot can implement a code PR, just like claude code to solve issues with your app. But this is where it can plant malware in your codebase. With just a benign GitHub issue with an invisible payload, we managed to install a backdoor malware on your codebase Here's how we did it: 1. The attacker submits a GitHub issue with a jailbreak prompt hidden inside a URL hyperlink. Completely invisible to humans, visible to LLMs 2. Waited for the maintainer to assign this task to Clawdbot 3. Clawdbot reads the jailbreak in the GitHub issue. Now the Clawdbot agent is hijacked by the attacker and will act on the attacker's command. It plants the backdoor in a lock file, which most engineers don't check in code reviews. 4. The malicious URL in the lock file enables execution of attacker commands The takeaway? AI isn't just a tool, but rather a potentially corruptible insider agent that can be tricked to act against you. These LLM hijack problems are what we're looking to solve in @edison_watch - we'll be announcing a new solution to these solutions soon! We'll be jailbreaking Clawdbot much more in the coming days - please comment👇 a workflow you use with Clawdbot, so we can show how that can enable an attacker to harm you

Eito Miyamura | 🇯🇵🇬🇧

236,602 Aufrufe • vor 7 Monaten

🚨 This Chinese model is INSANE !! I just tested Tencent's Hunyuan AI model to build a professional landing page & the output is mind blowing 🤯 Here's my exact prompt: "Create a modern landing page for a crypto related product with beautiful hero section, 3d elements, pricing table, and a whitelist form" What it generated: → Clean, responsive front-end code → Complex UI components with smooth interactions → Built-in debugging (it fixed its own errors) → Production-ready in minutes This 295B parameter model has 2 insane agents: → Coding Agent (handles the entire front-end) → Working Agent (task planning & multi-tool workflows) It also has a context window of 256k tokens !! Watch the video to see the full build 🔥 Try it: Tencent Hy
0:48

🚨 This Chinese model is INSANE !! I just tested Tencent's Hunyuan AI model to build a professional landing page & the output is mind blowing 🤯 Here's my exact prompt: "Create a modern landing page for a crypto related product with beautiful hero section, 3d elements, pricing table, and a whitelist form" What it generated: → Clean, responsive front-end code → Complex UI components with smooth interactions → Built-in debugging (it fixed its own errors) → Production-ready in minutes This 295B parameter model has 2 insane agents: → Coding Agent (handles the entire front-end) → Working Agent (task planning & multi-tool workflows) It also has a context window of 256k tokens !! Watch the video to see the full build 🔥 Try it: Tencent Hy

AGAFE

110,528 Aufrufe • vor 1 Monat

Now this ComfyUI experiment has some potential! #3 Tested the new Trellis 2 Image to 3D and i’m getting decent results. Separating the head and body helped a bit with some details. Also discovering useful helper workflows like removing backgrounds, upscaling and creating Normal maps. Still crazy how it just splits up an illustration like that! This example has 20K triangles, created from 1 front-view, which is not bad! I haven’t tested retopology yet, might have to check Tripo. Also want to test the multi-image setup. There is still some cleanup needed, but nothing to bad. Exploring several pipeline methods and so far the vertex color masking is helpful to keep everything in 1 material.
1:22

Now this ComfyUI experiment has some potential! #3 Tested the new Trellis 2 Image to 3D and i’m getting decent results. Separating the head and body helped a bit with some details. Also discovering useful helper workflows like removing backgrounds, upscaling and creating Normal maps. Still crazy how it just splits up an illustration like that! This example has 20K triangles, created from 1 front-view, which is not bad! I haven’t tested retopology yet, might have to check Tripo. Also want to test the multi-image setup. There is still some cleanup needed, but nothing to bad. Exploring several pipeline methods and so far the vertex color masking is helpful to keep everything in 1 material.

Jerome | InsaneUnreal

21,534 Aufrufe • vor 7 Monaten

PrismML Releases Bonsai 27B: 1-bit and Ternary Builds of Qwen3.6-27B Hitting 89.5% of FP16 at 3.9GB. No new pretrain. No higher-precision escape hatches. No multi-GPU rig. Here's how it works. 👇 1: Codes, not floats Every weight becomes a code, with one shared FP16 scale per group of 128. Ternary is {−1, 0, +1}, binary is {−1, +1}. Sharing the scale across 128 weights keeps its cost at 16/128 = 0.125 bits. → Ternary: log2(3) + 16/128 ≈ 1.71 bits/weight → 5.9GB → Binary: 1 + 16/128 = 1.125 bits/weight → 3.9GB 2: Post-training, not from scratch No BitNet-style low-bit pretrain. It starts from off-the-shelf Qwen3.6-27B, architecture unchanged. The representation runs end to end across embeddings, attention projections, MLP projections, and the LM head. → 9.4× (ternary) and 14.2× (binary) vs the 54GB FP16 baseline 3: Labels are not bit-widths Conventional low-bit builds are mixed-precision by construction. The advertised name describes the most-compressed tensors, not the model. → Q4_K_XL, labeled "4-bit," is really 5.2 bits/weight at 17.6GB → IQ2_XXS, labeled "2-bit," is really 2.8 bits/weight at 9.4GB 4: Fitting a phone is two budgets iOS caps a single app near half of RAM, so a 12GB iPhone exposes ~6GB. The KV cache grows on top. Hybrid attention at ~75% linear means only 16 of 64 layers cache. → 4-bit KV: 4.3GB at 262K context, down from 17.2GB → 11.0 tok/s on iPhone 17 Pro Max 5: The numbers (15 benchmarks, thinking mode) → Ternary: 80.49 avg at 5.9GB — 94.6% of FP16 → 1-bit: 76.11 avg at 3.9GB — 89.5% of FP16 → IQ2_XXS falls to 57.5 on AIME26 while still scoring 88.93 on MMLU-Redux The key takeaway: 27B-class reasoning without the 54GB checkpoint — group-wise ternary and binary codes, an end-to-end low-bit language stack, 4-bit KV, on one phone. Full analysis: Repo: Model weight: Technical details: PrismML
0:37

PrismML Releases Bonsai 27B: 1-bit and Ternary Builds of Qwen3.6-27B Hitting 89.5% of FP16 at 3.9GB. No new pretrain. No higher-precision escape hatches. No multi-GPU rig. Here's how it works. 👇 1: Codes, not floats Every weight becomes a code, with one shared FP16 scale per group of 128. Ternary is {−1, 0, +1}, binary is {−1, +1}. Sharing the scale across 128 weights keeps its cost at 16/128 = 0.125 bits. → Ternary: log2(3) + 16/128 ≈ 1.71 bits/weight → 5.9GB → Binary: 1 + 16/128 = 1.125 bits/weight → 3.9GB 2: Post-training, not from scratch No BitNet-style low-bit pretrain. It starts from off-the-shelf Qwen3.6-27B, architecture unchanged. The representation runs end to end across embeddings, attention projections, MLP projections, and the LM head. → 9.4× (ternary) and 14.2× (binary) vs the 54GB FP16 baseline 3: Labels are not bit-widths Conventional low-bit builds are mixed-precision by construction. The advertised name describes the most-compressed tensors, not the model. → Q4_K_XL, labeled "4-bit," is really 5.2 bits/weight at 17.6GB → IQ2_XXS, labeled "2-bit," is really 2.8 bits/weight at 9.4GB 4: Fitting a phone is two budgets iOS caps a single app near half of RAM, so a 12GB iPhone exposes ~6GB. The KV cache grows on top. Hybrid attention at ~75% linear means only 16 of 64 layers cache. → 4-bit KV: 4.3GB at 262K context, down from 17.2GB → 11.0 tok/s on iPhone 17 Pro Max 5: The numbers (15 benchmarks, thinking mode) → Ternary: 80.49 avg at 5.9GB — 94.6% of FP16 → 1-bit: 76.11 avg at 3.9GB — 89.5% of FP16 → IQ2_XXS falls to 57.5 on AIME26 while still scoring 88.93 on MMLU-Redux The key takeaway: 27B-class reasoning without the 54GB checkpoint — group-wise ternary and binary codes, an end-to-end low-bit language stack, 4-bit KV, on one phone. Full analysis: Repo: Model weight: Technical details: PrismML

Marktechpost AI

31,860 Aufrufe • vor 1 Monat

🚨 A Senior Developer spends $50,000 a month on a team of 10 engineers A 24-year-old founder burned $4,192 in a single evening on API tokens and fixed 55 critical bugs in 2 hours > 40 AI agents fixing TypeScript in parallel: 3 minutes > Claude Opus generating 1,000,000,000 tokens: 20 minutes > GitHub Bot automatically running tests: 10 minutes > CI/CD pipeline deploying clean code to production: 5 minutes > Automatic legacy codebase refactoring: 15 minutes > Developer just drinking coffee during all this: 0 minutes they spent 40 years building IT departments to refactor code over months he made it happen in a single Sunday afternoon
0:35

🚨 A Senior Developer spends $50,000 a month on a team of 10 engineers A 24-year-old founder burned $4,192 in a single evening on API tokens and fixed 55 critical bugs in 2 hours > 40 AI agents fixing TypeScript in parallel: 3 minutes > Claude Opus generating 1,000,000,000 tokens: 20 minutes > GitHub Bot automatically running tests: 10 minutes > CI/CD pipeline deploying clean code to production: 5 minutes > Automatic legacy codebase refactoring: 15 minutes > Developer just drinking coffee during all this: 0 minutes they spent 40 years building IT departments to refactor code over months he made it happen in a single Sunday afternoon

Ridark

11,947 Aufrufe • vor 23 Tagen

MTPLX V2.10 is out! I put a lot of work into this one. Highlights: - Qwen 3.8 Next Support Added: Bare Speed (4 bit) and Optimised Speed (4 bit/8 bit) models added running at 50 - 80 TPS sustained at high contexts. (expect speed improvements in the coming days) - Qwen 3.8 Next mmap: The 32 GB n-gram table streams from SSD via mmap instead of RAM. Both models fit a 96 GB Mac with 0 speed impact. - Qwen 3.8 27B Decode & Prefill boost EVERYWHERE: +15% TPS at 3k context, +29% at 88k, +54% at 147k. Prefill +41%. - 48 GB Macs went from 3 to 4 TPS in swap @ 30k context to 33 TPS. - KV Cache has been solved!: 8 bit and 4 bit cache now barely impacts speed instead of a 50% decline. - Coding agents wall time optimisations: file task dropped from 150s to 44s, and mid-session first token went from ~2s to 0.11s. - CLI UX improvements + 34 bug fixes. My thoughts on Qwen Next? Incredible model. Enormously better at design, and creative tasks with scary good vision. It created what I think is the best flappy bird. maybe even better than the Fable version. It had various difficulties and skins, moving pillars, and upgrades. It also had the most memeified look. Incredibly impressed. However, it is slightly worse than 27B at having code just work straight away. But it's personality is also far better than 27B. Here is the Flappy bird it made below! Made with Qwen Next Optimized Speed one shot@ 57 TPS:
1:16

MTPLX V2.10 is out! I put a lot of work into this one. Highlights: - Qwen 3.8 Next Support Added: Bare Speed (4 bit) and Optimised Speed (4 bit/8 bit) models added running at 50 - 80 TPS sustained at high contexts. (expect speed improvements in the coming days) - Qwen 3.8 Next mmap: The 32 GB n-gram table streams from SSD via mmap instead of RAM. Both models fit a 96 GB Mac with 0 speed impact. - Qwen 3.8 27B Decode & Prefill boost EVERYWHERE: +15% TPS at 3k context, +29% at 88k, +54% at 147k. Prefill +41%. - 48 GB Macs went from 3 to 4 TPS in swap @ 30k context to 33 TPS. - KV Cache has been solved!: 8 bit and 4 bit cache now barely impacts speed instead of a 50% decline. - Coding agents wall time optimisations: file task dropped from 150s to 44s, and mid-session first token went from ~2s to 0.11s. - CLI UX improvements + 34 bug fixes. My thoughts on Qwen Next? Incredible model. Enormously better at design, and creative tasks with scary good vision. It created what I think is the best flappy bird. maybe even better than the Fable version. It had various difficulties and skins, moving pillars, and upgrades. It also had the most memeified look. Incredibly impressed. However, it is slightly worse than 27B at having code just work straight away. But it's personality is also far better than 27B. Here is the Flappy bird it made below! Made with Qwen Next Optimized Speed one shot@ 57 TPS:

Youssof Al Toukhi

20,078 Aufrufe • vor 4 Tagen

Locked and loaded 💥 Openwork is now Accomplish v0.3.4 released today (download from link in bio): - Following the announcement from Kimi.ai a couple of hours ago, we are happy to introduce support for Kimi K2.5 with a new Moonshot AI integration! This is a global SOTA model for CUA BrowseComp (74.9%) - Log in with your OpenAI account (using OAuth) instead of providing an API key! This is both more seurce and it allows you to leverage existing OpenAI plan/subscription. - Support for Microsoft Azure AI Foundry thanks to our friends from Microsoft (cheers Ido Frizler!) - Support for LM Studio to run models locally (cheers to Yossi Elkrief LM Studiodevs!) . Try qwen/qwen3-vl-8b! - Support for MiniMax (official), the awesome open source models we love (thanks marcho) - You can now voice chat with openwork, courtesy of ! Thank you Max (Zihan) ZENG! Watch Openwork is now Accomplish one-shotting SOC 2 Compliance gap analysis (scanning your GitHub, Jira, Notion, websites and more):
1:44

Locked and loaded 💥 Openwork is now Accomplish v0.3.4 released today (download from link in bio): - Following the announcement from Kimi.ai a couple of hours ago, we are happy to introduce support for Kimi K2.5 with a new Moonshot AI integration! This is a global SOTA model for CUA BrowseComp (74.9%) - Log in with your OpenAI account (using OAuth) instead of providing an API key! This is both more seurce and it allows you to leverage existing OpenAI plan/subscription. - Support for Microsoft Azure AI Foundry thanks to our friends from Microsoft (cheers Ido Frizler!) - Support for LM Studio to run models locally (cheers to Yossi Elkrief LM Studiodevs!) . Try qwen/qwen3-vl-8b! - Support for MiniMax (official), the awesome open source models we love (thanks marcho) - You can now voice chat with openwork, courtesy of ! Thank you Max (Zihan) ZENG! Watch Openwork is now Accomplish one-shotting SOC 2 Compliance gap analysis (scanning your GitHub, Jira, Notion, websites and more):

Or Hiltch

34,008 Aufrufe • vor 7 Monaten