1 6 14

Tyler Williams PRO

unmodeled-tyler

https://unmodeledtyler.com

AI & ML interests

AI researcher/engineer. Open source advocate.

Recent Activity

posted an update about 2 hours ago

If you haven't played around with the Hugging Face Hub API yet, you should! I recently built this custom analytics dashboard to track how our models perform over time. I'm hoping this tool will help us understand how the community uses the models, which types of models people prefer over others, and insights about what's maybe working and what isn't. This is the second project I've built with the Hub API, and I can't wait to do more!

reacted to danielhanchen's post with 🚀 about 2 hours ago

You can now run Kimi K2 Thinking locally with our Dynamic 1-bit GGUFs: https://huggingface.co/unsloth/Kimi-K2-Thinking-GGUF We shrank the 1T model to 245GB (-62%) & retained ~85% of accuracy on Aider Polyglot. Run on >247GB RAM for fast inference. We also collaborated with the Moonshot AI Kimi team on a system prompt fix! 🥰 Guide + fix details: https://docs.unsloth.ai/models/kimi-k2-thinking-how-to-run-locally

reacted to danielhanchen's post with ❤️ about 2 hours ago

View all activity

Organizations

posted an update about 2 hours ago

Post

If you haven't played around with the Hugging Face Hub API yet, you should! I recently built this custom analytics dashboard to track how our models perform over time.

I'm hoping this tool will help us understand how the community uses the models, which types of models people prefer over others, and insights about what's maybe working and what isn't.

This is the second project I've built with the Hub API, and I can't wait to do more!

reacted to danielhanchen's post with 🚀❤️ about 2 hours ago

Post

2982

You can now run Kimi K2 Thinking locally with our Dynamic 1-bit GGUFs: unsloth/Kimi-K2-Thinking-GGUF

We shrank the 1T model to 245GB (-62%) & retained ~85% of accuracy on Aider Polyglot. Run on >247GB RAM for fast inference.

We also collaborated with the Moonshot AI Kimi team on a system prompt fix! 🥰

Guide + fix details: https://docs.unsloth.ai/models/kimi-k2-thinking-how-to-run-locally

reacted to jasoncorkill's post with 🚀 about 2 hours ago

Post

311

Do you remember https://thispersondoesnotexist.com/ ? It was one of the first cases where the future of generative media really hit us. Humans are incredibly good at recognizing and analyzing faces, so they are a very good litmus test for any generative image model.

But none of the current benchmarks measure the ability of models to generate humans independently. So we built our own. We measure the models ability to generate a diverse set of human faces and using over 20'000 human annotations we ranked all of the major models on their ability to generate faces. Find the full ranking here:
https://app.rapidata.ai/mri/benchmarks/68af24ae74482280b62f7596

We have release the full underlying data publicly here on huggingface: Rapidata/Face_Generation_Benchmark

reacted to mike-ravkine's post with 👀 2 days ago

Post

1356

measuring the information content of a reasoning trace seems like a straightforward reasoning LLM KPI, but how can we achieve this?

what if we keep it simple: gzip the resulting text and take the length of the compressed stream... "compressed bytes of information per output token" becomes the KPI

if we split across correct answers vs incorrect answers vs truncated answers and group by difficulty, a whole new world of analysis becomes not just possible but visually intuitive and almost trivial:

1) what is the model's overall reasoning efficiency? this is the slope of the scatterplot curve segments (there may be more then one..)

2) is the model able to apply more test-time compute towards more difficult variations of the task? the two on the left are not, the two on the right are.

3) when applying more test-time compute, is that compute useful? this is the curvature of the scatterplot trends - the two in the middle are 'losing their mojo' as answers get longer the information content falls down

4) is the model applying multiple approaches to the task? (right) do those approaches change with difficulty?

5) are truncations because we don't have enough context budget (left) or because the model has lost its mind and gone into a repeat loop (middle two) and does this happen across the board (middle left) or only when the problem is more difficult (middle right)

would love to hear your guys feedback on this kind of analysis, is anyone doing similar work?

this approach generates 12 plots per model (one for each task) so quite a bit of data and i've been hesitant to publish it so far, consider this post a toe tip.

reacted to branikita's post with 🚀 2 days ago

Post

2753

In November, with

Robonine are starting work on the next version of the SO ARM 102 manipulator. The version will be open source and agreed upon with @therobotbuilder the creator of the original manipulator.

