| { | |
| "best_metric": 0.7996415495872498, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.7561436672967864, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003780718336483932, | |
| "grad_norm": 25.16098976135254, | |
| "learning_rate": 1e-05, | |
| "loss": 4.6715, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003780718336483932, | |
| "eval_loss": 1.5147686004638672, | |
| "eval_runtime": 32.2459, | |
| "eval_samples_per_second": 13.831, | |
| "eval_steps_per_second": 3.473, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007561436672967864, | |
| "grad_norm": 27.060680389404297, | |
| "learning_rate": 2e-05, | |
| "loss": 4.6865, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011342155009451797, | |
| "grad_norm": 26.14573097229004, | |
| "learning_rate": 3e-05, | |
| "loss": 4.4814, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015122873345935728, | |
| "grad_norm": 22.122207641601562, | |
| "learning_rate": 4e-05, | |
| "loss": 4.1077, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01890359168241966, | |
| "grad_norm": 20.75678062438965, | |
| "learning_rate": 5e-05, | |
| "loss": 3.906, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.022684310018903593, | |
| "grad_norm": 14.588601112365723, | |
| "learning_rate": 6e-05, | |
| "loss": 4.1167, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.026465028355387523, | |
| "grad_norm": 14.807610511779785, | |
| "learning_rate": 7e-05, | |
| "loss": 3.8745, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.030245746691871456, | |
| "grad_norm": 14.569311141967773, | |
| "learning_rate": 8e-05, | |
| "loss": 4.0067, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.034026465028355386, | |
| "grad_norm": 11.986530303955078, | |
| "learning_rate": 9e-05, | |
| "loss": 3.772, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03780718336483932, | |
| "grad_norm": 14.717965126037598, | |
| "learning_rate": 0.0001, | |
| "loss": 3.762, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04158790170132325, | |
| "grad_norm": 13.476912498474121, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 4.0498, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.045368620037807186, | |
| "grad_norm": 14.233503341674805, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 3.7616, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04914933837429111, | |
| "grad_norm": 12.872749328613281, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 3.7019, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.052930056710775046, | |
| "grad_norm": 12.905577659606934, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 3.4222, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05671077504725898, | |
| "grad_norm": 13.652875900268555, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 3.4414, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06049149338374291, | |
| "grad_norm": 15.430229187011719, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 3.7592, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06427221172022685, | |
| "grad_norm": 11.641162872314453, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 3.6027, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06805293005671077, | |
| "grad_norm": 12.991557121276855, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 3.7312, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07183364839319471, | |
| "grad_norm": 12.716787338256836, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 3.8725, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07561436672967864, | |
| "grad_norm": 12.194295883178711, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 3.3815, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07939508506616257, | |
| "grad_norm": 10.72068977355957, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 3.0332, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0831758034026465, | |
| "grad_norm": 13.57220458984375, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 3.5567, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 11.49730396270752, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 3.6272, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09073724007561437, | |
| "grad_norm": 13.56385326385498, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 3.0481, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0945179584120983, | |
| "grad_norm": 11.78315544128418, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 3.3534, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09829867674858223, | |
| "grad_norm": 18.407567977905273, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 3.2758, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10207939508506617, | |
| "grad_norm": 13.258454322814941, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 3.6035, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.10586011342155009, | |
| "grad_norm": 10.505517959594727, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 3.2197, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10964083175803403, | |
| "grad_norm": 11.3196439743042, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.6479, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11342155009451796, | |
| "grad_norm": 12.575491905212402, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 3.5406, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11720226843100189, | |
| "grad_norm": 12.494804382324219, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 3.6458, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12098298676748583, | |
| "grad_norm": 14.734025001525879, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 3.5121, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.12476370510396975, | |
| "grad_norm": 10.00881290435791, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 3.2603, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1285444234404537, | |
| "grad_norm": 11.375361442565918, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 3.5365, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1323251417769376, | |
| "grad_norm": 10.176946640014648, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 3.3114, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13610586011342155, | |
| "grad_norm": 10.291150093078613, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 3.2982, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13988657844990549, | |
| "grad_norm": 13.161905288696289, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 3.9598, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.14366729678638943, | |
| "grad_norm": 11.773653030395508, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 3.2511, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.14744801512287334, | |
| "grad_norm": 12.06856632232666, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 3.7918, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15122873345935728, | |
| "grad_norm": 12.7516450881958, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 3.2789, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15500945179584122, | |
| "grad_norm": 11.407308578491211, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 3.479, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15879017013232513, | |
| "grad_norm": 11.646713256835938, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 3.302, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.16257088846880907, | |
| "grad_norm": 10.183646202087402, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 2.882, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.166351606805293, | |
| "grad_norm": 11.927104949951172, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 3.5429, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17013232514177692, | |
| "grad_norm": 10.855883598327637, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 3.4548, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 11.88790512084961, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 3.0145, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1776937618147448, | |
| "grad_norm": 13.963061332702637, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 3.2849, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.18147448015122875, | |
| "grad_norm": 13.203959465026855, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 3.8767, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.18525519848771266, | |
| "grad_norm": 16.107746124267578, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 3.171, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1890359168241966, | |
| "grad_norm": 15.196197509765625, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 3.0762, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1890359168241966, | |
| "eval_loss": 0.9335125088691711, | |
| "eval_runtime": 32.8629, | |
| "eval_samples_per_second": 13.572, | |
| "eval_steps_per_second": 3.408, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19281663516068054, | |
| "grad_norm": 21.29393196105957, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 4.0358, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.19659735349716445, | |
| "grad_norm": 11.304404258728027, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 3.7355, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2003780718336484, | |
| "grad_norm": 10.158336639404297, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 3.8875, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.20415879017013233, | |
| "grad_norm": 10.010252952575684, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 3.599, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.20793950850661624, | |
| "grad_norm": 9.562002182006836, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 3.869, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.21172022684310018, | |
| "grad_norm": 8.870956420898438, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 3.3619, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.21550094517958412, | |
| "grad_norm": 9.110651969909668, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 3.4124, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.21928166351606806, | |
| "grad_norm": 7.6697678565979, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 3.3309, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.22306238185255198, | |
| "grad_norm": 8.301525115966797, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 3.1453, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.22684310018903592, | |
| "grad_norm": 10.347870826721191, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 3.5897, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.23062381852551986, | |
| "grad_norm": 8.439586639404297, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 3.4303, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.23440453686200377, | |
| "grad_norm": 9.513819694519043, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 3.6442, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2381852551984877, | |
| "grad_norm": 8.744819641113281, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 3.0771, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.24196597353497165, | |
| "grad_norm": 11.595215797424316, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 3.8367, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.24574669187145556, | |
| "grad_norm": 8.577299118041992, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 3.3196, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2495274102079395, | |
| "grad_norm": 10.696349143981934, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 3.3712, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2533081285444234, | |
| "grad_norm": 9.022214889526367, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.317, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2570888468809074, | |
| "grad_norm": 8.869611740112305, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 3.3489, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 8.902680397033691, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 3.4113, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2646502835538752, | |
| "grad_norm": 8.611761093139648, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 3.4823, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2684310018903592, | |
| "grad_norm": 9.569182395935059, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 3.6526, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2722117202268431, | |
| "grad_norm": 8.870281219482422, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 3.2249, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.27599243856332706, | |
| "grad_norm": 9.278509140014648, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 3.4246, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.27977315689981097, | |
| "grad_norm": 8.466548919677734, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 3.4338, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2835538752362949, | |
| "grad_norm": 8.425887107849121, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 3.5268, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.28733459357277885, | |
| "grad_norm": 8.38651180267334, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 3.4466, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.29111531190926276, | |
| "grad_norm": 10.235416412353516, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 3.3785, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2948960302457467, | |
| "grad_norm": 10.151168823242188, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 3.4663, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.29867674858223064, | |
| "grad_norm": 9.136027336120605, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 3.2603, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.30245746691871456, | |
| "grad_norm": 8.610246658325195, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 3.442, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.30623818525519847, | |
| "grad_norm": 8.665996551513672, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 3.3577, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.31001890359168244, | |
| "grad_norm": 8.941457748413086, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 3.14, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.31379962192816635, | |
| "grad_norm": 14.6861572265625, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 3.138, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.31758034026465026, | |
| "grad_norm": 9.767927169799805, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 3.5077, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.32136105860113423, | |
| "grad_norm": 9.671030044555664, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 3.3803, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.32514177693761814, | |
| "grad_norm": 9.433391571044922, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.032, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.32892249527410206, | |
| "grad_norm": 9.82751178741455, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 3.3651, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.332703213610586, | |
| "grad_norm": 10.734698295593262, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 3.7739, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.33648393194706994, | |
| "grad_norm": 10.445623397827148, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 3.3833, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.34026465028355385, | |
| "grad_norm": 9.154644966125488, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 3.2674, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3440453686200378, | |
| "grad_norm": 12.329596519470215, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 3.5239, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 9.883134841918945, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 3.3821, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3516068052930057, | |
| "grad_norm": 10.114555358886719, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 3.6195, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3553875236294896, | |
| "grad_norm": 10.966856002807617, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 3.7073, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3591682419659735, | |
| "grad_norm": 10.266011238098145, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 3.2225, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3629489603024575, | |
| "grad_norm": 10.050321578979492, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 3.3454, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3667296786389414, | |
| "grad_norm": 10.897590637207031, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 3.317, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3705103969754253, | |
| "grad_norm": 10.815629005432129, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 3.197, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3742911153119093, | |
| "grad_norm": 11.034646987915039, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.9457, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 13.58685302734375, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 3.2583, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "eval_loss": 0.8677875995635986, | |
| "eval_runtime": 32.8636, | |
| "eval_samples_per_second": 13.571, | |
| "eval_steps_per_second": 3.408, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3818525519848771, | |
| "grad_norm": 9.579238891601562, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 3.5538, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3856332703213611, | |
| "grad_norm": 9.08609390258789, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 3.6579, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.389413988657845, | |
| "grad_norm": 8.470030784606934, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 3.5472, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3931947069943289, | |
| "grad_norm": 7.983551502227783, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 3.126, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.39697542533081287, | |
| "grad_norm": 8.211849212646484, | |
| "learning_rate": 5e-05, | |
| "loss": 3.2246, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4007561436672968, | |
| "grad_norm": 7.689363479614258, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 3.3606, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4045368620037807, | |
| "grad_norm": 7.449761390686035, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 3.4443, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.40831758034026466, | |
| "grad_norm": 7.9302778244018555, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 3.3722, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4120982986767486, | |
| "grad_norm": 8.147639274597168, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 3.4303, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4158790170132325, | |
| "grad_norm": 7.9510498046875, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 3.3212, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.41965973534971646, | |
| "grad_norm": 7.70236349105835, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 3.3949, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.42344045368620037, | |
| "grad_norm": 7.634131908416748, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 3.0981, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.42722117202268434, | |
| "grad_norm": 7.487667083740234, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 3.2295, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.43100189035916825, | |
| "grad_norm": 8.120771408081055, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 3.1201, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 9.001714706420898, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 3.4735, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.43856332703213613, | |
| "grad_norm": 7.907503604888916, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 3.2468, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.44234404536862004, | |
| "grad_norm": 8.025362014770508, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 3.4063, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.44612476370510395, | |
| "grad_norm": 7.450461387634277, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 2.9818, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4499054820415879, | |
| "grad_norm": 8.391257286071777, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 3.1225, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.45368620037807184, | |
| "grad_norm": 8.543011665344238, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 3.2363, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.45746691871455575, | |
| "grad_norm": 7.939721584320068, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 3.267, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4612476370510397, | |
| "grad_norm": 9.583046913146973, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 3.4252, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.46502835538752363, | |
| "grad_norm": 8.423657417297363, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 3.1331, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.46880907372400754, | |
| "grad_norm": 8.184349060058594, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 3.0419, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4725897920604915, | |
| "grad_norm": 10.645812034606934, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 3.1175, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4763705103969754, | |
| "grad_norm": 8.734024047851562, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 3.1994, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.48015122873345933, | |
| "grad_norm": 8.907480239868164, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 3.3357, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4839319470699433, | |
| "grad_norm": 9.246767044067383, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 3.1285, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4877126654064272, | |
| "grad_norm": 9.020038604736328, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 3.3502, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4914933837429111, | |
| "grad_norm": 9.485525131225586, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 3.4852, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4952741020793951, | |
| "grad_norm": 9.426743507385254, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 3.4174, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.499054820415879, | |
| "grad_norm": 8.460476875305176, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 3.1284, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.502835538752363, | |
| "grad_norm": 11.027420997619629, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 3.3026, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5066162570888468, | |
| "grad_norm": 8.760469436645508, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 3.3057, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5103969754253308, | |
| "grad_norm": 8.567980766296387, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 3.0398, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5141776937618148, | |
| "grad_norm": 8.330060005187988, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 2.8685, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5179584120982986, | |
| "grad_norm": 9.437362670898438, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 3.0373, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 7.858784198760986, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 2.936, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5255198487712666, | |
| "grad_norm": 9.211569786071777, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 3.3076, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5293005671077504, | |
| "grad_norm": 8.228165626525879, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 3.0484, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5330812854442344, | |
| "grad_norm": 8.578217506408691, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 3.1356, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5368620037807184, | |
| "grad_norm": 8.754902839660645, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 2.8195, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5406427221172023, | |
| "grad_norm": 9.32534408569336, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.5471, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5444234404536862, | |
| "grad_norm": 8.705456733703613, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 2.9405, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5482041587901701, | |
| "grad_norm": 11.633933067321777, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 3.3586, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5519848771266541, | |
| "grad_norm": 10.78929615020752, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 3.1784, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.555765595463138, | |
| "grad_norm": 9.891103744506836, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.952, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5595463137996219, | |
| "grad_norm": 12.48646354675293, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 3.1926, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5633270321361059, | |
| "grad_norm": 10.62431812286377, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.9827, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 10.362546920776367, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.9446, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "eval_loss": 0.8182804584503174, | |
| "eval_runtime": 32.8641, | |
| "eval_samples_per_second": 13.571, | |
| "eval_steps_per_second": 3.408, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5708884688090737, | |
| "grad_norm": 6.677933692932129, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 3.1274, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5746691871455577, | |
| "grad_norm": 6.997452259063721, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 3.1368, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5784499054820416, | |
| "grad_norm": 7.476685523986816, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 3.1754, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5822306238185255, | |
| "grad_norm": 6.509499549865723, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 2.9101, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5860113421550095, | |
| "grad_norm": 7.388987064361572, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 3.0561, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5897920604914934, | |
| "grad_norm": 6.738891124725342, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 3.2056, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5935727788279773, | |
| "grad_norm": 7.2047271728515625, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 3.1051, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5973534971644613, | |
| "grad_norm": 7.396187782287598, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 3.07, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6011342155009451, | |
| "grad_norm": 7.220828056335449, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 3.1337, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6049149338374291, | |
| "grad_norm": 7.189826965332031, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 3.1187, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 7.4274444580078125, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 3.0282, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6124763705103969, | |
| "grad_norm": 6.307836532592773, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.9411, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6162570888468809, | |
| "grad_norm": 7.574887275695801, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 3.1878, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6200378071833649, | |
| "grad_norm": 7.00742769241333, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 3.3242, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6238185255198487, | |
| "grad_norm": 6.812418460845947, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 2.628, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6275992438563327, | |
| "grad_norm": 7.7055511474609375, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 3.058, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6313799621928167, | |
| "grad_norm": 10.322258949279785, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 3.4627, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6351606805293005, | |
| "grad_norm": 7.693629264831543, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 3.4181, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6389413988657845, | |
| "grad_norm": 8.244001388549805, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 3.0583, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6427221172022685, | |
| "grad_norm": 7.373150825500488, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 3.3661, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6465028355387523, | |
| "grad_norm": 7.881206035614014, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 3.4858, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6502835538752363, | |
| "grad_norm": 7.820554733276367, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 3.4975, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6540642722117203, | |
| "grad_norm": 8.051074981689453, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 3.4249, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6578449905482041, | |
| "grad_norm": 9.448352813720703, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 3.302, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6616257088846881, | |
| "grad_norm": 7.202307224273682, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 3.0817, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.665406427221172, | |
| "grad_norm": 8.348740577697754, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 3.4199, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6691871455576559, | |
| "grad_norm": 7.092697620391846, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 3.0145, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6729678638941399, | |
| "grad_norm": 9.28335952758789, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 3.128, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6767485822306238, | |
| "grad_norm": 8.880681037902832, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 3.4113, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6805293005671077, | |
| "grad_norm": 7.5821309089660645, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 3.0677, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6843100189035917, | |
| "grad_norm": 8.36470890045166, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.8724, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6880907372400756, | |
| "grad_norm": 9.106186866760254, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 3.4504, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6918714555765595, | |
| "grad_norm": 8.229049682617188, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 3.0558, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 11.976222038269043, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 3.3652, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6994328922495274, | |
| "grad_norm": 7.686277389526367, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 2.8093, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7032136105860114, | |
| "grad_norm": 7.257493019104004, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 2.8761, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7069943289224953, | |
| "grad_norm": 8.06243896484375, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 2.9172, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7107750472589792, | |
| "grad_norm": 8.421680450439453, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 3.1263, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7145557655954632, | |
| "grad_norm": 7.971578121185303, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 3.0788, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.718336483931947, | |
| "grad_norm": 8.98708724975586, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 3.0703, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.722117202268431, | |
| "grad_norm": 9.273012161254883, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 3.7144, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.725897920604915, | |
| "grad_norm": 8.652299880981445, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 3.2203, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7296786389413988, | |
| "grad_norm": 9.335807800292969, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 3.4474, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7334593572778828, | |
| "grad_norm": 8.34329605102539, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 2.6767, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7372400756143668, | |
| "grad_norm": 9.559061050415039, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 3.4894, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7410207939508506, | |
| "grad_norm": 8.606451034545898, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 3.1745, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7448015122873346, | |
| "grad_norm": 9.714865684509277, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 3.285, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7485822306238186, | |
| "grad_norm": 9.648216247558594, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 2.9202, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7523629489603024, | |
| "grad_norm": 11.27525806427002, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 3.4114, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 11.783528327941895, | |
| "learning_rate": 0.0, | |
| "loss": 3.4886, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "eval_loss": 0.7996415495872498, | |
| "eval_runtime": 32.8411, | |
| "eval_samples_per_second": 13.581, | |
| "eval_steps_per_second": 3.41, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.861999239200768e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |