Labels

nc-review had nothing to raise

nc-review left non-blocking findings

nc-review found blocking findings

Benchmarks and judges: tests.json, match modes, judge providers and API keys, reports

Chatting with a fine-tuned model: nanotune chat, llama-server lifecycle

GitHub Actions and CI

Project setup and CLI surface: init, status, clean, config.json, flags, the TUI, model downloads an

Dataset management: data add/import/list/validate/split, train.jsonl and valid.jsonl

Documentation

Fusing adapters and exporting models: GGUF conversion, quantisation, llama.cpp

Fine-tuning with MLX: LoRA parameters, training runs, checkpoints, training logs

Something isn't working

Pull requests that update a dependency file

Improvements or additions to documentation

This issue or pull request already exists

New feature or request

Good for newcomers

Extra attention is needed

Severe user impact: fix before other work

This doesn't seem right

Pull requests that update javascript code

Queued for automated issue triage

Issue triage: not an actionable issue report

Speed, memory use or disk use

Issue triage: likely duplicates another issue

A speculative idea rather than committed work

Further information is requested

A security weakness or hardening

This will not be worked on