Hugging Face Fellows

non-profit

AI & ML interests

The Fellowship is a network of exceptional people from different backgrounds who contribute to open-source machine learning πŸ§™β€β™‚οΈπŸ¦Έβ€β™€οΈπŸ¦ΉπŸ§β€β™‚οΈ

Recent Activity

hugging-fellows's activity

merveΒ 
posted an update 1 day ago
merveΒ 
posted an update 7 days ago
view post
Post
2325
Aya by Cohere For AI can now see! πŸ‘€

C4AI community has built Maya 8B, a new open-source multilingual VLM built on SigLIP and Aya 8B 🌱 works on 8 languages! πŸ—£οΈ

The authors extend Llava dataset using Aya's translation capabilities with 558k examples!
ry it here kkr5155/maya_demo

Dataset maya-multimodal/pretrain

Model maya-multimodal/maya πŸ‘
kudos @nahidalam and team
  • 1 reply
Β·
clemΒ 
posted an update 8 days ago
view post
Post
1526
Coming back to Paris Friday to open our new Hugging Face office!

We're at capacity for the party but add your name in the waiting list as we're trying to privatize the passage du Caire for extra space for robots πŸ€–πŸ¦ΎπŸ¦Ώ

https://t.co/enkFXjWndJ
  • 1 reply
Β·
merveΒ 
posted an update 8 days ago
view post
Post
2905
Apollo is a new family of open-source video language models by Meta, where 3B model outperforms most 7B models and 7B outperforms most 30B models 🧢

✨ the models come in 1.5B https://huggingface.co./Apollo-LMMs/Apollo-1_5B-t32, 3B https://huggingface.co./Apollo-LMMs/Apollo-3B-t32 and 7B https://huggingface.co./Apollo-LMMs/Apollo-7B-t32 with A2.0 license, based on Qwen1.5 & Qwen2
✨ the authors also release a benchmark dataset https://huggingface.co./spaces/Apollo-LMMs/ApolloBench

The paper has a lot of experiments (they trained 84 models!) about what makes the video LMs work ⏯️

Try the demo for best setup here https://huggingface.co./spaces/Apollo-LMMs/Apollo-3B
they evaluate sampling strategies, scaling laws for models and datasets, video representation and more!
> The authors find out that whatever design decision was applied to small models also scale properly when the model and dataset are scaled πŸ“ˆ scaling dataset has diminishing returns for smaller models
> They evaluate frame sampling strategies, and find that FPS sampling is better than uniform sampling, and they find 8-32 tokens per frame optimal
> They also compare image encoders, they try a variation of models from shape optimized SigLIP to DINOv2
they find google/siglip-so400m-patch14-384 to be most powerful πŸ”₯
> they also compare freezing different parts of models, training all stages with some frozen parts give the best yield

They eventually release three models, where Apollo-3B outperforms most 7B models and Apollo 7B outperforms 30B models πŸ”₯
  • 3 replies
Β·
merveΒ 
posted an update 13 days ago
view post
Post
1669
A complete RAG pipeline includes a reranker, which ranks the documents to find the best document πŸ““
Same goes for multimodal RAG, multimodal rerankers which we can integrate to multimodal RAG pipelines!
Learn how to build a complete multimodal RAG pipeline with vidore/colqwen2-v1.0 as retriever, lightonai/MonoQwen2-VL-v0.1 as reranker, Qwen/Qwen2-VL-7B-Instruct as VLM in this notebook that runs on a GPU as small as L4 πŸ”₯ https://huggingface.co./learn/cookbook/multimodal_rag_using_document_retrieval_and_reranker_and_vlms
christopherΒ 
posted an update 17 days ago
view post
Post
1563
The folks at Foursquare released a dataset of 104.5 million places of interest ( foursquare/fsq-os-places) and here's all of them on a plot
Β·
merveΒ 
posted an update 17 days ago
view post
Post
5505
This week in open-source AI was insane 🀠 A small recapπŸ•ΊπŸ» merve/dec-6-releases-67545caebe9fc4776faac0a3

Multimodal πŸ–ΌοΈ
> Google shipped a PaliGemma 2, new iteration of PaliGemma with more sizes: 3B, 10B and 28B, with pre-trained and captioning variants πŸ‘
> OpenGVLab released InternVL2, seven new vision LMs in different sizes, with sota checkpoint with MIT license ✨
> Qwen team at Alibaba released the base models of Qwen2VL models with 2B, 7B and 72B ckpts

LLMs πŸ’¬
> Meta released a new iteration of Llama 70B, Llama3.2-70B trained further
> EuroLLM-9B-Instruct is a new multilingual LLM for European languages with Apache 2.0 license πŸ”₯
> Dataset: CohereForAI released GlobalMMLU, multilingual version of MMLU with 42 languages with Apache 2.0 license
> Dataset: QwQ-LongCoT-130K is a new dataset to train reasoning models
> Dataset: FineWeb2 just landed with multilinguality update! πŸ”₯ nearly 8TB pretraining data in many languages!

Image/Video Generation πŸ–ΌοΈ
> Tencent released HunyuanVideo, a new photorealistic video generation model
> OminiControl is a new editing/control framework for image generation models like Flux

Audio πŸ”Š
> Indic-Parler-TTS is a new text2speech model made by community
merveΒ 
posted an update 18 days ago
view post
Post
1500
New InternVL drop with a state-of-the-art 78B vision language model with MIT license πŸ”₯ https://huggingface.co./collections/OpenGVLab/internvl-25-673e1019b66e2218f68d7c1c
The release comes with seven new vision LMs based on InternViT 300M/6B and Qwen2.5 (0.5B, 3B, 32B, 72B) and InternLM2 (8B, 7B, 20B) in different sizes
78B model is of InternViT 6B and Qwen2.5-72B Instruct, can accomplish variety of tasks πŸ‘ Try here OpenGVLab/InternVL
christopherΒ 
posted an update 19 days ago
BramVanroyΒ 
posted an update 19 days ago
view post
Post
440
In the spirit of "Better late than never", I've finally written a brief overview paper for GEITje 7B Ultra. Initially released 10 months ago (oops), but still reaching around 1300 monthly downloads across the HF ecosystem (not including ollama).

GEITje 7B Ultra: A Conversational Model for Dutch (2412.04092)

While the paper discusses the model a little bit, I especially wanted to write about the datasets, which to this day seem an important asset for Dutch LLM training (SFT and preference tuning). We have a long way to go for Dutch, but publishing transparent and reproducible artefacts seems an important step to me, alongside having open discussions about data, bias, architectures.

In that spirit, thanks are in order for the creation of GEITje 7B Ultra and all related datasets:

- Michiel Buisman and UWV for providing the means to create the datasets
- Flemish Supercomputer Center (VSC) for the compute
- The Hugging Face Fellows and rest of the team for their discussions and insights
- The Dutch NLP community, notably @Rijgersberg for building the base GEITje model and the fruitful discussions we've had

More to come, step by step!

BramVanroy/geitje-7b-ultra-65c1ee010ad80fd1f6a8f208
ariG23498Β 
posted an update 20 days ago
merveΒ 
posted an update 23 days ago
view post
Post
2628
small but mighty πŸ”₯
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 πŸ’ https://github.com/huggingface/smollm/blob/main/finetuning/Smol_VLM_FT.ipynb
clemΒ 
posted an update 23 days ago
view post
Post
4090
Six predictions for AI in 2025 (and a review of how my 2024 predictions turned out):

- There will be the first major public protest related to AI
- A big company will see its market cap divided by two or more because of AI
- At least 100,000 personal AI robots will be pre-ordered
- China will start to lead the AI race (as a consequence of leading the open-source AI race).
- There will be big breakthroughs in AI for biology and chemistry.
- We will begin to see the economic and employment growth potential of AI, with 15M AI builders on Hugging Face.

How my predictions for 2024 turned out:

- A hyped AI company will go bankrupt or get acquired for a ridiculously low price
βœ… (Inflexion, AdeptAI,...)

- Open-source LLMs will reach the level of the best closed-source LLMs
βœ… with QwQ and dozens of others

- Big breakthroughs in AI for video, time-series, biology and chemistry
βœ… for video πŸ”΄for time-series, biology and chemistry

- We will talk much more about the cost (monetary and environmental) of AI
βœ…Monetary πŸ”΄Environmental (😒)

- A popular media will be mostly AI-generated
βœ… with NotebookLM by Google

- 10 millions AI builders on Hugging Face leading to no increase of unemployment
πŸ”œcurrently 7M of AI builders on Hugging Face
Β·
merveΒ 
posted an update 23 days ago
view post
Post
2864
Last week we were blessed with open-source models! A recap πŸ’
merve/nov-29-releases-674ccc255a57baf97b1e2d31

πŸ–ΌοΈ Multimodal
> At Hugging Face we released SmolVLM, a performant and efficient smol vision language model πŸ’—
> Show Lab released ShowUI-2B: new vision-language-action model to build GUI/web automation agents πŸ€–
> Rhymes AI has released the base model of Aria: Aria-Base-64K and Aria-Base-8K with their respective context length
> ViDoRe team released ColSmolVLM: A new ColPali-like retrieval model based on SmolVLM
> Dataset: Llava-CoT-o1-Instruct: new dataset labelled using Llava-CoT multimodal reasoning modelπŸ“–
> Dataset: LLaVA-CoT-100k dataset used to train Llava-CoT released by creators of Llava-CoT πŸ“•

πŸ’¬ LLMs
> Qwen team released QwQ-32B-Preview, state-of-the-art open-source reasoning model, broke the internet πŸ”₯
> AliBaba has released Marco-o1, a new open-source reasoning model πŸ’₯
> NVIDIA released Hymba 1.5B Base and Instruct, the new state-of-the-art SLMs with hybrid architecture (Mamba + transformer)

⏯️ Image/Video Generation
> Qwen2VL-Flux: new image generation model based on Qwen2VL image encoder, T5 and Flux for generation
> Lightricks released LTX-Video, a new DiT-based video generation model that can generate 24 FPS videos at 768x512 res ⏯️
> Dataset: Image Preferences is a new image generation preference dataset made with DIBT community effort of Argilla 🏷️

Audio
> OuteAI released OuteTTS-0.2-500M new multilingual text-to-speech model based on Qwen-2.5-0.5B trained on 5B audio prompt tokens
clemΒ 
posted an update 25 days ago
view post
Post
4350
Hugging Face is becoming the best place to share the most viral AI apps with spaces.

Kolors Virtual Try-on just crossed 6,000,000 unique visitors & is now the #5 most popular space. Congrats to the Kwai Kolors team!

Kwai-Kolors/Kolors-Virtual-Try-On
  • 2 replies
Β·