diff --git "a/checkpoint-1400/trainer_state.json" "b/checkpoint-1400/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1400/trainer_state.json" @@ -0,0 +1,9833 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.4047412546978896, + "eval_steps": 500, + "global_step": 1400, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0002891008962127783, + "grad_norm": 10.526118368353698, + "learning_rate": 5.714285714285715e-07, + "loss": 1.8059, + "step": 1 + }, + { + "epoch": 0.0005782017924255566, + "grad_norm": 10.546832449734852, + "learning_rate": 1.142857142857143e-06, + "loss": 1.8355, + "step": 2 + }, + { + "epoch": 0.0008673026886383347, + "grad_norm": 9.678077753237705, + "learning_rate": 1.7142857142857145e-06, + "loss": 1.8885, + "step": 3 + }, + { + "epoch": 0.0011564035848511131, + "grad_norm": 7.133178553946601, + "learning_rate": 2.285714285714286e-06, + "loss": 1.9111, + "step": 4 + }, + { + "epoch": 0.0014455044810638912, + "grad_norm": 4.135673123106036, + "learning_rate": 2.8571428571428573e-06, + "loss": 1.7868, + "step": 5 + }, + { + "epoch": 0.0017346053772766695, + "grad_norm": 3.0971146670050285, + "learning_rate": 3.428571428571429e-06, + "loss": 1.8928, + "step": 6 + }, + { + "epoch": 0.002023706273489448, + "grad_norm": 2.5830650058945004, + "learning_rate": 4.000000000000001e-06, + "loss": 1.8522, + "step": 7 + }, + { + "epoch": 0.0023128071697022263, + "grad_norm": 2.3055953581941364, + "learning_rate": 4.571428571428572e-06, + "loss": 1.7831, + "step": 8 + }, + { + "epoch": 0.0026019080659150044, + "grad_norm": 1.8810933130960161, + "learning_rate": 5.142857142857142e-06, + "loss": 1.8335, + "step": 9 + }, + { + "epoch": 0.0028910089621277824, + "grad_norm": 2.308546357873222, + "learning_rate": 5.7142857142857145e-06, + "loss": 1.7858, + "step": 10 + }, + { + "epoch": 0.003180109858340561, + "grad_norm": 1.9146380317808562, + "learning_rate": 6.285714285714286e-06, + "loss": 1.8005, + "step": 11 + }, + { + "epoch": 0.003469210754553339, + "grad_norm": 1.7733597852169312, + "learning_rate": 6.857142857142858e-06, + "loss": 1.7313, + "step": 12 + }, + { + "epoch": 0.0037583116507661175, + "grad_norm": 1.9163487932648622, + "learning_rate": 7.428571428571429e-06, + "loss": 1.8676, + "step": 13 + }, + { + "epoch": 0.004047412546978896, + "grad_norm": 1.81837480862002, + "learning_rate": 8.000000000000001e-06, + "loss": 1.7696, + "step": 14 + }, + { + "epoch": 0.004336513443191674, + "grad_norm": 1.7872590059217233, + "learning_rate": 8.571428571428571e-06, + "loss": 1.804, + "step": 15 + }, + { + "epoch": 0.004625614339404453, + "grad_norm": 1.875901273963478, + "learning_rate": 9.142857142857144e-06, + "loss": 1.7379, + "step": 16 + }, + { + "epoch": 0.004914715235617231, + "grad_norm": 1.6902352819283006, + "learning_rate": 9.714285714285715e-06, + "loss": 1.7908, + "step": 17 + }, + { + "epoch": 0.005203816131830009, + "grad_norm": 1.673145216450599, + "learning_rate": 1.0285714285714285e-05, + "loss": 1.8525, + "step": 18 + }, + { + "epoch": 0.005492917028042787, + "grad_norm": 1.6917810585273594, + "learning_rate": 1.0857142857142858e-05, + "loss": 1.7779, + "step": 19 + }, + { + "epoch": 0.005782017924255565, + "grad_norm": 1.5858495861042294, + "learning_rate": 1.1428571428571429e-05, + "loss": 1.6791, + "step": 20 + }, + { + "epoch": 0.006071118820468344, + "grad_norm": 2.004987370634698, + "learning_rate": 1.2e-05, + "loss": 1.8019, + "step": 21 + }, + { + "epoch": 0.006360219716681122, + "grad_norm": 1.5348568469917172, + "learning_rate": 1.2571428571428572e-05, + "loss": 1.8565, + "step": 22 + }, + { + "epoch": 0.0066493206128939, + "grad_norm": 1.8857981149629766, + "learning_rate": 1.3142857142857145e-05, + "loss": 1.7292, + "step": 23 + }, + { + "epoch": 0.006938421509106678, + "grad_norm": 1.4949439821061543, + "learning_rate": 1.3714285714285716e-05, + "loss": 1.7357, + "step": 24 + }, + { + "epoch": 0.007227522405319456, + "grad_norm": 1.693165357859836, + "learning_rate": 1.4285714285714287e-05, + "loss": 1.8103, + "step": 25 + }, + { + "epoch": 0.007516623301532235, + "grad_norm": 1.6138034582127923, + "learning_rate": 1.4857142857142858e-05, + "loss": 1.7634, + "step": 26 + }, + { + "epoch": 0.007805724197745013, + "grad_norm": 1.6357966310241434, + "learning_rate": 1.542857142857143e-05, + "loss": 1.7699, + "step": 27 + }, + { + "epoch": 0.008094825093957791, + "grad_norm": 1.5038647382371504, + "learning_rate": 1.6000000000000003e-05, + "loss": 1.7824, + "step": 28 + }, + { + "epoch": 0.00838392599017057, + "grad_norm": 1.4835637524801124, + "learning_rate": 1.6571428571428574e-05, + "loss": 1.685, + "step": 29 + }, + { + "epoch": 0.008673026886383347, + "grad_norm": 1.5071883679399785, + "learning_rate": 1.7142857142857142e-05, + "loss": 1.8, + "step": 30 + }, + { + "epoch": 0.008962127782596125, + "grad_norm": 1.7942262887809508, + "learning_rate": 1.7714285714285717e-05, + "loss": 1.8129, + "step": 31 + }, + { + "epoch": 0.009251228678808905, + "grad_norm": 1.6582976562001455, + "learning_rate": 1.8285714285714288e-05, + "loss": 1.8308, + "step": 32 + }, + { + "epoch": 0.009540329575021683, + "grad_norm": 1.5337861742601142, + "learning_rate": 1.885714285714286e-05, + "loss": 1.7456, + "step": 33 + }, + { + "epoch": 0.009829430471234461, + "grad_norm": 1.8002402808530826, + "learning_rate": 1.942857142857143e-05, + "loss": 1.9268, + "step": 34 + }, + { + "epoch": 0.01011853136744724, + "grad_norm": 1.7514076269246104, + "learning_rate": 2e-05, + "loss": 1.652, + "step": 35 + }, + { + "epoch": 0.010407632263660017, + "grad_norm": 1.8491103433261384, + "learning_rate": 1.9999995790774362e-05, + "loss": 1.8142, + "step": 36 + }, + { + "epoch": 0.010696733159872795, + "grad_norm": 1.8365212131962536, + "learning_rate": 1.9999983163100983e-05, + "loss": 1.7107, + "step": 37 + }, + { + "epoch": 0.010985834056085574, + "grad_norm": 1.5873275727358707, + "learning_rate": 1.9999962116990498e-05, + "loss": 1.6696, + "step": 38 + }, + { + "epoch": 0.011274934952298352, + "grad_norm": 1.7256221512542549, + "learning_rate": 1.999993265246062e-05, + "loss": 1.6901, + "step": 39 + }, + { + "epoch": 0.01156403584851113, + "grad_norm": 1.8002146086893285, + "learning_rate": 1.9999894769536163e-05, + "loss": 1.8886, + "step": 40 + }, + { + "epoch": 0.01185313674472391, + "grad_norm": 1.8367868066313766, + "learning_rate": 1.999984846824901e-05, + "loss": 1.7876, + "step": 41 + }, + { + "epoch": 0.012142237640936688, + "grad_norm": 1.5841989203084583, + "learning_rate": 1.999979374863814e-05, + "loss": 1.9172, + "step": 42 + }, + { + "epoch": 0.012431338537149466, + "grad_norm": 1.5522062286750515, + "learning_rate": 1.9999730610749623e-05, + "loss": 1.8207, + "step": 43 + }, + { + "epoch": 0.012720439433362244, + "grad_norm": 1.6373562793041219, + "learning_rate": 1.9999659054636612e-05, + "loss": 1.6358, + "step": 44 + }, + { + "epoch": 0.013009540329575022, + "grad_norm": 1.4771807978853968, + "learning_rate": 1.9999579080359337e-05, + "loss": 1.6684, + "step": 45 + }, + { + "epoch": 0.0132986412257878, + "grad_norm": 1.6255295191073982, + "learning_rate": 1.9999490687985134e-05, + "loss": 1.7113, + "step": 46 + }, + { + "epoch": 0.013587742122000578, + "grad_norm": 1.4924171597100626, + "learning_rate": 1.999939387758841e-05, + "loss": 1.7272, + "step": 47 + }, + { + "epoch": 0.013876843018213356, + "grad_norm": 1.536593102227256, + "learning_rate": 1.9999288649250667e-05, + "loss": 1.7532, + "step": 48 + }, + { + "epoch": 0.014165943914426134, + "grad_norm": 1.510648288143349, + "learning_rate": 1.9999175003060487e-05, + "loss": 1.8358, + "step": 49 + }, + { + "epoch": 0.014455044810638912, + "grad_norm": 1.4755302580340564, + "learning_rate": 1.9999052939113552e-05, + "loss": 1.8305, + "step": 50 + }, + { + "epoch": 0.014744145706851692, + "grad_norm": 1.659035076331237, + "learning_rate": 1.9998922457512608e-05, + "loss": 1.7916, + "step": 51 + }, + { + "epoch": 0.01503324660306447, + "grad_norm": 1.4194234922554345, + "learning_rate": 1.9998783558367506e-05, + "loss": 1.6693, + "step": 52 + }, + { + "epoch": 0.015322347499277248, + "grad_norm": 1.5709438854610118, + "learning_rate": 1.9998636241795184e-05, + "loss": 1.7873, + "step": 53 + }, + { + "epoch": 0.015611448395490026, + "grad_norm": 1.520579238616479, + "learning_rate": 1.999848050791965e-05, + "loss": 1.6949, + "step": 54 + }, + { + "epoch": 0.015900549291702804, + "grad_norm": 1.6871179175362505, + "learning_rate": 1.9998316356872015e-05, + "loss": 1.7697, + "step": 55 + }, + { + "epoch": 0.016189650187915582, + "grad_norm": 1.3949302668328034, + "learning_rate": 1.9998143788790462e-05, + "loss": 1.6176, + "step": 56 + }, + { + "epoch": 0.01647875108412836, + "grad_norm": 1.76393196989828, + "learning_rate": 1.9997962803820274e-05, + "loss": 1.7379, + "step": 57 + }, + { + "epoch": 0.01676785198034114, + "grad_norm": 2.3952825919806724, + "learning_rate": 1.9997773402113804e-05, + "loss": 1.8324, + "step": 58 + }, + { + "epoch": 0.017056952876553916, + "grad_norm": 1.5638827679672842, + "learning_rate": 1.9997575583830508e-05, + "loss": 1.783, + "step": 59 + }, + { + "epoch": 0.017346053772766695, + "grad_norm": 1.5321008746402438, + "learning_rate": 1.999736934913691e-05, + "loss": 1.7136, + "step": 60 + }, + { + "epoch": 0.017635154668979473, + "grad_norm": 1.46754278560031, + "learning_rate": 1.9997154698206636e-05, + "loss": 1.7567, + "step": 61 + }, + { + "epoch": 0.01792425556519225, + "grad_norm": 1.5842416386679679, + "learning_rate": 1.999693163122038e-05, + "loss": 1.757, + "step": 62 + }, + { + "epoch": 0.01821335646140503, + "grad_norm": 1.5267999996929678, + "learning_rate": 1.9996700148365936e-05, + "loss": 1.768, + "step": 63 + }, + { + "epoch": 0.01850245735761781, + "grad_norm": 1.6371904649667028, + "learning_rate": 1.9996460249838176e-05, + "loss": 1.8574, + "step": 64 + }, + { + "epoch": 0.01879155825383059, + "grad_norm": 1.4572549097707117, + "learning_rate": 1.9996211935839053e-05, + "loss": 1.8134, + "step": 65 + }, + { + "epoch": 0.019080659150043366, + "grad_norm": 1.6026953718941006, + "learning_rate": 1.9995955206577617e-05, + "loss": 1.7718, + "step": 66 + }, + { + "epoch": 0.019369760046256145, + "grad_norm": 1.4731386293380266, + "learning_rate": 1.9995690062269985e-05, + "loss": 1.7598, + "step": 67 + }, + { + "epoch": 0.019658860942468923, + "grad_norm": 1.4280673736098164, + "learning_rate": 1.9995416503139372e-05, + "loss": 1.7469, + "step": 68 + }, + { + "epoch": 0.0199479618386817, + "grad_norm": 1.4976095361519832, + "learning_rate": 1.9995134529416077e-05, + "loss": 1.7258, + "step": 69 + }, + { + "epoch": 0.02023706273489448, + "grad_norm": 1.3734046074406898, + "learning_rate": 1.9994844141337467e-05, + "loss": 1.6921, + "step": 70 + }, + { + "epoch": 0.020526163631107257, + "grad_norm": 1.5285691045400083, + "learning_rate": 1.9994545339148017e-05, + "loss": 1.755, + "step": 71 + }, + { + "epoch": 0.020815264527320035, + "grad_norm": 1.7237565941986015, + "learning_rate": 1.9994238123099264e-05, + "loss": 1.7254, + "step": 72 + }, + { + "epoch": 0.021104365423532813, + "grad_norm": 1.4511698572347005, + "learning_rate": 1.9993922493449833e-05, + "loss": 1.7339, + "step": 73 + }, + { + "epoch": 0.02139346631974559, + "grad_norm": 1.3034115810336904, + "learning_rate": 1.9993598450465446e-05, + "loss": 1.8015, + "step": 74 + }, + { + "epoch": 0.02168256721595837, + "grad_norm": 1.3220719840165054, + "learning_rate": 1.9993265994418887e-05, + "loss": 1.7794, + "step": 75 + }, + { + "epoch": 0.021971668112171147, + "grad_norm": 1.3514726546077938, + "learning_rate": 1.999292512559004e-05, + "loss": 1.781, + "step": 76 + }, + { + "epoch": 0.022260769008383925, + "grad_norm": 1.4854967526025784, + "learning_rate": 1.9992575844265857e-05, + "loss": 1.8188, + "step": 77 + }, + { + "epoch": 0.022549869904596703, + "grad_norm": 1.42527411671424, + "learning_rate": 1.9992218150740386e-05, + "loss": 1.7858, + "step": 78 + }, + { + "epoch": 0.02283897080080948, + "grad_norm": 1.3690019880914763, + "learning_rate": 1.9991852045314744e-05, + "loss": 1.69, + "step": 79 + }, + { + "epoch": 0.02312807169702226, + "grad_norm": 1.4796987548843166, + "learning_rate": 1.9991477528297136e-05, + "loss": 1.7501, + "step": 80 + }, + { + "epoch": 0.023417172593235037, + "grad_norm": 1.4237249543510602, + "learning_rate": 1.999109460000285e-05, + "loss": 1.7634, + "step": 81 + }, + { + "epoch": 0.02370627348944782, + "grad_norm": 1.3087042024971207, + "learning_rate": 1.9990703260754248e-05, + "loss": 1.8106, + "step": 82 + }, + { + "epoch": 0.023995374385660597, + "grad_norm": 1.4158584649264259, + "learning_rate": 1.999030351088078e-05, + "loss": 1.7503, + "step": 83 + }, + { + "epoch": 0.024284475281873375, + "grad_norm": 1.4266437173978106, + "learning_rate": 1.9989895350718972e-05, + "loss": 1.6435, + "step": 84 + }, + { + "epoch": 0.024573576178086153, + "grad_norm": 1.519614177980851, + "learning_rate": 1.9989478780612434e-05, + "loss": 1.7155, + "step": 85 + }, + { + "epoch": 0.02486267707429893, + "grad_norm": 1.4829974498829237, + "learning_rate": 1.9989053800911854e-05, + "loss": 1.7406, + "step": 86 + }, + { + "epoch": 0.02515177797051171, + "grad_norm": 1.5340468112410737, + "learning_rate": 1.9988620411974994e-05, + "loss": 1.6487, + "step": 87 + }, + { + "epoch": 0.025440878866724487, + "grad_norm": 1.3928928308705317, + "learning_rate": 1.9988178614166706e-05, + "loss": 1.7535, + "step": 88 + }, + { + "epoch": 0.025729979762937265, + "grad_norm": 1.4648291742355268, + "learning_rate": 1.9987728407858912e-05, + "loss": 1.6977, + "step": 89 + }, + { + "epoch": 0.026019080659150044, + "grad_norm": 1.5330205874026324, + "learning_rate": 1.9987269793430618e-05, + "loss": 1.8134, + "step": 90 + }, + { + "epoch": 0.02630818155536282, + "grad_norm": 1.4322119486638492, + "learning_rate": 1.9986802771267902e-05, + "loss": 1.7493, + "step": 91 + }, + { + "epoch": 0.0265972824515756, + "grad_norm": 1.3741952459535092, + "learning_rate": 1.9986327341763933e-05, + "loss": 1.7498, + "step": 92 + }, + { + "epoch": 0.026886383347788378, + "grad_norm": 1.4439173869710846, + "learning_rate": 1.998584350531894e-05, + "loss": 1.8364, + "step": 93 + }, + { + "epoch": 0.027175484244001156, + "grad_norm": 1.5238329839004898, + "learning_rate": 1.9985351262340242e-05, + "loss": 1.7025, + "step": 94 + }, + { + "epoch": 0.027464585140213934, + "grad_norm": 1.4211403871339836, + "learning_rate": 1.998485061324223e-05, + "loss": 1.8061, + "step": 95 + }, + { + "epoch": 0.027753686036426712, + "grad_norm": 1.4891220326396055, + "learning_rate": 1.998434155844638e-05, + "loss": 1.7756, + "step": 96 + }, + { + "epoch": 0.02804278693263949, + "grad_norm": 1.4672105325874756, + "learning_rate": 1.9983824098381226e-05, + "loss": 1.7642, + "step": 97 + }, + { + "epoch": 0.028331887828852268, + "grad_norm": 1.4102584299249543, + "learning_rate": 1.9983298233482396e-05, + "loss": 1.7302, + "step": 98 + }, + { + "epoch": 0.028620988725065046, + "grad_norm": 1.4570496244711266, + "learning_rate": 1.9982763964192586e-05, + "loss": 1.7299, + "step": 99 + }, + { + "epoch": 0.028910089621277824, + "grad_norm": 1.4536709099623053, + "learning_rate": 1.9982221290961568e-05, + "loss": 1.7647, + "step": 100 + }, + { + "epoch": 0.029199190517490606, + "grad_norm": 1.427202594686143, + "learning_rate": 1.9981670214246186e-05, + "loss": 1.7361, + "step": 101 + }, + { + "epoch": 0.029488291413703384, + "grad_norm": 1.4166737515686485, + "learning_rate": 1.9981110734510368e-05, + "loss": 1.7107, + "step": 102 + }, + { + "epoch": 0.029777392309916162, + "grad_norm": 1.4291514974937183, + "learning_rate": 1.99805428522251e-05, + "loss": 1.7162, + "step": 103 + }, + { + "epoch": 0.03006649320612894, + "grad_norm": 1.3691457808724619, + "learning_rate": 1.997996656786846e-05, + "loss": 1.8823, + "step": 104 + }, + { + "epoch": 0.030355594102341718, + "grad_norm": 1.38193263647762, + "learning_rate": 1.9979381881925587e-05, + "loss": 1.6658, + "step": 105 + }, + { + "epoch": 0.030644694998554496, + "grad_norm": 1.4331634763444776, + "learning_rate": 1.997878879488869e-05, + "loss": 1.7695, + "step": 106 + }, + { + "epoch": 0.030933795894767274, + "grad_norm": 1.4389434938173666, + "learning_rate": 1.9978187307257064e-05, + "loss": 1.7356, + "step": 107 + }, + { + "epoch": 0.031222896790980052, + "grad_norm": 1.4057807430118254, + "learning_rate": 1.9977577419537066e-05, + "loss": 1.7191, + "step": 108 + }, + { + "epoch": 0.03151199768719283, + "grad_norm": 1.4476225746298625, + "learning_rate": 1.9976959132242128e-05, + "loss": 1.8028, + "step": 109 + }, + { + "epoch": 0.03180109858340561, + "grad_norm": 1.3595624610249801, + "learning_rate": 1.997633244589275e-05, + "loss": 1.708, + "step": 110 + }, + { + "epoch": 0.032090199479618386, + "grad_norm": 1.5622440441344578, + "learning_rate": 1.997569736101651e-05, + "loss": 1.7138, + "step": 111 + }, + { + "epoch": 0.032379300375831165, + "grad_norm": 1.4451837327532462, + "learning_rate": 1.9975053878148045e-05, + "loss": 1.745, + "step": 112 + }, + { + "epoch": 0.03266840127204394, + "grad_norm": 1.3461945134165731, + "learning_rate": 1.9974401997829063e-05, + "loss": 1.792, + "step": 113 + }, + { + "epoch": 0.03295750216825672, + "grad_norm": 1.4916956994354813, + "learning_rate": 1.997374172060836e-05, + "loss": 1.7312, + "step": 114 + }, + { + "epoch": 0.0332466030644695, + "grad_norm": 1.3473635350533286, + "learning_rate": 1.997307304704178e-05, + "loss": 1.7257, + "step": 115 + }, + { + "epoch": 0.03353570396068228, + "grad_norm": 1.5308835567804706, + "learning_rate": 1.9972395977692243e-05, + "loss": 1.8833, + "step": 116 + }, + { + "epoch": 0.033824804856895055, + "grad_norm": 1.4933799858376915, + "learning_rate": 1.9971710513129735e-05, + "loss": 1.8031, + "step": 117 + }, + { + "epoch": 0.03411390575310783, + "grad_norm": 1.4666676384345598, + "learning_rate": 1.9971016653931315e-05, + "loss": 1.699, + "step": 118 + }, + { + "epoch": 0.03440300664932061, + "grad_norm": 1.376396676563015, + "learning_rate": 1.9970314400681098e-05, + "loss": 1.7561, + "step": 119 + }, + { + "epoch": 0.03469210754553339, + "grad_norm": 1.4309738039308064, + "learning_rate": 1.9969603753970285e-05, + "loss": 1.7812, + "step": 120 + }, + { + "epoch": 0.03498120844174617, + "grad_norm": 1.4395837496367765, + "learning_rate": 1.9968884714397116e-05, + "loss": 1.6863, + "step": 121 + }, + { + "epoch": 0.035270309337958945, + "grad_norm": 1.3440476522906626, + "learning_rate": 1.9968157282566917e-05, + "loss": 1.7026, + "step": 122 + }, + { + "epoch": 0.03555941023417172, + "grad_norm": 1.3253598540978417, + "learning_rate": 1.9967421459092076e-05, + "loss": 1.6773, + "step": 123 + }, + { + "epoch": 0.0358485111303845, + "grad_norm": 1.4654991309929606, + "learning_rate": 1.996667724459204e-05, + "loss": 1.6995, + "step": 124 + }, + { + "epoch": 0.03613761202659728, + "grad_norm": 1.3695366677193352, + "learning_rate": 1.9965924639693323e-05, + "loss": 1.8028, + "step": 125 + }, + { + "epoch": 0.03642671292281006, + "grad_norm": 1.5996594030035125, + "learning_rate": 1.9965163645029496e-05, + "loss": 1.6509, + "step": 126 + }, + { + "epoch": 0.036715813819022836, + "grad_norm": 1.3638136094831714, + "learning_rate": 1.9964394261241205e-05, + "loss": 1.7974, + "step": 127 + }, + { + "epoch": 0.03700491471523562, + "grad_norm": 1.5950089807963124, + "learning_rate": 1.9963616488976156e-05, + "loss": 1.7817, + "step": 128 + }, + { + "epoch": 0.0372940156114484, + "grad_norm": 1.3874690940551964, + "learning_rate": 1.9962830328889104e-05, + "loss": 1.7727, + "step": 129 + }, + { + "epoch": 0.03758311650766118, + "grad_norm": 1.428332455733304, + "learning_rate": 1.9962035781641877e-05, + "loss": 1.8162, + "step": 130 + }, + { + "epoch": 0.037872217403873955, + "grad_norm": 1.413377966510162, + "learning_rate": 1.996123284790336e-05, + "loss": 1.6724, + "step": 131 + }, + { + "epoch": 0.03816131830008673, + "grad_norm": 1.4114781128468488, + "learning_rate": 1.9960421528349503e-05, + "loss": 1.6688, + "step": 132 + }, + { + "epoch": 0.03845041919629951, + "grad_norm": 1.4667937866255885, + "learning_rate": 1.9959601823663305e-05, + "loss": 1.7404, + "step": 133 + }, + { + "epoch": 0.03873952009251229, + "grad_norm": 1.3864312896377629, + "learning_rate": 1.9958773734534834e-05, + "loss": 1.6797, + "step": 134 + }, + { + "epoch": 0.03902862098872507, + "grad_norm": 1.5863275696219379, + "learning_rate": 1.9957937261661215e-05, + "loss": 1.7093, + "step": 135 + }, + { + "epoch": 0.039317721884937845, + "grad_norm": 1.3943132521368478, + "learning_rate": 1.9957092405746628e-05, + "loss": 1.7466, + "step": 136 + }, + { + "epoch": 0.03960682278115062, + "grad_norm": 1.4418545458171168, + "learning_rate": 1.9956239167502304e-05, + "loss": 1.8363, + "step": 137 + }, + { + "epoch": 0.0398959236773634, + "grad_norm": 1.6597022379228474, + "learning_rate": 1.9955377547646546e-05, + "loss": 1.7726, + "step": 138 + }, + { + "epoch": 0.04018502457357618, + "grad_norm": 1.374590779273713, + "learning_rate": 1.9954507546904697e-05, + "loss": 1.7559, + "step": 139 + }, + { + "epoch": 0.04047412546978896, + "grad_norm": 1.2866499565497842, + "learning_rate": 1.995362916600917e-05, + "loss": 1.7478, + "step": 140 + }, + { + "epoch": 0.040763226366001735, + "grad_norm": 1.3549533524362687, + "learning_rate": 1.9952742405699425e-05, + "loss": 1.8389, + "step": 141 + }, + { + "epoch": 0.041052327262214514, + "grad_norm": 1.3512745611701633, + "learning_rate": 1.995184726672197e-05, + "loss": 1.7975, + "step": 142 + }, + { + "epoch": 0.04134142815842729, + "grad_norm": 1.3460191452855619, + "learning_rate": 1.995094374983038e-05, + "loss": 1.7245, + "step": 143 + }, + { + "epoch": 0.04163052905464007, + "grad_norm": 1.3305167255129993, + "learning_rate": 1.9950031855785276e-05, + "loss": 1.6294, + "step": 144 + }, + { + "epoch": 0.04191962995085285, + "grad_norm": 1.34284871707125, + "learning_rate": 1.9949111585354328e-05, + "loss": 1.7124, + "step": 145 + }, + { + "epoch": 0.042208730847065626, + "grad_norm": 1.3772925926259127, + "learning_rate": 1.9948182939312258e-05, + "loss": 1.7095, + "step": 146 + }, + { + "epoch": 0.042497831743278404, + "grad_norm": 1.384298315882825, + "learning_rate": 1.994724591844085e-05, + "loss": 1.8031, + "step": 147 + }, + { + "epoch": 0.04278693263949118, + "grad_norm": 1.3740208297445895, + "learning_rate": 1.994630052352893e-05, + "loss": 1.7814, + "step": 148 + }, + { + "epoch": 0.04307603353570396, + "grad_norm": 1.3187990568089156, + "learning_rate": 1.9945346755372367e-05, + "loss": 1.7486, + "step": 149 + }, + { + "epoch": 0.04336513443191674, + "grad_norm": 1.3347832866557865, + "learning_rate": 1.9944384614774095e-05, + "loss": 1.7225, + "step": 150 + }, + { + "epoch": 0.043654235328129516, + "grad_norm": 1.3850913268385951, + "learning_rate": 1.9943414102544083e-05, + "loss": 1.7608, + "step": 151 + }, + { + "epoch": 0.043943336224342294, + "grad_norm": 1.3180540802523992, + "learning_rate": 1.994243521949935e-05, + "loss": 1.7534, + "step": 152 + }, + { + "epoch": 0.04423243712055507, + "grad_norm": 1.3989923351107776, + "learning_rate": 1.9941447966463966e-05, + "loss": 1.7734, + "step": 153 + }, + { + "epoch": 0.04452153801676785, + "grad_norm": 1.2709853505702844, + "learning_rate": 1.9940452344269045e-05, + "loss": 1.7463, + "step": 154 + }, + { + "epoch": 0.04481063891298063, + "grad_norm": 1.3117490865184025, + "learning_rate": 1.9939448353752745e-05, + "loss": 1.7484, + "step": 155 + }, + { + "epoch": 0.045099739809193407, + "grad_norm": 1.3068520843456393, + "learning_rate": 1.9938435995760275e-05, + "loss": 1.7793, + "step": 156 + }, + { + "epoch": 0.045388840705406185, + "grad_norm": 1.2903723753260385, + "learning_rate": 1.993741527114388e-05, + "loss": 1.7654, + "step": 157 + }, + { + "epoch": 0.04567794160161896, + "grad_norm": 1.3281351334689, + "learning_rate": 1.993638618076285e-05, + "loss": 1.7574, + "step": 158 + }, + { + "epoch": 0.04596704249783174, + "grad_norm": 1.331538167128145, + "learning_rate": 1.9935348725483526e-05, + "loss": 1.7431, + "step": 159 + }, + { + "epoch": 0.04625614339404452, + "grad_norm": 1.4516409761685403, + "learning_rate": 1.9934302906179285e-05, + "loss": 1.7334, + "step": 160 + }, + { + "epoch": 0.0465452442902573, + "grad_norm": 1.2977322772074695, + "learning_rate": 1.9933248723730536e-05, + "loss": 1.6306, + "step": 161 + }, + { + "epoch": 0.046834345186470075, + "grad_norm": 1.357062910781076, + "learning_rate": 1.9932186179024744e-05, + "loss": 1.668, + "step": 162 + }, + { + "epoch": 0.04712344608268285, + "grad_norm": 1.3803460903193827, + "learning_rate": 1.9931115272956405e-05, + "loss": 1.7275, + "step": 163 + }, + { + "epoch": 0.04741254697889564, + "grad_norm": 1.315843862445011, + "learning_rate": 1.9930036006427058e-05, + "loss": 1.6444, + "step": 164 + }, + { + "epoch": 0.047701647875108416, + "grad_norm": 1.4617283691426304, + "learning_rate": 1.9928948380345274e-05, + "loss": 1.8109, + "step": 165 + }, + { + "epoch": 0.047990748771321194, + "grad_norm": 1.38383885394047, + "learning_rate": 1.9927852395626672e-05, + "loss": 1.7686, + "step": 166 + }, + { + "epoch": 0.04827984966753397, + "grad_norm": 1.458724964408903, + "learning_rate": 1.9926748053193898e-05, + "loss": 1.6962, + "step": 167 + }, + { + "epoch": 0.04856895056374675, + "grad_norm": 1.3411049270177544, + "learning_rate": 1.9925635353976634e-05, + "loss": 1.6548, + "step": 168 + }, + { + "epoch": 0.04885805145995953, + "grad_norm": 1.4316515117975288, + "learning_rate": 1.992451429891161e-05, + "loss": 1.7439, + "step": 169 + }, + { + "epoch": 0.049147152356172306, + "grad_norm": 1.3310049050553332, + "learning_rate": 1.9923384888942568e-05, + "loss": 1.7042, + "step": 170 + }, + { + "epoch": 0.049436253252385085, + "grad_norm": 1.411118381246596, + "learning_rate": 1.9922247125020307e-05, + "loss": 1.7172, + "step": 171 + }, + { + "epoch": 0.04972535414859786, + "grad_norm": 1.2882521449870807, + "learning_rate": 1.992110100810264e-05, + "loss": 1.7679, + "step": 172 + }, + { + "epoch": 0.05001445504481064, + "grad_norm": 1.3412263735063734, + "learning_rate": 1.991994653915443e-05, + "loss": 1.6375, + "step": 173 + }, + { + "epoch": 0.05030355594102342, + "grad_norm": 