diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7496 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.757632398753894,
+  "eval_steps": 151,
+  "global_step": 1057,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016628559551028891,
+      "grad_norm": 6.876781701391548,
+      "learning_rate": 2e-07,
+      "loss": 1.2322,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016628559551028891,
+      "eval_loss": 1.6686532497406006,
+      "eval_runtime": 4.7271,
+      "eval_samples_per_second": 2.327,
+      "eval_steps_per_second": 0.635,
+      "step": 1
+    },
+    {
+      "epoch": 0.0033257119102057782,
+      "grad_norm": 7.15147151737489,
+      "learning_rate": 4e-07,
+      "loss": 1.3856,
+      "step": 2
+    },
+    {
+      "epoch": 0.004988567865308668,
+      "grad_norm": 6.675277261967279,
+      "learning_rate": 6e-07,
+      "loss": 1.3114,
+      "step": 3
+    },
+    {
+      "epoch": 0.0066514238204115565,
+      "grad_norm": 6.276373071109286,
+      "learning_rate": 8e-07,
+      "loss": 1.2768,
+      "step": 4
+    },
+    {
+      "epoch": 0.008314279775514447,
+      "grad_norm": 6.004954010868464,
+      "learning_rate": 1e-06,
+      "loss": 1.3114,
+      "step": 5
+    },
+    {
+      "epoch": 0.009977135730617336,
+      "grad_norm": 6.210557996205687,
+      "learning_rate": 1.2e-06,
+      "loss": 1.3777,
+      "step": 6
+    },
+    {
+      "epoch": 0.011639991685720224,
+      "grad_norm": 6.22480460719047,
+      "learning_rate": 1.4e-06,
+      "loss": 1.1786,
+      "step": 7
+    },
+    {
+      "epoch": 0.013302847640823113,
+      "grad_norm": 4.853536830796908,
+      "learning_rate": 1.6e-06,
+      "loss": 1.432,
+      "step": 8
+    },
+    {
+      "epoch": 0.014965703595926003,
+      "grad_norm": 4.549188866460316,
+      "learning_rate": 1.8e-06,
+      "loss": 1.2102,
+      "step": 9
+    },
+    {
+      "epoch": 0.016628559551028894,
+      "grad_norm": 4.596321987039164,
+      "learning_rate": 2e-06,
+      "loss": 1.4431,
+      "step": 10
+    },
+    {
+      "epoch": 0.01829141550613178,
+      "grad_norm": 3.809686255936276,
+      "learning_rate": 1.9999984649975976e-06,
+      "loss": 1.1868,
+      "step": 11
+    },
+    {
+      "epoch": 0.01995427146123467,
+      "grad_norm": 3.831577778436731,
+      "learning_rate": 1.999993859995103e-06,
+      "loss": 1.2103,
+      "step": 12
+    },
+    {
+      "epoch": 0.021617127416337558,
+      "grad_norm": 3.462783579837251,
+      "learning_rate": 1.999986185006654e-06,
+      "loss": 1.2476,
+      "step": 13
+    },
+    {
+      "epoch": 0.02327998337144045,
+      "grad_norm": 3.3297536194169064,
+      "learning_rate": 1.999975440055812e-06,
+      "loss": 1.1819,
+      "step": 14
+    },
+    {
+      "epoch": 0.02494283932654334,
+      "grad_norm": 3.8443918916019637,
+      "learning_rate": 1.999961625175565e-06,
+      "loss": 1.1482,
+      "step": 15
+    },
+    {
+      "epoch": 0.026605695281646226,
+      "grad_norm": 4.209422853092109,
+      "learning_rate": 1.999944740408324e-06,
+      "loss": 1.2177,
+      "step": 16
+    },
+    {
+      "epoch": 0.028268551236749116,
+      "grad_norm": 3.6743224915375583,
+      "learning_rate": 1.9999247858059257e-06,
+      "loss": 1.3183,
+      "step": 17
+    },
+    {
+      "epoch": 0.029931407191852007,
+      "grad_norm": 3.387454350282322,
+      "learning_rate": 1.999901761429631e-06,
+      "loss": 1.1455,
+      "step": 18
+    },
+    {
+      "epoch": 0.0315942631469549,
+      "grad_norm": 2.9657179454835356,
+      "learning_rate": 1.9998756673501237e-06,
+      "loss": 1.1463,
+      "step": 19
+    },
+    {
+      "epoch": 0.03325711910205779,
+      "grad_norm": 3.455180779074787,
+      "learning_rate": 1.9998465036475145e-06,
+      "loss": 1.1571,
+      "step": 20
+    },
+    {
+      "epoch": 0.03491997505716067,
+      "grad_norm": 2.787618991649651,
+      "learning_rate": 1.9998142704113346e-06,
+      "loss": 1.2552,
+      "step": 21
+    },
+    {
+      "epoch": 0.03658283101226356,
+      "grad_norm": 2.765977452887127,
+      "learning_rate": 1.9997789677405414e-06,
+      "loss": 1.2326,
+      "step": 22
+    },
+    {
+      "epoch": 0.03824568696736645,
+      "grad_norm": 2.6518581076350554,
+      "learning_rate": 1.9997405957435133e-06,
+      "loss": 1.2298,
+      "step": 23
+    },
+    {
+      "epoch": 0.03990854292246934,
+      "grad_norm": 3.143557273823712,
+      "learning_rate": 1.999699154538053e-06,
+      "loss": 1.3306,
+      "step": 24
+    },
+    {
+      "epoch": 0.04157139887757223,
+      "grad_norm": 2.519680050881837,
+      "learning_rate": 1.999654644251385e-06,
+      "loss": 1.1641,
+      "step": 25
+    },
+    {
+      "epoch": 0.043234254832675116,
+      "grad_norm": 2.575750435555698,
+      "learning_rate": 1.9996070650201564e-06,
+      "loss": 1.0828,
+      "step": 26
+    },
+    {
+      "epoch": 0.044897110787778007,
+      "grad_norm": 2.5829804245352403,
+      "learning_rate": 1.9995564169904354e-06,
+      "loss": 1.1247,
+      "step": 27
+    },
+    {
+      "epoch": 0.0465599667428809,
+      "grad_norm": 2.4503279434629874,
+      "learning_rate": 1.9995027003177116e-06,
+      "loss": 1.1183,
+      "step": 28
+    },
+    {
+      "epoch": 0.04822282269798379,
+      "grad_norm": 2.2946133391864487,
+      "learning_rate": 1.9994459151668956e-06,
+      "loss": 1.2234,
+      "step": 29
+    },
+    {
+      "epoch": 0.04988567865308668,
+      "grad_norm": 2.403866063427151,
+      "learning_rate": 1.9993860617123183e-06,
+      "loss": 1.1327,
+      "step": 30
+    },
+    {
+      "epoch": 0.05154853460818957,
+      "grad_norm": 2.271628438557979,
+      "learning_rate": 1.99932314013773e-06,
+      "loss": 1.2817,
+      "step": 31
+    },
+    {
+      "epoch": 0.05321139056329245,
+      "grad_norm": 2.428340751451155,
+      "learning_rate": 1.9992571506362995e-06,
+      "loss": 1.1559,
+      "step": 32
+    },
+    {
+      "epoch": 0.05487424651839534,
+      "grad_norm": 2.2614420815326604,
+      "learning_rate": 1.999188093410616e-06,
+      "loss": 1.1504,
+      "step": 33
+    },
+    {
+      "epoch": 0.05653710247349823,
+      "grad_norm": 2.34799470197157,
+      "learning_rate": 1.9991159686726847e-06,
+      "loss": 1.2409,
+      "step": 34
+    },
+    {
+      "epoch": 0.05819995842860112,
+      "grad_norm": 2.2003715198087193,
+      "learning_rate": 1.9990407766439296e-06,
+      "loss": 1.1731,
+      "step": 35
+    },
+    {
+      "epoch": 0.05986281438370401,
+      "grad_norm": 2.0892124499601645,
+      "learning_rate": 1.99896251755519e-06,
+      "loss": 1.1715,
+      "step": 36
+    },
+    {
+      "epoch": 0.061525670338806904,
+      "grad_norm": 2.2608789864787218,
+      "learning_rate": 1.998881191646722e-06,
+      "loss": 1.1871,
+      "step": 37
+    },
+    {
+      "epoch": 0.0631885262939098,
+      "grad_norm": 2.4451085935787447,
+      "learning_rate": 1.9987967991681964e-06,
+      "loss": 1.31,
+      "step": 38
+    },
+    {
+      "epoch": 0.06485138224901268,
+      "grad_norm": 2.4506760337598776,
+      "learning_rate": 1.9987093403786983e-06,
+      "loss": 1.184,
+      "step": 39
+    },
+    {
+      "epoch": 0.06651423820411557,
+      "grad_norm": 2.5379516806093267,
+      "learning_rate": 1.9986188155467267e-06,
+      "loss": 1.1922,
+      "step": 40
+    },
+    {
+      "epoch": 0.06817709415921845,
+      "grad_norm": 2.35766565204526,
+      "learning_rate": 1.998525224950194e-06,
+      "loss": 1.1608,
+      "step": 41
+    },
+    {
+      "epoch": 0.06983995011432134,
+      "grad_norm": 2.2438854614295933,
+      "learning_rate": 1.9984285688764225e-06,
+      "loss": 1.1149,
+      "step": 42
+    },
+    {
+      "epoch": 0.07150280606942423,
+      "grad_norm": 2.5176050951814792,
+      "learning_rate": 1.998328847622148e-06,
+      "loss": 1.194,
+      "step": 43
+    },
+    {
+      "epoch": 0.07316566202452712,
+      "grad_norm": 3.5070711652473388,
+      "learning_rate": 1.998226061493514e-06,
+      "loss": 1.0549,
+      "step": 44
+    },
+    {
+      "epoch": 0.07482851797963001,
+      "grad_norm": 3.8844422731190065,
+      "learning_rate": 1.9981202108060757e-06,
+      "loss": 1.2055,
+      "step": 45
+    },
+    {
+      "epoch": 0.0764913739347329,
+      "grad_norm": 3.151088308194362,
+      "learning_rate": 1.9980112958847947e-06,
+      "loss": 1.3344,
+      "step": 46
+    },
+    {
+      "epoch": 0.0781542298898358,
+      "grad_norm": 2.8997950166217508,
+      "learning_rate": 1.9978993170640403e-06,
+      "loss": 1.2626,
+      "step": 47
+    },
+    {
+      "epoch": 0.07981708584493868,
+      "grad_norm": 2.7439550744638126,
+      "learning_rate": 1.9977842746875875e-06,
+      "loss": 1.2413,
+      "step": 48
+    },
+    {
+      "epoch": 0.08147994180004157,
+      "grad_norm": 2.252937617615512,
+      "learning_rate": 1.997666169108618e-06,
+      "loss": 1.1592,
+      "step": 49
+    },
+    {
+      "epoch": 0.08314279775514447,
+      "grad_norm": 2.171409977320052,
+      "learning_rate": 1.9975450006897158e-06,
+      "loss": 1.2077,
+      "step": 50
+    },
+    {
+      "epoch": 0.08480565371024736,
+      "grad_norm": 2.384700039420674,
+      "learning_rate": 1.9974207698028686e-06,
+      "loss": 1.2758,
+      "step": 51
+    },
+    {
+      "epoch": 0.08646850966535023,
+      "grad_norm": 2.2038886028156113,
+      "learning_rate": 1.9972934768294655e-06,
+      "loss": 1.0498,
+      "step": 52
+    },
+    {
+      "epoch": 0.08813136562045312,
+      "grad_norm": 2.104326907097657,
+      "learning_rate": 1.9971631221602976e-06,
+      "loss": 1.093,
+      "step": 53
+    },
+    {
+      "epoch": 0.08979422157555601,
+      "grad_norm": 2.174068419901149,
+      "learning_rate": 1.997029706195553e-06,
+      "loss": 1.2529,
+      "step": 54
+    },
+    {
+      "epoch": 0.0914570775306589,
+      "grad_norm": 2.232522077464942,
+      "learning_rate": 1.9968932293448205e-06,
+      "loss": 1.093,
+      "step": 55
+    },
+    {
+      "epoch": 0.0931199334857618,
+      "grad_norm": 3.0592613934800372,
+      "learning_rate": 1.996753692027084e-06,
+      "loss": 1.2318,
+      "step": 56
+    },
+    {
+      "epoch": 0.09478278944086468,
+      "grad_norm": 2.3919925670804103,
+      "learning_rate": 1.996611094670724e-06,
+      "loss": 1.1787,
+      "step": 57
+    },
+    {
+      "epoch": 0.09644564539596757,
+      "grad_norm": 2.839760642367617,
+      "learning_rate": 1.9964654377135153e-06,
+      "loss": 1.2441,
+      "step": 58
+    },
+    {
+      "epoch": 0.09810850135107047,
+      "grad_norm": 2.536098443986401,
+      "learning_rate": 1.996316721602625e-06,
+      "loss": 1.2302,
+      "step": 59
+    },
+    {
+      "epoch": 0.09977135730617336,
+      "grad_norm": 2.2895217740422447,
+      "learning_rate": 1.9961649467946124e-06,
+      "loss": 1.2022,
+      "step": 60
+    },
+    {
+      "epoch": 0.10143421326127625,
+      "grad_norm": 2.1513461743943414,
+      "learning_rate": 1.996010113755427e-06,
+      "loss": 1.1063,
+      "step": 61
+    },
+    {
+      "epoch": 0.10309706921637914,
+      "grad_norm": 2.1179800201775514,
+      "learning_rate": 1.995852222960407e-06,
+      "loss": 1.0463,
+      "step": 62
+    },
+    {
+      "epoch": 0.10475992517148203,
+      "grad_norm": 2.172434098998751,
+      "learning_rate": 1.9956912748942776e-06,
+      "loss": 1.1041,
+      "step": 63
+    },
+    {
+      "epoch": 0.1064227811265849,
+      "grad_norm": 2.254257836851709,
+      "learning_rate": 1.9955272700511504e-06,
+      "loss": 1.079,
+      "step": 64
+    },
+    {
+      "epoch": 0.1080856370816878,
+      "grad_norm": 2.3081819678701234,
+      "learning_rate": 1.9953602089345213e-06,
+      "loss": 1.2835,
+      "step": 65
+    },
+    {
+      "epoch": 0.10974849303679068,
+      "grad_norm": 2.4521650444864047,
+      "learning_rate": 1.9951900920572684e-06,
+      "loss": 1.0946,
+      "step": 66
+    },
+    {
+      "epoch": 0.11141134899189357,
+      "grad_norm": 2.301990359392983,
+      "learning_rate": 1.9950169199416512e-06,
+      "loss": 1.1454,
+      "step": 67
+    },
+    {
+      "epoch": 0.11307420494699646,
+      "grad_norm": 2.159086848434006,
+      "learning_rate": 1.994840693119309e-06,
+      "loss": 1.1554,
+      "step": 68
+    },
+    {
+      "epoch": 0.11473706090209936,
+      "grad_norm": 2.5488828016690834,
+      "learning_rate": 1.994661412131259e-06,
+      "loss": 1.1537,
+      "step": 69
+    },
+    {
+      "epoch": 0.11639991685720225,
+      "grad_norm": 2.171482670853498,
+      "learning_rate": 1.9944790775278954e-06,
+      "loss": 1.2015,
+      "step": 70
+    },
+    {
+      "epoch": 0.11806277281230514,
+      "grad_norm": 2.247147793101678,
+      "learning_rate": 1.994293689868985e-06,
+      "loss": 1.1658,
+      "step": 71
+    },
+    {
+      "epoch": 0.11972562876740803,
+      "grad_norm": 2.120880044321267,
+      "learning_rate": 1.99410524972367e-06,
+      "loss": 1.2776,
+      "step": 72
+    },
+    {
+      "epoch": 0.12138848472251092,
+      "grad_norm": 2.1241719868023434,
+      "learning_rate": 1.993913757670462e-06,
+      "loss": 1.1336,
+      "step": 73
+    },
+    {
+      "epoch": 0.12305134067761381,
+      "grad_norm": 2.326506113842894,
+      "learning_rate": 1.9937192142972426e-06,
+      "loss": 1.2217,
+      "step": 74
+    },
+    {
+      "epoch": 0.12471419663271668,
+      "grad_norm": 2.2329181207257536,
+      "learning_rate": 1.9935216202012607e-06,
+      "loss": 1.2696,
+      "step": 75
+    },
+    {
+      "epoch": 0.1263770525878196,
+      "grad_norm": 2.4936798260412947,
+      "learning_rate": 1.993320975989131e-06,
+      "loss": 1.1711,
+      "step": 76
+    },
+    {
+      "epoch": 0.12803990854292246,
+      "grad_norm": 1.9925320853249602,
+      "learning_rate": 1.993117282276833e-06,
+      "loss": 1.1406,
+      "step": 77
+    },
+    {
+      "epoch": 0.12970276449802537,
+      "grad_norm": 2.2228263048635073,
+      "learning_rate": 1.992910539689707e-06,
+      "loss": 1.1856,
+      "step": 78
+    },
+    {
+      "epoch": 0.13136562045312825,
+      "grad_norm": 2.316736137818131,
+      "learning_rate": 1.9927007488624534e-06,
+      "loss": 1.1939,
+      "step": 79
+    },
+    {
+      "epoch": 0.13302847640823115,
+      "grad_norm": 2.2236401181022227,
+      "learning_rate": 1.9924879104391306e-06,
+      "loss": 1.1508,
+      "step": 80
+    },
+    {
+      "epoch": 0.13469133236333403,
+      "grad_norm": 2.3513673044729053,
+      "learning_rate": 1.992272025073155e-06,
+      "loss": 1.1877,
+      "step": 81
+    },
+    {
+      "epoch": 0.1363541883184369,
+      "grad_norm": 2.1953606776371655,
+      "learning_rate": 1.9920530934272946e-06,
+      "loss": 1.0935,
+      "step": 82
+    },
+    {
+      "epoch": 0.1380170442735398,
+      "grad_norm": 2.1614518656924693,
+      "learning_rate": 1.9918311161736713e-06,
+      "loss": 1.0932,
+      "step": 83
+    },
+    {
+      "epoch": 0.13967990022864268,
+      "grad_norm": 2.2701167674984117,
+      "learning_rate": 1.9916060939937557e-06,
+      "loss": 1.2265,
+      "step": 84
+    },
+    {
+      "epoch": 0.1413427561837456,
+      "grad_norm": 2.186355070972481,
+      "learning_rate": 1.9913780275783674e-06,
+      "loss": 1.102,
+      "step": 85
+    },
+    {
+      "epoch": 0.14300561213884846,
+      "grad_norm": 2.1276524267117267,
+      "learning_rate": 1.991146917627671e-06,
+      "loss": 1.2096,
+      "step": 86
+    },
+    {
+      "epoch": 0.14466846809395137,
+      "grad_norm": 2.2049755692680653,
+      "learning_rate": 1.9909127648511754e-06,
+      "loss": 1.1638,
+      "step": 87
+    },
+    {
+      "epoch": 0.14633132404905425,
+      "grad_norm": 2.193976635251867,
+      "learning_rate": 1.990675569967731e-06,
+      "loss": 1.3051,
+      "step": 88
+    },
+    {
+      "epoch": 0.14799418000415715,
+      "grad_norm": 2.3179428851026205,
+      "learning_rate": 1.990435333705527e-06,
+      "loss": 1.1257,
+      "step": 89
+    },
+    {
+      "epoch": 0.14965703595926003,
+      "grad_norm": 2.9183136856468925,
+      "learning_rate": 1.9901920568020894e-06,
+      "loss": 1.0138,
+      "step": 90
+    },
+    {
+      "epoch": 0.15131989191436293,
+      "grad_norm": 2.0666844143324057,
+      "learning_rate": 1.9899457400042806e-06,
+      "loss": 1.0314,
+      "step": 91
+    },
+    {
+      "epoch": 0.1529827478694658,
+      "grad_norm": 2.146945135936375,
+      "learning_rate": 1.9896963840682935e-06,
+      "loss": 1.14,
+      "step": 92
+    },
+    {
+      "epoch": 0.15464560382456868,
+      "grad_norm": 2.0563546518727427,
+      "learning_rate": 1.989443989759652e-06,
+      "loss": 1.1322,
+      "step": 93
+    },
+    {
+      "epoch": 0.1563084597796716,
+      "grad_norm": 2.0942173553840124,
+      "learning_rate": 1.989188557853208e-06,
+      "loss": 1.1109,
+      "step": 94
+    },
+    {
+      "epoch": 0.15797131573477446,
+      "grad_norm": 2.077080441568994,
+      "learning_rate": 1.9889300891331387e-06,
+      "loss": 1.0046,
+      "step": 95
+    },
+    {
+      "epoch": 0.15963417168987737,
+      "grad_norm": 2.3184144062168803,
+      "learning_rate": 1.9886685843929446e-06,
+      "loss": 1.18,
+      "step": 96
+    },
+    {
+      "epoch": 0.16129702764498025,
+      "grad_norm": 2.116450161143651,
+      "learning_rate": 1.988404044435446e-06,
+      "loss": 1.1773,
+      "step": 97
+    },
+    {
+      "epoch": 0.16295988360008315,
+      "grad_norm": 2.195134284878133,
+      "learning_rate": 1.988136470072782e-06,
+      "loss": 1.1737,
+      "step": 98
+    },
+    {
+      "epoch": 0.16462273955518603,
+      "grad_norm": 2.2738180312076977,
+      "learning_rate": 1.987865862126408e-06,
+      "loss": 1.2176,
+      "step": 99
+    },
+    {
+      "epoch": 0.16628559551028893,
+      "grad_norm": 2.31803637733792,
+      "learning_rate": 1.98759222142709e-06,
+      "loss": 1.0916,
+      "step": 100
+    },
+    {
+      "epoch": 0.1679484514653918,
+      "grad_norm": 2.1944116623564676,
+      "learning_rate": 1.9873155488149076e-06,
+      "loss": 1.1836,
+      "step": 101
+    },
+    {
+      "epoch": 0.1696113074204947,
+      "grad_norm": 2.220799107434225,
+      "learning_rate": 1.9870358451392464e-06,
+      "loss": 1.0624,
+      "step": 102
+    },
+    {
+      "epoch": 0.1712741633755976,
+      "grad_norm": 2.2574992766238373,
+      "learning_rate": 1.9867531112587987e-06,
+      "loss": 1.0148,
+      "step": 103
+    },
+    {
+      "epoch": 0.17293701933070046,
+      "grad_norm": 2.2367520120502524,
+      "learning_rate": 1.9864673480415585e-06,
+      "loss": 1.2407,
+      "step": 104
+    },
+    {
+      "epoch": 0.17459987528580337,
+      "grad_norm": 2.4288409651842477,
+      "learning_rate": 1.98617855636482e-06,
+      "loss": 1.0532,
+      "step": 105
+    },
+    {
+      "epoch": 0.17626273124090625,
+      "grad_norm": 2.422107031056536,
+      "learning_rate": 1.9858867371151753e-06,
+      "loss": 1.1696,
+      "step": 106
+    },
+    {
+      "epoch": 0.17792558719600915,
+      "grad_norm": 2.1987680710210693,
+      "learning_rate": 1.9855918911885106e-06,
+      "loss": 1.0976,
+      "step": 107
+    },
+    {
+      "epoch": 0.17958844315111203,
+      "grad_norm": 2.0312525346517307,
+      "learning_rate": 1.985294019490005e-06,
+      "loss": 1.052,
+      "step": 108
+    },
+    {
+      "epoch": 0.18125129910621493,
+      "grad_norm": 2.2392770140952116,
+      "learning_rate": 1.9849931229341256e-06,
+      "loss": 1.0335,
+      "step": 109
+    },
+    {
+      "epoch": 0.1829141550613178,
+      "grad_norm": 2.4061693671071227,
+      "learning_rate": 1.984689202444626e-06,
+      "loss": 1.2724,
+      "step": 110
+    },
+    {
+      "epoch": 0.1845770110164207,
+      "grad_norm": 2.381783600731093,
+      "learning_rate": 1.984382258954544e-06,
+      "loss": 1.0571,
+      "step": 111
+    },
+    {
+      "epoch": 0.1862398669715236,
+      "grad_norm": 2.3568823162290076,
+      "learning_rate": 1.984072293406197e-06,
+      "loss": 1.1028,
+      "step": 112
+    },
+    {
+      "epoch": 0.1879027229266265,
+      "grad_norm": 3.889820858228295,
+      "learning_rate": 1.983759306751182e-06,
+      "loss": 1.2139,
+      "step": 113
+    },
+    {
+      "epoch": 0.18956557888172937,
+      "grad_norm": 2.2124619744454415,
+      "learning_rate": 1.9834432999503684e-06,
+      "loss": 1.0924,
+      "step": 114
+    },
+    {
+      "epoch": 0.19122843483683227,
+      "grad_norm": 2.3492979702975023,
+      "learning_rate": 1.9831242739738986e-06,
+      "loss": 1.2233,
+      "step": 115
+    },
+    {
+      "epoch": 0.19289129079193515,
+      "grad_norm": 2.6172379375651347,
+      "learning_rate": 1.982802229801184e-06,
+      "loss": 1.1797,
+      "step": 116
+    },
+    {
+      "epoch": 0.19455414674703803,
+      "grad_norm": 2.195119546109998,
+      "learning_rate": 1.9824771684209024e-06,
+      "loss": 1.0346,
+      "step": 117
+    },
+    {
+      "epoch": 0.19621700270214093,
+      "grad_norm": 2.1198330508663,
+      "learning_rate": 1.982149090830993e-06,
+      "loss": 1.2163,
+      "step": 118
+    },
+    {
+      "epoch": 0.1978798586572438,
+      "grad_norm": 2.23691299776538,
+      "learning_rate": 1.981817998038656e-06,
+      "loss": 1.1865,
+      "step": 119
+    },
+    {
+      "epoch": 0.1995427146123467,
+      "grad_norm": 2.088815523084208,
+      "learning_rate": 1.981483891060348e-06,
+      "loss": 1.1199,
+      "step": 120
+    },
+    {
+      "epoch": 0.2012055705674496,
+      "grad_norm": 2.273270712463945,
+      "learning_rate": 1.9811467709217785e-06,
+      "loss": 1.2325,
+      "step": 121
+    },
+    {
+      "epoch": 0.2028684265225525,
+      "grad_norm": 2.18183229638507,
+      "learning_rate": 1.980806638657908e-06,
+      "loss": 1.2246,
+      "step": 122
+    },
+    {
+      "epoch": 0.20453128247765537,
+      "grad_norm": 3.193814855632699,
+      "learning_rate": 1.980463495312945e-06,
+      "loss": 1.0395,
+      "step": 123
+    },
+    {
+      "epoch": 0.20619413843275827,
+      "grad_norm": 2.465032307682835,
+      "learning_rate": 1.9801173419403404e-06,
+      "loss": 1.171,
+      "step": 124
+    },
+    {
+      "epoch": 0.20785699438786115,
+      "grad_norm": 2.2673988159196665,
+      "learning_rate": 1.979768179602787e-06,
+      "loss": 1.0496,
+      "step": 125
+    },
+    {
+      "epoch": 0.20951985034296405,
+      "grad_norm": 2.1447719311727265,
+      "learning_rate": 1.9794160093722147e-06,
+      "loss": 1.0066,
+      "step": 126
+    },
+    {
+      "epoch": 0.21118270629806693,
+      "grad_norm": 2.760076601802052,
+      "learning_rate": 1.979060832329788e-06,
+      "loss": 1.1358,
+      "step": 127
+    },
+    {
+      "epoch": 0.2128455622531698,
+      "grad_norm": 2.2167665381661865,
+      "learning_rate": 1.978702649565902e-06,
+      "loss": 1.114,
+      "step": 128
+    },
+    {
+      "epoch": 0.2145084182082727,
+      "grad_norm": 2.2791559309596594,
+      "learning_rate": 1.9783414621801796e-06,
+      "loss": 1.0434,
+      "step": 129
+    },
+    {
+      "epoch": 0.2161712741633756,
+      "grad_norm": 2.432104553889974,
+      "learning_rate": 1.9779772712814675e-06,
+      "loss": 1.2542,
+      "step": 130
+    },
+    {
+      "epoch": 0.2178341301184785,
+      "grad_norm": 2.7306377502903985,
+      "learning_rate": 1.9776100779878343e-06,
+      "loss": 1.0987,
+      "step": 131
+    },
+    {
+      "epoch": 0.21949698607358137,
+      "grad_norm": 2.7879620153482114,
+      "learning_rate": 1.977239883426564e-06,
+      "loss": 1.1479,
+      "step": 132
+    },
