Csplk (Ci Splunk)

reacted to sergiopaniego's post with 🔥 about 1 hour ago

Post

835

The list of hands-on notebooks (some beginner-friendly!) to get started with fine-tuning using TRL keeps growing!!

• SFT
• GRPO
• Tool calling & agents
• RL environments with OpenEnv
• LLMs and VLMs
✨ Many run on FREE Colab, making it super easy to get started fast!

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

reacted to cedricbonhomme's post with 👍 about 15 hours ago

Post

738

With VLAgentIc, you can now use your local Qwen installation via Ollama and leverage the models CIRCL/vulnerability-severity-classification-roberta-base and CIRCL/cwe-parent-vulnerability-classification-roberta-base.

The project is available here:
https://github.com/vulnerability-lookup/VLAgentIc

The VLAI Severity and CWE classifiers are available on Hugging Face:
- CIRCL/vulnerability-severity-classification-roberta-base
- CIRCL/cwe-parent-vulnerability-classification-roberta-base

The concept of AI agents—combining models, tools, and orchestration—has become fairly standardized during the last year, but VLAgentIc brings something unique:

- Agents communicate over XMPP, enabling concurrent tasks and asynchronous messaging thanks to the SPADE framework.
- Built-in presence and discovery streamline interactions between components.
- Flexible behaviours make orchestrating AI-assisted security workflows seamless for future connections
- Last but not least, the VLAI Severity and VLAI CWE classifiers are now wrapped as LLM Tools and run entirely locally.

New, more comprehensive agent tools will soon be available, leveraging the Vulnerability-Lookup API and supporting the GCVE project.

The Human-in-the-Loop agent tool will be designed to notify you and request authorization whenever a query to an external service is about to be made—ensuring that, by default, all reasoning and processing stay local on your computer.

VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification (2507.03607)

reacted to Nymbo's post with 🔥 15 days ago

Post

1869

🚨 New tool for the Nymbo/Tools MCP server: The new Agent_Skills tool provides full support for Agent Skills (Claude Skills but open-source).

How it works: The tool exposes the standard discover/info/resources/validate actions. Skills live in /Skills under the same File_System root, and any bundled scripts run through Shell_Command, no new infrastructure required.

Agent_Skills(action="discover")  # List all available skills
Agent_Skills(action="info", skill_name="music-downloader")  # Full SKILL.md
Agent_Skills(action="resources", skill_name="music-downloader")  # Scripts, refs, assets

I've included a music-downloader skill as a working demo, it wraps yt-dlp for YouTube/SoundCloud audio extraction.

Caveat: On HF Spaces, Shell_Command works for most tasks, but some operations (like YouTube downloads) are restricted due to the container environment. For full functionality, run the server locally on your machine.

Try it out ~ https://www.nymbo.net/nymbot

replied to Nymbo's post about 1 month ago

Very handy

reacted to Nymbo's post with 🚀 about 1 month ago

Post

5117

🚀 I've just shipped a major update to the Nymbo/Tools MCP server: the Agent_Terminal, a single "master tool" that cuts token usage by over 90%!

Anthropic found 98.7% context savings using code execution with MCP, Cloudflare published similar findings. This is my open-source implementation of the same idea.

# The Problem

Traditional MCP exposes every tool definition directly to the model. With 12 tools, that's thousands of tokens consumed *before the conversation even starts*. Each tool call also passes intermediate results through the context window — a 10,000-row spreadsheet? That's all going into context just to sum a column.

# The Solution: One Tool to Rule Them All

Agent_Terminal wraps all 12 tools (Web_Search, Web_Fetch, File_System, Generate_Image, Generate_Speech, Generate_Video, Deep_Research, Memory_Manager, Obsidian_Vault, Shell_Command, Code_Interpreter) into a single Python code execution gateway.

Instead of the model making individual tool calls, it writes Python code that orchestrates the tools directly:

# Search for Bitcoin price
result = Web_Search("current price of bitcoin", max_results=3)
print(result)

Don't know what tools are available? The agent can discover them at runtime:

print(search_tools('image'))  # Find tools by keyword
print(usage('Generate_Image'))  # Get full docs for a specific tool

The individual direct tool calls are all still there, but they can be disabled if using the Agent_Terminal. Try it now - https://www.nymbo.net/nymbot

1 reply

·

reacted to sergiopaniego's post with 👍 about 1 month ago

Post

2594

we've just added several example scripts to TRL showing how to train models with GRPO using some of the new OpenEnv environments

train a model to interact with a browser (🎮 BrowserGym Env), play Wordle (🎮 Wordle Env) and moooore!

TRL (GRPO + vLLM) + OpenEnv! ⚡️

📝 go play with them: https://github.com/huggingface/trl/tree/main/examples/scripts/openenv

📝 examples list: https://huggingface.co/docs/trl/main/en/example_overview#scripts

reacted to thecollabagepatch's post with 👍 about 2 months ago

Post

408

hey musicians

hf continues to make the anti-suno device possible with gary4juce, the VST for your DAW that doesn't try to replace you.

v2 just released. https://thepatch.gumroad.com/l/gary4juce (pay what you want)

now you can use google's magenta-realtime model to generate 48k samples based on your input audio (or other model outputs...there's 4 to play with now).

just duplicate my hf space, turn on an L4/L40s and throw the url into the plugin.

i've got a few finetunes you can switch to as well. or you can push your finetune to the hub and play around.

the space: thecollabagepatch/magenta-retry (you can also use the html web tester to play around with realtime generation on the L40s)

13 replies

·

reacted to prithivMLmods's post with 🔥 about 2 months ago

Post

3278

Try the all-new trending Qwen-Image-Edit specialized adapter demos, including Photo-to-Anime, Light Restoration, Multi-Angle Edits, Relighting, and more — all in a single Hugging Face Space. Below is the demo link. 🤗🌠

⮞ Demo-Space: prithivMLmods/Qwen-Image-Edit-2509-LoRAs-Fast
⮞ How-to-Use: prithivMLmods/Qwen-Image-Edit-2509-LoRAs-Fast#2
⮞ Collection: https://huggingface.co/collections/prithivMLmods/image-generation-apps-collection

To know more about it, visit the app page or the respective model page!

3 replies

·

reacted to jbilcke-hf's post with 🚀 2 months ago

Post

4118

I made a code sniping agent to detect when new AI papers with code (and weights) are released, and then automatically create a Gradio demo on Hugging Face 🧙

Here are some examples generated 100% automatically:
https://huggingface.co/collections/jbilcke-hf/sniped

I call this agent CheatCode (https://github.com/jbilcke-hf/CheatCode) because it skips so many steps that it kinda feels like breaking the rules of the AI tech release game 😅

As with any experimental technology, there is still room for improvement 👩🏻‍🔬:

- Currently the demos are all generated in one go and not built or tested by the agent itself. A more robust version should loop over the deployed app to fix build/runtime issues.
- There is still a bit of human curation done to avoid making demos for things that can’t really be demonstrated on ZeroGPU (eg. tasks taking several minutes)
- Some papers can actually be showcased in a variety of ways, which isn’t really supported (see Demo 2)

reacted to prithivMLmods's post with ❤️ 2 months ago

Post

2287

Let’s have the comparison again with Multimodal OCR3:

nanonets/Nanonets-OCR2-3B vs allenai/olmOCR-2-7B-1025 vs rednote-hilab/dots.ocr vs datalab-to/chandra

Try it here @ prithivMLmods/Multimodal-OCR3

Collection: prithivMLmods/multimodal-implementations-67c9982ea04b39f0608badb0

reacted to Nymbo's post with 🚀 3 months ago

Post

1864

I've made some improvements to my custom Deep_Research tool in the Nymbo/Tools MCP server. I've added a second LLM process and it still takes less than 1 minute to complete!

The original version of my Deep_Research tool would basically dump up to 50 fetched webpages onto the Researcher model (Qwen3-235B), with only a little bit of context shown from each page.

