This is a pretty big update for sure. The models have improved significantly which is great for everyone involved, especially the end user. Those datasets look very promising as well!
Jordan Legg PRO
takarajordan
AI & ML interests
Chief AI Officer @takara.ai. Diffusion, Inference optimisation and all things MultiModal.
Recent Activity
Organizations
takarajordan's activity
Sounds interesting, I’ll check it out!
This is a really interesting post. I’ve been looking at the DeepSeek models for sure. This shows a pretty nice improvement, would love to see some example changes!
Very cool
posted
an
update
1 day ago
Post
1326
Takara takes 3rd place in the {tech:munich} AI hackathon with Fudeno!
A little over 2 weeks ago @aldigobbler and I set out to create the largest MultiModal SVG dataset ever created, we succeeded in this and when I was in Munich, Germany I took it one step further and made an entire app with it!
We fine-tuned Mistral Small, made a Next.JS application and blew some minds, taking 3rd place out of over 100 hackers. So cool!
If you want to see the dataset, please see below.
takara-ai/fudeno-instruct-4M
A little over 2 weeks ago @aldigobbler and I set out to create the largest MultiModal SVG dataset ever created, we succeeded in this and when I was in Munich, Germany I took it one step further and made an entire app with it!
We fine-tuned Mistral Small, made a Next.JS application and blew some minds, taking 3rd place out of over 100 hackers. So cool!
If you want to see the dataset, please see below.
takara-ai/fudeno-instruct-4M
replied to
their
post
3 months ago
Sir, basically I want to create a generative AI university helpdesk chatbot, and for this, I have created datasets myself and also fine-tuned models, but I am not getting satisfactory results. Sir, if you have time, could you please check my datasets in my profile and help me understand how I can improve my dataset and work on it so that my task gets completed? I would be very grateful to you.
I would enhance your dataset to use multi turn conversations if you can at all for llama2 you could do something like this:
<s>[INST] Is the BS Physics program a part-time or full-time course? [/INST] The BS Physics program is a full-time undergraduate program that requires regular on-campus attendance. </s><s>[INST] How many units per semester? [/INST] A typical semester load consists of 15-18 units. </s>
hope this helps! Again, please reach out to me on discord here: takarajordan_82155
gimme an invite! :D
reacted to
s3nh's
post with ❤️
3 months ago
Post
2012
Welcome back,
Small Language Models Enthusiasts and GPU Poor oss enjoyers lets connect.
Just created an organization which main target is to have fun with smaller models tuneable on consumer range GPUs, feel free to join and lets have some fun, much love ;3
https://huggingface.co/SmolTuners
Small Language Models Enthusiasts and GPU Poor oss enjoyers lets connect.
Just created an organization which main target is to have fun with smaller models tuneable on consumer range GPUs, feel free to join and lets have some fun, much love ;3
https://huggingface.co/SmolTuners
-
- 3 replies
Amazing work
reacted to
merve's
post with 🚀
3 months ago
Post
2837
Aya by Cohere For AI can now see! 👀
C4AI community has built Maya 8B, a new open-source multilingual VLM built on SigLIP and Aya 8B 🌱 works on 8 languages! 🗣️
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 👏
kudos @nahidalam and team
C4AI community has built Maya 8B, a new open-source multilingual VLM built on SigLIP and Aya 8B 🌱 works on 8 languages! 🗣️
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 👏
kudos @nahidalam and team
-
- 1 reply
reacted to
merve's
post with 🚀
3 months ago
Post
3562
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 📈 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 🔥
> 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 🔥
✨ 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 📈 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 🔥
> 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 🔥
- 8 replies
you guys are amazing!
reacted to
sayakpaul's
post with 🚀
3 months ago
Post
2222
In the past seven days, the Diffusers team has shipped:
1. Two new video models
2. One new image model
3. Two new quantization backends
4. Three new fine-tuning scripts
5. Multiple fixes and library QoL improvements
Coffee on me if someone can guess 1 - 4 correctly.
1. Two new video models
2. One new image model
3. Two new quantization backends
4. Three new fine-tuning scripts
5. Multiple fixes and library QoL improvements
Coffee on me if someone can guess 1 - 4 correctly.
-
- 1 reply
reacted to
lorraine2's
post with 🚀
3 months ago
Post
2006
🦙New NVIDIA paper: LLaMA-Mesh 🦙
We enable large language models to generate and understand 3D meshes by representing them as text and fine-tuning. This unifies the 3D and text modalities in a single model and preserves language abilities, unlocking conversational 3D creation with mesh understanding.