+    {
+      "epoch": 0.22115984202868427,
+      "grad_norm": 2.1514985446900354,
+      "learning_rate": 1.9768666887341564e-06,
+      "loss": 1.2583,
+      "step": 133
+    },
+    {
+      "epoch": 0.22282269798378715,
+      "grad_norm": 2.0415497832604057,
+      "learning_rate": 1.9764904950563213e-06,
+      "loss": 1.1969,
+      "step": 134
+    },
+    {
+      "epoch": 0.22448555393889005,
+      "grad_norm": 2.2681683904958403,
+      "learning_rate": 1.9761113035479745e-06,
+      "loss": 1.2112,
+      "step": 135
+    },
+    {
+      "epoch": 0.22614840989399293,
+      "grad_norm": 2.260381990967556,
+      "learning_rate": 1.975729115373236e-06,
+      "loss": 1.1884,
+      "step": 136
+    },
+    {
+      "epoch": 0.22781126584909583,
+      "grad_norm": 2.384237988040654,
+      "learning_rate": 1.9753439317054253e-06,
+      "loss": 0.9922,
+      "step": 137
+    },
+    {
+      "epoch": 0.2294741218041987,
+      "grad_norm": 2.2694287379916007,
+      "learning_rate": 1.9749557537270584e-06,
+      "loss": 1.0884,
+      "step": 138
+    },
+    {
+      "epoch": 0.2311369777593016,
+      "grad_norm": 2.358144314592439,
+      "learning_rate": 1.974564582629843e-06,
+      "loss": 1.1049,
+      "step": 139
+    },
+    {
+      "epoch": 0.2327998337144045,
+      "grad_norm": 2.1123073465297306,
+      "learning_rate": 1.9741704196146766e-06,
+      "loss": 1.0306,
+      "step": 140
+    },
+    {
+      "epoch": 0.23446268966950737,
+      "grad_norm": 2.1512707106124798,
+      "learning_rate": 1.973773265891641e-06,
+      "loss": 1.0528,
+      "step": 141
+    },
+    {
+      "epoch": 0.23612554562461027,
+      "grad_norm": 2.172534073331255,
+      "learning_rate": 1.973373122680001e-06,
+      "loss": 1.1115,
+      "step": 142
+    },
+    {
+      "epoch": 0.23778840157971315,
+      "grad_norm": 2.3733216420093473,
+      "learning_rate": 1.972969991208198e-06,
+      "loss": 1.1323,
+      "step": 143
+    },
+    {
+      "epoch": 0.23945125753481605,
+      "grad_norm": 2.4625937737389094,
+      "learning_rate": 1.9725638727138465e-06,
+      "loss": 1.0889,
+      "step": 144
+    },
+    {
+      "epoch": 0.24111411348991893,
+      "grad_norm": 2.2420911044484395,
+      "learning_rate": 1.9721547684437332e-06,
+      "loss": 1.0748,
+      "step": 145
+    },
+    {
+      "epoch": 0.24277696944502183,
+      "grad_norm": 2.274748664959236,
+      "learning_rate": 1.97174267965381e-06,
+      "loss": 1.2631,
+      "step": 146
+    },
+    {
+      "epoch": 0.2444398254001247,
+      "grad_norm": 2.251081234626043,
+      "learning_rate": 1.9713276076091916e-06,
+      "loss": 1.2066,
+      "step": 147
+    },
+    {
+      "epoch": 0.24610268135522761,
+      "grad_norm": 2.2539634056834137,
+      "learning_rate": 1.970909553584151e-06,
+      "loss": 1.0727,
+      "step": 148
+    },
+    {
+      "epoch": 0.2477655373103305,
+      "grad_norm": 2.1434067229672094,
+      "learning_rate": 1.9704885188621157e-06,
+      "loss": 1.1082,
+      "step": 149
+    },
+    {
+      "epoch": 0.24942839326543337,
+      "grad_norm": 2.236934300850519,
+      "learning_rate": 1.970064504735665e-06,
+      "loss": 1.1063,
+      "step": 150
+    },
+    {
+      "epoch": 0.2510912492205363,
+      "grad_norm": 2.760825374012749,
+      "learning_rate": 1.969637512506524e-06,
+      "loss": 1.2253,
+      "step": 151
+    },
+    {
+      "epoch": 0.2510912492205363,
+      "eval_loss": 1.4304277896881104,
+      "eval_runtime": 24.5872,
+      "eval_samples_per_second": 0.447,
+      "eval_steps_per_second": 0.122,
+      "step": 151
+    },
+    {
+      "epoch": 0.2527541051756392,
+      "grad_norm": 2.548532327058704,
+      "learning_rate": 1.9692075434855604e-06,
+      "loss": 1.1196,
+      "step": 152
+    },
+    {
+      "epoch": 0.254416961130742,
+      "grad_norm": 2.3149961972538464,
+      "learning_rate": 1.9687745989927824e-06,
+      "loss": 1.093,
+      "step": 153
+    },
+    {
+      "epoch": 0.25607981708584493,
+      "grad_norm": 2.2766910929781257,
+      "learning_rate": 1.96833868035733e-06,
+      "loss": 1.1225,
+      "step": 154
+    },
+    {
+      "epoch": 0.25774267304094783,
+      "grad_norm": 2.7395763625060674,
+      "learning_rate": 1.967899788917477e-06,
+      "loss": 1.1423,
+      "step": 155
+    },
+    {
+      "epoch": 0.25940552899605074,
+      "grad_norm": 2.389993560718361,
+      "learning_rate": 1.9674579260206213e-06,
+      "loss": 1.1823,
+      "step": 156
+    },
+    {
+      "epoch": 0.2610683849511536,
+      "grad_norm": 2.483941805273633,
+      "learning_rate": 1.967013093023285e-06,
+      "loss": 1.2279,
+      "step": 157
+    },
+    {
+      "epoch": 0.2627312409062565,
+      "grad_norm": 2.2477103547008914,
+      "learning_rate": 1.9665652912911065e-06,
+      "loss": 1.1962,
+      "step": 158
+    },
+    {
+      "epoch": 0.2643940968613594,
+      "grad_norm": 2.841612920075852,
+      "learning_rate": 1.9661145221988398e-06,
+      "loss": 1.1292,
+      "step": 159
+    },
+    {
+      "epoch": 0.2660569528164623,
+      "grad_norm": 2.5188774282226487,
+      "learning_rate": 1.9656607871303485e-06,
+      "loss": 1.0174,
+      "step": 160
+    },
+    {
+      "epoch": 0.26771980877156515,
+      "grad_norm": 2.601863007672941,
+      "learning_rate": 1.9652040874786007e-06,
+      "loss": 1.1803,
+      "step": 161
+    },
+    {
+      "epoch": 0.26938266472666805,
+      "grad_norm": 2.1729305988620027,
+      "learning_rate": 1.964744424645667e-06,
+      "loss": 1.0692,
+      "step": 162
+    },
+    {
+      "epoch": 0.27104552068177096,
+      "grad_norm": 2.4137099619779017,
+      "learning_rate": 1.964281800042714e-06,
+      "loss": 1.2412,
+      "step": 163
+    },
+    {
+      "epoch": 0.2727083766368738,
+      "grad_norm": 2.4257004197876437,
+      "learning_rate": 1.9638162150900025e-06,
+      "loss": 1.1754,
+      "step": 164
+    },
+    {
+      "epoch": 0.2743712325919767,
+      "grad_norm": 2.127187241617996,
+      "learning_rate": 1.96334767121688e-06,
+      "loss": 1.1212,
+      "step": 165
+    },
+    {
+      "epoch": 0.2760340885470796,
+      "grad_norm": 2.5505445228517796,
+      "learning_rate": 1.962876169861778e-06,
+      "loss": 1.1873,
+      "step": 166
+    },
+    {
+      "epoch": 0.2776969445021825,
+      "grad_norm": 2.362761021929743,
+      "learning_rate": 1.9624017124722085e-06,
+      "loss": 1.1519,
+      "step": 167
+    },
+    {
+      "epoch": 0.27935980045728537,
+      "grad_norm": 2.3485374311573874,
+      "learning_rate": 1.9619243005047574e-06,
+      "loss": 1.2356,
+      "step": 168
+    },
+    {
+      "epoch": 0.28102265641238827,
+      "grad_norm": 2.1257685683782066,
+      "learning_rate": 1.9614439354250824e-06,
+      "loss": 1.1825,
+      "step": 169
+    },
+    {
+      "epoch": 0.2826855123674912,
+      "grad_norm": 2.395738763591683,
+      "learning_rate": 1.960960618707906e-06,
+      "loss": 1.1009,
+      "step": 170
+    },
+    {
+      "epoch": 0.2843483683225941,
+      "grad_norm": 2.2654812048452855,
+      "learning_rate": 1.9604743518370133e-06,
+      "loss": 1.1115,
+      "step": 171
+    },
+    {
+      "epoch": 0.28601122427769693,
+      "grad_norm": 2.377019893527625,
+      "learning_rate": 1.959985136305246e-06,
+      "loss": 1.1096,
+      "step": 172
+    },
+    {
+      "epoch": 0.28767408023279983,
+      "grad_norm": 2.2583188742925597,
+      "learning_rate": 1.9594929736144973e-06,
+      "loss": 1.0579,
+      "step": 173
+    },
+    {
+      "epoch": 0.28933693618790274,
+      "grad_norm": 2.2955420850835644,
+      "learning_rate": 1.9589978652757096e-06,
+      "loss": 1.0044,
+      "step": 174
+    },
+    {
+      "epoch": 0.2909997921430056,
+      "grad_norm": 2.135658412712305,
+      "learning_rate": 1.9584998128088683e-06,
+      "loss": 0.9674,
+      "step": 175
+    },
+    {
+      "epoch": 0.2926626480981085,
+      "grad_norm": 2.3814784805947435,
+      "learning_rate": 1.9579988177429965e-06,
+      "loss": 0.9844,
+      "step": 176
+    },
+    {
+      "epoch": 0.2943255040532114,
+      "grad_norm": 4.249522359005645,
+      "learning_rate": 1.957494881616151e-06,
+      "loss": 1.0733,
+      "step": 177
+    },
+    {
+      "epoch": 0.2959883600083143,
+      "grad_norm": 2.1491559170675005,
+      "learning_rate": 1.956988005975419e-06,
+      "loss": 1.1137,
+      "step": 178
+    },
+    {
+      "epoch": 0.29765121596341715,
+      "grad_norm": 2.237150018318634,
+      "learning_rate": 1.9564781923769105e-06,
+      "loss": 1.1954,
+      "step": 179
+    },
+    {
+      "epoch": 0.29931407191852005,
+      "grad_norm": 2.051176936877328,
+      "learning_rate": 1.955965442385756e-06,
+      "loss": 1.132,
+      "step": 180
+    },
+    {
+      "epoch": 0.30097692787362296,
+      "grad_norm": 3.5719215857546045,
+      "learning_rate": 1.9554497575761e-06,
+      "loss": 1.2088,
+      "step": 181
+    },
+    {
+      "epoch": 0.30263978382872586,
+      "grad_norm": 2.0583047844957343,
+      "learning_rate": 1.9549311395310982e-06,
+      "loss": 1.0338,
+      "step": 182
+    },
+    {
+      "epoch": 0.3043026397838287,
+      "grad_norm": 2.356690350916363,
+      "learning_rate": 1.9544095898429094e-06,
+      "loss": 1.1253,
+      "step": 183
+    },
+    {
+      "epoch": 0.3059654957389316,
+      "grad_norm": 2.31830959115815,
+      "learning_rate": 1.9538851101126944e-06,
+      "loss": 0.9722,
+      "step": 184
+    },
+    {
+      "epoch": 0.3076283516940345,
+      "grad_norm": 2.2973614356210503,
+      "learning_rate": 1.9533577019506085e-06,
+      "loss": 1.1471,
+      "step": 185
+    },
+    {
+      "epoch": 0.30929120764913737,
+      "grad_norm": 2.2790981590004233,
+      "learning_rate": 1.952827366975797e-06,
+      "loss": 1.0407,
+      "step": 186
+    },
+    {
+      "epoch": 0.31095406360424027,
+      "grad_norm": 2.2618992906420714,
+      "learning_rate": 1.952294106816391e-06,
+      "loss": 1.2548,
+      "step": 187
+    },
+    {
+      "epoch": 0.3126169195593432,
+      "grad_norm": 2.644954828913663,
+      "learning_rate": 1.9517579231095015e-06,
+      "loss": 1.2529,
+      "step": 188
+    },
+    {
+      "epoch": 0.3142797755144461,
+      "grad_norm": 2.256019290547997,
+      "learning_rate": 1.9512188175012153e-06,
+      "loss": 1.1046,
+      "step": 189
+    },
+    {
+      "epoch": 0.31594263146954893,
+      "grad_norm": 2.5306772178751133,
+      "learning_rate": 1.950676791646589e-06,
+      "loss": 1.1444,
+      "step": 190
+    },
+    {
+      "epoch": 0.31760548742465183,
+      "grad_norm": 2.530764997510064,
+      "learning_rate": 1.9501318472096447e-06,
+      "loss": 1.1185,
+      "step": 191
+    },
+    {
+      "epoch": 0.31926834337975474,
+      "grad_norm": 2.308534558092756,
+      "learning_rate": 1.9495839858633648e-06,
+      "loss": 1.1669,
+      "step": 192
+    },
+    {
+      "epoch": 0.32093119933485764,
+      "grad_norm": 2.2116575569490924,
+      "learning_rate": 1.9490332092896857e-06,
+      "loss": 1.1689,
+      "step": 193
+    },
+    {
+      "epoch": 0.3225940552899605,
+      "grad_norm": 3.098403349489372,
+      "learning_rate": 1.9484795191794943e-06,
+      "loss": 1.0652,
+      "step": 194
+    },
+    {
+      "epoch": 0.3242569112450634,
+      "grad_norm": 2.2040383274861965,
+      "learning_rate": 1.947922917232622e-06,
+      "loss": 0.9792,
+      "step": 195
+    },
+    {
+      "epoch": 0.3259197672001663,
+      "grad_norm": 2.2311156638317557,
+      "learning_rate": 1.9473634051578394e-06,
+      "loss": 1.0935,
+      "step": 196
+    },
+    {
+      "epoch": 0.32758262315526915,
+      "grad_norm": 2.2800635132242997,
+      "learning_rate": 1.946800984672851e-06,
+      "loss": 1.2076,
+      "step": 197
+    },
+    {
+      "epoch": 0.32924547911037205,
+      "grad_norm": 3.654896049770491,
+      "learning_rate": 1.946235657504291e-06,
+      "loss": 1.1199,
+      "step": 198
+    },
+    {
+      "epoch": 0.33090833506547496,
+      "grad_norm": 3.21717248498429,
+      "learning_rate": 1.945667425387716e-06,
+      "loss": 1.0107,
+      "step": 199
+    },
+    {
+      "epoch": 0.33257119102057786,
+      "grad_norm": 2.401723766209091,
+      "learning_rate": 1.9450962900676014e-06,
+      "loss": 1.1726,
+      "step": 200
+    },
+    {
+      "epoch": 0.3342340469756807,
+      "grad_norm": 2.395261717837967,
+      "learning_rate": 1.9445222532973356e-06,
+      "loss": 1.1804,
+      "step": 201
+    },
+    {
+      "epoch": 0.3358969029307836,
+      "grad_norm": 2.1446239585083053,
+      "learning_rate": 1.943945316839214e-06,
+      "loss": 1.1057,
+      "step": 202
+    },
+    {
+      "epoch": 0.3375597588858865,
+      "grad_norm": 2.316344120227303,
+      "learning_rate": 1.9433654824644343e-06,
+      "loss": 1.2313,
+      "step": 203
+    },
+    {
+      "epoch": 0.3392226148409894,
+      "grad_norm": 2.260699038364008,
+      "learning_rate": 1.9427827519530917e-06,
+      "loss": 1.0259,
+      "step": 204
+    },
+    {
+      "epoch": 0.34088547079609227,
+      "grad_norm": 2.2098708691255027,
+      "learning_rate": 1.94219712709417e-06,
+      "loss": 1.162,
+      "step": 205
+    },
+    {
+      "epoch": 0.3425483267511952,
+      "grad_norm": 2.177000943321658,
+      "learning_rate": 1.9416086096855414e-06,
+      "loss": 1.1537,
+      "step": 206
+    },
+    {
+      "epoch": 0.3442111827062981,
+      "grad_norm": 2.1645641329995127,
+      "learning_rate": 1.9410172015339575e-06,
+      "loss": 1.0991,
+      "step": 207
+    },
+    {
+      "epoch": 0.34587403866140093,
+      "grad_norm": 2.352398281001825,
+      "learning_rate": 1.940422904455043e-06,
+      "loss": 1.1333,
+      "step": 208
+    },
+    {
+      "epoch": 0.34753689461650383,
+      "grad_norm": 2.3240858436164404,
+      "learning_rate": 1.939825720273294e-06,
+      "loss": 1.1263,
+      "step": 209
+    },
+    {
+      "epoch": 0.34919975057160674,
+      "grad_norm": 2.2225865374520404,
+      "learning_rate": 1.939225650822068e-06,
+      "loss": 1.1092,
+      "step": 210
+    },
+    {
+      "epoch": 0.35086260652670964,
+      "grad_norm": 2.344261138949479,
+      "learning_rate": 1.938622697943581e-06,
+      "loss": 1.1262,
+      "step": 211
+    },
+    {
+      "epoch": 0.3525254624818125,
+      "grad_norm": 2.0594249160281817,
+      "learning_rate": 1.938016863488902e-06,
+      "loss": 1.03,
+      "step": 212
+    },
+    {
+      "epoch": 0.3541883184369154,
+      "grad_norm": 2.197981511279412,
+      "learning_rate": 1.9374081493179453e-06,
+      "loss": 1.2617,
+      "step": 213
+    },
+    {
+      "epoch": 0.3558511743920183,
+      "grad_norm": 2.3896742142995717,
+      "learning_rate": 1.9367965572994663e-06,
+      "loss": 1.1666,
+      "step": 214
+    },
+    {
+      "epoch": 0.3575140303471212,
+      "grad_norm": 2.2197343813094075,
+      "learning_rate": 1.9361820893110554e-06,
+      "loss": 1.1399,
+      "step": 215
+    },
+    {
+      "epoch": 0.35917688630222405,
+      "grad_norm": 2.0777320432807684,
+      "learning_rate": 1.9355647472391325e-06,
+      "loss": 0.997,
+      "step": 216
+    },
+    {
+      "epoch": 0.36083974225732696,
+      "grad_norm": 2.1390003170764262,
+      "learning_rate": 1.93494453297894e-06,
+      "loss": 1.0987,
+      "step": 217
+    },
+    {
+      "epoch": 0.36250259821242986,
+      "grad_norm": 2.2336666277286485,
+      "learning_rate": 1.93432144843454e-06,
+      "loss": 1.1072,
+      "step": 218
+    },
+    {
+      "epoch": 0.36416545416753276,
+      "grad_norm": 2.5151519435391565,
+      "learning_rate": 1.933695495518804e-06,
+      "loss": 1.187,
+      "step": 219
+    },
+    {
+      "epoch": 0.3658283101226356,
+      "grad_norm": 2.237546644093813,
+      "learning_rate": 1.9330666761534104e-06,
+      "loss": 1.1992,
+      "step": 220
+    },
+    {
+      "epoch": 0.3674911660777385,
+      "grad_norm": 2.20747601132904,
+      "learning_rate": 1.932434992268838e-06,
+      "loss": 1.25,
+      "step": 221
+    },
+    {
+      "epoch": 0.3691540220328414,
+      "grad_norm": 2.0703921340607954,
+      "learning_rate": 1.9318004458043595e-06,
+      "loss": 0.9968,
+      "step": 222
+    },
+    {
+      "epoch": 0.37081687798794427,
+      "grad_norm": 2.6998540968138203,
+      "learning_rate": 1.9311630387080355e-06,
+      "loss": 1.1253,
+      "step": 223
+    },
+    {
+      "epoch": 0.3724797339430472,
+      "grad_norm": 2.2382556329258083,
+      "learning_rate": 1.9305227729367088e-06,
+      "loss": 1.033,
+      "step": 224
+    },
+    {
+      "epoch": 0.3741425898981501,
+      "grad_norm": 2.0744866884439426,
+      "learning_rate": 1.929879650455998e-06,
+      "loss": 0.9939,
+      "step": 225
+    },
+    {
+      "epoch": 0.375805445853253,
+      "grad_norm": 2.1380855568440293,
+      "learning_rate": 1.929233673240293e-06,
+      "loss": 1.1454,
+      "step": 226
+    },
+    {
+      "epoch": 0.37746830180835583,
+      "grad_norm": 2.3748389575645783,
+      "learning_rate": 1.928584843272746e-06,
+      "loss": 1.0883,
+      "step": 227
+    },
+    {
+      "epoch": 0.37913115776345874,
+      "grad_norm": 1.9756985785164076,
+      "learning_rate": 1.9279331625452694e-06,
+      "loss": 1.0773,
+      "step": 228
+    },
+    {
+      "epoch": 0.38079401371856164,
+      "grad_norm": 2.8312536305491416,
+      "learning_rate": 1.927278633058525e-06,
+      "loss": 1.1389,
+      "step": 229
+    },
+    {
+      "epoch": 0.38245686967366455,
+      "grad_norm": 2.130686929883375,
+      "learning_rate": 1.926621256821922e-06,
+      "loss": 1.0132,
+      "step": 230
+    },
+    {
+      "epoch": 0.3841197256287674,
+      "grad_norm": 2.305764167292808,
+      "learning_rate": 1.9259610358536085e-06,
+      "loss": 1.1054,
+      "step": 231
+    },
+    {
+      "epoch": 0.3857825815838703,
+      "grad_norm": 2.3978186298387683,
+      "learning_rate": 1.9252979721804657e-06,
+      "loss": 1.0411,
+      "step": 232
+    },
+    {
+      "epoch": 0.3874454375389732,
+      "grad_norm": 2.195123698256728,
+      "learning_rate": 1.9246320678381032e-06,
+      "loss": 1.1931,
+      "step": 233
+    },
+    {
+      "epoch": 0.38910829349407605,
+      "grad_norm": 2.6254493837547996,
+      "learning_rate": 1.92396332487085e-06,
+      "loss": 1.1892,
+      "step": 234
+    },
+    {
+      "epoch": 0.39077114944917896,
+      "grad_norm": 2.419866448691614,
+      "learning_rate": 1.9232917453317492e-06,
+      "loss": 1.0236,
+      "step": 235
+    },
+    {
+      "epoch": 0.39243400540428186,
+      "grad_norm": 2.1862355398588287,
+      "learning_rate": 1.9226173312825553e-06,
+      "loss": 1.0528,
+      "step": 236
+    },
+    {
+      "epoch": 0.39409686135938476,
+      "grad_norm": 2.3764854199802707,
+      "learning_rate": 1.921940084793721e-06,
+      "loss": 1.0602,
+      "step": 237
+    },
+    {
+      "epoch": 0.3957597173144876,
+      "grad_norm": 2.4103818910002057,
+      "learning_rate": 1.921260007944397e-06,
+      "loss": 0.9984,
+      "step": 238
+    },
+    {
+      "epoch": 0.3974225732695905,
+      "grad_norm": 2.1867489158332396,
+      "learning_rate": 1.920577102822422e-06,
+      "loss": 1.1426,
+      "step": 239
+    },
+    {
+      "epoch": 0.3990854292246934,
+      "grad_norm": 2.5978893922687964,
+      "learning_rate": 1.9198913715243182e-06,
+      "loss": 1.1274,
+      "step": 240
+    },
+    {
+      "epoch": 0.4007482851797963,
+      "grad_norm": 2.339459827714988,
+      "learning_rate": 1.9192028161552843e-06,
+      "loss": 1.0478,
+      "step": 241
+    },
+    {
+      "epoch": 0.4024111411348992,
+      "grad_norm": 2.3678596214542162,
+      "learning_rate": 1.9185114388291886e-06,
+      "loss": 1.0155,
+      "step": 242
+    },
+    {
+      "epoch": 0.4040739970900021,
+      "grad_norm": 2.1203964226087106,
+      "learning_rate": 1.9178172416685625e-06,
+      "loss": 1.0326,
+      "step": 243
+    },
+    {
+      "epoch": 0.405736853045105,
+      "grad_norm": 2.333908782402527,
+      "learning_rate": 1.9171202268045946e-06,
+      "loss": 1.0474,
+      "step": 244
+    },
+    {
+      "epoch": 0.40739970900020783,
+      "grad_norm": 2.9030943393253454,
+      "learning_rate": 1.9164203963771243e-06,
+      "loss": 1.1753,
+      "step": 245
+    },
+    {
+      "epoch": 0.40906256495531074,
+      "grad_norm": 2.4513503490604087,
+      "learning_rate": 1.915717752534634e-06,
+      "loss": 1.0477,
+      "step": 246
+    },
+    {
+      "epoch": 0.41072542091041364,
+      "grad_norm": 2.0267508782326735,
+      "learning_rate": 1.915012297434243e-06,
+      "loss": 1.0991,
+      "step": 247
+    },
+    {
+      "epoch": 0.41238827686551655,
+      "grad_norm": 2.184393317130635,
+      "learning_rate": 1.9143040332417036e-06,
+      "loss": 1.2479,
+      "step": 248
+    },
+    {
+      "epoch": 0.4140511328206194,
+      "grad_norm": 2.026607524191998,
+      "learning_rate": 1.9135929621313886e-06,
+      "loss": 1.0745,
+      "step": 249
+    },
+    {
+      "epoch": 0.4157139887757223,
+      "grad_norm": 2.66566941286477,
+      "learning_rate": 1.912879086286291e-06,
+      "loss": 0.9754,
+      "step": 250
+    },
+    {
+      "epoch": 0.4173768447308252,
+      "grad_norm": 2.1367244864183625,
+      "learning_rate": 1.9121624078980122e-06,
+      "loss": 1.1809,
+      "step": 251
+    },
+    {
+      "epoch": 0.4190397006859281,
+      "grad_norm": 2.38106027954001,
+      "learning_rate": 1.911442929166758e-06,
+      "loss": 1.105,
+      "step": 252
+    },
+    {
+      "epoch": 0.42070255664103096,
+      "grad_norm": 2.157838169653484,
+      "learning_rate": 1.910720652301333e-06,
+      "loss": 1.1357,
+      "step": 253
+    },
+    {
+      "epoch": 0.42236541259613386,
+      "grad_norm": 2.4900405664879135,
+      "learning_rate": 1.9099955795191292e-06,
+      "loss": 1.1238,
+      "step": 254
+    },
+    {
+      "epoch": 0.42402826855123676,
+      "grad_norm": 2.2488268195059717,
+      "learning_rate": 1.9092677130461245e-06,
+      "loss": 1.0831,
+      "step": 255
+    },
+    {
+      "epoch": 0.4256911245063396,
+      "grad_norm": 2.266782137690799,
+      "learning_rate": 1.9085370551168718e-06,
+      "loss": 1.0905,
+      "step": 256
+    },
+    {
+      "epoch": 0.4273539804614425,
+      "grad_norm": 3.7403186028520934,
+      "learning_rate": 1.9078036079744947e-06,
+      "loss": 1.101,
+      "step": 257
+    },
+    {
+      "epoch": 0.4290168364165454,
+      "grad_norm": 2.455666608661609,
+      "learning_rate": 1.9070673738706796e-06,
+      "loss": 1.1266,
+      "step": 258
+    },
+    {
+      "epoch": 0.4306796923716483,
+      "grad_norm": 2.515358176239364,
+      "learning_rate": 1.9063283550656687e-06,
+      "loss": 1.1149,
+      "step": 259
+    },
+    {
+      "epoch": 0.4323425483267512,
+      "grad_norm": 2.210842033312207,
+      "learning_rate": 1.905586553828253e-06,
+      "loss": 1.0168,
+      "step": 260
+    },
+    {
+      "epoch": 0.4340054042818541,
+      "grad_norm": 2.295182467603218,
+      "learning_rate": 1.9048419724357658e-06,
+      "loss": 1.173,
+      "step": 261
+    },
+    {
+      "epoch": 0.435668260236957,
+      "grad_norm": 2.1506193772075886,
+      "learning_rate": 1.9040946131740762e-06,
+      "loss": 1.084,
+      "step": 262
+    },
+    {
+      "epoch": 0.4373311161920599,
+      "grad_norm": 2.0996773232405403,
+      "learning_rate": 1.9033444783375804e-06,
+      "loss": 1.1712,
+      "step": 263
+    },
+    {
+      "epoch": 0.43899397214716274,
+      "grad_norm": 1.9545895155808832,
+      "learning_rate": 1.9025915702291954e-06,
+      "loss": 0.9843,
