mini-experiment - a appvoid Collection

appvoid 's Collections

favorite spaces

favorite models

favorite datasets

mini-experiment

mini-experiment

updated Nov 10

this represent an effort to discover optimal improvements on slms using curriculum learning as n+10k samples on a 250m parameters danube model

appvoid/arco-mini-run-1

Updated Nov 9 • 6

Note final loss: 2.353
appvoid/arco-mini-run-2

Updated Nov 9 • 3

Note final loss: 2.1243
appvoid/arco-mini-run-3

Updated Nov 9 • 2

Note final loss: 1.7117
appvoid/arco-mini-run-4

Updated Nov 9 • 3

Note final loss: 1.5411
appvoid/arco-mini-run-5

Updated Nov 9 • 4

Note final loss: 1.558 (high learning rate dropped performance, working on it)
appvoid/arco-mini-run-75k

Updated Nov 10 • 4

Note arco-mini-run-5 trained on 25k additional samples