Skip to content

T
Toy-MT-Introduction
Commit 65fd1ab3 by 姜雨帆
Options

update slides

parent 19043049
正在显示 包含 105 行增加 和 99 行删除
Section05-Neural-Networks-and-Language-Modeling/section05.tex
......@@ -3923,11 +3923,11 @@ NLP问题的\alert{隐含结构}假设 & 无隐含结构假设，\alert{端到
\node [anchor=north] (index0) at ([yshift=0.5em]w0.south) {\tiny(index)};
\node [anchor=north] (index1) at ([yshift=0.5em]w1.south) {\tiny(index)};
\node [anchor=north] (index2) at ([yshift=0.5em]w2.south) {\tiny(index)};
\node [anchor=south,draw,inner sep=3pt] (e0) at ([yshift=1em]w0.north) {\tiny{$e_0=w_{i-3} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e1) at ([yshift=1em]w1.north) {\tiny{$e_1=w_{i-2} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e2) at ([yshift=1em]w2.north) {\tiny{$e_2=w_{i-1} \textbf{C}$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (h0) at ([yshift=1.5em]e1.north) {\tiny{$h_0=\textrm{Tanh}([e_0,e_1,e_2] \textbf{H} + \textbf{d})$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (h1) at ([yshift=1.5em]h0.north) {\tiny{$y=\textrm{Softmax}(h_0 \textbf{U})$}};
\node [anchor=south,draw,inner sep=3pt] (e0) at ([yshift=1em]w0.north) {\tiny{$\textbf{e}_0=w_{i-3} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e1) at ([yshift=1em]w1.north) {\tiny{$\textbf{e}_1=w_{i-2} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e2) at ([yshift=1em]w2.north) {\tiny{$\textbf{e}_2=w_{i-1} \textbf{C}$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (h0) at ([yshift=1.5em]e1.north) {\tiny{$\textbf{h}_0=\textrm{Tanh}([\textbf{e}_0,\textbf{e}_1,\textbf{e}_2] \textbf{H} + \textbf{d})$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (h1) at ([yshift=1.5em]h0.north) {\tiny{$\textbf{y}=\textrm{Softmax}(h_0 \textbf{U})$}};
\node [anchor=south] (ylabel) at ([yshift=1em]h1.north) {\footnotesize{$\textrm{P}(w_i|w_{i-3}w_{i-2}w_{i-1})$}};
\draw [->] ([yshift=0.1em]w0.north) -- ([yshift=-0.1em]e0.south);
......@@ -4483,23 +4483,23 @@ $\textbf{V}, \textbf{U}, \textbf{W}$: 参数
\begin{tikzpicture}
\begin{scope}
\node [anchor=west] (w0) at (0,0) {\footnotesize{$w_{0}$}};
\node [anchor=west] (w1) at ([xshift=4em]w0.east) {\footnotesize{$w_{1}$}};
\node [anchor=west] (w2) at ([xshift=4em]w1.east) {\footnotesize{$w_{2}$}};
\node [anchor=west] (w3) at ([xshift=4em]w2.east) {\footnotesize{$w_{3}$}};
\node [anchor=west] (w1) at ([xshift=5em]w0.east) {\footnotesize{$w_{1}$}};
\node [anchor=west] (w2) at ([xshift=5em]w1.east) {\footnotesize{$w_{2}$}};
\node [anchor=west] (w3) at ([xshift=5em]w2.east) {\footnotesize{$w_{3}$}};
\node [anchor=north] (index0) at ([yshift=0.5em]w0.south) {\tiny(index)};
\node [anchor=north] (index1) at ([yshift=0.5em]w1.south) {\tiny(index)};
\node [anchor=north] (index2) at ([yshift=0.5em]w2.south) {\tiny(index)};
\node [anchor=north] (index3) at ([yshift=0.5em]w3.south) {\tiny(index)};
\node [anchor=south,draw,inner sep=3pt] (e0) at ([yshift=1em]w0.north) {\tiny{$e_0=w_{0} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e1) at ([yshift=1em]w1.north) {\tiny{$e_1=w_{1} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e2) at ([yshift=1em]w2.north) {\tiny{$e_2=w_{2} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (e3) at ([yshift=1em]w3.north) {\tiny{$e_3=w_{3} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt] (h0) at ([xshift=-0.5em, yshift=1.5em]e0.north) {\tiny{$h_{0}=\textrm{SelfAtt}(e_0,e_3)$}};
\node [anchor=south,draw,inner sep=3pt] (h1) at ([xshift=0.5em, yshift=1.5em]e1.north) {\tiny{$h_{1}=\textrm{SelfAtt}(e_1,e_3)$}};
\node [anchor=south,draw,inner sep=3pt] (h2) at ([xshift=1.5em, yshift=1.5em]e2.north) {\tiny{$h_{2}=\textrm{SelfAtt}(e_2,e_3)$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (f1) at ([xshift=0.5em, yshift=1.5em]h2.north) {\tiny{$f_3=\textrm{FNN}([h_0,h_1,h_2,e_3])$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (o1) at ([yshift=1em]f1.north) {\tiny{$y=\textrm{Softmax}(f_3 \textbf{U})$}};
\node [anchor=south,draw,inner sep=3pt] (e0) at ([yshift=1em]w0.north) {\tiny{$\textbf{e}_0=w_{0} \textbf{C} + \textrm{PE}(0)$}};
\node [anchor=south,draw,inner sep=3pt] (e1) at ([yshift=1em]w1.north) {\tiny{$\textbf{e}_1=w_{1} \textbf{C} + \textrm{PE}(1)$}};
\node [anchor=south,draw,inner sep=3pt] (e2) at ([yshift=1em]w2.north) {\tiny{$\textbf{e}_2=w_{2} \textbf{C} + \textrm{PE}(2)$}};
\node [anchor=south,draw,inner sep=3pt] (e3) at ([yshift=1em]w3.north) {\tiny{$\textbf{e}_3=w_{3} \textbf{C} + \textrm{PE}(3)$}};
\node [anchor=south,draw,inner sep=3pt] (h0) at ([xshift=-0.5em, yshift=1.5em]e0.north) {\tiny{$\textbf{h}_{0}=\textrm{SelfAtt}(\textbf{e}_0,\textbf{e}_3)$}};
\node [anchor=south,draw,inner sep=3pt] (h1) at ([xshift=0.5em, yshift=1.5em]e1.north) {\tiny{$\textbf{h}_{1}=\textrm{SelfAtt}(\textbf{e}_1,\textbf{e}_3)$}};
\node [anchor=south,draw,inner sep=3pt] (h2) at ([xshift=1.5em, yshift=1.5em]e2.north) {\tiny{$\textbf{h}_{2}=\textrm{SelfAtt}(\textbf{e}_2,\textbf{e}_3)$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (f1) at ([xshift=0.5em, yshift=1.5em]h2.north) {\tiny{$\textbf{f}_3=\textrm{FNN}([\textbf{h}_0,\textbf{h}_1,\textbf{h}_2,\textbf{e}_3])$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt] (o1) at ([yshift=1em]f1.north) {\tiny{$\textbf{y}=\textrm{Softmax}(f_3 \textbf{U})$}};
\node [anchor=south] (ylabel) at ([yshift=1em]o1.north) {\footnotesize{$\textrm{P}(w_4|w_{0}w_{1}w_{2}w_{3})$}};
\draw [->] ([yshift=0.1em]w0.north) -- ([yshift=-0.1em]e0.south);
......@@ -4521,23 +4521,23 @@ $\textbf{V}, \textbf{U}, \textbf{W}$: 参数
\draw [->] ([yshift=0.1em]o1.north) -- ([yshift=-0.1em]ylabel.south);
\visible<2->{
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e0) at ([yshift=1em]w0.north) {\tiny{$e_0=w_{0} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e1) at ([yshift=1em]w1.north) {\tiny{$e_1=w_{1} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e2) at ([yshift=1em]w2.north) {\tiny{$e_2=w_{2} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e3) at ([yshift=1em]w3.north) {\tiny{$e_3=w_{3} \textbf{C}$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e0) at ([yshift=1em]w0.north) {\tiny{$\textbf{e}_0=w_{0} \textbf{C} + \textrm{PE}(0)$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e1) at ([yshift=1em]w1.north) {\tiny{$\textbf{e}_1=w_{1} \textbf{C} + \textrm{PE}(1)$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e2) at ([yshift=1em]w2.north) {\tiny{$\textbf{e}_2=w_{2} \textbf{C} + \textrm{PE}(2)$}};
\node [anchor=south,draw,inner sep=3pt,fill=blue!20!white] (e3) at ([yshift=1em]w3.north) {\tiny{$\textbf{e}_3=w_{3} \textbf{C} + \textrm{PE}(3)$}};
}
\visible<2->{
\node [anchor=west] (embedinglabel0) at ([xshift=-5em,yshift=-2em]w0.south) {\scriptsize{{\blue \textbf{词的分布式表示}}}};
\node [anchor=north west] (embedinglabel1) at ([yshift=0.3em]embedinglabel0.south west) {\scriptsize{前面已经介绍过！}};
\node [anchor=north west] (embedinglabel2) at ([yshift=0.3em]embedinglabel1.south west) {\scriptsize{基于One-hot表示获得}};
\node [anchor=north west] (embedinglabel3) at ([yshift=0.3em]embedinglabel2.south west) {\scriptsize{新加入位置向量}};
\node [anchor=north west] (embedinglabel3) at ([yshift=0.3em]embedinglabel2.south west) {\scriptsize{新加入位置向量PE}};
}
\visible<3->{
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h0) at ([xshift=-0.5em, yshift=1.5em]e0.north) {\tiny{$h_{0}=\textrm{SelfAtt}(e_0,e_3)$}};
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h1) at ([xshift=0.5em, yshift=1.5em]e1.north) {\tiny{$h_{1}=\textrm{SelfAtt}(e_1,e_3)$}};
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h2) at ([xshift=1.5em, yshift=1.5em]e2.north) {\tiny{$h_{2}=\textrm{SelfAtt}(e_2,e_3)$}};
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h0) at ([xshift=-0.5em, yshift=1.5em]e0.north) {\tiny{$\textbf{h}_{0}=\textrm{SelfAtt}(\textbf{e}_0,\textbf{e}_3)$}};
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h1) at ([xshift=0.5em, yshift=1.5em]e1.north) {\tiny{$\textbf{h}_{1}=\textrm{SelfAtt}(\textbf{e}_1,\textbf{e}_3)$}};
\node [anchor=south,draw,inner sep=3pt,fill=ugreen!20!white] (h2) at ([xshift=1.5em, yshift=1.5em]e2.north) {\tiny{$\textbf{h}_{2}=\textrm{SelfAtt}(\textbf{e}_2,\textbf{e}_3)$}};
}
\visible<3->{
......@@ -4548,8 +4548,8 @@ $\textbf{V}, \textbf{U}, \textbf{W}$: 参数
}
\visible<4->{
\node [anchor=south,draw,minimum width=9em,inner sep=3pt,fill=orange!20!white] (f1) at ([xshift=0.5em, yshift=1.5em]h2.north) {\tiny{$f_3=\textrm{FNN}([h_0,h_1,h_2,e_3])$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt,fill=orange!20!white] (o1) at ([yshift=1em]f1.north) {\tiny{$y=\textrm{Softmax}(f_3 \textbf{U})$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt,fill=orange!20!white] (f1) at ([xshift=0.5em, yshift=1.5em]h2.north) {\tiny{$\textbf{f}_3=\textrm{FNN}([\textbf{h}_0,\textbf{h}_1,\textbf{h}_2,\textbf{e}_3])$}};
\node [anchor=south,draw,minimum width=9em,inner sep=3pt,fill=orange!20!white] (o1) at ([yshift=1em]f1.north) {\tiny{$\textbf{y}=\textrm{Softmax}(f_3 \textbf{U})$}};
}
\visible<4->{
......@@ -4731,7 +4731,7 @@ GPT-2 (Transformer) & Radford et al. & 2019 & 35.7
\node [anchor=south] (w2) at (o2.north) {\footnotesize{椅子}};
\node [anchor=north] (label) at (o1.south) {\footnotesize{单词的one-hot表示}};
\visible<3->{
\node [anchor=south,fill=red!20!white] (cosine) at (w1.north) {\footnotesize{$cosine(\textrm{`桌子'},\textrm{`椅子'})=0$}};
\node [anchor=south,fill=red!20!white] (cosine) at (w1.north) {\footnotesize{$\textrm{cosine}(\textrm{`桌子'},\textrm{`椅子'})=0$}};
}
\end{scope}
......@@ -4744,7 +4744,7 @@ GPT-2 (Transformer) & Radford et al. & 2019 & 35.7
\node [anchor=south] (w2) at (o2.north) {\footnotesize{椅子}};
\node [anchor=north] (label) at ([yshift=-2em]o1.south) {\footnotesize{单词的分布式表示(词嵌入)}};
\visible<3->{
\node [anchor=south,fill=red!20!white] (cosine) at (w1.north) {\footnotesize{$cosine(\textrm{`桌子'},\textrm{`椅子'})=0.5$}};
\node [anchor=south,fill=red!20!white] (cosine) at (w1.north) {\footnotesize{$\textrm{cosine}(\textrm{`桌子'},\textrm{`椅子'})=0.5$}};
}
\end{scope}
}
......@@ -5015,59 +5015,59 @@ GPT-2 (Transformer) & Radford et al. & 2019 & 35.7
\begin{tikzpicture}
\begin{scope}[scale=1.2]
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm0) at (0,0) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm1) at ([xshift=1em]Lstm0.east) {\scriptsize{Lstm}};
\node [anchor=west,inner sep=4pt] (sep) at ([xshift=1em]Lstm1.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm2) at ([xshift=1em]sep.east) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm0) at (0,0) {\scriptsize{LSTM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm1) at ([xshift=1em]Lstm0.east) {\scriptsize{LSTM}};
\node [anchor=west,inner sep=4pt] (sep) at ([xshift=0.5em]Lstm1.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm2) at ([xshift=0.5em]sep.east) {\scriptsize{LSTM}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white] (Lstm3) at ([yshift=1em]Lstm0.north) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm4) at ([xshift=1em]Lstm3.east) {\scriptsize{Lstm}};
\node [anchor=west,inner sep=4pt] (sep1) at ([xshift=1em]Lstm4.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm5) at ([xshift=1em]sep1.east) {\scriptsize{Lstm}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white] (Lstm3) at ([yshift=1em]Lstm0.north) {\scriptsize{LSTM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm4) at ([xshift=1em]Lstm3.east) {\scriptsize{LSTM}};
\node [anchor=west,inner sep=4pt] (sep1) at ([xshift=0.5em]Lstm4.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm5) at ([xshift=0.5em]sep1.east) {\scriptsize{LSTM}};
\node [rectangle,rounded corners,draw=black!50,densely dashed,inner sep=0.4em] [fit = (Lstm0) (Lstm2) (Lstm3) (Lstm5)] (inputshadow) {};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([xshift=-2em,yshift=-1em]Lstm2.south) {\scriptsize{$E_1$}};
\node [anchor=west,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([xshift=1em]e1.east) {\scriptsize{$E_2$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([xshift=-2em,yshift=-1em]Lstm2.south) {\scriptsize{$\textbf{x}_1$}};
\node [anchor=west,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([xshift=1em]e1.east) {\scriptsize{$\textbf{x}_2$}};
\node [anchor=west,inner sep=4pt] (sep5) at ([xshift=1em]e2.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([xshift=1em]sep5.east) {\scriptsize{$E_3$}};
\node [anchor=west,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([xshift=1em]sep5.east) {\scriptsize{$\textbf{x}_n$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([xshift=-2em,yshift=1em]Lstm5.north) {\scriptsize{$T_1$}};
\node [anchor=west,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([xshift=1em]t1.east) {\scriptsize{$T_2$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([xshift=-2em,yshift=1em]Lstm5.north) {\scriptsize{$\textbf{h}_1$}};
\node [anchor=west,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([xshift=1em]t1.east) {\scriptsize{$\textbf{h}_2$}};
\node [anchor=west,inner sep=4pt] (sep6) at ([xshift=1em]t2.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([xshift=1em]sep6.east) {\scriptsize{$T_3$}};
\node [anchor=west,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([xshift=1em]sep6.east) {\scriptsize{$\textbf{h}_n$}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm6) at ([xshift=1.5em]Lstm2.east) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm7) at ([xshift=1em]Lstm6.east) {\scriptsize{Lstm}};
\node [anchor=west,inner sep=4pt] (sep3) at ([xshift=1em]Lstm7.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm8) at ([xshift=1em]sep3.east) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm6) at ([xshift=1.5em]Lstm2.east) {\scriptsize{LSTM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm7) at ([xshift=1em]Lstm6.east) {\scriptsize{LSTM}};
\node [anchor=west,inner sep=4pt] (sep3) at ([xshift=0.5em]Lstm7.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm8) at ([xshift=0.5em]sep3.east) {\scriptsize{LSTM}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white] (Lstm9) at ([yshift=1em]Lstm6.north) {\scriptsize{Lstm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm10) at ([xshift=1em]Lstm9.east) {\scriptsize{Lstm}};
\node [anchor=west,inner sep=4pt] (sep4) at ([xshift=1em]Lstm10.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm11) at ([xshift=1em]sep4.east) {\scriptsize{Lstm}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white] (Lstm9) at ([yshift=1em]Lstm6.north) {\scriptsize{LSTM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm10) at ([xshift=1em]Lstm9.east) {\scriptsize{LSTM}};
\node [anchor=west,inner sep=4pt] (sep4) at ([xshift=0.5em]Lstm10.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white] (Lstm11) at ([xshift=0.5em]sep4.east) {\scriptsize{LSTM}};
\node [rectangle,rounded corners,draw=black!50,densely dashed,inner sep=0.4em] [fit = (Lstm6) (Lstm8) (Lstm9) (Lstm11)] (inputshadow) {};
\draw [->] ([xshift=0.1em]Lstm0.east) -- ([xshift=-0.1em]Lstm1.west);
\draw [->] ([xshift=0.1em]Lstm1.east) -- ([xshift=-0.1em]sep.west);
\draw [->] ([xshift=0.1em]sep.east) -- ([xshift=-0.1em]Lstm2.west);
\draw [->] ([xshift=0.1em]Lstm1.east) -- ([xshift=0.1em]sep.west);
\draw [->] ([xshift=-0.1em]sep.east) -- ([xshift=-0.1em]Lstm2.west);
\draw [->] ([xshift=0.1em]Lstm3.east) -- ([xshift=-0.1em]Lstm4.west);
\draw [->] ([xshift=0.1em]Lstm4.east) -- ([xshift=-0.1em]sep1.west);
\draw [->] ([xshift=0.1em]sep1.east) -- ([xshift=-0.1em]Lstm5.west);
\draw [->] ([xshift=0.1em]Lstm4.east) -- ([xshift=0.1em]sep1.west);
\draw [->] ([xshift=-0.1em]sep1.east) -- ([xshift=-0.1em]Lstm5.west);
\draw [->] ([yshift=0.1em]Lstm0.north) -- ([yshift=-0.1em]Lstm3.south);
\draw [->] ([yshift=0.1em]Lstm1.north) -- ([yshift=-0.1em]Lstm4.south);
\draw [->] ([yshift=0.1em]Lstm2.north) -- ([yshift=-0.1em]Lstm5.south);
\draw [->] ([xshift=0.1em]Lstm6.east) -- ([xshift=-0.1em]Lstm7.west);
\draw [->] ([xshift=0.1em]Lstm7.east) -- ([xshift=-0.1em]sep3.west);
\draw [->] ([xshift=0.1em]sep3.east) -- ([xshift=-0.1em]Lstm8.west);
\draw [->] ([xshift=0.1em]Lstm7.east) -- ([xshift=0.1em]sep3.west);
\draw [->] ([xshift=-0.1em]sep3.east) -- ([xshift=-0.1em]Lstm8.west);
\draw [->] ([xshift=0.1em]Lstm9.east) -- ([xshift=-0.1em]Lstm10.west);
\draw [->] ([xshift=0.1em]Lstm10.east) -- ([xshift=-0.1em]sep4.west);
\draw [->] ([xshift=0.1em]sep4.east) -- ([xshift=-0.1em]Lstm11.west);
\draw [->] ([xshift=0.1em]Lstm10.east) -- ([xshift=0.1em]sep4.west);
\draw [->] ([xshift=-0.1em]sep4.east) -- ([xshift=-0.1em]Lstm11.west);
\draw [->] ([yshift=0.1em]Lstm6.north) -- ([yshift=-0.1em]Lstm9.south);
\draw [->] ([yshift=0.1em]Lstm7.north) -- ([yshift=-0.1em]Lstm10.south);
......@@ -5108,35 +5108,38 @@ GPT-2 (Transformer) & Radford et al. & 2019 & 35.7
\begin{tikzpicture}
\begin{scope}[scale=1.2]
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm0) at (0,0) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm1) at ([xshift=1em]Trm0.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm2) at ([xshift=1em]Trm1.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm3) at ([xshift=1em]Trm2.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm0) at (0,0) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm1) at ([xshift=1em]Trm0.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm2) at ([xshift=1em]Trm1.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm3) at ([xshift=1em]Trm2.east) {\scriptsize{TRM}};
\node [anchor=west,inner sep=4pt] (sep) at ([xshift=1em]Trm3.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm4) at ([xshift=1em]sep.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm4) at ([xshift=1em]sep.east) {\scriptsize{TRM}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm5) at ([yshift=1em]Trm0.north) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm6) at ([xshift=1em]Trm5.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm7) at ([xshift=1em]Trm6.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm8) at ([xshift=1em]Trm7.east) {\scriptsize{Trm}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm5) at ([yshift=1em]Trm0.north) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm6) at ([xshift=1em]Trm5.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm7) at ([xshift=1em]Trm6.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm8) at ([xshift=1em]Trm7.east) {\scriptsize{TRM}};
\node [anchor=west,inner sep=4pt] (sep1) at ([xshift=1em]Trm8.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm9) at ([xshift=1em]sep1.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm9) at ([xshift=1em]sep1.east) {\scriptsize{TRM}};
\node [rectangle,rounded corners,draw=black!50,densely dashed,inner sep=0.4em] [fit = (Trm0) (Trm4) (Trm5) (Trm9)] (inputshadow) {};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([yshift=-1em]Trm0.south) {\scriptsize{$E_1$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([yshift=-1em]Trm1.south) {\scriptsize{$E_2$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([yshift=-1em]Trm2.south) {\scriptsize{$E_3$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e4) at ([yshift=-1em]Trm3.south) {\scriptsize{$E_4$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([yshift=-1em]Trm0.south) {\scriptsize{$\textbf{x}_1$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([yshift=-1em]Trm1.south) {\scriptsize{$\textbf{x}_2$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([yshift=-1em]Trm2.south) {\scriptsize{$\textbf{x}_3$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e4) at ([yshift=-1em]Trm3.south) {\scriptsize{$\textbf{x}_4$}};
\node [anchor=north,inner sep=4pt] (sep5) at ([yshift=-1em]sep.south) {\scriptsize{...}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e5) at ([yshift=-1em]Trm4.south) {\scriptsize{$E_5$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e5) at ([yshift=-1em]Trm4.south) {\scriptsize{$\textbf{x}_n$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([yshift=1em]Trm5.north) {\scriptsize{$T_1$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([yshift=1em]Trm6.north) {\scriptsize{$T_2$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([yshift=1em]Trm7.north) {\scriptsize{$T_3$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t4) at ([yshift=1em]Trm8.north) {\scriptsize{$T_4$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([yshift=1em]Trm5.north) {\scriptsize{$\textbf{h}_1$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([yshift=1em]Trm6.north) {\scriptsize{$\textbf{h}_2$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([yshift=1em]Trm7.north) {\scriptsize{$\textbf{h}_3$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t4) at ([yshift=1em]Trm8.north) {\scriptsize{$\textbf{h}_4$}};
\node [anchor=south,inner sep=4pt] (sep6) at ([yshift=1em]sep1.north) {\scriptsize{...}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t5) at ([yshift=1em]Trm9.north) {\scriptsize{$T_5$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t5) at ([yshift=1em]Trm9.north) {\scriptsize{$\textbf{h}_n$}};
\node [anchor=west,draw,inner sep=3pt,fill=blue!20!white,minimum width=1em] (Lt1) at ([yshift=1.5em]t1.west) {\tiny{TRM}};
\node [anchor=west] (Lt2) at ([xshift=-0.1em]Lt1.east) {\tiny{: Transformer}};
\draw [->] ([yshift=0.1em]e1.north) -- ([yshift=-0.1em]Trm0.south);
\draw [->] ([yshift=0.1em]e1.north) -- ([yshift=-0.1em]Trm1.south);
......@@ -5190,34 +5193,37 @@ GPT-2 (Transformer) & Radford et al. & 2019 & 35.7
\begin{scope}[scale=1.2]
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm0) at (0,0) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm1) at ([xshift=1em]Trm0.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm2) at ([xshift=1em]Trm1.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm3) at ([xshift=1em]Trm2.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm1) at ([xshift=1em]Trm0.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm2) at ([xshift=1em]Trm1.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm3) at ([xshift=1em]Trm2.east) {\scriptsize{TRM}};
\node [anchor=west,inner sep=4pt] (sep) at ([xshift=1em]Trm3.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm4) at ([xshift=1em]sep.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm4) at ([xshift=1em]sep.east) {\scriptsize{TRM}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm5) at ([yshift=1em]Trm0.north) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm6) at ([xshift=1em]Trm5.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm7) at ([xshift=1em]Trm6.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm8) at ([xshift=1em]Trm7.east) {\scriptsize{Trm}};
\node [anchor=south,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm5) at ([yshift=1em]Trm0.north) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm6) at ([xshift=1em]Trm5.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm7) at ([xshift=1em]Trm6.east) {\scriptsize{TRM}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm8) at ([xshift=1em]Trm7.east) {\scriptsize{TRM}};
\node [anchor=west,inner sep=4pt] (sep1) at ([xshift=1em]Trm8.east) {\scriptsize{...}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm9) at ([xshift=1em]sep1.east) {\scriptsize{Trm}};
\node [anchor=west,draw,inner sep=4pt,fill=blue!20!white,minimum width=3em] (Trm9) at ([xshift=1em]sep1.east) {\scriptsize{TRM}};
\node [rectangle,rounded corners,draw=black!50,densely dashed,inner sep=0.4em] [fit = (Trm0) (Trm4) (Trm5) (Trm9)] (inputshadow) {};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([yshift=-1em]Trm0.south) {\scriptsize{$E_1$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([yshift=-1em]Trm1.south) {\scriptsize{$E_2$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([yshift=-1em]Trm2.south) {\scriptsize{$E_3$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e4) at ([yshift=-1em]Trm3.south) {\scriptsize{$E_4$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e1) at ([yshift=-1em]Trm0.south) {\scriptsize{$\textbf{x}_1$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e2) at ([yshift=-1em]Trm1.south) {\scriptsize{$\textbf{x}_2$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e3) at ([yshift=-1em]Trm2.south) {\scriptsize{$\textbf{x}_3$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e4) at ([yshift=-1em]Trm3.south) {\scriptsize{$\textbf{x}_4$}};
\node [anchor=north,inner sep=4pt] (sep5) at ([yshift=-1em]sep.south) {\scriptsize{...}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e5) at ([yshift=-1em]Trm4.south) {\scriptsize{$E_5$}};
\node [anchor=north,draw,inner sep=4pt,fill=ugreen!20!white,minimum width=2em] (e5) at ([yshift=-1em]Trm4.south) {\scriptsize{$\textbf{x}_n$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([yshift=1em]Trm5.north) {\scriptsize{$T_1$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([yshift=1em]Trm6.north) {\scriptsize{$T_2$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([yshift=1em]Trm7.north) {\scriptsize{$T_3$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t4) at ([yshift=1em]Trm8.north) {\scriptsize{$T_4$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t1) at ([yshift=1em]Trm5.north) {\scriptsize{$\textbf{h}_1$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t2) at ([yshift=1em]Trm6.north) {\scriptsize{$\textbf{h}_2$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t3) at ([yshift=1em]Trm7.north) {\scriptsize{$\textbf{h}_3$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t4) at ([yshift=1em]Trm8.north) {\scriptsize{$\textbf{h}_4$}};
\node [anchor=south,inner sep=4pt] (sep6) at ([yshift=1em]sep1.north) {\scriptsize{...}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t5) at ([yshift=1em]Trm9.north) {\scriptsize{$T_5$}};
\node [anchor=south,draw,inner sep=4pt,fill=yellow!30,minimum width=2em] (t5) at ([yshift=1em]Trm9.north) {\scriptsize{$\textbf{h}_n$}};
\node [anchor=west,draw,inner sep=3pt,fill=blue!20!white,minimum width=1em] (Lt1) at ([yshift=1.5em]t1.west) {\tiny{TRM}};
\node [anchor=west] (Lt2) at ([xshift=-0.1em]Lt1.east) {\tiny{: Transformer}};
\draw [->] ([yshift=0.1em]e1.north) -- ([yshift=-0.1em]Trm0.south);
\draw [->] ([yshift=0.1em]e1.north) -- ([yshift=-0.1em]Trm1.south);
......
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论