🔎 Project Page: https://research.nvidia.com/labs/toronto-ai/LLaMA-Mesh/
🕹️ Interactive Demo: Zhengyi/LLaMA-Mesh (courtesy of HuggingFace and Gradio)
📖 Full Paper: https://arxiv.org/abs/2411.09595
👨💻Code: https://github.com/nv-tlabs/LLaMa-Mesh
💾 Model Checkpoint: Zhengyi/LLaMA-Mesh
🧩 Blender Addon: https://github.com/huggingface/meshgen (courtesy of Dylan Ebert)
🎥 5-min Overview Video: https://youtu.be/eZNazN-1lPo?si=-idQa5aaceVw0Bbj (courtesy of AI Papers Academy)
We enable large language models to generate and understand 3D meshes by representing them as text and fine-tuning. This unifies the 3D and text modalities in a single model and preserves language abilities, unlocking conversational 3D creation with mesh understanding.
🔎 Project Page: https://research.nvidia.com/labs/toronto-ai/LLaMA-Mesh/
🕹️ Interactive Demo: Zhengyi/LLaMA-Mesh (courtesy of HuggingFace and Gradio)
📖 Full Paper: https://arxiv.org/abs/2411.09595
👨💻Code: https://github.com/nv-tlabs/LLaMa-Mesh
💾 Model Checkpoint: Zhengyi/LLaMA-Mesh
🧩 Blender Addon: https://github.com/huggingface/meshgen (courtesy of Dylan Ebert)
🎥 5-min Overview Video: https://youtu.be/eZNazN-1lPo?si=-idQa5aaceVw0Bbj (courtesy of AI Papers Academy)
reacted to
DualityAI-RebekahBogdanoff's
post with ❤️🔥
3 months ago
Post
1978
Training YOLO with Synthetic Data from Duality AI's Falcon Simulation Software 🎮📊
Hello again! 👋 Duality.ai has released a second Google Colab and tutorial for training a YOLOv8 model using synthetic data from our Falcon simulation software!
https://falcon.duality.ai/secure/documentation/see-synth-work-no-specs?sidebarMode=learn#download-the-colab-notebook
Train using synthetic images of a soup can twin this time, and see it work on real-world images. 🥫🍜
The tutorial also walks you through how to add your own twin from our FalconCloud library, and our goal is to equip people like you to be able to create your own data for your own projects.
You'll have to create a free account to access the files, but once you do, you'll have access to not only this colab file, but also all of our lessons and our digital twin library. 🎓
Instructions for creating the synthetic data accessed by the colab notebook can be found here: https://falcon.duality.ai/secure/documentation/ex2-objdetection-newtwin?sidebarMode=learn
This method is a game-changer for cost-effective, scalable, and customizable datasets in computer vision.
Why Synthetic Data?🤔
- Precise Annotations: Get bounding boxes, segmentation masks, and more without manual effort.
- Customizable Scenarios: Get comprehensive data and cover all corner cases by simulating diverse conditions like lighting, weather, visual occlusions, and more.
What’s in the Notebook?📓
- Training & Evaluation: Train YOLOv8 with synthetic data and test its performance on real-world samples.
Let’s Discuss!💬
Check out our discord to see how people are using the Falcon simulation software to develop strong datasets and train robust models. https://discord.com/invite/dualityfalconcommunity
Hello again! 👋 Duality.ai has released a second Google Colab and tutorial for training a YOLOv8 model using synthetic data from our Falcon simulation software!
https://falcon.duality.ai/secure/documentation/see-synth-work-no-specs?sidebarMode=learn#download-the-colab-notebook
Train using synthetic images of a soup can twin this time, and see it work on real-world images. 🥫🍜
The tutorial also walks you through how to add your own twin from our FalconCloud library, and our goal is to equip people like you to be able to create your own data for your own projects.
You'll have to create a free account to access the files, but once you do, you'll have access to not only this colab file, but also all of our lessons and our digital twin library. 🎓
Instructions for creating the synthetic data accessed by the colab notebook can be found here: https://falcon.duality.ai/secure/documentation/ex2-objdetection-newtwin?sidebarMode=learn
This method is a game-changer for cost-effective, scalable, and customizable datasets in computer vision.
Why Synthetic Data?🤔
- Precise Annotations: Get bounding boxes, segmentation masks, and more without manual effort.
- Customizable Scenarios: Get comprehensive data and cover all corner cases by simulating diverse conditions like lighting, weather, visual occlusions, and more.
What’s in the Notebook?📓
- Training & Evaluation: Train YOLOv8 with synthetic data and test its performance on real-world samples.
