JSJavaScripts.com

/

Package

@metaharness/weight-eft

Fine-tune cheap open-source LLMs (GLM, Qwen, DeepSeek) on your AI coding agent's successful runs with LoRA (SFT + DPO) so your model cascade escalates to expensive frontier models (GPT, Claude) less often — cutting cost-per-resolved. Turns run history int

v0.1.1, 27 Jun 2026MITESM onlyTypes: Bundled

$ npm install @metaharness/weight-eft

npm GitHub Homepage

Key numbers

As of 25 Sept 2026

Weekly downloads: 16.3k; - over 12 months
Gzip bundle size: -; No Bundlephobia data
Dependencies: 0; Direct, from package.json
Last release: 2 months ago; v0.1.1
GitHub stars: 673; 85 forks
Open issues / PRs: 116 / 20
Contributors: -
Commits, last 52 weeks: 973

Downloads

No weekly download data in the snapshot.

16,317 downloads last week, 84,738 in the last month. No 52-week series has been fetched yet.

Maintenance

Releases, last 12 months: -
Latest version: 0.1.1, 27 Jun 20262 months ago
First published: -
Last push to GitHub: 24 Sept 2026same day as the snapshot
Repository archived: No
Open issues: 116
Open pull requests: 20
Commits, last 52 weeks: 973
Dependents on npm: 1

Major versions

-

Recent releases

-

Compatibility

Module format: ESM onlyFrom package.json type, main and exports fields
TypeScript types: Bundled
Node.js engine range: >=20.0.0
Runtimes: -
Unpacked size (npm): 105 kB
Dependencies (0): None

Install

npm install @metaharness/weight-eft

CDN

jsDelivrhttps://cdn.jsdelivr.net/npm/@metaharness/weight-eft@0.1.1/
unpkghttps://unpkg.com/@metaharness/weight-eft@0.1.1/
esm.shhttps://esm.sh/@metaharness/weight-eft@0.1.1

Sources: npm registry, GitHub API. Data fetched 25 Sept 2026.

Indexability: Not indexed (discovery only): no category.