Article
Mastering Long Contexts in LLMs with KVPress
nvidia
β’ β’ 75
i come in peace
akirathegreatest
Β·
AI & ML interests
None yet
Recent Activity
reacted to kasbsquall's post with π₯ 22 days ago
π UX Crime Scene β every interface hides a crime.
Drop a screenshot of ANY website or app, and THE INSPECTOR β a film-noir detective β works it as a crime scene: he circles each UX flaw on the real pixels, names the charge, and files a verdict with a letter grade. A UX audit that plays like a detective thriller.
But the verdict is just the opening statement. Now it goes further:
βοΈ THE TRIAL β put the interface on trial. The guilty UI elements take the stand and defend themselves while the Inspector rules from the evidence.
πΌοΈ THE RECONSTRUCTION β one click and FLUX.2 Klein rebuilds the worst element FIXED, live. Before/after, on the real pixels.
π THE VOICE β hear the verdict read aloud (Kokoro, local, no keys).
π¨ MOST WANTED β a public rogues' gallery. Book your case onto a shared board where the city's worst interfaces are ranked by their crimes. Booked by the public.
Three small models, all on Modal (scale-to-zero), none over 32B:
ποΈ Qwen2.5-VL-7B (vision agent) Β· πΌοΈ FLUX.2 Klein (reconstruction) Β· π Kokoro-82M (voice)
π Human-graded: 84% grounding / 92% valid charges.
βΆοΈ Trailer: https://youtu.be/6u58YIEPrkA
πΉ Full walkthrough: https://youtu.be/WyQbY0XJ_9E
π΅οΈ Try it: https://huggingface.co/spaces/build-small-hackathon/ux-crime-scene
Built solo for #BuildSmallHackathon (Gradio Γ Hugging Face). Open the case β the Inspector is waiting. reacted to sergiopaniego's post with π₯ 7 months ago
ICYMI, you can fine-tune open LLMs using Claude Code
just tell it:
βFine-tune Qwen3-0.6B on open-r1/codeforces-cotsβ
and Claude submits a real training job on HF GPUs using TRL.
it handles everything:
> dataset validation
> GPU selection
> training + Trackio monitoring
> job submission + cost estimation
when itβs done, your model is on the Hub, ready to use
read more about the process: https://huggingface.co/blog/hf-skills-training liked a model 7 months ago
viktor02/sbert_classification_ru_ai_textsOrganizations
None yet
upvoted an article over 1 year ago