Large language models (LLMs) are all the rage in the generative AI world these days, with the truly large ones like GPT, LLaMA, and others using tens or even hundreds of billions of parameters to churn out their text-based responses. These typically require glacier-melting amounts of computing hardware, but the “large” in “large language models” doesn’t really need to be that big for there to be a functional, useful model. LLMs designed for limited hardware or consumer-grade PCs are available now as well, but [Binh] wanted something even smaller and more portable, so he put an LLM on a USB stick.

This USB stick isn’t just a jump drive with a bit of memory on it, though. Inside the custom 3D printed case is a Raspberry Pi Zero W running llama.cpp, a lightweight, high-performance version of LLaMA. Getting it on this Pi wasn’t straightforward at all, though, as the latest version of llama.cpp is meant for ARMv8 and this particular Pi was running the ARMv6 instruction set. That meant that [Binh] needed to change the source code to remove the optimizations for the more modern ARM machines, but with a week’s worth of effort spent on it he finally got the model on the older Raspberry Pi.

Getting the model to run was just one part of this project. The rest of the build was ensuring that the LLM could run on any computer without drivers and be relatively simple to use. By setting up the USB device as a composite device which presents a filesystem to the host computer, all a user has to do to interact with the LLM is to create an empty text file with a filename, and the LLM will automatically fill the file with generated text. While it’s not blindingly fast, [Binh] believes this is the first plug-and-play USB-based LLM, and we’d have to agree. It’s not the least powerful computer to ever run an LLM, though. That honor goes to this project which is able to cram one on an ESP32.

Large language models (LLMs) are wholly dependent on the quality of the input data with which these models are trained. While suggestions that people eat rocks are funny to you and me, in the case of LLMs intended to help out medical professionals, any false claims or statements dripping out of such an LLM can have dire consequences, ranging from incorrect diagnoses to much worse. In a recent study published in Nature Medicine by [Daniel Alexander Alber] et al. the ease with which this data poisoning can occur is demonstrated.

According to their findings, only 0.001% of training tokens have to be replaced with medical misinformation to order to create models that are likely to produce medically erroneous statement. Most concerning is that such a corrupted model isn’t readily discovered using standard medical LLM benchmarks. There are filters for erroneous content, but these tend to be limited in scope due to the overhead. Post-training adjustments can be made, as can the addition of RAG, but none of this helps with the confident bull excrement due to corruption.

The mitigation approach that the researchers developed cross-references LLM output against biomedical knowledge graphs, to reduce the LLM mostly for generating natural language. In this approach LLM outputs are matched against the graphs and if LLM ‘facts’ cannot be verified, it’s marked as potential misinformation. In a test with 1,000 random passages detected issues with a claimed effectiveness of 91.9%.

Naturally, this does not guarantee that misinformation does not make it past these knowledge graphs, and largely leaves the original problem with LLMs in place, namely that their outputs can never be fully trusted. This study also makes it abundantly clear how easy it is to corrupt an LLM via the input training data, as well as underlining the broader problem that AI is making mistakes that we don’t expect.

In the world of large language models (LLMs) there tend to be relatively few upsets ever since OpenAI barged onto the scene with its transformer-based GPT models a few years ago, yet now it seems that Chinese company DeepSeek has upended the status quo. Its new DeepSeek-V3 model is not only open source, it also claims to have been trained for only a fraction of the effort required by competing models, while performing significantly better.

The full training of DeepSeek-V3’s 671B parameters is claimed to have only taken 2.788 M hours on NVidia H800 (Hopper-based) GPUs, which is almost a factor of ten less than others. Naturally this has the LLM industry somewhat up in a mild panic, but for those who are not investors in LLM companies or NVidia can partake in this new OSS model that has been released under the MIT license, along with the DeepSeek-R1 reasoning model.

Both of these models can be run locally, using both AMD and NVidia GPUs, as well as using the online APIs. If these models do indeed perform as efficiently as claimed, they stand to massively reduce the hardware and power required to not only train but also query LLMs.

In the olden days of the WWW you could just put a robots.txt file in the root of your website and crawling bots from search engines and kin would (generally) respect the rules in it. These days, however, we have especially web crawlers from large language model (LLM) companies happily ignoring such signs on the lawn before proceeding to hover up every scrap of content on websites. Naturally this makes a lot of people very angry, but what can you do about it? The answer by [Aaron B] is Nepenthes, described on the project page as a ‘tar pit for catching web crawlers’.

More commonly known as ‘pitcher plants’, nepenthes is a genus of carnivorous plants that use a fluid-filled cup to trap insects and small critters unfortunate enough to slip & slide down into it. In the case of this Lua-based project the idea is roughly the same. Configured as a trap behind a web server (e.g. /nepenthes), any web crawler that accesses it will be presented with an endless number of (randomly generated) pages with many URLs to follow. Page generating is deliberately quite slow to not soak up significant CPU time, while still giving the LLM scrapers plenty of random nonsense to chew on.

Considering that these web crawlers deemed adhering to the friendly sign on the lawn beneath them, the least we can do in response, is to hasten model collapse by feeding these LLM scrapers whatever rolls out of a simple (optionally Markov-based) text generator.

