benhaotang (Benhao Tang)

reacted to piercus's post with 👍 2 months ago

Post

3918

Starts erasing! 🎉 🎉 🎉
This is made with a one-step SD1.5 LBM [1] eraser !

Data is open. Data pipeline is open. Training code is open.
On our LBM fork : https://github.com/finegrain-ai/LBM

[1] LBM: Latent Bridge Matching for Fast Image-to-Image Translation (2503.07535)

1 reply

·

reacted to dhruv3006's post with 🚀 8 months ago

Post

3024

Lumier – Run macOS & Linux VMs in a Docker

Lumier is an open-source tool for running macOS virtual machines in Docker containers on Apple Silicon Macs.

When building virtualized environments for AI agents, we needed a reliable way to package and distribute macOS VMs. Inspired by projects like dockur/macos that made macOS running in Docker possible, we wanted to create something similar but optimized for Apple Silicon.

The existing solutions either didn't support M-series chips or relied on KVM/Intel emulation, which was slow and cumbersome. We realized we could leverage Apple's Virtualization Framework to create a much better experience.

Lumier takes a different approach: It uses Docker as a delivery mechanism (not for isolation) and connects to a lightweight virtualization service (lume) running on your Mac.

Lumier is 100% open-source under MIT license and part of C/ua.

Github : https://github.com/trycua/cua/tree/main/libs/lumier
Join the discussion here : https://discord.gg/fqrYJvNr4a

reacted to Jaward's post with 🧠 8 months ago

Post

3319

finally, a course that makes diffusion math much easier to grasp, well done 👍 https://diffusion.csail.mit.edu/

1 reply

·

reacted to abidlabs's post with 🔥 8 months ago

Post

5245

HOW TO ADD MCP SUPPORT TO ANY 🤗 SPACE

Gradio now supports MCP! If you want to convert an existing Space, like this one hexgrad/Kokoro-TTS, so that you can use it with Claude Desktop / Cursor / Cline / TinyAgents / or any LLM that supports MCP, here's all you need to do:

1. Duplicate the Space (in the Settings Tab)
2. Upgrade the Gradio sdk_version to 5.28 (in the README.md)
3. Set mcp_server=True in launch()
4. (Optionally) add docstrings to the function so that the LLM knows how to use it, like this:

def generate(text, speed=1):
    """
    Convert text to speech audio.

    Parameters:
        text (str): The input text to be converted to speech.
        speed (float, optional): Playback speed of the generated speech.

That's it! Now your LLM will be able to talk to you 🤯

reacted to fdaudens's post with 👍 8 months ago

Post

1887

Want to know which AI models are least likely to hallucinate — and how to keep yours from spiking hallucinations by 20%?

A new benchmark called Phare, by Giskard, tested leading models across multiple languages, revealing three key findings:

1️⃣ Popular models aren't necessarily factual. Some models ranking highest in user satisfaction benchmarks like LMArena are actually more prone to hallucination.

2️⃣ The way you ask matters - a lot. When users present claims confidently ("My teacher said..."), models are 15% less likely to correct misinformation vs. neutral framing ("I heard...").

3️⃣ Telling models to "be concise" can increase hallucination by up to 20%.

What's also cool is that the full dataset is public - use them to test your own models or dive deeper into the results! H/t @davidberenstein1957 for the link.

- Study: https://www.giskard.ai/knowledge/good-answers-are-not-necessarily-factual-answers-an-analysis-of-hallucination-in-leading-llms
- Leaderboard: https://phare.giskard.ai/
- Dataset: giskardai/phare

reacted to Reality123b's post with 👍 9 months ago

Post

2180

ok, there must be a problem. HF charged me 0.12$ for 3 inference requests to text models

8 replies

·

replied to Reality123b's post 9 months ago

look at someone being charged $300 here: https://old.reddit.com/r/huggingface/comments/1jkyj2a/huggingface_just_billed_me_300_on_top_of_the_9/
I am not saying that he used almost 400k request is not crazy, but before the price change it should be covered in 20k*30 per month, ig the OP has no idea about the price change and carried on his usage.
For my own experience I think many HF-inference API text model are at least 10x the price since last week, e.g. command-r-plus is even more expensive than using from cohere now
To run my own number, I only use text models, Feb I did 210 requests for 0.26, Mar I did 60 for 0.26 and then after price increase another 20 for 0.14... almost 0.01 per request now, mind you most of my request is below 1k tokens

reacted to Keltezaa's post with 🔥 9 months ago

Post

9842

Dear HF Staff and pro Users.

Why did you remove the "Regen" feature from the ZeroGPU feature?
Is this an error or intended?

