validation loss at iteration 1000 | lm loss value: 4.047245E+00 | lm loss PPL: 5.723952E+01 | validation loss at iteration 2000 | lm loss value: 2.185507E+00 | lm loss PPL: 8.895156E+00 | validation loss at iteration 3000 | lm loss value: 1.787350E+00 | lm loss PPL: 5.973599E+00 | validation loss at iteration 4000 | lm loss value: 1.590500E+00 | lm loss PPL: 4.906203E+00 | validation loss at iteration 5000 | lm loss value: 1.487892E+00 | lm loss PPL: 4.427752E+00 | validation loss at iteration 6000 | lm loss value: 1.415529E+00 | lm loss PPL: 4.118663E+00 | validation loss at iteration 7000 | lm loss value: 1.377813E+00 | lm loss PPL: 3.966218E+00 | validation loss at iteration 8000 | lm loss value: 1.326605E+00 | lm loss PPL: 3.768230E+00 | validation loss at iteration 9000 | lm loss value: 1.317780E+00 | lm loss PPL: 3.735120E+00 | validation loss at iteration 10000 | lm loss value: 1.284709E+00 | lm loss PPL: 3.613616E+00 | validation loss at iteration 11000 | lm loss value: 1.265982E+00 | lm loss PPL: 3.546573E+00 | validation loss at iteration 12000 | lm loss value: 1.234432E+00 | lm loss PPL: 3.436425E+00 | validation loss at iteration 13000 | lm loss value: 1.191512E+00 | lm loss PPL: 3.292055E+00 | validation loss at iteration 14000 | lm loss value: 1.219802E+00 | lm loss PPL: 3.386518E+00 | validation loss at iteration 15000 | lm loss value: 1.180529E+00 | lm loss PPL: 3.256095E+00 | validation loss at iteration 16000 | lm loss value: 1.180913E+00 | lm loss PPL: 3.257347E+00 | validation loss at iteration 17000 | lm loss value: 1.132317E+00 | lm loss PPL: 3.102838E+00 | validation loss at iteration 18000 | lm loss value: 1.105132E+00 | lm loss PPL: 3.019622E+00 | validation loss at iteration 19000 | lm loss value: 1.097962E+00 | lm loss PPL: 2.998051E+00 | validation loss at iteration 20000 | lm loss value: 1.152857E+00 | lm loss PPL: 3.167230E+00 | validation loss at iteration 21000 | lm loss value: 1.093314E+00 | lm loss PPL: 2.984148E+00 | validation loss at iteration 22000 | lm loss value: 1.071403E+00 | lm loss PPL: 2.919472E+00 | validation loss at iteration 23000 | lm loss value: 1.094638E+00 | lm loss PPL: 2.988099E+00 | validation loss at iteration 24000 | lm loss value: 1.113983E+00 | lm loss PPL: 3.046467E+00 | validation loss at iteration 25000 | lm loss value: 1.080887E+00 | lm loss PPL: 2.947294E+00 | validation loss at iteration 26000 | lm loss value: 1.085535E+00 | lm loss PPL: 2.961022E+00 | validation loss at iteration 27000 | lm loss value: 1.050126E+00 | lm loss PPL: 2.858011E+00 | validation loss at iteration 28000 | lm loss value: 1.059814E+00 | lm loss PPL: 2.885833E+00 | validation loss at iteration 29000 | lm loss value: 1.060280E+00 | lm loss PPL: 2.887180E+00 | validation loss at iteration 30000 | lm loss value: 1.036961E+00 | lm loss PPL: 2.820633E+00 | validation loss at iteration 31000 | lm loss value: 1.078923E+00 | lm loss PPL: 2.941509E+00 | validation loss at iteration 32000 | lm loss value: 1.018491E+00 | lm loss PPL: 2.769014E+00 | validation loss at iteration 33000 | lm loss value: 1.041267E+00 | lm loss PPL: 2.832805E+00 | validation loss at iteration 34000 | lm loss value: 1.000734E+00 | lm loss PPL: 2.720277E+00 | validation loss at iteration 35000 | lm loss value: 1.020639E+00 | lm loss PPL: 2.774968E+00 | validation loss at iteration 36000 | lm loss value: 1.056819E+00 | lm loss PPL: 2.877203E+00 | validation loss at iteration 37000 | lm loss value: 1.023292E+00 | lm loss PPL: 2.782340E+00 | validation loss at iteration 38000 | lm loss value: 1.043830E+00 | lm loss PPL: 2.840073E+00 | validation loss at iteration 39000 | lm loss value: 1.011889E+00 | lm loss PPL: 2.750793E+00 | validation loss at iteration 40000 | lm loss value: 1.013070E+00 | lm loss PPL: 2.754042E+00 | validation loss at iteration 41000 | lm loss value: 9.995539E-01 | lm loss PPL: 2.717069E+00 | validation loss at iteration 42000 | lm loss value: 9.886448E-01 | lm loss PPL: 2.687590E+00 | validation loss at iteration 43000 | lm loss value: 9.725732E-01 | lm loss PPL: 2.644741E+00 | validation loss at iteration 44000 | lm loss value: 9.968861E-01 | lm loss PPL: 2.709830E+00 | validation loss at iteration 45000 | lm loss value: 9.907036E-01 | lm loss PPL: 2.693129E+00 | validation loss at iteration 46000 | lm loss value: 9.670502E-01 | lm loss PPL: 2.630175E+00 | validation loss at iteration 47000 | lm loss value: 1.010807E+00 | lm loss PPL: 2.747816E+00 | validation loss at iteration 48000 | lm loss value: 9.514932E-01 | lm loss PPL: 2.589574E+00 | validation loss at iteration 49000 | lm loss value: 1.002607E+00 | lm loss PPL: 2.725377E+00 | validation loss at iteration 50000 | lm loss value: 9.625344E-01 | lm loss PPL: 2.618324E+00 | validation loss at iteration 51000 | lm loss value: 1.002735E+00 | lm loss PPL: 2.725726E+00 | validation loss at iteration 52000 | lm loss value: 9.746785E-01 | lm loss PPL: 2.650315E+00 | validation loss at iteration 53000 | lm loss value: 9.980860E-01 | lm loss PPL: 2.713084E+00 | validation loss at iteration 54000 | lm loss value: 9.323012E-01 | lm loss PPL: 2.540348E+00 | validation loss at iteration 55000 | lm loss value: 1.007919E+00 | lm loss PPL: 2.739894E+00 | validation loss at iteration 56000 | lm loss value: 9.554288E-01 | lm loss PPL: 2.599785E+00 | validation loss at iteration 57000 | lm loss value: 9.433770E-01 | lm loss PPL: 2.568641E+00 | validation loss at iteration 58000 | lm loss value: 9.538841E-01 | lm loss PPL: 2.595772E+00 | validation loss at iteration 59000 | lm loss value: 9.640460E-01 | lm loss PPL: 2.622285E+00 | validation loss at iteration 60000 | lm loss value: 9.728367E-01 | lm loss PPL: 2.645438E+00 | validation loss at iteration 61000 | lm loss value: 9.297212E-01 | lm loss PPL: 2.533803E+00 | validation loss at iteration 62000 | lm loss value: 9.485184E-01 | lm loss PPL: 2.581881E+00 | validation loss at iteration 63000 | lm loss value: 9.281455E-01 | lm loss PPL: 2.529813E+00 | validation loss at iteration 64000 | lm loss value: 9.680632E-01 | lm loss PPL: 2.632840E+00 | validation loss at iteration 65000 | lm loss value: 9.320363E-01 | lm loss PPL: 2.539676E+00 | validation loss at iteration 66000 | lm loss value: 9.814236E-01 | lm loss PPL: 2.668252E+00 | validation loss at iteration 67000 | lm loss value: 9.774494E-01 | lm loss PPL: 2.657669E+00 | validation loss at iteration 68000 | lm loss value: 9.262792E-01 | lm loss PPL: 2.525096E+00 | validation loss at iteration 69000 | lm loss value: 1.000724E+00 | lm loss PPL: 2.720250E+00 | validation loss at iteration 70000 | lm loss value: 9.246699E-01 | lm loss PPL: 2.521036E+00 | validation loss at iteration 71000 | lm loss value: 9.843366E-01 | lm loss PPL: 2.676036E+00 | validation loss at iteration 72000 | lm loss value: 9.372199E-01 | lm loss PPL: 2.552874E+00 | validation loss at iteration 73000 | lm loss value: 9.305505E-01 | lm loss PPL: 2.535905E+00 | validation loss at iteration 74000 | lm loss value: 9.204554E-01 | lm loss PPL: 2.510433E+00 | validation loss at iteration 75000 | lm loss value: 9.237205E-01 | lm loss PPL: 2.518644E+00 | validation loss at iteration 76000 | lm loss value: 9.331719E-01 | lm loss PPL: 2.542561E+00 | validation loss at iteration 77000 | lm loss value: 9.575729E-01 | lm loss PPL: 2.605365E+00 | validation loss at iteration 78000 | lm loss value: 9.410254E-01 | lm loss PPL: 2.562608E+00 | validation loss at iteration 79000 | lm loss value: 9.343886E-01 | lm loss PPL: 2.545657E+00 | validation loss at iteration 80000 | lm loss value: 9.202853E-01 | lm loss PPL: 2.510007E+00 | validation loss at iteration 81000 | lm loss value: 9.674646E-01 | lm loss PPL: 2.631265E+00 | validation loss at iteration 82000 | lm loss value: 9.274186E-01 | lm loss PPL: 2.527975E+00 | validation loss at iteration 83000 | lm loss value: 9.359873E-01 | lm loss PPL: 2.549730E+00 | validation loss at iteration 84000 | lm loss value: 9.343195E-01 | lm loss PPL: 2.545481E+00 | validation loss at iteration 85000 | lm loss value: 9.488503E-01 | lm loss PPL: 2.582739E+00 | validation loss at iteration 86000 | lm loss value: 9.489172E-01 | lm loss PPL: 2.582911E+00 | validation loss at iteration 87000 | lm loss value: 9.396169E-01 | lm loss PPL: 2.559001E+00 | validation loss at iteration 88000 | lm loss value: 9.078221E-01 | lm loss PPL: 2.478918E+00 | validation loss at iteration 89000 | lm loss value: 9.092955E-01 | lm loss PPL: 2.482573E+00 | validation loss at iteration 90000 | lm loss value: 9.106666E-01 | lm loss PPL: 2.485979E+00 | validation loss at iteration 91000 | lm loss value: 9.367995E-01 | lm loss PPL: 2.551801E+00 | validation loss at iteration 92000 | lm loss value: 8.820238E-01 | lm loss PPL: 2.415784E+00 | validation loss at iteration 93000 | lm loss value: 8.794007E-01 | lm loss PPL: 2.409455E+00 | validation loss at iteration 94000 | lm loss value: 8.992079E-01 | lm loss PPL: 2.457656E+00 | validation loss at iteration 95000 | lm loss value: 9.000293E-01 | lm loss PPL: 2.459675E+00 | validation loss at iteration 96000 | lm loss value: 9.076358E-01 | lm loss