# New "Filterer" Process

The new process includes another LLM call before the researcher process. The Filterer (also Qwen3-235B) gets the query summary and the original 50 pages with low context, and decides which pages are most relevant to the research topic. The Filterer then outputs the URLs to the relevant pages, which are then re-fetched (with more context) and sent to the Researcher.

# Researcher Context

The Researcher now gets only the relevant webpages, then begins writing the report. When testing with 50 initial results, the researcher would often end up with 10-20 results of relevant context.

It still takes less than a minute to accomplish everything, thanks entirely to Cerebras inference. It now takes about 35-45 seconds to complete once the tool is run.

It's also worth noting that both the Filterer and Researcher now are provided the current time/date before they see the content, reducing hallucinations caused by knowledge cutoffs.

reacted to vikhyatk's post with 🔥 4 months ago

Post

4831

Just released a preview of Moondream 3! moondream/moondream3-preview

This is a 9B parameter, 2B active MoE VLM with state of the art visual reasoning capabilities.

More details in the release blog post: https://moondream.ai/blog/moondream-3-preview

3 replies

·

reacted to prithivMLmods's post with ❤️ 4 months ago

Post

3407

Comparing: DeepCaption-VLA-7B, built on Qwen2.5-VL-7B-Instruct, is tailored for image captioning and vision-language attribution, focusing on precise, descriptive captions of visual properties, object attributes, and scene details. In contrast, Qwen2.5-VL-7B-Abliterated-Caption-it is fine-tuned for abliterated captioning, generating highly detailed descriptions across diverse visual categories.

Models🤗
✦ DeepCaption-VLA-7B : prithivMLmods/DeepCaption-VLA-7B
✦ Qwen2.5-VL-7B-Abliterated-Caption-it : prithivMLmods/Qwen2.5-VL-7B-Abliterated-Caption-it

Spaces⛵
➜ VisionScope-R2 : prithivMLmods/VisionScope-R2
➜ Qwen2.5-VL-Outpost : https://huggingface.co/spaces/prithivMLmods/Qwen2.5-VL-Outpost

Collection🗞️
DeepCaption attr. : prithivMLmods/deepcaption-attr-68b041172ebcb867e45c556a
VL Abliterated-Caption : prithivMLmods/vl-abliterated-caption-68a0443b63182e97a15c47a3
Multimodal VLMs - Until July'25 : prithivMLmods/multimodal-vlms-until-july25-688312e6b840e1e156f13027
Multimodal VLMs - Aug'25 : prithivMLmods/multimodal-vlms-until-july25-688312e6b840e1e156f13027

GitHub↗️
> DeepCaption-VLA-7B [4bit-notebook demo] : https://github.com/PRITHIVSAKTHIUR/Multimodal-Outpost-Notebooks/blob/main/DeepCaption-VLA-7B%5B4bit%20-%20notebook%20demo%5D/DeepCaption-VLA-7B.ipynb
> Qwen2.5-VL-3B-Abliterated-Caption-it(caption) : https://github.com/PRITHIVSAKTHIUR/Multimodal-Outpost-Notebooks/blob/main/Qwen2.5-VL-3B-Abliterated-Caption-it(caption)/Qwen2_5_VL_3B_Abliterated_Caption_it.ipynb

The community GPU grant was given by Hugging Face — special thanks to them. 🤗🚀

To know more about it, visit the app page or the respective model page!!

replied to ZennyKenny's post 5 months ago

Just use FHE and move on

reacted to fdaudens's post with 👍 6 months ago

Post

2626

You might not have heard of Moonshot AI — but within 24 hours, their new model Kimi K2 shot to the top of Hugging Face’s trending leaderboard.

So… who are they, and why does it matter?

Had a lot of fun co-writing this blog post with @xianbao , with key insights translated from Chinese, to unpack how this startup built a model that outperforms GPT-4.1, Claude Opus, and DeepSeek V3 on several major benchmarks.

🧵 A few standout facts:

1. From zero to $3.3B in 18 months:
Founded in March 2023, Moonshot is now backed by Alibaba, Tencent, Meituan, and HongShan.

2. A CEO who thinks from the end:
Yang Zhilin (31) previously worked at Meta AI, Google Brain, and Carnegie Mellon. His vision? Nothing less than AGI — still a rare ambition among Chinese AI labs.