I am now limited to 13 images per 24 hrs. Using my space.
When I upgraded to Pro, it was exclusively for the 5x more usage and the faster regen.

The reason I spend my hard earned money on your site was for this feature.
This is totally unacceptable.

########
Other Pro Users please reply and tag others
IF YOU AGREE or DISAGREE.
########
@Always-cheating ,@anonymous111110987654321 ,@Arshili @bedspirit @blackedguy @John6666 ,@DavidBaloches @E-07 ,@f-14 @mindfulpeoples @multimodalart

9 replies

·

replied to julien-c's post 9 months ago

This comment has been hidden

reacted to albertvillanova's post with 👍 10 months ago

Post

4191

🚀 New smolagents update: Safer Local Python Execution! 🦾🐍

With the latest release, we've added security checks to the local Python interpreter: every evaluation is now analyzed for dangerous builtins, modules, and functions. 🔒

Here's why this matters & what you need to know! 🧵👇

1️⃣ Why is local execution risky? ⚠️
AI agents that run arbitrary Python code can unintentionally (or maliciously) access system files, run unsafe commands, or exfiltrate data.

2️⃣ New Safety Layer in smolagents 🛡️
We now inspect every return value during execution:
✅ Allowed: Safe built-in types (e.g., numbers, strings, lists)
⛔ Blocked: Dangerous functions/modules (e.g., os.system, subprocess, exec, shutil)

3️⃣ Immediate Benefits 💡
- Prevent agents from accessing unsafe builtins
- Block unauthorized file or network access
- Reduce accidental security vulnerabilities

4️⃣ Security Disclaimer ⚠️
🚨 Despite these improvements, local Python execution is NEVER 100% safe. 🚨
If you need true isolation, use a remote sandboxed executor like Docker or E2B.

5️⃣ The Best Practice: Use Sandboxed Execution 🔐
For production-grade AI agents, we strongly recommend running code in a Docker or E2B sandbox to ensure complete isolation.

6️⃣ Upgrade Now & Stay Safe! 🚀
Check out the latest smolagents release and start building safer AI agents today.

🔗 https://github.com/huggingface/smolagents

What security measures do you take when running AI-generated code? Let’s discuss! 👇

#AI #smolagents #Python #Security

2 replies

·

replied to their post 11 months ago

OK, grok 3 deep research also failed on my benchmark...

And this is the final solution it gives me:

Use wsl --shutdown before hibernating; if it fails, try net stop LxssManager.

What? How about just tell me if WSL have problem, just do not using WSL... How can this be a solution when there is even an official troubleshooting guide that provide more solutions. This is the even worst than gemini and perplexity, at least they read the official guide, just got lost in github issue threads... Now I really want to know how OpenAI's compares to mine, if I have 200 dollars.

reacted to schuler's post with 😎 11 months ago

Post

3435

🔮 GPT-3 implemented in pure Free Pascal!
https://github.com/joaopauloschuler/gpt-3-for-pascal

This implementation follows the GPT-3 Small architecture from the landmark paper "Language Models are Few-Shot Learners":

┌─────────────────────────┐
│     Input Layer       │
├─────────────────────────┤
│ Token & Positional    │
│     Embedding         │
├─────────────────────────┤
│   12x Transformer     │
│      Blocks           │
│  - 12 heads           │
│  - 768 hidden dims    │
│  - 3072 intermediate  │
├─────────────────────────┤
│   Output Layer        │
└─────────────────────────┘

Clean Pascal Implementation

for CntLayer := 1 to {Layers=}12 do
begin
  Result.AddTransformerBlockCAI(
    {Heads=}12, 
    {intermediate dimensions=}4*768, 
    {NoForward=}true, 
    {HasNorm=}true, 
    false
  );
end;

reacted to louisbrulenaudet's post with 🤗 11 months ago

Post

3481

I am pleased to introduce my first project built upon Hugging Face’s smolagents framework, integrated with Alpaca for financial market analysis automation 🦙🤗

The project implements technical indicators such as the Relative Strength Index (RSI) and Bollinger Bands to provide momentum and volatility analysis. Market data is retrieved through the Alpaca API, enabling access to historical price information across various timeframes.

AI-powered insights are generated using Hugging Face’s inference API, facilitating the analysis of market trends through natural language processing with DuckDuckGo search integration for real-time sentiment analysis based on financial news 🦆

Link to the GitHub project: https://github.com/louisbrulenaudet/agentic-market-tool

posted an update 11 months ago

Post

2542

Try out my updated implementation of forked OpenDeepResearcher(link below) as an OpenAI compatible endpoint, but with full control, can be deployed completely free with Gemini api or completely locally with ollama, or pay-as-you-go in BYOK format, the AI agents will think dynamically based on the difficulties of given research, compatible with any OpenAI compatible configurable clients(Msty, Chatbox, even vscode AI Toolkit playground).