We are planning to:

- increase positioning accuracy by approximately 2x using Feetech STS 3250 motors
- increase working payload from 200g to 300g
- increase rigidity using parametric design optimization and stiffer plastic
- increase length to 550 mm
- increase folding angles
- use ISO 9409-1-50-4-M6 mounting standard for the gripper
- use a parallel gripper in the default version
- update the mounting plate for different camera types, M3 grid with 12.5 mm pitch
- add table mounting standard 80x80 M8

The number of degrees of freedom and basic kinematics will remain the same.

Are there other things missing for working with SO ARM 100?

- Any standard inputs/outputs, for example?
- Status indicators?
- Perhaps some types of mounting for third-party grippers are more preferable?
- Anything else?

posted an update 3 days ago

Post

1615

NEW DATASET: vanta-research/human-ai-collaboration-1

We are excited to share the first iteration of our dataset focused on human-AI collaborative tasks!

This dataset contains 3,050 lines of warm, collaborative, and natural conversational examples designed to teach the model how to effectively and efficiently problem solve back and forth with a human.

Additionally, the examples include <think> tags, showing the model proper internal reasoning.

VANTA Research is committed to contributing back to the open source community in order to make AI development more accessible, transparent, and beneficial for all.

reacted to DavidVivancos's post with 🔥 3 days ago

Post

861

Hi all!,

Neuraxon ( a novel Neural Growth & Computation Blueprint) is live in demo at HuggingFace DavidVivancos/Neuraxon

Paper: https://www.researchgate.net/publication/397331336_Neuraxon (on its way to arxiv too)

Code: https://github.com/DavidVivancos/Neuraxon

HuggingFace Model in the oven!

Hope you like it!
@DavidVivancos

2 replies

reacted to mike-ravkine's post with 🔥 4 days ago

Post

2704

There is no anxiety quite like powering up 2KW of basement compute after rewiring it all. Small bit of trouble with the horizontal 3090 because I misread my motherboard manual, but otherwise so far so good.. Next we see if I've built up enough cooling to hit my target TDP on those 3-slot nvlinked cards especially. The 4-slot bridges are much easier to work with but their prices went bananas and I couldn't acquire a second, so gotta get a little creative with intakes.

5 replies

reacted to wang12390's post with ❤️ 4 days ago

Post

1580

🧠 What Is SpeedPaint?

SpeedPaint is an AI Speed Painting software that simulates how an artist paints — step by step, layer by layer — but at machine speed.
Instead of generating a finished image instantly, it paints in motion, giving users a live-brush experience.

Miragic-AI/Miragic-Speed-Painting
https://miragic.ai/products/speed-painting

1 reply

reacted to flozi00's post with 🧠 4 days ago

Post

1891

I just got asked about the differences between Blackwell systems and Grace Blackwell systems. What's the difference and how much of a performance gap is there between them?

https://flozi.net/en/hardware/nvidia/benchmarks/b200-vs-gb200-efficiency-comparison

Here's a summary of the key points from the article:

GB200 (Grace Blackwell) is a Superchip: It integrates a Grace CPU and two Blackwell GPUs into a single package.
B200 is a GPU-only module: It's designed to be paired with x86 or ARM CPUs in more traditional server setups.

Performance and Efficiency:

Based on MLPerf Training v5.0 benchmarks, the article concludes:

GB200 systems are approximately 42% more efficient than B200 systems on average. This is especially true in large-scale deployments (100+ GPUs), where the GB200's integrated design and high-speed NVLink interconnect provide a significant advantage.

In smaller, single-node systems (e.g., 8 GPUs), the performance difference is much smaller, around 10-15%.

Use Cases:

Choose GB200 for large-scale AI clusters, training massive models, and when maximum efficiency is the top priority.

Choose B200 for smaller deployments, when you need the flexibility to choose your own CPU, or for mixed AI and HPC workloads.

reacted to AdinaY's post with 🚀 4 days ago

Post

2585

Kimi K2 Thinking is now live on the hub 🔥

moonshotai/Kimi-K2-Thinking

✨ 1T MoE for deep reasoning & tool use
✨ Native INT4 quantization = 2× faster inference
✨ 256K context window
✨ Modified MIT license

reacted to branikita's post with 🔥 4 days ago

Post

2223

Load test conducted on the Feetech STS3250 servo motor.

With a 2 kg load on a 100 mm arm, the motor operated near its limit. At higher acceleration settings, lifting performance decreased noticeably. The temperature increased from 40 °C to 70 °C within 8 minutes.
The test highlights the torque and thermal constraints under sustained load conditions.

