Difference between revisions of "LightR"

Latest revision as of 16:08, 1 March 2023

LightR is a complete overhaul from its predecessor, ScalarR, to specialize in 1v1 scenarios. See Version History for recent activities.

A plan with little weight,
But heavy with machine learning fate.

@@ Line 1: / Line 1: @@
 {{Navbox small
-| title = LightR Sub-pages
+| title = Sub-pages
 | parent = LightR
 | page1 = Version History
 }}
-This page is under construction. For recent activities, see [[/Version History|Version History]].
+'''LightR''' is a complete overhaul from its predecessor, [[ScalarR]], to specialize in 1v1 scenarios. See [[/Version History | Version History]] for recent activities.
+==== Description ====
+A plan with little weight,<br>
+But heavy with machine learning fate.
-; Design principle
-: Strategy light, machine learning heavy.
-; Planned experiments
-: Towards deep learning:
-:: Multiple hand-tuned danger models -> Expert model & gate model
-:: Hand-crafted features with naive KNN -> Search-based sequence model
-:: Slow networks -> Knowledge distill & quantization aware training
-:: Offline pre-training & online fine-tuning of everything above.
-: Towards differentiable programming:
-:: Directly optimizing max escape angle & prior probability of getting hit (distancing)
-:: Per-instance level optimization of the above (mea & distancing as part of network)
 __NOTOC__ __NOEDITSECTION__
 {{Template:Bot Categorizer|author=Xor|isMega=true|isOneOnOne=true|isMelee=false|isOpenSource=false|extends=Interface}}