{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9996676636756398, "eval_steps": 500, "global_step": 1504, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0006646726487205051, "grad_norm": 60.970770959791075, "learning_rate": 2.173913043478261e-07, "loss": 0.4433, "step": 1 }, { "epoch": 0.0013293452974410102, "grad_norm": 63.31886870971556, "learning_rate": 4.347826086956522e-07, "loss": 0.4178, "step": 2 }, { "epoch": 0.0019940179461615153, "grad_norm": 61.288021460030315, "learning_rate": 6.521739130434783e-07, "loss": 0.4224, "step": 3 }, { "epoch": 0.0026586905948820204, "grad_norm": 62.25609151967674, "learning_rate": 8.695652173913044e-07, "loss": 0.4122, "step": 4 }, { "epoch": 0.003323363243602526, "grad_norm": 64.75114100147859, "learning_rate": 1.0869565217391306e-06, "loss": 0.355, "step": 5 }, { "epoch": 0.003988035892323031, "grad_norm": 41.5167763703531, "learning_rate": 1.3043478260869566e-06, "loss": 0.1343, "step": 6 }, { "epoch": 0.004652708541043536, "grad_norm": 28.001674249743854, "learning_rate": 1.521739130434783e-06, "loss": 0.1004, "step": 7 }, { "epoch": 0.005317381189764041, "grad_norm": 9.275314268012675, "learning_rate": 1.7391304347826088e-06, "loss": 0.0598, "step": 8 }, { "epoch": 0.005982053838484547, "grad_norm": 8.241880087134492, "learning_rate": 1.956521739130435e-06, "loss": 0.0493, "step": 9 }, { "epoch": 0.006646726487205052, "grad_norm": 4.901672630865819, "learning_rate": 2.173913043478261e-06, "loss": 0.0397, "step": 10 }, { "epoch": 0.007311399135925557, "grad_norm": 1.681492858780252, "learning_rate": 2.391304347826087e-06, "loss": 0.0284, "step": 11 }, { "epoch": 0.007976071784646061, "grad_norm": 0.5770734524620795, "learning_rate": 2.6086956521739132e-06, "loss": 0.0293, "step": 12 }, { "epoch": 0.008640744433366567, "grad_norm": 0.8524918991841441, "learning_rate": 2.8260869565217393e-06, "loss": 0.0229, "step": 13 }, { "epoch": 0.009305417082087071, "grad_norm": 4.193040234296588, "learning_rate": 3.043478260869566e-06, "loss": 0.0351, "step": 14 }, { "epoch": 0.009970089730807577, "grad_norm": 3.3392975596493364, "learning_rate": 3.2608695652173914e-06, "loss": 0.0302, "step": 15 }, { "epoch": 0.010634762379528082, "grad_norm": 2.8026088809193337, "learning_rate": 3.4782608695652175e-06, "loss": 0.0273, "step": 16 }, { "epoch": 0.011299435028248588, "grad_norm": 1.4649396502944714, "learning_rate": 3.6956521739130436e-06, "loss": 0.0158, "step": 17 }, { "epoch": 0.011964107676969093, "grad_norm": 1.795591002624188, "learning_rate": 3.91304347826087e-06, "loss": 0.0223, "step": 18 }, { "epoch": 0.012628780325689598, "grad_norm": 1.0179266424228162, "learning_rate": 4.130434782608696e-06, "loss": 0.0169, "step": 19 }, { "epoch": 0.013293452974410104, "grad_norm": 2.0223769514509033, "learning_rate": 4.347826086956522e-06, "loss": 0.0237, "step": 20 }, { "epoch": 0.013958125623130608, "grad_norm": 2.6414726093929186, "learning_rate": 4.565217391304348e-06, "loss": 0.0196, "step": 21 }, { "epoch": 0.014622798271851114, "grad_norm": 2.762431670662566, "learning_rate": 4.782608695652174e-06, "loss": 0.0176, "step": 22 }, { "epoch": 0.015287470920571618, "grad_norm": 1.6868280272100167, "learning_rate": 5e-06, "loss": 0.0213, "step": 23 }, { "epoch": 0.015952143569292122, "grad_norm": 1.3162037504364383, "learning_rate": 5.2173913043478265e-06, "loss": 0.017, "step": 24 }, { "epoch": 0.01661681621801263, "grad_norm": 0.9325497342450657, "learning_rate": 5.4347826086956525e-06, "loss": 0.0155, "step": 25 }, { "epoch": 0.017281488866733134, "grad_norm": 0.908746299202415, "learning_rate": 5.652173913043479e-06, "loss": 0.0163, "step": 26 }, { "epoch": 0.01794616151545364, "grad_norm": 0.40531336084522424, "learning_rate": 5.8695652173913055e-06, "loss": 0.0124, "step": 27 }, { "epoch": 0.018610834164174143, "grad_norm": 0.882809214932779, "learning_rate": 6.086956521739132e-06, "loss": 0.0146, "step": 28 }, { "epoch": 0.01927550681289465, "grad_norm": 0.33582807675108245, "learning_rate": 6.304347826086958e-06, "loss": 0.0131, "step": 29 }, { "epoch": 0.019940179461615155, "grad_norm": 0.6531979846582778, "learning_rate": 6.521739130434783e-06, "loss": 0.0148, "step": 30 }, { "epoch": 0.02060485211033566, "grad_norm": 1.2589719995994493, "learning_rate": 6.739130434782609e-06, "loss": 0.0192, "step": 31 }, { "epoch": 0.021269524759056163, "grad_norm": 0.6454998366907141, "learning_rate": 6.956521739130435e-06, "loss": 0.0161, "step": 32 }, { "epoch": 0.02193419740777667, "grad_norm": 0.5289479235804806, "learning_rate": 7.173913043478261e-06, "loss": 0.0116, "step": 33 }, { "epoch": 0.022598870056497175, "grad_norm": 0.2914456221068029, "learning_rate": 7.391304347826087e-06, "loss": 0.0119, "step": 34 }, { "epoch": 0.02326354270521768, "grad_norm": 0.6093431877720631, "learning_rate": 7.608695652173914e-06, "loss": 0.0116, "step": 35 }, { "epoch": 0.023928215353938187, "grad_norm": 0.5338791381117465, "learning_rate": 7.82608695652174e-06, "loss": 0.0134, "step": 36 }, { "epoch": 0.02459288800265869, "grad_norm": 0.6971604882392116, "learning_rate": 8.043478260869566e-06, "loss": 0.0158, "step": 37 }, { "epoch": 0.025257560651379195, "grad_norm": 0.7487742105878584, "learning_rate": 8.260869565217392e-06, "loss": 0.0086, "step": 38 }, { "epoch": 0.0259222333000997, "grad_norm": 0.8145315066665602, "learning_rate": 8.478260869565218e-06, "loss": 0.0188, "step": 39 }, { "epoch": 0.026586905948820207, "grad_norm": 0.7657328527509938, "learning_rate": 8.695652173913044e-06, "loss": 0.0112, "step": 40 }, { "epoch": 0.02725157859754071, "grad_norm": 0.7965754798989254, "learning_rate": 8.91304347826087e-06, "loss": 0.0118, "step": 41 }, { "epoch": 0.027916251246261216, "grad_norm": 1.0182253880741658, "learning_rate": 9.130434782608697e-06, "loss": 0.0175, "step": 42 }, { "epoch": 0.028580923894981722, "grad_norm": 0.8159799700749677, "learning_rate": 9.347826086956523e-06, "loss": 0.0134, "step": 43 }, { "epoch": 0.029245596543702228, "grad_norm": 1.6698654957456858, "learning_rate": 9.565217391304349e-06, "loss": 0.0109, "step": 44 }, { "epoch": 0.02991026919242273, "grad_norm": 0.5935190769400233, "learning_rate": 9.782608695652175e-06, "loss": 0.014, "step": 45 }, { "epoch": 0.030574941841143236, "grad_norm": 0.20846862149424067, "learning_rate": 1e-05, "loss": 0.0096, "step": 46 }, { "epoch": 0.031239614489863742, "grad_norm": 1.8230375659276052, "learning_rate": 9.999988392878299e-06, "loss": 0.0182, "step": 47 }, { "epoch": 0.031904287138584245, "grad_norm": 0.8507271501848354, "learning_rate": 9.999953571567086e-06, "loss": 0.0157, "step": 48 }, { "epoch": 0.03256895978730475, "grad_norm": 0.8855889032923526, "learning_rate": 9.999895536228031e-06, "loss": 0.0135, "step": 49 }, { "epoch": 0.03323363243602526, "grad_norm": 1.1799803352308096, "learning_rate": 9.99981428713058e-06, "loss": 0.0149, "step": 50 }, { "epoch": 0.03389830508474576, "grad_norm": 0.7507208371373172, "learning_rate": 9.999709824651965e-06, "loss": 0.011, "step": 51 }, { "epoch": 0.03456297773346627, "grad_norm": 0.3417646829103957, "learning_rate": 9.999582149277188e-06, "loss": 0.0178, "step": 52 }, { "epoch": 0.035227650382186775, "grad_norm": 0.967206468925284, "learning_rate": 9.999431261599025e-06, "loss": 0.0133, "step": 53 }, { "epoch": 0.03589232303090728, "grad_norm": 0.9527765614550642, "learning_rate": 9.999257162318027e-06, "loss": 0.0149, "step": 54 }, { "epoch": 0.036556995679627786, "grad_norm": 0.28932661882565486, "learning_rate": 9.999059852242508e-06, "loss": 0.013, "step": 55 }, { "epoch": 0.037221668328348285, "grad_norm": 1.9364775309785114, "learning_rate": 9.99883933228855e-06, "loss": 0.02, "step": 56 }, { "epoch": 0.03788634097706879, "grad_norm": 1.0609733740408223, "learning_rate": 9.998595603479995e-06, "loss": 0.0179, "step": 57 }, { "epoch": 0.0385510136257893, "grad_norm": 0.49527347820893447, "learning_rate": 9.998328666948437e-06, "loss": 0.0094, "step": 58 }, { "epoch": 0.0392156862745098, "grad_norm": 0.4372324960151166, "learning_rate": 9.998038523933224e-06, "loss": 0.0096, "step": 59 }, { "epoch": 0.03988035892323031, "grad_norm": 0.6464316057194724, "learning_rate": 9.997725175781445e-06, "loss": 0.0117, "step": 60 }, { "epoch": 0.040545031571950815, "grad_norm": 0.6987667278536408, "learning_rate": 9.997388623947927e-06, "loss": 0.0119, "step": 61 }, { "epoch": 0.04120970422067132, "grad_norm": 1.7130914482404813, "learning_rate": 9.99702886999523e-06, "loss": 0.0228, "step": 62 }, { "epoch": 0.04187437686939183, "grad_norm": 0.4117960597890336, "learning_rate": 9.996645915593639e-06, "loss": 0.0133, "step": 63 }, { "epoch": 0.042539049518112326, "grad_norm": 1.2932996784542414, "learning_rate": 9.996239762521152e-06, "loss": 0.0115, "step": 64 }, { "epoch": 0.04320372216683283, "grad_norm": 1.3674099246411462, "learning_rate": 9.995810412663473e-06, "loss": 0.0117, "step": 65 }, { "epoch": 0.04386839481555334, "grad_norm": 1.6053889474918175, "learning_rate": 9.995357868014014e-06, "loss": 0.015, "step": 66 }, { "epoch": 0.044533067464273844, "grad_norm": 0.27477674665342927, "learning_rate": 9.994882130673869e-06, "loss": 0.0164, "step": 67 }, { "epoch": 0.04519774011299435, "grad_norm": 2.210389961890394, "learning_rate": 9.994383202851813e-06, "loss": 0.0204, "step": 68 }, { "epoch": 0.045862412761714856, "grad_norm": 3.023467310150046, "learning_rate": 9.993861086864294e-06, "loss": 0.024, "step": 69 }, { "epoch": 0.04652708541043536, "grad_norm": 2.208505829292387, "learning_rate": 9.993315785135417e-06, "loss": 0.018, "step": 70 }, { "epoch": 0.04719175805915587, "grad_norm": 0.8693257140907787, "learning_rate": 9.992747300196935e-06, "loss": 0.0153, "step": 71 }, { "epoch": 0.047856430707876374, "grad_norm": 0.5860107187265328, "learning_rate": 9.99215563468824e-06, "loss": 0.0147, "step": 72 }, { "epoch": 0.04852110335659687, "grad_norm": 0.49667771485582274, "learning_rate": 9.991540791356342e-06, "loss": 0.0126, "step": 73 }, { "epoch": 0.04918577600531738, "grad_norm": 2.3621689044782097, "learning_rate": 9.990902773055868e-06, "loss": 0.0219, "step": 74 }, { "epoch": 0.049850448654037885, "grad_norm": 1.7629413870031103, "learning_rate": 9.990241582749039e-06, "loss": 0.0174, "step": 75 }, { "epoch": 0.05051512130275839, "grad_norm": 0.37616185123766155, "learning_rate": 9.989557223505661e-06, "loss": 0.0106, "step": 76 }, { "epoch": 0.0511797939514789, "grad_norm": 0.27687180041583864, "learning_rate": 9.988849698503112e-06, "loss": 0.0129, "step": 77 }, { "epoch": 0.0518444666001994, "grad_norm": 0.5877900585958946, "learning_rate": 9.988119011026325e-06, "loss": 0.012, "step": 78 }, { "epoch": 0.05250913924891991, "grad_norm": 1.8535031653078293, "learning_rate": 9.987365164467767e-06, "loss": 0.0179, "step": 79 }, { "epoch": 0.053173811897640415, "grad_norm": 1.0003423175984858, "learning_rate": 9.986588162327436e-06, "loss": 0.0136, "step": 80 }, { "epoch": 0.053838484546360914, "grad_norm": 1.3365768362375101, "learning_rate": 9.985788008212837e-06, "loss": 0.0156, "step": 81 }, { "epoch": 0.05450315719508142, "grad_norm": 0.5805865152268904, "learning_rate": 9.98496470583896e-06, "loss": 0.0142, "step": 82 }, { "epoch": 0.055167829843801926, "grad_norm": 1.1904648845306414, "learning_rate": 9.98411825902828e-06, "loss": 0.0131, "step": 83 }, { "epoch": 0.05583250249252243, "grad_norm": 1.3016928681226674, "learning_rate": 9.983248671710714e-06, "loss": 0.0111, "step": 84 }, { "epoch": 0.05649717514124294, "grad_norm": 0.4011694177073853, "learning_rate": 9.98235594792363e-06, "loss": 0.0111, "step": 85 }, { "epoch": 0.057161847789963444, "grad_norm": 0.5837548332714089, "learning_rate": 9.981440091811805e-06, "loss": 0.0142, "step": 86 }, { "epoch": 0.05782652043868395, "grad_norm": 0.6865193702769102, "learning_rate": 9.980501107627425e-06, "loss": 0.0149, "step": 87 }, { "epoch": 0.058491193087404456, "grad_norm": 0.3514697264200902, "learning_rate": 9.979538999730047e-06, "loss": 0.0132, "step": 88 }, { "epoch": 0.05915586573612496, "grad_norm": 0.30265569565853434, "learning_rate": 9.978553772586596e-06, "loss": 0.011, "step": 89 }, { "epoch": 0.05982053838484546, "grad_norm": 0.5004793700925844, "learning_rate": 9.977545430771332e-06, "loss": 0.0146, "step": 90 }, { "epoch": 0.060485211033565967, "grad_norm": 0.26294836808361827, "learning_rate": 9.976513978965829e-06, "loss": 0.0099, "step": 91 }, { "epoch": 0.06114988368228647, "grad_norm": 0.9190030643877792, "learning_rate": 9.975459421958969e-06, "loss": 0.0132, "step": 92 }, { "epoch": 0.06181455633100698, "grad_norm": 0.2959912908827373, "learning_rate": 9.974381764646894e-06, "loss": 0.0153, "step": 93 }, { "epoch": 0.062479228979727484, "grad_norm": 0.5975731268605555, "learning_rate": 9.973281012033009e-06, "loss": 0.014, "step": 94 }, { "epoch": 0.06314390162844799, "grad_norm": 0.8113597767320757, "learning_rate": 9.972157169227937e-06, "loss": 0.0119, "step": 95 }, { "epoch": 0.06380857427716849, "grad_norm": 0.28144834234849603, "learning_rate": 9.971010241449514e-06, "loss": 0.0108, "step": 96 }, { "epoch": 0.064473246925889, "grad_norm": 0.4166786310414046, "learning_rate": 9.96984023402275e-06, "loss": 0.0115, "step": 97 }, { "epoch": 0.0651379195746095, "grad_norm": 0.6257073392967791, "learning_rate": 9.968647152379813e-06, "loss": 0.0148, "step": 98 }, { "epoch": 0.06580259222333001, "grad_norm": 0.40388258286019546, "learning_rate": 9.967431002060001e-06, "loss": 0.0155, "step": 99 }, { "epoch": 0.06646726487205051, "grad_norm": 0.36186955381710356, "learning_rate": 9.966191788709716e-06, "loss": 0.0115, "step": 100 }, { "epoch": 0.06713193752077103, "grad_norm": 0.31379970023288684, "learning_rate": 9.964929518082436e-06, "loss": 0.0134, "step": 101 }, { "epoch": 0.06779661016949153, "grad_norm": 0.35955264841421186, "learning_rate": 9.963644196038695e-06, "loss": 0.0107, "step": 102 }, { "epoch": 0.06846128281821202, "grad_norm": 0.7835116116021953, "learning_rate": 9.962335828546049e-06, "loss": 0.0137, "step": 103 }, { "epoch": 0.06912595546693254, "grad_norm": 0.7808060669603882, "learning_rate": 9.961004421679048e-06, "loss": 0.015, "step": 104 }, { "epoch": 0.06979062811565304, "grad_norm": 0.7171964438595184, "learning_rate": 9.959649981619214e-06, "loss": 0.0133, "step": 105 }, { "epoch": 0.07045530076437355, "grad_norm": 0.28242956765832666, "learning_rate": 9.958272514655006e-06, "loss": 0.01, "step": 106 }, { "epoch": 0.07111997341309405, "grad_norm": 0.45078880338883015, "learning_rate": 9.956872027181797e-06, "loss": 0.0126, "step": 107 }, { "epoch": 0.07178464606181456, "grad_norm": 1.2370065847479403, "learning_rate": 9.955448525701836e-06, "loss": 0.0135, "step": 108 }, { "epoch": 0.07244931871053506, "grad_norm": 0.3753757002709158, "learning_rate": 9.954002016824226e-06, "loss": 0.0112, "step": 109 }, { "epoch": 0.07311399135925557, "grad_norm": 0.32837638393260105, "learning_rate": 9.952532507264892e-06, "loss": 0.0134, "step": 110 }, { "epoch": 0.07377866400797607, "grad_norm": 0.3373764833711116, "learning_rate": 9.951040003846537e-06, "loss": 0.0135, "step": 111 }, { "epoch": 0.07444333665669657, "grad_norm": 0.6622672104842571, "learning_rate": 9.949524513498636e-06, "loss": 0.0112, "step": 112 }, { "epoch": 0.07510800930541708, "grad_norm": 1.2107157371433923, "learning_rate": 9.947986043257377e-06, "loss": 0.016, "step": 113 }, { "epoch": 0.07577268195413758, "grad_norm": 0.5978241379607916, "learning_rate": 9.946424600265647e-06, "loss": 0.0085, "step": 114 }, { "epoch": 0.0764373546028581, "grad_norm": 0.36182468361829123, "learning_rate": 9.944840191772987e-06, "loss": 0.0148, "step": 115 }, { "epoch": 0.0771020272515786, "grad_norm": 1.2675227954269332, "learning_rate": 9.943232825135568e-06, "loss": 0.0133, "step": 116 }, { "epoch": 0.07776669990029911, "grad_norm": 0.7700080328396309, "learning_rate": 9.941602507816149e-06, "loss": 0.0148, "step": 117 }, { "epoch": 0.0784313725490196, "grad_norm": 0.6726926682160788, "learning_rate": 9.939949247384046e-06, "loss": 0.0154, "step": 118 }, { "epoch": 0.07909604519774012, "grad_norm": 0.6089081317809364, "learning_rate": 9.938273051515098e-06, "loss": 0.0178, "step": 119 }, { "epoch": 0.07976071784646062, "grad_norm": 0.644474024420446, "learning_rate": 9.936573927991631e-06, "loss": 0.0148, "step": 120 }, { "epoch": 0.08042539049518112, "grad_norm": 0.9364477044735832, "learning_rate": 9.934851884702415e-06, "loss": 0.017, "step": 121 }, { "epoch": 0.08109006314390163, "grad_norm": 0.19304986064034851, "learning_rate": 9.933106929642635e-06, "loss": 0.0126, "step": 122 }, { "epoch": 0.08175473579262213, "grad_norm": 0.47354696444228556, "learning_rate": 9.931339070913859e-06, "loss": 0.0109, "step": 123 }, { "epoch": 0.08241940844134264, "grad_norm": 0.41017504287412127, "learning_rate": 9.929548316723983e-06, "loss": 0.0152, "step": 124 }, { "epoch": 0.08308408109006314, "grad_norm": 0.2492432717974285, "learning_rate": 9.92773467538721e-06, "loss": 0.0084, "step": 125 }, { "epoch": 0.08374875373878365, "grad_norm": 0.5761337571958625, "learning_rate": 9.925898155324001e-06, "loss": 0.0102, "step": 126 }, { "epoch": 0.08441342638750415, "grad_norm": 0.5354804682859692, "learning_rate": 9.924038765061042e-06, "loss": 0.0149, "step": 127 }, { "epoch": 0.08507809903622465, "grad_norm": 0.37188430864299515, "learning_rate": 9.922156513231197e-06, "loss": 0.0081, "step": 128 }, { "epoch": 0.08574277168494517, "grad_norm": 0.38511929720460997, "learning_rate": 9.920251408573482e-06, "loss": 0.0131, "step": 129 }, { "epoch": 0.08640744433366566, "grad_norm": 0.6930146020356339, "learning_rate": 9.918323459933006e-06, "loss": 0.0117, "step": 130 }, { "epoch": 0.08707211698238618, "grad_norm": 0.6303221804187904, "learning_rate": 9.916372676260943e-06, "loss": 0.0135, "step": 131 }, { "epoch": 0.08773678963110668, "grad_norm": 1.3789594959511349, "learning_rate": 9.914399066614488e-06, "loss": 0.0151, "step": 132 }, { "epoch": 0.08840146227982719, "grad_norm": 1.3078911807780915, "learning_rate": 9.912402640156812e-06, "loss": 0.0135, "step": 133 }, { "epoch": 0.08906613492854769, "grad_norm": 0.559643422966823, "learning_rate": 9.910383406157019e-06, "loss": 0.0105, "step": 134 }, { "epoch": 0.0897308075772682, "grad_norm": 0.5841937953300166, "learning_rate": 9.908341373990107e-06, "loss": 0.0103, "step": 135 }, { "epoch": 0.0903954802259887, "grad_norm": 0.8642063635516389, "learning_rate": 9.906276553136924e-06, "loss": 0.0111, "step": 136 }, { "epoch": 0.0910601528747092, "grad_norm": 0.7958234434494227, "learning_rate": 9.904188953184119e-06, "loss": 0.0142, "step": 137 }, { "epoch": 0.09172482552342971, "grad_norm": 0.2730360636246964, "learning_rate": 9.902078583824106e-06, "loss": 0.0126, "step": 138 }, { "epoch": 0.09238949817215021, "grad_norm": 0.35536366767523137, "learning_rate": 9.899945454855007e-06, "loss": 0.0137, "step": 139 }, { "epoch": 0.09305417082087072, "grad_norm": 0.6440704498816134, "learning_rate": 9.897789576180617e-06, "loss": 0.0158, "step": 140 }, { "epoch": 0.09371884346959122, "grad_norm": 0.4141234814182246, "learning_rate": 9.895610957810357e-06, "loss": 0.011, "step": 141 }, { "epoch": 0.09438351611831174, "grad_norm": 0.3911627182800861, "learning_rate": 9.893409609859221e-06, "loss": 0.0115, "step": 142 }, { "epoch": 0.09504818876703224, "grad_norm": 0.33514797480492575, "learning_rate": 9.891185542547735e-06, "loss": 0.013, "step": 143 }, { "epoch": 0.09571286141575275, "grad_norm": 0.22372917720722707, "learning_rate": 9.888938766201908e-06, "loss": 0.0115, "step": 144 }, { "epoch": 0.09637753406447325, "grad_norm": 0.5577788601787329, "learning_rate": 9.886669291253178e-06, "loss": 0.0154, "step": 145 }, { "epoch": 0.09704220671319375, "grad_norm": 0.7128789787341765, "learning_rate": 9.884377128238378e-06, "loss": 0.0085, "step": 146 }, { "epoch": 0.09770687936191426, "grad_norm": 2.7252320590941803, "learning_rate": 9.882062287799674e-06, "loss": 0.0125, "step": 147 }, { "epoch": 0.09837155201063476, "grad_norm": 0.3687053560173959, "learning_rate": 9.879724780684518e-06, "loss": 0.0105, "step": 148 }, { "epoch": 0.09903622465935527, "grad_norm": 0.22462741221254443, "learning_rate": 9.877364617745604e-06, "loss": 0.0106, "step": 149 }, { "epoch": 0.09970089730807577, "grad_norm": 1.0030046147147829, "learning_rate": 9.87498180994081e-06, "loss": 0.011, "step": 150 }, { "epoch": 0.10036556995679628, "grad_norm": 0.3131307429660239, "learning_rate": 9.872576368333152e-06, "loss": 0.0103, "step": 151 }, { "epoch": 0.10103024260551678, "grad_norm": 0.9056511820785378, "learning_rate": 9.87014830409073e-06, "loss": 0.0129, "step": 152 }, { "epoch": 0.1016949152542373, "grad_norm": 0.6625456889764615, "learning_rate": 9.86769762848668e-06, "loss": 0.012, "step": 153 }, { "epoch": 0.1023595879029578, "grad_norm": 0.5475498774739779, "learning_rate": 9.86522435289912e-06, "loss": 0.0119, "step": 154 }, { "epoch": 0.10302426055167829, "grad_norm": 0.5827615238833249, "learning_rate": 9.862728488811094e-06, "loss": 0.0191, "step": 155 }, { "epoch": 0.1036889332003988, "grad_norm": 0.37753366607575367, "learning_rate": 9.860210047810517e-06, "loss": 0.0126, "step": 156 }, { "epoch": 0.1043536058491193, "grad_norm": 0.32857074630518646, "learning_rate": 9.857669041590135e-06, "loss": 0.0113, "step": 157 }, { "epoch": 0.10501827849783982, "grad_norm": 0.7419212530147024, "learning_rate": 9.85510548194745e-06, "loss": 0.0111, "step": 158 }, { "epoch": 0.10568295114656032, "grad_norm": 0.4721111653667675, "learning_rate": 9.852519380784686e-06, "loss": 0.0141, "step": 159 }, { "epoch": 0.10634762379528083, "grad_norm": 0.4207051412951935, "learning_rate": 9.849910750108718e-06, "loss": 0.0125, "step": 160 }, { "epoch": 0.10701229644400133, "grad_norm": 0.21134456329404774, "learning_rate": 9.847279602031024e-06, "loss": 0.009, "step": 161 }, { "epoch": 0.10767696909272183, "grad_norm": 0.38448054271650456, "learning_rate": 9.844625948767624e-06, "loss": 0.0095, "step": 162 }, { "epoch": 0.10834164174144234, "grad_norm": 0.35541835353015216, "learning_rate": 9.841949802639031e-06, "loss": 0.0158, "step": 163 }, { "epoch": 0.10900631439016284, "grad_norm": 0.44289228243645407, "learning_rate": 9.839251176070184e-06, "loss": 0.0116, "step": 164 }, { "epoch": 0.10967098703888335, "grad_norm": 0.6467754208372888, "learning_rate": 9.836530081590401e-06, "loss": 0.0138, "step": 165 }, { "epoch": 0.11033565968760385, "grad_norm": 0.8286741013878848, "learning_rate": 9.833786531833311e-06, "loss": 0.0121, "step": 166 }, { "epoch": 0.11100033233632436, "grad_norm": 0.705185946340053, "learning_rate": 9.831020539536798e-06, "loss": 0.0123, "step": 167 }, { "epoch": 0.11166500498504486, "grad_norm": 1.1102218558063535, "learning_rate": 9.828232117542948e-06, "loss": 0.0135, "step": 168 }, { "epoch": 0.11232967763376538, "grad_norm": 0.8797965216959907, "learning_rate": 9.825421278797984e-06, "loss": 0.0118, "step": 169 }, { "epoch": 0.11299435028248588, "grad_norm": 0.49436345297389883, "learning_rate": 9.822588036352201e-06, "loss": 0.0112, "step": 170 }, { "epoch": 0.11365902293120637, "grad_norm": 0.535561655617538, "learning_rate": 9.819732403359916e-06, "loss": 0.0121, "step": 171 }, { "epoch": 0.11432369557992689, "grad_norm": 0.5240278610581294, "learning_rate": 9.816854393079402e-06, "loss": 0.0137, "step": 172 }, { "epoch": 0.11498836822864739, "grad_norm": 0.7848367301373668, "learning_rate": 9.813954018872826e-06, "loss": 0.0093, "step": 173 }, { "epoch": 0.1156530408773679, "grad_norm": 0.31424300530552746, "learning_rate": 9.811031294206184e-06, "loss": 0.0141, "step": 174 }, { "epoch": 0.1163177135260884, "grad_norm": 0.696939311710776, "learning_rate": 9.808086232649246e-06, "loss": 0.0136, "step": 175 }, { "epoch": 0.11698238617480891, "grad_norm": 0.6927849895100924, "learning_rate": 9.805118847875487e-06, "loss": 0.0126, "step": 176 }, { "epoch": 0.11764705882352941, "grad_norm": 0.7954905153164459, "learning_rate": 9.802129153662027e-06, "loss": 0.0153, "step": 177 }, { "epoch": 0.11831173147224992, "grad_norm": 0.46717250707308305, "learning_rate": 9.79911716388956e-06, "loss": 0.0084, "step": 178 }, { "epoch": 0.11897640412097042, "grad_norm": 0.19623670466343462, "learning_rate": 9.796082892542303e-06, "loss": 0.0109, "step": 179 }, { "epoch": 0.11964107676969092, "grad_norm": 0.2701821190756133, "learning_rate": 9.793026353707915e-06, "loss": 0.0095, "step": 180 }, { "epoch": 0.12030574941841143, "grad_norm": 0.7071097647338463, "learning_rate": 9.789947561577445e-06, "loss": 0.0117, "step": 181 }, { "epoch": 0.12097042206713193, "grad_norm": 0.6767990984513651, "learning_rate": 9.786846530445259e-06, "loss": 0.0133, "step": 182 }, { "epoch": 0.12163509471585245, "grad_norm": 0.6859072595345503, "learning_rate": 9.783723274708975e-06, "loss": 0.0093, "step": 183 }, { "epoch": 0.12229976736457294, "grad_norm": 0.7085061962563467, "learning_rate": 9.7805778088694e-06, "loss": 0.0139, "step": 184 }, { "epoch": 0.12296444001329346, "grad_norm": 0.2739957821512854, "learning_rate": 9.77741014753045e-06, "loss": 0.0085, "step": 185 }, { "epoch": 0.12362911266201396, "grad_norm": 0.6380426959775581, "learning_rate": 9.7742203053991e-06, "loss": 0.0116, "step": 186 }, { "epoch": 0.12429378531073447, "grad_norm": 0.8963922227499305, "learning_rate": 9.771008297285307e-06, "loss": 0.0102, "step": 187 }, { "epoch": 0.12495845795945497, "grad_norm": 0.45051647727613386, "learning_rate": 9.767774138101934e-06, "loss": 0.0151, "step": 188 }, { "epoch": 0.12562313060817548, "grad_norm": 0.798138226166225, "learning_rate": 9.764517842864696e-06, "loss": 0.0207, "step": 189 }, { "epoch": 0.12628780325689598, "grad_norm": 0.40271992611271096, "learning_rate": 9.761239426692077e-06, "loss": 0.0115, "step": 190 }, { "epoch": 0.12695247590561648, "grad_norm": 0.21819961547534378, "learning_rate": 9.757938904805269e-06, "loss": 0.012, "step": 191 }, { "epoch": 0.12761714855433698, "grad_norm": 0.29737806582054377, "learning_rate": 9.754616292528094e-06, "loss": 0.0133, "step": 192 }, { "epoch": 0.1282818212030575, "grad_norm": 0.40605514829010925, "learning_rate": 9.75127160528694e-06, "loss": 0.0136, "step": 193 }, { "epoch": 0.128946493851778, "grad_norm": 0.6613212213121648, "learning_rate": 9.747904858610681e-06, "loss": 0.0121, "step": 194 }, { "epoch": 0.1296111665004985, "grad_norm": 0.1643040372937212, "learning_rate": 9.744516068130616e-06, "loss": 0.0082, "step": 195 }, { "epoch": 0.130275839149219, "grad_norm": 0.30135094865623563, "learning_rate": 9.741105249580383e-06, "loss": 0.0081, "step": 196 }, { "epoch": 0.1309405117979395, "grad_norm": 0.5993217338396998, "learning_rate": 9.737672418795898e-06, "loss": 0.0096, "step": 197 }, { "epoch": 0.13160518444666003, "grad_norm": 0.8772668384555161, "learning_rate": 9.734217591715276e-06, "loss": 0.014, "step": 198 }, { "epoch": 0.13226985709538053, "grad_norm": 0.36455491449666355, "learning_rate": 9.730740784378755e-06, "loss": 0.017, "step": 199 }, { "epoch": 0.13293452974410103, "grad_norm": 1.1094255646566713, "learning_rate": 9.727242012928622e-06, "loss": 0.0114, "step": 200 }, { "epoch": 0.13359920239282153, "grad_norm": 0.933384762208318, "learning_rate": 9.72372129360915e-06, "loss": 0.012, "step": 201 }, { "epoch": 0.13426387504154205, "grad_norm": 0.48550244968200185, "learning_rate": 9.7201786427665e-06, "loss": 0.0144, "step": 202 }, { "epoch": 0.13492854769026255, "grad_norm": 0.9940596174478525, "learning_rate": 9.71661407684867e-06, "loss": 0.0164, "step": 203 }, { "epoch": 0.13559322033898305, "grad_norm": 0.2793266744201439, "learning_rate": 9.713027612405394e-06, "loss": 0.0116, "step": 204 }, { "epoch": 0.13625789298770355, "grad_norm": 0.35695536601946154, "learning_rate": 9.709419266088086e-06, "loss": 0.0102, "step": 205 }, { "epoch": 0.13692256563642405, "grad_norm": 0.5827355320041212, "learning_rate": 9.705789054649755e-06, "loss": 0.0141, "step": 206 }, { "epoch": 0.13758723828514458, "grad_norm": 0.3772682127306454, "learning_rate": 9.70213699494492e-06, "loss": 0.0088, "step": 207 }, { "epoch": 0.13825191093386507, "grad_norm": 0.7408045883983772, "learning_rate": 9.698463103929542e-06, "loss": 0.0161, "step": 208 }, { "epoch": 0.13891658358258557, "grad_norm": 0.39726542306882334, "learning_rate": 9.694767398660943e-06, "loss": 0.0117, "step": 209 }, { "epoch": 0.13958125623130607, "grad_norm": 0.49277023342427106, "learning_rate": 9.69104989629772e-06, "loss": 0.0098, "step": 210 }, { "epoch": 0.1402459288800266, "grad_norm": 0.9883447730549795, "learning_rate": 9.687310614099676e-06, "loss": 0.0136, "step": 211 }, { "epoch": 0.1409106015287471, "grad_norm": 0.15358461843546223, "learning_rate": 9.683549569427733e-06, "loss": 0.0076, "step": 212 }, { "epoch": 0.1415752741774676, "grad_norm": 0.6752402868770357, "learning_rate": 9.67976677974385e-06, "loss": 0.0112, "step": 213 }, { "epoch": 0.1422399468261881, "grad_norm": 0.287583569997248, "learning_rate": 9.67596226261095e-06, "loss": 0.0151, "step": 214 }, { "epoch": 0.1429046194749086, "grad_norm": 0.48200739129720765, "learning_rate": 9.672136035692827e-06, "loss": 0.0134, "step": 215 }, { "epoch": 0.14356929212362912, "grad_norm": 0.7517128399716646, "learning_rate": 9.668288116754077e-06, "loss": 0.012, "step": 216 }, { "epoch": 0.14423396477234962, "grad_norm": 0.39579158577415857, "learning_rate": 9.664418523660004e-06, "loss": 0.0099, "step": 217 }, { "epoch": 0.14489863742107012, "grad_norm": 0.46234364211673595, "learning_rate": 9.66052727437654e-06, "loss": 0.0103, "step": 218 }, { "epoch": 0.14556331006979062, "grad_norm": 0.22481121633452386, "learning_rate": 9.656614386970173e-06, "loss": 0.0105, "step": 219 }, { "epoch": 0.14622798271851115, "grad_norm": 0.8600068213528458, "learning_rate": 9.652679879607843e-06, "loss": 0.0135, "step": 220 }, { "epoch": 0.14689265536723164, "grad_norm": 0.33098071490836845, "learning_rate": 9.648723770556874e-06, "loss": 0.0135, "step": 221 }, { "epoch": 0.14755732801595214, "grad_norm": 0.17161327039725147, "learning_rate": 9.644746078184878e-06, "loss": 0.0097, "step": 222 }, { "epoch": 0.14822200066467264, "grad_norm": 0.5554530554345292, "learning_rate": 9.640746820959684e-06, "loss": 0.0107, "step": 223 }, { "epoch": 0.14888667331339314, "grad_norm": 0.7566953884363588, "learning_rate": 9.636726017449236e-06, "loss": 0.0138, "step": 224 }, { "epoch": 0.14955134596211367, "grad_norm": 0.2099504521607455, "learning_rate": 9.632683686321514e-06, "loss": 0.0103, "step": 225 }, { "epoch": 0.15021601861083417, "grad_norm": 0.28757578639404235, "learning_rate": 9.628619846344453e-06, "loss": 0.0133, "step": 226 }, { "epoch": 0.15088069125955467, "grad_norm": 0.5127074639282999, "learning_rate": 9.624534516385846e-06, "loss": 0.0074, "step": 227 }, { "epoch": 0.15154536390827517, "grad_norm": 0.3068181713741408, "learning_rate": 9.620427715413259e-06, "loss": 0.0109, "step": 228 }, { "epoch": 0.1522100365569957, "grad_norm": 0.431348351854592, "learning_rate": 9.616299462493952e-06, "loss": 0.0119, "step": 229 }, { "epoch": 0.1528747092057162, "grad_norm": 0.4975807194781063, "learning_rate": 9.612149776794776e-06, "loss": 0.015, "step": 230 }, { "epoch": 0.1535393818544367, "grad_norm": 0.44055872366506227, "learning_rate": 9.607978677582097e-06, "loss": 0.0098, "step": 231 }, { "epoch": 0.1542040545031572, "grad_norm": 0.3555873158934212, "learning_rate": 9.603786184221693e-06, "loss": 0.0136, "step": 232 }, { "epoch": 0.1548687271518777, "grad_norm": 0.17256693956815738, "learning_rate": 9.599572316178679e-06, "loss": 0.008, "step": 233 }, { "epoch": 0.15553339980059822, "grad_norm": 0.6278585943513464, "learning_rate": 9.595337093017403e-06, "loss": 0.0132, "step": 234 }, { "epoch": 0.15619807244931871, "grad_norm": 0.36132812067780773, "learning_rate": 9.591080534401371e-06, "loss": 0.0153, "step": 235 }, { "epoch": 0.1568627450980392, "grad_norm": 0.6194929851919586, "learning_rate": 9.586802660093138e-06, "loss": 0.0115, "step": 236 }, { "epoch": 0.1575274177467597, "grad_norm": 0.292173022973647, "learning_rate": 9.582503489954225e-06, "loss": 0.0122, "step": 237 }, { "epoch": 0.15819209039548024, "grad_norm": 0.22057710604709763, "learning_rate": 9.578183043945031e-06, "loss": 0.0097, "step": 238 }, { "epoch": 0.15885676304420074, "grad_norm": 0.3454260093325801, "learning_rate": 9.57384134212473e-06, "loss": 0.0105, "step": 239 }, { "epoch": 0.15952143569292124, "grad_norm": 0.2162918132784966, "learning_rate": 9.569478404651192e-06, "loss": 0.0103, "step": 240 }, { "epoch": 0.16018610834164174, "grad_norm": 0.3261237935534593, "learning_rate": 9.565094251780872e-06, "loss": 0.0114, "step": 241 }, { "epoch": 0.16085078099036224, "grad_norm": 0.2895025685786001, "learning_rate": 9.560688903868726e-06, "loss": 0.0101, "step": 242 }, { "epoch": 0.16151545363908276, "grad_norm": 0.15500567981634156, "learning_rate": 9.556262381368124e-06, "loss": 0.0086, "step": 243 }, { "epoch": 0.16218012628780326, "grad_norm": 0.2064406757020295, "learning_rate": 9.551814704830734e-06, "loss": 0.008, "step": 244 }, { "epoch": 0.16284479893652376, "grad_norm": 0.31144870417185716, "learning_rate": 9.547345894906447e-06, "loss": 0.0065, "step": 245 }, { "epoch": 0.16350947158524426, "grad_norm": 0.3686548835605602, "learning_rate": 9.542855972343274e-06, "loss": 0.0095, "step": 246 }, { "epoch": 0.16417414423396479, "grad_norm": 0.4220409188180881, "learning_rate": 9.538344957987245e-06, "loss": 0.0117, "step": 247 }, { "epoch": 0.16483881688268529, "grad_norm": 0.3854720063810567, "learning_rate": 9.533812872782315e-06, "loss": 0.0112, "step": 248 }, { "epoch": 0.16550348953140578, "grad_norm": 0.5273632407670993, "learning_rate": 9.52925973777027e-06, "loss": 0.016, "step": 249 }, { "epoch": 0.16616816218012628, "grad_norm": 0.8380294353506951, "learning_rate": 9.524685574090627e-06, "loss": 0.0093, "step": 250 }, { "epoch": 0.16683283482884678, "grad_norm": 0.37198832192228387, "learning_rate": 9.520090402980537e-06, "loss": 0.0108, "step": 251 }, { "epoch": 0.1674975074775673, "grad_norm": 0.9283673290694682, "learning_rate": 9.515474245774685e-06, "loss": 0.0116, "step": 252 }, { "epoch": 0.1681621801262878, "grad_norm": 0.6072809825448826, "learning_rate": 9.51083712390519e-06, "loss": 0.0114, "step": 253 }, { "epoch": 0.1688268527750083, "grad_norm": 0.28656877899519456, "learning_rate": 9.506179058901504e-06, "loss": 0.0094, "step": 254 }, { "epoch": 0.1694915254237288, "grad_norm": 0.4734528434094727, "learning_rate": 9.501500072390321e-06, "loss": 0.0118, "step": 255 }, { "epoch": 0.1701561980724493, "grad_norm": 0.6346082164755101, "learning_rate": 9.496800186095466e-06, "loss": 0.0096, "step": 256 }, { "epoch": 0.17082087072116983, "grad_norm": 1.2836090791228307, "learning_rate": 9.492079421837801e-06, "loss": 0.0153, "step": 257 }, { "epoch": 0.17148554336989033, "grad_norm": 0.2348493631436914, "learning_rate": 9.487337801535119e-06, "loss": 0.0111, "step": 258 }, { "epoch": 0.17215021601861083, "grad_norm": 0.3666371262438672, "learning_rate": 9.482575347202047e-06, "loss": 0.011, "step": 259 }, { "epoch": 0.17281488866733133, "grad_norm": 0.3092083173898237, "learning_rate": 9.477792080949938e-06, "loss": 0.0097, "step": 260 }, { "epoch": 0.17347956131605186, "grad_norm": 0.8686439595831802, "learning_rate": 9.472988024986775e-06, "loss": 0.0126, "step": 261 }, { "epoch": 0.17414423396477235, "grad_norm": 0.7860742657252004, "learning_rate": 9.468163201617063e-06, "loss": 0.0187, "step": 262 }, { "epoch": 0.17480890661349285, "grad_norm": 0.41866243226177396, "learning_rate": 9.463317633241724e-06, "loss": 0.0072, "step": 263 }, { "epoch": 0.17547357926221335, "grad_norm": 0.3386252993815702, "learning_rate": 9.458451342358002e-06, "loss": 0.0124, "step": 264 }, { "epoch": 0.17613825191093385, "grad_norm": 0.34075931515882407, "learning_rate": 9.453564351559348e-06, "loss": 0.0083, "step": 265 }, { "epoch": 0.17680292455965438, "grad_norm": 1.2013971234715064, "learning_rate": 9.44865668353532e-06, "loss": 0.0131, "step": 266 }, { "epoch": 0.17746759720837488, "grad_norm": 0.2910246669647025, "learning_rate": 9.44372836107148e-06, "loss": 0.0089, "step": 267 }, { "epoch": 0.17813226985709538, "grad_norm": 1.177370925091528, "learning_rate": 9.438779407049282e-06, "loss": 0.0137, "step": 268 }, { "epoch": 0.17879694250581588, "grad_norm": 0.3616215359721409, "learning_rate": 9.433809844445969e-06, "loss": 0.0092, "step": 269 }, { "epoch": 0.1794616151545364, "grad_norm": 0.5542577337457057, "learning_rate": 9.42881969633447e-06, "loss": 0.0085, "step": 270 }, { "epoch": 0.1801262878032569, "grad_norm": 0.28790832260143506, "learning_rate": 9.423808985883289e-06, "loss": 0.0108, "step": 271 }, { "epoch": 0.1807909604519774, "grad_norm": 0.29094811213738114, "learning_rate": 9.418777736356394e-06, "loss": 0.0113, "step": 272 }, { "epoch": 0.1814556331006979, "grad_norm": 0.9062547388326991, "learning_rate": 9.41372597111312e-06, "loss": 0.0132, "step": 273 }, { "epoch": 0.1821203057494184, "grad_norm": 0.8150638478196089, "learning_rate": 9.40865371360804e-06, "loss": 0.012, "step": 274 }, { "epoch": 0.18278497839813893, "grad_norm": 0.30597404629311237, "learning_rate": 9.403560987390885e-06, "loss": 0.0089, "step": 275 }, { "epoch": 0.18344965104685942, "grad_norm": 1.022326565626853, "learning_rate": 9.39844781610641e-06, "loss": 0.0117, "step": 276 }, { "epoch": 0.18411432369557992, "grad_norm": 0.47867283059459287, "learning_rate": 9.393314223494297e-06, "loss": 0.0104, "step": 277 }, { "epoch": 0.18477899634430042, "grad_norm": 0.6280778533774865, "learning_rate": 9.388160233389036e-06, "loss": 0.0141, "step": 278 }, { "epoch": 0.18544366899302095, "grad_norm": 0.8589918344000519, "learning_rate": 9.382985869719826e-06, "loss": 0.011, "step": 279 }, { "epoch": 0.18610834164174145, "grad_norm": 0.29000329893767934, "learning_rate": 9.377791156510456e-06, "loss": 0.0099, "step": 280 }, { "epoch": 0.18677301429046195, "grad_norm": 1.274686885764933, "learning_rate": 9.372576117879188e-06, "loss": 0.0147, "step": 281 }, { "epoch": 0.18743768693918245, "grad_norm": 1.048719478678211, "learning_rate": 9.367340778038663e-06, "loss": 0.0105, "step": 282 }, { "epoch": 0.18810235958790295, "grad_norm": 0.9566595733782226, "learning_rate": 9.362085161295768e-06, "loss": 0.0121, "step": 283 }, { "epoch": 0.18876703223662347, "grad_norm": 0.3255177915694972, "learning_rate": 9.35680929205154e-06, "loss": 0.0073, "step": 284 }, { "epoch": 0.18943170488534397, "grad_norm": 0.3957389778943825, "learning_rate": 9.351513194801038e-06, "loss": 0.0094, "step": 285 }, { "epoch": 0.19009637753406447, "grad_norm": 0.3736418298509006, "learning_rate": 9.346196894133239e-06, "loss": 0.0102, "step": 286 }, { "epoch": 0.19076105018278497, "grad_norm": 0.9287275379330988, "learning_rate": 9.340860414730927e-06, "loss": 0.014, "step": 287 }, { "epoch": 0.1914257228315055, "grad_norm": 1.2088552429518862, "learning_rate": 9.335503781370566e-06, "loss": 0.0134, "step": 288 }, { "epoch": 0.192090395480226, "grad_norm": 0.48462231329205135, "learning_rate": 9.330127018922195e-06, "loss": 0.0167, "step": 289 }, { "epoch": 0.1927550681289465, "grad_norm": 0.45196942733251494, "learning_rate": 9.324730152349305e-06, "loss": 0.0208, "step": 290 }, { "epoch": 0.193419740777667, "grad_norm": 0.26179557552703064, "learning_rate": 9.319313206708736e-06, "loss": 0.0105, "step": 291 }, { "epoch": 0.1940844134263875, "grad_norm": 0.2688354205211216, "learning_rate": 9.313876207150544e-06, "loss": 0.0126, "step": 292 }, { "epoch": 0.19474908607510802, "grad_norm": 0.34458442940664596, "learning_rate": 9.308419178917896e-06, "loss": 0.013, "step": 293 }, { "epoch": 0.19541375872382852, "grad_norm": 0.616587916648638, "learning_rate": 9.302942147346946e-06, "loss": 0.0118, "step": 294 }, { "epoch": 0.19607843137254902, "grad_norm": 0.63098677901715, "learning_rate": 9.297445137866726e-06, "loss": 0.0125, "step": 295 }, { "epoch": 0.19674310402126952, "grad_norm": 0.38353889099209665, "learning_rate": 9.291928175999018e-06, "loss": 0.0145, "step": 296 }, { "epoch": 0.19740777666999004, "grad_norm": 0.43109455932784146, "learning_rate": 9.286391287358243e-06, "loss": 0.0117, "step": 297 }, { "epoch": 0.19807244931871054, "grad_norm": 0.937750712577231, "learning_rate": 9.280834497651334e-06, "loss": 0.0131, "step": 298 }, { "epoch": 0.19873712196743104, "grad_norm": 0.17379806669445777, "learning_rate": 9.275257832677624e-06, "loss": 0.0104, "step": 299 }, { "epoch": 0.19940179461615154, "grad_norm": 0.2078103682572689, "learning_rate": 9.26966131832873e-06, "loss": 0.0126, "step": 300 }, { "epoch": 0.20006646726487204, "grad_norm": 0.46480655900318324, "learning_rate": 9.264044980588415e-06, "loss": 0.0118, "step": 301 }, { "epoch": 0.20073113991359257, "grad_norm": 0.3945901657429371, "learning_rate": 9.25840884553249e-06, "loss": 0.013, "step": 302 }, { "epoch": 0.20139581256231306, "grad_norm": 0.2861435884290271, "learning_rate": 9.252752939328675e-06, "loss": 0.0129, "step": 303 }, { "epoch": 0.20206048521103356, "grad_norm": 0.6116503662696815, "learning_rate": 9.247077288236488e-06, "loss": 0.0118, "step": 304 }, { "epoch": 0.20272515785975406, "grad_norm": 1.266176603756398, "learning_rate": 9.241381918607115e-06, "loss": 0.0124, "step": 305 }, { "epoch": 0.2033898305084746, "grad_norm": 0.44337541615652654, "learning_rate": 9.235666856883299e-06, "loss": 0.0069, "step": 306 }, { "epoch": 0.2040545031571951, "grad_norm": 0.5065447924774985, "learning_rate": 9.229932129599206e-06, "loss": 0.0148, "step": 307 }, { "epoch": 0.2047191758059156, "grad_norm": 0.46025045568912626, "learning_rate": 9.224177763380306e-06, "loss": 0.0113, "step": 308 }, { "epoch": 0.2053838484546361, "grad_norm": 0.7461824572174611, "learning_rate": 9.21840378494325e-06, "loss": 0.0093, "step": 309 }, { "epoch": 0.20604852110335659, "grad_norm": 0.35994200980884283, "learning_rate": 9.212610221095748e-06, "loss": 0.012, "step": 310 }, { "epoch": 0.2067131937520771, "grad_norm": 0.2101148723572065, "learning_rate": 9.20679709873644e-06, "loss": 0.0081, "step": 311 }, { "epoch": 0.2073778664007976, "grad_norm": 0.46794769192698826, "learning_rate": 9.200964444854772e-06, "loss": 0.0136, "step": 312 }, { "epoch": 0.2080425390495181, "grad_norm": 0.2343029106364292, "learning_rate": 9.195112286530874e-06, "loss": 0.0104, "step": 313 }, { "epoch": 0.2087072116982386, "grad_norm": 0.3515318424192329, "learning_rate": 9.189240650935433e-06, "loss": 0.0096, "step": 314 }, { "epoch": 0.20937188434695914, "grad_norm": 0.9014967022005104, "learning_rate": 9.183349565329562e-06, "loss": 0.0089, "step": 315 }, { "epoch": 0.21003655699567964, "grad_norm": 0.2950618379345221, "learning_rate": 9.177439057064684e-06, "loss": 0.0092, "step": 316 }, { "epoch": 0.21070122964440013, "grad_norm": 0.3731699572558992, "learning_rate": 9.17150915358239e-06, "loss": 0.0128, "step": 317 }, { "epoch": 0.21136590229312063, "grad_norm": 0.39371211020200947, "learning_rate": 9.165559882414325e-06, "loss": 0.0079, "step": 318 }, { "epoch": 0.21203057494184113, "grad_norm": 0.6272025806874811, "learning_rate": 9.159591271182058e-06, "loss": 0.0072, "step": 319 }, { "epoch": 0.21269524759056166, "grad_norm": 0.32491453802391024, "learning_rate": 9.153603347596946e-06, "loss": 0.0113, "step": 320 }, { "epoch": 0.21335992023928216, "grad_norm": 0.21680170033540036, "learning_rate": 9.147596139460013e-06, "loss": 0.0084, "step": 321 }, { "epoch": 0.21402459288800266, "grad_norm": 0.3520010563797755, "learning_rate": 9.141569674661816e-06, "loss": 0.0127, "step": 322 }, { "epoch": 0.21468926553672316, "grad_norm": 0.41941904209378145, "learning_rate": 9.13552398118232e-06, "loss": 0.0107, "step": 323 }, { "epoch": 0.21535393818544366, "grad_norm": 0.8630907179998242, "learning_rate": 9.129459087090765e-06, "loss": 0.008, "step": 324 }, { "epoch": 0.21601861083416418, "grad_norm": 0.2939661748778777, "learning_rate": 9.123375020545534e-06, "loss": 0.0127, "step": 325 }, { "epoch": 0.21668328348288468, "grad_norm": 0.8093514549410721, "learning_rate": 9.117271809794032e-06, "loss": 0.0096, "step": 326 }, { "epoch": 0.21734795613160518, "grad_norm": 0.758409064427298, "learning_rate": 9.111149483172539e-06, "loss": 0.0088, "step": 327 }, { "epoch": 0.21801262878032568, "grad_norm": 0.3592918728636694, "learning_rate": 9.105008069106093e-06, "loss": 0.0105, "step": 328 }, { "epoch": 0.2186773014290462, "grad_norm": 0.4725035440947185, "learning_rate": 9.098847596108351e-06, "loss": 0.0106, "step": 329 }, { "epoch": 0.2193419740777667, "grad_norm": 0.2780451677623958, "learning_rate": 9.092668092781454e-06, "loss": 0.014, "step": 330 }, { "epoch": 0.2200066467264872, "grad_norm": 0.8196827689316102, "learning_rate": 9.086469587815904e-06, "loss": 0.0108, "step": 331 }, { "epoch": 0.2206713193752077, "grad_norm": 0.46358625068189424, "learning_rate": 9.08025210999042e-06, "loss": 0.0127, "step": 332 }, { "epoch": 0.2213359920239282, "grad_norm": 1.1963218973895087, "learning_rate": 9.07401568817181e-06, "loss": 0.0123, "step": 333 }, { "epoch": 0.22200066467264873, "grad_norm": 0.8500930860523507, "learning_rate": 9.067760351314838e-06, "loss": 0.014, "step": 334 }, { "epoch": 0.22266533732136923, "grad_norm": 0.23029774816444434, "learning_rate": 9.061486128462087e-06, "loss": 0.0112, "step": 335 }, { "epoch": 0.22333000997008973, "grad_norm": 0.28192211343810847, "learning_rate": 9.055193048743824e-06, "loss": 0.0113, "step": 336 }, { "epoch": 0.22399468261881023, "grad_norm": 0.7620465169350527, "learning_rate": 9.048881141377863e-06, "loss": 0.0103, "step": 337 }, { "epoch": 0.22465935526753075, "grad_norm": 0.5443536511448932, "learning_rate": 9.04255043566944e-06, "loss": 0.0087, "step": 338 }, { "epoch": 0.22532402791625125, "grad_norm": 0.569846492101772, "learning_rate": 9.03620096101106e-06, "loss": 0.0109, "step": 339 }, { "epoch": 0.22598870056497175, "grad_norm": 0.33325499022680055, "learning_rate": 9.029832746882372e-06, "loss": 0.0132, "step": 340 }, { "epoch": 0.22665337321369225, "grad_norm": 0.31360893528789946, "learning_rate": 9.023445822850035e-06, "loss": 0.0081, "step": 341 }, { "epoch": 0.22731804586241275, "grad_norm": 0.21198492180143544, "learning_rate": 9.017040218567565e-06, "loss": 0.0095, "step": 342 }, { "epoch": 0.22798271851113328, "grad_norm": 0.7474064674412734, "learning_rate": 9.01061596377522e-06, "loss": 0.0134, "step": 343 }, { "epoch": 0.22864739115985377, "grad_norm": 0.21804269279272503, "learning_rate": 9.004173088299838e-06, "loss": 0.01, "step": 344 }, { "epoch": 0.22931206380857427, "grad_norm": 0.6475867863557033, "learning_rate": 8.997711622054717e-06, "loss": 0.0125, "step": 345 }, { "epoch": 0.22997673645729477, "grad_norm": 0.31127783794512365, "learning_rate": 8.991231595039464e-06, "loss": 0.0062, "step": 346 }, { "epoch": 0.2306414091060153, "grad_norm": 0.22842269717875766, "learning_rate": 8.98473303733987e-06, "loss": 0.0121, "step": 347 }, { "epoch": 0.2313060817547358, "grad_norm": 0.40610731356871477, "learning_rate": 8.97821597912775e-06, "loss": 0.0095, "step": 348 }, { "epoch": 0.2319707544034563, "grad_norm": 0.24025933765140492, "learning_rate": 8.97168045066082e-06, "loss": 0.0074, "step": 349 }, { "epoch": 0.2326354270521768, "grad_norm": 0.2507513577714693, "learning_rate": 8.96512648228255e-06, "loss": 0.0138, "step": 350 }, { "epoch": 0.2333000997008973, "grad_norm": 0.33282351330670684, "learning_rate": 8.958554104422024e-06, "loss": 0.0098, "step": 351 }, { "epoch": 0.23396477234961782, "grad_norm": 0.3152160221022002, "learning_rate": 8.951963347593797e-06, "loss": 0.0105, "step": 352 }, { "epoch": 0.23462944499833832, "grad_norm": 0.5128763476927577, "learning_rate": 8.945354242397756e-06, "loss": 0.0147, "step": 353 }, { "epoch": 0.23529411764705882, "grad_norm": 0.5184497421850583, "learning_rate": 8.938726819518978e-06, "loss": 0.0097, "step": 354 }, { "epoch": 0.23595879029577932, "grad_norm": 0.22729181658617056, "learning_rate": 8.932081109727582e-06, "loss": 0.0077, "step": 355 }, { "epoch": 0.23662346294449985, "grad_norm": 0.29795513531254586, "learning_rate": 8.925417143878594e-06, "loss": 0.0073, "step": 356 }, { "epoch": 0.23728813559322035, "grad_norm": 0.6958588648119022, "learning_rate": 8.918734952911801e-06, "loss": 0.0121, "step": 357 }, { "epoch": 0.23795280824194084, "grad_norm": 0.2218236215605131, "learning_rate": 8.9120345678516e-06, "loss": 0.0081, "step": 358 }, { "epoch": 0.23861748089066134, "grad_norm": 0.563047498692037, "learning_rate": 8.905316019806867e-06, "loss": 0.017, "step": 359 }, { "epoch": 0.23928215353938184, "grad_norm": 0.3186037846027852, "learning_rate": 8.898579339970806e-06, "loss": 0.0101, "step": 360 }, { "epoch": 0.23994682618810237, "grad_norm": 1.5189648261318371, "learning_rate": 8.891824559620801e-06, "loss": 0.0176, "step": 361 }, { "epoch": 0.24061149883682287, "grad_norm": 0.26786563410207703, "learning_rate": 8.885051710118274e-06, "loss": 0.014, "step": 362 }, { "epoch": 0.24127617148554337, "grad_norm": 0.8748035907760621, "learning_rate": 8.878260822908539e-06, "loss": 0.015, "step": 363 }, { "epoch": 0.24194084413426387, "grad_norm": 0.5267764536524051, "learning_rate": 8.871451929520662e-06, "loss": 0.0086, "step": 364 }, { "epoch": 0.2426055167829844, "grad_norm": 0.3679370256659736, "learning_rate": 8.864625061567303e-06, "loss": 0.0083, "step": 365 }, { "epoch": 0.2432701894317049, "grad_norm": 0.19517382421702098, "learning_rate": 8.857780250744574e-06, "loss": 0.0094, "step": 366 }, { "epoch": 0.2439348620804254, "grad_norm": 0.3274178987849208, "learning_rate": 8.8509175288319e-06, "loss": 0.0103, "step": 367 }, { "epoch": 0.2445995347291459, "grad_norm": 0.5437760823295965, "learning_rate": 8.844036927691857e-06, "loss": 0.0088, "step": 368 }, { "epoch": 0.2452642073778664, "grad_norm": 0.4789697861288138, "learning_rate": 8.837138479270037e-06, "loss": 0.0114, "step": 369 }, { "epoch": 0.24592888002658692, "grad_norm": 0.5441389645820673, "learning_rate": 8.83022221559489e-06, "loss": 0.0101, "step": 370 }, { "epoch": 0.24659355267530741, "grad_norm": 0.4010415785554097, "learning_rate": 8.823288168777586e-06, "loss": 0.015, "step": 371 }, { "epoch": 0.2472582253240279, "grad_norm": 0.31952561251441525, "learning_rate": 8.816336371011851e-06, "loss": 0.0106, "step": 372 }, { "epoch": 0.2479228979727484, "grad_norm": 0.39700805216169793, "learning_rate": 8.80936685457383e-06, "loss": 0.0142, "step": 373 }, { "epoch": 0.24858757062146894, "grad_norm": 0.6593115149327142, "learning_rate": 8.802379651821938e-06, "loss": 0.0121, "step": 374 }, { "epoch": 0.24925224327018944, "grad_norm": 0.4678738658206623, "learning_rate": 8.795374795196694e-06, "loss": 0.0117, "step": 375 }, { "epoch": 0.24991691591890994, "grad_norm": 0.4850606055915218, "learning_rate": 8.78835231722059e-06, "loss": 0.0094, "step": 376 }, { "epoch": 0.25058158856763046, "grad_norm": 0.26911040386767177, "learning_rate": 8.78131225049793e-06, "loss": 0.014, "step": 377 }, { "epoch": 0.25124626121635096, "grad_norm": 0.9418348960665115, "learning_rate": 8.774254627714676e-06, "loss": 0.0165, "step": 378 }, { "epoch": 0.25191093386507146, "grad_norm": 1.0045633115510086, "learning_rate": 8.767179481638303e-06, "loss": 0.0151, "step": 379 }, { "epoch": 0.25257560651379196, "grad_norm": 0.24951907950013252, "learning_rate": 8.760086845117648e-06, "loss": 0.0102, "step": 380 }, { "epoch": 0.25324027916251246, "grad_norm": 0.17956728755131446, "learning_rate": 8.752976751082742e-06, "loss": 0.0095, "step": 381 }, { "epoch": 0.25390495181123296, "grad_norm": 0.5177895217836687, "learning_rate": 8.74584923254468e-06, "loss": 0.0108, "step": 382 }, { "epoch": 0.25456962445995346, "grad_norm": 0.18771880731650165, "learning_rate": 8.738704322595453e-06, "loss": 0.0081, "step": 383 }, { "epoch": 0.25523429710867396, "grad_norm": 1.0095335305368216, "learning_rate": 8.731542054407793e-06, "loss": 0.0126, "step": 384 }, { "epoch": 0.25589896975739446, "grad_norm": 0.36502806171872737, "learning_rate": 8.72436246123503e-06, "loss": 0.0138, "step": 385 }, { "epoch": 0.256563642406115, "grad_norm": 0.25787011889294087, "learning_rate": 8.71716557641093e-06, "loss": 0.0113, "step": 386 }, { "epoch": 0.2572283150548355, "grad_norm": 0.2567733474614457, "learning_rate": 8.709951433349536e-06, "loss": 0.0149, "step": 387 }, { "epoch": 0.257892987703556, "grad_norm": 0.43404962933931723, "learning_rate": 8.702720065545024e-06, "loss": 0.0095, "step": 388 }, { "epoch": 0.2585576603522765, "grad_norm": 0.45023422702603766, "learning_rate": 8.695471506571543e-06, "loss": 0.0077, "step": 389 }, { "epoch": 0.259222333000997, "grad_norm": 0.4996924976589022, "learning_rate": 8.688205790083053e-06, "loss": 0.0083, "step": 390 }, { "epoch": 0.2598870056497175, "grad_norm": 0.4030334491475912, "learning_rate": 8.680922949813177e-06, "loss": 0.0147, "step": 391 }, { "epoch": 0.260551678298438, "grad_norm": 0.42487361717662014, "learning_rate": 8.673623019575042e-06, "loss": 0.0111, "step": 392 }, { "epoch": 0.2612163509471585, "grad_norm": 0.47898709390859606, "learning_rate": 8.666306033261115e-06, "loss": 0.0115, "step": 393 }, { "epoch": 0.261881023595879, "grad_norm": 0.34131185830655547, "learning_rate": 8.658972024843063e-06, "loss": 0.0098, "step": 394 }, { "epoch": 0.26254569624459956, "grad_norm": 0.4979203081696408, "learning_rate": 8.65162102837157e-06, "loss": 0.0104, "step": 395 }, { "epoch": 0.26321036889332006, "grad_norm": 0.42398071579391594, "learning_rate": 8.644253077976204e-06, "loss": 0.0088, "step": 396 }, { "epoch": 0.26387504154204056, "grad_norm": 0.5405134523432256, "learning_rate": 8.636868207865244e-06, "loss": 0.0089, "step": 397 }, { "epoch": 0.26453971419076105, "grad_norm": 0.41061409174976227, "learning_rate": 8.629466452325524e-06, "loss": 0.0107, "step": 398 }, { "epoch": 0.26520438683948155, "grad_norm": 0.46723617125018985, "learning_rate": 8.622047845722274e-06, "loss": 0.0134, "step": 399 }, { "epoch": 0.26586905948820205, "grad_norm": 0.7092658462268281, "learning_rate": 8.614612422498965e-06, "loss": 0.0151, "step": 400 }, { "epoch": 0.26653373213692255, "grad_norm": 0.3376115595460269, "learning_rate": 8.607160217177137e-06, "loss": 0.0095, "step": 401 }, { "epoch": 0.26719840478564305, "grad_norm": 0.35138933824561336, "learning_rate": 8.599691264356257e-06, "loss": 0.0128, "step": 402 }, { "epoch": 0.26786307743436355, "grad_norm": 0.358801121918168, "learning_rate": 8.592205598713539e-06, "loss": 0.0117, "step": 403 }, { "epoch": 0.2685277500830841, "grad_norm": 0.3920759040185174, "learning_rate": 8.584703255003796e-06, "loss": 0.0122, "step": 404 }, { "epoch": 0.2691924227318046, "grad_norm": 0.3628334215358763, "learning_rate": 8.577184268059276e-06, "loss": 0.0105, "step": 405 }, { "epoch": 0.2698570953805251, "grad_norm": 0.3034806663580783, "learning_rate": 8.569648672789496e-06, "loss": 0.0103, "step": 406 }, { "epoch": 0.2705217680292456, "grad_norm": 0.2335545293541909, "learning_rate": 8.562096504181086e-06, "loss": 0.0085, "step": 407 }, { "epoch": 0.2711864406779661, "grad_norm": 0.19423184842598806, "learning_rate": 8.554527797297622e-06, "loss": 0.0076, "step": 408 }, { "epoch": 0.2718511133266866, "grad_norm": 0.3416481152189461, "learning_rate": 8.546942587279465e-06, "loss": 0.0137, "step": 409 }, { "epoch": 0.2725157859754071, "grad_norm": 0.48961552246390344, "learning_rate": 8.539340909343597e-06, "loss": 0.0108, "step": 410 }, { "epoch": 0.2731804586241276, "grad_norm": 0.24911458267400743, "learning_rate": 8.531722798783457e-06, "loss": 0.0105, "step": 411 }, { "epoch": 0.2738451312728481, "grad_norm": 0.3616792302604157, "learning_rate": 8.524088290968781e-06, "loss": 0.0087, "step": 412 }, { "epoch": 0.27450980392156865, "grad_norm": 1.284981155801795, "learning_rate": 8.516437421345433e-06, "loss": 0.0124, "step": 413 }, { "epoch": 0.27517447657028915, "grad_norm": 0.5693005072050209, "learning_rate": 8.508770225435244e-06, "loss": 0.0176, "step": 414 }, { "epoch": 0.27583914921900965, "grad_norm": 0.3944381971204075, "learning_rate": 8.501086738835843e-06, "loss": 0.0107, "step": 415 }, { "epoch": 0.27650382186773015, "grad_norm": 0.609516123860261, "learning_rate": 8.493386997220497e-06, "loss": 0.0106, "step": 416 }, { "epoch": 0.27716849451645065, "grad_norm": 0.4942809285583405, "learning_rate": 8.485671036337939e-06, "loss": 0.0114, "step": 417 }, { "epoch": 0.27783316716517115, "grad_norm": 0.26860228762066785, "learning_rate": 8.477938892012209e-06, "loss": 0.0071, "step": 418 }, { "epoch": 0.27849783981389165, "grad_norm": 0.38431318734345415, "learning_rate": 8.470190600142486e-06, "loss": 0.0113, "step": 419 }, { "epoch": 0.27916251246261214, "grad_norm": 0.39653494764365277, "learning_rate": 8.462426196702912e-06, "loss": 0.0145, "step": 420 }, { "epoch": 0.27982718511133264, "grad_norm": 0.4999818013714175, "learning_rate": 8.45464571774244e-06, "loss": 0.0137, "step": 421 }, { "epoch": 0.2804918577600532, "grad_norm": 0.32150423393322547, "learning_rate": 8.446849199384655e-06, "loss": 0.0106, "step": 422 }, { "epoch": 0.2811565304087737, "grad_norm": 0.43625928006546033, "learning_rate": 8.439036677827613e-06, "loss": 0.0115, "step": 423 }, { "epoch": 0.2818212030574942, "grad_norm": 0.30735825848324666, "learning_rate": 8.43120818934367e-06, "loss": 0.0087, "step": 424 }, { "epoch": 0.2824858757062147, "grad_norm": 0.31453582147593395, "learning_rate": 8.423363770279311e-06, "loss": 0.0087, "step": 425 }, { "epoch": 0.2831505483549352, "grad_norm": 0.19779849085464432, "learning_rate": 8.415503457054988e-06, "loss": 0.008, "step": 426 }, { "epoch": 0.2838152210036557, "grad_norm": 0.39045756110991126, "learning_rate": 8.407627286164948e-06, "loss": 0.0124, "step": 427 }, { "epoch": 0.2844798936523762, "grad_norm": 0.38524411622428917, "learning_rate": 8.399735294177056e-06, "loss": 0.0191, "step": 428 }, { "epoch": 0.2851445663010967, "grad_norm": 0.30080223844735193, "learning_rate": 8.391827517732642e-06, "loss": 0.0136, "step": 429 }, { "epoch": 0.2858092389498172, "grad_norm": 0.6624891931019012, "learning_rate": 8.38390399354631e-06, "loss": 0.0147, "step": 430 }, { "epoch": 0.28647391159853774, "grad_norm": 0.2988303104566721, "learning_rate": 8.375964758405788e-06, "loss": 0.0112, "step": 431 }, { "epoch": 0.28713858424725824, "grad_norm": 0.47342069754821636, "learning_rate": 8.368009849171743e-06, "loss": 0.0099, "step": 432 }, { "epoch": 0.28780325689597874, "grad_norm": 0.792605112928616, "learning_rate": 8.360039302777614e-06, "loss": 0.0125, "step": 433 }, { "epoch": 0.28846792954469924, "grad_norm": 0.4412882780314945, "learning_rate": 8.352053156229439e-06, "loss": 0.0097, "step": 434 }, { "epoch": 0.28913260219341974, "grad_norm": 0.6754563553315797, "learning_rate": 8.344051446605692e-06, "loss": 0.0142, "step": 435 }, { "epoch": 0.28979727484214024, "grad_norm": 0.3241169841658398, "learning_rate": 8.336034211057098e-06, "loss": 0.0093, "step": 436 }, { "epoch": 0.29046194749086074, "grad_norm": 0.7780228945460835, "learning_rate": 8.32800148680647e-06, "loss": 0.0132, "step": 437 }, { "epoch": 0.29112662013958124, "grad_norm": 0.6831936931028925, "learning_rate": 8.31995331114853e-06, "loss": 0.011, "step": 438 }, { "epoch": 0.29179129278830174, "grad_norm": 0.760445970745941, "learning_rate": 8.31188972144974e-06, "loss": 0.0122, "step": 439 }, { "epoch": 0.2924559654370223, "grad_norm": 0.2387277082202268, "learning_rate": 8.303810755148127e-06, "loss": 0.0124, "step": 440 }, { "epoch": 0.2931206380857428, "grad_norm": 0.6530178367626034, "learning_rate": 8.295716449753109e-06, "loss": 0.0093, "step": 441 }, { "epoch": 0.2937853107344633, "grad_norm": 0.5627289642461663, "learning_rate": 8.28760684284532e-06, "loss": 0.0109, "step": 442 }, { "epoch": 0.2944499833831838, "grad_norm": 0.31706537962870157, "learning_rate": 8.279481972076437e-06, "loss": 0.0104, "step": 443 }, { "epoch": 0.2951146560319043, "grad_norm": 0.3239172163384193, "learning_rate": 8.27134187516901e-06, "loss": 0.0108, "step": 444 }, { "epoch": 0.2957793286806248, "grad_norm": 0.3806269393045076, "learning_rate": 8.263186589916273e-06, "loss": 0.0137, "step": 445 }, { "epoch": 0.2964440013293453, "grad_norm": 0.48123055782712915, "learning_rate": 8.255016154181982e-06, "loss": 0.0121, "step": 446 }, { "epoch": 0.2971086739780658, "grad_norm": 1.5483132046638926, "learning_rate": 8.246830605900237e-06, "loss": 0.009, "step": 447 }, { "epoch": 0.2977733466267863, "grad_norm": 0.34353844140531886, "learning_rate": 8.238629983075296e-06, "loss": 0.0143, "step": 448 }, { "epoch": 0.29843801927550684, "grad_norm": 0.5404915543494925, "learning_rate": 8.23041432378141e-06, "loss": 0.0155, "step": 449 }, { "epoch": 0.29910269192422734, "grad_norm": 0.40938442505083267, "learning_rate": 8.222183666162647e-06, "loss": 0.0126, "step": 450 }, { "epoch": 0.29976736457294784, "grad_norm": 0.3715257554780622, "learning_rate": 8.213938048432697e-06, "loss": 0.0123, "step": 451 }, { "epoch": 0.30043203722166834, "grad_norm": 0.5651278646110928, "learning_rate": 8.20567750887472e-06, "loss": 0.0069, "step": 452 }, { "epoch": 0.30109670987038883, "grad_norm": 0.2789217749494629, "learning_rate": 8.197402085841154e-06, "loss": 0.0099, "step": 453 }, { "epoch": 0.30176138251910933, "grad_norm": 0.3063850384150043, "learning_rate": 8.18911181775353e-06, "loss": 0.0134, "step": 454 }, { "epoch": 0.30242605516782983, "grad_norm": 0.7557480230146246, "learning_rate": 8.180806743102312e-06, "loss": 0.0159, "step": 455 }, { "epoch": 0.30309072781655033, "grad_norm": 0.15858503833790136, "learning_rate": 8.172486900446703e-06, "loss": 0.0085, "step": 456 }, { "epoch": 0.30375540046527083, "grad_norm": 0.23933761601257225, "learning_rate": 8.164152328414476e-06, "loss": 0.0135, "step": 457 }, { "epoch": 0.3044200731139914, "grad_norm": 0.30771960991648895, "learning_rate": 8.155803065701783e-06, "loss": 0.0102, "step": 458 }, { "epoch": 0.3050847457627119, "grad_norm": 0.28084340695901117, "learning_rate": 8.147439151072993e-06, "loss": 0.0094, "step": 459 }, { "epoch": 0.3057494184114324, "grad_norm": 0.4302601876799286, "learning_rate": 8.139060623360494e-06, "loss": 0.0114, "step": 460 }, { "epoch": 0.3064140910601529, "grad_norm": 0.19401937027992683, "learning_rate": 8.130667521464519e-06, "loss": 0.0089, "step": 461 }, { "epoch": 0.3070787637088734, "grad_norm": 0.44817990023681686, "learning_rate": 8.122259884352975e-06, "loss": 0.0108, "step": 462 }, { "epoch": 0.3077434363575939, "grad_norm": 0.5668528563734161, "learning_rate": 8.113837751061246e-06, "loss": 0.0106, "step": 463 }, { "epoch": 0.3084081090063144, "grad_norm": 0.8975048950597451, "learning_rate": 8.105401160692023e-06, "loss": 0.0118, "step": 464 }, { "epoch": 0.3090727816550349, "grad_norm": 0.3319306887880883, "learning_rate": 8.09695015241512e-06, "loss": 0.0101, "step": 465 }, { "epoch": 0.3097374543037554, "grad_norm": 0.5140411193910503, "learning_rate": 8.088484765467286e-06, "loss": 0.0094, "step": 466 }, { "epoch": 0.31040212695247593, "grad_norm": 0.4864559042130476, "learning_rate": 8.080005039152034e-06, "loss": 0.0108, "step": 467 }, { "epoch": 0.31106679960119643, "grad_norm": 0.3729635503725394, "learning_rate": 8.071511012839453e-06, "loss": 0.0097, "step": 468 }, { "epoch": 0.31173147224991693, "grad_norm": 0.32374458242313886, "learning_rate": 8.063002725966014e-06, "loss": 0.009, "step": 469 }, { "epoch": 0.31239614489863743, "grad_norm": 0.6485688187501434, "learning_rate": 8.054480218034415e-06, "loss": 0.0105, "step": 470 }, { "epoch": 0.31306081754735793, "grad_norm": 0.3046125855854896, "learning_rate": 8.045943528613362e-06, "loss": 0.0106, "step": 471 }, { "epoch": 0.3137254901960784, "grad_norm": 0.3601738458811542, "learning_rate": 8.037392697337418e-06, "loss": 0.0122, "step": 472 }, { "epoch": 0.3143901628447989, "grad_norm": 0.20844942856355533, "learning_rate": 8.028827763906798e-06, "loss": 0.0095, "step": 473 }, { "epoch": 0.3150548354935194, "grad_norm": 0.1657352756386714, "learning_rate": 8.020248768087189e-06, "loss": 0.0053, "step": 474 }, { "epoch": 0.3157195081422399, "grad_norm": 0.29952519086009544, "learning_rate": 8.011655749709575e-06, "loss": 0.0069, "step": 475 }, { "epoch": 0.3163841807909605, "grad_norm": 0.6057990427958913, "learning_rate": 8.003048748670034e-06, "loss": 0.013, "step": 476 }, { "epoch": 0.317048853439681, "grad_norm": 0.2813495851686625, "learning_rate": 7.994427804929573e-06, "loss": 0.0092, "step": 477 }, { "epoch": 0.3177135260884015, "grad_norm": 0.7155250597834967, "learning_rate": 7.985792958513932e-06, "loss": 0.0089, "step": 478 }, { "epoch": 0.318378198737122, "grad_norm": 0.5455272258190905, "learning_rate": 7.977144249513391e-06, "loss": 0.011, "step": 479 }, { "epoch": 0.3190428713858425, "grad_norm": 0.30317424211233834, "learning_rate": 7.968481718082601e-06, "loss": 0.012, "step": 480 }, { "epoch": 0.319707544034563, "grad_norm": 0.47539151269639346, "learning_rate": 7.95980540444038e-06, "loss": 0.0093, "step": 481 }, { "epoch": 0.3203722166832835, "grad_norm": 0.2173242191148894, "learning_rate": 7.951115348869546e-06, "loss": 0.0073, "step": 482 }, { "epoch": 0.32103688933200397, "grad_norm": 0.5601198104618945, "learning_rate": 7.942411591716706e-06, "loss": 0.0129, "step": 483 }, { "epoch": 0.32170156198072447, "grad_norm": 0.449087158007511, "learning_rate": 7.93369417339209e-06, "loss": 0.0083, "step": 484 }, { "epoch": 0.322366234629445, "grad_norm": 0.20358827965600385, "learning_rate": 7.924963134369353e-06, "loss": 0.0103, "step": 485 }, { "epoch": 0.3230309072781655, "grad_norm": 0.2804197027689223, "learning_rate": 7.916218515185386e-06, "loss": 0.01, "step": 486 }, { "epoch": 0.323695579926886, "grad_norm": 0.21137979568768858, "learning_rate": 7.907460356440133e-06, "loss": 0.0079, "step": 487 }, { "epoch": 0.3243602525756065, "grad_norm": 0.4262053657453829, "learning_rate": 7.898688698796403e-06, "loss": 0.0093, "step": 488 }, { "epoch": 0.325024925224327, "grad_norm": 0.2973515718681296, "learning_rate": 7.88990358297967e-06, "loss": 0.0104, "step": 489 }, { "epoch": 0.3256895978730475, "grad_norm": 0.694934200823646, "learning_rate": 7.881105049777902e-06, "loss": 0.0131, "step": 490 }, { "epoch": 0.326354270521768, "grad_norm": 0.5129657361426667, "learning_rate": 7.872293140041353e-06, "loss": 0.0166, "step": 491 }, { "epoch": 0.3270189431704885, "grad_norm": 0.3298061640218595, "learning_rate": 7.863467894682391e-06, "loss": 0.0088, "step": 492 }, { "epoch": 0.327683615819209, "grad_norm": 0.25416534596466156, "learning_rate": 7.854629354675292e-06, "loss": 0.0087, "step": 493 }, { "epoch": 0.32834828846792957, "grad_norm": 0.5648136194724173, "learning_rate": 7.84577756105606e-06, "loss": 0.0149, "step": 494 }, { "epoch": 0.32901296111665007, "grad_norm": 0.204272615149441, "learning_rate": 7.836912554922233e-06, "loss": 0.0076, "step": 495 }, { "epoch": 0.32967763376537057, "grad_norm": 0.17799867208323966, "learning_rate": 7.828034377432694e-06, "loss": 0.0102, "step": 496 }, { "epoch": 0.33034230641409107, "grad_norm": 0.5184612406773145, "learning_rate": 7.819143069807477e-06, "loss": 0.0118, "step": 497 }, { "epoch": 0.33100697906281157, "grad_norm": 0.33867929197688984, "learning_rate": 7.81023867332758e-06, "loss": 0.0095, "step": 498 }, { "epoch": 0.33167165171153207, "grad_norm": 0.21571360389307873, "learning_rate": 7.801321229334764e-06, "loss": 0.0095, "step": 499 }, { "epoch": 0.33233632436025257, "grad_norm": 0.2050339512795772, "learning_rate": 7.792390779231374e-06, "loss": 0.0095, "step": 500 }, { "epoch": 0.33300099700897307, "grad_norm": 0.43703162858153477, "learning_rate": 7.783447364480142e-06, "loss": 0.0138, "step": 501 }, { "epoch": 0.33366566965769356, "grad_norm": 0.30256871421988485, "learning_rate": 7.774491026603985e-06, "loss": 0.0076, "step": 502 }, { "epoch": 0.33433034230641406, "grad_norm": 0.7205885546658176, "learning_rate": 7.765521807185827e-06, "loss": 0.0148, "step": 503 }, { "epoch": 0.3349950149551346, "grad_norm": 0.5592593279206337, "learning_rate": 7.756539747868394e-06, "loss": 0.0128, "step": 504 }, { "epoch": 0.3356596876038551, "grad_norm": 0.512450644282298, "learning_rate": 7.747544890354031e-06, "loss": 0.0088, "step": 505 }, { "epoch": 0.3363243602525756, "grad_norm": 0.19914259076708438, "learning_rate": 7.738537276404498e-06, "loss": 0.0061, "step": 506 }, { "epoch": 0.3369890329012961, "grad_norm": 0.34264603764367224, "learning_rate": 7.729516947840786e-06, "loss": 0.0096, "step": 507 }, { "epoch": 0.3376537055500166, "grad_norm": 0.24214521709185532, "learning_rate": 7.720483946542913e-06, "loss": 0.0085, "step": 508 }, { "epoch": 0.3383183781987371, "grad_norm": 0.6007361020731513, "learning_rate": 7.711438314449741e-06, "loss": 0.0109, "step": 509 }, { "epoch": 0.3389830508474576, "grad_norm": 0.34469241082183033, "learning_rate": 7.702380093558766e-06, "loss": 0.0086, "step": 510 }, { "epoch": 0.3396477234961781, "grad_norm": 0.525829771732804, "learning_rate": 7.69330932592594e-06, "loss": 0.0121, "step": 511 }, { "epoch": 0.3403123961448986, "grad_norm": 0.4051135935833249, "learning_rate": 7.684226053665463e-06, "loss": 0.0105, "step": 512 }, { "epoch": 0.34097706879361916, "grad_norm": 0.4766264863702415, "learning_rate": 7.675130318949596e-06, "loss": 0.0123, "step": 513 }, { "epoch": 0.34164174144233966, "grad_norm": 0.4271732174526758, "learning_rate": 7.666022164008458e-06, "loss": 0.011, "step": 514 }, { "epoch": 0.34230641409106016, "grad_norm": 0.25241378574135087, "learning_rate": 7.656901631129832e-06, "loss": 0.006, "step": 515 }, { "epoch": 0.34297108673978066, "grad_norm": 0.49010559634428563, "learning_rate": 7.647768762658973e-06, "loss": 0.0107, "step": 516 }, { "epoch": 0.34363575938850116, "grad_norm": 0.199589455729234, "learning_rate": 7.638623600998409e-06, "loss": 0.0069, "step": 517 }, { "epoch": 0.34430043203722166, "grad_norm": 0.34199927633451555, "learning_rate": 7.629466188607741e-06, "loss": 0.0121, "step": 518 }, { "epoch": 0.34496510468594216, "grad_norm": 0.45761547602935065, "learning_rate": 7.620296568003449e-06, "loss": 0.0141, "step": 519 }, { "epoch": 0.34562977733466266, "grad_norm": 0.3006576977771372, "learning_rate": 7.6111147817586925e-06, "loss": 0.0103, "step": 520 }, { "epoch": 0.34629444998338316, "grad_norm": 0.22844414410638084, "learning_rate": 7.601920872503119e-06, "loss": 0.0101, "step": 521 }, { "epoch": 0.3469591226321037, "grad_norm": 0.6106381720132763, "learning_rate": 7.592714882922654e-06, "loss": 0.0121, "step": 522 }, { "epoch": 0.3476237952808242, "grad_norm": 0.3480161829431828, "learning_rate": 7.5834968557593155e-06, "loss": 0.0103, "step": 523 }, { "epoch": 0.3482884679295447, "grad_norm": 0.19957169893377527, "learning_rate": 7.574266833811009e-06, "loss": 0.0076, "step": 524 }, { "epoch": 0.3489531405782652, "grad_norm": 0.209959751517232, "learning_rate": 7.565024859931331e-06, "loss": 0.0077, "step": 525 }, { "epoch": 0.3496178132269857, "grad_norm": 0.23232856968639554, "learning_rate": 7.5557709770293664e-06, "loss": 0.0076, "step": 526 }, { "epoch": 0.3502824858757062, "grad_norm": 0.32119253379340007, "learning_rate": 7.5465052280694935e-06, "loss": 0.0076, "step": 527 }, { "epoch": 0.3509471585244267, "grad_norm": 0.2811745173719261, "learning_rate": 7.537227656071181e-06, "loss": 0.0097, "step": 528 }, { "epoch": 0.3516118311731472, "grad_norm": 0.4791498119473934, "learning_rate": 7.527938304108795e-06, "loss": 0.0105, "step": 529 }, { "epoch": 0.3522765038218677, "grad_norm": 0.5917609290831101, "learning_rate": 7.518637215311388e-06, "loss": 0.0114, "step": 530 }, { "epoch": 0.35294117647058826, "grad_norm": 0.4670496649161578, "learning_rate": 7.509324432862511e-06, "loss": 0.0093, "step": 531 }, { "epoch": 0.35360584911930876, "grad_norm": 0.5621782659311169, "learning_rate": 7.500000000000001e-06, "loss": 0.0117, "step": 532 }, { "epoch": 0.35427052176802926, "grad_norm": 0.8798948032508013, "learning_rate": 7.490663960015789e-06, "loss": 0.0119, "step": 533 }, { "epoch": 0.35493519441674976, "grad_norm": 0.339325429479605, "learning_rate": 7.4813163562556975e-06, "loss": 0.0087, "step": 534 }, { "epoch": 0.35559986706547025, "grad_norm": 0.35731424592022903, "learning_rate": 7.471957232119235e-06, "loss": 0.009, "step": 535 }, { "epoch": 0.35626453971419075, "grad_norm": 0.22515905276759726, "learning_rate": 7.4625866310594e-06, "loss": 0.0086, "step": 536 }, { "epoch": 0.35692921236291125, "grad_norm": 0.9630185262488252, "learning_rate": 7.453204596582473e-06, "loss": 0.0122, "step": 537 }, { "epoch": 0.35759388501163175, "grad_norm": 0.6945298851942855, "learning_rate": 7.443811172247822e-06, "loss": 0.0103, "step": 538 }, { "epoch": 0.35825855766035225, "grad_norm": 0.4226391903210947, "learning_rate": 7.434406401667695e-06, "loss": 0.0077, "step": 539 }, { "epoch": 0.3589232303090728, "grad_norm": 0.35818150479304384, "learning_rate": 7.424990328507017e-06, "loss": 0.011, "step": 540 }, { "epoch": 0.3595879029577933, "grad_norm": 0.8353613154149249, "learning_rate": 7.415562996483193e-06, "loss": 0.0109, "step": 541 }, { "epoch": 0.3602525756065138, "grad_norm": 0.8944476960014304, "learning_rate": 7.406124449365898e-06, "loss": 0.0095, "step": 542 }, { "epoch": 0.3609172482552343, "grad_norm": 0.6826581321816971, "learning_rate": 7.3966747309768775e-06, "loss": 0.0127, "step": 543 }, { "epoch": 0.3615819209039548, "grad_norm": 0.6865486768650689, "learning_rate": 7.387213885189746e-06, "loss": 0.0119, "step": 544 }, { "epoch": 0.3622465935526753, "grad_norm": 0.6578033790468896, "learning_rate": 7.377741955929776e-06, "loss": 0.009, "step": 545 }, { "epoch": 0.3629112662013958, "grad_norm": 0.7825103372622395, "learning_rate": 7.368258987173704e-06, "loss": 0.0105, "step": 546 }, { "epoch": 0.3635759388501163, "grad_norm": 0.2974677929905823, "learning_rate": 7.358765022949519e-06, "loss": 0.0079, "step": 547 }, { "epoch": 0.3642406114988368, "grad_norm": 0.8406057581848128, "learning_rate": 7.349260107336258e-06, "loss": 0.0127, "step": 548 }, { "epoch": 0.36490528414755735, "grad_norm": 0.7104386927886566, "learning_rate": 7.339744284463808e-06, "loss": 0.0153, "step": 549 }, { "epoch": 0.36556995679627785, "grad_norm": 0.19131288266198826, "learning_rate": 7.330217598512696e-06, "loss": 0.0059, "step": 550 }, { "epoch": 0.36623462944499835, "grad_norm": 0.47227167848071316, "learning_rate": 7.320680093713879e-06, "loss": 0.0102, "step": 551 }, { "epoch": 0.36689930209371885, "grad_norm": 0.6376475359173951, "learning_rate": 7.311131814348551e-06, "loss": 0.009, "step": 552 }, { "epoch": 0.36756397474243935, "grad_norm": 0.37694529213123096, "learning_rate": 7.30157280474793e-06, "loss": 0.0132, "step": 553 }, { "epoch": 0.36822864739115985, "grad_norm": 0.46939561103149463, "learning_rate": 7.292003109293048e-06, "loss": 0.0137, "step": 554 }, { "epoch": 0.36889332003988035, "grad_norm": 0.692309905775168, "learning_rate": 7.282422772414555e-06, "loss": 0.012, "step": 555 }, { "epoch": 0.36955799268860084, "grad_norm": 0.30492987182103687, "learning_rate": 7.2728318385925035e-06, "loss": 0.0144, "step": 556 }, { "epoch": 0.37022266533732134, "grad_norm": 0.2856525284147588, "learning_rate": 7.263230352356151e-06, "loss": 0.0102, "step": 557 }, { "epoch": 0.3708873379860419, "grad_norm": 0.20680039864816357, "learning_rate": 7.253618358283743e-06, "loss": 0.0105, "step": 558 }, { "epoch": 0.3715520106347624, "grad_norm": 0.43588697601101783, "learning_rate": 7.243995901002312e-06, "loss": 0.0123, "step": 559 }, { "epoch": 0.3722166832834829, "grad_norm": 0.5251581392438576, "learning_rate": 7.234363025187474e-06, "loss": 0.0108, "step": 560 }, { "epoch": 0.3728813559322034, "grad_norm": 0.45664528792412595, "learning_rate": 7.2247197755632124e-06, "loss": 0.0093, "step": 561 }, { "epoch": 0.3735460285809239, "grad_norm": 0.6103685308142055, "learning_rate": 7.215066196901676e-06, "loss": 0.0091, "step": 562 }, { "epoch": 0.3742107012296444, "grad_norm": 0.22998940357567568, "learning_rate": 7.205402334022971e-06, "loss": 0.0096, "step": 563 }, { "epoch": 0.3748753738783649, "grad_norm": 0.16847007275316825, "learning_rate": 7.19572823179495e-06, "loss": 0.0061, "step": 564 }, { "epoch": 0.3755400465270854, "grad_norm": 0.6616215209094138, "learning_rate": 7.186043935133005e-06, "loss": 0.0126, "step": 565 }, { "epoch": 0.3762047191758059, "grad_norm": 0.5288671713506298, "learning_rate": 7.176349488999859e-06, "loss": 0.0096, "step": 566 }, { "epoch": 0.37686939182452645, "grad_norm": 0.7988443891963547, "learning_rate": 7.166644938405361e-06, "loss": 0.0127, "step": 567 }, { "epoch": 0.37753406447324694, "grad_norm": 0.35447102070889436, "learning_rate": 7.156930328406268e-06, "loss": 0.0096, "step": 568 }, { "epoch": 0.37819873712196744, "grad_norm": 0.29078404959346027, "learning_rate": 7.147205704106046e-06, "loss": 0.0064, "step": 569 }, { "epoch": 0.37886340977068794, "grad_norm": 0.6502569688205814, "learning_rate": 7.137471110654656e-06, "loss": 0.01, "step": 570 }, { "epoch": 0.37952808241940844, "grad_norm": 0.5586458237706238, "learning_rate": 7.127726593248337e-06, "loss": 0.0108, "step": 571 }, { "epoch": 0.38019275506812894, "grad_norm": 0.2870302268653187, "learning_rate": 7.117972197129415e-06, "loss": 0.0061, "step": 572 }, { "epoch": 0.38085742771684944, "grad_norm": 0.35915045178879285, "learning_rate": 7.108207967586071e-06, "loss": 0.0051, "step": 573 }, { "epoch": 0.38152210036556994, "grad_norm": 0.4095410816701309, "learning_rate": 7.098433949952146e-06, "loss": 0.0099, "step": 574 }, { "epoch": 0.38218677301429044, "grad_norm": 0.6747149296750776, "learning_rate": 7.088650189606926e-06, "loss": 0.0117, "step": 575 }, { "epoch": 0.382851445663011, "grad_norm": 0.5789999982523566, "learning_rate": 7.07885673197493e-06, "loss": 0.0114, "step": 576 }, { "epoch": 0.3835161183117315, "grad_norm": 0.42685246555086065, "learning_rate": 7.069053622525697e-06, "loss": 0.0108, "step": 577 }, { "epoch": 0.384180790960452, "grad_norm": 0.5053574196210717, "learning_rate": 7.059240906773583e-06, "loss": 0.0133, "step": 578 }, { "epoch": 0.3848454636091725, "grad_norm": 0.3332957877727812, "learning_rate": 7.049418630277543e-06, "loss": 0.009, "step": 579 }, { "epoch": 0.385510136257893, "grad_norm": 0.5521284959380209, "learning_rate": 7.039586838640918e-06, "loss": 0.0118, "step": 580 }, { "epoch": 0.3861748089066135, "grad_norm": 0.3492766426453421, "learning_rate": 7.029745577511232e-06, "loss": 0.0127, "step": 581 }, { "epoch": 0.386839481555334, "grad_norm": 0.5377414655718215, "learning_rate": 7.019894892579969e-06, "loss": 0.008, "step": 582 }, { "epoch": 0.3875041542040545, "grad_norm": 0.14878035808350976, "learning_rate": 7.0100348295823706e-06, "loss": 0.009, "step": 583 }, { "epoch": 0.388168826852775, "grad_norm": 0.52529745526814, "learning_rate": 7.000165434297214e-06, "loss": 0.0103, "step": 584 }, { "epoch": 0.38883349950149554, "grad_norm": 0.2224555860564274, "learning_rate": 6.9902867525466116e-06, "loss": 0.0105, "step": 585 }, { "epoch": 0.38949817215021604, "grad_norm": 0.15021128763326028, "learning_rate": 6.980398830195785e-06, "loss": 0.0065, "step": 586 }, { "epoch": 0.39016284479893654, "grad_norm": 0.1877258462930616, "learning_rate": 6.9705017131528625e-06, "loss": 0.0095, "step": 587 }, { "epoch": 0.39082751744765704, "grad_norm": 0.2545920726128384, "learning_rate": 6.960595447368662e-06, "loss": 0.0096, "step": 588 }, { "epoch": 0.39149219009637753, "grad_norm": 0.3555982790161214, "learning_rate": 6.950680078836475e-06, "loss": 0.012, "step": 589 }, { "epoch": 0.39215686274509803, "grad_norm": 0.17594418785497495, "learning_rate": 6.940755653591859e-06, "loss": 0.0054, "step": 590 }, { "epoch": 0.39282153539381853, "grad_norm": 0.6987081904211158, "learning_rate": 6.930822217712418e-06, "loss": 0.0108, "step": 591 }, { "epoch": 0.39348620804253903, "grad_norm": 0.6486921438985525, "learning_rate": 6.920879817317588e-06, "loss": 0.0113, "step": 592 }, { "epoch": 0.39415088069125953, "grad_norm": 0.27284700713735915, "learning_rate": 6.910928498568435e-06, "loss": 0.0084, "step": 593 }, { "epoch": 0.3948155533399801, "grad_norm": 0.25769721114521166, "learning_rate": 6.900968307667424e-06, "loss": 0.0104, "step": 594 }, { "epoch": 0.3954802259887006, "grad_norm": 0.3299261578423499, "learning_rate": 6.890999290858213e-06, "loss": 0.0054, "step": 595 }, { "epoch": 0.3961448986374211, "grad_norm": 0.43910099238089556, "learning_rate": 6.88102149442544e-06, "loss": 0.0106, "step": 596 }, { "epoch": 0.3968095712861416, "grad_norm": 0.338574725547578, "learning_rate": 6.871034964694504e-06, "loss": 0.0099, "step": 597 }, { "epoch": 0.3974742439348621, "grad_norm": 0.26491883961369567, "learning_rate": 6.861039748031351e-06, "loss": 0.0108, "step": 598 }, { "epoch": 0.3981389165835826, "grad_norm": 0.21317555102009628, "learning_rate": 6.85103589084226e-06, "loss": 0.0072, "step": 599 }, { "epoch": 0.3988035892323031, "grad_norm": 0.32544598623336923, "learning_rate": 6.841023439573623e-06, "loss": 0.0088, "step": 600 }, { "epoch": 0.3994682618810236, "grad_norm": 0.45217378967864114, "learning_rate": 6.8310024407117405e-06, "loss": 0.0091, "step": 601 }, { "epoch": 0.4001329345297441, "grad_norm": 0.22695735797763986, "learning_rate": 6.820972940782592e-06, "loss": 0.0076, "step": 602 }, { "epoch": 0.40079760717846463, "grad_norm": 0.4553197085999024, "learning_rate": 6.810934986351629e-06, "loss": 0.0109, "step": 603 }, { "epoch": 0.40146227982718513, "grad_norm": 0.26812600555564403, "learning_rate": 6.800888624023552e-06, "loss": 0.0103, "step": 604 }, { "epoch": 0.40212695247590563, "grad_norm": 0.7024527429206964, "learning_rate": 6.7908339004421055e-06, "loss": 0.0129, "step": 605 }, { "epoch": 0.40279162512462613, "grad_norm": 0.34999655159427806, "learning_rate": 6.780770862289846e-06, "loss": 0.0092, "step": 606 }, { "epoch": 0.40345629777334663, "grad_norm": 0.45285032332411074, "learning_rate": 6.770699556287939e-06, "loss": 0.0088, "step": 607 }, { "epoch": 0.4041209704220671, "grad_norm": 0.5997018707452677, "learning_rate": 6.760620029195932e-06, "loss": 0.0096, "step": 608 }, { "epoch": 0.4047856430707876, "grad_norm": 0.5514074896434142, "learning_rate": 6.750532327811547e-06, "loss": 0.011, "step": 609 }, { "epoch": 0.4054503157195081, "grad_norm": 0.37176586119207755, "learning_rate": 6.740436498970453e-06, "loss": 0.0084, "step": 610 }, { "epoch": 0.4061149883682286, "grad_norm": 0.27177979114541556, "learning_rate": 6.730332589546057e-06, "loss": 0.0114, "step": 611 }, { "epoch": 0.4067796610169492, "grad_norm": 0.3155915334093962, "learning_rate": 6.72022064644928e-06, "loss": 0.0097, "step": 612 }, { "epoch": 0.4074443336656697, "grad_norm": 0.3630978188083772, "learning_rate": 6.710100716628345e-06, "loss": 0.0131, "step": 613 }, { "epoch": 0.4081090063143902, "grad_norm": 0.18765224364350858, "learning_rate": 6.699972847068554e-06, "loss": 0.0087, "step": 614 }, { "epoch": 0.4087736789631107, "grad_norm": 0.18599228269433404, "learning_rate": 6.6898370847920725e-06, "loss": 0.0079, "step": 615 }, { "epoch": 0.4094383516118312, "grad_norm": 0.20485696206177575, "learning_rate": 6.679693476857712e-06, "loss": 0.008, "step": 616 }, { "epoch": 0.4101030242605517, "grad_norm": 0.31466782644228947, "learning_rate": 6.669542070360709e-06, "loss": 0.0097, "step": 617 }, { "epoch": 0.4107676969092722, "grad_norm": 0.4686745434154618, "learning_rate": 6.659382912432506e-06, "loss": 0.0089, "step": 618 }, { "epoch": 0.41143236955799267, "grad_norm": 0.24238323375836263, "learning_rate": 6.649216050240539e-06, "loss": 0.0075, "step": 619 }, { "epoch": 0.41209704220671317, "grad_norm": 0.2670812329975021, "learning_rate": 6.639041530988009e-06, "loss": 0.0083, "step": 620 }, { "epoch": 0.4127617148554337, "grad_norm": 0.23529371570929705, "learning_rate": 6.62885940191367e-06, "loss": 0.0128, "step": 621 }, { "epoch": 0.4134263875041542, "grad_norm": 0.5824583044607933, "learning_rate": 6.618669710291607e-06, "loss": 0.0119, "step": 622 }, { "epoch": 0.4140910601528747, "grad_norm": 0.4446092567307767, "learning_rate": 6.608472503431013e-06, "loss": 0.0126, "step": 623 }, { "epoch": 0.4147557328015952, "grad_norm": 0.41347103785645517, "learning_rate": 6.5982678286759795e-06, "loss": 0.0103, "step": 624 }, { "epoch": 0.4154204054503157, "grad_norm": 0.3338845213578862, "learning_rate": 6.588055733405266e-06, "loss": 0.0097, "step": 625 }, { "epoch": 0.4160850780990362, "grad_norm": 0.36627748338652033, "learning_rate": 6.577836265032088e-06, "loss": 0.0071, "step": 626 }, { "epoch": 0.4167497507477567, "grad_norm": 0.3398309868685469, "learning_rate": 6.5676094710038885e-06, "loss": 0.0063, "step": 627 }, { "epoch": 0.4174144233964772, "grad_norm": 0.2028781075400023, "learning_rate": 6.557375398802124e-06, "loss": 0.0103, "step": 628 }, { "epoch": 0.4180790960451977, "grad_norm": 0.33930270206275065, "learning_rate": 6.547134095942045e-06, "loss": 0.009, "step": 629 }, { "epoch": 0.4187437686939183, "grad_norm": 0.2354065061956709, "learning_rate": 6.536885609972467e-06, "loss": 0.0097, "step": 630 }, { "epoch": 0.41940844134263877, "grad_norm": 0.3133176449040743, "learning_rate": 6.526629988475567e-06, "loss": 0.0113, "step": 631 }, { "epoch": 0.42007311399135927, "grad_norm": 0.33078518943415103, "learning_rate": 6.516367279066637e-06, "loss": 0.0119, "step": 632 }, { "epoch": 0.42073778664007977, "grad_norm": 0.38893906939213974, "learning_rate": 6.5060975293938865e-06, "loss": 0.0143, "step": 633 }, { "epoch": 0.42140245928880027, "grad_norm": 0.2547718557795754, "learning_rate": 6.495820787138209e-06, "loss": 0.0097, "step": 634 }, { "epoch": 0.42206713193752077, "grad_norm": 0.6357167916060998, "learning_rate": 6.485537100012964e-06, "loss": 0.0136, "step": 635 }, { "epoch": 0.42273180458624127, "grad_norm": 0.24172901416385553, "learning_rate": 6.475246515763753e-06, "loss": 0.0089, "step": 636 }, { "epoch": 0.42339647723496177, "grad_norm": 0.42791960967947706, "learning_rate": 6.4649490821682035e-06, "loss": 0.0103, "step": 637 }, { "epoch": 0.42406114988368226, "grad_norm": 0.12542724400280028, "learning_rate": 6.454644847035741e-06, "loss": 0.0056, "step": 638 }, { "epoch": 0.4247258225324028, "grad_norm": 0.3240087571703041, "learning_rate": 6.44433385820737e-06, "loss": 0.0136, "step": 639 }, { "epoch": 0.4253904951811233, "grad_norm": 0.4831422023803344, "learning_rate": 6.434016163555452e-06, "loss": 0.0126, "step": 640 }, { "epoch": 0.4260551678298438, "grad_norm": 0.26404807622144777, "learning_rate": 6.42369181098348e-06, "loss": 0.0103, "step": 641 }, { "epoch": 0.4267198404785643, "grad_norm": 0.7106358570106606, "learning_rate": 6.4133608484258625e-06, "loss": 0.0128, "step": 642 }, { "epoch": 0.4273845131272848, "grad_norm": 0.1909189463837318, "learning_rate": 6.403023323847695e-06, "loss": 0.0118, "step": 643 }, { "epoch": 0.4280491857760053, "grad_norm": 0.16305714417406786, "learning_rate": 6.392679285244538e-06, "loss": 0.0059, "step": 644 }, { "epoch": 0.4287138584247258, "grad_norm": 0.20980001282845573, "learning_rate": 6.382328780642202e-06, "loss": 0.0065, "step": 645 }, { "epoch": 0.4293785310734463, "grad_norm": 0.6345858900603542, "learning_rate": 6.371971858096509e-06, "loss": 0.0101, "step": 646 }, { "epoch": 0.4300432037221668, "grad_norm": 0.2526581898487362, "learning_rate": 6.3616085656930845e-06, "loss": 0.0084, "step": 647 }, { "epoch": 0.4307078763708873, "grad_norm": 0.5563764968794693, "learning_rate": 6.351238951547127e-06, "loss": 0.0093, "step": 648 }, { "epoch": 0.43137254901960786, "grad_norm": 0.3280061671425709, "learning_rate": 6.340863063803187e-06, "loss": 0.0102, "step": 649 }, { "epoch": 0.43203722166832836, "grad_norm": 0.29114680990388586, "learning_rate": 6.330480950634942e-06, "loss": 0.0071, "step": 650 }, { "epoch": 0.43270189431704886, "grad_norm": 0.7540844145468062, "learning_rate": 6.320092660244969e-06, "loss": 0.0118, "step": 651 }, { "epoch": 0.43336656696576936, "grad_norm": 0.7129618909686243, "learning_rate": 6.30969824086453e-06, "loss": 0.014, "step": 652 }, { "epoch": 0.43403123961448986, "grad_norm": 0.37761044935808713, "learning_rate": 6.2992977407533425e-06, "loss": 0.0147, "step": 653 }, { "epoch": 0.43469591226321036, "grad_norm": 0.6235565064106601, "learning_rate": 6.288891208199354e-06, "loss": 0.0143, "step": 654 }, { "epoch": 0.43536058491193086, "grad_norm": 0.4853294209376415, "learning_rate": 6.278478691518519e-06, "loss": 0.0071, "step": 655 }, { "epoch": 0.43602525756065136, "grad_norm": 0.3809871361215779, "learning_rate": 6.268060239054577e-06, "loss": 0.011, "step": 656 }, { "epoch": 0.43668993020937186, "grad_norm": 0.6387386149948802, "learning_rate": 6.25763589917883e-06, "loss": 0.0143, "step": 657 }, { "epoch": 0.4373546028580924, "grad_norm": 0.4015521779761784, "learning_rate": 6.247205720289907e-06, "loss": 0.014, "step": 658 }, { "epoch": 0.4380192755068129, "grad_norm": 0.4117265254888033, "learning_rate": 6.2367697508135505e-06, "loss": 0.0094, "step": 659 }, { "epoch": 0.4386839481555334, "grad_norm": 0.2308963938069441, "learning_rate": 6.22632803920239e-06, "loss": 0.0106, "step": 660 }, { "epoch": 0.4393486208042539, "grad_norm": 0.5565667568037039, "learning_rate": 6.215880633935709e-06, "loss": 0.0119, "step": 661 }, { "epoch": 0.4400132934529744, "grad_norm": 0.13383004160026224, "learning_rate": 6.205427583519233e-06, "loss": 0.0082, "step": 662 }, { "epoch": 0.4406779661016949, "grad_norm": 0.12559365714009277, "learning_rate": 6.19496893648489e-06, "loss": 0.0057, "step": 663 }, { "epoch": 0.4413426387504154, "grad_norm": 0.24802201422675454, "learning_rate": 6.184504741390596e-06, "loss": 0.0124, "step": 664 }, { "epoch": 0.4420073113991359, "grad_norm": 0.35523200372156655, "learning_rate": 6.174035046820029e-06, "loss": 0.0138, "step": 665 }, { "epoch": 0.4426719840478564, "grad_norm": 0.6446429153918144, "learning_rate": 6.163559901382391e-06, "loss": 0.0101, "step": 666 }, { "epoch": 0.44333665669657696, "grad_norm": 0.19204760848336078, "learning_rate": 6.153079353712201e-06, "loss": 0.0084, "step": 667 }, { "epoch": 0.44400132934529746, "grad_norm": 0.3606528722422487, "learning_rate": 6.142593452469057e-06, "loss": 0.0112, "step": 668 }, { "epoch": 0.44466600199401796, "grad_norm": 0.32923458085729596, "learning_rate": 6.132102246337408e-06, "loss": 0.011, "step": 669 }, { "epoch": 0.44533067464273846, "grad_norm": 0.2498741139091563, "learning_rate": 6.121605784026339e-06, "loss": 0.0087, "step": 670 }, { "epoch": 0.44599534729145895, "grad_norm": 0.2294721012914431, "learning_rate": 6.1111041142693354e-06, "loss": 0.0056, "step": 671 }, { "epoch": 0.44666001994017945, "grad_norm": 0.34052806371715416, "learning_rate": 6.100597285824061e-06, "loss": 0.0098, "step": 672 }, { "epoch": 0.44732469258889995, "grad_norm": 0.317963671335907, "learning_rate": 6.09008534747213e-06, "loss": 0.0082, "step": 673 }, { "epoch": 0.44798936523762045, "grad_norm": 0.4296825991195016, "learning_rate": 6.079568348018881e-06, "loss": 0.0071, "step": 674 }, { "epoch": 0.44865403788634095, "grad_norm": 0.47778547270400257, "learning_rate": 6.069046336293154e-06, "loss": 0.0093, "step": 675 }, { "epoch": 0.4493187105350615, "grad_norm": 0.5777658191227268, "learning_rate": 6.058519361147055e-06, "loss": 0.0086, "step": 676 }, { "epoch": 0.449983383183782, "grad_norm": 0.3622945985112122, "learning_rate": 6.047987471455735e-06, "loss": 0.0122, "step": 677 }, { "epoch": 0.4506480558325025, "grad_norm": 0.3245378324124243, "learning_rate": 6.037450716117167e-06, "loss": 0.0117, "step": 678 }, { "epoch": 0.451312728481223, "grad_norm": 0.457138027692893, "learning_rate": 6.02690914405191e-06, "loss": 0.0094, "step": 679 }, { "epoch": 0.4519774011299435, "grad_norm": 0.9939712798598156, "learning_rate": 6.01636280420289e-06, "loss": 0.0116, "step": 680 }, { "epoch": 0.452642073778664, "grad_norm": 0.9022525191572159, "learning_rate": 6.005811745535165e-06, "loss": 0.0167, "step": 681 }, { "epoch": 0.4533067464273845, "grad_norm": 0.7518128688850892, "learning_rate": 5.995256017035703e-06, "loss": 0.01, "step": 682 }, { "epoch": 0.453971419076105, "grad_norm": 0.290281391886538, "learning_rate": 5.984695667713157e-06, "loss": 0.0123, "step": 683 }, { "epoch": 0.4546360917248255, "grad_norm": 0.3062966481904787, "learning_rate": 5.974130746597628e-06, "loss": 0.0103, "step": 684 }, { "epoch": 0.45530076437354605, "grad_norm": 0.6339903493097341, "learning_rate": 5.9635613027404495e-06, "loss": 0.0087, "step": 685 }, { "epoch": 0.45596543702226655, "grad_norm": 0.7676221063606729, "learning_rate": 5.952987385213948e-06, "loss": 0.0113, "step": 686 }, { "epoch": 0.45663010967098705, "grad_norm": 0.8254164344197613, "learning_rate": 5.9424090431112204e-06, "loss": 0.018, "step": 687 }, { "epoch": 0.45729478231970755, "grad_norm": 0.41163678118289565, "learning_rate": 5.931826325545912e-06, "loss": 0.0103, "step": 688 }, { "epoch": 0.45795945496842805, "grad_norm": 0.2689991907397002, "learning_rate": 5.921239281651976e-06, "loss": 0.0101, "step": 689 }, { "epoch": 0.45862412761714855, "grad_norm": 0.14116303639980618, "learning_rate": 5.910647960583458e-06, "loss": 0.0089, "step": 690 }, { "epoch": 0.45928880026586905, "grad_norm": 0.21552921395180238, "learning_rate": 5.900052411514257e-06, "loss": 0.0103, "step": 691 }, { "epoch": 0.45995347291458955, "grad_norm": 0.7311396613947106, "learning_rate": 5.889452683637906e-06, "loss": 0.0124, "step": 692 }, { "epoch": 0.46061814556331004, "grad_norm": 0.692389828796664, "learning_rate": 5.878848826167335e-06, "loss": 0.0094, "step": 693 }, { "epoch": 0.4612828182120306, "grad_norm": 1.083315355203257, "learning_rate": 5.8682408883346535e-06, "loss": 0.0132, "step": 694 }, { "epoch": 0.4619474908607511, "grad_norm": 0.5552847101191409, "learning_rate": 5.8576289193909066e-06, "loss": 0.0093, "step": 695 }, { "epoch": 0.4626121635094716, "grad_norm": 0.25809558813205624, "learning_rate": 5.847012968605864e-06, "loss": 0.0064, "step": 696 }, { "epoch": 0.4632768361581921, "grad_norm": 0.4484954970574295, "learning_rate": 5.836393085267777e-06, "loss": 0.0107, "step": 697 }, { "epoch": 0.4639415088069126, "grad_norm": 0.3748808259569666, "learning_rate": 5.825769318683158e-06, "loss": 0.0137, "step": 698 }, { "epoch": 0.4646061814556331, "grad_norm": 0.5135640958070253, "learning_rate": 5.815141718176549e-06, "loss": 0.0114, "step": 699 }, { "epoch": 0.4652708541043536, "grad_norm": 0.6767031423114493, "learning_rate": 5.804510333090287e-06, "loss": 0.0094, "step": 700 }, { "epoch": 0.4659355267530741, "grad_norm": 0.5821188316816928, "learning_rate": 5.793875212784288e-06, "loss": 0.0113, "step": 701 }, { "epoch": 0.4666001994017946, "grad_norm": 0.4409806468106221, "learning_rate": 5.783236406635803e-06, "loss": 0.0125, "step": 702 }, { "epoch": 0.46726487205051515, "grad_norm": 0.21555953024189517, "learning_rate": 5.772593964039203e-06, "loss": 0.0074, "step": 703 }, { "epoch": 0.46792954469923564, "grad_norm": 0.5228984214582533, "learning_rate": 5.761947934405736e-06, "loss": 0.0109, "step": 704 }, { "epoch": 0.46859421734795614, "grad_norm": 0.4948415126737405, "learning_rate": 5.751298367163307e-06, "loss": 0.0093, "step": 705 }, { "epoch": 0.46925888999667664, "grad_norm": 0.3688000672546482, "learning_rate": 5.740645311756246e-06, "loss": 0.0083, "step": 706 }, { "epoch": 0.46992356264539714, "grad_norm": 0.5698623206821869, "learning_rate": 5.729988817645075e-06, "loss": 0.0108, "step": 707 }, { "epoch": 0.47058823529411764, "grad_norm": 0.21618485076241153, "learning_rate": 5.719328934306286e-06, "loss": 0.0096, "step": 708 }, { "epoch": 0.47125290794283814, "grad_norm": 0.18095593599382673, "learning_rate": 5.708665711232103e-06, "loss": 0.0078, "step": 709 }, { "epoch": 0.47191758059155864, "grad_norm": 0.23192973151623708, "learning_rate": 5.697999197930259e-06, "loss": 0.0105, "step": 710 }, { "epoch": 0.47258225324027914, "grad_norm": 0.2176232510829752, "learning_rate": 5.687329443923758e-06, "loss": 0.0106, "step": 711 }, { "epoch": 0.4732469258889997, "grad_norm": 0.3782269310072872, "learning_rate": 5.6766564987506564e-06, "loss": 0.0117, "step": 712 }, { "epoch": 0.4739115985377202, "grad_norm": 0.23650414365163921, "learning_rate": 5.665980411963822e-06, "loss": 0.0098, "step": 713 }, { "epoch": 0.4745762711864407, "grad_norm": 0.3560109224673454, "learning_rate": 5.655301233130711e-06, "loss": 0.0098, "step": 714 }, { "epoch": 0.4752409438351612, "grad_norm": 0.1930184705151546, "learning_rate": 5.644619011833134e-06, "loss": 0.0069, "step": 715 }, { "epoch": 0.4759056164838817, "grad_norm": 0.27914365709206773, "learning_rate": 5.633933797667028e-06, "loss": 0.0093, "step": 716 }, { "epoch": 0.4765702891326022, "grad_norm": 0.18875523840785524, "learning_rate": 5.623245640242228e-06, "loss": 0.0086, "step": 717 }, { "epoch": 0.4772349617813227, "grad_norm": 0.39399194099722185, "learning_rate": 5.612554589182228e-06, "loss": 0.0098, "step": 718 }, { "epoch": 0.4778996344300432, "grad_norm": 0.3149239229882813, "learning_rate": 5.601860694123962e-06, "loss": 0.0105, "step": 719 }, { "epoch": 0.4785643070787637, "grad_norm": 0.422704004266628, "learning_rate": 5.591164004717567e-06, "loss": 0.0087, "step": 720 }, { "epoch": 0.47922897972748424, "grad_norm": 0.3602805146680993, "learning_rate": 5.5804645706261515e-06, "loss": 0.009, "step": 721 }, { "epoch": 0.47989365237620474, "grad_norm": 0.291487301780777, "learning_rate": 5.569762441525572e-06, "loss": 0.0103, "step": 722 }, { "epoch": 0.48055832502492524, "grad_norm": 0.5282605238528107, "learning_rate": 5.559057667104191e-06, "loss": 0.0127, "step": 723 }, { "epoch": 0.48122299767364574, "grad_norm": 0.5788839886843503, "learning_rate": 5.548350297062659e-06, "loss": 0.0101, "step": 724 }, { "epoch": 0.48188767032236623, "grad_norm": 0.36809665261771013, "learning_rate": 5.537640381113673e-06, "loss": 0.0096, "step": 725 }, { "epoch": 0.48255234297108673, "grad_norm": 0.25887008834225905, "learning_rate": 5.526927968981754e-06, "loss": 0.0102, "step": 726 }, { "epoch": 0.48321701561980723, "grad_norm": 0.21720892588117346, "learning_rate": 5.516213110403009e-06, "loss": 0.0077, "step": 727 }, { "epoch": 0.48388168826852773, "grad_norm": 0.40849716443647854, "learning_rate": 5.505495855124905e-06, "loss": 0.0098, "step": 728 }, { "epoch": 0.48454636091724823, "grad_norm": 0.6027373145628082, "learning_rate": 5.4947762529060366e-06, "loss": 0.0116, "step": 729 }, { "epoch": 0.4852110335659688, "grad_norm": 0.4877466028567149, "learning_rate": 5.484054353515896e-06, "loss": 0.0099, "step": 730 }, { "epoch": 0.4858757062146893, "grad_norm": 0.38762455194794565, "learning_rate": 5.473330206734639e-06, "loss": 0.0083, "step": 731 }, { "epoch": 0.4865403788634098, "grad_norm": 0.20656803243125438, "learning_rate": 5.462603862352857e-06, "loss": 0.0088, "step": 732 }, { "epoch": 0.4872050515121303, "grad_norm": 0.19952617294361694, "learning_rate": 5.451875370171341e-06, "loss": 0.008, "step": 733 }, { "epoch": 0.4878697241608508, "grad_norm": 0.32916661603970343, "learning_rate": 5.44114478000086e-06, "loss": 0.0104, "step": 734 }, { "epoch": 0.4885343968095713, "grad_norm": 0.23315426449518953, "learning_rate": 5.43041214166192e-06, "loss": 0.0068, "step": 735 }, { "epoch": 0.4891990694582918, "grad_norm": 0.35075782360321667, "learning_rate": 5.419677504984534e-06, "loss": 0.0069, "step": 736 }, { "epoch": 0.4898637421070123, "grad_norm": 0.6404544044244168, "learning_rate": 5.408940919808e-06, "loss": 0.0101, "step": 737 }, { "epoch": 0.4905284147557328, "grad_norm": 0.5638294021557101, "learning_rate": 5.398202435980654e-06, "loss": 0.0102, "step": 738 }, { "epoch": 0.49119308740445333, "grad_norm": 0.24508390663968715, "learning_rate": 5.387462103359655e-06, "loss": 0.0112, "step": 739 }, { "epoch": 0.49185776005317383, "grad_norm": 0.27376956453984214, "learning_rate": 5.376719971810741e-06, "loss": 0.0104, "step": 740 }, { "epoch": 0.49252243270189433, "grad_norm": 0.6463299074535875, "learning_rate": 5.365976091208001e-06, "loss": 0.0097, "step": 741 }, { "epoch": 0.49318710535061483, "grad_norm": 0.5118177089324828, "learning_rate": 5.3552305114336515e-06, "loss": 0.0095, "step": 742 }, { "epoch": 0.49385177799933533, "grad_norm": 0.3264448188879537, "learning_rate": 5.344483282377788e-06, "loss": 0.0093, "step": 743 }, { "epoch": 0.4945164506480558, "grad_norm": 0.23644388833868601, "learning_rate": 5.333734453938174e-06, "loss": 0.0094, "step": 744 }, { "epoch": 0.4951811232967763, "grad_norm": 0.24025823089832044, "learning_rate": 5.32298407601999e-06, "loss": 0.0123, "step": 745 }, { "epoch": 0.4958457959454968, "grad_norm": 0.4402862063583305, "learning_rate": 5.312232198535614e-06, "loss": 0.007, "step": 746 }, { "epoch": 0.4965104685942173, "grad_norm": 0.3442328678549942, "learning_rate": 5.301478871404386e-06, "loss": 0.0074, "step": 747 }, { "epoch": 0.4971751412429379, "grad_norm": 0.6823323853092952, "learning_rate": 5.290724144552379e-06, "loss": 0.0138, "step": 748 }, { "epoch": 0.4978398138916584, "grad_norm": 0.5745037445778555, "learning_rate": 5.279968067912161e-06, "loss": 0.0138, "step": 749 }, { "epoch": 0.4985044865403789, "grad_norm": 0.2559359732479916, "learning_rate": 5.26921069142257e-06, "loss": 0.0109, "step": 750 }, { "epoch": 0.4991691591890994, "grad_norm": 0.3646088462874537, "learning_rate": 5.258452065028473e-06, "loss": 0.0097, "step": 751 }, { "epoch": 0.4998338318378199, "grad_norm": 0.2651599568983938, "learning_rate": 5.247692238680548e-06, "loss": 0.0073, "step": 752 }, { "epoch": 0.5004985044865404, "grad_norm": 0.4018366742171501, "learning_rate": 5.2369312623350395e-06, "loss": 0.0139, "step": 753 }, { "epoch": 0.5011631771352609, "grad_norm": 0.5344590495873953, "learning_rate": 5.2261691859535325e-06, "loss": 0.0105, "step": 754 }, { "epoch": 0.5018278497839814, "grad_norm": 0.3639429315478386, "learning_rate": 5.215406059502719e-06, "loss": 0.0063, "step": 755 }, { "epoch": 0.5024925224327019, "grad_norm": 0.24619091005080893, "learning_rate": 5.204641932954165e-06, "loss": 0.0078, "step": 756 }, { "epoch": 0.5031571950814224, "grad_norm": 1.7963343305256685, "learning_rate": 5.193876856284085e-06, "loss": 0.0109, "step": 757 }, { "epoch": 0.5038218677301429, "grad_norm": 0.2440032264782902, "learning_rate": 5.1831108794730964e-06, "loss": 0.0089, "step": 758 }, { "epoch": 0.5044865403788634, "grad_norm": 0.44972524967383337, "learning_rate": 5.172344052506003e-06, "loss": 0.0087, "step": 759 }, { "epoch": 0.5051512130275839, "grad_norm": 0.26245708880263857, "learning_rate": 5.161576425371554e-06, "loss": 0.0053, "step": 760 }, { "epoch": 0.5058158856763044, "grad_norm": 0.2978266650836691, "learning_rate": 5.150808048062211e-06, "loss": 0.0099, "step": 761 }, { "epoch": 0.5064805583250249, "grad_norm": 0.7291062229944648, "learning_rate": 5.14003897057392e-06, "loss": 0.0104, "step": 762 }, { "epoch": 0.5071452309737454, "grad_norm": 0.6453933717038971, "learning_rate": 5.1292692429058824e-06, "loss": 0.0098, "step": 763 }, { "epoch": 0.5078099036224659, "grad_norm": 0.5807828770462001, "learning_rate": 5.118498915060307e-06, "loss": 0.013, "step": 764 }, { "epoch": 0.5084745762711864, "grad_norm": 0.5360840353535625, "learning_rate": 5.107728037042204e-06, "loss": 0.0098, "step": 765 }, { "epoch": 0.5091392489199069, "grad_norm": 0.47334293259911675, "learning_rate": 5.096956658859122e-06, "loss": 0.014, "step": 766 }, { "epoch": 0.5098039215686274, "grad_norm": 0.1687174939000583, "learning_rate": 5.086184830520947e-06, "loss": 0.0063, "step": 767 }, { "epoch": 0.5104685942173479, "grad_norm": 0.7049415612865658, "learning_rate": 5.075412602039646e-06, "loss": 0.0114, "step": 768 }, { "epoch": 0.5111332668660684, "grad_norm": 0.4526398312602828, "learning_rate": 5.064640023429042e-06, "loss": 0.0107, "step": 769 }, { "epoch": 0.5117979395147889, "grad_norm": 0.6680987216631, "learning_rate": 5.053867144704594e-06, "loss": 0.0165, "step": 770 }, { "epoch": 0.5124626121635095, "grad_norm": 0.4111396230491692, "learning_rate": 5.043094015883143e-06, "loss": 0.017, "step": 771 }, { "epoch": 0.51312728481223, "grad_norm": 0.28301907133703696, "learning_rate": 5.032320686982697e-06, "loss": 0.008, "step": 772 }, { "epoch": 0.5137919574609505, "grad_norm": 0.30335498287275575, "learning_rate": 5.0215472080221925e-06, "loss": 0.0099, "step": 773 }, { "epoch": 0.514456630109671, "grad_norm": 0.1837083789004642, "learning_rate": 5.010773629021261e-06, "loss": 0.0056, "step": 774 }, { "epoch": 0.5151213027583915, "grad_norm": 0.16231366475320347, "learning_rate": 5e-06, "loss": 0.0075, "step": 775 }, { "epoch": 0.515785975407112, "grad_norm": 0.2916140148532633, "learning_rate": 4.989226370978741e-06, "loss": 0.0118, "step": 776 }, { "epoch": 0.5164506480558325, "grad_norm": 0.20626278252128077, "learning_rate": 4.97845279197781e-06, "loss": 0.0077, "step": 777 }, { "epoch": 0.517115320704553, "grad_norm": 0.21583759128470975, "learning_rate": 4.967679313017304e-06, "loss": 0.0121, "step": 778 }, { "epoch": 0.5177799933532735, "grad_norm": 0.46444971246093797, "learning_rate": 4.956905984116858e-06, "loss": 0.0103, "step": 779 }, { "epoch": 0.518444666001994, "grad_norm": 0.31507275281451314, "learning_rate": 4.946132855295407e-06, "loss": 0.0074, "step": 780 }, { "epoch": 0.5191093386507145, "grad_norm": 0.27082829681822324, "learning_rate": 4.9353599765709585e-06, "loss": 0.0096, "step": 781 }, { "epoch": 0.519774011299435, "grad_norm": 0.37005273845436243, "learning_rate": 4.924587397960357e-06, "loss": 0.0095, "step": 782 }, { "epoch": 0.5204386839481555, "grad_norm": 0.2643648982714299, "learning_rate": 4.913815169479054e-06, "loss": 0.0096, "step": 783 }, { "epoch": 0.521103356596876, "grad_norm": 0.25635014295235947, "learning_rate": 4.903043341140879e-06, "loss": 0.0081, "step": 784 }, { "epoch": 0.5217680292455965, "grad_norm": 0.4254240080114205, "learning_rate": 4.892271962957799e-06, "loss": 0.01, "step": 785 }, { "epoch": 0.522432701894317, "grad_norm": 0.9095957538501734, "learning_rate": 4.881501084939694e-06, "loss": 0.0126, "step": 786 }, { "epoch": 0.5230973745430375, "grad_norm": 0.18646865925644313, "learning_rate": 4.870730757094121e-06, "loss": 0.0076, "step": 787 }, { "epoch": 0.523762047191758, "grad_norm": 0.6272282856868566, "learning_rate": 4.8599610294260805e-06, "loss": 0.0094, "step": 788 }, { "epoch": 0.5244267198404786, "grad_norm": 0.3216322367525329, "learning_rate": 4.84919195193779e-06, "loss": 0.0105, "step": 789 }, { "epoch": 0.5250913924891991, "grad_norm": 0.5085096496860171, "learning_rate": 4.838423574628447e-06, "loss": 0.0101, "step": 790 }, { "epoch": 0.5257560651379196, "grad_norm": 0.5105382876121034, "learning_rate": 4.827655947493998e-06, "loss": 0.0094, "step": 791 }, { "epoch": 0.5264207377866401, "grad_norm": 0.3103491820455327, "learning_rate": 4.816889120526905e-06, "loss": 0.0102, "step": 792 }, { "epoch": 0.5270854104353606, "grad_norm": 0.2997496633274557, "learning_rate": 4.806123143715916e-06, "loss": 0.0096, "step": 793 }, { "epoch": 0.5277500830840811, "grad_norm": 0.45390548580535, "learning_rate": 4.795358067045835e-06, "loss": 0.0114, "step": 794 }, { "epoch": 0.5284147557328016, "grad_norm": 0.27942540850784636, "learning_rate": 4.784593940497282e-06, "loss": 0.0092, "step": 795 }, { "epoch": 0.5290794283815221, "grad_norm": 0.2594244211956754, "learning_rate": 4.773830814046469e-06, "loss": 0.0092, "step": 796 }, { "epoch": 0.5297441010302426, "grad_norm": 0.493600238510158, "learning_rate": 4.763068737664963e-06, "loss": 0.0116, "step": 797 }, { "epoch": 0.5304087736789631, "grad_norm": 0.14975954490405455, "learning_rate": 4.752307761319454e-06, "loss": 0.0084, "step": 798 }, { "epoch": 0.5310734463276836, "grad_norm": 0.19464500105989566, "learning_rate": 4.741547934971528e-06, "loss": 0.0096, "step": 799 }, { "epoch": 0.5317381189764041, "grad_norm": 0.46023091211688155, "learning_rate": 4.730789308577432e-06, "loss": 0.0104, "step": 800 }, { "epoch": 0.5324027916251246, "grad_norm": 0.2243571958019791, "learning_rate": 4.72003193208784e-06, "loss": 0.009, "step": 801 }, { "epoch": 0.5330674642738451, "grad_norm": 0.21459986565174058, "learning_rate": 4.7092758554476215e-06, "loss": 0.0077, "step": 802 }, { "epoch": 0.5337321369225656, "grad_norm": 0.24942831647994296, "learning_rate": 4.698521128595615e-06, "loss": 0.0098, "step": 803 }, { "epoch": 0.5343968095712861, "grad_norm": 0.33928927103917483, "learning_rate": 4.687767801464388e-06, "loss": 0.0095, "step": 804 }, { "epoch": 0.5350614822200066, "grad_norm": 0.23489720803376424, "learning_rate": 4.677015923980012e-06, "loss": 0.0122, "step": 805 }, { "epoch": 0.5357261548687271, "grad_norm": 0.2761460164751238, "learning_rate": 4.666265546061828e-06, "loss": 0.0102, "step": 806 }, { "epoch": 0.5363908275174477, "grad_norm": 0.4146470939529886, "learning_rate": 4.655516717622213e-06, "loss": 0.0126, "step": 807 }, { "epoch": 0.5370555001661682, "grad_norm": 0.15653134850885483, "learning_rate": 4.644769488566351e-06, "loss": 0.0068, "step": 808 }, { "epoch": 0.5377201728148887, "grad_norm": 0.32644930715201503, "learning_rate": 4.634023908791999e-06, "loss": 0.0124, "step": 809 }, { "epoch": 0.5383848454636092, "grad_norm": 0.3747837759026735, "learning_rate": 4.6232800281892604e-06, "loss": 0.013, "step": 810 }, { "epoch": 0.5390495181123297, "grad_norm": 0.2632600574425649, "learning_rate": 4.6125378966403465e-06, "loss": 0.0075, "step": 811 }, { "epoch": 0.5397141907610502, "grad_norm": 0.32145286934292666, "learning_rate": 4.601797564019347e-06, "loss": 0.0096, "step": 812 }, { "epoch": 0.5403788634097707, "grad_norm": 0.3007889606771815, "learning_rate": 4.591059080192003e-06, "loss": 0.0116, "step": 813 }, { "epoch": 0.5410435360584912, "grad_norm": 0.7501761137235431, "learning_rate": 4.580322495015466e-06, "loss": 0.0103, "step": 814 }, { "epoch": 0.5417082087072117, "grad_norm": 0.27305810384069584, "learning_rate": 4.569587858338081e-06, "loss": 0.0109, "step": 815 }, { "epoch": 0.5423728813559322, "grad_norm": 0.30907536215366266, "learning_rate": 4.558855219999141e-06, "loss": 0.0085, "step": 816 }, { "epoch": 0.5430375540046527, "grad_norm": 0.38238198045306393, "learning_rate": 4.548124629828661e-06, "loss": 0.0079, "step": 817 }, { "epoch": 0.5437022266533732, "grad_norm": 0.23247098587430676, "learning_rate": 4.5373961376471465e-06, "loss": 0.0086, "step": 818 }, { "epoch": 0.5443668993020937, "grad_norm": 0.19122719511288216, "learning_rate": 4.526669793265362e-06, "loss": 0.0079, "step": 819 }, { "epoch": 0.5450315719508142, "grad_norm": 0.42214690182176545, "learning_rate": 4.515945646484105e-06, "loss": 0.0128, "step": 820 }, { "epoch": 0.5456962445995347, "grad_norm": 0.2173423757608228, "learning_rate": 4.505223747093965e-06, "loss": 0.0079, "step": 821 }, { "epoch": 0.5463609172482552, "grad_norm": 0.478198989356843, "learning_rate": 4.494504144875097e-06, "loss": 0.0101, "step": 822 }, { "epoch": 0.5470255898969757, "grad_norm": 0.4869732775322962, "learning_rate": 4.483786889596993e-06, "loss": 0.0072, "step": 823 }, { "epoch": 0.5476902625456962, "grad_norm": 0.28732781113209366, "learning_rate": 4.473072031018248e-06, "loss": 0.0092, "step": 824 }, { "epoch": 0.5483549351944168, "grad_norm": 0.3525558801182026, "learning_rate": 4.462359618886327e-06, "loss": 0.0094, "step": 825 }, { "epoch": 0.5490196078431373, "grad_norm": 0.2586132866726273, "learning_rate": 4.451649702937343e-06, "loss": 0.0095, "step": 826 }, { "epoch": 0.5496842804918578, "grad_norm": 0.32483176902914424, "learning_rate": 4.440942332895811e-06, "loss": 0.0106, "step": 827 }, { "epoch": 0.5503489531405783, "grad_norm": 0.29129550032166013, "learning_rate": 4.43023755847443e-06, "loss": 0.0114, "step": 828 }, { "epoch": 0.5510136257892988, "grad_norm": 0.25162765093753825, "learning_rate": 4.4195354293738484e-06, "loss": 0.0089, "step": 829 }, { "epoch": 0.5516782984380193, "grad_norm": 0.29297038277981785, "learning_rate": 4.408835995282434e-06, "loss": 0.0084, "step": 830 }, { "epoch": 0.5523429710867398, "grad_norm": 0.3339100195105828, "learning_rate": 4.398139305876039e-06, "loss": 0.0101, "step": 831 }, { "epoch": 0.5530076437354603, "grad_norm": 0.24802426592485774, "learning_rate": 4.387445410817774e-06, "loss": 0.0094, "step": 832 }, { "epoch": 0.5536723163841808, "grad_norm": 0.28712596016845304, "learning_rate": 4.376754359757775e-06, "loss": 0.0123, "step": 833 }, { "epoch": 0.5543369890329013, "grad_norm": 0.25154754098177656, "learning_rate": 4.366066202332973e-06, "loss": 0.0089, "step": 834 }, { "epoch": 0.5550016616816218, "grad_norm": 0.5354915306406873, "learning_rate": 4.355380988166867e-06, "loss": 0.0107, "step": 835 }, { "epoch": 0.5556663343303423, "grad_norm": 0.42959774560151104, "learning_rate": 4.3446987668692905e-06, "loss": 0.0068, "step": 836 }, { "epoch": 0.5563310069790628, "grad_norm": 0.17756801994449914, "learning_rate": 4.33401958803618e-06, "loss": 0.0078, "step": 837 }, { "epoch": 0.5569956796277833, "grad_norm": 0.18349354402899126, "learning_rate": 4.323343501249346e-06, "loss": 0.01, "step": 838 }, { "epoch": 0.5576603522765038, "grad_norm": 0.16706059351707975, "learning_rate": 4.312670556076244e-06, "loss": 0.0063, "step": 839 }, { "epoch": 0.5583250249252243, "grad_norm": 0.41204581720405936, "learning_rate": 4.302000802069744e-06, "loss": 0.0124, "step": 840 }, { "epoch": 0.5589896975739448, "grad_norm": 1.104392885950872, "learning_rate": 4.291334288767899e-06, "loss": 0.013, "step": 841 }, { "epoch": 0.5596543702226653, "grad_norm": 0.6817385349482706, "learning_rate": 4.280671065693716e-06, "loss": 0.0112, "step": 842 }, { "epoch": 0.5603190428713859, "grad_norm": 0.2631667925919954, "learning_rate": 4.270011182354927e-06, "loss": 0.0123, "step": 843 }, { "epoch": 0.5609837155201064, "grad_norm": 0.30334319600256593, "learning_rate": 4.259354688243758e-06, "loss": 0.0069, "step": 844 }, { "epoch": 0.5616483881688269, "grad_norm": 0.3950534756993261, "learning_rate": 4.248701632836694e-06, "loss": 0.0103, "step": 845 }, { "epoch": 0.5623130608175474, "grad_norm": 0.20502915664290183, "learning_rate": 4.238052065594265e-06, "loss": 0.0097, "step": 846 }, { "epoch": 0.5629777334662679, "grad_norm": 0.24147181229565928, "learning_rate": 4.227406035960798e-06, "loss": 0.0082, "step": 847 }, { "epoch": 0.5636424061149884, "grad_norm": 0.29647869175711156, "learning_rate": 4.216763593364198e-06, "loss": 0.0054, "step": 848 }, { "epoch": 0.5643070787637089, "grad_norm": 0.3498499059651843, "learning_rate": 4.206124787215714e-06, "loss": 0.0094, "step": 849 }, { "epoch": 0.5649717514124294, "grad_norm": 0.2630055216263143, "learning_rate": 4.195489666909714e-06, "loss": 0.0122, "step": 850 }, { "epoch": 0.5656364240611499, "grad_norm": 0.2579029020729969, "learning_rate": 4.1848582818234525e-06, "loss": 0.01, "step": 851 }, { "epoch": 0.5663010967098704, "grad_norm": 0.4614764932028072, "learning_rate": 4.174230681316843e-06, "loss": 0.0057, "step": 852 }, { "epoch": 0.5669657693585909, "grad_norm": 0.3281393451875222, "learning_rate": 4.163606914732224e-06, "loss": 0.0097, "step": 853 }, { "epoch": 0.5676304420073114, "grad_norm": 0.23675069244402497, "learning_rate": 4.152987031394139e-06, "loss": 0.0082, "step": 854 }, { "epoch": 0.5682951146560319, "grad_norm": 0.6320922337449688, "learning_rate": 4.142371080609094e-06, "loss": 0.0117, "step": 855 }, { "epoch": 0.5689597873047524, "grad_norm": 0.40408722728314517, "learning_rate": 4.131759111665349e-06, "loss": 0.0107, "step": 856 }, { "epoch": 0.5696244599534729, "grad_norm": 0.3042580043785332, "learning_rate": 4.121151173832666e-06, "loss": 0.0091, "step": 857 }, { "epoch": 0.5702891326021934, "grad_norm": 0.31752488541754537, "learning_rate": 4.1105473163620956e-06, "loss": 0.0068, "step": 858 }, { "epoch": 0.5709538052509139, "grad_norm": 0.2704495201581175, "learning_rate": 4.099947588485744e-06, "loss": 0.008, "step": 859 }, { "epoch": 0.5716184778996344, "grad_norm": 0.1992923865651012, "learning_rate": 4.089352039416543e-06, "loss": 0.0078, "step": 860 }, { "epoch": 0.5722831505483549, "grad_norm": 0.26693430868699486, "learning_rate": 4.078760718348025e-06, "loss": 0.0111, "step": 861 }, { "epoch": 0.5729478231970755, "grad_norm": 0.24985634515043012, "learning_rate": 4.06817367445409e-06, "loss": 0.007, "step": 862 }, { "epoch": 0.573612495845796, "grad_norm": 0.30440267534054555, "learning_rate": 4.057590956888782e-06, "loss": 0.0132, "step": 863 }, { "epoch": 0.5742771684945165, "grad_norm": 0.2477677185167418, "learning_rate": 4.047012614786056e-06, "loss": 0.0107, "step": 864 }, { "epoch": 0.574941841143237, "grad_norm": 0.2020343995992476, "learning_rate": 4.036438697259551e-06, "loss": 0.0085, "step": 865 }, { "epoch": 0.5756065137919575, "grad_norm": 0.29670379949886366, "learning_rate": 4.025869253402372e-06, "loss": 0.01, "step": 866 }, { "epoch": 0.576271186440678, "grad_norm": 0.2774624781345721, "learning_rate": 4.015304332286845e-06, "loss": 0.0115, "step": 867 }, { "epoch": 0.5769358590893985, "grad_norm": 0.3027757357985481, "learning_rate": 4.004743982964298e-06, "loss": 0.0111, "step": 868 }, { "epoch": 0.577600531738119, "grad_norm": 0.24668362912618527, "learning_rate": 3.994188254464838e-06, "loss": 0.0098, "step": 869 }, { "epoch": 0.5782652043868395, "grad_norm": 0.36095815737268167, "learning_rate": 3.983637195797111e-06, "loss": 0.007, "step": 870 }, { "epoch": 0.57892987703556, "grad_norm": 0.5331721604147639, "learning_rate": 3.9730908559480904e-06, "loss": 0.0111, "step": 871 }, { "epoch": 0.5795945496842805, "grad_norm": 0.20758829619150207, "learning_rate": 3.9625492838828336e-06, "loss": 0.0062, "step": 872 }, { "epoch": 0.580259222333001, "grad_norm": 0.2377767071847957, "learning_rate": 3.952012528544267e-06, "loss": 0.0095, "step": 873 }, { "epoch": 0.5809238949817215, "grad_norm": 0.27409653775113685, "learning_rate": 3.941480638852948e-06, "loss": 0.0093, "step": 874 }, { "epoch": 0.581588567630442, "grad_norm": 0.17738784191066295, "learning_rate": 3.930953663706848e-06, "loss": 0.0068, "step": 875 }, { "epoch": 0.5822532402791625, "grad_norm": 0.3714159041651952, "learning_rate": 3.9204316519811194e-06, "loss": 0.0092, "step": 876 }, { "epoch": 0.582917912927883, "grad_norm": 0.3267204099709432, "learning_rate": 3.909914652527872e-06, "loss": 0.0104, "step": 877 }, { "epoch": 0.5835825855766035, "grad_norm": 0.19552154609858763, "learning_rate": 3.89940271417594e-06, "loss": 0.0111, "step": 878 }, { "epoch": 0.584247258225324, "grad_norm": 0.20664256551796079, "learning_rate": 3.888895885730666e-06, "loss": 0.0064, "step": 879 }, { "epoch": 0.5849119308740446, "grad_norm": 0.2927735516459094, "learning_rate": 3.878394215973663e-06, "loss": 0.01, "step": 880 }, { "epoch": 0.5855766035227651, "grad_norm": 0.487963157718782, "learning_rate": 3.867897753662593e-06, "loss": 0.0085, "step": 881 }, { "epoch": 0.5862412761714856, "grad_norm": 0.3749537401460589, "learning_rate": 3.857406547530944e-06, "loss": 0.0079, "step": 882 }, { "epoch": 0.5869059488202061, "grad_norm": 0.4581215686284471, "learning_rate": 3.8469206462878e-06, "loss": 0.0104, "step": 883 }, { "epoch": 0.5875706214689266, "grad_norm": 0.2950089373433228, "learning_rate": 3.836440098617611e-06, "loss": 0.0067, "step": 884 }, { "epoch": 0.5882352941176471, "grad_norm": 0.4136394200761863, "learning_rate": 3.825964953179975e-06, "loss": 0.0071, "step": 885 }, { "epoch": 0.5888999667663676, "grad_norm": 0.21385497069182782, "learning_rate": 3.815495258609404e-06, "loss": 0.0102, "step": 886 }, { "epoch": 0.5895646394150881, "grad_norm": 0.3093788375767867, "learning_rate": 3.805031063515112e-06, "loss": 0.0088, "step": 887 }, { "epoch": 0.5902293120638086, "grad_norm": 0.5003744530102889, "learning_rate": 3.7945724164807695e-06, "loss": 0.0128, "step": 888 }, { "epoch": 0.5908939847125291, "grad_norm": 0.27041093922808607, "learning_rate": 3.784119366064293e-06, "loss": 0.0115, "step": 889 }, { "epoch": 0.5915586573612496, "grad_norm": 0.5336871267415977, "learning_rate": 3.773671960797613e-06, "loss": 0.0116, "step": 890 }, { "epoch": 0.5922233300099701, "grad_norm": 0.2144546209253404, "learning_rate": 3.763230249186449e-06, "loss": 0.0088, "step": 891 }, { "epoch": 0.5928880026586906, "grad_norm": 0.46707851604413364, "learning_rate": 3.752794279710094e-06, "loss": 0.0097, "step": 892 }, { "epoch": 0.5935526753074111, "grad_norm": 0.35446204255226843, "learning_rate": 3.7423641008211715e-06, "loss": 0.0093, "step": 893 }, { "epoch": 0.5942173479561316, "grad_norm": 0.5784612626950911, "learning_rate": 3.7319397609454235e-06, "loss": 0.012, "step": 894 }, { "epoch": 0.5948820206048521, "grad_norm": 0.26859645790731096, "learning_rate": 3.721521308481483e-06, "loss": 0.0075, "step": 895 }, { "epoch": 0.5955466932535726, "grad_norm": 0.49761062258351685, "learning_rate": 3.711108791800648e-06, "loss": 0.0104, "step": 896 }, { "epoch": 0.5962113659022931, "grad_norm": 0.26040368163373034, "learning_rate": 3.700702259246659e-06, "loss": 0.0079, "step": 897 }, { "epoch": 0.5968760385510137, "grad_norm": 0.3838390114941637, "learning_rate": 3.690301759135471e-06, "loss": 0.0129, "step": 898 }, { "epoch": 0.5975407111997342, "grad_norm": 0.14685851440899328, "learning_rate": 3.6799073397550326e-06, "loss": 0.0088, "step": 899 }, { "epoch": 0.5982053838484547, "grad_norm": 0.25906040643336087, "learning_rate": 3.6695190493650608e-06, "loss": 0.0075, "step": 900 }, { "epoch": 0.5988700564971752, "grad_norm": 0.572878164503762, "learning_rate": 3.6591369361968127e-06, "loss": 0.008, "step": 901 }, { "epoch": 0.5995347291458957, "grad_norm": 0.23364634950324775, "learning_rate": 3.6487610484528736e-06, "loss": 0.0084, "step": 902 }, { "epoch": 0.6001994017946162, "grad_norm": 0.22480529672823044, "learning_rate": 3.638391434306917e-06, "loss": 0.0074, "step": 903 }, { "epoch": 0.6008640744433367, "grad_norm": 0.3383734925517731, "learning_rate": 3.6280281419034934e-06, "loss": 0.0124, "step": 904 }, { "epoch": 0.6015287470920572, "grad_norm": 0.18230017286267414, "learning_rate": 3.617671219357801e-06, "loss": 0.0081, "step": 905 }, { "epoch": 0.6021934197407777, "grad_norm": 0.22240001068311868, "learning_rate": 3.607320714755461e-06, "loss": 0.0122, "step": 906 }, { "epoch": 0.6028580923894982, "grad_norm": 0.17534060516407263, "learning_rate": 3.596976676152306e-06, "loss": 0.0089, "step": 907 }, { "epoch": 0.6035227650382187, "grad_norm": 0.3622140591320175, "learning_rate": 3.5866391515741383e-06, "loss": 0.0074, "step": 908 }, { "epoch": 0.6041874376869392, "grad_norm": 0.26745435375981724, "learning_rate": 3.5763081890165213e-06, "loss": 0.0069, "step": 909 }, { "epoch": 0.6048521103356597, "grad_norm": 0.25777233140619626, "learning_rate": 3.5659838364445505e-06, "loss": 0.0105, "step": 910 }, { "epoch": 0.6055167829843802, "grad_norm": 0.2628826969068799, "learning_rate": 3.555666141792632e-06, "loss": 0.01, "step": 911 }, { "epoch": 0.6061814556331007, "grad_norm": 0.3086547175647566, "learning_rate": 3.5453551529642593e-06, "loss": 0.0151, "step": 912 }, { "epoch": 0.6068461282818212, "grad_norm": 0.4738951847159867, "learning_rate": 3.535050917831797e-06, "loss": 0.0121, "step": 913 }, { "epoch": 0.6075108009305417, "grad_norm": 0.4486015642506278, "learning_rate": 3.5247534842362484e-06, "loss": 0.0075, "step": 914 }, { "epoch": 0.6081754735792622, "grad_norm": 0.17929132555986252, "learning_rate": 3.514462899987038e-06, "loss": 0.01, "step": 915 }, { "epoch": 0.6088401462279828, "grad_norm": 0.19116437425556015, "learning_rate": 3.504179212861793e-06, "loss": 0.0081, "step": 916 }, { "epoch": 0.6095048188767033, "grad_norm": 0.3889971142520261, "learning_rate": 3.493902470606115e-06, "loss": 0.0112, "step": 917 }, { "epoch": 0.6101694915254238, "grad_norm": 0.7631560447744277, "learning_rate": 3.4836327209333643e-06, "loss": 0.0081, "step": 918 }, { "epoch": 0.6108341641741443, "grad_norm": 0.2630811957200147, "learning_rate": 3.473370011524435e-06, "loss": 0.009, "step": 919 }, { "epoch": 0.6114988368228648, "grad_norm": 0.19188762326028141, "learning_rate": 3.463114390027533e-06, "loss": 0.0106, "step": 920 }, { "epoch": 0.6121635094715853, "grad_norm": 0.6317066442111138, "learning_rate": 3.4528659040579582e-06, "loss": 0.0187, "step": 921 }, { "epoch": 0.6128281821203058, "grad_norm": 0.31684400843017074, "learning_rate": 3.442624601197877e-06, "loss": 0.0101, "step": 922 }, { "epoch": 0.6134928547690263, "grad_norm": 0.2401026201762082, "learning_rate": 3.4323905289961136e-06, "loss": 0.0093, "step": 923 }, { "epoch": 0.6141575274177468, "grad_norm": 0.2765999499640504, "learning_rate": 3.422163734967913e-06, "loss": 0.0063, "step": 924 }, { "epoch": 0.6148222000664673, "grad_norm": 0.6782051800274294, "learning_rate": 3.4119442665947346e-06, "loss": 0.0099, "step": 925 }, { "epoch": 0.6154868727151878, "grad_norm": 0.7257259874024371, "learning_rate": 3.4017321713240226e-06, "loss": 0.0094, "step": 926 }, { "epoch": 0.6161515453639083, "grad_norm": 0.3046900368167348, "learning_rate": 3.3915274965689883e-06, "loss": 0.0088, "step": 927 }, { "epoch": 0.6168162180126288, "grad_norm": 0.23996154523745525, "learning_rate": 3.3813302897083955e-06, "loss": 0.0102, "step": 928 }, { "epoch": 0.6174808906613493, "grad_norm": 0.2885343362276256, "learning_rate": 3.3711405980863316e-06, "loss": 0.0099, "step": 929 }, { "epoch": 0.6181455633100698, "grad_norm": 0.36273522301876804, "learning_rate": 3.3609584690119924e-06, "loss": 0.0077, "step": 930 }, { "epoch": 0.6188102359587903, "grad_norm": 0.7278813431347978, "learning_rate": 3.350783949759462e-06, "loss": 0.013, "step": 931 }, { "epoch": 0.6194749086075108, "grad_norm": 0.49980696421231835, "learning_rate": 3.3406170875674944e-06, "loss": 0.0124, "step": 932 }, { "epoch": 0.6201395812562313, "grad_norm": 0.32381836245247797, "learning_rate": 3.330457929639293e-06, "loss": 0.0088, "step": 933 }, { "epoch": 0.6208042539049519, "grad_norm": 0.4133144492504386, "learning_rate": 3.3203065231422904e-06, "loss": 0.0104, "step": 934 }, { "epoch": 0.6214689265536724, "grad_norm": 0.16277475097572863, "learning_rate": 3.3101629152079296e-06, "loss": 0.0073, "step": 935 }, { "epoch": 0.6221335992023929, "grad_norm": 0.22635422915011666, "learning_rate": 3.3000271529314487e-06, "loss": 0.0093, "step": 936 }, { "epoch": 0.6227982718511134, "grad_norm": 0.21368932174418523, "learning_rate": 3.289899283371657e-06, "loss": 0.0081, "step": 937 }, { "epoch": 0.6234629444998339, "grad_norm": 0.2794426426952632, "learning_rate": 3.2797793535507204e-06, "loss": 0.0107, "step": 938 }, { "epoch": 0.6241276171485544, "grad_norm": 0.322128335795853, "learning_rate": 3.2696674104539444e-06, "loss": 0.0086, "step": 939 }, { "epoch": 0.6247922897972749, "grad_norm": 0.2421829073773608, "learning_rate": 3.259563501029548e-06, "loss": 0.0075, "step": 940 }, { "epoch": 0.6254569624459954, "grad_norm": 0.32130301213787865, "learning_rate": 3.2494676721884543e-06, "loss": 0.0118, "step": 941 }, { "epoch": 0.6261216350947159, "grad_norm": 0.347721069814857, "learning_rate": 3.2393799708040674e-06, "loss": 0.01, "step": 942 }, { "epoch": 0.6267863077434364, "grad_norm": 0.1792082240350679, "learning_rate": 3.2293004437120622e-06, "loss": 0.0084, "step": 943 }, { "epoch": 0.6274509803921569, "grad_norm": 0.31266061038755355, "learning_rate": 3.2192291377101547e-06, "loss": 0.0114, "step": 944 }, { "epoch": 0.6281156530408774, "grad_norm": 0.19925247258232015, "learning_rate": 3.2091660995578966e-06, "loss": 0.0081, "step": 945 }, { "epoch": 0.6287803256895979, "grad_norm": 0.40169324139993695, "learning_rate": 3.1991113759764493e-06, "loss": 0.008, "step": 946 }, { "epoch": 0.6294449983383184, "grad_norm": 0.2117190578788506, "learning_rate": 3.189065013648372e-06, "loss": 0.0089, "step": 947 }, { "epoch": 0.6301096709870389, "grad_norm": 0.23770522976838573, "learning_rate": 3.1790270592174088e-06, "loss": 0.0082, "step": 948 }, { "epoch": 0.6307743436357593, "grad_norm": 0.2176049091866862, "learning_rate": 3.1689975592882603e-06, "loss": 0.0116, "step": 949 }, { "epoch": 0.6314390162844798, "grad_norm": 0.24580378556214288, "learning_rate": 3.158976560426379e-06, "loss": 0.0075, "step": 950 }, { "epoch": 0.6321036889332003, "grad_norm": 0.2859953080386775, "learning_rate": 3.148964109157743e-06, "loss": 0.012, "step": 951 }, { "epoch": 0.632768361581921, "grad_norm": 0.23955087427163585, "learning_rate": 3.1389602519686515e-06, "loss": 0.0116, "step": 952 }, { "epoch": 0.6334330342306415, "grad_norm": 0.24367914742949504, "learning_rate": 3.128965035305497e-06, "loss": 0.0089, "step": 953 }, { "epoch": 0.634097706879362, "grad_norm": 0.23505495921007316, "learning_rate": 3.11897850557456e-06, "loss": 0.0122, "step": 954 }, { "epoch": 0.6347623795280825, "grad_norm": 0.36256136723764965, "learning_rate": 3.1090007091417884e-06, "loss": 0.0077, "step": 955 }, { "epoch": 0.635427052176803, "grad_norm": 0.5783020350703105, "learning_rate": 3.0990316923325782e-06, "loss": 0.0106, "step": 956 }, { "epoch": 0.6360917248255235, "grad_norm": 0.31649309477895776, "learning_rate": 3.089071501431567e-06, "loss": 0.0086, "step": 957 }, { "epoch": 0.636756397474244, "grad_norm": 0.696207916808104, "learning_rate": 3.0791201826824117e-06, "loss": 0.012, "step": 958 }, { "epoch": 0.6374210701229645, "grad_norm": 0.20769856628640398, "learning_rate": 3.0691777822875845e-06, "loss": 0.0103, "step": 959 }, { "epoch": 0.638085742771685, "grad_norm": 0.44868891079841433, "learning_rate": 3.0592443464081424e-06, "loss": 0.012, "step": 960 }, { "epoch": 0.6387504154204054, "grad_norm": 0.1909386876833034, "learning_rate": 3.049319921163526e-06, "loss": 0.0071, "step": 961 }, { "epoch": 0.639415088069126, "grad_norm": 0.3301173458668869, "learning_rate": 3.03940455263134e-06, "loss": 0.0136, "step": 962 }, { "epoch": 0.6400797607178464, "grad_norm": 0.27708704665407097, "learning_rate": 3.029498286847138e-06, "loss": 0.0075, "step": 963 }, { "epoch": 0.640744433366567, "grad_norm": 0.3759232406313754, "learning_rate": 3.019601169804216e-06, "loss": 0.0109, "step": 964 }, { "epoch": 0.6414091060152874, "grad_norm": 0.27177993471319756, "learning_rate": 3.009713247453391e-06, "loss": 0.0068, "step": 965 }, { "epoch": 0.6420737786640079, "grad_norm": 0.35348824046889704, "learning_rate": 2.999834565702787e-06, "loss": 0.0108, "step": 966 }, { "epoch": 0.6427384513127284, "grad_norm": 0.33342069949697306, "learning_rate": 2.9899651704176324e-06, "loss": 0.0094, "step": 967 }, { "epoch": 0.6434031239614489, "grad_norm": 0.573642069568304, "learning_rate": 2.980105107420031e-06, "loss": 0.0111, "step": 968 }, { "epoch": 0.6440677966101694, "grad_norm": 0.161625437702315, "learning_rate": 2.9702544224887687e-06, "loss": 0.0054, "step": 969 }, { "epoch": 0.64473246925889, "grad_norm": 0.24481655132663913, "learning_rate": 2.9604131613590825e-06, "loss": 0.0078, "step": 970 }, { "epoch": 0.6453971419076105, "grad_norm": 0.3681819548003033, "learning_rate": 2.9505813697224584e-06, "loss": 0.0088, "step": 971 }, { "epoch": 0.646061814556331, "grad_norm": 0.36813362352122053, "learning_rate": 2.9407590932264187e-06, "loss": 0.0099, "step": 972 }, { "epoch": 0.6467264872050515, "grad_norm": 0.3976928845492841, "learning_rate": 2.9309463774743047e-06, "loss": 0.0078, "step": 973 }, { "epoch": 0.647391159853772, "grad_norm": 0.2272118381342224, "learning_rate": 2.921143268025072e-06, "loss": 0.0074, "step": 974 }, { "epoch": 0.6480558325024925, "grad_norm": 0.2932799475251196, "learning_rate": 2.911349810393076e-06, "loss": 0.009, "step": 975 }, { "epoch": 0.648720505151213, "grad_norm": 0.31351018891128823, "learning_rate": 2.901566050047855e-06, "loss": 0.0129, "step": 976 }, { "epoch": 0.6493851777999335, "grad_norm": 0.20170901171224367, "learning_rate": 2.891792032413932e-06, "loss": 0.0095, "step": 977 }, { "epoch": 0.650049850448654, "grad_norm": 0.3421935906129705, "learning_rate": 2.8820278028705855e-06, "loss": 0.0071, "step": 978 }, { "epoch": 0.6507145230973745, "grad_norm": 0.3554264393007603, "learning_rate": 2.8722734067516637e-06, "loss": 0.0085, "step": 979 }, { "epoch": 0.651379195746095, "grad_norm": 0.2349606425846591, "learning_rate": 2.862528889345346e-06, "loss": 0.0096, "step": 980 }, { "epoch": 0.6520438683948155, "grad_norm": 0.3076753543848085, "learning_rate": 2.8527942958939537e-06, "loss": 0.0076, "step": 981 }, { "epoch": 0.652708541043536, "grad_norm": 0.14398183811473508, "learning_rate": 2.843069671593734e-06, "loss": 0.0061, "step": 982 }, { "epoch": 0.6533732136922565, "grad_norm": 0.2263995658621755, "learning_rate": 2.8333550615946415e-06, "loss": 0.0118, "step": 983 }, { "epoch": 0.654037886340977, "grad_norm": 0.5808365121043862, "learning_rate": 2.823650511000142e-06, "loss": 0.0112, "step": 984 }, { "epoch": 0.6547025589896975, "grad_norm": 0.29564543999897336, "learning_rate": 2.813956064866996e-06, "loss": 0.0075, "step": 985 }, { "epoch": 0.655367231638418, "grad_norm": 0.30933570800269505, "learning_rate": 2.8042717682050523e-06, "loss": 0.0064, "step": 986 }, { "epoch": 0.6560319042871385, "grad_norm": 0.40093514876726827, "learning_rate": 2.79459766597703e-06, "loss": 0.012, "step": 987 }, { "epoch": 0.6566965769358591, "grad_norm": 0.1954223940513839, "learning_rate": 2.784933803098326e-06, "loss": 0.0062, "step": 988 }, { "epoch": 0.6573612495845796, "grad_norm": 0.21071688734414556, "learning_rate": 2.7752802244367875e-06, "loss": 0.0072, "step": 989 }, { "epoch": 0.6580259222333001, "grad_norm": 0.41025794500212187, "learning_rate": 2.7656369748125277e-06, "loss": 0.0086, "step": 990 }, { "epoch": 0.6586905948820206, "grad_norm": 0.1935032363904584, "learning_rate": 2.7560040989976894e-06, "loss": 0.0068, "step": 991 }, { "epoch": 0.6593552675307411, "grad_norm": 0.40488179544340286, "learning_rate": 2.7463816417162615e-06, "loss": 0.0074, "step": 992 }, { "epoch": 0.6600199401794616, "grad_norm": 0.1838037586133697, "learning_rate": 2.736769647643852e-06, "loss": 0.0092, "step": 993 }, { "epoch": 0.6606846128281821, "grad_norm": 0.24819942058511169, "learning_rate": 2.7271681614074973e-06, "loss": 0.0089, "step": 994 }, { "epoch": 0.6613492854769026, "grad_norm": 0.25683540579208103, "learning_rate": 2.717577227585446e-06, "loss": 0.0094, "step": 995 }, { "epoch": 0.6620139581256231, "grad_norm": 0.18391189835613822, "learning_rate": 2.707996890706954e-06, "loss": 0.0095, "step": 996 }, { "epoch": 0.6626786307743436, "grad_norm": 0.15104364444426333, "learning_rate": 2.6984271952520723e-06, "loss": 0.0068, "step": 997 }, { "epoch": 0.6633433034230641, "grad_norm": 0.3940394230961875, "learning_rate": 2.68886818565145e-06, "loss": 0.0111, "step": 998 }, { "epoch": 0.6640079760717846, "grad_norm": 0.2609681211065955, "learning_rate": 2.679319906286122e-06, "loss": 0.0097, "step": 999 }, { "epoch": 0.6646726487205051, "grad_norm": 0.3645402376552464, "learning_rate": 2.6697824014873076e-06, "loss": 0.0099, "step": 1000 }, { "epoch": 0.6653373213692256, "grad_norm": 0.23654236854625932, "learning_rate": 2.660255715536193e-06, "loss": 0.0075, "step": 1001 }, { "epoch": 0.6660019940179461, "grad_norm": 0.24009513254503748, "learning_rate": 2.650739892663743e-06, "loss": 0.0089, "step": 1002 }, { "epoch": 0.6666666666666666, "grad_norm": 0.20154175339249858, "learning_rate": 2.641234977050484e-06, "loss": 0.0089, "step": 1003 }, { "epoch": 0.6673313393153871, "grad_norm": 0.18869581180280823, "learning_rate": 2.6317410128262953e-06, "loss": 0.0096, "step": 1004 }, { "epoch": 0.6679960119641076, "grad_norm": 0.19084928944347762, "learning_rate": 2.622258044070225e-06, "loss": 0.0071, "step": 1005 }, { "epoch": 0.6686606846128281, "grad_norm": 0.22525861430167277, "learning_rate": 2.6127861148102552e-06, "loss": 0.0081, "step": 1006 }, { "epoch": 0.6693253572615487, "grad_norm": 0.211563645002409, "learning_rate": 2.6033252690231238e-06, "loss": 0.0082, "step": 1007 }, { "epoch": 0.6699900299102692, "grad_norm": 0.5372954390814825, "learning_rate": 2.5938755506341036e-06, "loss": 0.0082, "step": 1008 }, { "epoch": 0.6706547025589897, "grad_norm": 0.22068958935151187, "learning_rate": 2.5844370035168077e-06, "loss": 0.0077, "step": 1009 }, { "epoch": 0.6713193752077102, "grad_norm": 0.24185511029717174, "learning_rate": 2.575009671492983e-06, "loss": 0.0063, "step": 1010 }, { "epoch": 0.6719840478564307, "grad_norm": 0.31560557307761405, "learning_rate": 2.565593598332308e-06, "loss": 0.0062, "step": 1011 }, { "epoch": 0.6726487205051512, "grad_norm": 0.5130502047257309, "learning_rate": 2.5561888277521797e-06, "loss": 0.0086, "step": 1012 }, { "epoch": 0.6733133931538717, "grad_norm": 0.38937425351118965, "learning_rate": 2.5467954034175303e-06, "loss": 0.0097, "step": 1013 }, { "epoch": 0.6739780658025922, "grad_norm": 0.21505699299118056, "learning_rate": 2.537413368940601e-06, "loss": 0.009, "step": 1014 }, { "epoch": 0.6746427384513127, "grad_norm": 0.2499664011990959, "learning_rate": 2.528042767880766e-06, "loss": 0.0118, "step": 1015 }, { "epoch": 0.6753074111000332, "grad_norm": 0.29296563693167743, "learning_rate": 2.5186836437443037e-06, "loss": 0.0095, "step": 1016 }, { "epoch": 0.6759720837487537, "grad_norm": 0.34769532892270855, "learning_rate": 2.5093360399842114e-06, "loss": 0.0065, "step": 1017 }, { "epoch": 0.6766367563974742, "grad_norm": 0.2727489794496769, "learning_rate": 2.5000000000000015e-06, "loss": 0.0075, "step": 1018 }, { "epoch": 0.6773014290461947, "grad_norm": 0.28787812689732917, "learning_rate": 2.4906755671374905e-06, "loss": 0.0087, "step": 1019 }, { "epoch": 0.6779661016949152, "grad_norm": 0.4992399231267387, "learning_rate": 2.4813627846886117e-06, "loss": 0.0085, "step": 1020 }, { "epoch": 0.6786307743436357, "grad_norm": 0.34835717822431245, "learning_rate": 2.4720616958912054e-06, "loss": 0.0054, "step": 1021 }, { "epoch": 0.6792954469923562, "grad_norm": 0.6266322829734265, "learning_rate": 2.4627723439288204e-06, "loss": 0.0119, "step": 1022 }, { "epoch": 0.6799601196410767, "grad_norm": 0.1957354028539535, "learning_rate": 2.453494771930508e-06, "loss": 0.0086, "step": 1023 }, { "epoch": 0.6806247922897972, "grad_norm": 0.2605729316787177, "learning_rate": 2.4442290229706344e-06, "loss": 0.0122, "step": 1024 }, { "epoch": 0.6812894649385178, "grad_norm": 0.23665821585690736, "learning_rate": 2.434975140068669e-06, "loss": 0.007, "step": 1025 }, { "epoch": 0.6819541375872383, "grad_norm": 0.2201415649312906, "learning_rate": 2.4257331661889923e-06, "loss": 0.0098, "step": 1026 }, { "epoch": 0.6826188102359588, "grad_norm": 0.3164784462208598, "learning_rate": 2.4165031442406857e-06, "loss": 0.0068, "step": 1027 }, { "epoch": 0.6832834828846793, "grad_norm": 0.2031926822498836, "learning_rate": 2.4072851170773493e-06, "loss": 0.0067, "step": 1028 }, { "epoch": 0.6839481555333998, "grad_norm": 0.5685953903499967, "learning_rate": 2.398079127496884e-06, "loss": 0.0083, "step": 1029 }, { "epoch": 0.6846128281821203, "grad_norm": 0.4124104323221366, "learning_rate": 2.3888852182413087e-06, "loss": 0.008, "step": 1030 }, { "epoch": 0.6852775008308408, "grad_norm": 0.31933378899670556, "learning_rate": 2.3797034319965517e-06, "loss": 0.0103, "step": 1031 }, { "epoch": 0.6859421734795613, "grad_norm": 0.2697847646102195, "learning_rate": 2.370533811392261e-06, "loss": 0.0123, "step": 1032 }, { "epoch": 0.6866068461282818, "grad_norm": 0.24910557974411127, "learning_rate": 2.361376399001592e-06, "loss": 0.0091, "step": 1033 }, { "epoch": 0.6872715187770023, "grad_norm": 0.2780938443797265, "learning_rate": 2.3522312373410276e-06, "loss": 0.0108, "step": 1034 }, { "epoch": 0.6879361914257228, "grad_norm": 0.3566072774983843, "learning_rate": 2.3430983688701693e-06, "loss": 0.0101, "step": 1035 }, { "epoch": 0.6886008640744433, "grad_norm": 0.4960859376704064, "learning_rate": 2.333977835991545e-06, "loss": 0.0101, "step": 1036 }, { "epoch": 0.6892655367231638, "grad_norm": 0.5317805068289205, "learning_rate": 2.3248696810504052e-06, "loss": 0.0098, "step": 1037 }, { "epoch": 0.6899302093718843, "grad_norm": 0.1938454533449783, "learning_rate": 2.315773946334537e-06, "loss": 0.0069, "step": 1038 }, { "epoch": 0.6905948820206048, "grad_norm": 0.6492464563794429, "learning_rate": 2.3066906740740626e-06, "loss": 0.0077, "step": 1039 }, { "epoch": 0.6912595546693253, "grad_norm": 0.23311160889060348, "learning_rate": 2.297619906441234e-06, "loss": 0.0079, "step": 1040 }, { "epoch": 0.6919242273180458, "grad_norm": 0.48555576709924847, "learning_rate": 2.288561685550261e-06, "loss": 0.0095, "step": 1041 }, { "epoch": 0.6925888999667663, "grad_norm": 0.37785034335609585, "learning_rate": 2.2795160534570866e-06, "loss": 0.0109, "step": 1042 }, { "epoch": 0.6932535726154869, "grad_norm": 0.5271114818146994, "learning_rate": 2.270483052159216e-06, "loss": 0.0099, "step": 1043 }, { "epoch": 0.6939182452642074, "grad_norm": 0.2746157855119046, "learning_rate": 2.261462723595503e-06, "loss": 0.0096, "step": 1044 }, { "epoch": 0.6945829179129279, "grad_norm": 0.2084867131464582, "learning_rate": 2.2524551096459703e-06, "loss": 0.0086, "step": 1045 }, { "epoch": 0.6952475905616484, "grad_norm": 0.20692844736475394, "learning_rate": 2.243460252131606e-06, "loss": 0.0094, "step": 1046 }, { "epoch": 0.6959122632103689, "grad_norm": 0.26891033391026026, "learning_rate": 2.2344781928141753e-06, "loss": 0.013, "step": 1047 }, { "epoch": 0.6965769358590894, "grad_norm": 0.5760814124778825, "learning_rate": 2.2255089733960162e-06, "loss": 0.0129, "step": 1048 }, { "epoch": 0.6972416085078099, "grad_norm": 0.30749673888826295, "learning_rate": 2.216552635519861e-06, "loss": 0.005, "step": 1049 }, { "epoch": 0.6979062811565304, "grad_norm": 0.29505761160797606, "learning_rate": 2.207609220768625e-06, "loss": 0.0091, "step": 1050 }, { "epoch": 0.6985709538052509, "grad_norm": 0.17156404897235497, "learning_rate": 2.1986787706652377e-06, "loss": 0.0054, "step": 1051 }, { "epoch": 0.6992356264539714, "grad_norm": 0.45828026748924755, "learning_rate": 2.1897613266724225e-06, "loss": 0.0084, "step": 1052 }, { "epoch": 0.6999002991026919, "grad_norm": 0.24937158859953384, "learning_rate": 2.1808569301925255e-06, "loss": 0.0086, "step": 1053 }, { "epoch": 0.7005649717514124, "grad_norm": 0.20009862827547395, "learning_rate": 2.171965622567308e-06, "loss": 0.0067, "step": 1054 }, { "epoch": 0.7012296444001329, "grad_norm": 0.2610053544028202, "learning_rate": 2.1630874450777684e-06, "loss": 0.008, "step": 1055 }, { "epoch": 0.7018943170488534, "grad_norm": 0.36206474630804264, "learning_rate": 2.1542224389439415e-06, "loss": 0.0109, "step": 1056 }, { "epoch": 0.7025589896975739, "grad_norm": 0.5153003321704763, "learning_rate": 2.1453706453247088e-06, "loss": 0.0106, "step": 1057 }, { "epoch": 0.7032236623462944, "grad_norm": 0.28180824613056, "learning_rate": 2.1365321053176103e-06, "loss": 0.0077, "step": 1058 }, { "epoch": 0.7038883349950149, "grad_norm": 0.18391422249578976, "learning_rate": 2.127706859958647e-06, "loss": 0.0056, "step": 1059 }, { "epoch": 0.7045530076437354, "grad_norm": 0.20132395241767914, "learning_rate": 2.1188949502220987e-06, "loss": 0.0094, "step": 1060 }, { "epoch": 0.705217680292456, "grad_norm": 0.19824157734039166, "learning_rate": 2.110096417020329e-06, "loss": 0.0101, "step": 1061 }, { "epoch": 0.7058823529411765, "grad_norm": 0.31171142557284354, "learning_rate": 2.1013113012035987e-06, "loss": 0.0122, "step": 1062 }, { "epoch": 0.706547025589897, "grad_norm": 0.26011746700717575, "learning_rate": 2.0925396435598665e-06, "loss": 0.0152, "step": 1063 }, { "epoch": 0.7072116982386175, "grad_norm": 0.19498909156207744, "learning_rate": 2.0837814848146166e-06, "loss": 0.0063, "step": 1064 }, { "epoch": 0.707876370887338, "grad_norm": 0.16478822992660008, "learning_rate": 2.075036865630649e-06, "loss": 0.0062, "step": 1065 }, { "epoch": 0.7085410435360585, "grad_norm": 0.24476062400704385, "learning_rate": 2.066305826607911e-06, "loss": 0.0093, "step": 1066 }, { "epoch": 0.709205716184779, "grad_norm": 0.2755871609128912, "learning_rate": 2.057588408283295e-06, "loss": 0.0084, "step": 1067 }, { "epoch": 0.7098703888334995, "grad_norm": 0.5967019937302827, "learning_rate": 2.0488846511304574e-06, "loss": 0.0116, "step": 1068 }, { "epoch": 0.71053506148222, "grad_norm": 0.22449805956333627, "learning_rate": 2.0401945955596206e-06, "loss": 0.0063, "step": 1069 }, { "epoch": 0.7111997341309405, "grad_norm": 0.3953906525587086, "learning_rate": 2.031518281917403e-06, "loss": 0.0111, "step": 1070 }, { "epoch": 0.711864406779661, "grad_norm": 0.20552810831043333, "learning_rate": 2.0228557504866098e-06, "loss": 0.0091, "step": 1071 }, { "epoch": 0.7125290794283815, "grad_norm": 0.19438287812600616, "learning_rate": 2.0142070414860704e-06, "loss": 0.0074, "step": 1072 }, { "epoch": 0.713193752077102, "grad_norm": 0.2346091558916181, "learning_rate": 2.0055721950704265e-06, "loss": 0.0081, "step": 1073 }, { "epoch": 0.7138584247258225, "grad_norm": 0.2799026020572236, "learning_rate": 1.9969512513299667e-06, "loss": 0.0094, "step": 1074 }, { "epoch": 0.714523097374543, "grad_norm": 0.2812263108958078, "learning_rate": 1.9883442502904284e-06, "loss": 0.0104, "step": 1075 }, { "epoch": 0.7151877700232635, "grad_norm": 0.5864755921529266, "learning_rate": 1.9797512319128113e-06, "loss": 0.0103, "step": 1076 }, { "epoch": 0.715852442671984, "grad_norm": 0.36431552265388817, "learning_rate": 1.9711722360932032e-06, "loss": 0.0093, "step": 1077 }, { "epoch": 0.7165171153207045, "grad_norm": 0.37856788599561675, "learning_rate": 1.962607302662582e-06, "loss": 0.0101, "step": 1078 }, { "epoch": 0.7171817879694251, "grad_norm": 0.25389342991422, "learning_rate": 1.954056471386639e-06, "loss": 0.0094, "step": 1079 }, { "epoch": 0.7178464606181456, "grad_norm": 0.25274846244933785, "learning_rate": 1.9455197819655874e-06, "loss": 0.01, "step": 1080 }, { "epoch": 0.7185111332668661, "grad_norm": 0.40880975408120696, "learning_rate": 1.936997274033986e-06, "loss": 0.01, "step": 1081 }, { "epoch": 0.7191758059155866, "grad_norm": 0.36741432876198077, "learning_rate": 1.928488987160549e-06, "loss": 0.0114, "step": 1082 }, { "epoch": 0.7198404785643071, "grad_norm": 0.45231503627975217, "learning_rate": 1.919994960847967e-06, "loss": 0.0121, "step": 1083 }, { "epoch": 0.7205051512130276, "grad_norm": 0.40540511927594813, "learning_rate": 1.9115152345327154e-06, "loss": 0.0135, "step": 1084 }, { "epoch": 0.7211698238617481, "grad_norm": 0.5985682335123047, "learning_rate": 1.9030498475848836e-06, "loss": 0.0156, "step": 1085 }, { "epoch": 0.7218344965104686, "grad_norm": 0.23782977919360973, "learning_rate": 1.8945988393079767e-06, "loss": 0.0084, "step": 1086 }, { "epoch": 0.7224991691591891, "grad_norm": 0.22204768766409977, "learning_rate": 1.8861622489387555e-06, "loss": 0.0081, "step": 1087 }, { "epoch": 0.7231638418079096, "grad_norm": 0.19671607473231634, "learning_rate": 1.877740115647026e-06, "loss": 0.0077, "step": 1088 }, { "epoch": 0.7238285144566301, "grad_norm": 0.31717712158835565, "learning_rate": 1.869332478535482e-06, "loss": 0.0075, "step": 1089 }, { "epoch": 0.7244931871053506, "grad_norm": 0.20974258133420504, "learning_rate": 1.8609393766395083e-06, "loss": 0.0071, "step": 1090 }, { "epoch": 0.7251578597540711, "grad_norm": 0.227446610653222, "learning_rate": 1.8525608489270076e-06, "loss": 0.0066, "step": 1091 }, { "epoch": 0.7258225324027916, "grad_norm": 0.38841195839936826, "learning_rate": 1.8441969342982163e-06, "loss": 0.0088, "step": 1092 }, { "epoch": 0.7264872050515121, "grad_norm": 0.3788166674198978, "learning_rate": 1.8358476715855262e-06, "loss": 0.011, "step": 1093 }, { "epoch": 0.7271518777002326, "grad_norm": 0.16872630511567938, "learning_rate": 1.8275130995532974e-06, "loss": 0.0075, "step": 1094 }, { "epoch": 0.7278165503489531, "grad_norm": 0.3109404863259487, "learning_rate": 1.8191932568976889e-06, "loss": 0.0116, "step": 1095 }, { "epoch": 0.7284812229976736, "grad_norm": 0.1780456890629778, "learning_rate": 1.8108881822464697e-06, "loss": 0.0044, "step": 1096 }, { "epoch": 0.7291458956463942, "grad_norm": 0.19005863968055675, "learning_rate": 1.8025979141588462e-06, "loss": 0.0073, "step": 1097 }, { "epoch": 0.7298105682951147, "grad_norm": 0.2530368239715967, "learning_rate": 1.7943224911252794e-06, "loss": 0.0075, "step": 1098 }, { "epoch": 0.7304752409438352, "grad_norm": 0.23367996596455318, "learning_rate": 1.7860619515673034e-06, "loss": 0.0108, "step": 1099 }, { "epoch": 0.7311399135925557, "grad_norm": 0.17488852313656292, "learning_rate": 1.7778163338373567e-06, "loss": 0.0091, "step": 1100 }, { "epoch": 0.7318045862412762, "grad_norm": 0.24169717195528131, "learning_rate": 1.7695856762185892e-06, "loss": 0.0111, "step": 1101 }, { "epoch": 0.7324692588899967, "grad_norm": 0.49770088570403187, "learning_rate": 1.7613700169247055e-06, "loss": 0.0074, "step": 1102 }, { "epoch": 0.7331339315387172, "grad_norm": 0.35656588808246886, "learning_rate": 1.7531693940997641e-06, "loss": 0.0073, "step": 1103 }, { "epoch": 0.7337986041874377, "grad_norm": 0.4646395799201426, "learning_rate": 1.744983845818019e-06, "loss": 0.0111, "step": 1104 }, { "epoch": 0.7344632768361582, "grad_norm": 0.30206693557737946, "learning_rate": 1.7368134100837286e-06, "loss": 0.0108, "step": 1105 }, { "epoch": 0.7351279494848787, "grad_norm": 0.19912927059280186, "learning_rate": 1.728658124830993e-06, "loss": 0.0095, "step": 1106 }, { "epoch": 0.7357926221335992, "grad_norm": 0.31365924671273226, "learning_rate": 1.7205180279235627e-06, "loss": 0.006, "step": 1107 }, { "epoch": 0.7364572947823197, "grad_norm": 0.24607295500620077, "learning_rate": 1.7123931571546826e-06, "loss": 0.0109, "step": 1108 }, { "epoch": 0.7371219674310402, "grad_norm": 0.3114568025543917, "learning_rate": 1.7042835502468936e-06, "loss": 0.0085, "step": 1109 }, { "epoch": 0.7377866400797607, "grad_norm": 0.2222091000938827, "learning_rate": 1.696189244851874e-06, "loss": 0.0084, "step": 1110 }, { "epoch": 0.7384513127284812, "grad_norm": 0.22538225229577186, "learning_rate": 1.6881102785502618e-06, "loss": 0.0124, "step": 1111 }, { "epoch": 0.7391159853772017, "grad_norm": 0.22344568489200187, "learning_rate": 1.680046688851471e-06, "loss": 0.0071, "step": 1112 }, { "epoch": 0.7397806580259222, "grad_norm": 0.16989939033750653, "learning_rate": 1.6719985131935308e-06, "loss": 0.0072, "step": 1113 }, { "epoch": 0.7404453306746427, "grad_norm": 0.4935050306495117, "learning_rate": 1.6639657889429017e-06, "loss": 0.0095, "step": 1114 }, { "epoch": 0.7411100033233633, "grad_norm": 0.19266789264953518, "learning_rate": 1.6559485533943092e-06, "loss": 0.0064, "step": 1115 }, { "epoch": 0.7417746759720838, "grad_norm": 0.446555497324972, "learning_rate": 1.647946843770562e-06, "loss": 0.0073, "step": 1116 }, { "epoch": 0.7424393486208043, "grad_norm": 0.34268526110184033, "learning_rate": 1.639960697222388e-06, "loss": 0.0081, "step": 1117 }, { "epoch": 0.7431040212695248, "grad_norm": 0.22360453110933848, "learning_rate": 1.6319901508282577e-06, "loss": 0.0117, "step": 1118 }, { "epoch": 0.7437686939182453, "grad_norm": 0.3892214760395955, "learning_rate": 1.624035241594213e-06, "loss": 0.0098, "step": 1119 }, { "epoch": 0.7444333665669658, "grad_norm": 0.17989440718297894, "learning_rate": 1.6160960064536907e-06, "loss": 0.0049, "step": 1120 }, { "epoch": 0.7450980392156863, "grad_norm": 0.3441734612638478, "learning_rate": 1.6081724822673611e-06, "loss": 0.0121, "step": 1121 }, { "epoch": 0.7457627118644068, "grad_norm": 0.18321831971977126, "learning_rate": 1.6002647058229437e-06, "loss": 0.0058, "step": 1122 }, { "epoch": 0.7464273845131273, "grad_norm": 0.27607588235363356, "learning_rate": 1.5923727138350548e-06, "loss": 0.0091, "step": 1123 }, { "epoch": 0.7470920571618478, "grad_norm": 0.6005890235792981, "learning_rate": 1.5844965429450133e-06, "loss": 0.0088, "step": 1124 }, { "epoch": 0.7477567298105683, "grad_norm": 0.3305883781990034, "learning_rate": 1.5766362297206917e-06, "loss": 0.0111, "step": 1125 }, { "epoch": 0.7484214024592888, "grad_norm": 0.18849564536836697, "learning_rate": 1.5687918106563326e-06, "loss": 0.0073, "step": 1126 }, { "epoch": 0.7490860751080093, "grad_norm": 0.3050537896965794, "learning_rate": 1.5609633221723885e-06, "loss": 0.0097, "step": 1127 }, { "epoch": 0.7497507477567298, "grad_norm": 0.23500705314353945, "learning_rate": 1.5531508006153462e-06, "loss": 0.0054, "step": 1128 }, { "epoch": 0.7504154204054503, "grad_norm": 0.1956593383266017, "learning_rate": 1.5453542822575624e-06, "loss": 0.0066, "step": 1129 }, { "epoch": 0.7510800930541708, "grad_norm": 0.3657603832823888, "learning_rate": 1.5375738032970894e-06, "loss": 0.01, "step": 1130 }, { "epoch": 0.7517447657028913, "grad_norm": 0.24669301872116672, "learning_rate": 1.5298093998575152e-06, "loss": 0.0076, "step": 1131 }, { "epoch": 0.7524094383516118, "grad_norm": 0.3086268545923642, "learning_rate": 1.52206110798779e-06, "loss": 0.0115, "step": 1132 }, { "epoch": 0.7530741110003324, "grad_norm": 0.3394309769943332, "learning_rate": 1.5143289636620606e-06, "loss": 0.0082, "step": 1133 }, { "epoch": 0.7537387836490529, "grad_norm": 0.4161383126186537, "learning_rate": 1.5066130027795046e-06, "loss": 0.0088, "step": 1134 }, { "epoch": 0.7544034562977734, "grad_norm": 0.6669551305883772, "learning_rate": 1.4989132611641576e-06, "loss": 0.01, "step": 1135 }, { "epoch": 0.7550681289464939, "grad_norm": 0.7583220265882658, "learning_rate": 1.491229774564758e-06, "loss": 0.0103, "step": 1136 }, { "epoch": 0.7557328015952144, "grad_norm": 0.16308272681711836, "learning_rate": 1.4835625786545666e-06, "loss": 0.0069, "step": 1137 }, { "epoch": 0.7563974742439349, "grad_norm": 0.25901031767584826, "learning_rate": 1.4759117090312197e-06, "loss": 0.006, "step": 1138 }, { "epoch": 0.7570621468926554, "grad_norm": 0.34703931374055236, "learning_rate": 1.4682772012165435e-06, "loss": 0.0086, "step": 1139 }, { "epoch": 0.7577268195413759, "grad_norm": 0.32760820780427374, "learning_rate": 1.4606590906564055e-06, "loss": 0.0106, "step": 1140 }, { "epoch": 0.7583914921900964, "grad_norm": 0.2464013647257441, "learning_rate": 1.453057412720536e-06, "loss": 0.0048, "step": 1141 }, { "epoch": 0.7590561648388169, "grad_norm": 0.24307084793057304, "learning_rate": 1.4454722027023798e-06, "loss": 0.0098, "step": 1142 }, { "epoch": 0.7597208374875374, "grad_norm": 0.28372569032423917, "learning_rate": 1.437903495818914e-06, "loss": 0.0063, "step": 1143 }, { "epoch": 0.7603855101362579, "grad_norm": 0.254849398965095, "learning_rate": 1.4303513272105057e-06, "loss": 0.0054, "step": 1144 }, { "epoch": 0.7610501827849784, "grad_norm": 0.3577358829627769, "learning_rate": 1.422815731940726e-06, "loss": 0.0089, "step": 1145 }, { "epoch": 0.7617148554336989, "grad_norm": 0.3262026271605971, "learning_rate": 1.4152967449962067e-06, "loss": 0.0101, "step": 1146 }, { "epoch": 0.7623795280824194, "grad_norm": 0.4396677613166942, "learning_rate": 1.4077944012864636e-06, "loss": 0.0121, "step": 1147 }, { "epoch": 0.7630442007311399, "grad_norm": 0.3735700658155585, "learning_rate": 1.400308735643745e-06, "loss": 0.0125, "step": 1148 }, { "epoch": 0.7637088733798604, "grad_norm": 0.2049654627255116, "learning_rate": 1.392839782822863e-06, "loss": 0.0078, "step": 1149 }, { "epoch": 0.7643735460285809, "grad_norm": 0.4072052028309622, "learning_rate": 1.3853875775010355e-06, "loss": 0.0074, "step": 1150 }, { "epoch": 0.7650382186773014, "grad_norm": 0.7007804259499075, "learning_rate": 1.3779521542777263e-06, "loss": 0.011, "step": 1151 }, { "epoch": 0.765702891326022, "grad_norm": 0.3262383334404254, "learning_rate": 1.370533547674477e-06, "loss": 0.0093, "step": 1152 }, { "epoch": 0.7663675639747425, "grad_norm": 0.5129541884543688, "learning_rate": 1.3631317921347564e-06, "loss": 0.0126, "step": 1153 }, { "epoch": 0.767032236623463, "grad_norm": 0.2398283391568401, "learning_rate": 1.3557469220237963e-06, "loss": 0.0118, "step": 1154 }, { "epoch": 0.7676969092721835, "grad_norm": 0.4622451475529138, "learning_rate": 1.3483789716284323e-06, "loss": 0.0082, "step": 1155 }, { "epoch": 0.768361581920904, "grad_norm": 0.33196169225105054, "learning_rate": 1.3410279751569399e-06, "loss": 0.0133, "step": 1156 }, { "epoch": 0.7690262545696245, "grad_norm": 0.1693889972824183, "learning_rate": 1.3336939667388865e-06, "loss": 0.007, "step": 1157 }, { "epoch": 0.769690927218345, "grad_norm": 0.2026575272845, "learning_rate": 1.3263769804249592e-06, "loss": 0.0077, "step": 1158 }, { "epoch": 0.7703555998670655, "grad_norm": 0.3317924315070264, "learning_rate": 1.3190770501868243e-06, "loss": 0.0122, "step": 1159 }, { "epoch": 0.771020272515786, "grad_norm": 0.2644567708590121, "learning_rate": 1.3117942099169485e-06, "loss": 0.0099, "step": 1160 }, { "epoch": 0.7716849451645065, "grad_norm": 0.5967241141158173, "learning_rate": 1.3045284934284596e-06, "loss": 0.0095, "step": 1161 }, { "epoch": 0.772349617813227, "grad_norm": 0.3065593215514762, "learning_rate": 1.297279934454978e-06, "loss": 0.0108, "step": 1162 }, { "epoch": 0.7730142904619475, "grad_norm": 0.16429767793033337, "learning_rate": 1.290048566650467e-06, "loss": 0.007, "step": 1163 }, { "epoch": 0.773678963110668, "grad_norm": 0.21304178810165386, "learning_rate": 1.2828344235890727e-06, "loss": 0.0095, "step": 1164 }, { "epoch": 0.7743436357593885, "grad_norm": 0.2560215473513446, "learning_rate": 1.2756375387649717e-06, "loss": 0.0072, "step": 1165 }, { "epoch": 0.775008308408109, "grad_norm": 0.5488560143414549, "learning_rate": 1.2684579455922086e-06, "loss": 0.0108, "step": 1166 }, { "epoch": 0.7756729810568295, "grad_norm": 0.3066562359844814, "learning_rate": 1.2612956774045486e-06, "loss": 0.0082, "step": 1167 }, { "epoch": 0.77633765370555, "grad_norm": 0.32119385817027957, "learning_rate": 1.25415076745532e-06, "loss": 0.0075, "step": 1168 }, { "epoch": 0.7770023263542705, "grad_norm": 0.2604879037873993, "learning_rate": 1.2470232489172591e-06, "loss": 0.0067, "step": 1169 }, { "epoch": 0.7776669990029911, "grad_norm": 0.24882729616305863, "learning_rate": 1.239913154882354e-06, "loss": 0.0112, "step": 1170 }, { "epoch": 0.7783316716517116, "grad_norm": 0.21009481894442797, "learning_rate": 1.2328205183616964e-06, "loss": 0.0077, "step": 1171 }, { "epoch": 0.7789963443004321, "grad_norm": 0.19419419958694167, "learning_rate": 1.2257453722853258e-06, "loss": 0.0082, "step": 1172 }, { "epoch": 0.7796610169491526, "grad_norm": 0.188812071213171, "learning_rate": 1.2186877495020705e-06, "loss": 0.0073, "step": 1173 }, { "epoch": 0.7803256895978731, "grad_norm": 0.2704368673679914, "learning_rate": 1.2116476827794104e-06, "loss": 0.0093, "step": 1174 }, { "epoch": 0.7809903622465936, "grad_norm": 0.3407872623102282, "learning_rate": 1.204625204803307e-06, "loss": 0.0089, "step": 1175 }, { "epoch": 0.7816550348953141, "grad_norm": 0.3265433343443895, "learning_rate": 1.1976203481780646e-06, "loss": 0.0108, "step": 1176 }, { "epoch": 0.7823197075440346, "grad_norm": 0.30023902808188774, "learning_rate": 1.1906331454261704e-06, "loss": 0.0105, "step": 1177 }, { "epoch": 0.7829843801927551, "grad_norm": 0.20094025972184332, "learning_rate": 1.183663628988152e-06, "loss": 0.0082, "step": 1178 }, { "epoch": 0.7836490528414756, "grad_norm": 0.15605301988534284, "learning_rate": 1.1767118312224151e-06, "loss": 0.0071, "step": 1179 }, { "epoch": 0.7843137254901961, "grad_norm": 0.3694149715312689, "learning_rate": 1.1697777844051105e-06, "loss": 0.0093, "step": 1180 }, { "epoch": 0.7849783981389166, "grad_norm": 0.4636740581877333, "learning_rate": 1.1628615207299642e-06, "loss": 0.0149, "step": 1181 }, { "epoch": 0.7856430707876371, "grad_norm": 0.1858060898847147, "learning_rate": 1.1559630723081456e-06, "loss": 0.0081, "step": 1182 }, { "epoch": 0.7863077434363576, "grad_norm": 0.29803163581129977, "learning_rate": 1.1490824711681026e-06, "loss": 0.0088, "step": 1183 }, { "epoch": 0.7869724160850781, "grad_norm": 0.3065257433728342, "learning_rate": 1.142219749255427e-06, "loss": 0.01, "step": 1184 }, { "epoch": 0.7876370887337986, "grad_norm": 0.39812607082429013, "learning_rate": 1.1353749384326984e-06, "loss": 0.0084, "step": 1185 }, { "epoch": 0.7883017613825191, "grad_norm": 0.2748580473755831, "learning_rate": 1.1285480704793378e-06, "loss": 0.0072, "step": 1186 }, { "epoch": 0.7889664340312396, "grad_norm": 0.2956252923934504, "learning_rate": 1.121739177091461e-06, "loss": 0.0133, "step": 1187 }, { "epoch": 0.7896311066799602, "grad_norm": 0.30918740567679376, "learning_rate": 1.1149482898817277e-06, "loss": 0.0063, "step": 1188 }, { "epoch": 0.7902957793286807, "grad_norm": 0.2730422456460517, "learning_rate": 1.1081754403792e-06, "loss": 0.0091, "step": 1189 }, { "epoch": 0.7909604519774012, "grad_norm": 0.31589699855317993, "learning_rate": 1.1014206600291938e-06, "loss": 0.0093, "step": 1190 }, { "epoch": 0.7916251246261217, "grad_norm": 0.24866108987599006, "learning_rate": 1.0946839801931337e-06, "loss": 0.0095, "step": 1191 }, { "epoch": 0.7922897972748422, "grad_norm": 0.3223160878548046, "learning_rate": 1.0879654321484012e-06, "loss": 0.0052, "step": 1192 }, { "epoch": 0.7929544699235627, "grad_norm": 0.3326808714284003, "learning_rate": 1.0812650470882025e-06, "loss": 0.0106, "step": 1193 }, { "epoch": 0.7936191425722832, "grad_norm": 0.2276420163603284, "learning_rate": 1.0745828561214056e-06, "loss": 0.009, "step": 1194 }, { "epoch": 0.7942838152210037, "grad_norm": 0.2392992345077749, "learning_rate": 1.067918890272419e-06, "loss": 0.0091, "step": 1195 }, { "epoch": 0.7949484878697242, "grad_norm": 0.22211020466619663, "learning_rate": 1.0612731804810228e-06, "loss": 0.01, "step": 1196 }, { "epoch": 0.7956131605184447, "grad_norm": 0.3292482115973677, "learning_rate": 1.054645757602245e-06, "loss": 0.0118, "step": 1197 }, { "epoch": 0.7962778331671652, "grad_norm": 0.6103408706172215, "learning_rate": 1.0480366524062041e-06, "loss": 0.0101, "step": 1198 }, { "epoch": 0.7969425058158857, "grad_norm": 0.3488478614465216, "learning_rate": 1.0414458955779772e-06, "loss": 0.0073, "step": 1199 }, { "epoch": 0.7976071784646062, "grad_norm": 0.28131540620029044, "learning_rate": 1.0348735177174507e-06, "loss": 0.0092, "step": 1200 }, { "epoch": 0.7982718511133267, "grad_norm": 0.18828436384580835, "learning_rate": 1.0283195493391823e-06, "loss": 0.0073, "step": 1201 }, { "epoch": 0.7989365237620472, "grad_norm": 0.24210605934382595, "learning_rate": 1.0217840208722518e-06, "loss": 0.0068, "step": 1202 }, { "epoch": 0.7996011964107677, "grad_norm": 0.25550490683459076, "learning_rate": 1.0152669626601314e-06, "loss": 0.0113, "step": 1203 }, { "epoch": 0.8002658690594882, "grad_norm": 0.17719340880485387, "learning_rate": 1.008768404960535e-06, "loss": 0.0089, "step": 1204 }, { "epoch": 0.8009305417082087, "grad_norm": 0.21500448130988142, "learning_rate": 1.0022883779452852e-06, "loss": 0.0078, "step": 1205 }, { "epoch": 0.8015952143569293, "grad_norm": 0.28112570405644205, "learning_rate": 9.958269117001635e-07, "loss": 0.0117, "step": 1206 }, { "epoch": 0.8022598870056498, "grad_norm": 0.29440783746871396, "learning_rate": 9.893840362247809e-07, "loss": 0.0176, "step": 1207 }, { "epoch": 0.8029245596543703, "grad_norm": 0.27188042308708893, "learning_rate": 9.829597814324354e-07, "loss": 0.007, "step": 1208 }, { "epoch": 0.8035892323030908, "grad_norm": 0.23038778297348897, "learning_rate": 9.76554177149966e-07, "loss": 0.0095, "step": 1209 }, { "epoch": 0.8042539049518113, "grad_norm": 0.4939674266493548, "learning_rate": 9.701672531176287e-07, "loss": 0.0102, "step": 1210 }, { "epoch": 0.8049185776005318, "grad_norm": 0.2379653531551173, "learning_rate": 9.637990389889418e-07, "loss": 0.0109, "step": 1211 }, { "epoch": 0.8055832502492523, "grad_norm": 0.24478407132403104, "learning_rate": 9.57449564330562e-07, "loss": 0.0113, "step": 1212 }, { "epoch": 0.8062479228979728, "grad_norm": 0.28349473125224656, "learning_rate": 9.511188586221376e-07, "loss": 0.0059, "step": 1213 }, { "epoch": 0.8069125955466933, "grad_norm": 0.28434605518662, "learning_rate": 9.448069512561774e-07, "loss": 0.0095, "step": 1214 }, { "epoch": 0.8075772681954138, "grad_norm": 0.16922176159982383, "learning_rate": 9.385138715379133e-07, "loss": 0.0055, "step": 1215 }, { "epoch": 0.8082419408441343, "grad_norm": 0.31299012173969426, "learning_rate": 9.322396486851626e-07, "loss": 0.0095, "step": 1216 }, { "epoch": 0.8089066134928548, "grad_norm": 0.1651906651789544, "learning_rate": 9.259843118281909e-07, "loss": 0.0074, "step": 1217 }, { "epoch": 0.8095712861415753, "grad_norm": 0.1838687063498829, "learning_rate": 9.197478900095824e-07, "loss": 0.0057, "step": 1218 }, { "epoch": 0.8102359587902958, "grad_norm": 0.5599887793109998, "learning_rate": 9.135304121840976e-07, "loss": 0.0088, "step": 1219 }, { "epoch": 0.8109006314390163, "grad_norm": 0.2270673307711792, "learning_rate": 9.073319072185466e-07, "loss": 0.0096, "step": 1220 }, { "epoch": 0.8115653040877367, "grad_norm": 0.391550489767566, "learning_rate": 9.011524038916503e-07, "loss": 0.0091, "step": 1221 }, { "epoch": 0.8122299767364572, "grad_norm": 0.22649918578616765, "learning_rate": 8.949919308939081e-07, "loss": 0.0082, "step": 1222 }, { "epoch": 0.8128946493851777, "grad_norm": 0.28731086463281247, "learning_rate": 8.888505168274625e-07, "loss": 0.0066, "step": 1223 }, { "epoch": 0.8135593220338984, "grad_norm": 0.3407894854383489, "learning_rate": 8.827281902059697e-07, "loss": 0.0094, "step": 1224 }, { "epoch": 0.8142239946826189, "grad_norm": 0.18966931868219372, "learning_rate": 8.766249794544662e-07, "loss": 0.0097, "step": 1225 }, { "epoch": 0.8148886673313394, "grad_norm": 0.23090383705250192, "learning_rate": 8.705409129092362e-07, "loss": 0.0092, "step": 1226 }, { "epoch": 0.8155533399800599, "grad_norm": 0.4149238687910955, "learning_rate": 8.644760188176815e-07, "loss": 0.0084, "step": 1227 }, { "epoch": 0.8162180126287804, "grad_norm": 0.6334986867496075, "learning_rate": 8.584303253381848e-07, "loss": 0.0132, "step": 1228 }, { "epoch": 0.8168826852775009, "grad_norm": 0.3886322626931735, "learning_rate": 8.524038605399887e-07, "loss": 0.0077, "step": 1229 }, { "epoch": 0.8175473579262214, "grad_norm": 0.5526457335937617, "learning_rate": 8.463966524030531e-07, "loss": 0.0095, "step": 1230 }, { "epoch": 0.8182120305749419, "grad_norm": 0.4247184685689392, "learning_rate": 8.404087288179425e-07, "loss": 0.0104, "step": 1231 }, { "epoch": 0.8188767032236623, "grad_norm": 0.20351069830152665, "learning_rate": 8.344401175856747e-07, "loss": 0.0085, "step": 1232 }, { "epoch": 0.8195413758723828, "grad_norm": 0.23075812657045539, "learning_rate": 8.284908464176128e-07, "loss": 0.0096, "step": 1233 }, { "epoch": 0.8202060485211033, "grad_norm": 0.39668656713147676, "learning_rate": 8.225609429353187e-07, "loss": 0.0084, "step": 1234 }, { "epoch": 0.8208707211698238, "grad_norm": 0.194637777981652, "learning_rate": 8.166504346704385e-07, "loss": 0.0103, "step": 1235 }, { "epoch": 0.8215353938185443, "grad_norm": 0.3055482179639377, "learning_rate": 8.107593490645677e-07, "loss": 0.0098, "step": 1236 }, { "epoch": 0.8222000664672648, "grad_norm": 0.1599439783111463, "learning_rate": 8.048877134691269e-07, "loss": 0.0064, "step": 1237 }, { "epoch": 0.8228647391159853, "grad_norm": 0.33615091102130434, "learning_rate": 7.990355551452294e-07, "loss": 0.0072, "step": 1238 }, { "epoch": 0.8235294117647058, "grad_norm": 0.3782897462134746, "learning_rate": 7.932029012635623e-07, "loss": 0.0089, "step": 1239 }, { "epoch": 0.8241940844134263, "grad_norm": 0.2878936740325695, "learning_rate": 7.873897789042523e-07, "loss": 0.0112, "step": 1240 }, { "epoch": 0.8248587570621468, "grad_norm": 0.4848427919599975, "learning_rate": 7.815962150567508e-07, "loss": 0.0117, "step": 1241 }, { "epoch": 0.8255234297108675, "grad_norm": 0.17647663244932313, "learning_rate": 7.758222366196955e-07, "loss": 0.0065, "step": 1242 }, { "epoch": 0.826188102359588, "grad_norm": 0.19208674990207186, "learning_rate": 7.700678704007947e-07, "loss": 0.0076, "step": 1243 }, { "epoch": 0.8268527750083084, "grad_norm": 0.27835870680776875, "learning_rate": 7.643331431167017e-07, "loss": 0.0069, "step": 1244 }, { "epoch": 0.827517447657029, "grad_norm": 0.35210132331899274, "learning_rate": 7.586180813928856e-07, "loss": 0.0098, "step": 1245 }, { "epoch": 0.8281821203057494, "grad_norm": 0.2847030336805179, "learning_rate": 7.529227117635135e-07, "loss": 0.0094, "step": 1246 }, { "epoch": 0.82884679295447, "grad_norm": 0.3595229385132859, "learning_rate": 7.47247060671325e-07, "loss": 0.007, "step": 1247 }, { "epoch": 0.8295114656031904, "grad_norm": 1.6577040072161255, "learning_rate": 7.415911544675109e-07, "loss": 0.0114, "step": 1248 }, { "epoch": 0.8301761382519109, "grad_norm": 0.25078868551860073, "learning_rate": 7.35955019411585e-07, "loss": 0.0093, "step": 1249 }, { "epoch": 0.8308408109006314, "grad_norm": 0.19820861279780877, "learning_rate": 7.303386816712715e-07, "loss": 0.0069, "step": 1250 }, { "epoch": 0.8315054835493519, "grad_norm": 0.6153166940464705, "learning_rate": 7.247421673223753e-07, "loss": 0.0093, "step": 1251 }, { "epoch": 0.8321701561980724, "grad_norm": 0.28152019188577093, "learning_rate": 7.191655023486682e-07, "loss": 0.0101, "step": 1252 }, { "epoch": 0.8328348288467929, "grad_norm": 0.38282709980068497, "learning_rate": 7.136087126417584e-07, "loss": 0.0106, "step": 1253 }, { "epoch": 0.8334995014955134, "grad_norm": 0.1985857495245599, "learning_rate": 7.080718240009826e-07, "loss": 0.0093, "step": 1254 }, { "epoch": 0.8341641741442339, "grad_norm": 0.19744125990666672, "learning_rate": 7.02554862133275e-07, "loss": 0.0069, "step": 1255 }, { "epoch": 0.8348288467929544, "grad_norm": 0.4053150874775869, "learning_rate": 6.970578526530547e-07, "loss": 0.0064, "step": 1256 }, { "epoch": 0.8354935194416749, "grad_norm": 0.1865298188473891, "learning_rate": 6.915808210821056e-07, "loss": 0.0065, "step": 1257 }, { "epoch": 0.8361581920903954, "grad_norm": 0.3459737085731713, "learning_rate": 6.86123792849458e-07, "loss": 0.0064, "step": 1258 }, { "epoch": 0.8368228647391159, "grad_norm": 0.39716978997497415, "learning_rate": 6.806867932912653e-07, "loss": 0.0079, "step": 1259 }, { "epoch": 0.8374875373878365, "grad_norm": 0.33561847883572204, "learning_rate": 6.752698476506953e-07, "loss": 0.0064, "step": 1260 }, { "epoch": 0.838152210036557, "grad_norm": 0.23923403392920115, "learning_rate": 6.698729810778065e-07, "loss": 0.0079, "step": 1261 }, { "epoch": 0.8388168826852775, "grad_norm": 0.18547366829938705, "learning_rate": 6.644962186294357e-07, "loss": 0.0075, "step": 1262 }, { "epoch": 0.839481555333998, "grad_norm": 0.20118166619585642, "learning_rate": 6.591395852690741e-07, "loss": 0.0072, "step": 1263 }, { "epoch": 0.8401462279827185, "grad_norm": 0.1421489993661855, "learning_rate": 6.53803105866761e-07, "loss": 0.0056, "step": 1264 }, { "epoch": 0.840810900631439, "grad_norm": 0.3047617695433436, "learning_rate": 6.484868051989646e-07, "loss": 0.0072, "step": 1265 }, { "epoch": 0.8414755732801595, "grad_norm": 0.34806396826696295, "learning_rate": 6.431907079484606e-07, "loss": 0.0154, "step": 1266 }, { "epoch": 0.84214024592888, "grad_norm": 0.24816193850425583, "learning_rate": 6.379148387042317e-07, "loss": 0.0107, "step": 1267 }, { "epoch": 0.8428049185776005, "grad_norm": 0.3232079151540652, "learning_rate": 6.326592219613376e-07, "loss": 0.0086, "step": 1268 }, { "epoch": 0.843469591226321, "grad_norm": 0.2212674506891991, "learning_rate": 6.274238821208129e-07, "loss": 0.01, "step": 1269 }, { "epoch": 0.8441342638750415, "grad_norm": 0.5341011828574356, "learning_rate": 6.222088434895462e-07, "loss": 0.013, "step": 1270 }, { "epoch": 0.844798936523762, "grad_norm": 0.2198757630894285, "learning_rate": 6.170141302801747e-07, "loss": 0.0084, "step": 1271 }, { "epoch": 0.8454636091724825, "grad_norm": 0.14265291148277198, "learning_rate": 6.118397666109643e-07, "loss": 0.0038, "step": 1272 }, { "epoch": 0.846128281821203, "grad_norm": 0.6171099758643466, "learning_rate": 6.066857765057055e-07, "loss": 0.0128, "step": 1273 }, { "epoch": 0.8467929544699235, "grad_norm": 0.621433132767458, "learning_rate": 6.015521838935906e-07, "loss": 0.0117, "step": 1274 }, { "epoch": 0.847457627118644, "grad_norm": 0.1713918002731256, "learning_rate": 5.964390126091163e-07, "loss": 0.0059, "step": 1275 }, { "epoch": 0.8481222997673645, "grad_norm": 0.21449463464160287, "learning_rate": 5.9134628639196e-07, "loss": 0.007, "step": 1276 }, { "epoch": 0.848786972416085, "grad_norm": 0.3385776079983938, "learning_rate": 5.862740288868829e-07, "loss": 0.0097, "step": 1277 }, { "epoch": 0.8494516450648056, "grad_norm": 0.1751389796927551, "learning_rate": 5.812222636436055e-07, "loss": 0.0059, "step": 1278 }, { "epoch": 0.8501163177135261, "grad_norm": 0.29202634371041664, "learning_rate": 5.76191014116711e-07, "loss": 0.0133, "step": 1279 }, { "epoch": 0.8507809903622466, "grad_norm": 0.36860647674214575, "learning_rate": 5.711803036655306e-07, "loss": 0.0087, "step": 1280 }, { "epoch": 0.8514456630109671, "grad_norm": 0.3484876300628792, "learning_rate": 5.661901555540328e-07, "loss": 0.012, "step": 1281 }, { "epoch": 0.8521103356596876, "grad_norm": 0.31367930557500423, "learning_rate": 5.612205929507209e-07, "loss": 0.0126, "step": 1282 }, { "epoch": 0.8527750083084081, "grad_norm": 0.45161498857266796, "learning_rate": 5.562716389285216e-07, "loss": 0.0115, "step": 1283 }, { "epoch": 0.8534396809571286, "grad_norm": 0.1788671259620612, "learning_rate": 5.513433164646815e-07, "loss": 0.0089, "step": 1284 }, { "epoch": 0.8541043536058491, "grad_norm": 0.4257909245869742, "learning_rate": 5.464356484406535e-07, "loss": 0.0117, "step": 1285 }, { "epoch": 0.8547690262545696, "grad_norm": 0.5959690806932872, "learning_rate": 5.415486576419992e-07, "loss": 0.0098, "step": 1286 }, { "epoch": 0.8554336989032901, "grad_norm": 0.36435381860833504, "learning_rate": 5.366823667582766e-07, "loss": 0.0065, "step": 1287 }, { "epoch": 0.8560983715520106, "grad_norm": 0.19623785183599504, "learning_rate": 5.318367983829393e-07, "loss": 0.0078, "step": 1288 }, { "epoch": 0.8567630442007311, "grad_norm": 0.4733871562296425, "learning_rate": 5.270119750132258e-07, "loss": 0.0103, "step": 1289 }, { "epoch": 0.8574277168494516, "grad_norm": 0.23129700586426916, "learning_rate": 5.222079190500634e-07, "loss": 0.0058, "step": 1290 }, { "epoch": 0.8580923894981721, "grad_norm": 0.191226124199867, "learning_rate": 5.174246527979532e-07, "loss": 0.0077, "step": 1291 }, { "epoch": 0.8587570621468926, "grad_norm": 0.37195425247389013, "learning_rate": 5.126621984648816e-07, "loss": 0.0135, "step": 1292 }, { "epoch": 0.8594217347956131, "grad_norm": 0.18690135945880493, "learning_rate": 5.079205781621999e-07, "loss": 0.0093, "step": 1293 }, { "epoch": 0.8600864074443336, "grad_norm": 0.22681216531688878, "learning_rate": 5.031998139045352e-07, "loss": 0.0077, "step": 1294 }, { "epoch": 0.8607510800930541, "grad_norm": 0.20037186847646823, "learning_rate": 4.984999276096803e-07, "loss": 0.0086, "step": 1295 }, { "epoch": 0.8614157527417746, "grad_norm": 0.21114773911943185, "learning_rate": 4.938209410984973e-07, "loss": 0.0086, "step": 1296 }, { "epoch": 0.8620804253904952, "grad_norm": 0.4411104511911419, "learning_rate": 4.891628760948114e-07, "loss": 0.0075, "step": 1297 }, { "epoch": 0.8627450980392157, "grad_norm": 0.28957408663001083, "learning_rate": 4.845257542253156e-07, "loss": 0.0085, "step": 1298 }, { "epoch": 0.8634097706879362, "grad_norm": 0.23438971826970809, "learning_rate": 4.799095970194628e-07, "loss": 0.0067, "step": 1299 }, { "epoch": 0.8640744433366567, "grad_norm": 0.587154153978012, "learning_rate": 4.753144259093734e-07, "loss": 0.0117, "step": 1300 }, { "epoch": 0.8647391159853772, "grad_norm": 0.25847473299696, "learning_rate": 4.707402622297319e-07, "loss": 0.0071, "step": 1301 }, { "epoch": 0.8654037886340977, "grad_norm": 0.21226560471973854, "learning_rate": 4.6618712721768654e-07, "loss": 0.0094, "step": 1302 }, { "epoch": 0.8660684612828182, "grad_norm": 0.22453730774857814, "learning_rate": 4.6165504201275635e-07, "loss": 0.006, "step": 1303 }, { "epoch": 0.8667331339315387, "grad_norm": 0.2858523502402271, "learning_rate": 4.5714402765672574e-07, "loss": 0.005, "step": 1304 }, { "epoch": 0.8673978065802592, "grad_norm": 0.2105900454562083, "learning_rate": 4.526541050935529e-07, "loss": 0.0107, "step": 1305 }, { "epoch": 0.8680624792289797, "grad_norm": 0.2504403695016181, "learning_rate": 4.481852951692672e-07, "loss": 0.0097, "step": 1306 }, { "epoch": 0.8687271518777002, "grad_norm": 0.3168145106008975, "learning_rate": 4.437376186318776e-07, "loss": 0.0113, "step": 1307 }, { "epoch": 0.8693918245264207, "grad_norm": 0.13482607013080378, "learning_rate": 4.393110961312735e-07, "loss": 0.0057, "step": 1308 }, { "epoch": 0.8700564971751412, "grad_norm": 0.30057439803737657, "learning_rate": 4.349057482191299e-07, "loss": 0.0088, "step": 1309 }, { "epoch": 0.8707211698238617, "grad_norm": 0.43314707886578907, "learning_rate": 4.30521595348809e-07, "loss": 0.0079, "step": 1310 }, { "epoch": 0.8713858424725822, "grad_norm": 0.19563364889253065, "learning_rate": 4.261586578752702e-07, "loss": 0.0093, "step": 1311 }, { "epoch": 0.8720505151213027, "grad_norm": 0.3859456161811097, "learning_rate": 4.2181695605497066e-07, "loss": 0.0096, "step": 1312 }, { "epoch": 0.8727151877700232, "grad_norm": 0.21364472633151707, "learning_rate": 4.1749651004577675e-07, "loss": 0.0078, "step": 1313 }, { "epoch": 0.8733798604187437, "grad_norm": 0.26177107889370144, "learning_rate": 4.131973399068645e-07, "loss": 0.0066, "step": 1314 }, { "epoch": 0.8740445330674643, "grad_norm": 0.2831138711053405, "learning_rate": 4.089194655986306e-07, "loss": 0.0073, "step": 1315 }, { "epoch": 0.8747092057161848, "grad_norm": 0.3644255305067573, "learning_rate": 4.046629069825975e-07, "loss": 0.01, "step": 1316 }, { "epoch": 0.8753738783649053, "grad_norm": 0.4101294175243439, "learning_rate": 4.0042768382132333e-07, "loss": 0.0077, "step": 1317 }, { "epoch": 0.8760385510136258, "grad_norm": 0.24609740940670535, "learning_rate": 3.9621381577830855e-07, "loss": 0.0082, "step": 1318 }, { "epoch": 0.8767032236623463, "grad_norm": 0.4000227377785715, "learning_rate": 3.9202132241790426e-07, "loss": 0.0091, "step": 1319 }, { "epoch": 0.8773678963110668, "grad_norm": 0.221223645188277, "learning_rate": 3.8785022320522394e-07, "loss": 0.0112, "step": 1320 }, { "epoch": 0.8780325689597873, "grad_norm": 0.3032336344068287, "learning_rate": 3.837005375060482e-07, "loss": 0.0089, "step": 1321 }, { "epoch": 0.8786972416085078, "grad_norm": 0.6465774259720459, "learning_rate": 3.7957228458674055e-07, "loss": 0.0105, "step": 1322 }, { "epoch": 0.8793619142572283, "grad_norm": 0.31190572399260413, "learning_rate": 3.7546548361415525e-07, "loss": 0.0053, "step": 1323 }, { "epoch": 0.8800265869059488, "grad_norm": 0.36312040480494495, "learning_rate": 3.7138015365554834e-07, "loss": 0.011, "step": 1324 }, { "epoch": 0.8806912595546693, "grad_norm": 0.1713100331318772, "learning_rate": 3.6731631367848684e-07, "loss": 0.0068, "step": 1325 }, { "epoch": 0.8813559322033898, "grad_norm": 0.2877925980140741, "learning_rate": 3.632739825507664e-07, "loss": 0.0108, "step": 1326 }, { "epoch": 0.8820206048521103, "grad_norm": 0.28339102121317306, "learning_rate": 3.592531790403159e-07, "loss": 0.0112, "step": 1327 }, { "epoch": 0.8826852775008308, "grad_norm": 0.255632244408717, "learning_rate": 3.5525392181512187e-07, "loss": 0.009, "step": 1328 }, { "epoch": 0.8833499501495513, "grad_norm": 0.31000803010277955, "learning_rate": 3.512762294431271e-07, "loss": 0.0089, "step": 1329 }, { "epoch": 0.8840146227982718, "grad_norm": 0.5356642520907057, "learning_rate": 3.473201203921578e-07, "loss": 0.012, "step": 1330 }, { "epoch": 0.8846792954469923, "grad_norm": 0.18440426000235735, "learning_rate": 3.4338561302982745e-07, "loss": 0.0071, "step": 1331 }, { "epoch": 0.8853439680957128, "grad_norm": 0.12540041819825556, "learning_rate": 3.3947272562345923e-07, "loss": 0.004, "step": 1332 }, { "epoch": 0.8860086407444334, "grad_norm": 0.4250985409615803, "learning_rate": 3.355814763399973e-07, "loss": 0.0084, "step": 1333 }, { "epoch": 0.8866733133931539, "grad_norm": 0.3623646451327974, "learning_rate": 3.317118832459243e-07, "loss": 0.0126, "step": 1334 }, { "epoch": 0.8873379860418744, "grad_norm": 0.360135238023425, "learning_rate": 3.278639643071735e-07, "loss": 0.0113, "step": 1335 }, { "epoch": 0.8880026586905949, "grad_norm": 0.32952721504108706, "learning_rate": 3.2403773738905185e-07, "loss": 0.0101, "step": 1336 }, { "epoch": 0.8886673313393154, "grad_norm": 0.2667603702145041, "learning_rate": 3.2023322025615143e-07, "loss": 0.0107, "step": 1337 }, { "epoch": 0.8893320039880359, "grad_norm": 0.28247791552348844, "learning_rate": 3.164504305722688e-07, "loss": 0.0129, "step": 1338 }, { "epoch": 0.8899966766367564, "grad_norm": 0.15625646332724677, "learning_rate": 3.1268938590032495e-07, "loss": 0.0062, "step": 1339 }, { "epoch": 0.8906613492854769, "grad_norm": 0.5060739242699384, "learning_rate": 3.0895010370228076e-07, "loss": 0.0101, "step": 1340 }, { "epoch": 0.8913260219341974, "grad_norm": 0.21446324793899632, "learning_rate": 3.0523260133905887e-07, "loss": 0.008, "step": 1341 }, { "epoch": 0.8919906945829179, "grad_norm": 0.2022900654333179, "learning_rate": 3.015368960704584e-07, "loss": 0.0088, "step": 1342 }, { "epoch": 0.8926553672316384, "grad_norm": 0.5141386462050389, "learning_rate": 2.9786300505508016e-07, "loss": 0.0077, "step": 1343 }, { "epoch": 0.8933200398803589, "grad_norm": 0.25142091580604636, "learning_rate": 2.942109453502451e-07, "loss": 0.0076, "step": 1344 }, { "epoch": 0.8939847125290794, "grad_norm": 0.43794909917261327, "learning_rate": 2.905807339119138e-07, "loss": 0.0084, "step": 1345 }, { "epoch": 0.8946493851777999, "grad_norm": 0.16340568299048047, "learning_rate": 2.8697238759460676e-07, "loss": 0.006, "step": 1346 }, { "epoch": 0.8953140578265204, "grad_norm": 0.23454354844351744, "learning_rate": 2.8338592315133253e-07, "loss": 0.0073, "step": 1347 }, { "epoch": 0.8959787304752409, "grad_norm": 0.39274982686886334, "learning_rate": 2.798213572335001e-07, "loss": 0.0096, "step": 1348 }, { "epoch": 0.8966434031239614, "grad_norm": 0.1539465320402893, "learning_rate": 2.762787063908523e-07, "loss": 0.0047, "step": 1349 }, { "epoch": 0.8973080757726819, "grad_norm": 0.5804042908524948, "learning_rate": 2.727579870713787e-07, "loss": 0.0118, "step": 1350 }, { "epoch": 0.8979727484214025, "grad_norm": 0.3804620515339711, "learning_rate": 2.6925921562124867e-07, "loss": 0.0087, "step": 1351 }, { "epoch": 0.898637421070123, "grad_norm": 0.5259100625593183, "learning_rate": 2.6578240828472604e-07, "loss": 0.0102, "step": 1352 }, { "epoch": 0.8993020937188435, "grad_norm": 0.21823229800828492, "learning_rate": 2.623275812041026e-07, "loss": 0.0082, "step": 1353 }, { "epoch": 0.899966766367564, "grad_norm": 0.2206204294695038, "learning_rate": 2.5889475041961767e-07, "loss": 0.0093, "step": 1354 }, { "epoch": 0.9006314390162845, "grad_norm": 0.6210681161404372, "learning_rate": 2.554839318693847e-07, "loss": 0.0097, "step": 1355 }, { "epoch": 0.901296111665005, "grad_norm": 0.23288096465397104, "learning_rate": 2.520951413893191e-07, "loss": 0.0086, "step": 1356 }, { "epoch": 0.9019607843137255, "grad_norm": 0.17222888697258087, "learning_rate": 2.487283947130609e-07, "loss": 0.0056, "step": 1357 }, { "epoch": 0.902625456962446, "grad_norm": 0.18003467949614746, "learning_rate": 2.453837074719062e-07, "loss": 0.0065, "step": 1358 }, { "epoch": 0.9032901296111665, "grad_norm": 0.4820087740609228, "learning_rate": 2.4206109519473164e-07, "loss": 0.0097, "step": 1359 }, { "epoch": 0.903954802259887, "grad_norm": 0.2323492870688366, "learning_rate": 2.3876057330792344e-07, "loss": 0.0084, "step": 1360 }, { "epoch": 0.9046194749086075, "grad_norm": 0.24413854187061568, "learning_rate": 2.3548215713530497e-07, "loss": 0.0074, "step": 1361 }, { "epoch": 0.905284147557328, "grad_norm": 0.22817527462630224, "learning_rate": 2.3222586189806716e-07, "loss": 0.0077, "step": 1362 }, { "epoch": 0.9059488202060485, "grad_norm": 0.25553386602777856, "learning_rate": 2.289917027146943e-07, "loss": 0.0068, "step": 1363 }, { "epoch": 0.906613492854769, "grad_norm": 0.2775593712558632, "learning_rate": 2.257796946009e-07, "loss": 0.0055, "step": 1364 }, { "epoch": 0.9072781655034895, "grad_norm": 0.41755011185783786, "learning_rate": 2.2258985246955078e-07, "loss": 0.0091, "step": 1365 }, { "epoch": 0.90794283815221, "grad_norm": 0.25570708270401765, "learning_rate": 2.1942219113060215e-07, "loss": 0.0099, "step": 1366 }, { "epoch": 0.9086075108009305, "grad_norm": 0.2924731133048024, "learning_rate": 2.1627672529102472e-07, "loss": 0.0069, "step": 1367 }, { "epoch": 0.909272183449651, "grad_norm": 0.23952324922004076, "learning_rate": 2.1315346955474102e-07, "loss": 0.0072, "step": 1368 }, { "epoch": 0.9099368560983716, "grad_norm": 0.3143770898780771, "learning_rate": 2.1005243842255552e-07, "loss": 0.0096, "step": 1369 }, { "epoch": 0.9106015287470921, "grad_norm": 0.32636204895537685, "learning_rate": 2.0697364629208626e-07, "loss": 0.0121, "step": 1370 }, { "epoch": 0.9112662013958126, "grad_norm": 0.2342022665142366, "learning_rate": 2.03917107457699e-07, "loss": 0.0103, "step": 1371 }, { "epoch": 0.9119308740445331, "grad_norm": 0.300998863801611, "learning_rate": 2.0088283611044034e-07, "loss": 0.0129, "step": 1372 }, { "epoch": 0.9125955466932536, "grad_norm": 0.18985213257992264, "learning_rate": 1.978708463379747e-07, "loss": 0.0084, "step": 1373 }, { "epoch": 0.9132602193419741, "grad_norm": 0.47628406829102193, "learning_rate": 1.9488115212451308e-07, "loss": 0.0095, "step": 1374 }, { "epoch": 0.9139248919906946, "grad_norm": 0.32323123793718106, "learning_rate": 1.919137673507543e-07, "loss": 0.0097, "step": 1375 }, { "epoch": 0.9145895646394151, "grad_norm": 0.23907467342245894, "learning_rate": 1.8896870579381666e-07, "loss": 0.008, "step": 1376 }, { "epoch": 0.9152542372881356, "grad_norm": 0.19280814398846127, "learning_rate": 1.860459811271753e-07, "loss": 0.0089, "step": 1377 }, { "epoch": 0.9159189099368561, "grad_norm": 0.24176747745973787, "learning_rate": 1.8314560692059836e-07, "loss": 0.0073, "step": 1378 }, { "epoch": 0.9165835825855766, "grad_norm": 0.2334961645283537, "learning_rate": 1.8026759664008464e-07, "loss": 0.0084, "step": 1379 }, { "epoch": 0.9172482552342971, "grad_norm": 0.21786995305611914, "learning_rate": 1.7741196364780055e-07, "loss": 0.0085, "step": 1380 }, { "epoch": 0.9179129278830176, "grad_norm": 0.2840523898409003, "learning_rate": 1.745787212020178e-07, "loss": 0.0108, "step": 1381 }, { "epoch": 0.9185776005317381, "grad_norm": 0.226048142965821, "learning_rate": 1.7176788245705178e-07, "loss": 0.0092, "step": 1382 }, { "epoch": 0.9192422731804586, "grad_norm": 0.35007830940392926, "learning_rate": 1.689794604632028e-07, "loss": 0.0145, "step": 1383 }, { "epoch": 0.9199069458291791, "grad_norm": 0.22762583575066672, "learning_rate": 1.6621346816668993e-07, "loss": 0.0083, "step": 1384 }, { "epoch": 0.9205716184778996, "grad_norm": 0.4335184335860766, "learning_rate": 1.6346991840959947e-07, "loss": 0.0116, "step": 1385 }, { "epoch": 0.9212362911266201, "grad_norm": 0.30726518900936856, "learning_rate": 1.6074882392981604e-07, "loss": 0.0081, "step": 1386 }, { "epoch": 0.9219009637753407, "grad_norm": 0.2130389539698074, "learning_rate": 1.5805019736097105e-07, "loss": 0.0066, "step": 1387 }, { "epoch": 0.9225656364240612, "grad_norm": 0.22518910562095648, "learning_rate": 1.5537405123237758e-07, "loss": 0.0095, "step": 1388 }, { "epoch": 0.9232303090727817, "grad_norm": 0.2644571085269553, "learning_rate": 1.5272039796897786e-07, "loss": 0.0074, "step": 1389 }, { "epoch": 0.9238949817215022, "grad_norm": 0.22407464278154024, "learning_rate": 1.500892498912826e-07, "loss": 0.0084, "step": 1390 }, { "epoch": 0.9245596543702227, "grad_norm": 0.2715702285577862, "learning_rate": 1.4748061921531497e-07, "loss": 0.0093, "step": 1391 }, { "epoch": 0.9252243270189432, "grad_norm": 0.44775054525017477, "learning_rate": 1.4489451805255072e-07, "loss": 0.0094, "step": 1392 }, { "epoch": 0.9258889996676637, "grad_norm": 0.3120170865886277, "learning_rate": 1.4233095840986756e-07, "loss": 0.0135, "step": 1393 }, { "epoch": 0.9265536723163842, "grad_norm": 0.5631823125245807, "learning_rate": 1.397899521894841e-07, "loss": 0.0078, "step": 1394 }, { "epoch": 0.9272183449651047, "grad_norm": 0.25260723896350507, "learning_rate": 1.3727151118890834e-07, "loss": 0.0069, "step": 1395 }, { "epoch": 0.9278830176138252, "grad_norm": 0.25514417133119477, "learning_rate": 1.3477564710088097e-07, "loss": 0.0109, "step": 1396 }, { "epoch": 0.9285476902625457, "grad_norm": 0.2718088879890349, "learning_rate": 1.3230237151332037e-07, "loss": 0.0096, "step": 1397 }, { "epoch": 0.9292123629112662, "grad_norm": 0.2709774446294237, "learning_rate": 1.2985169590927216e-07, "loss": 0.0076, "step": 1398 }, { "epoch": 0.9298770355599867, "grad_norm": 0.27971714331010683, "learning_rate": 1.2742363166685035e-07, "loss": 0.0105, "step": 1399 }, { "epoch": 0.9305417082087072, "grad_norm": 0.26093422463096994, "learning_rate": 1.250181900591918e-07, "loss": 0.0079, "step": 1400 }, { "epoch": 0.9312063808574277, "grad_norm": 0.40641911549379955, "learning_rate": 1.2263538225439686e-07, "loss": 0.0103, "step": 1401 }, { "epoch": 0.9318710535061482, "grad_norm": 0.4054184737547687, "learning_rate": 1.2027521931548214e-07, "loss": 0.0087, "step": 1402 }, { "epoch": 0.9325357261548687, "grad_norm": 0.26579166946650856, "learning_rate": 1.1793771220032669e-07, "loss": 0.0109, "step": 1403 }, { "epoch": 0.9332003988035892, "grad_norm": 0.2191487473782411, "learning_rate": 1.1562287176162201e-07, "loss": 0.0064, "step": 1404 }, { "epoch": 0.9338650714523098, "grad_norm": 0.22438879883955717, "learning_rate": 1.1333070874682217e-07, "loss": 0.0097, "step": 1405 }, { "epoch": 0.9345297441010303, "grad_norm": 0.3468025813553504, "learning_rate": 1.110612337980943e-07, "loss": 0.01, "step": 1406 }, { "epoch": 0.9351944167497508, "grad_norm": 0.4825662416744784, "learning_rate": 1.0881445745226538e-07, "loss": 0.0109, "step": 1407 }, { "epoch": 0.9358590893984713, "grad_norm": 0.3282549185507877, "learning_rate": 1.0659039014077943e-07, "loss": 0.0061, "step": 1408 }, { "epoch": 0.9365237620471918, "grad_norm": 0.24629567486926288, "learning_rate": 1.0438904218964319e-07, "loss": 0.0097, "step": 1409 }, { "epoch": 0.9371884346959123, "grad_norm": 0.4215719043805905, "learning_rate": 1.0221042381938329e-07, "loss": 0.0088, "step": 1410 }, { "epoch": 0.9378531073446328, "grad_norm": 0.19212652672923897, "learning_rate": 1.0005454514499413e-07, "loss": 0.0094, "step": 1411 }, { "epoch": 0.9385177799933533, "grad_norm": 0.38929297760718345, "learning_rate": 9.792141617589512e-08, "loss": 0.0069, "step": 1412 }, { "epoch": 0.9391824526420738, "grad_norm": 0.38410631274492574, "learning_rate": 9.581104681588072e-08, "loss": 0.0088, "step": 1413 }, { "epoch": 0.9398471252907943, "grad_norm": 0.2713883419731266, "learning_rate": 9.372344686307655e-08, "loss": 0.0096, "step": 1414 }, { "epoch": 0.9405117979395148, "grad_norm": 0.2749479294928266, "learning_rate": 9.165862600989339e-08, "loss": 0.0091, "step": 1415 }, { "epoch": 0.9411764705882353, "grad_norm": 0.13997930958187876, "learning_rate": 8.961659384298215e-08, "loss": 0.0058, "step": 1416 }, { "epoch": 0.9418411432369558, "grad_norm": 0.19156522457850245, "learning_rate": 8.759735984318896e-08, "loss": 0.0093, "step": 1417 }, { "epoch": 0.9425058158856763, "grad_norm": 0.48891629721262814, "learning_rate": 8.560093338551178e-08, "loss": 0.0106, "step": 1418 }, { "epoch": 0.9431704885343968, "grad_norm": 0.39922897974582106, "learning_rate": 8.362732373905724e-08, "loss": 0.0098, "step": 1419 }, { "epoch": 0.9438351611831173, "grad_norm": 0.21109107434978996, "learning_rate": 8.167654006699444e-08, "loss": 0.0069, "step": 1420 }, { "epoch": 0.9444998338318378, "grad_norm": 0.2332346634114929, "learning_rate": 7.974859142651892e-08, "loss": 0.01, "step": 1421 }, { "epoch": 0.9451645064805583, "grad_norm": 0.5853964147599253, "learning_rate": 7.784348676880271e-08, "loss": 0.0107, "step": 1422 }, { "epoch": 0.9458291791292789, "grad_norm": 0.17128230932959967, "learning_rate": 7.59612349389599e-08, "loss": 0.0082, "step": 1423 }, { "epoch": 0.9464938517779994, "grad_norm": 0.2938304819216332, "learning_rate": 7.410184467600002e-08, "loss": 0.0098, "step": 1424 }, { "epoch": 0.9471585244267199, "grad_norm": 0.29266126267822085, "learning_rate": 7.226532461279079e-08, "loss": 0.0103, "step": 1425 }, { "epoch": 0.9478231970754404, "grad_norm": 0.3485122648409076, "learning_rate": 7.04516832760177e-08, "loss": 0.0078, "step": 1426 }, { "epoch": 0.9484878697241609, "grad_norm": 0.3596435476197589, "learning_rate": 6.86609290861423e-08, "loss": 0.0069, "step": 1427 }, { "epoch": 0.9491525423728814, "grad_norm": 0.4972666140126574, "learning_rate": 6.689307035736503e-08, "loss": 0.0108, "step": 1428 }, { "epoch": 0.9498172150216019, "grad_norm": 0.22502294434114065, "learning_rate": 6.514811529758747e-08, "loss": 0.0109, "step": 1429 }, { "epoch": 0.9504818876703224, "grad_norm": 0.2188354845793532, "learning_rate": 6.34260720083707e-08, "loss": 0.0091, "step": 1430 }, { "epoch": 0.9511465603190429, "grad_norm": 0.2117565630575369, "learning_rate": 6.172694848490201e-08, "loss": 0.0117, "step": 1431 }, { "epoch": 0.9518112329677634, "grad_norm": 0.16143261780899512, "learning_rate": 6.005075261595495e-08, "loss": 0.008, "step": 1432 }, { "epoch": 0.9524759056164839, "grad_norm": 0.25476203410232795, "learning_rate": 5.8397492183853155e-08, "loss": 0.0057, "step": 1433 }, { "epoch": 0.9531405782652044, "grad_norm": 0.2818271954806162, "learning_rate": 5.6767174864434396e-08, "loss": 0.009, "step": 1434 }, { "epoch": 0.9538052509139249, "grad_norm": 0.24237909017799808, "learning_rate": 5.515980822701439e-08, "loss": 0.0101, "step": 1435 }, { "epoch": 0.9544699235626454, "grad_norm": 0.3552432189827661, "learning_rate": 5.357539973435522e-08, "loss": 0.0101, "step": 1436 }, { "epoch": 0.9551345962113659, "grad_norm": 0.23186978865636249, "learning_rate": 5.201395674262366e-08, "loss": 0.0051, "step": 1437 }, { "epoch": 0.9557992688600864, "grad_norm": 0.41374112140120983, "learning_rate": 5.047548650136513e-08, "loss": 0.0072, "step": 1438 }, { "epoch": 0.9564639415088069, "grad_norm": 0.5024415191209991, "learning_rate": 4.895999615346314e-08, "loss": 0.0108, "step": 1439 }, { "epoch": 0.9571286141575274, "grad_norm": 0.2796273839118238, "learning_rate": 4.746749273510987e-08, "loss": 0.0071, "step": 1440 }, { "epoch": 0.9577932868062479, "grad_norm": 0.3035503697430027, "learning_rate": 4.599798317577342e-08, "loss": 0.0072, "step": 1441 }, { "epoch": 0.9584579594549685, "grad_norm": 0.35090412706542284, "learning_rate": 4.455147429816453e-08, "loss": 0.0071, "step": 1442 }, { "epoch": 0.959122632103689, "grad_norm": 0.22677588589423187, "learning_rate": 4.31279728182038e-08, "loss": 0.0075, "step": 1443 }, { "epoch": 0.9597873047524095, "grad_norm": 0.461202717330117, "learning_rate": 4.172748534499449e-08, "loss": 0.0087, "step": 1444 }, { "epoch": 0.96045197740113, "grad_norm": 0.4009475426759403, "learning_rate": 4.0350018380787e-08, "loss": 0.0105, "step": 1445 }, { "epoch": 0.9611166500498505, "grad_norm": 0.22013580162867818, "learning_rate": 3.89955783209528e-08, "loss": 0.0092, "step": 1446 }, { "epoch": 0.961781322698571, "grad_norm": 0.26806002369562965, "learning_rate": 3.766417145395218e-08, "loss": 0.0109, "step": 1447 }, { "epoch": 0.9624459953472915, "grad_norm": 0.2684175058923534, "learning_rate": 3.635580396130489e-08, "loss": 0.0074, "step": 1448 }, { "epoch": 0.963110667996012, "grad_norm": 0.5228750139796844, "learning_rate": 3.507048191756401e-08, "loss": 0.0099, "step": 1449 }, { "epoch": 0.9637753406447325, "grad_norm": 0.40948934869786435, "learning_rate": 3.3808211290284886e-08, "loss": 0.0103, "step": 1450 }, { "epoch": 0.964440013293453, "grad_norm": 0.4762124447868586, "learning_rate": 3.256899793999957e-08, "loss": 0.0054, "step": 1451 }, { "epoch": 0.9651046859421735, "grad_norm": 0.24857389167204513, "learning_rate": 3.135284762018742e-08, "loss": 0.007, "step": 1452 }, { "epoch": 0.965769358590894, "grad_norm": 0.27281111676577097, "learning_rate": 3.015976597725068e-08, "loss": 0.0127, "step": 1453 }, { "epoch": 0.9664340312396145, "grad_norm": 0.2845804129461655, "learning_rate": 2.8989758550487244e-08, "loss": 0.0109, "step": 1454 }, { "epoch": 0.967098703888335, "grad_norm": 0.26838551477963307, "learning_rate": 2.7842830772064068e-08, "loss": 0.012, "step": 1455 }, { "epoch": 0.9677633765370555, "grad_norm": 0.4721587194708573, "learning_rate": 2.6718987966992683e-08, "loss": 0.0069, "step": 1456 }, { "epoch": 0.968428049185776, "grad_norm": 0.2225663113950291, "learning_rate": 2.561823535310648e-08, "loss": 0.0079, "step": 1457 }, { "epoch": 0.9690927218344965, "grad_norm": 0.27146373608555935, "learning_rate": 2.4540578041032383e-08, "loss": 0.0074, "step": 1458 }, { "epoch": 0.969757394483217, "grad_norm": 0.21599098741222889, "learning_rate": 2.3486021034170857e-08, "loss": 0.0111, "step": 1459 }, { "epoch": 0.9704220671319376, "grad_norm": 0.3856402402760231, "learning_rate": 2.2454569228670398e-08, "loss": 0.0106, "step": 1460 }, { "epoch": 0.9710867397806581, "grad_norm": 0.33093801802024314, "learning_rate": 2.1446227413404742e-08, "loss": 0.0091, "step": 1461 }, { "epoch": 0.9717514124293786, "grad_norm": 0.2520678371267694, "learning_rate": 2.0461000269953457e-08, "loss": 0.0081, "step": 1462 }, { "epoch": 0.9724160850780991, "grad_norm": 0.410541684100847, "learning_rate": 1.9498892372576405e-08, "loss": 0.008, "step": 1463 }, { "epoch": 0.9730807577268196, "grad_norm": 0.2654159944302283, "learning_rate": 1.855990818819542e-08, "loss": 0.0095, "step": 1464 }, { "epoch": 0.9737454303755401, "grad_norm": 0.33137468500833794, "learning_rate": 1.7644052076371544e-08, "loss": 0.0058, "step": 1465 }, { "epoch": 0.9744101030242606, "grad_norm": 0.4089555121812419, "learning_rate": 1.6751328289286717e-08, "loss": 0.0139, "step": 1466 }, { "epoch": 0.9750747756729811, "grad_norm": 0.2783833078816801, "learning_rate": 1.5881740971722126e-08, "loss": 0.01, "step": 1467 }, { "epoch": 0.9757394483217016, "grad_norm": 0.20684165854676587, "learning_rate": 1.5035294161039882e-08, "loss": 0.0031, "step": 1468 }, { "epoch": 0.9764041209704221, "grad_norm": 0.2754255644912798, "learning_rate": 1.4211991787164148e-08, "loss": 0.0078, "step": 1469 }, { "epoch": 0.9770687936191426, "grad_norm": 0.2182609254435047, "learning_rate": 1.3411837672564488e-08, "loss": 0.0065, "step": 1470 }, { "epoch": 0.9777334662678631, "grad_norm": 0.41637997230927803, "learning_rate": 1.2634835532233658e-08, "loss": 0.0094, "step": 1471 }, { "epoch": 0.9783981389165836, "grad_norm": 0.21084932780377647, "learning_rate": 1.1880988973676511e-08, "loss": 0.0071, "step": 1472 }, { "epoch": 0.9790628115653041, "grad_norm": 0.3630598726797507, "learning_rate": 1.1150301496888338e-08, "loss": 0.0106, "step": 1473 }, { "epoch": 0.9797274842140246, "grad_norm": 0.2230323874514755, "learning_rate": 1.044277649433989e-08, "loss": 0.0089, "step": 1474 }, { "epoch": 0.9803921568627451, "grad_norm": 0.6772716764451815, "learning_rate": 9.758417250962382e-09, "loss": 0.0115, "step": 1475 }, { "epoch": 0.9810568295114656, "grad_norm": 0.17428103883165974, "learning_rate": 9.09722694413362e-09, "loss": 0.0066, "step": 1476 }, { "epoch": 0.981721502160186, "grad_norm": 0.20398655612179167, "learning_rate": 8.459208643659122e-09, "loss": 0.0075, "step": 1477 }, { "epoch": 0.9823861748089067, "grad_norm": 0.492245237759995, "learning_rate": 7.844365311761581e-09, "loss": 0.0089, "step": 1478 }, { "epoch": 0.9830508474576272, "grad_norm": 0.2603920276555642, "learning_rate": 7.2526998030653104e-09, "loss": 0.009, "step": 1479 }, { "epoch": 0.9837155201063477, "grad_norm": 0.2303936415465965, "learning_rate": 6.6842148645840374e-09, "loss": 0.0074, "step": 1480 }, { "epoch": 0.9843801927550682, "grad_norm": 0.42265637248630394, "learning_rate": 6.1389131357070254e-09, "loss": 0.0074, "step": 1481 }, { "epoch": 0.9850448654037887, "grad_norm": 0.297536803451014, "learning_rate": 5.616797148187969e-09, "loss": 0.0089, "step": 1482 }, { "epoch": 0.9857095380525092, "grad_norm": 0.168931548680635, "learning_rate": 5.11786932613223e-09, "loss": 0.0063, "step": 1483 }, { "epoch": 0.9863742107012297, "grad_norm": 0.20287090250367135, "learning_rate": 4.642131985986287e-09, "loss": 0.0059, "step": 1484 }, { "epoch": 0.9870388833499502, "grad_norm": 0.40582488934168354, "learning_rate": 4.189587336526635e-09, "loss": 0.0116, "step": 1485 }, { "epoch": 0.9877035559986707, "grad_norm": 0.27364927584003845, "learning_rate": 3.760237478849793e-09, "loss": 0.0086, "step": 1486 }, { "epoch": 0.9883682286473912, "grad_norm": 0.6309116441674326, "learning_rate": 3.3540844063617574e-09, "loss": 0.0118, "step": 1487 }, { "epoch": 0.9890329012961117, "grad_norm": 0.30530064132867457, "learning_rate": 2.9711300047696734e-09, "loss": 0.0071, "step": 1488 }, { "epoch": 0.9896975739448322, "grad_norm": 0.35560754624600943, "learning_rate": 2.611376052073511e-09, "loss": 0.0094, "step": 1489 }, { "epoch": 0.9903622465935527, "grad_norm": 0.31990540878998597, "learning_rate": 2.274824218556071e-09, "loss": 0.012, "step": 1490 }, { "epoch": 0.9910269192422732, "grad_norm": 0.5825676082755873, "learning_rate": 1.9614760667763243e-09, "loss": 0.0102, "step": 1491 }, { "epoch": 0.9916915918909937, "grad_norm": 0.33811575307490965, "learning_rate": 1.6713330515627512e-09, "loss": 0.008, "step": 1492 }, { "epoch": 0.9923562645397142, "grad_norm": 0.210567439954917, "learning_rate": 1.4043965200055686e-09, "loss": 0.0106, "step": 1493 }, { "epoch": 0.9930209371884346, "grad_norm": 0.564170511276121, "learning_rate": 1.1606677114500698e-09, "loss": 0.0108, "step": 1494 }, { "epoch": 0.9936856098371551, "grad_norm": 0.5031417663076111, "learning_rate": 9.401477574932927e-10, "loss": 0.0099, "step": 1495 }, { "epoch": 0.9943502824858758, "grad_norm": 0.2971666199117056, "learning_rate": 7.428376819745841e-10, "loss": 0.0111, "step": 1496 }, { "epoch": 0.9950149551345963, "grad_norm": 0.18937913592511368, "learning_rate": 5.687384009755992e-10, "loss": 0.0072, "step": 1497 }, { "epoch": 0.9956796277833168, "grad_norm": 0.2987558672753604, "learning_rate": 4.178507228136397e-10, "loss": 0.0094, "step": 1498 }, { "epoch": 0.9963443004320373, "grad_norm": 0.23915956769350216, "learning_rate": 2.901753480361036e-10, "loss": 0.0088, "step": 1499 }, { "epoch": 0.9970089730807578, "grad_norm": 0.3291231973139481, "learning_rate": 1.8571286942048462e-10, "loss": 0.0103, "step": 1500 }, { "epoch": 0.9976736457294783, "grad_norm": 0.1764987251042461, "learning_rate": 1.0446377197104174e-10, "loss": 0.0058, "step": 1501 }, { "epoch": 0.9983383183781988, "grad_norm": 0.32243727144117085, "learning_rate": 4.642843291491339e-11, "loss": 0.0074, "step": 1502 }, { "epoch": 0.9990029910269193, "grad_norm": 0.2373646127033815, "learning_rate": 1.1607121701562307e-11, "loss": 0.0118, "step": 1503 }, { "epoch": 0.9996676636756398, "grad_norm": 0.23081223888003607, "learning_rate": 0.0, "loss": 0.0066, "step": 1504 }, { "epoch": 0.9996676636756398, "step": 1504, "total_flos": 109429382725632.0, "train_loss": 0.01188936162965392, "train_runtime": 35067.4375, "train_samples_per_second": 8.238, "train_steps_per_second": 0.043 } ], "logging_steps": 1.0, "max_steps": 1504, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 109429382725632.0, "train_batch_size": 12, "trial_name": null, "trial_params": null }