{ "best_global_step": 200, "best_metric": 2.8438220024108887, "best_model_checkpoint": null, "epoch": 0.7818411097099621, "eval_steps": 100, "global_step": 310, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0025220680958385876, "grad_norm": 119.76318359375, "kl_loss": -1.1687562835330993e-15, "kl_weight": 1000000.0, "learning_rate": 0.0, "loss": 2.6394360065460205, "step": 1, "total_loss": 2.6394360065460205 }, { "epoch": 0.005044136191677175, "grad_norm": 116.01831817626953, "kl_loss": -1.280914393650412e-14, "kl_weight": 1000000.0, "learning_rate": 0.0001, "loss": 3.2936160564422607, "step": 2, "total_loss": 3.2936160564422607 }, { "epoch": 0.007566204287515763, "grad_norm": 104.04817962646484, "kl_loss": 7.10318071028837e-09, "kl_weight": 1000000.0, "learning_rate": 0.0002, "loss": 3.084439992904663, "step": 3, "total_loss": 3.091543197631836 }, { "epoch": 0.01008827238335435, "grad_norm": 68.36679077148438, "kl_loss": 2.8489626657801637e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003, "loss": 3.105210304260254, "step": 4, "total_loss": 3.133699893951416 }, { "epoch": 0.012610340479192938, "grad_norm": 61.00284957885742, "kl_loss": 4.923957774849441e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004, "loss": 3.345022678375244, "step": 5, "total_loss": 3.3942623138427734 }, { "epoch": 0.015132408575031526, "grad_norm": 65.48960876464844, "kl_loss": 1.43211394743048e-07, "kl_weight": 1000000.0, "learning_rate": 0.0005, "loss": 2.3467514514923096, "step": 6, "total_loss": 2.4899628162384033 }, { "epoch": 0.017654476670870115, "grad_norm": 63.001102447509766, "kl_loss": 9.109995602329946e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004993662864385298, "loss": 2.5077083110809326, "step": 7, "total_loss": 2.5988082885742188 }, { "epoch": 0.0201765447667087, "grad_norm": 58.6073112487793, "kl_loss": 2.3511624647198914e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004987325728770596, "loss": 2.2668278217315674, "step": 8, "total_loss": 2.501944065093994 }, { "epoch": 0.02269861286254729, "grad_norm": 97.743896484375, "kl_loss": 2.1175161180053692e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004980988593155894, "loss": 2.352029800415039, "step": 9, "total_loss": 2.563781499862671 }, { "epoch": 0.025220680958385876, "grad_norm": 60.91500473022461, "kl_loss": 1.2846226127294358e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004974651457541192, "loss": 2.2376697063446045, "step": 10, "total_loss": 2.3661320209503174 }, { "epoch": 0.027742749054224466, "grad_norm": 55.095516204833984, "kl_loss": 1.4181343033214944e-07, "kl_weight": 1000000.0, "learning_rate": 0.000496831432192649, "loss": 2.8243818283081055, "step": 11, "total_loss": 2.9661953449249268 }, { "epoch": 0.03026481715006305, "grad_norm": 44.97727966308594, "kl_loss": 1.545683971926337e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004961977186311787, "loss": 2.4689197540283203, "step": 12, "total_loss": 2.623488187789917 }, { "epoch": 0.03278688524590164, "grad_norm": 51.62504196166992, "kl_loss": 2.2357993145760702e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004955640050697085, "loss": 2.2227847576141357, "step": 13, "total_loss": 2.446364641189575 }, { "epoch": 0.03530895334174023, "grad_norm": 42.21575927734375, "kl_loss": 1.6229765265052265e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004949302915082382, "loss": 2.4396450519561768, "step": 14, "total_loss": 2.601942777633667 }, { "epoch": 0.03783102143757881, "grad_norm": 40.02684783935547, "kl_loss": 1.4151250127270032e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004942965779467681, "loss": 2.509690761566162, "step": 15, "total_loss": 2.651203155517578 }, { "epoch": 0.0403530895334174, "grad_norm": 44.62814712524414, "kl_loss": 1.450005981951108e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004936628643852978, "loss": 2.4844541549682617, "step": 16, "total_loss": 2.6294548511505127 }, { "epoch": 0.04287515762925599, "grad_norm": 41.87761688232422, "kl_loss": 1.397227009647395e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004930291508238277, "loss": 2.7985713481903076, "step": 17, "total_loss": 2.938293933868408 }, { "epoch": 0.04539722572509458, "grad_norm": 39.647457122802734, "kl_loss": 1.0770181546604363e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004923954372623574, "loss": 2.1876273155212402, "step": 18, "total_loss": 2.2953290939331055 }, { "epoch": 0.04791929382093316, "grad_norm": 44.82719039916992, "kl_loss": 1.325549447983576e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004917617237008873, "loss": 2.344290256500244, "step": 19, "total_loss": 2.4768452644348145 }, { "epoch": 0.05044136191677175, "grad_norm": 35.45253372192383, "kl_loss": 1.3449634650442022e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004911280101394169, "loss": 2.393965244293213, "step": 20, "total_loss": 2.5284616947174072 }, { "epoch": 0.05296343001261034, "grad_norm": 36.362369537353516, "kl_loss": 1.552224659917556e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004904942965779467, "loss": 2.1951944828033447, "step": 21, "total_loss": 2.350416898727417 }, { "epoch": 0.05548549810844893, "grad_norm": 42.16935348510742, "kl_loss": 1.1523614062980414e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004898605830164765, "loss": 2.4038805961608887, "step": 22, "total_loss": 2.5191166400909424 }, { "epoch": 0.058007566204287514, "grad_norm": 39.14812088012695, "kl_loss": 1.1659390963814076e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004892268694550063, "loss": 2.6124515533447266, "step": 23, "total_loss": 2.7290453910827637 }, { "epoch": 0.0605296343001261, "grad_norm": 49.780704498291016, "kl_loss": 2.0625684271635691e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004885931558935361, "loss": 2.9930167198181152, "step": 24, "total_loss": 3.1992735862731934 }, { "epoch": 0.06305170239596469, "grad_norm": 53.23894500732422, "kl_loss": 1.4698964889703348e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004879594423320659, "loss": 2.4087769985198975, "step": 25, "total_loss": 2.5557665824890137 }, { "epoch": 0.06557377049180328, "grad_norm": 50.209110260009766, "kl_loss": 1.2840492047416774e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004873257287705957, "loss": 2.918276309967041, "step": 26, "total_loss": 3.0466811656951904 }, { "epoch": 0.06809583858764187, "grad_norm": 41.90302658081055, "kl_loss": 1.8274477042723447e-07, "kl_weight": 1000000.0, "learning_rate": 0.00048669201520912546, "loss": 2.272730588912964, "step": 27, "total_loss": 2.455475330352783 }, { "epoch": 0.07061790668348046, "grad_norm": 39.8343620300293, "kl_loss": 1.3182453528770566e-07, "kl_weight": 1000000.0, "learning_rate": 0.00048605830164765525, "loss": 2.6464455127716064, "step": 28, "total_loss": 2.7782700061798096 }, { "epoch": 0.07313997477931904, "grad_norm": 36.19342803955078, "kl_loss": 1.3146133426289452e-07, "kl_weight": 1000000.0, "learning_rate": 0.00048542458808618503, "loss": 2.6744461059570312, "step": 29, "total_loss": 2.8059074878692627 }, { "epoch": 0.07566204287515763, "grad_norm": 39.07732391357422, "kl_loss": 1.3345737670533708e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004847908745247148, "loss": 2.624727249145508, "step": 30, "total_loss": 2.7581846714019775 }, { "epoch": 0.07818411097099622, "grad_norm": 46.28368377685547, "kl_loss": 1.4280556115409127e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004841571609632446, "loss": 2.4123611450195312, "step": 31, "total_loss": 2.555166721343994 }, { "epoch": 0.0807061790668348, "grad_norm": 43.698612213134766, "kl_loss": 1.2306992402955075e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004835234474017744, "loss": 2.641366958618164, "step": 32, "total_loss": 2.764436960220337 }, { "epoch": 0.0832282471626734, "grad_norm": 45.80537796020508, "kl_loss": 1.3520984509796108e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004828897338403042, "loss": 2.5132896900177, "step": 33, "total_loss": 2.6484994888305664 }, { "epoch": 0.08575031525851198, "grad_norm": 48.67962646484375, "kl_loss": 1.237484639204922e-07, "kl_weight": 1000000.0, "learning_rate": 0.000482256020278834, "loss": 2.7271363735198975, "step": 34, "total_loss": 2.8508849143981934 }, { "epoch": 0.08827238335435057, "grad_norm": 49.93069839477539, "kl_loss": 1.2296362683628104e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004816223067173637, "loss": 2.2032339572906494, "step": 35, "total_loss": 2.326197624206543 }, { "epoch": 0.09079445145018916, "grad_norm": 34.86311721801758, "kl_loss": 1.3591355241260317e-07, "kl_weight": 1000000.0, "learning_rate": 0.00048098859315589355, "loss": 2.0896949768066406, "step": 36, "total_loss": 2.2256085872650146 }, { "epoch": 0.09331651954602774, "grad_norm": 37.413055419921875, "kl_loss": 1.403360414542476e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004803548795944233, "loss": 2.331239700317383, "step": 37, "total_loss": 2.4715757369995117 }, { "epoch": 0.09583858764186633, "grad_norm": 71.4142074584961, "kl_loss": 9.95134428194433e-08, "kl_weight": 1000000.0, "learning_rate": 0.00047972116603295313, "loss": 2.3557369709014893, "step": 38, "total_loss": 2.4552505016326904 }, { "epoch": 0.09836065573770492, "grad_norm": 43.588504791259766, "kl_loss": 9.627683539292775e-08, "kl_weight": 1000000.0, "learning_rate": 0.00047908745247148286, "loss": 