{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9991967871485944, "eval_steps": 500, "global_step": 311, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00321285140562249, "grad_norm": 9.468059539794922, "learning_rate": 1.0000000000000002e-06, "loss": 1.1591, "step": 1 }, { "epoch": 0.00642570281124498, "grad_norm": 12.971900939941406, "learning_rate": 2.0000000000000003e-06, "loss": 0.9396, "step": 2 }, { "epoch": 0.00963855421686747, "grad_norm": 9.142277717590332, "learning_rate": 3e-06, "loss": 1.0339, "step": 3 }, { "epoch": 0.01285140562248996, "grad_norm": 7.4195380210876465, "learning_rate": 4.000000000000001e-06, "loss": 1.0046, "step": 4 }, { "epoch": 0.01606425702811245, "grad_norm": 7.382334232330322, "learning_rate": 5e-06, "loss": 0.8551, "step": 5 }, { "epoch": 0.01927710843373494, "grad_norm": 5.827909469604492, "learning_rate": 6e-06, "loss": 0.8208, "step": 6 }, { "epoch": 0.02248995983935743, "grad_norm": 4.3395795822143555, "learning_rate": 7e-06, "loss": 0.6632, "step": 7 }, { "epoch": 0.02570281124497992, "grad_norm": 34.95582962036133, "learning_rate": 8.000000000000001e-06, "loss": 1.1221, "step": 8 }, { "epoch": 0.02891566265060241, "grad_norm": 5.787684440612793, "learning_rate": 9e-06, "loss": 0.7092, "step": 9 }, { "epoch": 0.0321285140562249, "grad_norm": 3.8266000747680664, "learning_rate": 1e-05, "loss": 0.8341, "step": 10 }, { "epoch": 0.035341365461847386, "grad_norm": 7.160825252532959, "learning_rate": 9.999727665400876e-06, "loss": 0.7178, "step": 11 }, { "epoch": 0.03855421686746988, "grad_norm": 3.6943607330322266, "learning_rate": 9.998910691269957e-06, "loss": 0.8316, "step": 12 }, { "epoch": 0.04176706827309237, "grad_norm": 4.6069159507751465, "learning_rate": 9.99754916660337e-06, "loss": 0.631, "step": 13 }, { "epoch": 0.04497991967871486, "grad_norm": 4.735833168029785, "learning_rate": 9.995643239717228e-06, "loss": 0.7483, "step": 14 }, { "epoch": 0.04819277108433735, "grad_norm": 3.0799083709716797, "learning_rate": 9.993193118231463e-06, "loss": 0.4765, "step": 15 }, { "epoch": 0.05140562248995984, "grad_norm": 2.8460216522216797, "learning_rate": 9.990199069047216e-06, "loss": 0.6183, "step": 16 }, { "epoch": 0.05461847389558233, "grad_norm": 3.5829732418060303, "learning_rate": 9.986661418317759e-06, "loss": 0.7606, "step": 17 }, { "epoch": 0.05783132530120482, "grad_norm": 2.786634683609009, "learning_rate": 9.982580551412972e-06, "loss": 0.6369, "step": 18 }, { "epoch": 0.06104417670682731, "grad_norm": 3.3990700244903564, "learning_rate": 9.977956912877356e-06, "loss": 0.5388, "step": 19 }, { "epoch": 0.0642570281124498, "grad_norm": 2.86079478263855, "learning_rate": 9.97279100638161e-06, "loss": 0.7238, "step": 20 }, { "epoch": 0.06746987951807229, "grad_norm": 2.6607353687286377, "learning_rate": 9.967083394667763e-06, "loss": 0.6505, "step": 21 }, { "epoch": 0.07068273092369477, "grad_norm": 2.682445526123047, "learning_rate": 9.960834699487873e-06, "loss": 0.4418, "step": 22 }, { "epoch": 0.07389558232931727, "grad_norm": 2.598595380783081, "learning_rate": 9.9540456015363e-06, "loss": 0.616, "step": 23 }, { "epoch": 0.07710843373493977, "grad_norm": 3.148313045501709, "learning_rate": 9.946716840375552e-06, "loss": 0.5655, "step": 24 }, { "epoch": 0.08032128514056225, "grad_norm": 3.6518056392669678, "learning_rate": 9.938849214355722e-06, "loss": 0.5541, "step": 25 }, { "epoch": 0.08353413654618475, "grad_norm": 2.7225069999694824, "learning_rate": 9.93044358052752e-06, "loss": 0.6182, "step": 26 }, { "epoch": 0.08674698795180723, "grad_norm": 2.9630682468414307, "learning_rate": 9.921500854548916e-06, "loss": 0.5794, "step": 27 }, { "epoch": 0.08995983935742972, "grad_norm": 2.659717082977295, "learning_rate": 9.912022010585385e-06, "loss": 0.5407, "step": 28 }, { "epoch": 0.09317269076305221, "grad_norm": 2.64546799659729, "learning_rate": 9.902008081203796e-06, "loss": 0.5833, "step": 29 }, { "epoch": 0.0963855421686747, "grad_norm": 3.679818868637085, "learning_rate": 9.89146015725993e-06, "loss": 0.6286, "step": 30 }, { "epoch": 0.09959839357429719, "grad_norm": 3.6220390796661377, "learning_rate": 9.880379387779637e-06, "loss": 0.717, "step": 31 }, { "epoch": 0.10281124497991968, "grad_norm": 2.7515664100646973, "learning_rate": 9.868766979833686e-06, "loss": 0.5431, "step": 32 }, { "epoch": 0.10602409638554217, "grad_norm": 2.658803939819336, "learning_rate": 9.856624198406262e-06, "loss": 0.6401, "step": 33 }, { "epoch": 0.10923694779116466, "grad_norm": 3.2038729190826416, "learning_rate": 9.84395236625717e-06, "loss": 0.5142, "step": 34 }, { "epoch": 0.11244979919678715, "grad_norm": 4.682802200317383, "learning_rate": 9.830752863777741e-06, "loss": 0.6152, "step": 35 }, { "epoch": 0.11566265060240964, "grad_norm": 2.9723594188690186, "learning_rate": 9.817027128840462e-06, "loss": 0.3418, "step": 36 }, { "epoch": 0.11887550200803212, "grad_norm": 3.891775369644165, "learning_rate": 9.802776656642341e-06, "loss": 0.6575, "step": 37 }, { "epoch": 0.12208835341365462, "grad_norm": 2.52054762840271, "learning_rate": 9.78800299954203e-06, "loss": 0.5537, "step": 38 }, { "epoch": 0.12530120481927712, "grad_norm": 2.726982593536377, "learning_rate": 9.772707766890726e-06, "loss": 0.6654, "step": 39 }, { "epoch": 