Hugging Face's logo

wac81
/
toy_retnet_1.3b_pretrain

Fill-Mask
Transformers
PyTorch
retnet
Inference Endpoints
Model card Files Community
wac81 commited on
Commit
8f7942f
•
1 Parent(s): 029c758

Upload folder using huggingface_hub

Browse files
Files changed (12) hide show
  1. .gitattributes +2 -0
  2. config.json +28 -0
  3. index.html +17 -0
  4. optimizer.pt +3 -0
  5. pytorch_model.bin +3 -0
  6. rng_state.pth +3 -0
  7. scheduler.pt +3 -0
  8. special_tokens_map.json +24 -0
  9. tokenizer.model +3 -0
  10. tokenizer_config.json +39 -0
  11. trainer_state.json +3634 -0
  12. training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ optimizer.pt filter=lfs diff=lfs merge=lfs -text
37
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "RetNetModelWithLMHead"
4
+ ],
5
+ "chunk_size": 512,
6
+ "eos_token_id": 50256,
7
+ "ffn_proj_size": 4096,
8
+ "forward_impl": "parallel",
9
+ "hidden_size": 2048,
10
+ "initializer_range": 0.02,
11
+ "is_decoder": true,
12
+ "model_type": "retnet",
13
+ "num_heads": 8,
14
+ "num_layers": 24,
15
+ "output_retentions": false,
16
+ "pad_token_id": 50256,
17
+ "qk_dim": 2048,
18
+ "torch_dtype": "float32",
19
+ "transformers_version": "4.29.0",
20
+ "unk_token_id": 50256,
21
+ "use_bias_in_mlp": true,
22
+ "use_bias_in_msr": false,
23
+ "use_bias_in_msr_out": false,
24
+ "use_cache": true,
25
+ "use_default_gamma": false,
26
+ "v_dim": 4096,
27
+ "vocab_size": 64000
28
+ }
index.html ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!doctype html><html itemscope="" itemtype="http://schema.org/WebPage" lang="en"><head><meta content="Search the world's information, including webpages, images, videos and more. Google has many special features to help you find exactly what you're looking for." name="description"><meta content="noodp" name="robots"><meta content="text/html; charset=UTF-8" http-equiv="Content-Type"><meta content="/logos/doodles/2023/seasonal-holidays-2023-6753651837110165-law.gif" itemprop="image"><meta content="Seasonal Holidays 2023" property="twitter:title"><meta content="Happy Holidays! #GoogleDoodle" property="twitter:description"><meta content="Happy Holidays! #GoogleDoodle" property="og:description"><meta content="summary_large_image" property="twitter:card"><meta content="@GoogleDoodles" property="twitter:site"><meta content="https://www.google.com/logos/doodles/2023/seasonal-holidays-2023-6753651837110165-2xa.gif" property="twitter:image"><meta content="https://www.google.com/logos/doodles/2023/seasonal-holidays-2023-6753651837110165-2xa.gif" property="og:image"><meta content="1000" property="og:image:width"><meta content="400" property="og:image:height"><meta content="https://www.google.com/logos/doodles/2023/seasonal-holidays-2023-6753651837110165-2xa.gif" property="og:url"><meta content="video.other" property="og:type"><title>Google</title><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){var _g={kEI:'gtWKZdObB-Sk5NoP7dCMmAc',kEXPI:'0,1365467,207,4804,1132070,870537,327189,668,380096,16115,28684,22430,1362,12313,2822,14764,4998,17075,6884,34432,2891,561,3365,4423,3405,606,6189,23654,30847,6397,8927,2025,1,16916,2652,4,57402,2215,2980,24067,6630,7593,1,42154,2,16395,342,21266,1758,6700,31122,4568,6255,23421,1252,33064,2,2,1,10957,13669,2006,8155,23351,20506,7,1922,9779,42459,20198,20137,14,82,16514,3801,363,2049,25097,5122,3030,11151,4665,1804,7734,18098,575,7220,13455,1633,8841,4654,29970,7593,5303,5203197,6942,2,1586,5992169,1210,2803214,3311,141,795,7474665,5,16495898,4044106,16672,39684,4,4199,3,1603,3,262,3,234,3,2121276,2585,23029351,8163,4636,8408,8028,6672,1966,3164,9858,3610,818,1224,6235,3117,5878,4240,10366,10078,421,2,5885,261,8952,2,1296,2052,27,6,5,15,265,1390,4145,663,209,439,663,1644,46,80,1238,162,1068,3158,1031,2042,1,1994,230,49,885,159,39,7075,2422,2,2371,1252,107,853,149,518,1553,312,45,611,1598,2,571,1,6,2816,823,9,2,3300,1545,769,1468,218,3,2060,89,55,111,39,271,44,2227,210,376,753,3367,5,2161,808,626,813,320,19,247,1,6,2521,208,1055,813,708,768,3,4,2,2,2,29,141,110,609,899,318,4,391,463,879,591,281,250,1208,525,460,934,1294,19,2712,310,147,813,585,88,312,1444,2,6,1,627,6,771,78,63,49,4,1001,128,336,601,18,3,210,1,1,105,497,265,9,132,304,1206,87,215,932,315,1565,358,762,401,73,178,1627,717,563,83,438,8,446,4,14,199,62,2963,1,150,15,34,3,45,1,363,310,1001,495,4,397,117,21,3,663,3439,3,21310435,57419,3692,359917,728,218,510,3495,5,274,482,382',kBL:'qnU1',kOPI:89978449};(function(){var a;(null==(a=window.google)?0:a.stvsc)?google.kEI=_g.kEI:window.google=_g;}).call(this);})();(function(){google.sn='webhp';google.kHL='en';})();(function(){
2
+ var h=this||self;function l(){return void 0!==window.google&&void 0!==window.google.kOPI&&0!==window.google.kOPI?window.google.kOPI:null};var m,n=[];function p(a){for(var b;a&&(!a.getAttribute||!(b=a.getAttribute("eid")));)a=a.parentNode;return b||m}function q(a){for(var b=null;a&&(!a.getAttribute||!(b=a.getAttribute("leid")));)a=a.parentNode;return b}function r(a){/^http:/i.test(a)&&"https:"===window.location.protocol&&(google.ml&&google.ml(Error("a"),!1,{src:a,glmm:1}),a="");return a}
3
+ function t(a,b,c,d,k){var e="";-1===b.search("&ei=")&&(e="&ei="+p(d),-1===b.search("&lei=")&&(d=q(d))&&(e+="&lei="+d));d="";var g=-1===b.search("&cshid=")&&"slh"!==a,f=[];f.push(["zx",Date.now().toString()]);h._cshid&&g&&f.push(["cshid",h._cshid]);c=c();null!=c&&f.push(["opi",c.toString()]);for(c=0;c<f.length;c++){if(0===c||0<c)d+="&";d+=f[c][0]+"="+f[c][1]}return"/"+(k||"gen_204")+"?atyp=i&ct="+String(a)+"&cad="+(b+e+d)};m=google.kEI;google.getEI=p;google.getLEI=q;google.ml=function(){return null};google.log=function(a,b,c,d,k,e){e=void 0===e?l:e;c||(c=t(a,b,e,d,k));if(c=r(c)){a=new Image;var g=n.length;n[g]=a;a.onerror=a.onload=a.onabort=function(){delete n[g]};a.src=c}};google.logUrl=function(a,b){b=void 0===b?l:b;return t("",a,b)};}).call(this);(function(){google.y={};google.sy=[];google.x=function(a,b){if(a)var c=a.id;else{do c=Math.random();while(google.y[c])}google.y[c]=[a,b];return!1};google.sx=function(a){google.sy.push(a)};google.lm=[];google.plm=function(a){google.lm.push.apply(google.lm,a)};google.lq=[];google.load=function(a,b,c){google.lq.push([[a],b,c])};google.loadAll=function(a,b){google.lq.push([a,b])};google.bx=!1;google.lx=function(){};var d=[];google.fce=function(a,b,c,e){d.push([a,b,c,e])};google.qce=d;}).call(this);google.f={};(function(){
4
+ document.documentElement.addEventListener("submit",function(b){var a;if(a=b.target){var c=a.getAttribute("data-submitfalse");a="1"===c||"q"===c&&!a.elements.q.value?!0:!1}else a=!1;a&&(b.preventDefault(),b.stopPropagation())},!0);document.documentElement.addEventListener("click",function(b){var a;a:{for(a=b.target;a&&a!==document.documentElement;a=a.parentElement)if("A"===a.tagName){a="1"===a.getAttribute("data-nohref");break a}a=!1}a&&b.preventDefault()},!0);}).call(this);</script><style>#gbar,#guser{font-size:13px;padding-top:1px !important;}#gbar{height:22px}#guser{padding-bottom:7px !important;text-align:right}.gbh,.gbd{border-top:1px solid #c9d7f1;font-size:1px}.gbh{height:0;position:absolute;top:24px;width:100%}@media all{.gb1{height:22px;margin-right:.5em;vertical-align:top}#gbar{float:left}}a.gb1,a.gb4{text-decoration:underline !important}a.gb1,a.gb4{color:#00c !important}.gbi .gb4{color:#dd8e27 !important}.gbf .gb4{color:#900 !important}
5
+ </style><style>body,td,a,p,.h{font-family:arial,sans-serif}body{margin:0;overflow-y:scroll}#gog{padding:3px 8px 0}td{line-height:.8em}.gac_m td{line-height:17px}form{margin-bottom:20px}.h{color:#1967d2}em{font-weight:bold;font-style:normal}.lst{height:25px;width:496px}.gsfi,.lst{font:18px arial,sans-serif}.gsfs{font:17px arial,sans-serif}.ds{display:inline-box;display:inline-block;margin:3px 0 4px;margin-left:4px}input{font-family:inherit}body{background:#fff;color:#000}a{color:#681da8;text-decoration:none}a:hover,a:active{text-decoration:underline}.fl a{color:#1967d2}a:visited{color:#681da8}.sblc{padding-top:5px}.sblc a{display:block;margin:2px 0;margin-left:13px;font-size:11px}.lsbb{background:#f8f9fa;border:solid 1px;border-color:#dadce0 #70757a #70757a #dadce0;height:30px}.lsbb{display:block}#WqQANb a{display:inline-block;margin:0 12px}.lsb{background:url(/images/nav_logo229.png) 0 -261px repeat-x;color:#000;border:none;cursor:pointer;height:30px;margin:0;outline:0;font:15px arial,sans-serif;vertical-align:top}.lsb:active{background:#dadce0}.lst:focus{outline:none}</style><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){window.google.erd={jsr:1,bv:1919,de:true};
6
+ var h=this||self;var k,l=null!=(k=h.mei)?k:1,n,p=null!=(n=h.sdo)?n:!0,q=0,r,t=google.erd,v=t.jsr;google.ml=function(a,b,d,m,e){e=void 0===e?2:e;b&&(r=a&&a.message);void 0===d&&(d={});d.cad="ple_"+google.ple+".aple_"+google.aple;if(google.dl)return google.dl(a,e,d),null;if(0>v){window.console&&console.error(a,d);if(-2===v)throw a;b=!1}else b=!a||!a.message||"Error loading script"===a.message||q>=l&&!m?!1:!0;if(!b)return null;q++;d=d||{};b=encodeURIComponent;var c="/gen_204?atyp=i&ei="+b(google.kEI);google.kEXPI&&(c+="&jexpid="+b(google.kEXPI));c+="&srcpg="+b(google.sn)+"&jsr="+b(t.jsr)+"&bver="+
7
+ b(t.bv);var f=a.lineNumber;void 0!==f&&(c+="&line="+f);var g=a.fileName;g&&(0<g.indexOf("-extension:/")&&(e=3),c+="&script="+b(g),f&&g===window.location.href&&(f=document.documentElement.outerHTML.split("\n")[f],c+="&cad="+b(f?f.substring(0,300):"No script found.")));google.ple&&1===google.ple&&(e=2);c+="&jsel="+e;for(var u in d)c+="&",c+=b(u),c+="=",c+=b(d[u]);c=c+"&emsg="+b(a.name+": "+a.message);c=c+"&jsst="+b(a.stack||"N/A");12288<=c.length&&(c=c.substr(0,12288));a=c;m||google.log(0,"",a);return a};window.onerror=function(a,b,d,m,e){r!==a&&(a=e instanceof Error?e:Error(a),void 0===d||"lineNumber"in a||(a.lineNumber=d),void 0===b||"fileName"in a||(a.fileName=b),google.ml(a,!1,void 0,!1,"SyntaxError"===a.name||"SyntaxError"===a.message.substring(0,11)||-1!==a.message.indexOf("Script error")?3:0));r=null;p&&q>=l&&(window.onerror=null)};})();</script></head><body bgcolor="#fff"><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){var src='/images/nav_logo229.png';var iesg=false;document.body.onload = function(){window.n && window.n();if (document.images){new Image().src=src;}
8
+ if (!iesg){document.f&&document.f.q.focus();document.gbqf&&document.gbqf.q.focus();}
9
+ }
10
+ })();</script><div id="mngb"><div id=gbar><nobr><b class=gb1>Search</b> <a class=gb1 href="https://www.google.com/imghp?hl=en&tab=wi">Images</a> <a class=gb1 href="http://maps.google.com/maps?hl=en&tab=wl">Maps</a> <a class=gb1 href="https://play.google.com/?hl=en&tab=w8">Play</a> <a class=gb1 href="https://www.youtube.com/?tab=w1">YouTube</a> <a class=gb1 href="https://news.google.com/?tab=wn">News</a> <a class=gb1 href="https://mail.google.com/mail/?tab=wm">Gmail</a> <a class=gb1 href="https://drive.google.com/?tab=wo">Drive</a> <a class=gb1 style="text-decoration:none" href="https://www.google.com/intl/en/about/products?tab=wh"><u>More</u> &raquo;</a></nobr></div><div id=guser width=100%><nobr><span id=gbn class=gbi></span><span id=gbf class=gbf></span><span id=gbe></span><a href="http://www.google.com/history/optout?hl=en" class=gb4>Web History</a> | <a href="/preferences?hl=en" class=gb4>Settings</a> | <a target=_top id=gb_70 href="https://accounts.google.com/ServiceLogin?hl=en&passive=true&continue=http://www.google.com/&ec=GAZAAQ" class=gb4>Sign in</a></nobr></div><div class=gbh style=left:0></div><div class=gbh style=right:0></div></div><center><br clear="all" id="lgpd"><div id="lga"><img alt="Seasonal Holidays 2023" border="0" height="200" src="/logos/doodles/2023/seasonal-holidays-2023-6753651837110165-law.gif" title="Seasonal Holidays 2023" width="500" id="hplogo"><br></div><form action="/search" name="f"><table cellpadding="0" cellspacing="0"><tr valign="top"><td width="25%">&nbsp;</td><td align="center" nowrap=""><input name="ie" value="ISO-8859-1" type="hidden"><input value="en" name="hl" type="hidden"><input name="source" type="hidden" value="hp"><input name="biw" type="hidden"><input name="bih" type="hidden"><div class="ds" style="height:32px;margin:4px 0"><input class="lst" style="margin:0;padding:5px 8px 0 6px;vertical-align:top;color:#000" autocomplete="off" value="" title="Google Search" maxlength="2048" name="q" size="57"></div><br style="line-height:0"><span class="ds"><span class="lsbb"><input class="lsb" value="Google Search" name="btnG" type="submit"></span></span><span class="ds"><span class="lsbb"><input class="lsb" id="tsuid_1" value="I'm Feeling Lucky" name="btnI" type="submit"><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){var id='tsuid_1';document.getElementById(id).onclick = function(){if (this.form.q.value){this.checked = 1;if (this.form.iflsig)this.form.iflsig.disabled = false;}