2.483755111694336, "step": 39, "total_loss": 2.5800318717956543 }, { "epoch": 0.1008827238335435, "grad_norm": 53.27906036376953, "kl_loss": 1.6940899172368518e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004784537389100127, "loss": 2.2950119972229004, "step": 40, "total_loss": 2.464421033859253 }, { "epoch": 0.1034047919293821, "grad_norm": 36.538055419921875, "kl_loss": 1.7988656964007532e-07, "kl_weight": 1000000.0, "learning_rate": 0.00047782002534854244, "loss": 2.2202811241149902, "step": 41, "total_loss": 2.40016770362854 }, { "epoch": 0.10592686002522068, "grad_norm": 37.166908264160156, "kl_loss": 1.6280961290249252e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004771863117870723, "loss": 2.3544418811798096, "step": 42, "total_loss": 2.517251491546631 }, { "epoch": 0.10844892812105927, "grad_norm": 42.741424560546875, "kl_loss": 1.2595839393725328e-07, "kl_weight": 1000000.0, "learning_rate": 0.000476552598225602, "loss": 2.299348831176758, "step": 43, "total_loss": 2.425307273864746 }, { "epoch": 0.11097099621689786, "grad_norm": 39.69258117675781, "kl_loss": 1.7368185467603325e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004759188846641318, "loss": 2.6060984134674072, "step": 44, "total_loss": 2.779780387878418 }, { "epoch": 0.11349306431273644, "grad_norm": 48.37623596191406, "kl_loss": 1.2949078609381104e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004752851711026616, "loss": 2.5570790767669678, "step": 45, "total_loss": 2.686569929122925 }, { "epoch": 0.11601513240857503, "grad_norm": 32.54016876220703, "kl_loss": 1.0888848578360921e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004746514575411914, "loss": 2.351015567779541, "step": 46, "total_loss": 2.4599039554595947 }, { "epoch": 0.11853720050441362, "grad_norm": 39.50113296508789, "kl_loss": 8.368018455939819e-08, "kl_weight": 1000000.0, "learning_rate": 0.00047401774397972117, "loss": 2.030034303665161, "step": 47, "total_loss": 2.1137144565582275 }, { "epoch": 0.1210592686002522, "grad_norm": 58.826534271240234, "kl_loss": 7.160872428357834e-08, "kl_weight": 1000000.0, "learning_rate": 0.00047338403041825096, "loss": 2.4345922470092773, "step": 48, "total_loss": 2.5062010288238525 }, { "epoch": 0.1235813366960908, "grad_norm": 49.829811096191406, "kl_loss": 1.4755499933016836e-07, "kl_weight": 1000000.0, "learning_rate": 0.00047275031685678075, "loss": 2.4012234210968018, "step": 49, "total_loss": 2.548778533935547 }, { "epoch": 0.12610340479192939, "grad_norm": 54.877708435058594, "kl_loss": 1.25692750430062e-07, "kl_weight": 1000000.0, "learning_rate": 0.00047211660329531054, "loss": 2.4835963249206543, "step": 50, "total_loss": 2.6092891693115234 }, { "epoch": 0.12862547288776796, "grad_norm": 57.37236785888672, "kl_loss": 1.1875315664156005e-07, "kl_weight": 1000000.0, "learning_rate": 0.00047148288973384027, "loss": 2.25846266746521, "step": 51, "total_loss": 2.37721586227417 }, { "epoch": 0.13114754098360656, "grad_norm": 46.449405670166016, "kl_loss": 9.858030125542427e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004708491761723701, "loss": 2.2844595909118652, "step": 52, "total_loss": 2.383039951324463 }, { "epoch": 0.13366960907944514, "grad_norm": 88.41195678710938, "kl_loss": 1.3280877908528055e-07, "kl_weight": 1000000.0, "learning_rate": 0.00047021546261089985, "loss": 2.73095440864563, "step": 53, "total_loss": 2.8637630939483643 }, { "epoch": 0.13619167717528374, "grad_norm": 46.997642517089844, "kl_loss": 1.318484947887555e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004695817490494297, "loss": 2.9694020748138428, "step": 54, "total_loss": 3.101250648498535 }, { "epoch": 0.13871374527112232, "grad_norm": 59.88271713256836, "kl_loss": 1.311551756089102e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004689480354879594, "loss": 2.339895009994507, "step": 55, "total_loss": 2.471050262451172 }, { "epoch": 0.14123581336696092, "grad_norm": 50.745269775390625, "kl_loss": 1.2313154229559586e-07, "kl_weight": 1000000.0, "learning_rate": 0.00046831432192648927, "loss": 2.3892390727996826, "step": 56, "total_loss": 2.5123705863952637 }, { "epoch": 0.1437578814627995, "grad_norm": 38.669708251953125, "kl_loss": 1.3242403440472117e-07, "kl_weight": 1000000.0, "learning_rate": 0.000467680608365019, "loss": 2.3022656440734863, "step": 57, "total_loss": 2.43468976020813 }, { "epoch": 0.14627994955863807, "grad_norm": 35.511695861816406, "kl_loss": 9.955392243909955e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004670468948035488, "loss": 2.4699792861938477, "step": 58, "total_loss": 2.569533109664917 }, { "epoch": 0.14880201765447668, "grad_norm": 42.08246994018555, "kl_loss": 1.3320465086508193e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004664131812420786, "loss": 2.288395643234253, "step": 59, "total_loss": 2.421600341796875 }, { "epoch": 0.15132408575031525, "grad_norm": 37.103790283203125, "kl_loss": 5.67616069702126e-08, "kl_weight": 1000000.0, "learning_rate": 0.00046577946768060837, "loss": 2.5267558097839355, "step": 60, "total_loss": 2.58351731300354 }, { "epoch": 0.15384615384615385, "grad_norm": 38.30869674682617, "kl_loss": 1.197651613438211e-07, "kl_weight": 1000000.0, "learning_rate": 0.00046514575411913815, "loss": 2.6164743900299072, "step": 61, "total_loss": 2.736239433288574 }, { "epoch": 0.15636822194199243, "grad_norm": 39.79656982421875, "kl_loss": 9.629880537431745e-08, "kl_weight": 1000000.0, "learning_rate": 0.00046451204055766794, "loss": 2.2402400970458984, "step": 62, "total_loss": 2.336538791656494 }, { "epoch": 0.15889029003783103, "grad_norm": 40.427127838134766, "kl_loss": 1.266524236598343e-07, "kl_weight": 1000000.0, "learning_rate": 0.00046387832699619773, "loss": 2.634199857711792, "step": 63, "total_loss": 2.760852336883545 }, { "epoch": 0.1614123581336696, "grad_norm": 45.250335693359375, "kl_loss": 1.2385322634145268e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004632446134347275, "loss": 2.8654606342315674, "step": 64, "total_loss": 2.989313840866089 }, { "epoch": 0.16393442622950818, "grad_norm": 41.70964050292969, "kl_loss": 8.554673769367582e-08, "kl_weight": 1000000.0, "learning_rate": 0.00046261089987325725, "loss": 2.4218544960021973, "step": 65, "total_loss": 2.50740122795105 }, { "epoch": 0.1664564943253468, "grad_norm": 48.33979415893555, "kl_loss": 8.96402525540907e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004619771863117871, "loss": 2.1502766609191895, "step": 66, "total_loss": 2.2399168014526367 }, { "epoch": 0.16897856242118536, "grad_norm": 49.24103546142578, "kl_loss": 1.0698733632352742e-07, "kl_weight": 1000000.0, "learning_rate": 0.00046134347275031683, "loss": 2.373239278793335, "step": 67, "total_loss": 2.480226516723633 }, { "epoch": 0.17150063051702397, "grad_norm": 47.207801818847656, "kl_loss": 1.1443621161788542e-07, "kl_weight": 1000000.0, "learning_rate": 0.00046070975918884667, "loss": 2.083575963973999, "step": 68, "total_loss": 2.198012113571167 }, { "epoch": 0.17402269861286254, "grad_norm": 55.49308395385742, "kl_loss": 7.693908798955817e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004600760456273764, "loss": 2.3746559619903564, "step": 69, "total_loss": 2.4515950679779053 }, { "epoch": 0.17654476670870115, "grad_norm": 35.293094635009766, "kl_loss": 1.1812133493549481e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045944233206590625, "loss": 2.300846576690674, "step": 70, "total_loss": 2.4189679622650146 }, { "epoch": 0.17906683480453972, "grad_norm": 40.90177536010742, "kl_loss": 1.278984171904085e-07, "kl_weight": 1000000.0, "learning_rate": 0.000458808618504436, "loss": 2.6642138957977295, "step": 71, "total_loss": 2.792112350463867 }, { "epoch": 0.18158890290037832, "grad_norm": 35.18415832519531, "kl_loss": 1.0427108065869106e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045817490494296577, "loss": 2.0037200450897217, "step": 72, "total_loss": 2.1079912185668945 }, { "epoch": 0.1841109709962169, "grad_norm": 58.994651794433594, "kl_loss": 1.1237546715392455e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045754119138149556, "loss": 2.1798439025878906, "step": 73, "total_loss": 2.292219400405884 }, { "epoch": 0.18663303909205547, "grad_norm": 58.826744079589844, "kl_loss": 1.2047219399846654e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045690747782002535, "loss": 2.182692527770996, "step": 74, "total_loss": 2.3031647205352783 }, { "epoch": 0.18915510718789408, "grad_norm": 48.294960021972656, "kl_loss": 9.133592726584538e-08, "kl_weight": 1000000.0, "learning_rate": 0.00045627376425855514, "loss": 2.640594005584717, "step": 75, "total_loss": 2.7319300174713135 }, { "epoch": 0.19167717528373265, "grad_norm": 42.91217803955078, "kl_loss": 1.3158961564840865e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004556400506970849, "loss": 2.20942759513855, "step": 76, "total_loss": 2.341017246246338 }, { "epoch": 0.19419924337957126, "grad_norm": 37.41611099243164, "kl_loss": 7.94594257058634e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004550063371356147, "loss": 1.8829660415649414, "step": 77, "total_loss": 1.9624254703521729 }, { "epoch": 0.19672131147540983, "grad_norm": 68.14502716064453, "kl_loss": 6.981721156762433e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004543726235741445, "loss": 2.176638603210449, "step": 78, "total_loss": 2.2464559078216553 }, { "epoch": 0.19924337957124844, "grad_norm": 37.93901824951172, "kl_loss": 1.4584460927835607e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045373891001267424, "loss": 2.3125131130218506, "step": 79, "total_loss": 2.458357810974121 }, { "epoch": 0.201765447667087, "grad_norm": 36.36228942871094, "kl_loss": 1.0491366708720307e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004531051964512041, "loss": 2.669814348220825, "step": 80, "total_loss": 2.7747280597686768 }, { "epoch": 0.2042875157629256, "grad_norm": 33.21589279174805, "kl_loss": 8.339785040334391e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004524714828897338, "loss": 2.1568803787231445, "step": 81, "total_loss": 2.2402782440185547 }, { "epoch": 0.2068095838587642, "grad_norm": 50.5767936706543, "kl_loss": 8.075436142007675e-08, "kl_weight": 1000000.0, "learning_rate": 0.00045183776932826366, "loss": 2.582383871078491, "step": 82, "total_loss": 2.6631381511688232 }, { "epoch": 0.20933165195460277, "grad_norm": 35.17644119262695, "kl_loss": 7.244600652711597e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004512040557667934, "loss": 2.234858274459839, "step": 83, "total_loss": 2.3073043823242188 }, { "epoch": 0.21185372005044137, "grad_norm": 42.40107727050781, "kl_loss": 1.1873476779555858e-07, "kl_weight": 1000000.0, "learning_rate": 0.00045057034220532323, "loss": 2.455256462097168, "step": 84, "total_loss": 2.573991298675537 }, { "epoch": 0.21437578814627994, "grad_norm": 36.42718505859375, "kl_loss": 1.1225206009157773e-07, "kl_weight": 1000000.0, "learning_rate": 0.00044993662864385297, "loss": 2.2919607162475586, "step": 85, "total_loss": 2.404212713241577 }, { "epoch": 0.21689785624211855, "grad_norm": 53.86962127685547, "kl_loss": 1.4159495265175792e-07, "kl_weight": 1000000.0, "learning_rate": 0.00044930291508238275, "loss": 2.3607466220855713, "step": 86, "total_loss": 2.5023415088653564 }, { "epoch": 0.21941992433795712, "grad_norm": 38.30321502685547, "kl_loss": 1.2286402295558219e-07, "kl_weight": 1000000.0, "learning_rate": 0.00044866920152091254, "loss": 2.518197774887085, "step": 87, "total_loss": 2.641061782836914 }, { "epoch": 0.22194199243379573, "grad_norm": 39.55595016479492, "kl_loss": 7.652983669004243e-08, "kl_weight": 1000000.0, "learning_rate": 0.00044803548795944233, "loss": 2.497284173965454, "step": 88, "total_loss": 2.5738139152526855 }, { "epoch": 0.2244640605296343, "grad_norm": 54.330936431884766, "kl_loss": 1.4088276145685086e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004474017743979721, "loss": 2.0035746097564697, "step": 89, "total_loss": 2.1444573402404785 }, { "epoch": 0.22698612862547288, "grad_norm": 115.77396392822266, "kl_loss": 7.670182355923316e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004467680608365019, "loss": 2.504946708679199, "step": 90, "total_loss": 2.581648588180542 }, { "epoch": 0.22950819672131148, "grad_norm": 106.39173126220703, "kl_loss": 1.4106633727806184e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004461343472750317, "loss": 2.12337064743042, "step": 91, "total_loss": 2.264436960220337 }, { "epoch": 0.23203026481715006, "grad_norm": 65.24938201904297, "kl_loss": 9.111739984746237e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004455006337135615, "loss": 2.0131237506866455, "step": 92, "total_loss": 2.104241132736206 }, { "epoch": 0.23455233291298866, "grad_norm": 44.74656295776367, "kl_loss": 1.4964980721288157e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004448669201520912, "loss": 2.265284776687622, "step": 93, "total_loss": 2.4149346351623535 }, { "epoch": 0.23707440100882723, "grad_norm": 91.94232940673828, "kl_loss": 9.062223682576587e-08, "kl_weight": 1000000.0, "learning_rate": 0.00044423320659062106, "loss": 2.1661198139190674, "step": 94, "total_loss": 2.256742000579834 }, { "epoch": 0.23959646910466584, "grad_norm": 93.24700164794922, "kl_loss": 9.902019826313335e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004435994930291508, "loss": 2.3344357013702393, "step": 95, "total_loss": 2.4334559440612793 }, { "epoch": 0.2421185372005044, "grad_norm": 121.62272644042969, "kl_loss": 7.655695810626639e-08, "kl_weight": 1000000.0, "learning_rate": 0.00044296577946768064, "loss": 2.6050221920013428, "step": 96, "total_loss": 2.681579113006592 }, { "epoch": 0.244640605296343, "grad_norm": 46.2025260925293, "kl_loss": 1.436859946579716e-07, "kl_weight": 1000000.0, "learning_rate": 0.00044233206590621037, "loss": 2.530095100402832, "step": 97, "total_loss": 2.673781156539917 }, { "epoch": 0.2471626733921816, "grad_norm": 113.35319519042969, "kl_loss": 1.0582026988004145e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004416983523447402, "loss": 2.698303461074829, "step": 98, "total_loss": 2.804123640060425 }, { "epoch": 0.24968474148802017, "grad_norm": 102.921875, "kl_loss": 1.3178673441416322e-07, "kl_weight": 1000000.0, "learning_rate": 0.00044106463878326995, "loss": 2.6428942680358887, "step": 99, "total_loss": 2.7746810913085938 }, { "epoch": 0.25220680958385877, "grad_norm": 79.49183654785156, "kl_loss": 1.0406408534890943e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004404309252217998, "loss": 1.7494869232177734, "step": 100, "total_loss": 1.8535510301589966 }, { "epoch": 0.25220680958385877, "eval_loss": 2.8745031356811523, "eval_runtime": 36.3348, "eval_samples_per_second": 19.403, "eval_steps_per_second": 2.449, "step": 100 }, { "epoch": 0.2547288776796974, "grad_norm": 52.9495849609375, "kl_loss": 8.112144200822513e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004397972116603295, "loss": 2.6726205348968506, "step": 101, "total_loss": 2.753741979598999 }, { "epoch": 0.2572509457755359, "grad_norm": 43.57243728637695, "kl_loss": 8.97685552558869e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004391634980988593, "loss": 2.557943105697632, "step": 102, "total_loss": 2.647711753845215 }, { "epoch": 0.2597730138713745, "grad_norm": 89.02836608886719, "kl_loss": 7.308499050395767e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004385297845373891, "loss": 2.3326590061187744, "step": 103, "total_loss": 2.4057440757751465 }, { "epoch": 0.26229508196721313, "grad_norm": 70.54704284667969, "kl_loss": 1.186173719247563e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004378960709759189, "loss": 2.48604154586792, "step": 104, "total_loss": 2.604658842086792 }, { "epoch": 0.2648171500630517, "grad_norm": 56.15437316894531, "kl_loss": 1.0662374450021161e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004372623574144487, "loss": 2.398261070251465, "step": 105, "total_loss": 2.504884719848633 }, { "epoch": 0.2673392181588903, "grad_norm": 50.12539291381836, "kl_loss": 1.2759885237301205e-07, "kl_weight": 1000000.0, "learning_rate": 0.00043662864385297847, "loss": 2.3349037170410156, "step": 106, "total_loss": 2.4625024795532227 }, { "epoch": 0.2698612862547289, "grad_norm": 58.65528106689453, "kl_loss": 6.978748245956012e-08, "kl_weight": 1000000.0, "learning_rate": 0.00043599493029150826, "loss": 2.520481586456299, "step": 107, "total_loss": 2.590269088745117 }, { "epoch": 0.2723833543505675, "grad_norm": 114.07057189941406, "kl_loss": 2.180455993538999e-07, "kl_weight": 1000000.0, "learning_rate": 0.00043536121673003804, "loss": 2.2573392391204834, "step": 108, "total_loss": 2.4753847122192383 }, { "epoch": 0.27490542244640603, "grad_norm": 42.89162063598633, "kl_loss": 1.3848097069057985e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004347275031685678, "loss": 2.557804822921753, "step": 109, "total_loss": 2.6962857246398926 }, { "epoch": 0.27742749054224464, "grad_norm": 34.359859466552734, "kl_loss": 7.828553094668678e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004340937896070976, "loss": 2.673866033554077, "step": 110, "total_loss": 2.7521514892578125 }, { "epoch": 0.27994955863808324, "grad_norm": 56.4673957824707, "kl_loss": 1.2365975976535992e-07, "kl_weight": 1000000.0, "learning_rate": 0.00043346007604562736, "loss": 2.275721311569214, "step": 111, "total_loss": 2.399381160736084 }, { "epoch": 0.28247162673392184, "grad_norm": 80.7730484008789, "kl_loss": 7.710952587558495e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004328263624841572, "loss": 2.3654656410217285, "step": 112, "total_loss": 2.442575216293335 }, { "epoch": 0.2849936948297604, "grad_norm": 78.50458526611328, "kl_loss": 1.1497648699787533e-07, "kl_weight": 1000000.0, "learning_rate": 0.00043219264892268693, "loss": 2.0045268535614014, "step": 113, "total_loss": 2.1195032596588135 }, { "epoch": 0.287515762925599, "grad_norm": 42.11017990112305, "kl_loss": 1.2359024026409315e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004315589353612168, "loss": 2.8311715126037598, "step": 114, "total_loss": 2.9547617435455322 }, { "epoch": 0.2900378310214376, "grad_norm": 46.80673599243164, "kl_loss": 1.1356344487012393e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004309252217997465, "loss": 2.4908668994903564, "step": 115, "total_loss": 2.6044304370880127 }, { "epoch": 0.29255989911727615, "grad_norm": 37.80940628051758, "kl_loss": 8.934256356951664e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004302915082382763, "loss": 2.5639941692352295, "step": 116, "total_loss": 2.653336763381958 }, { "epoch": 0.29508196721311475, "grad_norm": 40.135955810546875, "kl_loss": 9.139830581261776e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004296577946768061, "loss": 2.3605127334594727, "step": 117, "total_loss": 2.451910972595215 }, { "epoch": 0.29760403530895335, "grad_norm": 34.20405197143555, "kl_loss": 6.737117530519754e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004290240811153359, "loss": 2.2198715209960938, "step": 118, "total_loss": 2.2872426509857178 }, { "epoch": 0.30012610340479196, "grad_norm": 38.27109909057617, "kl_loss": 7.206848096075191e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042839036755386566, "loss": 2.442990779876709, "step": 119, "total_loss": 2.515059232711792 }, { "epoch": 0.3026481715006305, "grad_norm": 33.383628845214844, "kl_loss": 5.8401464997359653e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042775665399239545, "loss": 1.9709354639053345, "step": 120, "total_loss": 2.029336929321289 }, { "epoch": 0.3051702395964691, "grad_norm": 29.64325714111328, "kl_loss": 7.080717523422209e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042712294043092524, "loss": 2.5838398933410645, "step": 121, "total_loss": 2.6546471118927 }, { "epoch": 0.3076923076923077, "grad_norm": 31.308616638183594, "kl_loss": 7.158789117056585e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042648922686945503, "loss": 2.582289457321167, "step": 122, "total_loss": 2.6538772583007812 }, { "epoch": 0.31021437578814626, "grad_norm": 31.91105842590332, "kl_loss": 5.958595039601278e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042585551330798476, "loss": 2.362175703048706, "step": 123, "total_loss": 2.4217617511749268 }, { "epoch": 0.31273644388398486, "grad_norm": 48.74770736694336, "kl_loss": 1.0601585387348678e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004252217997465146, "loss": 2.6029539108276367, "step": 124, "total_loss": 2.708969831466675 }, { "epoch": 0.31525851197982346, "grad_norm": 37.71382141113281, "kl_loss": 7.022106984777565e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042458808618504434, "loss": 1.9780246019363403, "step": 125, "total_loss": 2.048245668411255 }, { "epoch": 0.31778058007566207, "grad_norm": 58.47407531738281, "kl_loss": 7.89053729022271e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004239543726235742, "loss": 3.0634782314300537, "step": 126, "total_loss": 3.142383575439453 }, { "epoch": 0.3203026481715006, "grad_norm": 49.75334167480469, "kl_loss": 9.529335898150748e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004233206590621039, "loss": 2.625107765197754, "step": 127, "total_loss": 2.7204010486602783 }, { "epoch": 0.3228247162673392, "grad_norm": 37.173675537109375, "kl_loss": 9.76136007579953e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042268694550063376, "loss": 2.3797764778137207, "step": 128, "total_loss": 2.4773900508880615 }, { "epoch": 0.3253467843631778, "grad_norm": 35.94040298461914, "kl_loss": 8.867663581213492e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004220532319391635, "loss": 2.19878888130188, "step": 129, "total_loss": 2.2874655723571777 }, { "epoch": 0.32786885245901637, "grad_norm": 30.7440128326416, "kl_loss": 8.357946512660419e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004214195183776933, "loss": 2.374424934387207, "step": 130, "total_loss": 2.4580044746398926 }, { "epoch": 0.33039092055485497, "grad_norm": 44.41523361206055, "kl_loss": 7.932466417059914e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042078580481622307, "loss": 2.270353078842163, "step": 131, "total_loss": 2.349677801132202 }, { "epoch": 0.3329129886506936, "grad_norm": 50.61003494262695, "kl_loss": 7.062011064817852e-08, "kl_weight": 1000000.0, "learning_rate": 0.00042015209125475286, "loss": 2.704482078552246, "step": 132, "total_loss": 2.775102138519287 }, { "epoch": 0.3354350567465322, "grad_norm": 52.31584930419922, "kl_loss": 8.649838179053404e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041951837769328265, "loss": 2.623465061187744, "step": 133, "total_loss": 2.709963321685791 }, { "epoch": 0.3379571248423707, "grad_norm": 44.85032653808594, "kl_loss": 7.965866899439789e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041888466413181243, "loss": 2.3895857334136963, "step": 134, "total_loss": 2.4692444801330566 }, { "epoch": 0.34047919293820933, "grad_norm": 52.73245620727539, "kl_loss": 9.069820805507334e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004182509505703422, "loss": 2.5411245822906494, "step": 135, "total_loss": 2.6318228244781494 }, { "epoch": 0.34300126103404793, "grad_norm": 54.24882507324219, "kl_loss": 9.488780960964505e-08, "kl_weight": 1000000.0, "learning_rate": 0.000417617237008872, "loss": 1.959804654121399, "step": 136, "total_loss": 2.054692506790161 }, { "epoch": 0.3455233291298865, "grad_norm": 80.58323669433594, "kl_loss": 1.1013327139153262e-07, "kl_weight": 1000000.0, "learning_rate": 0.00041698352344740175, "loss": 2.191988229751587, "step": 137, "total_loss": 2.30212140083313 }, { "epoch": 0.3480453972257251, "grad_norm": 94.20862579345703, "kl_loss": 4.3523357362573734e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004163498098859316, "loss": 2.3833768367767334, "step": 138, "total_loss": 2.4269001483917236 }, { "epoch": 0.3505674653215637, "grad_norm": 57.289390563964844, "kl_loss": 6.378383687888345e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004157160963244613, "loss": 2.5576775074005127, "step": 139, "total_loss": 2.6214613914489746 }, { "epoch": 0.3530895334174023, "grad_norm": 108.49940490722656, "kl_loss": 1.0705950614919857e-07, "kl_weight": 1000000.0, "learning_rate": 0.00041508238276299116, "loss": 2.5679421424865723, "step": 140, "total_loss": 2.675001621246338 }, { "epoch": 0.35561160151324084, "grad_norm": 45.74307632446289, "kl_loss": 1.295888125696365e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004144486692015209, "loss": 2.040994882583618, "step": 141, "total_loss": 2.170583724975586 }, { "epoch": 0.35813366960907944, "grad_norm": 43.9136962890625, "kl_loss": 1.1324662807510322e-07, "kl_weight": 1000000.0, "learning_rate": 0.00041381495564005074, "loss": 2.430799961090088, "step": 142, "total_loss": 2.544046640396118 }, { "epoch": 0.36065573770491804, "grad_norm": 37.08522415161133, "kl_loss": 1.0119717330780986e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004131812420785805, "loss": 2.4721150398254395, "step": 143, "total_loss": 2.573312282562256 }, { "epoch": 0.36317780580075665, "grad_norm": 30.434844970703125, "kl_loss": 8.321502775743284e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041254752851711026, "loss": 2.430828094482422, "step": 144, "total_loss": 2.514043092727661 }, { "epoch": 0.3656998738965952, "grad_norm": 31.82435417175293, "kl_loss": 7.403247082038433e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041191381495564005, "loss": 2.4657552242279053, "step": 145, "total_loss": 2.539787769317627 }, { "epoch": 0.3682219419924338, "grad_norm": 56.06254196166992, "kl_loss": 6.84136765016774e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041128010139416984, "loss": 2.2596001625061035, "step": 146, "total_loss": 2.3280138969421387 }, { "epoch": 0.3707440100882724, "grad_norm": 36.77891159057617, "kl_loss": 6.048455958307386e-08, "kl_weight": 1000000.0, "learning_rate": 0.00041064638783269963, "loss": 2.7930824756622314, "step": 147, "total_loss": 2.853567123413086 }, { "epoch": 0.37326607818411095, "grad_norm": 41.25841522216797, "kl_loss": 4.9962110892920464e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004100126742712294, "loss": 2.2069859504699707, "step": 148, "total_loss": 2.2569479942321777 }, { "epoch": 0.37578814627994955, "grad_norm": 57.461708068847656, "kl_loss": 9.180979532175115e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004093789607097592, "loss": 2.405407190322876, "step": 149, "total_loss": 2.4972169399261475 }, { "epoch": 0.37831021437578816, "grad_norm": 50.4982795715332, "kl_loss": 1.0024658081420057e-07, "kl_weight": 1000000.0, "learning_rate": 0.000408745247148289, "loss": 2.210681676864624, "step": 150, "total_loss": 2.3109283447265625 }, { "epoch": 0.38083228247162676, "grad_norm": 110.077880859375, "kl_loss": 1.407131406949702e-07, "kl_weight": 1000000.0, "learning_rate": 0.00040811153358681873, "loss": 2.104773759841919, "step": 151, "total_loss": 2.2454869747161865 }, { "epoch": 0.3833543505674653, "grad_norm": 54.86238098144531, "kl_loss": 9.496130815023207e-08, "kl_weight": 1000000.0, "learning_rate": 0.00040747782002534857, "loss": 1.9683960676193237, "step": 152, "total_loss": 2.063357353210449 }, { "epoch": 0.3858764186633039, "grad_norm": 50.60408401489258, "kl_loss": 1.2318479036821373e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004068441064638783, "loss": 2.5901904106140137, "step": 153, "total_loss": 2.7133750915527344 }, { "epoch": 0.3883984867591425, "grad_norm": 107.67759704589844, "kl_loss": 6.34291694723288e-08, "kl_weight": 1000000.0, "learning_rate": 0.00040621039290240815, "loss": 2.605423927307129, "step": 154, "total_loss": 2.6688530445098877 }, { "epoch": 0.39092055485498106, "grad_norm": 115.0508041381836, "kl_loss": 7.845541460937966e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004055766793409379, "loss": 2.3245625495910645, "step": 155, "total_loss": 2.403017997741699 }, { "epoch": 0.39344262295081966, "grad_norm": 62.95680618286133, "kl_loss": 7.661606815645428e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004049429657794677, "loss": 2.1633670330047607, "step": 156, "total_loss": 2.239983081817627 }, { "epoch": 0.39596469104665827, "grad_norm": 53.571048736572266, "kl_loss": 7.554017145139369e-08, "kl_weight": 1000000.0, "learning_rate": 0.00040430925221799746, "loss": 3.01509952545166, "step": 157, "total_loss": 3.090639591217041 }, { "epoch": 0.39848675914249687, "grad_norm": 45.94689178466797, "kl_loss": 7.277655100779157e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004036755386565273, "loss": 2.1610336303710938, "step": 158, "total_loss": 2.2338101863861084 }, { "epoch": 0.4010088272383354, "grad_norm": 120.42121887207031, "kl_loss": 1.1003065480963414e-07, "kl_weight": 1000000.0, "learning_rate": 0.00040304182509505703, "loss": 2.266493320465088, "step": 159, "total_loss": 2.376523971557617 }, { "epoch": 0.403530895334174, "grad_norm": 75.83570098876953, "kl_loss": 5.584304574313137e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004024081115335868, "loss": 2.5498852729797363, "step": 160, "total_loss": 2.6057283878326416 }, { "epoch": 0.4060529634300126, "grad_norm": 69.80756378173828, "kl_loss": 1.0243423531619555e-07, "kl_weight": 1000000.0, "learning_rate": 0.0004017743979721166, "loss": 2.326557159423828, "step": 161, "total_loss": 2.4289913177490234 }, { "epoch": 0.4085750315258512, "grad_norm": 41.94525909423828, "kl_loss": 7.603321705573762e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004011406844106464, "loss": 2.06467866897583, "step": 162, "total_loss": 2.140711784362793 }, { "epoch": 0.4110970996216898, "grad_norm": 84.79508209228516, "kl_loss": 9.042678073001298e-08, "kl_weight": 1000000.0, "learning_rate": 0.0004005069708491762, "loss": 2.454202651977539, "step": 163, "total_loss": 2.5446293354034424 }, { "epoch": 0.4136191677175284, "grad_norm": 60.46647644042969, "kl_loss": 7.448775107832262e-08, "kl_weight": 1000000.0, "learning_rate": 0.000399873257287706, "loss": 2.454906940460205, "step": 164, "total_loss": 2.5293946266174316 }, { "epoch": 0.416141235813367, "grad_norm": 52.88587188720703, "kl_loss": 7.54991447138309e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039923954372623577, "loss": 2.235410213470459, "step": 165, "total_loss": 2.3109092712402344 }, { "epoch": 0.41866330390920553, "grad_norm": 35.26350402832031, "kl_loss": 7.14880030727727e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039860583016476555, "loss": 2.3518946170806885, "step": 166, "total_loss": 2.423382520675659 }, { "epoch": 0.42118537200504413, "grad_norm": 43.75752258300781, "kl_loss": 7.024154768942026e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003979721166032953, "loss": 2.780740976333618, "step": 167, "total_loss": 2.850982427597046 }, { "epoch": 0.42370744010088274, "grad_norm": 66.19912719726562, "kl_loss": 1.1769069629963269e-07, "kl_weight": 1000000.0, "learning_rate": 0.00039733840304182513, "loss": 2.3344695568084717, "step": 168, "total_loss": 2.452160358428955 }, { "epoch": 0.4262295081967213, "grad_norm": 59.10941696166992, "kl_loss": 6.215353209881869e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039670468948035486, "loss": 2.570136785507202, "step": 169, "total_loss": 2.6322903633117676 }, { "epoch": 0.4287515762925599, "grad_norm": 52.37656784057617, "kl_loss": 1.028036678007993e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003960709759188847, "loss": 2.5251572132110596, "step": 170, "total_loss": 2.6279609203338623 }, { "epoch": 0.4312736443883985, "grad_norm": 40.00493240356445, "kl_loss": 7.369838073145729e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039543726235741444, "loss": 2.526984691619873, "step": 171, "total_loss": 2.6006829738616943 }, { "epoch": 0.4337957124842371, "grad_norm": 45.39324188232422, "kl_loss": 8.142234975139218e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003948035487959443, "loss": 2.818206548690796, "step": 172, "total_loss": 2.8996288776397705 }, { "epoch": 0.43631778058007564, "grad_norm": 95.10191345214844, "kl_loss": 9.187782978870018e-08, "kl_weight": 1000000.0, "learning_rate": 0.000394169835234474, "loss": 2.186061143875122, "step": 173, "total_loss": 2.2779390811920166 }, { "epoch": 0.43883984867591425, "grad_norm": 60.82707595825195, "kl_loss": 8.505615056719762e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003935361216730038, "loss": 2.61464262008667, "step": 174, "total_loss": 2.6996986865997314 }, { "epoch": 0.44136191677175285, "grad_norm": 34.89598846435547, "kl_loss": 6.178815681323613e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003929024081115336, "loss": 2.393256425857544, "step": 175, "total_loss": 2.4550445079803467 }, { "epoch": 0.44388398486759145, "grad_norm": 41.88828659057617, "kl_loss": 9.847246928984532e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003922686945500634, "loss": 2.18418550491333, "step": 176, "total_loss": 2.2826578617095947 }, { "epoch": 0.44640605296343, "grad_norm": 54.601078033447266, "kl_loss": 7.419349401516229e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039163498098859317, "loss": 2.479668140411377, "step": 177, "total_loss": 2.553861618041992 }, { "epoch": 0.4489281210592686, "grad_norm": 56.431785583496094, "kl_loss": 6.644057748417254e-08, "kl_weight": 1000000.0, "learning_rate": 0.00039100126742712296, "loss": 1.9767550230026245, "step": 178, "total_loss": 2.0431954860687256 }, { "epoch": 0.4514501891551072, "grad_norm": 35.20497131347656, "kl_loss": 1.1079168160677e-07, "kl_weight": 1000000.0, "learning_rate": 0.00039036755386565275, "loss": 2.156259059906006, "step": 179, "total_loss": 2.2670507431030273 }, { "epoch": 0.45397225725094575, "grad_norm": 41.2132453918457, "kl_loss": 8.99928167541475e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038973384030418254, "loss": 2.3539505004882812, "step": 180, "total_loss": 2.4439432621002197 }, { "epoch": 0.45649432534678436, "grad_norm": 52.909889221191406, "kl_loss": 7.597910212098213e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038910012674271227, "loss": 2.2199325561523438, "step": 181, "total_loss": 2.2959115505218506 }, { "epoch": 0.45901639344262296, "grad_norm": 36.9537239074707, "kl_loss": 7.003095703339568e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003884664131812421, "loss": 2.4476423263549805, "step": 182, "total_loss": 2.5176732540130615 }, { "epoch": 0.46153846153846156, "grad_norm": 37.37628173828125, "kl_loss": 6.331342206067347e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038783269961977185, "loss": 2.4311509132385254, "step": 183, "total_loss": 2.49446439743042 }, { "epoch": 0.4640605296343001, "grad_norm": 38.719974517822266, "kl_loss": 9.451380833525036e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003871989860583017, "loss": 2.2305736541748047, "step": 184, "total_loss": 2.325087547302246 }, { "epoch": 0.4665825977301387, "grad_norm": 42.09044647216797, "kl_loss": 9.232453379581784e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003865652724968314, "loss": 2.2907114028930664, "step": 185, "total_loss": 2.383035898208618 }, { "epoch": 0.4691046658259773, "grad_norm": 38.898799896240234, "kl_loss": 9.683325430387413e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038593155893536127, "loss": 2.2948708534240723, "step": 186, "total_loss": 2.3917040824890137 }, { "epoch": 0.47162673392181587, "grad_norm": 43.5036506652832, "kl_loss": 9.654700505734581e-08, "kl_weight": 1000000.0, "learning_rate": 0.000385297845373891, "loss": 2.246976852416992, "step": 187, "total_loss": 2.3435239791870117 }, { "epoch": 0.47414880201765447, "grad_norm": 61.16709518432617, "kl_loss": 7.404270263577928e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003846641318124208, "loss": 2.168644666671753, "step": 188, "total_loss": 2.242687463760376 }, { "epoch": 0.4766708701134931, "grad_norm": 36.72150802612305, "kl_loss": 1.0344436418563419e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003840304182509506, "loss": 2.2275993824005127, "step": 189, "total_loss": 2.3310437202453613 }, { "epoch": 0.4791929382093317, "grad_norm": 30.705520629882812, "kl_loss": 6.61248193978281e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038339670468948037, "loss": 2.0510165691375732, "step": 190, "total_loss": 2.1171414852142334 }, { "epoch": 0.4817150063051702, "grad_norm": 35.57632064819336, "kl_loss": 8.263459960744513e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038276299112801015, "loss": 2.4444973468780518, "step": 191, "total_loss": 2.527132034301758 }, { "epoch": 0.4842370744010088, "grad_norm": 36.82917022705078, "kl_loss": 7.212786101717938e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038212927756653994, "loss": 2.446396589279175, "step": 192, "total_loss": 2.518524408340454 }, { "epoch": 0.48675914249684743, "grad_norm": 30.78325843811035, "kl_loss": 1.2874077981450682e-07, "kl_weight": 1000000.0, "learning_rate": 0.00038149556400506973, "loss": 2.339862585067749, "step": 193, "total_loss": 2.4686033725738525 }, { "epoch": 0.489281210592686, "grad_norm": 52.84626388549805, "kl_loss": 7.056150508333303e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003808618504435995, "loss": 2.5392770767211914, "step": 194, "total_loss": 2.6098384857177734 }, { "epoch": 0.4918032786885246, "grad_norm": 56.915245056152344, "kl_loss": 6.626930115771756e-08, "kl_weight": 1000000.0, "learning_rate": 0.00038022813688212925, "loss": 2.524722099304199, "step": 