#Robotics #Engineering #ServoMotor #Testing #Feetech #Automation #Mechatronics #Hardware

reacted to wang12390's post with 🚀 5 days ago

Post

2166

Totally Free Image Generator, which is the best quality.

Transform your imagination into breathtaking visuals with our advanced AI technology. No skills required—just describe your vision and watch the magic happen

Huggingface Demo:
Miragic-AI/Miragic-AI-Image-Generator

Website:
https://miragic.ai/products/image-generator

reacted to codelion's post with 🔥 6 days ago

Post

3545

On this day in 2019, OpenAI released the final GPT-2 model as part of their staged release. I still remember that November well - so much was happening, but GPT-2's release felt like a watershed moment for the field. It showed us what was possible with carefully trained language models.

To recreate some of that GPT-2 magic, I recently tackled an interesting challenge: can you pretrain a language model with just 1 billion tokens - roughly 1/10th of what GPT-2 used - and still get comparable performance? After 50+ systematic experiments testing different dataset mixtures, the answer is yes.

The result is codelion/gpt-2-70m, which achieves over 90% of GPT-2's benchmark performance despite being trained on 10x less data. The key was finding the optimal dataset composition: 50% high-quality textbook PDFs, 30% filtered web content, and 20% educational resources. It even beats GPT-2 on TruthfulQA (47.31% vs 40.69%).

If you're interested in the full story of how we discovered this optimal mixture and why curriculum learning catastrophically failed, check out the complete article: https://huggingface.co/blog/codelion/optimal-dataset-mixing

Sometimes less really is more - when you mix it right.

1 reply

replied to codelion's post 6 days ago

This is so cool! What a fun project. Thanks for sharing!

reacted to Ihor's post with 👀 6 days ago

Post

1120

Hey builders 👷‍♀️

We’re Knowledgator, the team behind open-source NLP models like GLiNER, GLiClass, and many other used for zero-shot text classification and information extraction.

If you’ve explored them on Hugging Face or used our frameworks from GitHub, we’d love your input:
🧩 Which of our models, like GLiNER or zero-shot classifiers, do you find helpful in your practical workflows?
🧩 How’s the setup, performance, and accuracy been for you?
🧩 Anything confusing, buggy, or missing that would make your workflow smoother?

Your feedback helps us improve speed, clarity, and stability for everyone in the open-source community.

💬 Comment directly here or join the discussion. We read every one 😉:
GitHub: https://github.com/Knowledgator
Discord: https://discord.gg/GXRcAVJQ
HuggingFace:

knowledgator

📝 Want to shape our next release?
Click here to complete this 2-min survey: https://docs.google.com/forms/d/e/1FAIpQLSdyz2UMHrMDX8S9stpBk0wyfngtKSYzwk-02mN1VNYDdTw8OQ/viewform

reacted to sergiopaniego's post with 🔥 6 days ago

Post

5135

fine-tuning a 14B model with TRL + SFT on a free Colab (T4 GPU)?
thanks to the latest TRL optimizations, you actually can!
sharing a new notebook showing how to do it 😎

colab: https://colab.research.google.com/github/huggingface/trl/blob/main/examples/notebooks/sft_trl_lora_qlora.ipynb

notebooks in TRL: https://github.com/huggingface/trl/tree/main/examples/notebooks

2 replies

reacted to InezCornell's post with 🔥 7 days ago

Post

254

Radar data for experimentation: Access real radar datasets and expert support to build AI/ML solutions using radar data.

Ideal for companies looking to add radar capabilities to their services – access the data and guidance to help you get started quickly.

Free access to:
- Sample dataset as a CSV
- Accompanying README file

https://www.plextek.com/learn/radar-data-for-experimentation/

posted an update 10 days ago

Post

644

New Datasets Published:
vanta-research/poetic-imagery-small
vanta-research/excitement-small

We are open sourcing two of our datasets today, which were used in the training of Apollo Astralis 8B and 4B.

The first dataset, poetic-imagery-small is designed to give the model's responses a bit of "depth" to them in order to encourage curiosity and thought from the user.

Additionally, the excitement-small dataset is designed to teach the model how to use "excited" language conversationally. This dataset was used on both Apollo Astralis models, which effectively demonstrate general excitement during user interaction.

VANTA Research is an AI safety project which aims to research and develop language models aligned for all types of thinking. These datasets were created aligned with that mission, in addition to rigorous AI safety standards.

Tyler Williams PRO

AI & ML interests

Recent Activity

Organizations

unmodeled-tyler's activity