ngmi on Nostr: Sounds good on paper 🤔 Swift-Qwen3.8-27B is UkisAI's reasoning-efficient ...
Sounds good on paper 🤔
Swift-Qwen3.8-27B is UkisAI's reasoning-efficient derivative of Qwen3.8-27B, using 58.3% fewer thinking tokens while maintaining near-identical performance (<1% loss) and as a result getting a x1.95 speed-up on several tasks.
https://huggingface.co/ukisai/Swift-Qwen3.8-27B-GGUFPublished at
2026-09-21 11:46:42 UTCEvent JSON
{
"id": "be32267839a8d55e83dfd439d5ea1da992952644daacad02d355372bfb2980c9",
"pubkey": "a87c9bfc4570fe15afdda6c0095ade2d74fb14e3294bff37a965522983bda6a8",
"created_at": 1789991202,
"kind": 1,
"tags": [],
"content": "Sounds good on paper 🤔\n\nSwift-Qwen3.8-27B is UkisAI's reasoning-efficient derivative of Qwen3.8-27B, using 58.3% fewer thinking tokens while maintaining near-identical performance (\u003c1% loss) and as a result getting a x1.95 speed-up on several tasks.\n\nhttps://huggingface.co/ukisai/Swift-Qwen3.8-27B-GGUF",
"sig": "98a883638355f9a3fea7e11c9f9386423f37d85c351755c1aed92b41b5f2a557716728eaaec51b20441ee929e094e0b1f1237ef77c72f6453ce23d1f102552fc"
}
