Exp B B1: gpt2_b135_conv512k -> exp-b-b1/gpt2-denom-b135/metrics.csv

exp-b-b1/gpt2-denom-b135/metrics.csv

@@ -0,0 +1,31 @@
+step,tokens,timed_hours,gpu_hours,train_loss_ema,neutral_bpb,ownval_bpb,lr,wallclock_s
+17,8912896,0.001006,0.008044,10.4278,3.502167,3.163440,1.46e-05,29.7
+23,12058624,0.001868,0.014945,10.1919,3.458874,3.102812,1.97e-05,53.5
+30,15728640,0.002875,0.022997,9.9392,3.392975,3.030356,2.57e-05,77.8
+40,20971520,0.004310,0.034482,9.6003,3.323465,2.942193,3.43e-05,103.7
+54,28311552,0.006320,0.050561,9.1489,3.193023,2.802958,4.63e-05,131.7
+71,37224448,0.008762,0.070093,8.6080,3.031384,2.612106,6.09e-05,161.2
+95,49807360,0.012209,0.097670,7.8513,2.841267,2.383395,8.14e-05,194.3
+127,66584576,0.016804,0.134435,7.0722,2.690798,2.212879,1.09e-04,231.6
+169,88604672,0.022837,0.182699,6.5460,2.572810,2.095715,1.45e-04,274.2
+226,118489088,0.031024,0.248195,6.1902,2.453567,2.004249,1.94e-04,324.4
+301,157810688,0.041795,0.334361,5.9260,2.366380,1.924478,2.58e-04,384.0
+401,210239488,0.056149,0.449190,5.6055,2.263794,1.822086,3.44e-04,456.5
+535,280494080,0.075385,0.603076,5.2097,2.161257,1.695952,4.59e-04,546.5
+713,373817344,0.100946,0.807569,4.7629,1.953347,1.542335,6.00e-04,659.2
+951,498597888,0.135114,1.080913,4.1801,1.622045,1.371782,6.00e-04,802.9
+1269,665321472,0.180765,1.446120,3.8705,1.518161,1.275497,5.98e-04,988.0
+1692,887095296,0.241523,1.932183,3.6741,1.445093,1.215222,5.95e-04,1227.4
+2256,1182793728,0.322500,2.579998,3.5205,1.401400,1.166298,5.88e-04,1539.7
+3009,1577582592,0.430590,3.444719,3.4144,1.364974,1.131645,5.73e-04,1949.6
+4012,2103443456,0.574577,4.596618,3.3398,1.340601,1.102884,5.46e-04,2488.7
+5350,2804940800,0.766655,6.133242,3.2475,1.310521,1.078509,4.98e-04,3200.9
+7135,3740794880,1.022874,8.182989,3.1799,1.310101,1.056011,4.16e-04,4144.0
+9514,4988076032,1.364346,10.914771,3.1160,1.285669,1.034681,2.93e-04,5394.0
+12687,6651641856,1.819806,14.558446,3.0629,1.275780,1.014594,1.46e-04,7054.5
+15227,7983333376,2.184420,17.475357,3.0255,1.270975,1.004381,7.44e-05,8387.9
+15565,8160542720,2.232942,17.863534,3.0207,1.275941,1.003460,6.92e-05,8583.4
+15904,8338276352,2.281612,18.252893,3.0269,1.270124,1.002664,6.52e-05,8779.3
+16242,8515485696,2.330134,18.641068,3.0101,1.269679,1.002082,6.23e-05,8974.7
+16581,8693219328,2.378800,19.030404,3.0150,1.272916,1.001556,6.06e-05,9170.6
+16919,8870428672,2.427347,19.418779,3.0229,1.271722,1.001095,6.00e-05,9366.2