If you don't want to pay OpenAI $200 to use or want to take control of your deep research, check out here:
👉 https://github.com/benhaotang/OpenDeepResearcher-via-searxng

**Personal take**

Based on my testing against Perplexity's and Gemini's implementation with some Physics domain questions, mine is comparable and very competent at finding even the most rare articles or methods.

Also a funny benchmark of mine to test all these searching models, is to trouble shot a WSL2 hanging issue I experienced last year, with prompt:

> wsl2 in windows hangs in background with high vmmem cpu usage once in a while, especially after hibernation, no error logs captured in linux, also unable to shutdown in powershell, provide solutions

the final solution that took me a day last year to find is to patch the kernel with some steps documented in carlfriedrich's repo and wait Microsoft to solve it(it is buried deep in wsl issues). Out of the three, only my Deep Research agent has found this solution, Perplexity and Gemini just focus on other force restart or memory management methods. I am very impressed with how it has this kind of obscure and scarce trouble shooting ability.

**Limitations**

Some caveats to be done later:
- Multi-turn conversation is not yet supported, so no follow-up questions
- System message is only extra writing instructions, don't affect on search
- Small local model may have trouble citing source reliably, I am working on a fix to fact check all citation claims

1 reply

·

reacted to mitkox's post with 👍 11 months ago

Post

3777

llama.cpp is 26.8% faster than ollama.
I have upgraded both, and using the same settings, I am running the same DeepSeek R1 Distill 1.5B on the same hardware. It's an Apples to Apples comparison.

Total duration:
llama.cpp 6.85 sec <- 26.8% faster
ollama 8.69 sec

Breakdown by phase:
Model loading
llama.cpp 241 ms <- 2x faster
ollama 553 ms

Prompt processing
llama.cpp 416.04 tokens/s with an eval time 45.67 ms <- 10x faster
ollama 42.17 tokens/s with an eval time of 498 ms

Token generation
llama.cpp 137.79 tokens/s with an eval time 6.62 sec <- 13% faster
ollama 122.07 tokens/s with an eval time 7.64 sec

llama.cpp is LLM inference in C/C++; ollama adds abstraction layers and marketing.

Make sure you own your AI. AI in the cloud is not aligned with you; it's aligned with the company that owns it.

7 replies

·

reacted to sometimesanotion's post with 🔥 11 months ago

Post

2764

I've managed a #1 score of 41.22% average for 14B parameter models on the Open LLM Leaderboard. As of this writing, sometimesanotion/Lamarck-14B-v0.7 is #8 for all models up to 70B parameters.

It took a custom toolchain around Arcee AI's mergekit to manage the complex merges, gradients, and LoRAs required to make this happen. I really like seeing features of many quality finetunes in one solid generalist model.

8 replies

·

reacted to aaditya's post with 👍❤️ about 1 year ago

Post

5319

Last Week in Medical AI: Top Research Papers/Models 🔥
🏅 (December 15 – December 21, 2024)

Medical LLM & Other Models
- MedMax: Mixed-Modal Biomedical Assistant
- Advanced multimodal instruction tuning
- Enhanced biomedical knowledge integration
- Comprehensive assistant capabilities
- MGH Radiology Llama 70B
- Specialized radiology focus
- State-of-the-art performance
- Enhanced report generation capabilities
- HC-LLM: Historical Radiology Reports
- Context-aware report generation
- Historical data integration
- Improved accuracy in diagnostics

Frameworks & Methods
- ReflecTool: Reflection-Aware Clinical Agents
- Process-Supervised Clinical Notes
- Federated Learning with RAG
- Query Pipeline Optimization

Benchmarks & Evaluations
- Multi-OphthaLingua
- Multilingual ophthalmology benchmark
- Focus on LMICs healthcare
- Bias assessment framework
- ACE-M3 Evaluation Framework
- Multimodal medical model testing
- Comprehensive capability assessment
- Standardized evaluation metrics

LLM Applications
- Patient-Friendly Video Reports
- Medical Video QA Systems
- Gene Ontology Annotation
- Healthcare Recommendations

Special Focus: Medical Ethics & AI
- Clinical Trust Impact Study
- Mental Health AI Challenges
- Hospital Monitoring Ethics
- Radiology AI Integration

Now you can watch and listen to the latest Medical AI papers daily on our YouTube and Spotify channels as well!

- Full thread in detail:
https://x.com/OpenlifesciAI/status/1870504774162063760
- Youtube Link: youtu.be/SbFp4fnuxjo
- Spotify: https://t.co/QPmdrXuWP9