Hugging Face's logo
Dean Byrne's picture
Building on HF
10 1 44

Dean Byrne PRO

Quazim0t0
sometimesanotion's profile picture upmittal's profile picture Sakalti's profile picture
·

AI & ML interests

Joined HuggingFace: 1-20-2025 / DaisyChainAI🌼 / SmallLM's / San Francisco

Recent Activity

new activity about 1 hour ago
Intel/low_bit_open_llm_leaderboard:My models failed
reacted to goumsss's post with 👀 about 1 hour ago
🦁 I built a math app for my daughters — and they already love it. NumZoo is a mental math trainer for kids. Answer questions, earn cute AI-generated animal rewards. My girls picked their favorite animals, chose a magical place, and haven't stopped playing since. ▶️ Watch the demo: https://www.loom.com/share/c565cba49f2c4ad1bfb428e38ff4b629 🎮 Try it: https://huggingface.co/spaces/build-small-hackathon/numzoo Stack: - 🎨 FLUX.2-klein-4B — 4B params, Apache 2.0, 4 inference steps - 🖼️ Custom kawaii style LoRA (trained on FLUX.2-klein-base-4B) - ⚡ ZeroGPU · Gradio · Diffusers The LoRA: locking in a kawaii storybook style Out of the box, FLUX.2-klein is unpredictable — same prompt, wildly different styles. Not great when you want every reward to feel cozy and on-brand for a 6-year-old. So I fine-tuned a style LoRA: 54 hand-curated scenes (12 animals × 10 places), captioned with a NUMZOO. trigger and no style words — the model learns the look from the images alone. Rank 32, 1500 steps, trained on a Modal A100 in ~45 min. Published here: https://huggingface.co/goumsss/numzoo-flux2-klein-lora Two performance tricks that made it feel instant: - Pre-generation: image generation starts in the background the moment the quiz begins. By the time a reward is earned (3 correct answers), the image is already ready. Zero wait. - Cold-start fix: from_pretrained runs at module scope on CPU, outside the ZeroGPU budget. Inside @GPU we only do .to("cuda") — shaves the cold-start from potential timeout to ~1s. Built for the 🤗 Build Small Hackathon. Small model, real joy. 🐾
reacted to TerminalCalm's post with 👍 about 1 hour ago
A simple demo of TemperCheck, which is meant to quickly scan a screengrab of a social media profile and give the user a rough back-of-the-envelope estimation of how much of a temper they have. Made for the build small hackathon, relying on a Gemma4 edge model. Made for the build-small-hackathon with ZeroGPU, Gradio, and Gemma. Try it: https://huggingface.co/spaces/build-small-hackathon/TemperCheck
View all activity

Organizations

Build Small Hackathon's profile picture DaisyChainAI's profile picture