0.1285140562248996, "grad_norm": 2.5964579582214355, "learning_rate": 9.756892624856848e-06, "loss": 0.5041, "step": 40 }, { "epoch": 0.13172690763052208, "grad_norm": 3.1076455116271973, "learning_rate": 9.740559296244543e-06, "loss": 0.4663, "step": 41 }, { "epoch": 0.13493975903614458, "grad_norm": 3.2797796726226807, "learning_rate": 9.723709560306009e-06, "loss": 0.5047, "step": 42 }, { "epoch": 0.13815261044176708, "grad_norm": 2.475914478302002, "learning_rate": 9.706345252547681e-06, "loss": 0.6025, "step": 43 }, { "epoch": 0.14136546184738955, "grad_norm": 2.9697823524475098, "learning_rate": 9.688468264530278e-06, "loss": 0.3588, "step": 44 }, { "epoch": 0.14457831325301204, "grad_norm": 2.9819068908691406, "learning_rate": 9.670080543662742e-06, "loss": 0.5415, "step": 45 }, { "epoch": 0.14779116465863454, "grad_norm": 2.63847017288208, "learning_rate": 9.651184092990109e-06, "loss": 0.548, "step": 46 }, { "epoch": 0.15100401606425704, "grad_norm": 2.9994335174560547, "learning_rate": 9.631780970975311e-06, "loss": 0.5151, "step": 47 }, { "epoch": 0.15421686746987953, "grad_norm": 2.440392017364502, "learning_rate": 9.611873291274927e-06, "loss": 0.4567, "step": 48 }, { "epoch": 0.157429718875502, "grad_norm": 2.510368585586548, "learning_rate": 9.591463222508947e-06, "loss": 0.5311, "step": 49 }, { "epoch": 0.1606425702811245, "grad_norm": 2.9306483268737793, "learning_rate": 9.570552988024527e-06, "loss": 0.3788, "step": 50 }, { "epoch": 0.163855421686747, "grad_norm": 3.0692126750946045, "learning_rate": 9.5491448656538e-06, "loss": 0.5438, "step": 51 }, { "epoch": 0.1670682730923695, "grad_norm": 2.9109346866607666, "learning_rate": 9.527241187465735e-06, "loss": 0.5757, "step": 52 }, { "epoch": 0.17028112449799196, "grad_norm": 2.6207594871520996, "learning_rate": 9.504844339512096e-06, "loss": 0.5567, "step": 53 }, { "epoch": 0.17349397590361446, "grad_norm": 2.2211403846740723, "learning_rate": 9.481956761567531e-06, "loss": 0.3586, "step": 54 }, { "epoch": 0.17670682730923695, "grad_norm": 2.620298147201538, "learning_rate": 9.458580946863784e-06, "loss": 0.4068, "step": 55 }, { "epoch": 0.17991967871485945, "grad_norm": 3.2399895191192627, "learning_rate": 9.434719441818106e-06, "loss": 0.4569, "step": 56 }, { "epoch": 0.18313253012048192, "grad_norm": 3.0689821243286133, "learning_rate": 9.410374845755862e-06, "loss": 0.4875, "step": 57 }, { "epoch": 0.18634538152610441, "grad_norm": 2.3414645195007324, "learning_rate": 9.385549810627374e-06, "loss": 0.4211, "step": 58 }, { "epoch": 0.1895582329317269, "grad_norm": 2.688450336456299, "learning_rate": 9.36024704071904e-06, "loss": 0.3738, "step": 59 }, { "epoch": 0.1927710843373494, "grad_norm": 2.5549943447113037, "learning_rate": 9.334469292358736e-06, "loss": 0.4483, "step": 60 }, { "epoch": 0.19598393574297188, "grad_norm": 3.2717857360839844, "learning_rate": 9.308219373615574e-06, "loss": 0.4543, "step": 61 }, { "epoch": 0.19919678714859437, "grad_norm": 2.5511434078216553, "learning_rate": 9.28150014399399e-06, "loss": 0.5196, "step": 62 }, { "epoch": 0.20240963855421687, "grad_norm": 4.289268970489502, "learning_rate": 9.25431451412226e-06, "loss": 0.4566, "step": 63 }, { "epoch": 0.20562248995983937, "grad_norm": 2.4556353092193604, "learning_rate": 9.226665445435428e-06, "loss": 0.4854, "step": 64 }, { "epoch": 0.20883534136546184, "grad_norm": 2.211134910583496, "learning_rate": 9.19855594985271e-06, "loss": 0.4626, "step": 65 }, { "epoch": 0.21204819277108433, "grad_norm": 2.233396530151367, "learning_rate": 9.16998908944939e-06, "loss": 0.4688, "step": 66 }, { "epoch": 0.21526104417670683, "grad_norm": 2.952958106994629, "learning_rate": 9.14096797612326e-06, "loss": 0.5239, "step": 67 }, { "epoch": 0.21847389558232932, "grad_norm": 2.30307936668396, "learning_rate": 9.111495771255623e-06, "loss": 0.4765, "step": 68 }, { "epoch": 0.2216867469879518, "grad_norm": 3.2021255493164062, "learning_rate": 9.081575685366919e-06, "loss": 0.579, "step": 69 }, { "epoch": 0.2248995983935743, "grad_norm": 2.920560359954834, "learning_rate": 9.051210977766987e-06, "loss": 0.4866, "step": 70 }, { "epoch": 0.2281124497991968, "grad_norm": 2.718421697616577, "learning_rate": 9.020404956200016e-06, "loss": 0.4484, "step": 71 }, { "epoch": 0.23132530120481928, "grad_norm": 2.929945707321167, "learning_rate": 8.989160976484218e-06, "loss": 0.4432, "step": 72 }, { "epoch": 0.23453815261044178, "grad_norm": 3.0811638832092285, "learning_rate": 8.957482442146271e-06, "loss": 0.5614, "step": 73 }, { "epoch": 0.23775100401606425, "grad_norm": 3.6012563705444336, "learning_rate": 8.925372804050554e-06, "loss": 0.4146, "step": 74 }, { "epoch": 0.24096385542168675, "grad_norm": 2.4589638710021973, "learning_rate": 8.892835560023236e-06, "loss": 0.4544, "step": 75 }, { "epoch": 0.24417670682730924, "grad_norm": 2.542167901992798, "learning_rate": 8.85987425447124e-06, "loss": 0.5399, "step": 76 }, { "epoch": 0.24738955823293174, "grad_norm": 2.2393405437469482, "learning_rate": 8.826492477996138e-06, "loss": 0.4348, "step": 77 }, { "epoch": 0.25060240963855424, "grad_norm": 2.5719082355499268, "learning_rate": 8.792693867003017e-06, "loss": 0.5533, "step": 78 }, { "epoch": 0.25381526104417673, "grad_norm": 2.2429287433624268, "learning_rate": 8.758482103304348e-06, "loss": 0.5089, "step": 79 }, { "epoch": 0.2570281124497992, "grad_norm": 2.215193033218384, "learning_rate": 8.72386091371891e-06, "loss": 0.3864, "step": 80 }, { "epoch": 0.26024096385542167, "grad_norm": 2.946692943572998, "learning_rate": 8.688834069665819e-06, "loss": 0.4929, "step": 81 }, { "epoch": 0.26345381526104417, "grad_norm": 2.1577095985412598, "learning_rate": 8.653405386753688e-06, "loss": 0.3006, "step": 82 }, { "epoch": 0.26666666666666666, "grad_norm": 2.4236855506896973, "learning_rate": 8.617578724364984e-06, "loss": 0.3955, "step": 83 }, { "epoch": 0.26987951807228916, "grad_norm": 2.1099393367767334, "learning_rate": 8.581357985235595e-06, "loss": 0.289, "step": 84 }, { "epoch": 0.27309236947791166, "grad_norm": 2.321336030960083, "learning_rate": 8.544747115029717e-06, "loss": 0.3508, "step": 85 }, { "epoch": 0.27630522088353415, "grad_norm": 3.0077357292175293, "learning_rate": 8.50775010191001e-06, "loss": 0.38, "step": 86 }, { "epoch": 0.27951807228915665, "grad_norm": 2.451876640319824, "learning_rate": 8.470370976103171e-06, "loss": 0.3444, "step": 87 }, { "epoch": 0.2827309236947791, "grad_norm": 2.5099685192108154, "learning_rate": 8.432613809460895e-06, "loss": 0.5357, "step": 88 }, { "epoch": 0.2859437751004016, "grad_norm": 2.212970018386841, "learning_rate": 8.394482715016318e-06, "loss": 0.361, "step": 89 }, { "epoch": 0.2891566265060241, "grad_norm": 2.779324531555176, "learning_rate": 8.355981846535972e-06, "loss": 0.5175, "step": 90 }, { "epoch": 0.2923694779116466, "grad_norm": 2.486966133117676, "learning_rate": 8.317115398067289e-06, "loss": 0.4769, "step": 91 }, { "epoch": 0.2955823293172691, "grad_norm": 2.3630478382110596, "learning_rate": 8.27788760348173e-06, "loss": 0.3273, "step": 92 }, { "epoch": 0.2987951807228916, "grad_norm": 3.0468597412109375, "learning_rate": 8.238302736013587e-06, "loss": 0.465, "step": 93 }, { "epoch": 0.30200803212851407, "grad_norm": 2.1398801803588867, "learning_rate": 8.198365107794457e-06, "loss": 0.5196, "step": 94 }, { "epoch": 0.30522088353413657, "grad_norm": 2.3869407176971436, "learning_rate": 8.158079069383535e-06, "loss": 0.3714, "step": 95 }, { "epoch": 0.30843373493975906, "grad_norm": 2.6658105850219727, "learning_rate": 8.117449009293668e-06, "loss": 0.4704, "step": 96 }, { "epoch": 0.3116465863453815, "grad_norm": 3.014535665512085, "learning_rate": 8.076479353513308e-06, "loss": 0.4581, "step": 97 }, { "epoch": 0.314859437751004, "grad_norm": 2.575618028640747, "learning_rate": 8.035174565024362e-06, "loss": 0.4222, "step": 98 }, { "epoch": 0.3180722891566265, "grad_norm": 2.499056577682495, "learning_rate": 7.993539143316044e-06, "loss": 0.3373, "step": 99 }, { "epoch": 0.321285140562249, "grad_norm": 1.9054265022277832, "learning_rate": 7.951577623894701e-06, "loss": 0.4234, "step": 100 }, { "epoch": 0.3244979919678715, "grad_norm": 2.496711015701294, "learning_rate": 7.909294577789765e-06, "loss": 0.3642, "step": 101 }, { "epoch": 0.327710843373494, "grad_norm": 2.4711644649505615, "learning_rate": 7.866694611055796e-06, "loss": 0.3368, "step": 102 }, { "epoch": 0.3309236947791165, "grad_norm": 2.246197462081909, "learning_rate": 7.823782364270743e-06, "loss": 0.3013, "step": 103 }, { "epoch": 0.334136546184739, "grad_norm": 2.3415684700012207, "learning_rate": 7.780562512030414e-06, "loss": 0.4313, "step": 104 }, { "epoch": 0.3373493975903614, "grad_norm": 2.7522289752960205, "learning_rate": 7.737039762439263e-06, "loss": 0.5178, "step": 105 }, { "epoch": 0.3405622489959839, "grad_norm": 2.612412452697754, "learning_rate": 7.693218856597515e-06, "loss": 0.4683, "step": 106 }, { "epoch": 0.3437751004016064, "grad_norm": 2.491758108139038, "learning_rate": 7.649104568084701e-06, "loss": 0.4714, "step": 107 }, { "epoch": 0.3469879518072289, "grad_norm": 2.803365707397461, "learning_rate": 7.604701702439652e-06, "loss": 0.4583, "step": 108 }, { "epoch": 0.3502008032128514, "grad_norm": 2.161888360977173, "learning_rate": 7.560015096637015e-06, "loss": 0.3207, "step": 109 }, { "epoch": 0.3534136546184739, "grad_norm": 2.9406943321228027, "learning_rate": 7.515049618560337e-06, "loss": 0.4199, "step": 110 }, { "epoch": 0.3566265060240964, "grad_norm": 2.4751358032226562, "learning_rate": 7.469810166471802e-06, "loss": 0.3597, "step": 111 }, { "epoch": 0.3598393574297189, "grad_norm": 2.4176955223083496, "learning_rate": 7.424301668478626e-06, "loss": 0.4643, "step": 112 }, { "epoch": 0.36305220883534134, "grad_norm": 1.987687110900879, "learning_rate": 7.378529081996233e-06, "loss": 0.3283, "step": 113 }, { "epoch": 0.36626506024096384, "grad_norm": 2.1835029125213623, "learning_rate": 7.332497393208221e-06, "loss": 0.3521, "step": 114 }, { "epoch": 0.36947791164658633, "grad_norm": 2.5302932262420654, "learning_rate": 7.286211616523193e-06, "loss": 0.397, "step": 115 }, { "epoch": 0.37269076305220883, "grad_norm": 2.3871219158172607, "learning_rate": 7.239676794028526e-06, "loss": 0.4685, "step": 116 }, { "epoch": 0.3759036144578313, "grad_norm": 2.5505502223968506, "learning_rate": 7.192897994941111e-06, "loss": 0.4473, "step": 117 }, { "epoch": 0.3791164658634538, "grad_norm": 2.5716140270233154, "learning_rate": 7.145880315055145e-06, "loss": 0.2596, "step": 118 }, { "epoch": 0.3823293172690763, "grad_norm": 2.9478588104248047, "learning_rate": 7.098628876187031e-06, "loss": 0.3903, "step": 119 }, { "epoch": 0.3855421686746988, "grad_norm": 2.7252838611602783, "learning_rate": 7.051148825617435e-06, "loss": 0.3011, "step": 120 }, { "epoch": 0.3887550200803213, "grad_norm": 1.9095062017440796, "learning_rate": 7.003445335530572e-06, "loss": 0.242, "step": 121 }, { "epoch": 0.39196787148594375, "grad_norm": 2.5876998901367188, "learning_rate": 6.95552360245078e-06, "loss": 0.4669, "step": 122 }, { "epoch": 0.39518072289156625, "grad_norm": 2.438539505004883, "learning_rate": 6.9073888466764495e-06, "loss": 0.2939, "step": 123 }, { "epoch": 0.39839357429718875, "grad_norm": 2.326950788497925, "learning_rate": 6.859046311711344e-06, "loss": 0.3136, "step": 124 }, { "epoch": 0.40160642570281124, "grad_norm": 2.442441463470459, "learning_rate": 6.810501263693416e-06, "loss": 0.241, "step": 125 }, { "epoch": 0.40481927710843374, "grad_norm": 2.1230807304382324, "learning_rate": 6.761758990821143e-06, "loss": 0.341, "step": 126 }, { "epoch": 0.40803212851405624, "grad_norm": 2.29844069480896, "learning_rate": 6.712824802777465e-06, "loss": 0.2936, "step": 127 }, { "epoch": 0.41124497991967873, "grad_norm": 2.681148052215576, "learning_rate": 6.66370403015137e-06, "loss": 0.271, "step": 128 }, { "epoch": 0.41445783132530123, "grad_norm": 2.411924123764038, "learning_rate": 6.614402023857231e-06, "loss": 0.4166, "step": 129 }, { "epoch": 0.41767068273092367, "grad_norm": 2.795295238494873, "learning_rate": 6.564924154551895e-06, "loss": 0.3683, "step": 130 }, { "epoch": 0.42088353413654617, "grad_norm": 2.0868473052978516, "learning_rate": 6.515275812049644e-06, "loss": 0.2563, "step": 131 }, { "epoch": 0.42409638554216866, "grad_norm": 2.622237205505371, "learning_rate": 6.4654624047350575e-06, "loss": 0.3725, "step": 132 }, { "epoch": 0.42730923694779116, "grad_norm": 2.958596706390381, "learning_rate": 6.41548935897386e-06, "loss": 0.2818, "step": 133 }, { "epoch": 0.43052208835341366, "grad_norm": 2.81485915184021, "learning_rate": 6.365362118521807e-06, "loss": 0.3887, "step": 134 }, { "epoch": 0.43373493975903615, "grad_norm": 2.5471572875976562, "learning_rate": 6.31508614393167e-06, "loss": 0.3455, "step": 135 }, { "epoch": 0.43694779116465865, "grad_norm": 5.213260173797607, "learning_rate": 6.264666911958404e-06, "loss": 0.3154, "step": 136 }, { "epoch": 0.44016064257028115, "grad_norm": 2.910991668701172, "learning_rate": 6.214109914962542e-06, "loss": 0.3963, "step": 137 }, { "epoch": 0.4433734939759036, "grad_norm": 2.244249105453491, "learning_rate": 6.1634206603118844e-06, "loss": 0.2974, "step": 138 }, { "epoch": 0.4465863453815261, "grad_norm": 2.3065643310546875, "learning_rate": 6.112604669781572e-06, "loss": 0.344, "step": 139 }, { "epoch": 0.4497991967871486, "grad_norm": 2.094327211380005, "learning_rate": 6.06166747895257e-06, "loss": 0.3327, "step": 140 }, { "epoch": 0.4530120481927711, "grad_norm": 2.5069355964660645, "learning_rate": 6.0106146366086514e-06, "loss": 0.2542, "step": 141 }, { "epoch": 0.4562248995983936, "grad_norm": 2.227370262145996, "learning_rate": 5.959451704131962e-06, "loss": 0.3063, "step": 142 }, { "epoch": 0.45943775100401607, "grad_norm": 3.291332244873047, "learning_rate": 5.908184254897183e-06, "loss": 0.3719, "step": 143 }, { "epoch": 0.46265060240963857, "grad_norm": 2.1085941791534424, "learning_rate": 5.856817873664409e-06, "loss": 0.3306, "step": 144 }, { "epoch": 0.46586345381526106, "grad_norm": 2.1606924533843994, "learning_rate": 5.8053581559707754e-06, "loss": 0.3847, "step": 145 }, { "epoch": 0.46907630522088356, "grad_norm": 3.6136634349823, "learning_rate": 5.753810707520918e-06, "loss": 0.409, "step": 146 }, { "epoch": 0.472289156626506, "grad_norm": 2.9589033126831055, "learning_rate": 5.702181143576323e-06, "loss": 0.519, "step": 147 }, { "epoch": 0.4755020080321285, "grad_norm": 2.941072463989258, "learning_rate": 5.6504750883436275e-06, "loss": 0.3586, "step": 148 }, { "epoch": 0.478714859437751, "grad_norm": 2.1762218475341797, "learning_rate": 5.5986981743619615e-06, "loss": 0.3692, "step": 149 }, { "epoch": 0.4819277108433735, "grad_norm": 1.9385762214660645, "learning_rate": 5.546856041889374e-06, "loss": 0.3044, "step": 150 }, { "epoch": 0.485140562248996, "grad_norm": 2.7357289791107178, "learning_rate": 5.494954338288404e-06, "loss": 0.4027, "step": 151 }, { "epoch": 0.4883534136546185, "grad_norm": 2.4241161346435547, "learning_rate": 5.442998717410916e-06, "loss": 0.3591, "step": 152 }, { "epoch": 0.491566265060241, "grad_norm": 5.211270809173584, "learning_rate": 5.390994838982178e-06, "loss": 0.3662, "step": 153 }, { "epoch": 0.4947791164658635, "grad_norm": 2.5617518424987793, "learning_rate": 5.338948367984347e-06, "loss": 0.3067, "step": 154 }, { "epoch": 0.4979919678714859, "grad_norm": 2.434497356414795, "learning_rate": 5.286864974039349e-06, "loss": 0.3023, "step": 155 }, { "epoch": 0.5012048192771085, "grad_norm": 2.5631253719329834, "learning_rate": 5.234750330791268e-06, "loss": 0.3823, "step": 156 }, { "epoch": 0.5044176706827309, "grad_norm": 2.003878593444824, "learning_rate": 5.182610115288296e-06, "loss": 0.361, "step": 157 }, { "epoch": 0.5076305220883535, "grad_norm": 2.133862018585205, "learning_rate": 5.1304500073643045e-06, "loss": 0.2718, "step": 158 }, { "epoch": 0.5108433734939759, "grad_norm": 2.344472646713257, "learning_rate": 5.078275689020129e-06, "loss": 0.2921, "step": 159 }, { "epoch": 0.5140562248995983, "grad_norm": 2.4135284423828125, "learning_rate": 5.026092843804599e-06, "loss": 0.3341, "step": 160 }, { "epoch": 0.5172690763052209, "grad_norm": 2.5881218910217285, "learning_rate": 4.973907156195405e-06, "loss": 0.2986, "step": 161 }, { "epoch": 0.5204819277108433, "grad_norm": 2.7781827449798584, "learning_rate": 4.921724310979872e-06, "loss": 0.4795, "step": 162 }, { "epoch": 0.5236947791164659, "grad_norm": 2.816742181777954, "learning_rate": 4.869549992635697e-06, "loss": 0.4024, "step": 163 }, { "epoch": 0.5269076305220883, "grad_norm": 2.479278326034546, "learning_rate": 4.817389884711706e-06, "loss": 0.381, "step": 164 }, { "epoch": 0.5301204819277109, "grad_norm": 2.0837013721466064, "learning_rate": 4.765249669208733e-06, "loss": 0.2775, "step": 165 }, { "epoch": 0.5333333333333333, "grad_norm": 3.396789312362671, "learning_rate": 4.713135025960652e-06, "loss": 0.3469, "step": 166 }, { "epoch": 0.5365461847389559, "grad_norm": 2.179645299911499, "learning_rate": 4.661051632015655e-06, "loss": 0.375, "step": 167 }, { "epoch": 0.5397590361445783, "grad_norm": 2.4666926860809326, "learning_rate": 4.609005161017824e-06, "loss": 0.3723, "step": 168 }, { "epoch": 0.5429718875502008, "grad_norm": 2.0646708011627197, "learning_rate": 4.557001282589086e-06, "loss": 0.3036, "step": 169 }, { "epoch": 0.5461847389558233, "grad_norm": 2.0979270935058594, "learning_rate": 4.505045661711596e-06, "loss": 0.312, "step": 170 }, { "epoch": 0.5493975903614458, "grad_norm": 2.0291059017181396, "learning_rate": 4.4531439581106295e-06, "loss": 0.3151, "step": 171 }, { "epoch": 0.5526104417670683, "grad_norm": 1.8637864589691162, "learning_rate": 4.401301825638039e-06, "loss": 0.2561, "step": 172 }, { "epoch": 0.5558232931726907, "grad_norm": 2.0765888690948486, "learning_rate": 4.349524911656373e-06, "loss": 0.3333, "step": 173 }, { "epoch": 0.5590361445783133, "grad_norm": 2.5530149936676025, "learning_rate": 4.297818856423679e-06, "loss": 0.3819, "step": 174 }, { "epoch": 0.5622489959839357, "grad_norm": 2.2014477252960205, "learning_rate": 4.2461892924790825e-06, "loss": 0.254, "step": 175 }, { "epoch": 0.5654618473895582, "grad_norm": 3.6043734550476074, "learning_rate": 4.194641844029227e-06, "loss": 0.4335, "step": 176 }, { "epoch": 0.5686746987951807, "grad_norm": 2.4686837196350098, "learning_rate": 4.143182126335594e-06, "loss": 0.2313, "step": 177 }, { "epoch": 0.5718875502008032, "grad_norm": 1.9996862411499023, "learning_rate": 4.091815745102818e-06, "loss": 0.296, "step": 178 }, { "epoch": 0.5751004016064257, "grad_norm": 2.37868595123291, "learning_rate": 4.040548295868039e-06, "loss": 0.2864, "step": 179 }, { "epoch": 0.5783132530120482, "grad_norm": 2.2569093704223633, "learning_rate": 3.9893853633913485e-06, "loss": 0.2444, "step": 180 }, { "epoch": 0.5815261044176707, "grad_norm": 3.60836124420166, "learning_rate": 3.938332521047434e-06, "loss": 0.3296, "step": 181 }, { "epoch": 0.5847389558232932, "grad_norm": 2.706819772720337, "learning_rate": 3.887395330218429e-06, "loss": 0.1887, "step": 182 }, { "epoch": 0.5879518072289157, "grad_norm": 2.409383535385132, "learning_rate": 3.836579339688116e-06, "loss": 0.29, "step": 183 }, { "epoch": 0.5911646586345382, "grad_norm": 2.9502406120300293, "learning_rate": 3.7858900850374596e-06, "loss": 0.3951, "step": 184 }, { "epoch": 0.5943775100401606, "grad_norm": 2.052128314971924, "learning_rate": 3.7353330880415963e-06, "loss": 0.3078, "step": 185 }, { "epoch": 0.5975903614457831, "grad_norm": 2.404327630996704, "learning_rate": 3.6849138560683305e-06, "loss": 0.4112, "step": 186 }, { "epoch": 0.6008032128514056, "grad_norm": 2.3815231323242188, "learning_rate": 3.634637881478196e-06, "loss": 0.322, "step": 187 }, { "epoch": 0.6040160642570281, "grad_norm": 2.2108402252197266, "learning_rate": 3.5845106410261417e-06, "loss": 0.3744, "step": 188 }, { "epoch": 0.6072289156626506, "grad_norm": 2.1574432849884033, "learning_rate": 3.534537595264944e-06, "loss": 0.2951, "step": 189 }, { "epoch": 0.6104417670682731, "grad_norm": 2.283750057220459, "learning_rate": 3.4847241879503574e-06, "loss": 0.2728, "step": 190 }, { "epoch": 0.6136546184738956, "grad_norm": 2.0512266159057617, "learning_rate": 3.435075845448105e-06, "loss": 0.2585, "step": 191 }, { "epoch": 0.6168674698795181, "grad_norm": 2.368990182876587, "learning_rate": 3.3855979761427705e-06, "loss": 0.4226, "step": 192 }, { "epoch": 0.6200803212851406, "grad_norm": 2.43058705329895, "learning_rate": 3.3362959698486307e-06, "loss": 0.3779, "step": 193 }, { "epoch": 0.623293172690763, "grad_norm": 2.8462181091308594, "learning_rate": 3.287175197222537e-06, "loss": 0.4504, "step": 194 }, { "epoch": 0.6265060240963856, "grad_norm": 2.358142614364624, "learning_rate": 3.2382410091788567e-06, "loss": 0.3824, "step": 195 }, { "epoch": 0.629718875502008, "grad_norm": 1.9749348163604736, "learning_rate": 3.189498736306584e-06, "loss": 0.3309, "step": 196 }, { "epoch": 0.6329317269076306, "grad_norm": 2.166022300720215, "learning_rate": 3.140953688288658e-06, "loss": 0.1816, "step": 197 }, { "epoch": 0.636144578313253, "grad_norm": 2.078129291534424, "learning_rate": 3.0926111533235526e-06, "loss": 0.3417, "step": 198 }, { "epoch": 0.6393574297188755, "grad_norm": 2.0340514183044434, "learning_rate": 3.044476397549221e-06, "loss": 0.2881, "step": 199 }, { "epoch": 0.642570281124498, "grad_norm": 2.3322885036468506, "learning_rate": 2.9965546644694287e-06, "loss": 0.2418, "step": 200 }, { "epoch": 0.6457831325301204, "grad_norm": 2.1144306659698486, "learning_rate": 2.948851174382565e-06, "loss": 0.2758, "step": 201 }, { "epoch": 0.648995983935743, "grad_norm": 2.089343309402466, "learning_rate": 2.9013711238129693e-06, "loss": 0.3059, "step": 202 }, { "epoch": 0.6522088353413654, "grad_norm": 2.4294567108154297, "learning_rate": 2.8541196849448582e-06, "loss": 0.3653, "step": 203 }, { "epoch": 0.655421686746988, "grad_norm": 2.432567596435547, "learning_rate": 2.8071020050588927e-06, "loss": 0.2259, "step": 204 }, { "epoch": 0.6586345381526104, "grad_norm": 2.2744345664978027, "learning_rate": 2.760323205971476e-06, "loss": 0.3401, "step": 205 }, { "epoch": 0.661847389558233, "grad_norm": 2.8238182067871094, "learning_rate": 2.7137883834768076e-06, "loss": 0.2912, "step": 206 }, { "epoch": 0.6650602409638554, "grad_norm": 2.3627312183380127, "learning_rate": 2.6675026067917808e-06, "loss": 0.2491, "step": 207 }, { "epoch": 0.668273092369478, "grad_norm": 2.5185694694519043, "learning_rate": 2.621470918003768e-06, "loss": 0.2879, "step": 208 }, { "epoch": 0.6714859437751004, "grad_norm": 2.401811361312866, "learning_rate": 2.5756983315213748e-06, "loss": 0.3885, "step": 209 }, { "epoch": 0.6746987951807228, "grad_norm": 2.2231993675231934, "learning_rate": 2.5301898335281994e-06, "loss": 0.2833, "step": 210 }, { "epoch": 0.6779116465863454, "grad_norm": 2.2760813236236572, "learning_rate": 2.4849503814396624e-06, "loss": 0.2532, "step": 211 }, { "epoch": 0.6811244979919678, "grad_norm": 2.149466037750244, "learning_rate": 2.439984903362988e-06, "loss": 0.2805, "step": 212 }, { "epoch": 0.6843373493975904, "grad_norm": 2.3873984813690186, "learning_rate": 2.3952982975603494e-06, "loss": 0.3158, "step": 213 }, { "epoch": 0.6875502008032128, "grad_norm": 2.5702414512634277, "learning_rate": 2.3508954319153e-06, "loss": 0.3221, "step": 214 }, { "epoch": 0.6907630522088354, "grad_norm": 2.643507719039917, "learning_rate": 2.306781143402485e-06, "loss": 0.3544, "step": 215 }, { "epoch": 0.6939759036144578, "grad_norm": 2.201829671859741, "learning_rate": 2.2629602375607373e-06, "loss": 0.3529, "step": 216 }, { "epoch": 0.6971887550200804, "grad_norm": 2.1155354976654053, "learning_rate": 2.219437487969588e-06, "loss": 0.2572, "step": 217 }, { "epoch": 0.7004016064257028, "grad_norm": 2.038470506668091, "learning_rate": 2.1762176357292582e-06, "loss": 0.2295, "step": 218 }, { "epoch": 0.7036144578313253, "grad_norm": 2.023308753967285, "learning_rate": 2.1333053889442033e-06, "loss": 0.3592, "step": 219 }, { "epoch": 0.7068273092369478, "grad_norm": 2.5169191360473633, "learning_rate": 2.0907054222102367e-06, "loss": 0.3218, "step": 220 }, { "epoch": 0.7100401606425703, "grad_norm": 2.5444135665893555, "learning_rate": 2.048422376105299e-06, "loss": 0.3331, "step": 221 }, { "epoch": 0.7132530120481928, "grad_norm": 2.6340672969818115, "learning_rate": 2.0064608566839584e-06, "loss": 0.4002, "step": 222 }, { "epoch": 0.7164658634538152, "grad_norm": 2.0738048553466797, "learning_rate": 1.964825434975639e-06, "loss": 0.2908, "step": 223 }, { "epoch": 0.7196787148594378, "grad_norm": 2.0317463874816895, "learning_rate": 1.923520646486695e-06, "loss": 0.2293, "step": 224 }, { "epoch": 0.7228915662650602, "grad_norm": 2.2780821323394775, "learning_rate": 1.8825509907063328e-06, "loss": 0.2623, "step": 225 }, { "epoch": 0.7261044176706827, "grad_norm": 2.3251006603240967, "learning_rate": 1.8419209306164653e-06, "loss": 0.2582, "step": 226 }, { "epoch": 0.7293172690763052, "grad_norm": 2.357585906982422, "learning_rate": 1.8016348922055448e-06, "loss": 0.2822, "step": 227 }, { "epoch": 0.7325301204819277, "grad_norm": 2.6229782104492188, "learning_rate": 1.7616972639864166e-06, "loss": 0.3912, "step": 228 }, { "epoch": 0.7357429718875502, "grad_norm": 2.2909278869628906, "learning_rate": 1.7221123965182712e-06, "loss": 0.3465, "step": 229 }, { "epoch": 0.7389558232931727, "grad_norm": 1.9987268447875977, "learning_rate": 1.6828846019327128e-06, "loss": 0.3453, "step": 230 }, { "epoch": 0.7421686746987952, "grad_norm": 2.4425132274627686, "learning_rate": 1.6440181534640277e-06, "loss": 0.2378, "step": 231 }, { "epoch": 0.7453815261044177, "grad_norm": 1.847144365310669, "learning_rate": 1.6055172849836826e-06, "loss": 0.2161, "step": 232 }, { "epoch": 0.7485943775100402, "grad_norm": 2.3810346126556396, "learning_rate": 1.567386190539107e-06, "loss": 0.2798, "step": 233 }, { "epoch": 0.7518072289156627, "grad_norm": 1.8688887357711792, "learning_rate": 1.5296290238968303e-06, "loss": 0.2168, "step": 234 }, { "epoch": 0.7550200803212851, "grad_norm": 3.1742520332336426, "learning_rate": 1.4922498980899907e-06, "loss": 0.3661, "step": 235 }, { "epoch": 0.7582329317269076, "grad_norm": 2.355210542678833, "learning_rate": 1.4552528849702852e-06, "loss": 0.2874, "step": 236 }, { "epoch": 0.7614457831325301, "grad_norm": 3.493511915206909, "learning_rate": 1.4186420147644053e-06, "loss": 0.3802, "step": 237 }, { "epoch": 0.7646586345381526, "grad_norm": 1.8001739978790283, "learning_rate": 1.3824212756350196e-06, "loss": 0.2334, "step": 238 }, { "epoch": 0.7678714859437751, "grad_norm": 1.9317985773086548, "learning_rate": 1.3465946132463125e-06, "loss": 0.245, "step": 239 }, { "epoch": 0.7710843373493976, "grad_norm": 2.0644335746765137, "learning_rate": 1.3111659303341824e-06, "loss": 0.243, "step": 240 }, { "epoch": 0.7742971887550201, "grad_norm": 2.091796398162842, "learning_rate": 1.2761390862810907e-06, "loss": 0.2382, "step": 241 }, { "epoch": 0.7775100401606426, "grad_norm": 2.3172249794006348, "learning_rate": 1.2415178966956531e-06, "loss": 0.2797, "step": 242 }, { "epoch": 0.7807228915662651, "grad_norm": 2.269488573074341, "learning_rate": 1.2073061329969843e-06, "loss": 0.3524, "step": 243 }, { "epoch": 0.7839357429718875, "grad_norm": 2.3127260208129883, "learning_rate": 1.1735075220038634e-06, "loss": 0.2378, "step": 244 }, { "epoch": 0.7871485943775101, "grad_norm": 1.958678126335144, "learning_rate": 1.1401257455287612e-06, "loss": 0.2539, "step": 245 }, { "epoch": 0.7903614457831325, "grad_norm": 2.530526876449585, "learning_rate": 1.107164439976764e-06, "loss": 0.2905, "step": 246 }, { "epoch": 0.793574297188755, "grad_norm": 2.522650957107544, "learning_rate": 1.0746271959494453e-06, "loss": 0.3653, "step": 247 }, { "epoch": 0.7967871485943775, "grad_norm": 2.9361941814422607, "learning_rate": 1.04251755785373e-06, "loss": 0.3054, "step": 248 }, { "epoch": 0.8, "grad_norm": 2.346240758895874, "learning_rate": 1.0108390235157828e-06, "loss": 0.2623, "step": 249 }, { "epoch": 0.8032128514056225, "grad_norm": 2.452631711959839, "learning_rate": 9.795950437999852e-07, "loss": 0.2113, "step": 250 }, { "epoch": 0.8064257028112449, "grad_norm": 2.9312245845794678, "learning_rate": 9.487890222330137e-07, "loss": 0.4234, "step": 251 }, { "epoch": 0.8096385542168675, "grad_norm": 2.1248555183410645, "learning_rate": 9.184243146330829e-07, "loss": 0.2872, "step": 252 }, { "epoch": 0.8128514056224899, "grad_norm": 2.393862247467041, "learning_rate": 8.885042287443785e-07, "loss": 0.2849, "step": 253 }, { "epoch": 0.8160642570281125, "grad_norm": 2.453737258911133, "learning_rate": 8.590320238767425e-07, "loss": 0.2962, "step": 254 }, { "epoch": 0.8192771084337349, "grad_norm": 2.9770538806915283, "learning_rate": 8.30010910550611e-07, "loss": 0.3318, "step": 255 }, { "epoch": 0.8224899598393575, "grad_norm": 1.8912490606307983, "learning_rate": 8.014440501472909e-07, "loss": 0.2446, "step": 256 }, { "epoch": 0.8257028112449799, "grad_norm": 3.3345978260040283, "learning_rate": 7.733345545645726e-07, "loss": 0.3895, "step": 257 }, { "epoch": 0.8289156626506025, "grad_norm": 2.2788500785827637, "learning_rate": 7.456854858777418e-07, "loss": 0.2767, "step": 258 }, { "epoch": 0.8321285140562249, "grad_norm": 2.039923906326294, "learning_rate": 7.184998560060114e-07, "loss": 0.1987, "step": 259 }, { "epoch": 0.8353413654618473, "grad_norm": 2.3319668769836426, "learning_rate": 6.917806263844268e-07, "loss": 0.2265, "step": 260 }, { "epoch": 0.8385542168674699, "grad_norm": 3.085174560546875, "learning_rate": 6.655307076412637e-07, "loss": 0.3313, "step": 261 }, { "epoch": 0.8417670682730923, "grad_norm": 1.9811396598815918, "learning_rate": 6.397529592809615e-07, "loss": 0.2447, "step": 262 }, { "epoch": 0.8449799196787149, "grad_norm": 1.8202673196792603, "learning_rate": 6.14450189372628e-07, "loss": 0.2427, "step": 263 }, { "epoch": 0.8481927710843373, "grad_norm": 2.299744129180908, "learning_rate": 5.896251542441395e-07, "loss": 0.2412, "step": 264 }, { "epoch": 0.8514056224899599, "grad_norm": 2.096524953842163, "learning_rate": 5.652805581818943e-07, "loss": 0.2329, "step": 265 }, { "epoch": 0.8546184738955823, "grad_norm": 2.4148600101470947, "learning_rate": 5.414190531362162e-07, "loss": 0.2183, "step": 266 }, { "epoch": 0.8578313253012049, "grad_norm": 1.961333155632019, "learning_rate": 5.180432384324691e-07, "loss": 0.2278, "step": 267 }, { "epoch": 0.8610441767068273, "grad_norm": 1.9496490955352783, "learning_rate": 4.951556604879049e-07, "loss": 0.1662, "step": 268 }, { "epoch": 0.8642570281124498, "grad_norm": 2.195952892303467, "learning_rate": 4.727588125342669e-07, "loss": 0.2645, "step": 269 }, { "epoch": 0.8674698795180723, "grad_norm": 1.8171138763427734, "learning_rate": 4.508551343462014e-07, "loss": 0.1446, "step": 270 }, { "epoch": 0.8706827309236947, "grad_norm": 2.3251218795776367, "learning_rate": 4.29447011975474e-07, "loss": 0.2679, "step": 271 }, { "epoch": 0.8738955823293173, "grad_norm": 2.06396746635437, "learning_rate": 4.0853677749105426e-07, "loss": 0.2793, "step": 272 }, { "epoch": 0.8771084337349397, "grad_norm": 2.0605921745300293, "learning_rate": 3.8812670872507454e-07, "loss": 0.2798, "step": 273 }, { "epoch": 0.8803212851405623, "grad_norm": 2.0419671535491943, "learning_rate": 3.6821902902469066e-07, "loss": 0.3494, "step": 274 }, { "epoch": 0.8835341365461847, "grad_norm": 1.9494353532791138, "learning_rate": 