1.2875141274832036, + "learning_rate": 1.991878371914755e-05, + "loss": 1.6816, + "step": 174 + }, + { + "epoch": 0.0505926568372362, + "grad_norm": 1.2707786543811448, + "learning_rate": 1.991761254906092e-05, + "loss": 1.7382, + "step": 175 + }, + { + "epoch": 0.050881757733448975, + "grad_norm": 1.2815480979153002, + "learning_rate": 1.9916433029880485e-05, + "loss": 1.7519, + "step": 176 + }, + { + "epoch": 0.05117085862966175, + "grad_norm": 1.358987922639111, + "learning_rate": 1.991524516259921e-05, + "loss": 1.7895, + "step": 177 + }, + { + "epoch": 0.05145995952587453, + "grad_norm": 1.2688791830287245, + "learning_rate": 1.9914048948217105e-05, + "loss": 1.776, + "step": 178 + }, + { + "epoch": 0.05174906042208731, + "grad_norm": 1.2813810920416182, + "learning_rate": 1.9912844387741194e-05, + "loss": 1.7085, + "step": 179 + }, + { + "epoch": 0.05203816131830009, + "grad_norm": 1.2916088597948596, + "learning_rate": 1.991163148218553e-05, + "loss": 1.7124, + "step": 180 + }, + { + "epoch": 0.052327262214512865, + "grad_norm": 1.310527847083729, + "learning_rate": 1.9910410232571187e-05, + "loss": 1.6333, + "step": 181 + }, + { + "epoch": 0.05261636311072564, + "grad_norm": 1.3292526744086122, + "learning_rate": 1.9909180639926274e-05, + "loss": 1.663, + "step": 182 + }, + { + "epoch": 0.05290546400693842, + "grad_norm": 1.4107465041661023, + "learning_rate": 1.990794270528592e-05, + "loss": 1.6939, + "step": 183 + }, + { + "epoch": 0.0531945649031512, + "grad_norm": 1.5697948856294517, + "learning_rate": 1.9906696429692265e-05, + "loss": 1.79, + "step": 184 + }, + { + "epoch": 0.05348366579936398, + "grad_norm": 1.359687672105227, + "learning_rate": 1.9905441814194482e-05, + "loss": 1.6024, + "step": 185 + }, + { + "epoch": 0.053772766695576756, + "grad_norm": 1.435220148856358, + "learning_rate": 1.990417885984877e-05, + "loss": 1.6539, + "step": 186 + }, + { + "epoch": 0.054061867591789534, + "grad_norm": 1.4702625299572514, + "learning_rate": 1.990290756771834e-05, + "loss": 1.7306, + "step": 187 + }, + { + "epoch": 0.05435096848800231, + "grad_norm": 1.5247866204720824, + "learning_rate": 1.9901627938873416e-05, + "loss": 1.7609, + "step": 188 + }, + { + "epoch": 0.05464006938421509, + "grad_norm": 1.3031717794857007, + "learning_rate": 1.9900339974391252e-05, + "loss": 1.7389, + "step": 189 + }, + { + "epoch": 0.05492917028042787, + "grad_norm": 1.3511696476383066, + "learning_rate": 1.9899043675356114e-05, + "loss": 1.7297, + "step": 190 + }, + { + "epoch": 0.055218271176640646, + "grad_norm": 1.4509301331653846, + "learning_rate": 1.9897739042859285e-05, + "loss": 1.737, + "step": 191 + }, + { + "epoch": 0.055507372072853424, + "grad_norm": 1.3380070550795329, + "learning_rate": 1.9896426077999062e-05, + "loss": 1.6662, + "step": 192 + }, + { + "epoch": 0.0557964729690662, + "grad_norm": 1.3684181423939343, + "learning_rate": 1.989510478188076e-05, + "loss": 1.7297, + "step": 193 + }, + { + "epoch": 0.05608557386527898, + "grad_norm": 1.4948293977771248, + "learning_rate": 1.9893775155616704e-05, + "loss": 1.7733, + "step": 194 + }, + { + "epoch": 0.05637467476149176, + "grad_norm": 6.056711246922273, + "learning_rate": 1.989243720032624e-05, + "loss": 1.8127, + "step": 195 + }, + { + "epoch": 0.056663775657704536, + "grad_norm": 1.4031162820625223, + "learning_rate": 1.989109091713571e-05, + "loss": 1.7476, + "step": 196 + }, + { + "epoch": 0.056952876553917314, + "grad_norm": 1.388743000455743, + "learning_rate": 1.988973630717848e-05, + "loss": 1.7256, + "step": 197 + }, + { + "epoch": 0.05724197745013009, + "grad_norm": 1.3228310217145096, + "learning_rate": 1.9888373371594923e-05, + "loss": 1.7153, + "step": 198 + }, + { + "epoch": 0.05753107834634287, + "grad_norm": 1.349390586328493, + "learning_rate": 1.9887002111532413e-05, + "loss": 1.7913, + "step": 199 + }, + { + "epoch": 0.05782017924255565, + "grad_norm": 1.3218835310967434, + "learning_rate": 1.9885622528145346e-05, + "loss": 1.893, + "step": 200 + }, + { + "epoch": 0.058109280138768434, + "grad_norm": 1.3066393320266039, + "learning_rate": 1.9884234622595117e-05, + "loss": 1.7445, + "step": 201 + }, + { + "epoch": 0.05839838103498121, + "grad_norm": 1.3431846081510803, + "learning_rate": 1.988283839605013e-05, + "loss": 1.764, + "step": 202 + }, + { + "epoch": 0.05868748193119399, + "grad_norm": 1.2716836951252142, + "learning_rate": 1.988143384968578e-05, + "loss": 1.689, + "step": 203 + }, + { + "epoch": 0.05897658282740677, + "grad_norm": 1.3463201818447268, + "learning_rate": 1.9880020984684486e-05, + "loss": 1.6615, + "step": 204 + }, + { + "epoch": 0.059265683723619546, + "grad_norm": 1.3067391536442128, + "learning_rate": 1.9878599802235662e-05, + "loss": 1.683, + "step": 205 + }, + { + "epoch": 0.059554784619832324, + "grad_norm": 1.407390031033922, + "learning_rate": 1.987717030353572e-05, + "loss": 1.7617, + "step": 206 + }, + { + "epoch": 0.0598438855160451, + "grad_norm": 1.2785798130045682, + "learning_rate": 1.9875732489788082e-05, + "loss": 1.6828, + "step": 207 + }, + { + "epoch": 0.06013298641225788, + "grad_norm": 1.581333677513783, + "learning_rate": 1.9874286362203162e-05, + "loss": 1.776, + "step": 208 + }, + { + "epoch": 0.06042208730847066, + "grad_norm": 1.3170624490173282, + "learning_rate": 1.987283192199837e-05, + "loss": 1.7548, + "step": 209 + }, + { + "epoch": 0.060711188204683436, + "grad_norm": 1.6773819921976063, + "learning_rate": 1.9871369170398126e-05, + "loss": 1.7935, + "step": 210 + }, + { + "epoch": 0.061000289100896214, + "grad_norm": 1.4063651245701678, + "learning_rate": 1.9869898108633834e-05, + "loss": 1.7806, + "step": 211 + }, + { + "epoch": 0.06128938999710899, + "grad_norm": 1.3241937165367996, + "learning_rate": 1.986841873794391e-05, + "loss": 1.7262, + "step": 212 + }, + { + "epoch": 0.06157849089332177, + "grad_norm": 1.49389208627189, + "learning_rate": 1.9866931059573748e-05, + "loss": 1.7103, + "step": 213 + }, + { + "epoch": 0.06186759178953455, + "grad_norm": 1.219442276366396, + "learning_rate": 1.9865435074775744e-05, + "loss": 1.6194, + "step": 214 + }, + { + "epoch": 0.062156692685747326, + "grad_norm": 1.3453821292169061, + "learning_rate": 1.9863930784809284e-05, + "loss": 1.8086, + "step": 215 + }, + { + "epoch": 0.062445793581960105, + "grad_norm": 1.2961743317843437, + "learning_rate": 1.986241819094075e-05, + "loss": 1.6352, + "step": 216 + }, + { + "epoch": 0.06273489447817288, + "grad_norm": 1.3729904036841547, + "learning_rate": 1.986089729444351e-05, + "loss": 1.7041, + "step": 217 + }, + { + "epoch": 0.06302399537438566, + "grad_norm": 1.3443809779979903, + "learning_rate": 1.9859368096597925e-05, + "loss": 1.7332, + "step": 218 + }, + { + "epoch": 0.06331309627059845, + "grad_norm": 1.3367070791840931, + "learning_rate": 1.9857830598691337e-05, + "loss": 1.639, + "step": 219 + }, + { + "epoch": 0.06360219716681122, + "grad_norm": 1.3799415387024292, + "learning_rate": 1.9856284802018087e-05, + "loss": 1.7675, + "step": 220 + }, + { + "epoch": 0.063891298063024, + "grad_norm": 1.3464389753341524, + "learning_rate": 1.9854730707879502e-05, + "loss": 1.6702, + "step": 221 + }, + { + "epoch": 0.06418039895923677, + "grad_norm": 1.3867842033565003, + "learning_rate": 1.9853168317583874e-05, + "loss": 1.7216, + "step": 222 + }, + { + "epoch": 0.06446949985544956, + "grad_norm": 1.2892516322419894, + "learning_rate": 1.9851597632446508e-05, + "loss": 1.6755, + "step": 223 + }, + { + "epoch": 0.06475860075166233, + "grad_norm": 1.4307425363074875, + "learning_rate": 1.9850018653789665e-05, + "loss": 1.7618, + "step": 224 + }, + { + "epoch": 0.06504770164787511, + "grad_norm": 1.288672937098248, + "learning_rate": 1.984843138294261e-05, + "loss": 1.756, + "step": 225 + }, + { + "epoch": 0.06533680254408789, + "grad_norm": 1.3540206108809039, + "learning_rate": 1.9846835821241573e-05, + "loss": 1.7373, + "step": 226 + }, + { + "epoch": 0.06562590344030067, + "grad_norm": 1.3108759943656667, + "learning_rate": 1.9845231970029774e-05, + "loss": 1.6429, + "step": 227 + }, + { + "epoch": 0.06591500433651344, + "grad_norm": 1.3549691917787245, + "learning_rate": 1.9843619830657404e-05, + "loss": 1.6692, + "step": 228 + }, + { + "epoch": 0.06620410523272623, + "grad_norm": 1.350566839669386, + "learning_rate": 1.984199940448164e-05, + "loss": 1.6499, + "step": 229 + }, + { + "epoch": 0.066493206128939, + "grad_norm": 1.3992430914185707, + "learning_rate": 1.9840370692866624e-05, + "loss": 1.7279, + "step": 230 + }, + { + "epoch": 0.06678230702515178, + "grad_norm": 1.336273448437935, + "learning_rate": 1.9838733697183482e-05, + "loss": 1.7378, + "step": 231 + }, + { + "epoch": 0.06707140792136455, + "grad_norm": 1.3714334108906678, + "learning_rate": 1.9837088418810312e-05, + "loss": 1.6845, + "step": 232 + }, + { + "epoch": 0.06736050881757734, + "grad_norm": 1.331192814515942, + "learning_rate": 1.983543485913218e-05, + "loss": 1.7605, + "step": 233 + }, + { + "epoch": 0.06764960971379011, + "grad_norm": 1.3590047127242209, + "learning_rate": 1.983377301954113e-05, + "loss": 1.7027, + "step": 234 + }, + { + "epoch": 0.0679387106100029, + "grad_norm": 1.2628103031697302, + "learning_rate": 1.983210290143617e-05, + "loss": 1.6682, + "step": 235 + }, + { + "epoch": 0.06822781150621567, + "grad_norm": 1.31345700201241, + "learning_rate": 1.9830424506223286e-05, + "loss": 1.7787, + "step": 236 + }, + { + "epoch": 0.06851691240242845, + "grad_norm": 1.367333629142093, + "learning_rate": 1.9828737835315422e-05, + "loss": 1.7182, + "step": 237 + }, + { + "epoch": 0.06880601329864122, + "grad_norm": 1.3814327986016506, + "learning_rate": 1.9827042890132498e-05, + "loss": 1.8058, + "step": 238 + }, + { + "epoch": 0.06909511419485401, + "grad_norm": 1.2616372891530419, + "learning_rate": 1.982533967210139e-05, + "loss": 1.7131, + "step": 239 + }, + { + "epoch": 0.06938421509106678, + "grad_norm": 1.3695087479555836, + "learning_rate": 1.982362818265595e-05, + "loss": 1.7163, + "step": 240 + }, + { + "epoch": 0.06967331598727956, + "grad_norm": 1.3557868876434347, + "learning_rate": 1.9821908423236983e-05, + "loss": 1.7421, + "step": 241 + }, + { + "epoch": 0.06996241688349233, + "grad_norm": 1.392442344450402, + "learning_rate": 1.982018039529226e-05, + "loss": 1.747, + "step": 242 + }, + { + "epoch": 0.07025151777970512, + "grad_norm": 1.4471159007355705, + "learning_rate": 1.9818444100276517e-05, + "loss": 1.6452, + "step": 243 + }, + { + "epoch": 0.07054061867591789, + "grad_norm": 1.323270395385573, + "learning_rate": 1.981669953965144e-05, + "loss": 1.795, + "step": 244 + }, + { + "epoch": 0.07082971957213068, + "grad_norm": 1.32959668728154, + "learning_rate": 1.9814946714885682e-05, + "loss": 1.8027, + "step": 245 + }, + { + "epoch": 0.07111882046834345, + "grad_norm": 1.3027587139671195, + "learning_rate": 1.981318562745485e-05, + "loss": 1.7326, + "step": 246 + }, + { + "epoch": 0.07140792136455623, + "grad_norm": 1.328453385919723, + "learning_rate": 1.9811416278841507e-05, + "loss": 1.6839, + "step": 247 + }, + { + "epoch": 0.071697022260769, + "grad_norm": 1.2271866822510065, + "learning_rate": 1.980963867053517e-05, + "loss": 1.6021, + "step": 248 + }, + { + "epoch": 0.07198612315698179, + "grad_norm": 1.2908284006992323, + "learning_rate": 1.9807852804032306e-05, + "loss": 1.6339, + "step": 249 + }, + { + "epoch": 0.07227522405319456, + "grad_norm": 1.373522694157661, + "learning_rate": 1.9806058680836343e-05, + "loss": 1.6902, + "step": 250 + }, + { + "epoch": 0.07256432494940734, + "grad_norm": 1.2634235256940958, + "learning_rate": 1.9804256302457653e-05, + "loss": 1.6517, + "step": 251 + }, + { + "epoch": 0.07285342584562012, + "grad_norm": 1.3278784508700443, + "learning_rate": 1.9802445670413562e-05, + "loss": 1.7972, + "step": 252 + }, + { + "epoch": 0.0731425267418329, + "grad_norm": 1.3552386657179452, + "learning_rate": 1.980062678622834e-05, + "loss": 1.726, + "step": 253 + }, + { + "epoch": 0.07343162763804567, + "grad_norm": 1.3742079147545718, + "learning_rate": 1.9798799651433204e-05, + "loss": 1.7317, + "step": 254 + }, + { + "epoch": 0.07372072853425846, + "grad_norm": 1.4074238143735376, + "learning_rate": 1.9796964267566322e-05, + "loss": 1.7275, + "step": 255 + }, + { + "epoch": 0.07400982943047124, + "grad_norm": 1.286876584367334, + "learning_rate": 1.9795120636172804e-05, + "loss": 1.6061, + "step": 256 + }, + { + "epoch": 0.07429893032668401, + "grad_norm": 1.3521349118231711, + "learning_rate": 1.9793268758804693e-05, + "loss": 1.7237, + "step": 257 + }, + { + "epoch": 0.0745880312228968, + "grad_norm": 1.2816583167182087, + "learning_rate": 1.979140863702099e-05, + "loss": 1.6783, + "step": 258 + }, + { + "epoch": 0.07487713211910957, + "grad_norm": 1.3634926357253074, + "learning_rate": 1.978954027238763e-05, + "loss": 1.6784, + "step": 259 + }, + { + "epoch": 0.07516623301532235, + "grad_norm": 1.4038229521295733, + "learning_rate": 1.9787663666477486e-05, + "loss": 1.6535, + "step": 260 + }, + { + "epoch": 0.07545533391153512, + "grad_norm": 1.2842218606303095, + "learning_rate": 1.9785778820870366e-05, + "loss": 1.7058, + "step": 261 + }, + { + "epoch": 0.07574443480774791, + "grad_norm": 1.347020075001843, + "learning_rate": 1.9783885737153022e-05, + "loss": 1.7832, + "step": 262 + }, + { + "epoch": 0.07603353570396068, + "grad_norm": 1.26053283819741, + "learning_rate": 1.9781984416919138e-05, + "loss": 1.644, + "step": 263 + }, + { + "epoch": 0.07632263660017347, + "grad_norm": 1.2495419382455348, + "learning_rate": 1.9780074861769328e-05, + "loss": 1.6807, + "step": 264 + }, + { + "epoch": 0.07661173749638624, + "grad_norm": 1.7139411191235945, + "learning_rate": 1.977815707331114e-05, + "loss": 1.7222, + "step": 265 + }, + { + "epoch": 0.07690083839259902, + "grad_norm": 1.4547863480275938, + "learning_rate": 1.9776231053159063e-05, + "loss": 1.7774, + "step": 266 + }, + { + "epoch": 0.07718993928881179, + "grad_norm": 1.3437347227693044, + "learning_rate": 1.97742968029345e-05, + "loss": 1.7208, + "step": 267 + }, + { + "epoch": 0.07747904018502458, + "grad_norm": 1.354739938343438, + "learning_rate": 1.9772354324265796e-05, + "loss": 1.6918, + "step": 268 + }, + { + "epoch": 0.07776814108123735, + "grad_norm": 1.2544728177655773, + "learning_rate": 1.977040361878821e-05, + "loss": 1.6351, + "step": 269 + }, + { + "epoch": 0.07805724197745013, + "grad_norm": 1.3417761983959458, + "learning_rate": 1.9768444688143938e-05, + "loss": 1.6564, + "step": 270 + }, + { + "epoch": 0.0783463428736629, + "grad_norm": 1.3422116229415715, + "learning_rate": 1.9766477533982094e-05, + "loss": 1.8033, + "step": 271 + }, + { + "epoch": 0.07863544376987569, + "grad_norm": 1.2706457122008528, + "learning_rate": 1.976450215795872e-05, + "loss": 1.6505, + "step": 272 + }, + { + "epoch": 0.07892454466608846, + "grad_norm": 1.267352186352676, + "learning_rate": 1.9762518561736782e-05, + "loss": 1.6591, + "step": 273 + }, + { + "epoch": 0.07921364556230125, + "grad_norm": 1.4466694988557338, + "learning_rate": 1.976052674698615e-05, + "loss": 1.7952, + "step": 274 + }, + { + "epoch": 0.07950274645851402, + "grad_norm": 1.321832538613218, + "learning_rate": 1.975852671538363e-05, + "loss": 1.7061, + "step": 275 + }, + { + "epoch": 0.0797918473547268, + "grad_norm": 1.2829050605849048, + "learning_rate": 1.9756518468612934e-05, + "loss": 1.6605, + "step": 276 + }, + { + "epoch": 0.08008094825093957, + "grad_norm": 1.2987764926867447, + "learning_rate": 1.9754502008364702e-05, + "loss": 1.7047, + "step": 277 + }, + { + "epoch": 0.08037004914715236, + "grad_norm": 1.2552473311885344, + "learning_rate": 1.9752477336336473e-05, + "loss": 1.7031, + "step": 278 + }, + { + "epoch": 0.08065915004336513, + "grad_norm": 1.1985068633897615, + "learning_rate": 1.9750444454232714e-05, + "loss": 1.7382, + "step": 279 + }, + { + "epoch": 0.08094825093957791, + "grad_norm": 1.314553485624065, + "learning_rate": 1.9748403363764795e-05, + "loss": 1.6491, + "step": 280 + }, + { + "epoch": 0.08123735183579069, + "grad_norm": 1.294998222736541, + "learning_rate": 1.9746354066650995e-05, + "loss": 1.7044, + "step": 281 + }, + { + "epoch": 0.08152645273200347, + "grad_norm": 1.3584436916798586, + "learning_rate": 1.974429656461651e-05, + "loss": 1.6672, + "step": 282 + }, + { + "epoch": 0.08181555362821624, + "grad_norm": 1.3102650044903212, + "learning_rate": 1.9742230859393432e-05, + "loss": 1.6954, + "step": 283 + }, + { + "epoch": 0.08210465452442903, + "grad_norm": 2.693221641429579, + "learning_rate": 1.974015695272077e-05, + "loss": 1.6296, + "step": 284 + }, + { + "epoch": 0.0823937554206418, + "grad_norm": 1.3351424022555927, + "learning_rate": 1.9738074846344427e-05, + "loss": 1.6594, + "step": 285 + }, + { + "epoch": 0.08268285631685458, + "grad_norm": 1.2766910430053877, + "learning_rate": 1.973598454201722e-05, + "loss": 1.6983, + "step": 286 + }, + { + "epoch": 0.08297195721306735, + "grad_norm": 1.4668406259400848, + "learning_rate": 1.973388604149886e-05, + "loss": 1.6698, + "step": 287 + }, + { + "epoch": 0.08326105810928014, + "grad_norm": 1.3104546778761357, + "learning_rate": 1.9731779346555957e-05, + "loss": 1.6723, + "step": 288 + }, + { + "epoch": 0.08355015900549291, + "grad_norm": 1.531005942756331, + "learning_rate": 1.9729664458962024e-05, + "loss": 1.7863, + "step": 289 + }, + { + "epoch": 0.0838392599017057, + "grad_norm": 1.3653225817314174, + "learning_rate": 1.9727541380497465e-05, + "loss": 1.7107, + "step": 290 + }, + { + "epoch": 0.08412836079791847, + "grad_norm": 1.3892591267410797, + "learning_rate": 1.972541011294959e-05, + "loss": 1.6403, + "step": 291 + }, + { + "epoch": 0.08441746169413125, + "grad_norm": 1.4118648874854935, + "learning_rate": 1.972327065811259e-05, + "loss": 1.6871, + "step": 292 + }, + { + "epoch": 0.08470656259034404, + "grad_norm": 1.4198103761024161, + "learning_rate": 1.972112301778756e-05, + "loss": 1.7095, + "step": 293 + }, + { + "epoch": 0.08499566348655681, + "grad_norm": 1.3896634687986935, + "learning_rate": 1.9718967193782475e-05, + "loss": 1.6347, + "step": 294 + }, + { + "epoch": 0.08528476438276959, + "grad_norm": 1.3431516951928908, + "learning_rate": 1.971680318791221e-05, + "loss": 1.7874, + "step": 295 + }, + { + "epoch": 0.08557386527898236, + "grad_norm": 1.3768126854675709, + "learning_rate": 1.9714631001998517e-05, + "loss": 1.6698, + "step": 296 + }, + { + "epoch": 0.08586296617519515, + "grad_norm": 1.452732246792645, + "learning_rate": 1.9712450637870048e-05, + "loss": 1.7896, + "step": 297 + }, + { + "epoch": 0.08615206707140792, + "grad_norm": 1.3268182615670008, + "learning_rate": 1.9710262097362323e-05, + "loss": 1.7162, + "step": 298 + }, + { + "epoch": 0.0864411679676207, + "grad_norm": 1.3451587005617647, + "learning_rate": 1.9708065382317763e-05, + "loss": 1.6132, + "step": 299 + }, + { + "epoch": 0.08673026886383348, + "grad_norm": 1.3567905514483718, + "learning_rate": 1.9705860494585653e-05, + "loss": 1.7181, + "step": 300 + }, + { + "epoch": 0.08701936976004626, + "grad_norm": 1.4178400258624095, + "learning_rate": 1.9703647436022174e-05, + "loss": 1.7197, + "step": 301 + }, + { + "epoch": 0.08730847065625903, + "grad_norm": 1.3176335169067535, + "learning_rate": 1.970142620849038e-05, + "loss": 1.6433, + "step": 302 + }, + { + "epoch": 0.08759757155247182, + "grad_norm": 1.3848318848405352, + "learning_rate": 1.9699196813860188e-05, + "loss": 1.7975, + "step": 303 + }, + { + "epoch": 0.08788667244868459, + "grad_norm": 1.4792413426843254, + "learning_rate": 1.9696959254008416e-05, + "loss": 1.7221, + "step": 304 + }, + { + "epoch": 0.08817577334489737, + "grad_norm": 1.3289471852066919, + "learning_rate": 1.9694713530818738e-05, + "loss": 1.7631, + "step": 305 + }, + { + "epoch": 0.08846487424111014, + "grad_norm": 1.4455763431259683, + "learning_rate": 1.9692459646181706e-05, + "loss": 1.758, + "step": 306 + }, + { + "epoch": 0.08875397513732293, + "grad_norm": 1.4156278432006244, + "learning_rate": 1.969019760199474e-05, + "loss": 1.7106, + "step": 307 + }, + { + "epoch": 0.0890430760335357, + "grad_norm": 1.335012985289393, + "learning_rate": 1.9687927400162134e-05, + "loss": 1.6303, + "step": 308 + }, + { + "epoch": 0.08933217692974849, + "grad_norm": 1.3515063736435737, + "learning_rate": 1.9685649042595045e-05, + "loss": 1.6907, + "step": 309 + }, + { + "epoch": 0.08962127782596126, + "grad_norm": 1.2906659764446389, + "learning_rate": 1.96833625312115e-05, + "loss": 1.7601, + "step": 310 + }, + { + "epoch": 0.08991037872217404, + "grad_norm": 1.4128720604920837, + "learning_rate": 1.968106786793638e-05, + "loss": 1.685, + "step": 311 + }, + { + "epoch": 0.09019947961838681, + "grad_norm": 1.3417671224975973, + "learning_rate": 1.967876505470144e-05, + "loss": 1.8098, + "step": 312 + }, + { + "epoch": 0.0904885805145996, + "grad_norm": 1.2524980838386586, + "learning_rate": 1.9676454093445293e-05, + "loss": 1.7347, + "step": 313 + }, + { + "epoch": 0.09077768141081237, + "grad_norm": 1.2903112596909276, + "learning_rate": 1.9674134986113407e-05, + "loss": 1.7183, + "step": 314 + }, + { + "epoch": 0.09106678230702515, + "grad_norm": 1.315346772447713, + "learning_rate": 1.9671807734658122e-05, + "loss": 1.666, + "step": 315 + }, + { + "epoch": 0.09135588320323793, + "grad_norm": 1.2815003343295048, + "learning_rate": 1.966947234103861e-05, + "loss": 1.6706, + "step": 316 + }, + { + "epoch": 0.09164498409945071, + "grad_norm": 1.3032057492518452, + "learning_rate": 1.9667128807220918e-05, + "loss": 1.6852, + "step": 317 + }, + { + "epoch": 0.09193408499566348, + "grad_norm": 1.399838923906023, + "learning_rate": 1.9664777135177936e-05, + "loss": 1.6786, + "step": 318 + }, + { + "epoch": 0.09222318589187627, + "grad_norm": 1.3183521495611081, + "learning_rate": 1.9662417326889408e-05, + "loss": 1.7656, + "step": 319 + }, + { + "epoch": 0.09251228678808904, + "grad_norm": 1.2919964291273878, + "learning_rate": 1.966004938434193e-05, + "loss": 1.7994, + "step": 320 + }, + { + "epoch": 0.09280138768430182, + "grad_norm": 1.331071824643648, + "learning_rate": 1.965767330952894e-05, + "loss": 1.7315, + "step": 321 + }, + { + "epoch": 0.0930904885805146, + "grad_norm": 1.3127780153887916, + "learning_rate": 1.9655289104450724e-05, + "loss": 1.58, + "step": 322 + }, + { + "epoch": 0.09337958947672738, + "grad_norm": 1.427592788785929, + "learning_rate": 1.9652896771114416e-05, + "loss": 1.6926, + "step": 323 + }, + { + "epoch": 0.09366869037294015, + "grad_norm": 1.3536283403103335, + "learning_rate": 1.9650496311533985e-05, + "loss": 1.6208, + "step": 324 + }, + { + "epoch": 0.09395779126915293, + "grad_norm": 1.3049249531971474, + "learning_rate": 1.9648087727730254e-05, + "loss": 1.7007, + "step": 325 + }, + { + "epoch": 0.0942468921653657, + "grad_norm": 8.074839924042823, + "learning_rate": 1.9645671021730876e-05, + "loss": 1.7401, + "step": 326 + }, + { + "epoch": 0.09453599306157849, + "grad_norm": 1.689958269781092, + "learning_rate": 1.9643246195570337e-05, + "loss": 1.7612, + "step": 327 + }, + { + "epoch": 0.09482509395779128, + "grad_norm": 1.3759192019092044, + "learning_rate": 1.9640813251289968e-05, + "loss": 1.6256, + "step": 328 + }, + { + "epoch": 0.09511419485400405, + "grad_norm": 1.2852662388113483, + "learning_rate": 1.9638372190937938e-05, + "loss": 1.5961, + "step": 329 + }, + { + "epoch": 0.09540329575021683, + "grad_norm": 1.3397296142416237, + "learning_rate": 1.963592301656923e-05, + "loss": 1.7228, + "step": 330 + }, + { + "epoch": 0.0956923966464296, + "grad_norm": 1.3050235554989664, + "learning_rate": 1.963346573024568e-05, + "loss": 1.631, + "step": 331 + }, + { + "epoch": 0.09598149754264239, + "grad_norm": 1.2687361679426568, + "learning_rate": 1.9631000334035936e-05, + "loss": 1.727, + "step": 332 + }, + { + "epoch": 0.09627059843885516, + "grad_norm": 1.4434478962876411, + "learning_rate": 1.962852683001548e-05, + "loss": 1.6118, + "step": 333 + }, + { + "epoch": 0.09655969933506794, + "grad_norm": 1.3835831592110455, + "learning_rate": 1.9626045220266626e-05, + "loss": 1.7184, + "step": 334 + }, + { + "epoch": 0.09684880023128072, + "grad_norm": 1.285345165652217, + "learning_rate": 1.96235555068785e-05, + "loss": 1.7396, + "step": 335 + }, + { + "epoch": 0.0971379011274935, + "grad_norm": 1.2741876410509978, + "learning_rate": 1.9621057691947052e-05, + "loss": 1.6827, + "step": 336 + }, + { + "epoch": 0.09742700202370627, + "grad_norm": 1.3414110861817643, + "learning_rate": 1.961855177757506e-05, + "loss": 1.7649, + "step": 337 + }, + { + "epoch": 0.09771610291991906, + "grad_norm": 1.3155148016487777, + "learning_rate": 1.9616037765872115e-05, + "loss": 1.6339, + "step": 338 + }, + { + "epoch": 0.09800520381613183, + "grad_norm": 1.3422815166211728, + "learning_rate": 1.9613515658954627e-05, + "loss": 1.6248, + "step": 339 + }, + { + "epoch": 0.09829430471234461, + "grad_norm": 1.2928851291446535, + "learning_rate": 1.9610985458945813e-05, + "loss": 1.6883, + "step": 340 + }, + { + "epoch": 0.09858340560855738, + "grad_norm": 1.310011271967797, + "learning_rate": 1.9608447167975716e-05, + "loss": 1.5808, + "step": 341 + }, + { + "epoch": 0.09887250650477017, + "grad_norm": 1.3065951865270455, + "learning_rate": 1.960590078818118e-05, + "loss": 1.6558, + "step": 342 + }, + { + "epoch": 0.09916160740098294, + "grad_norm": 1.3349828850162693, + "learning_rate": 1.960334632170587e-05, + "loss": 1.7289, + "step": 343 + }, + { + "epoch": 0.09945070829719573, + "grad_norm": 1.2943452424584643, + "learning_rate": 1.960078377070024e-05, + "loss": 1.613, + "step": 344 + }, + { + "epoch": 0.0997398091934085, + "grad_norm": 1.4861917307646695, + "learning_rate": 1.9598213137321568e-05, + "loss": 1.7105, + "step": 345 + }, + { + "epoch": 0.10002891008962128, + "grad_norm": 1.3722730686933504, + "learning_rate": 1.959563442373393e-05, + "loss": 1.7553, + "step": 346 + }, + { + "epoch": 0.10031801098583405, + "grad_norm": 1.3453966326217088, + "learning_rate": 1.95930476321082e-05, + "loss": 1.737, + "step": 347 + }, + { + "epoch": 0.10060711188204684, + "grad_norm": 1.3523784458893544, + "learning_rate": 1.9590452764622056e-05, + "loss": 1.6319, + "step": 348 + }, + { + "epoch": 0.10089621277825961, + "grad_norm": 1.3502487223737636, + "learning_rate": 1.9587849823459976e-05, + "loss": 1.6941, + "step": 