Let’s Discuss!💬
Check out our discord to see how people are using the Falcon simulation software to develop strong datasets and train robust models. https://discord.com/invite/dualityfalconcommunity
- 2 replies
reacted to
prithivMLmods's
post with ❤️
3 months ago
Post
3388
🎄 Here Before - Xmas🎅✨
🧑🏻🎄Models
+ [ Xmas 2D Illustration ] : strangerzonehf/Flux-Xmas-Illustration-LoRA
+ [ Xmas 3D Art ] : strangerzonehf/Flux-Xmas-3D-LoRA
+ [ Xmas Chocolate ] : strangerzonehf/Flux-Xmas-Chocolate-LoRA
+ [ Xmas Isometric Kit ] : strangerzonehf/Flux-Xmas-Isometric-Kit-LoRA
+ [ Xmas Realpix ] : strangerzonehf/Flux-Xmas-Realpix-LoRA
+ [ Xmas Anime ] : strangerzonehf/Flux-Anime-Xmas-LoRA
❄️Collections
+ [ Xmas Art ] : strangerzonehf/christmas-pack-6758b199487adafaddb68f82
+ [ Stranger Zone Collection ] : prithivMLmods/stranger-zone-collections-org-6737118adcf2cb40d66d0c7e
🥶Page
+ [ Stranger Zone ] : https://huggingface.co/strangerzonehf
.
.
.
@prithivMLmods 🤗
🧑🏻🎄Models
+ [ Xmas 2D Illustration ] : strangerzonehf/Flux-Xmas-Illustration-LoRA
+ [ Xmas 3D Art ] : strangerzonehf/Flux-Xmas-3D-LoRA
+ [ Xmas Chocolate ] : strangerzonehf/Flux-Xmas-Chocolate-LoRA
+ [ Xmas Isometric Kit ] : strangerzonehf/Flux-Xmas-Isometric-Kit-LoRA
+ [ Xmas Realpix ] : strangerzonehf/Flux-Xmas-Realpix-LoRA
+ [ Xmas Anime ] : strangerzonehf/Flux-Anime-Xmas-LoRA
❄️Collections
+ [ Xmas Art ] : strangerzonehf/christmas-pack-6758b199487adafaddb68f82
+ [ Stranger Zone Collection ] : prithivMLmods/stranger-zone-collections-org-6737118adcf2cb40d66d0c7e
🥶Page
+ [ Stranger Zone ] : https://huggingface.co/strangerzonehf
.
.
.
@prithivMLmods 🤗
reacted to
davidberenstein1957's
post with 🔥
3 months ago
Post
4238
Introducing the Synthetic Data Generator, a user-friendly application that takes a no-code approach to creating custom datasets with Large Language Models (LLMs). The best part: A simple step-by-step process, making dataset creation a non-technical breeze, allowing anyone to create datasets and models in minutes and without any code.
Blog: https://huggingface.co/blog/synthetic-data-generator
Space: argilla/synthetic-data-generator
Blog: https://huggingface.co/blog/synthetic-data-generator
Space: argilla/synthetic-data-generator
-
- 4 replies
reacted to
cutechicken's
post with ❤️
3 months ago
Post
2958
🚀 RAGOndevice: High-Performance Local AI Document Analysis Assistant
💫 Core Value
RAGOndevice is a high-performance AI system running locally without cloud dependency. Using CohereForAI's optimized 7B model, it enables professional-grade document analysis on standard PCs. ✨
🌟 Ondevice AI Advantages
1. 🔋 Efficient Resource Utilization
🎯 Optimized 7B Model: Runs on standard PCs
⚡ Local Processing: Instant response without cloud
💻 Low-Spec Compatible: Performs well on regular GPUs
🔄 Optimized Memory: Ensures stable operation
2. 🛡️ Data Security & Cost Efficiency
🔒 Complete Privacy: No external data transmission
🌐 Offline Operation: No internet required
💰 No Subscription: One-time installation
⚙️ Resource Optimization: Uses existing hardware
🎮 Key Features
1. 📊 Powerful Document Analysis
📁 Multi-Format Support: TXT, CSV, PDF, Parquet
🧠 Intelligent Analysis: Automatic structure recognition
👁️ OCR Support: Advanced PDF text extraction
💬 Real-time Chat: Natural language interaction
2. 🔍 Local RAG System
🎯 Efficient Search: TF-IDF based local search
🧩 Context Understanding: Accurate information retrieval
📚 Wikipedia Integration: Rich background knowledge
🎯 Use Cases
🏢 Enterprise: Secure confidential document processing
🔬 Personal Research: Private data analysis
📚 Education: Personal learning material analysis
💻 Development: Local codebase analysis
⭐ Differentiators
🏃♂️ Independent Operation: Zero cloud dependency
⚡ Instant Response: No network latency
🔐 Complete Security: Full data control
💎 Cost Efficiency: No ongoing costs
🔮 Future Plans
🚀 Enhanced model optimization
📚 Local knowledge base expansion
⚡ Hardware optimization
📁 Extended file support
🌟 RAGOndevice democratizes high-performance AI, providing the optimal local AI solution for security-sensitive environments. 🚀
🔥 Power of Local AI: Experience enterprise-grade AI capabilities right on your device!
VIDraft/RAGOndevice
💫 Core Value
RAGOndevice is a high-performance AI system running locally without cloud dependency. Using CohereForAI's optimized 7B model, it enables professional-grade document analysis on standard PCs. ✨
🌟 Ondevice AI Advantages
1. 🔋 Efficient Resource Utilization
🎯 Optimized 7B Model: Runs on standard PCs
⚡ Local Processing: Instant response without cloud
💻 Low-Spec Compatible: Performs well on regular GPUs
🔄 Optimized Memory: Ensures stable operation
2. 🛡️ Data Security & Cost Efficiency
🔒 Complete Privacy: No external data transmission
🌐 Offline Operation: No internet required
💰 No Subscription: One-time installation
⚙️ Resource Optimization: Uses existing hardware
🎮 Key Features
1. 📊 Powerful Document Analysis
📁 Multi-Format Support: TXT, CSV, PDF, Parquet
🧠 Intelligent Analysis: Automatic structure recognition
👁️ OCR Support: Advanced PDF text extraction
💬 Real-time Chat: Natural language interaction
2. 🔍 Local RAG System
🎯 Efficient Search: TF-IDF based local search
🧩 Context Understanding: Accurate information retrieval
📚 Wikipedia Integration: Rich background knowledge
🎯 Use Cases
🏢 Enterprise: Secure confidential document processing
🔬 Personal Research: Private data analysis
📚 Education: Personal learning material analysis
💻 Development: Local codebase analysis
⭐ Differentiators
🏃♂️ Independent Operation: Zero cloud dependency
⚡ Instant Response: No network latency
🔐 Complete Security: Full data control
💎 Cost Efficiency: No ongoing costs
🔮 Future Plans
🚀 Enhanced model optimization
📚 Local knowledge base expansion
⚡ Hardware optimization
📁 Extended file support
🌟 RAGOndevice democratizes high-performance AI, providing the optimal local AI solution for security-sensitive environments. 🚀
🔥 Power of Local AI: Experience enterprise-grade AI capabilities right on your device!
VIDraft/RAGOndevice
reacted to
lewtun's
post with 🔥
3 months ago
Post
6947
We outperform Llama 70B with Llama 3B on hard math by scaling test-time compute 🔥
How? By combining step-wise reward models with tree search algorithms :)
We show that smol models can match or exceed the performance of their much larger siblings when given enough "time to think"
We're open sourcing the full recipe and sharing a detailed blog post.
In our blog post we cover:
📈 Compute-optimal scaling: How we implemented DeepMind's recipe to boost the mathematical capabilities of open models at test-time.
🎄 Diverse Verifier Tree Search (DVTS): An unpublished extension we developed to the verifier-guided tree search technique. This simple yet effective method improves diversity and delivers better performance, particularly at large test-time compute budgets.
🧭 Search and Learn: A lightweight toolkit for implementing search strategies with LLMs and built for speed with vLLM
Here's the links:
- Blog post: HuggingFaceH4/blogpost-scaling-test-time-compute
- Code: https://github.com/huggingface/search-and-learn
Enjoy!
How? By combining step-wise reward models with tree search algorithms :)
We show that smol models can match or exceed the performance of their much larger siblings when given enough "time to think"
We're open sourcing the full recipe and sharing a detailed blog post.
In our blog post we cover:
📈 Compute-optimal scaling: How we implemented DeepMind's recipe to boost the mathematical capabilities of open models at test-time.
🎄 Diverse Verifier Tree Search (DVTS): An unpublished extension we developed to the verifier-guided tree search technique. This simple yet effective method improves diversity and delivers better performance, particularly at large test-time compute budgets.
🧭 Search and Learn: A lightweight toolkit for implementing search strategies with LLMs and built for speed with vLLM
Here's the links:
- Blog post: HuggingFaceH4/blogpost-scaling-test-time-compute
- Code: https://github.com/huggingface/search-and-learn
Enjoy!
- 2 replies