open_clip
Dreambooth-Stable-Diffusion
Our great sponsors
open_clip | Dreambooth-Stable-Diffusion | |
---|---|---|
27 | 47 | |
8,452 | 7,383 | |
8.2% | - | |
8.2 | 0.0 | |
17 days ago | over 1 year ago | |
Jupyter Notebook | Jupyter Notebook | |
GNU General Public License v3.0 or later | MIT 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.
open_clip
-
A History of CLIP Model Training Data Advances
While OpenAI’s CLIP model has garnered a lot of attention, it is far from the only game in town—and far from the best! On the OpenCLIP leaderboard, for instance, the largest and most capable CLIP model from OpenAI ranks just 41st(!) in its average zero-shot accuracy across 38 datasets.
-
How to Build a Semantic Search Engine for Emojis
Whenever I’m working on semantic search applications that connect images and text, I start with a family of models known as contrastive language image pre-training (CLIP). These models are trained on image-text pairs to generate similar vector representations or embeddings for images and their captions, and dissimilar vectors when images are paired with other text strings. There are multiple CLIP-style models, including OpenCLIP and MetaCLIP, but for simplicity we’ll focus on the original CLIP model from OpenAI. No model is perfect, and at a fundamental level there is no right way to compare images and text, but CLIP certainly provides a good starting point.
-
Database of 16,000 Artists Used to Train Midjourney AI Goes Viral
It is a misconception that Adobe's models have not been trained on copyrighted work. Nobody should be repeating their marketing claims.
Adobe has not shown how they train the text encoders in Firefly, or what images were used for the text-based conditioning (i.e. "text to image") part of their image generation model. They are almost certainly using CLIP or T5, which are trained on LAION2b, an image dataset with the very problems they are trying to address, C4 (a text dataset similarly encumbered) and similar.
I welcome anyone who works at Adobe to simply answer this question of how they trained the text encoders for text conditioning and put it to rest. There is absolutely nothing sensitive about the issue, unless it exposes them in a lie.
So no chance. I think it's a big fat lie. They'd have to have made some other scientific breakthrough, which they didn't.
Using information from https://openai.com/research/clip and https://github.com/mlfoundations/open_clip, it's possible to investigate the likelihood that using just their stock image dataset, can they make a working text encoder?
It's certainly not impossible, but it's impracticable. On 248m images (roughly the size of Adobe Stock), CLIP gets 37% on ImageNet, and on the 2000m from LAION, it performs 71-80%. And even with 2000m images, CLIP is substantially worse performing than the approach that Imagen uses for "text comprehension," which relies on essentially many billions more images and text tokens.
-
MetaCLIP – Meta AI Research
https://github.com/mlfoundations/open_clip/blob/main/docs/op...
-
COMFYUI SDXL WORKFLOW INBOUND! Q&A NOW OPEN! (WIP EARLY ACCESS WORKFLOW INCLUDED!)
in the modal card it says: pretrained text encoders (OpenCLIP-ViT/G and CLIP-ViT/L).
-
Is Nicholas Renotte a good guide for a person who knows nothing about ML?
also, if you describe your task a bit more, we might be able to direct you to a fairly out-of-the-box solution, e.g. you might be able to use one of the pretrained models supported by https://github.com/mlfoundations/open_clip without any additional training
-
Generate Image from Vector Embedding
It says on the Stable Diffusion Github repo that it uses the “OpenCLIP-ViT/H” https://github.com/mlfoundations/open_clip model as a text encoder, and from my prior experience with CLIP, I have found that it is very easy to generate image and text embeddings (because CLIP is a multimodal model).
-
What's up in the Python community? – April 2023
https://replicate.com/pharmapsychotic/clip-interrogator
using:
cfg.apply_low_vram_defaults()
interrogate_fast()
I tried lighter models like vit32/laion400 and others etc all are very very slow to load or use (model list: https://github.com/mlfoundations/open_clip)
I'm desperately looking for something more modest and light.
-
Low accuracy on my CNN model.
A library that is very useful for this kind of application is timm. You may also find the feature representation provided by a CLIP model particularly powerful.
- Looking for OpenAI CLIP alternative
Dreambooth-Stable-Diffusion
- Where can I train my own LoRA?
-
I am having an error with ControlNet (RuntimeError: CUDA error: CUBLAS_STATUS_ALLOC_FAILED when calling `cublasCreate(handle)`)
I did search online for an answer, but I am a PC noob, I didn't know what to do when I found this solution in this link: https://github.com/XavierXiao/Dreambooth-Stable-Diffusion/issues/113
- True to life photorealism v2
-
How can to create a custom image generation model?
Do you know some projects or guided tutorials that could help me? How many drawings with the desired style I should then have to give to train the AI model? I found Dreambooth on Stable Diffusion but it seams to be for another use case.
-
How to Make Your Own Anime (Linux/Mac Tutorial follow along)
This seems to be an issue with the code and or the environment itself. There is an open bug for this where some suggestions are p provided by others on how to fix. https://github.com/XavierXiao/Dreambooth-Stable-Diffusion/issues/47
-
AI generated portraits of Myself as different classes: Looking for opinion!
Could you provide some more detail on how this works? Did you just use this GitHub repository or did you put together your own implementation?
- Looking for an AI model to transform a video of me (full body) into an animated avatar. Does something like this exist?
-
Ray Liotta as Tommy Vercetti from GTA Vice City
I think the best way to do this would be to train Dreambooth on a number of photos of Ray Liotta first, and use Stable Diffusion instead. https://github.com/XavierXiao/Dreambooth-Stable-Diffusion
-
Luddites don't have a issue with AI, just that it "steals" from them (it doesn't). But they also have a issue with using your own child's drawings as a reference.
Dreambooth. There are other ways, but that is the gold standard. It takes even more Vram than regular stable diffusion, so if you don't have a very beefy card (e.g. 4090 with 25 GB VRAM) various websites let you do it onlin for a small fee. You then download a new model that has all the old stuff (e.g.the 4 gigabyte SD 1.5 file) plus your new images. Like I said, there are other ways that are easier, but when people show great results they are usually talking about Dreambooth.
-
Bunch of misinformation being spread in this thread
THE CODE (unofficial implementation, for the exact wording stating how little images you need read the paper) is designed with extremely little data in mind. I don't know how else to phrase it dude, do you think the training is a magic black box that runs with snail neurons? If you train a dreambooth model the jupyter ide makes calls to python files, those are the files. That is the code
What are some alternatives?
CLIP - CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
xformers - Hackable and optimized Transformers building blocks, supporting a composable construction.
DALLE-pytorch - Implementation / replication of DALL-E, OpenAI's Text to Image Transformer, in Pytorch
stable-diffusion-webui - Stable Diffusion web UI
taming-transformers - Taming Transformers for High-Resolution Image Synthesis
SHARK - SHARK - High Performance Machine Learning Distribution
bitsandbytes - Accessible large language models via k-bit quantization for PyTorch.
stable-diffusion - Optimized Stable Diffusion modified to run on lower GPU VRAM
clip-retrieval - Easily compute clip embeddings and build a clip retrieval system with them
StableTuner - Finetuning SD in style.
stablediffusion - High-Resolution Image Synthesis with Latent Diffusion Models
Dreambooth-SD-optimized - Implementation of Dreambooth (https://arxiv.org/abs/2208.12242) with Stable Diffusion