New Checkpoint Released for Krea 2 Turbo 4-Step Distillation LoRA
TimeTruth2490 · reddit · 2026-08-23
The author released a new checkpoint, chk00014000, for the Krea 2 Turbo 4-step distillation LoRA. This version reduces prediction error by an additional 6% compared to chk00010000 and 23% compared to chk00006000 on the same held-out set.
Training Methodology:
- The project does not rely on simply training longer; it follows a loop of "train -> assess -> adapt recipe -> retrain".
- Checkpoints are published only when they show measurable improvement and no regression in full resolution sweeps.
- Specific adjustments were made for optimizer issues (e.g., learning rate decay and weight decay) to solve model drift after convergence.
More from Multimodal
- AssemblyAI's Universal-3.6 Pro speech-to-text lands on OpenRouter at $0.30/hr — AssemblyAI · 2026-10-16
- FLUX 3 Image Klein lands on fal: box-drawn layouts, multi-reference placement, style transfer — bfl_ai · 2026-10-16
- SGLang ships Day-0 support for FLUX 3 Image Klein: 10 reference images, BF16/FP8/NVFP4 variants — ying11231 · 2026-10-16
- Krea 3 Scores 4.81 on CPI-Bench, Second Only to GPT Image 2.5 — testingcatalog · 2026-10-16
- Higgsfield teases Kling 4.0 rollout with a monkey-driven demo video — azed_ai · 2026-10-16
- Halcyon early-fusion image models match late-fusion with 275x less pretraining compute, open-sourced — davidwromero · 2026-10-16