laion.ai
LAVIS
laion.ai | LAVIS | |
---|---|---|
25 | 18 | |
107 | 8,985 | |
4.7% | 2.7% | |
8.5 | 5.4 | |
6 days ago | 16 days ago | |
HTML | Jupyter Notebook | |
MIT License | BSD 3-clause "New" or "Revised" License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
laion.ai
-
How Open is Generative AI? Part 2
LAION (Large-scale Artificial Intelligence Open Network), a German non-profit established in 2020, is dedicated to advancing open-source models and datasets (primarily under Apache 2 and MIT licenses) to foster open research and the evolution of benevolent AI. Their datasets, encompassing both images and text, have been pivotal in the training of renowned text-to-image models like Stable Diffusion.
-
How artists are sabotaging AI to take revenge on image generators
> there is going to be a "pre-GPT" internet training set from 2022
Well, yeah, there are several here, and I think all the major image generators are using some combination of them as their starting points: https://laion.ai/
> As AI increases as an overall % of all online posts and activity it will death spiral on model quality.
Nope, it will just mean that it will be more expensive to source additional training data on top of the massive trove of existing "clean" (from intentional poisoning) data (much of which isn't perfectly captioned and human work on improving captioning can improve its utility in model training, as can more advanced models with more advanced text encoders, etc.)
If poisoning was widespread, it wouldn't impact "big model" quality much -- they aren't grabbing new random data on the internet for continuous training. It might drive up the expense of community fine tuning, which often does depend on sourcing representative imagery for target styles or concepts from, among other places, the internet.
-
[D] Why is most Open Source AI happening outside the USA?
Also don't forget https://laion.ai/ from Germany. They focus more on datasets, but still.
-
OpenAI is too cheap to beat
I think the weird thing about this is that it's completely true right now but in X months it may be totally outdated advice.
For example, efforts like OpenMOE https://github.com/XueFuzhao/OpenMoE or similar will probably eventually lead to very competitive performance and cost-effectiveness for open source models. At least in terms of competing with GPT-3.5 for many applications.
Also see https://laion.ai/
I also believe that within say 1-3 years there will be a different type of training approach that does not require such large datasets or manual human feedback.
-
MJ images sources?
Billions. MJ's initial training dataset was from LAION: https://laion.ai/ . Not sure which version, and I am pretty sure additional data has been added since MJ v1, but MJ doesn't release anything more exact. However my guess is: more billions, lol.
-
AI tools apps in one place sorted by category
Missing LAION and OpenAssistant: https://laion.ai/
- GPT detectors are biased against non-native English writers
-
Model Suggestions
As far as I am concerned weights of llama are not allowed for commercial use, but if you are willing to do full training and change it's all weights it would probably be fine. There was a discussion on this topic on forums and no one was sure, you can research it. Also you can take a look at laion.ai and dolly from databricks, they are open source and are allowed for commercial use, if they meet your needs.
- HuggingChat, the first open source alternative to ChatGPT
-
Hugging Face releases its own version of ChatGPT
that's OpenAssistant's / LAION AI model, HuggingFace provided the infrastructure.
LAVIS
- FLaNK AI for 11 March 2024
- FLaNK 04 March 2024
-
[D] Why is most Open Source AI happening outside the USA?
For multimodal, there's China (*many), then Salesforce.
-
Need help for a colab notebook running Lavis blip2_instruct_vicuna13b?
Been trying for all day to get a working inference for this example: https://github.com/salesforce/LAVIS/tree/main/projects/instructblip
-
most sane web3 job listing
There's also been big breakthroughs in computer vision. Not that long ago it was hard to recognize if a photo contained a bird; that's solved now by models like CLIP, Yolo, or Segment Anything. Now research has moved on to generating 3D scenes from images or interactively answering questions about images.
-
I work at a non-tech company and have been asked to make software that is impossible. How do I explain it to my boss?
The new hotness is multimodal vision-language models like InstructBLIP that can interactively answer questions about images. Check out the examples in the github repo, I would not have thought this was possible a few years ago.
-
Two-minute Daily AI Update (Date: 5/15/2023)
Salesforce’s BLIP family has a new member– InstructBLIP, a vision-language instruction-tuning framework using BLIP-2 models. It has achieved state-of-the-art zero-shot generalization performance on a wide range of vision-language tasks, substantially outperforming BLIP-2 and Flamingo. (Source)
-
InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning
Github
-
Can I use my own art as a training set?
Most of my workflows are self-made. For captioning I used Blip-2 in a custom script I made that automates the process by going into directories and their sub-directories and creates a .txt file beside each image. This way I can keep my images organized in their proper directories, without having to put dump them all in a single place.
- FLiP Stack Weekly for 13-Feb-2023
What are some alternatives?
llm-foundry - LLM training code for Databricks foundation models
pytorch-widedeep - A flexible package for multimodal-deep-learning to combine tabular data with text and images using Wide and Deep models in Pytorch
llama - Inference code for Llama models
CLIP-Caption-Reward - PyTorch code for "Fine-grained Image Captioning with CLIP Reward" (Findings of NAACL 2022)
stable-diffusion-webui - Stable Diffusion web UI
sparseml - Libraries for applying sparsification recipes to neural networks with a few lines of code, enabling faster and smaller models
qlora - QLoRA: Efficient Finetuning of Quantized LLMs
robo-vln - Pytorch code for ICRA'21 paper: "Hierarchical Cross-Modal Agent for Robotics Vision-and-Language Navigation"
gpt-neox - An implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries
DeepViewAgg - [CVPR'22 Best Paper Finalist] Official PyTorch implementation of the method presented in "Learning Multi-View Aggregation In the Wild for Large-Scale 3D Semantic Segmentation"
Open-Assistant - OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.
linkis - Apache Linkis builds a computation middleware layer to facilitate connection, governance and orchestration between the upper applications and the underlying data engines.