Friedrich Marty

Smorty100

https://gitlab.com/users/Marty_Friedrich/projects

AI & ML interests

I'm most interested in content rerouting between LLM and VLLM agens for automation possibilities. Using templates for each agent which is then filled in by another agents inputs seems really useful.

Recent Activity

liked a model 1 day ago

nvidia/Llama-3_1-Nemotron-Ultra-253B-v1

replied to etemiz's post 4 days ago

It looks like Llama 4 team gamed the LMArena benchmarks by making their Maverick model output emojis, longer responses and ultra high enthusiasm! Is that ethical or not? They could certainly do a better job by working with teams like llama.cpp, just like Qwen team did with Qwen 3 before releasing the model. In 2024 I started playing with LLMs just before the release of Llama 3. I think Meta contributed a lot to this field and still contributing. Most LLM fine tuning tools are based on their models and also the inference tool llama.cpp has their name on it. The Llama 4 is fast and maybe not the greatest in real performance but still deserves respect. But my enthusiasm towards Llama models is probably because they rank highest on my AHA Leaderboard: https://sheet.zoho.com/sheet/open/mz41j09cc640a29ba47729fed784a263c1d08 Looks like they did a worse job compared to Llama 3.1 this time. Llama 3.1 has been on top for a while. Ranking high on my leaderboard is not correlated to technological progress or parameter size. In fact if LLM training is getting away from human alignment thanks to synthetic datasets or something else (?), it could be easily inversely correlated to technological progress. It seems there is a correlation regarding the location of the builders (in the West or East). Western models are ranking higher. This has become more visible as the leaderboard progressed, in the past there was less correlation. And Europeans seem to be in the middle! Whether you like positive vibes from AI or not, maybe the times are getting closer where humans may be susceptible to being gamed by an AI? What do you think?

replied to danielhanchen's post 8 days ago

You can now run Llama 4 on your own local device! 🦙 Run our Dynamic 1.78-bit and 2.71-bit Llama 4 GGUFs: https://huggingface.co/unsloth/Llama-4-Scout-17B-16E-Instruct-GGUF You can run them on llama.cpp and other inference engines. See our guide here: https://docs.unsloth.ai/basics/tutorial-how-to-run-and-fine-tune-llama-4

View all activity

Organizations

None yet

Smorty100's activity

liked a model 1 day ago

nvidia/Llama-3_1-Nemotron-Ultra-253B-v1

Text Generation • Updated about 16 hours ago • 12.5k • • 240

replied to etemiz's post 4 days ago

Where do u have that info about qwen 3 from? I'm really curious now...

replied to danielhanchen's post 8 days ago

maybe i am not understanding this correctly, but in the picture you say that you menaged to squeeze the model to a size of 34GB, but then you say that now it can run on GPUs with "just" 24 GB of VRAM.

Was this a typo or am I maybe missing something here?

reacted to AtAndDev's post with 😔 9 days ago

Post

2877

Llama 4 is out...

3 replies

liked a model 16 days ago

teapotai/teapotllm

Text2Text Generation • Updated 10 days ago • 8.59k • 165

liked a model 23 days ago

zzzrw/DeepMesh

Updated 27 days ago • 40

New activity in mistralai/Mistral-Small-3.1-24B-Instruct-2503 23 days ago

Can't determine properly which is greater between 9.9 and 9.11

#38 opened 25 days ago by

sniffski

replied to clem's post 23 days ago

This really comes down to what type of robots.

Are we talking lil hobbyist robots, about 20-30 centimeters?

or are we talking

larger human-sized robots with figure-2 level movement options?

for a smol hobbyist one, I'd be able to pay anything sub 300€, if it's also open-hardware that is (if something breaks, we can replace it).

For a big, human-sized robot, the pricing is more difficult. I would pay anything up to 5000€, even though I'm currently unable to.

I believe it would make most sense to compute all the AI stuff over the network and not inside the bot (better for weight and battery life).

If this is real, I'd be super excited either way.

reacted to clem's post with ❤️🔥 23 days ago

Post

3714

Should we assemble affordable open-source robots at Hugging Face for the community. Would you buy them? At what price?

8 replies

reacted to Jaward's post with 🚀 27 days ago

Post

2114

Nvidia brings blue (from starwars droids) to life 🤯, supercute with flawless dexterity and droid voice. It's the result of their colab research with Google DeepMind and Disney, revealed as part of their new opensource physics engine for robotics simulation: NEWTON - which enables robots to learn how to complete complex tasks with greater precision.

ReadMore: https://developer.nvidia.com/blog/announcing-newton-an-open-source-physics-engine-for-robotics-simulation?ncid=so-twit-820797-vt48