MochiCore
Adapter training that fits on one GPU
The trainer the rest of the lab runs on. Adapter families, quantised paths and a config format you can read out loud.
- Language
- Python
- Stars
- 3.6k
- Licence
- MIT
- Category
- Training
Twelve open-source projects across research, training, data, evaluation and infrastructure. All freely available, most of them still moving.
12 projects
Adapter training that fits on one GPU
The trainer the rest of the lab runs on. Adapter families, quantised paths and a config format you can read out loud.
Flow-matching generation with region control
A text-to-image stack built around region conditioning, so a prompt can say where something belongs rather than only what it is.
Multi-label tagging for illustration
A tagger over a 12k-tag vocabulary that holds up on the long tail, where general-purpose vision models quietly collapse into the fifty most common labels.
The dataset everything else is trained on
An 18M-image illustration corpus with tags, captions and quality bands, plus the whole filtering pipeline that produced it.
A registry for artefacts you actually own
Self-hostable model and dataset registry with content addressing, so a checkpoint on your cluster and one on ours are the same object or provably are not.
Small detectors for illustration
Face, head, person and hand detection tuned for drawn images, small enough to run on a laptop CPU in real time.
Vector search that fits in a process
An embedded vector index for the very common case where your corpus is large but not that large.
Benchmarks with their failure cases attached
An evaluation harness that refuses to emit a single number without the error analysis that explains it.
Annotation without the enterprise tax
A browser annotation tool for tagging, boxes and preference pairs, built because the alternatives all wanted a licence conversation.
Job scheduling for one cluster and four GPUs
A scheduler sized for small labs: queue a run, get it placed, see why it did not start when it did not start.
Prompt expansion that respects your tags
A small language model that turns a handful of tags into a full prompt without inventing subject matter you did not ask for.
A game about training a model badly
A small browser game where you tune hyperparameters under time pressure. Educational by accident, mostly built for fun.