[Fraens] has been re-making industrial machines in fantastic 3D-printable versions for a few years now, and we’ve loved watching his creations get progressively more intricate. But with this nearly completely 3D-printable needle loom, he’s pushing right up against the edge of the possible.

The needle loom is a lot like the flying shuttle loom that started the Industrial Revolution, except for making belts or ribbons. It’s certainly among the most complex 3D-printed machines that we’ve ever seen, and [Fraens] himself says that it is pushing the limits of what’s doable in plastic — for more consistent webbing, he’d make some parts out of metal. But that’s quibbling; this thing is amazing.

There are mechanical details galore here. For instance, check out the cam-chain that raises, holds, and lowers arms to make the pattern. Equally important are the adjustable friction brakes on the rollers that hold the warp, that create a controlled constant tension on the strings.  (Don’t ask us, we had to Wikipedia it!) We can see that design coming in handy in some of our own projects.

On the aesthetic front, the simple but consistent choice of three colors for gears, arms, and frame make the build look super tidy. And the accents of two-color printing on the end caps is just the cherry on the top.

This is no small project, with eight-beds-worth of printed parts, plus all the screws, bearings, washers, etc. The models are for pay, but if you’re going to actually make this, that’s just a tiny fraction of the investment, and we think it’s going to a good home.

We are still thinking of making [Fraens]’s vibratory rock tumbler design, but check out all of his work if you’re interested in nice 3D-printed mechanical designs.

Not all 3D scanning is alike, and the right workflow can depend on the object involved. [Ding Dong Drift] demonstrates this in his 3D scan of a project car. His goal is to design custom attachments, and designing parts gets a lot easier with an accurate 3D model of the surface you want to stick them on. But it’s not as simple as just scanning the whole vehicle. His advice? Don’t try to use or edit the 3D scan directly as a model. Use it as a reference instead.

To do this, [Ding Dong Drift] scans the car’s back end and uses it as a reference for further CAD work. The 3D scan is essentially a big point cloud and the resulting model has a very high number of polygons. While it is dimensionally accurate, it’s also fragmented (the scanner only captures what it can see, after all) and not easy to work with in terms of part design.

In [Ding Dong Drift]’s case, he already has a 3D model of this particular car. He uses the 3D scan to fine-tune the model so that he can ensure it matches his actual car where it counts. That way, he’s confident that any parts he designs will fit perfectly.

3D scanning has a lot of value when parts have to fit other parts closely and there isn’t a flat surface or a right angle to be found. We saw how useful it was when photogrammetry was used to scan the interior of a van to help convert it to an off-grid camper. Things have gotten better since then, and handheld scanners that make dimensionally accurate scans are even more useful.

Code Llama is a suite of large language models released by Meta AI for generating and enhancing code. It includes foundation models for general coding, Python specializations, and models tailored for following instructions. Key features include state-of-the-art performance, code infilling, large context support up to 100K tokens, and zero-shot ability to follow instructions for programming […]

Source

ChatGLM-6B is an open-source, bilingual conversational AI LLM based on the General Language Model (GLM) framework. It has 6.2 billion parameters and can be deployed locally with only 6GB of GPU memory. This model allows for natural language processing in both Chinese and English, question answering, task-oriented dialogue, and easy integration via API and demo […]

Source

Codestral is a powerful 22B parameter AI model from Mistral AI. This open-weight model is designed specifically for code generation across over 80 programming languages including Python, Java, C++, JavaScript and more. Codestral offers impressive performance, outperforming other models on benchmarks like HumanEval and RepoBench with its large 32k token context window. The model is […]

Source

Animate Anything is an AI animation tool that lets you rig and animate your 3D models. Simply upload your static 3D models and the AI will automatically rig and animate them, ready for games, VR, and metaverse worlds. The tool supports multiple 3D formats (FBX, GLB, GLTF+BIN) and rigged models can be imported into Unity, […]

Source

Contlo’s Brand AI Model allows businesses to orchestrate all their marketing activities with their brand’s own AI model by contextually generating personalized marketing creatives including emails, images, copies, and more. To use the tool, provide the AI with information such as brand story, identity, design language, tonality, and aesthetic attributes that represent the brand. The […]

Source

3DFY.ai is a text-to-3D tool that makes creating high-quality 3D models as easy as writing a text prompt. Just type in an item or object, such as a tree or furniture, and the AI will generate a 3D asset that matches your prompt. The 3D model can then be imported into most popular 3D modelling […]

Source

TensorFlow is a popular end-to-end open source machine learning platform. It provides tools to prepare data, build models, and deploy models in production. Developers are able to utilize pre-trained models or create their own custom ones. TensorFlow supports on-device, in-browser, on-server and cloud deployment. It also has an active community forum where you can connect, […]

Source

Mistral AI is a large language model and chat assistant tool. You can access the chatbot via the Mitral website by clicking on “Talk to le Chat“, or if you prefer a local setup then you can download and run the model files on your own hardware. The creators of Mistral describe it as an […]

Source