+      "step": 264
+    },
+    {
+      "epoch": 0.44065682810226564,
+      "grad_norm": 2.2269341512176224,
+      "learning_rate": 1.9018358911603535e-06,
+      "loss": 1.1599,
+      "step": 265
+    },
+    {
+      "epoch": 0.44231968405736855,
+      "grad_norm": 2.1685909801965297,
+      "learning_rate": 1.901077443450993e-06,
+      "loss": 1.0454,
+      "step": 266
+    },
+    {
+      "epoch": 0.4439825400124714,
+      "grad_norm": 2.377103399607552,
+      "learning_rate": 1.9003162294295513e-06,
+      "loss": 1.0963,
+      "step": 267
+    },
+    {
+      "epoch": 0.4456453959675743,
+      "grad_norm": 2.5151408653095904,
+      "learning_rate": 1.8995522514329601e-06,
+      "loss": 1.104,
+      "step": 268
+    },
+    {
+      "epoch": 0.4473082519226772,
+      "grad_norm": 2.094881644352632,
+      "learning_rate": 1.8987855118066348e-06,
+      "loss": 1.2278,
+      "step": 269
+    },
+    {
+      "epoch": 0.4489711078777801,
+      "grad_norm": 2.2344009413987376,
+      "learning_rate": 1.8980160129044695e-06,
+      "loss": 1.1049,
+      "step": 270
+    },
+    {
+      "epoch": 0.45063396383288296,
+      "grad_norm": 2.297839277269806,
+      "learning_rate": 1.8972437570888307e-06,
+      "loss": 1.1538,
+      "step": 271
+    },
+    {
+      "epoch": 0.45229681978798586,
+      "grad_norm": 2.238645887183212,
+      "learning_rate": 1.8964687467305463e-06,
+      "loss": 1.1412,
+      "step": 272
+    },
+    {
+      "epoch": 0.45395967574308876,
+      "grad_norm": 2.145666305518761,
+      "learning_rate": 1.895690984208902e-06,
+      "loss": 1.0315,
+      "step": 273
+    },
+    {
+      "epoch": 0.45562253169819167,
+      "grad_norm": 2.310288742114033,
+      "learning_rate": 1.894910471911633e-06,
+      "loss": 1.0277,
+      "step": 274
+    },
+    {
+      "epoch": 0.4572853876532945,
+      "grad_norm": 2.155602139621284,
+      "learning_rate": 1.894127212234916e-06,
+      "loss": 1.0503,
+      "step": 275
+    },
+    {
+      "epoch": 0.4589482436083974,
+      "grad_norm": 2.6510720645982944,
+      "learning_rate": 1.8933412075833607e-06,
+      "loss": 1.0964,
+      "step": 276
+    },
+    {
+      "epoch": 0.4606110995635003,
+      "grad_norm": 2.077628159647317,
+      "learning_rate": 1.8925524603700062e-06,
+      "loss": 1.0037,
+      "step": 277
+    },
+    {
+      "epoch": 0.4622739555186032,
+      "grad_norm": 2.117391684386515,
+      "learning_rate": 1.8917609730163103e-06,
+      "loss": 1.0695,
+      "step": 278
+    },
+    {
+      "epoch": 0.4639368114737061,
+      "grad_norm": 2.372444846141058,
+      "learning_rate": 1.8909667479521425e-06,
+      "loss": 1.1472,
+      "step": 279
+    },
+    {
+      "epoch": 0.465599667428809,
+      "grad_norm": 2.4078311689148064,
+      "learning_rate": 1.8901697876157776e-06,
+      "loss": 1.2753,
+      "step": 280
+    },
+    {
+      "epoch": 0.4672625233839119,
+      "grad_norm": 3.21629533072984,
+      "learning_rate": 1.8893700944538881e-06,
+      "loss": 1.1071,
+      "step": 281
+    },
+    {
+      "epoch": 0.46892537933901474,
+      "grad_norm": 2.4295013417380287,
+      "learning_rate": 1.8885676709215353e-06,
+      "loss": 1.1355,
+      "step": 282
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 2.275602301735013,
+      "learning_rate": 1.8877625194821636e-06,
+      "loss": 1.2554,
+      "step": 283
+    },
+    {
+      "epoch": 0.47225109124922054,
+      "grad_norm": 2.3766077332166957,
+      "learning_rate": 1.8869546426075917e-06,
+      "loss": 1.0717,
+      "step": 284
+    },
+    {
+      "epoch": 0.47391394720432345,
+      "grad_norm": 2.132945677416556,
+      "learning_rate": 1.8861440427780058e-06,
+      "loss": 1.1736,
+      "step": 285
+    },
+    {
+      "epoch": 0.4755768031594263,
+      "grad_norm": 2.045707307029965,
+      "learning_rate": 1.8853307224819506e-06,
+      "loss": 0.9087,
+      "step": 286
+    },
+    {
+      "epoch": 0.4772396591145292,
+      "grad_norm": 2.163246456875578,
+      "learning_rate": 1.8845146842163238e-06,
+      "loss": 1.1869,
+      "step": 287
+    },
+    {
+      "epoch": 0.4789025150696321,
+      "grad_norm": 2.230799208427525,
+      "learning_rate": 1.8836959304863669e-06,
+      "loss": 1.0081,
+      "step": 288
+    },
+    {
+      "epoch": 0.48056537102473496,
+      "grad_norm": 2.430756467607878,
+      "learning_rate": 1.8828744638056573e-06,
+      "loss": 1.1908,
+      "step": 289
+    },
+    {
+      "epoch": 0.48222822697983786,
+      "grad_norm": 2.2702841588026117,
+      "learning_rate": 1.882050286696102e-06,
+      "loss": 1.1872,
+      "step": 290
+    },
+    {
+      "epoch": 0.48389108293494076,
+      "grad_norm": 2.214741995886969,
+      "learning_rate": 1.881223401687929e-06,
+      "loss": 1.101,
+      "step": 291
+    },
+    {
+      "epoch": 0.48555393889004367,
+      "grad_norm": 2.1908978532802,
+      "learning_rate": 1.8803938113196784e-06,
+      "loss": 1.0547,
+      "step": 292
+    },
+    {
+      "epoch": 0.4872167948451465,
+      "grad_norm": 2.1957825231056844,
+      "learning_rate": 1.8795615181381974e-06,
+      "loss": 1.0209,
+      "step": 293
+    },
+    {
+      "epoch": 0.4888796508002494,
+      "grad_norm": 2.4659476575942056,
+      "learning_rate": 1.8787265246986298e-06,
+      "loss": 1.0354,
+      "step": 294
+    },
+    {
+      "epoch": 0.4905425067553523,
+      "grad_norm": 2.499181314884686,
+      "learning_rate": 1.877888833564409e-06,
+      "loss": 1.1167,
+      "step": 295
+    },
+    {
+      "epoch": 0.49220536271045523,
+      "grad_norm": 2.297165687585788,
+      "learning_rate": 1.8770484473072517e-06,
+      "loss": 1.2812,
+      "step": 296
+    },
+    {
+      "epoch": 0.4938682186655581,
+      "grad_norm": 2.155148132032728,
+      "learning_rate": 1.8762053685071471e-06,
+      "loss": 0.9955,
+      "step": 297
+    },
+    {
+      "epoch": 0.495531074620661,
+      "grad_norm": 2.2105051332330183,
+      "learning_rate": 1.8753595997523513e-06,
+      "loss": 1.0817,
+      "step": 298
+    },
+    {
+      "epoch": 0.4971939305757639,
+      "grad_norm": 2.2489919353382937,
+      "learning_rate": 1.8745111436393785e-06,
+      "loss": 1.116,
+      "step": 299
+    },
+    {
+      "epoch": 0.49885678653086674,
+      "grad_norm": 2.245142666798669,
+      "learning_rate": 1.8736600027729933e-06,
+      "loss": 1.2594,
+      "step": 300
+    },
+    {
+      "epoch": 0.5005196424859697,
+      "grad_norm": 2.4738621609715263,
+      "learning_rate": 1.8728061797662016e-06,
+      "loss": 1.2273,
+      "step": 301
+    },
+    {
+      "epoch": 0.5021824984410725,
+      "grad_norm": 2.1040097093458745,
+      "learning_rate": 1.8719496772402447e-06,
+      "loss": 1.1395,
+      "step": 302
+    },
+    {
+      "epoch": 0.5021824984410725,
+      "eval_loss": 1.4084999561309814,
+      "eval_runtime": 24.7725,
+      "eval_samples_per_second": 0.444,
+      "eval_steps_per_second": 0.121,
+      "step": 302
+    },
+    {
+      "epoch": 0.5038453543961754,
+      "grad_norm": 2.4241733699272565,
+      "learning_rate": 1.8710904978245894e-06,
+      "loss": 1.0195,
+      "step": 303
+    },
+    {
+      "epoch": 0.5055082103512784,
+      "grad_norm": 2.0299543366481254,
+      "learning_rate": 1.8702286441569203e-06,
+      "loss": 0.9248,
+      "step": 304
+    },
+    {
+      "epoch": 0.5071710663063812,
+      "grad_norm": 2.0464859041290984,
+      "learning_rate": 1.8693641188831328e-06,
+      "loss": 1.0278,
+      "step": 305
+    },
+    {
+      "epoch": 0.508833922261484,
+      "grad_norm": 2.2925599147135043,
+      "learning_rate": 1.8684969246573232e-06,
+      "loss": 1.1409,
+      "step": 306
+    },
+    {
+      "epoch": 0.510496778216587,
+      "grad_norm": 2.4406740096357598,
+      "learning_rate": 1.8676270641417821e-06,
+      "loss": 1.139,
+      "step": 307
+    },
+    {
+      "epoch": 0.5121596341716899,
+      "grad_norm": 2.2502338928571692,
+      "learning_rate": 1.8667545400069856e-06,
+      "loss": 1.2704,
+      "step": 308
+    },
+    {
+      "epoch": 0.5138224901267928,
+      "grad_norm": 2.6218475196698274,
+      "learning_rate": 1.8658793549315868e-06,
+      "loss": 1.1125,
+      "step": 309
+    },
+    {
+      "epoch": 0.5154853460818957,
+      "grad_norm": 2.129495598032236,
+      "learning_rate": 1.8650015116024077e-06,
+      "loss": 1.0833,
+      "step": 310
+    },
+    {
+      "epoch": 0.5171482020369985,
+      "grad_norm": 4.815386485319442,
+      "learning_rate": 1.8641210127144328e-06,
+      "loss": 1.0958,
+      "step": 311
+    },
+    {
+      "epoch": 0.5188110579921015,
+      "grad_norm": 2.1912355764465383,
+      "learning_rate": 1.8632378609707967e-06,
+      "loss": 1.1466,
+      "step": 312
+    },
+    {
+      "epoch": 0.5204739139472043,
+      "grad_norm": 2.6871843307281504,
+      "learning_rate": 1.8623520590827797e-06,
+      "loss": 1.0443,
+      "step": 313
+    },
+    {
+      "epoch": 0.5221367699023072,
+      "grad_norm": 2.229171454549482,
+      "learning_rate": 1.8614636097697983e-06,
+      "loss": 1.2176,
+      "step": 314
+    },
+    {
+      "epoch": 0.5237996258574101,
+      "grad_norm": 2.3499709800000343,
+      "learning_rate": 1.8605725157593957e-06,
+      "loss": 1.0718,
+      "step": 315
+    },
+    {
+      "epoch": 0.525462481812513,
+      "grad_norm": 2.128381081935293,
+      "learning_rate": 1.8596787797872353e-06,
+      "loss": 1.0168,
+      "step": 316
+    },
+    {
+      "epoch": 0.5271253377676158,
+      "grad_norm": 2.4014872942261847,
+      "learning_rate": 1.85878240459709e-06,
+      "loss": 1.1539,
+      "step": 317
+    },
+    {
+      "epoch": 0.5287881937227188,
+      "grad_norm": 2.265996337760467,
+      "learning_rate": 1.857883392940837e-06,
+      "loss": 1.1284,
+      "step": 318
+    },
+    {
+      "epoch": 0.5304510496778216,
+      "grad_norm": 2.1606204777164826,
+      "learning_rate": 1.8569817475784457e-06,
+      "loss": 1.1039,
+      "step": 319
+    },
+    {
+      "epoch": 0.5321139056329246,
+      "grad_norm": 2.314221527250599,
+      "learning_rate": 1.8560774712779719e-06,
+      "loss": 1.1662,
+      "step": 320
+    },
+    {
+      "epoch": 0.5337767615880274,
+      "grad_norm": 2.2036586985620064,
+      "learning_rate": 1.8551705668155479e-06,
+      "loss": 1.0452,
+      "step": 321
+    },
+    {
+      "epoch": 0.5354396175431303,
+      "grad_norm": 2.749989537581693,
+      "learning_rate": 1.8542610369753753e-06,
+      "loss": 1.1471,
+      "step": 322
+    },
+    {
+      "epoch": 0.5371024734982333,
+      "grad_norm": 2.368917726575034,
+      "learning_rate": 1.8533488845497146e-06,
+      "loss": 1.3107,
+      "step": 323
+    },
+    {
+      "epoch": 0.5387653294533361,
+      "grad_norm": 2.1977699456063298,
+      "learning_rate": 1.8524341123388787e-06,
+      "loss": 0.9482,
+      "step": 324
+    },
+    {
+      "epoch": 0.540428185408439,
+      "grad_norm": 2.3335155676558625,
+      "learning_rate": 1.851516723151222e-06,
+      "loss": 1.0011,
+      "step": 325
+    },
+    {
+      "epoch": 0.5420910413635419,
+      "grad_norm": 2.2010547018844857,
+      "learning_rate": 1.850596719803134e-06,
+      "loss": 1.2211,
+      "step": 326
+    },
+    {
+      "epoch": 0.5437538973186448,
+      "grad_norm": 2.077916314192133,
+      "learning_rate": 1.8496741051190298e-06,
+      "loss": 1.1015,
+      "step": 327
+    },
+    {
+      "epoch": 0.5454167532737476,
+      "grad_norm": 2.164449806731655,
+      "learning_rate": 1.84874888193134e-06,
+      "loss": 0.9801,
+      "step": 328
+    },
+    {
+      "epoch": 0.5470796092288506,
+      "grad_norm": 2.098902846890324,
+      "learning_rate": 1.847821053080505e-06,
+      "loss": 0.9848,
+      "step": 329
+    },
+    {
+      "epoch": 0.5487424651839534,
+      "grad_norm": 2.3723967966499253,
+      "learning_rate": 1.8468906214149636e-06,
+      "loss": 1.1727,
+      "step": 330
+    },
+    {
+      "epoch": 0.5504053211390564,
+      "grad_norm": 2.28191518531219,
+      "learning_rate": 1.8459575897911453e-06,
+      "loss": 1.2235,
+      "step": 331
+    },
+    {
+      "epoch": 0.5520681770941592,
+      "grad_norm": 2.2820229639301406,
+      "learning_rate": 1.845021961073462e-06,
+      "loss": 1.2018,
+      "step": 332
+    },
+    {
+      "epoch": 0.5537310330492621,
+      "grad_norm": 2.138213918755252,
+      "learning_rate": 1.844083738134298e-06,
+      "loss": 1.2856,
+      "step": 333
+    },
+    {
+      "epoch": 0.555393889004365,
+      "grad_norm": 2.158628371398021,
+      "learning_rate": 1.8431429238540027e-06,
+      "loss": 1.0644,
+      "step": 334
+    },
+    {
+      "epoch": 0.5570567449594679,
+      "grad_norm": 2.3156623015320377,
+      "learning_rate": 1.84219952112088e-06,
+      "loss": 1.0024,
+      "step": 335
+    },
+    {
+      "epoch": 0.5587196009145707,
+      "grad_norm": 2.1704636280326923,
+      "learning_rate": 1.8412535328311812e-06,
+      "loss": 1.1603,
+      "step": 336
+    },
+    {
+      "epoch": 0.5603824568696737,
+      "grad_norm": 2.324581658968828,
+      "learning_rate": 1.8403049618890948e-06,
+      "loss": 1.2107,
+      "step": 337
+    },
+    {
+      "epoch": 0.5620453128247765,
+      "grad_norm": 2.1819124113613504,
+      "learning_rate": 1.8393538112067376e-06,
+      "loss": 1.0166,
+      "step": 338
+    },
+    {
+      "epoch": 0.5637081687798794,
+      "grad_norm": 2.14128534614559,
+      "learning_rate": 1.8384000837041476e-06,
+      "loss": 1.0348,
+      "step": 339
+    },
+    {
+      "epoch": 0.5653710247349824,
+      "grad_norm": 2.2165793831097034,
+      "learning_rate": 1.8374437823092722e-06,
+      "loss": 1.1544,
+      "step": 340
+    },
+    {
+      "epoch": 0.5670338806900852,
+      "grad_norm": 2.0867661916904456,
+      "learning_rate": 1.8364849099579618e-06,
+      "loss": 1.074,
+      "step": 341
+    },
+    {
+      "epoch": 0.5686967366451882,
+      "grad_norm": 2.278047999899843,
+      "learning_rate": 1.8355234695939585e-06,
+      "loss": 1.1037,
+      "step": 342
+    },
+    {
+      "epoch": 0.570359592600291,
+      "grad_norm": 2.15233238553855,
+      "learning_rate": 1.8345594641688894e-06,
+      "loss": 1.0763,
+      "step": 343
+    },
+    {
+      "epoch": 0.5720224485553939,
+      "grad_norm": 2.1861191898315067,
+      "learning_rate": 1.8335928966422553e-06,
+      "loss": 1.273,
+      "step": 344
+    },
+    {
+      "epoch": 0.5736853045104968,
+      "grad_norm": 2.191986681142867,
+      "learning_rate": 1.8326237699814238e-06,
+      "loss": 1.1571,
+      "step": 345
+    },
+    {
+      "epoch": 0.5753481604655997,
+      "grad_norm": 2.2910127446336737,
+      "learning_rate": 1.8316520871616178e-06,
+      "loss": 1.0243,
+      "step": 346
+    },
+    {
+      "epoch": 0.5770110164207025,
+      "grad_norm": 2.088235032433896,
+      "learning_rate": 1.8306778511659085e-06,
+      "loss": 1.0289,
+      "step": 347
+    },
+    {
+      "epoch": 0.5786738723758055,
+      "grad_norm": 2.213333845589627,
+      "learning_rate": 1.829701064985205e-06,
+      "loss": 1.1229,
+      "step": 348
+    },
+    {
+      "epoch": 0.5803367283309083,
+      "grad_norm": 2.3961880269672626,
+      "learning_rate": 1.8287217316182457e-06,
+      "loss": 0.9937,
+      "step": 349
+    },
+    {
+      "epoch": 0.5819995842860112,
+      "grad_norm": 2.052444834442602,
+      "learning_rate": 1.8277398540715887e-06,
+      "loss": 1.1965,
+      "step": 350
+    },
+    {
+      "epoch": 0.5836624402411141,
+      "grad_norm": 2.410438103487435,
+      "learning_rate": 1.8267554353596024e-06,
+      "loss": 1.1488,
+      "step": 351
+    },
+    {
+      "epoch": 0.585325296196217,
+      "grad_norm": 2.3043831544316835,
+      "learning_rate": 1.8257684785044576e-06,
+      "loss": 0.9968,
+      "step": 352
+    },
+    {
+      "epoch": 0.5869881521513199,
+      "grad_norm": 2.116368023415211,
+      "learning_rate": 1.8247789865361163e-06,
+      "loss": 1.0944,
+      "step": 353
+    },
+    {
+      "epoch": 0.5886510081064228,
+      "grad_norm": 2.0486087552163146,
+      "learning_rate": 1.8237869624923234e-06,
+      "loss": 1.127,
+      "step": 354
+    },
+    {
+      "epoch": 0.5903138640615256,
+      "grad_norm": 2.2869039816178036,
+      "learning_rate": 1.8227924094185978e-06,
+      "loss": 1.0497,
+      "step": 355
+    },
+    {
+      "epoch": 0.5919767200166286,
+      "grad_norm": 2.1915186885351448,
+      "learning_rate": 1.821795330368222e-06,
+      "loss": 1.1584,
+      "step": 356
+    },
+    {
+      "epoch": 0.5936395759717314,
+      "grad_norm": 2.4441273557487566,
+      "learning_rate": 1.8207957284022337e-06,
+      "loss": 1.1206,
+      "step": 357
+    },
+    {
+      "epoch": 0.5953024319268343,
+      "grad_norm": 2.3161936585249334,
+      "learning_rate": 1.8197936065894155e-06,
+      "loss": 1.119,
+      "step": 358
+    },
+    {
+      "epoch": 0.5969652878819373,
+      "grad_norm": 2.303219346593548,
+      "learning_rate": 1.8187889680062863e-06,
+      "loss": 1.2847,
+      "step": 359
+    },
+    {
+      "epoch": 0.5986281438370401,
+      "grad_norm": 2.1218629296829152,
+      "learning_rate": 1.8177818157370912e-06,
+      "loss": 1.0495,
+      "step": 360
+    },
+    {
+      "epoch": 0.600290999792143,
+      "grad_norm": 2.055562774852349,
+      "learning_rate": 1.816772152873793e-06,
+      "loss": 1.1212,
+      "step": 361
+    },
+    {
+      "epoch": 0.6019538557472459,
+      "grad_norm": 2.1036164422163695,
+      "learning_rate": 1.8157599825160607e-06,
+      "loss": 0.9293,
+      "step": 362
+    },
+    {
+      "epoch": 0.6036167117023488,
+      "grad_norm": 2.2429881312173596,
+      "learning_rate": 1.8147453077712634e-06,
+      "loss": 1.134,
+      "step": 363
+    },
+    {
+      "epoch": 0.6052795676574517,
+      "grad_norm": 2.147826876347206,
+      "learning_rate": 1.813728131754456e-06,
+      "loss": 1.0184,
+      "step": 364
+    },
+    {
+      "epoch": 0.6069424236125546,
+      "grad_norm": 2.221807143035665,
+      "learning_rate": 1.8127084575883748e-06,
+      "loss": 1.1232,
+      "step": 365
+    },
+    {
+      "epoch": 0.6086052795676574,
+      "grad_norm": 2.608531400881515,
+      "learning_rate": 1.811686288403424e-06,
+      "loss": 1.166,
+      "step": 366
+    },
+    {
+      "epoch": 0.6102681355227604,
+      "grad_norm": 2.3488837184353937,
+      "learning_rate": 1.8106616273376681e-06,
+      "loss": 1.0413,
+      "step": 367
+    },
+    {
+      "epoch": 0.6119309914778632,
+      "grad_norm": 2.2777707823295668,
+      "learning_rate": 1.8096344775368211e-06,
+      "loss": 1.0941,
+      "step": 368
+    },
+    {
+      "epoch": 0.6135938474329661,
+      "grad_norm": 2.2605130570731253,
+      "learning_rate": 1.808604842154238e-06,
+      "loss": 1.0537,
+      "step": 369
+    },
+    {
+      "epoch": 0.615256703388069,
+      "grad_norm": 2.292940209107732,
+      "learning_rate": 1.807572724350905e-06,
+      "loss": 1.0708,
+      "step": 370
+    },
+    {
+      "epoch": 0.6169195593431719,
+      "grad_norm": 2.10745264157216,
+      "learning_rate": 1.8065381272954276e-06,
+      "loss": 0.9477,
+      "step": 371
+    },
+    {
+      "epoch": 0.6185824152982747,
+      "grad_norm": 2.1624162140056176,
+      "learning_rate": 1.8055010541640243e-06,
+      "loss": 1.0009,
+      "step": 372
+    },
+    {
+      "epoch": 0.6202452712533777,
+      "grad_norm": 2.4110174672424347,
+      "learning_rate": 1.8044615081405151e-06,
+      "loss": 1.082,
+      "step": 373
+    },
+    {
+      "epoch": 0.6219081272084805,
+      "grad_norm": 2.3314352780690166,
+      "learning_rate": 1.8034194924163103e-06,
+      "loss": 1.2243,
+      "step": 374
+    },
+    {
+      "epoch": 0.6235709831635835,
+      "grad_norm": 2.2871677222090137,
+      "learning_rate": 1.802375010190404e-06,
+      "loss": 0.9866,
+      "step": 375
+    },
+    {
+      "epoch": 0.6252338391186864,
+      "grad_norm": 2.3308466745130465,
+      "learning_rate": 1.8013280646693612e-06,
+      "loss": 1.0981,
+      "step": 376
+    },
+    {
+      "epoch": 0.6268966950737892,
+      "grad_norm": 2.211137881920983,
+      "learning_rate": 1.8002786590673096e-06,
+      "loss": 1.0953,
+      "step": 377
+    },
+    {
+      "epoch": 0.6285595510288922,
+      "grad_norm": 2.2653592375727727,
+      "learning_rate": 1.7992267966059298e-06,
+      "loss": 1.0843,
+      "step": 378
+    },
+    {
+      "epoch": 0.630222406983995,
+      "grad_norm": 2.0429297450824024,
+      "learning_rate": 1.7981724805144443e-06,
+      "loss": 1.0853,
+      "step": 379
+    },
+    {
+      "epoch": 0.6318852629390979,
+      "grad_norm": 2.2106979108018794,
+      "learning_rate": 1.7971157140296088e-06,
+      "loss": 1.0946,
+      "step": 380
+    },
+    {
+      "epoch": 0.6335481188942008,
+      "grad_norm": 2.2495780703298025,
+      "learning_rate": 1.7960565003957016e-06,
+      "loss": 1.1758,
+      "step": 381
+    },
+    {
+      "epoch": 0.6352109748493037,
+      "grad_norm": 2.442736100176031,
+      "learning_rate": 1.7949948428645133e-06,
+      "loss": 1.0848,
+      "step": 382
+    },
+    {
+      "epoch": 0.6368738308044065,
+      "grad_norm": 2.1142990741185286,
+      "learning_rate": 1.7939307446953378e-06,
+      "loss": 1.0653,
+      "step": 383
+    },
+    {
+      "epoch": 0.6385366867595095,
+      "grad_norm": 2.0910930966792782,
+      "learning_rate": 1.7928642091549612e-06,
+      "loss": 0.9725,
+      "step": 384
+    },
+    {
+      "epoch": 0.6401995427146123,
+      "grad_norm": 2.278583439400143,
+      "learning_rate": 1.7917952395176535e-06,
+      "loss": 1.0259,
+      "step": 385
+    },
+    {
+      "epoch": 0.6418623986697153,
+      "grad_norm": 2.545286223573369,
+      "learning_rate": 1.790723839065156e-06,
+      "loss": 1.1821,
+      "step": 386
+    },
+    {
+      "epoch": 0.6435252546248181,
+      "grad_norm": 2.261488639334733,
+      "learning_rate": 1.7896500110866737e-06,
+      "loss": 0.939,
+      "step": 387
+    },
+    {
+      "epoch": 0.645188110579921,
+      "grad_norm": 2.7678067419857597,
+      "learning_rate": 1.7885737588788632e-06,
+      "loss": 1.0408,
+      "step": 388
+    },
+    {
+      "epoch": 0.6468509665350239,
+      "grad_norm": 2.11635682407254,
+      "learning_rate": 1.787495085745824e-06,
+      "loss": 1.1696,
+      "step": 389
+    },
+    {
+      "epoch": 0.6485138224901268,
+      "grad_norm": 2.3192534757366703,
+      "learning_rate": 1.7864139949990882e-06,
+      "loss": 0.9589,
+      "step": 390
+    },
+    {
+      "epoch": 0.6501766784452296,
+      "grad_norm": 2.150754996425321,
+      "learning_rate": 1.7853304899576091e-06,
+      "loss": 1.1067,
+      "step": 391
+    },
+    {
+      "epoch": 0.6518395344003326,
+      "grad_norm": 2.2383513192031574,
+      "learning_rate": 1.784244573947753e-06,
+      "loss": 1.3417,
+      "step": 392
+    },
+    {
+      "epoch": 0.6535023903554354,
+      "grad_norm": 2.290710629399966,
+      "learning_rate": 1.7831562503032865e-06,
+      "loss": 1.1442,
+      "step": 393
+    },
+    {
+      "epoch": 0.6551652463105383,
+      "grad_norm": 3.3410043284621143,
+      "learning_rate": 1.7820655223653689e-06,
+      "loss": 1.2796,
+      "step": 394
+    },
+    {
+      "epoch": 0.6568281022656413,
+      "grad_norm": 2.1116250875932727,
+      "learning_rate": 1.7809723934825402e-06,
+      "loss": 1.0086,
+      "step": 395
+    },
+    {
+      "epoch": 0.6584909582207441,
+      "grad_norm": 2.242523763487976,
+      "learning_rate": 1.7798768670107113e-06,
+      "loss": 1.0907,
+      "step": 396
+    },
+    {
