{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2415, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00041407867494824016, "grad_norm": 24.186954732535973, "learning_rate": 4.132231404958678e-08, "loss": 1.427, "step": 1 }, { "epoch": 0.002070393374741201, "grad_norm": 24.196130418494608, "learning_rate": 2.066115702479339e-07, "loss": 1.4047, "step": 5 }, { "epoch": 0.004140786749482402, "grad_norm": 16.3762981570342, "learning_rate": 4.132231404958678e-07, "loss": 1.3799, "step": 10 }, { "epoch": 0.006211180124223602, "grad_norm": 8.425206725694215, "learning_rate": 6.198347107438018e-07, "loss": 1.2594, "step": 15 }, { "epoch": 0.008281573498964804, "grad_norm": 10.05371068200267, "learning_rate": 8.264462809917356e-07, "loss": 1.1298, "step": 20 }, { "epoch": 0.010351966873706004, "grad_norm": 4.721175355009842, "learning_rate": 1.0330578512396695e-06, "loss": 1.0296, "step": 25 }, { "epoch": 0.012422360248447204, "grad_norm": 3.3182689234643394, "learning_rate": 1.2396694214876035e-06, "loss": 0.9653, "step": 30 }, { "epoch": 0.014492753623188406, "grad_norm": 3.504456631337904, "learning_rate": 1.4462809917355372e-06, "loss": 0.9407, "step": 35 }, { "epoch": 0.016563146997929608, "grad_norm": 2.9960316182859783, "learning_rate": 1.6528925619834712e-06, "loss": 0.9251, "step": 40 }, { "epoch": 0.018633540372670808, "grad_norm": 3.1240932779362622, "learning_rate": 1.859504132231405e-06, "loss": 0.9188, "step": 45 }, { "epoch": 0.020703933747412008, "grad_norm": 3.2513730400449203, "learning_rate": 2.066115702479339e-06, "loss": 0.8944, "step": 50 }, { "epoch": 0.022774327122153208, "grad_norm": 3.1786210930130308, "learning_rate": 2.2727272727272728e-06, "loss": 0.9026, "step": 55 }, { "epoch": 0.024844720496894408, "grad_norm": 3.1076677023399464, "learning_rate": 2.479338842975207e-06, "loss": 0.9026, "step": 60 }, { "epoch": 0.026915113871635612, "grad_norm": 3.0823234166313824, "learning_rate": 2.6859504132231405e-06, "loss": 0.8895, "step": 65 }, { "epoch": 0.028985507246376812, "grad_norm": 3.005596270571828, "learning_rate": 2.8925619834710743e-06, "loss": 0.8802, "step": 70 }, { "epoch": 0.031055900621118012, "grad_norm": 3.3025364718701056, "learning_rate": 3.0991735537190086e-06, "loss": 0.8779, "step": 75 }, { "epoch": 0.033126293995859216, "grad_norm": 2.9484146765559256, "learning_rate": 3.3057851239669424e-06, "loss": 0.8556, "step": 80 }, { "epoch": 0.035196687370600416, "grad_norm": 2.991384604651969, "learning_rate": 3.5123966942148763e-06, "loss": 0.8768, "step": 85 }, { "epoch": 0.037267080745341616, "grad_norm": 3.2763200828654044, "learning_rate": 3.71900826446281e-06, "loss": 0.8693, "step": 90 }, { "epoch": 0.039337474120082816, "grad_norm": 2.9849839023650913, "learning_rate": 3.925619834710744e-06, "loss": 0.8663, "step": 95 }, { "epoch": 0.041407867494824016, "grad_norm": 3.0799688517985406, "learning_rate": 4.132231404958678e-06, "loss": 0.8449, "step": 100 }, { "epoch": 0.043478260869565216, "grad_norm": 3.164277927549159, "learning_rate": 4.338842975206612e-06, "loss": 0.8556, "step": 105 }, { "epoch": 0.045548654244306416, "grad_norm": 3.054698785813227, "learning_rate": 4.5454545454545455e-06, "loss": 0.8406, "step": 110 }, { "epoch": 0.047619047619047616, "grad_norm": 3.1517792822498296, "learning_rate": 4.75206611570248e-06, "loss": 0.845, "step": 115 }, { "epoch": 0.049689440993788817, "grad_norm": 3.1935189527896988, "learning_rate": 4.958677685950414e-06, "loss": 0.8501, "step": 120 }, { "epoch": 0.051759834368530024, "grad_norm": 3.2963890157762696, "learning_rate": 5.165289256198347e-06, "loss": 0.8274, "step": 125 }, { "epoch": 0.053830227743271224, "grad_norm": 3.1324161276204303, "learning_rate": 5.371900826446281e-06, "loss": 0.8291, "step": 130 }, { "epoch": 0.055900621118012424, "grad_norm": 3.010837639139227, "learning_rate": 5.578512396694216e-06, "loss": 0.8415, "step": 135 }, { "epoch": 0.057971014492753624, "grad_norm": 3.0560184585554295, "learning_rate": 5.785123966942149e-06, "loss": 0.8273, "step": 140 }, { "epoch": 0.060041407867494824, "grad_norm": 3.1627396501144025, "learning_rate": 5.991735537190083e-06, "loss": 0.8352, "step": 145 }, { "epoch": 0.062111801242236024, "grad_norm": 3.3250094405823707, "learning_rate": 6.198347107438017e-06, "loss": 0.8289, "step": 150 }, { "epoch": 0.06418219461697723, "grad_norm": 3.2642262323401003, "learning_rate": 6.404958677685951e-06, "loss": 0.8111, "step": 155 }, { "epoch": 0.06625258799171843, "grad_norm": 2.980902395191929, "learning_rate": 6.611570247933885e-06, "loss": 0.819, "step": 160 }, { "epoch": 0.06832298136645963, "grad_norm": 3.0419508898389886, "learning_rate": 6.818181818181818e-06, "loss": 0.8177, "step": 165 }, { "epoch": 0.07039337474120083, "grad_norm": 3.1843886993600385, "learning_rate": 7.0247933884297525e-06, "loss": 0.8312, "step": 170 }, { "epoch": 0.07246376811594203, "grad_norm": 3.1854141596433867, "learning_rate": 7.231404958677687e-06, "loss": 0.813, "step": 175 }, { "epoch": 0.07453416149068323, "grad_norm": 3.2876165638717443, "learning_rate": 7.43801652892562e-06, "loss": 0.8384, "step": 180 }, { "epoch": 0.07660455486542443, "grad_norm": 2.9781450388628192, "learning_rate": 7.644628099173555e-06, "loss": 0.8094, "step": 185 }, { "epoch": 0.07867494824016563, "grad_norm": 3.348906096067214, "learning_rate": 7.851239669421489e-06, "loss": 0.8049, "step": 190 }, { "epoch": 0.08074534161490683, "grad_norm": 3.1092493369410783, "learning_rate": 8.057851239669421e-06, "loss": 0.826, "step": 195 }, { "epoch": 0.08281573498964803, "grad_norm": 3.056952614753785, "learning_rate": 8.264462809917356e-06, "loss": 0.8139, "step": 200 }, { "epoch": 0.08488612836438923, "grad_norm": 3.1015829113018234, "learning_rate": 8.47107438016529e-06, "loss": 0.8271, "step": 205 }, { "epoch": 0.08695652173913043, "grad_norm": 3.0740038363605926, "learning_rate": 8.677685950413224e-06, "loss": 0.801, "step": 210 }, { "epoch": 0.08902691511387163, "grad_norm": 2.9037810245033997, "learning_rate": 8.884297520661158e-06, "loss": 0.8095, "step": 215 }, { "epoch": 0.09109730848861283, "grad_norm": 3.027706124969457, "learning_rate": 9.090909090909091e-06, "loss": 0.8125, "step": 220 }, { "epoch": 0.09316770186335403, "grad_norm": 3.1151507400215745, "learning_rate": 9.297520661157025e-06, "loss": 0.797, "step": 225 }, { "epoch": 0.09523809523809523, "grad_norm": 3.053150958204543, "learning_rate": 9.50413223140496e-06, "loss": 0.7984, "step": 230 }, { "epoch": 0.09730848861283643, "grad_norm": 3.0112118668387047, "learning_rate": 9.710743801652894e-06, "loss": 0.8126, "step": 235 }, { "epoch": 0.09937888198757763, "grad_norm": 3.117727242870845, "learning_rate": 9.917355371900828e-06, "loss": 0.8048, "step": 240 }, { "epoch": 0.10144927536231885, "grad_norm": 2.8274397177377812, "learning_rate": 9.999952971391373e-06, "loss": 0.7992, "step": 245 }, { "epoch": 0.10351966873706005, "grad_norm": 3.006340482443734, "learning_rate": 9.999665577542406e-06, "loss": 0.7823, "step": 250 }, { "epoch": 0.10559006211180125, "grad_norm": 2.9753685361571915, "learning_rate": 9.999116931848504e-06, "loss": 0.804, "step": 255 }, { "epoch": 0.10766045548654245, "grad_norm": 2.8643733992658995, "learning_rate": 9.99830706297852e-06, "loss": 0.7747, "step": 260 }, { "epoch": 0.10973084886128365, "grad_norm": 2.8078822435477777, "learning_rate": 9.997236013251234e-06, "loss": 0.7961, "step": 265 }, { "epoch": 0.11180124223602485, "grad_norm": 2.912717313270744, "learning_rate": 9.995903838633133e-06, "loss": 0.783, "step": 270 }, { "epoch": 0.11387163561076605, "grad_norm": 2.8217001344431036, "learning_rate": 9.994310608735492e-06, "loss": 0.7999, "step": 275 }, { "epoch": 0.11594202898550725, "grad_norm": 2.958269779951464, "learning_rate": 9.99245640681072e-06, "loss": 0.7617, "step": 280 }, { "epoch": 0.11801242236024845, "grad_norm": 3.5988430245555416, "learning_rate": 9.99034132974804e-06, "loss": 0.79, "step": 285 }, { "epoch": 0.12008281573498965, "grad_norm": 2.7976040573963856, "learning_rate": 9.987965488068398e-06, "loss": 0.7506, "step": 290 }, { "epoch": 0.12215320910973085, "grad_norm": 2.8825061372075935, "learning_rate": 9.985329005918702e-06, "loss": 0.7841, "step": 295 }, { "epoch": 0.12422360248447205, "grad_norm": 3.05421834044982, "learning_rate": 9.982432021065334e-06, "loss": 0.7831, "step": 300 }, { "epoch": 0.12629399585921325, "grad_norm": 2.7477317097995537, "learning_rate": 9.979274684886943e-06, "loss": 0.7734, "step": 305 }, { "epoch": 0.12836438923395446, "grad_norm": 2.949511742636124, "learning_rate": 9.975857162366547e-06, "loss": 0.7577, "step": 310 }, { "epoch": 0.13043478260869565, "grad_norm": 3.0377639689629885, "learning_rate": 9.972179632082899e-06, "loss": 0.7725, "step": 315 }, { "epoch": 0.13250517598343686, "grad_norm": 3.1353289951358754, "learning_rate": 9.968242286201171e-06, "loss": 0.7619, "step": 320 }, { "epoch": 0.13457556935817805, "grad_norm": 3.1332684478011084, "learning_rate": 9.964045330462896e-06, "loss": 0.7611, "step": 325 }, { "epoch": 0.13664596273291926, "grad_norm": 3.156332097898071, "learning_rate": 9.959588984175228e-06, "loss": 0.7764, "step": 330 }, { "epoch": 0.13871635610766045, "grad_norm": 2.900772606537427, "learning_rate": 9.954873480199483e-06, "loss": 0.7571, "step": 335 }, { "epoch": 0.14078674948240166, "grad_norm": 2.975845497976506, "learning_rate": 9.94989906493896e-06, "loss": 0.7407, "step": 340 }, { "epoch": 0.14285714285714285, "grad_norm": 3.1365253794712364, "learning_rate": 9.94466599832608e-06, "loss": 0.776, "step": 345 }, { "epoch": 0.14492753623188406, "grad_norm": 2.7718028032418522, "learning_rate": 9.939174553808793e-06, "loss": 0.7441, "step": 350 }, { "epoch": 0.14699792960662525, "grad_norm": 2.8486682818864284, "learning_rate": 9.933425018336292e-06, "loss": 0.7476, "step": 355 }, { "epoch": 0.14906832298136646, "grad_norm": 2.9252854160968735, "learning_rate": 9.927417692344025e-06, "loss": 0.752, "step": 360 }, { "epoch": 0.15113871635610765, "grad_norm": 2.9773109330044902, "learning_rate": 9.921152889737985e-06, "loss": 0.7349, "step": 365 }, { "epoch": 0.15320910973084886, "grad_norm": 2.910786730363286, "learning_rate": 9.914630937878315e-06, "loss": 0.76, "step": 370 }, { "epoch": 0.15527950310559005, "grad_norm": 2.620946999109397, "learning_rate": 9.907852177562201e-06, "loss": 0.7246, "step": 375 }, { "epoch": 0.15734989648033126, "grad_norm": 2.7006556546764915, "learning_rate": 9.900816963006063e-06, "loss": 0.7548, "step": 380 }, { "epoch": 0.15942028985507245, "grad_norm": 2.8388206457825054, "learning_rate": 9.893525661827043e-06, "loss": 0.7334, "step": 385 }, { "epoch": 0.16149068322981366, "grad_norm": 2.7395620691398856, "learning_rate": 9.885978655023805e-06, "loss": 0.7487, "step": 390 }, { "epoch": 0.16356107660455488, "grad_norm": 2.7273846238672754, "learning_rate": 9.87817633695661e-06, "loss": 0.7304, "step": 395 }, { "epoch": 0.16563146997929606, "grad_norm": 2.8179008423909524, "learning_rate": 9.87011911532673e-06, "loss": 0.7419, "step": 400 }, { "epoch": 0.16770186335403728, "grad_norm": 3.149985443780922, "learning_rate": 9.861807411155126e-06, "loss": 0.7371, "step": 405 }, { "epoch": 0.16977225672877846, "grad_norm": 2.8768198170816026, "learning_rate": 9.853241658760457e-06, "loss": 0.7411, "step": 410 }, { "epoch": 0.17184265010351968, "grad_norm": 2.7229415130760146, "learning_rate": 9.844422305736383e-06, "loss": 0.7206, "step": 415 }, { "epoch": 0.17391304347826086, "grad_norm": 2.851284715864324, "learning_rate": 9.835349812928178e-06, "loss": 0.7168, "step": 420 }, { "epoch": 0.17598343685300208, "grad_norm": 2.757115924608802, "learning_rate": 9.826024654408645e-06, "loss": 0.7058, "step": 425 }, { "epoch": 0.17805383022774326, "grad_norm": 3.1745359708751812, "learning_rate": 9.816447317453353e-06, "loss": 0.718, "step": 430 }, { "epoch": 0.18012422360248448, "grad_norm": 2.911212797883499, "learning_rate": 9.80661830251516e-06, "loss": 0.7092, "step": 435 }, { "epoch": 0.18219461697722567, "grad_norm": 2.79176343971146, "learning_rate": 9.796538123198077e-06, "loss": 0.718, "step": 440 }, { "epoch": 0.18426501035196688, "grad_norm": 3.0726347632388755, "learning_rate": 9.786207306230422e-06, "loss": 0.7034, "step": 445 }, { "epoch": 0.18633540372670807, "grad_norm": 2.69142104953591, "learning_rate": 9.775626391437303e-06, "loss": 0.7029, "step": 450 }, { "epoch": 0.18840579710144928, "grad_norm": 2.8242682148315366, "learning_rate": 9.764795931712396e-06, "loss": 0.7148, "step": 455 }, { "epoch": 0.19047619047619047, "grad_norm": 3.107425827164937, "learning_rate": 9.753716492989076e-06, "loss": 0.7065, "step": 460 }, { "epoch": 0.19254658385093168, "grad_norm": 2.942574456055828, "learning_rate": 9.742388654210822e-06, "loss": 0.7061, "step": 465 }, { "epoch": 0.19461697722567287, "grad_norm": 2.7983523129620926, "learning_rate": 9.730813007300984e-06, "loss": 0.7009, "step": 470 }, { "epoch": 0.19668737060041408, "grad_norm": 2.748714187875416, "learning_rate": 9.718990157131841e-06, "loss": 0.6953, "step": 475 }, { "epoch": 0.19875776397515527, "grad_norm": 2.7787928902685732, "learning_rate": 9.706920721492995e-06, "loss": 0.6792, "step": 480 }, { "epoch": 0.20082815734989648, "grad_norm": 2.8759174624811843, "learning_rate": 9.694605331059094e-06, "loss": 0.688, "step": 485 }, { "epoch": 0.2028985507246377, "grad_norm": 2.839118884637647, "learning_rate": 9.682044629356874e-06, "loss": 0.688, "step": 490 }, { "epoch": 0.20496894409937888, "grad_norm": 2.7979296188735674, "learning_rate": 9.669239272731538e-06, "loss": 0.6781, "step": 495 }, { "epoch": 0.2070393374741201, "grad_norm": 2.9252774557866124, "learning_rate": 9.656189930312443e-06, "loss": 0.6664, "step": 500 }, { "epoch": 0.20910973084886128, "grad_norm": 2.809436333693126, "learning_rate": 9.642897283978157e-06, "loss": 0.6406, "step": 505 }, { "epoch": 0.2111801242236025, "grad_norm": 2.937022586016644, "learning_rate": 9.629362028320808e-06, "loss": 0.6732, "step": 510 }, { "epoch": 0.21325051759834368, "grad_norm": 2.9843908796613356, "learning_rate": 9.615584870609809e-06, "loss": 0.6745, "step": 515 }, { "epoch": 0.2153209109730849, "grad_norm": 3.0213679386695427, "learning_rate": 9.601566530754882e-06, "loss": 0.7024, "step": 520 }, { "epoch": 0.21739130434782608, "grad_norm": 2.7454641085709515, "learning_rate": 9.587307741268452e-06, "loss": 0.6654, "step": 525 }, { "epoch": 0.2194616977225673, "grad_norm": 2.699259252813139, "learning_rate": 9.572809247227366e-06, "loss": 0.6816, "step": 530 }, { "epoch": 0.22153209109730848, "grad_norm": 2.856456590780441, "learning_rate": 9.558071806233955e-06, "loss": 0.6807, "step": 535 }, { "epoch": 0.2236024844720497, "grad_norm": 2.8884926905246817, "learning_rate": 9.54309618837646e-06, "loss": 0.675, "step": 540 }, { "epoch": 0.22567287784679088, "grad_norm": 2.7159778958415655, "learning_rate": 9.52788317618878e-06, "loss": 0.6693, "step": 545 }, { "epoch": 0.2277432712215321, "grad_norm": 2.7094266982916024, "learning_rate": 9.512433564609578e-06, "loss": 0.6584, "step": 550 }, { "epoch": 0.22981366459627328, "grad_norm": 2.684727550045969, "learning_rate": 9.496748160940762e-06, "loss": 0.6521, "step": 555 }, { "epoch": 0.2318840579710145, "grad_norm": 2.8152646867143747, "learning_rate": 9.480827784805278e-06, "loss": 0.6631, "step": 560 }, { "epoch": 0.23395445134575568, "grad_norm": 2.746998219194102, "learning_rate": 9.464673268104299e-06, "loss": 0.6649, "step": 565 }, { "epoch": 0.2360248447204969, "grad_norm": 2.6681124619837417, "learning_rate": 9.448285454973739e-06, "loss": 0.6641, "step": 570 }, { "epoch": 0.23809523809523808, "grad_norm": 2.7076660662242773, "learning_rate": 9.431665201740154e-06, "loss": 0.661, "step": 575 }, { "epoch": 0.2401656314699793, "grad_norm": 2.79748527458811, "learning_rate": 9.414813376876003e-06, "loss": 0.6544, "step": 580 }, { "epoch": 0.2422360248447205, "grad_norm": 2.6352068976041223, "learning_rate": 9.397730860954242e-06, "loss": 0.6656, "step": 585 }, { "epoch": 0.2443064182194617, "grad_norm": 2.697353214501619, "learning_rate": 9.38041854660234e-06, "loss": 0.6455, "step": 590 }, { "epoch": 0.2463768115942029, "grad_norm": 2.7822792953780184, "learning_rate": 9.362877338455611e-06, "loss": 0.6572, "step": 595 }, { "epoch": 0.2484472049689441, "grad_norm": 2.76479005730922, "learning_rate": 9.345108153109963e-06, "loss": 0.657, "step": 600 }, { "epoch": 0.2505175983436853, "grad_norm": 