349 + }, + { + "epoch": 0.1011853136744724, + "grad_norm": 1.398518128752155, + "learning_rate": 1.9585238810813233e-05, + "loss": 1.6475, + "step": 350 + }, + { + "epoch": 0.10147441457068516, + "grad_norm": 1.200679760577495, + "learning_rate": 1.9582619728879898e-05, + "loss": 1.7337, + "step": 351 + }, + { + "epoch": 0.10176351546689795, + "grad_norm": 1.3585311474803605, + "learning_rate": 1.9579992579864827e-05, + "loss": 1.7192, + "step": 352 + }, + { + "epoch": 0.10205261636311072, + "grad_norm": 1.4501829885257667, + "learning_rate": 1.9577357365979673e-05, + "loss": 1.7227, + "step": 353 + }, + { + "epoch": 0.1023417172593235, + "grad_norm": 1.2783896023412555, + "learning_rate": 1.9574714089442884e-05, + "loss": 1.7873, + "step": 354 + }, + { + "epoch": 0.10263081815553628, + "grad_norm": 1.2457865420249044, + "learning_rate": 1.9572062752479684e-05, + "loss": 1.6078, + "step": 355 + }, + { + "epoch": 0.10291991905174906, + "grad_norm": 1.298867064640292, + "learning_rate": 1.956940335732209e-05, + "loss": 1.7645, + "step": 356 + }, + { + "epoch": 0.10320901994796183, + "grad_norm": 1.2599727613590195, + "learning_rate": 1.95667359062089e-05, + "loss": 1.6698, + "step": 357 + }, + { + "epoch": 0.10349812084417462, + "grad_norm": 1.2508214829214046, + "learning_rate": 1.9564060401385697e-05, + "loss": 1.6279, + "step": 358 + }, + { + "epoch": 0.10378722174038739, + "grad_norm": 1.1832707796953907, + "learning_rate": 1.956137684510484e-05, + "loss": 1.5799, + "step": 359 + }, + { + "epoch": 0.10407632263660017, + "grad_norm": 1.3001383458399352, + "learning_rate": 1.9558685239625467e-05, + "loss": 1.6116, + "step": 360 + }, + { + "epoch": 0.10436542353281295, + "grad_norm": 1.2428470322542244, + "learning_rate": 1.9555985587213495e-05, + "loss": 1.6716, + "step": 361 + }, + { + "epoch": 0.10465452442902573, + "grad_norm": 1.3093773655524557, + "learning_rate": 1.955327789014161e-05, + "loss": 1.6915, + "step": 362 + }, + { + "epoch": 0.1049436253252385, + "grad_norm": 1.383021031949293, + "learning_rate": 1.9550562150689277e-05, + "loss": 1.7317, + "step": 363 + }, + { + "epoch": 0.10523272622145129, + "grad_norm": 1.329573981449402, + "learning_rate": 1.9547838371142726e-05, + "loss": 1.5913, + "step": 364 + }, + { + "epoch": 0.10552182711766407, + "grad_norm": 1.4166951368370655, + "learning_rate": 1.9545106553794962e-05, + "loss": 1.6004, + "step": 365 + }, + { + "epoch": 0.10581092801387684, + "grad_norm": 1.3377620749480776, + "learning_rate": 1.9542366700945746e-05, + "loss": 1.7907, + "step": 366 + }, + { + "epoch": 0.10610002891008963, + "grad_norm": 1.3388694333443414, + "learning_rate": 1.953961881490161e-05, + "loss": 1.6979, + "step": 367 + }, + { + "epoch": 0.1063891298063024, + "grad_norm": 1.3581005844418126, + "learning_rate": 1.9536862897975852e-05, + "loss": 1.6602, + "step": 368 + }, + { + "epoch": 0.10667823070251518, + "grad_norm": 1.3242684005167475, + "learning_rate": 1.9534098952488527e-05, + "loss": 1.7131, + "step": 369 + }, + { + "epoch": 0.10696733159872795, + "grad_norm": 1.2765833725967193, + "learning_rate": 1.9531326980766444e-05, + "loss": 1.7917, + "step": 370 + }, + { + "epoch": 0.10725643249494074, + "grad_norm": 1.2957317184999853, + "learning_rate": 1.952854698514318e-05, + "loss": 1.6569, + "step": 371 + }, + { + "epoch": 0.10754553339115351, + "grad_norm": 1.3050908294406351, + "learning_rate": 1.952575896795906e-05, + "loss": 1.5815, + "step": 372 + }, + { + "epoch": 0.1078346342873663, + "grad_norm": 1.4007267378768093, + "learning_rate": 1.952296293156116e-05, + "loss": 1.6437, + "step": 373 + }, + { + "epoch": 0.10812373518357907, + "grad_norm": 1.288112071905893, + "learning_rate": 1.952015887830331e-05, + "loss": 1.7148, + "step": 374 + }, + { + "epoch": 0.10841283607979185, + "grad_norm": 1.4132340648150645, + "learning_rate": 1.9517346810546094e-05, + "loss": 1.7236, + "step": 375 + }, + { + "epoch": 0.10870193697600462, + "grad_norm": 1.3520649123502502, + "learning_rate": 1.951452673065683e-05, + "loss": 1.7027, + "step": 376 + }, + { + "epoch": 0.10899103787221741, + "grad_norm": 1.344289195938231, + "learning_rate": 1.951169864100959e-05, + "loss": 1.6753, + "step": 377 + }, + { + "epoch": 0.10928013876843018, + "grad_norm": 1.2727655897790644, + "learning_rate": 1.950886254398519e-05, + "loss": 1.7595, + "step": 378 + }, + { + "epoch": 0.10956923966464296, + "grad_norm": 1.3488163579726409, + "learning_rate": 1.9506018441971186e-05, + "loss": 1.6687, + "step": 379 + }, + { + "epoch": 0.10985834056085574, + "grad_norm": 1.3474523993403258, + "learning_rate": 1.9503166337361863e-05, + "loss": 1.6676, + "step": 380 + }, + { + "epoch": 0.11014744145706852, + "grad_norm": 1.2758765009101956, + "learning_rate": 1.9500306232558263e-05, + "loss": 1.6571, + "step": 381 + }, + { + "epoch": 0.11043654235328129, + "grad_norm": 1.2196542542467945, + "learning_rate": 1.9497438129968144e-05, + "loss": 1.6551, + "step": 382 + }, + { + "epoch": 0.11072564324949408, + "grad_norm": 1.301338171513902, + "learning_rate": 1.9494562032006e-05, + "loss": 1.6548, + "step": 383 + }, + { + "epoch": 0.11101474414570685, + "grad_norm": 1.3203283874481997, + "learning_rate": 1.9491677941093075e-05, + "loss": 1.7338, + "step": 384 + }, + { + "epoch": 0.11130384504191963, + "grad_norm": 1.3024684784620577, + "learning_rate": 1.9488785859657314e-05, + "loss": 1.5528, + "step": 385 + }, + { + "epoch": 0.1115929459381324, + "grad_norm": 1.2625026147962017, + "learning_rate": 1.9485885790133402e-05, + "loss": 1.6967, + "step": 386 + }, + { + "epoch": 0.11188204683434519, + "grad_norm": 1.49068356270609, + "learning_rate": 1.9482977734962753e-05, + "loss": 1.7096, + "step": 387 + }, + { + "epoch": 0.11217114773055796, + "grad_norm": 1.3130185709588313, + "learning_rate": 1.9480061696593502e-05, + "loss": 1.7174, + "step": 388 + }, + { + "epoch": 0.11246024862677075, + "grad_norm": 1.2742915647508892, + "learning_rate": 1.9477137677480493e-05, + "loss": 1.6696, + "step": 389 + }, + { + "epoch": 0.11274934952298352, + "grad_norm": 1.3627324489936457, + "learning_rate": 1.9474205680085302e-05, + "loss": 1.6542, + "step": 390 + }, + { + "epoch": 0.1130384504191963, + "grad_norm": 1.3437349437643114, + "learning_rate": 1.947126570687622e-05, + "loss": 1.6747, + "step": 391 + }, + { + "epoch": 0.11332755131540907, + "grad_norm": 1.2678997121287945, + "learning_rate": 1.9468317760328244e-05, + "loss": 1.6918, + "step": 392 + }, + { + "epoch": 0.11361665221162186, + "grad_norm": 1.2929245484043537, + "learning_rate": 1.9465361842923087e-05, + "loss": 1.7217, + "step": 393 + }, + { + "epoch": 0.11390575310783463, + "grad_norm": 1.2499932038294317, + "learning_rate": 1.9462397957149182e-05, + "loss": 1.6128, + "step": 394 + }, + { + "epoch": 0.11419485400404741, + "grad_norm": 1.221557019735438, + "learning_rate": 1.9459426105501653e-05, + "loss": 1.6467, + "step": 395 + }, + { + "epoch": 0.11448395490026018, + "grad_norm": 1.2693539163839065, + "learning_rate": 1.945644629048234e-05, + "loss": 1.7283, + "step": 396 + }, + { + "epoch": 0.11477305579647297, + "grad_norm": 1.21274399309353, + "learning_rate": 1.9453458514599794e-05, + "loss": 1.7008, + "step": 397 + }, + { + "epoch": 0.11506215669268574, + "grad_norm": 1.2218147838259905, + "learning_rate": 1.945046278036925e-05, + "loss": 1.7437, + "step": 398 + }, + { + "epoch": 0.11535125758889853, + "grad_norm": 1.2968685896469103, + "learning_rate": 1.9447459090312656e-05, + "loss": 1.6983, + "step": 399 + }, + { + "epoch": 0.1156403584851113, + "grad_norm": 1.2334993392446711, + "learning_rate": 1.9444447446958658e-05, + "loss": 1.7061, + "step": 400 + }, + { + "epoch": 0.11592945938132408, + "grad_norm": 1.2984129611641557, + "learning_rate": 1.9441427852842584e-05, + "loss": 1.663, + "step": 401 + }, + { + "epoch": 0.11621856027753687, + "grad_norm": 1.2475965365995239, + "learning_rate": 1.943840031050647e-05, + "loss": 1.6427, + "step": 402 + }, + { + "epoch": 0.11650766117374964, + "grad_norm": 1.2694332854835886, + "learning_rate": 1.943536482249904e-05, + "loss": 1.7047, + "step": 403 + }, + { + "epoch": 0.11679676206996242, + "grad_norm": 1.2646968431962804, + "learning_rate": 1.94323213913757e-05, + "loss": 1.6885, + "step": 404 + }, + { + "epoch": 0.1170858629661752, + "grad_norm": 1.2293159012021218, + "learning_rate": 1.9429270019698553e-05, + "loss": 1.757, + "step": 405 + }, + { + "epoch": 0.11737496386238798, + "grad_norm": 1.2484085036943657, + "learning_rate": 1.9426210710036374e-05, + "loss": 1.6741, + "step": 406 + }, + { + "epoch": 0.11766406475860075, + "grad_norm": 1.43839701132394, + "learning_rate": 1.9423143464964633e-05, + "loss": 1.6768, + "step": 407 + }, + { + "epoch": 0.11795316565481354, + "grad_norm": 1.292908155403034, + "learning_rate": 1.9420068287065476e-05, + "loss": 1.6126, + "step": 408 + }, + { + "epoch": 0.1182422665510263, + "grad_norm": 1.370089960450461, + "learning_rate": 1.9416985178927724e-05, + "loss": 1.731, + "step": 409 + }, + { + "epoch": 0.11853136744723909, + "grad_norm": 1.3262176752311388, + "learning_rate": 1.941389414314688e-05, + "loss": 1.6579, + "step": 410 + }, + { + "epoch": 0.11882046834345186, + "grad_norm": 1.339305591546624, + "learning_rate": 1.9410795182325113e-05, + "loss": 1.7495, + "step": 411 + }, + { + "epoch": 0.11910956923966465, + "grad_norm": 1.341821261680916, + "learning_rate": 1.940768829907127e-05, + "loss": 1.7005, + "step": 412 + }, + { + "epoch": 0.11939867013587742, + "grad_norm": 1.276085151807024, + "learning_rate": 1.9404573496000867e-05, + "loss": 1.6043, + "step": 413 + }, + { + "epoch": 0.1196877710320902, + "grad_norm": 1.302570146214278, + "learning_rate": 1.940145077573608e-05, + "loss": 1.6722, + "step": 414 + }, + { + "epoch": 0.11997687192830297, + "grad_norm": 1.2582811301580585, + "learning_rate": 1.9398320140905765e-05, + "loss": 1.6383, + "step": 415 + }, + { + "epoch": 0.12026597282451576, + "grad_norm": 1.258733310784144, + "learning_rate": 1.9395181594145428e-05, + "loss": 1.5806, + "step": 416 + }, + { + "epoch": 0.12055507372072853, + "grad_norm": 1.2988890119500425, + "learning_rate": 1.9392035138097235e-05, + "loss": 1.7048, + "step": 417 + }, + { + "epoch": 0.12084417461694132, + "grad_norm": 1.29122939937491, + "learning_rate": 1.938888077541002e-05, + "loss": 1.5844, + "step": 418 + }, + { + "epoch": 0.12113327551315409, + "grad_norm": 1.2756402426612268, + "learning_rate": 1.9385718508739263e-05, + "loss": 1.6935, + "step": 419 + }, + { + "epoch": 0.12142237640936687, + "grad_norm": 1.3542547094512076, + "learning_rate": 1.9382548340747105e-05, + "loss": 1.6781, + "step": 420 + }, + { + "epoch": 0.12171147730557964, + "grad_norm": 1.3120014515081435, + "learning_rate": 1.937937027410234e-05, + "loss": 1.6601, + "step": 421 + }, + { + "epoch": 0.12200057820179243, + "grad_norm": 1.2955651895370563, + "learning_rate": 1.9376184311480404e-05, + "loss": 1.59, + "step": 422 + }, + { + "epoch": 0.1222896790980052, + "grad_norm": 1.3192882851086605, + "learning_rate": 1.9372990455563384e-05, + "loss": 1.6291, + "step": 423 + }, + { + "epoch": 0.12257877999421798, + "grad_norm": 1.4080370752052034, + "learning_rate": 1.9369788709040014e-05, + "loss": 1.5843, + "step": 424 + }, + { + "epoch": 0.12286788089043076, + "grad_norm": 1.309890861391924, + "learning_rate": 1.936657907460567e-05, + "loss": 1.7576, + "step": 425 + }, + { + "epoch": 0.12315698178664354, + "grad_norm": 1.3296661949737203, + "learning_rate": 1.936336155496236e-05, + "loss": 1.6412, + "step": 426 + }, + { + "epoch": 0.12344608268285631, + "grad_norm": 1.2137542022109604, + "learning_rate": 1.9360136152818746e-05, + "loss": 1.6677, + "step": 427 + }, + { + "epoch": 0.1237351835790691, + "grad_norm": 1.304478996513997, + "learning_rate": 1.9356902870890114e-05, + "loss": 1.6157, + "step": 428 + }, + { + "epoch": 0.12402428447528187, + "grad_norm": 1.2362598408529617, + "learning_rate": 1.9353661711898382e-05, + "loss": 1.6493, + "step": 429 + }, + { + "epoch": 0.12431338537149465, + "grad_norm": 1.3868990501411094, + "learning_rate": 1.9350412678572114e-05, + "loss": 1.7838, + "step": 430 + }, + { + "epoch": 0.12460248626770742, + "grad_norm": 1.3811144656508765, + "learning_rate": 1.934715577364648e-05, + "loss": 1.6164, + "step": 431 + }, + { + "epoch": 0.12489158716392021, + "grad_norm": 1.1886802269263785, + "learning_rate": 1.93438909998633e-05, + "loss": 1.6917, + "step": 432 + }, + { + "epoch": 0.125180688060133, + "grad_norm": 1.373861158324828, + "learning_rate": 1.9340618359971003e-05, + "loss": 1.7273, + "step": 433 + }, + { + "epoch": 0.12546978895634575, + "grad_norm": 1.360887411910304, + "learning_rate": 1.9337337856724647e-05, + "loss": 1.6129, + "step": 434 + }, + { + "epoch": 0.12575888985255854, + "grad_norm": 1.2310796622388105, + "learning_rate": 1.9334049492885906e-05, + "loss": 1.686, + "step": 435 + }, + { + "epoch": 0.12604799074877132, + "grad_norm": 1.2150119003782842, + "learning_rate": 1.9330753271223073e-05, + "loss": 1.7097, + "step": 436 + }, + { + "epoch": 0.1263370916449841, + "grad_norm": 1.3137735986745411, + "learning_rate": 1.932744919451106e-05, + "loss": 1.7312, + "step": 437 + }, + { + "epoch": 0.1266261925411969, + "grad_norm": 1.19581010090534, + "learning_rate": 1.9324137265531383e-05, + "loss": 1.7073, + "step": 438 + }, + { + "epoch": 0.12691529343740965, + "grad_norm": 1.1864958252257658, + "learning_rate": 1.9320817487072174e-05, + "loss": 1.6634, + "step": 439 + }, + { + "epoch": 0.12720439433362243, + "grad_norm": 1.2519130079679632, + "learning_rate": 1.9317489861928173e-05, + "loss": 1.6686, + "step": 440 + }, + { + "epoch": 0.12749349522983522, + "grad_norm": 1.2098081719210536, + "learning_rate": 1.9314154392900728e-05, + "loss": 1.697, + "step": 441 + }, + { + "epoch": 0.127782596126048, + "grad_norm": 1.297112805486722, + "learning_rate": 1.9310811082797784e-05, + "loss": 1.5979, + "step": 442 + }, + { + "epoch": 0.12807169702226076, + "grad_norm": 1.3006902970607843, + "learning_rate": 1.930745993443389e-05, + "loss": 1.5987, + "step": 443 + }, + { + "epoch": 0.12836079791847355, + "grad_norm": 1.3888684660127295, + "learning_rate": 1.9304100950630197e-05, + "loss": 1.6047, + "step": 444 + }, + { + "epoch": 0.12864989881468633, + "grad_norm": 1.3325945840074427, + "learning_rate": 1.9300734134214443e-05, + "loss": 1.6813, + "step": 445 + }, + { + "epoch": 0.12893899971089912, + "grad_norm": 1.3053169239838074, + "learning_rate": 1.9297359488020974e-05, + "loss": 1.8416, + "step": 446 + }, + { + "epoch": 0.12922810060711187, + "grad_norm": 1.1934876374596244, + "learning_rate": 1.9293977014890716e-05, + "loss": 1.7005, + "step": 447 + }, + { + "epoch": 0.12951720150332466, + "grad_norm": 1.2640391776060138, + "learning_rate": 1.929058671767119e-05, + "loss": 1.6578, + "step": 448 + }, + { + "epoch": 0.12980630239953744, + "grad_norm": 1.2646711203792806, + "learning_rate": 1.9287188599216492e-05, + "loss": 1.6699, + "step": 449 + }, + { + "epoch": 0.13009540329575023, + "grad_norm": 1.2320011875752572, + "learning_rate": 1.928378266238732e-05, + "loss": 1.6525, + "step": 450 + }, + { + "epoch": 0.13038450419196299, + "grad_norm": 1.34573408732214, + "learning_rate": 1.9280368910050943e-05, + "loss": 1.7015, + "step": 451 + }, + { + "epoch": 0.13067360508817577, + "grad_norm": 1.2353671099343633, + "learning_rate": 1.9276947345081213e-05, + "loss": 1.6397, + "step": 452 + }, + { + "epoch": 0.13096270598438856, + "grad_norm": 1.2197831661100311, + "learning_rate": 1.9273517970358557e-05, + "loss": 1.6836, + "step": 453 + }, + { + "epoch": 0.13125180688060134, + "grad_norm": 1.328367741401711, + "learning_rate": 1.9270080788769978e-05, + "loss": 1.7178, + "step": 454 + }, + { + "epoch": 0.1315409077768141, + "grad_norm": 1.2668567068996344, + "learning_rate": 1.9266635803209047e-05, + "loss": 1.7255, + "step": 455 + }, + { + "epoch": 0.13183000867302688, + "grad_norm": 1.2322035718739746, + "learning_rate": 1.9263183016575912e-05, + "loss": 1.612, + "step": 456 + }, + { + "epoch": 0.13211910956923967, + "grad_norm": 1.244606536596135, + "learning_rate": 1.9259722431777286e-05, + "loss": 1.7994, + "step": 457 + }, + { + "epoch": 0.13240821046545245, + "grad_norm": 1.2384126698619529, + "learning_rate": 1.9256254051726443e-05, + "loss": 1.6401, + "step": 458 + }, + { + "epoch": 0.1326973113616652, + "grad_norm": 1.2780953635907069, + "learning_rate": 1.925277787934322e-05, + "loss": 1.5826, + "step": 459 + }, + { + "epoch": 0.132986412257878, + "grad_norm": 1.2798396535904, + "learning_rate": 1.9249293917554017e-05, + "loss": 1.674, + "step": 460 + }, + { + "epoch": 0.13327551315409078, + "grad_norm": 1.2799203698146353, + "learning_rate": 1.924580216929179e-05, + "loss": 1.6331, + "step": 461 + }, + { + "epoch": 0.13356461405030357, + "grad_norm": 1.2111633406900368, + "learning_rate": 1.9242302637496057e-05, + "loss": 1.6682, + "step": 462 + }, + { + "epoch": 0.13385371494651632, + "grad_norm": 1.2112859487688281, + "learning_rate": 1.9238795325112867e-05, + "loss": 1.6484, + "step": 463 + }, + { + "epoch": 0.1341428158427291, + "grad_norm": 1.288135140665184, + "learning_rate": 1.9235280235094852e-05, + "loss": 1.626, + "step": 464 + }, + { + "epoch": 0.1344319167389419, + "grad_norm": 1.308513915152716, + "learning_rate": 1.923175737040116e-05, + "loss": 1.6597, + "step": 465 + }, + { + "epoch": 0.13472101763515468, + "grad_norm": 1.3262569041157928, + "learning_rate": 1.9228226733997503e-05, + "loss": 1.7642, + "step": 466 + }, + { + "epoch": 0.13501011853136743, + "grad_norm": 1.3424290380477188, + "learning_rate": 1.9224688328856127e-05, + "loss": 1.7465, + "step": 467 + }, + { + "epoch": 0.13529921942758022, + "grad_norm": 1.2651279914803906, + "learning_rate": 1.9221142157955825e-05, + "loss": 1.6126, + "step": 468 + }, + { + "epoch": 0.135588320323793, + "grad_norm": 1.3558655043001309, + "learning_rate": 1.921758822428192e-05, + "loss": 1.7556, + "step": 469 + }, + { + "epoch": 0.1358774212200058, + "grad_norm": 1.402001055732072, + "learning_rate": 1.921402653082628e-05, + "loss": 1.7992, + "step": 470 + }, + { + "epoch": 0.13616652211621855, + "grad_norm": 1.3182689558852307, + "learning_rate": 1.9210457080587285e-05, + "loss": 1.6262, + "step": 471 + }, + { + "epoch": 0.13645562301243133, + "grad_norm": 1.2723760551262082, + "learning_rate": 1.9206879876569877e-05, + "loss": 1.7064, + "step": 472 + }, + { + "epoch": 0.13674472390864412, + "grad_norm": 1.2954400935028445, + "learning_rate": 1.92032949217855e-05, + "loss": 1.6374, + "step": 473 + }, + { + "epoch": 0.1370338248048569, + "grad_norm": 1.2402948267042866, + "learning_rate": 1.9199702219252128e-05, + "loss": 1.5951, + "step": 474 + }, + { + "epoch": 0.1373229257010697, + "grad_norm": 1.2739630063945762, + "learning_rate": 1.919610177199426e-05, + "loss": 1.7167, + "step": 475 + }, + { + "epoch": 0.13761202659728244, + "grad_norm": 1.4113213475938031, + "learning_rate": 1.9192493583042922e-05, + "loss": 1.5751, + "step": 476 + }, + { + "epoch": 0.13790112749349523, + "grad_norm": 1.2872821986905296, + "learning_rate": 1.9188877655435645e-05, + "loss": 1.65, + "step": 477 + }, + { + "epoch": 0.13819022838970801, + "grad_norm": 1.2480655610937714, + "learning_rate": 1.918525399221648e-05, + "loss": 1.6882, + "step": 478 + }, + { + "epoch": 0.1384793292859208, + "grad_norm": 1.2681694886992483, + "learning_rate": 1.9181622596435993e-05, + "loss": 1.7528, + "step": 479 + }, + { + "epoch": 0.13876843018213356, + "grad_norm": 1.226404564427387, + "learning_rate": 1.917798347115125e-05, + "loss": 1.5918, + "step": 480 + }, + { + "epoch": 0.13905753107834634, + "grad_norm": 1.2687390154564733, + "learning_rate": 1.9174336619425842e-05, + "loss": 1.7533, + "step": 481 + }, + { + "epoch": 0.13934663197455913, + "grad_norm": 1.2254053559809934, + "learning_rate": 1.9170682044329845e-05, + "loss": 1.5968, + "step": 482 + }, + { + "epoch": 0.1396357328707719, + "grad_norm": 1.2732869875336201, + "learning_rate": 1.9167019748939847e-05, + "loss": 1.6944, + "step": 483 + }, + { + "epoch": 0.13992483376698467, + "grad_norm": 1.2541243190490785, + "learning_rate": 1.9163349736338935e-05, + "loss": 1.6237, + "step": 484 + }, + { + "epoch": 0.14021393466319745, + "grad_norm": 1.3291823171118335, + "learning_rate": 1.915967200961669e-05, + "loss": 1.5639, + "step": 485 + }, + { + "epoch": 0.14050303555941024, + "grad_norm": 1.2528167731427482, + "learning_rate": 1.9155986571869184e-05, + "loss": 1.7458, + "step": 486 + }, + { + "epoch": 0.14079213645562302, + "grad_norm": 1.2107095174484985, + "learning_rate": 1.9152293426198997e-05, + "loss": 1.6838, + "step": 487 + }, + { + "epoch": 0.14108123735183578, + "grad_norm": 1.2070210238330805, + "learning_rate": 1.9148592575715175e-05, + "loss": 1.6957, + "step": 488 + }, + { + "epoch": 0.14137033824804857, + "grad_norm": 1.3501466427051918, + "learning_rate": 1.9144884023533262e-05, + "loss": 1.6848, + "step": 489 + }, + { + "epoch": 0.14165943914426135, + "grad_norm": 1.289837942023481, + "learning_rate": 1.914116777277529e-05, + "loss": 1.6398, + "step": 490 + }, + { + "epoch": 0.14194854004047414, + "grad_norm": 1.2072886210991958, + "learning_rate": 1.9137443826569758e-05, + "loss": 1.6803, + "step": 491 + }, + { + "epoch": 0.1422376409366869, + "grad_norm": 1.2261256594899907, + "learning_rate": 1.9133712188051664e-05, + "loss": 1.8143, + "step": 492 + }, + { + "epoch": 0.14252674183289968, + "grad_norm": 1.191555412915695, + "learning_rate": 1.912997286036246e-05, + "loss": 1.7273, + "step": 493 + }, + { + "epoch": 0.14281584272911246, + "grad_norm": 1.2668769555610104, + "learning_rate": 1.9126225846650083e-05, + "loss": 1.6729, + "step": 494 + }, + { + "epoch": 0.14310494362532525, + "grad_norm": 1.2310680992910144, + "learning_rate": 1.912247115006894e-05, + "loss": 1.5509, + "step": 495 + }, + { + "epoch": 0.143394044521538, + "grad_norm": 1.2540456207457256, + "learning_rate": 1.91187087737799e-05, + "loss": 1.6512, + "step": 496 + }, + { + "epoch": 0.1436831454177508, + "grad_norm": 1.2983726556931725, + "learning_rate": 1.9114938720950307e-05, + "loss": 1.6434, + "step": 497 + }, + { + "epoch": 0.14397224631396358, + "grad_norm": 1.2424942795398506, + "learning_rate": 1.9111160994753957e-05, + "loss": 1.712, + "step": 498 + }, + { + "epoch": 0.14426134721017636, + "grad_norm": 1.2549250776078968, + "learning_rate": 1.9107375598371112e-05, + "loss": 1.5818, + "step": 499 + }, + { + "epoch": 0.14455044810638912, + "grad_norm": 1.171699859687026, + "learning_rate": 1.910358253498849e-05, + "loss": 1.6075, + "step": 500 + }, + { + "epoch": 0.1448395490026019, + "grad_norm": 1.2660287905231888, + "learning_rate": 1.9099781807799263e-05, + "loss": 1.6403, + "step": 501 + }, + { + "epoch": 0.1451286498988147, + "grad_norm": 1.2138215200444276, + "learning_rate": 1.9095973420003056e-05, + "loss": 1.6214, + "step": 502 + }, + { + "epoch": 0.14541775079502747, + "grad_norm": 1.2379532082411215, + "learning_rate": 1.909215737480594e-05, + "loss": 1.5915, + "step": 503 + }, + { + "epoch": 0.14570685169124023, + "grad_norm": 1.2691642883216934, + "learning_rate": 1.9088333675420432e-05, + "loss": 1.6913, + "step": 504 + }, + { + "epoch": 0.14599595258745302, + "grad_norm": 1.3365845490222084, + "learning_rate": 1.9084502325065494e-05, + "loss": 1.7174, + "step": 505 + }, + { + "epoch": 0.1462850534836658, + "grad_norm": 1.2933988368966916, + "learning_rate": 1.9080663326966536e-05, + "loss": 1.6947, + "step": 506 + }, + { + "epoch": 0.14657415437987859, + "grad_norm": 1.1997391314057113, + "learning_rate": 1.9076816684355398e-05, + "loss": 1.6454, + "step": 507 + }, + { + "epoch": 0.14686325527609134, + "grad_norm": 1.2253205782269874, + "learning_rate": 1.9072962400470348e-05, + "loss": 1.6475, + "step": 508 + }, + { + "epoch": 0.14715235617230413, + "grad_norm": 1.149822466017807, + "learning_rate": 1.9069100478556112e-05, + "loss": 1.5374, + "step": 509 + }, + { + "epoch": 0.1474414570685169, + "grad_norm": 1.2247546090753016, + "learning_rate": 1.906523092186382e-05, + "loss": 1.7436, + "step": 510 + }, + { + "epoch": 0.1477305579647297, + "grad_norm": 1.1999484360154251, + "learning_rate": 1.9061353733651036e-05, + "loss": 1.6747, + "step": 511 + }, + { + "epoch": 0.14801965886094248, + "grad_norm": 1.1947780812062576, + "learning_rate": 1.905746891718176e-05, + "loss": 1.59, + "step": 512 + }, + { + "epoch": 0.14830875975715524, + "grad_norm": 1.2483481364125735, + "learning_rate": 1.9053576475726406e-05, + "loss": 1.6343, + "step": 513 + }, + { + "epoch": 0.14859786065336802, + "grad_norm": 1.2303596001336339, + "learning_rate": 1.9049676412561803e-05, + "loss": 1.701, + "step": 514 + }, + { + "epoch": 0.1488869615495808, + "grad_norm": 1.2116845736311206, + "learning_rate": 1.9045768730971198e-05, + "loss": 1.7507, + "step": 515 + }, + { + "epoch": 0.1491760624457936, + "grad_norm": 1.299089146099033, + "learning_rate": 1.904185343424426e-05, + "loss": 1.7314, + "step": 516 + }, + { + "epoch": 0.14946516334200635, + "grad_norm": 1.2754215966625346, + "learning_rate": 1.9037930525677058e-05, + "loss": 1.6955, + "step": 517 + }, + { + "epoch": 0.14975426423821914, + "grad_norm": 1.2864098329743727, + "learning_rate": 1.9034000008572073e-05, + "loss": 1.6926, + "step": 518 + }, + { + "epoch": 0.15004336513443192, + "grad_norm": 1.2920675731077655, + "learning_rate": 1.9030061886238195e-05, + "loss": 1.6528, + "step": 519 + }, + { + "epoch": 0.1503324660306447, + "grad_norm": 1.2265090425552296, + "learning_rate": 1.9026116161990712e-05, + "loss": 1.647, + "step": 520 + }, + { + "epoch": 0.15062156692685746, + "grad_norm": 1.305532335335852, + "learning_rate": 1.902216283915131e-05, + "loss": 1.6085, + "step": 521 + }, + { + "epoch": 0.15091066782307025, + "grad_norm": 1.2996647218299158, + "learning_rate": 1.901820192104808e-05, + "loss": 1.6858, + "step": 522 + }, + { + "epoch": 0.15119976871928303, + "grad_norm": 1.2728823004232528, + "learning_rate": 1.9014233411015496e-05, + "loss": 1.6152, + "step": 523 + }, + { + "epoch": 0.15148886961549582, + "grad_norm": 1.237846119035289, + "learning_rate": 1.901025731239443e-05, + "loss": 1.7852, + "step": 524 + }, + { + "epoch": 0.15177797051170858, + "grad_norm": 1.263777772165666, + "learning_rate": 1.9006273628532146e-05, + "loss": 1.6714, + "step": 525 + }, + { + "epoch": 0.15206707140792136, + "grad_norm": 1.1966029829151998, + "learning_rate": 1.9002282362782284e-05, + "loss": 1.5772, + "step": 526 + }, + { + "epoch": 0.15235617230413415, + "grad_norm": 1.2573815442286682, + "learning_rate": 1.8998283518504874e-05, + "loss": 1.6586, + "step": 527 + }, + { + "epoch": 0.15264527320034693, + "grad_norm": 1.3315550643097303, + "learning_rate": 1.899427709906632e-05, + "loss": 1.5776, + "step": 528 + }, + { + "epoch": 0.1529343740965597, + "grad_norm": 1.27455167369078, + "learning_rate": 1.899026310783941e-05, + "loss": 1.6383, + "step": 529 + }, + { + "epoch": 0.15322347499277247, + "grad_norm": 1.2177716057481967, + "learning_rate": 1.8986241548203305e-05, + "loss": 1.7259, + "step": 530 + }, + { + "epoch": 0.15351257588898526, + "grad_norm": 1.4172388350158833, + "learning_rate": 1.898221242354353e-05, + "loss": 1.6084, + "step": 531 + }, + { + "epoch": 0.15380167678519804, + "grad_norm": 1.2977973632114062, + "learning_rate": 1.8978175737251984e-05, + "loss": 1.7148, + "step": 532 + }, + { + "epoch": 0.1540907776814108, + "grad_norm": 1.3288320048402222, + "learning_rate": 1.897413149272694e-05, + "loss": 1.7092, + "step": 533 + }, + { + "epoch": 0.15437987857762359, + "grad_norm": 1.362022094342137, + "learning_rate": 1.8970079693373014e-05, + "loss": 1.6727, + "step": 534 + }, + { + "epoch": 0.15466897947383637, + "grad_norm": 1.2868416836773502, + "learning_rate": 1.8966020342601204e-05, + "loss": 1.6544, + "step": 535 + }, + { + "epoch": 0.15495808037004916, + "grad_norm": 1.2255959793529978, + "learning_rate": 1.8961953443828853e-05, + "loss": 1.595, + "step": 536 + }, + { + "epoch": 0.1552471812662619, + "grad_norm": 1.2265187293683357, + "learning_rate": 1.8957879000479647e-05, + "loss": 1.6067, + "step": 537 + }, + { + "epoch": 0.1555362821624747, + "grad_norm": 1.198940426728034, + "learning_rate": 1.8953797015983654e-05, + "loss": 1.6146, + "step": 538 + }, + { + "epoch": 0.15582538305868748, + "grad_norm": 1.190556858810557, + "learning_rate": 1.8949707493777262e-05, + "loss": 1.699, + "step": 539 + }, + { + "epoch": 0.15611448395490027, + "grad_norm": 1.1336765732790504, + "learning_rate": 1.894561043730322e-05, + "loss": 1.5966, + "step": 540 + }, + { + "epoch": 0.15640358485111303, + "grad_norm": 1.1667137603744793, + "learning_rate": 1.8941505850010612e-05, + "loss": 1.6644, + "step": 541 + }, + { + "epoch": 0.1566926857473258, + "grad_norm": 1.3524401243741684, + "learning_rate": 1.8937393735354865e-05, + "loss": 1.6905, + "step": 542 + }, + { + "epoch": 0.1569817866435386, + "grad_norm": 1.2665686572396084, + "learning_rate": 1.8933274096797745e-05, + "loss": 1.5214, + "step": 543 + }, + { + "epoch": 0.15727088753975138, + "grad_norm": 1.2333567103961114, + "learning_rate": 1.892914693780735e-05, + "loss": 1.6044, + "step": 544 + }, + { + "epoch": 0.15755998843596414, + "grad_norm": 1.2700672132910338, + "learning_rate": 1.8925012261858103e-05, + "loss": 1.6295, + "step": 545 + }, + { + "epoch": 0.15784908933217692, + "grad_norm": 1.3269513720335189, + "learning_rate": 1.892087007243077e-05, + "loss": 1.5688, + "step": 546 + }, + { + "epoch": 0.1581381902283897, + "grad_norm": 1.444704489940057, + "learning_rate": 1.8916720373012425e-05, + "loss": 1.6048, + "step": 547 + }, + { + "epoch": 0.1584272911246025, + "grad_norm": 1.2362771720222352, + "learning_rate": 1.8912563167096478e-05, + "loss": 1.6548, + "step": 548 + }, + { + "epoch": 0.15871639202081528, + "grad_norm": 1.3841466139534755, + "learning_rate": 1.8908398458182647e-05, + "loss": 1.5571, + "step": 549 + }, + { + "epoch": 0.15900549291702804, + "grad_norm": 1.3571467399932051, + "learning_rate": 1.8904226249776977e-05, + "loss": 1.741, + "step": 550 + }, + { + "epoch": 0.15929459381324082, + "grad_norm": 1.3013296148852647, + "learning_rate": 1.8900046545391817e-05, + "loss": 1.6925, + "step": 551 + }, + { + "epoch": 0.1595836947094536, + "grad_norm": 1.4245052366468887, + "learning_rate": 1.8895859348545832e-05, + "loss": 1.6373, + "step": 552 + }, + { + "epoch": 0.1598727956056664, + "grad_norm": 1.4862334656509957, + "learning_rate": 1.8891664662763998e-05, + "loss": 1.7684, + "step": 553 + }, + { + "epoch": 0.16016189650187915, + "grad_norm": 1.4591040164361495, + "learning_rate": 1.8887462491577585e-05, + "loss": 1.6048, + "step": 554 + }, + { + "epoch": 0.16045099739809193, + "grad_norm": 1.3905790801269649, + "learning_rate": 1.8883252838524167e-05, + "loss": 1.7057, + "step": 555 + }, + { + "epoch": 0.16074009829430472, + "grad_norm": 1.296590703552253, + "learning_rate": 1.8879035707147634e-05, + "loss": 1.5708, + "step": 556 + }, + { + "epoch": 0.1610291991905175, + "grad_norm": 1.4224940976099907, + "learning_rate": 1.887481110099814e-05, + "loss": 1.6613, + "step": 557 + }, + { + "epoch": 0.16131830008673026, + "grad_norm": 1.209416335942298, + "learning_rate": 1.8870579023632163e-05, + "loss": 1.6734, + "step": 558 + }, + { + "epoch": 0.16160740098294304, + "grad_norm": 1.3856482467534985, + "learning_rate": 1.886633947861245e-05, + "loss": 1.6075, + "step": 559 + }, + { + "epoch": 0.16189650187915583, + "grad_norm": 1.298197860303536, + "learning_rate": 1.8862092469508043e-05, + "loss": 1.7309, + "step": 560 + }, + { + "epoch": 0.16218560277536861, + "grad_norm": 1.2884008702670706, + "learning_rate": 1.8857837999894265e-05, + "loss": 1.7367, + "step": 561 + }, + { + "epoch": 0.16247470367158137, + "grad_norm": 1.3295149830237742, + "learning_rate": 1.8853576073352724e-05, + "loss": 1.6373, + "step": 562 + }, + { + "epoch": 0.16276380456779416, + "grad_norm": 1.244943878603924, + "learning_rate": 1.8849306693471297e-05, + "loss": 1.7586, + "step": 563 + }, + { + "epoch": 0.16305290546400694, + "grad_norm": 1.4571917293027963, + "learning_rate": 1.884502986384414e-05, + "loss": 1.6308, + "step": 564 + }, + { + "epoch": 0.16334200636021973, + "grad_norm": 1.2273508652400458, + "learning_rate": 1.884074558807169e-05, + "loss": 1.7033, + "step": 565 + }, + { + "epoch": 0.16363110725643248, + "grad_norm": 1.3596367181968156, + "learning_rate": 1.8836453869760633e-05, + "loss": 1.7688, + "step": 566 + }, + { + "epoch": 0.16392020815264527, + "grad_norm": 1.1782222979069559, + "learning_rate": 1.883215471252394e-05, + "loss": 1.6864, + "step": 567 + }, + { + "epoch": 0.16420930904885805, + "grad_norm": 1.1876724035966852, + "learning_rate": 1.882784811998083e-05, + "loss": 1.651, + "step": 568 + }, + { + "epoch": 0.16449840994507084, + "grad_norm": 1.2807588292559342, + "learning_rate": 1.8823534095756787e-05, + "loss": 1.7127, + "step": 569 + }, + { + "epoch": 0.1647875108412836, + "grad_norm": 1.1959776867333827, + "learning_rate": 1.881921264348355e-05, + "loss": 1.6117, + "step": 570 + }, + { + "epoch": 0.16507661173749638, + "grad_norm": 1.1977710735486093, + "learning_rate": 1.881488376679912e-05, + "loss": 1.6682, + "step": 571 + }, + { + "epoch": 0.16536571263370917, + "grad_norm": 1.2133712471676765, + "learning_rate": 1.8810547469347737e-05, + "loss": 1.701, + "step": 572 + }, + { + "epoch": 0.16565481352992195, + "grad_norm": 1.1615194296116027, + "learning_rate": 1.880620375477989e-05, + "loss": 1.6957, + "step": 573 + }, + { + "epoch": 0.1659439144261347, + "grad_norm": 1.2434508263741582, + "learning_rate": 1.880185262675231e-05, + "loss": 1.6334, + "step": 574 + }, + { + "epoch": 0.1662330153223475, + "grad_norm": 1.1996625528866651, + "learning_rate": 1.8797494088927984e-05, + "loss": 1.7306, + "step": 575 + }, + { + "epoch": 0.16652211621856028, + "grad_norm": 1.25369862209721, + "learning_rate": 1.8793128144976114e-05, + "loss": 1.6611, + "step": 576 + }, + { + "epoch": 0.16681121711477306, + "grad_norm": 1.2478485036785194, + "learning_rate": 1.878875479857216e-05, + "loss": 1.6418, + "step": 577 + }, + { + "epoch": 0.16710031801098582, + "grad_norm": 1.3329254058696818, + "learning_rate": 1.8784374053397792e-05, + "loss": 1.6929, + "step": 578 + }, + { + "epoch": 0.1673894189071986, + "grad_norm": 1.3690716864028842, + "learning_rate": 1.8779985913140927e-05, + "loss": 1.6979, + "step": 579 + }, + { + "epoch": 0.1676785198034114, + "grad_norm": 1.2637256897353144, + "learning_rate": 1.877559038149569e-05, + "loss": 1.6229, + "step": 580 + }, + { + "epoch": 0.16796762069962418, + "grad_norm": 1.3196301095363738, + "learning_rate": 1.8771187462162452e-05, + "loss": 1.6806, + "step": 581 + }, + { + "epoch": 0.16825672159583693, + "grad_norm": 1.2751675322567242, + "learning_rate": 1.8766777158847777e-05, + "loss": 1.7331, + "step": 582 + }, + { + "epoch": 0.16854582249204972, + "grad_norm": 1.2217555466284777, + "learning_rate": 1.8762359475264464e-05, + "loss": 1.6439, + "step": 583 + }, + { + "epoch": 0.1688349233882625, + "grad_norm": 1.2138119129265534, + "learning_rate": 1.8757934415131513e-05, + "loss": 1.6229, + "step": 584 + }, + { + "epoch": 0.1691240242844753, + "grad_norm": 1.2481504467799762, + "learning_rate": 1.8753501982174142e-05, + "loss": 1.6527, + "step": 585 + }, + { + "epoch": 0.16941312518068807, + "grad_norm": 1.1551738364013855, + "learning_rate": 1.8749062180123778e-05, + "loss": 1.6202, + "step": 586 + }, + { + "epoch": 0.16970222607690083, + "grad_norm": 1.221158314927715, + "learning_rate": 1.8744615012718038e-05, + "loss": 1.674, + "step": 587 + }, + { + "epoch": 0.16999132697311362, + "grad_norm": 8.33538007172163, + "learning_rate": 1.8740160483700754e-05, + "loss": 1.7745, + "step": 588 + }, + { + "epoch": 0.1702804278693264, + "grad_norm": 1.2594866391024326, + "learning_rate": 1.8735698596821948e-05, + "loss": 1.7114, + "step": 589 + }, + { + "epoch": 0.17056952876553919, + "grad_norm": 1.1762832459510764, + "learning_rate": 1.8731229355837834e-05, + "loss": 1.6011, + "step": 590 + }, + { + "epoch": 0.17085862966175194, + "grad_norm": 1.2410876448625578, + "learning_rate": 1.8726752764510827e-05, + "loss": 1.5289, + "step": 591 + }, + { + "epoch": 0.17114773055796473, + "grad_norm": 1.2509478918932246, + "learning_rate": 1.8722268826609524e-05, + "loss": 1.6921, + "step": 592 + }, + { + "epoch": 0.1714368314541775, + "grad_norm": 1.2876270619453627, + "learning_rate": 1.87177775459087e-05, + "loss": 1.6632, + "step": 593 + }, + { + "epoch": 0.1717259323503903, + "grad_norm": 1.236444222374534, + "learning_rate": 1.8713278926189324e-05, + "loss": 1.628, + "step": 594 + }, + { + "epoch": 0.17201503324660306, + "grad_norm": 1.3325085992804107, + "learning_rate": 1.8708772971238528e-05, + "loss": 1.7139, + "step": 595 + }, + { + "epoch": 0.17230413414281584, + "grad_norm": 1.3054803605234377, + "learning_rate": 1.8704259684849636e-05, + "loss": 1.6715, + "step": 596 + }, + { + "epoch": 0.17259323503902863, + "grad_norm": 1.2114622828234745, + "learning_rate": 1.869973907082214e-05, + "loss": 1.6289, + "step": 597 + }, + { + "epoch": 0.1728823359352414, + "grad_norm": 1.3255943035899682, + "learning_rate": 1.8695211132961686e-05, + "loss": 1.5875, + "step": 598 + }, + { + "epoch": 0.17317143683145417, + "grad_norm": 1.2436093130787584, + "learning_rate": 1.86906758750801e-05, + "loss": 1.6891, + "step": 599 + }, + { + "epoch": 0.17346053772766695, + "grad_norm": 1.1789555197737465, + "learning_rate": 1.8686133300995374e-05, + "loss": 1.7097, + "step": 600 + }, + { + "epoch": 0.17374963862387974, + "grad_norm": 1.3205013238390153, + "learning_rate": 1.8681583414531644e-05, + "loss": 1.6392, + "step": 601 + }, + { + "epoch": 0.17403873952009252, + "grad_norm": 1.1648007546672043, + "learning_rate": 1.8677026219519212e-05, + "loss": 1.675, + "step": 602 + }, + { + "epoch": 0.17432784041630528, + "grad_norm": 1.2633263020507286, + "learning_rate": 1.8672461719794532e-05, + "loss": 1.5957, + "step": 603 + }, + { + "epoch": 0.17461694131251806, + "grad_norm": 1.2570430438589553, + "learning_rate": 1.8667889919200203e-05, + "loss": 1.7026, + "step": 604 + }, + { + "epoch": 0.17490604220873085, + "grad_norm": 1.203370562367749, + "learning_rate": 1.8663310821584976e-05, + "loss": 1.7054, + "step": 605 + }, + { + "epoch": 0.17519514310494363, + "grad_norm": 1.2387868298918792, + "learning_rate": 1.865872443080374e-05, + "loss": 1.7107, + "step": 606 + }, + { + "epoch": 0.1754842440011564, + "grad_norm": 1.197249289822212, + "learning_rate": 1.8654130750717526e-05, + "loss": 1.5961, + "step": 607 + }, + { + "epoch": 0.17577334489736918, + "grad_norm": 1.1922137775006185, + "learning_rate": 1.8649529785193502e-05, + "loss": 1.6349, + "step": 608 + }, + { + "epoch": 0.17606244579358196, + "grad_norm": 1.238013689830228, + "learning_rate": 1.8644921538104967e-05, + "loss": 1.6487, + "step": 609 + }, + { + "epoch": 0.17635154668979475, + "grad_norm": 1.198711220073165, + "learning_rate": 1.864030601333135e-05, + "loss": 1.6152, + "step": 610 + }, + { + "epoch": 0.1766406475860075, + "grad_norm": 1.1608845614417607, + "learning_rate": 1.8635683214758213e-05, + "loss": 1.6114, + "step": 611 + }, + { + "epoch": 0.1769297484822203, + "grad_norm": 1.223527743047107, + "learning_rate": 1.8631053146277233e-05, + "loss": 1.5324, + "step": 612 + }, + { + "epoch": 0.17721884937843307, + "grad_norm": 1.2244849270886753, + "learning_rate": 1.8626415811786215e-05, + "loss": 1.7019, + "step": 613 + }, + { + "epoch": 0.17750795027464586, + "grad_norm": 1.2422110644907864, + "learning_rate": 1.8621771215189065e-05, + "loss": 1.6713, + "step": 614 + }, + { + "epoch": 0.17779705117085862, + "grad_norm": 1.210470089961754, + "learning_rate": 1.861711936039583e-05, + "loss": 1.6569, + "step": 615 + }, + { + "epoch": 0.1780861520670714, + "grad_norm": 1.207776683125208, + "learning_rate": 1.8612460251322637e-05, + "loss": 1.5186, + "step": 616 + }, + { + "epoch": 0.1783752529632842, + "grad_norm": 1.2462777938743828, + "learning_rate": 1.8607793891891744e-05, + "loss": 1.7027, + "step": 617 + }, + { + "epoch": 0.17866435385949697, + "grad_norm": 1.2187274782133524, + "learning_rate": 1.8603120286031496e-05, + "loss": 1.7697, + "step": 618 + }, + { + "epoch": 0.17895345475570976, + "grad_norm": 1.2542678538432044, + "learning_rate": 1.859843943767635e-05, + "loss": 1.7017, + "step": 619 + }, + { + "epoch": 0.1792425556519225, + "grad_norm": 1.1778318765224127, + "learning_rate": 1.8593751350766854e-05, + "loss": 1.6475, + "step": 620 + }, + { + "epoch": 0.1795316565481353, + "grad_norm": 1.2711298413207304, + "learning_rate": 1.858905602924965e-05, + "loss": 1.6394, + "step": 621 + }, + { + "epoch": 0.17982075744434808, + "grad_norm": 1.232207257276827, + "learning_rate": 1.8584353477077476e-05, + "loss": 1.6074, + "step": 622 + }, + { + "epoch": 0.18010985834056087, + "grad_norm": 1.2540297946919436, + "learning_rate": 1.857964369820915e-05, + "loss": 1.6683, + "step": 623 + }, + { + "epoch": 0.18039895923677363, + "grad_norm": 1.2340677553615835, + "learning_rate": 1.857492669660957e-05, + "loss": 1.6716, + "step": 624 + }, + { + "epoch": 0.1806880601329864, + "grad_norm": 1.2065806455145902, + "learning_rate": 1.857020247624973e-05, + "loss": 1.5365, + "step": 625 + }, + { + "epoch": 0.1809771610291992, + "grad_norm": 1.2263400735274184, + "learning_rate": 1.856547104110669e-05, + "loss": 1.5683, + "step": 626 + }, + { + "epoch": 0.18126626192541198, + "grad_norm": 1.347386958745164, + "learning_rate": 1.8560732395163585e-05, + "loss": 1.7436, + "step": 627 + }, + { + "epoch": 0.18155536282162474, + "grad_norm": 1.3051300929582412, + "learning_rate": 1.8555986542409615e-05, + "loss": 1.7476, + "step": 628 + }, + { + "epoch": 0.18184446371783752, + "grad_norm": 1.3479071566939564, + "learning_rate": 1.855123348684006e-05, + "loss": 1.6623, + "step": 629 + }, + { + "epoch": 0.1821335646140503, + "grad_norm": 1.2568804583760171, + "learning_rate": 1.8546473232456255e-05, + "loss": 1.6444, + "step": 630 + }, + { + "epoch": 0.1824226655102631, + "grad_norm": 1.2426342867282067, + "learning_rate": 1.85417057832656e-05, + "loss": 1.5964, + "step": 631 + }, + { + "epoch": 0.18271176640647585, + "grad_norm": 1.2638177775731838, + "learning_rate": 1.853693114328154e-05, + "loss": 1.5322, + "step": 632 + }, + { + "epoch": 0.18300086730268864, + "grad_norm": 1.164069058129512, + "learning_rate": 1.8532149316523587e-05, + "loss": 1.5849, + "step": 633 + }, + { + "epoch": 0.18328996819890142, + "grad_norm": 1.297248418393498, + "learning_rate": 1.8527360307017306e-05, + "loss": 1.7388, + "step": 634 + }, + { + "epoch": 0.1835790690951142, + "grad_norm": 1.1341179063142641, + "learning_rate": 1.852256411879429e-05, + "loss": 1.7354, + "step": 635 + }, + { + "epoch": 0.18386816999132696, + "grad_norm": 1.215237098350405, + "learning_rate": 1.8517760755892197e-05, + "loss": 1.6552, + "step": 636 + }, + { + "epoch": 0.18415727088753975, + "grad_norm": 1.1982757446916235, + "learning_rate": 1.8512950222354704e-05, + "loss": 1.6313, + "step": 637 + }, + { + "epoch": 0.18444637178375253, + "grad_norm": 1.2063061953012788, + "learning_rate": 1.850813252223154e-05, + "loss": 1.6108, + "step": 638 + }, + { + "epoch": 0.18473547267996532, + "grad_norm": 1.2120232512101812, + "learning_rate": 1.8503307659578466e-05, + "loss": 1.6357, + "step": 639 + }, + { + "epoch": 0.18502457357617808, + "grad_norm": 1.2479587047666476, + "learning_rate": 1.8498475638457265e-05, + "loss": 1.6444, + "step": 640 + }, + { + "epoch": 0.18531367447239086, + "grad_norm": 1.2758952934748575, + "learning_rate": 1.849363646293575e-05, + "loss": 1.632, + "step": 641 + }, + { + "epoch": 0.18560277536860365, + "grad_norm": 1.1490779391882489, + "learning_rate": 1.8488790137087764e-05, + "loss": 1.6316, + "step": 642 + }, + { + "epoch": 0.18589187626481643, + "grad_norm": 1.195110289142274, + "learning_rate": 1.8483936664993152e-05, + "loss": 1.6018, + "step": 643 + }, + { + "epoch": 0.1861809771610292, + "grad_norm": 1.283174200034784, + "learning_rate": 1.847907605073779e-05, + "loss": 1.6488, + "step": 644 + }, + { + "epoch": 0.18647007805724197, + "grad_norm": 1.2662030686633299, + "learning_rate": 1.847420829841357e-05, + "loss": 1.6538, + "step": 645 + }, + { + "epoch": 0.18675917895345476, + "grad_norm": 1.2237194592018654, + "learning_rate": 1.8469333412118375e-05, + "loss": 1.6171, + "step": 646 + }, + { + "epoch": 0.18704827984966754, + "grad_norm": 1.155988769874379, + "learning_rate": 1.846445139595611e-05, + "loss": 1.5969, + "step": 647 + }, + { + "epoch": 0.1873373807458803, + "grad_norm": 1.2416911437703613, + "learning_rate": 1.845956225403667e-05, + "loss": 1.6744, + "step": 648 + }, + { + "epoch": 0.18762648164209308, + "grad_norm": 1.1875358817766466, + "learning_rate": 1.8454665990475967e-05, + "loss": 1.5515, + "step": 649 + }, + { + "epoch": 0.18791558253830587, + "grad_norm": 1.1818241708803934, + "learning_rate": 1.844976260939589e-05, + "loss": 1.5487, + "step": 650 + }, + { + "epoch": 0.18820468343451865, + "grad_norm": 1.1840647734438647, + "learning_rate": 1.8444852114924325e-05, + "loss": 1.6148, + "step": 651 + }, + { + "epoch": 0.1884937843307314, + "grad_norm": 1.2706827502060694, + "learning_rate": 1.8439934511195147e-05, + "loss": 1.6112, + "step": 652 + }, + { + "epoch": 0.1887828852269442, + "grad_norm": 1.220131649175792, + "learning_rate": 1.843500980234822e-05, + "loss": 1.6449, + "step": 653 + }, + { + "epoch": 0.18907198612315698, + "grad_norm": 1.103922527658955, + "learning_rate": 1.8430077992529388e-05, + "loss": 1.6444, + "step": 654 + }, + { + "epoch": 0.18936108701936977, + "grad_norm": 1.1899809619524633, + "learning_rate": 1.8425139085890466e-05, + "loss": 1.5967, + "step": 655 + }, + { + "epoch": 0.18965018791558255, + "grad_norm": 1.2244806367070153, + "learning_rate": 1.8420193086589252e-05, + "loss": 1.6504, + "step": 656 + }, + { + "epoch": 0.1899392888117953, + "grad_norm": 1.1912067775204556, + "learning_rate": 1.841523999878951e-05, + "loss": 1.718, + "step": 657 + }, + { + "epoch": 0.1902283897080081, + "grad_norm": 1.2117478146838403, + "learning_rate": 1.8410279826660976e-05, + "loss": 1.5804, + "step": 658 + }, + { + "epoch": 0.19051749060422088, + "grad_norm": 1.1341524351911334, + "learning_rate": 1.840531257437934e-05, + "loss": 1.6287, + "step": 659 + }, + { + "epoch": 0.19080659150043366, + "grad_norm": 1.1914356537636899, + "learning_rate": 1.8400338246126265e-05, + "loss": 1.6563, + "step": 660 + }, + { + "epoch": 0.19109569239664642, + "grad_norm": 1.4105624776944174, + "learning_rate": 1.839535684608936e-05, + "loss": 1.7139, + "step": 661 + }, + { + "epoch": 0.1913847932928592, + "grad_norm": 1.1645723203046823, + "learning_rate": 1.8390368378462197e-05, + "loss": 1.7166, + "step": 662 + }, + { + "epoch": 0.191673894189072, + "grad_norm": 1.2936969076152187, + "learning_rate": 1.8385372847444296e-05, + "loss": 1.6482, + "step": 663 + }, + { + "epoch": 0.19196299508528478, + "grad_norm": 1.1569033962294648, + "learning_rate": 1.838037025724111e-05, + "loss": 1.6843, + "step": 664 + }, + { + "epoch": 0.19225209598149753, + "grad_norm": 1.1405851915704683, + "learning_rate": 1.837536061206405e-05, + "loss": 1.6033, + "step": 665 + }, + { + "epoch": 0.19254119687771032, + "grad_norm": 1.2542382381366322, + "learning_rate": 1.8370343916130467e-05, + "loss": 1.6706, + "step": 666 + }, + { + "epoch": 0.1928302977739231, + "grad_norm": 1.2159546224332614, + "learning_rate": 1.8365320173663638e-05, + "loss": 1.6918, + "step": 667 + }, + { + "epoch": 0.1931193986701359, + "grad_norm": 1.2532817994557603, + "learning_rate": 1.8360289388892773e-05, + "loss": 1.6683, + "step": 668 + }, + { + "epoch": 0.19340849956634865, + "grad_norm": 1.2395321203333016, + "learning_rate": 1.8355251566053016e-05, + "loss": 1.629, + "step": 669 + }, + { + "epoch": 0.19369760046256143, + "grad_norm": 1.1689572668191017, + "learning_rate": 1.8350206709385437e-05, + "loss": 1.6462, + "step": 670 + }, + { + "epoch": 0.19398670135877422, + "grad_norm": 1.1981131507977363, + "learning_rate": 1.8345154823137015e-05, + "loss": 1.5874, + "step": 671 + }, + { + "epoch": 0.194275802254987, + "grad_norm": 1.1854854067275211, + "learning_rate": 1.834009591156067e-05, + "loss": 1.5968, + "step": 672 + }, + { + "epoch": 0.19456490315119976, + "grad_norm": 1.1769715429764516, + "learning_rate": 1.8335029978915207e-05, + "loss": 1.5856, + "step": 673 + }, + { + "epoch": 0.19485400404741254, + "grad_norm": 1.1759193873525724, + "learning_rate": 1.8329957029465367e-05, + "loss": 1.676, + "step": 674 + }, + { + "epoch": 0.19514310494362533, + "grad_norm": 1.2034916407402159, + "learning_rate": 1.8324877067481782e-05, + "loss": 1.6385, + "step": 675 + }, + { + "epoch": 0.1954322058398381, + "grad_norm": 1.1498736100928149, + "learning_rate": 1.8319790097240998e-05, + "loss": 1.5622, + "step": 676 + }, + { + "epoch": 0.19572130673605087, + "grad_norm": 1.1833854630281373, + "learning_rate": 1.8314696123025456e-05, + "loss": 1.5906, + "step": 677 + }, + { + "epoch": 0.19601040763226366, + "grad_norm": 1.3064848599083938, + "learning_rate": 1.8309595149123486e-05, + "loss": 1.5784, + "step": 678 + }, + { + "epoch": 0.19629950852847644, + "grad_norm": 1.207598354497605, + "learning_rate": 1.8304487179829325e-05, + "loss": 1.6724, + "step": 679 + }, + { + "epoch": 0.19658860942468923, + "grad_norm": 1.3657596862796557, + "learning_rate": 1.829937221944309e-05, + "loss": 1.6808, + "step": 680 + }, + { + "epoch": 0.19687771032090198, + "grad_norm": 1.2300109610049414, + "learning_rate": 1.8294250272270787e-05, + "loss": 1.701, + "step": 681 + }, + { + "epoch": 0.19716681121711477, + "grad_norm": 1.1932414325288254, + "learning_rate": 1.82891213426243e-05, + "loss": 1.6844, + "step": 682 + }, + { + "epoch": 0.19745591211332755, + "grad_norm": 1.2040899317538816, + "learning_rate": 1.8283985434821394e-05, + "loss": 1.6674, + "step": 683 + }, + { + "epoch": 0.19774501300954034, + "grad_norm": 1.20051829187703, + "learning_rate": 1.827884255318571e-05, + "loss": 1.6988, + "step": 684 + }, + { + "epoch": 0.1980341139057531, + "grad_norm": 1.1808862977245562, + "learning_rate": 1.8273692702046754e-05, + "loss": 1.5823, + "step": 685 + }, + { + "epoch": 0.19832321480196588, + "grad_norm": 1.226726509340714, + "learning_rate": 1.8268535885739905e-05, + "loss": 1.6371, + "step": 686 + }, + { + "epoch": 0.19861231569817867, + "grad_norm": 1.254783885869565, + "learning_rate": 1.8263372108606404e-05, + "loss": 1.5849, + "step": 687 + }, + { + "epoch": 0.19890141659439145, + "grad_norm": 1.2210257028134839, + "learning_rate": 1.8258201374993355e-05, + "loss": 1.7316, + "step": 688 + }, + { + "epoch": 0.1991905174906042, + "grad_norm": 1.27099587011664, + "learning_rate": 1.8253023689253707e-05, + "loss": 1.592, + "step": 689 + }, + { + "epoch": 0.199479618386817, + "grad_norm": 1.366689582202721, + "learning_rate": 1.8247839055746276e-05, + "loss": 1.5789, + "step": 690 + }, + { + "epoch": 0.19976871928302978, + "grad_norm": 1.1755442721898772, + "learning_rate": 1.8242647478835717e-05, + "loss": 1.6069, + "step": 691 + }, + { + "epoch": 0.20005782017924256, + "grad_norm": 1.2600721970141062, + "learning_rate": 1.8237448962892537e-05, + "loss": 1.7384, + "step": 692 + }, + { + "epoch": 0.20034692107545535, + "grad_norm": 1.22621954725729, + "learning_rate": 1.823224351229308e-05, + "loss": 1.6226, + "step": 693 + }, + { + "epoch": 0.2006360219716681, + "grad_norm": 1.2182097639154552, + "learning_rate": 1.822703113141953e-05, + "loss": 1.613, + "step": 694 + }, + { + "epoch": 0.2009251228678809, + "grad_norm": 1.2785953541313604, + "learning_rate": 1.82218118246599e-05, + "loss": 1.7014, + "step": 695 + }, + { + "epoch": 0.20121422376409367, + "grad_norm": 1.2338324878833378, + "learning_rate": 1.821658559640804e-05, + "loss": 1.6298, + "step": 696 + }, + { + "epoch": 0.20150332466030646, + "grad_norm": 1.1768453406356614, + "learning_rate": 1.821135245106363e-05, + "loss": 1.5619, + "step": 697 + }, + { + "epoch": 0.20179242555651922, + "grad_norm": 1.1771764186726315, + "learning_rate": 1.820611239303216e-05, + "loss": 1.68, + "step": 698 + }, + { + "epoch": 0.202081526452732, + "grad_norm": 1.1751259370580003, + "learning_rate": 1.8200865426724955e-05, + "loss": 1.6353, + "step": 699 + }, + { + "epoch": 0.2023706273489448, + "grad_norm": 1.234717870608629, + "learning_rate": 1.819561155655914e-05, + "loss": 1.6222, + "step": 700 + }, + { + "epoch": 0.20265972824515757, + "grad_norm": 1.3036974701284907, + "learning_rate": 1.819035078695767e-05, + "loss": 1.6272, + "step": 701 + }, + { + "epoch": 0.20294882914137033, + "grad_norm": 1.201996899703672, + "learning_rate": 1.8185083122349288e-05, + "loss": 1.6977, + "step": 702 + }, + { + "epoch": 0.20323793003758311, + "grad_norm": 1.2513029094302637, + "learning_rate": 1.8179808567168556e-05, + "loss": 1.7145, + "step": 703 + }, + { + "epoch": 0.2035270309337959, + "grad_norm": 1.2829160939047408, + "learning_rate": 1.8174527125855836e-05, + "loss": 1.7475, + "step": 704 + }, + { + "epoch": 0.20381613183000868, + "grad_norm": 1.2634189347114115, + "learning_rate": 1.8169238802857275e-05, + "loss": 1.7567, + "step": 705 + }, + { + "epoch": 0.20410523272622144, + "grad_norm": 1.1500402066275863, + "learning_rate": 1.8163943602624834e-05, + "loss": 1.568, + "step": 706 + }, + { + "epoch": 0.20439433362243423, + "grad_norm": 1.2208940822026542, + "learning_rate": 1.815864152961624e-05, + "loss": 1.6101, + "step": 707 + }, + { + "epoch": 0.204683434518647, + "grad_norm": 1.2920340236775505, + "learning_rate": 1.8153332588295023e-05, + "loss": 1.6039, + "step": 708 + }, + { + "epoch": 0.2049725354148598, + "grad_norm": 1.3082533794058193, + "learning_rate": 1.814801678313049e-05, + "loss": 1.66, + "step": 709 + }, + { + "epoch": 0.20526163631107255, + "grad_norm": 1.2031595547172316, + "learning_rate": 1.8142694118597727e-05, + "loss": 1.6835, + "step": 710 + }, + { + "epoch": 0.20555073720728534, + "grad_norm": 1.1799641826600815, + "learning_rate": 1.8137364599177587e-05, + "loss": 1.6254, + "step": 711 + }, + { + "epoch": 0.20583983810349812, + "grad_norm": 1.208426243281953, + "learning_rate": 1.8132028229356703e-05, + "loss": 1.6405, + "step": 712 + }, + { + "epoch": 0.2061289389997109, + "grad_norm": 1.1403284646294907, + "learning_rate": 1.8126685013627475e-05, + "loss": 1.7065, + "step": 713 + }, + { + "epoch": 0.20641803989592367, + "grad_norm": 1.2249537048855292, + "learning_rate": 1.8121334956488058e-05, + "loss": 1.6447, + "step": 714 + }, + { + "epoch": 0.20670714079213645, + "grad_norm": 1.189614154915905, + "learning_rate": 1.8115978062442375e-05, + "loss": 1.5539, + "step": 715 + }, + { + "epoch": 0.20699624168834924, + "grad_norm": 1.1643490997811285, + "learning_rate": 1.81106143360001e-05, + "loss": 1.6068, + "step": 716 + }, + { + "epoch": 0.20728534258456202, + "grad_norm": 1.1646757540028276, + "learning_rate": 1.8105243781676663e-05, + "loss": 1.5891, + "step": 717 + }, + { + "epoch": 0.20757444348077478, + "grad_norm": 1.2441086225997955, + "learning_rate": 1.809986640399323e-05, + "loss": 1.529, + "step": 718 + }, + { + "epoch": 0.20786354437698756, + "grad_norm": 1.3159032556741779, + "learning_rate": 1.809448220747673e-05, + "loss": 1.5943, + "step": 719 + }, + { + "epoch": 0.20815264527320035, + "grad_norm": 1.1977613211999474, + "learning_rate": 1.8089091196659822e-05, + "loss": 1.6186, + "step": 720 + }, + { + "epoch": 0.20844174616941313, + "grad_norm": 1.1611286191966523, + "learning_rate": 1.8083693376080896e-05, + "loss": 1.609, + "step": 721 + }, + { + "epoch": 0.2087308470656259, + "grad_norm": 1.2479677009723977, + "learning_rate": 1.807828875028408e-05, + "loss": 1.6632, + "step": 722 + }, + { + "epoch": 0.20901994796183868, + "grad_norm": 1.2198060446129206, + "learning_rate": 1.8072877323819247e-05, + "loss": 1.6938, + "step": 723 + }, + { + "epoch": 0.20930904885805146, + "grad_norm": 1.1959812561872514, + "learning_rate": 1.806745910124196e-05, + "loss": 1.6715, + "step": 724 + }, + { + "epoch": 0.20959814975426425, + "grad_norm": 1.1644575837628937, + "learning_rate": 1.8062034087113538e-05, + "loss": 1.654, + "step": 725 + }, + { + "epoch": 0.209887250650477, + "grad_norm": 1.1952595776641282, + "learning_rate": 1.8056602286000995e-05, + "loss": 1.7361, + "step": 726 + }, + { + "epoch": 0.2101763515466898, + "grad_norm": 1.1187248069836293, + "learning_rate": 1.8051163702477067e-05, + "loss": 1.4882, + "step": 727 + }, + { + "epoch": 0.21046545244290257, + "grad_norm": 1.23860503080743, + "learning_rate": 1.80457183411202e-05, + "loss": 1.6494, + "step": 728 + }, + { + "epoch": 0.21075455333911536, + "grad_norm": 1.1198629149288368, + "learning_rate": 1.8040266206514548e-05, + "loss": 1.6625, + "step": 729 + }, + { + "epoch": 0.21104365423532814, + "grad_norm": 1.1312334064032668, + "learning_rate": 1.8034807303249956e-05, + "loss": 1.644, + "step": 730 + }, + { + "epoch": 0.2113327551315409, + "grad_norm": 1.178023621775733, + "learning_rate": 1.8029341635921985e-05, + "loss": 1.6237, + "step": 731 + }, + { + "epoch": 0.21162185602775369, + "grad_norm": 1.1831551646009177, + "learning_rate": 1.8023869209131876e-05, + "loss": 1.684, + "step": 732 + }, + { + "epoch": 0.21191095692396647, + "grad_norm": 1.1269798393522688, + "learning_rate": 1.8018390027486564e-05, + "loss": 1.6614, + "step": 733 + }, + { + "epoch": 0.21220005782017926, + "grad_norm": 1.163617004294811, + "learning_rate": 1.8012904095598672e-05, + "loss": 1.5863, + "step": 734 + }, + { + "epoch": 0.212489158716392, + "grad_norm": 1.1666239101175062, + "learning_rate": 1.8007411418086504e-05, + "loss": 1.6286, + "step": 735 + }, + { + "epoch": 0.2127782596126048, + "grad_norm": 1.1783363837507064, + "learning_rate": 1.8001911999574048e-05, + "loss": 1.6024, + "step": 736 + }, + { + "epoch": 0.21306736050881758, + "grad_norm": 1.1324429879608466, + "learning_rate": 1.7996405844690955e-05, + "loss": 1.6086, + "step": 737 + }, + { + "epoch": 0.21335646140503037, + "grad_norm": 1.1857513831899675, + "learning_rate": 1.7990892958072562e-05, + "loss": 1.6637, + "step": 738 + }, + { + "epoch": 0.21364556230124312, + "grad_norm": 1.20546977750714, + "learning_rate": 1.798537334435986e-05, + "loss": 1.651, + "step": 739 + }, + { + "epoch": 0.2139346631974559, + "grad_norm": 1.214143713272185, + "learning_rate": 1.7979847008199515e-05, + "loss": 1.5663, + "step": 740 + }, + { + "epoch": 0.2142237640936687, + "grad_norm": 1.2482881773540793, + "learning_rate": 1.7974313954243846e-05, + "loss": 1.6634, + "step": 741 + }, + { + "epoch": 0.21451286498988148, + "grad_norm": 1.3788124938610393, + "learning_rate": 1.796877418715082e-05, + "loss": 1.7858, + "step": 742 + }, + { + "epoch": 0.21480196588609424, + "grad_norm": 1.210426850026346, + "learning_rate": 1.796322771158407e-05, + "loss": 1.5901, + "step": 743 + }, + { + "epoch": 0.21509106678230702, + "grad_norm": 1.1818842755255796, + "learning_rate": 1.7957674532212862e-05, + "loss": 1.5516, + "step": 744 + }, + { + "epoch": 0.2153801676785198, + "grad_norm": 1.2331450769496057, + "learning_rate": 1.7952114653712123e-05, + "loss": 1.6224, + "step": 745 + }, + { + "epoch": 0.2156692685747326, + "grad_norm": 1.1415088042700048, + "learning_rate": 1.7946548080762402e-05, + "loss": 1.7262, + "step": 746 + }, + { + "epoch": 0.21595836947094535, + "grad_norm": 1.0990551729161935, + "learning_rate": 1.7940974818049898e-05, + "loss": 1.674, + "step": 747 + }, + { + "epoch": 0.21624747036715813, + "grad_norm": 1.217412085009134, + "learning_rate": 1.7935394870266425e-05, + "loss": 1.6138, + "step": 748 + }, + { + "epoch": 0.21653657126337092, + "grad_norm": 1.167638566385817, + "learning_rate": 1.7929808242109444e-05, + "loss": 1.6414, + "step": 749 + }, + { + "epoch": 0.2168256721595837, + "grad_norm": 1.2003093360263695, + "learning_rate": 1.7924214938282026e-05, + "loss": 1.6289, + "step": 750 + }, + { + "epoch": 0.21711477305579646, + "grad_norm": 1.2685172755141476, + "learning_rate": 1.7918614963492872e-05, + "loss": 1.7019, + "step": 751 + }, + { + "epoch": 0.21740387395200925, + "grad_norm": 1.173005539771057, + "learning_rate": 1.7913008322456283e-05, + "loss": 1.6812, + "step": 752 + }, + { + "epoch": 0.21769297484822203, + "grad_norm": 1.2159228008333007, + "learning_rate": 1.7907395019892196e-05, + "loss": 1.6077, + "step": 753 + }, + { + "epoch": 0.21798207574443482, + "grad_norm": 1.1742389210088622, + "learning_rate": 1.7901775060526132e-05, + "loss": 1.622, + "step": 754 + }, + { + "epoch": 0.21827117664064757, + "grad_norm": 1.2128889995971577, + "learning_rate": 1.789614844908923e-05, + "loss": 1.6743, + "step": 755 + }, + { + "epoch": 0.21856027753686036, + "grad_norm": 1.1667752847351143, + "learning_rate": 1.7890515190318222e-05, + "loss": 1.5754, + "step": 756 + }, + { + "epoch": 0.21884937843307314, + "grad_norm": 1.1295223300664317, + "learning_rate": 1.7884875288955448e-05, + "loss": 1.6738, + "step": 757 + }, + { + "epoch": 0.21913847932928593, + "grad_norm": 1.261068879100486, + "learning_rate": 1.7879228749748824e-05, + "loss": 1.6835, + "step": 758 + }, + { + "epoch": 0.2194275802254987, + "grad_norm": 1.1728226481913466, + "learning_rate": 1.7873575577451863e-05, + "loss": 1.6891, + "step": 759 + }, + { + "epoch": 0.21971668112171147, + "grad_norm": 1.1787350920697617, + "learning_rate": 1.7867915776823663e-05, + "loss": 1.6257, + "step": 760 + }, + { + "epoch": 0.22000578201792426, + "grad_norm": 1.1620635102842405, + "learning_rate": 1.7862249352628897e-05, + "loss": 1.6228, + "step": 761 + }, + { + "epoch": 0.22029488291413704, + "grad_norm": 1.174838883761586, + "learning_rate": 1.785657630963782e-05, + "loss": 1.6795, + "step": 762 + }, + { + "epoch": 0.2205839838103498, + "grad_norm": 1.2540807508404137, + "learning_rate": 1.785089665262625e-05, + "loss": 1.6921, + "step": 763 + }, + { + "epoch": 0.22087308470656258, + "grad_norm": 1.1328480065063387, + "learning_rate": 1.784521038637558e-05, + "loss": 1.6261, + "step": 764 + }, + { + "epoch": 0.22116218560277537, + "grad_norm": 1.1913711901780004, + "learning_rate": 1.783951751567277e-05, + "loss": 1.7268, + "step": 765 + }, + { + "epoch": 0.22145128649898815, + "grad_norm": 1.1168162011581586, + "learning_rate": 1.7833818045310335e-05, + "loss": 1.6462, + "step": 766 + }, + { + "epoch": 0.22174038739520094, + "grad_norm": 1.1597448679252735, + "learning_rate": 1.782811198008634e-05, + "loss": 1.6369, + "step": 767 + }, + { + "epoch": 0.2220294882914137, + "grad_norm": 1.1983400772535562, + "learning_rate": 1.782239932480441e-05, + "loss": 1.6555, + "step": 768 + }, + { + "epoch": 0.22231858918762648, + "grad_norm": 1.1771360817965915, + "learning_rate": 1.7816680084273724e-05, + "loss": 1.6722, + "step": 769 + }, + { + "epoch": 0.22260769008383927, + "grad_norm": 1.202362947433072, + "learning_rate": 1.7810954263308987e-05, + "loss": 1.6489, + "step": 770 + }, + { + "epoch": 0.22289679098005205, + "grad_norm": 1.1514930888062975, + "learning_rate": 1.780522186673046e-05, + "loss": 1.5065, + "step": 771 + }, + { + "epoch": 0.2231858918762648, + "grad_norm": 1.2543833283472365, + "learning_rate": 1.779948289936393e-05, + "loss": 1.6361, + "step": 772 + }, + { + "epoch": 0.2234749927724776, + "grad_norm": 1.1103075126565838, + "learning_rate": 1.779373736604072e-05, + "loss": 1.5877, + "step": 773 + }, + { + "epoch": 0.22376409366869038, + "grad_norm": 1.2145730232558314, + "learning_rate": 1.7787985271597678e-05, + "loss": 1.586, + "step": 774 + }, + { + "epoch": 0.22405319456490316, + "grad_norm": 1.2583750794569701, + "learning_rate": 1.7782226620877178e-05, + "loss": 1.6928, + "step": 775 + }, + { + "epoch": 0.22434229546111592, + "grad_norm": 1.192600045488377, + "learning_rate": 1.777646141872711e-05, + "loss": 1.6711, + "step": 776 + }, + { + "epoch": 0.2246313963573287, + "grad_norm": 1.2325135805846623, + "learning_rate": 1.7770689670000882e-05, + "loss": 1.6051, + "step": 777 + }, + { + "epoch": 0.2249204972535415, + "grad_norm": 1.2445267162996139, + "learning_rate": 1.7764911379557415e-05, + "loss": 1.6704, + "step": 778 + }, + { + "epoch": 0.22520959814975428, + "grad_norm": 1.164473435396523, + "learning_rate": 1.7759126552261133e-05, + "loss": 1.6788, + "step": 779 + }, + { + "epoch": 0.22549869904596703, + "grad_norm": 1.3267375597230986, + "learning_rate": 1.7753335192981963e-05, + "loss": 1.6578, + "step": 780 + }, + { + "epoch": 0.22578779994217982, + "grad_norm": 1.138551289696566, + "learning_rate": 1.7747537306595336e-05, + "loss": 1.6774, + "step": 781 + }, + { + "epoch": 0.2260769008383926, + "grad_norm": 1.2577334291491808, + "learning_rate": 1.774173289798217e-05, + "loss": 1.6842, + "step": 782 + }, + { + "epoch": 0.2263660017346054, + "grad_norm": 1.1496680909323838, + "learning_rate": 1.7735921972028883e-05, + "loss": 1.628, + "step": 783 + }, + { + "epoch": 0.22665510263081814, + "grad_norm": 1.179850018613425, + "learning_rate": 1.773010453362737e-05, + "loss": 1.6444, + "step": 784 + }, + { + "epoch": 0.22694420352703093, + "grad_norm": 1.2654130215050725, + "learning_rate": 1.7724280587675017e-05, + "loss": 1.4592, + "step": 785 + }, + { + "epoch": 0.22723330442324371, + "grad_norm": 1.1907665269492804, + "learning_rate": 1.7718450139074685e-05, + "loss": 1.58, + "step": 786 + }, + { + "epoch": 0.2275224053194565, + "grad_norm": 1.217021995188848, + "learning_rate": 1.7712613192734703e-05, + "loss": 1.575, + "step": 787 + }, + { + "epoch": 0.22781150621566926, + "grad_norm": 1.1808590593353434, + "learning_rate": 1.7706769753568882e-05, + "loss": 1.6172, + "step": 788 + }, + { + "epoch": 0.22810060711188204, + "grad_norm": 1.196849426701046, + "learning_rate": 1.770091982649649e-05, + "loss": 1.6436, + "step": 789 + }, + { + "epoch": 0.22838970800809483, + "grad_norm": 1.1412097565678294, + "learning_rate": 1.7695063416442263e-05, + "loss": 1.6423, + "step": 790 + }, + { + "epoch": 0.2286788089043076, + "grad_norm": 1.1352763403847086, + "learning_rate": 1.7689200528336384e-05, + "loss": 1.6991, + "step": 791 + }, + { + "epoch": 0.22896790980052037, + "grad_norm": 1.1832029724179747, + "learning_rate": 1.7683331167114503e-05, + "loss": 1.6004, + "step": 792 + }, + { + "epoch": 0.22925701069673315, + "grad_norm": 1.190404880041848, + "learning_rate": 1.767745533771771e-05, + "loss": 1.6232, + "step": 793 + }, + { + "epoch": 0.22954611159294594, + "grad_norm": 1.1425994526274166, + "learning_rate": 1.7671573045092546e-05, + "loss": 1.6601, + "step": 794 + }, + { + "epoch": 0.22983521248915872, + "grad_norm": 4.306902746171457, + "learning_rate": 1.766568429419099e-05, + "loss": 1.6119, + "step": 795 + }, + { + "epoch": 0.23012431338537148, + "grad_norm": 1.2116557080614034, + "learning_rate": 1.7659789089970457e-05, + "loss": 1.553, + "step": 796 + }, + { + "epoch": 0.23041341428158427, + "grad_norm": 1.133998906444825, + "learning_rate": 1.7653887437393792e-05, + "loss": 1.5345, + "step": 797 + }, + { + "epoch": 0.23070251517779705, + "grad_norm": 1.2512101232473924, + "learning_rate": 1.764797934142928e-05, + "loss": 1.6369, + "step": 798 + }, + { + "epoch": 0.23099161607400984, + "grad_norm": 1.1088774895792184, + "learning_rate": 1.7642064807050617e-05, + "loss": 1.6432, + "step": 799 + }, + { + "epoch": 0.2312807169702226, + "grad_norm": 1.2101553368194318, + "learning_rate": 1.7636143839236934e-05, + "loss": 1.6479, + "step": 800 + }, + { + "epoch": 0.23156981786643538, + "grad_norm": 1.191660567807356, + "learning_rate": 1.763021644297276e-05, + "loss": 1.5965, + "step": 801 + }, + { + "epoch": 0.23185891876264816, + "grad_norm": 1.143724831739925, + "learning_rate": 1.7624282623248043e-05, + "loss": 1.6417, + "step": 802 + }, + { + "epoch": 0.23214801965886095, + "grad_norm": 1.143027015736506, + "learning_rate": 1.7618342385058147e-05, + "loss": 1.6015, + "step": 803 + }, + { + "epoch": 0.23243712055507373, + "grad_norm": 1.1346799180515472, + "learning_rate": 1.761239573340383e-05, + "loss": 1.6608, + "step": 804 + }, + { + "epoch": 0.2327262214512865, + "grad_norm": 1.1117892300100156, + "learning_rate": 1.760644267329125e-05, + "loss": 1.6395, + "step": 805 + }, + { + "epoch": 0.23301532234749928, + "grad_norm": 1.180786293910612, + "learning_rate": 1.7600483209731963e-05, + "loss": 1.7766, + "step": 806 + }, + { + "epoch": 0.23330442324371206, + "grad_norm": 1.1174984255388902, + "learning_rate": 1.7594517347742918e-05, + "loss": 1.5313, + "step": 807 + }, + { + "epoch": 0.23359352413992485, + "grad_norm": 1.17896186676479, + "learning_rate": 1.7588545092346438e-05, + "loss": 1.7089, + "step": 808 + }, + { + "epoch": 0.2338826250361376, + "grad_norm": 1.149670644485711, + "learning_rate": 1.7582566448570245e-05, + "loss": 1.6145, + "step": 809 + }, + { + "epoch": 0.2341717259323504, + "grad_norm": 1.2037877986724042, + "learning_rate": 1.7576581421447428e-05, + "loss": 1.663, + "step": 810 + }, + { + "epoch": 0.23446082682856317, + "grad_norm": 1.1979774494369526, + "learning_rate": 1.7570590016016455e-05, + "loss": 1.6597, + "step": 811 + }, + { + "epoch": 0.23474992772477596, + "grad_norm": 1.198242854855127, + "learning_rate": 1.7564592237321162e-05, + "loss": 1.6047, + "step": 812 + }, + { + "epoch": 0.23503902862098872, + "grad_norm": 1.2006909754615687, + "learning_rate": 1.7558588090410747e-05, + "loss": 1.7242, + "step": 813 + }, + { + "epoch": 0.2353281295172015, + "grad_norm": 1.1396286336459736, + "learning_rate": 1.7552577580339768e-05, + "loss": 1.5758, + "step": 814 + }, + { + "epoch": 0.23561723041341429, + "grad_norm": 1.1676518798162923, + "learning_rate": 1.7546560712168154e-05, + "loss": 1.6103, + "step": 815 + }, + { + "epoch": 0.23590633130962707, + "grad_norm": 1.1417358311927566, + "learning_rate": 1.7540537490961168e-05, + "loss": 1.6361, + "step": 816 + }, + { + "epoch": 0.23619543220583983, + "grad_norm": 1.1842710928884372, + "learning_rate": 1.753450792178943e-05, + "loss": 1.6522, + "step": 817 + }, + { + "epoch": 0.2364845331020526, + "grad_norm": 1.2334219525738506, + "learning_rate": 1.752847200972891e-05, + "loss": 1.6686, + "step": 818 + }, + { + "epoch": 0.2367736339982654, + "grad_norm": 1.1595381193664076, + "learning_rate": 1.7522429759860903e-05, + "loss": 1.6584, + "step": 819 + }, + { + "epoch": 0.23706273489447818, + "grad_norm": 1.1998920618458444, + "learning_rate": 1.751638117727205e-05, + "loss": 1.5457, + "step": 820 + }, + { + "epoch": 0.23735183579069094, + "grad_norm": 1.2098085468599757, + "learning_rate": 1.7510326267054325e-05, + "loss": 1.611, + "step": 821 + }, + { + "epoch": 0.23764093668690373, + "grad_norm": 1.145177051100393, + "learning_rate": 1.7504265034305023e-05, + "loss": 1.5325, + "step": 822 + }, + { + "epoch": 0.2379300375831165, + "grad_norm": 1.1017627071573495, + "learning_rate": 1.7498197484126756e-05, + "loss": 1.5977, + "step": 823 + }, + { + "epoch": 0.2382191384793293, + "grad_norm": 1.2445774010012118, + "learning_rate": 1.749212362162747e-05, + "loss": 1.7187, + "step": 824 + }, + { + "epoch": 0.23850823937554205, + "grad_norm": 1.2507033146298203, + "learning_rate": 1.7486043451920415e-05, + "loss": 1.6508, + "step": 825 + }, + { + "epoch": 0.23879734027175484, + "grad_norm": 1.1841673566917597, + "learning_rate": 1.747995698012415e-05, + "loss": 1.6084, + "step": 826 + }, + { + "epoch": 0.23908644116796762, + "grad_norm": 1.1396731986156217, + "learning_rate": 1.7473864211362544e-05, + "loss": 1.6588, + "step": 827 + }, + { + "epoch": 0.2393755420641804, + "grad_norm": 1.204843676859692, + "learning_rate": 1.7467765150764763e-05, + "loss": 1.6455, + "step": 828 + }, + { + "epoch": 0.23966464296039316, + "grad_norm": 1.2736891694731587, + "learning_rate": 1.7461659803465273e-05, + "loss": 1.638, + "step": 829 + }, + { + "epoch": 0.23995374385660595, + "grad_norm": 1.2082297185144963, + "learning_rate": 1.7455548174603828e-05, + "loss": 1.6357, + "step": 830 + }, + { + "epoch": 0.24024284475281873, + "grad_norm": 1.1833363832715826, + "learning_rate": 1.7449430269325474e-05, + "loss": 1.6351, + "step": 831 + }, + { + "epoch": 0.24053194564903152, + "grad_norm": 1.1586148600596027, + "learning_rate": 1.7443306092780543e-05, + "loss": 1.6456, + "step": 832 + }, + { + "epoch": 0.24082104654524428, + "grad_norm": 1.1151118614534092, + "learning_rate": 1.743717565012464e-05, + "loss": 1.6143, + "step": 833 + }, + { + "epoch": 0.24111014744145706, + "grad_norm": 1.1764063110634089, + "learning_rate": 1.7431038946518648e-05, + "loss": 1.6603, + "step": 834 + }, + { + "epoch": 0.24139924833766985, + "grad_norm": 1.1417362079110664, + "learning_rate": 1.7424895987128723e-05, + "loss": 1.6167, + "step": 835 + }, + { + "epoch": 0.24168834923388263, + "grad_norm": 1.1257893632889384, + "learning_rate": 1.7418746777126284e-05, + "loss": 1.62, + "step": 836 + }, + { + "epoch": 0.2419774501300954, + "grad_norm": 1.3179549442602967, + "learning_rate": 1.741259132168801e-05, + "loss": 1.5686, + "step": 837 + }, + { + "epoch": 0.24226655102630817, + "grad_norm": 1.1244904370304307, + "learning_rate": 1.7406429625995853e-05, + "loss": 1.6366, + "step": 838 + }, + { + "epoch": 0.24255565192252096, + "grad_norm": 1.2065775280910245, + "learning_rate": 1.7400261695236996e-05, + "loss": 1.7549, + "step": 839 + }, + { + "epoch": 0.24284475281873374, + "grad_norm": 1.1396498478056365, + "learning_rate": 1.739408753460388e-05, + "loss": 1.552, + "step": 840 + }, + { + "epoch": 0.24313385371494653, + "grad_norm": 1.1240397902793318, + "learning_rate": 1.73879071492942e-05, + "loss": 1.6687, + "step": 841 + }, + { + "epoch": 0.2434229546111593, + "grad_norm": 1.179163925851606, + "learning_rate": 1.738172054451088e-05, + "loss": 1.6504, + "step": 842 + }, + { + "epoch": 0.24371205550737207, + "grad_norm": 1.1030610833520664, + "learning_rate": 1.737552772546208e-05, + "loss": 1.5912, + "step": 843 + }, + { + "epoch": 0.24400115640358486, + "grad_norm": 1.1838602443986301, + "learning_rate": 1.73693286973612e-05, + "loss": 1.5732, + "step": 844 + }, + { + "epoch": 0.24429025729979764, + "grad_norm": 1.143992740058496, + "learning_rate": 1.7363123465426853e-05, + "loss": 1.6071, + "step": 845 + }, + { + "epoch": 0.2445793581960104, + "grad_norm": 1.174590060660084, + "learning_rate": 1.7356912034882893e-05, + "loss": 1.5759, + "step": 846 + }, + { + "epoch": 0.24486845909222318, + "grad_norm": 1.1329692315930469, + "learning_rate": 1.735069441095838e-05, + "loss": 1.6088, + "step": 847 + }, + { + "epoch": 0.24515755998843597, + "grad_norm": 1.1350450062350554, + "learning_rate": 1.7344470598887585e-05, + "loss": 1.7202, + "step": 848 + }, + { + "epoch": 0.24544666088464875, + "grad_norm": 1.181601786302941, + "learning_rate": 1.733824060391e-05, + "loss": 1.7271, + "step": 849 + }, + { + "epoch": 0.2457357617808615, + "grad_norm": 1.1583792292333024, + "learning_rate": 1.7332004431270312e-05, + "loss": 1.5625, + "step": 850 + }, + { + "epoch": 0.2460248626770743, + "grad_norm": 1.1568878076104787, + "learning_rate": 1.7325762086218415e-05, + "loss": 1.6503, + "step": 851 + }, + { + "epoch": 0.24631396357328708, + "grad_norm": 1.2339640973176655, + "learning_rate": 1.7319513574009398e-05, + "loss": 1.7399, + "step": 852 + }, + { + "epoch": 0.24660306446949987, + "grad_norm": 1.2009335734006834, + "learning_rate": 1.731325889990354e-05, + "loss": 1.5677, + "step": 853 + }, + { + "epoch": 0.24689216536571262, + "grad_norm": 1.1688999648294327, + "learning_rate": 1.73069980691663e-05, + "loss": 1.6136, + "step": 854 + }, + { + "epoch": 0.2471812662619254, + "grad_norm": 1.1635714362599143, + "learning_rate": 1.730073108706834e-05, + "loss": 1.6131, + "step": 855 + }, + { + "epoch": 0.2474703671581382, + "grad_norm": 1.2123420765200588, + "learning_rate": 1.7294457958885482e-05, + "loss": 1.7155, + "step": 856 + }, + { + "epoch": 0.24775946805435098, + "grad_norm": 1.1434228753959954, + "learning_rate": 1.7288178689898725e-05, + "loss": 1.7325, + "step": 857 + }, + { + "epoch": 0.24804856895056374, + "grad_norm": 1.2246804177847264, + "learning_rate": 1.728189328539425e-05, + "loss": 1.6505, + "step": 858 + }, + { + "epoch": 0.24833766984677652, + "grad_norm": 1.1585117368059001, + "learning_rate": 1.7275601750663387e-05, + "loss": 1.5769, + "step": 859 + }, + { + "epoch": 0.2486267707429893, + "grad_norm": 1.3882387244074696, + "learning_rate": 1.7269304091002633e-05, + "loss": 1.6452, + "step": 860 + }, + { + "epoch": 0.2489158716392021, + "grad_norm": 1.160203748747366, + "learning_rate": 1.726300031171365e-05, + "loss": 1.5732, + "step": 861 + }, + { + "epoch": 0.24920497253541485, + "grad_norm": 1.157523734644445, + "learning_rate": 1.7256690418103238e-05, + "loss": 1.6401, + "step": 862 + }, + { + "epoch": 0.24949407343162763, + "grad_norm": 1.1890152342635811, + "learning_rate": 1.725037441548335e-05, + "loss": 1.6668, + "step": 863 + }, + { + "epoch": 0.24978317432784042, + "grad_norm": 1.2257536935692366, + "learning_rate": 1.7244052309171087e-05, + "loss": 1.6516, + "step": 864 + }, + { + "epoch": 0.2500722752240532, + "grad_norm": 1.2023469701707792, + "learning_rate": 1.7237724104488675e-05, + "loss": 1.5962, + "step": 865 + }, + { + "epoch": 0.250361376120266, + "grad_norm": 1.2280068128337949, + "learning_rate": 1.723138980676349e-05, + "loss": 1.6962, + "step": 866 + }, + { + "epoch": 0.25065047701647875, + "grad_norm": 1.217449764349956, + "learning_rate": 1.7225049421328024e-05, + "loss": 1.6377, + "step": 867 + }, + { + "epoch": 0.2509395779126915, + "grad_norm": 1.2841045771510446, + "learning_rate": 1.7218702953519904e-05, + "loss": 1.6582, + "step": 868 + }, + { + "epoch": 0.2512286788089043, + "grad_norm": 1.2601843986629024, + "learning_rate": 1.7212350408681872e-05, + "loss": 1.6102, + "step": 869 + }, + { + "epoch": 0.2515177797051171, + "grad_norm": 1.145255827066236, + "learning_rate": 1.7205991792161783e-05, + "loss": 1.6581, + "step": 870 + }, + { + "epoch": 0.2518068806013299, + "grad_norm": 1.2362667478605363, + "learning_rate": 1.7199627109312615e-05, + "loss": 1.6043, + "step": 871 + }, + { + "epoch": 0.25209598149754264, + "grad_norm": 1.2401716127785924, + "learning_rate": 1.7193256365492436e-05, + "loss": 1.632, + "step": 872 + }, + { + "epoch": 0.2523850823937554, + "grad_norm": 1.119783771494612, + "learning_rate": 1.7186879566064433e-05, + "loss": 1.5714, + "step": 873 + }, + { + "epoch": 0.2526741832899682, + "grad_norm": 1.103085959008315, + "learning_rate": 1.718049671639688e-05, + "loss": 1.6384, + "step": 874 + }, + { + "epoch": 0.25296328418618097, + "grad_norm": 1.133467746444829, + "learning_rate": 1.717410782186315e-05, + "loss": 1.6366, + "step": 875 + }, + { + "epoch": 0.2532523850823938, + "grad_norm": 1.175009710061539, + "learning_rate": 1.7167712887841705e-05, + "loss": 1.6428, + "step": 876 + }, + { + "epoch": 0.25354148597860654, + "grad_norm": 1.1563638245548544, + "learning_rate": 1.716131191971608e-05, + "loss": 1.5624, + "step": 877 + }, + { + "epoch": 0.2538305868748193, + "grad_norm": 1.116419831991319, + "learning_rate": 1.715490492287491e-05, + "loss": 1.5829, + "step": 878 + }, + { + "epoch": 0.2541196877710321, + "grad_norm": 1.1370821345076967, + "learning_rate": 1.7148491902711884e-05, + "loss": 1.6631, + "step": 879 + }, + { + "epoch": 0.25440878866724487, + "grad_norm": 1.1611842225988047, + "learning_rate": 1.7142072864625777e-05, + "loss": 1.5913, + "step": 880 + }, + { + "epoch": 0.2546978895634576, + "grad_norm": 1.1019030805813086, + "learning_rate": 1.7135647814020425e-05, + "loss": 1.6758, + "step": 881 + }, + { + "epoch": 0.25498699045967044, + "grad_norm": 1.04613261420445, + "learning_rate": 1.7129216756304723e-05, + "loss": 1.5948, + "step": 882 + }, + { + "epoch": 0.2552760913558832, + "grad_norm": 1.1936376971157356, + "learning_rate": 1.7122779696892628e-05, + "loss": 1.6552, + "step": 883 + }, + { + "epoch": 0.255565192252096, + "grad_norm": 1.1593207052719288, + "learning_rate": 1.7116336641203147e-05, + "loss": 1.6959, + "step": 884 + }, + { + "epoch": 0.25585429314830876, + "grad_norm": 1.1061563744857061, + "learning_rate": 1.7109887594660336e-05, + "loss": 1.6163, + "step": 885 + }, + { + "epoch": 0.2561433940445215, + "grad_norm": 1.134065834530409, + "learning_rate": 1.710343256269329e-05, + "loss": 1.6726, + "step": 886 + }, + { + "epoch": 0.25643249494073433, + "grad_norm": 1.1359039495462206, + "learning_rate": 1.7096971550736153e-05, + "loss": 1.5318, + "step": 887 + }, + { + "epoch": 0.2567215958369471, + "grad_norm": 1.3387153646158823, + "learning_rate": 1.7090504564228087e-05, + "loss": 1.6655, + "step": 888 + }, + { + "epoch": 0.25701069673315985, + "grad_norm": 1.093747361454369, + "learning_rate": 1.7084031608613297e-05, + "loss": 1.7196, + "step": 889 + }, + { + "epoch": 0.25729979762937266, + "grad_norm": 1.114071184877325, + "learning_rate": 1.7077552689341012e-05, + "loss": 1.5488, + "step": 890 + }, + { + "epoch": 0.2575888985255854, + "grad_norm": 1.1616291244384729, + "learning_rate": 1.7071067811865477e-05, + "loss": 1.5431, + "step": 891 + }, + { + "epoch": 0.25787799942179823, + "grad_norm": 1.1354100602957258, + "learning_rate": 1.7064576981645954e-05, + "loss": 1.6055, + "step": 892 + }, + { + "epoch": 0.258167100318011, + "grad_norm": 1.1522440785073058, + "learning_rate": 1.7058080204146718e-05, + "loss": 1.629, + "step": 893 + }, + { + "epoch": 0.25845620121422375, + "grad_norm": 1.1026615179112005, + "learning_rate": 1.7051577484837043e-05, + "loss": 1.5596, + "step": 894 + }, + { + "epoch": 0.25874530211043656, + "grad_norm": 1.1687730289765201, + "learning_rate": 1.7045068829191218e-05, + "loss": 1.6721, + "step": 895 + }, + { + "epoch": 0.2590344030066493, + "grad_norm": 1.1074104029137306, + "learning_rate": 1.7038554242688526e-05, + "loss": 1.6789, + "step": 896 + }, + { + "epoch": 0.2593235039028621, + "grad_norm": 1.1026942328183036, + "learning_rate": 1.7032033730813236e-05, + "loss": 1.558, + "step": 897 + }, + { + "epoch": 0.2596126047990749, + "grad_norm": 1.1646081156981725, + "learning_rate": 1.7025507299054605e-05, + "loss": 1.601, + "step": 898 + }, + { + "epoch": 0.25990170569528764, + "grad_norm": 1.1271463774298178, + "learning_rate": 1.7018974952906885e-05, + "loss": 1.6538, + "step": 899 + }, + { + "epoch": 0.26019080659150046, + "grad_norm": 1.1091390731718525, + "learning_rate": 1.7012436697869295e-05, + "loss": 1.5819, + "step": 900 + }, + { + "epoch": 0.2604799074877132, + "grad_norm": 1.173447822417614, + "learning_rate": 1.7005892539446035e-05, + "loss": 1.6451, + "step": 901 + }, + { + "epoch": 0.26076900838392597, + "grad_norm": 1.1350959293058749, + "learning_rate": 1.699934248314627e-05, + "loss": 1.6962, + "step": 902 + }, + { + "epoch": 0.2610581092801388, + "grad_norm": 1.1162197648231467, + "learning_rate": 1.699278653448414e-05, + "loss": 1.6747, + "step": 903 + }, + { + "epoch": 0.26134721017635154, + "grad_norm": 1.2139920201386476, + "learning_rate": 1.698622469897873e-05, + "loss": 1.7454, + "step": 904 + }, + { + "epoch": 0.2616363110725643, + "grad_norm": 1.0734616221718876, + "learning_rate": 1.6979656982154095e-05, + "loss": 1.6061, + "step": 905 + }, + { + "epoch": 0.2619254119687771, + "grad_norm": 1.1328541465259365, + "learning_rate": 1.6973083389539233e-05, + "loss": 1.585, + "step": 906 + }, + { + "epoch": 0.26221451286498987, + "grad_norm": 1.1568303460362723, + "learning_rate": 1.6966503926668094e-05, + "loss": 1.6825, + "step": 907 + }, + { + "epoch": 0.2625036137612027, + "grad_norm": 1.196990001143863, + "learning_rate": 1.6959918599079562e-05, + "loss": 1.6538, + "step": 908 + }, + { + "epoch": 0.26279271465741544, + "grad_norm": 1.1342440267026168, + "learning_rate": 1.6953327412317468e-05, + "loss": 1.6075, + "step": 909 + }, + { + "epoch": 0.2630818155536282, + "grad_norm": 1.1596789456175167, + "learning_rate": 1.6946730371930565e-05, + "loss": 1.6929, + "step": 910 + }, + { + "epoch": 0.263370916449841, + "grad_norm": 1.1373065436703054, + "learning_rate": 1.6940127483472543e-05, + "loss": 1.6481, + "step": 911 + }, + { + "epoch": 0.26366001734605377, + "grad_norm": 1.1610606493879128, + "learning_rate": 1.6933518752502013e-05, + "loss": 1.7379, + "step": 912 + }, + { + "epoch": 0.2639491182422666, + "grad_norm": 1.125687311211295, + "learning_rate": 1.69269041845825e-05, + "loss": 1.5849, + "step": 913 + }, + { + "epoch": 0.26423821913847934, + "grad_norm": 1.1489558599701768, + "learning_rate": 1.692028378528244e-05, + "loss": 1.7226, + "step": 914 + }, + { + "epoch": 0.2645273200346921, + "grad_norm": 1.188601400208746, + "learning_rate": 1.6913657560175197e-05, + "loss": 1.6326, + "step": 915 + }, + { + "epoch": 0.2648164209309049, + "grad_norm": 1.113219452373414, + "learning_rate": 1.6907025514839013e-05, + "loss": 1.6478, + "step": 916 + }, + { + "epoch": 0.26510552182711766, + "grad_norm": 1.1593190572859267, + "learning_rate": 1.6900387654857055e-05, + "loss": 1.6093, + "step": 917 + }, + { + "epoch": 0.2653946227233304, + "grad_norm": 1.2134229994183705, + "learning_rate": 1.6893743985817366e-05, + "loss": 1.6903, + "step": 918 + }, + { + "epoch": 0.26568372361954323, + "grad_norm": 1.1993364856075077, + "learning_rate": 1.688709451331289e-05, + "loss": 1.6092, + "step": 919 + }, + { + "epoch": 0.265972824515756, + "grad_norm": 1.0919639751411407, + "learning_rate": 1.688043924294145e-05, + "loss": 1.6078, + "step": 920 + }, + { + "epoch": 0.2662619254119688, + "grad_norm": 1.2013320201783475, + "learning_rate": 1.687377818030575e-05, + "loss": 1.6899, + "step": 921 + }, + { + "epoch": 0.26655102630818156, + "grad_norm": 1.1806015487593893, + "learning_rate": 1.6867111331013384e-05, + "loss": 1.6197, + "step": 922 + }, + { + "epoch": 0.2668401272043943, + "grad_norm": 1.1020975960157413, + "learning_rate": 1.6860438700676797e-05, + "loss": 1.5893, + "step": 923 + }, + { + "epoch": 0.26712922810060713, + "grad_norm": 1.094951519307646, + "learning_rate": 1.685376029491331e-05, + "loss": 1.5254, + "step": 924 + }, + { + "epoch": 0.2674183289968199, + "grad_norm": 1.1163275611243435, + "learning_rate": 1.684707611934511e-05, + "loss": 1.5477, + "step": 925 + }, + { + "epoch": 0.26770742989303264, + "grad_norm": 1.2307806882838923, + "learning_rate": 1.6840386179599242e-05, + "loss": 1.6667, + "step": 926 + }, + { + "epoch": 0.26799653078924546, + "grad_norm": 1.171373289157806, + "learning_rate": 1.683369048130759e-05, + "loss": 1.6172, + "step": 927 + }, + { + "epoch": 0.2682856316854582, + "grad_norm": 1.1886579501994778, + "learning_rate": 1.6826989030106902e-05, + "loss": 1.6133, + "step": 928 + }, + { + "epoch": 0.268574732581671, + "grad_norm": 1.1934353770717314, + "learning_rate": 1.6820281831638757e-05, + "loss": 1.559, + "step": 929 + }, + { + "epoch": 0.2688638334778838, + "grad_norm": 1.1614491161854386, + "learning_rate": 1.6813568891549584e-05, + "loss": 1.5769, + "step": 930 + }, + { + "epoch": 0.26915293437409654, + "grad_norm": 1.1397915171658068, + "learning_rate": 1.680685021549063e-05, + "loss": 1.5537, + "step": 931 + }, + { + "epoch": 0.26944203527030935, + "grad_norm": 1.149992178338205, + "learning_rate": 1.6800125809117984e-05, + "loss": 1.6231, + "step": 932 + }, + { + "epoch": 0.2697311361665221, + "grad_norm": 1.1401360403507486, + "learning_rate": 1.6793395678092555e-05, + "loss": 1.6182, + "step": 933 + }, + { + "epoch": 0.27002023706273487, + "grad_norm": 1.1209347016808902, + "learning_rate": 1.6786659828080078e-05, + "loss": 1.6145, + "step": 934 + }, + { + "epoch": 0.2703093379589477, + "grad_norm": 1.159654576616978, + "learning_rate": 1.6779918264751082e-05, + "loss": 1.707, + "step": 935 + }, + { + "epoch": 0.27059843885516044, + "grad_norm": 1.1184081607191674, + "learning_rate": 1.6773170993780926e-05, + "loss": 1.6816, + "step": 936 + }, + { + "epoch": 0.27088753975137325, + "grad_norm": 1.0985004602757438, + "learning_rate": 1.676641802084977e-05, + "loss": 1.6155, + "step": 937 + }, + { + "epoch": 0.271176640647586, + "grad_norm": 1.1086581131090065, + "learning_rate": 1.6759659351642563e-05, + "loss": 1.5973, + "step": 938 + }, + { + "epoch": 0.27146574154379877, + "grad_norm": 1.1262795371219783, + "learning_rate": 1.6752894991849066e-05, + "loss": 1.6819, + "step": 939 + }, + { + "epoch": 0.2717548424400116, + "grad_norm": 1.1042181463660192, + "learning_rate": 1.674612494716382e-05, + "loss": 1.6242, + "step": 940 + }, + { + "epoch": 0.27204394333622434, + "grad_norm": 1.1225263911796381, + "learning_rate": 1.6739349223286153e-05, + "loss": 1.688, + "step": 941 + }, + { + "epoch": 0.2723330442324371, + "grad_norm": 1.0507156428201587, + "learning_rate": 1.673256782592018e-05, + "loss": 1.6293, + "step": 942 + }, + { + "epoch": 0.2726221451286499, + "grad_norm": 1.1687937216638025, + "learning_rate": 1.6725780760774775e-05, + "loss": 1.5753, + "step": 943 + }, + { + "epoch": 0.27291124602486266, + "grad_norm": 1.0675977484748183, + "learning_rate": 1.671898803356361e-05, + "loss": 1.5865, + "step": 944 + }, + { + "epoch": 0.2732003469210755, + "grad_norm": 1.1223391674880248, + "learning_rate": 1.67121896500051e-05, + "loss": 1.6529, + "step": 945 + }, + { + "epoch": 0.27348944781728823, + "grad_norm": 1.1326572315570196, + "learning_rate": 1.6705385615822436e-05, + "loss": 1.6463, + "step": 946 + }, + { + "epoch": 0.273778548713501, + "grad_norm": 1.2086326669717427, + "learning_rate": 1.669857593674356e-05, + "loss": 1.6377, + "step": 947 + }, + { + "epoch": 0.2740676496097138, + "grad_norm": 1.193825641421145, + "learning_rate": 1.6691760618501163e-05, + "loss": 1.5547, + "step": 948 + }, + { + "epoch": 0.27435675050592656, + "grad_norm": 1.2866576631032545, + "learning_rate": 1.6684939666832695e-05, + "loss": 1.6659, + "step": 949 + }, + { + "epoch": 0.2746458514021394, + "grad_norm": 1.2166777753299653, + "learning_rate": 1.6678113087480334e-05, + "loss": 1.5869, + "step": 950 + }, + { + "epoch": 0.27493495229835213, + "grad_norm": 1.192007342337803, + "learning_rate": 1.6671280886191005e-05, + "loss": 1.6139, + "step": 951 + }, + { + "epoch": 0.2752240531945649, + "grad_norm": 1.1110607953589973, + "learning_rate": 1.6664443068716365e-05, + "loss": 1.5512, + "step": 952 + }, + { + "epoch": 0.2755131540907777, + "grad_norm": 1.1519768318705965, + "learning_rate": 1.6657599640812793e-05, + "loss": 1.5454, + "step": 953 + }, + { + "epoch": 0.27580225498699046, + "grad_norm": 1.1479317865206893, + "learning_rate": 1.6650750608241403e-05, + "loss": 1.6038, + "step": 954 + }, + { + "epoch": 0.2760913558832032, + "grad_norm": 1.0950679969123849, + "learning_rate": 1.664389597676801e-05, + "loss": 1.5197, + "step": 955 + }, + { + "epoch": 0.27638045677941603, + "grad_norm": 1.1421306294859326, + "learning_rate": 1.6637035752163166e-05, + "loss": 1.7071, + "step": 956 + }, + { + "epoch": 0.2766695576756288, + "grad_norm": 1.0604403244112353, + "learning_rate": 1.66301699402021e-05, + "loss": 1.6066, + "step": 957 + }, + { + "epoch": 0.2769586585718416, + "grad_norm": 1.1297323126616883, + "learning_rate": 1.6623298546664773e-05, + "loss": 1.7718, + "step": 958 + }, + { + "epoch": 0.27724775946805436, + "grad_norm": 1.0910744186175767, + "learning_rate": 1.6616421577335834e-05, + "loss": 1.6962, + "step": 959 + }, + { + "epoch": 0.2775368603642671, + "grad_norm": 1.12747611212375, + "learning_rate": 1.6609539038004624e-05, + "loss": 1.7292, + "step": 960 + }, + { + "epoch": 0.2778259612604799, + "grad_norm": 1.093202597412715, + "learning_rate": 1.6602650934465178e-05, + "loss": 1.6183, + "step": 961 + }, + { + "epoch": 0.2781150621566927, + "grad_norm": 1.072162280018483, + "learning_rate": 1.659575727251621e-05, + "loss": 1.5427, + "step": 962 + }, + { + "epoch": 0.27840416305290544, + "grad_norm": 1.1781301893929088, + "learning_rate": 1.658885805796111e-05, + "loss": 1.6114, + "step": 963 + }, + { + "epoch": 0.27869326394911825, + "grad_norm": 1.153766159493101, + "learning_rate": 1.6581953296607963e-05, + "loss": 1.6674, + "step": 964 + }, + { + "epoch": 0.278982364845331, + "grad_norm": 1.1586767056723473, + "learning_rate": 1.657504299426949e-05, + "loss": 1.5837, + "step": 965 + }, + { + "epoch": 0.2792714657415438, + "grad_norm": 1.0755813818211648, + "learning_rate": 1.656812715676311e-05, + "loss": 1.6153, + "step": 966 + }, + { + "epoch": 0.2795605666377566, + "grad_norm": 1.2638124252742915, + "learning_rate": 1.6561205789910885e-05, + "loss": 1.5646, + "step": 967 + }, + { + "epoch": 0.27984966753396934, + "grad_norm": 1.0904862968899787, + "learning_rate": 1.655427889953953e-05, + "loss": 1.6286, + "step": 968 + }, + { + "epoch": 0.28013876843018215, + "grad_norm": 1.1573153681833863, + "learning_rate": 1.6547346491480414e-05, + "loss": 1.6379, + "step": 969 + }, + { + "epoch": 0.2804278693263949, + "grad_norm": 1.1816273912861235, + "learning_rate": 1.654040857156955e-05, + "loss": 1.5763, + "step": 970 + }, + { + "epoch": 0.28071697022260766, + "grad_norm": 1.1643158633524044, + "learning_rate": 1.6533465145647598e-05, + "loss": 1.6681, + "step": 971 + }, + { + "epoch": 0.2810060711188205, + "grad_norm": 1.156818680500087, + "learning_rate": 1.6526516219559843e-05, + "loss": 1.5255, + "step": 972 + }, + { + "epoch": 0.28129517201503323, + "grad_norm": 1.1426079298705811, + "learning_rate": 1.6519561799156202e-05, + "loss": 1.5607, + "step": 973 + }, + { + "epoch": 0.28158427291124605, + "grad_norm": 1.0995905315062742, + "learning_rate": 1.6512601890291226e-05, + "loss": 1.5456, + "step": 974 + }, + { + "epoch": 0.2818733738074588, + "grad_norm": 1.1899775942969868, + "learning_rate": 1.6505636498824074e-05, + "loss": 1.6684, + "step": 975 + }, + { + "epoch": 0.28216247470367156, + "grad_norm": 1.155528745251187, + "learning_rate": 1.6498665630618532e-05, + "loss": 1.5868, + "step": 976 + }, + { + "epoch": 0.2824515755998844, + "grad_norm": 1.1411174440976906, + "learning_rate": 1.649168929154299e-05, + "loss": 1.4712, + "step": 977 + }, + { + "epoch": 0.28274067649609713, + "grad_norm": 1.0860228511069425, + "learning_rate": 1.6484707487470447e-05, + "loss": 1.662, + "step": 978 + }, + { + "epoch": 0.2830297773923099, + "grad_norm": 1.1466851119233112, + "learning_rate": 1.6477720224278493e-05, + "loss": 1.6142, + "step": 979 + }, + { + "epoch": 0.2833188782885227, + "grad_norm": 1.3078811221895994, + "learning_rate": 1.6470727507849332e-05, + "loss": 1.6702, + "step": 980 + }, + { + "epoch": 0.28360797918473546, + "grad_norm": 1.1996795619215384, + "learning_rate": 1.6463729344069736e-05, + "loss": 1.6115, + "step": 981 + }, + { + "epoch": 0.28389708008094827, + "grad_norm": 1.2042463666525114, + "learning_rate": 1.645672573883109e-05, + "loss": 1.6333, + "step": 982 + }, + { + "epoch": 0.28418618097716103, + "grad_norm": 1.1063042405661339, + "learning_rate": 1.6449716698029335e-05, + "loss": 1.6195, + "step": 983 + }, + { + "epoch": 0.2844752818733738, + "grad_norm": 1.1072288580101002, + "learning_rate": 1.6442702227564996e-05, + "loss": 1.6116, + "step": 984 + }, + { + "epoch": 0.2847643827695866, + "grad_norm": 1.2243676300508977, + "learning_rate": 1.643568233334318e-05, + "loss": 1.5821, + "step": 985 + }, + { + "epoch": 0.28505348366579936, + "grad_norm": 1.1664544235823486, + "learning_rate": 1.642865702127354e-05, + "loss": 1.671, + "step": 986 + }, + { + "epoch": 0.28534258456201217, + "grad_norm": 1.1167634055353126, + "learning_rate": 1.6421626297270316e-05, + "loss": 1.6361, + "step": 987 + }, + { + "epoch": 0.2856316854582249, + "grad_norm": 1.1612972833032917, + "learning_rate": 1.6414590167252273e-05, + "loss": 1.5859, + "step": 988 + }, + { + "epoch": 0.2859207863544377, + "grad_norm": 1.213672115046863, + "learning_rate": 1.6407548637142746e-05, + "loss": 1.5757, + "step": 989 + }, + { + "epoch": 0.2862098872506505, + "grad_norm": 1.1300532059127562, + "learning_rate": 1.640050171286962e-05, + "loss": 1.6489, + "step": 990 + }, + { + "epoch": 0.28649898814686325, + "grad_norm": 1.1475160145568735, + "learning_rate": 1.6393449400365307e-05, + "loss": 1.5792, + "step": 991 + }, + { + "epoch": 0.286788089043076, + "grad_norm": 1.171976265527746, + "learning_rate": 1.6386391705566768e-05, + "loss": 1.5027, + "step": 992 + }, + { + "epoch": 0.2870771899392888, + "grad_norm": 1.1698090854087668, + "learning_rate": 1.6379328634415484e-05, + "loss": 1.681, + "step": 993 + }, + { + "epoch": 0.2873662908355016, + "grad_norm": 1.1317960093108423, + "learning_rate": 1.6372260192857465e-05, + "loss": 1.603, + "step": 994 + }, + { + "epoch": 0.2876553917317144, + "grad_norm": 1.179188272729989, + "learning_rate": 1.636518638684325e-05, + "loss": 1.6851, + "step": 995 + }, + { + "epoch": 0.28794449262792715, + "grad_norm": 1.1079735999761406, + "learning_rate": 1.6358107222327884e-05, + "loss": 1.6323, + "step": 996 + }, + { + "epoch": 0.2882335935241399, + "grad_norm": 1.1148610947340667, + "learning_rate": 1.635102270527093e-05, + "loss": 1.5511, + "step": 997 + }, + { + "epoch": 0.2885226944203527, + "grad_norm": 1.184333908207028, + "learning_rate": 1.6343932841636455e-05, + "loss": 1.6631, + "step": 998 + }, + { + "epoch": 0.2888117953165655, + "grad_norm": 1.1154370991272007, + "learning_rate": 1.6336837637393025e-05, + "loss": 1.6557, + "step": 999 + }, + { + "epoch": 0.28910089621277824, + "grad_norm": 1.1055845000132498, + "learning_rate": 1.63297370985137e-05, + "loss": 1.623, + "step": 1000 + }, + { + "epoch": 0.28938999710899105, + "grad_norm": 1.1478774668713727, + "learning_rate": 1.6322631230976035e-05, + "loss": 1.6303, + "step": 1001 + }, + { + "epoch": 0.2896790980052038, + "grad_norm": 1.1260832092736042, + "learning_rate": 1.6315520040762073e-05, + "loss": 1.6627, + "step": 1002 + }, + { + "epoch": 0.2899681989014166, + "grad_norm": 1.1328217900445974, + "learning_rate": 1.6308403533858334e-05, + "loss": 1.6185, + "step": 1003 + }, + { + "epoch": 0.2902572997976294, + "grad_norm": 1.1217019928925396, + "learning_rate": 1.6301281716255814e-05, + "loss": 1.5887, + "step": 1004 + }, + { + "epoch": 0.29054640069384213, + "grad_norm": 1.0662817504674387, + "learning_rate": 1.629415459394998e-05, + "loss": 1.6385, + "step": 1005 + }, + { + "epoch": 0.29083550159005495, + "grad_norm": 1.0503730461501608, + "learning_rate": 1.6287022172940762e-05, + "loss": 1.6164, + "step": 1006 + }, + { + "epoch": 0.2911246024862677, + "grad_norm": 1.0733421544977206, + "learning_rate": 1.627988445923256e-05, + "loss": 1.6689, + "step": 1007 + }, + { + "epoch": 0.29141370338248046, + "grad_norm": 1.0667066239811298, + "learning_rate": 1.6272741458834224e-05, + "loss": 1.6697, + "step": 1008 + }, + { + "epoch": 0.2917028042786933, + "grad_norm": 1.087739247251025, + "learning_rate": 1.6265593177759046e-05, + "loss": 1.6863, + "step": 1009 + }, + { + "epoch": 0.29199190517490603, + "grad_norm": 1.0713166761734145, + "learning_rate": 1.625843962202478e-05, + "loss": 1.5898, + "step": 1010 + }, + { + "epoch": 0.29228100607111884, + "grad_norm": 1.1902495194222464, + "learning_rate": 1.6251280797653606e-05, + "loss": 1.6618, + "step": 1011 + }, + { + "epoch": 0.2925701069673316, + "grad_norm": 1.125478902548264, + "learning_rate": 1.624411671067215e-05, + "loss": 1.5907, + "step": 1012 + }, + { + "epoch": 0.29285920786354436, + "grad_norm": 1.1078237145953016, + "learning_rate": 1.623694736711146e-05, + "loss": 1.7451, + "step": 1013 + }, + { + "epoch": 0.29314830875975717, + "grad_norm": 1.1051877093548326, + "learning_rate": 1.6229772773007016e-05, + "loss": 1.6274, + "step": 1014 + }, + { + "epoch": 0.2934374096559699, + "grad_norm": 1.121714398039267, + "learning_rate": 1.6222592934398715e-05, + "loss": 1.4989, + "step": 1015 + }, + { + "epoch": 0.2937265105521827, + "grad_norm": 1.0925330870668526, + "learning_rate": 1.6215407857330862e-05, + "loss": 1.5753, + "step": 1016 + }, + { + "epoch": 0.2940156114483955, + "grad_norm": 1.086137363905831, + "learning_rate": 1.620821754785219e-05, + "loss": 1.5898, + "step": 1017 + }, + { + "epoch": 0.29430471234460825, + "grad_norm": 1.1237702975359138, + "learning_rate": 1.6201022012015823e-05, + "loss": 1.7232, + "step": 1018 + }, + { + "epoch": 0.29459381324082107, + "grad_norm": 1.1254688780372208, + "learning_rate": 1.619382125587928e-05, + "loss": 1.6816, + "step": 1019 + }, + { + "epoch": 0.2948829141370338, + "grad_norm": 1.123440253692758, + "learning_rate": 1.6186615285504496e-05, + "loss": 1.6185, + "step": 1020 + }, + { + "epoch": 0.2951720150332466, + "grad_norm": 1.0947286679961508, + "learning_rate": 1.617940410695777e-05, + "loss": 1.6673, + "step": 1021 + }, + { + "epoch": 0.2954611159294594, + "grad_norm": 1.1147257497620007, + "learning_rate": 1.6172187726309802e-05, + "loss": 1.5691, + "step": 1022 + }, + { + "epoch": 0.29575021682567215, + "grad_norm": 1.2128326700223733, + "learning_rate": 1.6164966149635664e-05, + "loss": 1.5773, + "step": 1023 + }, + { + "epoch": 0.29603931772188496, + "grad_norm": 1.1251444723700124, + "learning_rate": 1.615773938301481e-05, + "loss": 1.6696, + "step": 1024 + }, + { + "epoch": 0.2963284186180977, + "grad_norm": 1.1144709330002471, + "learning_rate": 1.615050743253106e-05, + "loss": 1.5903, + "step": 1025 + }, + { + "epoch": 0.2966175195143105, + "grad_norm": 1.114572930209085, + "learning_rate": 1.6143270304272588e-05, + "loss": 1.5259, + "step": 1026 + }, + { + "epoch": 0.2969066204105233, + "grad_norm": 1.1041968816612577, + "learning_rate": 1.613602800433194e-05, + "loss": 1.594, + "step": 1027 + }, + { + "epoch": 0.29719572130673605, + "grad_norm": 1.1309087615970796, + "learning_rate": 1.6128780538806012e-05, + "loss": 1.6098, + "step": 1028 + }, + { + "epoch": 0.2974848222029488, + "grad_norm": 1.0655925882162895, + "learning_rate": 1.6121527913796045e-05, + "loss": 1.5776, + "step": 1029 + }, + { + "epoch": 0.2977739230991616, + "grad_norm": 1.153738047328653, + "learning_rate": 1.611427013540763e-05, + "loss": 1.6305, + "step": 1030 + }, + { + "epoch": 0.2980630239953744, + "grad_norm": 1.1277739002652585, + "learning_rate": 1.6107007209750687e-05, + "loss": 1.5917, + "step": 1031 + }, + { + "epoch": 0.2983521248915872, + "grad_norm": 1.1292106867822904, + "learning_rate": 1.6099739142939477e-05, + "loss": 1.644, + "step": 1032 + }, + { + "epoch": 0.29864122578779995, + "grad_norm": 1.1236415518463128, + "learning_rate": 1.6092465941092585e-05, + "loss": 1.6763, + "step": 1033 + }, + { + "epoch": 0.2989303266840127, + "grad_norm": 1.119641797142331, + "learning_rate": 1.6085187610332925e-05, + "loss": 1.6655, + "step": 1034 + }, + { + "epoch": 0.2992194275802255, + "grad_norm": 1.1300850404882916, + "learning_rate": 1.607790415678772e-05, + "loss": 1.6187, + "step": 1035 + }, + { + "epoch": 0.2995085284764383, + "grad_norm": 1.1047280483132156, + "learning_rate": 1.6070615586588516e-05, + "loss": 1.6469, + "step": 1036 + }, + { + "epoch": 0.29979762937265103, + "grad_norm": 1.152691255287852, + "learning_rate": 1.6063321905871152e-05, + "loss": 1.665, + "step": 1037 + }, + { + "epoch": 0.30008673026886384, + "grad_norm": 1.169423494531767, + "learning_rate": 1.605602312077578e-05, + "loss": 1.6282, + "step": 1038 + }, + { + "epoch": 0.3003758311650766, + "grad_norm": 1.2677425058482785, + "learning_rate": 1.6048719237446852e-05, + "loss": 1.6314, + "step": 1039 + }, + { + "epoch": 0.3006649320612894, + "grad_norm": 1.1668503548868798, + "learning_rate": 1.60414102620331e-05, + "loss": 1.6291, + "step": 1040 + }, + { + "epoch": 0.30095403295750217, + "grad_norm": 1.1367727503019929, + "learning_rate": 1.6034096200687552e-05, + "loss": 1.4852, + "step": 1041 + }, + { + "epoch": 0.30124313385371493, + "grad_norm": 1.1323984808855299, + "learning_rate": 1.6026777059567515e-05, + "loss": 1.5893, + "step": 1042 + }, + { + "epoch": 0.30153223474992774, + "grad_norm": 1.1609097650981914, + "learning_rate": 1.601945284483457e-05, + "loss": 1.5886, + "step": 1043 + }, + { + "epoch": 0.3018213356461405, + "grad_norm": 1.1415713259757136, + "learning_rate": 1.601212356265458e-05, + "loss": 1.7002, + "step": 1044 + }, + { + "epoch": 0.30211043654235326, + "grad_norm": 1.227231895248573, + "learning_rate": 1.6004789219197656e-05, + "loss": 1.5416, + "step": 1045 + }, + { + "epoch": 0.30239953743856607, + "grad_norm": 1.163043113422494, + "learning_rate": 1.599744982063818e-05, + "loss": 1.6469, + "step": 1046 + }, + { + "epoch": 0.3026886383347788, + "grad_norm": 1.1126723523837783, + "learning_rate": 1.5990105373154793e-05, + "loss": 1.6411, + "step": 1047 + }, + { + "epoch": 0.30297773923099164, + "grad_norm": 1.122459840411915, + "learning_rate": 1.5982755882930382e-05, + "loss": 1.6519, + "step": 1048 + }, + { + "epoch": 0.3032668401272044, + "grad_norm": 1.5840082766344463, + "learning_rate": 1.5975401356152074e-05, + "loss": 1.6135, + "step": 1049 + }, + { + "epoch": 0.30355594102341715, + "grad_norm": 1.1225058692910097, + "learning_rate": 1.596804179901125e-05, + "loss": 1.5674, + "step": 1050 + }, + { + "epoch": 0.30384504191962997, + "grad_norm": 1.1620815072786224, + "learning_rate": 1.5960677217703512e-05, + "loss": 1.5796, + "step": 1051 + }, + { + "epoch": 0.3041341428158427, + "grad_norm": 1.1236988795621041, + "learning_rate": 1.59533076184287e-05, + "loss": 1.6219, + "step": 1052 + }, + { + "epoch": 0.3044232437120555, + "grad_norm": 1.1118641361809816, + "learning_rate": 1.594593300739087e-05, + "loss": 1.6241, + "step": 1053 + }, + { + "epoch": 0.3047123446082683, + "grad_norm": 1.1947464614628307, + "learning_rate": 1.5938553390798306e-05, + "loss": 1.5793, + "step": 1054 + }, + { + "epoch": 0.30500144550448105, + "grad_norm": 1.0683464597665702, + "learning_rate": 1.59311687748635e-05, + "loss": 1.59, + "step": 1055 + }, + { + "epoch": 0.30529054640069386, + "grad_norm": 1.148680753466922, + "learning_rate": 1.5923779165803163e-05, + "loss": 1.5368, + "step": 1056 + }, + { + "epoch": 0.3055796472969066, + "grad_norm": 1.0975947872426266, + "learning_rate": 1.591638456983819e-05, + "loss": 1.5634, + "step": 1057 + }, + { + "epoch": 0.3058687481931194, + "grad_norm": 1.0758466082542466, + "learning_rate": 1.5908984993193696e-05, + "loss": 1.5855, + "step": 1058 + }, + { + "epoch": 0.3061578490893322, + "grad_norm": 1.140426199545744, + "learning_rate": 1.590158044209897e-05, + "loss": 1.6559, + "step": 1059 + }, + { + "epoch": 0.30644694998554495, + "grad_norm": 1.1060228312495028, + "learning_rate": 1.58941709227875e-05, + "loss": 1.6378, + "step": 1060 + }, + { + "epoch": 0.30673605088175776, + "grad_norm": 1.113526257443253, + "learning_rate": 1.588675644149695e-05, + "loss": 1.6179, + "step": 1061 + }, + { + "epoch": 0.3070251517779705, + "grad_norm": 1.2173138936643453, + "learning_rate": 1.587933700446918e-05, + "loss": 1.6806, + "step": 1062 + }, + { + "epoch": 0.3073142526741833, + "grad_norm": 1.1400857891987628, + "learning_rate": 1.587191261795019e-05, + "loss": 1.6742, + "step": 1063 + }, + { + "epoch": 0.3076033535703961, + "grad_norm": 1.1305997594773096, + "learning_rate": 1.5864483288190174e-05, + "loss": 1.5272, + "step": 1064 + }, + { + "epoch": 0.30789245446660884, + "grad_norm": 1.228007225532647, + "learning_rate": 1.5857049021443467e-05, + "loss": 1.7073, + "step": 1065 + }, + { + "epoch": 0.3081815553628216, + "grad_norm": 1.162916715072549, + "learning_rate": 1.584960982396858e-05, + "loss": 1.633, + "step": 1066 + }, + { + "epoch": 0.3084706562590344, + "grad_norm": 1.1597744213298864, + "learning_rate": 1.5842165702028157e-05, + "loss": 1.4521, + "step": 1067 + }, + { + "epoch": 0.30875975715524717, + "grad_norm": 1.091472479025558, + "learning_rate": 1.5834716661889004e-05, + "loss": 1.5902, + "step": 1068 + }, + { + "epoch": 0.30904885805146, + "grad_norm": 1.1086742754929058, + "learning_rate": 1.5827262709822056e-05, + "loss": 1.5939, + "step": 1069 + }, + { + "epoch": 0.30933795894767274, + "grad_norm": 1.1593052089546423, + "learning_rate": 1.581980385210238e-05, + "loss": 1.5925, + "step": 1070 + }, + { + "epoch": 0.3096270598438855, + "grad_norm": 1.1650634297424989, + "learning_rate": 1.5812340095009192e-05, + "loss": 1.6855, + "step": 1071 + }, + { + "epoch": 0.3099161607400983, + "grad_norm": 1.099899286022252, + "learning_rate": 1.580487144482581e-05, + "loss": 1.5522, + "step": 1072 + }, + { + "epoch": 0.31020526163631107, + "grad_norm": 1.1052999796757328, + "learning_rate": 1.5797397907839678e-05, + "loss": 1.5938, + "step": 1073 + }, + { + "epoch": 0.3104943625325238, + "grad_norm": 1.1060803844176064, + "learning_rate": 1.5789919490342367e-05, + "loss": 1.5835, + "step": 1074 + }, + { + "epoch": 0.31078346342873664, + "grad_norm": 1.1187707670977645, + "learning_rate": 1.578243619862954e-05, + "loss": 1.5327, + "step": 1075 + }, + { + "epoch": 0.3110725643249494, + "grad_norm": 1.1521574761443172, + "learning_rate": 1.5774948039000966e-05, + "loss": 1.5998, + "step": 1076 + }, + { + "epoch": 0.3113616652211622, + "grad_norm": 1.109080218315599, + "learning_rate": 1.5767455017760524e-05, + "loss": 1.6479, + "step": 1077 + }, + { + "epoch": 0.31165076611737497, + "grad_norm": 1.1370919220061753, + "learning_rate": 1.5759957141216175e-05, + "loss": 1.5246, + "step": 1078 + }, + { + "epoch": 0.3119398670135877, + "grad_norm": 1.113530924054974, + "learning_rate": 1.5752454415679968e-05, + "loss": 1.6073, + "step": 1079 + }, + { + "epoch": 0.31222896790980054, + "grad_norm": 1.0973243348785886, + "learning_rate": 1.5744946847468036e-05, + "loss": 1.694, + "step": 1080 + }, + { + "epoch": 0.3125180688060133, + "grad_norm": 1.1647355097669179, + "learning_rate": 1.5737434442900585e-05, + "loss": 1.6135, + "step": 1081 + }, + { + "epoch": 0.31280716970222605, + "grad_norm": 1.1314628251550698, + "learning_rate": 1.5729917208301905e-05, + "loss": 1.7268, + "step": 1082 + }, + { + "epoch": 0.31309627059843886, + "grad_norm": 1.1557900599799165, + "learning_rate": 1.5722395150000336e-05, + "loss": 1.6416, + "step": 1083 + }, + { + "epoch": 0.3133853714946516, + "grad_norm": 1.0866193980050511, + "learning_rate": 1.5714868274328293e-05, + "loss": 1.5172, + "step": 1084 + }, + { + "epoch": 0.31367447239086443, + "grad_norm": 1.158830944861789, + "learning_rate": 1.5707336587622232e-05, + "loss": 1.6342, + "step": 1085 + }, + { + "epoch": 0.3139635732870772, + "grad_norm": 1.1695982419960425, + "learning_rate": 1.569980009622267e-05, + "loss": 1.6118, + "step": 1086 + }, + { + "epoch": 0.31425267418328995, + "grad_norm": 1.095428236784029, + "learning_rate": 1.5692258806474168e-05, + "loss": 1.5944, + "step": 1087 + }, + { + "epoch": 0.31454177507950276, + "grad_norm": 1.1019964357061058, + "learning_rate": 1.568471272472532e-05, + "loss": 1.573, + "step": 1088 + }, + { + "epoch": 0.3148308759757155, + "grad_norm": 1.1256111784675789, + "learning_rate": 1.567716185732876e-05, + "loss": 1.604, + "step": 1089 + }, + { + "epoch": 0.3151199768719283, + "grad_norm": 1.1148637100070335, + "learning_rate": 1.566960621064115e-05, + "loss": 1.617, + "step": 1090 + }, + { + "epoch": 0.3154090777681411, + "grad_norm": 1.0804672597638105, + "learning_rate": 1.566204579102317e-05, + "loss": 1.5047, + "step": 1091 + }, + { + "epoch": 0.31569817866435385, + "grad_norm": 1.1419136802627774, + "learning_rate": 1.5654480604839532e-05, + "loss": 1.5378, + "step": 1092 + }, + { + "epoch": 0.31598727956056666, + "grad_norm": 1.1498293898506018, + "learning_rate": 1.564691065845894e-05, + "loss": 1.5626, + "step": 1093 + }, + { + "epoch": 0.3162763804567794, + "grad_norm": 1.1143430311083644, + "learning_rate": 1.5639335958254123e-05, + "loss": 1.595, + "step": 1094 + }, + { + "epoch": 0.3165654813529922, + "grad_norm": 1.0931970046054698, + "learning_rate": 1.5631756510601802e-05, + "loss": 1.4775, + "step": 1095 + }, + { + "epoch": 0.316854582249205, + "grad_norm": 1.105514673067743, + "learning_rate": 1.5624172321882703e-05, + "loss": 1.6032, + "step": 1096 + }, + { + "epoch": 0.31714368314541774, + "grad_norm": 1.0931739627772648, + "learning_rate": 1.5616583398481533e-05, + "loss": 1.6174, + "step": 1097 + }, + { + "epoch": 0.31743278404163056, + "grad_norm": 1.1211144837250133, + "learning_rate": 1.560898974678699e-05, + "loss": 1.7499, + "step": 1098 + }, + { + "epoch": 0.3177218849378433, + "grad_norm": 1.1209699013931562, + "learning_rate": 1.5601391373191757e-05, + "loss": 1.4995, + "step": 1099 + }, + { + "epoch": 0.31801098583405607, + "grad_norm": 1.0922080473768103, + "learning_rate": 1.559378828409249e-05, + "loss": 1.5329, + "step": 1100 + }, + { + "epoch": 0.3183000867302689, + "grad_norm": 1.1365989497096085, + "learning_rate": 1.5586180485889803e-05, + "loss": 1.6751, + "step": 1101 + }, + { + "epoch": 0.31858918762648164, + "grad_norm": 1.1489329053278985, + "learning_rate": 1.5578567984988292e-05, + "loss": 1.6953, + "step": 1102 + }, + { + "epoch": 0.3188782885226944, + "grad_norm": 1.1566994177367973, + "learning_rate": 1.5570950787796498e-05, + "loss": 1.6297, + "step": 1103 + }, + { + "epoch": 0.3191673894189072, + "grad_norm": 1.1580906580853507, + "learning_rate": 1.5563328900726925e-05, + "loss": 1.5271, + "step": 1104 + }, + { + "epoch": 0.31945649031511997, + "grad_norm": 1.0995229482590019, + "learning_rate": 1.5555702330196024e-05, + "loss": 1.6119, + "step": 1105 + }, + { + "epoch": 0.3197455912113328, + "grad_norm": 1.0799883018230365, + "learning_rate": 1.554807108262418e-05, + "loss": 1.6782, + "step": 1106 + }, + { + "epoch": 0.32003469210754554, + "grad_norm": 1.1073029882869814, + "learning_rate": 1.5540435164435726e-05, + "loss": 1.5397, + "step": 1107 + }, + { + "epoch": 0.3203237930037583, + "grad_norm": 1.1329749767157702, + "learning_rate": 1.5532794582058922e-05, + "loss": 1.7132, + "step": 1108 + }, + { + "epoch": 0.3206128938999711, + "grad_norm": 1.1204291824946706, + "learning_rate": 1.5525149341925953e-05, + "loss": 1.7194, + "step": 1109 + }, + { + "epoch": 0.32090199479618386, + "grad_norm": 1.1098026934922063, + "learning_rate": 1.5517499450472927e-05, + "loss": 1.5843, + "step": 1110 + }, + { + "epoch": 0.3211910956923966, + "grad_norm": 1.1014699968821742, + "learning_rate": 1.550984491413987e-05, + "loss": 1.6016, + "step": 1111 + }, + { + "epoch": 0.32148019658860943, + "grad_norm": 1.130886787299436, + "learning_rate": 1.5502185739370717e-05, + "loss": 1.5793, + "step": 1112 + }, + { + "epoch": 0.3217692974848222, + "grad_norm": 1.1290955593656395, + "learning_rate": 1.5494521932613307e-05, + "loss": 1.6501, + "step": 1113 + }, + { + "epoch": 0.322058398381035, + "grad_norm": 1.0883652426866905, + "learning_rate": 1.548685350031937e-05, + "loss": 1.5592, + "step": 1114 + }, + { + "epoch": 0.32234749927724776, + "grad_norm": 1.1464074008441614, + "learning_rate": 1.547918044894455e-05, + "loss": 1.6295, + "step": 1115 + }, + { + "epoch": 0.3226366001734605, + "grad_norm": 1.0848408826768916, + "learning_rate": 1.5471502784948364e-05, + "loss": 1.6202, + "step": 1116 + }, + { + "epoch": 0.32292570106967333, + "grad_norm": 1.150507963992238, + "learning_rate": 1.5463820514794215e-05, + "loss": 1.5854, + "step": 1117 + }, + { + "epoch": 0.3232148019658861, + "grad_norm": 1.0976974430882895, + "learning_rate": 1.545613364494938e-05, + "loss": 1.5758, + "step": 1118 + }, + { + "epoch": 0.32350390286209885, + "grad_norm": 1.1069482166625944, + "learning_rate": 1.5448442181885026e-05, + "loss": 1.6208, + "step": 1119 + }, + { + "epoch": 0.32379300375831166, + "grad_norm": 1.2077191144956074, + "learning_rate": 1.5440746132076155e-05, + "loss": 1.5925, + "step": 1120 + }, + { + "epoch": 0.3240821046545244, + "grad_norm": 1.1541833705064752, + "learning_rate": 1.5433045502001668e-05, + "loss": 1.6509, + "step": 1121 + }, + { + "epoch": 0.32437120555073723, + "grad_norm": 1.1100819739423928, + "learning_rate": 1.542534029814429e-05, + "loss": 1.5918, + "step": 1122 + }, + { + "epoch": 0.32466030644695, + "grad_norm": 1.1765209537252967, + "learning_rate": 1.5417630526990613e-05, + "loss": 1.6111, + "step": 1123 + }, + { + "epoch": 0.32494940734316274, + "grad_norm": 1.092648262023873, + "learning_rate": 1.540991619503107e-05, + "loss": 1.5288, + "step": 1124 + }, + { + "epoch": 0.32523850823937556, + "grad_norm": 1.1276180188457035, + "learning_rate": 1.540219730875994e-05, + "loss": 1.5382, + "step": 1125 + }, + { + "epoch": 0.3255276091355883, + "grad_norm": 1.571590596776389, + "learning_rate": 1.5394473874675325e-05, + "loss": 1.6576, + "step": 1126 + }, + { + "epoch": 0.32581671003180107, + "grad_norm": 1.118162272885637, + "learning_rate": 1.5386745899279155e-05, + "loss": 1.5767, + "step": 1127 + }, + { + "epoch": 0.3261058109280139, + "grad_norm": 1.0863276376658753, + "learning_rate": 1.5379013389077193e-05, + "loss": 1.5751, + "step": 1128 + }, + { + "epoch": 0.32639491182422664, + "grad_norm": 1.1100894550632194, + "learning_rate": 1.537127635057902e-05, + "loss": 1.4374, + "step": 1129 + }, + { + "epoch": 0.32668401272043945, + "grad_norm": 1.086813502970389, + "learning_rate": 1.5363534790298012e-05, + "loss": 1.6202, + "step": 1130 + }, + { + "epoch": 0.3269731136166522, + "grad_norm": 1.1464824890024201, + "learning_rate": 1.5355788714751378e-05, + "loss": 1.5852, + "step": 1131 + }, + { + "epoch": 0.32726221451286497, + "grad_norm": 1.1665253096605663, + "learning_rate": 1.5348038130460106e-05, + "loss": 1.6235, + "step": 1132 + }, + { + "epoch": 0.3275513154090778, + "grad_norm": 1.137364189884065, + "learning_rate": 1.534028304394899e-05, + "loss": 1.5424, + "step": 1133 + }, + { + "epoch": 0.32784041630529054, + "grad_norm": 1.1333000047594932, + "learning_rate": 1.5332523461746607e-05, + "loss": 1.5362, + "step": 1134 + }, + { + "epoch": 0.32812951720150335, + "grad_norm": 1.189175584600463, + "learning_rate": 1.5324759390385326e-05, + "loss": 1.6452, + "step": 1135 + }, + { + "epoch": 0.3284186180977161, + "grad_norm": 1.8321030345914788, + "learning_rate": 1.53169908364013e-05, + "loss": 1.7045, + "step": 1136 + }, + { + "epoch": 0.32870771899392887, + "grad_norm": 1.1179090119797042, + "learning_rate": 1.530921780633444e-05, + "loss": 1.5984, + "step": 1137 + }, + { + "epoch": 0.3289968198901417, + "grad_norm": 1.1414953551831817, + "learning_rate": 1.5301440306728433e-05, + "loss": 1.6185, + "step": 1138 + }, + { + "epoch": 0.32928592078635444, + "grad_norm": 1.036221101321721, + "learning_rate": 1.5293658344130734e-05, + "loss": 1.5744, + "step": 1139 + }, + { + "epoch": 0.3295750216825672, + "grad_norm": 1.5416356009448273, + "learning_rate": 1.5285871925092545e-05, + "loss": 1.6635, + "step": 1140 + }, + { + "epoch": 0.32986412257878, + "grad_norm": 1.2162733684436655, + "learning_rate": 1.527808105616883e-05, + "loss": 1.6407, + "step": 1141 + }, + { + "epoch": 0.33015322347499276, + "grad_norm": 1.1033030184596553, + "learning_rate": 1.5270285743918294e-05, + "loss": 1.6466, + "step": 1142 + }, + { + "epoch": 0.3304423243712056, + "grad_norm": 1.109124489223676, + "learning_rate": 1.526248599490338e-05, + "loss": 1.7746, + "step": 1143 + }, + { + "epoch": 0.33073142526741833, + "grad_norm": 1.092288021297295, + "learning_rate": 1.5254681815690271e-05, + "loss": 1.526, + "step": 1144 + }, + { + "epoch": 0.3310205261636311, + "grad_norm": 1.1575352020094662, + "learning_rate": 1.5246873212848874e-05, + "loss": 1.5878, + "step": 1145 + }, + { + "epoch": 0.3313096270598439, + "grad_norm": 1.1405141791110793, + "learning_rate": 1.5239060192952821e-05, + "loss": 1.5774, + "step": 1146 + }, + { + "epoch": 0.33159872795605666, + "grad_norm": 1.122962325771197, + "learning_rate": 1.5231242762579474e-05, + "loss": 1.5913, + "step": 1147 + }, + { + "epoch": 0.3318878288522694, + "grad_norm": 1.2373901697470753, + "learning_rate": 1.5223420928309889e-05, + "loss": 1.6305, + "step": 1148 + }, + { + "epoch": 0.33217692974848223, + "grad_norm": 1.091007710090704, + "learning_rate": 1.5215594696728849e-05, + "loss": 1.6023, + "step": 1149 + }, + { + "epoch": 0.332466030644695, + "grad_norm": 1.1020358370583065, + "learning_rate": 1.5207764074424818e-05, + "loss": 1.5833, + "step": 1150 + }, + { + "epoch": 0.3327551315409078, + "grad_norm": 1.1069948831247647, + "learning_rate": 1.5199929067989975e-05, + "loss": 1.6207, + "step": 1151 + }, + { + "epoch": 0.33304423243712056, + "grad_norm": 1.15867755196439, + "learning_rate": 1.5192089684020179e-05, + "loss": 1.6841, + "step": 1152 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 1.06051478046549, + "learning_rate": 1.518424592911498e-05, + "loss": 1.6747, + "step": 1153 + }, + { + "epoch": 0.3336224342295461, + "grad_norm": 1.0941722623539405, + "learning_rate": 1.5176397809877599e-05, + "loss": 1.5649, + "step": 1154 + }, + { + "epoch": 0.3339115351257589, + "grad_norm": 1.065308694755837, + "learning_rate": 1.5168545332914942e-05, + "loss": 1.5981, + "step": 1155 + }, + { + "epoch": 0.33420063602197164, + "grad_norm": 1.103949052624009, + "learning_rate": 1.5160688504837577e-05, + "loss": 1.6367, + "step": 1156 + }, + { + "epoch": 0.33448973691818445, + "grad_norm": 1.108624129747703, + "learning_rate": 1.5152827332259738e-05, + "loss": 1.5322, + "step": 1157 + }, + { + "epoch": 0.3347788378143972, + "grad_norm": 1.171699585963367, + "learning_rate": 1.5144961821799314e-05, + "loss": 1.6257, + "step": 1158 + }, + { + "epoch": 0.33506793871061, + "grad_norm": 1.1189026609089276, + "learning_rate": 1.5137091980077842e-05, + "loss": 1.6154, + "step": 1159 + }, + { + "epoch": 0.3353570396068228, + "grad_norm": 1.0695939123636227, + "learning_rate": 1.5129217813720516e-05, + "loss": 1.6156, + "step": 1160 + }, + { + "epoch": 0.33564614050303554, + "grad_norm": 1.1139701552322105, + "learning_rate": 1.5121339329356163e-05, + "loss": 1.6639, + "step": 1161 + }, + { + "epoch": 0.33593524139924835, + "grad_norm": 1.0899522793859586, + "learning_rate": 1.5113456533617243e-05, + "loss": 1.5977, + "step": 1162 + }, + { + "epoch": 0.3362243422954611, + "grad_norm": 1.1231652645265944, + "learning_rate": 1.5105569433139859e-05, + "loss": 1.5887, + "step": 1163 + }, + { + "epoch": 0.33651344319167387, + "grad_norm": 1.1311200256557958, + "learning_rate": 1.5097678034563716e-05, + "loss": 1.6756, + "step": 1164 + }, + { + "epoch": 0.3368025440878867, + "grad_norm": 1.0808514264048, + "learning_rate": 1.5089782344532155e-05, + "loss": 1.649, + "step": 1165 + }, + { + "epoch": 0.33709164498409944, + "grad_norm": 1.1031304339670052, + "learning_rate": 1.5081882369692129e-05, + "loss": 1.6388, + "step": 1166 + }, + { + "epoch": 0.33738074588031225, + "grad_norm": 1.0793600842313416, + "learning_rate": 1.507397811669418e-05, + "loss": 1.5254, + "step": 1167 + }, + { + "epoch": 0.337669846776525, + "grad_norm": 1.2081295628157755, + "learning_rate": 1.5066069592192483e-05, + "loss": 1.6703, + "step": 1168 + }, + { + "epoch": 0.33795894767273776, + "grad_norm": 1.0883959306735564, + "learning_rate": 1.5058156802844774e-05, + "loss": 1.579, + "step": 1169 + }, + { + "epoch": 0.3382480485689506, + "grad_norm": 1.1328433977876502, + "learning_rate": 1.5050239755312405e-05, + "loss": 1.5702, + "step": 1170 + }, + { + "epoch": 0.33853714946516333, + "grad_norm": 1.1355213706257294, + "learning_rate": 1.5042318456260305e-05, + "loss": 1.6607, + "step": 1171 + }, + { + "epoch": 0.33882625036137615, + "grad_norm": 1.1058142406765221, + "learning_rate": 1.5034392912356973e-05, + "loss": 1.6631, + "step": 1172 + }, + { + "epoch": 0.3391153512575889, + "grad_norm": 1.063531469536154, + "learning_rate": 1.50264631302745e-05, + "loss": 1.6077, + "step": 1173 + }, + { + "epoch": 0.33940445215380166, + "grad_norm": 1.127116435978979, + "learning_rate": 1.5018529116688526e-05, + "loss": 1.585, + "step": 1174 + }, + { + "epoch": 0.3396935530500145, + "grad_norm": 1.0953876410839178, + "learning_rate": 1.5010590878278266e-05, + "loss": 1.5929, + "step": 1175 + }, + { + "epoch": 0.33998265394622723, + "grad_norm": 1.0550371976819637, + "learning_rate": 1.5002648421726487e-05, + "loss": 1.5637, + "step": 1176 + }, + { + "epoch": 0.34027175484244, + "grad_norm": 1.0884184120480185, + "learning_rate": 1.4994701753719505e-05, + "loss": 1.5982, + "step": 1177 + }, + { + "epoch": 0.3405608557386528, + "grad_norm": 1.114897410077523, + "learning_rate": 1.4986750880947184e-05, + "loss": 1.6062, + "step": 1178 + }, + { + "epoch": 0.34084995663486556, + "grad_norm": 1.066891980414234, + "learning_rate": 1.497879581010293e-05, + "loss": 1.606, + "step": 1179 + }, + { + "epoch": 0.34113905753107837, + "grad_norm": 1.0880910736641207, + "learning_rate": 1.4970836547883681e-05, + "loss": 1.5708, + "step": 1180 + }, + { + "epoch": 0.34142815842729113, + "grad_norm": 1.1167707203657147, + "learning_rate": 1.49628731009899e-05, + "loss": 1.5817, + "step": 1181 + }, + { + "epoch": 0.3417172593235039, + "grad_norm": 1.1339896953914148, + "learning_rate": 1.4954905476125578e-05, + "loss": 1.6754, + "step": 1182 + }, + { + "epoch": 0.3420063602197167, + "grad_norm": 1.1010220576391134, + "learning_rate": 1.494693367999822e-05, + "loss": 1.5575, + "step": 1183 + }, + { + "epoch": 0.34229546111592946, + "grad_norm": 1.2739889022885835, + "learning_rate": 1.4938957719318845e-05, + "loss": 1.5497, + "step": 1184 + }, + { + "epoch": 0.3425845620121422, + "grad_norm": 1.148232650591867, + "learning_rate": 1.4930977600801977e-05, + "loss": 1.5948, + "step": 1185 + }, + { + "epoch": 0.342873662908355, + "grad_norm": 1.1837885057183477, + "learning_rate": 1.4922993331165639e-05, + "loss": 1.6017, + "step": 1186 + }, + { + "epoch": 0.3431627638045678, + "grad_norm": 1.1050078395263963, + "learning_rate": 1.4915004917131345e-05, + "loss": 1.6225, + "step": 1187 + }, + { + "epoch": 0.3434518647007806, + "grad_norm": 1.0565673964453821, + "learning_rate": 1.4907012365424112e-05, + "loss": 1.6723, + "step": 1188 + }, + { + "epoch": 0.34374096559699335, + "grad_norm": 1.0426217055278246, + "learning_rate": 1.4899015682772422e-05, + "loss": 1.5567, + "step": 1189 + }, + { + "epoch": 0.3440300664932061, + "grad_norm": 1.0750133699973452, + "learning_rate": 1.489101487590825e-05, + "loss": 1.5984, + "step": 1190 + }, + { + "epoch": 0.3443191673894189, + "grad_norm": 1.0444023205829724, + "learning_rate": 1.488300995156703e-05, + "loss": 1.524, + "step": 1191 + }, + { + "epoch": 0.3446082682856317, + "grad_norm": 1.1113297257235197, + "learning_rate": 1.4875000916487674e-05, + "loss": 1.6477, + "step": 1192 + }, + { + "epoch": 0.34489736918184444, + "grad_norm": 1.0700189539478588, + "learning_rate": 1.4866987777412548e-05, + "loss": 1.6747, + "step": 1193 + }, + { + "epoch": 0.34518647007805725, + "grad_norm": 1.0791046704423155, + "learning_rate": 1.485897054108747e-05, + "loss": 1.5364, + "step": 1194 + }, + { + "epoch": 0.34547557097427, + "grad_norm": 1.1152714216930488, + "learning_rate": 1.4850949214261717e-05, + "loss": 1.6436, + "step": 1195 + }, + { + "epoch": 0.3457646718704828, + "grad_norm": 1.1163524885739, + "learning_rate": 1.4842923803687999e-05, + "loss": 1.6719, + "step": 1196 + }, + { + "epoch": 0.3460537727666956, + "grad_norm": 1.1470433328057497, + "learning_rate": 1.483489431612247e-05, + "loss": 1.6618, + "step": 1197 + }, + { + "epoch": 0.34634287366290833, + "grad_norm": 1.1152658216672549, + "learning_rate": 1.482686075832472e-05, + "loss": 1.6082, + "step": 1198 + }, + { + "epoch": 0.34663197455912115, + "grad_norm": 1.087512082279524, + "learning_rate": 1.4818823137057751e-05, + "loss": 1.6302, + "step": 1199 + }, + { + "epoch": 0.3469210754553339, + "grad_norm": 1.1269711018498454, + "learning_rate": 1.4810781459088009e-05, + "loss": 1.6265, + "step": 1200 + }, + { + "epoch": 0.34721017635154666, + "grad_norm": 1.1738148200035237, + "learning_rate": 1.4802735731185327e-05, + "loss": 1.7348, + "step": 1201 + }, + { + "epoch": 0.3474992772477595, + "grad_norm": 1.0479144362222594, + "learning_rate": 1.4794685960122967e-05, + "loss": 1.5731, + "step": 1202 + }, + { + "epoch": 0.34778837814397223, + "grad_norm": 1.1183507786337057, + "learning_rate": 1.4786632152677596e-05, + "loss": 1.6627, + "step": 1203 + }, + { + "epoch": 0.34807747904018504, + "grad_norm": 1.0560834768468748, + "learning_rate": 1.4778574315629262e-05, + "loss": 1.5118, + "step": 1204 + }, + { + "epoch": 0.3483665799363978, + "grad_norm": 1.1504897519151327, + "learning_rate": 1.4770512455761425e-05, + "loss": 1.6384, + "step": 1205 + }, + { + "epoch": 0.34865568083261056, + "grad_norm": 1.128971905615821, + "learning_rate": 1.476244657986092e-05, + "loss": 1.6184, + "step": 1206 + }, + { + "epoch": 0.34894478172882337, + "grad_norm": 1.1323411355011963, + "learning_rate": 1.4754376694717958e-05, + "loss": 1.742, + "step": 1207 + }, + { + "epoch": 0.34923388262503613, + "grad_norm": 1.0865319922492744, + "learning_rate": 1.4746302807126142e-05, + "loss": 1.5589, + "step": 1208 + }, + { + "epoch": 0.34952298352124894, + "grad_norm": 1.1013771826316385, + "learning_rate": 1.473822492388243e-05, + "loss": 1.5835, + "step": 1209 + }, + { + "epoch": 0.3498120844174617, + "grad_norm": 1.133500512314171, + "learning_rate": 1.4730143051787151e-05, + "loss": 1.5905, + "step": 1210 + }, + { + "epoch": 0.35010118531367446, + "grad_norm": 1.0512972757957857, + "learning_rate": 1.4722057197643986e-05, + "loss": 1.5644, + "step": 1211 + }, + { + "epoch": 0.35039028620988727, + "grad_norm": 1.0982556984479055, + "learning_rate": 1.4713967368259981e-05, + "loss": 1.617, + "step": 1212 + }, + { + "epoch": 0.3506793871061, + "grad_norm": 1.1056551102940748, + "learning_rate": 1.4705873570445506e-05, + "loss": 1.586, + "step": 1213 + }, + { + "epoch": 0.3509684880023128, + "grad_norm": 1.1194856012781043, + "learning_rate": 1.4697775811014294e-05, + "loss": 1.6349, + "step": 1214 + }, + { + "epoch": 0.3512575888985256, + "grad_norm": 1.1635585093324126, + "learning_rate": 1.4689674096783404e-05, + "loss": 1.7592, + "step": 1215 + }, + { + "epoch": 0.35154668979473835, + "grad_norm": 1.0761660420536625, + "learning_rate": 1.4681568434573221e-05, + "loss": 1.6587, + "step": 1216 + }, + { + "epoch": 0.35183579069095117, + "grad_norm": 1.0862955186287286, + "learning_rate": 1.4673458831207463e-05, + "loss": 1.6006, + "step": 1217 + }, + { + "epoch": 0.3521248915871639, + "grad_norm": 1.0784753202499149, + "learning_rate": 1.4665345293513154e-05, + "loss": 1.5591, + "step": 1218 + }, + { + "epoch": 0.3524139924833767, + "grad_norm": 1.1153278904671513, + "learning_rate": 1.4657227828320637e-05, + "loss": 1.5676, + "step": 1219 + }, + { + "epoch": 0.3527030933795895, + "grad_norm": 1.0768215016106648, + "learning_rate": 1.4649106442463565e-05, + "loss": 1.5335, + "step": 1220 + }, + { + "epoch": 0.35299219427580225, + "grad_norm": 1.1314763789684028, + "learning_rate": 1.4640981142778883e-05, + "loss": 1.6171, + "step": 1221 + }, + { + "epoch": 0.353281295172015, + "grad_norm": 1.0990493963931165, + "learning_rate": 1.4632851936106838e-05, + "loss": 1.6048, + "step": 1222 + }, + { + "epoch": 0.3535703960682278, + "grad_norm": 1.085058079934556, + "learning_rate": 1.462471882929096e-05, + "loss": 1.6296, + "step": 1223 + }, + { + "epoch": 0.3538594969644406, + "grad_norm": 1.132646612791557, + "learning_rate": 1.4616581829178065e-05, + "loss": 1.6962, + "step": 1224 + }, + { + "epoch": 0.3541485978606534, + "grad_norm": 1.08766264140374, + "learning_rate": 1.4608440942618252e-05, + "loss": 1.5844, + "step": 1225 + }, + { + "epoch": 0.35443769875686615, + "grad_norm": 1.0655552716344114, + "learning_rate": 1.4600296176464878e-05, + "loss": 1.6268, + "step": 1226 + }, + { + "epoch": 0.3547267996530789, + "grad_norm": 1.0715746687267362, + "learning_rate": 1.4592147537574585e-05, + "loss": 