Oct 28 7 days ago

1. 13:56

   Add ZeroGPU stub so Space boots⁠ Running

   ReallyFloppyPenguin +1: Claude Opus 5.5
   Verified signature

   2ed55bb
2. 13:56

   WebGPU only: drop ZeroGPU backend, auto-scroll code tail⁠

   ReallyFloppyPenguin +1: Claude Opus 5.5
   Verified signature

   3d29673
3. 12:18

   Warn when no WebGPU adapter (CPU fallback)

   ReallyFloppyPenguin
   Verified signature

   e8582c5
4. 12:14

   Live 3D preview while generating, spinner + timing stats

   ReallyFloppyPenguin
   Verified signature

   ec59de0
5. 12:01

   WebGPU: clear partial model cache and retry once

   ReallyFloppyPenguin
   Verified signature

   8305ea5
6. 12:00

   Request 160s effective ZeroGPU time (107 * 1.5 factor)

   ReallyFloppyPenguin
   Verified signature

   677a745
7. 11:15

   Prefill ```html fence to skip chatty preamble

   ReallyFloppyPenguin
   Verified signature

   d1c5106
8. 11:10

   Add in-browser WebGPU backend (default) via wllama

   ReallyFloppyPenguin
   Verified signature

   67152c8
9. 10:12

   Set ZeroGPU duration to 160s

   ReallyFloppyPenguin
   Verified signature

   4d13480
10. 10:10

    Set ZeroGPU duration to 215s

    ReallyFloppyPenguin
    Verified signature

    64917ec
11. 10:01

    Preload CUDA 12 runtime libs for llama.cpp

    ReallyFloppyPenguin
    Verified signature

    dd41abd
12. 09:59

    Add full model card example prompts with labels

    ReallyFloppyPenguin
    Verified signature

    e19e7af
13. 09:57

    Initial 3D World Generator Space

    ReallyFloppyPenguin
    Verified signature

    2263bae
14. 09:56

    initial commit

    ReallyFloppyPenguin
    Verified signature

    79e334e