Hippius
maincontainerPublic

tired-cascade/gen-d4871f4eba85-top8

sha256:e060ccfdf982ff7eee2da32522e4533e69babf6ba03a245e51a4a92650c6ecdd·Indexed 26d ago

Layers

5

Total size

54.0 KB

Files

5

Quantization

README.md

928 B

gen-d4871f4eba85-top8

Top-eight fork of gen-d4871f4eba85. It applies the pruning idea from forever-real_tired3-5-top8, but selects families from this parent's own configured weights.

Active top-level families:

  • k14: parent weight 0.310099
  • k15: parent weight 0.210951
  • k03: parent weight 0.185441
  • k00: parent weight 0.042055
  • k21: parent weight 0.040000
  • k22: parent weight 0.040000
  • k10: parent weight 0.033967
  • k01: parent weight 0.032350

Their parent weights sum to 0.894863 and are renormalized to one without changing their relative proportions.

Sequence length remains fixed at 4096 and the parent curriculum remains disabled. Unselected kernel functions remain only where useful as implementation helpers for selected families (for example, k10 and k14 use k07 internally); they are absent from _FAMILIES, dispatch, and configured weights, so they cannot be selected as top-level training families.

Files

5 items
  • generator.py

    356515e7e27d

    48.4 KB

  • tests/test_top8.py

    3ef50fd40583

    2.3 KB

  • config.json

    3f20c0ed0b1c

    2.2 KB

  • README.md

    8cade9b95c31

    928 B

  • requirements.txt

    bce394a998f0

    186 B