Halcyon trains text-to-image models without language pretraining
The authors report Halcyon-1.6B beating a larger JiT baseline on gFD with 275 times less pretraining compute.
40
AI model
Halcyon-0.52B
- Mentions · 7d0
- Mentions · 30d2
- All mentions2
- First seenOct 18, 2026
- Last mentioned11d ago
Timeline
2 articles · newest firstHalcyon trains text-to-image models without language pretraining
Authors say Halcyon-1.6B beats a JiT baseline on gFD with 275 times less pretraining compute.
44
Appears with
- Halcyon-1.6Bai model · 2
- Halcyonai model · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.