11
+ else top.location='/doodles/';};})();</script><input value="AO6bgOgAAAAAZYrjkiExfeQRSEHVK7gSxtVMLH-7PGYL" name="iflsig" type="hidden"></span></span></td><td class="fl sblc" align="left" nowrap="" width="25%"><a href="/advanced_search?hl=en&amp;authuser=0">Advanced search</a></td></tr></table><input id="gbv" name="gbv" type="hidden" value="1"><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){var a,b="1";if(document&&document.getElementById)if("undefined"!=typeof XMLHttpRequest)b="2";else if("undefined"!=typeof ActiveXObject){var c,d,e=["MSXML2.XMLHTTP.6.0","MSXML2.XMLHTTP.3.0","MSXML2.XMLHTTP","Microsoft.XMLHTTP"];for(c=0;d=e[c++];)try{new ActiveXObject(d),b="2"}catch(h){}}a=b;if("2"==a&&-1==location.search.indexOf("&gbv=2")){var f=google.gbvu,g=document.getElementById("gbv");g&&(g.value=a);f&&window.setTimeout(function(){location.href=f},0)};}).call(this);</script></form><div id="gac_scont"></div><div style="font-size:83%;min-height:3.5em"><br></div><span id="footer"><div style="font-size:10pt"><div style="margin:19px auto;text-align:center" id="WqQANb"><a href="/intl/en/ads/">Advertising</a><a href="/services/">Business Solutions</a><a href="/intl/en/about.html">About Google</a></div></div><p style="font-size:8pt;color:#70757a">&copy; 2023 - <a href="/intl/en/policies/privacy/">Privacy</a> - <a href="/intl/en/policies/terms/">Terms</a></p></span></center><script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){window.google.cdo={height:757,width:1440};(function(){var a=window.innerWidth,b=window.innerHeight;if(!a||!b){var c=window.document,d="CSS1Compat"==c.compatMode?c.documentElement:c.body;a=d.clientWidth;b=d.clientHeight}
12
+ if(a&&b&&(a!=google.cdo.width||b!=google.cdo.height)){var e=google,f=e.log,g="/client_204?&atyp=i&biw="+a+"&bih="+b+"&ei="+google.kEI,h="",k=[],l=void 0!==window.google&&void 0!==window.google.kOPI&&0!==window.google.kOPI?window.google.kOPI:null;null!=l&&k.push(["opi",l.toString()]);for(var m=0;m<k.length;m++){if(0===m||0<m)h+="&";h+=k[m][0]+"="+k[m][1]}f.call(e,"","",g+h)};}).call(this);})();</script> <script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){google.xjs={ck:'xjs.hp.WxzFigac95U.L.X.O',combam:'AAAAAAAAAAAAAAAAAAAAAAAAAAAIAAAAAAA4AAACEAAAAAAAAAAAgABAdAQAwAIAXA',cs:'ACT90oG5GJwlApB6iXGEJqQQ5_y30Cmyng',cssam:'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACEAAAAAAAAAAAgABA',cssopt:false,csss:'ACT90oE1ELKT2MAkt-pRU4h2AmQxsBvWPA',excm:[],sepam:false,sepcss:false};})();</script> <script nonce="9VlSQlLTebcCdRc-6mXofg">(function(){var u='/xjs/_/js/k\x3dxjs.hp.en.6YapfFsyys0.O/am\x3dAAAAAAAAAAAAAAAAAAAAAAAAAAAIAAAAAAA4AAACEAAAAAAAAAAAgABAdAQAwAIAXA/d\x3d1/ed\x3d1/rs\x3dACT90oFa9UWkgkEahQVHgbKWj02U79yPSQ/m\x3dsb_he,d,cEt90b,SNUn3,qddgKe,sTsDMc,dtl0hd,eHDfl';var amd=0;
13
+ var e=this||self,f=function(a){return a};var g;var h=function(a){this.g=a};h.prototype.toString=function(){return this.g+""};var k={};var l=function(){var a=document;var b="SCRIPT";"application/xhtml+xml"===a.contentType&&(b=b.toLowerCase());return a.createElement(b)};
14
+ function m(a,b){a.src=b instanceof h&&b.constructor===h?b.g:"type_error:TrustedResourceUrl";var c,d;(c=(b=null==(d=(c=(a.ownerDocument&&a.ownerDocument.defaultView||window).document).querySelector)?void 0:d.call(c,"script[nonce]"))?b.nonce||b.getAttribute("nonce")||"":"")&&a.setAttribute("nonce",c)};function n(a){a=null===a?"null":void 0===a?"undefined":a;if(void 0===g){var b=null;var c=e.trustedTypes;if(c&&c.createPolicy){try{b=c.createPolicy("goog#html",{createHTML:f,createScript:f,createScriptURL:f})}catch(d){e.console&&e.console.error(d.message)}g=b}else g=b}a=(b=g)?b.createScriptURL(a):a;return new h(a,k)};void 0===google.ps&&(google.ps=[]);function p(){var a=u,b=function(){};google.lx=google.stvsc?b:function(){q(a);google.lx=b};google.bx||google.lx()}function r(a,b){b&&m(a,n(b));var c=a.onload;a.onload=function(d){c&&c(d);google.ps=google.ps.filter(function(t){return a!==t})};google.ps.push(a);document.body.appendChild(a)}google.as=r;function q(a){google.timers&&google.timers.load&&google.tick&&google.tick("load","xjsls");var b=l();b.onerror=function(){google.ple=1};b.onload=function(){google.ple=0};google.xjsus=void 0;r(b,a);google.aple=-1;google.psa=!0};google.xjsu=u;e._F_jsUrl=u;setTimeout(function(){0<amd?google.caft(function(){return p()},amd):p()},0);})();window._ = window._ || {};window._DumpException = _._DumpException = function(e){throw e;};window._s = window._s || {};_s._DumpException = _._DumpException;window._qs = window._qs || {};_qs._DumpException = _._DumpException;(function(){var t=[1,0,0,0,0,8192,0,8388654,134217744,33554432,524288,72976,469762752,1];window._F_toggles = window._xjs_toggles = t;})();function _F_installCss(c){}
15
+ (function(){google.jl={blt:'none',chnk:0,dw:false,dwu:true,emtn:0,end:0,ico:false,ikb:0,ine:false,injs:'none',injt:0,injth:0,injv2:false,lls:'default',pdt:0,rep:0,snet:true,strt:0,ubm:false,uwp:true};})();(function(){var pmc='{\x22d\x22:{},\x22sb_he\x22:{\x22agen\x22:false,\x22cgen\x22:false,\x22client\x22:\x22heirloom-hp\x22,\x22dh\x22:true,\x22ds\x22:\x22\x22,\x22fl\x22:true,\x22host\x22:\x22google.com\x22,\x22jsonp\x22:true,\x22msgs\x22:{\x22cibl\x22:\x22Clear Search\x22,\x22dym\x22:\x22Did you mean:\x22,\x22lcky\x22:\x22I\\u0026#39;m Feeling Lucky\x22,\x22lml\x22:\x22Learn more\x22,\x22psrc\x22:\x22This search was removed from your \\u003Ca href\x3d\\\x22/history\\\x22\\u003EWeb History\\u003C/a\\u003E\x22,\x22psrl\x22:\x22Remove\x22,\x22sbit\x22:\x22Search by image\x22,\x22srch\x22:\x22Google Search\x22},\x22ovr\x22:{},\x22pq\x22:\x22\x22,\x22rfs\x22:[],\x22sbas\x22:\x220 3px 8px 0 rgba(0,0,0,0.2),0 0 0 1px rgba(0,0,0,0.08)\x22,\x22stok\x22:\x22Bnt3al-Htt2q_NpQTgsFPz0XxCQ\x22}}';google.pmc=JSON.parse(pmc);})();(function(){var b=function(a){var c=0;return function(){return c<a.length?{done:!1,value:a[c++]}:{done:!0}}};
16
+ var e=this||self;var g,h;a:{for(var k=["CLOSURE_FLAGS"],l=e,n=0;n<k.length;n++)if(l=l[k[n]],null==l){h=null;break a}h=l}var p=h&&h[610401301];g=null!=p?p:!1;var q,r=e.navigator;q=r?r.userAgentData||null:null;function t(a){return g?q?q.brands.some(function(c){return(c=c.brand)&&-1!=c.indexOf(a)}):!1:!1}function u(a){var c;a:{if(c=e.navigator)if(c=c.userAgent)break a;c=""}return-1!=c.indexOf(a)};function v(){return g?!!q&&0<q.brands.length:!1}function w(){return u("Safari")&&!(x()||(v()?0:u("Coast"))||(v()?0:u("Opera"))||(v()?0:u("Edge"))||(v()?t("Microsoft Edge"):u("Edg/"))||(v()?t("Opera"):u("OPR"))||u("Firefox")||u("FxiOS")||u("Silk")||u("Android"))}function x(){return v()?t("Chromium"):(u("Chrome")||u("CriOS"))&&!(v()?0:u("Edge"))||u("Silk")}function y(){return u("Android")&&!(x()||u("Firefox")||u("FxiOS")||(v()?0:u("Opera"))||u("Silk"))};var z=v()?!1:u("Trident")||u("MSIE");y();x();w();Object.freeze(new function(){});Object.freeze(new function(){});var A=!z&&!w(),D=function(a){if(/-[a-z]/.test("ved"))return null;if(A&&a.dataset){if(y()&&!("ved"in a.dataset))return null;a=a.dataset.ved;return void 0===a?null:a}return a.getAttribute("data-"+"ved".replace(/([A-Z])/g,"-$1").toLowerCase())};var E=[],F=null;function G(a){a=a.target;var c=performance.now(),f=[],H=f.concat,d=E;if(!(d instanceof Array)){var m="undefined"!=typeof Symbol&&Symbol.iterator&&d[Symbol.iterator];if(m)d=m.call(d);else if("number"==typeof d.length)d={next:b(d)};else throw Error("a`"+String(d));for(var B=[];!(m=d.next()).done;)B.push(m.value);d=B}E=H.call(f,d,[c]);if(a&&a instanceof HTMLElement)if(a===F){if(c=4<=E.length)c=5>(E[E.length-1]-E[E.length-4])/1E3;if(c){c=google.getEI(a);a.hasAttribute("data-ved")?f=a?D(a)||"":"":f=(f=
17
+ a.closest("[data-ved]"))?D(f)||"":"";f=f||"";if(a.hasAttribute("jsname"))a=a.getAttribute("jsname");else{var C;a=null==(C=a.closest("[jsname]"))?void 0:C.getAttribute("jsname")}google.log("rcm","&ei="+c+"&ved="+f+"&jsname="+(a||""))}}else F=a,E=[c]}window.document.addEventListener("DOMContentLoaded",function(){document.body.addEventListener("click",G)});}).call(this);</script></body></html>
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37a21747294a72f8b0179a6b18f3725e53771003b7e025c0274003b9798f767f
3
+ size 10715166669
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba455c9d564745898593e9bf423b362aeded4b7b00943708c6ef5bd94128b900
3
+ size 5357702585
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2278a87cdf86c3f9219223c847f6b27f6b7f15b8226b617f38936e8ff2cbcde
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a33122775d3d1c6f825078cf641ca5b6a22239f6742050b9bbc45a4afded537
3
+ size 627
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<unk>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4be54af290d93c113bcbf421115ae9eed9d6340408f564898f1e966dc738ef01
3
+ size 1136699
tokenizer_config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_eos_token": false,
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "baichuan-inc/baichuan-7B--tokenization_baichuan.BaiChuanTokenizer",
7
+ null
8
+ ]
9
+ },
10
+ "bos_token": {
11
+ "__type": "AddedToken",
12
+ "content": "<s>",
13
+ "lstrip": false,
14
+ "normalized": true,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "clean_up_tokenization_spaces": false,
19
+ "eos_token": {
20
+ "__type": "AddedToken",
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false
26
+ },
27
+ "model_max_length": 1000000000000000019884624838656,
28
+ "pad_token": null,
29
+ "sp_model_kwargs": {},
30
+ "tokenizer_class": "BaiChuanTokenizer",
31
+ "unk_token": {
32
+ "__type": "AddedToken",
33
+ "content": "<unk>",
34
+ "lstrip": false,
35
+ "normalized": true,
36
+ "rstrip": false,
37
+ "single_word": false
38
+ }
39
+ }
trainer_state.json ADDED
@@ -0,0 +1,3634 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9058960565963444,
5
+ "global_step": 301500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.99749613653868e-05,
13
+ "loss": 4.5704,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.9949922730773584e-05,
19
+ "loss": 3.9367,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.0,
24
+ "learning_rate": 4.992488409616038e-05,
25
+ "loss": 4.3169,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.9899845461547165e-05,
31
+ "loss": 4.4226,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.987480682693396e-05,
37
+ "loss": 4.3777,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.984976819232076e-05,
43
+ "loss": 4.3272,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 4.982472955770754e-05,
49
+ "loss": 4.2964,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.01,
54
+ "learning_rate": 4.979969092309434e-05,
55
+ "loss": 4.2587,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.01,
60
+ "learning_rate": 4.9774652288481124e-05,
61
+ "loss": 4.2217,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.974961365386792e-05,
67
+ "loss": 4.2056,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 4.972457501925471e-05,
73
+ "loss": 4.2108,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 4.96995363846415e-05,
79
+ "loss": 4.135,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.02,
84
+ "learning_rate": 4.96744977500283e-05,
85
+ "loss": 4.1552,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.02,
90
+ "learning_rate": 4.9649459115415084e-05,
91
+ "loss": 4.0666,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.02,
96
+ "learning_rate": 4.962442048080188e-05,
97
+ "loss": 4.1164,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.02,
102
+ "learning_rate": 4.959938184618867e-05,
103
+ "loss": 4.0798,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.03,
108
+ "learning_rate": 4.957434321157546e-05,
109
+ "loss": 4.0284,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.03,
114
+ "learning_rate": 4.954930457696226e-05,
115
+ "loss": 4.046,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.03,
120
+ "learning_rate": 4.9524265942349043e-05,
121
+ "loss": 4.0595,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.03,
126
+ "learning_rate": 4.949922730773584e-05,
127
+ "loss": 4.0243,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.03,
132
+ "learning_rate": 4.947418867312263e-05,
133
+ "loss": 4.0064,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.03,
138
+ "learning_rate": 4.944915003850942e-05,
139
+ "loss": 3.9907,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.03,
144
+ "learning_rate": 4.942411140389621e-05,
145
+ "loss": 3.9488,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.04,
150
+ "learning_rate": 4.9399072769283e-05,
151
+ "loss": 3.95,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.04,
156
+ "learning_rate": 4.93740341346698e-05,
157
+ "loss": 3.9148,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.04,
162
+ "learning_rate": 4.934899550005659e-05,
163
+ "loss": 3.917,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.04,
168
+ "learning_rate": 4.932395686544338e-05,
169
+ "loss": 3.8958,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.04,
174
+ "learning_rate": 4.929891823083017e-05,
175
+ "loss": 3.9248,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.04,
180
+ "learning_rate": 4.927387959621696e-05,
181
+ "loss": 3.8752,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.05,
186
+ "learning_rate": 4.924884096160376e-05,
187
+ "loss": 3.8548,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.05,
192
+ "learning_rate": 4.922380232699055e-05,
193
+ "loss": 3.8882,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.05,
198
+ "learning_rate": 4.919876369237734e-05,
199
+ "loss": 3.8457,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.05,
204
+ "learning_rate": 4.917372505776413e-05,
205
+ "loss": 3.8472,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.05,
210
+ "learning_rate": 4.914868642315092e-05,
211
+ "loss": 3.8743,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.05,
216
+ "learning_rate": 4.912364778853771e-05,
217
+ "loss": 3.8541,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.05,
222
+ "learning_rate": 4.909860915392451e-05,
223
+ "loss": 3.823,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.06,
228
+ "learning_rate": 4.90735705193113e-05,
229
+ "loss": 3.841,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.06,
234
+ "learning_rate": 4.904853188469809e-05,
235
+ "loss": 3.8262,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.06,
240
+ "learning_rate": 4.902349325008488e-05,
241
+ "loss": 3.7779,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.06,
246
+ "learning_rate": 4.899845461547167e-05,
247
+ "loss": 3.7952,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.06,
252
+ "learning_rate": 4.897341598085847e-05,
253
+ "loss": 3.8115,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.06,
258
+ "learning_rate": 4.894837734624526e-05,
259
+ "loss": 3.7978,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.06,
264
+ "learning_rate": 4.892333871163205e-05,
265
+ "loss": 3.7757,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.07,
270
+ "learning_rate": 4.889830007701884e-05,
271
+ "loss": 3.7953,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.07,
276
+ "learning_rate": 4.887326144240563e-05,
277
+ "loss": 3.7988,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.07,
282
+ "learning_rate": 4.884822280779243e-05,
283
+ "loss": 3.779,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.07,
288
+ "learning_rate": 4.882318417317922e-05,
289
+ "loss": 3.7545,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.07,
294
+ "learning_rate": 4.879814553856601e-05,
295
+ "loss": 3.7502,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.07,
300
+ "learning_rate": 4.87731069039528e-05,
301
+ "loss": 3.762,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.08,
306
+ "learning_rate": 4.874806826933959e-05,
307
+ "loss": 3.7711,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.08,
312
+ "learning_rate": 4.872302963472639e-05,
313
+ "loss": 3.7314,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.08,
318
+ "learning_rate": 4.869799100011317e-05,
319
+ "loss": 3.7427,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.08,
324
+ "learning_rate": 4.867295236549997e-05,
325
+ "loss": 3.7024,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.08,
330
+ "learning_rate": 4.864791373088676e-05,
331
+ "loss": 3.7356,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.08,
336
+ "learning_rate": 4.862287509627355e-05,
337
+ "loss": 3.7117,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.08,
342
+ "learning_rate": 4.859783646166035e-05,
343
+ "loss": 3.7235,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.09,
348
+ "learning_rate": 4.857279782704713e-05,
349
+ "loss": 3.6791,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.09,
354
+ "learning_rate": 4.854775919243393e-05,
355
+ "loss": 3.696,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.09,
360
+ "learning_rate": 4.852272055782072e-05,
361
+ "loss": 3.6852,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.09,
366
+ "learning_rate": 4.849768192320751e-05,
367
+ "loss": 3.6896,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.09,
372
+ "learning_rate": 4.847264328859431e-05,
373
+ "loss": 3.6759,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.09,
378
+ "learning_rate": 4.844760465398109e-05,
379
+ "loss": 3.7012,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.09,
384
+ "learning_rate": 4.842256601936789e-05,
385
+ "loss": 3.6378,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.1,
390
+ "learning_rate": 4.839752738475467e-05,
391
+ "loss": 3.6746,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.1,
396
+ "learning_rate": 4.837248875014147e-05,
397
+ "loss": 3.68,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.1,
402
+ "learning_rate": 4.834745011552827e-05,
403
+ "loss": 3.6775,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.1,
408
+ "learning_rate": 4.832241148091505e-05,
409
+ "loss": 3.6471,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.1,
414
+ "learning_rate": 4.829737284630185e-05,
415
+ "loss": 3.662,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.1,
420
+ "learning_rate": 4.827233421168863e-05,
421
+ "loss": 3.6359,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.11,
426
+ "learning_rate": 4.824729557707543e-05,
427
+ "loss": 3.6494,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.11,
432
+ "learning_rate": 4.822225694246222e-05,
433
+ "loss": 3.6187,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.11,
438
+ "learning_rate": 4.819721830784901e-05,
439
+ "loss": 3.6115,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.11,
444
+ "learning_rate": 4.817217967323581e-05,
445
+ "loss": 3.6062,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.11,
450
+ "learning_rate": 4.814714103862259e-05,
451
+ "loss": 3.6109,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.11,
456
+ "learning_rate": 4.812210240400939e-05,
457
+ "loss": 3.6239,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.11,
462
+ "learning_rate": 4.809706376939618e-05,
463
+ "loss": 3.646,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.12,
468
+ "learning_rate": 4.807202513478297e-05,
469
+ "loss": 3.5898,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.12,
474
+ "learning_rate": 4.804698650016977e-05,
475
+ "loss": 3.6008,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.12,
480
+ "learning_rate": 4.802194786555655e-05,
481
+ "loss": 3.6265,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.12,
486
+ "learning_rate": 4.799690923094335e-05,
487
+ "loss": 3.6111,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.12,
492
+ "learning_rate": 4.797187059633014e-05,
493
+ "loss": 3.5766,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.12,
498
+ "learning_rate": 4.794683196171693e-05,
499
+ "loss": 3.5916,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.12,
504
+ "learning_rate": 4.792179332710373e-05,
505
+ "loss": 3.641,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.13,
510
+ "learning_rate": 4.789675469249051e-05,
511
+ "loss": 3.5743,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.13,
516
+ "learning_rate": 4.787171605787731e-05,
517
+ "loss": 3.6153,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.13,
522
+ "learning_rate": 4.78466774232641e-05,
523
+ "loss": 3.6073,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.13,
528
+ "learning_rate": 4.782163878865089e-05,
529
+ "loss": 3.5881,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.13,
534
+ "learning_rate": 4.779660015403769e-05,
535
+ "loss": 3.5649,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.13,
540
+ "learning_rate": 4.777156151942447e-05,
541
+ "loss": 3.5602,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.14,
546
+ "learning_rate": 4.774652288481127e-05,
547
+ "loss": 3.5707,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.14,
552
+ "learning_rate": 4.772148425019806e-05,
553
+ "loss": 3.5546,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.14,
558
+ "learning_rate": 4.769644561558485e-05,
559
+ "loss": 3.5327,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.14,
564
+ "learning_rate": 4.767140698097164e-05,
565
+ "loss": 3.5489,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.14,
570
+ "learning_rate": 4.764636834635843e-05,
571
+ "loss": 3.5559,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.14,
576
+ "learning_rate": 4.762132971174523e-05,
577
+ "loss": 3.5514,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.14,
582
+ "learning_rate": 4.759629107713202e-05,
583
+ "loss": 3.5295,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.15,
588
+ "learning_rate": 4.757125244251881e-05,
589
+ "loss": 3.5198,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.15,
594
+ "learning_rate": 4.75462138079056e-05,
595
+ "loss": 3.5231,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.15,
600
+ "learning_rate": 4.752117517329239e-05,
601
+ "loss": 3.5695,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.15,
606
+ "learning_rate": 4.749613653867919e-05,
607
+ "loss": 3.581,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.15,
612
+ "learning_rate": 4.747109790406598e-05,
613
+ "loss": 3.5228,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.15,
618
+ "learning_rate": 4.744605926945277e-05,
619
+ "loss": 3.5382,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.15,
624
+ "learning_rate": 4.742102063483956e-05,
625
+ "loss": 3.5384,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.16,
630
+ "learning_rate": 4.739598200022635e-05,
631
+ "loss": 3.4945,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.16,
636
+ "learning_rate": 4.737094336561314e-05,
637
+ "loss": 3.503,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.16,
642
+ "learning_rate": 4.734590473099994e-05,
643
+ "loss": 3.523,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.16,
648
+ "learning_rate": 4.732086609638673e-05,
649
+ "loss": 3.5115,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.16,
654
+ "learning_rate": 4.729582746177352e-05,
655
+ "loss": 3.5014,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.16,
660
+ "learning_rate": 4.727078882716031e-05,
661
+ "loss": 3.5253,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.17,
666
+ "learning_rate": 4.72457501925471e-05,
667
+ "loss": 3.4913,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.17,
672
+ "learning_rate": 4.72207115579339e-05,
673
+ "loss": 3.5152,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.17,
678
+ "learning_rate": 4.719567292332069e-05,
679
+ "loss": 3.4847,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.17,
684
+ "learning_rate": 4.717063428870748e-05,
685
+ "loss": 3.5086,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.17,
690
+ "learning_rate": 4.714559565409427e-05,
691
+ "loss": 3.5071,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.17,
696
+ "learning_rate": 4.712055701948106e-05,
697
+ "loss": 3.4918,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.17,
702
+ "learning_rate": 4.709551838486786e-05,
703
+ "loss": 3.4677,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.18,
708
+ "learning_rate": 4.707047975025464e-05,
709
+ "loss": 3.4594,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.18,
714
+ "learning_rate": 4.704544111564144e-05,
715
+ "loss": 3.5236,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.18,
720
+ "learning_rate": 4.702040248102823e-05,
721
+ "loss": 3.4743,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.18,
726
+ "learning_rate": 4.699536384641502e-05,
727
+ "loss": 3.478,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.18,
732
+ "learning_rate": 4.6970325211801817e-05,
733
+ "loss": 3.458,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.18,
738
+ "learning_rate": 4.69452865771886e-05,
739
+ "loss": 3.464,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.18,
744
+ "learning_rate": 4.69202479425754e-05,
745
+ "loss": 3.4501,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.19,
750
+ "learning_rate": 4.689520930796219e-05,
751
+ "loss": 3.4772,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.19,
756
+ "learning_rate": 4.687017067334898e-05,
757
+ "loss": 3.4522,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.19,
762
+ "learning_rate": 4.6845132038735776e-05,
763
+ "loss": 3.4415,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.19,
768
+ "learning_rate": 4.682009340412256e-05,
769
+ "loss": 3.4953,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.19,
774
+ "learning_rate": 4.679505476950936e-05,
775
+ "loss": 3.4713,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.19,
780
+ "learning_rate": 4.677001613489615e-05,
781
+ "loss": 3.4588,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.2,
786
+ "learning_rate": 4.674497750028294e-05,
787
+ "loss": 3.4765,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.2,
792
+ "learning_rate": 4.671993886566973e-05,
793
+ "loss": 3.4408,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.2,
798
+ "learning_rate": 4.669490023105652e-05,
799
+ "loss": 3.4454,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.2,
804
+ "learning_rate": 4.666986159644332e-05,
805
+ "loss": 3.4629,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.2,
810
+ "learning_rate": 4.66448229618301e-05,
811
+ "loss": 3.4751,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.2,
816
+ "learning_rate": 4.66197843272169e-05,
817
+ "loss": 3.446,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.2,
822
+ "learning_rate": 4.659474569260369e-05,
823
+ "loss": 3.4132,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.21,
828
+ "learning_rate": 4.656970705799048e-05,
829
+ "loss": 3.4266,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.21,
834
+ "learning_rate": 4.6544668423377277e-05,
835
+ "loss": 3.4197,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.21,
840
+ "learning_rate": 4.651962978876406e-05,
841
+ "loss": 3.444,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.21,
846
+ "learning_rate": 4.649459115415086e-05,
847
+ "loss": 3.4453,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.21,
852
+ "learning_rate": 4.646955251953765e-05,
853
+ "loss": 3.4215,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.21,
858
+ "learning_rate": 4.644451388492444e-05,
859
+ "loss": 3.4524,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.21,
864
+ "learning_rate": 4.6419475250311236e-05,
865
+ "loss": 3.4432,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.22,
870
+ "learning_rate": 4.639443661569802e-05,
871
+ "loss": 3.4105,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.22,
876
+ "learning_rate": 4.636939798108482e-05,
877
+ "loss": 3.415,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.22,
882
+ "learning_rate": 4.634435934647161e-05,
883
+ "loss": 3.4441,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.22,
888
+ "learning_rate": 4.63193207118584e-05,
889
+ "loss": 3.4117,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.22,
894
+ "learning_rate": 4.6294282077245196e-05,
895
+ "loss": 3.4101,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.22,
900
+ "learning_rate": 4.626924344263198e-05,
901
+ "loss": 3.3981,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.23,
906
+ "learning_rate": 4.624420480801878e-05,
907
+ "loss": 3.4246,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.23,
912
+ "learning_rate": 4.621916617340557e-05,
913
+ "loss": 3.423,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.23,
918
+ "learning_rate": 4.619412753879236e-05,
919
+ "loss": 3.4062,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.23,
924
+ "learning_rate": 4.6169088904179155e-05,
925
+ "loss": 3.4059,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.23,
930
+ "learning_rate": 4.614405026956594e-05,
931
+ "loss": 3.3988,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.23,
936
+ "learning_rate": 4.6119011634952736e-05,
937
+ "loss": 3.4308,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.23,
942
+ "learning_rate": 4.609397300033953e-05,
943
+ "loss": 3.3904,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.24,
948
+ "learning_rate": 4.606893436572632e-05,
949
+ "loss": 3.3847,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.24,
954
+ "learning_rate": 4.6043895731113115e-05,
955
+ "loss": 3.3941,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.24,
960
+ "learning_rate": 4.60188570964999e-05,
961
+ "loss": 3.3834,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.24,
966
+ "learning_rate": 4.5993818461886696e-05,
967
+ "loss": 3.4271,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.24,
972
+ "learning_rate": 4.596877982727349e-05,
973
+ "loss": 3.3714,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.24,
978
+ "learning_rate": 4.594374119266028e-05,
979
+ "loss": 3.3709,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.24,
984
+ "learning_rate": 4.591870255804707e-05,
985
+ "loss": 3.4005,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.25,
990
+ "learning_rate": 4.589366392343386e-05,
991
+ "loss": 3.3783,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.25,
996
+ "learning_rate": 4.5868625288820656e-05,
997
+ "loss": 3.385,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.25,
1002
+ "learning_rate": 4.5843586654207446e-05,
1003
+ "loss": 3.3812,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.25,
1008
+ "learning_rate": 4.581854801959424e-05,
1009
+ "loss": 3.362,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.25,
1014
+ "learning_rate": 4.579350938498103e-05,
1015
+ "loss": 3.3715,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.25,
1020
+ "learning_rate": 4.576847075036782e-05,
1021
+ "loss": 3.387,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.26,
1026
+ "learning_rate": 4.5743432115754615e-05,
1027
+ "loss": 3.3741,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.26,
1032
+ "learning_rate": 4.5718393481141406e-05,
1033
+ "loss": 3.3461,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.26,
1038
+ "learning_rate": 4.5693354846528196e-05,
1039
+ "loss": 3.3659,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 0.26,
1044
+ "learning_rate": 4.566831621191499e-05,
1045
+ "loss": 3.3693,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 0.26,
1050
+ "learning_rate": 4.564327757730178e-05,
1051
+ "loss": 3.3776,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 0.26,
1056
+ "learning_rate": 4.561823894268857e-05,
1057
+ "loss": 3.3391,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 0.26,
1062
+ "learning_rate": 4.5593200308075365e-05,
1063
+ "loss": 3.3516,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 0.27,
1068
+ "learning_rate": 4.5568161673462156e-05,
1069
+ "loss": 3.3808,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 0.27,
1074
+ "learning_rate": 4.5543123038848947e-05,
1075
+ "loss": 3.3933,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 0.27,
1080
+ "learning_rate": 4.551808440423574e-05,
1081
+ "loss": 3.3717,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 0.27,
1086
+ "learning_rate": 4.549304576962253e-05,
1087
+ "loss": 3.3625,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 0.27,
1092
+ "learning_rate": 4.5468007135009325e-05,
1093
+ "loss": 3.3391,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 0.27,
1098
+ "learning_rate": 4.5442968500396116e-05,
1099
+ "loss": 3.3361,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 0.27,
1104
+ "learning_rate": 4.5417929865782906e-05,
1105
+ "loss": 3.3201,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 0.28,
1110
+ "learning_rate": 4.53928912311697e-05,
1111
+ "loss": 3.3444,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 0.28,
1116
+ "learning_rate": 4.536785259655649e-05,
1117
+ "loss": 3.3624,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 0.28,
1122
+ "learning_rate": 4.5342813961943285e-05,
1123
+ "loss": 3.3289,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 0.28,
1128
+ "learning_rate": 4.531777532733007e-05,
1129
+ "loss": 3.342,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 0.28,
1134
+ "learning_rate": 4.5292736692716866e-05,
1135
+ "loss": 3.3372,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 0.28,
1140
+ "learning_rate": 4.5267698058103656e-05,
1141
+ "loss": 3.3458,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 0.29,
1146
+ "learning_rate": 4.524265942349045e-05,
1147
+ "loss": 3.386,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 0.29,
1152
+ "learning_rate": 4.521762078887724e-05,
1153
+ "loss": 3.336,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 0.29,
1158
+ "learning_rate": 4.519258215426403e-05,
1159
+ "loss": 3.3324,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 0.29,
1164
+ "learning_rate": 4.5167543519650825e-05,
1165
+ "loss": 3.3054,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 0.29,
1170
+ "learning_rate": 4.5142504885037616e-05,
1171
+ "loss": 3.3409,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 0.29,
1176
+ "learning_rate": 4.5117466250424406e-05,
1177
+ "loss": 3.3233,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 0.29,
1182
+ "learning_rate": 4.50924276158112e-05,
1183
+ "loss": 3.3296,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 0.3,
1188
+ "learning_rate": 4.506738898119799e-05,
1189
+ "loss": 3.3135,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 0.3,
1194
+ "learning_rate": 4.5042350346584785e-05,
1195
+ "loss": 3.3307,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 0.3,
1200
+ "learning_rate": 4.5017311711971576e-05,
1201
+ "loss": 3.3117,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 0.3,
1206
+ "learning_rate": 4.4992273077358366e-05,
1207
+ "loss": 3.3407,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 0.3,
1212
+ "learning_rate": 4.496723444274516e-05,
1213
+ "loss": 3.3376,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 0.3,
1218
+ "learning_rate": 4.494219580813195e-05,
1219
+ "loss": 3.3178,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 0.3,
1224
+ "learning_rate": 4.4917157173518745e-05,
1225
+ "loss": 3.3416,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 0.31,
1230
+ "learning_rate": 4.489211853890553e-05,
1231
+ "loss": 3.2907,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 0.31,
1236
+ "learning_rate": 4.4867079904292326e-05,
1237
+ "loss": 3.3315,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 0.31,
1242
+ "learning_rate": 4.4842041269679116e-05,
1243
+ "loss": 3.3199,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 0.31,
1248
+ "learning_rate": 4.481700263506591e-05,
1249
+ "loss": 3.3082,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 0.31,
1254
+ "learning_rate": 4.4791964000452704e-05,
1255
+ "loss": 3.3224,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 0.31,
1260
+ "learning_rate": 4.476692536583949e-05,
1261
+ "loss": 3.3023,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 0.32,
1266
+ "learning_rate": 4.4741886731226285e-05,
1267
+ "loss": 3.3068,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 0.32,
1272
+ "learning_rate": 4.4716848096613076e-05,
1273
+ "loss": 3.3125,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 0.32,
1278
+ "learning_rate": 4.4691809461999866e-05,
1279
+ "loss": 3.3026,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 0.32,
1284
+ "learning_rate": 4.4666770827386664e-05,
1285
+ "loss": 3.302,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 0.32,
1290
+ "learning_rate": 4.464173219277345e-05,
1291
+ "loss": 3.299,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 0.32,
1296
+ "learning_rate": 4.4616693558160245e-05,
1297
+ "loss": 3.2802,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 0.32,
1302
+ "learning_rate": 4.4591654923547035e-05,
1303
+ "loss": 3.2993,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 0.33,
1308
+ "learning_rate": 4.4566616288933826e-05,
1309
+ "loss": 3.2822,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 0.33,
1314
+ "learning_rate": 4.454157765432062e-05,
1315
+ "loss": 3.2963,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 0.33,
1320
+ "learning_rate": 4.451653901970741e-05,
1321
+ "loss": 3.3129,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 0.33,
1326
+ "learning_rate": 4.4491500385094204e-05,
1327
+ "loss": 3.3025,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 0.33,
1332
+ "learning_rate": 4.4466461750480995e-05,
1333
+ "loss": 3.3245,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 0.33,
1338
+ "learning_rate": 4.4441423115867786e-05,
1339
+ "loss": 3.2676,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 0.34,
1344
+ "learning_rate": 4.441638448125458e-05,
1345
+ "loss": 3.317,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 0.34,
1350
+ "learning_rate": 4.439134584664137e-05,
1351
+ "loss": 3.3008,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 0.34,
1356
+ "learning_rate": 4.4366307212028164e-05,
1357
+ "loss": 3.3093,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 0.34,
1362
+ "learning_rate": 4.4341268577414955e-05,
1363
+ "loss": 3.3066,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 0.34,
1368
+ "learning_rate": 4.4316229942801745e-05,
1369
+ "loss": 3.2671,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 0.34,
1374
+ "learning_rate": 4.4291191308188536e-05,
1375
+ "loss": 3.2873,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 0.34,
1380
+ "learning_rate": 4.4266152673575326e-05,
1381
+ "loss": 3.2857,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 0.35,
1386
+ "learning_rate": 4.4241114038962124e-05,
1387
+ "loss": 3.2992,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 0.35,
1392
+ "learning_rate": 4.4216075404348914e-05,
1393
+ "loss": 3.2959,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 0.35,
1398
+ "learning_rate": 4.4191036769735705e-05,
1399
+ "loss": 3.2688,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 0.35,
1404
+ "learning_rate": 4.4165998135122495e-05,
1405
+ "loss": 3.2958,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 0.35,
1410
+ "learning_rate": 4.4140959500509286e-05,
1411
+ "loss": 3.2933,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 0.35,
1416
+ "learning_rate": 4.411592086589608e-05,
1417
+ "loss": 3.2796,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 0.35,
1422
+ "learning_rate": 4.4090882231282874e-05,
1423
+ "loss": 3.2577,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 0.36,
1428
+ "learning_rate": 4.4065843596669664e-05,
1429
+ "loss": 3.2797,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 0.36,
1434
+ "learning_rate": 4.4040804962056455e-05,
1435
+ "loss": 3.3004,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 0.36,
1440
+ "learning_rate": 4.4015766327443246e-05,
1441
+ "loss": 3.2539,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 0.36,
1446
+ "learning_rate": 4.399072769283004e-05,
1447
+ "loss": 3.2525,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 0.36,
1452
+ "learning_rate": 4.3965689058216833e-05,
1453
+ "loss": 3.2668,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 0.36,
1458
+ "learning_rate": 4.3940650423603624e-05,
1459
+ "loss": 3.2693,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 0.37,
1464
+ "learning_rate": 4.3915611788990415e-05,
1465
+ "loss": 3.2792,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 0.37,
1470
+ "learning_rate": 4.3890573154377205e-05,
1471
+ "loss": 3.2649,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 0.37,
1476
+ "learning_rate": 4.3865534519763996e-05,
1477
+ "loss": 3.2556,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 0.37,
1482
+ "learning_rate": 4.3840495885150786e-05,
1483
+ "loss": 3.2509,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 0.37,
1488
+ "learning_rate": 4.3815457250537584e-05,
1489
+ "loss": 3.2616,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 0.37,
1494
+ "learning_rate": 4.3790418615924374e-05,
1495
+ "loss": 3.2654,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 0.37,
1500
+ "learning_rate": 4.3765379981311165e-05,
1501
+ "loss": 3.2628,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 0.38,
1506
+ "learning_rate": 4.3740341346697955e-05,
1507
+ "loss": 3.2462,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 0.38,
1512
+ "learning_rate": 4.3715302712084746e-05,
1513
+ "loss": 3.2517,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 0.38,
1518
+ "learning_rate": 4.369026407747154e-05,
1519
+ "loss": 3.2897,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 0.38,
1524
+ "learning_rate": 4.3665225442858334e-05,
1525
+ "loss": 3.2707,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 0.38,
1530
+ "learning_rate": 4.3640186808245124e-05,
1531
+ "loss": 3.269,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 0.38,
1536
+ "learning_rate": 4.3615148173631915e-05,
1537
+ "loss": 3.2331,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 0.38,
1542
+ "learning_rate": 4.3590109539018705e-05,
1543
+ "loss": 3.2644,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 0.39,
1548
+ "learning_rate": 4.3565070904405496e-05,
1549
+ "loss": 3.2406,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 0.39,
1554
+ "learning_rate": 4.354003226979229e-05,
1555
+ "loss": 3.2499,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 0.39,
1560
+ "learning_rate": 4.3514993635179084e-05,
1561
+ "loss": 3.2325,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 0.39,
1566
+ "learning_rate": 4.3489955000565875e-05,
1567
+ "loss": 3.2631,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 0.39,
1572
+ "learning_rate": 4.3464916365952665e-05,
1573
+ "loss": 3.2447,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 0.39,
1578
+ "learning_rate": 4.3439877731339456e-05,
1579
+ "loss": 3.2349,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 0.4,
1584
+ "learning_rate": 4.341483909672625e-05,
1585
+ "loss": 3.2372,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 0.4,
1590
+ "learning_rate": 4.3389800462113044e-05,
1591
+ "loss": 3.2456,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 0.4,
1596
+ "learning_rate": 4.3364761827499834e-05,
1597
+ "loss": 3.2477,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 0.4,
1602
+ "learning_rate": 4.3339723192886625e-05,
1603
+ "loss": 3.243,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 0.4,
1608
+ "learning_rate": 4.3314684558273415e-05,
1609
+ "loss": 3.2287,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 0.4,
1614
+ "learning_rate": 4.328964592366021e-05,
1615
+ "loss": 3.2613,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 0.4,
1620
+ "learning_rate": 4.3264607289046996e-05,
1621
+ "loss": 3.2242,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 0.41,
1626
+ "learning_rate": 4.3239568654433794e-05,
1627
+ "loss": 3.2481,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 0.41,
1632
+ "learning_rate": 4.3214530019820584e-05,
1633
+ "loss": 3.2482,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 0.41,
1638
+ "learning_rate": 4.3189491385207375e-05,
1639
+ "loss": 3.2439,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 0.41,
1644
+ "learning_rate": 4.316445275059417e-05,
1645
+ "loss": 3.2321,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 0.41,
1650
+ "learning_rate": 4.3139414115980956e-05,
1651
+ "loss": 3.2422,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 0.41,
1656
+ "learning_rate": 4.311437548136775e-05,
1657
+ "loss": 3.249,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 0.41,
1662
+ "learning_rate": 4.3089336846754544e-05,
1663
+ "loss": 3.2349,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 0.42,
1668
+ "learning_rate": 4.3064298212141334e-05,
1669
+ "loss": 3.218,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 0.42,
1674
+ "learning_rate": 4.303925957752813e-05,
1675
+ "loss": 3.2232,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 0.42,
1680
+ "learning_rate": 4.3014220942914916e-05,
1681
+ "loss": 3.2183,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 0.42,
1686
+ "learning_rate": 4.298918230830171e-05,
1687
+ "loss": 3.2547,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 0.42,
1692
+ "learning_rate": 4.2964143673688503e-05,
1693
+ "loss": 3.2168,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 0.42,
1698
+ "learning_rate": 4.2939105039075294e-05,
1699
+ "loss": 3.2378,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 0.43,
1704
+ "learning_rate": 4.291406640446209e-05,
1705
+ "loss": 3.2035,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 0.43,
1710
+ "learning_rate": 4.2889027769848875e-05,
1711
+ "loss": 3.2112,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 0.43,
1716
+ "learning_rate": 4.286398913523567e-05,
1717
+ "loss": 3.1962,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 0.43,
1722
+ "learning_rate": 4.283895050062246e-05,
1723
+ "loss": 3.2278,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 0.43,
1728
+ "learning_rate": 4.2813911866009254e-05,
1729
+ "loss": 3.2453,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 0.43,
1734
+ "learning_rate": 4.278887323139605e-05,
1735
+ "loss": 3.2265,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 0.43,
1740
+ "learning_rate": 4.2763834596782835e-05,
1741
+ "loss": 3.1941,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 0.44,
1746
+ "learning_rate": 4.273879596216963e-05,
1747
+ "loss": 3.2519,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 0.44,
1752
+ "learning_rate": 4.271375732755642e-05,
1753
+ "loss": 3.1941,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 0.44,
1758
+ "learning_rate": 4.268871869294321e-05,
1759
+ "loss": 3.2421,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 0.44,
1764
+ "learning_rate": 4.266368005833001e-05,
1765
+ "loss": 3.2396,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 0.44,
1770
+ "learning_rate": 4.2638641423716794e-05,
1771
+ "loss": 3.2021,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 0.44,
1776
+ "learning_rate": 4.261360278910359e-05,
1777
+ "loss": 3.2229,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 0.44,
1782
+ "learning_rate": 4.258856415449038e-05,
1783
+ "loss": 3.2251,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 0.45,
1788
+ "learning_rate": 4.256352551987717e-05,
1789
+ "loss": 3.1941,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 0.45,
1794
+ "learning_rate": 4.2538486885263963e-05,
1795
+ "loss": 3.2002,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 0.45,
1800
+ "learning_rate": 4.2513448250650754e-05,
1801
+ "loss": 3.2177,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 0.45,
1806
+ "learning_rate": 4.248840961603755e-05,
1807
+ "loss": 3.2386,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 0.45,
1812
+ "learning_rate": 4.246337098142434e-05,
1813
+ "loss": 3.1961,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 0.45,
1818
+ "learning_rate": 4.243833234681113e-05,
1819
+ "loss": 3.1943,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 0.46,
1824
+ "learning_rate": 4.241329371219792e-05,
1825
+ "loss": 3.1818,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 0.46,
1830
+ "learning_rate": 4.2388255077584714e-05,
1831
+ "loss": 3.2139,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 0.46,
1836
+ "learning_rate": 4.236321644297151e-05,
1837
+ "loss": 3.1935,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 0.46,
1842
+ "learning_rate": 4.2338177808358295e-05,
1843
+ "loss": 3.206,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 0.46,
1848
+ "learning_rate": 4.231313917374509e-05,
1849
+ "loss": 3.2269,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 0.46,
1854
+ "learning_rate": 4.228810053913188e-05,
1855
+ "loss": 3.2269,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 0.46,
1860
+ "learning_rate": 4.226306190451867e-05,
1861
+ "loss": 3.1957,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 0.47,
1866
+ "learning_rate": 4.2238023269905464e-05,
1867
+ "loss": 3.2306,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 0.47,
1872
+ "learning_rate": 4.2212984635292254e-05,
1873
+ "loss": 3.224,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 0.47,
1878
+ "learning_rate": 4.218794600067905e-05,
1879
+ "loss": 3.2183,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 0.47,
1884
+ "learning_rate": 4.216290736606584e-05,
1885
+ "loss": 3.2007,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 0.47,
1890
+ "learning_rate": 4.213786873145263e-05,
1891
+ "loss": 3.1717,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 0.47,
1896
+ "learning_rate": 4.211283009683942e-05,
1897
+ "loss": 3.1949,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 0.47,
1902
+ "learning_rate": 4.2087791462226214e-05,
1903
+ "loss": 3.1852,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 0.48,
1908
+ "learning_rate": 4.206275282761301e-05,
1909
+ "loss": 3.1967,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 0.48,
1914
+ "learning_rate": 4.20377141929998e-05,
1915
+ "loss": 3.1857,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 0.48,
1920
+ "learning_rate": 4.201267555838659e-05,
1921
+ "loss": 3.184,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 0.48,
1926
+ "learning_rate": 4.198763692377338e-05,
1927
+ "loss": 3.1921,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 0.48,
1932
+ "learning_rate": 4.1962598289160173e-05,
1933
+ "loss": 3.1994,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 0.48,
1938
+ "learning_rate": 4.193755965454697e-05,
1939
+ "loss": 3.1853,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 0.49,
1944
+ "learning_rate": 4.191252101993376e-05,
1945
+ "loss": 3.207,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 0.49,
1950
+ "learning_rate": 4.188748238532055e-05,
1951
+ "loss": 3.1822,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 0.49,
1956
+ "learning_rate": 4.186244375070734e-05,
1957
+ "loss": 3.1786,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 0.49,
1962
+ "learning_rate": 4.183740511609413e-05,
1963
+ "loss": 3.1686,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 0.49,
1968
+ "learning_rate": 4.1812366481480924e-05,
1969
+ "loss": 3.2091,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 0.49,
1974
+ "learning_rate": 4.178732784686772e-05,
1975
+ "loss": 3.1702,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 0.49,
1980
+ "learning_rate": 4.176228921225451e-05,
1981
+ "loss": 3.1915,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 0.5,
1986
+ "learning_rate": 4.17372505776413e-05,
1987
+ "loss": 3.1912,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 0.5,
1992
+ "learning_rate": 4.171221194302809e-05,
1993
+ "loss": 3.2141,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 0.5,
1998
+ "learning_rate": 4.168717330841488e-05,
1999
+ "loss": 3.1828,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 0.5,
2004
+ "learning_rate": 4.166213467380168e-05,
2005
+ "loss": 3.1877,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 0.5,
2010
+ "learning_rate": 4.163709603918847e-05,
2011
+ "loss": 3.1859,
2012
+ "step": 167000
2013
+ },
2014
+ {
2015
+ "epoch": 0.5,
2016
+ "learning_rate": 4.161205740457526e-05,
2017
+ "loss": 3.1695,
2018
+ "step": 167500
2019
+ },
2020
+ {
2021
+ "epoch": 0.5,
2022
+ "learning_rate": 4.158701876996205e-05,
2023
+ "loss": 3.227,
2024
+ "step": 168000
2025
+ },
2026
+ {
2027
+ "epoch": 0.51,
2028
+ "learning_rate": 4.156198013534884e-05,
2029
+ "loss": 3.1862,
2030
+ "step": 168500
2031
+ },
2032
+ {
2033
+ "epoch": 0.51,
2034
+ "learning_rate": 4.153694150073564e-05,
2035
+ "loss": 3.2209,
2036
+ "step": 169000
2037
+ },
2038
+ {
2039
+ "epoch": 0.51,
2040
+ "learning_rate": 4.1511902866122424e-05,
2041
+ "loss": 3.1829,
2042
+ "step": 169500
2043
+ },
2044
+ {
2045
+ "epoch": 0.51,
2046
+ "learning_rate": 4.148686423150922e-05,
2047
+ "loss": 3.1981,
2048
+ "step": 170000
2049
+ },
2050
+ {
2051
+ "epoch": 0.51,
2052
+ "learning_rate": 4.146182559689601e-05,
2053
+ "loss": 3.1563,
2054
+ "step": 170500
2055
+ },
2056
+ {
2057
+ "epoch": 0.51,
2058
+ "learning_rate": 4.14367869622828e-05,
2059
+ "loss": 3.1743,
2060
+ "step": 171000
2061
+ },
2062
+ {
2063
+ "epoch": 0.52,
2064
+ "learning_rate": 4.14117483276696e-05,
2065
+ "loss": 3.1754,
2066
+ "step": 171500
2067
+ },
2068
+ {
2069
+ "epoch": 0.52,
2070
+ "learning_rate": 4.1386709693056384e-05,
2071
+ "loss": 3.1729,
2072
+ "step": 172000
2073
+ },
2074
+ {
2075
+ "epoch": 0.52,
2076
+ "learning_rate": 4.136167105844318e-05,
2077
+ "loss": 3.1476,
2078
+ "step": 172500
2079
+ },
2080
+ {
2081
+ "epoch": 0.52,
2082
+ "learning_rate": 4.133663242382997e-05,
2083
+ "loss": 3.1561,
2084
+ "step": 173000
2085
+ },
2086
+ {
2087
+ "epoch": 0.52,
2088
+ "learning_rate": 4.131159378921676e-05,
2089
+ "loss": 3.1844,
2090
+ "step": 173500
2091
+ },
2092
+ {
2093
+ "epoch": 0.52,
2094
+ "learning_rate": 4.128655515460356e-05,
2095
+ "loss": 3.1756,
2096
+ "step": 174000
2097
+ },
2098
+ {
2099
+ "epoch": 0.52,
2100
+ "learning_rate": 4.126151651999034e-05,
2101
+ "loss": 3.181,
2102
+ "step": 174500
2103
+ },
2104
+ {
2105
+ "epoch": 0.53,
2106
+ "learning_rate": 4.123647788537714e-05,
2107
+ "loss": 3.1479,
2108
+ "step": 175000
2109
+ },
2110
+ {
2111
+ "epoch": 0.53,
2112
+ "learning_rate": 4.121143925076393e-05,
2113
+ "loss": 3.1654,
2114
+ "step": 175500
2115
+ },
2116
+ {
2117
+ "epoch": 0.53,
2118
+ "learning_rate": 4.118640061615072e-05,
2119
+ "loss": 3.1903,
2120
+ "step": 176000
2121
+ },
2122
+ {
2123
+ "epoch": 0.53,
2124
+ "learning_rate": 4.116136198153752e-05,
2125
+ "loss": 3.1761,
2126
+ "step": 176500
2127
+ },
2128
+ {
2129
+ "epoch": 0.53,
2130
+ "learning_rate": 4.11363233469243e-05,
2131
+ "loss": 3.1882,
2132
+ "step": 177000
2133
+ },
2134
+ {
2135
+ "epoch": 0.53,
2136
+ "learning_rate": 4.11112847123111e-05,
2137
+ "loss": 3.175,
2138
+ "step": 177500
2139
+ },
2140
+ {
2141
+ "epoch": 0.53,
2142
+ "learning_rate": 4.108624607769789e-05,
2143
+ "loss": 3.1517,
2144
+ "step": 178000
2145
+ },
2146
+ {
2147
+ "epoch": 0.54,
2148
+ "learning_rate": 4.106120744308468e-05,
2149
+ "loss": 3.1663,
2150
+ "step": 178500
2151
+ },
2152
+ {
2153
+ "epoch": 0.54,
2154
+ "learning_rate": 4.103616880847148e-05,
2155
+ "loss": 3.1669,
2156
+ "step": 179000
2157
+ },
2158
+ {
2159
+ "epoch": 0.54,
2160
+ "learning_rate": 4.101113017385826e-05,
2161
+ "loss": 3.1764,
2162
+ "step": 179500
2163
+ },
2164
+ {
2165
+ "epoch": 0.54,
2166
+ "learning_rate": 4.098609153924506e-05,
2167
+ "loss": 3.1554,
2168
+ "step": 180000
2169
+ },
2170
+ {
2171
+ "epoch": 0.54,
2172
+ "learning_rate": 4.096105290463185e-05,
2173
+ "loss": 3.1488,
2174
+ "step": 180500
2175
+ },
2176
+ {
2177
+ "epoch": 0.54,
2178
+ "learning_rate": 4.093601427001864e-05,
2179
+ "loss": 3.1657,
2180
+ "step": 181000
2181
+ },
2182
+ {
2183
+ "epoch": 0.55,
2184
+ "learning_rate": 4.091097563540544e-05,
2185
+ "loss": 3.1415,
2186
+ "step": 181500
2187
+ },
2188
+ {
2189
+ "epoch": 0.55,
2190
+ "learning_rate": 4.088593700079222e-05,
2191
+ "loss": 3.1519,
2192
+ "step": 182000
2193
+ },
2194
+ {
2195
+ "epoch": 0.55,
2196
+ "learning_rate": 4.086089836617902e-05,
2197
+ "loss": 3.1763,
2198
+ "step": 182500
2199
+ },
2200
+ {
2201
+ "epoch": 0.55,
2202
+ "learning_rate": 4.08358597315658e-05,
2203
+ "loss": 3.1373,
2204
+ "step": 183000
2205
+ },
2206
+ {
2207
+ "epoch": 0.55,
2208
+ "learning_rate": 4.08108210969526e-05,
2209
+ "loss": 3.1619,
2210
+ "step": 183500
2211
+ },
2212
+ {
2213
+ "epoch": 0.55,
2214
+ "learning_rate": 4.078578246233939e-05,
2215
+ "loss": 3.1682,
2216
+ "step": 184000
2217
+ },
2218
+ {
2219
+ "epoch": 0.55,
2220
+ "learning_rate": 4.076074382772618e-05,
2221
+ "loss": 3.1474,
2222
+ "step": 184500
2223
+ },
2224
+ {
2225
+ "epoch": 0.56,
2226
+ "learning_rate": 4.073570519311298e-05,
2227
+ "loss": 3.2058,
2228
+ "step": 185000
2229
+ },
2230
+ {
2231
+ "epoch": 0.56,
2232
+ "learning_rate": 4.071066655849976e-05,
2233
+ "loss": 3.1558,
2234
+ "step": 185500
2235
+ },
2236
+ {
2237
+ "epoch": 0.56,
2238
+ "learning_rate": 4.068562792388656e-05,
2239
+ "loss": 3.1667,
2240
+ "step": 186000
2241
+ },
2242
+ {
2243
+ "epoch": 0.56,
2244
+ "learning_rate": 4.066058928927335e-05,
2245
+ "loss": 3.1724,
2246
+ "step": 186500
2247
+ },
2248
+ {
2249
+ "epoch": 0.56,
2250
+ "learning_rate": 4.063555065466014e-05,
2251
+ "loss": 3.1421,
2252
+ "step": 187000
2253
+ },
2254
+ {
2255
+ "epoch": 0.56,
2256
+ "learning_rate": 4.061051202004694e-05,
2257
+ "loss": 3.1602,
2258
+ "step": 187500
2259
+ },
2260
+ {
2261
+ "epoch": 0.56,
2262
+ "learning_rate": 4.058547338543372e-05,
2263
+ "loss": 3.1505,
2264
+ "step": 188000
2265
+ },
2266
+ {
2267
+ "epoch": 0.57,
2268
+ "learning_rate": 4.056043475082052e-05,
2269
+ "loss": 3.1426,
2270
+ "step": 188500
2271
+ },
2272
+ {
2273
+ "epoch": 0.57,
2274
+ "learning_rate": 4.053539611620731e-05,
2275
+ "loss": 3.1507,
2276
+ "step": 189000
2277
+ },
2278
+ {
2279
+ "epoch": 0.57,
2280
+ "learning_rate": 4.05103574815941e-05,
2281
+ "loss": 3.1579,
2282
+ "step": 189500
2283
+ },
2284
+ {
2285
+ "epoch": 0.57,
2286
+ "learning_rate": 4.048531884698089e-05,
2287
+ "loss": 3.1455,
2288
+ "step": 190000
2289
+ },
2290
+ {
2291
+ "epoch": 0.57,
2292
+ "learning_rate": 4.046028021236768e-05,
2293
+ "loss": 3.1776,
2294
+ "step": 190500
2295
+ },
2296
+ {
2297
+ "epoch": 0.57,
2298
+ "learning_rate": 4.043524157775448e-05,
2299
+ "loss": 3.1343,
2300
+ "step": 191000
2301
+ },
2302
+ {
2303
+ "epoch": 0.58,
2304
+ "learning_rate": 4.041020294314127e-05,
2305
+ "loss": 3.1487,
2306
+ "step": 191500
2307
+ },
2308
+ {
2309
+ "epoch": 0.58,
2310
+ "learning_rate": 4.038516430852806e-05,
2311
+ "loss": 3.15,
2312
+ "step": 192000
2313
+ },
2314
+ {
2315
+ "epoch": 0.58,
2316
+ "learning_rate": 4.036012567391485e-05,
2317
+ "loss": 3.1351,
2318
+ "step": 192500
2319
+ },
2320
+ {
2321
+ "epoch": 0.58,
2322
+ "learning_rate": 4.033508703930164e-05,
2323
+ "loss": 3.1242,
2324
+ "step": 193000
2325
+ },
2326
+ {
2327
+ "epoch": 0.58,
2328
+ "learning_rate": 4.031004840468844e-05,
2329
+ "loss": 3.1655,
2330
+ "step": 193500
2331
+ },
2332
+ {
2333
+ "epoch": 0.58,
2334
+ "learning_rate": 4.028500977007523e-05,
2335
+ "loss": 3.1386,
2336
+ "step": 194000
2337
+ },
2338
+ {
2339
+ "epoch": 0.58,
2340
+ "learning_rate": 4.025997113546202e-05,
2341
+ "loss": 3.1384,
2342
+ "step": 194500
2343
+ },
2344
+ {
2345
+ "epoch": 0.59,
2346
+ "learning_rate": 4.023493250084881e-05,
2347
+ "loss": 3.1433,
2348
+ "step": 195000
2349
+ },
2350
+ {
2351
+ "epoch": 0.59,
2352
+ "learning_rate": 4.02098938662356e-05,
2353
+ "loss": 3.1139,
2354
+ "step": 195500
2355
+ },
2356
+ {
2357
+ "epoch": 0.59,
2358
+ "learning_rate": 4.018485523162239e-05,
2359
+ "loss": 3.1579,
2360
+ "step": 196000
2361
+ },
2362
+ {
2363
+ "epoch": 0.59,
2364
+ "learning_rate": 4.015981659700919e-05,
2365
+ "loss": 3.1283,
2366
+ "step": 196500
2367
+ },
2368
+ {
2369
+ "epoch": 0.59,
2370
+ "learning_rate": 4.013477796239598e-05,
2371
+ "loss": 3.138,
2372
+ "step": 197000
2373
+ },
2374
+ {
2375
+ "epoch": 0.59,
2376
+ "learning_rate": 4.010973932778277e-05,
2377
+ "loss": 3.1325,
2378
+ "step": 197500
2379
+ },
2380
+ {
2381
+ "epoch": 0.59,
2382
+ "learning_rate": 4.008470069316956e-05,
2383
+ "loss": 3.1043,
2384
+ "step": 198000
2385
+ },
2386
+ {
2387
+ "epoch": 0.6,
2388
+ "learning_rate": 4.005966205855635e-05,
2389
+ "loss": 3.1572,
2390
+ "step": 198500
2391
+ },
2392
+ {
2393
+ "epoch": 0.6,
2394
+ "learning_rate": 4.003462342394315e-05,
2395
+ "loss": 3.1539,
2396
+ "step": 199000
2397
+ },
2398
+ {
2399
+ "epoch": 0.6,
2400
+ "learning_rate": 4.000958478932994e-05,
2401
+ "loss": 3.1408,
2402
+ "step": 199500
2403
+ },
2404
+ {
2405
+ "epoch": 0.6,
2406
+ "learning_rate": 3.998454615471673e-05,
2407
+ "loss": 3.1204,
2408
+ "step": 200000
2409
+ },
2410
+ {
2411
+ "epoch": 0.6,
2412
+ "learning_rate": 3.995950752010352e-05,
2413
+ "loss": 3.1164,
2414
+ "step": 200500
2415
+ },
2416
+ {
2417
+ "epoch": 0.6,
2418
+ "learning_rate": 3.993446888549031e-05,
2419
+ "loss": 3.1518,
2420
+ "step": 201000
2421
+ },
2422
+ {
2423
+ "epoch": 0.61,
2424
+ "learning_rate": 3.990943025087711e-05,
2425
+ "loss": 3.1399,
2426
+ "step": 201500
2427
+ },
2428
+ {
2429
+ "epoch": 0.61,
2430
+ "learning_rate": 3.98843916162639e-05,
2431
+ "loss": 3.116,
2432
+ "step": 202000
2433
+ },
2434
+ {
2435
+ "epoch": 0.61,
2436
+ "learning_rate": 3.985935298165069e-05,
2437
+ "loss": 3.1405,
2438
+ "step": 202500
2439
+ },
2440
+ {
2441
+ "epoch": 0.61,
2442
+ "learning_rate": 3.983431434703748e-05,
2443
+ "loss": 3.116,
2444
+ "step": 203000
2445
+ },
2446
+ {
2447
+ "epoch": 0.61,
2448
+ "learning_rate": 3.980927571242427e-05,
2449
+ "loss": 3.1357,
2450
+ "step": 203500
2451
+ },
2452
+ {
2453
+ "epoch": 0.61,
2454
+ "learning_rate": 3.978423707781107e-05,
2455
+ "loss": 3.149,
2456
+ "step": 204000
2457
+ },
2458
+ {
2459
+ "epoch": 0.61,
2460
+ "learning_rate": 3.975919844319785e-05,
2461
+ "loss": 3.1418,
2462
+ "step": 204500
2463
+ },
2464
+ {
2465
+ "epoch": 0.62,
2466
+ "learning_rate": 3.973415980858465e-05,
2467
+ "loss": 3.1229,
2468
+ "step": 205000
2469
+ },
2470
+ {
2471
+ "epoch": 0.62,
2472
+ "learning_rate": 3.970912117397144e-05,
2473
+ "loss": 3.141,
2474
+ "step": 205500
2475
+ },
2476
+ {
2477
+ "epoch": 0.62,
2478
+ "learning_rate": 3.968408253935823e-05,
2479
+ "loss": 3.1376,
2480
+ "step": 206000
2481
+ },
2482
+ {
2483
+ "epoch": 0.62,
2484
+ "learning_rate": 3.965904390474503e-05,
2485
+ "loss": 3.1284,
2486
+ "step": 206500
2487
+ },
2488
+ {
2489
+ "epoch": 0.62,
2490
+ "learning_rate": 3.963400527013181e-05,
2491
+ "loss": 3.1405,
2492
+ "step": 207000
2493
+ },
2494
+ {
2495
+ "epoch": 0.62,
2496
+ "learning_rate": 3.960896663551861e-05,
2497
+ "loss": 3.1543,
2498
+ "step": 207500
2499
+ },
2500
+ {
2501
+ "epoch": 0.62,
2502
+ "learning_rate": 3.95839280009054e-05,
2503
+ "loss": 3.1693,
2504
+ "step": 208000
2505
+ },
2506
+ {
2507
+ "epoch": 0.63,
2508
+ "learning_rate": 3.955888936629219e-05,
2509
+ "loss": 3.116,
2510
+ "step": 208500
2511
+ },
2512
+ {
2513
+ "epoch": 0.63,
2514
+ "learning_rate": 3.953385073167899e-05,
2515
+ "loss": 3.125,
2516
+ "step": 209000
2517
+ },
2518
+ {
2519
+ "epoch": 0.63,
2520
+ "learning_rate": 3.950881209706577e-05,
2521
+ "loss": 3.1362,
2522
+ "step": 209500
2523
+ },
2524
+ {
2525
+ "epoch": 0.63,
2526
+ "learning_rate": 3.948377346245257e-05,
2527
+ "loss": 3.0917,
2528
+ "step": 210000
2529
+ },
2530
+ {
2531
+ "epoch": 0.63,
2532
+ "learning_rate": 3.945873482783936e-05,
2533
+ "loss": 3.121,
2534
+ "step": 210500
2535
+ },
2536
+ {
2537
+ "epoch": 0.63,
2538
+ "learning_rate": 3.943369619322615e-05,
2539
+ "loss": 3.0947,
2540
+ "step": 211000
2541
+ },
2542
+ {
2543
+ "epoch": 0.64,
2544
+ "learning_rate": 3.940865755861295e-05,
2545
+ "loss": 3.1065,
2546
+ "step": 211500
2547
+ },
2548
+ {
2549
+ "epoch": 0.64,
2550
+ "learning_rate": 3.938361892399973e-05,
2551
+ "loss": 3.1274,
2552
+ "step": 212000
2553
+ },
2554
+ {
2555
+ "epoch": 0.64,
2556
+ "learning_rate": 3.935858028938653e-05,
2557
+ "loss": 3.1071,
2558
+ "step": 212500
2559
+ },
2560
+ {
2561
+ "epoch": 0.64,
2562
+ "learning_rate": 3.933354165477331e-05,
2563
+ "loss": 3.1405,
2564
+ "step": 213000
2565
+ },
2566
+ {
2567
+ "epoch": 0.64,
2568
+ "learning_rate": 3.930850302016011e-05,
2569
+ "loss": 3.1544,
2570
+ "step": 213500
2571
+ },
2572
+ {
2573
+ "epoch": 0.64,
2574
+ "learning_rate": 3.9283464385546906e-05,
2575
+ "loss": 3.1427,
2576
+ "step": 214000
2577
+ },
2578
+ {
2579
+ "epoch": 0.64,
2580
+ "learning_rate": 3.925842575093369e-05,
2581
+ "loss": 3.1102,
2582
+ "step": 214500
2583
+ },
2584
+ {
2585
+ "epoch": 0.65,
2586
+ "learning_rate": 3.923338711632049e-05,
2587
+ "loss": 3.1475,
2588
+ "step": 215000
2589
+ },
2590
+ {
2591
+ "epoch": 0.65,
2592
+ "learning_rate": 3.920834848170727e-05,
2593
+ "loss": 3.125,
2594
+ "step": 215500
2595
+ },
2596
+ {
2597
+ "epoch": 0.65,
2598
+ "learning_rate": 3.918330984709407e-05,
2599
+ "loss": 3.1,
2600
+ "step": 216000
2601
+ },
2602
+ {
2603
+ "epoch": 0.65,
2604
+ "learning_rate": 3.9158271212480866e-05,
2605
+ "loss": 3.1193,
2606
+ "step": 216500
2607
+ },
2608
+ {
2609
+ "epoch": 0.65,
2610
+ "learning_rate": 3.913323257786765e-05,
2611
+ "loss": 3.1084,
2612
+ "step": 217000
2613
+ },
2614
+ {
2615
+ "epoch": 0.65,
2616
+ "learning_rate": 3.910819394325445e-05,
2617
+ "loss": 3.1027,
2618
+ "step": 217500
2619
+ },
2620
+ {
2621
+ "epoch": 0.66,
2622
+ "learning_rate": 3.908315530864123e-05,
2623
+ "loss": 3.1362,
2624
+ "step": 218000
2625
+ },
2626
+ {
2627
+ "epoch": 0.66,
2628
+ "learning_rate": 3.905811667402803e-05,
2629
+ "loss": 3.1049,
2630
+ "step": 218500
2631
+ },
2632
+ {
2633
+ "epoch": 0.66,
2634
+ "learning_rate": 3.903307803941482e-05,
2635
+ "loss": 3.1278,
2636
+ "step": 219000
2637
+ },
2638
+ {
2639
+ "epoch": 0.66,
2640
+ "learning_rate": 3.900803940480161e-05,
2641
+ "loss": 3.1058,
2642
+ "step": 219500
2643
+ },
2644
+ {
2645
+ "epoch": 0.66,
2646
+ "learning_rate": 3.8983000770188407e-05,
2647
+ "loss": 3.0892,
2648
+ "step": 220000
2649
+ },
2650
+ {
2651
+ "epoch": 0.66,
2652
+ "learning_rate": 3.895796213557519e-05,
2653
+ "loss": 3.1202,
2654
+ "step": 220500
2655
+ },
2656
+ {
2657
+ "epoch": 0.66,
2658
+ "learning_rate": 3.893292350096199e-05,
2659
+ "loss": 3.1286,
2660
+ "step": 221000
2661
+ },
2662
+ {
2663
+ "epoch": 0.67,
2664
+ "learning_rate": 3.890788486634878e-05,
2665
+ "loss": 3.1254,
2666
+ "step": 221500
2667
+ },
2668
+ {
2669
+ "epoch": 0.67,
2670
+ "learning_rate": 3.888284623173557e-05,
2671
+ "loss": 3.1017,
2672
+ "step": 222000
2673
+ },
2674
+ {
2675
+ "epoch": 0.67,
2676
+ "learning_rate": 3.8857807597122366e-05,
2677
+ "loss": 3.0974,
2678
+ "step": 222500
2679
+ },
2680
+ {
2681
+ "epoch": 0.67,
2682
+ "learning_rate": 3.883276896250915e-05,
2683
+ "loss": 3.0832,
2684
+ "step": 223000
2685
+ },
2686
+ {
2687
+ "epoch": 0.67,
2688
+ "learning_rate": 3.880773032789595e-05,
2689
+ "loss": 3.1195,
2690
+ "step": 223500
2691
+ },
2692
+ {
2693
+ "epoch": 0.67,
2694
+ "learning_rate": 3.878269169328274e-05,
2695
+ "loss": 3.1234,
2696
+ "step": 224000
2697
+ },
2698
+ {
2699
+ "epoch": 0.67,
2700
+ "learning_rate": 3.875765305866953e-05,
2701
+ "loss": 3.0852,
2702
+ "step": 224500
2703
+ },
2704
+ {
2705
+ "epoch": 0.68,
2706
+ "learning_rate": 3.873261442405632e-05,
2707
+ "loss": 3.1366,
2708
+ "step": 225000
2709
+ },
2710
+ {
2711
+ "epoch": 0.68,
2712
+ "learning_rate": 3.870757578944311e-05,
2713
+ "loss": 3.1166,
2714
+ "step": 225500
2715
+ },
2716
+ {
2717
+ "epoch": 0.68,
2718
+ "learning_rate": 3.868253715482991e-05,
2719
+ "loss": 3.1384,
2720
+ "step": 226000
2721
+ },
2722
+ {
2723
+ "epoch": 0.68,
2724
+ "learning_rate": 3.86574985202167e-05,
2725
+ "loss": 3.1121,
2726
+ "step": 226500
2727
+ },
2728
+ {
2729
+ "epoch": 0.68,
2730
+ "learning_rate": 3.863245988560349e-05,
2731
+ "loss": 3.1149,
2732
+ "step": 227000
2733
+ },
2734
+ {
2735
+ "epoch": 0.68,
2736
+ "learning_rate": 3.860742125099028e-05,
2737
+ "loss": 3.1072,
2738
+ "step": 227500
2739
+ },
2740
+ {
2741
+ "epoch": 0.69,
2742
+ "learning_rate": 3.858238261637707e-05,
2743
+ "loss": 3.0932,
2744
+ "step": 228000
2745
+ },
2746
+ {
2747
+ "epoch": 0.69,
2748
+ "learning_rate": 3.8557343981763866e-05,
2749
+ "loss": 3.0846,
2750
+ "step": 228500
2751
+ },
2752
+ {
2753
+ "epoch": 0.69,
2754
+ "learning_rate": 3.853230534715066e-05,
2755
+ "loss": 3.0819,
2756
+ "step": 229000
2757
+ },
2758
+ {
2759
+ "epoch": 0.69,
2760
+ "learning_rate": 3.850726671253745e-05,
2761
+ "loss": 3.104,
2762
+ "step": 229500
2763
+ },
2764
+ {
2765
+ "epoch": 0.69,
2766
+ "learning_rate": 3.848222807792424e-05,
2767
+ "loss": 3.1016,
2768
+ "step": 230000
2769
+ },
2770
+ {
2771
+ "epoch": 0.69,
2772
+ "learning_rate": 3.845718944331103e-05,
2773
+ "loss": 3.075,
2774
+ "step": 230500
2775
+ },
2776
+ {
2777
+ "epoch": 0.69,
2778
+ "learning_rate": 3.843215080869782e-05,
2779
+ "loss": 3.0945,
2780
+ "step": 231000
2781
+ },
2782
+ {
2783
+ "epoch": 0.7,
2784
+ "learning_rate": 3.840711217408462e-05,
2785
+ "loss": 3.0949,
2786
+ "step": 231500
2787
+ },
2788
+ {
2789
+ "epoch": 0.7,
2790
+ "learning_rate": 3.838207353947141e-05,
2791
+ "loss": 3.1097,
2792
+ "step": 232000
2793
+ },
2794
+ {
2795
+ "epoch": 0.7,
2796
+ "learning_rate": 3.83570349048582e-05,
2797
+ "loss": 3.0881,
2798
+ "step": 232500
2799
+ },
2800
+ {
2801
+ "epoch": 0.7,
2802
+ "learning_rate": 3.833199627024499e-05,
2803
+ "loss": 3.1087,
2804
+ "step": 233000
2805
+ },
2806
+ {
2807
+ "epoch": 0.7,
2808
+ "learning_rate": 3.830695763563178e-05,
2809
+ "loss": 3.1164,
2810
+ "step": 233500
2811
+ },
2812
+ {
2813
+ "epoch": 0.7,
2814
+ "learning_rate": 3.8281919001018576e-05,
2815
+ "loss": 3.1171,
2816
+ "step": 234000
2817
+ },
2818
+ {
2819
+ "epoch": 0.7,
2820
+ "learning_rate": 3.825688036640537e-05,
2821
+ "loss": 3.0979,
2822
+ "step": 234500
2823
+ },
2824
+ {
2825
+ "epoch": 0.71,
2826
+ "learning_rate": 3.823184173179216e-05,
2827
+ "loss": 3.0912,
2828
+ "step": 235000
2829
+ },
2830
+ {
2831
+ "epoch": 0.71,
2832
+ "learning_rate": 3.820680309717895e-05,
2833
+ "loss": 3.1041,
2834
+ "step": 235500
2835
+ },
2836
+ {
2837
+ "epoch": 0.71,
2838
+ "learning_rate": 3.818176446256574e-05,
2839
+ "loss": 3.0904,
2840
+ "step": 236000
2841
+ },
2842
+ {
2843
+ "epoch": 0.71,
2844
+ "learning_rate": 3.8156725827952536e-05,
2845
+ "loss": 3.0874,
2846
+ "step": 236500
2847
+ },
2848
+ {
2849
+ "epoch": 0.71,
2850
+ "learning_rate": 3.8131687193339326e-05,
2851
+ "loss": 3.0746,
2852
+ "step": 237000
2853
+ },
2854
+ {
2855
+ "epoch": 0.71,
2856
+ "learning_rate": 3.810664855872612e-05,
2857
+ "loss": 3.1111,
2858
+ "step": 237500
2859
+ },
2860
+ {
2861
+ "epoch": 0.72,
2862
+ "learning_rate": 3.808160992411291e-05,
2863
+ "loss": 3.0794,
2864
+ "step": 238000
2865
+ },
2866
+ {
2867
+ "epoch": 0.72,
2868
+ "learning_rate": 3.80565712894997e-05,
2869
+ "loss": 3.0809,
2870
+ "step": 238500
2871
+ },
2872
+ {
2873
+ "epoch": 0.72,
2874
+ "learning_rate": 3.8031532654886495e-05,
2875
+ "loss": 3.0907,
2876
+ "step": 239000
2877
+ },
2878
+ {
2879
+ "epoch": 0.72,
2880
+ "learning_rate": 3.800649402027328e-05,
2881
+ "loss": 3.0692,
2882
+ "step": 239500
2883
+ },
2884
+ {
2885
+ "epoch": 0.72,
2886
+ "learning_rate": 3.7981455385660077e-05,
2887
+ "loss": 3.1191,
2888
+ "step": 240000
2889
+ },
2890
+ {
2891
+ "epoch": 0.72,
2892
+ "learning_rate": 3.795641675104687e-05,
2893
+ "loss": 3.0884,
2894
+ "step": 240500
2895
+ },
2896
+ {
2897
+ "epoch": 0.72,
2898
+ "learning_rate": 3.793137811643366e-05,
2899
+ "loss": 3.0955,
2900
+ "step": 241000
2901
+ },
2902
+ {
2903
+ "epoch": 0.73,
2904
+ "learning_rate": 3.7906339481820455e-05,
2905
+ "loss": 3.1308,
2906
+ "step": 241500
2907
+ },
2908
+ {
2909
+ "epoch": 0.73,
2910
+ "learning_rate": 3.788130084720724e-05,
2911
+ "loss": 3.0821,
2912
+ "step": 242000
2913
+ },
2914
+ {
2915
+ "epoch": 0.73,
2916
+ "learning_rate": 3.7856262212594036e-05,
2917
+ "loss": 3.1044,
2918
+ "step": 242500
2919
+ },
2920
+ {
2921
+ "epoch": 0.73,
2922
+ "learning_rate": 3.783122357798083e-05,
2923
+ "loss": 3.0543,
2924
+ "step": 243000
2925
+ },
2926
+ {
2927
+ "epoch": 0.73,
2928
+ "learning_rate": 3.780618494336762e-05,
2929
+ "loss": 3.0943,
2930
+ "step": 243500
2931
+ },
2932
+ {
2933
+ "epoch": 0.73,
2934
+ "learning_rate": 3.7781146308754415e-05,
2935
+ "loss": 3.0787,
2936
+ "step": 244000
2937
+ },
2938
+ {
2939
+ "epoch": 0.73,
2940
+ "learning_rate": 3.77561076741412e-05,
2941
+ "loss": 3.0943,
2942
+ "step": 244500
2943
+ },
2944
+ {
2945
+ "epoch": 0.74,
2946
+ "learning_rate": 3.7731069039527996e-05,
2947
+ "loss": 3.0755,
2948
+ "step": 245000
2949
+ },
2950
+ {
2951
+ "epoch": 0.74,
2952
+ "learning_rate": 3.770603040491478e-05,
2953
+ "loss": 3.0912,
2954
+ "step": 245500
2955
+ },
2956
+ {
2957
+ "epoch": 0.74,
2958
+ "learning_rate": 3.768099177030158e-05,
2959
+ "loss": 3.0616,
2960
+ "step": 246000
2961
+ },
2962
+ {
2963
+ "epoch": 0.74,
2964
+ "learning_rate": 3.7655953135688374e-05,
2965
+ "loss": 3.0884,
2966
+ "step": 246500
2967
+ },
2968
+ {
2969
+ "epoch": 0.74,
2970
+ "learning_rate": 3.763091450107516e-05,
2971
+ "loss": 3.0833,
2972
+ "step": 247000
2973
+ },
2974
+ {
2975
+ "epoch": 0.74,
2976
+ "learning_rate": 3.7605875866461955e-05,
2977
+ "loss": 3.0648,
2978
+ "step": 247500
2979
+ },
2980
+ {
2981
+ "epoch": 0.75,
2982
+ "learning_rate": 3.758083723184874e-05,
2983
+ "loss": 3.1001,
2984
+ "step": 248000
2985
+ },
2986
+ {
2987
+ "epoch": 0.75,
2988
+ "learning_rate": 3.7555798597235537e-05,
2989
+ "loss": 3.0793,
2990
+ "step": 248500
2991
+ },
2992
+ {
2993
+ "epoch": 0.75,
2994
+ "learning_rate": 3.7530759962622334e-05,
2995
+ "loss": 3.0914,
2996
+ "step": 249000
2997
+ },
2998
+ {
2999
+ "epoch": 0.75,
3000
+ "learning_rate": 3.750572132800912e-05,
3001
+ "loss": 3.0715,
3002
+ "step": 249500
3003
+ },
3004
+ {
3005
+ "epoch": 0.75,
3006
+ "learning_rate": 3.7480682693395915e-05,
3007
+ "loss": 3.0778,
3008
+ "step": 250000
3009
+ },
3010
+ {
3011
+ "epoch": 0.75,
3012
+ "learning_rate": 3.74556440587827e-05,
3013
+ "loss": 3.0749,
3014
+ "step": 250500
3015
+ },
3016
+ {
3017
+ "epoch": 0.75,
3018
+ "learning_rate": 3.7430605424169496e-05,
3019
+ "loss": 3.0795,
3020
+ "step": 251000
3021
+ },
3022
+ {
3023
+ "epoch": 0.76,
3024
+ "learning_rate": 3.740556678955629e-05,
3025
+ "loss": 3.0838,
3026
+ "step": 251500
3027
+ },
3028
+ {
3029
+ "epoch": 0.76,
3030
+ "learning_rate": 3.738052815494308e-05,
3031
+ "loss": 3.0874,
3032
+ "step": 252000
3033
+ },
3034
+ {
3035
+ "epoch": 0.76,
3036
+ "learning_rate": 3.7355489520329875e-05,
3037
+ "loss": 3.1254,
3038
+ "step": 252500
3039
+ },
3040
+ {
3041
+ "epoch": 0.76,
3042
+ "learning_rate": 3.733045088571666e-05,
3043
+ "loss": 3.0686,
3044
+ "step": 253000
3045
+ },
3046
+ {
3047
+ "epoch": 0.76,
3048
+ "learning_rate": 3.7305412251103456e-05,
3049
+ "loss": 3.0645,
3050
+ "step": 253500
3051
+ },
3052
+ {
3053
+ "epoch": 0.76,
3054
+ "learning_rate": 3.7280373616490246e-05,
3055
+ "loss": 3.0608,
3056
+ "step": 254000
3057
+ },
3058
+ {
3059
+ "epoch": 0.76,
3060
+ "learning_rate": 3.725533498187704e-05,
3061
+ "loss": 3.0896,
3062
+ "step": 254500
3063
+ },
3064
+ {
3065
+ "epoch": 0.77,
3066
+ "learning_rate": 3.7230296347263834e-05,
3067
+ "loss": 3.0622,
3068
+ "step": 255000
3069
+ },
3070
+ {
3071
+ "epoch": 0.77,
3072
+ "learning_rate": 3.720525771265062e-05,
3073
+ "loss": 3.0935,
3074
+ "step": 255500
3075
+ },
3076
+ {
3077
+ "epoch": 0.77,
3078
+ "learning_rate": 3.7180219078037415e-05,
3079
+ "loss": 3.051,
3080
+ "step": 256000
3081
+ },
3082
+ {
3083
+ "epoch": 0.77,
3084
+ "learning_rate": 3.7155180443424206e-05,
3085
+ "loss": 3.05,
3086
+ "step": 256500
3087
+ },
3088
+ {
3089
+ "epoch": 0.77,
3090
+ "learning_rate": 3.7130141808810996e-05,
3091
+ "loss": 3.0711,
3092
+ "step": 257000
3093
+ },
3094
+ {
3095
+ "epoch": 0.77,
3096
+ "learning_rate": 3.7105103174197794e-05,
3097
+ "loss": 3.086,
3098
+ "step": 257500
3099
+ },
3100
+ {
3101
+ "epoch": 0.78,
3102
+ "learning_rate": 3.708006453958458e-05,
3103
+ "loss": 3.0711,
3104
+ "step": 258000
3105
+ },
3106
+ {
3107
+ "epoch": 0.78,
3108
+ "learning_rate": 3.7055025904971375e-05,
3109
+ "loss": 3.0573,
3110
+ "step": 258500
3111
+ },
3112
+ {
3113
+ "epoch": 0.78,
3114
+ "learning_rate": 3.7029987270358165e-05,
3115
+ "loss": 3.0937,
3116
+ "step": 259000
3117
+ },
3118
+ {
3119
+ "epoch": 0.78,
3120
+ "learning_rate": 3.7004948635744956e-05,
3121
+ "loss": 3.0803,
3122
+ "step": 259500
3123
+ },
3124
+ {
3125
+ "epoch": 0.78,
3126
+ "learning_rate": 3.6979910001131747e-05,
3127
+ "loss": 3.0894,
3128
+ "step": 260000
3129
+ },
3130
+ {
3131
+ "epoch": 0.78,
3132
+ "learning_rate": 3.695487136651854e-05,
3133
+ "loss": 3.0443,
3134
+ "step": 260500
3135
+ },
3136
+ {
3137
+ "epoch": 0.78,
3138
+ "learning_rate": 3.6929832731905335e-05,
3139
+ "loss": 3.0307,
3140
+ "step": 261000
3141
+ },
3142
+ {
3143
+ "epoch": 0.79,
3144
+ "learning_rate": 3.6904794097292125e-05,
3145
+ "loss": 3.0806,
3146
+ "step": 261500
3147
+ },
3148
+ {
3149
+ "epoch": 0.79,
3150
+ "learning_rate": 3.6879755462678916e-05,
3151
+ "loss": 3.0864,
3152
+ "step": 262000
3153
+ },
3154
+ {
3155
+ "epoch": 0.79,
3156
+ "learning_rate": 3.6854716828065706e-05,
3157
+ "loss": 3.0655,
3158
+ "step": 262500
3159
+ },
3160
+ {
3161
+ "epoch": 0.79,
3162
+ "learning_rate": 3.68296781934525e-05,
3163
+ "loss": 3.0769,
3164
+ "step": 263000
3165
+ },
3166
+ {
3167
+ "epoch": 0.79,
3168
+ "learning_rate": 3.6804639558839294e-05,
3169
+ "loss": 3.0842,
3170
+ "step": 263500
3171
+ },
3172
+ {
3173
+ "epoch": 0.79,
3174
+ "learning_rate": 3.6779600924226085e-05,
3175
+ "loss": 3.0599,
3176
+ "step": 264000
3177
+ },
3178
+ {
3179
+ "epoch": 0.79,
3180
+ "learning_rate": 3.6754562289612875e-05,
3181
+ "loss": 3.0444,
3182
+ "step": 264500
3183
+ },
3184
+ {
3185
+ "epoch": 0.8,
3186
+ "learning_rate": 3.6729523654999666e-05,
3187
+ "loss": 3.0919,
3188
+ "step": 265000
3189
+ },
3190
+ {
3191
+ "epoch": 0.8,
3192
+ "learning_rate": 3.6704485020386456e-05,
3193
+ "loss": 3.0803,
3194
+ "step": 265500
3195
+ },
3196
+ {
3197
+ "epoch": 0.8,
3198
+ "learning_rate": 3.667944638577325e-05,
3199
+ "loss": 3.0625,
3200
+ "step": 266000
3201
+ },
3202
+ {
3203
+ "epoch": 0.8,
3204
+ "learning_rate": 3.6654407751160044e-05,
3205
+ "loss": 3.0659,
3206
+ "step": 266500
3207
+ },
3208
+ {
3209
+ "epoch": 0.8,
3210
+ "learning_rate": 3.6629369116546835e-05,
3211
+ "loss": 3.058,
3212
+ "step": 267000
3213
+ },
3214
+ {
3215
+ "epoch": 0.8,
3216
+ "learning_rate": 3.6604330481933625e-05,
3217
+ "loss": 3.039,
3218
+ "step": 267500
3219
+ },
3220
+ {
3221
+ "epoch": 0.81,
3222
+ "learning_rate": 3.6579291847320416e-05,
3223
+ "loss": 3.0609,
3224
+ "step": 268000
3225
+ },
3226
+ {
3227
+ "epoch": 0.81,
3228
+ "learning_rate": 3.6554253212707207e-05,
3229
+ "loss": 3.0642,
3230
+ "step": 268500
3231
+ },
3232
+ {
3233
+ "epoch": 0.81,
3234
+ "learning_rate": 3.6529214578094004e-05,
3235
+ "loss": 3.0689,
3236
+ "step": 269000
3237
+ },
3238
+ {
3239
+ "epoch": 0.81,
3240
+ "learning_rate": 3.6504175943480794e-05,
3241
+ "loss": 3.0742,
3242
+ "step": 269500
3243
+ },
3244
+ {
3245
+ "epoch": 0.81,
3246
+ "learning_rate": 3.6479137308867585e-05,
3247
+ "loss": 3.0593,
3248
+ "step": 270000
3249
+ },
3250
+ {
3251
+ "epoch": 0.81,
3252
+ "learning_rate": 3.6454098674254376e-05,
3253
+ "loss": 3.0275,
3254
+ "step": 270500
3255
+ },
3256
+ {
3257
+ "epoch": 0.81,
3258
+ "learning_rate": 3.6429060039641166e-05,
3259
+ "loss": 3.0494,
3260
+ "step": 271000
3261
+ },
3262
+ {
3263
+ "epoch": 0.82,
3264
+ "learning_rate": 3.6404021405027963e-05,
3265
+ "loss": 3.0539,
3266
+ "step": 271500
3267
+ },
3268
+ {
3269
+ "epoch": 0.82,
3270
+ "learning_rate": 3.637898277041475e-05,
3271
+ "loss": 3.0219,
3272
+ "step": 272000
3273
+ },
3274
+ {
3275
+ "epoch": 0.82,
3276
+ "learning_rate": 3.6353944135801545e-05,
3277
+ "loss": 3.0594,
3278
+ "step": 272500
3279
+ },
3280
+ {
3281
+ "epoch": 0.82,
3282
+ "learning_rate": 3.6328905501188335e-05,
3283
+ "loss": 3.0226,
3284
+ "step": 273000
3285
+ },
3286
+ {
3287
+ "epoch": 0.82,
3288
+ "learning_rate": 3.6303866866575126e-05,
3289
+ "loss": 3.062,
3290
+ "step": 273500
3291
+ },
3292
+ {
3293
+ "epoch": 0.82,
3294
+ "learning_rate": 3.627882823196192e-05,
3295
+ "loss": 3.0278,
3296
+ "step": 274000
3297
+ },
3298
+ {
3299
+ "epoch": 0.82,
3300
+ "learning_rate": 3.625378959734871e-05,
3301
+ "loss": 3.0481,
3302
+ "step": 274500
3303
+ },
3304
+ {
3305
+ "epoch": 0.83,
3306
+ "learning_rate": 3.6228750962735504e-05,
3307
+ "loss": 3.0711,
3308
+ "step": 275000
3309
+ },
3310
+ {
3311
+ "epoch": 0.83,
3312
+ "learning_rate": 3.6203712328122295e-05,
3313
+ "loss": 3.0537,
3314
+ "step": 275500
3315
+ },
3316
+ {
3317
+ "epoch": 0.83,
3318
+ "learning_rate": 3.6178673693509085e-05,
3319
+ "loss": 3.0285,
3320
+ "step": 276000
3321
+ },
3322
+ {
3323
+ "epoch": 0.83,
3324
+ "learning_rate": 3.615363505889588e-05,
3325
+ "loss": 3.059,
3326
+ "step": 276500
3327
+ },
3328
+ {
3329
+ "epoch": 0.83,
3330
+ "learning_rate": 3.6128596424282666e-05,
3331
+ "loss": 3.0548,
3332
+ "step": 277000
3333
+ },
3334
+ {
3335
+ "epoch": 0.83,
3336
+ "learning_rate": 3.6103557789669464e-05,
3337
+ "loss": 3.0681,
3338
+ "step": 277500
3339
+ },
3340
+ {
3341
+ "epoch": 0.84,
3342
+ "learning_rate": 3.6078519155056254e-05,
3343
+ "loss": 3.0542,
3344
+ "step": 278000
3345
+ },
3346
+ {
3347
+ "epoch": 0.84,
3348
+ "learning_rate": 3.6053480520443045e-05,
3349
+ "loss": 3.0432,
3350
+ "step": 278500
3351
+ },
3352
+ {
3353
+ "epoch": 0.84,
3354
+ "learning_rate": 3.602844188582984e-05,
3355
+ "loss": 3.0262,
3356
+ "step": 279000
3357
+ },
3358
+ {
3359
+ "epoch": 0.84,
3360
+ "learning_rate": 3.6003403251216626e-05,
3361
+ "loss": 3.0477,
3362
+ "step": 279500
3363
+ },
3364
+ {
3365
+ "epoch": 0.84,
3366
+ "learning_rate": 3.5978364616603423e-05,
3367
+ "loss": 3.0659,
3368
+ "step": 280000
3369
+ },
3370
+ {
3371
+ "epoch": 0.84,
3372
+ "learning_rate": 3.595332598199021e-05,
3373
+ "loss": 3.0485,
3374
+ "step": 280500
3375
+ },
3376
+ {
3377
+ "epoch": 0.84,
3378
+ "learning_rate": 3.5928287347377005e-05,
3379
+ "loss": 3.0876,
3380
+ "step": 281000
3381
+ },
3382
+ {
3383
+ "epoch": 0.85,
3384
+ "learning_rate": 3.5903248712763795e-05,
3385
+ "loss": 3.0533,
3386
+ "step": 281500
3387
+ },
3388
+ {
3389
+ "epoch": 0.85,
3390
+ "learning_rate": 3.5878210078150586e-05,
3391
+ "loss": 3.0515,
3392
+ "step": 282000
3393
+ },
3394
+ {
3395
+ "epoch": 0.85,
3396
+ "learning_rate": 3.585317144353738e-05,
3397
+ "loss": 3.0581,
3398
+ "step": 282500
3399
+ },
3400
+ {
3401
+ "epoch": 0.85,
3402
+ "learning_rate": 3.582813280892417e-05,
3403
+ "loss": 3.0733,
3404
+ "step": 283000
3405
+ },
3406
+ {
3407
+ "epoch": 0.85,
3408
+ "learning_rate": 3.5803094174310964e-05,
3409
+ "loss": 3.0407,
3410
+ "step": 283500
3411
+ },
3412
+ {
3413
+ "epoch": 0.85,
3414
+ "learning_rate": 3.5778055539697755e-05,
3415
+ "loss": 3.0176,
3416
+ "step": 284000
3417
+ },
3418
+ {
3419
+ "epoch": 0.85,
3420
+ "learning_rate": 3.5753016905084545e-05,
3421
+ "loss": 3.0331,
3422
+ "step": 284500
3423
+ },
3424
+ {
3425
+ "epoch": 0.86,
3426
+ "learning_rate": 3.572797827047134e-05,
3427
+ "loss": 3.0587,
3428
+ "step": 285000
3429
+ },
3430
+ {
3431
+ "epoch": 0.86,
3432
+ "learning_rate": 3.5702939635858126e-05,
3433
+ "loss": 3.036,
3434
+ "step": 285500
3435
+ },
3436
+ {
3437
+ "epoch": 0.86,
3438
+ "learning_rate": 3.5677901001244924e-05,
3439
+ "loss": 3.0488,
3440
+ "step": 286000
3441
+ },
3442
+ {
3443
+ "epoch": 0.86,
3444
+ "learning_rate": 3.5652862366631714e-05,
3445
+ "loss": 3.0591,
3446
+ "step": 286500
3447
+ },
3448
+ {
3449
+ "epoch": 0.86,
3450
+ "learning_rate": 3.5627823732018505e-05,
3451
+ "loss": 3.0318,
3452
+ "step": 287000
3453
+ },
3454
+ {
3455
+ "epoch": 0.86,
3456
+ "learning_rate": 3.56027850974053e-05,
3457
+ "loss": 3.0268,
3458
+ "step": 287500
3459
+ },
3460
+ {
3461
+ "epoch": 0.87,
3462
+ "learning_rate": 3.5577746462792086e-05,
3463
+ "loss": 3.0655,
3464
+ "step": 288000
3465
+ },
3466
+ {
3467
+ "epoch": 0.87,
3468
+ "learning_rate": 3.555270782817888e-05,
3469
+ "loss": 3.0387,
3470
+ "step": 288500
3471
+ },
3472
+ {
3473
+ "epoch": 0.87,
3474
+ "learning_rate": 3.5527669193565674e-05,
3475
+ "loss": 3.0587,
3476
+ "step": 289000
3477
+ },
3478
+ {
3479
+ "epoch": 0.87,
3480
+ "learning_rate": 3.5502630558952464e-05,
3481
+ "loss": 3.0486,
3482
+ "step": 289500
3483
+ },
3484
+ {
3485
+ "epoch": 0.87,
3486
+ "learning_rate": 3.547759192433926e-05,
3487
+ "loss": 3.078,
3488
+ "step": 290000
3489
+ },
3490
+ {
3491
+ "epoch": 0.87,
3492
+ "learning_rate": 3.5452553289726046e-05,
3493
+ "loss": 3.0346,
3494
+ "step": 290500
3495
+ },
3496
+ {
3497
+ "epoch": 0.87,
3498
+ "learning_rate": 3.542751465511284e-05,
3499
+ "loss": 3.0631,
3500
+ "step": 291000
3501
+ },
3502
+ {
3503
+ "epoch": 0.88,
3504
+ "learning_rate": 3.5402476020499634e-05,
3505
+ "loss": 3.0286,
3506
+ "step": 291500
3507
+ },
3508
+ {
3509
+ "epoch": 0.88,
3510
+ "learning_rate": 3.5377437385886424e-05,
3511
+ "loss": 3.0451,
3512
+ "step": 292000
3513
+ },
3514
+ {
3515
+ "epoch": 0.88,
3516
+ "learning_rate": 3.5352398751273215e-05,
3517
+ "loss": 3.0457,
3518
+ "step": 292500
3519
+ },
3520
+ {
3521
+ "epoch": 0.88,
3522
+ "learning_rate": 3.5327360116660005e-05,
3523
+ "loss": 3.0523,
3524
+ "step": 293000
3525
+ },
3526
+ {
3527
+ "epoch": 0.88,
3528
+ "learning_rate": 3.53023214820468e-05,
3529
+ "loss": 3.0505,
3530
+ "step": 293500
3531
+ },
3532
+ {
3533
+ "epoch": 0.88,
3534
+ "learning_rate": 3.527728284743359e-05,
3535
+ "loss": 3.0391,
3536
+ "step": 294000
3537
+ },
3538
+ {
3539
+ "epoch": 0.88,
3540
+ "learning_rate": 3.5252244212820384e-05,
3541
+ "loss": 3.032,
3542
+ "step": 294500
3543
+ },
3544
+ {
3545
+ "epoch": 0.89,
3546
+ "learning_rate": 3.5227205578207174e-05,
3547
+ "loss": 3.048,
3548
+ "step": 295000
3549
+ },
3550
+ {
3551
+ "epoch": 0.89,
3552
+ "learning_rate": 3.5202166943593965e-05,
3553
+ "loss": 3.0276,
3554
+ "step": 295500
3555
+ },
3556
+ {
3557
+ "epoch": 0.89,
3558
+ "learning_rate": 3.517712830898076e-05,
3559
+ "loss": 3.0132,
3560
+ "step": 296000
3561
+ },
3562
+ {
3563
+ "epoch": 0.89,
3564
+ "learning_rate": 3.515208967436755e-05,
3565
+ "loss": 3.0059,
3566
+ "step": 296500
3567
+ },
3568
+ {
3569
+ "epoch": 0.89,
3570
+ "learning_rate": 3.512705103975434e-05,
3571
+ "loss": 3.0293,
3572
+ "step": 297000
3573
+ },
3574
+ {
3575
+ "epoch": 0.89,
3576
+ "learning_rate": 3.5102012405141134e-05,
3577
+ "loss": 3.0349,
3578
+ "step": 297500
3579
+ },
3580
+ {
3581
+ "epoch": 0.9,
3582
+ "learning_rate": 3.5076973770527924e-05,
3583
+ "loss": 3.0338,
3584
+ "step": 298000
3585
+ },
3586
+ {
3587
+ "epoch": 0.9,
3588
+ "learning_rate": 3.505193513591472e-05,
3589
+ "loss": 3.0429,
3590
+ "step": 298500
3591
+ },
3592
+ {
3593
+ "epoch": 0.9,
3594
+ "learning_rate": 3.502689650130151e-05,
3595
+ "loss": 3.017,
3596
+ "step": 299000
3597
+ },
3598
+ {
3599
+ "epoch": 0.9,
3600
+ "learning_rate": 3.50018578666883e-05,
3601
+ "loss": 3.0174,
3602
+ "step": 299500
3603
+ },
3604
+ {
3605
+ "epoch": 0.9,
3606
+ "learning_rate": 3.4976819232075093e-05,
3607
+ "loss": 3.0668,
3608
+ "step": 300000
3609
+ },
3610
+ {
3611
+ "epoch": 0.9,
3612
+ "learning_rate": 3.4951780597461884e-05,
3613
+ "loss": 3.0236,
3614
+ "step": 300500
3615
+ },
3616
+ {
3617
+ "epoch": 0.9,
3618
+ "learning_rate": 3.4926741962848675e-05,
3619
+ "loss": 3.0195,
3620
+ "step": 301000
3621
+ },
3622
+ {
3623
+ "epoch": 0.91,
3624
+ "learning_rate": 3.490170332823547e-05,
3625
+ "loss": 3.0443,
3626
+ "step": 301500
3627
+ }
3628
+ ],
3629
+ "max_steps": 998457,
3630
+ "num_train_epochs": 3,
3631
+ "total_flos": 4.46482559989836e+18,
3632
+ "trial_name": null,
3633
+ "trial_params": null
3634
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ebe01baaac4ffc3f7330dfc964505a07db58347091297709440d24260203bcd
3
+ size 3899