+      "epoch": 0.6601538141758471,
+      "grad_norm": 2.151880661955345,
+      "learning_rate": 1.7787789463131535e-06,
+      "loss": 0.9459,
+      "step": 397
+    },
+    {
+      "epoch": 0.6618166701309499,
+      "grad_norm": 2.1313600148722047,
+      "learning_rate": 1.777678634760489e-06,
+      "loss": 1.1144,
+      "step": 398
+    },
+    {
+      "epoch": 0.6634795260860528,
+      "grad_norm": 2.445945952206475,
+      "learning_rate": 1.7765759357306793e-06,
+      "loss": 1.1141,
+      "step": 399
+    },
+    {
+      "epoch": 0.6651423820411557,
+      "grad_norm": 2.1049458345635417,
+      "learning_rate": 1.7754708526090155e-06,
+      "loss": 0.9842,
+      "step": 400
+    },
+    {
+      "epoch": 0.6668052379962586,
+      "grad_norm": 2.4326727008731135,
+      "learning_rate": 1.7743633887881088e-06,
+      "loss": 1.1459,
+      "step": 401
+    },
+    {
+      "epoch": 0.6684680939513614,
+      "grad_norm": 2.2415440353250022,
+      "learning_rate": 1.7732535476678776e-06,
+      "loss": 1.0153,
+      "step": 402
+    },
+    {
+      "epoch": 0.6701309499064644,
+      "grad_norm": 2.6676880307885034,
+      "learning_rate": 1.77214133265554e-06,
+      "loss": 1.216,
+      "step": 403
+    },
+    {
+      "epoch": 0.6717938058615672,
+      "grad_norm": 2.259873210911179,
+      "learning_rate": 1.7710267471656013e-06,
+      "loss": 1.1607,
+      "step": 404
+    },
+    {
+      "epoch": 0.6734566618166701,
+      "grad_norm": 2.1060724600194627,
+      "learning_rate": 1.7699097946198443e-06,
+      "loss": 1.0631,
+      "step": 405
+    },
+    {
+      "epoch": 0.675119517771773,
+      "grad_norm": 2.1821450546259067,
+      "learning_rate": 1.7687904784473186e-06,
+      "loss": 1.0213,
+      "step": 406
+    },
+    {
+      "epoch": 0.6767823737268759,
+      "grad_norm": 2.043314834472435,
+      "learning_rate": 1.7676688020843305e-06,
+      "loss": 1.1173,
+      "step": 407
+    },
+    {
+      "epoch": 0.6784452296819788,
+      "grad_norm": 2.342119646043276,
+      "learning_rate": 1.7665447689744317e-06,
+      "loss": 1.0894,
+      "step": 408
+    },
+    {
+      "epoch": 0.6801080856370817,
+      "grad_norm": 2.0782278008731687,
+      "learning_rate": 1.7654183825684091e-06,
+      "loss": 1.1245,
+      "step": 409
+    },
+    {
+      "epoch": 0.6817709415921845,
+      "grad_norm": 2.0556118504588388,
+      "learning_rate": 1.7642896463242744e-06,
+      "loss": 1.0274,
+      "step": 410
+    },
+    {
+      "epoch": 0.6834337975472875,
+      "grad_norm": 2.4517557799237957,
+      "learning_rate": 1.7631585637072535e-06,
+      "loss": 1.1762,
+      "step": 411
+    },
+    {
+      "epoch": 0.6850966535023904,
+      "grad_norm": 2.2629874321423635,
+      "learning_rate": 1.7620251381897751e-06,
+      "loss": 0.9806,
+      "step": 412
+    },
+    {
+      "epoch": 0.6867595094574932,
+      "grad_norm": 2.33240157213656,
+      "learning_rate": 1.7608893732514615e-06,
+      "loss": 1.156,
+      "step": 413
+    },
+    {
+      "epoch": 0.6884223654125962,
+      "grad_norm": 2.179297842815191,
+      "learning_rate": 1.7597512723791162e-06,
+      "loss": 1.0117,
+      "step": 414
+    },
+    {
+      "epoch": 0.690085221367699,
+      "grad_norm": 2.1949726299406045,
+      "learning_rate": 1.7586108390667142e-06,
+      "loss": 1.0476,
+      "step": 415
+    },
+    {
+      "epoch": 0.6917480773228019,
+      "grad_norm": 2.3336572022760884,
+      "learning_rate": 1.7574680768153915e-06,
+      "loss": 1.1267,
+      "step": 416
+    },
+    {
+      "epoch": 0.6934109332779048,
+      "grad_norm": 2.3983506966023542,
+      "learning_rate": 1.7563229891334336e-06,
+      "loss": 1.034,
+      "step": 417
+    },
+    {
+      "epoch": 0.6950737892330077,
+      "grad_norm": 2.0848820780573476,
+      "learning_rate": 1.7551755795362654e-06,
+      "loss": 0.9663,
+      "step": 418
+    },
+    {
+      "epoch": 0.6967366451881106,
+      "grad_norm": 2.2371795871187063,
+      "learning_rate": 1.7540258515464395e-06,
+      "loss": 1.1261,
+      "step": 419
+    },
+    {
+      "epoch": 0.6983995011432135,
+      "grad_norm": 2.1100352921599437,
+      "learning_rate": 1.7528738086936269e-06,
+      "loss": 1.0711,
+      "step": 420
+    },
+    {
+      "epoch": 0.7000623570983163,
+      "grad_norm": 2.147207122561456,
+      "learning_rate": 1.7517194545146036e-06,
+      "loss": 1.0874,
+      "step": 421
+    },
+    {
+      "epoch": 0.7017252130534193,
+      "grad_norm": 2.1810803049772645,
+      "learning_rate": 1.750562792553244e-06,
+      "loss": 1.0132,
+      "step": 422
+    },
+    {
+      "epoch": 0.7033880690085221,
+      "grad_norm": 2.155938991100877,
+      "learning_rate": 1.7494038263605049e-06,
+      "loss": 1.121,
+      "step": 423
+    },
+    {
+      "epoch": 0.705050924963625,
+      "grad_norm": 2.368347963131549,
+      "learning_rate": 1.7482425594944182e-06,
+      "loss": 1.1172,
+      "step": 424
+    },
+    {
+      "epoch": 0.7067137809187279,
+      "grad_norm": 2.2271099712267013,
+      "learning_rate": 1.7470789955200786e-06,
+      "loss": 1.0263,
+      "step": 425
+    },
+    {
+      "epoch": 0.7083766368738308,
+      "grad_norm": 2.229936898551502,
+      "learning_rate": 1.7459131380096336e-06,
+      "loss": 1.0274,
+      "step": 426
+    },
+    {
+      "epoch": 0.7100394928289336,
+      "grad_norm": 2.158588860368013,
+      "learning_rate": 1.744744990542271e-06,
+      "loss": 1.1544,
+      "step": 427
+    },
+    {
+      "epoch": 0.7117023487840366,
+      "grad_norm": 2.161831442303464,
+      "learning_rate": 1.7435745567042094e-06,
+      "loss": 1.0847,
+      "step": 428
+    },
+    {
+      "epoch": 0.7133652047391394,
+      "grad_norm": 2.169173816530374,
+      "learning_rate": 1.7424018400886858e-06,
+      "loss": 1.042,
+      "step": 429
+    },
+    {
+      "epoch": 0.7150280606942424,
+      "grad_norm": 2.2018091926987964,
+      "learning_rate": 1.7412268442959465e-06,
+      "loss": 1.0173,
+      "step": 430
+    },
+    {
+      "epoch": 0.7166909166493453,
+      "grad_norm": 2.2558519988159045,
+      "learning_rate": 1.7400495729332337e-06,
+      "loss": 1.2228,
+      "step": 431
+    },
+    {
+      "epoch": 0.7183537726044481,
+      "grad_norm": 2.286849860794016,
+      "learning_rate": 1.7388700296147763e-06,
+      "loss": 1.1118,
+      "step": 432
+    },
+    {
+      "epoch": 0.7200166285595511,
+      "grad_norm": 2.376710954298681,
+      "learning_rate": 1.737688217961778e-06,
+      "loss": 1.221,
+      "step": 433
+    },
+    {
+      "epoch": 0.7216794845146539,
+      "grad_norm": 2.165270182256038,
+      "learning_rate": 1.7365041416024063e-06,
+      "loss": 1.0928,
+      "step": 434
+    },
+    {
+      "epoch": 0.7233423404697568,
+      "grad_norm": 2.039272996802367,
+      "learning_rate": 1.7353178041717814e-06,
+      "loss": 1.0993,
+      "step": 435
+    },
+    {
+      "epoch": 0.7250051964248597,
+      "grad_norm": 3.179799411220361,
+      "learning_rate": 1.7341292093119648e-06,
+      "loss": 1.1437,
+      "step": 436
+    },
+    {
+      "epoch": 0.7266680523799626,
+      "grad_norm": 2.438107959779604,
+      "learning_rate": 1.732938360671948e-06,
+      "loss": 1.0467,
+      "step": 437
+    },
+    {
+      "epoch": 0.7283309083350655,
+      "grad_norm": 2.227901031363515,
+      "learning_rate": 1.7317452619076426e-06,
+      "loss": 1.2184,
+      "step": 438
+    },
+    {
+      "epoch": 0.7299937642901684,
+      "grad_norm": 2.484925764202615,
+      "learning_rate": 1.7305499166818679e-06,
+      "loss": 1.0869,
+      "step": 439
+    },
+    {
+      "epoch": 0.7316566202452712,
+      "grad_norm": 2.328525821533016,
+      "learning_rate": 1.7293523286643383e-06,
+      "loss": 1.1372,
+      "step": 440
+    },
+    {
+      "epoch": 0.7333194762003742,
+      "grad_norm": 2.3975391729646014,
+      "learning_rate": 1.7281525015316559e-06,
+      "loss": 1.164,
+      "step": 441
+    },
+    {
+      "epoch": 0.734982332155477,
+      "grad_norm": 2.045522380294782,
+      "learning_rate": 1.726950438967295e-06,
+      "loss": 0.9614,
+      "step": 442
+    },
+    {
+      "epoch": 0.7366451881105799,
+      "grad_norm": 2.233756657460333,
+      "learning_rate": 1.7257461446615939e-06,
+      "loss": 1.2181,
+      "step": 443
+    },
+    {
+      "epoch": 0.7383080440656828,
+      "grad_norm": 2.2137748173388694,
+      "learning_rate": 1.724539622311742e-06,
+      "loss": 1.143,
+      "step": 444
+    },
+    {
+      "epoch": 0.7399709000207857,
+      "grad_norm": 2.384585810876256,
+      "learning_rate": 1.723330875621768e-06,
+      "loss": 1.0785,
+      "step": 445
+    },
+    {
+      "epoch": 0.7416337559758885,
+      "grad_norm": 2.3185815551148585,
+      "learning_rate": 1.7221199083025305e-06,
+      "loss": 1.0606,
+      "step": 446
+    },
+    {
+      "epoch": 0.7432966119309915,
+      "grad_norm": 2.1918576903211053,
+      "learning_rate": 1.7209067240717055e-06,
+      "loss": 1.0824,
+      "step": 447
+    },
+    {
+      "epoch": 0.7449594678860944,
+      "grad_norm": 2.353863396233425,
+      "learning_rate": 1.7196913266537736e-06,
+      "loss": 1.1912,
+      "step": 448
+    },
+    {
+      "epoch": 0.7466223238411973,
+      "grad_norm": 2.0380687806247253,
+      "learning_rate": 1.7184737197800113e-06,
+      "loss": 1.0981,
+      "step": 449
+    },
+    {
+      "epoch": 0.7482851797963002,
+      "grad_norm": 2.5650954285686245,
+      "learning_rate": 1.717253907188477e-06,
+      "loss": 1.0268,
+      "step": 450
+    },
+    {
+      "epoch": 0.749948035751403,
+      "grad_norm": 2.3190693587689464,
+      "learning_rate": 1.7160318926240014e-06,
+      "loss": 1.1847,
+      "step": 451
+    },
+    {
+      "epoch": 0.751610891706506,
+      "grad_norm": 2.1059332821710783,
+      "learning_rate": 1.7148076798381754e-06,
+      "loss": 1.0832,
+      "step": 452
+    },
+    {
+      "epoch": 0.7532737476616088,
+      "grad_norm": 3.056086058439594,
+      "learning_rate": 1.713581272589338e-06,
+      "loss": 1.1343,
+      "step": 453
+    },
+    {
+      "epoch": 0.7532737476616088,
+      "eval_loss": 1.3934024572372437,
+      "eval_runtime": 24.6818,
+      "eval_samples_per_second": 0.446,
+      "eval_steps_per_second": 0.122,
+      "step": 453
+    },
+    {
+      "epoch": 0.7549366036167117,
+      "grad_norm": 2.4488905684680455,
+      "learning_rate": 1.7123526746425649e-06,
+      "loss": 1.0994,
+      "step": 454
+    },
+    {
+      "epoch": 0.7565994595718146,
+      "grad_norm": 2.6403911690075215,
+      "learning_rate": 1.7111218897696585e-06,
+      "loss": 1.0764,
+      "step": 455
+    },
+    {
+      "epoch": 0.7582623155269175,
+      "grad_norm": 2.1327294883754577,
+      "learning_rate": 1.7098889217491336e-06,
+      "loss": 1.1461,
+      "step": 456
+    },
+    {
+      "epoch": 0.7599251714820203,
+      "grad_norm": 2.0936769499266,
+      "learning_rate": 1.7086537743662084e-06,
+      "loss": 1.1827,
+      "step": 457
+    },
+    {
+      "epoch": 0.7615880274371233,
+      "grad_norm": 2.0368197792938814,
+      "learning_rate": 1.707416451412791e-06,
+      "loss": 1.0709,
+      "step": 458
+    },
+    {
+      "epoch": 0.7632508833922261,
+      "grad_norm": 2.1159264469009154,
+      "learning_rate": 1.7061769566874688e-06,
+      "loss": 1.1597,
+      "step": 459
+    },
+    {
+      "epoch": 0.7649137393473291,
+      "grad_norm": 2.2371503881264134,
+      "learning_rate": 1.7049352939954966e-06,
+      "loss": 1.1034,
+      "step": 460
+    },
+    {
+      "epoch": 0.7665765953024319,
+      "grad_norm": 2.167369770231308,
+      "learning_rate": 1.7036914671487849e-06,
+      "loss": 1.1398,
+      "step": 461
+    },
+    {
+      "epoch": 0.7682394512575348,
+      "grad_norm": 2.682964801749785,
+      "learning_rate": 1.7024454799658883e-06,
+      "loss": 1.1015,
+      "step": 462
+    },
+    {
+      "epoch": 0.7699023072126377,
+      "grad_norm": 2.3132284135399903,
+      "learning_rate": 1.7011973362719929e-06,
+      "loss": 0.9757,
+      "step": 463
+    },
+    {
+      "epoch": 0.7715651631677406,
+      "grad_norm": 2.226364690832632,
+      "learning_rate": 1.6999470398989066e-06,
+      "loss": 1.0315,
+      "step": 464
+    },
+    {
+      "epoch": 0.7732280191228434,
+      "grad_norm": 2.333679960095519,
+      "learning_rate": 1.6986945946850445e-06,
+      "loss": 1.1647,
+      "step": 465
+    },
+    {
+      "epoch": 0.7748908750779464,
+      "grad_norm": 2.1371349082975883,
+      "learning_rate": 1.6974400044754198e-06,
+      "loss": 1.0292,
+      "step": 466
+    },
+    {
+      "epoch": 0.7765537310330493,
+      "grad_norm": 2.044267477405619,
+      "learning_rate": 1.6961832731216305e-06,
+      "loss": 1.0265,
+      "step": 467
+    },
+    {
+      "epoch": 0.7782165869881521,
+      "grad_norm": 2.4049976847050987,
+      "learning_rate": 1.694924404481848e-06,
+      "loss": 1.0263,
+      "step": 468
+    },
+    {
+      "epoch": 0.7798794429432551,
+      "grad_norm": 2.1513258273801097,
+      "learning_rate": 1.6936634024208045e-06,
+      "loss": 1.0255,
+      "step": 469
+    },
+    {
+      "epoch": 0.7815422988983579,
+      "grad_norm": 2.267086285542123,
+      "learning_rate": 1.692400270809783e-06,
+      "loss": 1.1037,
+      "step": 470
+    },
+    {
+      "epoch": 0.7832051548534609,
+      "grad_norm": 2.207206827718745,
+      "learning_rate": 1.6911350135266034e-06,
+      "loss": 1.2229,
+      "step": 471
+    },
+    {
+      "epoch": 0.7848680108085637,
+      "grad_norm": 2.3370445622010267,
+      "learning_rate": 1.6898676344556116e-06,
+      "loss": 1.0456,
+      "step": 472
+    },
+    {
+      "epoch": 0.7865308667636666,
+      "grad_norm": 2.694579739504598,
+      "learning_rate": 1.6885981374876675e-06,
+      "loss": 1.1304,
+      "step": 473
+    },
+    {
+      "epoch": 0.7881937227187695,
+      "grad_norm": 2.4654672927693024,
+      "learning_rate": 1.6873265265201329e-06,
+      "loss": 1.1218,
+      "step": 474
+    },
+    {
+      "epoch": 0.7898565786738724,
+      "grad_norm": 2.4497013813355504,
+      "learning_rate": 1.6860528054568596e-06,
+      "loss": 1.188,
+      "step": 475
+    },
+    {
+      "epoch": 0.7915194346289752,
+      "grad_norm": 2.2437903296852517,
+      "learning_rate": 1.684776978208177e-06,
+      "loss": 1.0714,
+      "step": 476
+    },
+    {
+      "epoch": 0.7931822905840782,
+      "grad_norm": 2.430879750782421,
+      "learning_rate": 1.6834990486908816e-06,
+      "loss": 1.1451,
+      "step": 477
+    },
+    {
+      "epoch": 0.794845146539181,
+      "grad_norm": 2.1676565673117265,
+      "learning_rate": 1.6822190208282226e-06,
+      "loss": 1.0458,
+      "step": 478
+    },
+    {
+      "epoch": 0.7965080024942839,
+      "grad_norm": 2.179125948082596,
+      "learning_rate": 1.6809368985498918e-06,
+      "loss": 1.125,
+      "step": 479
+    },
+    {
+      "epoch": 0.7981708584493868,
+      "grad_norm": 2.277708398927927,
+      "learning_rate": 1.679652685792011e-06,
+      "loss": 1.0762,
+      "step": 480
+    },
+    {
+      "epoch": 0.7998337144044897,
+      "grad_norm": 2.2097528761591834,
+      "learning_rate": 1.6783663864971191e-06,
+      "loss": 1.0587,
+      "step": 481
+    },
+    {
+      "epoch": 0.8014965703595927,
+      "grad_norm": 1.9951583919167484,
+      "learning_rate": 1.6770780046141614e-06,
+      "loss": 1.0065,
+      "step": 482
+    },
+    {
+      "epoch": 0.8031594263146955,
+      "grad_norm": 2.1163449529859815,
+      "learning_rate": 1.6757875440984765e-06,
+      "loss": 0.9654,
+      "step": 483
+    },
+    {
+      "epoch": 0.8048222822697984,
+      "grad_norm": 2.169660817074707,
+      "learning_rate": 1.6744950089117845e-06,
+      "loss": 1.109,
+      "step": 484
+    },
+    {
+      "epoch": 0.8064851382249013,
+      "grad_norm": 2.5007409181660596,
+      "learning_rate": 1.6732004030221743e-06,
+      "loss": 1.0073,
+      "step": 485
+    },
+    {
+      "epoch": 0.8081479941800042,
+      "grad_norm": 2.382682682731147,
+      "learning_rate": 1.6719037304040921e-06,
+      "loss": 1.117,
+      "step": 486
+    },
+    {
+      "epoch": 0.809810850135107,
+      "grad_norm": 2.1404776200804827,
+      "learning_rate": 1.6706049950383299e-06,
+      "loss": 1.1485,
+      "step": 487
+    },
+    {
+      "epoch": 0.81147370609021,
+      "grad_norm": 1.9674458939644788,
+      "learning_rate": 1.6693042009120104e-06,
+      "loss": 0.9832,
+      "step": 488
+    },
+    {
+      "epoch": 0.8131365620453128,
+      "grad_norm": 2.428436601155877,
+      "learning_rate": 1.6680013520185786e-06,
+      "loss": 1.1419,
+      "step": 489
+    },
+    {
+      "epoch": 0.8147994180004157,
+      "grad_norm": 2.252054044332263,
+      "learning_rate": 1.6666964523577866e-06,
+      "loss": 1.1008,
+      "step": 490
+    },
+    {
+      "epoch": 0.8164622739555186,
+      "grad_norm": 2.3185844238230167,
+      "learning_rate": 1.6653895059356827e-06,
+      "loss": 1.0961,
+      "step": 491
+    },
+    {
+      "epoch": 0.8181251299106215,
+      "grad_norm": 2.245128772330143,
+      "learning_rate": 1.6640805167645984e-06,
+      "loss": 1.0219,
+      "step": 492
+    },
+    {
+      "epoch": 0.8197879858657244,
+      "grad_norm": 2.2252839355087404,
+      "learning_rate": 1.6627694888631374e-06,
+      "loss": 1.1808,
+      "step": 493
+    },
+    {
+      "epoch": 0.8214508418208273,
+      "grad_norm": 2.4802130292227487,
+      "learning_rate": 1.6614564262561608e-06,
+      "loss": 1.0952,
+      "step": 494
+    },
+    {
+      "epoch": 0.8231136977759301,
+      "grad_norm": 2.0600513070759594,
+      "learning_rate": 1.6601413329747778e-06,
+      "loss": 1.0461,
+      "step": 495
+    },
+    {
+      "epoch": 0.8247765537310331,
+      "grad_norm": 2.129782322403382,
+      "learning_rate": 1.6588242130563308e-06,
+      "loss": 1.0565,
+      "step": 496
+    },
+    {
+      "epoch": 0.8264394096861359,
+      "grad_norm": 2.24596861189698,
+      "learning_rate": 1.657505070544384e-06,
+      "loss": 1.2183,
+      "step": 497
+    },
+    {
+      "epoch": 0.8281022656412388,
+      "grad_norm": 2.2144169319672686,
+      "learning_rate": 1.6561839094887123e-06,
+      "loss": 1.056,
+      "step": 498
+    },
+    {
+      "epoch": 0.8297651215963417,
+      "grad_norm": 2.1326715488699346,
+      "learning_rate": 1.6548607339452852e-06,
+      "loss": 1.081,
+      "step": 499
+    },
+    {
+      "epoch": 0.8314279775514446,
+      "grad_norm": 2.400366586745486,
+      "learning_rate": 1.6535355479762584e-06,
+      "loss": 1.0388,
+      "step": 500
+    },
+    {
+      "epoch": 0.8330908335065474,
+      "grad_norm": 2.101276111938611,
+      "learning_rate": 1.6522083556499595e-06,
+      "loss": 1.13,
+      "step": 501
+    },
+    {
+      "epoch": 0.8347536894616504,
+      "grad_norm": 2.1474700452589017,
+      "learning_rate": 1.6508791610408751e-06,
+      "loss": 1.0743,
+      "step": 502
+    },
+    {
+      "epoch": 0.8364165454167533,
+      "grad_norm": 2.449496114081372,
+      "learning_rate": 1.6495479682296393e-06,
+      "loss": 1.1763,
+      "step": 503
+    },
+    {
+      "epoch": 0.8380794013718562,
+      "grad_norm": 2.2397418249566705,
+      "learning_rate": 1.6482147813030202e-06,
+      "loss": 1.0881,
+      "step": 504
+    },
+    {
+      "epoch": 0.8397422573269591,
+      "grad_norm": 2.38564204606019,
+      "learning_rate": 1.646879604353908e-06,
+      "loss": 0.9649,
+      "step": 505
+    },
+    {
+      "epoch": 0.8414051132820619,
+      "grad_norm": 2.51940590317821,
+      "learning_rate": 1.6455424414813024e-06,
+      "loss": 1.2114,
+      "step": 506
+    },
+    {
+      "epoch": 0.8430679692371649,
+      "grad_norm": 2.3181783821555175,
+      "learning_rate": 1.6442032967903e-06,
+      "loss": 1.1246,
+      "step": 507
+    },
+    {
+      "epoch": 0.8447308251922677,
+      "grad_norm": 2.159422324003234,
+      "learning_rate": 1.6428621743920812e-06,
+      "loss": 1.0351,
+      "step": 508
+    },
+    {
+      "epoch": 0.8463936811473706,
+      "grad_norm": 2.411802254120254,
+      "learning_rate": 1.6415190784038982e-06,
+      "loss": 1.0776,
+      "step": 509
+    },
+    {
+      "epoch": 0.8480565371024735,
+      "grad_norm": 2.0734882608633085,
+      "learning_rate": 1.6401740129490622e-06,
+      "loss": 1.1148,
+      "step": 510
+    },
+    {
+      "epoch": 0.8497193930575764,
+      "grad_norm": 2.318932880803621,
+      "learning_rate": 1.638826982156931e-06,
+      "loss": 1.0853,
+      "step": 511
+    },
+    {
+      "epoch": 0.8513822490126792,
+      "grad_norm": 2.1981900335907247,
+      "learning_rate": 1.637477990162895e-06,
+      "loss": 1.0151,
+      "step": 512
+    },
+    {
+      "epoch": 0.8530451049677822,
+      "grad_norm": 2.121049624802221,
+      "learning_rate": 1.6361270411083665e-06,
+      "loss": 1.0518,
+      "step": 513
+    },
+    {
+      "epoch": 0.854707960922885,
+      "grad_norm": 2.191890004201356,
+      "learning_rate": 1.6347741391407653e-06,
+      "loss": 1.0426,
+      "step": 514
+    },
+    {
+      "epoch": 0.856370816877988,
+      "grad_norm": 2.194218873534906,
+      "learning_rate": 1.6334192884135073e-06,
+      "loss": 1.1111,
+      "step": 515
+    },
+    {
+      "epoch": 0.8580336728330908,
+      "grad_norm": 2.0367103549252983,
+      "learning_rate": 1.6320624930859904e-06,
+      "loss": 1.0521,
+      "step": 516
+    },
+    {
+      "epoch": 0.8596965287881937,
+      "grad_norm": 2.5679919755305973,
+      "learning_rate": 1.630703757323583e-06,
+      "loss": 1.0547,
+      "step": 517
+    },
+    {
+      "epoch": 0.8613593847432967,
+      "grad_norm": 2.2355473603338676,
+      "learning_rate": 1.62934308529761e-06,
+      "loss": 1.0582,
+      "step": 518
+    },
+    {
+      "epoch": 0.8630222406983995,
+      "grad_norm": 2.04568580636059,
+      "learning_rate": 1.6279804811853418e-06,
+      "loss": 0.9508,
+      "step": 519
+    },
+    {
+      "epoch": 0.8646850966535023,
+      "grad_norm": 2.137101454934,
+      "learning_rate": 1.6266159491699787e-06,
+      "loss": 1.05,
+      "step": 520
+    },
+    {
+      "epoch": 0.8663479526086053,
+      "grad_norm": 2.3456035912178126,
+      "learning_rate": 1.6252494934406413e-06,
+      "loss": 1.0513,
+      "step": 521
+    },
+    {
+      "epoch": 0.8680108085637082,
+      "grad_norm": 2.3170057893277645,
+      "learning_rate": 1.6238811181923548e-06,
+      "loss": 1.0833,
+      "step": 522
+    },
+    {
+      "epoch": 0.869673664518811,
+      "grad_norm": 2.03138518708311,
+      "learning_rate": 1.6225108276260384e-06,
+      "loss": 1.0733,
+      "step": 523
+    },
+    {
+      "epoch": 0.871336520473914,
+      "grad_norm": 2.1936879157391225,
+      "learning_rate": 1.6211386259484898e-06,
+      "loss": 1.1578,
+      "step": 524
+    },
+    {
+      "epoch": 0.8729993764290168,
+      "grad_norm": 2.162845347014475,
+      "learning_rate": 1.6197645173723755e-06,
+      "loss": 1.0716,
+      "step": 525
+    },
+    {
+      "epoch": 0.8746622323841198,
+      "grad_norm": 2.2797022460412335,
+      "learning_rate": 1.6183885061162149e-06,
+      "loss": 1.1435,
+      "step": 526
+    },
+    {
+      "epoch": 0.8763250883392226,
+      "grad_norm": 2.8402768433436965,
+      "learning_rate": 1.6170105964043693e-06,
+      "loss": 1.0986,
+      "step": 527
+    },
+    {
+      "epoch": 0.8779879442943255,
+      "grad_norm": 2.2040976685633726,
+      "learning_rate": 1.6156307924670287e-06,
+      "loss": 1.1769,
+      "step": 528
+    },
+    {
+      "epoch": 0.8796508002494284,
