|
|
\definecolor{hf1}{HTML}{FFD220} |
|
|
\definecolor{hf2}{HTML}{FF8360} |
|
|
\definecolor{hf3}{HTML}{2D728F} |
|
|
\definecolor{hf4}{HTML}{B5DFCA} |
|
|
\definecolor{hf5}{HTML}{BABFD1} |
|
|
|
|
|
\newcommand{\highlight}[1]{\textcolor{hf2}{#1}} |
|
|
|
|
|
\newcommand{\lerobot}{\texttt{lerobot}} |
|
|
\newcommand{\FK}{\text{FK}} |
|
|
\newcommand{\targetvel}{\dot {p}^*} |
|
|
\newcommand{\targetpos}{p^*} |
|
|
|
|
|
\newcommand{\statespace}{\mathcal S} |
|
|
\newcommand{\actionspace}{\mathcal A} |
|
|
\newcommand{\obsspace}{\mathcal O} |
|
|
\newcommand{\dynamics}{\mathcal D} |
|
|
\newcommand{\stateplusone}{s_{t+1}} |
|
|
\newcommand{\state}{s_t} |
|
|
\newcommand{\action}{a_t} |
|
|
\newcommand{\transition}{(\state, \action, \stateplusone)} |
|
|
\newcommand{\sars}{(\state, \action, r_t, \stateplusone)} |
|
|
\newcommand{\transitiongiven}{(\stateplusone \vert \state, \action)} |
|
|
\newcommand{\transitionprob}{\mathbb P \transitiongiven} |
|
|
\newcommand{\trajectory}{(s_0, a_0, r_0, s_1, a_1, r_1, \dots, s_{T-1}, a_{T-1}, r_{T-1}, s_T)} |
|
|
\newcommand{\Jpi}{J (\pi_\theta) } |
|
|
\newcommand{\qfunction}{\(Q\)-function} |
|
|
\newcommand{\qopt}{\( Q^* \)} |
|
|
|
|
|
\newcommand{\supp}[1]{\operatorname{supp}({#1})} |
|
|
\newcommand{\DKL}{\text{D}_{\text{KL}}} |
|
|
|
|
|
\newcommand{\actionchunk}{\mathbf{A}} |
|
|
\newcommand{\actionexpert}{\mathbf{v}_\theta} |
|
|
\newcommand{\pizero}{\( \pi_0 \)} |
|
|
|
|
|
|
|
|
\newtcolorbox{callout}[2][]{ |
|
|
enhanced, breakable, |
|
|
colback=hfbackground, opacityback=0.85, |
|
|
colframe=ai2accent, boxrule=0.6pt, arc=2mm, |
|
|
left=8pt, right=8pt, top=8pt, bottom=8pt, |
|
|
before skip=10pt, after skip=10pt, |
|
|
fonttitle=\sffamily\bfseries, |
|
|
title={\sffamily\bfseries #2}, |
|
|
#1 |
|
|
} |
|
|
|
|
|
|
|
|
\newenvironment{tldr}[1][TL;DR]{\begin{callout}{#1}}{\end{callout}} |
|
|
|
|
|
\newcommand{\lerobotdataset}{\texttt{LeRobotDataset}} |
