| { | |
| "best_metric": 0.796931266784668, | |
| "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": 25.868711471557617, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1679, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007561436672967864, | |
| "grad_norm": 26.996562957763672, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1712, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011342155009451797, | |
| "grad_norm": 26.536226272583008, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1189, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015122873345935728, | |
| "grad_norm": 22.53629493713379, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0266, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01890359168241966, | |
| "grad_norm": 21.505897521972656, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9793, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.022684310018903593, | |
| "grad_norm": 14.928448677062988, | |
| "learning_rate": 6e-05, | |
| "loss": 1.0265, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.026465028355387523, | |
| "grad_norm": 14.477608680725098, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9681, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.030245746691871456, | |
| "grad_norm": 15.048117637634277, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9996, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.034026465028355386, | |
| "grad_norm": 12.123842239379883, | |
| "learning_rate": 9e-05, | |
| "loss": 0.9432, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03780718336483932, | |
| "grad_norm": 14.511091232299805, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9389, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04158790170132325, | |
| "grad_norm": 13.797380447387695, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 1.0154, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.045368620037807186, | |
| "grad_norm": 14.586050033569336, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.9381, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04914933837429111, | |
| "grad_norm": 13.497241973876953, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 0.9256, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.052930056710775046, | |
| "grad_norm": 14.387834548950195, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.8607, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05671077504725898, | |
| "grad_norm": 13.219511032104492, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.8533, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06049149338374291, | |
| "grad_norm": 14.26727294921875, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.9375, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06427221172022685, | |
| "grad_norm": 11.609099388122559, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.9001, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06805293005671077, | |
| "grad_norm": 13.114840507507324, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.933, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07183364839319471, | |
| "grad_norm": 14.351373672485352, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.9689, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07561436672967864, | |
| "grad_norm": 12.573498725891113, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.8473, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07939508506616257, | |
| "grad_norm": 11.286185264587402, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.764, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0831758034026465, | |
| "grad_norm": 12.673500061035156, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.8825, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 11.740640640258789, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.9075, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09073724007561437, | |
| "grad_norm": 11.874197959899902, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.7685, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0945179584120983, | |
| "grad_norm": 11.308283805847168, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.8379, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09829867674858223, | |
| "grad_norm": 12.97791576385498, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 0.8118, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10207939508506617, | |
| "grad_norm": 12.41085147857666, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.8996, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.10586011342155009, | |
| "grad_norm": 10.894973754882812, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.8036, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10964083175803403, | |
| "grad_norm": 22.103853225708008, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.9205, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11342155009451796, | |
| "grad_norm": 11.814549446105957, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.8935, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11720226843100189, | |
| "grad_norm": 12.61363410949707, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.9126, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12098298676748583, | |
| "grad_norm": 12.015908241271973, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.8745, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.12476370510396975, | |
| "grad_norm": 11.367657661437988, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.8188, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.1285444234404537, | |
| "grad_norm": 12.085219383239746, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.8813, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1323251417769376, | |
| "grad_norm": 10.321392059326172, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.8272, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13610586011342155, | |
| "grad_norm": 10.407423973083496, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 0.8204, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13988657844990549, | |
| "grad_norm": 13.754597663879395, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 0.9912, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.14366729678638943, | |
| "grad_norm": 12.137901306152344, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 0.8117, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.14744801512287334, | |
| "grad_norm": 12.281770706176758, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.9461, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15122873345935728, | |
| "grad_norm": 11.265958786010742, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.8125, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15500945179584122, | |
| "grad_norm": 12.097908973693848, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.8829, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15879017013232513, | |
| "grad_norm": 11.462668418884277, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.8174, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.16257088846880907, | |
| "grad_norm": 10.441878318786621, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.7266, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.166351606805293, | |
| "grad_norm": 11.729705810546875, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.8822, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17013232514177692, | |
| "grad_norm": 10.593080520629883, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.8647, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 12.002582550048828, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.7542, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1776937618147448, | |
| "grad_norm": 13.187206268310547, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.8168, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.18147448015122875, | |
| "grad_norm": 15.53892993927002, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9638, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.18525519848771266, | |
| "grad_norm": 20.015409469604492, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.7975, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1890359168241966, | |
| "grad_norm": 14.737130165100098, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.7617, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1890359168241966, | |
| "eval_loss": 0.937868058681488, | |
| "eval_runtime": 23.8905, | |
| "eval_samples_per_second": 18.669, | |
| "eval_steps_per_second": 4.688, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19281663516068054, | |
| "grad_norm": 19.0587158203125, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 1.0027, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.19659735349716445, | |
| "grad_norm": 10.542956352233887, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 0.9294, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2003780718336484, | |
| "grad_norm": 10.02282428741455, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 0.9812, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.20415879017013233, | |
| "grad_norm": 15.777267456054688, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.8996, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.20793950850661624, | |
| "grad_norm": 9.673592567443848, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.9667, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.21172022684310018, | |
| "grad_norm": 8.31556510925293, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.8437, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.21550094517958412, | |
| "grad_norm": 8.866250038146973, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.8598, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.21928166351606806, | |
| "grad_norm": 8.169628143310547, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.8421, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.22306238185255198, | |
| "grad_norm": 7.955894947052002, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.78, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.22684310018903592, | |
| "grad_norm": 9.585834503173828, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.8945, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.23062381852551986, | |
| "grad_norm": 10.034743309020996, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.866, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.23440453686200377, | |
| "grad_norm": 9.564702033996582, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.9144, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2381852551984877, | |
| "grad_norm": 8.519525527954102, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.7744, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.24196597353497165, | |
| "grad_norm": 9.579198837280273, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.9486, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.24574669187145556, | |
| "grad_norm": 8.82320785522461, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.834, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2495274102079395, | |
| "grad_norm": 25.461153030395508, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.8452, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2533081285444234, | |
| "grad_norm": 9.499732971191406, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.8343, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2570888468809074, | |
| "grad_norm": 9.04475212097168, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.8387, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 9.186306953430176, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.8606, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2646502835538752, | |
| "grad_norm": 8.687911033630371, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.8632, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2684310018903592, | |
| "grad_norm": 9.258337020874023, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.9112, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2722117202268431, | |
| "grad_norm": 9.268692016601562, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.7968, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.27599243856332706, | |
| "grad_norm": 9.389679908752441, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.8604, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.27977315689981097, | |
| "grad_norm": 8.71633243560791, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.8564, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2835538752362949, | |
| "grad_norm": 8.666472434997559, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.8845, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.28733459357277885, | |
| "grad_norm": 8.605114936828613, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.857, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.29111531190926276, | |
| "grad_norm": 9.577071189880371, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.8482, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2948960302457467, | |
| "grad_norm": 9.829218864440918, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.8762, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.29867674858223064, | |
| "grad_norm": 8.901524543762207, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.8257, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.30245746691871456, | |
| "grad_norm": 8.700294494628906, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.8635, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.30623818525519847, | |
| "grad_norm": 8.539888381958008, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.845, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.31001890359168244, | |
| "grad_norm": 8.737793922424316, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.7761, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.31379962192816635, | |
| "grad_norm": 10.041154861450195, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.78, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.31758034026465026, | |
| "grad_norm": 10.532041549682617, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.8732, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.32136105860113423, | |
| "grad_norm": 10.783872604370117, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.8432, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.32514177693761814, | |
| "grad_norm": 8.684202194213867, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.7461, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.32892249527410206, | |
| "grad_norm": 12.615586280822754, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.852, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.332703213610586, | |
| "grad_norm": 13.347943305969238, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.9457, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.33648393194706994, | |
| "grad_norm": 11.482415199279785, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.847, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.34026465028355385, | |
| "grad_norm": 9.020286560058594, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.8173, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3440453686200378, | |
| "grad_norm": 9.47546100616455, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.8804, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 9.721048355102539, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.8527, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3516068052930057, | |
| "grad_norm": 10.509419441223145, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.8958, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3553875236294896, | |
| "grad_norm": 10.52464771270752, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.9308, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3591682419659735, | |
| "grad_norm": 12.20820426940918, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.7969, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3629489603024575, | |
| "grad_norm": 10.435171127319336, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.8355, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3667296786389414, | |
| "grad_norm": 10.847406387329102, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.8236, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3705103969754253, | |
| "grad_norm": 10.55009937286377, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.7918, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3742911153119093, | |
| "grad_norm": 10.11078929901123, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.7335, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "grad_norm": 15.384814262390137, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.8123, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3780718336483932, | |
| "eval_loss": 0.8648684620857239, | |
| "eval_runtime": 23.8735, | |
| "eval_samples_per_second": 18.682, | |
| "eval_steps_per_second": 4.691, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3818525519848771, | |
| "grad_norm": 8.453338623046875, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.8779, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3856332703213611, | |
| "grad_norm": 9.364800453186035, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.9183, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.389413988657845, | |
| "grad_norm": 7.999811172485352, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.8824, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3931947069943289, | |
| "grad_norm": 8.631146430969238, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.773, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.39697542533081287, | |
| "grad_norm": 8.38680362701416, | |
| "learning_rate": 5e-05, | |
| "loss": 0.806, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4007561436672968, | |
| "grad_norm": 8.239096641540527, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.8421, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4045368620037807, | |
| "grad_norm": 8.71033000946045, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.8713, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.40831758034026466, | |
| "grad_norm": 7.696841716766357, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.8395, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4120982986767486, | |
| "grad_norm": 8.243152618408203, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.8478, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4158790170132325, | |
| "grad_norm": 7.823327541351318, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.827, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.41965973534971646, | |
| "grad_norm": 8.201346397399902, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.8584, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.42344045368620037, | |
| "grad_norm": 7.314600467681885, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.7623, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.42722117202268434, | |
| "grad_norm": 7.454195499420166, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.809, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.43100189035916825, | |
| "grad_norm": 7.83469295501709, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.7791, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 8.487276077270508, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.8794, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.43856332703213613, | |
| "grad_norm": 7.3819403648376465, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.8056, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.44234404536862004, | |
| "grad_norm": 7.734245777130127, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.8643, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.44612476370510395, | |
| "grad_norm": 7.622498989105225, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.7474, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4499054820415879, | |
| "grad_norm": 7.614246368408203, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.7816, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.45368620037807184, | |
| "grad_norm": 8.39334774017334, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.8036, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.45746691871455575, | |
| "grad_norm": 7.695074558258057, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.8185, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4612476370510397, | |
| "grad_norm": 8.945220947265625, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.8366, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.46502835538752363, | |
| "grad_norm": 9.416232109069824, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.7891, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.46880907372400754, | |
| "grad_norm": 7.6796793937683105, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.7673, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4725897920604915, | |
| "grad_norm": 9.226118087768555, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.7863, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4763705103969754, | |
| "grad_norm": 9.392522811889648, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.8126, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.48015122873345933, | |
| "grad_norm": 8.817024230957031, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.8353, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4839319470699433, | |
| "grad_norm": 9.890538215637207, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.8016, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4877126654064272, | |
| "grad_norm": 8.743091583251953, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.8398, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4914933837429111, | |
| "grad_norm": 8.72990894317627, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.8724, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4952741020793951, | |
| "grad_norm": 8.343745231628418, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.8495, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.499054820415879, | |
| "grad_norm": 8.648836135864258, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.7893, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.502835538752363, | |
| "grad_norm": 9.886215209960938, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.8258, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5066162570888468, | |
| "grad_norm": 8.574617385864258, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.818, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5103969754253308, | |
| "grad_norm": 8.148327827453613, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.7673, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5141776937618148, | |
| "grad_norm": 8.997401237487793, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.7205, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5179584120982986, | |
| "grad_norm": 8.486831665039062, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.7618, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 8.21071720123291, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.7401, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5255198487712666, | |
| "grad_norm": 9.185858726501465, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.829, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5293005671077504, | |
| "grad_norm": 10.219714164733887, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.7614, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5330812854442344, | |
| "grad_norm": 9.078256607055664, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.7828, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5368620037807184, | |
| "grad_norm": 8.314924240112305, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.7112, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5406427221172023, | |
| "grad_norm": 9.759888648986816, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.8955, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5444234404536862, | |
| "grad_norm": 8.717164993286133, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.7285, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5482041587901701, | |
| "grad_norm": 11.066908836364746, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.839, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5519848771266541, | |
| "grad_norm": 12.829963684082031, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.8001, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.555765595463138, | |
| "grad_norm": 9.667625427246094, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.7371, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5595463137996219, | |
| "grad_norm": 12.464801788330078, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.811, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5633270321361059, | |
| "grad_norm": 11.538606643676758, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.7504, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "grad_norm": 10.46953296661377, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.7313, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5671077504725898, | |
| "eval_loss": 0.8185301423072815, | |
| "eval_runtime": 23.8735, | |
| "eval_samples_per_second": 18.682, | |
| "eval_steps_per_second": 4.691, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5708884688090737, | |
| "grad_norm": 7.325110912322998, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 0.782, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5746691871455577, | |
| "grad_norm": 7.679309844970703, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 0.7868, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5784499054820416, | |
| "grad_norm": 7.797189235687256, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 0.7972, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5822306238185255, | |
| "grad_norm": 6.593117713928223, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 0.7352, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5860113421550095, | |
| "grad_norm": 6.538660526275635, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 0.7535, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5897920604914934, | |
| "grad_norm": 6.678139686584473, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 0.7951, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5935727788279773, | |
| "grad_norm": 6.2444000244140625, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 0.7729, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5973534971644613, | |
| "grad_norm": 6.7908806800842285, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 0.773, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6011342155009451, | |
| "grad_norm": 7.167297840118408, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 0.7844, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6049149338374291, | |
| "grad_norm": 7.551186561584473, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 0.7814, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 7.515049457550049, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 0.7622, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6124763705103969, | |
| "grad_norm": 6.513700485229492, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.7393, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6162570888468809, | |
| "grad_norm": 7.5850677490234375, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 0.7926, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6200378071833649, | |
| "grad_norm": 6.940384864807129, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 0.8321, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6238185255198487, | |
| "grad_norm": 6.8749518394470215, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 0.6635, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6275992438563327, | |
| "grad_norm": 8.078300476074219, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 0.7592, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6313799621928167, | |
| "grad_norm": 8.392534255981445, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 0.8723, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6351606805293005, | |
| "grad_norm": 7.7992777824401855, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 0.852, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6389413988657845, | |
| "grad_norm": 7.612271785736084, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 0.7654, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6427221172022685, | |
| "grad_norm": 7.638868808746338, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 0.8484, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6465028355387523, | |
| "grad_norm": 7.941742420196533, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 0.865, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6502835538752363, | |
| "grad_norm": 7.6441121101379395, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 0.863, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6540642722117203, | |
| "grad_norm": 8.091294288635254, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 0.8611, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6578449905482041, | |
| "grad_norm": 9.733619689941406, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 0.8203, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6616257088846881, | |
| "grad_norm": 7.44538688659668, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 0.7759, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.665406427221172, | |
| "grad_norm": 8.235344886779785, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 0.8583, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6691871455576559, | |
| "grad_norm": 7.185950756072998, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 0.7537, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6729678638941399, | |
| "grad_norm": 8.205653190612793, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 0.7697, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6767485822306238, | |
| "grad_norm": 8.178430557250977, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 0.8505, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6805293005671077, | |
| "grad_norm": 8.145241737365723, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 0.7712, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6843100189035917, | |
| "grad_norm": 8.158324241638184, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.7297, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6880907372400756, | |
| "grad_norm": 8.506850242614746, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 0.8544, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6918714555765595, | |
| "grad_norm": 7.831125736236572, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 0.7568, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 9.162361145019531, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 0.8469, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6994328922495274, | |
| "grad_norm": 7.752386569976807, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 0.7049, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7032136105860114, | |
| "grad_norm": 7.178589820861816, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 0.7126, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7069943289224953, | |
| "grad_norm": 8.789344787597656, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 0.7266, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7107750472589792, | |
| "grad_norm": 8.555848121643066, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 0.7912, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7145557655954632, | |
| "grad_norm": 7.807133674621582, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 0.759, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.718336483931947, | |
| "grad_norm": 8.937325477600098, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 0.7737, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.722117202268431, | |
| "grad_norm": 9.433504104614258, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 0.9284, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.725897920604915, | |
| "grad_norm": 9.529987335205078, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 0.8082, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7296786389413988, | |
| "grad_norm": 9.436301231384277, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 0.8659, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7334593572778828, | |
| "grad_norm": 8.199891090393066, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 0.6689, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7372400756143668, | |
| "grad_norm": 9.403008460998535, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 0.8685, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7410207939508506, | |
| "grad_norm": 8.600671768188477, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 0.7835, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7448015122873346, | |
| "grad_norm": 10.223450660705566, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 0.8211, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7485822306238186, | |
| "grad_norm": 11.230772018432617, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 0.7293, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7523629489603024, | |
| "grad_norm": 10.848817825317383, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 0.859, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "grad_norm": 11.677827835083008, | |
| "learning_rate": 0.0, | |
| "loss": 0.8752, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7561436672967864, | |
| "eval_loss": 0.796931266784668, | |
| "eval_runtime": 23.8844, | |
| "eval_samples_per_second": 18.673, | |
| "eval_steps_per_second": 4.689, | |
| "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 | |
| } | |