shl0th (Saad Farooq)

liked a model 5 days ago

Qwen/Qwen2.5-Omni-7B

Any-to-Any • Updated 3 days ago • 128k • 1.35k

upvoted a collection 5 days ago

Llama 4

Collection

Meta's new Llama 4 multimodal models, Scout & Maverick. Includes Dynamic GGUFs, 16-bit & Dynamic 4-bit uploads. Run & fine-tune them with Unsloth! • 15 items • Updated 5 days ago • 42

upvoted a paper 6 days ago

SmolVLM: Redefining small and efficient multimodal models

Paper • 2504.05299 • Published 6 days ago • 156

reacted to jsulz's post with 🔥 6 days ago

Post

3553

Huge week for

xet-team as Llama 4 is the first major model on Hugging Face uploaded with Xet providing the backing! Every byte downloaded comes through our infrastructure.

Using Xet on Hugging Face is the fastest way to download and iterate on open source models and we've proved it with Llama 4 giving a boost of ~25% across all models.

We expect builders on the Hub to see even more improvements, helping power innovation across the community.

With the models on our infrastructure, we can peer in and see how well our dedupe performs across the Llama 4 family. On average, we're seeing ~25% dedupe, providing huge savings to the community who iterate on these state-of-the-art models. The attached image shows a few selected models and how they perform on Xet.

Thanks to the

meta-llama team for launching on Xet!

liked a model 12 days ago

all-hands/openhands-lm-32b-v0.1

Text Generation • Updated 10 days ago • 66.8k • 341

liked a model 21 days ago

deepseek-ai/DeepSeek-V3-0324

Text Generation • Updated 18 days ago • 208k • • 2.56k

upvoted a paper 21 days ago

MambaVision: A Hybrid Mamba-Transformer Vision Backbone

Paper • 2407.08083 • Published Jul 10, 2024 • 33

upvoted a collection 21 days ago

MambaVision

Collection

MambaVision: A Hybrid Mamba-Transformer Vision Backbone. Includes both 1K and 21K pretrained models. • 13 items • Updated 3 days ago • 31

liked 2 models 23 days ago

manycore-research/SpatialLM-Llama-1B

Text Generation • Updated 24 days ago • 18.1k • 936

manycore-research/SpatialLM-Qwen-0.5B

Text Generation • Updated 24 days ago • 1.53k • 88

reacted to tonyassi's post with 👍 27 days ago

Post

MANIFESTO
After working in fashion e-commerce for years I've come to the conclusion that in e-commerce we do not sell clothes... we sell images of clothes. Compressed, digital versions of physical products. As Roland Barthes pointed out in The Fashion System, a product image is a symbol or metaphor of a product. Media--in this case images--mediates the space between customer and product; viewer and object. Images can be altered, changed, corrupted, photoshopped, edited, deleted, or imagined. E-commerce products (or e-commerce photos) can thought of as a possibility space of digital pixels. AI/ML can analyze, manipulate, and create within this "possibility space of pixels"--thus it can be observed that there are opportunities to intervene in the physical fashion world through the imagination of artificial intelligence. Not to replace human creativity--but to augment it. To make it ART-ificial. Art is an artificial representation of reality. AI images are an artificial representation of reality. The sewing machine greatly increased the efficiency of clothing production. Similarly, AI has greatly increased efficiency image production, in our case product photo production. The fashion design paradigm of the past century (design->produce->photograph) has been flipped on this head. Instead of going from physical clothing to digital image via photography--we can go from digital image to physical clothing via stable diffusion. We are writing the chapter of Understanding Media that Marshall McLuhan never imagined. Virtual production hasn't replaced the physical production; it has simply made it out of style.