Infrastructure●●●Banger
We fine-tuned an AI model for log search – Accuracy 50% to 80%
Fixes AI log search blindness by fine-tuning embeddings on operational data.
Big BrainDark HorseSolve My Problem
rkorlimarla
101mo ago
Back to browse
A SaaS tagline critic that runs in the browser
by trashhalo·May 28, 2026·2 points·1 comment
AI Analysis
●●SolidBig BrainWizardry
336 MB model runs fully client-side via WASM - fine-tuning broke 0.80 accuracy ceiling.
Strengths
- •Fine-tuned entire embedding body not just linear head - achieved 0.90 pairwise accuracy.
- •Zero backend architecture - model downloads once then works fully offline.
- •OpenRouter PKCE OAuth brainstorm is 100% client-side with no app registration needed.
Weaknesses
- •Niche single-purpose tool - tagline rating is narrow use case.
- •336 MB download is heavy for a browser-based utility.
Category
Target Audience
SaaS founders and marketers writing landing page copy
Similar To
Coolors · Namelix · Looka
Similar Projects
AI/ML●●Solid
I built a local AI-powered Ouija board with a fine-tuned 3B model
Fine-tuned 3B model as a Ouija board with crisis detection—clever execution, zero practical use.
Rabbit HoleEye CandyCozy
SurceBeats
653mo ago
AI/ML●Mid
OpenAI CLIP fine tuned on Galaxy morphology
Galaxy classification model, but model card has mostly empty fields.
Niche Gem
mjupp1
102mo ago
AI/ML●●Solid
Pre-training, fine-tuning, and evals platform
Eval-synthesize-train loop automates custom model development better than manual fine-tuning.
SlickBold Bet
oli_kitty
402mo ago
AI/ML●●Solid
Fine-tuned 3B outperforms Claude Haiku on constrained generation
Fine-tuned 3B Qwen matches Haiku on jokes, validating small models for constrained agent tasks.
Big BrainNiche Gem
serendip-ml
102mo ago
AI/ML●●Solid
Flint – A 30B model fine-tuned for less repetition
Fine-tuned Qwen 30B that prioritizes output diversity over convergent accuracy.
Niche GemSolve My Problem
thmsmxwll
621mo ago