+      "grad_norm": 2.2121586340352724,
+      "learning_rate": 1.614249098540197e-06,
+      "loss": 1.0384,
+      "step": 529
+    },
+    {
+      "epoch": 0.8813136562045313,
+      "grad_norm": 2.1737348089144772,
+      "learning_rate": 1.6128655188656818e-06,
+      "loss": 1.1178,
+      "step": 530
+    },
+    {
+      "epoch": 0.8829765121596341,
+      "grad_norm": 2.045692860407394,
+      "learning_rate": 1.6114800576910789e-06,
+      "loss": 1.0526,
+      "step": 531
+    },
+    {
+      "epoch": 0.8846393681147371,
+      "grad_norm": 2.4254673272096094,
+      "learning_rate": 1.610092719269761e-06,
+      "loss": 1.1918,
+      "step": 532
+    },
+    {
+      "epoch": 0.8863022240698399,
+      "grad_norm": 2.2186502378953463,
+      "learning_rate": 1.6087035078608636e-06,
+      "loss": 1.0816,
+      "step": 533
+    },
+    {
+      "epoch": 0.8879650800249428,
+      "grad_norm": 2.1899563735102316,
+      "learning_rate": 1.6073124277292726e-06,
+      "loss": 1.077,
+      "step": 534
+    },
+    {
+      "epoch": 0.8896279359800457,
+      "grad_norm": 2.3420002228619183,
+      "learning_rate": 1.6059194831456105e-06,
+      "loss": 1.046,
+      "step": 535
+    },
+    {
+      "epoch": 0.8912907919351486,
+      "grad_norm": 2.217136239350302,
+      "learning_rate": 1.6045246783862237e-06,
+      "loss": 1.0914,
+      "step": 536
+    },
+    {
+      "epoch": 0.8929536478902516,
+      "grad_norm": 2.6270358774502,
+      "learning_rate": 1.6031280177331704e-06,
+      "loss": 1.2156,
+      "step": 537
+    },
+    {
+      "epoch": 0.8946165038453544,
+      "grad_norm": 2.4382518238543596,
+      "learning_rate": 1.6017295054742044e-06,
+      "loss": 1.0995,
+      "step": 538
+    },
+    {
+      "epoch": 0.8962793598004573,
+      "grad_norm": 2.3688360741829917,
+      "learning_rate": 1.6003291459027653e-06,
+      "loss": 1.0843,
+      "step": 539
+    },
+    {
+      "epoch": 0.8979422157555602,
+      "grad_norm": 2.208126828921951,
+      "learning_rate": 1.5989269433179644e-06,
+      "loss": 1.1147,
+      "step": 540
+    },
+    {
+      "epoch": 0.8996050717106631,
+      "grad_norm": 2.2413007358436188,
+      "learning_rate": 1.5975229020245697e-06,
+      "loss": 1.1003,
+      "step": 541
+    },
+    {
+      "epoch": 0.9012679276657659,
+      "grad_norm": 2.226637939011591,
+      "learning_rate": 1.596117026332995e-06,
+      "loss": 0.9256,
+      "step": 542
+    },
+    {
+      "epoch": 0.9029307836208689,
+      "grad_norm": 2.117810788415696,
+      "learning_rate": 1.5947093205592851e-06,
+      "loss": 1.159,
+      "step": 543
+    },
+    {
+      "epoch": 0.9045936395759717,
+      "grad_norm": 2.0662799008829476,
+      "learning_rate": 1.593299789025104e-06,
+      "loss": 1.0951,
+      "step": 544
+    },
+    {
+      "epoch": 0.9062564955310746,
+      "grad_norm": 2.22383521867496,
+      "learning_rate": 1.5918884360577201e-06,
+      "loss": 1.0437,
+      "step": 545
+    },
+    {
+      "epoch": 0.9079193514861775,
+      "grad_norm": 2.1293539340887735,
+      "learning_rate": 1.5904752659899933e-06,
+      "loss": 1.1039,
+      "step": 546
+    },
+    {
+      "epoch": 0.9095822074412804,
+      "grad_norm": 2.1965800318638724,
+      "learning_rate": 1.5890602831603632e-06,
+      "loss": 1.0456,
+      "step": 547
+    },
+    {
+      "epoch": 0.9112450633963833,
+      "grad_norm": 2.084489657850369,
+      "learning_rate": 1.5876434919128334e-06,
+      "loss": 1.063,
+      "step": 548
+    },
+    {
+      "epoch": 0.9129079193514862,
+      "grad_norm": 2.167327544689684,
+      "learning_rate": 1.5862248965969603e-06,
+      "loss": 1.0414,
+      "step": 549
+    },
+    {
+      "epoch": 0.914570775306589,
+      "grad_norm": 2.0664162881784587,
+      "learning_rate": 1.584804501567838e-06,
+      "loss": 1.0985,
+      "step": 550
+    },
+    {
+      "epoch": 0.916233631261692,
+      "grad_norm": 2.319234065867708,
+      "learning_rate": 1.583382311186086e-06,
+      "loss": 0.9869,
+      "step": 551
+    },
+    {
+      "epoch": 0.9178964872167948,
+      "grad_norm": 2.2524367417375113,
+      "learning_rate": 1.581958329817836e-06,
+      "loss": 0.9564,
+      "step": 552
+    },
+    {
+      "epoch": 0.9195593431718977,
+      "grad_norm": 2.2424034311871184,
+      "learning_rate": 1.5805325618347169e-06,
+      "loss": 1.065,
+      "step": 553
+    },
+    {
+      "epoch": 0.9212221991270007,
+      "grad_norm": 2.337350211180452,
+      "learning_rate": 1.5791050116138438e-06,
+      "loss": 1.0277,
+      "step": 554
+    },
+    {
+      "epoch": 0.9228850550821035,
+      "grad_norm": 2.1419920030575206,
+      "learning_rate": 1.577675683537803e-06,
+      "loss": 0.9841,
+      "step": 555
+    },
+    {
+      "epoch": 0.9245479110372063,
+      "grad_norm": 2.2368638321185865,
+      "learning_rate": 1.5762445819946383e-06,
+      "loss": 1.1817,
+      "step": 556
+    },
+    {
+      "epoch": 0.9262107669923093,
+      "grad_norm": 2.1713305118749955,
+      "learning_rate": 1.5748117113778379e-06,
+      "loss": 1.0755,
+      "step": 557
+    },
+    {
+      "epoch": 0.9278736229474122,
+      "grad_norm": 2.288987960510324,
+      "learning_rate": 1.5733770760863219e-06,
+      "loss": 1.1393,
+      "step": 558
+    },
+    {
+      "epoch": 0.9295364789025151,
+      "grad_norm": 2.2628198071243526,
+      "learning_rate": 1.5719406805244274e-06,
+      "loss": 1.0763,
+      "step": 559
+    },
+    {
+      "epoch": 0.931199334857618,
+      "grad_norm": 2.164816303310072,
+      "learning_rate": 1.570502529101896e-06,
+      "loss": 1.1296,
+      "step": 560
+    },
+    {
+      "epoch": 0.9328621908127208,
+      "grad_norm": 2.198211513279556,
+      "learning_rate": 1.569062626233859e-06,
+      "loss": 1.1306,
+      "step": 561
+    },
+    {
+      "epoch": 0.9345250467678238,
+      "grad_norm": 2.0700768610981766,
+      "learning_rate": 1.5676209763408252e-06,
+      "loss": 1.0939,
+      "step": 562
+    },
+    {
+      "epoch": 0.9361879027229266,
+      "grad_norm": 2.2514400348833927,
+      "learning_rate": 1.5661775838486673e-06,
+      "loss": 1.1552,
+      "step": 563
+    },
+    {
+      "epoch": 0.9378507586780295,
+      "grad_norm": 2.090942303284215,
+      "learning_rate": 1.5647324531886064e-06,
+      "loss": 0.9839,
+      "step": 564
+    },
+    {
+      "epoch": 0.9395136146331324,
+      "grad_norm": 2.140109882341939,
+      "learning_rate": 1.5632855887972007e-06,
+      "loss": 1.2005,
+      "step": 565
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 1.9572836049893243,
+      "learning_rate": 1.5618369951163316e-06,
+      "loss": 1.0442,
+      "step": 566
+    },
+    {
+      "epoch": 0.9428393265433381,
+      "grad_norm": 2.247679565843514,
+      "learning_rate": 1.5603866765931874e-06,
+      "loss": 1.0675,
+      "step": 567
+    },
+    {
+      "epoch": 0.9445021824984411,
+      "grad_norm": 2.16539672713286,
+      "learning_rate": 1.558934637680254e-06,
+      "loss": 1.0969,
+      "step": 568
+    },
+    {
+      "epoch": 0.9461650384535439,
+      "grad_norm": 2.136321120880899,
+      "learning_rate": 1.5574808828352977e-06,
+      "loss": 1.0832,
+      "step": 569
+    },
+    {
+      "epoch": 0.9478278944086469,
+      "grad_norm": 2.35022238962833,
+      "learning_rate": 1.556025416521352e-06,
+      "loss": 1.0707,
+      "step": 570
+    },
+    {
+      "epoch": 0.9494907503637497,
+      "grad_norm": 2.2709844918364186,
+      "learning_rate": 1.5545682432067063e-06,
+      "loss": 1.1081,
+      "step": 571
+    },
+    {
+      "epoch": 0.9511536063188526,
+      "grad_norm": 2.2603031964486706,
+      "learning_rate": 1.5531093673648897e-06,
+      "loss": 1.1301,
+      "step": 572
+    },
+    {
+      "epoch": 0.9528164622739556,
+      "grad_norm": 2.168121102440009,
+      "learning_rate": 1.5516487934746575e-06,
+      "loss": 1.054,
+      "step": 573
+    },
+    {
+      "epoch": 0.9544793182290584,
+      "grad_norm": 2.2875944656650526,
+      "learning_rate": 1.5501865260199794e-06,
+      "loss": 1.0521,
+      "step": 574
+    },
+    {
+      "epoch": 0.9561421741841613,
+      "grad_norm": 2.034798392016362,
+      "learning_rate": 1.5487225694900222e-06,
+      "loss": 0.9311,
+      "step": 575
+    },
+    {
+      "epoch": 0.9578050301392642,
+      "grad_norm": 2.3324178914152967,
+      "learning_rate": 1.547256928379141e-06,
+      "loss": 1.2792,
+      "step": 576
+    },
+    {
+      "epoch": 0.9594678860943671,
+      "grad_norm": 2.12207756608058,
+      "learning_rate": 1.54578960718686e-06,
+      "loss": 1.0741,
+      "step": 577
+    },
+    {
+      "epoch": 0.9611307420494699,
+      "grad_norm": 2.3421431613806085,
+      "learning_rate": 1.5443206104178627e-06,
+      "loss": 1.126,
+      "step": 578
+    },
+    {
+      "epoch": 0.9627935980045729,
+      "grad_norm": 2.2122877182310714,
+      "learning_rate": 1.5428499425819764e-06,
+      "loss": 1.135,
+      "step": 579
+    },
+    {
+      "epoch": 0.9644564539596757,
+      "grad_norm": 2.2047878214189485,
+      "learning_rate": 1.5413776081941578e-06,
+      "loss": 1.1399,
+      "step": 580
+    },
+    {
+      "epoch": 0.9661193099147787,
+      "grad_norm": 2.236152079611479,
+      "learning_rate": 1.5399036117744811e-06,
+      "loss": 1.1442,
+      "step": 581
+    },
+    {
+      "epoch": 0.9677821658698815,
+      "grad_norm": 2.1601189484933476,
+      "learning_rate": 1.538427957848122e-06,
+      "loss": 1.0039,
+      "step": 582
+    },
+    {
+      "epoch": 0.9694450218249844,
+      "grad_norm": 2.060582773193502,
+      "learning_rate": 1.5369506509453455e-06,
+      "loss": 1.0841,
+      "step": 583
+    },
+    {
+      "epoch": 0.9711078777800873,
+      "grad_norm": 2.456082310651417,
+      "learning_rate": 1.5354716956014909e-06,
+      "loss": 1.1203,
+      "step": 584
+    },
+    {
+      "epoch": 0.9727707337351902,
+      "grad_norm": 2.047282112632501,
+      "learning_rate": 1.5339910963569583e-06,
+      "loss": 0.9737,
+      "step": 585
+    },
+    {
+      "epoch": 0.974433589690293,
+      "grad_norm": 2.0171083884034444,
+      "learning_rate": 1.5325088577571937e-06,
+      "loss": 1.0301,
+      "step": 586
+    },
+    {
+      "epoch": 0.976096445645396,
+      "grad_norm": 2.3643508794229513,
+      "learning_rate": 1.5310249843526774e-06,
+      "loss": 1.1398,
+      "step": 587
+    },
+    {
+      "epoch": 0.9777593016004988,
+      "grad_norm": 2.167267117059674,
+      "learning_rate": 1.5295394806989076e-06,
+      "loss": 1.1058,
+      "step": 588
+    },
+    {
+      "epoch": 0.9794221575556017,
+      "grad_norm": 2.1579109019121674,
+      "learning_rate": 1.5280523513563884e-06,
+      "loss": 0.9808,
+      "step": 589
+    },
+    {
+      "epoch": 0.9810850135107047,
+      "grad_norm": 2.2214933213880776,
+      "learning_rate": 1.526563600890613e-06,
+      "loss": 1.2226,
+      "step": 590
+    },
+    {
+      "epoch": 0.9827478694658075,
+      "grad_norm": 2.2694359795676085,
+      "learning_rate": 1.525073233872053e-06,
+      "loss": 1.0298,
+      "step": 591
+    },
+    {
+      "epoch": 0.9844107254209105,
+      "grad_norm": 2.115482333445717,
+      "learning_rate": 1.5235812548761424e-06,
+      "loss": 1.0619,
+      "step": 592
+    },
+    {
+      "epoch": 0.9860735813760133,
+      "grad_norm": 2.2939214488584883,
+      "learning_rate": 1.5220876684832638e-06,
+      "loss": 1.2193,
+      "step": 593
+    },
+    {
+      "epoch": 0.9877364373311162,
+      "grad_norm": 2.145954515190259,
+      "learning_rate": 1.5205924792787344e-06,
+      "loss": 1.1042,
+      "step": 594
+    },
+    {
+      "epoch": 0.9893992932862191,
+      "grad_norm": 2.1711020269771675,
+      "learning_rate": 1.5190956918527925e-06,
+      "loss": 1.0768,
+      "step": 595
+    },
+    {
+      "epoch": 0.991062149241322,
+      "grad_norm": 2.349827442265106,
+      "learning_rate": 1.517597310800582e-06,
+      "loss": 1.1737,
+      "step": 596
+    },
+    {
+      "epoch": 0.9927250051964248,
+      "grad_norm": 2.1677338271015074,
+      "learning_rate": 1.516097340722141e-06,
+      "loss": 1.1894,
+      "step": 597
+    },
+    {
+      "epoch": 0.9943878611515278,
+      "grad_norm": 2.37315650850326,
+      "learning_rate": 1.5145957862223843e-06,
+      "loss": 0.9953,
+      "step": 598
+    },
+    {
+      "epoch": 0.9960507171066306,
+      "grad_norm": 2.115914446964703,
+      "learning_rate": 1.5130926519110914e-06,
+      "loss": 1.1486,
+      "step": 599
+    },
+    {
+      "epoch": 0.9977135730617335,
+      "grad_norm": 2.2944478230358034,
+      "learning_rate": 1.5115879424028918e-06,
+      "loss": 1.01,
+      "step": 600
+    },
+    {
+      "epoch": 0.9993764290168364,
+      "grad_norm": 2.137749774673581,
+      "learning_rate": 1.5100816623172512e-06,
+      "loss": 1.1942,
+      "step": 601
+    },
+    {
+      "epoch": 1.0010392849719394,
+      "grad_norm": 2.1518759789580555,
+      "learning_rate": 1.5085738162784565e-06,
+      "loss": 1.0448,
+      "step": 602
+    },
+    {
+      "epoch": 1.0027021409270422,
+      "grad_norm": 2.0692500287896634,
+      "learning_rate": 1.5070644089156024e-06,
+      "loss": 1.0378,
+      "step": 603
+    },
+    {
+      "epoch": 1.0016614745586707,
+      "grad_norm": 2.230842725197309,
+      "learning_rate": 1.5055534448625764e-06,
+      "loss": 0.9973,
+      "step": 604
+    },
+    {
+      "epoch": 1.0016614745586707,
+      "eval_loss": 1.389662504196167,
+      "eval_runtime": 24.637,
+      "eval_samples_per_second": 0.446,
+      "eval_steps_per_second": 0.122,
+      "step": 604
+    },
+    {
+      "epoch": 1.0033229491173417,
+      "grad_norm": 2.009928207068121,
+      "learning_rate": 1.5040409287580457e-06,
+      "loss": 1.0293,
+      "step": 605
+    },
+    {
+      "epoch": 1.0049844236760124,
+      "grad_norm": 2.120439791656414,
+      "learning_rate": 1.5025268652454421e-06,
+      "loss": 1.048,
+      "step": 606
+    },
+    {
+      "epoch": 1.0066458982346833,
+      "grad_norm": 2.04371411144112,
+      "learning_rate": 1.501011258972948e-06,
+      "loss": 1.0411,
+      "step": 607
+    },
+    {
+      "epoch": 1.008307372793354,
+      "grad_norm": 1.9287047292797497,
+      "learning_rate": 1.4994941145934815e-06,
+      "loss": 0.8436,
+      "step": 608
+    },
+    {
+      "epoch": 1.009968847352025,
+      "grad_norm": 2.0850070547651067,
+      "learning_rate": 1.4979754367646833e-06,
+      "loss": 0.9103,
+      "step": 609
+    },
+    {
+      "epoch": 1.0116303219106957,
+      "grad_norm": 2.0786145289365203,
+      "learning_rate": 1.4964552301489018e-06,
+      "loss": 0.9403,
+      "step": 610
+    },
+    {
+      "epoch": 1.0132917964693666,
+      "grad_norm": 2.2212349830956652,
+      "learning_rate": 1.494933499413178e-06,
+      "loss": 1.1803,
+      "step": 611
+    },
+    {
+      "epoch": 1.0149532710280373,
+      "grad_norm": 2.2692136037661843,
+      "learning_rate": 1.4934102492292336e-06,
+      "loss": 0.9701,
+      "step": 612
+    },
+    {
+      "epoch": 1.0166147455867083,
+      "grad_norm": 2.0584997416327253,
+      "learning_rate": 1.491885484273453e-06,
+      "loss": 1.0582,
+      "step": 613
+    },
+    {
+      "epoch": 1.018276220145379,
+      "grad_norm": 2.259916957406716,
+      "learning_rate": 1.4903592092268726e-06,
+      "loss": 0.9255,
+      "step": 614
+    },
+    {
+      "epoch": 1.01993769470405,
+      "grad_norm": 2.848368582871536,
+      "learning_rate": 1.4888314287751638e-06,
+      "loss": 0.9602,
+      "step": 615
+    },
+    {
+      "epoch": 1.0215991692627207,
+      "grad_norm": 2.760714314244588,
+      "learning_rate": 1.48730214760862e-06,
+      "loss": 0.9578,
+      "step": 616
+    },
+    {
+      "epoch": 1.0232606438213916,
+      "grad_norm": 2.23754202122278,
+      "learning_rate": 1.4857713704221419e-06,
+      "loss": 1.0313,
+      "step": 617
+    },
+    {
+      "epoch": 1.0249221183800623,
+      "grad_norm": 2.348557579974889,
+      "learning_rate": 1.4842391019152225e-06,
+      "loss": 1.0304,
+      "step": 618
+    },
+    {
+      "epoch": 1.026583592938733,
+      "grad_norm": 2.3247250951347485,
+      "learning_rate": 1.482705346791934e-06,
+      "loss": 0.9455,
+      "step": 619
+    },
+    {
+      "epoch": 1.028245067497404,
+      "grad_norm": 2.0361498175354997,
+      "learning_rate": 1.481170109760911e-06,
+      "loss": 0.9218,
+      "step": 620
+    },
+    {
+      "epoch": 1.0299065420560747,
+      "grad_norm": 2.214215564027277,
+      "learning_rate": 1.4796333955353395e-06,
+      "loss": 1.0163,
+      "step": 621
+    },
+    {
+      "epoch": 1.0315680166147456,
+      "grad_norm": 2.1719690959578037,
+      "learning_rate": 1.4780952088329394e-06,
+      "loss": 1.0134,
+      "step": 622
+    },
+    {
+      "epoch": 1.0332294911734163,
+      "grad_norm": 2.2204516984962357,
+      "learning_rate": 1.476555554375951e-06,
+      "loss": 0.9959,
+      "step": 623
+    },
+    {
+      "epoch": 1.0348909657320873,
+      "grad_norm": 2.184066671975778,
+      "learning_rate": 1.4750144368911207e-06,
+      "loss": 0.9662,
+      "step": 624
+    },
+    {
+      "epoch": 1.036552440290758,
+      "grad_norm": 2.265473071872864,
+      "learning_rate": 1.4734718611096874e-06,
+      "loss": 0.893,
+      "step": 625
+    },
+    {
+      "epoch": 1.038213914849429,
+      "grad_norm": 2.2673393121172083,
+      "learning_rate": 1.4719278317673654e-06,
+      "loss": 1.0177,
+      "step": 626
+    },
+    {
+      "epoch": 1.0398753894080996,
+      "grad_norm": 2.3668732645254806,
+      "learning_rate": 1.4703823536043324e-06,
+      "loss": 0.983,
+      "step": 627
+    },
+    {
+      "epoch": 1.0415368639667706,
+      "grad_norm": 2.131498764332704,
+      "learning_rate": 1.468835431365214e-06,
+      "loss": 1.0071,
+      "step": 628
+    },
+    {
+      "epoch": 1.0431983385254413,
+      "grad_norm": 2.149621804291588,
+      "learning_rate": 1.4672870697990686e-06,
+      "loss": 0.9003,
+      "step": 629
+    },
+    {
+      "epoch": 1.0448598130841122,
+      "grad_norm": 2.159649238111743,
+      "learning_rate": 1.4657372736593736e-06,
+      "loss": 1.0028,
+      "step": 630
+    },
+    {
+      "epoch": 1.046521287642783,
+      "grad_norm": 2.1986785581325,
+      "learning_rate": 1.464186047704011e-06,
+      "loss": 1.1123,
+      "step": 631
+    },
+    {
+      "epoch": 1.0481827622014537,
+      "grad_norm": 2.2105863636515726,
+      "learning_rate": 1.4626333966952518e-06,
+      "loss": 0.9999,
+      "step": 632
+    },
+    {
+      "epoch": 1.0498442367601246,
+      "grad_norm": 2.233749646735216,
+      "learning_rate": 1.4610793253997419e-06,
+      "loss": 0.9993,
+      "step": 633
+    },
+    {
+      "epoch": 1.0515057113187953,
+      "grad_norm": 2.143359223529294,
+      "learning_rate": 1.4595238385884878e-06,
+      "loss": 0.9324,
+      "step": 634
+    },
+    {
+      "epoch": 1.0531671858774663,
+      "grad_norm": 2.1270783200542325,
+      "learning_rate": 1.4579669410368412e-06,
+      "loss": 0.913,
+      "step": 635
+    },
+    {
+      "epoch": 1.054828660436137,
+      "grad_norm": 2.311976174877509,
+      "learning_rate": 1.4564086375244854e-06,
+      "loss": 1.0857,
+      "step": 636
+    },
+    {
+      "epoch": 1.056490134994808,
+      "grad_norm": 2.1242450037351333,
+      "learning_rate": 1.4548489328354194e-06,
+      "loss": 1.0415,
+      "step": 637
+    },
+    {
+      "epoch": 1.0581516095534786,
+      "grad_norm": 2.1489145088861554,
+      "learning_rate": 1.4532878317579443e-06,
+      "loss": 1.0571,
+      "step": 638
+    },
+    {
+      "epoch": 1.0598130841121496,
+      "grad_norm": 2.1748830811019055,
+      "learning_rate": 1.451725339084648e-06,
+      "loss": 1.1093,
+      "step": 639
+    },
+    {
+      "epoch": 1.0614745586708203,
+      "grad_norm": 2.038403380979606,
+      "learning_rate": 1.4501614596123897e-06,
+      "loss": 0.8422,
+      "step": 640
+    },
+    {
+      "epoch": 1.0631360332294912,
+      "grad_norm": 2.2259593245678806,
+      "learning_rate": 1.4485961981422882e-06,
+      "loss": 1.125,
+      "step": 641
+    },
+    {
+      "epoch": 1.064797507788162,
+      "grad_norm": 2.14202683520297,
+      "learning_rate": 1.4470295594797028e-06,
+      "loss": 0.9507,
+      "step": 642
+    },
+    {
+      "epoch": 1.066458982346833,
+      "grad_norm": 2.304225430064475,
+      "learning_rate": 1.4454615484342222e-06,
+      "loss": 0.947,
+      "step": 643
+    },
+    {
+      "epoch": 1.0681204569055036,
+      "grad_norm": 2.469583985925475,
+      "learning_rate": 1.4438921698196477e-06,
+      "loss": 0.9516,
+      "step": 644
+    },
+    {
+      "epoch": 1.0697819314641746,
+      "grad_norm": 2.079662445475448,
+      "learning_rate": 1.4423214284539787e-06,
+      "loss": 0.9201,
+      "step": 645
+    },
+    {
+      "epoch": 1.0714434060228453,
+      "grad_norm": 2.2706631611703108,
+      "learning_rate": 1.4407493291593992e-06,
+      "loss": 0.9743,
+      "step": 646
+    },
+    {
+      "epoch": 1.073104880581516,
+      "grad_norm": 2.3017848193416053,
+      "learning_rate": 1.439175876762262e-06,
+      "loss": 1.0577,
+      "step": 647
+    },
+    {
+      "epoch": 1.074766355140187,
+      "grad_norm": 2.118285473386722,
+      "learning_rate": 1.4376010760930727e-06,
+      "loss": 0.8594,
+      "step": 648
+    },
+    {
+      "epoch": 1.0764278296988576,
+      "grad_norm": 2.071039252515359,
+      "learning_rate": 1.4360249319864775e-06,
+      "loss": 0.8875,
+      "step": 649
+    },
+    {
+      "epoch": 1.0780893042575286,
+      "grad_norm": 2.0620465316960934,
+      "learning_rate": 1.434447449281246e-06,
+      "loss": 1.0298,
+      "step": 650
+    },
+    {
+      "epoch": 1.0797507788161993,
+      "grad_norm": 2.3185372556623025,
+      "learning_rate": 1.432868632820258e-06,
+      "loss": 0.9762,
+      "step": 651
+    },
+    {
+      "epoch": 1.0814122533748702,
+      "grad_norm": 2.6906645034252317,
+      "learning_rate": 1.4312884874504873e-06,
+      "loss": 1.0237,
+      "step": 652
+    },
+    {
+      "epoch": 1.083073727933541,
+      "grad_norm": 2.1616057770804815,
+      "learning_rate": 1.4297070180229881e-06,
+      "loss": 0.9281,
+      "step": 653
+    },
+    {
+      "epoch": 1.084735202492212,
+      "grad_norm": 2.117360636691568,
+      "learning_rate": 1.4281242293928787e-06,
+      "loss": 0.8539,
+      "step": 654
+    },
+    {
+      "epoch": 1.0863966770508826,
+      "grad_norm": 2.3065809797011667,
+      "learning_rate": 1.4265401264193284e-06,
+      "loss": 1.0164,
+      "step": 655
+    },
+    {
+      "epoch": 1.0880581516095535,
+      "grad_norm": 2.2835922952642407,
+      "learning_rate": 1.4249547139655408e-06,
+      "loss": 0.9527,
+      "step": 656
+    },
+    {
+      "epoch": 1.0897196261682243,
+      "grad_norm": 2.2551385893064353,
+      "learning_rate": 1.4233679968987392e-06,
+      "loss": 1.189,
+      "step": 657
+    },
+    {
+      "epoch": 1.0913811007268952,
+      "grad_norm": 2.087220306176795,
+      "learning_rate": 1.421779980090153e-06,
+      "loss": 0.9813,
+      "step": 658
+    },
+    {
+      "epoch": 1.093042575285566,
+      "grad_norm": 2.291653419238335,
+      "learning_rate": 1.4201906684150019e-06,
+      "loss": 1.0983,
+      "step": 659
+    },
+    {
+      "epoch": 1.0947040498442369,
+      "grad_norm": 2.174357913738222,
+      "learning_rate": 1.4186000667524794e-06,
+      "loss": 1.0493,
+      "step": 660
+    },
+    {
+      "epoch": 1.0963655244029076,
+      "grad_norm": 2.3975146777189793,
+      "learning_rate": 1.417008179985741e-06,
+      "loss": 1.0219,
+      "step": 661
+    },
+    {
+      "epoch": 1.0980269989615783,
+      "grad_norm": 2.0940561543065135,
+      "learning_rate": 1.4154150130018865e-06,
+      "loss": 1.09,
+      "step": 662
+    },
+    {
+      "epoch": 1.0996884735202492,
+      "grad_norm": 2.376622784313192,
+      "learning_rate": 1.4138205706919459e-06,
+      "loss": 0.9597,
+      "step": 663
+    },
+    {
+      "epoch": 1.10134994807892,
+      "grad_norm": 2.504249815738343,
+      "learning_rate": 1.4122248579508655e-06,
+      "loss": 1.0957,
+      "step": 664
+    },
+    {
+      "epoch": 1.1030114226375909,
+      "grad_norm": 2.126205812177635,
+      "learning_rate": 1.41062787967749e-06,
+      "loss": 0.9934,
+      "step": 665
+    },
+    {
+      "epoch": 1.1046728971962616,
+      "grad_norm": 2.169506367867318,
+      "learning_rate": 1.4090296407745514e-06,
+      "loss": 0.9573,
+      "step": 666
+    },
+    {
+      "epoch": 1.1063343717549325,
+      "grad_norm": 2.2665210150054893,
+      "learning_rate": 1.4074301461486504e-06,
+      "loss": 1.0109,
+      "step": 667
+    },
+    {
+      "epoch": 1.1079958463136033,
+      "grad_norm": 2.154696910395812,
+      "learning_rate": 1.4058294007102431e-06,
+      "loss": 0.9593,
+      "step": 668
+    },
+    {
+      "epoch": 1.1096573208722742,
+      "grad_norm": 2.16252295450875,
+      "learning_rate": 1.4042274093736256e-06,
+      "loss": 1.0709,
+      "step": 669
+    },
+    {
+      "epoch": 1.111318795430945,
+      "grad_norm": 2.0289971880838116,
+      "learning_rate": 1.4026241770569196e-06,
+      "loss": 0.9739,
+      "step": 670
+    },
+    {
+      "epoch": 1.1129802699896159,
+      "grad_norm": 2.060249988117431,
+      "learning_rate": 1.4010197086820552e-06,
+      "loss": 0.9788,
+      "step": 671
+    },
+    {
+      "epoch": 1.1146417445482866,
+      "grad_norm": 2.169402054591631,
+      "learning_rate": 1.3994140091747586e-06,
+      "loss": 1.0504,
+      "step": 672
+    },
+    {
+      "epoch": 1.1163032191069575,
+      "grad_norm": 2.5824180613309844,
+      "learning_rate": 1.3978070834645348e-06,
+      "loss": 1.0102,
+      "step": 673
+    },
+    {
+      "epoch": 1.1179646936656282,
+      "grad_norm": 2.2309401801547173,
+      "learning_rate": 1.3961989364846532e-06,
+      "loss": 0.8809,
+      "step": 674
+    },
+    {
+      "epoch": 1.1196261682242992,
+      "grad_norm": 2.1738166442416134,
+      "learning_rate": 1.3945895731721331e-06,
+      "loss": 0.9891,
+      "step": 675
+    },
+    {
+      "epoch": 1.1212876427829699,
+      "grad_norm": 2.264956154758612,
+      "learning_rate": 1.3929789984677277e-06,
+      "loss": 1.03,
+      "step": 676
+    },
+    {
+      "epoch": 1.1229491173416406,
+      "grad_norm": 2.2123647976197796,
+      "learning_rate": 1.3913672173159085e-06,
+      "loss": 1.0692,
+      "step": 677
+    },
+    {
+      "epoch": 1.1246105919003115,
+      "grad_norm": 2.364489969539715,
+      "learning_rate": 1.3897542346648523e-06,
+      "loss": 1.0008,
+      "step": 678
+    },
+    {
+      "epoch": 1.1262720664589823,
+      "grad_norm": 2.1920721564053407,
+      "learning_rate": 1.3881400554664227e-06,
+      "loss": 0.9403,
+      "step": 679
+    },
+    {
+      "epoch": 1.1279335410176532,
+      "grad_norm": 2.0979599935589643,
+      "learning_rate": 1.386524684676158e-06,
+      "loss": 1.0091,
+      "step": 680
+    },
+    {
+      "epoch": 1.129595015576324,
+      "grad_norm": 2.278663731270055,
+      "learning_rate": 1.3849081272532544e-06,
+      "loss": 1.0526,
+      "step": 681
+    },
+    {
+      "epoch": 1.1312564901349949,
+      "grad_norm": 2.1605440095876753,
+      "learning_rate": 1.3832903881605507e-06,
+      "loss": 1.0356,
+      "step": 682
+    },
+    {
+      "epoch": 1.1329179646936656,
+      "grad_norm": 2.0718604939036136,
+      "learning_rate": 1.381671472364514e-06,
+      "loss": 0.9792,
+      "step": 683
+    },
+    {
+      "epoch": 1.1345794392523365,
+      "grad_norm": 2.1463194438640265,
+      "learning_rate": 1.380051384835223e-06,
+      "loss": 0.9814,
+      "step": 684
+    },
+    {
+      "epoch": 1.1362409138110072,
+      "grad_norm": 2.091677003972167,
+      "learning_rate": 1.3784301305463545e-06,
+      "loss": 1.0427,
+      "step": 685
+    },
+    {
+      "epoch": 1.1379023883696782,
+      "grad_norm": 2.2700678556630494,
+      "learning_rate": 1.3768077144751674e-06,
+      "loss": 0.9369,
+      "step": 686
+    },
+    {
+      "epoch": 1.1395638629283489,
+      "grad_norm": 2.0625458347908943,
+      "learning_rate": 1.3751841416024862e-06,
+      "loss": 0.8808,
+      "step": 687
+    },
+    {
+      "epoch": 1.1412253374870198,
+      "grad_norm": 2.3876992754405912,
+      "learning_rate": 1.3735594169126878e-06,
+      "loss": 0.9346,
+      "step": 688
+    },
+    {
+      "epoch": 1.1428868120456905,
+      "grad_norm": 8.47258353126556,
+      "learning_rate": 1.3719335453936844e-06,
+      "loss": 0.9763,
+      "step": 689
+    },
+    {
+      "epoch": 1.1445482866043615,
+      "grad_norm": 2.1087038863817478,
+      "learning_rate": 1.37030653203691e-06,
+      "loss": 1.159,
+      "step": 690
+    },
+    {
+      "epoch": 1.1462097611630322,
+      "grad_norm": 2.127868180262951,
+      "learning_rate": 1.3686783818373026e-06,
+      "loss": 1.0057,
+      "step": 691
+    },
+    {
+      "epoch": 1.147871235721703,
+      "grad_norm": 2.196072659767744,
+      "learning_rate": 1.367049099793292e-06,
+      "loss": 0.9821,
+      "step": 692
+    },
+    {
+      "epoch": 1.1495327102803738,
+      "grad_norm": 2.199287353415817,
+      "learning_rate": 1.3654186909067815e-06,
+      "loss": 1.0796,
+      "step": 693
+    },
+    {
+      "epoch": 1.1511941848390446,
+      "grad_norm": 2.265025353172786,
+      "learning_rate": 1.3637871601831338e-06,
+      "loss": 0.8848,
+      "step": 694
+    },
+    {
+      "epoch": 1.1528556593977155,
+      "grad_norm": 2.199148208887257,
+      "learning_rate": 1.3621545126311569e-06,
+      "loss": 0.8418,
+      "step": 695
+    },
+    {
+      "epoch": 1.1545171339563862,
+      "grad_norm": 2.1180720226692418,
+      "learning_rate": 1.3605207532630863e-06,
+      "loss": 0.8909,
+      "step": 696
+    },
+    {
+      "epoch": 1.1561786085150572,
+      "grad_norm": 2.2219661239549513,
+      "learning_rate": 1.358885887094571e-06,
+      "loss": 1.1277,
+      "step": 697
+    },
+    {
+      "epoch": 1.1578400830737279,
+      "grad_norm": 2.3215590968893474,
+      "learning_rate": 1.3572499191446578e-06,
+      "loss": 0.9389,
+      "step": 698
+    },
+    {
+      "epoch": 1.1595015576323988,
+      "grad_norm": 2.0426465674972407,
+      "learning_rate": 1.355612854435776e-06,
+      "loss": 0.9844,
+      "step": 699
+    },
+    {
+      "epoch": 1.1611630321910695,
+      "grad_norm": 2.3030330331075555,
+      "learning_rate": 1.3539746979937233e-06,
+      "loss": 0.997,
+      "step": 700
+    },
+    {
+      "epoch": 1.1628245067497405,
+      "grad_norm": 2.2497492700696475,
+      "learning_rate": 1.3523354548476466e-06,
+      "loss": 0.8244,
+      "step": 701
+    },
+    {
+      "epoch": 1.1644859813084112,
+      "grad_norm": 2.585474730795241,
+      "learning_rate": 1.3506951300300309e-06,
+      "loss": 1.0063,
+      "step": 702
+    },
+    {
+      "epoch": 1.1661474558670821,
+      "grad_norm": 2.041245335602706,
+      "learning_rate": 1.3490537285766808e-06,
+      "loss": 1.0547,
+      "step": 703
+    },
+    {
+      "epoch": 1.1678089304257528,
+      "grad_norm": 2.477977632976126,
+      "learning_rate": 1.347411255526707e-06,
+      "loss": 1.0055,
+      "step": 704
+    },
+    {
+      "epoch": 1.1694704049844238,
+      "grad_norm": 2.2319941565239425,
+      "learning_rate": 1.3457677159225095e-06,
+      "loss": 0.9844,
+      "step": 705
+    },
+    {
+      "epoch": 1.1711318795430945,
+      "grad_norm": 2.1658358037506784,
+      "learning_rate": 1.3441231148097628e-06,
+      "loss": 1.0095,
+      "step": 706
+    },
+    {
+      "epoch": 1.1727933541017652,
+      "grad_norm": 2.046058505379159,
+      "learning_rate": 1.3424774572374004e-06,
+      "loss": 1.0553,
+      "step": 707
+    },
+    {
+      "epoch": 1.1744548286604362,
+      "grad_norm": 2.1891816920862364,
+      "learning_rate": 1.340830748257599e-06,
+      "loss": 0.9626,
+      "step": 708
+    },
+    {
+      "epoch": 1.1761163032191069,
+      "grad_norm": 2.866582842225965,
+      "learning_rate": 1.3391829929257623e-06,
+      "loss": 1.0702,
+      "step": 709
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 2.076967565191604,
+      "learning_rate": 1.337534196300508e-06,
+      "loss": 0.9365,
+      "step": 710
+    },
+    {
+      "epoch": 1.1794392523364485,
+      "grad_norm": 2.1835016754122063,
+      "learning_rate": 1.3358843634436495e-06,
+      "loss": 1.0592,
+      "step": 711
+    },
+    {
+      "epoch": 1.1811007268951195,
+      "grad_norm": 2.219299561319535,
+      "learning_rate": 1.3342334994201814e-06,
+      "loss": 1.0805,
+      "step": 712
+    },
+    {
+      "epoch": 1.1827622014537902,
+      "grad_norm": 2.147024039114467,
+      "learning_rate": 1.332581609298264e-06,
+      "loss": 0.9914,
+      "step": 713
+    },
+    {
+      "epoch": 1.1844236760124611,
+      "grad_norm": 2.0325479434123457,
+      "learning_rate": 1.3309286981492082e-06,
+      "loss": 0.8889,
+      "step": 714
+    },
+    {
+      "epoch": 1.1860851505711318,
+      "grad_norm": 2.5059245553533858,
+      "learning_rate": 1.3292747710474592e-06,
+      "loss": 1.1349,
+      "step": 715
+    },
+    {
+      "epoch": 1.1877466251298028,
+      "grad_norm": 2.2414243259263524,
+      "learning_rate": 1.327619833070581e-06,
+      "loss": 0.8834,
+      "step": 716
+    },
+    {
+      "epoch": 1.1894080996884735,
+      "grad_norm": 2.2835729231057336,
+      "learning_rate": 1.3259638892992411e-06,
+      "loss": 1.0753,
+      "step": 717
+    },
+    {
+      "epoch": 1.1910695742471444,
+      "grad_norm": 2.1983108569959553,
+      "learning_rate": 1.3243069448171951e-06,
+      "loss": 0.9993,
+      "step": 718
+    },
+    {
+      "epoch": 1.1927310488058152,
+      "grad_norm": 2.1493768529520776,
+      "learning_rate": 1.3226490047112702e-06,
+      "loss": 0.9992,
+      "step": 719
+    },
+    {
+      "epoch": 1.194392523364486,
+      "grad_norm": 2.1982177635859474,
+      "learning_rate": 1.3209900740713506e-06,
+      "loss": 1.0959,
+      "step": 720
+    },
+    {
+      "epoch": 1.1960539979231568,
+      "grad_norm": 2.8542968450301105,
+      "learning_rate": 1.3193301579903615e-06,
+      "loss": 1.0538,
+      "step": 721
+    },
+    {
+      "epoch": 1.1977154724818275,
+      "grad_norm": 2.1619403628202876,
+      "learning_rate": 1.317669261564253e-06,
+      "loss": 1.0251,
+      "step": 722
+    },
+    {
+      "epoch": 1.1993769470404985,
+      "grad_norm": 2.418169263907194,
+      "learning_rate": 1.3160073898919852e-06,
+      "loss": 1.0045,
+      "step": 723
+    },
+    {
+      "epoch": 1.2010384215991692,
+      "grad_norm": 2.255661060936539,
+      "learning_rate": 1.3143445480755122e-06,
+      "loss": 0.8968,
+      "step": 724
+    },
+    {
+      "epoch": 1.2026998961578401,
+      "grad_norm": 2.272919278531151,
+      "learning_rate": 1.3126807412197664e-06,
+      "loss": 0.9584,
+      "step": 725
+    },
+    {
+      "epoch": 1.2043613707165108,
+      "grad_norm": 2.1840708120633594,
+      "learning_rate": 1.3110159744326426e-06,
+      "loss": 1.0289,
+      "step": 726
+    },
+    {
+      "epoch": 1.2060228452751818,
+      "grad_norm": 2.19563599663801,
+      "learning_rate": 1.3093502528249828e-06,
+      "loss": 0.9625,
+      "step": 727
+    },
+    {
+      "epoch": 1.2076843198338525,
+      "grad_norm": 3.020937698837689,
+      "learning_rate": 1.307683581510561e-06,
+      "loss": 0.8759,
+      "step": 728
+    },
+    {
+      "epoch": 1.2093457943925234,
+      "grad_norm": 2.246654200444507,
+      "learning_rate": 1.3060159656060653e-06,
+      "loss": 1.2153,
+      "step": 729
+    },
+    {
+      "epoch": 1.2110072689511941,
+      "grad_norm": 2.0769588680295774,
+      "learning_rate": 1.304347410231085e-06,
+      "loss": 1.0093,
+      "step": 730
+    },
+    {
+      "epoch": 1.212668743509865,
+      "grad_norm": 2.288118936663974,
+      "learning_rate": 1.3026779205080931e-06,
+      "loss": 1.0355,
+      "step": 731
+    },
+    {
+      "epoch": 1.2143302180685358,
+      "grad_norm": 2.184081985214361,
+      "learning_rate": 1.3010075015624308e-06,
+      "loss": 0.9306,
+      "step": 732
+    },
+    {
+      "epoch": 1.2159916926272065,
+      "grad_norm": 2.1299774896655106,
+      "learning_rate": 1.2993361585222927e-06,
+      "loss": 0.9958,
+      "step": 733
+    },
+    {
+      "epoch": 1.2176531671858775,
+      "grad_norm": 2.175753135707468,
+      "learning_rate": 1.2976638965187094e-06,
+      "loss": 0.9293,
+      "step": 734
+    },
+    {
+      "epoch": 1.2193146417445484,
+      "grad_norm": 2.2923713334477536,
+      "learning_rate": 1.295990720685534e-06,
+      "loss": 1.0258,
+      "step": 735
+    },
+    {
+      "epoch": 1.2209761163032191,
+      "grad_norm": 2.2957849147388525,
+      "learning_rate": 1.294316636159424e-06,
+      "loss": 0.9638,
+      "step": 736
+    },
+    {
+      "epoch": 1.2226375908618898,
+      "grad_norm": 2.250938246343808,
+      "learning_rate": 1.2926416480798267e-06,
+      "loss": 0.8653,
+      "step": 737
+    },
+    {
+      "epoch": 1.2242990654205608,
+      "grad_norm": 2.142325631732433,
+      "learning_rate": 1.2909657615889638e-06,
+      "loss": 0.9697,
+      "step": 738
+    },
+    {
+      "epoch": 1.2259605399792315,
+      "grad_norm": 2.039460360724837,
+      "learning_rate": 1.289288981831815e-06,
+      "loss": 0.9789,
+      "step": 739
+    },
+    {
+      "epoch": 1.2276220145379024,
+      "grad_norm": 2.2252231947797054,
+      "learning_rate": 1.2876113139561018e-06,
+      "loss": 1.1306,
+      "step": 740
+    },
+    {
+      "epoch": 1.2292834890965731,
+      "grad_norm": 2.101017091432986,
+      "learning_rate": 1.285932763112273e-06,
+      "loss": 0.9276,
+      "step": 741
+    },
+    {
+      "epoch": 1.230944963655244,
+      "grad_norm": 2.1947947768274334,
+      "learning_rate": 1.2842533344534875e-06,
+      "loss": 0.9702,
+      "step": 742
+    },
+    {
+      "epoch": 1.2326064382139148,
+      "grad_norm": 2.0326132196001545,
+      "learning_rate": 1.2825730331355995e-06,
+      "loss": 0.868,
+      "step": 743
+    },
+    {
+      "epoch": 1.2342679127725857,
+      "grad_norm": 2.148348415668642,
+      "learning_rate": 1.2808918643171423e-06,
+      "loss": 0.9215,
+      "step": 744
+    },
+    {
+      "epoch": 1.2359293873312565,
+      "grad_norm": 2.2147832136292327,
+      "learning_rate": 1.279209833159312e-06,
+      "loss": 0.9085,
+      "step": 745
+    },
+    {
+      "epoch": 1.2375908618899274,
+      "grad_norm": 2.2716999685750796,
+      "learning_rate": 1.2775269448259524e-06,
+      "loss": 0.9814,
+      "step": 746
+    },
+    {
+      "epoch": 1.2392523364485981,
+      "grad_norm": 2.294321169355599,
+      "learning_rate": 1.275843204483539e-06,
+      "loss": 1.0459,
+      "step": 747
+    },
+    {
+      "epoch": 1.2409138110072688,
+      "grad_norm": 2.0155276917734435,
+      "learning_rate": 1.2741586173011623e-06,
+      "loss": 0.9093,
+      "step": 748
+    },
+    {
+      "epoch": 1.2425752855659398,
+      "grad_norm": 2.176244110291096,
+      "learning_rate": 1.2724731884505134e-06,
+      "loss": 1.0437,
+      "step": 749
+    },
+    {
+      "epoch": 1.2442367601246107,
+      "grad_norm": 2.060993253717314,
+      "learning_rate": 1.2707869231058665e-06,
+      "loss": 0.9021,
+      "step": 750
+    },
+    {
+      "epoch": 1.2458982346832814,
+      "grad_norm": 2.2567394830685688,
+      "learning_rate": 1.2690998264440651e-06,
+      "loss": 1.0458,
+      "step": 751
+    },
+    {
+      "epoch": 1.2475597092419521,
+      "grad_norm": 2.1356542645695917,
+      "learning_rate": 1.2674119036445034e-06,
+      "loss": 0.8728,
+      "step": 752
+    },
+    {
+      "epoch": 1.249221183800623,
+      "grad_norm": 2.146513267279053,
+      "learning_rate": 1.2657231598891125e-06,
+      "loss": 0.9412,
+      "step": 753
+    },
+    {
+      "epoch": 1.2508826583592938,
+      "grad_norm": 2.237050919851434,
+      "learning_rate": 1.2640336003623442e-06,
+      "loss": 0.9418,
+      "step": 754
+    },
+    {
+      "epoch": 1.2525441329179647,
+      "grad_norm": 2.309743390797678,
+      "learning_rate": 1.2623432302511542e-06,
+      "loss": 1.1053,
+      "step": 755
+    },
+    {
+      "epoch": 1.2525441329179647,
+      "eval_loss": 1.3914175033569336,
+      "eval_runtime": 24.4622,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.123,
+      "step": 755
+    },
+    {
+      "epoch": 1.257528556593977,
+      "grad_norm": 2.194516385421637,
+      "learning_rate": 1.260652054744987e-06,
+      "loss": 1.0752,
+      "step": 756
+    },
+    {
+      "epoch": 1.259190031152648,
+      "grad_norm": 2.0798826656496514,
+      "learning_rate": 1.258960079035759e-06,
+      "loss": 1.0474,
+      "step": 757
+    },
+    {
+      "epoch": 1.2608515057113188,
+      "grad_norm": 2.4630029006380663,
+      "learning_rate": 1.2572673083178447e-06,
+      "loss": 1.0878,
+      "step": 758
+    },
+    {
+      "epoch": 1.2625129802699897,
+      "grad_norm": 2.059087102554878,
+      "learning_rate": 1.2555737477880575e-06,
+      "loss": 0.9406,
+      "step": 759
+    },
+    {
+      "epoch": 1.2641744548286604,
+      "grad_norm": 2.0994176763702717,
+      "learning_rate": 1.2538794026456365e-06,
+      "loss": 0.7893,
+      "step": 760
+    },
+    {
+      "epoch": 1.2658359293873311,
+      "grad_norm": 2.2656607881570534,
+      "learning_rate": 1.2521842780922298e-06,
+      "loss": 0.9353,
+      "step": 761
+    },
+    {
+      "epoch": 1.267497403946002,
+      "grad_norm": 2.2731202307003735,
+      "learning_rate": 1.2504883793318777e-06,
+      "loss": 0.9712,
+      "step": 762
+    },
+    {
+      "epoch": 1.269158878504673,
+      "grad_norm": 2.199059936980419,
+      "learning_rate": 1.2487917115709973e-06,
+      "loss": 0.8983,
+      "step": 763
+    },
+    {
+      "epoch": 1.2708203530633437,
+      "grad_norm": 2.3268110070781787,
+      "learning_rate": 1.2470942800183674e-06,
+      "loss": 0.881,
+      "step": 764
+    },
+    {
+      "epoch": 1.2724818276220144,
+      "grad_norm": 2.4916547720738245,
+      "learning_rate": 1.2453960898851105e-06,
+      "loss": 1.0081,
+      "step": 765
+    },
+    {
+      "epoch": 1.2741433021806854,
+      "grad_norm": 2.2962920389351407,
+      "learning_rate": 1.2436971463846788e-06,
+      "loss": 0.9745,
+      "step": 766
+    },
+    {
+      "epoch": 1.275804776739356,
+      "grad_norm": 2.2101054126420863,
+      "learning_rate": 1.2419974547328364e-06,
+      "loss": 0.9563,
+      "step": 767
+    },
+    {
+      "epoch": 1.277466251298027,
+      "grad_norm": 2.2891187380085913,
+      "learning_rate": 1.2402970201476457e-06,
+      "loss": 0.9303,
+      "step": 768
+    },
+    {
+      "epoch": 1.2791277258566978,
+      "grad_norm": 2.1360488073197024,
+      "learning_rate": 1.2385958478494484e-06,
+      "loss": 0.8754,
+      "step": 769
+    },
+    {
+      "epoch": 1.2807892004153687,
+      "grad_norm": 2.19618235102571,
+      "learning_rate": 1.236893943060852e-06,
+      "loss": 0.9035,
+      "step": 770
+    },
+    {
+      "epoch": 1.2824506749740394,
+      "grad_norm": 2.134112006128728,
+      "learning_rate": 1.235191311006712e-06,
+      "loss": 0.9439,
+      "step": 771
+    },
+    {
+      "epoch": 1.2841121495327104,
+      "grad_norm": 3.1612461680428816,
+      "learning_rate": 1.2334879569141172e-06,
+      "loss": 0.8783,
+      "step": 772
+    },
+    {
+      "epoch": 1.285773624091381,
+      "grad_norm": 1.9962401932268041,
+      "learning_rate": 1.231783886012373e-06,
+      "loss": 0.9263,
+      "step": 773
+    },
+    {
+      "epoch": 1.287435098650052,
+      "grad_norm": 2.1827109534797695,
+      "learning_rate": 1.230079103532985e-06,
+      "loss": 0.9417,
+      "step": 774
+    },
+    {
+      "epoch": 1.2890965732087227,
+      "grad_norm": 2.7468270854753336,
+      "learning_rate": 1.228373614709644e-06,
+      "loss": 1.004,
+      "step": 775
+    },
+    {
+      "epoch": 1.2907580477673934,
+      "grad_norm": 2.1747453190202646,
+      "learning_rate": 1.2266674247782086e-06,
+      "loss": 0.9531,
+      "step": 776
+    },
+    {
+      "epoch": 1.2924195223260644,
+      "grad_norm": 2.085633930937514,
+      "learning_rate": 1.2249605389766895e-06,
+      "loss": 1.0289,
+      "step": 777
+    },
+    {
+      "epoch": 1.2940809968847353,
+      "grad_norm": 2.2021065356232317,
+      "learning_rate": 1.223252962545235e-06,
+      "loss": 0.8765,
+      "step": 778
+    },
+    {
+      "epoch": 1.295742471443406,
+      "grad_norm": 2.1829312571834896,
+      "learning_rate": 1.2215447007261133e-06,
+      "loss": 1.0022,
+      "step": 779
+    },
+    {
+      "epoch": 1.2974039460020768,
+      "grad_norm": 2.1283332483638286,
+      "learning_rate": 1.2198357587636956e-06,
+      "loss": 0.9697,
+      "step": 780
+    },
+    {
+      "epoch": 1.2990654205607477,
+      "grad_norm": 2.525909866186781,
+      "learning_rate": 1.2181261419044426e-06,
+      "loss": 0.9726,
+      "step": 781
+    },
+    {
+      "epoch": 1.3007268951194184,
+      "grad_norm": 2.1648814032733186,
+      "learning_rate": 1.2164158553968855e-06,
+      "loss": 1.0231,
+      "step": 782
+    },
+    {
+      "epoch": 1.3023883696780894,
+      "grad_norm": 2.2334029411110556,
+      "learning_rate": 1.2147049044916128e-06,
+      "loss": 1.1608,
+      "step": 783
+    },
+    {
+      "epoch": 1.30404984423676,
+      "grad_norm": 2.3547357866951093,
+      "learning_rate": 1.2129932944412518e-06,
+      "loss": 0.9692,
+      "step": 784
+    },
+    {
+      "epoch": 1.305711318795431,
+      "grad_norm": 2.2065719120739726,
+      "learning_rate": 1.2112810305004535e-06,
+      "loss": 0.8479,
+      "step": 785
+    },
+    {
+      "epoch": 1.3073727933541017,
+      "grad_norm": 2.125915368197623,
+      "learning_rate": 1.2095681179258764e-06,
+      "loss": 0.9621,
+      "step": 786
+    },
+    {
+      "epoch": 1.3090342679127727,
+      "grad_norm": 2.645809087395215,
+      "learning_rate": 1.2078545619761702e-06,
+      "loss": 1.0085,
+      "step": 787
+    },
+    {
+      "epoch": 1.3106957424714434,
+      "grad_norm": 2.250819574066223,
+      "learning_rate": 1.2061403679119601e-06,
+      "loss": 0.9068,
+      "step": 788
+    },
+    {
+      "epoch": 1.3123572170301143,
+      "grad_norm": 2.21749937163358,
+      "learning_rate": 1.2044255409958303e-06,
+      "loss": 0.9283,
+      "step": 789
+    },
+    {
+      "epoch": 1.314018691588785,
+      "grad_norm": 2.2743487946390992,
+      "learning_rate": 1.2027100864923075e-06,
+      "loss": 0.9947,
+      "step": 790
+    },
+    {
+      "epoch": 1.3156801661474558,
+      "grad_norm": 2.6923629420708646,
+      "learning_rate": 1.200994009667845e-06,
+      "loss": 0.9471,
+      "step": 791
+    },
+    {
+      "epoch": 1.3173416407061267,
+      "grad_norm": 2.3195724880810102,
+      "learning_rate": 1.1992773157908072e-06,
+      "loss": 0.9601,
+      "step": 792
+    },
+    {
+      "epoch": 1.3190031152647976,
+      "grad_norm": 2.0807605681167405,
+      "learning_rate": 1.1975600101314525e-06,
+      "loss": 0.8885,
+      "step": 793
+    },
+    {
+      "epoch": 1.3206645898234683,
+      "grad_norm": 2.0900202951082147,
+      "learning_rate": 1.1958420979619175e-06,
+      "loss": 1.0104,
+      "step": 794
+    },
+    {
+      "epoch": 1.322326064382139,
+      "grad_norm": 2.183167054296833,
+      "learning_rate": 1.1941235845562005e-06,
+      "loss": 0.9351,
+      "step": 795
+    },
+    {
+      "epoch": 1.32398753894081,
+      "grad_norm": 2.094498837050807,
+      "learning_rate": 1.1924044751901464e-06,
+      "loss": 0.995,
+      "step": 796
+    },
+    {
+      "epoch": 1.3256490134994807,
+      "grad_norm": 2.1534518031095957,
+      "learning_rate": 1.1906847751414291e-06,
+      "loss": 1.0239,
+      "step": 797
+    },
+    {
+      "epoch": 1.3273104880581517,
+      "grad_norm": 2.4401111022025557,
+      "learning_rate": 1.188964489689536e-06,
+      "loss": 1.0308,
+      "step": 798
+    },
+    {
+      "epoch": 1.3289719626168224,
+      "grad_norm": 2.6081491691522873,
+      "learning_rate": 1.1872436241157518e-06,
+      "loss": 1.1202,
+      "step": 799
+    },
+    {
+      "epoch": 1.3306334371754933,
+      "grad_norm": 2.3979641911118383,
+      "learning_rate": 1.1855221837031418e-06,
+      "loss": 0.9712,
+      "step": 800
+    },
+    {
+      "epoch": 1.332294911734164,
+      "grad_norm": 2.3288896242545563,
+      "learning_rate": 1.1838001737365363e-06,
+      "loss": 0.9857,
+      "step": 801
+    },
+    {
+      "epoch": 1.333956386292835,
+      "grad_norm": 2.2743604591828714,
+      "learning_rate": 1.1820775995025146e-06,
+      "loss": 1.0457,
+      "step": 802
+    },
+    {
+      "epoch": 1.3356178608515057,
+      "grad_norm": 2.2419016690025106,
+      "learning_rate": 1.1803544662893875e-06,
+      "loss": 1.0037,
+      "step": 803
+    },
+    {
+      "epoch": 1.3372793354101766,
+      "grad_norm": 2.097257183297578,
+      "learning_rate": 1.1786307793871823e-06,
+      "loss": 1.0137,
+      "step": 804
+    },
+    {
+      "epoch": 1.3389408099688473,
+      "grad_norm": 2.0525680314921115,
+      "learning_rate": 1.1769065440876263e-06,
+      "loss": 0.941,
+      "step": 805
+    },
+    {
+      "epoch": 1.340602284527518,
+      "grad_norm": 2.1484754300130797,
+      "learning_rate": 1.1751817656841297e-06,
+      "loss": 0.8931,
+      "step": 806
+    },
+    {
+      "epoch": 1.342263759086189,
+      "grad_norm": 2.049973872619327,
+      "learning_rate": 1.1734564494717708e-06,
+      "loss": 0.9391,
+      "step": 807
+    },
+    {
+      "epoch": 1.34392523364486,
+      "grad_norm": 2.1411969844449117,
+      "learning_rate": 1.171730600747279e-06,
+      "loss": 0.8776,
+      "step": 808
+    },
+    {
+      "epoch": 1.3455867082035307,
+      "grad_norm": 2.333385888491211,
+      "learning_rate": 1.1700042248090174e-06,
+      "loss": 0.8865,
+      "step": 809
+    },
+    {
+      "epoch": 1.3472481827622014,
+      "grad_norm": 2.172814651117749,
+      "learning_rate": 1.1682773269569692e-06,
+      "loss": 1.1343,
+      "step": 810
+    },
+    {
+      "epoch": 1.3489096573208723,
+      "grad_norm": 2.106719631129538,
+      "learning_rate": 1.1665499124927182e-06,
+      "loss": 0.9706,
+      "step": 811
+    },
+    {
+      "epoch": 1.350571131879543,
+      "grad_norm": 2.2297462633284812,
+      "learning_rate": 1.164821986719436e-06,
+      "loss": 0.9913,
+      "step": 812
+    },
+    {
+      "epoch": 1.352232606438214,
+      "grad_norm": 2.617400755249897,
+      "learning_rate": 1.1630935549418626e-06,
+      "loss": 0.9445,
+      "step": 813
+    },
+    {
+      "epoch": 1.3538940809968847,
+      "grad_norm": 2.1827702013845895,
+      "learning_rate": 1.161364622466292e-06,
+      "loss": 0.9375,
+      "step": 814
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 2.2016299775994876,
+      "learning_rate": 1.159635194600555e-06,
+      "loss": 0.9031,
+      "step": 815
+    },
+    {
+      "epoch": 1.3572170301142263,
+      "grad_norm": 2.070277398005119,
+      "learning_rate": 1.157905276654004e-06,
+      "loss": 0.9009,
+      "step": 816
+    },
+    {
+      "epoch": 1.358878504672897,
+      "grad_norm": 2.141179928339341,
+      "learning_rate": 1.1561748739374944e-06,
+      "loss": 0.9099,
+      "step": 817
+    },
+    {
+      "epoch": 1.360539979231568,
+      "grad_norm": 2.007914206117487,
+      "learning_rate": 1.1544439917633716e-06,
+      "loss": 0.892,
+      "step": 818
+    },
+    {
+      "epoch": 1.362201453790239,
+      "grad_norm": 2.0648321350070526,
+      "learning_rate": 1.1527126354454525e-06,
+      "loss": 0.958,
+      "step": 819
+    },
+    {
+      "epoch": 1.3638629283489097,
+      "grad_norm": 2.0018244208752565,
+      "learning_rate": 1.1509808102990085e-06,
+      "loss": 0.9431,
+      "step": 820
+    },
+    {
+      "epoch": 1.3655244029075804,
+      "grad_norm": 2.361320004069871,
+      "learning_rate": 1.1492485216407513e-06,
+      "loss": 1.0479,
+      "step": 821
+    },
+    {
+      "epoch": 1.3671858774662513,
+      "grad_norm": 2.6806488863596862,
+      "learning_rate": 1.1475157747888158e-06,
+      "loss": 0.9181,
+      "step": 822
+    },
+    {
+      "epoch": 1.3688473520249222,
+      "grad_norm": 2.1942518382547713,
+      "learning_rate": 1.145782575062743e-06,
+      "loss": 0.9427,
+      "step": 823
+    },
+    {
+      "epoch": 1.370508826583593,
+      "grad_norm": 2.122601010531213,
+      "learning_rate": 1.1440489277834645e-06,
+      "loss": 1.0671,
+      "step": 824
+    },
+    {
+      "epoch": 1.3721703011422637,
+      "grad_norm": 1.9965619962296717,
+      "learning_rate": 1.1423148382732853e-06,
+      "loss": 0.9456,
+      "step": 825
+    },
+    {
+      "epoch": 1.3738317757009346,
+      "grad_norm": 2.318013669509405,
+      "learning_rate": 1.1405803118558687e-06,
+      "loss": 0.9946,
+      "step": 826
+    },
+    {
+      "epoch": 1.3754932502596053,
+      "grad_norm": 2.1436011385350056,
+      "learning_rate": 1.1388453538562195e-06,
+      "loss": 0.9548,
+      "step": 827
+    },
+    {
+      "epoch": 1.3771547248182763,
+      "grad_norm": 2.1480524834330588,
+      "learning_rate": 1.137109969600667e-06,
+      "loss": 1.0541,
+      "step": 828
+    },
+    {
+      "epoch": 1.378816199376947,
+      "grad_norm": 2.417891031584686,
+      "learning_rate": 1.1353741644168487e-06,
+      "loss": 0.9857,
+      "step": 829
+    },
+    {
+      "epoch": 1.380477673935618,
+      "grad_norm": 2.1577692143799516,
+      "learning_rate": 1.1336379436336953e-06,
+      "loss": 1.0296,
+      "step": 830
+    },
+    {
+      "epoch": 1.3821391484942886,
+      "grad_norm": 2.3755322611919807,
+      "learning_rate": 1.131901312581413e-06,
+      "loss": 1.0684,
+      "step": 831
+    },
+    {
+      "epoch": 1.3838006230529594,
+      "grad_norm": 2.193724004809641,
+      "learning_rate": 1.1301642765914672e-06,
+      "loss": 0.995,
+      "step": 832
+    },
+    {
+      "epoch": 1.3854620976116303,
+      "grad_norm": 2.084892517254851,
+      "learning_rate": 1.1284268409965671e-06,
+      "loss": 1.0584,
+      "step": 833
+    },
+    {
+      "epoch": 1.3871235721703012,
+      "grad_norm": 2.107802237296406,
+      "learning_rate": 1.1266890111306483e-06,
+      "loss": 0.9933,
+      "step": 834
+    },
+    {
+      "epoch": 1.388785046728972,
+      "grad_norm": 2.2944459376176116,
+      "learning_rate": 1.1249507923288561e-06,
+      "loss": 0.8813,
+      "step": 835
+    },
+    {
+      "epoch": 1.3904465212876427,
+      "grad_norm": 2.4275620341058697,
+      "learning_rate": 1.1232121899275313e-06,
+      "loss": 1.0109,
+      "step": 836
+    },
+    {
+      "epoch": 1.3921079958463136,
+      "grad_norm": 2.213400646410023,
+      "learning_rate": 1.1214732092641914e-06,
+      "loss": 1.0679,
+      "step": 837
+    },
+    {
+      "epoch": 1.3937694704049846,
+      "grad_norm": 2.1158008518205724,
+      "learning_rate": 1.1197338556775155e-06,
+      "loss": 0.945,
+      "step": 838
+    },
+    {
+      "epoch": 1.3954309449636553,
+      "grad_norm": 2.2576033887254674,
+      "learning_rate": 1.1179941345073277e-06,
+      "loss": 1.1281,
+      "step": 839
+    },
+    {
+      "epoch": 1.397092419522326,
+      "grad_norm": 2.174956086888558,
+      "learning_rate": 1.1162540510945798e-06,
+      "loss": 0.9392,
+      "step": 840
+    },
+    {
+      "epoch": 1.398753894080997,
+      "grad_norm": 2.2102299571149473,
+      "learning_rate": 1.1145136107813361e-06,
+      "loss": 0.8779,
+      "step": 841
+    },
+    {
+      "epoch": 1.4004153686396676,
+      "grad_norm": 2.293662364545717,
+      "learning_rate": 1.1127728189107574e-06,
+      "loss": 0.8608,
+      "step": 842
+    },
+    {
+      "epoch": 1.4020768431983386,
+      "grad_norm": 2.2613684494376756,
+      "learning_rate": 1.111031680827083e-06,
+      "loss": 0.8828,
+      "step": 843
+    },
+    {
+      "epoch": 1.4037383177570093,
+      "grad_norm": 2.1319924720800696,
+      "learning_rate": 1.1092902018756148e-06,
+      "loss": 1.0049,
+      "step": 844
+    },
+    {
+      "epoch": 1.4053997923156802,
+      "grad_norm": 2.28021887844631,
+      "learning_rate": 1.1075483874027018e-06,
+      "loss": 0.9074,
+      "step": 845
+    },
+    {
+      "epoch": 1.407061266874351,
+      "grad_norm": 2.1484668932462445,
+      "learning_rate": 1.1058062427557228e-06,
+      "loss": 1.0847,
+      "step": 846
+    },
+    {
+      "epoch": 1.4087227414330217,
+      "grad_norm": 2.2205863989831416,
+      "learning_rate": 1.10406377328307e-06,
+      "loss": 0.9719,
+      "step": 847
+    },
+    {
+      "epoch": 1.4103842159916926,
+      "grad_norm": 2.1508101547003564,
+      "learning_rate": 1.1023209843341332e-06,
+      "loss": 1.0049,
+      "step": 848
+    },
+    {
+      "epoch": 1.4120456905503636,
+      "grad_norm": 2.3302734390720397,
+      "learning_rate": 1.1005778812592832e-06,
+      "loss": 1.1868,
+      "step": 849
+    },
+    {
+      "epoch": 1.4137071651090343,
+      "grad_norm": 2.1781701265290327,
+      "learning_rate": 1.0988344694098544e-06,
+      "loss": 0.8906,
+      "step": 850
+    },
+    {
+      "epoch": 1.415368639667705,
+      "grad_norm": 2.226671789203206,
+      "learning_rate": 1.0970907541381294e-06,
+      "loss": 0.929,
+      "step": 851
+    },
+    {
+      "epoch": 1.417030114226376,
+      "grad_norm": 2.5515388989279866,
+      "learning_rate": 1.095346740797323e-06,
+      "loss": 0.8837,
+      "step": 852
+    },
+    {
+      "epoch": 1.4186915887850469,
+      "grad_norm": 2.236679949130665,
+      "learning_rate": 1.0936024347415642e-06,
+      "loss": 0.9804,
+      "step": 853
+    },
+    {
+      "epoch": 1.4203530633437176,
+      "grad_norm": 2.9612492685412852,
+      "learning_rate": 1.091857841325881e-06,
+      "loss": 0.9525,
+      "step": 854
+    },
+    {
+      "epoch": 1.4220145379023883,
+      "grad_norm": 2.5666504315216137,
+      "learning_rate": 1.0901129659061837e-06,
+      "loss": 0.8627,
+      "step": 855
+    },
+    {
+      "epoch": 1.4236760124610592,
+      "grad_norm": 2.4015395133256106,
+      "learning_rate": 1.0883678138392475e-06,
+      "loss": 0.9089,
+      "step": 856
+    },
+    {
+      "epoch": 1.42533748701973,
+      "grad_norm": 2.3067956889994097,
+      "learning_rate": 1.0866223904826989e-06,
+      "loss": 0.9487,
+      "step": 857
+    },
+    {
+      "epoch": 1.426998961578401,
+      "grad_norm": 2.0391869701955647,
+      "learning_rate": 1.084876701194995e-06,
+      "loss": 1.0349,
+      "step": 858
+    },
+    {
+      "epoch": 1.4286604361370716,
+      "grad_norm": 2.1857273166825095,
+      "learning_rate": 1.0831307513354112e-06,
+      "loss": 0.9983,
+      "step": 859
+    },
+    {
+      "epoch": 1.4303219106957425,
+      "grad_norm": 2.2108513988117107,
+      "learning_rate": 1.0813845462640206e-06,
+      "loss": 0.9867,
+      "step": 860
+    },
+    {
+      "epoch": 1.4319833852544133,
+      "grad_norm": 2.4469400036513838,
+      "learning_rate": 1.0796380913416823e-06,
+      "loss": 0.9024,
+      "step": 861
+    },
+    {
+      "epoch": 1.433644859813084,
+      "grad_norm": 2.1429468615368905,
+      "learning_rate": 1.0778913919300209e-06,
+      "loss": 0.8963,
+      "step": 862
+    },
+    {
+      "epoch": 1.435306334371755,
+      "grad_norm": 2.176033522508349,
+      "learning_rate": 1.0761444533914124e-06,
+      "loss": 0.9561,
+      "step": 863
+    },
+    {
+      "epoch": 1.4369678089304259,
+      "grad_norm": 2.0784290252253137,
+      "learning_rate": 1.0743972810889654e-06,
+      "loss": 0.906,
+      "step": 864
+    },
+    {
+      "epoch": 1.4386292834890966,
+      "grad_norm": 2.2109329665454087,
+      "learning_rate": 1.0726498803865088e-06,
+      "loss": 0.9316,
+      "step": 865
+    },
+    {
+      "epoch": 1.4402907580477673,
+      "grad_norm": 2.160142006014489,
+      "learning_rate": 1.0709022566485697e-06,
+      "loss": 1.0291,
+      "step": 866
+    },
+    {
+      "epoch": 1.4419522326064382,
+      "grad_norm": 2.236888922467273,
+      "learning_rate": 1.069154415240362e-06,
+      "loss": 0.9812,
+      "step": 867
+    },
+    {
+      "epoch": 1.4436137071651092,
+      "grad_norm": 2.4886237263035276,
+      "learning_rate": 1.067406361527768e-06,
+      "loss": 1.0427,
+      "step": 868
+    },
+    {
+      "epoch": 1.44527518172378,
+      "grad_norm": 2.2522311274987823,
+      "learning_rate": 1.0656581008773197e-06,
+      "loss": 0.8864,
+      "step": 869
+    },
+    {
+      "epoch": 1.4469366562824506,
+      "grad_norm": 2.126057360159879,
+      "learning_rate": 1.0639096386561864e-06,
+      "loss": 1.0294,
+      "step": 870
+    },
+    {
+      "epoch": 1.4485981308411215,
+      "grad_norm": 2.257463390997248,
+      "learning_rate": 1.0621609802321553e-06,
+      "loss": 0.9633,
+      "step": 871
+    },
+    {
+      "epoch": 1.4502596053997923,
+      "grad_norm": 2.26321915407214,
+      "learning_rate": 1.0604121309736163e-06,
+      "loss": 0.8698,
+      "step": 872
+    },
+    {
+      "epoch": 1.4519210799584632,
+      "grad_norm": 2.061806930610862,
+      "learning_rate": 1.058663096249545e-06,
+      "loss": 1.0377,
+      "step": 873
+    },
+    {
+      "epoch": 1.453582554517134,
+      "grad_norm": 2.1295799083708147,
+      "learning_rate": 1.0569138814294862e-06,
+      "loss": 0.9534,
+      "step": 874
+    },
+    {
+      "epoch": 1.4552440290758049,
+      "grad_norm": 2.3106775542374294,
+      "learning_rate": 1.055164491883538e-06,
+      "loss": 0.9403,
+      "step": 875
+    },
+    {
+      "epoch": 1.4569055036344756,
+      "grad_norm": 2.2302178209419132,
+      "learning_rate": 1.0534149329823347e-06,
+      "loss": 0.9301,
+      "step": 876
+    },
+    {
+      "epoch": 1.4585669781931463,
+      "grad_norm": 2.2949573702055117,
+      "learning_rate": 1.0516652100970306e-06,
+      "loss": 0.9528,
+      "step": 877
+    },
+    {
+      "epoch": 1.4602284527518172,
+      "grad_norm": 2.10604631236836,
+      "learning_rate": 1.0499153285992832e-06,
+      "loss": 0.9069,
+      "step": 878
+    },
+    {
+      "epoch": 1.4618899273104882,
+      "grad_norm": 2.1660045153149934,
+      "learning_rate": 1.0481652938612372e-06,
+      "loss": 0.9408,
+      "step": 879
+    },
+    {
+      "epoch": 1.4635514018691589,
+      "grad_norm": 2.2748273606386995,
+      "learning_rate": 1.0464151112555076e-06,
+      "loss": 1.0554,
+      "step": 880
+    },
+    {
+      "epoch": 1.4652128764278296,
+      "grad_norm": 2.2711524480515406,
+      "learning_rate": 1.0446647861551632e-06,
+      "loss": 1.0468,
+      "step": 881
+    },
+    {
+      "epoch": 1.4668743509865005,
+      "grad_norm": 2.2946241383411987,
+      "learning_rate": 1.042914323933711e-06,
+      "loss": 0.8936,
+      "step": 882
+    },
+    {
+      "epoch": 1.4685358255451713,
+      "grad_norm": 2.390089218019007,
+      "learning_rate": 1.0411637299650781e-06,
+      "loss": 1.0086,
+      "step": 883
+    },
+    {
+      "epoch": 1.4701973001038422,
+      "grad_norm": 2.340221690538989,
+      "learning_rate": 1.0394130096235965e-06,
+      "loss": 1.0351,
+      "step": 884
+    },
+    {
+      "epoch": 1.471858774662513,
+      "grad_norm": 2.297583872743856,
+      "learning_rate": 1.0376621682839856e-06,
+      "loss": 0.9497,
+      "step": 885
+    },
+    {
+      "epoch": 1.4735202492211839,
+      "grad_norm": 2.305498668474036,
+      "learning_rate": 1.0359112113213374e-06,
+      "loss": 0.9291,
+      "step": 886
+    },
+    {
+      "epoch": 1.4751817237798546,
+      "grad_norm": 2.182291435160706,
+      "learning_rate": 1.0341601441110981e-06,
+      "loss": 0.9932,
+      "step": 887
+    },
+    {
+      "epoch": 1.4768431983385255,
+      "grad_norm": 2.0985924867543924,
+      "learning_rate": 1.0324089720290522e-06,
+      "loss": 0.9702,
+      "step": 888
+    },
+    {
+      "epoch": 1.4785046728971962,
+      "grad_norm": 2.2916738661032863,
+      "learning_rate": 1.0306577004513064e-06,
+      "loss": 1.0412,
+      "step": 889
+    },
+    {
+      "epoch": 1.4801661474558672,
+      "grad_norm": 2.3135550693056115,
+      "learning_rate": 1.0289063347542726e-06,
+      "loss": 0.8467,
+      "step": 890
+    },
+    {
+      "epoch": 1.4818276220145379,
+      "grad_norm": 2.3188283888371393,
+      "learning_rate": 1.0271548803146525e-06,
+      "loss": 0.9271,
+      "step": 891
+    },
+    {
+      "epoch": 1.4834890965732086,
+      "grad_norm": 2.2139777839358845,
+      "learning_rate": 1.0254033425094196e-06,
+      "loss": 0.9119,
+      "step": 892
+    },
+    {
+      "epoch": 1.4851505711318795,
+      "grad_norm": 2.3212272419519713,
+      "learning_rate": 1.0236517267158026e-06,
+      "loss": 1.0185,
+      "step": 893
+    },
+    {
+      "epoch": 1.4868120456905505,
+      "grad_norm": 2.0589993974559757,
+      "learning_rate": 1.0219000383112713e-06,
+      "loss": 1.022,
+      "step": 894
+    },
+    {
+      "epoch": 1.4884735202492212,
+      "grad_norm": 2.5364209766355015,
+      "learning_rate": 1.020148282673517e-06,
+      "loss": 0.9629,
+      "step": 895
+    },
+    {
+      "epoch": 1.490134994807892,
+      "grad_norm": 2.1254602155332982,
+      "learning_rate": 1.0183964651804382e-06,
+      "loss": 0.9765,
+      "step": 896
+    },
+    {
+      "epoch": 1.4917964693665628,
+      "grad_norm": 2.2144306649179,
+      "learning_rate": 1.0166445912101228e-06,
+      "loss": 0.9865,
+      "step": 897
+    },
+    {
+      "epoch": 1.4934579439252336,
+      "grad_norm": 2.6976218997101173,
+      "learning_rate": 1.0148926661408327e-06,
+      "loss": 1.1091,
+      "step": 898
+    },
+    {
+      "epoch": 1.4951194184839045,
+      "grad_norm": 2.1447116841130347,
+      "learning_rate": 1.0131406953509855e-06,
+      "loss": 0.9854,
+      "step": 899
+    },
+    {
+      "epoch": 1.4967808930425752,
+      "grad_norm": 2.313826573818675,
+      "learning_rate": 1.0113886842191408e-06,
+      "loss": 1.0696,
+      "step": 900
+    },
+    {
+      "epoch": 1.4984423676012462,
+      "grad_norm": 2.398689146513669,
+      "learning_rate": 1.0096366381239806e-06,
+      "loss": 0.9913,
+      "step": 901
+    },
+    {
+      "epoch": 1.5001038421599169,
+      "grad_norm": 2.01368084597207,
+      "learning_rate": 1.0078845624442953e-06,
+      "loss": 0.8734,
+      "step": 902
+    },
+    {
+      "epoch": 1.5017653167185876,
+      "grad_norm": 2.162541083177183,
+      "learning_rate": 1.0061324625589655e-06,
+      "loss": 0.8432,
+      "step": 903
+    },
+    {
+      "epoch": 1.5034267912772585,
+      "grad_norm": 2.066277760482693,
+      "learning_rate": 1.004380343846946e-06,
+      "loss": 0.9469,
+      "step": 904
+    },
+    {
+      "epoch": 1.5050882658359295,
+      "grad_norm": 2.291540588940334,
+      "learning_rate": 1.0026282116872498e-06,
+      "loss": 0.8913,
+      "step": 905
+    },
+    {
+      "epoch": 1.5067497403946002,
+      "grad_norm": 2.3475144052615833,
+      "learning_rate": 1.000876071458931e-06,
+      "loss": 0.8889,
+      "step": 906
+    },
+    {
+      "epoch": 1.5067497403946002,
+      "eval_loss": 1.3885624408721924,
+      "eval_runtime": 24.6265,
+      "eval_samples_per_second": 0.447,
+      "eval_steps_per_second": 0.122,
+      "step": 906
+    },
+    {
+      "epoch": 1.508411214953271,
+      "grad_norm": 3.065303881065278,
+      "learning_rate": 9.99123928541069e-07,
+      "loss": 0.9061,
+      "step": 907
+    },
+    {
+      "epoch": 1.5100726895119418,
+      "grad_norm": 2.191485923954429,
+      "learning_rate": 9.973717883127503e-07,
+      "loss": 0.9699,
+      "step": 908
+    },
+    {
+      "epoch": 1.5117341640706128,
+      "grad_norm": 2.3278723421202505,
+      "learning_rate": 9.95619656153054e-07,
+      "loss": 1.0169,
+      "step": 909
+    },
+    {
+      "epoch": 1.5133956386292835,
+      "grad_norm": 2.3678043287491586,
+      "learning_rate": 9.938675374410346e-07,
+      "loss": 0.9757,
+      "step": 910
+    },
+    {
+      "epoch": 1.5150571131879542,
+      "grad_norm": 2.1901568875128836,
+      "learning_rate": 9.921154375557046e-07,
+      "loss": 1.0352,
+      "step": 911
+    },
+    {
+      "epoch": 1.5167185877466252,
+      "grad_norm": 2.30893338243549,
+      "learning_rate": 9.903633618760193e-07,
+      "loss": 0.8631,
+      "step": 912
+    },
+    {
+      "epoch": 1.518380062305296,
+      "grad_norm": 2.2873579587086366,
+      "learning_rate": 9.886113157808594e-07,
+      "loss": 0.8675,
+      "step": 913
+    },
+    {
+      "epoch": 1.5200415368639668,
+      "grad_norm": 2.085982086942701,
+      "learning_rate": 9.868593046490144e-07,
+      "loss": 0.947,
+      "step": 914
+    },
+    {
+      "epoch": 1.5217030114226375,
+      "grad_norm": 2.060117879856327,
+      "learning_rate": 9.851073338591675e-07,
+      "loss": 1.0063,
+      "step": 915
+    },
+    {
+      "epoch": 1.5233644859813085,
+      "grad_norm": 2.43994470111981,
+      "learning_rate": 9.833554087898773e-07,
+      "loss": 0.9692,
+      "step": 916
+    },
+    {
+      "epoch": 1.5250259605399792,
+      "grad_norm": 2.2294605522286948,
+      "learning_rate": 9.81603534819562e-07,
+      "loss": 0.9935,
+      "step": 917
+    },
+    {
+      "epoch": 1.52668743509865,
+      "grad_norm": 2.1907032716227883,
+      "learning_rate": 9.798517173264831e-07,
+      "loss": 0.9507,
+      "step": 918
+    },
+    {
+      "epoch": 1.5283489096573208,
+      "grad_norm": 2.1045414815637504,
+      "learning_rate": 9.780999616887288e-07,
+      "loss": 0.9059,
+      "step": 919
+    },
+    {
+      "epoch": 1.5300103842159918,
+      "grad_norm": 2.3012223647608323,
+      "learning_rate": 9.763482732841975e-07,
+      "loss": 1.0022,
+      "step": 920
+    },
+    {
+      "epoch": 1.5316718587746625,
+      "grad_norm": 2.272854650284857,
+      "learning_rate": 9.74596657490581e-07,
+      "loss": 1.1475,
+      "step": 921
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 2.5844520190272853,
+      "learning_rate": 9.728451196853476e-07,
+      "loss": 0.9708,
+      "step": 922
+    },
+    {
+      "epoch": 1.5349948078920042,
+      "grad_norm": 2.1870847904085626,
+      "learning_rate": 9.710936652457275e-07,
+      "loss": 1.0149,
+      "step": 923
+    },
+    {
+      "epoch": 1.536656282450675,
+      "grad_norm": 2.167131744822391,
+      "learning_rate": 9.693422995486938e-07,
+      "loss": 0.9241,
+      "step": 924
+    },
+    {
+      "epoch": 1.5383177570093458,
+      "grad_norm": 2.103729458011538,
+      "learning_rate": 9.675910279709475e-07,
+      "loss": 0.8776,
+      "step": 925
+    },
+    {
+      "epoch": 1.5399792315680165,
+      "grad_norm": 2.2605948836777827,
+      "learning_rate": 9.658398558889018e-07,
+      "loss": 0.9223,
+      "step": 926
+    },
+    {
+      "epoch": 1.5416407061266875,
+      "grad_norm": 2.1973564828092282,
+      "learning_rate": 9.640887886786623e-07,
+      "loss": 0.9611,
+      "step": 927
+    },
+    {
+      "epoch": 1.5433021806853584,
+      "grad_norm": 2.1500165200756998,
+      "learning_rate": 9.62337831716014e-07,
+      "loss": 0.9845,
+      "step": 928
+    },
+    {
+      "epoch": 1.5449636552440291,
+      "grad_norm": 2.1276385579139716,
+      "learning_rate": 9.605869903764036e-07,
+      "loss": 0.8422,
+      "step": 929
+    },
+    {
+      "epoch": 1.5466251298026998,
+      "grad_norm": 2.16875527803873,
+      "learning_rate": 9.588362700349218e-07,
+      "loss": 0.8773,
+      "step": 930
+    },
+    {
+      "epoch": 1.5482866043613708,
+      "grad_norm": 2.175442832799032,
+      "learning_rate": 9.570856760662888e-07,
+      "loss": 0.9159,
+      "step": 931
+    },
+    {
+      "epoch": 1.5499480789200415,
+      "grad_norm": 2.0911533410140652,
+      "learning_rate": 9.553352138448365e-07,
+      "loss": 0.9862,
+      "step": 932
+    },
+    {
+      "epoch": 1.5516095534787122,
+      "grad_norm": 2.4217133148552907,
+      "learning_rate": 9.535848887444925e-07,
+      "loss": 1.0448,
+      "step": 933
+    },
+    {
+      "epoch": 1.5532710280373832,
+      "grad_norm": 2.156898906325239,
+      "learning_rate": 9.518347061387627e-07,
+      "loss": 1.0456,
+      "step": 934
+    },
+    {
+      "epoch": 1.554932502596054,
+      "grad_norm": 2.1174530253263852,
+      "learning_rate": 9.500846714007168e-07,
+      "loss": 0.9817,
+      "step": 935
+    },
+    {
+      "epoch": 1.5565939771547248,
+      "grad_norm": 2.2295767742183337,
+      "learning_rate": 9.483347899029695e-07,
+      "loss": 1.1012,
+      "step": 936
+    },
+    {
+      "epoch": 1.5582554517133955,
+      "grad_norm": 2.045078620592537,
+      "learning_rate": 9.465850670176653e-07,
+      "loss": 0.8389,
+      "step": 937
+    },
+    {
+      "epoch": 1.5599169262720665,
+      "grad_norm": 2.2882936479206313,
+      "learning_rate": 9.44835508116462e-07,
+      "loss": 0.9324,
+      "step": 938
+    },
+    {
+      "epoch": 1.5615784008307374,
+      "grad_norm": 2.274529967263343,
+      "learning_rate": 9.430861185705137e-07,
+      "loss": 1.0327,
+      "step": 939
+    },
+    {
+      "epoch": 1.5632398753894081,
+      "grad_norm": 2.1598121426869565,
+      "learning_rate": 9.41336903750455e-07,
+      "loss": 0.9787,
+      "step": 940
+    },
+    {
+      "epoch": 1.5649013499480788,
+      "grad_norm": 2.2018912955453733,
+      "learning_rate": 9.395878690263836e-07,
+      "loss": 0.9555,
+      "step": 941
+    },
+    {
+      "epoch": 1.5665628245067498,
+      "grad_norm": 2.089375883352287,
+      "learning_rate": 9.378390197678447e-07,
+      "loss": 0.9679,
+      "step": 942
+    },
+    {
+      "epoch": 1.5682242990654207,
+      "grad_norm": 2.203509678408647,
+      "learning_rate": 9.360903613438137e-07,
+      "loss": 0.9715,
+      "step": 943
+    },
+    {
+      "epoch": 1.5698857736240914,
+      "grad_norm": 2.1347102317454367,
+      "learning_rate": 9.343418991226803e-07,
+      "loss": 0.7945,
+      "step": 944
+    },
+    {
+      "epoch": 1.5715472481827621,
+      "grad_norm": 2.378865551913575,
+      "learning_rate": 9.325936384722321e-07,
+      "loss": 0.9477,
+      "step": 945
+    },
+    {
+      "epoch": 1.573208722741433,
+      "grad_norm": 2.4467581458666428,
+      "learning_rate": 9.308455847596377e-07,
+      "loss": 1.0578,
+      "step": 946
+    },
+    {
+      "epoch": 1.5748701973001038,
+      "grad_norm": 2.203361322184244,
+      "learning_rate": 9.290977433514305e-07,
+      "loss": 0.884,
+      "step": 947
+    },
+    {
+      "epoch": 1.5765316718587745,
+      "grad_norm": 2.1727119561544597,
+      "learning_rate": 9.273501196134914e-07,
+      "loss": 0.9208,
+      "step": 948
+    },
+    {
+      "epoch": 1.5781931464174455,
+      "grad_norm": 2.3340178077448837,
+      "learning_rate": 9.256027189110344e-07,
+      "loss": 0.9412,
+      "step": 949
+    },
+    {
+      "epoch": 1.5798546209761164,
+      "grad_norm": 2.2245206192765807,
+      "learning_rate": 9.23855546608588e-07,
+      "loss": 0.9677,
+      "step": 950
+    },
+    {
+      "epoch": 1.5815160955347871,
+      "grad_norm": 2.3713993791607786,
+      "learning_rate": 9.221086080699792e-07,
+      "loss": 0.8594,
+      "step": 951
+    },
+    {
+      "epoch": 1.5831775700934578,
+      "grad_norm": 2.30874801141752,
+      "learning_rate": 9.203619086583178e-07,
+      "loss": 1.1003,
+      "step": 952
+    },
+    {
+      "epoch": 1.5848390446521288,
+      "grad_norm": 2.3037415512544492,
+      "learning_rate": 9.186154537359794e-07,
+      "loss": 0.991,
+      "step": 953
+    },
+    {
+      "epoch": 1.5865005192107997,
+      "grad_norm": 2.360666046738533,
+      "learning_rate": 9.168692486645893e-07,
+      "loss": 0.9452,
+      "step": 954
+    },
+    {
+      "epoch": 1.5881619937694704,
+      "grad_norm": 2.1674212494443736,
+      "learning_rate": 9.15123298805005e-07,
+      "loss": 1.0289,
+      "step": 955
+    },
+    {
+      "epoch": 1.5898234683281411,
+      "grad_norm": 2.2039402749504693,
+      "learning_rate": 9.133776095173013e-07,
+      "loss": 0.8989,
+      "step": 956
+    },
+    {
+      "epoch": 1.591484942886812,
+      "grad_norm": 2.3319779317051994,
+      "learning_rate": 9.116321861607523e-07,
+      "loss": 1.0259,
+      "step": 957
+    },
+    {
+      "epoch": 1.593146417445483,
+      "grad_norm": 2.397971695824514,
+      "learning_rate": 9.098870340938168e-07,
+      "loss": 0.9254,
+      "step": 958
+    },
+    {
+      "epoch": 1.5948078920041537,
+      "grad_norm": 2.1189888494560956,
+      "learning_rate": 9.081421586741188e-07,
+      "loss": 0.9747,
+      "step": 959
+    },
+    {
+      "epoch": 1.5964693665628245,
+      "grad_norm": 2.4628203972860994,
+      "learning_rate": 9.063975652584357e-07,
+      "loss": 1.1281,
+      "step": 960
+    },
+    {
+      "epoch": 1.5981308411214954,
+      "grad_norm": 2.2290027206829452,
+      "learning_rate": 9.046532592026768e-07,
+      "loss": 1.0188,
+      "step": 961
+    },
+    {
+      "epoch": 1.599792315680166,
+      "grad_norm": 2.101837850760124,
+      "learning_rate": 9.029092458618705e-07,
+      "loss": 1.0343,
+      "step": 962
+    },
+    {
+      "epoch": 1.6014537902388368,
+      "grad_norm": 2.1955065161912586,
+      "learning_rate": 9.011655305901457e-07,
+      "loss": 0.9229,
+      "step": 963
+    },
+    {
+      "epoch": 1.6031152647975078,
+      "grad_norm": 2.208465293699187,
+      "learning_rate": 8.994221187407167e-07,
+      "loss": 0.9811,
+      "step": 964
+    },
+    {
+      "epoch": 1.6047767393561787,
+      "grad_norm": 2.189756306702255,
+      "learning_rate": 8.976790156658665e-07,
+      "loss": 0.9948,
+      "step": 965
+    },
+    {
+      "epoch": 1.6064382139148494,
+      "grad_norm": 2.21101125223709,
+      "learning_rate": 8.959362267169299e-07,
+      "loss": 0.9206,
+      "step": 966
+    },
+    {
+      "epoch": 1.6080996884735201,
+      "grad_norm": 2.1552977090716476,
+      "learning_rate": 8.941937572442773e-07,
+      "loss": 0.9684,
+      "step": 967
+    },
+    {
+      "epoch": 1.609761163032191,
+      "grad_norm": 2.0424533004826593,
+      "learning_rate": 8.924516125972983e-07,
+      "loss": 0.9515,
+      "step": 968
+    },
+    {
+      "epoch": 1.611422637590862,
+      "grad_norm": 2.2773490170655437,
+      "learning_rate": 8.907097981243851e-07,
+      "loss": 1.0617,
+      "step": 969
+    },
+    {
+      "epoch": 1.6130841121495327,
+      "grad_norm": 2.219673265679085,
+      "learning_rate": 8.88968319172917e-07,
+      "loss": 1.0573,
+      "step": 970
+    },
+    {
+      "epoch": 1.6147455867082035,
+      "grad_norm": 2.23001657969037,
+      "learning_rate": 8.872271810892424e-07,
+      "loss": 1.0252,
+      "step": 971
+    },
+    {
+      "epoch": 1.6164070612668744,
+      "grad_norm": 2.2328021310208968,
+      "learning_rate": 8.854863892186639e-07,
+      "loss": 0.9826,
+      "step": 972
+    },
+    {
+      "epoch": 1.6180685358255453,
+      "grad_norm": 2.5771699055938475,
+      "learning_rate": 8.837459489054203e-07,
+      "loss": 0.9947,
+      "step": 973
+    },
+    {
+      "epoch": 1.619730010384216,
+      "grad_norm": 2.798440468259025,
+      "learning_rate": 8.820058654926725e-07,
+      "loss": 0.9452,
+      "step": 974
+    },
+    {
+      "epoch": 1.6213914849428868,
+      "grad_norm": 2.1504820505748325,
+      "learning_rate": 8.802661443224844e-07,
+      "loss": 1.0738,
+      "step": 975
+    },
+    {
+      "epoch": 1.6230529595015577,
+      "grad_norm": 2.133788375605966,
+      "learning_rate": 8.785267907358084e-07,
+      "loss": 1.0597,
+      "step": 976
+    },
+    {
+      "epoch": 1.6247144340602284,
+      "grad_norm": 2.1928865638126434,
+      "learning_rate": 8.767878100724688e-07,
+      "loss": 0.9914,
+      "step": 977
+    },
+    {
+      "epoch": 1.6263759086188991,
+      "grad_norm": 2.67304969048529,
+      "learning_rate": 8.750492076711439e-07,
+      "loss": 0.9834,
+      "step": 978
+    },
+    {
+      "epoch": 1.62803738317757,
+      "grad_norm": 2.242441368839289,
+      "learning_rate": 8.73310988869352e-07,
+      "loss": 0.9394,
+      "step": 979
+    },
+    {
+      "epoch": 1.629698857736241,
+      "grad_norm": 2.155656747647904,
+      "learning_rate": 8.715731590034329e-07,
+      "loss": 1.083,
+      "step": 980
+    },
+    {
+      "epoch": 1.6313603322949117,
+      "grad_norm": 2.1738442975670798,
+      "learning_rate": 8.698357234085327e-07,
+      "loss": 0.937,
+      "step": 981
+    },
+    {
+      "epoch": 1.6330218068535824,
+      "grad_norm": 2.1949307066634187,
+      "learning_rate": 8.680986874185872e-07,
+      "loss": 0.9225,
+      "step": 982
+    },
+    {
+      "epoch": 1.6346832814122534,
+      "grad_norm": 2.209794807763499,
+      "learning_rate": 8.663620563663046e-07,
+      "loss": 0.9265,
+      "step": 983
+    },
+    {
+      "epoch": 1.6363447559709243,
+      "grad_norm": 2.27946429282308,
+      "learning_rate": 8.646258355831513e-07,
+      "loss": 0.9482,
+      "step": 984
+    },
+    {
+      "epoch": 1.638006230529595,
+      "grad_norm": 2.4532997949631086,
+      "learning_rate": 8.628900303993334e-07,
+      "loss": 1.0124,
+      "step": 985
+    },
+    {
+      "epoch": 1.6396677050882658,
+      "grad_norm": 2.0841134326683264,
+      "learning_rate": 8.611546461437808e-07,
+      "loss": 1.0067,
+      "step": 986
+    },
+    {
+      "epoch": 1.6413291796469367,
+      "grad_norm": 2.0640910431827395,
+      "learning_rate": 8.594196881441314e-07,
+      "loss": 0.7401,
+      "step": 987
+    },
+    {
+      "epoch": 1.6429906542056076,
+      "grad_norm": 2.442342265002308,
+      "learning_rate": 8.576851617267149e-07,
+      "loss": 0.9464,
+      "step": 988
+    },
+    {
+      "epoch": 1.6446521287642781,
+      "grad_norm": 2.2202450152134725,
+      "learning_rate": 8.559510722165359e-07,
+      "loss": 0.9864,
+      "step": 989
+    },
+    {
+      "epoch": 1.646313603322949,
+      "grad_norm": 2.270663471909469,
+      "learning_rate": 8.542174249372572e-07,
+      "loss": 1.1029,
+      "step": 990
+    },
+    {
+      "epoch": 1.64797507788162,
+      "grad_norm": 2.2174023151217566,
+      "learning_rate": 8.524842252111843e-07,
+      "loss": 0.9365,
+      "step": 991
+    },
+    {
+      "epoch": 1.6496365524402907,
+      "grad_norm": 2.344777176470885,
+      "learning_rate": 8.507514783592486e-07,
+      "loss": 1.014,
+      "step": 992
+    },
+    {
+      "epoch": 1.6512980269989614,
+      "grad_norm": 2.422227525332542,
+      "learning_rate": 8.490191897009915e-07,
+      "loss": 0.979,
+      "step": 993
+    },
+    {
+      "epoch": 1.6529595015576324,
+      "grad_norm": 2.258033290502269,
+      "learning_rate": 8.472873645545474e-07,
+      "loss": 1.0524,
+      "step": 994
+    },
+    {
+      "epoch": 1.6546209761163033,
+      "grad_norm": 2.303002200199465,
+      "learning_rate": 8.45556008236628e-07,
+      "loss": 0.933,
+      "step": 995
+    },
+    {
+      "epoch": 1.656282450674974,
+      "grad_norm": 2.2109316644818495,
+      "learning_rate": 8.438251260625055e-07,
+      "loss": 0.9229,
+      "step": 996
+    },
+    {
+      "epoch": 1.6579439252336448,
+      "grad_norm": 2.2713621854438357,
+      "learning_rate": 8.420947233459962e-07,
+      "loss": 0.997,
+      "step": 997
+    },
+    {
+      "epoch": 1.6596053997923157,
+      "grad_norm": 2.1886965102085965,
+      "learning_rate": 8.403648053994447e-07,
+      "loss": 0.9646,
+      "step": 998
+    },
+    {
+      "epoch": 1.6612668743509866,
+      "grad_norm": 2.3286768659520765,
+      "learning_rate": 8.386353775337078e-07,
+      "loss": 1.0782,
+      "step": 999
+    },
+    {
+      "epoch": 1.6629283489096574,
+      "grad_norm": 2.2131507866973443,
+      "learning_rate": 8.369064450581372e-07,
+      "loss": 0.9542,
+      "step": 1000
+    },
+    {
+      "epoch": 1.664589823468328,
+      "grad_norm": 2.2038565852379777,
+      "learning_rate": 8.351780132805639e-07,
+      "loss": 1.009,
+      "step": 1001
+    },
+    {
+      "epoch": 1.666251298026999,
+      "grad_norm": 2.2824651606443798,
+      "learning_rate": 8.334500875072817e-07,
+      "loss": 0.9784,
+      "step": 1002
+    },
+    {
+      "epoch": 1.66791277258567,
+      "grad_norm": 2.34183081355438,
+      "learning_rate": 8.317226730430309e-07,
+      "loss": 1.0175,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6695742471443404,
+      "grad_norm": 2.1288437772823134,
+      "learning_rate": 8.299957751909826e-07,
+      "loss": 0.8847,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6712357217030114,
+      "grad_norm": 2.4568293153056215,
+      "learning_rate": 8.282693992527212e-07,
+      "loss": 0.9493,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6728971962616823,
+      "grad_norm": 2.320949328340056,
+      "learning_rate": 8.265435505282292e-07,
+      "loss": 0.9644,
+      "step": 1006
+    },
+    {
+      "epoch": 1.674558670820353,
+      "grad_norm": 2.607704387361572,
+      "learning_rate": 8.248182343158705e-07,
+      "loss": 0.9348,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6762201453790238,
+      "grad_norm": 2.5655569259629742,
+      "learning_rate": 8.230934559123739e-07,
+      "loss": 1.1464,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6778816199376947,
+      "grad_norm": 2.2568110375917105,
+      "learning_rate": 8.213692206128178e-07,
+      "loss": 1.0082,
+      "step": 1009
+    },
+    {
+      "epoch": 1.6795430944963656,
+      "grad_norm": 2.329387559326929,
+      "learning_rate": 8.196455337106126e-07,
+      "loss": 1.0236,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6812045690550363,
+      "grad_norm": 2.1931737355029477,
+      "learning_rate": 8.179224004974856e-07,
+      "loss": 0.9806,
+      "step": 1011
+    },
+    {
+      "epoch": 1.682866043613707,
+      "grad_norm": 2.7512987063931384,
+      "learning_rate": 8.161998262634636e-07,
+      "loss": 1.0763,
+      "step": 1012
+    },
+    {
+      "epoch": 1.684527518172378,
+      "grad_norm": 2.3447221138047123,
+      "learning_rate": 8.144778162968583e-07,
+      "loss": 0.9142,
+      "step": 1013
+    },
+    {
+      "epoch": 1.686188992731049,
+      "grad_norm": 2.1872618111159152,
+      "learning_rate": 8.127563758842483e-07,
+      "loss": 0.9387,
+      "step": 1014
+    },
+    {
+      "epoch": 1.6878504672897197,
+      "grad_norm": 2.2470446091718,
+      "learning_rate": 8.11035510310464e-07,
+      "loss": 0.9224,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6895119418483904,
+      "grad_norm": 2.5029397601882417,
+      "learning_rate": 8.093152248585709e-07,
+      "loss": 0.9544,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6911734164070613,
+      "grad_norm": 2.1902389788861547,
+      "learning_rate": 8.075955248098535e-07,
+      "loss": 0.9536,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6928348909657323,
+      "grad_norm": 2.0810637251240225,
+      "learning_rate": 8.058764154437996e-07,
+      "loss": 1.0527,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6944963655244027,
+      "grad_norm": 2.3552600183690227,
+      "learning_rate": 8.041579020380828e-07,
+      "loss": 0.9047,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6961578400830737,
+      "grad_norm": 2.434839618311,
+      "learning_rate": 8.024399898685478e-07,
+      "loss": 0.9539,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6978193146417446,
+      "grad_norm": 2.2927568687484237,
+      "learning_rate": 8.007226842091929e-07,
+      "loss": 0.9926,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6994807892004153,
+      "grad_norm": 2.3228074078887775,
+      "learning_rate": 7.990059903321552e-07,
+      "loss": 1.0054,
+      "step": 1022
+    },
+    {
+      "epoch": 1.701142263759086,
+      "grad_norm": 2.110477821804407,
+      "learning_rate": 7.972899135076928e-07,
+      "loss": 0.919,
+      "step": 1023
+    },
+    {
+      "epoch": 1.702803738317757,
+      "grad_norm": 2.153197204937706,
+      "learning_rate": 7.9557445900417e-07,
+      "loss": 0.9917,
+      "step": 1024
+    },
+    {
+      "epoch": 1.704465212876428,
+      "grad_norm": 2.215029615261696,
+      "learning_rate": 7.938596320880401e-07,
+      "loss": 0.7933,
+      "step": 1025
+    },
+    {
+      "epoch": 1.7061266874350987,
+      "grad_norm": 2.3403292049764666,
+      "learning_rate": 7.9214543802383e-07,
+      "loss": 0.9624,
+      "step": 1026
+    },
+    {
+      "epoch": 1.7077881619937694,
+      "grad_norm": 2.1684163037426183,
+      "learning_rate": 7.904318820741238e-07,
+      "loss": 0.9629,
+      "step": 1027
+    },
+    {
+      "epoch": 1.7094496365524403,
+      "grad_norm": 2.263783250114826,
+      "learning_rate": 7.887189694995464e-07,
+      "loss": 0.8771,
+      "step": 1028
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 2.3185082936393657,
+      "learning_rate": 7.87006705558748e-07,
+      "loss": 0.9736,
+      "step": 1029
+    },
+    {
+      "epoch": 1.712772585669782,
+      "grad_norm": 2.1436040295655907,
+      "learning_rate": 7.85295095508387e-07,
+      "loss": 0.9073,
+      "step": 1030
+    },
+    {
+      "epoch": 1.7144340602284527,
+      "grad_norm": 2.31861818593136,
+      "learning_rate": 7.835841446031143e-07,
+      "loss": 0.8523,
+      "step": 1031
+    },
+    {
+      "epoch": 1.7160955347871236,
+      "grad_norm": 2.2422376300228946,
+      "learning_rate": 7.818738580955575e-07,
+      "loss": 0.9703,
+      "step": 1032
+    },
+    {
+      "epoch": 1.7177570093457943,
+      "grad_norm": 2.2643194712270307,
+      "learning_rate": 7.801642412363041e-07,
+      "loss": 1.0062,
+      "step": 1033
+    },
+    {
+      "epoch": 1.719418483904465,
+      "grad_norm": 2.237623999857119,
+      "learning_rate": 7.784552992738866e-07,
+      "loss": 0.8801,
+      "step": 1034
+    },
+    {
+      "epoch": 1.721079958463136,
+      "grad_norm": 2.282669507043474,
+      "learning_rate": 7.767470374547646e-07,
+      "loss": 1.0902,
+      "step": 1035
+    },
+    {
+      "epoch": 1.722741433021807,
+      "grad_norm": 2.2928400993422704,
+      "learning_rate": 7.750394610233105e-07,
+      "loss": 0.8845,
+      "step": 1036
+    },
+    {
+      "epoch": 1.7244029075804777,
+      "grad_norm": 2.198326988103989,
+      "learning_rate": 7.733325752217916e-07,
+      "loss": 0.9141,
+      "step": 1037
+    },
+    {
+      "epoch": 1.7260643821391484,
+      "grad_norm": 2.2329846089880716,
+      "learning_rate": 7.716263852903561e-07,
+      "loss": 0.8736,
+      "step": 1038
+    },
+    {
+      "epoch": 1.7277258566978193,
+      "grad_norm": 2.1112717332498416,
+      "learning_rate": 7.699208964670148e-07,
+      "loss": 0.9301,
+      "step": 1039
+    },
+    {
+      "epoch": 1.7293873312564902,
+      "grad_norm": 2.3471201925073943,
+      "learning_rate": 7.68216113987627e-07,
+      "loss": 1.1426,
+      "step": 1040
+    },
+    {
+      "epoch": 1.731048805815161,
+      "grad_norm": 2.2076292469918837,
+      "learning_rate": 7.665120430858828e-07,
+      "loss": 1.055,
+      "step": 1041
+    },
+    {
+      "epoch": 1.7327102803738317,
+      "grad_norm": 2.2641331824435107,
+      "learning_rate": 7.648086889932878e-07,
+      "loss": 0.9148,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7343717549325026,
+      "grad_norm": 2.011124865835,
+      "learning_rate": 7.631060569391481e-07,
+      "loss": 0.9435,
+      "step": 1043
+    },
+    {
+      "epoch": 1.7360332294911736,
+      "grad_norm": 2.5541303942545523,
+      "learning_rate": 7.614041521505517e-07,
+      "loss": 1.051,
+      "step": 1044
+    },
+    {
+      "epoch": 1.7376947040498443,
+      "grad_norm": 2.135958705290272,
+      "learning_rate": 7.597029798523544e-07,
+      "loss": 0.8675,
+      "step": 1045
+    },
+    {
+      "epoch": 1.739356178608515,
+      "grad_norm": 2.1710241061601208,
+      "learning_rate": 7.580025452671635e-07,
+      "loss": 0.9826,
+      "step": 1046
+    },
+    {
+      "epoch": 1.741017653167186,
+      "grad_norm": 2.108624585141848,
+      "learning_rate": 7.563028536153212e-07,
+      "loss": 0.9617,
+      "step": 1047
+    },
+    {
+      "epoch": 1.7426791277258566,
+      "grad_norm": 2.106050127640713,
+      "learning_rate": 7.546039101148895e-07,
+      "loss": 1.0632,
+      "step": 1048
+    },
+    {
+      "epoch": 1.7443406022845274,
+      "grad_norm": 2.187342834226081,
+      "learning_rate": 7.529057199816325e-07,
+      "loss": 1.0858,
+      "step": 1049
+    },
+    {
+      "epoch": 1.7460020768431983,
+      "grad_norm": 2.342073364970068,
+      "learning_rate": 7.512082884290025e-07,
+      "loss": 0.9126,
+      "step": 1050
+    },
+    {
+      "epoch": 1.7476635514018692,
+      "grad_norm": 2.3328737377816613,
+      "learning_rate": 7.495116206681222e-07,
+      "loss": 0.9681,
+      "step": 1051
+    },
+    {
+      "epoch": 1.74932502596054,
+      "grad_norm": 2.149584367750816,
+      "learning_rate": 7.478157219077702e-07,
+      "loss": 1.0266,
+      "step": 1052
+    },
+    {
+      "epoch": 1.7509865005192107,
+      "grad_norm": 2.4000332654502867,
+      "learning_rate": 7.461205973543635e-07,
+      "loss": 0.9125,
+      "step": 1053
+    },
+    {
+      "epoch": 1.7526479750778816,
+      "grad_norm": 2.2222345997916446,
+      "learning_rate": 7.444262522119427e-07,
+      "loss": 0.9607,
+      "step": 1054
+    },
+    {
+      "epoch": 1.7543094496365526,
+      "grad_norm": 2.2094198456747254,
+      "learning_rate": 7.427326916821557e-07,
+      "loss": 1.0043,
+      "step": 1055
+    },
+    {
+      "epoch": 1.7559709241952233,
+      "grad_norm": 2.1757968861860615,
+      "learning_rate": 7.410399209642409e-07,
+      "loss": 0.8917,
+      "step": 1056
+    },
+    {
+      "epoch": 1.757632398753894,
+      "grad_norm": 2.493378250255824,
+      "learning_rate": 7.393479452550132e-07,
+      "loss": 1.0598,
+      "step": 1057
+    },
+    {
+      "epoch": 1.757632398753894,
+      "eval_loss": 1.3888019323349,
+      "eval_runtime": 24.567,
+      "eval_samples_per_second": 0.448,
+      "eval_steps_per_second": 0.122,
+      "step": 1057
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1803,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 151,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.775024226258125e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}