Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
2
1
35
Nemat
Amin24
Follow
Mohammadreza's profile picture
1 follower
ยท
0 following
AI & ML interests
Applied Linguistics, Language Learning
Recent Activity
reacted
to
merve
's
post
with ๐
12 days ago
Google's SigLIP is another alternative to openai's CLIP, and it just got merged to ๐คtransformers and it's super easy to use! To celebrate this, I have created a repository including notebooks and bunch of Spaces on various SigLIP based projects ๐ฅณ Search for art ๐ https://huggingface.co/spaces/merve/draw_to_search_art Compare SigLIP with CLIP ๐ https://huggingface.co/spaces/merve/compare_clip_siglip How does SigLIP work? SigLIP an vision-text pre-training technique based on contrastive learning. It jointly trains an image encoder and text encoder such that the dot product of embeddings are most similar for the appropriate text-image pairs The image below is taken from CLIP, where this contrastive pre-training takes place with softmax, but SigLIP replaces softmax with sigmoid. ๐ Highlights from the paper on why you should use it โจ ๐ผ๏ธ๐ Authors used medium sized B/16 ViT for image encoder and B-sized transformer for text encoder ๐ More performant than CLIP on zero-shot ๐ฃ๏ธ Authors trained a multilingual model too! โก๏ธ Super efficient, sigmoid is enabling up to 1M items per batch, but the authors chose 32k because the performance saturates after that It's super easy to use thanks to transformers ๐ ```python from transformers import pipeline from PIL import Image import requests # load pipe image_classifier = pipeline(task="zero-shot-image-classification", model="google/siglip-base-patch16-256-i18n") # load image url = 'http://images.cocodataset.org/val2017/000000039769.jpg' image = Image.open(requests.get(url, stream=True).raw) # inference outputs = image_classifier(image, candidate_labels=["2 cats", "a plane", "a remote"]) outputs = [{"score": round(output["score"], 4), "label": output["label"] } for output in outputs] print(outputs) ``` For all the SigLIP notebooks on similarity search and indexing, you can check this [repository](https://github.com/merveenoyan/siglip) out. ๐ค
liked
a model
5 months ago
XLabs-AI/flux-ip-adapter
liked
a model
5 months ago
InstantX/FLUX.1-dev-Controlnet-Union
View all activity
Organizations
None yet
models
3
Sort:ย Recently updated
Amin24/setnence_gen
Updated
Dec 2, 2023
Amin24/word_example_generator
Updated
Nov 27, 2023
โข
4
Amin24/distilbert-base-uncased-finetuned-emotion
Text Classification
โข
Updated
Nov 25, 2023
โข
10
datasets
None public yet