diff --git a/build/convnet-min.js b/build/convnet-min.js
deleted file mode 100644
index 24354824..00000000
--- a/build/convnet-min.js
+++ /dev/null
@@ -1 +0,0 @@
-var convnetjs=convnetjs||{REVISION:"ALPHA"};(function(c){var j=false;var d=0;var k=function(){if(j){j=false;return d}var p=2*Math.random()-1;var o=2*Math.random()-1;var q=p*p+o*o;if(q==0||q>1){return k()}var s=Math.sqrt(-2*Math.log(q)/q);d=o*s;j=true;return p*s};var h=function(p,o){return Math.random()*(o-p)+p};var f=function(p,o){return Math.floor(Math.random()*(o-p)+p)};var b=function(p,o){return p+k()*o};var e=function(q){if(typeof(q)==="undefined"||isNaN(q)){return[]}if(typeof ArrayBuffer==="undefined"){var o=new Array(q);for(var p=0;p<q;p++){o[p]=0}return o}else{return new Float64Array(q)}};var m=function(o,p){for(var q=0,r=o.length;q<r;q++){if(o[q]===p){return true}}return false};var n=function(p){var o=[];for(var q=0,r=p.length;q<r;q++){if(!m(o,p[q])){o.push(p[q])}}return o};var a=function(p){if(p.length===0){return{}}var o=p[0];var r=p[0];var q=0;var t=0;var u=p.length;for(var s=1;s<u;s++){if(p[s]>o){o=p[s];q=s}if(p[s]<r){r=p[s];t=s}}return{maxi:q,maxv:o,mini:t,minv:r,dv:o-r}};var i=function(u){var r=u,p=0,o;var t=[];for(var s=0;s<u;s++){t[s]=s}while(r--){p=Math.floor(Math.random()*(r+1));o=t[r];t[r]=t[p];t[p]=o}return t};var g=function(o,u){var s=h(0,1);var r=0;for(var q=0,t=o.length;q<t;q++){r+=u[q];if(s<r){return o[q]}}};var l=function(p,q,o){return typeof p[q]!=="undefined"?p[q]:o};c.randf=h;c.randi=f;c.randn=b;c.zeros=e;c.maxmin=a;c.randperm=i;c.weightedSample=g;c.arrUnique=n;c.arrContains=m;c.getopt=l})(convnetjs);(function(b){var a=function(k,g,f,j){if(Object.prototype.toString.call(k)==="[object Array]"){this.sx=1;this.sy=1;this.depth=k.length;this.w=b.zeros(this.depth);this.dw=b.zeros(this.depth);for(var d=0;d<this.depth;d++){this.w[d]=k[d]}}else{this.sx=k;this.sy=g;this.depth=f;var h=k*g*f;this.w=b.zeros(h);this.dw=b.zeros(h);if(typeof j==="undefined"){var e=Math.sqrt(1/(k*g*f));for(var d=0;d<h;d++){this.w[d]=b.randn(0,e)}}else{for(var d=0;d<h;d++){this.w[d]=j}}}};a.prototype={get:function(c,g,f){var e=((this.sx*g)+c)*this.depth+f;return this.w[e]},set:function(c,h,g,f){var e=((this.sx*h)+c)*this.depth+g;this.w[e]=f},add:function(c,h,g,f){var e=((this.sx*h)+c)*this.depth+g;this.w[e]+=f},get_grad:function(c,g,f){var e=((this.sx*g)+c)*this.depth+f;return this.dw[e]},set_grad:function(c,h,g,f){var e=((this.sx*h)+c)*this.depth+g;this.dw[e]=f},add_grad:function(c,h,g,f){var e=((this.sx*h)+c)*this.depth+g;this.dw[e]+=f},cloneAndZero:function(){return new a(this.sx,this.sy,this.depth,0)},clone:function(){var c=new a(this.sx,this.sy,this.depth,0);var e=this.w.length;for(var d=0;d<e;d++){c.w[d]=this.w[d]}return c},addFrom:function(c){for(var d=0;d<this.w.length;d++){this.w[d]+=c.w[d]}},addFromScaled:function(d,c){for(var e=0;e<this.w.length;e++){this.w[e]+=c*d.w[e]}},setConst:function(c){for(var d=0;d<this.w.length;d++){this.w[d]=c}},toJSON:function(){var c={};c.sx=this.sx;c.sy=this.sy;c.depth=this.depth;c.w=this.w;return c},fromJSON:function(d){this.sx=d.sx;this.sy=d.sy;this.depth=d.depth;var e=this.sx*this.sy*this.depth;this.w=b.zeros(e);this.dw=b.zeros(e);for(var c=0;c<e;c++){this.w[c]=d.w[c]}}};b.Vol=a})(convnetjs);(function(c){var a=c.Vol;var b=function(f,h,n,m,g){if(typeof(g)==="undefined"){var g=false}if(typeof(n)==="undefined"){var n=c.randi(0,f.sx-h)}if(typeof(m)==="undefined"){var m=c.randi(0,f.sy-h)}var e;if(h!==f.sx||n!==0||m!==0){e=new a(h,h,f.depth,0);for(var l=0;l<h;l++){for(var k=0;k<h;k++){if(l+n<0||l+n>=f.sx||k+m<0||k+m>=f.sy){continue}for(var j=0;j<f.depth;j++){e.set(l,k,j,f.get(l+n,k+m,j))}}}}else{e=f}if(g){var i=e.cloneAndZero();for(var l=0;l<e.sx;l++){for(var k=0;k<e.sy;k++){for(var j=0;j<e.depth;j++){i.set(l,k,j,e.get(e.sx-l-1,k,j))}}}e=i}return e};var d=function(o,n){if(typeof(n)==="undefined"){var n=false}var h=document.createElement("canvas");h.width=o.width;h.height=o.height;var u=h.getContext("2d");try{u.drawImage(o,0,0)}catch(q){if(q.name==="NS_ERROR_NOT_AVAILABLE"){return false}else{throw q}}try{var v=u.getImageData(0,0,h.width,h.height)}catch(q){if(q.name==="IndexSizeError"){return false}else{throw q}}var g=v.data;var k=o.width;var s=o.height;var t=[];for(var m=0;m<g.length;m++){t.push(g[m]/255-0.5)}var r=new a(k,s,4,0);r.w=t;if(n){var f=new a(k,s,1,0);for(var m=0;m<k;m++){for(var l=0;l<s;l++){f.set(m,l,0,r.get(m,l,0))}}r=f}return r};c.augment=b;c.img_to_vol=d})(convnetjs);(function(c){var a=c.Vol;var d=function(g){var g=g||{};this.out_depth=g.filters;this.sx=g.sx;this.in_depth=g.in_depth;this.in_sx=g.in_sx;this.in_sy=g.in_sy;this.sy=typeof g.sy!=="undefined"?g.sy:this.sx;this.stride=typeof g.stride!=="undefined"?g.stride:1;this.pad=typeof g.pad!=="undefined"?g.pad:0;this.l1_decay_mul=typeof g.l1_decay_mul!=="undefined"?g.l1_decay_mul:0;this.l2_decay_mul=typeof g.l2_decay_mul!=="undefined"?g.l2_decay_mul:1;this.out_sx=Math.floor((this.in_sx+this.pad*2-this.sx)/this.stride+1);this.out_sy=Math.floor((this.in_sy+this.pad*2-this.sy)/this.stride+1);this.layer_type="conv";var e=typeof g.bias_pref!=="undefined"?g.bias_pref:0;this.filters=[];for(var f=0;f<this.out_depth;f++){this.filters.push(new a(this.sx,this.sy,this.in_depth))}this.biases=new a(1,1,this.out_depth,e)};d.prototype={forward:function(l,s){this.in_act=l;var j=new a(this.out_sx,this.out_sy,this.out_depth,0);for(var o=0;o<this.out_depth;o++){var n=this.filters[o];var r=-this.pad;var p=-this.pad;for(var e=0;e<this.out_sx;r+=this.stride,e++){p=-this.pad;for(var t=0;t<this.out_sy;p+=this.stride,t++){var q=0;for(var m=0;m<n.sx;m++){for(var k=0;k<n.sy;k++){for(var i=0;i<n.depth;i++){var g=p+k;var h=r+m;if(g>=0&&g<l.sy&&h>=0&&h<l.sx){q+=n.w[((n.sx*k)+m)*n.depth+i]*l.w[((l.sx*g)+h)*l.depth+i]}}}}q+=this.biases.w[o];j.set(e,t,o,q)}}}this.out_act=j;return this.out_act},backward:function(){var k=this.in_act;k.dw=c.zeros(k.w.length);for(var n=0;n<this.out_depth;n++){var m=this.filters[n];var q=-this.pad;var p=-this.pad;for(var e=0;e<this.out_sx;q+=this.stride,e++){p=-this.pad;for(var t=0;t<this.out_sy;p+=this.stride,t++){var o=this.out_act.get_grad(e,t,n);for(var l=0;l<m.sx;l++){for(var j=0;j<m.sy;j++){for(var i=0;i<m.depth;i++){var g=p+j;var h=q+l;if(g>=0&&g<k.sy&&h>=0&&h<k.sx){var s=((k.sx*g)+h)*k.depth+i;var r=((m.sx*j)+l)*m.depth+i;m.dw[r]+=k.w[s]*o;k.dw[s]+=m.w[r]*o}}}}this.biases.dw[n]+=o}}}},getParamsAndGrads:function(){var e=[];for(var f=0;f<this.out_depth;f++){e.push({params:this.filters[f].w,grads:this.filters[f].dw,l2_decay_mul:this.l2_decay_mul,l1_decay_mul:this.l1_decay_mul})}e.push({params:this.biases.w,grads:this.biases.dw,l1_decay_mul:0,l2_decay_mul:0});return e},toJSON:function(){var f={};f.sx=this.sx;f.sy=this.sy;f.stride=this.stride;f.in_depth=this.in_depth;f.out_depth=this.out_depth;f.out_sx=this.out_sx;f.out_sy=this.out_sy;f.layer_type=this.layer_type;f.l1_decay_mul=this.l1_decay_mul;f.l2_decay_mul=this.l2_decay_mul;f.pad=this.pad;f.filters=[];for(var e=0;e<this.filters.length;e++){f.filters.push(this.filters[e].toJSON())}f.biases=this.biases.toJSON();return f},fromJSON:function(g){this.out_depth=g.out_depth;this.out_sx=g.out_sx;this.out_sy=g.out_sy;this.layer_type=g.layer_type;this.sx=g.sx;this.sy=g.sy;this.stride=g.stride;this.in_depth=g.in_depth;this.filters=[];this.l1_decay_mul=typeof g.l1_decay_mul!=="undefined"?g.l1_decay_mul:1;this.l2_decay_mul=typeof g.l2_decay_mul!=="undefined"?g.l2_decay_mul:1;this.pad=typeof g.pad!=="undefined"?g.pad:0;for(var f=0;f<g.filters.length;f++){var e=new a(0,0,0,0);e.fromJSON(g.filters[f]);this.filters.push(e)}this.biases=new a(0,0,0,0);this.biases.fromJSON(g.biases)}};var b=function(g){var g=g||{};this.out_depth=typeof g.num_neurons!=="undefined"?g.num_neurons:g.filters;this.l1_decay_mul=typeof g.l1_decay_mul!=="undefined"?g.l1_decay_mul:0;this.l2_decay_mul=typeof g.l2_decay_mul!=="undefined"?g.l2_decay_mul:1;this.num_inputs=g.in_sx*g.in_sy*g.in_depth;this.out_sx=1;this.out_sy=1;this.layer_type="fc";var e=typeof g.bias_pref!=="undefined"?g.bias_pref:0;this.filters=[];for(var f=0;f<this.out_depth;f++){this.filters.push(new a(1,1,this.num_inputs))}this.biases=new a(1,1,this.out_depth,e)};b.prototype={forward:function(h,l){this.in_act=h;var f=new a(1,1,this.out_depth,0);var k=h.w;for(var j=0;j<this.out_depth;j++){var g=0;var e=this.filters[j].w;for(var m=0;m<this.num_inputs;m++){g+=k[m]*e[m]}g+=this.biases.w[j];f.w[j]=g}this.out_act=f;return this.out_act},backward:function(){var e=this.in_act;e.dw=c.zeros(e.w.length);for(var f=0;f<this.out_depth;f++){var h=this.filters[f];var g=this.out_act.dw[f];for(var j=0;j<this.num_inputs;j++){e.dw[j]+=h.w[j]*g;h.dw[j]+=e.w[j]*g}this.biases.dw[f]+=g}},getParamsAndGrads:function(){var e=[];for(var f=0;f<this.out_depth;f++){e.push({params:this.filters[f].w,grads:this.filters[f].dw,l1_decay_mul:this.l1_decay_mul,l2_decay_mul:this.l2_decay_mul})}e.push({params:this.biases.w,grads:this.biases.dw,l1_decay_mul:0,l2_decay_mul:0});return e},toJSON:function(){var f={};f.out_depth=this.out_depth;f.out_sx=this.out_sx;f.out_sy=this.out_sy;f.layer_type=this.layer_type;f.num_inputs=this.num_inputs;f.l1_decay_mul=this.l1_decay_mul;f.l2_decay_mul=this.l2_decay_mul;f.filters=[];for(var e=0;e<this.filters.length;e++){f.filters.push(this.filters[e].toJSON())}f.biases=this.biases.toJSON();return f},fromJSON:function(g){this.out_depth=g.out_depth;this.out_sx=g.out_sx;this.out_sy=g.out_sy;this.layer_type=g.layer_type;this.num_inputs=g.num_inputs;this.l1_decay_mul=typeof g.l1_decay_mul!=="undefined"?g.l1_decay_mul:1;this.l2_decay_mul=typeof g.l2_decay_mul!=="undefined"?g.l2_decay_mul:1;this.filters=[];for(var f=0;f<g.filters.length;f++){var e=new a(0,0,0,0);e.fromJSON(g.filters[f]);this.filters.push(e)}this.biases=new a(0,0,0,0);this.biases.fromJSON(g.biases)}};c.ConvLayer=d;c.FullyConnLayer=b})(convnetjs);(function(c){var a=c.Vol;var b=function(d){var d=d||{};this.sx=d.sx;this.in_depth=d.in_depth;this.in_sx=d.in_sx;this.in_sy=d.in_sy;this.sy=typeof d.sy!=="undefined"?d.sy:this.sx;this.stride=typeof d.stride!=="undefined"?d.stride:2;this.pad=typeof d.pad!=="undefined"?d.pad:0;this.out_depth=this.in_depth;this.out_sx=Math.floor((this.in_sx+this.pad*2-this.sx)/this.stride+1);this.out_sy=Math.floor((this.in_sy+this.pad*2-this.sy)/this.stride+1);this.layer_type="pool";this.switchx=c.zeros(this.out_sx*this.out_sy*this.out_depth);this.switchy=c.zeros(this.out_sx*this.out_sy*this.out_depth)};b.prototype={forward:function(l,u){this.in_act=l;var h=new a(this.out_sx,this.out_sy,this.out_depth,0);var i=0;for(var p=0;p<this.out_depth;p++){var s=-this.pad;var q=-this.pad;for(var e=0;e<this.out_sx;s+=this.stride,e++){q=-this.pad;for(var w=0;w<this.out_sy;q+=this.stride,w++){var r=-99999;var o=-1,k=-1;for(var m=0;m<this.sx;m++){for(var j=0;j<this.sy;j++){var f=q+j;var g=s+m;if(f>=0&&f<l.sy&&g>=0&&g<l.sx){var t=l.get(g,f,p);if(t>r){r=t;o=g;k=f}}}}this.switchx[i]=o;this.switchy[i]=k;i++;h.set(e,w,p,r)}}}this.out_act=h;return this.out_act},backward:function(){var h=this.in_act;h.dw=c.zeros(h.w.length);var f=this.out_act;var g=0;for(var j=0;j<this.out_depth;j++){var l=-this.pad;var k=-this.pad;for(var e=0;e<this.out_sx;l+=this.stride,e++){k=-this.pad;for(var m=0;m<this.out_sy;k+=this.stride,m++){var i=this.out_act.get_grad(e,m,j);h.add_grad(this.switchx[g],this.switchy[g],j,i);g++}}}},getParamsAndGrads:function(){return[]},toJSON:function(){var d={};d.sx=this.sx;d.sy=this.sy;d.stride=this.stride;d.in_depth=this.in_depth;d.out_depth=this.out_depth;d.out_sx=this.out_sx;d.out_sy=this.out_sy;d.layer_type=this.layer_type;d.pad=this.pad;return d},fromJSON:function(d){this.out_depth=d.out_depth;this.out_sx=d.out_sx;this.out_sy=d.out_sy;this.layer_type=d.layer_type;this.sx=d.sx;this.sy=d.sy;this.stride=d.stride;this.in_depth=d.in_depth;this.pad=typeof d.pad!=="undefined"?d.pad:0;this.switchx=c.zeros(this.out_sx*this.out_sy*this.out_depth);this.switchy=c.zeros(this.out_sx*this.out_sy*this.out_depth)}};c.PoolLayer=b})(convnetjs);(function(c){var a=c.Vol;var b=function(d){var d=d||{};this.out_sx=typeof d.out_sx!=="undefined"?d.out_sx:d.in_sx;this.out_sy=typeof d.out_sy!=="undefined"?d.out_sy:d.in_sy;this.out_depth=typeof d.out_depth!=="undefined"?d.out_depth:d.in_depth;this.layer_type="input"};b.prototype={forward:function(d,e){this.in_act=d;this.out_act=d;return this.out_act},backward:function(){},getParamsAndGrads:function(){return[]},toJSON:function(){var d={};d.out_depth=this.out_depth;d.out_sx=this.out_sx;d.out_sy=this.out_sy;d.layer_type=this.layer_type;return d},fromJSON:function(d){this.out_depth=d.out_depth;this.out_sx=d.out_sx;this.out_sy=d.out_sy;this.layer_type=d.layer_type}};c.InputLayer=b})(convnetjs);(function(e){var a=e.Vol;var c=function(f){var f=f||{};this.num_inputs=f.in_sx*f.in_sy*f.in_depth;this.out_depth=this.num_inputs;this.out_sx=1;this.out_sy=1;this.layer_type="softmax"};c.prototype={forward:function(h,o){this.in_act=h;var f=new a(1,1,this.out_depth,0);var j=h.w;var k=h.w[0];for(var l=1;l<this.out_depth;l++){if(j[l]>k){k=j[l]}}var n=e.zeros(this.out_depth);var g=0;for(var l=0;l<this.out_depth;l++){var m=Math.exp(j[l]-k);g+=m;n[l]=m}for(var l=0;l<this.out_depth;l++){n[l]/=g;f.w[l]=n[l]}this.es=n;this.out_act=f;return this.out_act},backward:function(k){var f=this.in_act;f.dw=e.zeros(f.w.length);for(var h=0;h<this.out_depth;h++){var g=h===k?1:0;var j=-(g-this.es[h]);f.dw[h]=j}return -Math.log(this.es[k])},getParamsAndGrads:function(){return[]},toJSON:function(){var f={};f.out_depth=this.out_depth;f.out_sx=this.out_sx;f.out_sy=this.out_sy;f.layer_type=this.layer_type;f.num_inputs=this.num_inputs;return f},fromJSON:function(f){this.out_depth=f.out_depth;this.out_sx=f.out_sx;this.out_sy=f.out_sy;this.layer_type=f.layer_type;this.num_inputs=f.num_inputs}};var d=function(f){var f=f||{};this.num_inputs=f.in_sx*f.in_sy*f.in_depth;this.out_depth=this.num_inputs;this.out_sx=1;this.out_sy=1;this.layer_type="regression"};d.prototype={forward:function(f,g){this.in_act=f;this.out_act=f;return f},backward:function(l){var f=this.in_act;f.dw=e.zeros(f.w.length);var k=0;if(l instanceof Array||l instanceof Float64Array){for(var j=0;j<this.out_depth;j++){var g=f.w[j]-l[j];f.dw[j]=g;k+=2*g*g}}else{var j=l.dim;var h=l.val;var g=f.w[j]-h;f.dw[j]=g;k+=2*g*g}return k},getParamsAndGrads:function(){return[]},toJSON:function(){var f={};f.out_depth=this.out_depth;f.out_sx=this.out_sx;f.out_sy=this.out_sy;f.layer_type=this.layer_type;f.num_inputs=this.num_inputs;return f},fromJSON:function(f){this.out_depth=f.out_depth;this.out_sx=f.out_sx;this.out_sy=f.out_sy;this.layer_type=f.layer_type;this.num_inputs=f.num_inputs}};var b=function(f){var f=f||{};this.num_inputs=f.in_sx*f.in_sy*f.in_depth;this.out_depth=this.num_inputs;this.out_sx=1;this.out_sy=1;this.layer_type="svm"};b.prototype={forward:function(f,g){this.in_act=f;this.out_act=f;return f},backward:function(l){var g=this.in_act;g.dw=e.zeros(g.w.length);var f=g.w[l];var k=1;var j=0;for(var h=0;h<this.out_depth;h++){if(-f+g.w[h]+k>0){g.dw[h]+=1;g.dw[l]-=1;j+=-f+g.w[h]+k}}return j},getParamsAndGrads:function(){return[]},toJSON:function(){var f={};f.out_depth=this.out_depth;f.out_sx=this.out_sx;f.out_sy=this.out_sy;f.layer_type=this.layer_type;f.num_inputs=this.num_inputs;return f},fromJSON:function(f){this.out_depth=f.out_depth;this.out_sx=f.out_sx;this.out_sy=f.out_sy;this.layer_type=f.layer_type;this.num_inputs=f.num_inputs}};e.RegressionLayer=d;e.SoftmaxLayer=c;e.SVMLayer=b})(convnetjs);(function(d){var a=d.Vol;var e=function(h){var h=h||{};this.out_sx=h.in_sx;this.out_sy=h.in_sy;this.out_depth=h.in_depth;this.layer_type="relu"};e.prototype={forward:function(j,l){this.in_act=j;var h=j.clone();var m=j.w.length;var n=h.w;for(var k=0;k<m;k++){if(n[k]<0){n[k]=0}}this.out_act=h;return this.out_act},backward:function(){var j=this.in_act;var h=this.out_act;var l=j.w.length;j.dw=d.zeros(l);for(var k=0;k<l;k++){if(h.w[k]<=0){j.dw[k]=0}else{j.dw[k]=h.dw[k]}}},getParamsAndGrads:function(){return[]},toJSON:function(){var h={};h.out_depth=this.out_depth;h.out_sx=this.out_sx;h.out_sy=this.out_sy;h.layer_type=this.layer_type;return h},fromJSON:function(h){this.out_depth=h.out_depth;this.out_sx=h.out_sx;this.out_sy=h.out_sy;this.layer_type=h.layer_type}};var g=function(h){var h=h||{};this.out_sx=h.in_sx;this.out_sy=h.in_sy;this.out_depth=h.in_depth;this.layer_type="sigmoid"};g.prototype={forward:function(j,m){this.in_act=j;var h=j.cloneAndZero();var n=j.w.length;var o=h.w;var l=j.w;for(var k=0;k<n;k++){o[k]=1/(1+Math.exp(-l[k]))}this.out_act=h;return this.out_act},backward:function(){var j=this.in_act;var h=this.out_act;var m=j.w.length;j.dw=d.zeros(m);for(var k=0;k<m;k++){var l=h.w[k];j.dw[k]=l*(1-l)*h.dw[k]}},getParamsAndGrads:function(){return[]},toJSON:function(){var h={};h.out_depth=this.out_depth;h.out_sx=this.out_sx;h.out_sy=this.out_sy;h.layer_type=this.layer_type;return h},fromJSON:function(h){this.out_depth=h.out_depth;this.out_sx=h.out_sx;this.out_sy=h.out_sy;this.layer_type=h.layer_type}};var f=function(h){var h=h||{};this.group_size=typeof h.group_size!=="undefined"?h.group_size:2;this.out_sx=h.in_sx;this.out_sy=h.in_sy;this.out_depth=Math.floor(h.in_depth/this.group_size);this.layer_type="maxout";this.switches=d.zeros(this.out_sx*this.out_sy*this.out_depth)};f.prototype={forward:function(l,w){this.in_act=l;var q=this.out_depth;var v=new a(this.out_sx,this.out_sy,this.out_depth,0);if(this.out_sx===1&&this.out_sy===1){for(var p=0;p<q;p++){var m=p*this.group_size;var u=l.w[m];var r=0;for(var o=1;o<this.group_size;o++){var h=l.w[m+o];if(h>u){u=h;r=o}}v.w[p]=u;this.switches[p]=m+r}}else{var k=0;for(var t=0;t<l.sx;t++){for(var s=0;s<l.sy;s++){for(var p=0;p<q;p++){var m=p*this.group_size;var u=l.get(t,s,m);var r=0;for(var o=1;o<this.group_size;o++){var h=l.get(t,s,m+o);if(h>u){u=h;r=o}}v.set(t,s,p,u);this.switches[k]=m+r;k++}}}}this.out_act=v;return this.out_act},backward:function(){var k=this.in_act;var j=this.out_act;var o=this.out_depth;k.dw=d.zeros(k.w.length);if(this.out_sx===1&&this.out_sy===1){for(var l=0;l<o;l++){var m=j.dw[l];k.dw[this.switches[l]]=m}}else{var q=0;for(var h=0;h<j.sx;h++){for(var p=0;p<j.sy;p++){for(var l=0;l<o;l++){var m=j.get_grad(h,p,l);k.set_grad(h,p,this.switches[q],m);q++}}}}},getParamsAndGrads:function(){return[]},toJSON:function(){var h={};h.out_depth=this.out_depth;h.out_sx=this.out_sx;h.out_sy=this.out_sy;h.layer_type=this.layer_type;h.group_size=this.group_size;return h},fromJSON:function(h){this.out_depth=h.out_depth;this.out_sx=h.out_sx;this.out_sy=h.out_sy;this.layer_type=h.layer_type;this.group_size=h.group_size;this.switches=d.zeros(this.group_size)}};function c(h){var i=Math.exp(2*h);return(i-1)/(i+1)}var b=function(h){var h=h||{};this.out_sx=h.in_sx;this.out_sy=h.in_sy;this.out_depth=h.in_depth;this.layer_type="tanh"};b.prototype={forward:function(j,l){this.in_act=j;var h=j.cloneAndZero();var m=j.w.length;for(var k=0;k<m;k++){h.w[k]=c(j.w[k])}this.out_act=h;return this.out_act},backward:function(){var j=this.in_act;var h=this.out_act;var m=j.w.length;j.dw=d.zeros(m);for(var k=0;k<m;k++){var l=h.w[k];j.dw[k]=(1-l*l)*h.dw[k]}},getParamsAndGrads:function(){return[]},toJSON:function(){var h={};h.out_depth=this.out_depth;h.out_sx=this.out_sx;h.out_sy=this.out_sy;h.layer_type=this.layer_type;return h},fromJSON:function(h){this.out_depth=h.out_depth;this.out_sx=h.out_sx;this.out_sy=h.out_sy;this.layer_type=h.layer_type}};d.TanhLayer=b;d.MaxoutLayer=f;d.ReluLayer=e;d.SigmoidLayer=g})(convnetjs);(function(c){var a=c.Vol;var b=function(d){var d=d||{};this.out_sx=d.in_sx;this.out_sy=d.in_sy;this.out_depth=d.in_depth;this.layer_type="dropout";this.drop_prob=typeof d.drop_prob!=="undefined"?d.drop_prob:0.5;this.dropped=c.zeros(this.out_sx*this.out_sy*this.out_depth)};b.prototype={forward:function(e,g){this.in_act=e;if(typeof(g)==="undefined"){g=false}var d=e.clone();var h=e.w.length;if(g){for(var f=0;f<h;f++){if(Math.random()<this.drop_prob){d.w[f]=0;this.dropped[f]=true}else{this.dropped[f]=false}}}else{for(var f=0;f<h;f++){d.w[f]*=this.drop_prob}}this.out_act=d;return this.out_act},backward:function(){var d=this.in_act;var f=this.out_act;var g=d.w.length;d.dw=c.zeros(g);for(var e=0;e<g;e++){if(!(this.dropped[e])){d.dw[e]=f.dw[e]}}},getParamsAndGrads:function(){return[]},toJSON:function(){var d={};d.out_depth=this.out_depth;d.out_sx=this.out_sx;d.out_sy=this.out_sy;d.layer_type=this.layer_type;d.drop_prob=this.drop_prob;return d},fromJSON:function(d){this.out_depth=d.out_depth;this.out_sx=d.out_sx;this.out_sy=d.out_sy;this.layer_type=d.layer_type;this.drop_prob=d.drop_prob}};c.DropoutLayer=b})(convnetjs);(function(c){var a=c.Vol;var b=function(d){var d=d||{};this.k=d.k;this.n=d.n;this.alpha=d.alpha;this.beta=d.beta;this.out_sx=d.in_sx;this.out_sy=d.in_sy;this.out_depth=d.in_depth;this.layer_type="lrn";if(this.n%2===0){console.log("WARNING n should be odd for LRN layer")}};b.prototype={forward:function(f,p){this.in_act=f;var e=f.cloneAndZero();this.S_cache_=f.cloneAndZero();var k=Math.floor(this.n/2);for(var n=0;n<f.sx;n++){for(var m=0;m<f.sy;m++){for(var h=0;h<f.depth;h++){var l=f.get(n,m,h);var o=0;for(var g=Math.max(0,h-k);g<=Math.min(h+k,f.depth-1);g++){var d=f.get(n,m,g);o+=d*d}o*=this.alpha/this.n;o+=this.k;this.S_cache_.set(n,m,h,o);o=Math.pow(o,this.beta);e.set(n,m,h,l/o)}}}this.out_act=e;return this.out_act},backward:function(){var f=this.in_act;f.dw=c.zeros(f.w.length);var d=this.out_act;var n=Math.floor(this.n/2);for(var r=0;r<f.sx;r++){for(var q=0;q<f.sy;q++){for(var l=0;l<f.depth;l++){var p=this.out_act.get_grad(r,q,l);var k=this.S_cache_.get(r,q,l);var e=Math.pow(k,this.beta);var s=e*e;for(var h=Math.max(0,l-n);h<=Math.min(l+n,f.depth-1);h++){var o=f.get(r,q,h);var m=-o*this.beta*Math.pow(k,this.beta-1)*this.alpha/this.n*2*o;if(h===l){m+=e}m/=s;m*=p;f.add_grad(r,q,h,m)}}}}},getParamsAndGrads:function(){return[]},toJSON:function(){var d={};d.k=this.k;d.n=this.n;d.alpha=this.alpha;d.beta=this.beta;d.out_sx=this.out_sx;d.out_sy=this.out_sy;d.out_depth=this.out_depth;d.layer_type=this.layer_type;return d},fromJSON:function(d){this.k=d.k;this.n=d.n;this.alpha=d.alpha;this.beta=d.beta;this.out_sx=d.out_sx;this.out_sy=d.out_sy;this.out_depth=d.out_depth;this.layer_type=d.layer_type}};c.LocalResponseNormalizationLayer=b})(convnetjs);(function(c){var a=c.Vol;var b=function(d){var d=d||{};this.out_sx=d.in_sx;this.out_sy=d.in_sy;this.out_depth=d.in_depth+d.in_depth*d.in_depth;this.layer_type="quadtransform"};b.prototype={forward:function(d,n){this.in_act=d;var j=this.out_depth;var e=d.depth;var m=new a(this.out_sx,this.out_sy,this.out_depth,0);for(var l=0;l<d.sx;l++){for(var k=0;k<d.sy;k++){for(var h=0;h<j;h++){if(h<e){m.set(l,k,h,d.get(l,k,h))}else{var g=Math.floor((h-e)/e);var f=(h-e)-g*e;m.set(l,k,h,d.get(l,k,g)*d.get(l,k,f))}}}}this.out_act=m;return this.out_act},backward:function(){var d=this.in_act;d.dw=c.zeros(d.w.length);var n=this.out_act;var j=this.out_depth;var e=d.depth;for(var m=0;m<d.sx;m++){for(var l=0;l<d.sy;l++){for(var h=0;h<j;h++){var k=n.get_grad(m,l,h);if(h<e){d.add_grad(m,l,h,k)}else{var g=Math.floor((h-e)/e);var f=(h-e)-g*e;d.add_grad(m,l,g,d.get(m,l,f)*k);d.add_grad(m,l,f,d.get(m,l,g)*k)}}}}},getParamsAndGrads:function(){return[]},toJSON:function(){var d={};d.out_depth=this.out_depth;d.out_sx=this.out_sx;d.out_sy=this.out_sy;d.layer_type=this.layer_type;return d},fromJSON:function(d){this.out_depth=d.out_depth;this.out_sx=d.out_sx;this.out_sy=d.out_sy;this.layer_type=d.layer_type}};c.QuadTransformLayer=b})(convnetjs);(function(c){var a=c.Vol;var b=function(d){this.layers=[]};b.prototype={makeLayers:function(d){if(d.length<2){console.log("ERROR! For now at least have input and softmax layers.")}if(d[0].type!=="input"){console.log("ERROR! For now first layer should be input.")}var e=function(){var l=[];for(var k=0;k<d.length;k++){var m=d[k];if(m.type==="softmax"||m.type==="svm"){l.push({type:"fc",num_neurons:m.num_classes})}if(m.type==="regression"){l.push({type:"fc",num_neurons:m.num_neurons})}if((m.type==="fc"||m.type==="conv")&&typeof(m.bias_pref)==="undefined"){m.bias_pref=0;if(typeof m.activation!=="undefined"&&m.activation==="relu"){m.bias_pref=0.1}}if(typeof m.tensor!=="undefined"){if(m.tensor){l.push({type:"quadtransform"})}}l.push(m);if(typeof m.activation!=="undefined"){if(m.activation==="relu"){l.push({type:"relu"})}else{if(m.activation==="sigmoid"){l.push({type:"sigmoid"})}else{if(m.activation==="tanh"){l.push({type:"tanh"})}else{if(m.activation==="maxout"){var j=m.group_size!=="undefined"?m.group_size:2;l.push({type:"maxout",group_size:j})}else{console.log("ERROR unsupported activation "+m.activation)}}}}}if(typeof m.drop_prob!=="undefined"&&m.type!=="dropout"){l.push({type:"dropout",drop_prob:m.drop_prob})}}return l};d=e(d);this.layers=[];for(var f=0;f<d.length;f++){var h=d[f];if(f>0){var g=this.layers[f-1];h.in_sx=g.out_sx;h.in_sy=g.out_sy;h.in_depth=g.out_depth}switch(h.type){case"fc":this.layers.push(new c.FullyConnLayer(h));break;case"lrn":this.layers.push(new c.LocalResponseNormalizationLayer(h));break;case"dropout":this.layers.push(new c.DropoutLayer(h));break;case"input":this.layers.push(new c.InputLayer(h));break;case"softmax":this.layers.push(new c.SoftmaxLayer(h));break;case"regression":this.layers.push(new c.RegressionLayer(h));break;case"conv":this.layers.push(new c.ConvLayer(h));break;case"pool":this.layers.push(new c.PoolLayer(h));break;case"relu":this.layers.push(new c.ReluLayer(h));break;case"sigmoid":this.layers.push(new c.SigmoidLayer(h));break;case"tanh":this.layers.push(new c.TanhLayer(h));break;case"maxout":this.layers.push(new c.MaxoutLayer(h));break;case"quadtransform":this.layers.push(new c.QuadTransformLayer(h));break;case"svm":this.layers.push(new c.SVMLayer(h));break;default:console.log("ERROR: UNRECOGNIZED LAYER TYPE!")}}},forward:function(e,g){if(typeof(g)==="undefined"){g=false}var d=this.layers[0].forward(e,g);for(var f=1;f<this.layers.length;f++){d=this.layers[f].forward(d,g)}return d},getCostLoss:function(d,g){this.forward(d,false);var f=this.layers.length;var e=this.layers[f-1].backward(g);return e},backward:function(g){var f=this.layers.length;var e=this.layers[f-1].backward(g);for(var d=f-2;d>=0;d--){this.layers[d].backward()}return e},getParamsAndGrads:function(){var d=[];for(var f=0;f<this.layers.length;f++){var g=this.layers[f].getParamsAndGrads();for(var e=0;e<g.length;e++){d.push(g[e])}}return d},getPrediction:function(){var g=this.layers[this.layers.length-1];var h=g.out_act.w;var d=h[0];var e=0;for(var f=1;f<h.length;f++){if(h[f]>d){d=h[f];e=f}}return e},toJSON:function(){var e={};e.layers=[];for(var d=0;d<this.layers.length;d++){e.layers.push(this.layers[d].toJSON())}return e},fromJSON:function(h){this.layers=[];for(var g=0;g<h.layers.length;g++){var e=h.layers[g];var f=e.layer_type;var d;if(f==="input"){d=new c.InputLayer()}if(f==="relu"){d=new c.ReluLayer()}if(f==="sigmoid"){d=new c.SigmoidLayer()}if(f==="tanh"){d=new c.TanhLayer()}if(f==="dropout"){d=new c.DropoutLayer()}if(f==="conv"){d=new c.ConvLayer()}if(f==="pool"){d=new c.PoolLayer()}if(f==="lrn"){d=new c.LocalResponseNormalizationLayer()}if(f==="softmax"){d=new c.SoftmaxLayer()}if(f==="regression"){d=new c.RegressionLayer()}if(f==="fc"){d=new c.FullyConnLayer()}if(f==="maxout"){d=new c.MaxoutLayer()}if(f==="quadtransform"){d=new c.QuadTransformLayer()}if(f==="svm"){d=new c.SVMLayer()}d.fromJSON(e);this.layers.push(d)}}};c.Net=b})(convnetjs);(function(b){var a=b.Vol;var c=function(e,d){this.net=e;var d=d||{};this.learning_rate=typeof d.learning_rate!=="undefined"?d.learning_rate:0.01;this.l1_decay=typeof d.l1_decay!=="undefined"?d.l1_decay:0;this.l2_decay=typeof d.l2_decay!=="undefined"?d.l2_decay:0;this.batch_size=typeof d.batch_size!=="undefined"?d.batch_size:1;this.method=typeof d.method!=="undefined"?d.method:"sgd";this.momentum=typeof d.momentum!=="undefined"?d.momentum:0.9;this.ro=typeof d.ro!=="undefined"?d.ro:0.95;this.eps=typeof d.eps!=="undefined"?d.eps:0.000001;this.k=0;this.gsum=[];this.xsum=[]};c.prototype={train:function(s,r){var h=new Date().getTime();this.net.forward(s,true);var f=new Date().getTime();var q=f-h;var h=new Date().getTime();var A=this.net.backward(r);var k=0;var d=0;var f=new Date().getTime();var G=f-h;this.k++;if(this.k%this.batch_size===0){var e=this.net.getParamsAndGrads();if(this.gsum.length===0&&(this.method!=="sgd"||this.momentum>0)){for(var E=0;E<e.length;E++){this.gsum.push(b.zeros(e[E].params.length));if(this.method==="adadelta"){this.xsum.push(b.zeros(e[E].params.length))}else{this.xsum.push([])}}}for(var E=0;E<e.length;E++){var H=e[E];var w=H.params;var F=H.grads;var z=typeof H.l2_decay_mul!=="undefined"?H.l2_decay_mul:1;var I=typeof H.l1_decay_mul!=="undefined"?H.l1_decay_mul:1;var l=this.l2_decay*z;var n=this.l1_decay*I;var u=w.length;for(var B=0;B<u;B++){k+=l*w[B]*w[B]/2;d+=n*Math.abs(w[B]);var D=n*(w[B]>0?1:-1);var o=l*(w[B]);var t=(o+D+F[B])/this.batch_size;var m=this.gsum[E];var C=this.xsum[E];if(this.method==="adagrad"){m[B]=m[B]+t*t;var v=-this.learning_rate/Math.sqrt(m[B]+this.eps)*t;w[B]+=v}else{if(this.method==="windowgrad"){m[B]=this.ro*m[B]+(1-this.ro)*t*t;var v=-this.learning_rate/Math.sqrt(m[B]+this.eps)*t;w[B]+=v}else{if(this.method==="adadelta"){m[B]=this.ro*m[B]+(1-this.ro)*t*t;var v=-Math.sqrt((C[B]+this.eps)/(m[B]+this.eps))*t;C[B]=this.ro*C[B]+(1-this.ro)*v*v;w[B]+=v}else{if(this.momentum>0){var v=this.momentum*m[B]-this.learning_rate*t;m[B]=v;w[B]+=v}else{w[B]+=-this.learning_rate*t}}}}F[B]=0}}}return{fwd_time:q,bwd_time:G,l2_decay_loss:k,l1_decay_loss:d,cost_loss:A,softmax_loss:A,loss:A+d+k}}};b.Trainer=c;b.SGDTrainer=c})(convnetjs);(function(c){var e=c.randf;var d=c.randi;var j=c.Net;var g=c.Trainer;var b=c.maxmin;var h=c.randperm;var f=c.weightedSample;var i=c.getopt;var k=c.arrUnique;var a=function(m,n,l){var l=l||{};if(typeof m==="undefined"){m=[]}if(typeof n==="undefined"){n=[]}this.data=m;this.labels=n;this.train_ratio=i(l,"train_ratio",0.7);this.num_folds=i(l,"num_folds",10);this.num_candidates=i(l,"num_candidates",50);this.num_epochs=i(l,"num_epochs",50);this.ensemble_size=i(l,"ensemble_size",10);this.batch_size_min=i(l,"batch_size_min",10);this.batch_size_max=i(l,"batch_size_max",300);this.l2_decay_min=i(l,"l2_decay_min",-4);this.l2_decay_max=i(l,"l2_decay_max",2);this.learning_rate_min=i(l,"learning_rate_min",-4);this.learning_rate_max=i(l,"learning_rate_max",0);this.momentum_min=i(l,"momentum_min",0.9);this.momentum_max=i(l,"momentum_max",0.9);this.neurons_min=i(l,"neurons_min",5);this.neurons_max=i(l,"neurons_max",30);this.folds=[];this.candidates=[];this.evaluated_candidates=[];this.unique_labels=k(n);this.iter=0;this.foldix=0;this.finish_fold_callback=null;this.finish_batch_callback=null;if(this.data.length>0){this.sampleFolds();this.sampleCandidates()}};a.prototype={sampleFolds:function(){var o=this.data.length;var m=Math.floor(this.train_ratio*o);this.folds=[];for(var l=0;l<this.num_folds;l++){var n=h(o);this.folds.push({train_ix:n.slice(0,m),test_ix:n.slice(m,o)})}},sampleCandidate:function(){var A=this.data[0].w.length;var z=this.unique_labels.length;var s=[];s.push({type:"input",out_sx:1,out_sy:1,out_depth:A});var l=f([0,1,2,3],[0.2,0.3,0.3,0.2]);for(var m=0;m<l;m++){var n=d(this.neurons_min,this.neurons_max);var v=["tanh","maxout","relu"][d(0,3)];if(e(0,1)<0.5){var r=Math.random();s.push({type:"fc",num_neurons:n,activation:v,drop_prob:r})}else{s.push({type:"fc",num_neurons:n,activation:v})}}s.push({type:"softmax",num_classes:z});var x=new j();x.makeLayers(s);var C=d(this.batch_size_min,this.batch_size_max);var o=Math.pow(10,e(this.l2_decay_min,this.l2_decay_max));var t=Math.pow(10,e(this.learning_rate_min,this.learning_rate_max));var p=e(this.momentum_min,this.momentum_max);var y=e(0,1);var u;if(y<0.33){u={method:"adadelta",batch_size:C,l2_decay:o}}else{if(y<0.66){u={method:"adagrad",learning_rate:t,batch_size:C,l2_decay:o}}else{u={method:"sgd",learning_rate:t,momentum:p,batch_size:C,l2_decay:o}}}var B=new g(x,u);var w={};w.acc=[];w.accv=0;w.layer_defs=s;w.trainer_def=u;w.net=x;w.trainer=B;return w},sampleCandidates:function(){this.candidates=[];for(var l=0;l<this.num_candidates;l++){var m=this.sampleCandidate();this.candidates.push(m)}},step:function(){this.iter++;var r=this.folds[this.foldix];var p=r.train_ix[d(0,r.train_ix.length)];for(var q=0;q<this.candidates.length;q++){var u=this.data[p];var o=this.labels[p];this.candidates[q].trainer.train(u,o)}var n=this.num_epochs*r.train_ix.length;if(this.iter>=n){var m=this.evalValErrors();for(var q=0;q<this.candidates.length;q++){var s=this.candidates[q];s.acc.push(m[q]);s.accv+=m[q]}this.iter=0;this.foldix++;if(this.finish_fold_callback!==null){this.finish_fold_callback()}if(this.foldix>=this.folds.length){for(var q=0;q<this.candidates.length;q++){this.evaluated_candidates.push(this.candidates[q])}this.evaluated_candidates.sort(function(w,l){return(w.accv/w.acc.length)>(l.accv/l.acc.length)?-1:1});if(this.evaluated_candidates.length>3*this.ensemble_size){this.evaluated_candidates=this.evaluated_candidates.slice(0,3*this.ensemble_size)}if(this.finish_batch_callback!==null){this.finish_batch_callback()}this.sampleCandidates();this.foldix=0}else{for(var q=0;q<this.candidates.length;q++){var s=this.candidates[q];var t=new j();t.makeLayers(s.layer_defs);var v=new g(t,s.trainer_def);s.net=t;s.trainer=v}}}},evalValErrors:function(){var t=[];var r=this.folds[this.foldix];for(var p=0;p<this.candidates.length;p++){var s=this.candidates[p].net;var w=0;for(var m=0;m<r.test_ix.length;m++){var u=this.data[r.test_ix[m]];var o=this.labels[r.test_ix[m]];s.forward(u);var n=s.getPrediction();w+=(n===o?1:0)}w/=r.test_ix.length;t.push(w)}return t},predict_soft:function(q){var m=Math.min(this.ensemble_size,this.evaluated_candidates.length);if(m===0){return new convnetjs.Vol(0,0,0)}var p,t;for(var o=0;o<m;o++){var r=this.evaluated_candidates[o].net;var l=r.forward(q);if(o===0){p=l;t=l.w.length}else{for(var s=0;s<t;s++){p.w[s]+=l.w[s]}}}for(var s=0;s<t;s++){p.w[s]/=t}return p},predict:function(n){var m=this.predict_soft(n);if(m.w.length!==0){var l=b(m.w);var o=l.maxi}else{var o=-1}return o},toJSON:function(){var l=Math.min(this.ensemble_size,this.evaluated_candidates.length);var n={};n.nets=[];for(var m=0;m<l;m++){n.nets.push(this.evaluated_candidates[m].net.toJSON())}return n},fromJSON:function(m){this.ensemble_size=m.nets.length;this.evaluated_candidates=[];for(var l=0;l<this.ensemble_size;l++){var n=new j();n.fromJSON(m.nets[l]);var o={};o.net=n;this.evaluated_candidates.push(o)}},onFinishFold:function(l){this.finish_fold_callback=l},onFinishBatch:function(l){this.finish_batch_callback=l}};c.MagicNet=a})(convnetjs);(function(a){if(typeof module==="undefined"||typeof module.exports==="undefined"){window.jsfeat=a}else{module.exports=a}})(convnetjs);
\ No newline at end of file
diff --git a/build/convnet.js b/build/convnet.js
deleted file mode 100644
index 16b2a3e6..00000000
--- a/build/convnet.js
+++ /dev/null
@@ -1,2205 +0,0 @@
-var convnetjs = convnetjs || { REVISION: 'ALPHA' };
-(function(global) {
-  "use strict";
-
-  // Random number utilities
-  var return_v = false;
-  var v_val = 0.0;
-  var gaussRandom = function() {
-    if(return_v) { 
-      return_v = false;
-      return v_val; 
-    }
-    var u = 2*Math.random()-1;
-    var v = 2*Math.random()-1;
-    var r = u*u + v*v;
-    if(r == 0 || r > 1) return gaussRandom();
-    var c = Math.sqrt(-2*Math.log(r)/r);
-    v_val = v*c; // cache this
-    return_v = true;
-    return u*c;
-  }
-  var randf = function(a, b) { return Math.random()*(b-a)+a; }
-  var randi = function(a, b) { return Math.floor(Math.random()*(b-a)+a); }
-  var randn = function(mu, std){ return mu+gaussRandom()*std; }
-
-  // Array utilities
-  var zeros = function(n) {
-    if(typeof(n)==='undefined' || isNaN(n)) { return []; }
-    if(typeof ArrayBuffer === 'undefined') {
-      // lacking browser support
-      var arr = new Array(n);
-      for(var i=0;i<n;i++) { arr[i]= 0; }
-      return arr;
-    } else {
-      return new Float64Array(n);
-    }
-  }
-
-  var arrContains = function(arr, elt) {
-    for(var i=0,n=arr.length;i<n;i++) {
-      if(arr[i]===elt) return true;
-    }
-    return false;
-  }
-
-  var arrUnique = function(arr) {
-    var b = [];
-    for(var i=0,n=arr.length;i<n;i++) {
-      if(!arrContains(b, arr[i])) {
-        b.push(arr[i]);
-      }
-    }
-    return b;
-  }
-
-  // return max and min of a given non-empty array.
-  var maxmin = function(w) {
-    if(w.length === 0) { return {}; } // ... ;s
-    var maxv = w[0];
-    var minv = w[0];
-    var maxi = 0;
-    var mini = 0;
-    var n = w.length;
-    for(var i=1;i<n;i++) {
-      if(w[i] > maxv) { maxv = w[i]; maxi = i; } 
-      if(w[i] < minv) { minv = w[i]; mini = i; } 
-    }
-    return {maxi: maxi, maxv: maxv, mini: mini, minv: minv, dv:maxv-minv};
-  }
-
-  // create random permutation of numbers, in range [0...n-1]
-  var randperm = function(n) {
-    var i = n,
-        j = 0,
-        temp;
-    var array = [];
-    for(var q=0;q<n;q++)array[q]=q;
-    while (i--) {
-        j = Math.floor(Math.random() * (i+1));
-        temp = array[i];
-        array[i] = array[j];
-        array[j] = temp;
-    }
-    return array;
-  }
-
-  // sample from list lst according to probabilities in list probs
-  // the two lists are of same size, and probs adds up to 1
-  var weightedSample = function(lst, probs) {
-    var p = randf(0, 1.0);
-    var cumprob = 0.0;
-    for(var k=0,n=lst.length;k<n;k++) {
-      cumprob += probs[k];
-      if(p < cumprob) { return lst[k]; }
-    }
-  }
-
-  // syntactic sugar function for getting default parameter values
-  var getopt = function(opt, field_name, default_value) {
-    return typeof opt[field_name] !== 'undefined' ? opt[field_name] : default_value;
-  }
-
-  global.randf = randf;
-  global.randi = randi;
-  global.randn = randn;
-  global.zeros = zeros;
-  global.maxmin = maxmin;
-  global.randperm = randperm;
-  global.weightedSample = weightedSample;
-  global.arrUnique = arrUnique;
-  global.arrContains = arrContains;
-  global.getopt = getopt;
-  
-})(convnetjs);
-(function(global) {
-  "use strict";
-
-  // Vol is the basic building block of all data in a net.
-  // it is essentially just a 3D volume of numbers, with a
-  // width (sx), height (sy), and depth (depth).
-  // it is used to hold data for all filters, all volumes,
-  // all weights, and also stores all gradients w.r.t. 
-  // the data. c is optionally a value to initialize the volume
-  // with. If c is missing, fills the Vol with random numbers.
-  var Vol = function(sx, sy, depth, c) {
-    // this is how you check if a variable is an array. Oh, Javascript :)
-    if(Object.prototype.toString.call(sx) === '[object Array]') {
-      // we were given a list in sx, assume 1D volume and fill it up
-      this.sx = 1;
-      this.sy = 1;
-      this.depth = sx.length;
-      // we have to do the following copy because we want to use
-      // fast typed arrays, not an ordinary javascript array
-      this.w = global.zeros(this.depth);
-      this.dw = global.zeros(this.depth);
-      for(var i=0;i<this.depth;i++) {
-        this.w[i] = sx[i];
-      }
-    } else {
-      // we were given dimensions of the vol
-      this.sx = sx;
-      this.sy = sy;
-      this.depth = depth;
-      var n = sx*sy*depth;
-      this.w = global.zeros(n);
-      this.dw = global.zeros(n);
-      if(typeof c === 'undefined') {
-        // weight normalization is done to equalize the output
-        // variance of every neuron, otherwise neurons with a lot
-        // of incoming connections have outputs of larger variance
-        var scale = Math.sqrt(1.0/(sx*sy*depth));
-        for(var i=0;i<n;i++) { 
-          this.w[i] = global.randn(0.0, scale);
-        }
-      } else {
-        for(var i=0;i<n;i++) { 
-          this.w[i] = c;
-        }
-      }
-    }
-  }
-
-  Vol.prototype = {
-    get: function(x, y, d) { 
-      var ix=((this.sx * y)+x)*this.depth+d;
-      return this.w[ix];
-    },
-    set: function(x, y, d, v) { 
-      var ix=((this.sx * y)+x)*this.depth+d;
-      this.w[ix] = v; 
-    },
-    add: function(x, y, d, v) { 
-      var ix=((this.sx * y)+x)*this.depth+d;
-      this.w[ix] += v; 
-    },
-    get_grad: function(x, y, d) { 
-      var ix = ((this.sx * y)+x)*this.depth+d;
-      return this.dw[ix]; 
-    },
-    set_grad: function(x, y, d, v) { 
-      var ix = ((this.sx * y)+x)*this.depth+d;
-      this.dw[ix] = v; 
-    },
-    add_grad: function(x, y, d, v) { 
-      var ix = ((this.sx * y)+x)*this.depth+d;
-      this.dw[ix] += v; 
-    },
-    cloneAndZero: function() { return new Vol(this.sx, this.sy, this.depth, 0.0)},
-    clone: function() {
-      var V = new Vol(this.sx, this.sy, this.depth, 0.0);
-      var n = this.w.length;
-      for(var i=0;i<n;i++) { V.w[i] = this.w[i]; }
-      return V;
-    },
-    addFrom: function(V) { for(var k=0;k<this.w.length;k++) { this.w[k] += V.w[k]; }},
-    addFromScaled: function(V, a) { for(var k=0;k<this.w.length;k++) { this.w[k] += a*V.w[k]; }},
-    setConst: function(a) { for(var k=0;k<this.w.length;k++) { this.w[k] = a; }},
-
-    toJSON: function() {
-      // todo: we may want to only save d most significant digits to save space
-      var json = {}
-      json.sx = this.sx; 
-      json.sy = this.sy;
-      json.depth = this.depth;
-      json.w = this.w;
-      return json;
-      // we wont back up gradients to save space
-    },
-    fromJSON: function(json) {
-      this.sx = json.sx;
-      this.sy = json.sy;
-      this.depth = json.depth;
-
-      var n = this.sx*this.sy*this.depth;
-      this.w = global.zeros(n);
-      this.dw = global.zeros(n);
-      // copy over the elements.
-      for(var i=0;i<n;i++) {
-        this.w[i] = json.w[i];
-      }
-    }
-  }
-
-  global.Vol = Vol;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-
-  // Volume utilities
-  // intended for use with data augmentation
-  // crop is the size of output
-  // dx,dy are offset wrt incoming volume, of the shift
-  // fliplr is boolean on whether we also want to flip left<->right
-  var augment = function(V, crop, dx, dy, fliplr) {
-    // note assumes square outputs of size crop x crop
-    if(typeof(fliplr)==='undefined') var fliplr = false;
-    if(typeof(dx)==='undefined') var dx = global.randi(0, V.sx - crop);
-    if(typeof(dy)==='undefined') var dy = global.randi(0, V.sy - crop);
-    
-    // randomly sample a crop in the input volume
-    var W;
-    if(crop !== V.sx || dx!==0 || dy!==0) {
-      W = new Vol(crop, crop, V.depth, 0.0);
-      for(var x=0;x<crop;x++) {
-        for(var y=0;y<crop;y++) {
-          if(x+dx<0 || x+dx>=V.sx || y+dy<0 || y+dy>=V.sy) continue; // oob
-          for(var d=0;d<V.depth;d++) {
-           W.set(x,y,d,V.get(x+dx,y+dy,d)); // copy data over
-          }
-        }
-      }
-    } else {
-      W = V;
-    }
-
-    if(fliplr) {
-      // flip volume horziontally
-      var W2 = W.cloneAndZero();
-      for(var x=0;x<W.sx;x++) {
-        for(var y=0;y<W.sy;y++) {
-          for(var d=0;d<W.depth;d++) {
-           W2.set(x,y,d,W.get(W.sx - x - 1,y,d)); // copy data over
-          }
-        }
-      }
-      W = W2; //swap
-    }
-    return W;
-  }
-
-  // img is a DOM element that contains a loaded image
-  // returns a Vol of size (W, H, 4). 4 is for RGBA
-  var img_to_vol = function(img, convert_grayscale) {
-
-    if(typeof(convert_grayscale)==='undefined') var convert_grayscale = false;
-
-    var canvas = document.createElement('canvas');
-    canvas.width = img.width;
-    canvas.height = img.height;
-    var ctx = canvas.getContext("2d");
-
-    // due to a Firefox bug
-    try {
-      ctx.drawImage(img, 0, 0);
-    } catch (e) {
-      if (e.name === "NS_ERROR_NOT_AVAILABLE") {
-        // sometimes happens, lets just abort
-        return false;
-      } else {
-        throw e;
-      }
-    }
-
-    try {
-      var img_data = ctx.getImageData(0, 0, canvas.width, canvas.height);
-    } catch (e) {
-      if(e.name === 'IndexSizeError') {
-        return false; // not sure what causes this sometimes but okay abort
-      } else {
-        throw e;
-      }
-    }
-
-    // prepare the input: get pixels and normalize them
-    var p = img_data.data;
-    var W = img.width;
-    var H = img.height;
-    var pv = []
-    for(var i=0;i<p.length;i++) {
-      pv.push(p[i]/255.0-0.5); // normalize image pixels to [-0.5, 0.5]
-    }
-    var x = new Vol(W, H, 4, 0.0); //input volume (image)
-    x.w = pv;
-
-    if(convert_grayscale) {
-      // flatten into depth=1 array
-      var x1 = new Vol(W, H, 1, 0.0);
-      for(var i=0;i<W;i++) {
-        for(var j=0;j<H;j++) {
-          x1.set(i,j,0,x.get(i,j,0));
-        }
-      }
-      x = x1;
-    }
-
-    return x;
-  }
-  
-  global.augment = augment;
-  global.img_to_vol = img_to_vol;
-
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-
-  // This file contains all layers that do dot products with input,
-  // but usually in a different connectivity pattern and weight sharing
-  // schemes: 
-  // - FullyConn is fully connected dot products 
-  // - ConvLayer does convolutions (so weight sharing spatially)
-  // putting them together in one file because they are very similar
-  var ConvLayer = function(opt) {
-    var opt = opt || {};
-
-    // required
-    this.out_depth = opt.filters;
-    this.sx = opt.sx; // filter size. Should be odd if possible, it's cleaner.
-    this.in_depth = opt.in_depth;
-    this.in_sx = opt.in_sx;
-    this.in_sy = opt.in_sy;
-    
-    // optional
-    this.sy = typeof opt.sy !== 'undefined' ? opt.sy : this.sx;
-    this.stride = typeof opt.stride !== 'undefined' ? opt.stride : 1; // stride at which we apply filters to input volume
-    this.pad = typeof opt.pad !== 'undefined' ? opt.pad : 0; // amount of 0 padding to add around borders of input volume
-    this.l1_decay_mul = typeof opt.l1_decay_mul !== 'undefined' ? opt.l1_decay_mul : 0.0;
-    this.l2_decay_mul = typeof opt.l2_decay_mul !== 'undefined' ? opt.l2_decay_mul : 1.0;
-
-    // computed
-    // note we are doing floor, so if the strided convolution of the filter doesnt fit into the input
-    // volume exactly, the output volume will be trimmed and not contain the (incomplete) computed
-    // final application.
-    this.out_sx = Math.floor((this.in_sx + this.pad * 2 - this.sx) / this.stride + 1);
-    this.out_sy = Math.floor((this.in_sy + this.pad * 2 - this.sy) / this.stride + 1);
-    this.layer_type = 'conv';
-
-    // initializations
-    var bias = typeof opt.bias_pref !== 'undefined' ? opt.bias_pref : 0.0;
-    this.filters = [];
-    for(var i=0;i<this.out_depth;i++) { this.filters.push(new Vol(this.sx, this.sy, this.in_depth)); }
-    this.biases = new Vol(1, 1, this.out_depth, bias);
-  }
-  ConvLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-
-      var A = new Vol(this.out_sx, this.out_sy, this.out_depth, 0.0);
-      for(var d=0;d<this.out_depth;d++) {
-        var f = this.filters[d];
-        var x = -this.pad;
-        var y = -this.pad;
-        for(var ax=0; ax<this.out_sx; x+=this.stride,ax++) {
-          y = -this.pad;
-          for(var ay=0; ay<this.out_sy; y+=this.stride,ay++) {
-
-            // convolve centered at this particular location
-            // could be bit more efficient, going for correctness first
-            var a = 0.0;
-            for(var fx=0;fx<f.sx;fx++) {
-              for(var fy=0;fy<f.sy;fy++) {
-                for(var fd=0;fd<f.depth;fd++) {
-                  var oy = y+fy; // coordinates in the original input array coordinates
-                  var ox = x+fx;
-                  if(oy>=0 && oy<V.sy && ox>=0 && ox<V.sx) {
-                    //a += f.get(fx, fy, fd) * V.get(ox, oy, fd);
-                    // avoid function call overhead for efficiency, compromise modularity :(
-                    a += f.w[((f.sx * fy)+fx)*f.depth+fd] * V.w[((V.sx * oy)+ox)*V.depth+fd];
-                  }
-                }
-              }
-            }
-            a += this.biases.w[d];
-            A.set(ax, ay, d, a);
-          }
-        }
-      }
-      this.out_act = A;
-      return this.out_act;
-    },
-    backward: function() { 
-
-      // compute gradient wrt weights, biases and input data
-      var V = this.in_act;
-      V.dw = global.zeros(V.w.length); // zero out gradient wrt bottom data, we're about to fill it
-      for(var d=0;d<this.out_depth;d++) {
-        var f = this.filters[d];
-        var x = -this.pad;
-        var y = -this.pad;
-        for(var ax=0; ax<this.out_sx; x+=this.stride,ax++) {
-          y = -this.pad;
-          for(var ay=0; ay<this.out_sy; y+=this.stride,ay++) {
-            // convolve and add up the gradients. 
-            // could be more efficient, going for correctness first
-            var chain_grad = this.out_act.get_grad(ax,ay,d); // gradient from above, from chain rule
-            for(var fx=0;fx<f.sx;fx++) {
-              for(var fy=0;fy<f.sy;fy++) {
-                for(var fd=0;fd<f.depth;fd++) {
-                  var oy = y+fy;
-                  var ox = x+fx;
-                  if(oy>=0 && oy<V.sy && ox>=0 && ox<V.sx) {
-                    // forward prop calculated: a += f.get(fx, fy, fd) * V.get(ox, oy, fd);
-                    //f.add_grad(fx, fy, fd, V.get(ox, oy, fd) * chain_grad);
-                    //V.add_grad(ox, oy, fd, f.get(fx, fy, fd) * chain_grad);
-
-                    // avoid function call overhead and use Vols directly for efficiency
-                    var ix1 = ((V.sx * oy)+ox)*V.depth+fd;
-                    var ix2 = ((f.sx * fy)+fx)*f.depth+fd;
-                    f.dw[ix2] += V.w[ix1]*chain_grad;
-                    V.dw[ix1] += f.w[ix2]*chain_grad;
-                  }
-                }
-              }
-            }
-            this.biases.dw[d] += chain_grad;
-          }
-        }
-      }
-    },
-    getParamsAndGrads: function() {
-      var response = [];
-      for(var i=0;i<this.out_depth;i++) {
-        response.push({params: this.filters[i].w, grads: this.filters[i].dw, l2_decay_mul: this.l2_decay_mul, l1_decay_mul: this.l1_decay_mul});
-      }
-      response.push({params: this.biases.w, grads: this.biases.dw, l1_decay_mul: 0.0, l2_decay_mul: 0.0});
-      return response;
-    },
-    toJSON: function() {
-      var json = {};
-      json.sx = this.sx; // filter size in x, y dims
-      json.sy = this.sy;
-      json.stride = this.stride;
-      json.in_depth = this.in_depth;
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.l1_decay_mul = this.l1_decay_mul;
-      json.l2_decay_mul = this.l2_decay_mul;
-      json.pad = this.pad;
-      json.filters = [];
-      for(var i=0;i<this.filters.length;i++) {
-        json.filters.push(this.filters[i].toJSON());
-      }
-      json.biases = this.biases.toJSON();
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.sx = json.sx; // filter size in x, y dims
-      this.sy = json.sy;
-      this.stride = json.stride;
-      this.in_depth = json.in_depth; // depth of input volume
-      this.filters = [];
-      this.l1_decay_mul = typeof json.l1_decay_mul !== 'undefined' ? json.l1_decay_mul : 1.0;
-      this.l2_decay_mul = typeof json.l2_decay_mul !== 'undefined' ? json.l2_decay_mul : 1.0;
-      this.pad = typeof json.pad !== 'undefined' ? json.pad : 0;
-      for(var i=0;i<json.filters.length;i++) {
-        var v = new Vol(0,0,0,0);
-        v.fromJSON(json.filters[i]);
-        this.filters.push(v);
-      }
-      this.biases = new Vol(0,0,0,0);
-      this.biases.fromJSON(json.biases);
-    }
-  }
-
-  var FullyConnLayer = function(opt) {
-    var opt = opt || {};
-
-    // required
-    // ok fine we will allow 'filters' as the word as well
-    this.out_depth = typeof opt.num_neurons !== 'undefined' ? opt.num_neurons : opt.filters;
-
-    // optional 
-    this.l1_decay_mul = typeof opt.l1_decay_mul !== 'undefined' ? opt.l1_decay_mul : 0.0;
-    this.l2_decay_mul = typeof opt.l2_decay_mul !== 'undefined' ? opt.l2_decay_mul : 1.0;
-
-    // computed
-    this.num_inputs = opt.in_sx * opt.in_sy * opt.in_depth;
-    this.out_sx = 1;
-    this.out_sy = 1;
-    this.layer_type = 'fc';
-
-    // initializations
-    var bias = typeof opt.bias_pref !== 'undefined' ? opt.bias_pref : 0.0;
-    this.filters = [];
-    for(var i=0;i<this.out_depth ;i++) { this.filters.push(new Vol(1, 1, this.num_inputs)); }
-    this.biases = new Vol(1, 1, this.out_depth, bias);
-  }
-
-  FullyConnLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var A = new Vol(1, 1, this.out_depth, 0.0);
-      var Vw = V.w;
-      for(var i=0;i<this.out_depth;i++) {
-        var a = 0.0;
-        var wi = this.filters[i].w;
-        for(var d=0;d<this.num_inputs;d++) {
-          a += Vw[d] * wi[d]; // for efficiency use Vols directly for now
-        }
-        a += this.biases.w[i];
-        A.w[i] = a;
-      }
-      this.out_act = A;
-      return this.out_act;
-    },
-    backward: function() {
-      var V = this.in_act;
-      V.dw = global.zeros(V.w.length); // zero out the gradient in input Vol
-      
-      // compute gradient wrt weights and data
-      for(var i=0;i<this.out_depth;i++) {
-        var tfi = this.filters[i];
-        var chain_grad = this.out_act.dw[i];
-        for(var d=0;d<this.num_inputs;d++) {
-          V.dw[d] += tfi.w[d]*chain_grad; // grad wrt input data
-          tfi.dw[d] += V.w[d]*chain_grad; // grad wrt params
-        }
-        this.biases.dw[i] += chain_grad;
-      }
-    },
-    getParamsAndGrads: function() {
-      var response = [];
-      for(var i=0;i<this.out_depth;i++) {
-        response.push({params: this.filters[i].w, grads: this.filters[i].dw, l1_decay_mul: this.l1_decay_mul, l2_decay_mul: this.l2_decay_mul});
-      }
-      response.push({params: this.biases.w, grads: this.biases.dw, l1_decay_mul: 0.0, l2_decay_mul: 0.0});
-      return response;
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.num_inputs = this.num_inputs;
-      json.l1_decay_mul = this.l1_decay_mul;
-      json.l2_decay_mul = this.l2_decay_mul;
-      json.filters = [];
-      for(var i=0;i<this.filters.length;i++) {
-        json.filters.push(this.filters[i].toJSON());
-      }
-      json.biases = this.biases.toJSON();
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.num_inputs = json.num_inputs;
-      this.l1_decay_mul = typeof json.l1_decay_mul !== 'undefined' ? json.l1_decay_mul : 1.0;
-      this.l2_decay_mul = typeof json.l2_decay_mul !== 'undefined' ? json.l2_decay_mul : 1.0;
-      this.filters = [];
-      for(var i=0;i<json.filters.length;i++) {
-        var v = new Vol(0,0,0,0);
-        v.fromJSON(json.filters[i]);
-        this.filters.push(v);
-      }
-      this.biases = new Vol(0,0,0,0);
-      this.biases.fromJSON(json.biases);
-    }
-  }
-
-  global.ConvLayer = ConvLayer;
-  global.FullyConnLayer = FullyConnLayer;
-  
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  var PoolLayer = function(opt) {
-
-    var opt = opt || {};
-
-    // required
-    this.sx = opt.sx; // filter size
-    this.in_depth = opt.in_depth;
-    this.in_sx = opt.in_sx;
-    this.in_sy = opt.in_sy;
-
-    // optional
-    this.sy = typeof opt.sy !== 'undefined' ? opt.sy : this.sx;
-    this.stride = typeof opt.stride !== 'undefined' ? opt.stride : 2;
-    this.pad = typeof opt.pad !== 'undefined' ? opt.pad : 0; // amount of 0 padding to add around borders of input volume
-
-    // computed
-    this.out_depth = this.in_depth;
-    this.out_sx = Math.floor((this.in_sx + this.pad * 2 - this.sx) / this.stride + 1);
-    this.out_sy = Math.floor((this.in_sy + this.pad * 2 - this.sy) / this.stride + 1);
-    this.layer_type = 'pool';
-    // store switches for x,y coordinates for where the max comes from, for each output neuron
-    this.switchx = global.zeros(this.out_sx*this.out_sy*this.out_depth);
-    this.switchy = global.zeros(this.out_sx*this.out_sy*this.out_depth);
-  }
-
-  PoolLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-
-      var A = new Vol(this.out_sx, this.out_sy, this.out_depth, 0.0);
-      
-      var n=0; // a counter for switches
-      for(var d=0;d<this.out_depth;d++) {
-        var x = -this.pad;
-        var y = -this.pad;
-        for(var ax=0; ax<this.out_sx; x+=this.stride,ax++) {
-          y = -this.pad;
-          for(var ay=0; ay<this.out_sy; y+=this.stride,ay++) {
-
-            // convolve centered at this particular location
-            var a = -99999; // hopefully small enough ;\
-            var winx=-1,winy=-1;
-            for(var fx=0;fx<this.sx;fx++) {
-              for(var fy=0;fy<this.sy;fy++) {
-                var oy = y+fy;
-                var ox = x+fx;
-                if(oy>=0 && oy<V.sy && ox>=0 && ox<V.sx) {
-                  var v = V.get(ox, oy, d);
-                  // perform max pooling and store pointers to where
-                  // the max came from. This will speed up backprop 
-                  // and can help make nice visualizations in future
-                  if(v > a) { a = v; winx=ox; winy=oy;}
-                }
-              }
-            }
-            this.switchx[n] = winx;
-            this.switchy[n] = winy;
-            n++;
-            A.set(ax, ay, d, a);
-          }
-        }
-      }
-      this.out_act = A;
-      return this.out_act;
-    },
-    backward: function() { 
-      // pooling layers have no parameters, so simply compute 
-      // gradient wrt data here
-      var V = this.in_act;
-      V.dw = global.zeros(V.w.length); // zero out gradient wrt data
-      var A = this.out_act; // computed in forward pass 
-
-      var n = 0;
-      for(var d=0;d<this.out_depth;d++) {
-        var x = -this.pad;
-        var y = -this.pad;
-        for(var ax=0; ax<this.out_sx; x+=this.stride,ax++) {
-          y = -this.pad;
-          for(var ay=0; ay<this.out_sy; y+=this.stride,ay++) {
-
-            var chain_grad = this.out_act.get_grad(ax,ay,d);
-            V.add_grad(this.switchx[n], this.switchy[n], d, chain_grad);
-            n++;
-
-          }
-        }
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.sx = this.sx;
-      json.sy = this.sy;
-      json.stride = this.stride;
-      json.in_depth = this.in_depth;
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.pad = this.pad;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.sx = json.sx;
-      this.sy = json.sy;
-      this.stride = json.stride;
-      this.in_depth = json.in_depth;
-      this.pad = typeof json.pad !== 'undefined' ? json.pad : 0; // backwards compatibility
-      this.switchx = global.zeros(this.out_sx*this.out_sy*this.out_depth); // need to re-init these appropriately
-      this.switchy = global.zeros(this.out_sx*this.out_sy*this.out_depth);
-    }
-  }
-
-  global.PoolLayer = PoolLayer;
-
-})(convnetjs);
-
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  var InputLayer = function(opt) {
-    var opt = opt || {};
-
-    // this is a bit silly but lets allow people to specify either ins or outs
-    this.out_sx = typeof opt.out_sx !== 'undefined' ? opt.out_sx : opt.in_sx;
-    this.out_sy = typeof opt.out_sy !== 'undefined' ? opt.out_sy : opt.in_sy;
-    this.out_depth = typeof opt.out_depth !== 'undefined' ? opt.out_depth : opt.in_depth;
-    this.layer_type = 'input';
-  }
-  InputLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      this.out_act = V;
-      return this.out_act; // dummy identity function for now
-    },
-    backward: function() { },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-    }
-  }
-
-  global.InputLayer = InputLayer;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  // Layers that implement a loss. Currently these are the layers that 
-  // can initiate a backward() pass. In future we probably want a more 
-  // flexible system that can accomodate multiple losses to do multi-task
-  // learning, and stuff like that. But for now, one of the layers in this
-  // file must be the final layer in a Net.
-
-  // This is a classifier, with N discrete classes from 0 to N-1
-  // it gets a stream of N incoming numbers and computes the softmax
-  // function (exponentiate and normalize to sum to 1 as probabilities should)
-  var SoftmaxLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.num_inputs = opt.in_sx * opt.in_sy * opt.in_depth;
-    this.out_depth = this.num_inputs;
-    this.out_sx = 1;
-    this.out_sy = 1;
-    this.layer_type = 'softmax';
-  }
-
-  SoftmaxLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-
-      var A = new Vol(1, 1, this.out_depth, 0.0);
-
-      // compute max activation
-      var as = V.w;
-      var amax = V.w[0];
-      for(var i=1;i<this.out_depth;i++) {
-        if(as[i] > amax) amax = as[i];
-      }
-
-      // compute exponentials (carefully to not blow up)
-      var es = global.zeros(this.out_depth);
-      var esum = 0.0;
-      for(var i=0;i<this.out_depth;i++) {
-        var e = Math.exp(as[i] - amax);
-        esum += e;
-        es[i] = e;
-      }
-
-      // normalize and output to sum to one
-      for(var i=0;i<this.out_depth;i++) {
-        es[i] /= esum;
-        A.w[i] = es[i];
-      }
-
-      this.es = es; // save these for backprop
-      this.out_act = A;
-      return this.out_act;
-    },
-    backward: function(y) {
-
-      // compute and accumulate gradient wrt weights and bias of this layer
-      var x = this.in_act;
-      x.dw = global.zeros(x.w.length); // zero out the gradient of input Vol
-
-      for(var i=0;i<this.out_depth;i++) {
-        var indicator = i === y ? 1.0 : 0.0;
-        var mul = -(indicator - this.es[i]);
-        x.dw[i] = mul;
-      }
-
-      // loss is the class negative log likelihood
-      return -Math.log(this.es[y]);
-    },
-    getParamsAndGrads: function() { 
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.num_inputs = this.num_inputs;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.num_inputs = json.num_inputs;
-    }
-  }
-
-  // implements an L2 regression cost layer,
-  // so penalizes \sum_i(||x_i - y_i||^2), where x is its input
-  // and y is the user-provided array of "correct" values.
-  var RegressionLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.num_inputs = opt.in_sx * opt.in_sy * opt.in_depth;
-    this.out_depth = this.num_inputs;
-    this.out_sx = 1;
-    this.out_sy = 1;
-    this.layer_type = 'regression';
-  }
-
-  RegressionLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      this.out_act = V;
-      return V; // identity function
-    },
-    // y is a list here of size num_inputs
-    backward: function(y) { 
-
-      // compute and accumulate gradient wrt weights and bias of this layer
-      var x = this.in_act;
-      x.dw = global.zeros(x.w.length); // zero out the gradient of input Vol
-      var loss = 0.0;
-      if(y instanceof Array || y instanceof Float64Array) {
-        for(var i=0;i<this.out_depth;i++) {
-          var dy = x.w[i] - y[i];
-          x.dw[i] = dy;
-          loss += 2*dy*dy;
-        }
-      } else {
-        // assume it is a struct with entries .dim and .val
-        // and we pass gradient only along dimension dim to be equal to val
-        var i = y.dim;
-        var yi = y.val;
-        var dy = x.w[i] - yi;
-        x.dw[i] = dy;
-        loss += 2*dy*dy;
-      }
-      return loss;
-    },
-    getParamsAndGrads: function() { 
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.num_inputs = this.num_inputs;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.num_inputs = json.num_inputs;
-    }
-  }
-
-  var SVMLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.num_inputs = opt.in_sx * opt.in_sy * opt.in_depth;
-    this.out_depth = this.num_inputs;
-    this.out_sx = 1;
-    this.out_sy = 1;
-    this.layer_type = 'svm';
-  }
-
-  SVMLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      this.out_act = V; // nothing to do, output raw scores
-      return V;
-    },
-    backward: function(y) {
-
-      // compute and accumulate gradient wrt weights and bias of this layer
-      var x = this.in_act;
-      x.dw = global.zeros(x.w.length); // zero out the gradient of input Vol
-
-      var yscore = x.w[y]; // score of ground truth
-      var margin = 1.0;
-      var loss = 0.0;
-      for(var i=0;i<this.out_depth;i++) {
-        if(-yscore + x.w[i] + margin > 0) {
-          // violating example, apply loss
-          // I love hinge loss, by the way. Truly.
-          // Seriously, compare this SVM code with Softmax forward AND backprop code above
-          // it's clear which one is superior, not only in code, simplicity
-          // and beauty, but also in practice.
-          x.dw[i] += 1;
-          x.dw[y] -= 1;
-          loss += -yscore + x.w[i] + margin;
-        }
-      }
-
-      return loss;
-    },
-    getParamsAndGrads: function() { 
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.num_inputs = this.num_inputs;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type;
-      this.num_inputs = json.num_inputs;
-    }
-  }
-  
-  global.RegressionLayer = RegressionLayer;
-  global.SoftmaxLayer = SoftmaxLayer;
-  global.SVMLayer = SVMLayer;
-
-})(convnetjs);
-
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  // Implements ReLU nonlinearity elementwise
-  // x -> max(0, x)
-  // the output is in [0, inf)
-  var ReluLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = opt.in_depth;
-    this.layer_type = 'relu';
-  }
-  ReluLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var V2 = V.clone();
-      var N = V.w.length;
-      var V2w = V2.w;
-      for(var i=0;i<N;i++) { 
-        if(V2w[i] < 0) V2w[i] = 0; // threshold at 0
-      }
-      this.out_act = V2;
-      return this.out_act;
-    },
-    backward: function() {
-      var V = this.in_act; // we need to set dw of this
-      var V2 = this.out_act;
-      var N = V.w.length;
-      V.dw = global.zeros(N); // zero out gradient wrt data
-      for(var i=0;i<N;i++) {
-        if(V2.w[i] <= 0) V.dw[i] = 0; // threshold
-        else V.dw[i] = V2.dw[i];
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-    }
-  }
-
-  // Implements Sigmoid nnonlinearity elementwise
-  // x -> 1/(1+e^(-x))
-  // so the output is between 0 and 1.
-  var SigmoidLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = opt.in_depth;
-    this.layer_type = 'sigmoid';
-  }
-  SigmoidLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var V2 = V.cloneAndZero();
-      var N = V.w.length;
-      var V2w = V2.w;
-      var Vw = V.w;
-      for(var i=0;i<N;i++) { 
-        V2w[i] = 1.0/(1.0+Math.exp(-Vw[i]));
-      }
-      this.out_act = V2;
-      return this.out_act;
-    },
-    backward: function() {
-      var V = this.in_act; // we need to set dw of this
-      var V2 = this.out_act;
-      var N = V.w.length;
-      V.dw = global.zeros(N); // zero out gradient wrt data
-      for(var i=0;i<N;i++) {
-        var v2wi = V2.w[i];
-        V.dw[i] =  v2wi * (1.0 - v2wi) * V2.dw[i];
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-    }
-  }
-
-  // Implements Maxout nnonlinearity that computes
-  // x -> max(x)
-  // where x is a vector of size group_size. Ideally of course,
-  // the input size should be exactly divisible by group_size
-  var MaxoutLayer = function(opt) {
-    var opt = opt || {};
-
-    // required
-    this.group_size = typeof opt.group_size !== 'undefined' ? opt.group_size : 2;
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = Math.floor(opt.in_depth / this.group_size);
-    this.layer_type = 'maxout';
-
-    this.switches = global.zeros(this.out_sx*this.out_sy*this.out_depth); // useful for backprop
-  }
-  MaxoutLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var N = this.out_depth; 
-      var V2 = new Vol(this.out_sx, this.out_sy, this.out_depth, 0.0);
-
-      // optimization branch. If we're operating on 1D arrays we dont have
-      // to worry about keeping track of x,y,d coordinates inside
-      // input volumes. In convnets we do :(
-      if(this.out_sx === 1 && this.out_sy === 1) {
-        for(var i=0;i<N;i++) {
-          var ix = i * this.group_size; // base index offset
-          var a = V.w[ix];
-          var ai = 0;
-          for(var j=1;j<this.group_size;j++) {
-            var a2 = V.w[ix+j];
-            if(a2 > a) {
-              a = a2;
-              ai = j;
-            }
-          }
-          V2.w[i] = a;
-          this.switches[i] = ix + ai;
-        }
-      } else {
-        var n=0; // counter for switches
-        for(var x=0;x<V.sx;x++) {
-          for(var y=0;y<V.sy;y++) {
-            for(var i=0;i<N;i++) {
-              var ix = i * this.group_size;
-              var a = V.get(x, y, ix);
-              var ai = 0;
-              for(var j=1;j<this.group_size;j++) {
-                var a2 = V.get(x, y, ix+j);
-                if(a2 > a) {
-                  a = a2;
-                  ai = j;
-                }
-              }
-              V2.set(x,y,i,a);
-              this.switches[n] = ix + ai;
-              n++;
-            }
-          }
-        }
-
-      }
-      this.out_act = V2;
-      return this.out_act;
-    },
-    backward: function() {
-      var V = this.in_act; // we need to set dw of this
-      var V2 = this.out_act;
-      var N = this.out_depth;
-      V.dw = global.zeros(V.w.length); // zero out gradient wrt data
-
-      // pass the gradient through the appropriate switch
-      if(this.out_sx === 1 && this.out_sy === 1) {
-        for(var i=0;i<N;i++) {
-          var chain_grad = V2.dw[i];
-          V.dw[this.switches[i]] = chain_grad;
-        }
-      } else {
-        // bleh okay, lets do this the hard way
-        var n=0; // counter for switches
-        for(var x=0;x<V2.sx;x++) {
-          for(var y=0;y<V2.sy;y++) {
-            for(var i=0;i<N;i++) {
-              var chain_grad = V2.get_grad(x,y,i);
-              V.set_grad(x,y,this.switches[n],chain_grad);
-              n++;
-            }
-          }
-        }
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.group_size = this.group_size;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-      this.group_size = json.group_size;
-      this.switches = global.zeros(this.group_size);
-    }
-  }
-
-  // a helper function, since tanh is not yet part of ECMAScript. Will be in v6.
-  function tanh(x) {
-    var y = Math.exp(2 * x);
-    return (y - 1) / (y + 1);
-  }
-  // Implements Tanh nnonlinearity elementwise
-  // x -> tanh(x) 
-  // so the output is between -1 and 1.
-  var TanhLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = opt.in_depth;
-    this.layer_type = 'tanh';
-  }
-  TanhLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var V2 = V.cloneAndZero();
-      var N = V.w.length;
-      for(var i=0;i<N;i++) { 
-        V2.w[i] = tanh(V.w[i]);
-      }
-      this.out_act = V2;
-      return this.out_act;
-    },
-    backward: function() {
-      var V = this.in_act; // we need to set dw of this
-      var V2 = this.out_act;
-      var N = V.w.length;
-      V.dw = global.zeros(N); // zero out gradient wrt data
-      for(var i=0;i<N;i++) {
-        var v2wi = V2.w[i];
-        V.dw[i] = (1.0 - v2wi * v2wi) * V2.dw[i];
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-    }
-  }
-  
-  global.TanhLayer = TanhLayer;
-  global.MaxoutLayer = MaxoutLayer;
-  global.ReluLayer = ReluLayer;
-  global.SigmoidLayer = SigmoidLayer;
-
-})(convnetjs);
-
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-
-  // An inefficient dropout layer
-  // Note this is not most efficient implementation since the layer before
-  // computed all these activations and now we're just going to drop them :(
-  // same goes for backward pass. Also, if we wanted to be efficient at test time
-  // we could equivalently be clever and upscale during train and copy pointers during test
-  // todo: make more efficient.
-  var DropoutLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = opt.in_depth;
-    this.layer_type = 'dropout';
-    this.drop_prob = typeof opt.drop_prob !== 'undefined' ? opt.drop_prob : 0.5;
-    this.dropped = global.zeros(this.out_sx*this.out_sy*this.out_depth);
-  }
-  DropoutLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      if(typeof(is_training)==='undefined') { is_training = false; } // default is prediction mode
-      var V2 = V.clone();
-      var N = V.w.length;
-      if(is_training) {
-        // do dropout
-        for(var i=0;i<N;i++) {
-          if(Math.random()<this.drop_prob) { V2.w[i]=0; this.dropped[i] = true; } // drop!
-          else {this.dropped[i] = false;}
-        }
-      } else {
-        // scale the activations during prediction
-        for(var i=0;i<N;i++) { V2.w[i]*=this.drop_prob; }
-      }
-      this.out_act = V2;
-      return this.out_act; // dummy identity function for now
-    },
-    backward: function() {
-      var V = this.in_act; // we need to set dw of this
-      var chain_grad = this.out_act;
-      var N = V.w.length;
-      V.dw = global.zeros(N); // zero out gradient wrt data
-      for(var i=0;i<N;i++) {
-        if(!(this.dropped[i])) { 
-          V.dw[i] = chain_grad.dw[i]; // copy over the gradient
-        }
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      json.drop_prob = this.drop_prob;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-      this.drop_prob = json.drop_prob;
-    }
-  }
-  
-
-  global.DropoutLayer = DropoutLayer;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  // a bit experimental layer for now. I think it works but I'm not 100%
-  // the gradient check is a bit funky. I'll look into this a bit later.
-  // Local Response Normalization in window, along depths of volumes
-  var LocalResponseNormalizationLayer = function(opt) {
-    var opt = opt || {};
-
-    // required
-    this.k = opt.k;
-    this.n = opt.n;
-    this.alpha = opt.alpha;
-    this.beta = opt.beta;
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    this.out_depth = opt.in_depth;
-    this.layer_type = 'lrn';
-
-    // checks
-    if(this.n%2 === 0) { console.log('WARNING n should be odd for LRN layer'); }
-  }
-  LocalResponseNormalizationLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-
-      var A = V.cloneAndZero();
-      this.S_cache_ = V.cloneAndZero();
-      var n2 = Math.floor(this.n/2);
-      for(var x=0;x<V.sx;x++) {
-        for(var y=0;y<V.sy;y++) {
-          for(var i=0;i<V.depth;i++) {
-
-            var ai = V.get(x,y,i);
-
-            // normalize in a window of size n
-            var den = 0.0;
-            for(var j=Math.max(0,i-n2);j<=Math.min(i+n2,V.depth-1);j++) {
-              var aa = V.get(x,y,j);
-              den += aa*aa;
-            }
-            den *= this.alpha / this.n;
-            den += this.k;
-            this.S_cache_.set(x,y,i,den); // will be useful for backprop
-            den = Math.pow(den, this.beta);
-            A.set(x,y,i,ai/den);
-          }
-        }
-      }
-
-      this.out_act = A;
-      return this.out_act; // dummy identity function for now
-    },
-    backward: function() { 
-      // evaluate gradient wrt data
-      var V = this.in_act; // we need to set dw of this
-      V.dw = global.zeros(V.w.length); // zero out gradient wrt data
-      var A = this.out_act; // computed in forward pass 
-
-      var n2 = Math.floor(this.n/2);
-      for(var x=0;x<V.sx;x++) {
-        for(var y=0;y<V.sy;y++) {
-          for(var i=0;i<V.depth;i++) {
-
-            var chain_grad = this.out_act.get_grad(x,y,i);
-            var S = this.S_cache_.get(x,y,i);
-            var SB = Math.pow(S, this.beta);
-            var SB2 = SB*SB;
-
-            // normalize in a window of size n
-            for(var j=Math.max(0,i-n2);j<=Math.min(i+n2,V.depth-1);j++) {              
-              var aj = V.get(x,y,j); 
-              var g = -aj*this.beta*Math.pow(S,this.beta-1)*this.alpha/this.n*2*aj;
-              if(j===i) g+= SB;
-              g /= SB2;
-              g *= chain_grad;
-              V.add_grad(x,y,j,g);
-            }
-
-          }
-        }
-      }
-    },
-    getParamsAndGrads: function() { return []; },
-    toJSON: function() {
-      var json = {};
-      json.k = this.k;
-      json.n = this.n;
-      json.alpha = this.alpha; // normalize by size
-      json.beta = this.beta;
-      json.out_sx = this.out_sx; 
-      json.out_sy = this.out_sy;
-      json.out_depth = this.out_depth;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.k = json.k;
-      this.n = json.n;
-      this.alpha = json.alpha; // normalize by size
-      this.beta = json.beta;
-      this.out_sx = json.out_sx; 
-      this.out_sy = json.out_sy;
-      this.out_depth = json.out_depth;
-      this.layer_type = json.layer_type;
-    }
-  }
-  
-
-  global.LocalResponseNormalizationLayer = LocalResponseNormalizationLayer;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-
-  // transforms x-> [x, x_i*x_j forall i,j]
-  // so the fully connected layer afters will essentially be doing tensor multiplies
-  var QuadTransformLayer = function(opt) {
-    var opt = opt || {};
-
-    // computed
-    this.out_sx = opt.in_sx;
-    this.out_sy = opt.in_sy;
-    // linear terms, and then quadratic terms, of which there are 1/2*n*(n+1),
-    // (offdiagonals and the diagonal total) and arithmetic series.
-    // Actually never mind, lets not be fancy here yet and just include
-    // terms x_ix_j and x_jx_i twice. Half as efficient but much less
-    // headache.
-    this.out_depth = opt.in_depth + opt.in_depth * opt.in_depth;
-    this.layer_type = 'quadtransform';
-
-  }
-  QuadTransformLayer.prototype = {
-    forward: function(V, is_training) {
-      this.in_act = V;
-      var N = this.out_depth;
-      var Ni = V.depth;
-      var V2 = new Vol(this.out_sx, this.out_sy, this.out_depth, 0.0);
-      for(var x=0;x<V.sx;x++) {
-        for(var y=0;y<V.sy;y++) {
-          for(var i=0;i<N;i++) {
-            if(i<Ni) {
-              V2.set(x,y,i,V.get(x,y,i)); // copy these over (linear terms)
-            } else {
-              var i0 = Math.floor((i-Ni)/Ni);
-              var i1 = (i-Ni) - i0*Ni;
-              V2.set(x,y,i,V.get(x,y,i0) * V.get(x,y,i1)); // quadratic
-            }
-          }
-        }
-      }
-      this.out_act = V2;
-      return this.out_act; // dummy identity function for now
-    },
-    backward: function() {
-      var V = this.in_act;
-      V.dw = global.zeros(V.w.length); // zero out gradient wrt data
-      var V2 = this.out_act;
-      var N = this.out_depth;
-      var Ni = V.depth;
-      for(var x=0;x<V.sx;x++) {
-        for(var y=0;y<V.sy;y++) {
-          for(var i=0;i<N;i++) {
-            var chain_grad = V2.get_grad(x,y,i);
-            if(i<Ni) {
-              V.add_grad(x,y,i,chain_grad);
-            } else {
-              var i0 = Math.floor((i-Ni)/Ni);
-              var i1 = (i-Ni) - i0*Ni;
-              V.add_grad(x,y,i0,V.get(x,y,i1)*chain_grad);
-              V.add_grad(x,y,i1,V.get(x,y,i0)*chain_grad);
-            }
-          }
-        }
-      }
-    },
-    getParamsAndGrads: function() {
-      return [];
-    },
-    toJSON: function() {
-      var json = {};
-      json.out_depth = this.out_depth;
-      json.out_sx = this.out_sx;
-      json.out_sy = this.out_sy;
-      json.layer_type = this.layer_type;
-      return json;
-    },
-    fromJSON: function(json) {
-      this.out_depth = json.out_depth;
-      this.out_sx = json.out_sx;
-      this.out_sy = json.out_sy;
-      this.layer_type = json.layer_type; 
-    }
-  }
-  
-
-  global.QuadTransformLayer = QuadTransformLayer;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-  
-  // Net manages a set of layers
-  // For now constraints: Simple linear order of layers, first layer input last layer a cost layer
-  var Net = function(options) {
-    this.layers = [];
-  }
-
-  Net.prototype = {
-    
-    // takes a list of layer definitions and creates the network layer objects
-    makeLayers: function(defs) {
-
-      // few checks for now
-      if(defs.length<2) {console.log('ERROR! For now at least have input and softmax layers.');}
-      if(defs[0].type !== 'input') {console.log('ERROR! For now first layer should be input.');}
-
-      // desugar syntactic for adding activations and dropouts
-      var desugar = function() {
-        var new_defs = [];
-        for(var i=0;i<defs.length;i++) {
-          var def = defs[i];
-          
-          if(def.type==='softmax' || def.type==='svm') {
-            // add an fc layer here, there is no reason the user should
-            // have to worry about this and we almost always want to
-            new_defs.push({type:'fc', num_neurons: def.num_classes});
-          }
-
-          if(def.type==='regression') {
-            // add an fc layer here, there is no reason the user should
-            // have to worry about this and we almost always want to
-            new_defs.push({type:'fc', num_neurons: def.num_neurons});
-          }
-
-          if((def.type==='fc' || def.type==='conv') 
-              && typeof(def.bias_pref) === 'undefined'){
-            def.bias_pref = 0.0;
-            if(typeof def.activation !== 'undefined' && def.activation === 'relu') {
-              def.bias_pref = 0.1; // relus like a bit of positive bias to get gradients early
-              // otherwise it's technically possible that a relu unit will never turn on (by chance)
-              // and will never get any gradient and never contribute any computation. Dead relu.
-            }
-          }
-          
-          if(typeof def.tensor !== 'undefined') {
-            // apply quadratic transform so that the upcoming multiply will include
-            // quadratic terms, equivalent to doing a tensor product
-            if(def.tensor) {
-              new_defs.push({type: 'quadtransform'});
-            }
-          }
-
-          new_defs.push(def);
-
-          if(typeof def.activation !== 'undefined') {
-            if(def.activation==='relu') { new_defs.push({type:'relu'}); }
-            else if (def.activation==='sigmoid') { new_defs.push({type:'sigmoid'}); }
-            else if (def.activation==='tanh') { new_defs.push({type:'tanh'}); }
-            else if (def.activation==='maxout') {
-              // create maxout activation, and pass along group size, if provided
-              var gs = def.group_size !== 'undefined' ? def.group_size : 2;
-              new_defs.push({type:'maxout', group_size:gs});
-            }
-            else { console.log('ERROR unsupported activation ' + def.activation); }
-          }
-          if(typeof def.drop_prob !== 'undefined' && def.type !== 'dropout') {
-            new_defs.push({type:'dropout', drop_prob: def.drop_prob});
-          }
-
-        }
-        return new_defs;
-      }
-      defs = desugar(defs);
-
-      // create the layers
-      this.layers = [];
-      for(var i=0;i<defs.length;i++) {
-        var def = defs[i];
-        if(i>0) {
-          var prev = this.layers[i-1];
-          def.in_sx = prev.out_sx;
-          def.in_sy = prev.out_sy;
-          def.in_depth = prev.out_depth;
-        }
-
-        switch(def.type) {
-          case 'fc': this.layers.push(new global.FullyConnLayer(def)); break;
-          case 'lrn': this.layers.push(new global.LocalResponseNormalizationLayer(def)); break;
-          case 'dropout': this.layers.push(new global.DropoutLayer(def)); break;
-          case 'input': this.layers.push(new global.InputLayer(def)); break;
-          case 'softmax': this.layers.push(new global.SoftmaxLayer(def)); break;
-          case 'regression': this.layers.push(new global.RegressionLayer(def)); break;
-          case 'conv': this.layers.push(new global.ConvLayer(def)); break;
-          case 'pool': this.layers.push(new global.PoolLayer(def)); break;
-          case 'relu': this.layers.push(new global.ReluLayer(def)); break;
-          case 'sigmoid': this.layers.push(new global.SigmoidLayer(def)); break;
-          case 'tanh': this.layers.push(new global.TanhLayer(def)); break;
-          case 'maxout': this.layers.push(new global.MaxoutLayer(def)); break;
-          case 'quadtransform': this.layers.push(new global.QuadTransformLayer(def)); break;
-          case 'svm': this.layers.push(new global.SVMLayer(def)); break;
-          default: console.log('ERROR: UNRECOGNIZED LAYER TYPE!');
-        }
-      }
-    },
-
-    // forward prop the network. A trainer will pass in is_training = true
-    forward: function(V, is_training) {
-      if(typeof(is_training)==='undefined') is_training = false;
-      var act = this.layers[0].forward(V, is_training);
-      for(var i=1;i<this.layers.length;i++) {
-        act = this.layers[i].forward(act, is_training);
-      }
-      return act;
-    },
-
-    getCostLoss: function(V, y) {
-      this.forward(V, false);
-      var N = this.layers.length;
-      var loss = this.layers[N-1].backward(y);
-      return loss;
-    },
-    
-    // backprop: compute gradients wrt all parameters
-    backward: function(y) {
-      var N = this.layers.length;
-      var loss = this.layers[N-1].backward(y); // last layer assumed softmax
-      for(var i=N-2;i>=0;i--) { // first layer assumed input
-        this.layers[i].backward();
-      }
-      return loss;
-    },
-    getParamsAndGrads: function() {
-      // accumulate parameters and gradients for the entire network
-      var response = [];
-      for(var i=0;i<this.layers.length;i++) {
-        var layer_reponse = this.layers[i].getParamsAndGrads();
-        for(var j=0;j<layer_reponse.length;j++) {
-          response.push(layer_reponse[j]);
-        }
-      }
-      return response;
-    },
-    getPrediction: function() {
-      var S = this.layers[this.layers.length-1]; // softmax layer
-      var p = S.out_act.w;
-      var maxv = p[0];
-      var maxi = 0;
-      for(var i=1;i<p.length;i++) {
-        if(p[i] > maxv) { maxv = p[i]; maxi = i;}
-      }
-      return maxi;
-    },
-    toJSON: function() {
-      var json = {};
-      json.layers = [];
-      for(var i=0;i<this.layers.length;i++) {
-        json.layers.push(this.layers[i].toJSON());
-      }
-      return json;
-    },
-    fromJSON: function(json) {
-      this.layers = [];
-      for(var i=0;i<json.layers.length;i++) {
-        var Lj = json.layers[i]
-        var t = Lj.layer_type;
-        var L;
-        if(t==='input') { L = new global.InputLayer(); }
-        if(t==='relu') { L = new global.ReluLayer(); }
-        if(t==='sigmoid') { L = new global.SigmoidLayer(); }
-        if(t==='tanh') { L = new global.TanhLayer(); }
-        if(t==='dropout') { L = new global.DropoutLayer(); }
-        if(t==='conv') { L = new global.ConvLayer(); }
-        if(t==='pool') { L = new global.PoolLayer(); }
-        if(t==='lrn') { L = new global.LocalResponseNormalizationLayer(); }
-        if(t==='softmax') { L = new global.SoftmaxLayer(); }
-        if(t==='regression') { L = new global.RegressionLayer(); }
-        if(t==='fc') { L = new global.FullyConnLayer(); }
-        if(t==='maxout') { L = new global.MaxoutLayer(); }
-        if(t==='quadtransform') { L = new global.QuadTransformLayer(); }
-        if(t==='svm') { L = new global.SVMLayer(); }
-        L.fromJSON(Lj);
-        this.layers.push(L);
-      }
-    }
-  }
-  
-
-  global.Net = Net;
-})(convnetjs);
-(function(global) {
-  "use strict";
-  var Vol = global.Vol; // convenience
-
-  var Trainer = function(net, options) {
-
-    this.net = net;
-
-    var options = options || {};
-    this.learning_rate = typeof options.learning_rate !== 'undefined' ? options.learning_rate : 0.01;
-    this.l1_decay = typeof options.l1_decay !== 'undefined' ? options.l1_decay : 0.0;
-    this.l2_decay = typeof options.l2_decay !== 'undefined' ? options.l2_decay : 0.0;
-    this.batch_size = typeof options.batch_size !== 'undefined' ? options.batch_size : 1;
-    this.method = typeof options.method !== 'undefined' ? options.method : 'sgd'; // sgd/adagrad/adadelta/windowgrad
-
-    this.momentum = typeof options.momentum !== 'undefined' ? options.momentum : 0.9;
-    this.ro = typeof options.ro !== 'undefined' ? options.ro : 0.95; // used in adadelta
-    this.eps = typeof options.eps !== 'undefined' ? options.eps : 1e-6; // used in adadelta
-
-    this.k = 0; // iteration counter
-    this.gsum = []; // last iteration gradients (used for momentum calculations)
-    this.xsum = []; // used in adadelta
-  }
-
-  Trainer.prototype = {
-    train: function(x, y) {
-
-      var start = new Date().getTime();
-      this.net.forward(x, true); // also set the flag that lets the net know we're just training
-      var end = new Date().getTime();
-      var fwd_time = end - start;
-
-      var start = new Date().getTime();
-      var cost_loss = this.net.backward(y);
-      var l2_decay_loss = 0.0;
-      var l1_decay_loss = 0.0;
-      var end = new Date().getTime();
-      var bwd_time = end - start;
-      
-      this.k++;
-      if(this.k % this.batch_size === 0) {
-
-        var pglist = this.net.getParamsAndGrads();
-
-        // initialize lists for accumulators. Will only be done once on first iteration
-        if(this.gsum.length === 0 && (this.method !== 'sgd' || this.momentum > 0.0)) {
-          // only vanilla sgd doesnt need either lists
-          // momentum needs gsum
-          // adagrad needs gsum
-          // adadelta needs gsum and xsum
-          for(var i=0;i<pglist.length;i++) {
-            this.gsum.push(global.zeros(pglist[i].params.length));
-            if(this.method === 'adadelta') {
-              this.xsum.push(global.zeros(pglist[i].params.length));
-            } else {
-              this.xsum.push([]); // conserve memory
-            }
-          }
-        }
-
-        // perform an update for all sets of weights
-        for(var i=0;i<pglist.length;i++) {
-          var pg = pglist[i]; // param, gradient, other options in future (custom learning rate etc)
-          var p = pg.params;
-          var g = pg.grads;
-
-          // learning rate for some parameters.
-          var l2_decay_mul = typeof pg.l2_decay_mul !== 'undefined' ? pg.l2_decay_mul : 1.0;
-          var l1_decay_mul = typeof pg.l1_decay_mul !== 'undefined' ? pg.l1_decay_mul : 1.0;
-          var l2_decay = this.l2_decay * l2_decay_mul;
-          var l1_decay = this.l1_decay * l1_decay_mul;
-
-          var plen = p.length;
-          for(var j=0;j<plen;j++) {
-            l2_decay_loss += l2_decay*p[j]*p[j]/2; // accumulate weight decay loss
-            l1_decay_loss += l1_decay*Math.abs(p[j]);
-            var l1grad = l1_decay * (p[j] > 0 ? 1 : -1);
-            var l2grad = l2_decay * (p[j]);
-
-            var gij = (l2grad + l1grad + g[j]) / this.batch_size; // raw batch gradient
-
-            var gsumi = this.gsum[i];
-            var xsumi = this.xsum[i];
-            if(this.method === 'adagrad') {
-              // adagrad update
-              gsumi[j] = gsumi[j] + gij * gij;
-              var dx = - this.learning_rate / Math.sqrt(gsumi[j] + this.eps) * gij;
-              p[j] += dx;
-            } else if(this.method === 'windowgrad') {
-              // this is adagrad but with a moving window weighted average
-              // so the gradient is not accumulated over the entire history of the run. 
-              // it's also referred to as Idea #1 in Zeiler paper on Adadelta. Seems reasonable to me!
-              gsumi[j] = this.ro * gsumi[j] + (1-this.ro) * gij * gij;
-              var dx = - this.learning_rate / Math.sqrt(gsumi[j] + this.eps) * gij; // eps added for better conditioning
-              p[j] += dx;
-            } else if(this.method === 'adadelta') {
-              // assume adadelta if not sgd or adagrad
-              gsumi[j] = this.ro * gsumi[j] + (1-this.ro) * gij * gij;
-              var dx = - Math.sqrt((xsumi[j] + this.eps)/(gsumi[j] + this.eps)) * gij;
-              xsumi[j] = this.ro * xsumi[j] + (1-this.ro) * dx * dx; // yes, xsum lags behind gsum by 1.
-              p[j] += dx;
-            } else {
-              // assume SGD
-              if(this.momentum > 0.0) {
-                // momentum update
-                var dx = this.momentum * gsumi[j] - this.learning_rate * gij; // step
-                gsumi[j] = dx; // back this up for next iteration of momentum
-                p[j] += dx; // apply corrected gradient
-              } else {
-                // vanilla sgd
-                p[j] +=  - this.learning_rate * gij;
-              }
-            }
-            g[j] = 0.0; // zero out gradient so that we can begin accumulating anew
-          }
-        }
-      }
-
-      // appending softmax_loss for backwards compatibility, but from now on we will always use cost_loss
-      // in future, TODO: have to completely redo the way loss is done around the network as currently 
-      // loss is a bit of a hack. Ideally, user should specify arbitrary number of loss functions on any layer
-      // and it should all be computed correctly and automatically. 
-      return {fwd_time: fwd_time, bwd_time: bwd_time, 
-              l2_decay_loss: l2_decay_loss, l1_decay_loss: l1_decay_loss,
-              cost_loss: cost_loss, softmax_loss: cost_loss, 
-              loss: cost_loss + l1_decay_loss + l2_decay_loss}
-    }
-  }
-  
-  global.Trainer = Trainer;
-  global.SGDTrainer = Trainer; // backwards compatibility
-})(convnetjs);
-
-(function(global) {
-  "use strict";
-
-  // used utilities, make explicit local references
-  var randf = global.randf;
-  var randi = global.randi;
-  var Net = global.Net;
-  var Trainer = global.Trainer;
-  var maxmin = global.maxmin;
-  var randperm = global.randperm;
-  var weightedSample = global.weightedSample;
-  var getopt = global.getopt;
-  var arrUnique = global.arrUnique;
-
-  /*
-  A MagicNet takes data: a list of convnetjs.Vol(), and labels
-  which for now are assumed to be class indeces 0..K. MagicNet then:
-  - creates data folds for cross-validation
-  - samples candidate networks
-  - evaluates candidate networks on all data folds
-  - produces predictions by model-averaging the best networks
-  */
-  var MagicNet = function(data, labels, opt) {
-    var opt = opt || {};
-    if(typeof data === 'undefined') { data = []; }
-    if(typeof labels === 'undefined') { labels = []; }
-
-    // required inputs
-    this.data = data; // store these pointers to data
-    this.labels = labels;
-
-    // optional inputs
-    this.train_ratio = getopt(opt, 'train_ratio', 0.7);
-    this.num_folds = getopt(opt, 'num_folds', 10);
-    this.num_candidates = getopt(opt, 'num_candidates', 50); // we evaluate several in parallel
-    // how many epochs of data to train every network? for every fold?
-    // higher values mean higher accuracy in final results, but more expensive
-    this.num_epochs = getopt(opt, 'num_epochs', 50); 
-    // number of best models to average during prediction. Usually higher = better
-    this.ensemble_size = getopt(opt, 'ensemble_size', 10);
-
-    // candidate parameters
-    this.batch_size_min = getopt(opt, 'batch_size_min', 10);
-    this.batch_size_max = getopt(opt, 'batch_size_max', 300);
-    this.l2_decay_min = getopt(opt, 'l2_decay_min', -4);
-    this.l2_decay_max = getopt(opt, 'l2_decay_max', 2);
-    this.learning_rate_min = getopt(opt, 'learning_rate_min', -4);
-    this.learning_rate_max = getopt(opt, 'learning_rate_max', 0);
-    this.momentum_min = getopt(opt, 'momentum_min', 0.9);
-    this.momentum_max = getopt(opt, 'momentum_max', 0.9);
-    this.neurons_min = getopt(opt, 'neurons_min', 5);
-    this.neurons_max = getopt(opt, 'neurons_max', 30);
-
-    // computed
-    this.folds = []; // data fold indices, gets filled by sampleFolds()
-    this.candidates = []; // candidate networks that are being currently evaluated
-    this.evaluated_candidates = []; // history of all candidates that were fully evaluated on all folds
-    this.unique_labels = arrUnique(labels);
-    this.iter = 0; // iteration counter, goes from 0 -> num_epochs * num_training_data
-    this.foldix = 0; // index of active fold
-
-    // callbacks
-    this.finish_fold_callback = null;
-    this.finish_batch_callback = null;
-
-    // initializations
-    if(this.data.length > 0) {
-      this.sampleFolds();
-      this.sampleCandidates();
-    }
-  };
-
-  MagicNet.prototype = {
-
-    // sets this.folds to a sampling of this.num_folds folds
-    sampleFolds: function() {
-      var N = this.data.length;
-      var num_train = Math.floor(this.train_ratio * N);
-      this.folds = []; // flush folds, if any
-      for(var i=0;i<this.num_folds;i++) {
-        var p = randperm(N);
-        this.folds.push({train_ix: p.slice(0, num_train), test_ix: p.slice(num_train, N)});
-      }
-    },
-
-    // returns a random candidate network
-    sampleCandidate: function() {
-      var input_depth = this.data[0].w.length;
-      var num_classes = this.unique_labels.length;
-
-      // sample network topology and hyperparameters
-      var layer_defs = [];
-      layer_defs.push({type:'input', out_sx:1, out_sy:1, out_depth: input_depth});
-      var nl = weightedSample([0,1,2,3], [0.2, 0.3, 0.3, 0.2]); // prefer nets with 1,2 hidden layers
-      for(var q=0;q<nl;q++) {
-        var ni = randi(this.neurons_min, this.neurons_max);
-        var act = ['tanh','maxout','relu'][randi(0,3)];
-        if(randf(0,1)<0.5) {
-          var dp = Math.random();
-          layer_defs.push({type:'fc', num_neurons: ni, activation: act, drop_prob: dp});
-        } else {
-          layer_defs.push({type:'fc', num_neurons: ni, activation: act});
-        }
-      }
-      layer_defs.push({type:'softmax', num_classes: num_classes});
-      var net = new Net();
-      net.makeLayers(layer_defs);
-
-      // sample training hyperparameters
-      var bs = randi(this.batch_size_min, this.batch_size_max); // batch size
-      var l2 = Math.pow(10, randf(this.l2_decay_min, this.l2_decay_max)); // l2 weight decay
-      var lr = Math.pow(10, randf(this.learning_rate_min, this.learning_rate_max)); // learning rate
-      var mom = randf(this.momentum_min, this.momentum_max); // momentum. Lets just use 0.9, works okay usually ;p
-      var tp = randf(0,1); // trainer type
-      var trainer_def;
-      if(tp<0.33) {
-        trainer_def = {method:'adadelta', batch_size:bs, l2_decay:l2};
-      } else if(tp<0.66) {
-        trainer_def = {method:'adagrad', learning_rate: lr, batch_size:bs, l2_decay:l2};
-      } else {
-        trainer_def = {method:'sgd', learning_rate: lr, momentum: mom, batch_size:bs, l2_decay:l2};
-      }
-      
-      var trainer = new Trainer(net, trainer_def);
-
-      var cand = {};
-      cand.acc = [];
-      cand.accv = 0; // this will maintained as sum(acc) for convenience
-      cand.layer_defs = layer_defs;
-      cand.trainer_def = trainer_def;
-      cand.net = net;
-      cand.trainer = trainer;
-      return cand;
-    },
-
-    // sets this.candidates with this.num_candidates candidate nets
-    sampleCandidates: function() {
-      this.candidates = []; // flush, if any
-      for(var i=0;i<this.num_candidates;i++) {
-        var cand = this.sampleCandidate();
-        this.candidates.push(cand);
-      }
-    },
-
-    step: function() {
-      
-      // run an example through current candidate
-      this.iter++;
-
-      // step all candidates on a random data point
-      var fold = this.folds[this.foldix]; // active fold
-      var dataix = fold.train_ix[randi(0, fold.train_ix.length)];
-      for(var k=0;k<this.candidates.length;k++) {
-        var x = this.data[dataix];
-        var l = this.labels[dataix];
-        this.candidates[k].trainer.train(x, l);
-      }
-
-      // process consequences: sample new folds, or candidates
-      var lastiter = this.num_epochs * fold.train_ix.length;
-      if(this.iter >= lastiter) {
-        // finished evaluation of this fold. Get final validation
-        // accuracies, record them, and go on to next fold.
-        var val_acc = this.evalValErrors();
-        for(var k=0;k<this.candidates.length;k++) {
-          var c = this.candidates[k];
-          c.acc.push(val_acc[k]);
-          c.accv += val_acc[k];
-        }
-        this.iter = 0; // reset step number
-        this.foldix++; // increment fold
-
-        if(this.finish_fold_callback !== null) {
-          this.finish_fold_callback();
-        }
-
-        if(this.foldix >= this.folds.length) {
-          // we finished all folds as well! Record these candidates
-          // and sample new ones to evaluate.
-          for(var k=0;k<this.candidates.length;k++) {
-            this.evaluated_candidates.push(this.candidates[k]);
-          }
-          // sort evaluated candidates according to accuracy achieved
-          this.evaluated_candidates.sort(function(a, b) { 
-            return (a.accv / a.acc.length) 
-                 > (b.accv / b.acc.length) 
-                 ? -1 : 1;
-          });
-          // and clip only to the top few ones (lets place limit at 3*ensemble_size)
-          // otherwise there are concerns with keeping these all in memory 
-          // if MagicNet is being evaluated for a very long time
-          if(this.evaluated_candidates.length > 3 * this.ensemble_size) {
-            this.evaluated_candidates = this.evaluated_candidates.slice(0, 3 * this.ensemble_size);
-          }
-          if(this.finish_batch_callback !== null) {
-            this.finish_batch_callback();
-          }
-          this.sampleCandidates(); // begin with new candidates
-          this.foldix = 0; // reset this
-        } else {
-          // we will go on to another fold. reset all candidates nets
-          for(var k=0;k<this.candidates.length;k++) {
-            var c = this.candidates[k];
-            var net = new Net();
-            net.makeLayers(c.layer_defs);
-            var trainer = new Trainer(net, c.trainer_def);
-            c.net = net;
-            c.trainer = trainer;
-          }
-        }
-      }
-    },
-
-    evalValErrors: function() {
-      // evaluate candidates on validation data and return performance of current networks
-      // as simple list
-      var vals = [];
-      var fold = this.folds[this.foldix]; // active fold
-      for(var k=0;k<this.candidates.length;k++) {
-        var net = this.candidates[k].net;
-        var v = 0.0;
-        for(var q=0;q<fold.test_ix.length;q++) {
-          var x = this.data[fold.test_ix[q]];
-          var l = this.labels[fold.test_ix[q]];
-          net.forward(x);
-          var yhat = net.getPrediction();
-          v += (yhat === l ? 1.0 : 0.0); // 0 1 loss
-        }
-        v /= fold.test_ix.length; // normalize
-        vals.push(v);
-      }
-      return vals;
-    },
-
-    // returns prediction scores for given test data point, as Vol
-    // uses an averaged prediction from the best ensemble_size models
-    // x is a Vol.
-    predict_soft: function(data) {
-      // forward prop the best networks
-      // and accumulate probabilities at last layer into a an output Vol
-      var nv = Math.min(this.ensemble_size, this.evaluated_candidates.length);
-      if(nv === 0) { return new convnetjs.Vol(0,0,0); } // not sure what to do here? we're not ready yet
-      var xout, n;
-      for(var j=0;j<nv;j++) {
-        var net = this.evaluated_candidates[j].net;
-        var x = net.forward(data);
-        if(j===0) { 
-          xout = x; 
-          n = x.w.length; 
-        } else {
-          // add it on
-          for(var d=0;d<n;d++) {
-            xout.w[d] += x.w[d];
-          }
-        }
-      }
-      // produce average
-      for(var d=0;d<n;d++) {
-        xout.w[d] /= n;
-      }
-      return xout;
-    },
-
-    predict: function(data) {
-      var xout = this.predict_soft(data);
-      if(xout.w.length !== 0) {
-        var stats = maxmin(xout.w);
-        var predicted_label = stats.maxi; 
-      } else {
-        var predicted_label = -1; // error out
-      }
-      return predicted_label;
-
-    },
-
-    toJSON: function() {
-      // dump the top ensemble_size networks as a list
-      var nv = Math.min(this.ensemble_size, this.evaluated_candidates.length);
-      var json = {};
-      json.nets = [];
-      for(var i=0;i<nv;i++) {
-        json.nets.push(this.evaluated_candidates[i].net.toJSON());
-      }
-      return json;
-    },
-
-    fromJSON: function(json) {
-      this.ensemble_size = json.nets.length;
-      this.evaluated_candidates = [];
-      for(var i=0;i<this.ensemble_size;i++) {
-        var net = new Net();
-        net.fromJSON(json.nets[i]);
-        var dummy_candidate = {};
-        dummy_candidate.net = net;
-        this.evaluated_candidates.push(dummy_candidate);
-      }
-    },
-
-    // callback functions
-    // called when a fold is finished, while evaluating a batch
-    onFinishFold: function(f) { this.finish_fold_callback = f; },
-    // called when a batch of candidates has finished evaluating
-    onFinishBatch: function(f) { this.finish_batch_callback = f; }
-    
-  };
-
-  global.MagicNet = MagicNet;
-})(convnetjs);
-(function(lib) {
-  "use strict";
-  if (typeof module === "undefined" || typeof module.exports === "undefined") {
-    window.jsfeat = lib; // in ordinary browser attach library to window
-  } else {
-    module.exports = lib; // in nodejs
-  }
-})(convnetjs);