2.641137724563326, "learning_rate": 9.327111919073988e-06, "loss": 0.6322, "step": 605 }, { "epoch": 0.2525879917184265, "grad_norm": 2.7457640915517962, "learning_rate": 9.308889576720453e-06, "loss": 0.6263, "step": 610 }, { "epoch": 0.2546583850931677, "grad_norm": 2.8809597496679715, "learning_rate": 9.290442078237154e-06, "loss": 0.6447, "step": 615 }, { "epoch": 0.2567287784679089, "grad_norm": 2.649088481356515, "learning_rate": 9.271770387577168e-06, "loss": 0.6513, "step": 620 }, { "epoch": 0.2587991718426501, "grad_norm": 2.723263824128368, "learning_rate": 9.252875480408479e-06, "loss": 0.6363, "step": 625 }, { "epoch": 0.2608695652173913, "grad_norm": 2.643341522308482, "learning_rate": 9.233758344062996e-06, "loss": 0.6636, "step": 630 }, { "epoch": 0.2629399585921325, "grad_norm": 2.760214611828711, "learning_rate": 9.21441997748496e-06, "loss": 0.6149, "step": 635 }, { "epoch": 0.2650103519668737, "grad_norm": 2.7138292470479155, "learning_rate": 9.194861391178749e-06, "loss": 0.6409, "step": 640 }, { "epoch": 0.2670807453416149, "grad_norm": 2.7960171070921853, "learning_rate": 9.175083607156067e-06, "loss": 0.6291, "step": 645 }, { "epoch": 0.2691511387163561, "grad_norm": 2.6781478740703535, "learning_rate": 9.155087658882555e-06, "loss": 0.6294, "step": 650 }, { "epoch": 0.2712215320910973, "grad_norm": 2.78377570516491, "learning_rate": 9.134874591223773e-06, "loss": 0.6424, "step": 655 }, { "epoch": 0.2732919254658385, "grad_norm": 2.6278416215911915, "learning_rate": 9.114445460390605e-06, "loss": 0.6256, "step": 660 }, { "epoch": 0.2753623188405797, "grad_norm": 2.9435581582969244, "learning_rate": 9.093801333884076e-06, "loss": 0.6396, "step": 665 }, { "epoch": 0.2774327122153209, "grad_norm": 2.717501103920811, "learning_rate": 9.072943290439566e-06, "loss": 0.6334, "step": 670 }, { "epoch": 0.2795031055900621, "grad_norm": 2.722616988464195, "learning_rate": 9.051872419970439e-06, "loss": 0.6278, "step": 675 }, { "epoch": 0.2815734989648033, "grad_norm": 2.718385633125168, "learning_rate": 9.03058982351109e-06, "loss": 0.604, "step": 680 }, { "epoch": 0.2836438923395445, "grad_norm": 2.6781823046558384, "learning_rate": 9.009096613159426e-06, "loss": 0.6092, "step": 685 }, { "epoch": 0.2857142857142857, "grad_norm": 2.7824837364041133, "learning_rate": 8.987393912018732e-06, "loss": 0.6145, "step": 690 }, { "epoch": 0.28778467908902694, "grad_norm": 2.7837043936680623, "learning_rate": 8.965482854139003e-06, "loss": 0.5975, "step": 695 }, { "epoch": 0.2898550724637681, "grad_norm": 2.716842889874823, "learning_rate": 8.943364584457675e-06, "loss": 0.619, "step": 700 }, { "epoch": 0.2919254658385093, "grad_norm": 2.5861342696674106, "learning_rate": 8.921040258739804e-06, "loss": 0.6001, "step": 705 }, { "epoch": 0.2939958592132505, "grad_norm": 2.810107936750979, "learning_rate": 8.898511043517668e-06, "loss": 0.6183, "step": 710 }, { "epoch": 0.29606625258799174, "grad_norm": 2.7398243739315036, "learning_rate": 8.875778116029816e-06, "loss": 0.6021, "step": 715 }, { "epoch": 0.2981366459627329, "grad_norm": 2.70859978292154, "learning_rate": 8.85284266415955e-06, "loss": 0.5953, "step": 720 }, { "epoch": 0.3002070393374741, "grad_norm": 2.783835934538832, "learning_rate": 8.829705886372845e-06, "loss": 0.6013, "step": 725 }, { "epoch": 0.3022774327122153, "grad_norm": 2.5763975267274994, "learning_rate": 8.806368991655747e-06, "loss": 0.5878, "step": 730 }, { "epoch": 0.30434782608695654, "grad_norm": 2.697730711173743, "learning_rate": 8.782833199451177e-06, "loss": 0.6062, "step": 735 }, { "epoch": 0.3064182194616977, "grad_norm": 2.9279171282668024, "learning_rate": 8.759099739595215e-06, "loss": 0.5935, "step": 740 }, { "epoch": 0.3084886128364389, "grad_norm": 2.5900198753888373, "learning_rate": 8.735169852252848e-06, "loss": 0.5846, "step": 745 }, { "epoch": 0.3105590062111801, "grad_norm": 2.6427817212845337, "learning_rate": 8.71104478785315e-06, "loss": 0.5852, "step": 750 }, { "epoch": 0.31262939958592134, "grad_norm": 2.6213352415697724, "learning_rate": 8.686725807023955e-06, "loss": 0.5939, "step": 755 }, { "epoch": 0.3146997929606625, "grad_norm": 2.695782898397081, "learning_rate": 8.662214180525982e-06, "loss": 0.5855, "step": 760 }, { "epoch": 0.3167701863354037, "grad_norm": 2.6027809243052262, "learning_rate": 8.637511189186425e-06, "loss": 0.5807, "step": 765 }, { "epoch": 0.3188405797101449, "grad_norm": 2.769447274560436, "learning_rate": 8.612618123832033e-06, "loss": 0.5815, "step": 770 }, { "epoch": 0.32091097308488614, "grad_norm": 2.7990707978795246, "learning_rate": 8.587536285221656e-06, "loss": 0.5758, "step": 775 }, { "epoch": 0.32298136645962733, "grad_norm": 2.584119759399477, "learning_rate": 8.562266983978278e-06, "loss": 0.5939, "step": 780 }, { "epoch": 0.3250517598343685, "grad_norm": 2.8164096331343647, "learning_rate": 8.536811540520529e-06, "loss": 0.5866, "step": 785 }, { "epoch": 0.32712215320910976, "grad_norm": 4.038986317699086, "learning_rate": 8.511171284993686e-06, "loss": 0.5538, "step": 790 }, { "epoch": 0.32919254658385094, "grad_norm": 2.8362127741980334, "learning_rate": 8.485347557200177e-06, "loss": 0.5678, "step": 795 }, { "epoch": 0.33126293995859213, "grad_norm": 2.772547092346996, "learning_rate": 8.459341706529557e-06, "loss": 0.5762, "step": 800 }, { "epoch": 0.3333333333333333, "grad_norm": 2.523559738817573, "learning_rate": 8.43315509188801e-06, "loss": 0.5616, "step": 805 }, { "epoch": 0.33540372670807456, "grad_norm": 2.734190622994657, "learning_rate": 8.406789081627337e-06, "loss": 0.5747, "step": 810 }, { "epoch": 0.33747412008281574, "grad_norm": 2.5647167102786743, "learning_rate": 8.380245053473452e-06, "loss": 0.579, "step": 815 }, { "epoch": 0.33954451345755693, "grad_norm": 2.7181506919040817, "learning_rate": 8.353524394454388e-06, "loss": 0.5645, "step": 820 }, { "epoch": 0.3416149068322981, "grad_norm": 2.532384721512014, "learning_rate": 8.326628500827826e-06, "loss": 0.5816, "step": 825 }, { "epoch": 0.34368530020703936, "grad_norm": 2.7095413968694593, "learning_rate": 8.299558778008137e-06, "loss": 0.5685, "step": 830 }, { "epoch": 0.34575569358178054, "grad_norm": 2.6888726345421694, "learning_rate": 8.272316640492932e-06, "loss": 0.5608, "step": 835 }, { "epoch": 0.34782608695652173, "grad_norm": 2.6165321145769216, "learning_rate": 8.244903511789158e-06, "loss": 0.5589, "step": 840 }, { "epoch": 0.3498964803312629, "grad_norm": 2.688022401858765, "learning_rate": 8.217320824338713e-06, "loss": 0.567, "step": 845 }, { "epoch": 0.35196687370600416, "grad_norm": 2.6698972691855003, "learning_rate": 8.189570019443597e-06, "loss": 0.5608, "step": 850 }, { "epoch": 0.35403726708074534, "grad_norm": 2.5607516214652395, "learning_rate": 8.161652547190593e-06, "loss": 0.5544, "step": 855 }, { "epoch": 0.35610766045548653, "grad_norm": 2.5093610647637807, "learning_rate": 8.133569866375497e-06, "loss": 0.556, "step": 860 }, { "epoch": 0.3581780538302277, "grad_norm": 2.5888370930695754, "learning_rate": 8.105323444426891e-06, "loss": 0.5382, "step": 865 }, { "epoch": 0.36024844720496896, "grad_norm": 2.502429725724875, "learning_rate": 8.076914757329467e-06, "loss": 0.5515, "step": 870 }, { "epoch": 0.36231884057971014, "grad_norm": 2.513132604961173, "learning_rate": 8.048345289546895e-06, "loss": 0.5504, "step": 875 }, { "epoch": 0.36438923395445133, "grad_norm": 2.613023773895591, "learning_rate": 8.01961653394426e-06, "loss": 0.5514, "step": 880 }, { "epoch": 0.36645962732919257, "grad_norm": 2.8379598375055024, "learning_rate": 7.990729991710046e-06, "loss": 0.5359, "step": 885 }, { "epoch": 0.36853002070393376, "grad_norm": 2.6353250882091075, "learning_rate": 7.9616871722777e-06, "loss": 0.5607, "step": 890 }, { "epoch": 0.37060041407867494, "grad_norm": 2.695376888325521, "learning_rate": 7.932489593246764e-06, "loss": 0.5425, "step": 895 }, { "epoch": 0.37267080745341613, "grad_norm": 2.59578776213556, "learning_rate": 7.903138780303556e-06, "loss": 0.5513, "step": 900 }, { "epoch": 0.3747412008281574, "grad_norm": 2.6357999566585075, "learning_rate": 7.873636267141463e-06, "loss": 0.5189, "step": 905 }, { "epoch": 0.37681159420289856, "grad_norm": 2.610338652616046, "learning_rate": 7.843983595380793e-06, "loss": 0.5564, "step": 910 }, { "epoch": 0.37888198757763975, "grad_norm": 2.7200267334644566, "learning_rate": 7.814182314488225e-06, "loss": 0.5291, "step": 915 }, { "epoch": 0.38095238095238093, "grad_norm": 2.55575507245088, "learning_rate": 7.784233981695835e-06, "loss": 0.5491, "step": 920 }, { "epoch": 0.3830227743271222, "grad_norm": 2.741846594136525, "learning_rate": 7.754140161919732e-06, "loss": 0.5344, "step": 925 }, { "epoch": 0.38509316770186336, "grad_norm": 2.6008608315377555, "learning_rate": 7.72390242767828e-06, "loss": 0.5124, "step": 930 }, { "epoch": 0.38716356107660455, "grad_norm": 2.8320649567171197, "learning_rate": 7.693522359009931e-06, "loss": 0.5248, "step": 935 }, { "epoch": 0.38923395445134573, "grad_norm": 2.603319566659037, "learning_rate": 7.663001543390657e-06, "loss": 0.525, "step": 940 }, { "epoch": 0.391304347826087, "grad_norm": 2.560918488979843, "learning_rate": 7.63234157565101e-06, "loss": 0.5381, "step": 945 }, { "epoch": 0.39337474120082816, "grad_norm": 2.708900648839639, "learning_rate": 7.601544057892769e-06, "loss": 0.5188, "step": 950 }, { "epoch": 0.39544513457556935, "grad_norm": 2.557453498959152, "learning_rate": 7.570610599405242e-06, "loss": 0.5403, "step": 955 }, { "epoch": 0.39751552795031053, "grad_norm": 2.568997745239728, "learning_rate": 7.539542816581157e-06, "loss": 0.5306, "step": 960 }, { "epoch": 0.3995859213250518, "grad_norm": 2.849320989394085, "learning_rate": 7.508342332832213e-06, "loss": 0.5137, "step": 965 }, { "epoch": 0.40165631469979296, "grad_norm": 2.4808674765715066, "learning_rate": 7.477010778504241e-06, "loss": 0.5207, "step": 970 }, { "epoch": 0.40372670807453415, "grad_norm": 2.560284660092028, "learning_rate": 7.445549790792021e-06, "loss": 0.5192, "step": 975 }, { "epoch": 0.4057971014492754, "grad_norm": 2.553120554037133, "learning_rate": 7.413961013653725e-06, "loss": 0.5114, "step": 980 }, { "epoch": 0.4078674948240166, "grad_norm": 2.495494520049296, "learning_rate": 7.3822460977250145e-06, "loss": 0.5136, "step": 985 }, { "epoch": 0.40993788819875776, "grad_norm": 2.6954833669770726, "learning_rate": 7.350406700232794e-06, "loss": 0.5264, "step": 990 }, { "epoch": 0.41200828157349895, "grad_norm": 2.7381082998519983, "learning_rate": 7.318444484908606e-06, "loss": 0.5052, "step": 995 }, { "epoch": 0.4140786749482402, "grad_norm": 2.550035768584111, "learning_rate": 7.286361121901706e-06, "loss": 0.5133, "step": 1000 }, { "epoch": 0.4161490683229814, "grad_norm": 3.142045365448196, "learning_rate": 7.254158287691775e-06, "loss": 0.51, "step": 1005 }, { "epoch": 0.41821946169772256, "grad_norm": 2.558804554803966, "learning_rate": 7.221837665001335e-06, "loss": 0.5168, "step": 1010 }, { "epoch": 0.42028985507246375, "grad_norm": 2.5530037450130343, "learning_rate": 7.189400942707804e-06, "loss": 0.506, "step": 1015 }, { "epoch": 0.422360248447205, "grad_norm": 2.5368973829874775, "learning_rate": 7.1568498157552576e-06, "loss": 0.5105, "step": 1020 }, { "epoch": 0.4244306418219462, "grad_norm": 2.4485501120829034, "learning_rate": 7.124185985065856e-06, "loss": 0.5219, "step": 1025 }, { "epoch": 0.42650103519668736, "grad_norm": 2.5431441032943836, "learning_rate": 7.091411157450965e-06, "loss": 0.5068, "step": 1030 }, { "epoch": 0.42857142857142855, "grad_norm": 2.463753095314535, "learning_rate": 7.0585270455219654e-06, "loss": 0.4885, "step": 1035 }, { "epoch": 0.4306418219461698, "grad_norm": 2.6318223236200873, "learning_rate": 7.025535367600771e-06, "loss": 0.5078, "step": 1040 }, { "epoch": 0.432712215320911, "grad_norm": 2.600283059347969, "learning_rate": 6.992437847630031e-06, "loss": 0.5008, "step": 1045 }, { "epoch": 0.43478260869565216, "grad_norm": 2.508163120217184, "learning_rate": 6.95923621508305e-06, "loss": 0.4939, "step": 1050 }, { "epoch": 0.43685300207039335, "grad_norm": 2.660009117404832, "learning_rate": 6.92593220487342e-06, "loss": 0.4967, "step": 1055 }, { "epoch": 0.4389233954451346, "grad_norm": 2.524467102115331, "learning_rate": 6.892527557264358e-06, "loss": 0.5016, "step": 1060 }, { "epoch": 0.4409937888198758, "grad_norm": 2.5265295780956722, "learning_rate": 6.859024017777779e-06, "loss": 0.5029, "step": 1065 }, { "epoch": 0.44306418219461696, "grad_norm": 2.765020322815882, "learning_rate": 6.825423337103074e-06, "loss": 0.4803, "step": 1070 }, { "epoch": 0.4451345755693582, "grad_norm": 2.704943538082429, "learning_rate": 6.791727271005642e-06, "loss": 0.4967, "step": 1075 }, { "epoch": 0.4472049689440994, "grad_norm": 2.609473963979335, "learning_rate": 6.757937580235138e-06, "loss": 0.4907, "step": 1080 }, { "epoch": 0.4492753623188406, "grad_norm": 2.551608044305943, "learning_rate": 6.724056030433464e-06, "loss": 0.4928, "step": 1085 }, { "epoch": 0.45134575569358176, "grad_norm": 2.5603760994887255, "learning_rate": 6.690084392042514e-06, "loss": 0.4904, "step": 1090 }, { "epoch": 0.453416149068323, "grad_norm": 2.489337193865874, "learning_rate": 6.656024440211662e-06, "loss": 0.4751, "step": 1095 }, { "epoch": 0.4554865424430642, "grad_norm": 2.476219428614687, "learning_rate": 6.621877954704996e-06, "loss": 0.4808, "step": 1100 }, { "epoch": 0.4575569358178054, "grad_norm": 2.51661088671244, "learning_rate": 6.5876467198083235e-06, "loss": 0.4838, "step": 1105 }, { "epoch": 0.45962732919254656, "grad_norm": 2.625540683850569, "learning_rate": 6.553332524235937e-06, "loss": 0.4857, "step": 1110 }, { "epoch": 0.4616977225672878, "grad_norm": 2.6692887392725533, "learning_rate": 6.518937161037144e-06, "loss": 0.4766, "step": 1115 }, { "epoch": 0.463768115942029, "grad_norm": 2.451852906338064, "learning_rate": 6.484462427502572e-06, "loss": 0.4689, "step": 1120 }, { "epoch": 0.4658385093167702, "grad_norm": 2.528928504026905, "learning_rate": 6.44991012507026e-06, "loss": 0.47, "step": 1125 }, { "epoch": 0.46790890269151136, "grad_norm": 2.6023195877779797, "learning_rate": 6.415282059231518e-06, "loss": 0.472, "step": 1130 }, { "epoch": 0.4699792960662526, "grad_norm": 2.515973456448443, "learning_rate": 6.380580039436586e-06, "loss": 0.4524, "step": 1135 }, { "epoch": 0.4720496894409938, "grad_norm": 2.4273434418887607, "learning_rate": 6.345805879000087e-06, "loss": 0.4662, "step": 1140 }, { "epoch": 0.474120082815735, "grad_norm": 2.5041098379950375, "learning_rate": 6.31096139500627e-06, "loss": 0.4684, "step": 1145 }, { "epoch": 0.47619047619047616, "grad_norm": 2.890849402884173, "learning_rate": 6.2760484082140604e-06, "loss": 0.4817, "step": 1150 }, { "epoch": 0.4782608695652174, "grad_norm": 2.4406551900826843, "learning_rate": 6.24106874296192e-06, "loss": 0.479, "step": 1155 }, { "epoch": 0.4803312629399586, "grad_norm": 2.581296500428043, "learning_rate": 6.20602422707252e-06, "loss": 0.4626, "step": 1160 }, { "epoch": 0.4824016563146998, "grad_norm": 2.462136348090071, "learning_rate": 6.1709166917572264e-06, "loss": 0.4796, "step": 1165 }, { "epoch": 0.484472049689441, "grad_norm": 2.6803789589303646, "learning_rate": 6.135747971520412e-06, "loss": 0.4728, "step": 1170 }, { "epoch": 0.4865424430641822, "grad_norm": 2.4077364719321985, "learning_rate": 6.100519904063597e-06, "loss": 0.4623, "step": 1175 }, { "epoch": 0.4886128364389234, "grad_norm": 2.4101411199940315, "learning_rate": 6.0652343301894345e-06, "loss": 0.4657, "step": 1180 }, { "epoch": 0.4906832298136646, "grad_norm": 2.5188929556410415, "learning_rate": 6.029893093705492e-06, "loss": 0.46, "step": 1185 }, { "epoch": 0.4927536231884058, "grad_norm": 2.3677717501653563, "learning_rate": 5.99449804132794e-06, "loss": 0.4757, "step": 1190 }, { "epoch": 0.494824016563147, "grad_norm": 2.5710734020272157, "learning_rate": 5.959051022585025e-06, "loss": 0.473, "step": 1195 }, { "epoch": 0.4968944099378882, "grad_norm": 2.836194471253979, "learning_rate": 5.923553889720447e-06, "loss": 0.4504, "step": 1200 }, { "epoch": 0.4989648033126294, "grad_norm": 2.5393687325157073, "learning_rate": 5.888008497596553e-06, "loss": 0.4609, "step": 1205 }, { "epoch": 0.5010351966873706, "grad_norm": 2.4510897506943845, "learning_rate": 5.852416703597431e-06, "loss": 0.4696, "step": 1210 }, { "epoch": 0.5031055900621118, "grad_norm": 2.4297048949306435, "learning_rate": 5.816780367531841e-06, "loss": 0.4469, "step": 1215 }, { "epoch": 0.505175983436853, "grad_norm": 2.456463228901228, "learning_rate": 5.781101351536041e-06, "loss": 0.4541, "step": 1220 }, { "epoch": 0.5072463768115942, "grad_norm": 2.44277926327073, "learning_rate": 5.745381519976477e-06, "loss": 0.4504, "step": 1225 }, { "epoch": 0.5093167701863354, "grad_norm": 2.5010485725693328, "learning_rate": 5.7096227393523716e-06, "loss": 0.4613, "step": 1230 }, { "epoch": 0.5113871635610766, "grad_norm": 2.6469453128675613, "learning_rate": 5.673826878198181e-06, "loss": 0.4493, "step": 1235 }, { "epoch": 0.5134575569358178, "grad_norm": 2.4138425627962636, "learning_rate": 5.637995806985961e-06, "loss": 0.4546, "step": 1240 }, { "epoch": 0.515527950310559, "grad_norm": 2.7188944082485924, "learning_rate": 5.602131398027637e-06, "loss": 0.4527, "step": 1245 }, { "epoch": 0.5175983436853002, "grad_norm": 2.3956639841305893, "learning_rate": 5.566235525377155e-06, "loss": 0.4456, "step": 1250 }, { "epoch": 0.5196687370600414, "grad_norm": 2.667940884594386, "learning_rate": 5.530310064732559e-06, "loss": 0.4475, "step": 1255 }, { "epoch": 0.5217391304347826, "grad_norm": 2.568790625515596, "learning_rate": 5.494356893337985e-06, "loss": 0.4603, "step": 1260 }, { "epoch": 0.5238095238095238, "grad_norm": 2.537093225337417, "learning_rate": 5.4583778898855576e-06, "loss": 0.4517, "step": 1265 }, { "epoch": 0.525879917184265, "grad_norm": 2.377214097836386, "learning_rate": 5.422374934417228e-06, "loss": 0.4334, "step": 1270 }, { "epoch": 0.5279503105590062, "grad_norm": 2.5322732831198307, "learning_rate": 5.386349908226538e-06, "loss": 0.443, "step": 1275 }, { "epoch": 0.5300207039337475, "grad_norm": 2.275812013448177, "learning_rate": 5.350304693760301e-06, "loss": 0.452, "step": 1280 }, { "epoch": 0.5320910973084886, "grad_norm": 2.490540393564678, "learning_rate": 5.314241174520251e-06, "loss": 0.4571, "step": 1285 }, { "epoch": 0.5341614906832298, "grad_norm": 2.303916018471547, "learning_rate": 5.2781612349646175e-06, "loss": 0.4437, "step": 1290 }, { "epoch": 0.5362318840579711, "grad_norm": 2.4742224611265367, "learning_rate": 5.242066760409653e-06, "loss": 0.4419, "step": 1295 }, { "epoch": 0.5383022774327122, "grad_norm": 2.6048926561003323, "learning_rate": 5.205959636931121e-06, "loss": 0.4334, "step": 1300 }, { "epoch": 0.5403726708074534, "grad_norm": 2.4996195423198975, "learning_rate": 5.169841751265738e-06, "loss": 0.4477, "step": 1305 }, { "epoch": 0.5424430641821946, "grad_norm": 2.436615222988275, "learning_rate": 5.13371499071259e-06, "loss": 0.4227, "step": 1310 }, { "epoch": 0.5445134575569358, "grad_norm": 2.4895563915580468, "learning_rate": 5.09758124303451e-06, "loss": 0.4382, "step": 1315 }, { "epoch": 0.546583850931677, "grad_norm": 2.457090055709984, "learning_rate": 5.0614423963594295e-06, "loss": 0.4383, "step": 1320 }, { "epoch": 0.5486542443064182, "grad_norm": 2.46529676282277, "learning_rate": 5.0253003390817264e-06, "loss": 0.4404, "step": 1325 }, { "epoch": 0.5507246376811594, "grad_norm": 2.509514530174065, "learning_rate": 4.989156959763551e-06, "loss": 0.432, "step": 1330 }, { "epoch": 0.5527950310559007, "grad_norm": 2.3489886489521568, "learning_rate": 4.953014147036121e-06, "loss": 0.4238, "step": 1335 }, { "epoch": 0.5548654244306418, "grad_norm": 2.548572605165313, "learning_rate": 4.9168737895010615e-06, "loss": 0.4385, "step": 1340 }, { "epoch": 0.556935817805383, "grad_norm": 2.5754573965566188, "learning_rate": 4.880737775631698e-06, "loss": 0.4188, "step": 1345 }, { "epoch": 0.5590062111801242, "grad_norm": 2.4663169631010633, "learning_rate": 4.844607993674382e-06, "loss": 0.4277, "step": 1350 }, { "epoch": 0.5610766045548654, "grad_norm": 2.5893970305781124, "learning_rate": 4.808486331549824e-06, "loss": 0.4246, "step": 1355 }, { "epoch": 0.5631469979296067, "grad_norm": 2.5461127745316987, "learning_rate": 4.772374676754444e-06, "loss": 0.4269, "step": 1360 }, { "epoch": 0.5652173913043478, "grad_norm": 2.4510878027098784, "learning_rate": 4.736274916261741e-06, "loss": 0.4227, "step": 1365 }, { "epoch": 0.567287784679089, "grad_norm": 2.4337437356788882, "learning_rate": 4.700188936423683e-06, "loss": 0.4301, "step": 1370 }, { "epoch": 0.5693581780538303, "grad_norm": 2.616623513135904, "learning_rate": 4.664118622872157e-06, "loss": 0.4234, "step": 1375 }, { "epoch": 0.5714285714285714, "grad_norm": 2.3671956167402874, "learning_rate": 4.628065860420417e-06, "loss": 0.4269, "step": 1380 }, { "epoch": 0.5734989648033126, "grad_norm": 2.484279489532884, "learning_rate": 4.592032532964611e-06, "loss": 0.4354, "step": 1385 }, { "epoch": 0.5755693581780539, "grad_norm": 2.4601006150341416, "learning_rate": 4.556020523385326e-06, "loss": 0.4245, "step": 1390 }, { "epoch": 0.577639751552795, "grad_norm": 2.6097660668896454, "learning_rate": 4.520031713449215e-06, "loss": 0.4295, "step": 1395 }, { "epoch": 0.5797101449275363, "grad_norm": 2.4072399455791658, "learning_rate": 4.484067983710653e-06, "loss": 0.4234, "step": 1400 }, { "epoch": 0.5817805383022774, "grad_norm": 2.448489377931204, "learning_rate": 4.448131213413485e-06, "loss": 0.4203, "step": 1405 }, { "epoch": 0.5838509316770186, "grad_norm": 2.443669450647813, "learning_rate": 4.4122232803928145e-06, "loss": 0.4197, "step": 1410 }, { "epoch": 0.5859213250517599, "grad_norm": 2.4564289974274485, "learning_rate": 4.376346060976888e-06, "loss": 0.4136, "step": 1415 }, { "epoch": 0.587991718426501, "grad_norm": 2.5200444579378933, "learning_rate": 4.340501429889053e-06, "loss": 0.4151, "step": 1420 }, { "epoch": 0.5900621118012422, "grad_norm": 2.48165117649947, "learning_rate": 4.30469126014978e-06, "loss": 0.402, "step": 1425 }, { "epoch": 0.5921325051759835, "grad_norm": 2.4558613956242845, "learning_rate": 4.268917422978811e-06, "loss": 0.4059, "step": 1430 }, { "epoch": 0.5942028985507246, "grad_norm": 2.478521562259034, "learning_rate": 4.233181787697364e-06, "loss": 0.4234, "step": 1435 }, { "epoch": 0.5962732919254659, "grad_norm": 2.3579770809913656, "learning_rate": 4.197486221630467e-06, "loss": 0.4177, "step": 1440 }, { "epoch": 0.598343685300207, "grad_norm": 2.440943795986475, "learning_rate": 4.161832590009371e-06, "loss": 0.4136, "step": 1445 }, { "epoch": 0.6004140786749482, "grad_norm": 2.3924731794257434, "learning_rate": 4.1262227558740915e-06, "loss": 0.4097, "step": 1450 }, { "epoch": 0.6024844720496895, "grad_norm": 2.3853646352098914, "learning_rate": 4.090658579976058e-06, "loss": 0.4125, "step": 1455 }, { "epoch": 0.6045548654244306, "grad_norm": 2.476730861427627, "learning_rate": 4.055141920680883e-06, "loss": 0.411, "step": 1460 }, { "epoch": 0.6066252587991718, "grad_norm": 2.338110722995676, "learning_rate": 4.019674633871246e-06, "loss": 0.4053, "step": 1465 }, { "epoch": 0.6086956521739131, "grad_norm": 2.3987001096487726, "learning_rate": 3.984258572849926e-06, "loss": 0.4136, "step": 1470 }, { "epoch": 0.6107660455486542, "grad_norm": 2.3639908244617422, "learning_rate": 3.9488955882429605e-06, "loss": 0.4093, "step": 1475 }, { "epoch": 0.6128364389233955, "grad_norm": 2.422313766220588, "learning_rate": 3.9135875279029366e-06, "loss": 0.4104, "step": 1480 }, { "epoch": 0.6149068322981367, "grad_norm": 2.4728098446592854, "learning_rate": 3.878336236812432e-06, "loss": 0.3814, "step": 1485 }, { "epoch": 0.6169772256728778, "grad_norm": 2.3708630478506176, "learning_rate": 3.84314355698762e-06, "loss": 0.4105, "step": 1490 }, { "epoch": 0.6190476190476191, "grad_norm": 2.4275761087290304, "learning_rate": 3.8080113273820024e-06, "loss": 0.4058, "step": 1495 }, { "epoch": 0.6211180124223602, "grad_norm": 2.4421724912048077, "learning_rate": 3.7729413837903285e-06, "loss": 0.3913, "step": 1500 }, { "epoch": 0.6231884057971014, "grad_norm": 2.4517011088536, "learning_rate": 3.7379355587526543e-06, "loss": 0.3962, "step": 1505 }, { "epoch": 0.6252587991718427, "grad_norm": 2.4069249372010946, "learning_rate": 3.702995681458605e-06, "loss": 0.3941, "step": 1510 }, { "epoch": 0.6273291925465838, "grad_norm": 2.323725479679109, "learning_rate": 3.668123577651773e-06, "loss": 0.3984, "step": 1515 }, { "epoch": 0.629399585921325, "grad_norm": 2.3190346458499884, "learning_rate": 3.633321069534325e-06, "loss": 0.3977, "step": 1520 }, { "epoch": 0.6314699792960663, "grad_norm": 2.335078659342428, "learning_rate": 3.5985899756717872e-06, "loss": 0.4047, "step": 1525 }, { "epoch": 0.6335403726708074, "grad_norm": 2.455638307213381, "learning_rate": 3.563932110898015e-06, "loss": 0.3912, "step": 1530 }, { "epoch": 0.6356107660455487, "grad_norm": 2.3702634454791993, "learning_rate": 3.5293492862203595e-06, "loss": 0.3847, "step": 1535 }, { "epoch": 0.6376811594202898, "grad_norm": 2.3580779842802753, "learning_rate": 3.494843308725032e-06, "loss": 0.3955, "step": 1540 }, { "epoch": 0.639751552795031, "grad_norm": 2.328047105272821, "learning_rate": 3.460415981482691e-06, "loss": 0.3915, "step": 1545 }, { "epoch": 0.6418219461697723, "grad_norm": 2.4764402089768476, "learning_rate": 3.426069103454209e-06, "loss": 0.3886, "step": 1550 }, { "epoch": 0.6438923395445134, "grad_norm": 2.454108156460191, "learning_rate": 3.3918044693966802e-06, "loss": 0.3842, "step": 1555 }, { "epoch": 0.6459627329192547, "grad_norm": 2.441202869593845, "learning_rate": 3.357623869769628e-06, "loss": 0.3724, "step": 1560 }, { "epoch": 0.6480331262939959, "grad_norm": 2.4755133325255256, "learning_rate": 3.3235290906414575e-06, "loss": 0.3822, "step": 1565 }, { "epoch": 0.650103519668737, "grad_norm": 2.4311073492384336, "learning_rate": 3.289521913596117e-06, "loss": 0.3908, "step": 1570 }, { "epoch": 0.6521739130434783, "grad_norm": 2.381193945822083, "learning_rate": 3.255604115640014e-06, "loss": 0.3825, "step": 1575 }, { "epoch": 0.6542443064182195, "grad_norm": 2.350792909332454, "learning_rate": 3.2217774691091437e-06, "loss": 0.3946, "step": 1580 }, { "epoch": 0.6563146997929606, "grad_norm": 2.390554035306188, "learning_rate": 3.1880437415764936e-06, "loss": 0.3993, "step": 1585 }, { "epoch": 0.6583850931677019, "grad_norm": 2.329131133970712, "learning_rate": 3.1544046957596746e-06, "loss": 0.3884, "step": 1590 }, { "epoch": 0.660455486542443, "grad_norm": 2.5542279992917902, "learning_rate": 3.1208620894288105e-06, "loss": 0.3908, "step": 1595 }, { "epoch": 0.6625258799171843, "grad_norm": 2.3701687553879243, "learning_rate": 3.087417675314687e-06, "loss": 0.3823, "step": 1600 }, { "epoch": 0.6645962732919255, "grad_norm": 2.225525354523903, "learning_rate": 3.0540732010171704e-06, "loss": 0.3733, "step": 1605 }, { "epoch": 0.6666666666666666, "grad_norm": 2.366938192597044, "learning_rate": 3.0208304089138873e-06, "loss": 0.383, "step": 1610 }, { "epoch": 0.6687370600414079, "grad_norm": 2.4030474398123953, "learning_rate": 2.9876910360691707e-06, "loss": 0.3919, "step": 1615 }, { "epoch": 0.6708074534161491, "grad_norm": 2.4354010216355992, "learning_rate": 2.9546568141433007e-06, "loss": 0.3868, "step": 1620 }, { "epoch": 0.6728778467908902, "grad_norm": 2.356177911915648, "learning_rate": 2.921729469302018e-06, "loss": 0.3803, "step": 1625 }, { "epoch": 0.6749482401656315, "grad_norm": 2.455281456899517, "learning_rate": 2.8889107221263203e-06, "loss": 0.3713, "step": 1630 }, { "epoch": 0.6770186335403726, "grad_norm": 2.221742782748604, "learning_rate": 2.856202287522556e-06, "loss": 0.3766, "step": 1635 }, { "epoch": 0.6790890269151139, "grad_norm": 2.6070082952771787, "learning_rate": 2.8236058746328203e-06, "loss": 0.3743, "step": 1640 }, { "epoch": 0.6811594202898551, "grad_norm": 2.3423694128380292, "learning_rate": 2.7911231867456345e-06, "loss": 0.3684, "step": 1645 }, { "epoch": 0.6832298136645962, "grad_norm": 2.242939946393628, "learning_rate": 2.7587559212069554e-06, "loss": 0.4181, "step": 1650 }, { "epoch": 0.6853002070393375, "grad_norm": 2.4538446272076775, "learning_rate": 2.72650576933147e-06, "loss": 0.3695, "step": 1655 }, { "epoch": 0.6873706004140787, "grad_norm": 2.248569538212587, "learning_rate": 2.6943744163142292e-06, "loss": 0.3634, "step": 1660 }, { "epoch": 0.6894409937888198, "grad_norm": 2.4019909678857494, "learning_rate": 2.662363541142585e-06, "loss": 0.3766, "step": 1665 }, { "epoch": 0.6915113871635611, "grad_norm": 2.571206708691807, "learning_rate": 2.6304748165084503e-06, "loss": 0.375, "step": 1670 }, { "epoch": 0.6935817805383023, "grad_norm": 2.434809314229936, "learning_rate": 2.5987099087209035e-06, "loss": 0.3693, "step": 1675 }, { "epoch": 0.6956521739130435, "grad_norm": 2.339716705035725, "learning_rate": 2.5670704776191158e-06, "loss": 0.3673, "step": 1680 }, { "epoch": 0.6977225672877847, "grad_norm": 2.442360671231122, "learning_rate": 2.5355581764856186e-06, "loss": 0.37, "step": 1685 }, { "epoch": 0.6997929606625258, "grad_norm": 2.356305225228509, "learning_rate": 2.5041746519599036e-06, "loss": 0.3722, "step": 1690 }, { "epoch": 0.7018633540372671, "grad_norm": 2.400871546850089, "learning_rate": 2.4729215439523945e-06, "loss": 0.378, "step": 1695 }, { "epoch": 0.7039337474120083, "grad_norm": 2.485733684370957, "learning_rate": 2.441800485558748e-06, "loss": 0.368, "step": 1700 }, { "epoch": 0.7060041407867494, "grad_norm": 2.4153165474956295, "learning_rate": 2.410813102974512e-06, "loss": 0.374, "step": 1705 }, { "epoch": 0.7080745341614907, "grad_norm": 2.417914099527471, "learning_rate": 2.3799610154101603e-06, "loss": 0.3649, "step": 1710 }, { "epoch": 0.7101449275362319, "grad_norm": 2.421566323056135, "learning_rate": 2.3492458350064805e-06, "loss": 0.3722, "step": 1715 }, { "epoch": 0.7122153209109731, "grad_norm": 110.14431209902904, "learning_rate": 2.3186691667503315e-06, "loss": 0.376, "step": 1720 }, { "epoch": 0.7142857142857143, "grad_norm": 2.498987084667395, "learning_rate": 2.288232608390778e-06, "loss": 0.367, "step": 1725 }, { "epoch": 0.7163561076604554, "grad_norm": 2.377271205692582, "learning_rate": 2.2579377503555967e-06, "loss": 0.364, "step": 1730 }, { "epoch": 0.7184265010351967, "grad_norm": 2.40592016919288, "learning_rate": 2.227786175668181e-06, "loss": 0.3661, "step": 1735 }, { "epoch": 0.7204968944099379, "grad_norm": 2.4853659060167748, "learning_rate": 2.1977794598648106e-06, "loss": 0.3566, "step": 1740 }, { "epoch": 0.722567287784679, "grad_norm": 2.5196941482373556, "learning_rate": 2.167919170912333e-06, "loss": 0.3679, "step": 1745 }, { "epoch": 0.7246376811594203, "grad_norm": 2.3428574169962872, "learning_rate": 2.138206869126222e-06, "loss": 0.3534, "step": 1750 }, { "epoch": 0.7267080745341615, "grad_norm": 2.3854557500475675, "learning_rate": 2.108644107089055e-06, "loss": 0.3584, "step": 1755 }, { "epoch": 0.7287784679089027, "grad_norm": 2.5623905943162635, "learning_rate": 2.0792324295693782e-06, "loss": 0.3663, "step": 1760 }, { "epoch": 0.7308488612836439, "grad_norm": 2.1463966743523235, "learning_rate": 2.0499733734409884e-06, "loss": 0.3689, "step": 1765 }, { "epoch": 0.7329192546583851, "grad_norm": 2.3434379762931234, "learning_rate": 2.020868467602623e-06, "loss": 0.3501, "step": 1770 }, { "epoch": 0.7349896480331263, "grad_norm": 2.466102573223308, "learning_rate": 1.9919192328980703e-06, "loss": 0.3406, "step": 1775 }, { "epoch": 0.7370600414078675, "grad_norm": 2.441564402693419, "learning_rate": 1.9631271820367058e-06, "loss": 0.3555, "step": 1780 }, { "epoch": 0.7391304347826086, "grad_norm": 2.474880211520262, "learning_rate": 1.9344938195144353e-06, "loss": 0.351, "step": 1785 }, { "epoch": 0.7412008281573499, "grad_norm": 2.482728988847221, "learning_rate": 1.9060206415350896e-06, "loss": 0.3638, "step": 1790 }, { "epoch": 0.7432712215320911, "grad_norm": 2.3721058207362, "learning_rate": 1.8777091359322386e-06, "loss": 0.3553, "step": 1795 }, { "epoch": 0.7453416149068323, "grad_norm": 2.3780878490180735, "learning_rate": 1.8495607820914451e-06, "loss": 0.3563, "step": 1800 }, { "epoch": 0.7474120082815735, "grad_norm": 2.5716332485081383, "learning_rate": 1.8215770508729602e-06, "loss": 0.3569, "step": 1805 }, { "epoch": 0.7494824016563147, "grad_norm": 2.353530194343243, "learning_rate": 1.7937594045348666e-06, "loss": 0.3394, "step": 1810 }, { "epoch": 0.7515527950310559, "grad_norm": 2.3569526754259504, "learning_rate": 1.7661092966566728e-06, "loss": 0.3481, "step": 1815 }, { "epoch": 0.7536231884057971, "grad_norm": 2.649114709642696, "learning_rate": 1.7386281720633569e-06, "loss": 0.3643, "step": 1820 }, { "epoch": 0.7556935817805382, "grad_norm": 2.460060042283697, "learning_rate": 1.7113174667498612e-06, "loss": 0.3576, "step": 1825 }, { "epoch": 0.7577639751552795, "grad_norm": 2.339978220484055, "learning_rate": 1.6841786078060674e-06, "loss": 0.3558, "step": 1830 }, { "epoch": 0.7598343685300207, "grad_norm": 2.4953672860958975, "learning_rate": 1.6572130133422204e-06, "loss": 0.3457, "step": 1835 }, { "epoch": 0.7619047619047619, "grad_norm": 2.426133939030405, "learning_rate": 1.630422092414823e-06, "loss": 0.3482, "step": 1840 }, { "epoch": 0.7639751552795031, "grad_norm": 2.459777432451568, "learning_rate": 1.6038072449530118e-06, "loss": 0.35, "step": 1845 }, { "epoch": 0.7660455486542443, "grad_norm": 2.368741592685829, "learning_rate": 1.5773698616854065e-06, "loss": 0.3467, "step": 1850 }, { "epoch": 0.7681159420289855, "grad_norm": 2.2896780011923488, "learning_rate": 1.5511113240674364e-06, "loss": 0.358, "step": 1855 }, { "epoch": 0.7701863354037267, "grad_norm": 2.3730136202177143, "learning_rate": 1.5250330042091543e-06, "loss": 0.3481, "step": 1860 }, { "epoch": 0.772256728778468, "grad_norm": 2.464479876236929, "learning_rate": 1.4991362648035368e-06, "loss": 0.3431, "step": 1865 }, { "epoch": 0.7743271221532091, "grad_norm": 2.391982798119773, "learning_rate": 1.4734224590552844e-06, "loss": 0.3368, "step": 1870 }, { "epoch": 0.7763975155279503, "grad_norm": 2.431774830394956, "learning_rate": 1.4478929306101025e-06, "loss": 0.3509, "step": 1875 }, { "epoch": 0.7784679089026915, "grad_norm": 2.2880451270645614, "learning_rate": 1.4225490134844967e-06, "loss": 0.3376, "step": 1880 }, { "epoch": 0.7805383022774327, "grad_norm": 2.541490709582719, "learning_rate": 1.3973920319960654e-06, "loss": 0.3422, "step": 1885 }, { "epoch": 0.782608695652174, "grad_norm": 2.486644761227762, "learning_rate": 1.3724233006942972e-06, "loss": 0.3486, "step": 1890 }, { "epoch": 0.7846790890269151, "grad_norm": 2.222907731738531, "learning_rate": 1.347644124291883e-06, "loss": 0.3401, "step": 1895 }, { "epoch": 0.7867494824016563, "grad_norm": 2.503229276614086, "learning_rate": 1.323055797596532e-06, "loss": 0.3423, "step": 1900 }, { "epoch": 0.7888198757763976, "grad_norm": 2.44333238565998, "learning_rate": 1.2986596054433255e-06, "loss": 0.3428, "step": 1905 }, { "epoch": 0.7908902691511387, "grad_norm": 2.2546076468429592, "learning_rate": 1.2744568226275662e-06, "loss": 0.3367, "step": 1910 }, { "epoch": 0.7929606625258799, "grad_norm": 2.241016793849244, "learning_rate": 1.2504487138381782e-06, "loss": 0.3301, "step": 1915 }, { "epoch": 0.7950310559006211, "grad_norm": 2.472605862322622, "learning_rate": 1.2266365335916086e-06, "loss": 0.3378, "step": 1920 }, { "epoch": 0.7971014492753623, "grad_norm": 2.235182498951894, "learning_rate": 1.203021526166287e-06, "loss": 0.3387, "step": 1925 }, { "epoch": 0.7991718426501035, "grad_norm": 2.3511754008831516, "learning_rate": 1.179604925537598e-06, "loss": 0.3328, "step": 1930 }, { "epoch": 0.8012422360248447, "grad_norm": 2.4302698271655863, "learning_rate": 1.1563879553134072e-06, "loss": 0.3479, "step": 1935 }, { "epoch": 0.8033126293995859, "grad_norm": 2.4051091564094724, "learning_rate": 1.1333718286701162e-06, "loss": 0.345, "step": 1940 }, { "epoch": 0.8053830227743272, "grad_norm": 2.4216508859806742, "learning_rate": 1.110557748289275e-06, "loss": 0.3362, "step": 1945 }, { "epoch": 0.8074534161490683, "grad_norm": 2.518070754468628, "learning_rate": 1.0879469062947369e-06, "loss": 0.3406, "step": 1950 }, { "epoch": 0.8095238095238095, "grad_norm": 2.4384803076253476, "learning_rate": 1.0655404841903648e-06, "loss": 0.3558, "step": 1955 }, { "epoch": 0.8115942028985508, "grad_norm": 2.6344448059599412, "learning_rate": 1.0433396527982893e-06, "loss": 0.3346, "step": 1960 }, { "epoch": 0.8136645962732919, "grad_norm": 2.3007277716039396, "learning_rate": 1.0213455721977356e-06, "loss": 0.3347, "step": 1965 }, { "epoch": 0.8157349896480331, "grad_norm": 2.3043896070313137, "learning_rate": 9.995593916643981e-07, "loss": 0.3273, "step": 1970 }, { "epoch": 0.8178053830227743, "grad_norm": 2.4269750741096465, "learning_rate": 9.779822496103902e-07, "loss": 0.3288, "step": 1975 }, { "epoch": 0.8198757763975155, "grad_norm": 2.370989013074573, "learning_rate": 9.566152735247576e-07, "loss": 0.3322, "step": 1980 }, { "epoch": 0.8219461697722568, "grad_norm": 2.2322391420672254, "learning_rate": 9.354595799145627e-07, "loss": 0.3302, "step": 1985 }, { "epoch": 0.8240165631469979, "grad_norm": 2.3138733868403896, "learning_rate": 9.145162742465424e-07, "loss": 0.3281, "step": 1990 }, { "epoch": 0.8260869565217391, "grad_norm": 2.218401312393179, "learning_rate": 8.937864508893396e-07, "loss": 0.3263, "step": 1995 }, { "epoch": 0.8281573498964804, "grad_norm": 2.501783210452481, "learning_rate": 8.732711930563259e-07, "loss": 0.3383, "step": 2000 }, { "epoch": 0.8302277432712215, "grad_norm": 2.339035923113422, "learning_rate": 8.529715727489912e-07, "loss": 0.3331, "step": 2005 }, { "epoch": 0.8322981366459627, "grad_norm": 2.419266862744491, "learning_rate": 8.32888650700936e-07, "loss": 0.3326, "step": 2010 }, { "epoch": 0.8343685300207039, "grad_norm": 2.3535716478661994, "learning_rate": 8.130234763224342e-07, "loss": 0.3383, "step": 2015 }, { "epoch": 0.8364389233954451, "grad_norm": 2.348419640739673, "learning_rate": 7.93377087645607e-07, "loss": 0.3277, "step": 2020 }, { "epoch": 0.8385093167701864, "grad_norm": 2.5091270230983422, "learning_rate": 7.739505112701751e-07, "loss": 0.3375, "step": 2025 }, { "epoch": 0.8405797101449275, "grad_norm": 2.4359136230459524, "learning_rate": 7.547447623098191e-07, "loss": 0.3258, "step": 2030 }, { "epoch": 0.8426501035196687, "grad_norm": 2.432506753046153, "learning_rate": 7.357608443391312e-07, "loss": 0.3291, "step": 2035 }, { "epoch": 0.84472049689441, "grad_norm": 2.5646276059325706, "learning_rate": 7.169997493411774e-07, "loss": 0.3344, "step": 2040 }, { "epoch": 0.8467908902691511, "grad_norm": 2.3486928663224, "learning_rate": 6.984624576556647e-07, "loss": 0.3271, "step": 2045 }, { "epoch": 0.8488612836438924, "grad_norm": 2.203166381807686, "learning_rate": 6.801499379277115e-07, "loss": 0.3167, "step": 2050 }, { "epoch": 0.8509316770186336, "grad_norm": 2.2436664305021385, "learning_rate": 6.620631470572314e-07, "loss": 0.3307, "step": 2055 }, { "epoch": 0.8530020703933747, "grad_norm": 2.3594907816627875, "learning_rate": 6.442030301489338e-07, "loss": 0.3164, "step": 2060 }, { "epoch": 0.855072463768116, "grad_norm": 2.256989647657127, "learning_rate": 6.265705204629402e-07, "loss": 0.3133, "step": 2065 }, { "epoch": 0.8571428571428571, "grad_norm": 2.3571020694555798, "learning_rate": 6.091665393660112e-07, "loss": 0.3203, "step": 2070 }, { "epoch": 0.8592132505175983, "grad_norm": 2.3091661794248157, "learning_rate": 5.919919962834069e-07, "loss": 0.3224, "step": 2075 }, { "epoch": 0.8612836438923396, "grad_norm": 2.2740550403647894, "learning_rate": 5.750477886513656e-07, "loss": 0.3285, "step": 2080 }, { "epoch": 0.8633540372670807, "grad_norm": 2.256740351456034, "learning_rate": 5.58334801870209e-07, "loss": 0.3178, "step": 2085 }, { "epoch": 0.865424430641822, "grad_norm": 2.385870564024283, "learning_rate": 5.418539092580727e-07, "loss": 0.329, "step": 2090 }, { "epoch": 0.8674948240165632, "grad_norm": 2.215079196146054, "learning_rate": 5.256059720052787e-07, "loss": 0.3303, "step": 2095 }, { "epoch": 0.8695652173913043, "grad_norm": 2.2942059659033833, "learning_rate": 5.095918391293303e-07, "loss": 0.3279, "step": 2100 }, { "epoch": 0.8716356107660456, "grad_norm": 2.5461731820432076, "learning_rate": 4.938123474305473e-07, "loss": 0.3222, "step": 2105 }, { "epoch": 0.8737060041407867, "grad_norm": 2.4819509581487846, "learning_rate": 4.782683214483413e-07, "loss": 0.3185, "step": 2110 }, { "epoch": 0.8757763975155279, "grad_norm": 2.357273267066927, "learning_rate": 4.629605734181331e-07, "loss": 0.316, "step": 2115 }, { "epoch": 0.8778467908902692, "grad_norm": 2.26836858721712, "learning_rate": 4.478899032289047e-07, "loss": 0.3191, "step": 2120 }, { "epoch": 0.8799171842650103, "grad_norm": 2.339531232436072, "learning_rate": 4.3305709838140755e-07, "loss": 0.3221, "step": 2125 }, { "epoch": 0.8819875776397516, "grad_norm": 2.354666153598989, "learning_rate": 4.1846293394700645e-07, "loss": 