Post
266
New interactive viz from AI World showing OpenAI's new open model gpt-oss-120b breaking into the top 50 most liked models of all time on the Hub in under a day! ☄️☄️☄️
AI & ML interests
Exploring smol models (for text, vision and video) and high quality web and synthetic datasets
Recent Activity
View all activity
Prevent Thinking
1
#35 opened 1 day ago
by
somya13
BrigitteTousi
posted
an
update
2 days ago
Post
1921
GPT-4.1-mini level model right in your iPhone 🤯
openbmb/MiniCPM-V-4 is only 4B while surpassing GPT-4.1-mini in vision benchmarks 🔥
allows commercial use as well!
openbmb/MiniCPM-V-4 is only 4B while surpassing GPT-4.1-mini in vision benchmarks 🔥
allows commercial use as well!
Post
2492
The next generation of AI-powered websites is going to be WILD! 🤯
In-browser tool calling & MCP is finally here, allowing LLMs to interact with websites programmatically.
To show what's possible, I built a demo using Liquid AI's new LFM2 model, powered by 🤗 Transformers.js: LiquidAI/LFM2-WebGPU
As always, the demo is open source (which you can find under the "Files" tab), so I'm excited to see how the community builds upon this! 🚀
In-browser tool calling & MCP is finally here, allowing LLMs to interact with websites programmatically.
To show what's possible, I built a demo using Liquid AI's new LFM2 model, powered by 🤗 Transformers.js: LiquidAI/LFM2-WebGPU
As always, the demo is open source (which you can find under the "Files" tab), so I'm excited to see how the community builds upon this! 🚀
- 1 reply
Post
2457
Well, it took just 2 hours for
openai/gpt-oss-120b to hit #1 on Hugging Face. Don’t remember seeing anything rise that fast!
- 1 reply
Post
898
we're all sleeping on this OCR model
rednote-hilab/dots.ocr 🔥
dots.ocr is a new 3B model with sota performance, support for 100 languages & allowing commercial use! 🤯
single e2e model to extract image, convert tables, formula, and more into markdown 📝
try it MohamedRashad/Dots-OCR
dots.ocr is a new 3B model with sota performance, support for 100 languages & allowing commercial use! 🤯
single e2e model to extract image, convert tables, formula, and more into markdown 📝
try it MohamedRashad/Dots-OCR
Post
510
massive releases and tons of Flux 1. Krea LoRas past week!
here's some of the picks, find more models in collection 🫡 merve/releases-august-2-6890c14248203522b7d0267f
LLMs 💬
> Tencent dropped tencent/Hunyuan-7B-Instruct
> Qwen released Qwen/Qwen3-Coder-30B-A3B-Instruct, 30B MoE with 3B params for coding (OS)
vision/multimodal
> RedNote released rednote-hilab/dots.ocr - 3B OCR model (OS)
> Cohere released CohereLabs/command-a-vision-07-2025 - 112B (dense!) VLM for 6 languages
> StepFun-AI shipped stepfun-ai/step3 - 321B MoE VLM (OS)
> Skywork shipped Skywork/Skywork-UniPic-1.5B - new any-to-any model (image+text → image+text) (OS)
here's some of the picks, find more models in collection 🫡 merve/releases-august-2-6890c14248203522b7d0267f
LLMs 💬
> Tencent dropped tencent/Hunyuan-7B-Instruct
> Qwen released Qwen/Qwen3-Coder-30B-A3B-Instruct, 30B MoE with 3B params for coding (OS)
vision/multimodal
> RedNote released rednote-hilab/dots.ocr - 3B OCR model (OS)
> Cohere released CohereLabs/command-a-vision-07-2025 - 112B (dense!) VLM for 6 languages
> StepFun-AI shipped stepfun-ai/step3 - 321B MoE VLM (OS)
> Skywork shipped Skywork/Skywork-UniPic-1.5B - new any-to-any model (image+text → image+text) (OS)
Post
2104
Cohere just dropped
CohereLabs/command-a-vision-07-2025, a 112B (dense!) vision LM
> based on SigLIP2 & Command-A
> built for enterprise use cases 🔥
> use with Inference Providers or transformers 🤗
read their blog https://huggingface.co/blog/CohereLabs/introducing-command-a-vision-07-2025
> based on SigLIP2 & Command-A
> built for enterprise use cases 🔥
> use with Inference Providers or transformers 🤗
read their blog https://huggingface.co/blog/CohereLabs/introducing-command-a-vision-07-2025
- 2 replies
Post
3519
past week in open AI was insane 🔥 here's some of picks, find more here
merve/releases-july-25-688768ca47fe3693407e02d1
💬 LLMs & VLMs
> Qwen/Qwen3-235B-A22B-Thinking-2507 had a new update (OS)
> Qwen/Qwen3-Coder-480B-A35B-Instruct is out with 480B total 35B active params 🤯 (OS)
> AllenAI dropped an update to allenai/olmOCR-7B-0725 📝
> InternLM released internlm/Intern-S1 - 235B Qwen3 MoE + 6B InternViT encoder (OS)
> OmniSVG/OmniSVG is a new SVG generation VLM (OS)
🖼️ image/video/3D generation
> WanAI released Wan2.2 series - both T2V and I2V 14B models for high-quality video generation (OS) multimodalart/wan-22-688767e313337b434ed55112
> Tencent dropped tencent/HunyuanWorld-1 - image-to-3D scene generation model
💬 LLMs & VLMs
> Qwen/Qwen3-235B-A22B-Thinking-2507 had a new update (OS)
> Qwen/Qwen3-Coder-480B-A35B-Instruct is out with 480B total 35B active params 🤯 (OS)
> AllenAI dropped an update to allenai/olmOCR-7B-0725 📝
> InternLM released internlm/Intern-S1 - 235B Qwen3 MoE + 6B InternViT encoder (OS)
> OmniSVG/OmniSVG is a new SVG generation VLM (OS)
🖼️ image/video/3D generation
> WanAI released Wan2.2 series - both T2V and I2V 14B models for high-quality video generation (OS) multimodalart/wan-22-688767e313337b434ed55112
> Tencent dropped tencent/HunyuanWorld-1 - image-to-3D scene generation model
- 1 reply
Post
4320
🤯 241B VLM with apache-2.0 license
internlm/Intern-S1
internlm released Intern-S1: multimodal reasoning model based on 235B MoE Qwen3 and 6B InternViT 😍
benchmarks look great (👑 best model ✅ best open model)
internlm released Intern-S1: multimodal reasoning model based on 235B MoE Qwen3 and 6B InternViT 😍
benchmarks look great (👑 best model ✅ best open model)
BrigitteTousi
posted
an
update
15 days ago
Post
508
This is what Hugging Face is all about. We want everyone, hobbyists, researchers and industry alike, to be able to contribute to AI because everyone is affected by it. Kudos to HF's
@irenesolaiman
for spreading the word!🔥🤗
Post
2841
Introducing Voxtral WebGPU: State-of-the-art audio transcription directly in your browser! 🤯
🗣️ Transcribe videos, meeting notes, songs and more
🔐 Runs on-device, meaning no data is sent to a server
🌎 Multilingual (8 languages)
🤗 Completely free (forever) & open source
That's right, we're running Mistral's new Voxtral-Mini-3B model 100% locally in-browser on WebGPU, powered by Transformers.js and ONNX Runtime Web! 🔥
Try it out yourself! 👇
webml-community/Voxtral-WebGPU
🗣️ Transcribe videos, meeting notes, songs and more
🔐 Runs on-device, meaning no data is sent to a server
🌎 Multilingual (8 languages)
🤗 Completely free (forever) & open source
That's right, we're running Mistral's new Voxtral-Mini-3B model 100% locally in-browser on WebGPU, powered by Transformers.js and ONNX Runtime Web! 🔥
Try it out yourself! 👇
webml-community/Voxtral-WebGPU
Post
2724
Many VLMs claim to process hours of video. But can they follow the story?🤔
Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!⏳
We test three skills that matter for real-world use:
🔎 Localized Retrieval: Find a specific action.
🧩 Information Synthesis: Piece together scattered clues.
🏃 Fine-Grained Perception: Analyze detailed motion (e.g., count how many times a person swings an axe).
The results are in, and they're revealing. Only Gemini 2.5 pro handles 1-hour-long videos.
Performance drops sharply with duration, proving that long video understanding is still challenging. We've found the breaking points—now the community can start fixing them.📈
Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI.
📖 Blog:
https://huggingface.co/blog/timescope-video-lmm-benchmark
👩💻 Leaderboard & Demo: Apollo-LMMs/TimeScope
📊 Dataset: Apollo-LMMs/TimeScope
⚙️ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!⏳
We test three skills that matter for real-world use:
🔎 Localized Retrieval: Find a specific action.
🧩 Information Synthesis: Piece together scattered clues.
🏃 Fine-Grained Perception: Analyze detailed motion (e.g., count how many times a person swings an axe).
The results are in, and they're revealing. Only Gemini 2.5 pro handles 1-hour-long videos.
Performance drops sharply with duration, proving that long video understanding is still challenging. We've found the breaking points—now the community can start fixing them.📈
Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI.
📖 Blog:
https://huggingface.co/blog/timescope-video-lmm-benchmark
👩💻 Leaderboard & Demo: Apollo-LMMs/TimeScope
📊 Dataset: Apollo-LMMs/TimeScope
⚙️ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
Post
782
so many open LLMs and image LoRAs dropped past week, here's some picks for you 🫡
merve/releases-july-18-687e3fbd2ab9b39c51f9238b
LLMs
> ByteDance released a bunch of translation models called Seed-X-RM (7B) ByteDance-Seed/Seed-X-RM-7B
> NVIDIA released reasoning models of which 32B surpassing the giant Qwen3-235B with cc-by-4.0 license 👏 nvidia/openreasoning-nemotron-687730dae0170059860f1f01
> LG released a new EXAONE model (32B) LGAI-EXAONE/EXAONE-4.0-32B
VLMs/any-to-any
> vidore/colqwen-omni-v0.1 is a new any-to-any retriever (MIT)
> HiDream-ai/HiDream-E1-1 is image+text in image+text out model (MIT)
LoRAs
> There's a bunch of LoRAs based on Flux Kontext, gotta check out the collection 🤠
LLMs
> ByteDance released a bunch of translation models called Seed-X-RM (7B) ByteDance-Seed/Seed-X-RM-7B
> NVIDIA released reasoning models of which 32B surpassing the giant Qwen3-235B with cc-by-4.0 license 👏 nvidia/openreasoning-nemotron-687730dae0170059860f1f01
> LG released a new EXAONE model (32B) LGAI-EXAONE/EXAONE-4.0-32B
VLMs/any-to-any
> vidore/colqwen-omni-v0.1 is a new any-to-any retriever (MIT)
> HiDream-ai/HiDream-E1-1 is image+text in image+text out model (MIT)
LoRAs
> There's a bunch of LoRAs based on Flux Kontext, gotta check out the collection 🤠
Post
4535
Kimi K2 tech report is full of gems as always. Here are my notes on it:
> MuonClip: Pretty crazy how after 70k the training stabilizes and the QK-clip is basically inactive. There is also no loss in perf with QK-clip which is not trivial at all (at small scale but with aggressive threshold). Also a cool explanation of why muon makes the logit explode in appendix E (tl;dr is that muon makes the singular value of the update matrix higher)
> Sparsity scaling laws to justify their ratio, they have a very solid training infra that allows the model to be trained at this sparsity level, they could have increased even more but as sparsity increases the training becomes less efficient.
> They diminish the number of attention heads to make it more efficient for long context since attention heads are a big bottleneck for long context. They also remove 2 of the 3 "first dense" layers in the dsv3 arch.
With the sparsity and attention heads (divided by 2) they achieve 83% increased flops compared to deepseek v3 arch at 128k.
> Data: Rephrasing is KEY. They do a lot more synthetic data generation and rephrase their corpus to have different styles, for longer documents they do it by chunk. I'm (half) surprised by the fact that ONLY 1 epoch (assuming same number of training tokens I think?) of data rephrased 10 times has better accuracy than 10 epochs of the same data rephrased once.
> They do rewriting for Math and Knowledge, for Math they apply the ShallowMath recipe and instruct the model to rephrase in a "learning note" style
> They talk about diversity and probably have some internal stuff/eval to test that, as always still a bit unclear for me how to properly measure that.
The infra is also very nice, quick summary:
> PP=16 (1F1B schedule, a bit custom), EP=16, zero1
> No FP8 computation but for storage of specific layers, selective recomputation for inexpensive block, activation offloading to CPU
> MuonClip: Pretty crazy how after 70k the training stabilizes and the QK-clip is basically inactive. There is also no loss in perf with QK-clip which is not trivial at all (at small scale but with aggressive threshold). Also a cool explanation of why muon makes the logit explode in appendix E (tl;dr is that muon makes the singular value of the update matrix higher)
> Sparsity scaling laws to justify their ratio, they have a very solid training infra that allows the model to be trained at this sparsity level, they could have increased even more but as sparsity increases the training becomes less efficient.
> They diminish the number of attention heads to make it more efficient for long context since attention heads are a big bottleneck for long context. They also remove 2 of the 3 "first dense" layers in the dsv3 arch.
With the sparsity and attention heads (divided by 2) they achieve 83% increased flops compared to deepseek v3 arch at 128k.
> Data: Rephrasing is KEY. They do a lot more synthetic data generation and rephrase their corpus to have different styles, for longer documents they do it by chunk. I'm (half) surprised by the fact that ONLY 1 epoch (assuming same number of training tokens I think?) of data rephrased 10 times has better accuracy than 10 epochs of the same data rephrased once.
> They do rewriting for Math and Knowledge, for Math they apply the ShallowMath recipe and instruct the model to rephrase in a "learning note" style
> They talk about diversity and probably have some internal stuff/eval to test that, as always still a bit unclear for me how to properly measure that.
The infra is also very nice, quick summary:
> PP=16 (1F1B schedule, a bit custom), EP=16, zero1
> No FP8 computation but for storage of specific layers, selective recomputation for inexpensive block, activation offloading to CPU
Post
2786
Now it's possible to do RAG with any-to-any models 🔥
Learn how to search in a video dataset and generate using Tevatron/OmniEmbed-v0.1-multivent an all modality retriever, and Qwen/Qwen2.5-Omni-7B, any-to-any model in this notebook 🤝 merve/smol-vision
Learn how to search in a video dataset and generate using Tevatron/OmniEmbed-v0.1-multivent an all modality retriever, and Qwen/Qwen2.5-Omni-7B, any-to-any model in this notebook 🤝 merve/smol-vision
Post
2153
AudioRAG is becoming real! Just built a demo with ColQwen-Omni that does semantic search on raw audio, no transcription needed.
Drop in a podcast, ask your question, and it finds the exact chunks where it happens. You can also get a written answer.
What’s exciting: it skips transcription, making it faster and better at capturing emotion, ambient sound, and tone, surfacing results text search would miss.
- Demo: fdaudens/colqwen-omni-demo
- Blog post from ColQwen team: https://huggingface.co/blog/manu/colqwen-omni-omnimodal-retrieval
Drop in a podcast, ask your question, and it finds the exact chunks where it happens. You can also get a written answer.
What’s exciting: it skips transcription, making it faster and better at capturing emotion, ambient sound, and tone, surfacing results text search would miss.
- Demo: fdaudens/colqwen-omni-demo
- Blog post from ColQwen team: https://huggingface.co/blog/manu/colqwen-omni-omnimodal-retrieval
- 1 reply
Post
2055
all modality RAG 🔥
ColQwen-Omni is a new multimodal retrieval model that can retrieve anything (videos, audios, documents and more!)
use with transformers 🤗
read the blog https://huggingface.co/blog/manu/colqwen-omni-omnimodal-retrieval
model repository vidore/colqwen-omni-v0.1
ColQwen-Omni is a new multimodal retrieval model that can retrieve anything (videos, audios, documents and more!)
use with transformers 🤗
read the blog https://huggingface.co/blog/manu/colqwen-omni-omnimodal-retrieval
model repository vidore/colqwen-omni-v0.1
Post
2607
Fine-tune Gemma3n on videos with audios inside with Colab A100 🔥
Just dropped the notebook where you can learn how to fine-tune Gemma3n on images+audio+text at the same time!
keep in mind, it's made for educational purposes 🫡 we do LoRA, audio resampling & video downsampling to be able to train <40GB VRAM
stretch modalities and unfreeze layers as you wish! 🙏🏻 merve/smol-vision
Just dropped the notebook where you can learn how to fine-tune Gemma3n on images+audio+text at the same time!
keep in mind, it's made for educational purposes 🫡 we do LoRA, audio resampling & video downsampling to be able to train <40GB VRAM
stretch modalities and unfreeze layers as you wish! 🙏🏻 merve/smol-vision
- 1 reply
Post
2525
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
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