🤝 Open to Collab

LH-Tech AI

LH-Tech-AI

·

https://lh-tech.de/ai/

LH-Tech-AI

AI & ML interests

Small AI and ML models. Trained by myself. Completely OpenSource. For you. | Reddit: https://www.reddit.com/user/LH-Tech_AI/

Recent Activity

liked a Space about 10 hours ago

multimodalart/MiniT2I

liked a model about 11 hours ago

bench-labs/pixelmodel-v1

reacted to wop's post with 👀 about 11 hours ago

# PixelModel v1 Last month we released PixelModel — a neural network whose weights are literally the pixels of a PNG. It was a toy: 202,752 parameters, welded to 32×32 output, trained on six solid-color swatches. It scored FID 566.84 on the Tiny-T2I-Leaderboard, mostly by producing the same yellow noise for every prompt. Today we're releasing PixelModel v1. It is 8.5× smaller — 23,747 parameters — and it beats v0 on both benchmark metrics while being trained on 20,000 real MS-COCO caption/image pairs instead of six color swatches. The entire model now fits in a 160×149 PNG. That image is not a visualization of the model. It is the model. All 23,747 weights, one per pixel. ## links https://huggingface.co/bench-labs Blog post [read it here ⇗](https://huggingface.co/spaces/bench-labs/blog?post=pixelmodel-v1.html) See us on the [Leaderboard ⇗](https://huggingface.co/spaces/FlameF0X/Tiny-T2I-Leaderboard) Model card [here](https://huggingface.co/bench-labs/pixelmodel-v1) ## The catch A 23K-parameter model does not draw sandwiches. With ~1 parameter per training image, the loss-minimizing behavior is to output the average of all plausible images for a caption — caption-conditioned color, light, and layout statistics. Food prompts come out warm and brown; sky prompts come out cool and bright. That is the ceiling for this size class, and we'd rather show it than crop around it. # cherry on top 🍒 The model generates 600 images (cpu) in 5 (five) seconds. Thats 5000 images in 24 seconds on cpu. The model trained on cpu for just 30 minutes.

View all activity

Organizations

LH-Tech-AI 's models 25

LH-Tech-AI/OpenMythos

Text Generation • 27B • Updated Jun 16 • 47 • 1

LH-Tech-AI/Quark-v2-0.5M

Text Generation • 466k • Updated May 9 • 74 • 7

LH-Tech-AI/Quark-0.5M

Text Generation • 466k • Updated May 9 • 287 • 2

LH-Tech-AI/NanoCalc-1M

Updated May 7 • 6

LH-Tech-AI/Flare-TTS-28M

Text-to-Speech • Updated May 6 • 3 • 9

LH-Tech-AI/Flare-TTS-v1.5

Text-to-Speech • Updated May 5 • 11

LH-Tech-AI/TinyMozart_85M

Text-to-Audio • Updated May 4 • 6

LH-Tech-AI/htmLLM-50M-Base

Text Generation • Updated May 4 • 8 • 3

LH-Tech-AI/Apex-1.5-Coder-Instruct-350M

Text Generation • 0.4B • Updated May 4 • 129 • 2

LH-Tech-AI/Apex-1.5-Instruct-350M

Text Generation • 0.4B • Updated May 4 • 305 • 8

LH-Tech-AI/Apex-1-Instruct-350M

Text Generation • Updated May 4 • 21 • 11

LH-Tech-AI/TinyMozart_v2_85M

Text-to-Audio • Updated May 3 • 20

LH-Tech-AI/Spark-5M-Base-v4

Text Generation • 4.98M • Updated May 2 • 56 • 3

LH-Tech-AI/CatGen-v2

Unconditional Image Generation • Updated Apr 30 • 3

LH-Tech-AI/FaceGen-v1

Unconditional Image Generation • Updated Apr 30 • 2

LH-Tech-AI/Shield-82M

Token Classification • 81.6M • Updated Apr 25 • 12 • 6

LH-Tech-AI/Crest-20M-Base

Text Generation • Updated Apr 20 • 1

LH-Tech-AI/Pin

Text Generation • Updated Apr 15

LH-Tech-AI/GyroScope

Image Classification • 11.2M • Updated Apr 12 • 55 • 9

LH-Tech-AI/Excerp_v1

Question Answering • Updated Apr 11 • 1

LH-Tech-AI/eMail-LM

Text Generation • Updated Apr 10 • 2

LH-Tech-AI/VibeCheck_v1

Text Classification • Updated Apr 5 • 2

LH-Tech-AI/CritiqueCore_v1

Text Classification • Updated Apr 5 • 1

LH-Tech-AI/Apex-1.6-Instruct-350M

Text Generation • 0.4B • Updated Mar 31 • 34 • 5

LH-Tech-AI/htmLLM-124M

Text Generation • Updated Mar 19 • 6 • 1