{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiceltpq7lz2ojjk5dyzprctpjr46y7c5fcawxsnu7csbxlklxazjm",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mhpkjlslc3b2"
  },
  "path": "/t/m-courtyard-v0-5-1-training-history-run-comparison-and-reliable-result-persistence/174540#post_1",
  "publishedAt": "2026-03-23T05:37:47.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "GitHub",
    "Release M-Courtyard v0.5.1 · Mcourtyard/m-courtyard"
  ],
  "textContent": "We’ve released M-Courtyard v0.5.1 for macOS Apple Silicon.\n\nThis update focuses on experiment traceability for local fine-tuning:\n\n  * Added a dedicated Training History workspace\n  * Added side-by-side run comparison\n  * Persisted training results reliably from the Rust backend\n  * Fixed missing / inconsistent loss curves in history and comparison mode\n  * Improved status accuracy for completed and stopped runs\n\n\n\nThe goal of this release is to make local fine-tuning workflows more reproducible and easier to iterate on without leaving the desktop app.\n\nGitHub Release:\n\nGitHub\n\n### Release M-Courtyard v0.5.1 · Mcourtyard/m-courtyard\n\nM-Courtyard v0.5.1 Download Platform Chip File macOS 14+ Apple Silicon (M1/M2/M3/M4) M-Courtyard__aarch64.dmg What's New See CHANGELOG for details. Installation Note ⚠️ Since th...",
  "title": "M-Courtyard v0.5.1: Training History, Run Comparison, and Reliable Result Persistence"
}