PPL: 2.478456E+00 | validation loss at iteration 97000 | lm loss value: 9.157909E-01 | lm loss PPL: 2.498751E+00 | validation loss at iteration 98000 | lm loss value: 9.177882E-01 | lm loss PPL: 2.503746E+00 | validation loss at iteration 99000 | lm loss value: 8.743231E-01 | lm loss PPL: 2.397252E+00 | validation loss at iteration 100000 | lm loss value: 9.003542E-01 | lm loss PPL: 2.460474E+00 | validation loss at iteration 101000 | lm loss value: 9.037837E-01 | lm loss PPL: 2.468927E+00 | validation loss at iteration 102000 | lm loss value: 9.110614E-01 | lm loss PPL: 2.486961E+00 | validation loss at iteration 103000 | lm loss value: 8.854871E-01 | lm loss PPL: 2.424165E+00 | validation loss at iteration 104000 | lm loss value: 8.715252E-01 | lm loss PPL: 2.390554E+00 | validation loss at iteration 105000 | lm loss value: 8.881654E-01 | lm loss PPL: 2.430666E+00 | validation loss at iteration 106000 | lm loss value: 8.831617E-01 | lm loss PPL: 2.418534E+00 | validation loss at iteration 107000 | lm loss value: 8.493679E-01 | lm loss PPL: 2.338168E+00 | validation loss at iteration 108000 | lm loss value: 8.896254E-01 | lm loss PPL: 2.434218E+00 | validation loss at iteration 109000 | lm loss value: 8.991036E-01 | lm loss PPL: 2.457399E+00 | validation loss at iteration 110000 | lm loss value: 8.493870E-01 | lm loss PPL: 2.338213E+00 | validation loss at iteration 111000 | lm loss value: 9.029331E-01 | lm loss PPL: 2.466828E+00 | validation loss at iteration 112000 | lm loss value: 9.135931E-01 | lm loss PPL: 2.493265E+00 | validation loss at iteration 113000 | lm loss value: 8.464906E-01 | lm loss PPL: 2.331450E+00 | validation loss at iteration 114000 | lm loss value: 8.688712E-01 | lm loss PPL: 2.384218E+00 | validation loss at iteration 115000 | lm loss value: 8.965376E-01 | lm loss PPL: 2.451102E+00 | validation loss at iteration 116000 | lm loss value: 8.591920E-01 | lm loss PPL: 2.361252E+00 | validation loss at iteration 117000 | lm loss value: 8.762812E-01 | lm loss PPL: 2.401951E+00 | validation loss at iteration 118000 | lm loss value: 8.955140E-01 | lm loss PPL: 2.448594E+00 | validation loss at iteration 119000 | lm loss value: 8.566702E-01 | lm loss PPL: 2.355305E+00 | validation loss at iteration 120000 | lm loss value: 9.021574E-01 | lm loss PPL: 2.464915E+00 | validation loss at iteration 121000 | lm loss value: 8.607898E-01 | lm loss PPL: 2.365028E+00 | validation loss at iteration 122000 | lm loss value: 8.921415E-01 | lm loss PPL: 2.440350E+00 | validation loss at iteration 123000 | lm loss value: 8.685483E-01 | lm loss PPL: 2.383448E+00 | validation loss at iteration 124000 | lm loss value: 8.563555E-01 | lm loss PPL: 2.354564E+00 | validation loss at iteration 125000 | lm loss value: 8.691928E-01 | lm loss PPL: 2.384985E+00 | validation loss at iteration 126000 | lm loss value: 8.823726E-01 | lm loss PPL: 2.416626E+00 | validation loss at iteration 127000 | lm loss value: 8.888642E-01 | lm loss PPL: 2.432365E+00 | validation loss at iteration 128000 | lm loss value: 8.553858E-01 | lm loss PPL: 2.352282E+00 | validation loss at iteration 129000 | lm loss value: 8.803064E-01 | lm loss PPL: 2.411638E+00 | validation loss at iteration 130000 | lm loss value: 8.815665E-01 | lm loss PPL: 2.414679E+00 | validation loss at iteration 131000 | lm loss value: 8.972144E-01 | lm loss PPL: 2.452761E+00 | validation loss at iteration 132000 | lm loss value: 8.760530E-01 | lm loss PPL: 2.401403E+00 | validation loss at iteration 133000 | lm loss value: 8.767945E-01 | lm loss PPL: 2.403184E+00 | validation loss at iteration 134000 | lm loss value: 8.598248E-01 | lm loss PPL: 2.362747E+00 | validation loss at iteration 135000 | lm loss value: 8.532380E-01 | lm loss PPL: 2.347235E+00 | validation loss at iteration 136000 | lm loss value: 8.939791E-01 | lm loss PPL: 2.444839E+00 | validation loss at iteration 137000 | lm loss value: 8.720918E-01 | lm loss PPL: 2.391909E+00 | validation loss at iteration 138000 | lm loss value: 8.366207E-01 | lm loss PPL: 2.308553E+00 | validation loss at iteration 139000 | lm loss value: 8.419310E-01 | lm loss PPL: 2.320844E+00 | validation loss at iteration 140000 | lm loss value: 8.422236E-01 | lm loss PPL: 2.321523E+00 | validation loss at iteration 141000 | lm loss value: 8.807864E-01 | lm loss PPL: 2.412796E+00 | validation loss at iteration 142000 | lm loss value: 8.914757E-01 | lm loss PPL: 2.438726E+00 | validation loss at iteration 143000 | lm loss value: 8.452522E-01 | lm loss PPL: 2.328565E+00 | validation loss at iteration 144000 | lm loss value: 8.482807E-01 | lm loss PPL: 2.335628E+00 | validation loss at iteration 145000 | lm loss value: 9.038057E-01 | lm loss PPL: 2.468982E+00 | validation loss at iteration 146000 | lm loss value: 8.714234E-01 | lm loss PPL: 2.390311E+00 | validation loss at iteration 147000 | lm loss value: 8.717233E-01 | lm loss PPL: 2.391028E+00 | validation loss at iteration 148000 | lm loss value: 8.682492E-01 | lm loss PPL: 2.382736E+00 | validation loss at iteration 149000 | lm loss value: 8.238327E-01 | lm loss PPL: 2.279219E+00 | validation loss at iteration 150000 | lm loss value: 8.812104E-01 | lm loss PPL: 2.413820E+00 | validation loss at iteration 151000 | lm loss value: 8.554614E-01 | lm loss PPL: 2.352460E+00 | validation loss at iteration 152000 | lm loss value: 8.718519E-01 | lm loss PPL: 2.391335E+00 | validation loss at iteration 153000 | lm loss value: 8.499790E-01 | lm loss PPL: 2.339598E+00 | validation loss at iteration 154000 | lm loss value: 8.652266E-01 | lm loss PPL: 2.375544E+00 | validation loss at iteration 155000 | lm loss value: 8.305758E-01 | lm loss PPL: 2.294640E+00 | validation loss at iteration 156000 | lm loss value: 8.519611E-01 | lm loss PPL: 2.344240E+00 | validation loss at iteration 157000 | lm loss value: 8.321554E-01 | lm loss PPL: 2.298267E+00 | validation loss at iteration 158000 | lm loss value: 8.338564E-01 | lm loss PPL: 2.302180E+00 | validation loss at iteration 159000 | lm loss value: 8.424427E-01 | lm loss PPL: 2.322032E+00 | validation loss at iteration 160000 | lm loss value: 8.711592E-01 | lm loss PPL: 2.389679E+00 | validation loss at iteration 161000 | lm loss value: 8.498114E-01 | lm loss PPL: 2.339206E+00 | validation loss at iteration 162000 | lm loss value: 8.293511E-01 | lm loss PPL: 2.291831E+00 | validation loss at iteration 163000 | lm loss value: 8.704237E-01 | lm loss PPL: 2.387922E+00 | validation loss at iteration 164000 | lm loss value: 8.688011E-01 | lm loss PPL: 2.384051E+00 | validation loss at iteration 165000 | lm loss value: 8.645583E-01 | lm loss PPL: 2.373957E+00 | validation loss at iteration 166000 | lm loss value: 8.356828E-01 | lm loss PPL: 2.306388E+00 | validation loss at iteration 167000 | lm loss value: 8.408365E-01 | lm loss PPL: 2.318305E+00 | validation loss at iteration 168000 | lm loss value: 8.476049E-01 | lm loss PPL: 2.334050E+00 | validation loss at iteration 169000 | lm loss value: 8.857953E-01 | lm loss PPL: 2.424912E+00 | validation loss at iteration 170000 | lm loss value: 8.577092E-01 | lm loss PPL: 2.357753E+00 | validation loss at iteration 171000 | lm loss value: 8.361296E-01 | lm loss PPL: 2.307419E+00 | validation loss at iteration 172000 | lm loss value: 8.309502E-01 | lm loss PPL: 2.295499E+00 | validation loss at iteration 173000 | lm loss value: 8.516952E-01 | lm loss PPL: 2.343616E+00 | validation loss at iteration 174000 | lm loss value: 7.988870E-01 | lm loss PPL: 2.223065E+00 | validation loss at iteration 175000 | lm loss value: 8.682228E-01 | lm loss PPL: 2.382673E+00 | validation loss at iteration 176000 | lm loss value: 8.417258E-01 | lm loss PPL: 2.320368E+00 | validation loss at iteration 177000 | lm loss value: 8.676408E-01 | lm loss PPL: 2.381286E+00 | validation loss at iteration 178000 | lm loss value: 7.898926E-01 | lm loss PPL: 2.203160E+00 | validation loss at iteration 179000 | lm loss value: 8.614833E-01 | lm loss PPL: 2.366669E+00 | validation loss at iteration 180000 | lm loss value: 8.447755E-01 | lm loss PPL: 2.327455E+00 | validation loss at iteration 181000 | lm loss value: 8.293174E-01 | lm loss PPL: 2.291754E+00 | validation loss at iteration 182000 | lm loss value: 8.206556E-01 | lm loss PPL: 2.271989E+00 | validation loss at iteration 183000 | lm loss value: 8.457853E-01 | lm loss PPL: 2.329807E+00 | validation loss at iteration 184000 | lm loss value: 8.312611E-01 | lm loss PPL: 2.296213E+00 | validation loss at iteration 185000 | lm loss value: 8.487062E-01 | lm loss PPL: 2.336622E+00 | validation loss at iteration 186000 | lm loss value: 8.379432E-01 | lm loss PPL: 2.311608E+00 | validation loss at iteration 187000 | lm loss value: 8.376940E-01 | lm loss PPL: 2.311032E+00 | validation loss at iteration 188000 | lm loss value: 8.518372E-01 | lm loss PPL: 2.343949E+00 | validation loss at iteration 189000 | lm loss value: 8.531367E-01 | lm loss PPL: 2.346997E+00 | validation loss at iteration 190000 | lm loss value: 8.561355E-01 | lm loss PPL: 2.354046E+00 | validation loss at iteration 191000 | lm loss value: 8.319536E-01 | lm loss PPL: 2.297803E+00 | validation loss at iteration 192000 | lm loss value: 8.302749E-01 | lm loss PPL: 2.293949E+00 | validation loss at iteration 193000 | lm loss value: 8.469012E-01 | lm loss PPL: 2.332408E+00 | validation loss at iteration 194000 | lm loss value: 8.401285E-01 | lm loss PPL: 2.316665E+00 | validation loss at iteration 195000 | lm loss value: 8.577501E-01 | lm loss PPL: 2.357850E+00 | validation loss at iteration 196000 | lm loss value: 8.208975E-01 | lm loss PPL: 2.272539E+00 | validation loss at iteration 197000 | lm loss value: 8.341989E-01 | lm loss PPL: 2.302968E+00 | validation loss at iteration 198000 | lm loss value: 8.299168E-01 | lm loss PPL: 2.293128E+00 | validation loss at iteration 199000 | lm loss value: 8.579417E-01 | lm loss PPL: 2.358302E+00 | validation loss at iteration 200000 | lm loss value: 8.325400E-01 | lm loss PPL: 2.299151E+00 | validation loss at iteration 201000 | lm loss value: 8.019601E-01 | lm loss PPL: 2.229908E+00 | validation loss at iteration 202000 | lm loss value: 8.153200E-01 | lm loss PPL: 2.259899E+00 | validation loss at iteration 203000 | lm loss value: 8.423535E-01 | lm loss PPL: 2.321825E+00 | validation loss at iteration 204000 | lm loss value: 8.340357E-01 | lm loss PPL: 2.302593E+00 | validation loss at iteration 205000 | lm loss value: 8.332598E-01 | lm loss PPL: 2.300807E+00 | validation loss at iteration 206000 | lm loss value: 7.922948E-01 | lm loss PPL: 2.208459E+00 | validation loss at iteration 207000 | lm loss value: 8.265033E-01 | lm loss PPL: 2.285314E+00 | validation loss at iteration 208000 | lm loss value: 8.677109E-01 | lm loss PPL: 2.381453E+00 | validation loss at iteration 209000 | lm loss value: 8.215567E-01 | lm loss PPL: 2.274037E+00 | validation loss at iteration 210000 | lm loss value: 8.438946E-01 | lm loss PPL: 2.325406E+00 | validation loss at iteration 211000 | lm loss value: 8.155533E-01 | lm loss PPL: 2.260426E+00 | validation loss at iteration 212000 | lm loss value: 7.956911E-01 | lm loss PPL: 2.215972E+00 | validation loss at iteration 213000 | lm loss value: 8.311703E-01 | lm loss PPL: 2.296004E+00 | validation loss at iteration 214000 | lm loss value: 7.970093E-01 | lm loss PPL: 2.218895E+00 | validation loss at iteration 215000 | lm loss value: 8.351642E-01 | lm loss PPL: 2.305193E+00 | validation loss at iteration 216000 | lm loss value: 8.030192E-01 | lm loss PPL: 2.232270E+00 | validation loss at iteration 217000 | lm loss value: 8.183990E-01 | lm loss PPL: 2.266868E+00 | validation loss at iteration 218000 | lm loss value: 8.007969E-01 | lm loss PPL: 2.227315E+00 | validation loss at iteration 219000 | lm loss value: 8.362185E-01 | lm loss PPL: 2.307624E+00 | validation loss at iteration 220000 | lm loss value: 8.252020E-01 | lm loss PPL: 2.282342E+00 | validation loss at iteration 221000 | lm loss value: 8.064855E-01 | lm loss PPL: 2.240022E+00 | validation loss at iteration 222000 | lm loss value: 7.977690E-01 | lm loss PPL: 2.220581E+00 | validation loss at iteration 223000 | lm loss value: 8.034332E-01 | lm loss PPL: 2.233195E+00 | validation loss at iteration 224000 | lm loss value: 8.057780E-01 | lm loss PPL: 2.238437E+00 | validation loss at iteration 225000 | lm loss value: 8.205453E-01 | lm loss PPL: 2.271738E+00 | validation loss at iteration 226000 | lm loss value: 8.430458E-01 | lm loss PPL: 2.323433E+00 | validation loss at iteration 227000 | lm loss value: 8.513870E-01 | lm loss PPL: 2.342894E+00 | validation loss at iteration 228000 | lm loss value: 7.814319E-01 | lm loss PPL: 2.184598E+00 | validation loss at iteration 229000 | lm loss value: 8.306801E-01 | lm loss PPL: 2.294879E+00 | validation loss at iteration 230000 | lm loss value: 8.265758E-01 | lm loss PPL: 2.285479E+00 | validation loss at iteration 231000 | lm loss value: 8.270227E-01 | lm loss PPL: 2.286501E+00 | validation loss at iteration 232000 | lm loss value: 7.936721E-01 | lm loss PPL: 2.211502E+00 | validation loss at iteration 233000 | lm loss value: 8.039231E-01 | lm loss PPL: 2.234289E+00 | validation loss at iteration 234000 | lm loss value: 8.168525E-01 | lm loss PPL: 2.263365E+00 | validation loss at iteration 235000 | lm loss value: 8.152491E-01 | lm loss PPL: 2.259739E+00 | validation loss at iteration 236000 | lm loss value: 7.900113E-01 | lm loss PPL: 2.203421E+00 | validation loss at iteration 237000 | lm loss value: 7.972509E-01 | lm loss PPL: 2.219431E+00 | validation loss at iteration 238000 | lm loss value: 7.963144E-01 | lm loss PPL: 2.217354E+00 | validation loss at iteration 239000 | lm loss value: 7.958148E-01 | lm loss PPL: 2.216246E+00 | validation loss at iteration 240000 | lm loss value: 7.948017E-01 | lm loss PPL: 2.214002E+00 | validation loss at iteration 241000 | lm loss value: 8.044324E-01 | lm loss PPL: 2.235427E+00 | validation loss at iteration 242000 | lm loss value: 8.443128E-01 | lm loss PPL: 2.326379E+00 | validation loss at iteration 243000 | lm loss value: 8.121519E-01 | lm loss PPL: 2.252750E+00 | validation loss at iteration 244000 | lm loss value: 8.027102E-01 | lm loss PPL: 2.231581E+00 | validation loss at iteration 245000 | lm loss value: 8.176475E-01 | lm loss PPL: 2.265165E+00 | validation loss at iteration 246000 | lm loss value: 8.326839E-01 | lm loss PPL: 2.299482E+00 | validation loss at iteration 247000 | lm loss value: 7.904744E-01 | lm loss PPL: 2.204442E+00 | validation loss at iteration 248000 | lm loss value: 7.863216E-01 | lm loss PPL: 2.195306E+00 | validation loss at iteration 249000 | lm loss value: 8.280783E-01 | lm loss PPL: 2.288916E+00 | validation loss at iteration 250000 | lm loss value: 8.046969E-01 | lm loss PPL: 2.236019E+00 | validation loss at iteration 251000 | lm loss value: 8.181589E-01 | lm loss PPL: 2.266324E+00 | validation loss at iteration 252000 | lm loss value: 8.048475E-01 | lm loss PPL: 2.236356E+00 | validation loss at iteration 253000 | lm loss value: 8.045262E-01 | lm loss PPL: 2.235637E+00 | validation loss at iteration 254000 | lm loss value: 8.145036E-01 | lm loss PPL: 2.258055E+00 | validation loss at iteration 255000 | lm loss value: 7.766764E-01 | lm loss PPL: 2.174234E+00 | validation loss at iteration 256000 | lm loss value: 8.553532E-01 | lm loss PPL: 2.352205E+00 | validation loss at iteration 257000 | lm loss value: 8.286043E-01 | lm loss PPL: 2.290120E+00 | validation loss at iteration 258000 | lm loss value: 8.128422E-01 | lm loss PPL: 2.254306E+00 | validation loss at iteration 259000 | lm loss value: 8.026119E-01 | lm loss PPL: 2.231362E+00 | validation loss at iteration 260000 | lm loss value: 8.201897E-01 | lm loss PPL: 2.270930E+00 | validation loss at iteration 261000 | lm loss value: 8.238492E-01 | lm loss PPL: 2.279256E+00 | validation loss at iteration 262000 | lm loss value: 7.662678E-01 | lm loss PPL: 2.151721E+00 | validation loss at iteration 263000 | lm loss value: 8.229710E-01 | lm loss PPL: 2.277256E+00 | validation loss at iteration 264000 | lm loss value: 7.985713E-01 | lm loss PPL: 2.222364E+00 | validation loss at iteration 265000 | lm loss value: 8.242220E-01 | lm loss PPL: 2.280106E+00 | validation loss at iteration 266000 | lm loss value: 8.084802E-01 | lm loss PPL: 2.244494E+00 | validation loss at iteration 267000 | lm loss value: 8.005447E-01 | lm loss PPL: 2.226754E+00 | validation loss at iteration 268000 | lm loss value: 8.019740E-01 | lm loss PPL: 2.229938E+00 | validation loss at iteration 269000 | lm loss value: 7.903088E-01 | lm loss PPL: 2.204077E+00 | validation loss at iteration 270000 | lm loss value: 7.973325E-01 | lm loss PPL: 2.219612E+00 | validation loss at iteration 271000 | lm loss value: 8.002435E-01 | lm loss PPL: 2.226083E+00 | validation loss at iteration 272000 | lm loss value: 7.713711E-01 | lm loss PPL: 2.162729E+00 | validation loss at iteration 273000 | lm loss value: 7.985616E-01 | lm loss PPL: 2.222342E+00 | validation loss at iteration 274000 | lm loss value: 8.163761E-01 | lm loss PPL: 2.262287E+00 | validation loss at iteration 275000 | lm loss value: 8.009943E-01 | lm loss PPL: 2.227755E+00 | validation loss at iteration 276000 | lm loss value: 8.184980E-01 | lm loss PPL: 2.267092E+00 | validation loss at iteration 277000 | lm loss value: 7.876898E-01 | lm loss PPL: 2.198312E+00 | validation loss at iteration 278000 | lm loss value: 8.265983E-01 | lm loss PPL: 2.285531E+00 | validation loss at iteration 279000 | lm loss value: 8.048902E-01 | lm loss PPL: 2.236451E+00 | validation loss at iteration 280000 | lm loss value: 7.798821E-01 | lm loss PPL: 2.181215E+00 | validation loss at iteration 281000 | lm loss value: 8.136095E-01 | lm loss PPL: 2.256037E+00 | validation loss at iteration 282000 | lm loss value: 7.977873E-01 | lm loss PPL: 2.220622E+00 | validation loss at iteration 283000 | lm loss value: 7.911233E-01 | lm loss PPL: 2.205873E+00 | validation loss at iteration 284000 | lm loss value: 7.918561E-01 | lm loss PPL: 2.207490E+00 | validation loss at iteration 285000 | lm loss value: 8.146284E-01 | lm loss PPL: 2.258336E+00 | validation loss at iteration 286000 | lm loss value: 8.027362E-01 | lm loss PPL: 2.231639E+00 | validation loss at iteration 287000 | lm loss value: 7.775673E-01 | lm loss PPL: 2.176172E+00 | validation loss at iteration 288000 | lm loss value: 8.026530E-01 | lm loss PPL: 2.231453E+00 | validation loss at iteration 289000 | lm loss value: 7.976859E-01 | lm loss PPL: 2.220397E+00 | validation loss at iteration 290000 | lm loss value: 8.045716E-01 | lm loss PPL: 2.235739E+00 | validation loss at iteration 291000 | lm loss value: 8.153024E-01 | lm loss PPL: 2.259859E+00 | validation loss at iteration 292000 | lm loss value: 7.349713E-01 | lm loss PPL: 2.085422E+00 | validation loss at iteration 293000 | lm loss value: 7.722172E-01 | lm loss PPL: 2.164560E+00 | validation loss at iteration 294000 | lm loss value: 7.996628E-01 | lm loss PPL: 2.224791E+00 | validation loss at iteration 295000 | lm loss value: 7.946648E-01 | lm loss PPL: 2.213699E+00 | validation loss at iteration 296000 | lm loss value: 7.913095E-01 | lm loss PPL: 2.206284E+00 | validation loss at iteration 297000 | lm loss value: 7.690780E-01 | lm loss PPL: 2.157776E+00 | validation loss at iteration 298000 | lm loss value: 8.013525E-01 | lm loss PPL: 2.228553E+00 | validation loss at iteration 299000 | lm loss value: 7.648926E-01 | lm loss PPL: 2.148764E+00 | validation loss at iteration 300000 | lm loss value: 7.856612E-01 | lm loss PPL: 2.193857E+00 | validation loss at iteration 301000 | lm loss value: 8.112941E-01 | lm loss PPL: 2.250819E+00 | validation loss at iteration 302000 | lm loss value: 7.772269E-01 | lm loss PPL: 2.175431E+00 | validation loss at iteration 303000 | lm loss value: 8.127834E-01 | lm loss PPL: 2.254174E+00 | validation loss at iteration 304000 | lm loss value: 7.857931E-01 | lm loss PPL: 2.194146E+00 | validation loss at iteration 305000 | lm loss value: 7.785457E-01 | lm loss PPL: 2.178302E+00 | validation loss at iteration 306000 | lm loss value: 7.664601E-01 | lm loss PPL: 2.152134E+00 | validation loss at iteration 307000 | lm loss value: 7.973075E-01 | lm loss PPL: 2.219557E+00 | validation loss at iteration 308000 | lm loss value: 7.548591E-01 | lm loss PPL: 2.127312E+00 | validation loss at iteration 309000 | lm loss value: 8.145952E-01 | lm loss PPL: 2.258261E+00 | validation loss at iteration 310000 | lm loss value: 7.848760E-01 | lm loss PPL: 2.192135E+00 | validation loss at iteration 311000 | lm loss value: 7.729430E-01 | lm loss PPL: 2.166132E+00 | validation loss at iteration 312000 | lm loss value: 7.926192E-01 | lm loss PPL: 2.209175E+00 | validation loss at iteration 313000 | lm loss value: 8.150043E-01 | lm loss PPL: 2.259185E+00 | validation loss at iteration 314000 | lm loss value: 7.902986E-01 | lm loss PPL: 2.204054E+00 | validation loss at iteration 315000 | lm loss value: 7.947099E-01 | lm loss PPL: 2.213799E+00 | validation loss at iteration 316000 | lm loss value: 7.765861E-01 | lm loss PPL: 2.174038E+00 | validation loss at iteration 317000 | lm loss value: 7.842644E-01 | lm loss PPL: 2.190795E+00 | validation loss at iteration 318000 | lm loss value: 7.421570E-01 | lm loss PPL: 2.100461E+00 | validation loss at iteration 319000 | lm loss value: 8.035652E-01 | lm loss PPL: 2.233490E+00 | validation loss at iteration 320000 | lm loss value: 7.848793E-01 | lm loss PPL: 2.192142E+00 | validation loss at iteration 321000 | lm loss value: 8.008446E-01 | lm loss PPL: 2.227421E+00 | validation loss at iteration 322000 | lm loss value: 7.766518E-01 | lm loss PPL: 2.174180E+00 | validation loss at iteration 323000 | lm loss value: 8.176198E-01 | lm loss PPL: 2.265102E+00 | validation loss at iteration 324000 | lm loss value: 7.732455E-01 | lm loss PPL: 2.166787E+00 | validation loss at iteration 325000 | lm loss value: 7.820513E-01 | lm loss PPL: 2.185952E+00 | validation loss at iteration 326000 | lm loss value: 7.853588E-01 | lm loss PPL: 2.193194E+00 | validation loss at iteration 327000 | lm loss value: 8.187327E-01 | lm loss PPL: 2.267624E+00 | validation loss at iteration 328000 | lm loss value: 8.106675E-01 | lm loss PPL: 2.249409E+00 | validation loss at iteration 329000 | lm loss value: 8.011547E-01 | lm loss PPL: 2.228112E+00 | validation loss at iteration 330000 | lm loss value: 8.073539E-01 | lm loss PPL: 2.241968E+00 | validation loss at iteration 331000 | lm loss value: 7.967555E-01 | lm loss PPL: 2.218332E+00 | validation loss at iteration 332000 | lm loss value: 7.752457E-01 | lm loss PPL: 2.171125E+00 | validation loss at iteration 333000 | lm loss value: 7.679848E-01 | lm loss PPL: 2.155418E+00 | validation loss at iteration 334000 | lm loss value: 8.151882E-01 | lm loss PPL: 2.259601E+00 | validation loss at iteration 335000 | lm loss value: 8.059840E-01 | lm loss PPL: 2.238899E+00 | validation loss at iteration 336000 | lm loss value: 8.077911E-01 | lm loss PPL: 2.242948E+00 | validation loss at iteration 337000 | lm loss value: 7.709336E-01 | lm loss PPL: 2.161784E+00 | validation loss at iteration 338000 | lm loss value: 7.868972E-01 | lm loss PPL: 2.196570E+00 | validation loss at iteration 339000 | lm loss value: 7.943464E-01 | lm loss PPL: 2.212994E+00 | validation loss at iteration 340000 | lm loss value: 8.109882E-01 | lm loss PPL: 2.250131E+00 | validation loss at iteration 341000 | lm loss value: 7.671905E-01 | lm loss PPL: 2.153707E+00 | validation loss at iteration 342000 | lm loss value: 7.837185E-01 | lm loss PPL: 2.189599E+00 | validation loss at iteration 343000 | lm loss value: 7.876207E-01 | lm loss PPL: 2.198160E+00 | validation loss at iteration 344000 | lm loss value: 7.512331E-01 | lm loss PPL: 2.119612E+00 | validation loss at iteration 345000 | lm loss value: 8.105088E-01 | lm loss PPL: 2.249052E+00 | validation loss at iteration 346000 | lm loss value: 7.987944E-01 | lm loss PPL: 2.222860E+00 | validation loss at iteration 347000 | lm loss value: 8.104767E-01 | lm loss PPL: 2.248980E+00 | validation loss at iteration 348000 | lm loss value: 7.909063E-01 | lm loss PPL: 2.205394E+00 | validation loss at iteration 349000 | lm loss value: 8.057012E-01 | lm loss PPL: 2.238265E+00 | validation loss at iteration 350000 | lm loss value: 7.901012E-01 | lm loss PPL: 2.203619E+00 | validation loss at iteration 351000 | lm loss value: 7.870669E-01 | lm loss PPL: 2.196943E+00 | validation loss at iteration 352000 | lm loss value: 8.009290E-01 | lm loss PPL: 2.227609E+00 | validation loss at iteration 353000 | lm loss value: 7.871494E-01 | lm loss PPL: 2.197124E+00 | validation loss at iteration 354000 | lm loss value: 7.847652E-01 | lm loss PPL: 2.191892E+00 | validation loss at iteration 355000 | lm loss value: 7.835929E-01 | lm loss PPL: 2.189324E+00 | validation loss at iteration 356000 | lm loss value: 7.964563E-01 | lm loss PPL: 2.217668E+00 | validation loss at iteration 357000 | lm loss value: 8.025600E-01 | lm loss PPL: 2.231246E+00 | validation loss at iteration 358000 | lm loss value: 7.923281E-01 | lm loss PPL: 2.208532E+00 | validation loss at iteration 359000 | lm loss value: 7.879450E-01 | lm loss PPL: 2.198873E+00 | validation loss at iteration 360000 | lm loss value: 7.714848E-01 | lm loss PPL: 2.162975E+00 | validation loss at iteration 361000 | lm loss value: 8.127524E-01 | lm loss PPL: 2.254104E+00 | validation loss at iteration 362000 | lm loss value: 7.760996E-01 | lm loss PPL: 2.172980E+00 | validation loss at iteration 363000 | lm loss value: 8.194810E-01 | lm loss PPL: 2.269322E+00 | validation loss at iteration 364000 | lm loss value: 8.117533E-01 | lm loss PPL: 2.251853E+00 | validation loss at iteration 365000 | lm loss value: 7.660697E-01 | lm loss PPL: 2.151294E+00 | validation loss at iteration 366000 | lm loss value: 7.699630E-01 | lm loss PPL: 2.159686E+00 | validation loss at iteration 367000 | lm loss value: 8.171715E-01 | lm loss PPL: 2.264087E+00 | validation loss at iteration 368000 | lm loss value: 8.024047E-01 | lm loss PPL: 2.230899E+00 | validation loss at iteration 369000 | lm loss value: 7.770532E-01 | lm loss PPL: 2.175053E+00 | validation loss at iteration 370000 | lm loss value: 7.698528E-01 | lm loss PPL: 2.159448E+00 | validation loss at iteration 371000 | lm loss value: 7.806970E-01 | lm loss PPL: 2.182993E+00 | validation loss at iteration 372000 | lm loss value: 7.587687E-01 | lm loss PPL: 2.135645E+00 | validation loss at iteration 373000 | lm loss value: 8.126971E-01 | lm loss PPL: 2.253979E+00 | validation loss at iteration 374000 | lm loss value: 7.822093E-01 | lm loss PPL: 2.186297E+00 | validation loss at iteration 375000 | lm loss value: 7.951312E-01 | lm loss PPL: 2.214732E+00 | validation loss at iteration 376000 | lm loss value: 8.060457E-01 | lm loss PPL: 2.239037E+00 | validation loss at iteration 377000 | lm loss value: 8.027647E-01 | lm loss PPL: 2.231702E+00 | validation loss at iteration 378000 | lm loss value: 7.751684E-01 | lm loss PPL: 2.170958E+00 | validation loss at iteration 379000 | lm loss value: 7.692026E-01 | lm loss PPL: 2.158045E+00 | validation loss at iteration 380000 | lm loss value: 7.753606E-01 | lm loss PPL: 2.171375E+00 |