Hugging Face's logo
Friedrich Marty's picture
21 3 32

Friedrich Marty

Smorty100
NickyNicky's profile picture Maisal02's profile picture
·
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

reacted to Reality123b's post with 😔 about 8 hours ago
https://huggingface.co/posts/Reality123b/533143502736808 Since many of you upvoted that post, I'm open-sourcing this on 19th February 2025. I don't know, but, this may be the "smartest AI on earth". im not totally sure. also, i need some kind of help with the UI coz i suck at that.
reacted to lxasqjc's post with 👍 4 days ago
⚡ Can Stable Diffusion's visual expertise enhance Llama-3.2? 🚀 Lavender: efficiently fine-tunes advanced vision-language models by aligning their text-vision attention with Stable Diffusion. Paper: https://huggingface.co/papers/2502.06814 🔑 Key Highlights: ✅ Significant Gains: +30% on 20 tasks, +68% on OOD WorldMedQA ✅ Data-Efficient: Needs only 0.13M samples (~2.5% of typical VLM datasets) ✅ Low Compute: Finetunes in ~1 day on 8 NVIDIA A10G GPUs ✅ Model-Agnostic: Works with Llama-3.2-11B, MiniCPM-Llama3-v2.5 & more ✅ Precise Alignment: Transfers strong text-vision alignment from Stable Diffusion ✅ Open-Source: Code, data & finetuned models will be available 👥 Discuss live at: https://www.alphaxiv.org/abs/2502.06814 🔗 Project Page: https://astrazeneca.github.io/vlm/
updated a Space 5 days ago
Smorty100/First_agent_template
View all activity

Organizations

None yet

Smorty100's activity

reacted to Reality123b's post with 😔 about 8 hours ago
view post
Post
561
https://huggingface.co/posts/Reality123b/533143502736808
Since many of you upvoted that post, I'm open-sourcing this on 19th February 2025.

I don't know, but, this may be the "smartest AI on earth". im not totally sure.
also, i need some kind of help with the UI coz i suck at that.
reacted to lxasqjc's post with 👍 4 days ago
view post
Post
1662
⚡ Can Stable Diffusion's visual expertise enhance Llama-3.2?
🚀 Lavender: efficiently fine-tunes advanced vision-language models by aligning their text-vision attention with Stable Diffusion.
Paper: Diffusion Instruction Tuning (2502.06814)
🔑 Key Highlights:
✅ Significant Gains: +30% on 20 tasks, +68% on OOD WorldMedQA
✅ Data-Efficient: Needs only 0.13M samples (~2.5% of typical VLM datasets)
✅ Low Compute: Finetunes in ~1 day on 8 NVIDIA A10G GPUs
✅ Model-Agnostic: Works with Llama-3.2-11B, MiniCPM-Llama3-v2.5 & more
✅ Precise Alignment: Transfers strong text-vision alignment from Stable Diffusion
✅ Open-Source: Code, data & finetuned models will be available

👥 Discuss live at: https://www.alphaxiv.org/abs/2502.06814
🔗 Project Page: https://astrazeneca.github.io/vlm/

  • 1 reply
·
updated a Space 5 days ago
reacted to lewtun's post with ❤️ 6 days ago
view post
Post
4314
Introducing OpenR1-Math-220k!

open-r1/OpenR1-Math-220k

The community has been busy distilling DeepSeek-R1 from inference providers, but we decided to have a go at doing it ourselves from scratch 💪

What’s new compared to existing reasoning datasets?

♾ Based on AI-MO/NuminaMath-1.5: we focus on math reasoning traces and generate answers for problems in NuminaMath 1.5, an improved version of the popular NuminaMath-CoT dataset.

🐳 800k R1 reasoning traces: We generate two answers for 400k problems using DeepSeek R1. The filtered dataset contains 220k problems with correct reasoning traces.

📀 512 H100s running locally: Instead of relying on an API, we leverage vLLM and SGLang to run generations locally on our science cluster, generating 180k reasoning traces per day.

⏳ Automated filtering: We apply Math Verify to only retain problems with at least one correct answer. We also leverage Llama3.3-70B-Instruct as a judge to retrieve more correct examples (e.g for cases with malformed answers that can’t be verified with a rules-based parser)

📊 We match the performance of DeepSeek-Distill-Qwen-7B by finetuning Qwen-7B-Math-Instruct on our dataset.

🔎 Read our blog post for all the nitty gritty details: https://huggingface.co/blog/open-r1/update-2
replied to nroggendorff's post 6 days ago
view reply

this is so real...

just like - adress the things u don't like, don't tell it to us through ur weird games.

it'd be fun if it were treated like in kindergarden where u throw a ball around and say a thing. but it's not somehow... but no, these activities are not as self-reflective as u would hope they'd be ;(
reacted to nroggendorff's post with 👍 6 days ago
view post
Post
2608
Dearest None-yet Team,

I couldn't help but notice that our productivity has room for improvement. To address this, we will be engaging in a company-wide morale-building activity designed to boost teamwork, enthusiasm, and *most importantly* results.