1.7251, + "step": 1227 + }, + { + "epoch": 0.3550159005492917, + "grad_norm": 1.153862024978196, + "learning_rate": 1.4583995032807259e-05, + "loss": 1.5546, + "step": 1228 + }, + { + "epoch": 0.3553050014455045, + "grad_norm": 1.0940461925978813, + "learning_rate": 1.4575838669026045e-05, + "loss": 1.5783, + "step": 1229 + }, + { + "epoch": 0.35559410234171723, + "grad_norm": 1.088784612051587, + "learning_rate": 1.4567678453097345e-05, + "loss": 1.6016, + "step": 1230 + }, + { + "epoch": 0.35588320323793005, + "grad_norm": 1.1233954959369794, + "learning_rate": 1.4559514391890788e-05, + "loss": 1.642, + "step": 1231 + }, + { + "epoch": 0.3561723041341428, + "grad_norm": 1.0795613752451976, + "learning_rate": 1.4551346492279257e-05, + "loss": 1.5359, + "step": 1232 + }, + { + "epoch": 0.3564614050303556, + "grad_norm": 1.1164982431881545, + "learning_rate": 1.4543174761138852e-05, + "loss": 1.537, + "step": 1233 + }, + { + "epoch": 0.3567505059265684, + "grad_norm": 1.0439396898471973, + "learning_rate": 1.4534999205348909e-05, + "loss": 1.5957, + "step": 1234 + }, + { + "epoch": 0.35703960682278113, + "grad_norm": 1.094478459597531, + "learning_rate": 1.4526819831791983e-05, + "loss": 1.5261, + "step": 1235 + }, + { + "epoch": 0.35732870771899394, + "grad_norm": 1.0523765417826885, + "learning_rate": 1.451863664735383e-05, + "loss": 1.7109, + "step": 1236 + }, + { + "epoch": 0.3576178086152067, + "grad_norm": 1.0697972681513481, + "learning_rate": 1.4510449658923431e-05, + "loss": 1.5079, + "step": 1237 + }, + { + "epoch": 0.3579069095114195, + "grad_norm": 1.0776837202479659, + "learning_rate": 1.4502258873392963e-05, + "loss": 1.5373, + "step": 1238 + }, + { + "epoch": 0.35819601040763227, + "grad_norm": 1.0948923626052913, + "learning_rate": 1.4494064297657795e-05, + "loss": 1.5953, + "step": 1239 + }, + { + "epoch": 0.358485111303845, + "grad_norm": 1.0851502014434444, + "learning_rate": 1.4485865938616495e-05, + "loss": 1.6227, + "step": 1240 + }, + { + "epoch": 0.35877421220005784, + "grad_norm": 1.1036122017363585, + "learning_rate": 1.4477663803170807e-05, + "loss": 1.6181, + "step": 1241 + }, + { + "epoch": 0.3590633130962706, + "grad_norm": 1.073741412650114, + "learning_rate": 1.446945789822566e-05, + "loss": 1.6449, + "step": 1242 + }, + { + "epoch": 0.35935241399248335, + "grad_norm": 1.1071745305977874, + "learning_rate": 1.446124823068916e-05, + "loss": 1.6455, + "step": 1243 + }, + { + "epoch": 0.35964151488869617, + "grad_norm": 1.1469504663679086, + "learning_rate": 1.445303480747257e-05, + "loss": 1.5989, + "step": 1244 + }, + { + "epoch": 0.3599306157849089, + "grad_norm": 1.1129037111462516, + "learning_rate": 1.4444817635490322e-05, + "loss": 1.6613, + "step": 1245 + }, + { + "epoch": 0.36021971668112174, + "grad_norm": 1.0968054047932645, + "learning_rate": 1.4436596721660002e-05, + "loss": 1.601, + "step": 1246 + }, + { + "epoch": 0.3605088175773345, + "grad_norm": 1.0927563331941559, + "learning_rate": 1.4428372072902344e-05, + "loss": 1.7164, + "step": 1247 + }, + { + "epoch": 0.36079791847354725, + "grad_norm": 1.1272073503734452, + "learning_rate": 1.4420143696141233e-05, + "loss": 1.5526, + "step": 1248 + }, + { + "epoch": 0.36108701936976006, + "grad_norm": 1.0615706091543697, + "learning_rate": 1.4411911598303688e-05, + "loss": 1.5344, + "step": 1249 + }, + { + "epoch": 0.3613761202659728, + "grad_norm": 1.0434496363748602, + "learning_rate": 1.4403675786319856e-05, + "loss": 1.5325, + "step": 1250 + }, + { + "epoch": 0.3616652211621856, + "grad_norm": 1.1158624940443362, + "learning_rate": 1.4395436267123017e-05, + "loss": 1.6085, + "step": 1251 + }, + { + "epoch": 0.3619543220583984, + "grad_norm": 1.1545246739231232, + "learning_rate": 1.4387193047649572e-05, + "loss": 1.6314, + "step": 1252 + }, + { + "epoch": 0.36224342295461115, + "grad_norm": 1.1500739858412437, + "learning_rate": 1.437894613483903e-05, + "loss": 1.5861, + "step": 1253 + }, + { + "epoch": 0.36253252385082396, + "grad_norm": 1.021310863594055, + "learning_rate": 1.4370695535634024e-05, + "loss": 1.509, + "step": 1254 + }, + { + "epoch": 0.3628216247470367, + "grad_norm": 1.086222036655496, + "learning_rate": 1.4362441256980271e-05, + "loss": 1.6284, + "step": 1255 + }, + { + "epoch": 0.3631107256432495, + "grad_norm": 1.120109032505505, + "learning_rate": 1.4354183305826601e-05, + "loss": 1.5861, + "step": 1256 + }, + { + "epoch": 0.3633998265394623, + "grad_norm": 1.0968623723093054, + "learning_rate": 1.4345921689124928e-05, + "loss": 1.6583, + "step": 1257 + }, + { + "epoch": 0.36368892743567505, + "grad_norm": 1.0757768358923314, + "learning_rate": 1.4337656413830252e-05, + "loss": 1.6732, + "step": 1258 + }, + { + "epoch": 0.3639780283318878, + "grad_norm": 1.1072185273450617, + "learning_rate": 1.4329387486900659e-05, + "loss": 1.5613, + "step": 1259 + }, + { + "epoch": 0.3642671292281006, + "grad_norm": 1.1152570699608948, + "learning_rate": 1.4321114915297298e-05, + "loss": 1.7207, + "step": 1260 + }, + { + "epoch": 0.3645562301243134, + "grad_norm": 1.1181703343533083, + "learning_rate": 1.4312838705984403e-05, + "loss": 1.5774, + "step": 1261 + }, + { + "epoch": 0.3648453310205262, + "grad_norm": 1.1234819294312621, + "learning_rate": 1.4304558865929251e-05, + "loss": 1.5807, + "step": 1262 + }, + { + "epoch": 0.36513443191673894, + "grad_norm": 1.1065221175707145, + "learning_rate": 1.429627540210219e-05, + "loss": 1.5762, + "step": 1263 + }, + { + "epoch": 0.3654235328129517, + "grad_norm": 1.061140154200443, + "learning_rate": 1.4287988321476615e-05, + "loss": 1.6927, + "step": 1264 + }, + { + "epoch": 0.3657126337091645, + "grad_norm": 1.0999054498119034, + "learning_rate": 1.4279697631028961e-05, + "loss": 1.5921, + "step": 1265 + }, + { + "epoch": 0.36600173460537727, + "grad_norm": 1.1096047939134852, + "learning_rate": 1.4271403337738706e-05, + "loss": 1.5883, + "step": 1266 + }, + { + "epoch": 0.36629083550159003, + "grad_norm": 1.117311000405351, + "learning_rate": 1.4263105448588362e-05, + "loss": 1.4585, + "step": 1267 + }, + { + "epoch": 0.36657993639780284, + "grad_norm": 1.1164044045072312, + "learning_rate": 1.4254803970563464e-05, + "loss": 1.6306, + "step": 1268 + }, + { + "epoch": 0.3668690372940156, + "grad_norm": 1.0628514308268988, + "learning_rate": 1.4246498910652573e-05, + "loss": 1.5824, + "step": 1269 + }, + { + "epoch": 0.3671581381902284, + "grad_norm": 1.181064137470451, + "learning_rate": 1.4238190275847262e-05, + "loss": 1.6634, + "step": 1270 + }, + { + "epoch": 0.36744723908644117, + "grad_norm": 1.0887767912574196, + "learning_rate": 1.4229878073142114e-05, + "loss": 1.5741, + "step": 1271 + }, + { + "epoch": 0.3677363399826539, + "grad_norm": 1.140612642718485, + "learning_rate": 1.4221562309534722e-05, + "loss": 1.6199, + "step": 1272 + }, + { + "epoch": 0.36802544087886674, + "grad_norm": 1.1783854536403464, + "learning_rate": 1.4213242992025662e-05, + "loss": 1.6202, + "step": 1273 + }, + { + "epoch": 0.3683145417750795, + "grad_norm": 1.2403828595576905, + "learning_rate": 1.4204920127618517e-05, + "loss": 1.652, + "step": 1274 + }, + { + "epoch": 0.3686036426712923, + "grad_norm": 1.14066458994363, + "learning_rate": 1.419659372331985e-05, + "loss": 1.7794, + "step": 1275 + }, + { + "epoch": 0.36889274356750507, + "grad_norm": 1.093343920812435, + "learning_rate": 1.41882637861392e-05, + "loss": 1.5567, + "step": 1276 + }, + { + "epoch": 0.3691818444637178, + "grad_norm": 1.0808707505493302, + "learning_rate": 1.4179930323089088e-05, + "loss": 1.5926, + "step": 1277 + }, + { + "epoch": 0.36947094535993064, + "grad_norm": 1.0509905823110333, + "learning_rate": 1.4171593341184997e-05, + "loss": 1.5042, + "step": 1278 + }, + { + "epoch": 0.3697600462561434, + "grad_norm": 1.1944099137159172, + "learning_rate": 1.4163252847445375e-05, + "loss": 1.6389, + "step": 1279 + }, + { + "epoch": 0.37004914715235615, + "grad_norm": 1.0550018481082393, + "learning_rate": 1.4154908848891626e-05, + "loss": 1.6389, + "step": 1280 + }, + { + "epoch": 0.37033824804856896, + "grad_norm": 1.144736846632199, + "learning_rate": 1.4146561352548109e-05, + "loss": 1.5658, + "step": 1281 + }, + { + "epoch": 0.3706273489447817, + "grad_norm": 1.1625366943570545, + "learning_rate": 1.4138210365442113e-05, + "loss": 1.6057, + "step": 1282 + }, + { + "epoch": 0.37091644984099453, + "grad_norm": 1.1240282635958534, + "learning_rate": 1.4129855894603885e-05, + "loss": 1.6278, + "step": 1283 + }, + { + "epoch": 0.3712055507372073, + "grad_norm": 1.120499316106146, + "learning_rate": 1.4121497947066593e-05, + "loss": 1.5968, + "step": 1284 + }, + { + "epoch": 0.37149465163342005, + "grad_norm": 1.1248421514320672, + "learning_rate": 1.4113136529866331e-05, + "loss": 1.6972, + "step": 1285 + }, + { + "epoch": 0.37178375252963286, + "grad_norm": 1.1229415098527773, + "learning_rate": 1.4104771650042122e-05, + "loss": 1.5947, + "step": 1286 + }, + { + "epoch": 0.3720728534258456, + "grad_norm": 1.0792013166290457, + "learning_rate": 1.4096403314635897e-05, + "loss": 1.5431, + "step": 1287 + }, + { + "epoch": 0.3723619543220584, + "grad_norm": 1.1584266878751586, + "learning_rate": 1.4088031530692497e-05, + "loss": 1.579, + "step": 1288 + }, + { + "epoch": 0.3726510552182712, + "grad_norm": 1.1181226193149765, + "learning_rate": 1.4079656305259673e-05, + "loss": 1.5989, + "step": 1289 + }, + { + "epoch": 0.37294015611448394, + "grad_norm": 1.139645588978396, + "learning_rate": 1.407127764538806e-05, + "loss": 1.5868, + "step": 1290 + }, + { + "epoch": 0.37322925701069676, + "grad_norm": 1.126265249898816, + "learning_rate": 1.40628955581312e-05, + "loss": 1.6933, + "step": 1291 + }, + { + "epoch": 0.3735183579069095, + "grad_norm": 1.0295263942219146, + "learning_rate": 1.4054510050545505e-05, + "loss": 1.587, + "step": 1292 + }, + { + "epoch": 0.3738074588031223, + "grad_norm": 1.1795604406769806, + "learning_rate": 1.4046121129690278e-05, + "loss": 1.6097, + "step": 1293 + }, + { + "epoch": 0.3740965596993351, + "grad_norm": 1.1403420688505401, + "learning_rate": 1.4037728802627693e-05, + "loss": 1.6267, + "step": 1294 + }, + { + "epoch": 0.37438566059554784, + "grad_norm": 1.0993080419061012, + "learning_rate": 1.4029333076422782e-05, + "loss": 1.671, + "step": 1295 + }, + { + "epoch": 0.3746747614917606, + "grad_norm": 1.1780757010441563, + "learning_rate": 1.4020933958143455e-05, + "loss": 1.7137, + "step": 1296 + }, + { + "epoch": 0.3749638623879734, + "grad_norm": 1.1243870924066819, + "learning_rate": 1.4012531454860464e-05, + "loss": 1.7021, + "step": 1297 + }, + { + "epoch": 0.37525296328418617, + "grad_norm": 1.0999261176344182, + "learning_rate": 1.4004125573647414e-05, + "loss": 1.5531, + "step": 1298 + }, + { + "epoch": 0.375542064180399, + "grad_norm": 1.1292673013365175, + "learning_rate": 1.399571632158076e-05, + "loss": 1.5297, + "step": 1299 + }, + { + "epoch": 0.37583116507661174, + "grad_norm": 1.0839707027340657, + "learning_rate": 1.3987303705739787e-05, + "loss": 1.6223, + "step": 1300 + }, + { + "epoch": 0.3761202659728245, + "grad_norm": 1.1689455340590824, + "learning_rate": 1.3978887733206614e-05, + "loss": 1.602, + "step": 1301 + }, + { + "epoch": 0.3764093668690373, + "grad_norm": 1.037131778360143, + "learning_rate": 1.3970468411066187e-05, + "loss": 1.5195, + "step": 1302 + }, + { + "epoch": 0.37669846776525007, + "grad_norm": 1.099003065082928, + "learning_rate": 1.3962045746406273e-05, + "loss": 1.5807, + "step": 1303 + }, + { + "epoch": 0.3769875686614628, + "grad_norm": 1.1281144891295398, + "learning_rate": 1.395361974631745e-05, + "loss": 1.6533, + "step": 1304 + }, + { + "epoch": 0.37727666955767564, + "grad_norm": 1.1355144650861142, + "learning_rate": 1.3945190417893106e-05, + "loss": 1.5925, + "step": 1305 + }, + { + "epoch": 0.3775657704538884, + "grad_norm": 1.095859210884497, + "learning_rate": 1.3936757768229428e-05, + "loss": 1.6053, + "step": 1306 + }, + { + "epoch": 0.3778548713501012, + "grad_norm": 1.1282184354527713, + "learning_rate": 1.39283218044254e-05, + "loss": 1.6052, + "step": 1307 + }, + { + "epoch": 0.37814397224631396, + "grad_norm": 1.0831491499806725, + "learning_rate": 1.3919882533582804e-05, + "loss": 1.5087, + "step": 1308 + }, + { + "epoch": 0.3784330731425267, + "grad_norm": 1.116233937615999, + "learning_rate": 1.3911439962806192e-05, + "loss": 1.5589, + "step": 1309 + }, + { + "epoch": 0.37872217403873953, + "grad_norm": 1.0633282902202397, + "learning_rate": 1.3902994099202904e-05, + "loss": 1.6164, + "step": 1310 + }, + { + "epoch": 0.3790112749349523, + "grad_norm": 1.0952692825870283, + "learning_rate": 1.3894544949883046e-05, + "loss": 1.7074, + "step": 1311 + }, + { + "epoch": 0.3793003758311651, + "grad_norm": 1.0824385592753576, + "learning_rate": 1.3886092521959497e-05, + "loss": 1.5499, + "step": 1312 + }, + { + "epoch": 0.37958947672737786, + "grad_norm": 1.0480969133383606, + "learning_rate": 1.3877636822547894e-05, + "loss": 1.6255, + "step": 1313 + }, + { + "epoch": 0.3798785776235906, + "grad_norm": 1.1345684718964264, + "learning_rate": 1.386917785876662e-05, + "loss": 1.6102, + "step": 1314 + }, + { + "epoch": 0.38016767851980343, + "grad_norm": 1.1080824270324598, + "learning_rate": 1.3860715637736817e-05, + "loss": 1.5263, + "step": 1315 + }, + { + "epoch": 0.3804567794160162, + "grad_norm": 1.059203635427344, + "learning_rate": 1.3852250166582364e-05, + "loss": 1.6089, + "step": 1316 + }, + { + "epoch": 0.38074588031222895, + "grad_norm": 1.087510027682475, + "learning_rate": 1.3843781452429875e-05, + "loss": 1.541, + "step": 1317 + }, + { + "epoch": 0.38103498120844176, + "grad_norm": 1.086247443147945, + "learning_rate": 1.3835309502408696e-05, + "loss": 1.632, + "step": 1318 + }, + { + "epoch": 0.3813240821046545, + "grad_norm": 1.1441906470265728, + "learning_rate": 1.3826834323650899e-05, + "loss": 1.6891, + "step": 1319 + }, + { + "epoch": 0.38161318300086733, + "grad_norm": 1.1166198316687277, + "learning_rate": 1.381835592329127e-05, + "loss": 1.5583, + "step": 1320 + }, + { + "epoch": 0.3819022838970801, + "grad_norm": 1.0406271039215775, + "learning_rate": 1.380987430846731e-05, + "loss": 1.5878, + "step": 1321 + }, + { + "epoch": 0.38219138479329284, + "grad_norm": 1.1782259448538297, + "learning_rate": 1.3801389486319223e-05, + "loss": 1.5041, + "step": 1322 + }, + { + "epoch": 0.38248048568950566, + "grad_norm": 1.119306255364395, + "learning_rate": 1.3792901463989918e-05, + "loss": 1.668, + "step": 1323 + }, + { + "epoch": 0.3827695865857184, + "grad_norm": 1.0695978855808004, + "learning_rate": 1.3784410248624996e-05, + "loss": 1.6487, + "step": 1324 + }, + { + "epoch": 0.38305868748193117, + "grad_norm": 1.0616454533037485, + "learning_rate": 1.377591584737274e-05, + "loss": 1.5401, + "step": 1325 + }, + { + "epoch": 0.383347788378144, + "grad_norm": 1.0774865869206818, + "learning_rate": 1.3767418267384128e-05, + "loss": 1.6066, + "step": 1326 + }, + { + "epoch": 0.38363688927435674, + "grad_norm": 1.060694994399504, + "learning_rate": 1.3758917515812798e-05, + "loss": 1.6073, + "step": 1327 + }, + { + "epoch": 0.38392599017056955, + "grad_norm": 1.0903641939801798, + "learning_rate": 1.3750413599815074e-05, + "loss": 1.6114, + "step": 1328 + }, + { + "epoch": 0.3842150910667823, + "grad_norm": 1.114947234018338, + "learning_rate": 1.3741906526549928e-05, + "loss": 1.6328, + "step": 1329 + }, + { + "epoch": 0.38450419196299507, + "grad_norm": 1.1067463745265371, + "learning_rate": 1.3733396303179005e-05, + "loss": 1.6358, + "step": 1330 + }, + { + "epoch": 0.3847932928592079, + "grad_norm": 1.1577228624063416, + "learning_rate": 1.3724882936866596e-05, + "loss": 1.5758, + "step": 1331 + }, + { + "epoch": 0.38508239375542064, + "grad_norm": 1.1558637576676867, + "learning_rate": 1.3716366434779628e-05, + "loss": 1.572, + "step": 1332 + }, + { + "epoch": 0.3853714946516334, + "grad_norm": 1.1143169382091267, + "learning_rate": 1.3707846804087686e-05, + "loss": 1.6113, + "step": 1333 + }, + { + "epoch": 0.3856605955478462, + "grad_norm": 1.130544176598161, + "learning_rate": 1.3699324051962974e-05, + "loss": 1.5409, + "step": 1334 + }, + { + "epoch": 0.38594969644405897, + "grad_norm": 1.0366446457013492, + "learning_rate": 1.3690798185580337e-05, + "loss": 1.5622, + "step": 1335 + }, + { + "epoch": 0.3862387973402718, + "grad_norm": 1.1288006310703098, + "learning_rate": 1.3682269212117223e-05, + "loss": 1.6841, + "step": 1336 + }, + { + "epoch": 0.38652789823648454, + "grad_norm": 1.1063560145710902, + "learning_rate": 1.3673737138753713e-05, + "loss": 1.6939, + "step": 1337 + }, + { + "epoch": 0.3868169991326973, + "grad_norm": 1.1459823915842076, + "learning_rate": 1.3665201972672492e-05, + "loss": 1.6378, + "step": 1338 + }, + { + "epoch": 0.3871061000289101, + "grad_norm": 1.1039738007856277, + "learning_rate": 1.3656663721058844e-05, + "loss": 1.5349, + "step": 1339 + }, + { + "epoch": 0.38739520092512286, + "grad_norm": 1.1237038119054343, + "learning_rate": 1.3648122391100663e-05, + "loss": 1.5913, + "step": 1340 + }, + { + "epoch": 0.3876843018213356, + "grad_norm": 1.0705587484764698, + "learning_rate": 1.3639577989988417e-05, + "loss": 1.5015, + "step": 1341 + }, + { + "epoch": 0.38797340271754843, + "grad_norm": 1.1425898222093245, + "learning_rate": 1.3631030524915171e-05, + "loss": 1.638, + "step": 1342 + }, + { + "epoch": 0.3882625036137612, + "grad_norm": 1.138358366597531, + "learning_rate": 1.362248000307657e-05, + "loss": 1.5921, + "step": 1343 + }, + { + "epoch": 0.388551604509974, + "grad_norm": 1.0937645241618539, + "learning_rate": 1.3613926431670824e-05, + "loss": 1.479, + "step": 1344 + }, + { + "epoch": 0.38884070540618676, + "grad_norm": 1.113613288350207, + "learning_rate": 1.360536981789872e-05, + "loss": 1.618, + "step": 1345 + }, + { + "epoch": 0.3891298063023995, + "grad_norm": 1.1367651938150571, + "learning_rate": 1.3596810168963599e-05, + "loss": 1.5512, + "step": 1346 + }, + { + "epoch": 0.38941890719861233, + "grad_norm": 1.079517622534508, + "learning_rate": 1.358824749207136e-05, + "loss": 1.5395, + "step": 1347 + }, + { + "epoch": 0.3897080080948251, + "grad_norm": 1.143875281810773, + "learning_rate": 1.3579681794430454e-05, + "loss": 1.6185, + "step": 1348 + }, + { + "epoch": 0.3899971089910379, + "grad_norm": 1.1732842322479438, + "learning_rate": 1.3571113083251865e-05, + "loss": 1.6675, + "step": 1349 + }, + { + "epoch": 0.39028620988725066, + "grad_norm": 1.1246289227681265, + "learning_rate": 1.356254136574913e-05, + "loss": 1.6558, + "step": 1350 + }, + { + "epoch": 0.3905753107834634, + "grad_norm": 1.165298085541357, + "learning_rate": 1.3553966649138297e-05, + "loss": 1.5468, + "step": 1351 + }, + { + "epoch": 0.3908644116796762, + "grad_norm": 1.1344138977255307, + "learning_rate": 1.3545388940637958e-05, + "loss": 1.5932, + "step": 1352 + }, + { + "epoch": 0.391153512575889, + "grad_norm": 1.0690031880578144, + "learning_rate": 1.3536808247469213e-05, + "loss": 1.5697, + "step": 1353 + }, + { + "epoch": 0.39144261347210174, + "grad_norm": 1.0975749907120596, + "learning_rate": 1.352822457685567e-05, + "loss": 1.6245, + "step": 1354 + }, + { + "epoch": 0.39173171436831455, + "grad_norm": 1.0927297214562863, + "learning_rate": 1.3519637936023462e-05, + "loss": 1.6026, + "step": 1355 + }, + { + "epoch": 0.3920208152645273, + "grad_norm": 1.0537160535977657, + "learning_rate": 1.3511048332201204e-05, + "loss": 1.4856, + "step": 1356 + }, + { + "epoch": 0.3923099161607401, + "grad_norm": 1.0920768107193122, + "learning_rate": 1.3502455772620012e-05, + "loss": 1.4882, + "step": 1357 + }, + { + "epoch": 0.3925990170569529, + "grad_norm": 1.0757741953402, + "learning_rate": 1.3493860264513492e-05, + "loss": 1.6938, + "step": 1358 + }, + { + "epoch": 0.39288811795316564, + "grad_norm": 1.1184693762965427, + "learning_rate": 1.348526181511773e-05, + "loss": 1.6371, + "step": 1359 + }, + { + "epoch": 0.39317721884937845, + "grad_norm": 1.0887972408619184, + "learning_rate": 1.3476660431671288e-05, + "loss": 1.5472, + "step": 1360 + }, + { + "epoch": 0.3934663197455912, + "grad_norm": 1.1259522274979648, + "learning_rate": 1.34680561214152e-05, + "loss": 1.5498, + "step": 1361 + }, + { + "epoch": 0.39375542064180397, + "grad_norm": 1.1133474449428702, + "learning_rate": 1.3459448891592962e-05, + "loss": 1.5723, + "step": 1362 + }, + { + "epoch": 0.3940445215380168, + "grad_norm": 1.1499692550396208, + "learning_rate": 1.3450838749450529e-05, + "loss": 1.484, + "step": 1363 + }, + { + "epoch": 0.39433362243422954, + "grad_norm": 1.0715825700296768, + "learning_rate": 1.3442225702236307e-05, + "loss": 1.5612, + "step": 1364 + }, + { + "epoch": 0.39462272333044235, + "grad_norm": 1.077029585331996, + "learning_rate": 1.3433609757201147e-05, + "loss": 1.5889, + "step": 1365 + }, + { + "epoch": 0.3949118242266551, + "grad_norm": 1.1501089531692315, + "learning_rate": 1.342499092159834e-05, + "loss": 1.4946, + "step": 1366 + }, + { + "epoch": 0.39520092512286786, + "grad_norm": 1.096740007082763, + "learning_rate": 1.3416369202683615e-05, + "loss": 1.5497, + "step": 1367 + }, + { + "epoch": 0.3954900260190807, + "grad_norm": 1.0987454926481957, + "learning_rate": 1.3407744607715119e-05, + "loss": 1.5969, + "step": 1368 + }, + { + "epoch": 0.39577912691529343, + "grad_norm": 1.060961357189651, + "learning_rate": 1.3399117143953426e-05, + "loss": 1.5441, + "step": 1369 + }, + { + "epoch": 0.3960682278115062, + "grad_norm": 1.1032994325908025, + "learning_rate": 1.3390486818661526e-05, + "loss": 1.6306, + "step": 1370 + }, + { + "epoch": 0.396357328707719, + "grad_norm": 1.1047680776182227, + "learning_rate": 1.3381853639104817e-05, + "loss": 1.5242, + "step": 1371 + }, + { + "epoch": 0.39664642960393176, + "grad_norm": 1.1022023946061226, + "learning_rate": 1.33732176125511e-05, + "loss": 1.6319, + "step": 1372 + }, + { + "epoch": 0.3969355305001446, + "grad_norm": 1.0789182866424254, + "learning_rate": 1.3364578746270564e-05, + "loss": 1.6542, + "step": 1373 + }, + { + "epoch": 0.39722463139635733, + "grad_norm": 1.13836495763387, + "learning_rate": 1.3355937047535806e-05, + "loss": 1.5158, + "step": 1374 + }, + { + "epoch": 0.3975137322925701, + "grad_norm": 1.1151693144482542, + "learning_rate": 1.3347292523621795e-05, + "loss": 1.6513, + "step": 1375 + }, + { + "epoch": 0.3978028331887829, + "grad_norm": 1.0532473348160287, + "learning_rate": 1.3338645181805876e-05, + "loss": 1.5503, + "step": 1376 + }, + { + "epoch": 0.39809193408499566, + "grad_norm": 1.119386196223983, + "learning_rate": 1.3329995029367782e-05, + "loss": 1.5879, + "step": 1377 + }, + { + "epoch": 0.3983810349812084, + "grad_norm": 1.0973825255595782, + "learning_rate": 1.3321342073589595e-05, + "loss": 1.5115, + "step": 1378 + }, + { + "epoch": 0.39867013587742123, + "grad_norm": 1.0693127903553905, + "learning_rate": 1.331268632175576e-05, + "loss": 1.6588, + "step": 1379 + }, + { + "epoch": 0.398959236773634, + "grad_norm": 1.0823638073171304, + "learning_rate": 1.330402778115309e-05, + "loss": 1.589, + "step": 1380 + }, + { + "epoch": 0.3992483376698468, + "grad_norm": 1.0805775803416322, + "learning_rate": 1.3295366459070724e-05, + "loss": 1.5052, + "step": 1381 + }, + { + "epoch": 0.39953743856605956, + "grad_norm": 1.0443203230920821, + "learning_rate": 1.328670236280016e-05, + "loss": 1.4888, + "step": 1382 + }, + { + "epoch": 0.3998265394622723, + "grad_norm": 1.1511681114820926, + "learning_rate": 1.3278035499635226e-05, + "loss": 1.6123, + "step": 1383 + }, + { + "epoch": 0.4001156403584851, + "grad_norm": 1.0863214241630808, + "learning_rate": 1.3269365876872078e-05, + "loss": 1.5493, + "step": 1384 + }, + { + "epoch": 0.4004047412546979, + "grad_norm": 1.0838040505052533, + "learning_rate": 1.3260693501809193e-05, + "loss": 1.6153, + "step": 1385 + }, + { + "epoch": 0.4006938421509107, + "grad_norm": 1.1500580767514397, + "learning_rate": 1.3252018381747367e-05, + "loss": 1.5897, + "step": 1386 + }, + { + "epoch": 0.40098294304712345, + "grad_norm": 1.0845655171544064, + "learning_rate": 1.3243340523989714e-05, + "loss": 1.5783, + "step": 1387 + }, + { + "epoch": 0.4012720439433362, + "grad_norm": 1.1970928864429935, + "learning_rate": 1.3234659935841641e-05, + "loss": 1.6129, + "step": 1388 + }, + { + "epoch": 0.401561144839549, + "grad_norm": 1.101894564995156, + "learning_rate": 1.3225976624610856e-05, + "loss": 1.5264, + "step": 1389 + }, + { + "epoch": 0.4018502457357618, + "grad_norm": 1.1047369726724903, + "learning_rate": 1.3217290597607373e-05, + "loss": 1.6217, + "step": 1390 + }, + { + "epoch": 0.40213934663197454, + "grad_norm": 1.1482189531968385, + "learning_rate": 1.3208601862143468e-05, + "loss": 1.6808, + "step": 1391 + }, + { + "epoch": 0.40242844752818735, + "grad_norm": 1.1210248814188173, + "learning_rate": 1.3199910425533721e-05, + "loss": 1.5766, + "step": 1392 + }, + { + "epoch": 0.4027175484244001, + "grad_norm": 1.1685854469228043, + "learning_rate": 1.3191216295094971e-05, + "loss": 1.4749, + "step": 1393 + }, + { + "epoch": 0.4030066493206129, + "grad_norm": 1.0797246071856748, + "learning_rate": 1.318251947814633e-05, + "loss": 1.6377, + "step": 1394 + }, + { + "epoch": 0.4032957502168257, + "grad_norm": 1.0635773691483372, + "learning_rate": 1.3173819982009173e-05, + "loss": 1.4949, + "step": 1395 + }, + { + "epoch": 0.40358485111303843, + "grad_norm": 1.115897033244827, + "learning_rate": 1.3165117814007124e-05, + "loss": 1.6583, + "step": 1396 + }, + { + "epoch": 0.40387395200925125, + "grad_norm": 1.0989308444609356, + "learning_rate": 1.3156412981466064e-05, + "loss": 1.649, + "step": 1397 + }, + { + "epoch": 0.404163052905464, + "grad_norm": 1.0797697429688853, + "learning_rate": 1.3147705491714113e-05, + "loss": 1.5827, + "step": 1398 + }, + { + "epoch": 0.40445215380167676, + "grad_norm": 1.0834261852077045, + "learning_rate": 1.3138995352081634e-05, + "loss": 1.5503, + "step": 1399 + }, + { + "epoch": 0.4047412546978896, + "grad_norm": 1.0787814057307044, + "learning_rate": 1.3130282569901205e-05, + "loss": 1.5232, + "step": 1400 + } + ], + "logging_steps": 1, + "max_steps": 3459, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 700, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 146565758976000.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}