195, "total_loss": 2.590991497039795 }, { "epoch": 0.4943253467843632, "grad_norm": 34.66722106933594, "kl_loss": 5.858741047859439e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003795944233206591, "loss": 2.3640778064727783, "step": 196, "total_loss": 2.4226651191711426 }, { "epoch": 0.4968474148802018, "grad_norm": 44.76052474975586, "kl_loss": 1.2576073515901953e-07, "kl_weight": 1000000.0, "learning_rate": 0.00037896070975918883, "loss": 2.3642749786376953, "step": 197, "total_loss": 2.4900357723236084 }, { "epoch": 0.49936948297604034, "grad_norm": 47.74071502685547, "kl_loss": 7.279945890559247e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003783269961977187, "loss": 2.1854708194732666, "step": 198, "total_loss": 2.258270263671875 }, { "epoch": 0.501891551071879, "grad_norm": 31.107900619506836, "kl_loss": 6.880613057091978e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003776932826362484, "loss": 2.2795894145965576, "step": 199, "total_loss": 2.348395586013794 }, { "epoch": 0.5044136191677175, "grad_norm": 38.996551513671875, "kl_loss": 7.437692772782611e-08, "kl_weight": 1000000.0, "learning_rate": 0.00037705956907477825, "loss": 2.4008922576904297, "step": 200, "total_loss": 2.475269079208374 }, { "epoch": 0.5044136191677175, "eval_loss": 2.8438220024108887, "eval_runtime": 36.2993, "eval_samples_per_second": 19.422, "eval_steps_per_second": 2.452, "step": 200 }, { "epoch": 0.5069356872635561, "grad_norm": 35.33005905151367, "kl_loss": 5.263793667609207e-08, "kl_weight": 1000000.0, "learning_rate": 0.000376425855513308, "loss": 2.8244736194610596, "step": 201, "total_loss": 2.8771116733551025 }, { "epoch": 0.5094577553593947, "grad_norm": 33.417179107666016, "kl_loss": 6.945005281977501e-08, "kl_weight": 1000000.0, "learning_rate": 0.00037579214195183777, "loss": 2.493217706680298, "step": 202, "total_loss": 2.5626678466796875 }, { "epoch": 0.5119798234552333, "grad_norm": 35.52710723876953, "kl_loss": 5.157462723559547e-08, "kl_weight": 1000000.0, "learning_rate": 0.00037515842839036756, "loss": 2.3386712074279785, "step": 203, "total_loss": 2.3902459144592285 }, { "epoch": 0.5145018915510718, "grad_norm": 33.11330032348633, "kl_loss": 8.693313446883622e-08, "kl_weight": 1000000.0, "learning_rate": 0.00037452471482889735, "loss": 2.5677270889282227, "step": 204, "total_loss": 2.654660224914551 }, { "epoch": 0.5170239596469105, "grad_norm": 45.05632019042969, "kl_loss": 7.886416142355301e-08, "kl_weight": 1000000.0, "learning_rate": 0.00037389100126742714, "loss": 2.4564504623413086, "step": 205, "total_loss": 2.5353145599365234 }, { "epoch": 0.519546027742749, "grad_norm": 53.712032318115234, "kl_loss": 9.798601752208924e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003732572877059569, "loss": 2.365539789199829, "step": 206, "total_loss": 2.4635257720947266 }, { "epoch": 0.5220680958385876, "grad_norm": 38.51359558105469, "kl_loss": 6.502447291723001e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003726235741444867, "loss": 2.337630033493042, "step": 207, "total_loss": 2.4026544094085693 }, { "epoch": 0.5245901639344263, "grad_norm": 52.29638671875, "kl_loss": 1.1507498243190639e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003719898605830165, "loss": 2.099738597869873, "step": 208, "total_loss": 2.214813470840454 }, { "epoch": 0.5271122320302648, "grad_norm": 44.06270980834961, "kl_loss": 1.1070725491890698e-07, "kl_weight": 1000000.0, "learning_rate": 0.00037135614702154624, "loss": 2.2641494274139404, "step": 209, "total_loss": 2.37485671043396 }, { "epoch": 0.5296343001261034, "grad_norm": 45.88140869140625, "kl_loss": 6.309808497917402e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003707224334600761, "loss": 2.3526082038879395, "step": 210, "total_loss": 2.4157063961029053 }, { "epoch": 0.532156368221942, "grad_norm": 37.57832717895508, "kl_loss": 9.964617220248329e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003700887198986058, "loss": 2.160687208175659, "step": 211, "total_loss": 2.260333299636841 }, { "epoch": 0.5346784363177806, "grad_norm": 39.08332443237305, "kl_loss": 9.133309930575706e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036945500633713566, "loss": 2.6020138263702393, "step": 212, "total_loss": 2.6933469772338867 }, { "epoch": 0.5372005044136192, "grad_norm": 37.06446838378906, "kl_loss": 1.0182624521348771e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003688212927756654, "loss": 2.288355588912964, "step": 213, "total_loss": 2.39018177986145 }, { "epoch": 0.5397225725094578, "grad_norm": 38.347320556640625, "kl_loss": 9.616571361448223e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036818757921419523, "loss": 2.241528034210205, "step": 214, "total_loss": 2.337693691253662 }, { "epoch": 0.5422446406052963, "grad_norm": 47.67724609375, "kl_loss": 9.903801156951886e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036755386565272497, "loss": 2.4759681224823, "step": 215, "total_loss": 2.5750062465667725 }, { "epoch": 0.544766708701135, "grad_norm": 42.321475982666016, "kl_loss": 7.528672085754806e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036692015209125476, "loss": 2.662468433380127, "step": 216, "total_loss": 2.737755060195923 }, { "epoch": 0.5472887767969735, "grad_norm": 49.61969757080078, "kl_loss": 6.641187155764783e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036628643852978454, "loss": 2.2327988147735596, "step": 217, "total_loss": 2.299210786819458 }, { "epoch": 0.5498108448928121, "grad_norm": 30.943737030029297, "kl_loss": 5.8987577489233445e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036565272496831433, "loss": 2.481475353240967, "step": 218, "total_loss": 2.5404629707336426 }, { "epoch": 0.5523329129886507, "grad_norm": 31.544883728027344, "kl_loss": 5.3747619688238046e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003650190114068441, "loss": 1.8688888549804688, "step": 219, "total_loss": 1.9226365089416504 }, { "epoch": 0.5548549810844893, "grad_norm": 40.07229232788086, "kl_loss": 5.666829139272522e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003643852978453739, "loss": 2.1646535396575928, "step": 220, "total_loss": 2.2213218212127686 }, { "epoch": 0.5573770491803278, "grad_norm": 41.914249420166016, "kl_loss": 7.765168419382462e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003637515842839037, "loss": 2.4201087951660156, "step": 221, "total_loss": 2.497760534286499 }, { "epoch": 0.5598991172761665, "grad_norm": 35.645835876464844, "kl_loss": 6.774806848852677e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003631178707224335, "loss": 2.434053421020508, "step": 222, "total_loss": 2.5018014907836914 }, { "epoch": 0.562421185372005, "grad_norm": 29.000551223754883, "kl_loss": 5.608683295577066e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003624841571609633, "loss": 2.1331686973571777, "step": 223, "total_loss": 2.189255475997925 }, { "epoch": 0.5649432534678437, "grad_norm": 80.6998519897461, "kl_loss": 1.0609716838416716e-07, "kl_weight": 1000000.0, "learning_rate": 0.00036185044359949306, "loss": 2.1721668243408203, "step": 224, "total_loss": 2.278264045715332 }, { "epoch": 0.5674653215636822, "grad_norm": 39.576663970947266, "kl_loss": 6.589240086896098e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003612167300380228, "loss": 2.0701348781585693, "step": 225, "total_loss": 2.1360273361206055 }, { "epoch": 0.5699873896595208, "grad_norm": 41.5588493347168, "kl_loss": 7.04123763739517e-08, "kl_weight": 1000000.0, "learning_rate": 0.00036058301647655264, "loss": 2.371951103210449, "step": 226, "total_loss": 2.4423635005950928 }, { "epoch": 0.5725094577553594, "grad_norm": 41.82505416870117, "kl_loss": 8.056593969740788e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003599493029150824, "loss": 2.121464967727661, "step": 227, "total_loss": 2.202030897140503 }, { "epoch": 0.575031525851198, "grad_norm": 49.150333404541016, "kl_loss": 9.009213641775204e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003593155893536122, "loss": 2.234654426574707, "step": 228, "total_loss": 2.324746608734131 }, { "epoch": 0.5775535939470365, "grad_norm": 59.28157043457031, "kl_loss": 9.32024164512768e-08, "kl_weight": 1000000.0, "learning_rate": 0.00035868187579214195, "loss": 2.3375024795532227, "step": 229, "total_loss": 2.4307048320770264 }, { "epoch": 0.5800756620428752, "grad_norm": 66.21656799316406, "kl_loss": 8.425497810549132e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003580481622306718, "loss": 2.195544481277466, "step": 230, "total_loss": 2.279799461364746 }, { "epoch": 0.5825977301387137, "grad_norm": 44.063621520996094, "kl_loss": 6.732751245408508e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003574144486692015, "loss": 2.1953518390655518, "step": 231, "total_loss": 2.2626793384552 }, { "epoch": 0.5851197982345523, "grad_norm": 57.83623504638672, "kl_loss": 7.052550898833942e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003567807351077313, "loss": 2.7928688526153564, "step": 232, "total_loss": 2.863394260406494 }, { "epoch": 0.587641866330391, "grad_norm": 60.28070831298828, "kl_loss": 7.187722417256737e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003561470215462611, "loss": 2.0700807571411133, "step": 233, "total_loss": 2.141957998275757 }, { "epoch": 0.5901639344262295, "grad_norm": 48.02665328979492, "kl_loss": 7.593448714260376e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003555133079847909, "loss": 2.2719221115112305, "step": 234, "total_loss": 2.3478565216064453 }, { "epoch": 0.592686002522068, "grad_norm": 40.39185333251953, "kl_loss": 6.198138891022609e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003548795944233207, "loss": 2.460548162460327, "step": 235, "total_loss": 2.5225296020507812 }, { "epoch": 0.5952080706179067, "grad_norm": 51.92955780029297, "kl_loss": 6.760530624205785e-08, "kl_weight": 1000000.0, "learning_rate": 0.00035424588086185047, "loss": 2.1937878131866455, "step": 236, "total_loss": 2.2613930702209473 }, { "epoch": 0.5977301387137453, "grad_norm": 60.31737518310547, "kl_loss": 9.117999155705547e-08, "kl_weight": 1000000.0, "learning_rate": 0.00035361216730038026, "loss": 2.5255377292633057, "step": 237, "total_loss": 2.61671781539917 }, { "epoch": 0.6002522068095839, "grad_norm": 45.97258377075195, "kl_loss": 4.57544189202963e-08, "kl_weight": 1000000.0, "learning_rate": 0.00035297845373891005, "loss": 2.1498923301696777, "step": 238, "total_loss": 2.1956467628479004 }, { "epoch": 0.6027742749054225, "grad_norm": 33.029720306396484, "kl_loss": 6.883194458850994e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003523447401774398, "loss": 2.6470816135406494, "step": 239, "total_loss": 2.7159135341644287 }, { "epoch": 0.605296343001261, "grad_norm": 65.16218566894531, "kl_loss": 7.724134576392316e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003517110266159696, "loss": 2.0316035747528076, "step": 240, "total_loss": 2.1088449954986572 }, { "epoch": 0.6078184110970997, "grad_norm": 54.9919319152832, "kl_loss": 1.0422303375889896e-07, "kl_weight": 1000000.0, "learning_rate": 0.00035107731305449936, "loss": 2.039462089538574, "step": 241, "total_loss": 2.1436851024627686 }, { "epoch": 0.6103404791929382, "grad_norm": 46.679317474365234, "kl_loss": 7.876125351913288e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003504435994930292, "loss": 2.1784236431121826, "step": 242, "total_loss": 2.2571849822998047 }, { "epoch": 0.6128625472887768, "grad_norm": 42.02393341064453, "kl_loss": 8.838826914825404e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034980988593155893, "loss": 2.3769657611846924, "step": 243, "total_loss": 2.4653539657592773 }, { "epoch": 0.6153846153846154, "grad_norm": 42.681602478027344, "kl_loss": 7.536393553664311e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003491761723700888, "loss": 2.6341843605041504, "step": 244, "total_loss": 2.709548234939575 }, { "epoch": 0.617906683480454, "grad_norm": 60.763973236083984, "kl_loss": 9.037336212713853e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003485424588086185, "loss": 1.9615586996078491, "step": 245, "total_loss": 2.0519320964813232 }, { "epoch": 0.6204287515762925, "grad_norm": 53.96251678466797, "kl_loss": 6.376885153258627e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003479087452471483, "loss": 2.3010590076446533, "step": 246, "total_loss": 2.364827871322632 }, { "epoch": 0.6229508196721312, "grad_norm": 54.281349182128906, "kl_loss": 6.555458753609855e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003472750316856781, "loss": 2.544086456298828, "step": 247, "total_loss": 2.6096410751342773 }, { "epoch": 0.6254728877679697, "grad_norm": 37.844505310058594, "kl_loss": 6.108597006004857e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003466413181242079, "loss": 2.4314613342285156, "step": 248, "total_loss": 2.4925472736358643 }, { "epoch": 0.6279949558638083, "grad_norm": 121.8286361694336, "kl_loss": 7.796773360269071e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034600760456273766, "loss": 2.589028835296631, "step": 249, "total_loss": 2.666996479034424 }, { "epoch": 0.6305170239596469, "grad_norm": 41.6161003112793, "kl_loss": 7.061287732312849e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034537389100126745, "loss": 2.3637168407440186, "step": 250, "total_loss": 2.4343297481536865 }, { "epoch": 0.6330390920554855, "grad_norm": 53.98466873168945, "kl_loss": 1.245601310984057e-07, "kl_weight": 1000000.0, "learning_rate": 0.00034474017743979724, "loss": 2.216432571411133, "step": 251, "total_loss": 2.3409926891326904 }, { "epoch": 0.6355611601513241, "grad_norm": 60.336673736572266, "kl_loss": 7.335739127256602e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034410646387832703, "loss": 2.193324327468872, "step": 252, "total_loss": 2.266681671142578 }, { "epoch": 0.6380832282471627, "grad_norm": 86.27165222167969, "kl_loss": 9.020882174581857e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034347275031685676, "loss": 2.4519028663635254, "step": 253, "total_loss": 2.54211163520813 }, { "epoch": 0.6406052963430012, "grad_norm": 48.824710845947266, "kl_loss": 1.2054276510298223e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003428390367553866, "loss": 2.4704232215881348, "step": 254, "total_loss": 2.590965986251831 }, { "epoch": 0.6431273644388399, "grad_norm": 41.089054107666016, "kl_loss": 6.186154877241279e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034220532319391634, "loss": 2.1385762691497803, "step": 255, "total_loss": 2.2004377841949463 }, { "epoch": 0.6456494325346784, "grad_norm": 57.07784652709961, "kl_loss": 7.342136854049386e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003415716096324462, "loss": 1.7062774896621704, "step": 256, "total_loss": 1.7796988487243652 }, { "epoch": 0.648171500630517, "grad_norm": 54.759586334228516, "kl_loss": 5.807413572256337e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003409378960709759, "loss": 2.3140125274658203, "step": 257, "total_loss": 2.372086763381958 }, { "epoch": 0.6506935687263556, "grad_norm": 41.672672271728516, "kl_loss": 7.009721514350531e-08, "kl_weight": 1000000.0, "learning_rate": 0.00034030418250950576, "loss": 2.1182868480682373, "step": 258, "total_loss": 2.1883840560913086 }, { "epoch": 0.6532156368221942, "grad_norm": 40.1267204284668, "kl_loss": 7.896883857938519e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003396704689480355, "loss": 2.3563849925994873, "step": 259, "total_loss": 2.4353537559509277 }, { "epoch": 0.6557377049180327, "grad_norm": 42.59463119506836, "kl_loss": 4.561388067259031e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003390367553865653, "loss": 1.9189460277557373, "step": 260, "total_loss": 1.9645599126815796 }, { "epoch": 0.6582597730138714, "grad_norm": 96.17916107177734, "kl_loss": 9.542584678001731e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033840304182509507, "loss": 2.1902644634246826, "step": 261, "total_loss": 2.2856903076171875 }, { "epoch": 0.6607818411097099, "grad_norm": 65.56468963623047, "kl_loss": 6.830008913993879e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033776932826362486, "loss": 2.250452756881714, "step": 262, "total_loss": 2.3187527656555176 }, { "epoch": 0.6633039092055486, "grad_norm": 34.11521911621094, "kl_loss": 7.455038542047987e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033713561470215465, "loss": 2.158390522003174, "step": 263, "total_loss": 2.232940912246704 }, { "epoch": 0.6658259773013872, "grad_norm": 49.9361457824707, "kl_loss": 6.679385933239246e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033650190114068443, "loss": 2.509239435195923, "step": 264, "total_loss": 2.576033353805542 }, { "epoch": 0.6683480453972257, "grad_norm": 36.359127044677734, "kl_loss": 5.8454563856003006e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003358681875792142, "loss": 1.8529467582702637, "step": 265, "total_loss": 1.9114012718200684 }, { "epoch": 0.6708701134930644, "grad_norm": 55.48241424560547, "kl_loss": 6.64626540469726e-08, "kl_weight": 1000000.0, "learning_rate": 0.000335234474017744, "loss": 2.074885606765747, "step": 266, "total_loss": 2.141348361968994 }, { "epoch": 0.6733921815889029, "grad_norm": 48.15768814086914, "kl_loss": 6.835133348204181e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033460076045627375, "loss": 2.9065771102905273, "step": 267, "total_loss": 2.974928379058838 }, { "epoch": 0.6759142496847415, "grad_norm": 39.49137878417969, "kl_loss": 8.859954192530495e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003339670468948036, "loss": 2.86928129196167, "step": 268, "total_loss": 2.957880735397339 }, { "epoch": 0.6784363177805801, "grad_norm": 58.989952087402344, "kl_loss": 5.006348047231768e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003333333333333333, "loss": 1.9880236387252808, "step": 269, "total_loss": 2.0380871295928955 }, { "epoch": 0.6809583858764187, "grad_norm": 52.51384353637695, "kl_loss": 9.513845355968442e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033269961977186316, "loss": 2.5748720169067383, "step": 270, "total_loss": 2.670010566711426 }, { "epoch": 0.6834804539722572, "grad_norm": 91.88314056396484, "kl_loss": 9.713623683182959e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003320659062103929, "loss": 2.5494627952575684, "step": 271, "total_loss": 2.646599054336548 }, { "epoch": 0.6860025220680959, "grad_norm": 37.61174011230469, "kl_loss": 7.669238755170227e-08, "kl_weight": 1000000.0, "learning_rate": 0.00033143219264892274, "loss": 2.308610439300537, "step": 272, "total_loss": 2.385302782058716 }, { "epoch": 0.6885245901639344, "grad_norm": 77.4565658569336, "kl_loss": 