A quick practice run is always beneficial Model: darlenedaniellericabo... on IG #blacktapeproject #BTP #BlackTape #OriginalBTP #theblacktapeproject #blacktapeart #theblacktape #models #model #modelgirl #modellifestyle #fashion #fashiondesign #fashiondesigner #StaySexyMyFriendsshow more

Visit our website: #blacktapeproject #BTP #BlackTape #OriginalBTP #theblacktapeproject #blacktapeart... #theblacktape #models #model #fashion #fashionblogger #fashiondesign #Fashiondesigner #StaySexyMyFriendsshow more

NYC Fashion week Feb 2024 next! VIP parties, model... casting and fashion show! #blacktapeproject #BTP #BlackTape #OriginalBTP #theblacktapeproject #blacktapeart #theblacktape #models #model #modelgirl #modeltiktok #modellifestyle #fashion #fashionblogger #fashiondesignshow more

Art Basel was sick! Heading to NYC fashion week... in February! We’re doing castings and a show. Private parties and more! Dragon - deemarieditt #blacktapeproject #BTP #BlackTape #OriginalBTP #theblacktapeproject #blacktapeart #theblacktape #models #model #fashionshow more

Join the dragons & myself as we party and... showcase at NYC Fashion week Feb! Casting announcement soon! 🐉 - daniellenicollee #modelcasting #casting #NYc casting #NYC #blacktapeproject #BTP #BlackTape #OriginalBTP #theblacktapeproject #blacktapeart #theblacktape #models #modelshow more

Announcing our latest open medical AI models for developers:... MedGemma 1.5, which is small enough to run offline & improves performance on 3D imaging (CT & MRI), & MedASR, a speech-to-text model for medical dictation. Both available on Hugging Face + Vertex AI. #MedGemma #HealthAI #GenerativeAIshow more

I like my Laravel seeders fast. I use recycle()... to dodge N+1 model creations. Another trick: when a model needs an image saved on disk, generating a unique one for each record is painfully slow. And honestly, images aren’t immediately essential. So I dispatch that work to a job. My local projects always run Horizon, so images get generated in the background while I keep coding. A local environment you can reset instantly with random data is a game-changer.show more

A peanut-sized Chinese model just dethroned Gemini at reading... documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.show more

A peanut-sized Chinese model just dethroned Gemini at reading... documents. GLM-OCR is a 0.9B parameter vision-language model. It scores 94.62 on OmniDocBench V1.5, ranking #1 overall. For context, it outperforms models 100x its size. 100% open-source. It works in two stages. 1. A layout engine detects every region in a document. 2. Each region gets read in parallel. The model predicts multiple tokens per step instead of one. That's what makes it so fast at small size. It handles things most OCR tools struggle with: > Complex tables and nested layouts > Handwritten text and stamps > Math formulas and code blocks > Mixed image-and-text documents You can run it locally through Ollama. It fits on edge devices with limited compute. Every expensive OCR API just got a free competitor.show more

Got a 16GB GPU? You can run all of... these right now. Tested 4 Qwen3.5-based models on ToolCall-15 & BugFind-15: Models: - Qwen3.5:9b Q8 (Official) - Qwopus v3 Q8 by Jackrong - OmniCoder-9B by Tesslate - Qwen3.5-9b-Sushi-Coder by bigatuna Summary: - ToolCall-15: Qwopus v3 went perfect 30/30, Sushicoder beat base Qwen3.5 - BugFind-15: Omnicoder flipped the script and took #1 at 83% No single model won both, that's the fun part. Open source community is cooking.show more

Most AI agent setups treat every message the same.... Simple question? top-tier model. complex task? top-tier model. Your token bill just keeps climbing. I tested OpenSquilla this week on a real document drafting workflow, and the routing caught me off guard. It judges each message's complexity locally, then picks the model tier that fits. Simple tasks go to cheaper models. Complex ones still get the heavy lifting done. You're not paying reasoning tokens for a "hello." I ran a longer workflow, and the context didn't collapse the way it usually does. It distills important information before compression, so you're not starting from scratch mid-session. If you run agents regularly, the bill adds up faster than you think. This is built specifically for that problem. They're running the 10M Token Bill Challenge right now. worth joining if you want to see what smart routing actually saves you in practice. #10MTokenChallenge OpenSquillashow more

Microsoft made 100B parameter models run on a single... CPU. bitnet.cpp: The official inference framework for 1-bit LLMs. The math behind 1-bit LLMs is what makes them revolutionary. Traditional LLMs use 16-bit floating point weights. Every parameter is a number like 0.0023847 or -1.4729. When you run inference, you multiply these floats together. Billions of times. That's why you need GPUs, they're optimized for floating point matrix multiplication. BitNet b1.58 uses ternary weights: {-1, 0, 1}. That's not a simplification. That's a fundamental change in the math. When your weights are only -1, 0, or 1: → Multiply by 1 = keep the value → Multiply by -1 = flip the sign → Multiply by 0 = skip entirely Matrix multiplication becomes addition and subtraction. No floating point operations. No GPU required. This is why bitnet.cpp achieves: → 2.37x to 6.17x speedup on x86 CPUs → 1.37x to 5.07x speedup on ARM CPUs → 71.9% to 82.2% energy reduction on x86 → 55.4% to 70.0% energy reduction on ARM The speedups scale with model size. Larger models see bigger gains because there are more operations to simplify. A 100B parameter model running at human reading speed (5-7 tokens/second) on a single CPU. That's not optimization. That's a different paradigm. Why 1.58 bits? Because log₂(3) ≈ 1.58. Three possible values = 1.58 bits of information per weight. The key insight: These models aren't quantized after training. They're trained from scratch with ternary weights. The model learns to work within the constraint. No precision loss. No quality tradeoff.show more

Don't train the model, evolve the harness. I read... a brilliant blog post from Hugging Face where they took a frozen open model scoring 0% on a hard legal agent benchmark, left its weights alone, and let an automated loop rewrite only the code around it. That code layer is the harness, the runtime wrapper that feeds the model context, runs its tool calls, and decides when a run ends. By the time the loop finished, the system had essentially matched Sonnet 4.6 on the benchmark's headline metric, at roughly 7x lower cost per task. Zero weights changed. The gain existed because of where the model was failing. The judge only grades files saved in the right place under the exact requested filename, and the model kept doing the legal analysis correctly, then saving it under the wrong name, dropping it in a scratch folder, or never writing it at all. So the 0% was never measuring legal reasoning. It was measuring the harness. Hand-tuning that layer is slow and model-specific, so they automated it. A Claude proposer adds exactly one mechanism per iteration, and an outer loop keeps it only if it clearly beats the current best, so accepted mechanisms compound. What the loop discovered says a lot about where agents actually fail. → The biggest single gain was file handling, not intelligence. An automatic step that lands the deliverable exactly where the judge expects it beat every prompt change, with zero extra model tokens. → Code fixes transferred across models, prompt playbooks did not. The same harness lifted a smaller model from the same family by 14 points, but the tuned prompts hurt a different model family on tasks it could already finish. → The harness mattered more than anything else. Same model, same judge, same tasks, and five different harnesses scored anywhere between 3.5% and 80.1%. The gains do eventually flatten, and the remaining misses look like real capability gaps. At some point the wrapper runs out of tricks and the model has to carry the work. But the lesson holds. A benchmark score measures the model and its harness together, and until the harness is fixed, it's impossible to know which one failed. I highly recommend reading this: I also wrote a deep dive on agent harness engineering a while back, covering the orchestration loop, tools, memory, context management, and everything that turns a stateless LLM into a capable agent. The article is quoted below.show more

Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR... is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!show more

Always a thrill to see customer results like this.... Perplexity is a pioneering generative AI startup that’s reimagining the future of search (catch their app in action below). They accelerated their foundation model training up to 40% using Amazon SageMaker HyperPod. And by deploying models on our best-in-class infrastructure, they’re able to handle over 100k queries per hour without compromising latency or the user experience. Proud to support innovators like Perplexity in pushing the boundaries of AI. Happy searching!show more

90% of "AI developers" just download pre packaged GGUF... files from Hugging Face, hit run, and call it a day. The top 10% know how to pull the raw safetensors, run the math, and quantize massive models into Q4_K_M themselves. If you think llama.cpp can only execute models, you’re missing the best part of the open source ecosystem. It’s a high performance optimization suite. Manually stripping 69% of the VRAM footprint off a brand new model architecture is where real infrastructure value is made. If you want to actually master local inference and deploy models like Google’s massive Gemma 4 12B it on consumer NVIDIA hardware using llama.cpp, you need to learn this pipeline. Let's build it. I just took the raw 22.7 GB Gemma 4 baseline and manually compressed it down to a 7.02 GB Q4_K_M GGUF artifact using llama.cpp. That is a 69% reduction in footprint. No quality loss. No VRAM bottlenecks. Just native, hardware accelerated C++ inference running a full 2,50,000 token context window on a dual NVIDIA Tesla T4 setup. Stop melting your VRAM on unoptimized weights and stop relying on other people's pipelines. Own your stack. I mapped this entire architecture from dynamic binary fetching to raw quantization and real time GPU streaming into a single, bulletproof notebook. Notebook link is in the comments below. Bookmark this blueprint for your next deployment and tell me which quantization works best for your workflow and model.show more

One thing I think AI developers underestimate: Your model... isn't always your biggest risk. Your infrastructure is. A lot of AI apps work great in development… Until the provider you're relying on has an outage, rate limits your requests, or a model suddenly becomes unavailable. If your entire product depends on one AI provider, you've created a single point of failure. That's why I like the approach CometAPI is taking. Instead of locking your app to one vendor, it gives you access to 500+ AI models through a single OpenAI-compatible endpoint. That means you can: → switch models without rewriting your integration → build fallback routes for better reliability → compare models as new ones launch → avoid getting locked into a single provider For teams building with Claude Code, Cursor, AI agents, SaaS products, or automation tools, that flexibility matters a lot once you're in production. They're also shipping new models quickly. Recent additions include: • GPT-5.6 • Happy Horse 1.1 • Kling Video Plus recently released: • Gemini 3.1 Flash Lite Image • Claude Sonnet 5 Worth checking out if you're building AI products that need to stay reliable as the ecosystem moves fast. Try it here: #AI #LLM #Developers #OpenAI #Claude #Cursor #ClaudeCode #BuildInPublicshow more

my 8 GB VRAM gaming laptop is absolutely going... to hate me for this. but I still did it. ran a 31b dense model (Gemma 4 31b Q4) with only 8 GB VRAM last week I ran Gemma 4 26B A4B a mixture of experts model on my RTX 4060 and hit 25–28 tokens/sec using llama.cpp's new MTP support. smooth. snappy. but MoE has a secret: it only activates 4B parameters per token despite having 26B total. that's why it flies. so the real question started haunting me. what if I throw a full, no tricks, every parameter fires on every token, 31B DENSE model at the same machine? # Hardware: GPU: NVIDIA RTX 4060, 8 GB VRAM RAM: 16 GB CPU: Intel Core i7 H Laptop. Gaming. Modest. The model: gemma-4-31B-it-qat-UD-Q4_K_XL.gguf (model's unsloth huggingface link in the comments) This is Google DeepMind's flagship dense model in the Gemma 4 family that can run on single consumer GPU. It packs a hybrid attention architecture, supports up to 256K context natively, and is QAT (Quantization Aware Training) optimized, meaning it retains far more quality than standard post training quants at the same bit depth. This is NOT the MoE. This is 31 BILLION dense parameters, every single one of them loaded. # the flags I used: -m gemma-4-31B-it-qat-UD-Q4_K_XL.gguf -cnv --spec-type draft-mtp --spec-draft-model mtp-gemma-4-31B-it.gguf --spec-draft-n-max 8 --spec-draft-p-min 0.6 -c 6000 -v Multi Token Prediction (MTP) is still active here. Separate draft GGUF required, same as the 26B setup. # Results: → Decode: ~3 tokens/sec → Prefill: ~2 tokens/sec → Context: 6000 tokens → Hardware crying quietly in the corner: yes so is 3 tps actually usable? For real time back and forth chat? Not ideal. You're not having a fluid conversation at 3 tps. but slow ≠ useless. And this is where it gets genuinely interesting. think about how senior devs actually work in a real team. But when something is architectural, deeply complex, or needs serious reasoning? they walk down the hall and escalate to the senior. That's exactly the local AI agent architecture this unlocks: → Fast orchestrator model (Gemma 4 26B MoE at 25+ tps) handles routing, simple queries, tool calls, memory. The junior dev. → Gemma 4 31B dense is the senior, called only when the fast model genuinely hits a wall. Hard multi step reasoning. Complex code generation. Deep architectural decisions. The agentic loop stays fast. Only the hard hops touch the 31B. That's a legitimate production grade local AI architecture on a budget hardware. (requires 2 8gb gpus) other workflows where 3 tps is completely fine: - overnight batch jobs. summarize documents, extract structured data, review code. Fire it off. Sleep. wake up to results. - One shot deep reasoning - Silent code audit loops, you write and test, the 31B reviews diffs and flags issues in the background between your sprints - Any workflow where output quality > output speed A few weeks ago, nobody was running a 30B+ dense model on a single consumer GPU with 8 GB VRAM. At all. Now we're doing it on an Intel i7-H gaming laptop with a NVIDIA RTX 4060, thanks to llama.cpp + QAT quants + MTP speculative drafting. Google DeepMind said the Gemma 4 31B targets "consumer GPUs and workstations." They were not exaggerating. The hardware bar to run serious frontier class models locally keeps dropping. the tools are here. the models are here. you just have to be willing to abuse your laptop a little. what workflows would you actually run on a local 3 tps 31B dense model? genuinely curious. drop it below.show more

🧪 My GEN-3 Prompting Process I get a lot... of questions on how I find my prompts when using Image-2-Video in Runway. Here is a quick breakdown of the general thought process. If you have any further questions, let's chat in the comments below. 1️⃣ I always start by doing 3x generations without any prompts and additional settings. 2️⃣ I analyze those 3x generations and identify patterns. What did the model always do well, where did it fail. 3️⃣ I then use prompting and the different controllability features to eliminate where the model struggled on its own. --- General Tips --- *️⃣ There are some tokens which work universally. "Muted colors, low contrast" are great to preserve the colors of the original input image. "Static camera, natural movement" works fantastically to get cinematic shots. *️⃣ My I2V prompts are on the shorter side. It's usually a sentence describing the scene and then individual modifiers like the ones mentioned above to fix certain camera/lighting/movement artifacts. *️⃣ Start small and prompt engineer in steps. This is very much an iterative process which rewards you for understanding model behavior and knowing how to craft a visual architecture with words. --- Disclaimer --- Please note that this approach is more suitable for a professional workflow. Therefore, I recommend it for users on the unlimited plan who don't need to worry about credits.show more

HOLY MOLY running a 35B model locally on a... MacBook shouldn’t be THIS FAST 🤯 Spent my weekend in atomic.chat testing Qwen 35B vs. Qwen 27B on my local machine. I had them generate a fully animated HTML/Canvas car mini-game (demo below), ... and both models breezed through the physics and parallax scrolling without a hitch! The secret sauce here is the Atomic Chat app. Because it's perfectly optimized for Mac and uses Google's new TurboQuant under the hood, you can run heavy open-source models flawlessly while keeping top-tier output quality 👊 Other perks: → ZERO setup required → Access 1,000+ models completely free → 100% offline and private → Zero API limits ... and MUCH more! I dropped the prompt I used in the 🧵↓ Spin it up locally and let me know what you get!show more

Qwen 3.7-max beats Opus 4.7 and GPT-5.5 We tested... three frontier models on a real agentic task: write a Tetris bot that plays the game and trains itself. Each model could read its own code, run benchmarks, and rewrite itself across 10 iterations. Then we compared the final bots head to head. Qwen 3.7-Max: training cost $1.32, bot improvement +56% Claude Opus 4.7: training cost $12.15, bot improvement +28% GPT-5.5: training cost $2.85, bot improvement +7% Qwen won on every dimension - biggest jump, 9× cheaper than Claude, 2× cheaper than GPT. Long agentic loops is where Qwen Max actually delivers.show more