3. A trillion-parameter model that’s surprisingly efficient:
Kimi K2 uses a mixture-of-experts architecture (32B active params per inference) and dominates on coding/math benchmarks.

4. The secret weapon: Muon optimizer:
A new training method that doubles efficiency, cuts memory in half, and ran 15.5T tokens with zero failures. Big implications.

Most importantly, their move from closed to open source signals a broader shift in China’s AI scene — following Baidu’s pivot. But as Yang puts it: “Users are the only real leaderboard.”

👇 Check out the full post to explore what Kimi K2 can do, how to try it, and why it matters for the future of open-source LLMs:
https://huggingface.co/blog/fdaudens/moonshot-ai-kimi-k2-explained

reacted to sergiopaniego's post with 🚀 6 months ago

Post

2028

Updated my HF Space for vibe testing smol VLMs on object detection, visual grounding, keypoint detection & counting! 👓

🆕 Compare Qwen2.5 VL 3B vs Moondream 2B side-by-side with annotated images & text outputs.

Try examples or test your own images! 🏃

📱Space: sergiopaniego/vlm_object_understanding

reacted to AdinaY's post with 🚀 6 months ago

Post

2017

The Chinese Open Source Heatmap is live 🔥
You can now track the companies/ research labs/ communities powering China’s open source AI movement.

zh-ai-community/model-release-heatmap-zh

Some highlights:

✨Giant Tech are investing more in open source.
-Alibaba: Full stack open ecosystem
-Tecent: Hunyuan image/video/3D
-Bytedance: Catching up fast in 2025
-Baidu: New player in open LLM

✨New players emerging post–DeepSeek moment.
-Xiaomi
-Red Note
-Bilibili
-MiniMax
-Moonshot AI

✨Startup list is shifting fast! Those who find a direction aligned with their strengths are the ones who endure.
-DeepSeek
-MiniMax
-StepFun
-Moonshot AI
-Zhipu AI
-OpenBMB

✨Research Lab & Community are making key contributions.
-BAAI
-Shanghai AI Lab
-OpenMOSS
-MAP

4 replies

·

replied to wewittc's post 6 months ago

I think it's nice that Zero GPU is something that provides so much to so many people across the hugging face community to build and use things that would otherwise be unavailable / unfair barriers of entry to do so at no cost to the average user.

zeroGPU has max execution time parameters with sensible defaults and ranges so it should not take to much quota even if left unchecked. Don't forget you will get more quota in no time at all.

reacted to merve's post with 🚀 7 months ago

Post

2400

y'all have been asking my opinion on how OCR models compare to each other 👀
I will leave three apps to compare newest models by @prithivMLmods instead ⤵️
> compare Nanonets-OCR-s, Qwen2-VL-OCR-2B-Instruct, RolmOCR, Aya-Vision prithivMLmods/Multimodal-OCR
> SmolDocling, Nanonets-OCR-s, MonkeyOCR, Typhoon-OCR-7B prithivMLmods/Multimodal-OCR2
> docscopeOCR, MonkeyOCR, coreOCR prithivMLmods/core-OCR

1 reply

·

reacted to cbensimon's post with 🔥 7 months ago

Post

4182

🚀 ZeroGPU now supports PyTorch native quantization via torchao

While it hasn’t been battle-tested yet, Int8WeightOnlyConfig is already working flawlessly in our tests.

Let us know if you run into any issues — and we’re excited to see what the community will build!

import spaces
from diffusers import FluxPipeline
from torchao.quantization.quant_api import Int8WeightOnlyConfig, quantize_

pipeline = FluxPipeline.from_pretrained(...).to('cuda')
quantize_(pipeline.transformer, Int8WeightOnlyConfig()) # Or any other component(s)

@spaces.GPU
def generate(prompt: str):
    return pipeline(prompt).images[0]

5 replies

·

Ci Splunk PRO

AI & ML interests

Recent Activity

Organizations

Ci Splunk PRO

AI & ML interests

Recent Activity

Organizations

Csplk's activity