I know you're all as excited as I am for this fun and absolutely required initiative. Participation is not just encouraged, it's mandatory. Think of it as a team-bonding experience you never signed up for but will absolutely tolerate.

More details to follow, but for now, mark your calendars and prepare for an engaging experience that will definitely make us all better, stronger, and more synchronized, or at least give us something to talk about later.

Looking forward to seeing you all there!

Best,
Me
·
New activity in huggingchat/chat-ui 7 days ago

[MODELS] Discussion

599
#372 opened 12 months ago by
victor
reacted to schuler's post with 👍🤯 7 days ago
view post
Post
7175
📢 New Research Alert: Making Language Models Smaller & Smarter!

Thrilled to share the latest technical report demonstrating how to reduce language model parameters by 77% while maintaining performance.

The secret? Grouped pointwise convolutions. Yes. We brought a method from computer vision to the transformers arena.

🔑 Key Findings:
• 77% parameter reduction.
• Maintained model capabilities.
• Improved generalization.

Paper: https://www.researchgate.net/publication/388835829_SAVING_77_OF_THE_PARAMETERS_IN_LARGE_LANGUAGE_MODELS_TECHNICAL_REPORT
Code: https://github.com/joaopauloschuler/less-parameters-llm
  • 2 replies
·
liked a model 10 days ago
replied to prithivMLmods's post 18 days ago
upvoted a collection 21 days ago
liked a model 22 days ago
reacted to etemiz's post with 😎👀 29 days ago
view post
Post
1117
Updated the Hoopoe model which is taking faith related and religious texts in.

etemiz/Hoopoe-8B-Llama-3.1

Faith score went from 8% to 54%. Expect more updates and increase in the score. I also did the instruct fine tuning before adding faith to the model. So some of the improvements may be there because I started with llama 3.1 base and not the instruct.

Here are some comparisons with original Llama 3.1:
replied to MonsterMMORPG's post about 1 month ago
view reply

woah, some of these demo images look actually gud... i kinda lost hope for image diffusers there for a minute, but this is impressive. the one with the leaves cought me offguard.

And on such smol GPUs now too? that is super cool!
liked a model about 1 month ago
reacted to Severian's post with 👍👀 about 1 month ago
view post
Post
3894
Interesting Solution to the Problem of Misguided Attention

So I've been fascinated by the problem of Misguided Attention for a few weeks. I am trying to build an inference algorithm to help LLMs address that issue; but in the process, I found a cool short-term fix I call "Mindful Attention" using just prompt-engineering.

Have you ever thought about how our brains filter reality through layers of past experiences, concepts, and mental images? For example, when you look at an oak tree, are you truly seeing that oak tree in all its unique details, or are you overlaying it with a generalized idea of "oak tree"? This phenomenon inspired the new approach.

LLMs often fall into a similar trap, hence the Misguided Attention problem. They process input not as it’s uniquely presented but through patterns and templates they’ve seen before. This leads to responses that can feel "off," like missing the point of a carefully crafted prompt or defaulting to familiar but irrelevant solutions.

I wanted to address this head-on by encouraging LLMs to slow down, focus, and engage directly with the input—free of assumptions. This is the core of the Mindful Attention Directive, a prompt designed to steer models away from over-generalization and back into the moment.

You can read more about the broader issue here: https://github.com/cpldcpu/MisguidedAttention

And if you want to try this mindful approach in action, check out the LLM I’ve set up for testing: https://hf.co/chat/assistant/677e7ebcb0f26b87340f032e. It works about 80% of the time to counteract these issues, and the results are pretty cool.

I'll add the Gist with the full prompt. I admit, it is quite verbose but it's the most effective one I have landed on yet. I am working on a smaller version that can be appended to any System Prompt to harness the Mindful Attention. Feel free to experiment to find a better version for the community!

Here is the Gist: https://gist.github.com/severian42/6dd96a94e546a38642278aeb4537cfb3
reacted to Nitral-AI's post with 😔 about 1 month ago
view post
Post
4590
That moment when you spend 5 days up babysitting trains, only for colab pro + to randomly disconnect the environment at every chance with 0 error indication of any kind (it just disconnects without an error). Nuke the session from the interface, but continue to eat my colab credits while it reports to wandb. 0 way of saving the models when this happens since it nukes the code preset up to auto-execute. And since the sessions 'exist' but also at the same time doesn't exist i cant close it. And have to wait till they auto timeout after 24hrs. Guess, i won't be using colab for 'quick' test trains anymore. Thanks google for scheming the very little model training budget i had for the month.
·