3.4881590700989175e-07, "loss": 0.3164, "step": 275 }, { "epoch": 0.8867469879518072, "grad_norm": 2.2515294551849365, "learning_rate": 3.299194563372604e-07, "loss": 0.2364, "step": 276 }, { "epoch": 0.8899598393574297, "grad_norm": 2.4932332038879395, "learning_rate": 3.1153173546972395e-07, "loss": 0.1982, "step": 277 }, { "epoch": 0.8931726907630522, "grad_norm": 1.9155004024505615, "learning_rate": 2.9365474745231935e-07, "loss": 0.2864, "step": 278 }, { "epoch": 0.8963855421686747, "grad_norm": 2.346914529800415, "learning_rate": 2.7629043969399193e-07, "loss": 0.3323, "step": 279 }, { "epoch": 0.8995983935742972, "grad_norm": 1.957244634628296, "learning_rate": 2.594407037554586e-07, "loss": 0.1917, "step": 280 }, { "epoch": 0.9028112449799197, "grad_norm": 2.4275577068328857, "learning_rate": 2.431073751431529e-07, "loss": 0.3431, "step": 281 }, { "epoch": 0.9060240963855422, "grad_norm": 2.3854384422302246, "learning_rate": 2.2729223310927473e-07, "loss": 0.1988, "step": 282 }, { "epoch": 0.9092369477911647, "grad_norm": 2.0022900104522705, "learning_rate": 2.1199700045797077e-07, "loss": 0.2806, "step": 283 }, { "epoch": 0.9124497991967871, "grad_norm": 2.0023062229156494, "learning_rate": 1.9722334335766092e-07, "loss": 0.3024, "step": 284 }, { "epoch": 0.9156626506024096, "grad_norm": 1.8861989974975586, "learning_rate": 1.829728711595391e-07, "loss": 0.2559, "step": 285 }, { "epoch": 0.9188755020080321, "grad_norm": 2.1966798305511475, "learning_rate": 1.6924713622225975e-07, "loss": 0.2724, "step": 286 }, { "epoch": 0.9220883534136546, "grad_norm": 1.6553095579147339, "learning_rate": 1.5604763374283073e-07, "loss": 0.1017, "step": 287 }, { "epoch": 0.9253012048192771, "grad_norm": 2.2511794567108154, "learning_rate": 1.4337580159373864e-07, "loss": 0.2722, "step": 288 }, { "epoch": 0.9285140562248996, "grad_norm": 2.169060230255127, "learning_rate": 1.3123302016631477e-07, "loss": 0.265, "step": 289 }, { "epoch": 0.9317269076305221, "grad_norm": 2.9961092472076416, "learning_rate": 1.196206122203647e-07, "loss": 0.3931, "step": 290 }, { "epoch": 0.9349397590361446, "grad_norm": 2.435908794403076, "learning_rate": 1.0853984274007246e-07, "loss": 0.2906, "step": 291 }, { "epoch": 0.9381526104417671, "grad_norm": 2.8427674770355225, "learning_rate": 9.799191879620474e-08, "loss": 0.2227, "step": 292 }, { "epoch": 0.9413654618473896, "grad_norm": 2.3079280853271484, "learning_rate": 8.797798941461655e-08, "loss": 0.2599, "step": 293 }, { "epoch": 0.944578313253012, "grad_norm": 2.447566032409668, "learning_rate": 7.84991454510864e-08, "loss": 0.25, "step": 294 }, { "epoch": 0.9477911646586346, "grad_norm": 2.5724740028381348, "learning_rate": 6.955641947248127e-08, "loss": 0.236, "step": 295 }, { "epoch": 0.951004016064257, "grad_norm": 2.4259979724884033, "learning_rate": 6.115078564427946e-08, "loss": 0.2004, "step": 296 }, { "epoch": 0.9542168674698795, "grad_norm": 1.9986400604248047, "learning_rate": 5.3283159624448745e-08, "loss": 0.2437, "step": 297 }, { "epoch": 0.957429718875502, "grad_norm": 2.7118661403656006, "learning_rate": 4.5954398463700647e-08, "loss": 0.4067, "step": 298 }, { "epoch": 0.9606425702811245, "grad_norm": 2.030395984649658, "learning_rate": 3.916530051212841e-08, "loss": 0.2347, "step": 299 }, { "epoch": 0.963855421686747, "grad_norm": 1.5845026969909668, "learning_rate": 3.2916605332238284e-08, "loss": 0.153, "step": 300 }, { "epoch": 0.9670682730923694, "grad_norm": 2.0269935131073, "learning_rate": 2.7208993618390578e-08, "loss": 0.1627, "step": 301 }, { "epoch": 0.970281124497992, "grad_norm": 1.7263789176940918, "learning_rate": 2.2043087122644023e-08, "loss": 0.1601, "step": 302 }, { "epoch": 0.9734939759036144, "grad_norm": 1.7540363073349, "learning_rate": 1.741944858702771e-08, "loss": 0.2004, "step": 303 }, { "epoch": 0.976706827309237, "grad_norm": 2.1071643829345703, "learning_rate": 1.333858168224178e-08, "loss": 0.2089, "step": 304 }, { "epoch": 0.9799196787148594, "grad_norm": 2.682478904724121, "learning_rate": 9.800930952786336e-09, "loss": 0.3269, "step": 305 }, { "epoch": 0.983132530120482, "grad_norm": 2.5605387687683105, "learning_rate": 6.806881768539053e-09, "loss": 0.2146, "step": 306 }, { "epoch": 0.9863453815261044, "grad_norm": 1.8436514139175415, "learning_rate": 4.356760282773209e-09, "loss": 0.2533, "step": 307 }, { "epoch": 0.989558232931727, "grad_norm": 2.426079034805298, "learning_rate": 2.4508333966305473e-09, "loss": 0.3066, "step": 308 }, { "epoch": 0.9927710843373494, "grad_norm": 2.5514280796051025, "learning_rate": 1.089308730043981e-09, "loss": 0.2522, "step": 309 }, { "epoch": 0.9959839357429718, "grad_norm": 3.108274221420288, "learning_rate": 2.723345991245685e-10, "loss": 0.3181, "step": 310 }, { "epoch": 0.9991967871485944, "grad_norm": 1.9873707294464111, "learning_rate": 0.0, "loss": 0.2276, "step": 311 }, { "epoch": 0.9991967871485944, "step": 311, "total_flos": 9.654917969372774e+16, "train_loss": 0.38514137028497897, "train_runtime": 6718.4685, "train_samples_per_second": 1.483, "train_steps_per_second": 0.046 } ], "logging_steps": 1.0, "max_steps": 311, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 9.654917969372774e+16, "train_batch_size": 1, "trial_name": null, "trial_params": null }