{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreihrhynly5q7qsr7elrzlzdyos4rah5az3ias2ae24vqhyj3bxehki",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mlengtuyac72"
  },
  "path": "/t/can-you-train-and-release-lora-for-different-tasks/1380529#post_1",
  "publishedAt": "2026-05-08T19:19:54.000Z",
  "site": "https://community.openai.com",
  "textContent": "Codex is awesome, but it’s specialized in coding, but not all of us wish to use it as a coding agent.\n\nI personally have used it for creative writings, but it just sucks at it, horrible!\n\nI have played AI for a little bit myself, so I understand how easy it is to train a LoRA. For a big company like Open AI, training a LoRA will be just like walking in a park, with eyes closed.\n\nSo I was thinking, is it possible for you to train a LoRA for different tasks and name it “Expert” or something.\n\nLike, Expert for creative writing, Expert for travel planning, so forth and so on.\n\nSo users could choose whichever tasks they prefer.\n\nI think this is a good suggestion because, honestly, “Use Codex only for coding” is really waste of an awesome app.\n\nAnd to be frank and blunt, that’s not a good marketing strategy either.\n\nThank you very much for your time and consideration.",
  "title": "Can you train and release LoRA for different tasks?"
}