8.696480335856904e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003307984790874525, "loss": 2.1501646041870117, "step": 273, "total_loss": 2.2371294498443604 }, { "epoch": 0.691046658259773, "grad_norm": 46.31070327758789, "kl_loss": 7.047064087828403e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003301647655259822, "loss": 2.3588707447052, "step": 274, "total_loss": 2.4293413162231445 }, { "epoch": 0.6935687263556116, "grad_norm": 99.26569366455078, "kl_loss": 6.073241820558906e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032953105196451205, "loss": 2.3441028594970703, "step": 275, "total_loss": 2.4048352241516113 }, { "epoch": 0.6960907944514502, "grad_norm": 50.96487808227539, "kl_loss": 1.0178317921827329e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003288973384030418, "loss": 2.205005168914795, "step": 276, "total_loss": 2.306788444519043 }, { "epoch": 0.6986128625472888, "grad_norm": 37.70353698730469, "kl_loss": 5.64434969874128e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032826362484157163, "loss": 2.7925455570220947, "step": 277, "total_loss": 2.8489890098571777 }, { "epoch": 0.7011349306431274, "grad_norm": 40.202232360839844, "kl_loss": 5.6257210445664896e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032762991128010136, "loss": 2.7806427478790283, "step": 278, "total_loss": 2.836899995803833 }, { "epoch": 0.7036569987389659, "grad_norm": 54.97629928588867, "kl_loss": 1.4827625705038372e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003269961977186312, "loss": 2.2128868103027344, "step": 279, "total_loss": 2.3611631393432617 }, { "epoch": 0.7061790668348046, "grad_norm": 31.53950309753418, "kl_loss": 8.161551079410856e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032636248415716094, "loss": 2.1168041229248047, "step": 280, "total_loss": 2.1984195709228516 }, { "epoch": 0.7087011349306431, "grad_norm": 32.2765998840332, "kl_loss": 6.324378887256898e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003257287705956908, "loss": 2.573775291442871, "step": 281, "total_loss": 2.637019157409668 }, { "epoch": 0.7112232030264817, "grad_norm": 34.755958557128906, "kl_loss": 5.3531053367805725e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003250950570342205, "loss": 2.1479485034942627, "step": 282, "total_loss": 2.20147967338562 }, { "epoch": 0.7137452711223203, "grad_norm": 49.57518768310547, "kl_loss": 5.849909356925309e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003244613434727503, "loss": 2.9778120517730713, "step": 283, "total_loss": 3.036311149597168 }, { "epoch": 0.7162673392181589, "grad_norm": 40.9071159362793, "kl_loss": 7.603016882740121e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003238276299112801, "loss": 2.1856653690338135, "step": 284, "total_loss": 2.261695623397827 }, { "epoch": 0.7187894073139974, "grad_norm": 45.15980911254883, "kl_loss": 9.367033726448426e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003231939163498099, "loss": 2.082435131072998, "step": 285, "total_loss": 2.176105499267578 }, { "epoch": 0.7213114754098361, "grad_norm": 36.46816635131836, "kl_loss": 8.986472010974467e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032256020278833967, "loss": 2.1399319171905518, "step": 286, "total_loss": 2.2297966480255127 }, { "epoch": 0.7238335435056746, "grad_norm": 37.271339416503906, "kl_loss": 8.904514459118218e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032192648922686946, "loss": 2.3471741676330566, "step": 287, "total_loss": 2.4362192153930664 }, { "epoch": 0.7263556116015133, "grad_norm": 32.7809944152832, "kl_loss": 7.45906518773154e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032129277566539925, "loss": 3.058664321899414, "step": 288, "total_loss": 3.1332550048828125 }, { "epoch": 0.7288776796973518, "grad_norm": 38.48100280761719, "kl_loss": 9.533075484569054e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032065906210392904, "loss": 2.259277105331421, "step": 289, "total_loss": 2.3546078205108643 }, { "epoch": 0.7313997477931904, "grad_norm": 61.44423294067383, "kl_loss": 5.4314426733981236e-08, "kl_weight": 1000000.0, "learning_rate": 0.00032002534854245877, "loss": 2.1022424697875977, "step": 290, "total_loss": 2.1565568447113037 }, { "epoch": 0.733921815889029, "grad_norm": 60.17184829711914, "kl_loss": 7.423577130794001e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003193916349809886, "loss": 2.7498912811279297, "step": 291, "total_loss": 2.824126958847046 }, { "epoch": 0.7364438839848676, "grad_norm": 56.09250259399414, "kl_loss": 7.274695690284716e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031875792141951835, "loss": 2.5288584232330322, "step": 292, "total_loss": 2.6016054153442383 }, { "epoch": 0.7389659520807061, "grad_norm": 104.29945373535156, "kl_loss": 1.4982573759425577e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003181242078580482, "loss": 2.3595778942108154, "step": 293, "total_loss": 2.509403705596924 }, { "epoch": 0.7414880201765448, "grad_norm": 59.080657958984375, "kl_loss": 7.656793599153389e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003174904942965779, "loss": 2.3922691345214844, "step": 294, "total_loss": 2.468837022781372 }, { "epoch": 0.7440100882723834, "grad_norm": 66.89306640625, "kl_loss": 6.386262896285189e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031685678073510777, "loss": 2.6482093334198, "step": 295, "total_loss": 2.7120718955993652 }, { "epoch": 0.7465321563682219, "grad_norm": 59.871437072753906, "kl_loss": 5.6765372846712125e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003162230671736375, "loss": 2.3313560485839844, "step": 296, "total_loss": 2.3881213665008545 }, { "epoch": 0.7490542244640606, "grad_norm": 30.758756637573242, "kl_loss": 1.0433349473260023e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003155893536121673, "loss": 2.3151373863220215, "step": 297, "total_loss": 2.41947078704834 }, { "epoch": 0.7515762925598991, "grad_norm": 48.252742767333984, "kl_loss": 7.452045736044965e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003149556400506971, "loss": 2.5079312324523926, "step": 298, "total_loss": 2.582451581954956 }, { "epoch": 0.7540983606557377, "grad_norm": 78.13578796386719, "kl_loss": 6.663464091616333e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031432192648922686, "loss": 2.5382885932922363, "step": 299, "total_loss": 2.6049232482910156 }, { "epoch": 0.7566204287515763, "grad_norm": 92.22211456298828, "kl_loss": 7.616570485424745e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031368821292775665, "loss": 2.1022658348083496, "step": 300, "total_loss": 2.178431510925293 }, { "epoch": 0.7566204287515763, "eval_loss": 2.853464365005493, "eval_runtime": 36.4132, "eval_samples_per_second": 19.361, "eval_steps_per_second": 2.444, "step": 300 }, { "epoch": 0.7591424968474149, "grad_norm": 63.4229621887207, "kl_loss": 5.6403411718974894e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031305449936628644, "loss": 2.7314107418060303, "step": 301, "total_loss": 2.787814140319824 }, { "epoch": 0.7616645649432535, "grad_norm": 85.6334228515625, "kl_loss": 9.44590681228874e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031242078580481623, "loss": 2.34684157371521, "step": 302, "total_loss": 2.441300630569458 }, { "epoch": 0.7641866330390921, "grad_norm": 61.75638961791992, "kl_loss": 7.220323539058882e-08, "kl_weight": 1000000.0, "learning_rate": 0.000311787072243346, "loss": 2.4858498573303223, "step": 303, "total_loss": 2.5580530166625977 }, { "epoch": 0.7667087011349306, "grad_norm": 87.85055541992188, "kl_loss": 8.759780456557564e-08, "kl_weight": 1000000.0, "learning_rate": 0.00031115335868187575, "loss": 2.034895896911621, "step": 304, "total_loss": 2.1224937438964844 }, { "epoch": 0.7692307692307693, "grad_norm": 112.65933227539062, "kl_loss": 1.2029458673623594e-07, "kl_weight": 1000000.0, "learning_rate": 0.0003105196451204056, "loss": 2.3612492084503174, "step": 305, "total_loss": 2.481543779373169 }, { "epoch": 0.7717528373266078, "grad_norm": 65.05411529541016, "kl_loss": 1.085965877223316e-07, "kl_weight": 1000000.0, "learning_rate": 0.00030988593155893533, "loss": 2.211334228515625, "step": 306, "total_loss": 2.3199307918548584 }, { "epoch": 0.7742749054224464, "grad_norm": 56.37029266357422, "kl_loss": 5.6032622097745843e-08, "kl_weight": 1000000.0, "learning_rate": 0.00030925221799746517, "loss": 2.299423933029175, "step": 307, "total_loss": 2.355456590652466 }, { "epoch": 0.776796973518285, "grad_norm": 73.94780731201172, "kl_loss": 8.045584820592921e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003086185044359949, "loss": 2.075188398361206, "step": 308, "total_loss": 2.155644178390503 }, { "epoch": 0.7793190416141236, "grad_norm": 48.057498931884766, "kl_loss": 8.641636384254525e-08, "kl_weight": 1000000.0, "learning_rate": 0.00030798479087452475, "loss": 2.218684673309326, "step": 309, "total_loss": 2.305100917816162 }, { "epoch": 0.7818411097099621, "grad_norm": 37.80225372314453, "kl_loss": 5.5079230065757656e-08, "kl_weight": 1000000.0, "learning_rate": 0.0003073510773130545, "loss": 2.448129177093506, "step": 310, "total_loss": 2.5032083988189697 } ], "logging_steps": 1, "max_steps": 794, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 5, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.06586701824e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }