MHGanainy/8-clusters-balanced-lex-best-v2-3

all_results.json

@@ -1,13 +1,13 @@
 {
  "epoch": 1.0,
- "eval_loss": 1.9266319274902344,
- "eval_runtime": 63.6699,
- "eval_samples_per_second": 15.549,
- "eval_steps_per_second": 1.948,
- "perplexity": 6.866344904696836,
+ "eval_loss": 1.9266563653945923,
+ "eval_runtime": 63.5112,
+ "eval_samples_per_second": 15.588,
+ "eval_steps_per_second": 1.952,
+ "perplexity": 6.866512705827251,
  "total_flos": 1.5640093507584e+17,
- "train_loss": 2.1807824082961527,
- "train_runtime": 2484.8584,
- "train_samples_per_second": 6.911,
- "train_steps_per_second": 3.456
+ "train_loss": 2.1807946249230077,
+ "train_runtime": 2485.1722,
+ "train_samples_per_second": 6.91,
+ "train_steps_per_second": 3.455
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 1.0,
- "eval_loss": 1.9266319274902344,
- "eval_runtime": 63.6699,
- "eval_samples_per_second": 15.549,
- "eval_steps_per_second": 1.948,
- "perplexity": 6.866344904696836
+ "eval_loss": 1.9266563653945923,
+ "eval_runtime": 63.5112,
+ "eval_samples_per_second": 15.588,
+ "eval_steps_per_second": 1.952,
+ "perplexity": 6.866512705827251
 }

train_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 1.0,
  "total_flos": 1.5640093507584e+17,
- "train_loss": 2.1807824082961527,
- "train_runtime": 2484.8584,
- "train_samples_per_second": 6.911,
- "train_steps_per_second": 3.456
+ "train_loss": 2.1807946249230077,
+ "train_runtime": 2485.1722,
+ "train_samples_per_second": 6.91,
+ "train_steps_per_second": 3.455
 }

trainer_state.json

@@ -10,615 +10,615 @@
  "log_history": [
  {
  "epoch": 0.01164551065564225,
- "grad_norm": 0.14791734516620636,
+ "grad_norm": 0.14460527896881104,
  "learning_rate": 2.331002331002331e-06,
  "loss": 2.4953,
  "step": 100
  },
  {
  "epoch": 0.0232910213112845,
- "grad_norm": 0.2441452145576477,
+ "grad_norm": 0.24249859154224396,
  "learning_rate": 4.662004662004662e-06,
- "loss": 2.4575,
+ "loss": 2.4574,
  "step": 200
  },
  {
  "epoch": 0.03493653196692675,
- "grad_norm": 0.22649115324020386,
+ "grad_norm": 0.22802527248859406,
  "learning_rate": 6.993006993006993e-06,
  "loss": 2.4409,
  "step": 300
  },
  {
  "epoch": 0.046582042622569,
- "grad_norm": 0.3321034610271454,
+ "grad_norm": 0.33553868532180786,
  "learning_rate": 9.324009324009324e-06,
  "loss": 2.4497,
  "step": 400
  },
  {
  "epoch": 0.05822755327821125,
- "grad_norm": 0.36618897318840027,
+ "grad_norm": 0.36703944206237793,
  "learning_rate": 1.1655011655011657e-05,
- "loss": 2.3973,
+ "loss": 2.3974,
  "step": 500
  },
  {
  "epoch": 0.0698730639338535,
- "grad_norm": 0.42568501830101013,
+ "grad_norm": 0.42922335863113403,
  "learning_rate": 1.3986013986013986e-05,
- "loss": 2.356,
+ "loss": 2.3563,
  "step": 600
  },
  {
  "epoch": 0.08151857458949575,
- "grad_norm": 0.4560384452342987,
+ "grad_norm": 0.4571892023086548,
  "learning_rate": 1.6317016317016318e-05,
- "loss": 2.317,
+ "loss": 2.3174,
  "step": 700
  },
  {
  "epoch": 0.093164085245138,
- "grad_norm": 0.5321765542030334,
+ "grad_norm": 0.533064067363739,
  "learning_rate": 1.8648018648018647e-05,
- "loss": 2.3544,
+ "loss": 2.3549,
  "step": 800
  },
  {
  "epoch": 0.10480959590078025,
- "grad_norm": 0.5255011916160583,
+ "grad_norm": 0.5255058407783508,
  "learning_rate": 1.999854282682324e-05,
- "loss": 2.3049,
+ "loss": 2.3053,
  "step": 900
  },
  {
  "epoch": 0.1164551065564225,
- "grad_norm": 0.563510537147522,
+ "grad_norm": 0.5629482865333557,
  "learning_rate": 1.9983347507285766e-05,
- "loss": 2.2863,
+ "loss": 2.2866,
  "step": 1000
  },
  {
  "epoch": 0.12810061721206475,
- "grad_norm": 0.5928713083267212,
+ "grad_norm": 0.5937526226043701,
  "learning_rate": 1.9951660332101616e-05,
- "loss": 2.2718,
+ "loss": 2.2721,
  "step": 1100
  },
  {
  "epoch": 0.139746127867707,
- "grad_norm": 0.7078022956848145,
+ "grad_norm": 0.703377366065979,
  "learning_rate": 1.9903533646470504e-05,
- "loss": 2.2645,
+ "loss": 2.2648,
  "step": 1200
  },
  {
  "epoch": 0.15139163852334925,
- "grad_norm": 0.612659752368927,
+ "grad_norm": 0.6113013625144958,
  "learning_rate": 1.9839046952618667e-05,
- "loss": 2.2507,
+ "loss": 2.2509,
  "step": 1300
  },
  {
  "epoch": 0.1630371491789915,
- "grad_norm": 0.6351534724235535,
+ "grad_norm": 0.6333953738212585,
  "learning_rate": 1.9758306778466264e-05,
- "loss": 2.2474,
+ "loss": 2.2475,
  "step": 1400
  },
  {
  "epoch": 0.17468265983463374,
- "grad_norm": 0.616267740726471,
+ "grad_norm": 0.6210408806800842,
  "learning_rate": 1.966144650165e-05,
- "loss": 2.2574,
+ "loss": 2.2575,
  "step": 1500
  },
  {
  "epoch": 0.186328170490276,
- "grad_norm": 0.6416229009628296,
+ "grad_norm": 0.6377599239349365,
  "learning_rate": 1.9548626129191778e-05,
- "loss": 2.249,
+ "loss": 2.2492,
  "step": 1600
  },
  {
  "epoch": 0.19797368114591826,
- "grad_norm": 0.6149903535842896,
+ "grad_norm": 0.6116755604743958,
  "learning_rate": 1.9420032033177225e-05,
- "loss": 2.2497,
+ "loss": 2.2498,
  "step": 1700
  },
  {
  "epoch": 0.2096191918015605,
- "grad_norm": 0.9221161007881165,
+ "grad_norm": 0.9220359921455383,
  "learning_rate": 1.927587664288089e-05,
- "loss": 2.223,
+ "loss": 2.2231,
  "step": 1800
  },
  {
  "epoch": 0.22126470245720276,
- "grad_norm": 0.6987215876579285,
+ "grad_norm": 0.6930053234100342,
  "learning_rate": 1.911639809384656e-05,
- "loss": 2.1861,
+ "loss": 2.1863,
  "step": 1900
  },
  {
  "epoch": 0.232910213112845,
- "grad_norm": 1.0042574405670166,
+ "grad_norm": 0.991111695766449,
  "learning_rate": 1.8941859834502484e-05,
- "loss": 2.1883,
+ "loss": 2.1885,
  "step": 2000
  },
  {
  "epoch": 0.24455572376848725,
- "grad_norm": 0.975431501865387,
+ "grad_norm": 0.9630569815635681,
  "learning_rate": 1.8752550190961288e-05,
- "loss": 2.2078,
+ "loss": 2.208,
  "step": 2100
  },
  {
  "epoch": 0.2562012344241295,
- "grad_norm": 0.8907233476638794,
+ "grad_norm": 0.8858104348182678,
  "learning_rate": 1.8548781890723614e-05,
- "loss": 2.2046,
+ "loss": 2.2048,
  "step": 2200
  },
  {
  "epoch": 0.26784674507977174,
- "grad_norm": 0.9513066411018372,
+ "grad_norm": 0.9373981952667236,
  "learning_rate": 1.8330891546072095e-05,
- "loss": 2.2023,
+ "loss": 2.2026,
  "step": 2300
  },
  {
  "epoch": 0.279492255735414,
- "grad_norm": 0.6928815841674805,
+ "grad_norm": 0.6907739639282227,
  "learning_rate": 1.809923909800931e-05,
  "loss": 2.1932,
  "step": 2400
  },
  {
  "epoch": 0.29113776639105626,
- "grad_norm": 0.7470653057098389,
+ "grad_norm": 0.7434536814689636,
  "learning_rate": 1.7854207221658092e-05,
- "loss": 2.2031,
+ "loss": 2.2032,
  "step": 2500
  },
  {
  "epoch": 0.3027832770466985,
- "grad_norm": 0.7841106653213501,
+ "grad_norm": 0.7883014678955078,
  "learning_rate": 1.7596200694106552e-05,
- "loss": 2.1867,
+ "loss": 2.1869,
  "step": 2600
  },
  {
  "epoch": 0.31442878770234073,
- "grad_norm": 0.9168167114257812,
+ "grad_norm": 0.9146271347999573,
  "learning_rate": 1.7325645725742056e-05,
- "loss": 2.1939,
+ "loss": 2.1941,
  "step": 2700
  },
  {
  "epoch": 0.326074298357983,
- "grad_norm": 0.9133324027061462,
+ "grad_norm": 0.9042865633964539,
  "learning_rate": 1.7042989256178744e-05,
- "loss": 2.1522,
+ "loss": 2.1523,
  "step": 2800
  },
  {
  "epoch": 0.33771980901362525,
- "grad_norm": 0.985498309135437,
+ "grad_norm": 0.9685861468315125,
  "learning_rate": 1.6748698215941704e-05,
- "loss": 2.1735,
+ "loss": 2.1736,
  "step": 2900
  },
  {
  "epoch": 0.3493653196692675,
- "grad_norm": 0.7974863052368164,
+ "grad_norm": 0.7863497138023376,
  "learning_rate": 1.6443258755127393e-05,
- "loss": 2.1939,
+ "loss": 2.1938,
  "step": 3000
  },
  {
  "epoch": 0.36101083032490977,
- "grad_norm": 0.6346118450164795,
+ "grad_norm": 0.631289005279541,
  "learning_rate": 1.6127175440314596e-05,
  "loss": 2.1204,
  "step": 3100
  },
  {
  "epoch": 0.372656340980552,
- "grad_norm": 0.8495335578918457,
+ "grad_norm": 0.8423033952713013,
  "learning_rate": 1.5800970421052487e-05,
- "loss": 2.1964,
+ "loss": 2.1962,
  "step": 3200
  },
  {
  "epoch": 0.38430185163619424,
- "grad_norm": 1.029366135597229,
+ "grad_norm": 1.026877522468567,
  "learning_rate": 1.546518256730277e-05,
- "loss": 2.1838,
+ "loss": 2.1839,
  "step": 3300
  },
  {
  "epoch": 0.3959473622918365,
- "grad_norm": 1.0092447996139526,
+ "grad_norm": 1.0023020505905151,
  "learning_rate": 1.5120366579260734e-05,
  "loss": 2.1607,
  "step": 3400
  },
  {
  "epoch": 0.40759287294747876,
- "grad_norm": 0.8872857689857483,
+ "grad_norm": 0.8861306309700012,
  "learning_rate": 1.4767092071025792e-05,
- "loss": 2.1836,
+ "loss": 2.1838,
  "step": 3500
  },
  {
  "epoch": 0.419238383603121,
- "grad_norm": 0.8824167251586914,
+ "grad_norm": 0.8740290403366089,
  "learning_rate": 1.4405942629635174e-05,
- "loss": 2.1207,
+ "loss": 2.1205,
  "step": 3600
  },
  {
  "epoch": 0.4308838942587632,
- "grad_norm": 0.7451480627059937,
+ "grad_norm": 0.7423445582389832,
  "learning_rate": 1.4037514851015241e-05,
  "loss": 2.215,
  "step": 3700
  },
  {
  "epoch": 0.4425294049144055,
- "grad_norm": 0.9333692193031311,
+ "grad_norm": 0.9221410155296326,
  "learning_rate": 1.3662417354442924e-05,
  "loss": 2.1836,
  "step": 3800
  },
  {
  "epoch": 0.45417491557004774,
- "grad_norm": 0.9870197176933289,
+ "grad_norm": 0.9722244143486023,
  "learning_rate": 1.3281269777145354e-05,
- "loss": 2.1935,
+ "loss": 2.1934,
  "step": 3900
  },
  {
  "epoch": 0.46582042622569,
- "grad_norm": 1.1925898790359497,
+ "grad_norm": 1.1752955913543701,
  "learning_rate": 1.2894701750698541e-05,
- "loss": 2.1383,
+ "loss": 2.1382,
  "step": 4000
  },
  {
  "epoch": 0.47746593688133226,
- "grad_norm": 0.8721115589141846,
+ "grad_norm": 0.8646457195281982,
  "learning_rate": 1.2503351860916024e-05,
- "loss": 2.1506,
+ "loss": 2.1507,
  "step": 4100
  },
  {
  "epoch": 0.4891114475369745,
- "grad_norm": 1.3675851821899414,
+ "grad_norm": 1.307706594467163,
  "learning_rate": 1.2107866592945686e-05,
- "loss": 2.1562,
+ "loss": 2.1564,
  "step": 4200
  },
  {
  "epoch": 0.5007569581926168,
- "grad_norm": 0.9814177751541138,
+ "grad_norm": 0.9780316948890686,
  "learning_rate": 1.1708899263317381e-05,
- "loss": 2.1394,
+ "loss": 2.1395,
  "step": 4300
  },
  {
  "epoch": 0.512402468848259,
- "grad_norm": 0.8111391067504883,
+ "grad_norm": 0.8423807621002197,
  "learning_rate": 1.1307108940705536e-05,
  "loss": 2.1605,
  "step": 4400
  },
  {
  "epoch": 0.5240479795039013,
- "grad_norm": 0.9435734152793884,
+ "grad_norm": 0.9365679621696472,
  "learning_rate": 1.090315935718958e-05,
  "loss": 2.1108,
  "step": 4500
  },
  {
  "epoch": 0.5356934901595435,
- "grad_norm": 0.8855921626091003,
+ "grad_norm": 0.8869751691818237,
  "learning_rate": 1.0497717811810748e-05,
- "loss": 2.1364,
+ "loss": 2.1365,
  "step": 4600
  },
  {
  "epoch": 0.5473390008151857,
- "grad_norm": 0.7305698990821838,
+ "grad_norm": 0.7278423309326172,
  "learning_rate": 1.0091454068236455e-05,
  "loss": 2.1498,
  "step": 4700
  },
  {
  "epoch": 0.558984511470828,
- "grad_norm": 1.123586654663086,
+ "grad_norm": 1.1241728067398071,
  "learning_rate": 9.685039248353284e-06,
- "loss": 2.1634,
+ "loss": 2.1635,
  "step": 4800
  },
  {
  "epoch": 0.5706300221264703,
- "grad_norm": 0.857729434967041,
+ "grad_norm": 0.8585646748542786,
  "learning_rate": 9.279144723616279e-06,
- "loss": 2.0721,
+ "loss": 2.072,
  "step": 4900
  },
  {
  "epoch": 0.5822755327821125,
- "grad_norm": 0.7783474326133728,
+ "grad_norm": 0.768528938293457,
  "learning_rate": 8.874441005985965e-06,
- "loss": 2.1569,
+ "loss": 2.1568,
  "step": 5000
  },
  {
  "epoch": 0.5939210434377548,
- "grad_norm": 1.0050169229507446,
+ "grad_norm": 1.002091407775879,
  "learning_rate": 8.47159664028521e-06,
  "loss": 2.1631,
  "step": 5100
  },
  {
  "epoch": 0.605566554093397,
- "grad_norm": 0.7871978282928467,
+ "grad_norm": 0.7817184329032898,
  "learning_rate": 8.07127709980564e-06,
- "loss": 2.1207,
+ "loss": 2.1203,
  "step": 5200
  },
  {
  "epoch": 0.6172120647490392,
- "grad_norm": 1.3452588319778442,
+ "grad_norm": 1.3114322423934937,
  "learning_rate": 7.674143686988085e-06,
  "loss": 2.0871,
  "step": 5300
  },
  {
  "epoch": 0.6288575754046815,
- "grad_norm": 0.7304858565330505,
+ "grad_norm": 0.7261312007904053,
  "learning_rate": 7.280852440992941e-06,
  "loss": 2.1289,
  "step": 5400
  },
  {
  "epoch": 0.6405030860603237,
- "grad_norm": 0.9213933944702148,
+ "grad_norm": 0.9209310412406921,
  "learning_rate": 6.89205305396518e-06,
- "loss": 2.1067,
+ "loss": 2.1063,
  "step": 5500
  },
  {
  "epoch": 0.652148596715966,
- "grad_norm": 0.9295869469642639,
+ "grad_norm": 0.9243987798690796,
  "learning_rate": 6.508387797784227e-06,
- "loss": 2.0959,
+ "loss": 2.0956,
  "step": 5600
  },
  {
  "epoch": 0.6637941073716083,
- "grad_norm": 0.9920222759246826,
+ "grad_norm": 0.9812811017036438,
  "learning_rate": 6.130490463071604e-06,
- "loss": 2.0919,
+ "loss": 2.0918,
  "step": 5700
  },
  {
  "epoch": 0.6754396180272505,
- "grad_norm": 0.8308998942375183,
+ "grad_norm": 0.8238828182220459,
  "learning_rate": 5.758985312209124e-06,
- "loss": 2.1206,
+ "loss": 2.1203,
  "step": 5800
  },
  {
  "epoch": 0.6870851286828927,
- "grad_norm": 1.1122294664382935,
+ "grad_norm": 1.1043486595153809,
  "learning_rate": 5.394486048097099e-06,
- "loss": 2.116,
+ "loss": 2.1159,
  "step": 5900
  },
  {
  "epoch": 0.698730639338535,
- "grad_norm": 0.7995828986167908,
+ "grad_norm": 0.7944552898406982,
  "learning_rate": 5.037594800356142e-06,
- "loss": 2.1159,
+ "loss": 2.1157,
  "step": 6000
  },
  {
  "epoch": 0.7103761499941772,
- "grad_norm": 1.1691644191741943,
+ "grad_norm": 1.1705522537231445,
  "learning_rate": 4.688901130647314e-06,
  "loss": 2.1396,
  "step": 6100
  },
  {
  "epoch": 0.7220216606498195,
- "grad_norm": 0.8228808045387268,
+ "grad_norm": 0.8150696754455566,
  "learning_rate": 4.348981058753708e-06,
  "loss": 2.0819,
  "step": 6200
  },
  {
  "epoch": 0.7336671713054618,
- "grad_norm": 1.0278650522232056,
+ "grad_norm": 1.0273561477661133,
  "learning_rate": 4.018396111032394e-06,
- "loss": 2.1227,
+ "loss": 2.1226,
  "step": 6300
  },
  {
  "epoch": 0.745312681961104,
- "grad_norm": 0.9175333976745605,
+ "grad_norm": 0.9231303334236145,
  "learning_rate": 3.697692392808545e-06,
- "loss": 2.1037,
+ "loss": 2.1035,
  "step": 6400
  },
  {
  "epoch": 0.7569581926167462,
- "grad_norm": 0.7555139064788818,
+ "grad_norm": 0.7557884454727173,
  "learning_rate": 3.387399686244144e-06,
- "loss": 2.0953,
+ "loss": 2.0954,
  "step": 6500
  },
  {
  "epoch": 0.7686037032723885,
- "grad_norm": 0.8671744465827942,
+ "grad_norm": 0.8633888959884644,
  "learning_rate": 3.0880305751715402e-06,
- "loss": 2.1201,
+ "loss": 2.1199,
  "step": 6600
  },
  {
  "epoch": 0.7802492139280307,
- "grad_norm": 0.8459142446517944,
+ "grad_norm": 0.8371317982673645,
  "learning_rate": 2.800079598337505e-06,
- "loss": 2.1136,
+ "loss": 2.1135,
  "step": 6700
  },
  {
  "epoch": 0.791894724583673,
- "grad_norm": 0.763583242893219,
+ "grad_norm": 0.7599378228187561,
  "learning_rate": 2.524022432456664e-06,
- "loss": 2.1386,
+ "loss": 2.1385,
  "step": 6800
  },
  {
  "epoch": 0.8035402352393153,
- "grad_norm": 0.8807295560836792,
+ "grad_norm": 0.8745434880256653,
  "learning_rate": 2.260315106423807e-06,
- "loss": 2.0779,
+ "loss": 2.0778,
  "step": 6900
  },
  {
  "epoch": 0.8151857458949575,
- "grad_norm": 0.8740783333778381,
+ "grad_norm": 0.871006429195404,
  "learning_rate": 2.0093932479830935e-06,
  "loss": 2.1367,
  "step": 7000
  },
  {
  "epoch": 0.8268312565505997,
- "grad_norm": 1.039287805557251,
+ "grad_norm": 1.034953236579895,
  "learning_rate": 1.7716713640987526e-06,
- "loss": 2.1166,
+ "loss": 2.1163,
  "step": 7100
  },
  {
  "epoch": 0.838476767206242,
- "grad_norm": 0.8560532331466675,
+ "grad_norm": 0.8476694226264954,
  "learning_rate": 1.5475421562158854e-06,
  "loss": 2.1459,
  "step": 7200
  },
  {
  "epoch": 0.8501222778618842,
- "grad_norm": 0.9102568030357361,
+ "grad_norm": 0.9011842012405396,
  "learning_rate": 1.3373758715426444e-06,
- "loss": 2.1095,
+ "loss": 2.1094,
  "step": 7300
  },
  {
  "epoch": 0.8617677885175264,
- "grad_norm": 0.8189859390258789,
+ "grad_norm": 0.8178461790084839,
  "learning_rate": 1.141519691425379e-06,
- "loss": 2.0852,
+ "loss": 2.0853,
  "step": 7400
  },
  {
  "epoch": 0.8734132991731688,
- "grad_norm": 0.9001794457435608,
+ "grad_norm": 0.898789644241333,
  "learning_rate": 9.60297157827106e-07,
- "loss": 2.1364,
+ "loss": 2.1359,
  "step": 7500
  },
  {
  "epoch": 0.885058809828811,
- "grad_norm": 0.8623350262641907,
+ "grad_norm": 0.8626888990402222,
  "learning_rate": 7.94007638856753e-07,
- "loss": 2.1407,
+ "loss": 2.1405,
  "step": 7600
  },
  {
  "epoch": 0.8967043204844533,
- "grad_norm": 0.8134225010871887,
+ "grad_norm": 0.8167033791542053,
  "learning_rate": 6.429258342320677e-07,
- "loss": 2.1512,
+ "loss": 2.151,
  "step": 7700
  },
  {
  "epoch": 0.9083498311400955,
- "grad_norm": 0.8605279326438904,
+ "grad_norm": 0.855806827545166,
  "learning_rate": 5.073013214931377e-07,
- "loss": 2.1837,
+ "loss": 2.1835,
  "step": 7800
  },
  {
  "epoch": 0.9199953417957377,
- "grad_norm": 0.8916401267051697,
+ "grad_norm": 0.8845811486244202,
  "learning_rate": 3.8735814371615554e-07,
- "loss": 2.1317,
+ "loss": 2.1314,
  "step": 7900
  },
  {
  "epoch": 0.93164085245138,
- "grad_norm": 0.8076800107955933,
+ "grad_norm": 0.7994192838668823,
  "learning_rate": 2.8329443940849577e-07,
  "loss": 2.1349,
  "step": 8000
  },
  {
  "epoch": 0.9432863631070223,
- "grad_norm": 0.7975759506225586,
+ "grad_norm": 0.792962372303009,
  "learning_rate": 1.9528211519649341e-07,
  "loss": 2.094,
  "step": 8100
  },
  {
  "epoch": 0.9549318737626645,
- "grad_norm": 1.0410187244415283,
+ "grad_norm": 1.0241106748580933,
  "learning_rate": 1.234665618466202e-07,
- "loss": 2.1501,
+ "loss": 2.1502,
  "step": 8200
  },
  {
  "epoch": 0.9665773844183068,
- "grad_norm": 0.8723394274711609,
+ "grad_norm": 0.8663437366485596,
  "learning_rate": 6.796641408917671e-08,
  "loss": 2.0692,
  "step": 8300
  },
  {
  "epoch": 0.978222895073949,
- "grad_norm": 0.7656298279762268,
+ "grad_norm": 0.7561269402503967,
  "learning_rate": 2.8873354641258955e-08,
- "loss": 2.0986,
+ "loss": 2.0988,
  "step": 8400
  },
  {
  "epoch": 0.9898684057295912,
- "grad_norm": 1.4604876041412354,
- "learning_rate": 6.396496618182868e-09,
+ "grad_norm": 1.3960742950439453,
+ "learning_rate": 6.2519627527379836e-09,
  "loss": 2.1445,
  "step": 8500
  },
  {
  "epoch": 1.0,
- "eval_loss": 1.9266319274902344,
- "eval_runtime": 63.649,
- "eval_samples_per_second": 15.554,
- "eval_steps_per_second": 1.948,
+ "eval_loss": 1.9266563653945923,
+ "eval_runtime": 63.7186,
+ "eval_samples_per_second": 15.537,
+ "eval_steps_per_second": 1.946,
  "step": 8587
  },
  {
  "epoch": 1.0,
  "step": 8587,
  "total_flos": 1.5640093507584e+17,
- "train_loss": 2.1807824082961527,
- "train_runtime": 2484.8584,
- "train_samples_per_second": 6.911,
- "train_steps_per_second": 3.456
+ "train_loss": 2.1807946249230077,
+ "train_runtime": 2485.1722,
+ "train_samples_per_second": 6.91,
+ "train_steps_per_second": 3.455
  }
  ],
  "logging_steps": 100,