sergiopaniego (Sergio Paniego)

posted an update 3 days ago

Post

213

Did you miss this? 👓

🧙‍♂️vLLM + transformers integration just got upgraded with direct VLM support.

Select a VLM + model_impl=transformers and play via vLLM!

replied to their post 4 days ago

Currently, not in the todo list but could be a nice addition! We also accept external contributions :)

posted an update 4 days ago

Post

2483

We just released TRL v0.20 with major multimodal upgrades!

👁️ VLM support for GRPO (highly requested by the community!)
🎞️ New GSPO trainer (from @Qwen , released last week, VLM-ready)
🐙 New MPO trainer (multimodal by design, as in the paper)

📝 Full release notes here: https://github.com/huggingface/trl/releases/tag/v0.20.0

posted an update 10 days ago

Post

1151

Yet Another New Multimodal Fine-Tuning Recipe 🥧

🧑‍🍳 In this @HuggingFace Face Cookbook notebook, we demonstrate how to align a multimodal model (VLM) using Mixed Preference Optimization (MPO) using trl.

💡 This recipe is powered by the new MPO support in trl, enabled through a recent upgrade to the DPO trainer!

We align the multimodal model using multiple optimization objectives (losses), guided by a preference dataset (chosen vs. rejected multimodal pairs).

Check it out! ➡️ https://huggingface.co/learn/cookbook/fine_tuning_vlm_mpo

2 replies

·

posted an update 15 days ago

Post

1637

🧑‍🍳 New Multimodal Fine-Tuning Recipe 🧑‍🍳

⚡️ In this new @huggingface Cookbook recipe, I walk you though the process of fine tuning a Visual Language Model (VLM) for Object Detection with Visual Grounding, using TRL.

🔍 Object detection typically involves detecting categories in images (e.g., vase).

By combining it with visual grounding, we add contextual understanding so instead of detecting just "vase", we can detect "middle vase" in an image.

VLMs are super powerful!

In this case, I use PaliGemma 2 which already supports object detection and extend it to also add visual grounding.

🤗 Check it out here: https://huggingface.co/learn/cookbook/fine_tuning_vlm_object_detection_grounding

posted an update 16 days ago

Post

1604

Multiple NEW notebooks and scripts added to the Hugging Face Gemma recipes repo!

Thanks to the community 🫶, we're adding more and more recipes using Gemma 💎

Fine tuning for all modalities, function calling, RAG...

Repo: https://github.com/huggingface/huggingface-gemma-recipes

We're also open to new ideas from the community 🤗!

1 reply

·

posted an update 19 days ago

Post

383

Loved this paper! ♥️

Authors benchmark multimodal models on vision tasks (detection, segmentation...) using clever prompting tricks.

📄 Results: VLMs are solid generalists but still lag behind SOTA task-specific models — especially on geometric tasks vs. semantic ones.

paper: How Well Does GPT-4o Understand Vision? Evaluating Multimodal Foundation Models on Standard Computer Vision Tasks (2507.01955)

posted an update 19 days ago

Post

255

You can already play with two of the latest most impressive models on HF via @novita-ai as Inference Provider 🚨

🌌 Kimi K2: 1T params model, MoE beast for coding, reasoning and agentic tasks
🔮 GLM-4.1V-9B-Thinking: VLM + deep reasoning model

Kimi K2: moonshotai/Kimi-K2-Instruct
GLM-4.1V-9B-Thinking: https://huggingface.co/THUDM/GLM-4.1V-9B-Thinking

posted an update 19 days ago

Post

219

Over 1K already on @huggingface !!

posted an update 24 days ago

Post

1584

Test SmolLM3, the newest fully open model released by @HuggingFaceTB !

It's smol (3B), multilingual (6 languages), comes with dual mode reasoning (think/no_think modes) and supports long-context (128k).

Try it now in the notebook below!! ⬇️

Colab notebook: https://colab.research.google.com/github/sergiopaniego/samples/blob/main/smollm3_3b_inference.ipynb
notebook: https://github.com/sergiopaniego/samples/blob/main/smollm3_3b_inference.ipynb
blog: https://huggingface.co/blog/smollm3

posted an update 30 days ago

Post

1990

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

posted an update about 1 month ago

Post

1059

📣 CALL FOR CONTRIBUTORS! 📣

Following last week’s full release of Gemma 3n, we launched a dedicated recipes repo to explore and share use cases. We already added some! 🧑‍🍳

Now we’re inviting the community to contribute and showcase how these models shine! ✨

Let them cook.

Check it out: https://github.com/huggingface/huggingface-gemma-recipes/issues/4

1 reply

·

posted an update about 1 month ago

Post

469

One of my favorite perks of the Hugging Face Pro plan: ✨Dev mode✨

Connect your HF Space to VS Code and just build — with hot reload out of the box.

Game changer for fast prototyping. 💻

Google Colab made AI accessible. Now HF Spaces are doing it too! 😍

💡 New Hugging Face pricing: http://hf.co/pricing
💡 More details: https://huggingface.co/learn/cookbook/en/enterprise_cookbook_dev_spaces

reacted to merve's post with ❤️🚀 about 1 month ago

Post

3644

Releases of the past week are here merve/releases-june-13-6852c3c1eaf1e0c24c958860

Here's our picks 🤓
So many interesting models released past week in open AI! 🤖

🖼️ Computer Vision/VLMs
> nanonets/Nanonets-OCR-s is the new state-of-the-art OCR model that can handle checkboxes, watermarks, tables (OS)
> Meta released facebook/v-jepa-2-6841bad8413014e185b497a6, new sota video embeddings with two new classification models (OS)
> ByteDance-Seed/SeedVR2-3B is a new 3B video restoration model (OS)

Audio
> Stepfun released stepfun-ai/Step-Audio-AQAA, new large (137B 🤯) audio language model that takes in audio and generates audio (OS)

🤖 Robotics
> nvidia released nvidia/GR00T-N1.5-3B, new open foundation vision language action model

3D
> tencent/Hunyuan3D-2.1 is the new version of Hunyuan by Tencent that can generate 3D assets from text and image prompts

reacted to cbensimon's post with 🔥 3 months ago

Post

5908

🚀 ZeroGPU medium size is now available as a power-user feature

Nothing too fancy for now—ZeroGPU Spaces still default to large (70GB VRAM)—but this paves the way for:
- 💰 size-based quotas / pricing (medium will offer significantly more usage than large)
- 🦣 the upcoming xlarge size (141GB VRAM)

You can as of now control GPU size via a Space variable. Accepted values:
- auto (future default)
- medium
- large (current default)

The auto mode checks total CUDA tensor size during startup:
- More than 30GB → large
- Otherwise → medium

3 replies

·

reacted to merve's post with 🔥 4 months ago

Post

4504

sooo many open AI releases past week, let's summarize! 🤗
merve/april-11-releases-67fcd78be33d241c0977b9d2

multimodal
> Moonshot AI released Kimi VL Thinking, first working open-source multimodal reasoning model and Kimi VL Instruct, both 16B MoEs with 3B active params (OS)
> InternVL3 released based on Qwen2.5VL, 7 ckpts with various sizes (1B to 78B)

LLMs
> NVIDIA released Llama-3_1-Nemotron-Ultra-253B-v1 an LLM built on Llama 405B for reasoning, chat and tool use
> Agentica released DeepCoder-14B-Preview, fine-tuned version of DeepSeek-R1-Distilled-Qwen-14B on problem-test pairs, along with the compiled dataset
> Zyphra/ZR1-1.5B is a new small reasoning LLM built on R1-Distill-1.5B (OS)
> Skywork-OR1-32B-Preview is a new reasoning model by Skywork

Image Generation
> HiDream releases three new models, HiDream I1 Dev, I1 Full, and I1 fast for image generation (OS)

*OS ones have Apache 2.0 or MIT licenses

4 replies

·

reacted to m-ric's post with 🤗 4 months ago

Post

5193

smolagents now support vLLM! 🥳

As one of the most popular local inference solutions, the community had been asking us to integrate vLLM: after a heavy refactoring of our LLM classes, we've just released smolagents 1.11.0, with a brand new VLLMModel class.

Go try it and tell us what you think!

https://github.com/huggingface/smolagents/blob/45b2c86857b7f7657daaa74e4d17d347e9e2c4a4/src/smolagents/models.py#L497

reacted to burtenshaw's post with 🔥 5 months ago

Post

6487

Now the Hugging Face agent course is getting real! With frameworks like smolagents, LlamaIndex, and LangChain.

🔗 Follow the org for updates

agents-course

This week we are releasing the first framework unit in the course and it’s on smolagents. This is what the unit covers:

- why should you use smolagents vs another library?
- how to build agents that use code
- build multiagents systems
- use vision language models for browser use

The team has been working flat out on this for a few weeks. Led by @sergiopaniego and supported by smolagents author @m-ric .

reacted to merve's post with 🚀 8 months ago

Post

2698

small but mighty 🔥
you can fine-tune SmolVLM on an L4 with batch size of 4 and it will only take 16.4 GB VRAM 🫰🏻 also with gradient accumulation simulated batch size is 16 ✨
I made a notebook that includes all the goodies: QLoRA, gradient accumulation, gradient checkpointing with explanations on how they work 💝 https://github.com/huggingface/smollm/blob/main/finetuning/Smol_VLM_FT.ipynb

Sergio Paniego PRO

AI & ML interests

Recent Activity

Organizations

Sergio Paniego PRO

AI & ML interests

Recent Activity

Organizations

sergiopaniego's activity