0.3291, "step": 2130 }, { "epoch": 0.8840579710144928, "grad_norm": 2.5023494902865484, "learning_rate": 4.041081725271856e-07, "loss": 0.3319, "step": 2135 }, { "epoch": 0.8861283643892339, "grad_norm": 2.660892225845494, "learning_rate": 3.8999356421369426e-07, "loss": 0.3273, "step": 2140 }, { "epoch": 0.8881987577639752, "grad_norm": 2.3866941174661993, "learning_rate": 3.7611984654935707e-07, "loss": 0.3218, "step": 2145 }, { "epoch": 0.8902691511387164, "grad_norm": 2.3348475610176, "learning_rate": 3.6248774448952695e-07, "loss": 0.3191, "step": 2150 }, { "epoch": 0.8923395445134575, "grad_norm": 2.441065668351126, "learning_rate": 3.490979703642117e-07, "loss": 0.336, "step": 2155 }, { "epoch": 0.8944099378881988, "grad_norm": 2.273245811499413, "learning_rate": 3.35951223840848e-07, "loss": 0.3136, "step": 2160 }, { "epoch": 0.8964803312629399, "grad_norm": 2.359301186608738, "learning_rate": 3.2304819188773815e-07, "loss": 0.3153, "step": 2165 }, { "epoch": 0.8985507246376812, "grad_norm": 2.4349069583068226, "learning_rate": 3.1038954873816027e-07, "loss": 0.325, "step": 2170 }, { "epoch": 0.9006211180124224, "grad_norm": 2.221649283047898, "learning_rate": 2.9797595585512986e-07, "loss": 0.3162, "step": 2175 }, { "epoch": 0.9026915113871635, "grad_norm": 2.4365346639466976, "learning_rate": 2.858080618968423e-07, "loss": 0.3211, "step": 2180 }, { "epoch": 0.9047619047619048, "grad_norm": 2.2949548389538226, "learning_rate": 2.738865026827714e-07, "loss": 0.3083, "step": 2185 }, { "epoch": 0.906832298136646, "grad_norm": 2.354756885436125, "learning_rate": 2.6221190116045126e-07, "loss": 0.3213, "step": 2190 }, { "epoch": 0.9089026915113871, "grad_norm": 2.385659076635714, "learning_rate": 2.507848673729224e-07, "loss": 0.3175, "step": 2195 }, { "epoch": 0.9109730848861284, "grad_norm": 2.4203042137247333, "learning_rate": 2.3960599842685394e-07, "loss": 0.3243, "step": 2200 }, { "epoch": 0.9130434782608695, "grad_norm": 2.4504849617910796, "learning_rate": 2.2867587846134188e-07, "loss": 0.3164, "step": 2205 }, { "epoch": 0.9151138716356108, "grad_norm": 2.3686013787646165, "learning_rate": 2.179950786173879e-07, "loss": 0.3174, "step": 2210 }, { "epoch": 0.917184265010352, "grad_norm": 2.271535894048282, "learning_rate": 2.07564157008055e-07, "loss": 0.3223, "step": 2215 }, { "epoch": 0.9192546583850931, "grad_norm": 2.3541613140984894, "learning_rate": 1.9738365868930188e-07, "loss": 0.3207, "step": 2220 }, { "epoch": 0.9213250517598344, "grad_norm": 2.3941528051254504, "learning_rate": 1.8745411563150128e-07, "loss": 0.3161, "step": 2225 }, { "epoch": 0.9233954451345756, "grad_norm": 2.261376584339555, "learning_rate": 1.7777604669164727e-07, "loss": 0.3138, "step": 2230 }, { "epoch": 0.9254658385093167, "grad_norm": 2.3190443101075973, "learning_rate": 1.6834995758623685e-07, "loss": 0.3155, "step": 2235 }, { "epoch": 0.927536231884058, "grad_norm": 2.4743221595041147, "learning_rate": 1.5917634086484778e-07, "loss": 0.3194, "step": 2240 }, { "epoch": 0.9296066252587992, "grad_norm": 2.3971910739689695, "learning_rate": 1.5025567588439925e-07, "loss": 0.3185, "step": 2245 }, { "epoch": 0.9316770186335404, "grad_norm": 2.3405920624750096, "learning_rate": 1.415884287841074e-07, "loss": 0.3274, "step": 2250 }, { "epoch": 0.9337474120082816, "grad_norm": 2.479303603019371, "learning_rate": 1.331750524611225e-07, "loss": 0.3145, "step": 2255 }, { "epoch": 0.9358178053830227, "grad_norm": 2.426253073867152, "learning_rate": 1.2501598654686532e-07, "loss": 0.3147, "step": 2260 }, { "epoch": 0.937888198757764, "grad_norm": 2.3156992646831136, "learning_rate": 1.1711165738405749e-07, "loss": 0.3205, "step": 2265 }, { "epoch": 0.9399585921325052, "grad_norm": 2.5003604779718707, "learning_rate": 1.0946247800443899e-07, "loss": 0.318, "step": 2270 }, { "epoch": 0.9420289855072463, "grad_norm": 2.4811177398484157, "learning_rate": 1.0206884810718964e-07, "loss": 0.3211, "step": 2275 }, { "epoch": 0.9440993788819876, "grad_norm": 2.3564288178465156, "learning_rate": 9.493115403803876e-08, "loss": 0.3215, "step": 2280 }, { "epoch": 0.9461697722567288, "grad_norm": 2.275026862745937, "learning_rate": 8.804976876908178e-08, "loss": 0.3184, "step": 2285 }, { "epoch": 0.94824016563147, "grad_norm": 2.363783479958005, "learning_rate": 8.142505187928861e-08, "loss": 0.3299, "step": 2290 }, { "epoch": 0.9503105590062112, "grad_norm": 2.3282452854981224, "learning_rate": 7.505734953571486e-08, "loss": 0.3104, "step": 2295 }, { "epoch": 0.9523809523809523, "grad_norm": 2.5588066429543237, "learning_rate": 6.894699447541231e-08, "loss": 0.3216, "step": 2300 }, { "epoch": 0.9544513457556936, "grad_norm": 2.3198859845159956, "learning_rate": 6.309430598804234e-08, "loss": 0.3097, "step": 2305 }, { "epoch": 0.9565217391304348, "grad_norm": 2.41344387465687, "learning_rate": 5.7499589899193134e-08, "loss": 0.3153, "step": 2310 }, { "epoch": 0.9585921325051759, "grad_norm": 2.5822922013181175, "learning_rate": 5.216313855439858e-08, "loss": 0.3193, "step": 2315 }, { "epoch": 0.9606625258799172, "grad_norm": 2.422135687600398, "learning_rate": 4.7085230803860513e-08, "loss": 0.3185, "step": 2320 }, { "epoch": 0.9627329192546584, "grad_norm": 2.4718246510689803, "learning_rate": 4.2266131987880876e-08, "loss": 0.3295, "step": 2325 }, { "epoch": 0.9648033126293996, "grad_norm": 2.495970339302664, "learning_rate": 3.770609392299285e-08, "loss": 0.3132, "step": 2330 }, { "epoch": 0.9668737060041408, "grad_norm": 2.4117713089348682, "learning_rate": 3.340535488880525e-08, "loss": 0.3098, "step": 2335 }, { "epoch": 0.968944099378882, "grad_norm": 2.4964819297620653, "learning_rate": 2.9364139615550846e-08, "loss": 0.3168, "step": 2340 }, { "epoch": 0.9710144927536232, "grad_norm": 2.3391947122526577, "learning_rate": 2.5582659272341827e-08, "loss": 0.3196, "step": 2345 }, { "epoch": 0.9730848861283644, "grad_norm": 2.506425184281738, "learning_rate": 2.206111145613865e-08, "loss": 0.3099, "step": 2350 }, { "epoch": 0.9751552795031055, "grad_norm": 2.179329803243906, "learning_rate": 1.8799680181421087e-08, "loss": 0.3177, "step": 2355 }, { "epoch": 0.9772256728778468, "grad_norm": 2.3142698292838753, "learning_rate": 1.5798535870575893e-08, "loss": 0.3234, "step": 2360 }, { "epoch": 0.979296066252588, "grad_norm": 2.383754154510841, "learning_rate": 1.30578353449895e-08, "loss": 0.3127, "step": 2365 }, { "epoch": 0.9813664596273292, "grad_norm": 2.370736953831053, "learning_rate": 1.0577721816854015e-08, "loss": 0.3091, "step": 2370 }, { "epoch": 0.9834368530020704, "grad_norm": 2.517651342295925, "learning_rate": 8.35832488168431e-09, "loss": 0.3173, "step": 2375 }, { "epoch": 0.9855072463768116, "grad_norm": 2.3821914184277295, "learning_rate": 6.399760511546227e-09, "loss": 0.3207, "step": 2380 }, { "epoch": 0.9875776397515528, "grad_norm": 2.315898281851513, "learning_rate": 4.7021310489953065e-09, "loss": 0.3185, "step": 2385 }, { "epoch": 0.989648033126294, "grad_norm": 2.4911092964288644, "learning_rate": 3.265525201731623e-09, "loss": 0.3188, "step": 2390 }, { "epoch": 0.9917184265010351, "grad_norm": 2.332775121742113, "learning_rate": 2.090018037960717e-09, "loss": 0.3149, "step": 2395 }, { "epoch": 0.9937888198757764, "grad_norm": 2.269380119205785, "learning_rate": 1.1756709824750633e-09, "loss": 0.321, "step": 2400 }, { "epoch": 0.9958592132505176, "grad_norm": 2.1851128802901725, "learning_rate": 5.225318134399749e-10, "loss": 0.3055, "step": 2405 }, { "epoch": 0.9979296066252588, "grad_norm": 2.347230158447268, "learning_rate": 1.306346599011521e-10, "loss": 0.3134, "step": 2410 }, { "epoch": 1.0, "grad_norm": 2.5576033751474125, "learning_rate": 0.0, "loss": 0.3164, "step": 2415 }, { "epoch": 1.0, "eval_runtime": 2.7619, "eval_samples_per_second": 3.621, "eval_steps_per_second": 1.086, "step": 2415 }, { "epoch": 1.0, "step": 2415, "total_flos": 252825934233600.0, "train_loss": 0.5146465574979289, "train_runtime": 25682.4245, "train_samples_per_second": 1.504, "train_steps_per_second": 0.094 } ], "logging_steps": 5, "max_steps": 2415, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 100, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 252825934233600.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }