diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4578 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.990024577128813,
+  "eval_steps": 500,
+  "global_step": 648,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004626283070695388,
+      "grad_norm": 0.6155030131340027,
+      "learning_rate": 9.98456790123457e-06,
+      "loss": 17.5936,
+      "step": 1
+    },
+    {
+      "epoch": 0.009252566141390776,
+      "grad_norm": 0.6313583254814148,
+      "learning_rate": 9.969135802469136e-06,
+      "loss": 21.7773,
+      "step": 2
+    },
+    {
+      "epoch": 0.013878849212086165,
+      "grad_norm": 0.4570765793323517,
+      "learning_rate": 9.953703703703704e-06,
+      "loss": 17.1664,
+      "step": 3
+    },
+    {
+      "epoch": 0.018505132282781552,
+      "grad_norm": 0.3836732506752014,
+      "learning_rate": 9.938271604938273e-06,
+      "loss": 18.3922,
+      "step": 4
+    },
+    {
+      "epoch": 0.02313141535347694,
+      "grad_norm": 0.3599857985973358,
+      "learning_rate": 9.92283950617284e-06,
+      "loss": 16.9993,
+      "step": 5
+    },
+    {
+      "epoch": 0.02775769842417233,
+      "grad_norm": 0.4342029094696045,
+      "learning_rate": 9.907407407407408e-06,
+      "loss": 17.4883,
+      "step": 6
+    },
+    {
+      "epoch": 0.032383981494867715,
+      "grad_norm": 0.40308186411857605,
+      "learning_rate": 9.891975308641975e-06,
+      "loss": 18.2562,
+      "step": 7
+    },
+    {
+      "epoch": 0.037010264565563104,
+      "grad_norm": 0.39312881231307983,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 17.9318,
+      "step": 8
+    },
+    {
+      "epoch": 0.04163654763625849,
+      "grad_norm": 0.6749095916748047,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 18.5244,
+      "step": 9
+    },
+    {
+      "epoch": 0.04626283070695388,
+      "grad_norm": 0.3622874617576599,
+      "learning_rate": 9.84567901234568e-06,
+      "loss": 16.4386,
+      "step": 10
+    },
+    {
+      "epoch": 0.05088911377764927,
+      "grad_norm": 0.24345389008522034,
+      "learning_rate": 9.830246913580249e-06,
+      "loss": 17.5414,
+      "step": 11
+    },
+    {
+      "epoch": 0.05551539684834466,
+      "grad_norm": 0.34016355872154236,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 16.9405,
+      "step": 12
+    },
+    {
+      "epoch": 0.06014167991904005,
+      "grad_norm": 0.25097739696502686,
+      "learning_rate": 9.799382716049384e-06,
+      "loss": 14.4692,
+      "step": 13
+    },
+    {
+      "epoch": 0.06476796298973543,
+      "grad_norm": 0.2524467408657074,
+      "learning_rate": 9.783950617283952e-06,
+      "loss": 16.2185,
+      "step": 14
+    },
+    {
+      "epoch": 0.06939424606043082,
+      "grad_norm": 0.3705744743347168,
+      "learning_rate": 9.768518518518519e-06,
+      "loss": 16.3573,
+      "step": 15
+    },
+    {
+      "epoch": 0.07402052913112621,
+      "grad_norm": 0.2848168909549713,
+      "learning_rate": 9.753086419753087e-06,
+      "loss": 16.331,
+      "step": 16
+    },
+    {
+      "epoch": 0.0786468122018216,
+      "grad_norm": 0.300182044506073,
+      "learning_rate": 9.737654320987654e-06,
+      "loss": 15.6623,
+      "step": 17
+    },
+    {
+      "epoch": 0.08327309527251699,
+      "grad_norm": 0.36958807706832886,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 18.5857,
+      "step": 18
+    },
+    {
+      "epoch": 0.08789937834321238,
+      "grad_norm": 0.3009611666202545,
+      "learning_rate": 9.706790123456791e-06,
+      "loss": 17.0655,
+      "step": 19
+    },
+    {
+      "epoch": 0.09252566141390776,
+      "grad_norm": 0.24450023472309113,
+      "learning_rate": 9.691358024691358e-06,
+      "loss": 15.3945,
+      "step": 20
+    },
+    {
+      "epoch": 0.09715194448460315,
+      "grad_norm": 0.3174358010292053,
+      "learning_rate": 9.675925925925926e-06,
+      "loss": 15.3045,
+      "step": 21
+    },
+    {
+      "epoch": 0.10177822755529854,
+      "grad_norm": 0.34466007351875305,
+      "learning_rate": 9.660493827160495e-06,
+      "loss": 16.7186,
+      "step": 22
+    },
+    {
+      "epoch": 0.10640451062599393,
+      "grad_norm": 0.294209361076355,
+      "learning_rate": 9.645061728395062e-06,
+      "loss": 16.2893,
+      "step": 23
+    },
+    {
+      "epoch": 0.11103079369668932,
+      "grad_norm": 0.3418446183204651,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 16.1827,
+      "step": 24
+    },
+    {
+      "epoch": 0.11565707676738471,
+      "grad_norm": 0.2584611177444458,
+      "learning_rate": 9.614197530864198e-06,
+      "loss": 15.7934,
+      "step": 25
+    },
+    {
+      "epoch": 0.1202833598380801,
+      "grad_norm": 0.3344869911670685,
+      "learning_rate": 9.598765432098767e-06,
+      "loss": 14.2338,
+      "step": 26
+    },
+    {
+      "epoch": 0.12490964290877549,
+      "grad_norm": 0.28123536705970764,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 16.2229,
+      "step": 27
+    },
+    {
+      "epoch": 0.12953592597947086,
+      "grad_norm": 0.22181656956672668,
+      "learning_rate": 9.567901234567902e-06,
+      "loss": 13.7744,
+      "step": 28
+    },
+    {
+      "epoch": 0.13416220905016626,
+      "grad_norm": 0.45139142870903015,
+      "learning_rate": 9.55246913580247e-06,
+      "loss": 15.4782,
+      "step": 29
+    },
+    {
+      "epoch": 0.13878849212086164,
+      "grad_norm": 0.21577508747577667,
+      "learning_rate": 9.537037037037037e-06,
+      "loss": 15.6752,
+      "step": 30
+    },
+    {
+      "epoch": 0.14341477519155704,
+      "grad_norm": 0.29512590169906616,
+      "learning_rate": 9.521604938271606e-06,
+      "loss": 14.7965,
+      "step": 31
+    },
+    {
+      "epoch": 0.14804105826225242,
+      "grad_norm": 0.28356751799583435,
+      "learning_rate": 9.506172839506174e-06,
+      "loss": 14.9097,
+      "step": 32
+    },
+    {
+      "epoch": 0.15266734133294782,
+      "grad_norm": 0.25781649351119995,
+      "learning_rate": 9.490740740740741e-06,
+      "loss": 14.494,
+      "step": 33
+    },
+    {
+      "epoch": 0.1572936244036432,
+      "grad_norm": 0.1967274248600006,
+      "learning_rate": 9.47530864197531e-06,
+      "loss": 15.0075,
+      "step": 34
+    },
+    {
+      "epoch": 0.1619199074743386,
+      "grad_norm": 0.42894116044044495,
+      "learning_rate": 9.459876543209878e-06,
+      "loss": 14.8055,
+      "step": 35
+    },
+    {
+      "epoch": 0.16654619054503397,
+      "grad_norm": 0.2417590320110321,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 14.7145,
+      "step": 36
+    },
+    {
+      "epoch": 0.17117247361572938,
+      "grad_norm": 0.15463685989379883,
+      "learning_rate": 9.429012345679013e-06,
+      "loss": 14.4273,
+      "step": 37
+    },
+    {
+      "epoch": 0.17579875668642475,
+      "grad_norm": 0.21183425188064575,
+      "learning_rate": 9.413580246913581e-06,
+      "loss": 16.6545,
+      "step": 38
+    },
+    {
+      "epoch": 0.18042503975712013,
+      "grad_norm": 0.41926464438438416,
+      "learning_rate": 9.398148148148148e-06,
+      "loss": 13.4609,
+      "step": 39
+    },
+    {
+      "epoch": 0.18505132282781553,
+      "grad_norm": 0.21585555374622345,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 14.1515,
+      "step": 40
+    },
+    {
+      "epoch": 0.1896776058985109,
+      "grad_norm": 0.28051456809043884,
+      "learning_rate": 9.367283950617285e-06,
+      "loss": 14.0035,
+      "step": 41
+    },
+    {
+      "epoch": 0.1943038889692063,
+      "grad_norm": 0.18281345069408417,
+      "learning_rate": 9.351851851851854e-06,
+      "loss": 14.6411,
+      "step": 42
+    },
+    {
+      "epoch": 0.19893017203990168,
+      "grad_norm": 0.1747712790966034,
+      "learning_rate": 9.33641975308642e-06,
+      "loss": 14.4986,
+      "step": 43
+    },
+    {
+      "epoch": 0.20355645511059708,
+      "grad_norm": 0.2613980174064636,
+      "learning_rate": 9.320987654320989e-06,
+      "loss": 14.2572,
+      "step": 44
+    },
+    {
+      "epoch": 0.20818273818129246,
+      "grad_norm": 0.19127604365348816,
+      "learning_rate": 9.305555555555557e-06,
+      "loss": 14.3084,
+      "step": 45
+    },
+    {
+      "epoch": 0.21280902125198786,
+      "grad_norm": 0.24545663595199585,
+      "learning_rate": 9.290123456790124e-06,
+      "loss": 13.3381,
+      "step": 46
+    },
+    {
+      "epoch": 0.21743530432268324,
+      "grad_norm": 0.25891631841659546,
+      "learning_rate": 9.274691358024692e-06,
+      "loss": 13.1076,
+      "step": 47
+    },
+    {
+      "epoch": 0.22206158739337864,
+      "grad_norm": 0.23256707191467285,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 13.3061,
+      "step": 48
+    },
+    {
+      "epoch": 0.22668787046407401,
+      "grad_norm": 0.2427562177181244,
+      "learning_rate": 9.243827160493828e-06,
+      "loss": 16.0646,
+      "step": 49
+    },
+    {
+      "epoch": 0.23131415353476942,
+      "grad_norm": 0.22760462760925293,
+      "learning_rate": 9.228395061728396e-06,
+      "loss": 12.937,
+      "step": 50
+    },
+    {
+      "epoch": 0.2359404366054648,
+      "grad_norm": 0.27145224809646606,
+      "learning_rate": 9.212962962962963e-06,
+      "loss": 13.8299,
+      "step": 51
+    },
+    {
+      "epoch": 0.2405667196761602,
+      "grad_norm": 0.3120661973953247,
+      "learning_rate": 9.197530864197531e-06,
+      "loss": 14.074,
+      "step": 52
+    },
+    {
+      "epoch": 0.24519300274685557,
+      "grad_norm": 0.21604514122009277,
+      "learning_rate": 9.1820987654321e-06,
+      "loss": 13.6731,
+      "step": 53
+    },
+    {
+      "epoch": 0.24981928581755097,
+      "grad_norm": 0.22970032691955566,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 14.0123,
+      "step": 54
+    },
+    {
+      "epoch": 0.2544455688882464,
+      "grad_norm": 0.24079963564872742,
+      "learning_rate": 9.151234567901235e-06,
+      "loss": 13.5458,
+      "step": 55
+    },
+    {
+      "epoch": 0.2590718519589417,
+      "grad_norm": 0.21451319754123688,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 12.5504,
+      "step": 56
+    },
+    {
+      "epoch": 0.2636981350296371,
+      "grad_norm": 0.2508305609226227,
+      "learning_rate": 9.120370370370372e-06,
+      "loss": 14.8642,
+      "step": 57
+    },
+    {
+      "epoch": 0.26832441810033253,
+      "grad_norm": 0.24022799730300903,
+      "learning_rate": 9.10493827160494e-06,
+      "loss": 15.1011,
+      "step": 58
+    },
+    {
+      "epoch": 0.2729507011710279,
+      "grad_norm": 0.3794403076171875,
+      "learning_rate": 9.089506172839507e-06,
+      "loss": 14.3993,
+      "step": 59
+    },
+    {
+      "epoch": 0.2775769842417233,
+      "grad_norm": 0.28563258051872253,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 14.294,
+      "step": 60
+    },
+    {
+      "epoch": 0.2822032673124187,
+      "grad_norm": 0.28061386942863464,
+      "learning_rate": 9.058641975308642e-06,
+      "loss": 13.033,
+      "step": 61
+    },
+    {
+      "epoch": 0.2868295503831141,
+      "grad_norm": 0.3351198136806488,
+      "learning_rate": 9.04320987654321e-06,
+      "loss": 14.5075,
+      "step": 62
+    },
+    {
+      "epoch": 0.29145583345380943,
+      "grad_norm": 0.23749324679374695,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 13.7952,
+      "step": 63
+    },
+    {
+      "epoch": 0.29608211652450483,
+      "grad_norm": 0.16919538378715515,
+      "learning_rate": 9.012345679012346e-06,
+      "loss": 14.7146,
+      "step": 64
+    },
+    {
+      "epoch": 0.30070839959520024,
+      "grad_norm": 0.18566973507404327,
+      "learning_rate": 8.996913580246914e-06,
+      "loss": 13.9967,
+      "step": 65
+    },
+    {
+      "epoch": 0.30533468266589564,
+      "grad_norm": 0.25144919753074646,
+      "learning_rate": 8.981481481481483e-06,
+      "loss": 13.0544,
+      "step": 66
+    },
+    {
+      "epoch": 0.309960965736591,
+      "grad_norm": 0.1971070021390915,
+      "learning_rate": 8.96604938271605e-06,
+      "loss": 12.7932,
+      "step": 67
+    },
+    {
+      "epoch": 0.3145872488072864,
+      "grad_norm": 0.2236046940088272,
+      "learning_rate": 8.950617283950618e-06,
+      "loss": 14.4338,
+      "step": 68
+    },
+    {
+      "epoch": 0.3192135318779818,
+      "grad_norm": 0.20062977075576782,
+      "learning_rate": 8.935185185185186e-06,
+      "loss": 13.6965,
+      "step": 69
+    },
+    {
+      "epoch": 0.3238398149486772,
+      "grad_norm": 0.3102545738220215,
+      "learning_rate": 8.919753086419753e-06,
+      "loss": 14.174,
+      "step": 70
+    },
+    {
+      "epoch": 0.32846609801937254,
+      "grad_norm": 0.282172828912735,
+      "learning_rate": 8.904320987654322e-06,
+      "loss": 14.012,
+      "step": 71
+    },
+    {
+      "epoch": 0.33309238109006795,
+      "grad_norm": 0.31374603509902954,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 12.9434,
+      "step": 72
+    },
+    {
+      "epoch": 0.33771866416076335,
+      "grad_norm": 0.24876756966114044,
+      "learning_rate": 8.873456790123458e-06,
+      "loss": 13.0801,
+      "step": 73
+    },
+    {
+      "epoch": 0.34234494723145875,
+      "grad_norm": 0.2828700840473175,
+      "learning_rate": 8.858024691358025e-06,
+      "loss": 12.3676,
+      "step": 74
+    },
+    {
+      "epoch": 0.3469712303021541,
+      "grad_norm": 0.23395580053329468,
+      "learning_rate": 8.842592592592594e-06,
+      "loss": 12.8463,
+      "step": 75
+    },
+    {
+      "epoch": 0.3515975133728495,
+      "grad_norm": 0.26694121956825256,
+      "learning_rate": 8.827160493827162e-06,
+      "loss": 11.5822,
+      "step": 76
+    },
+    {
+      "epoch": 0.3562237964435449,
+      "grad_norm": 0.2129782736301422,
+      "learning_rate": 8.811728395061729e-06,
+      "loss": 12.6834,
+      "step": 77
+    },
+    {
+      "epoch": 0.36085007951424025,
+      "grad_norm": 0.20400603115558624,
+      "learning_rate": 8.796296296296297e-06,
+      "loss": 14.2065,
+      "step": 78
+    },
+    {
+      "epoch": 0.36547636258493565,
+      "grad_norm": 0.45560839772224426,
+      "learning_rate": 8.780864197530866e-06,
+      "loss": 13.1033,
+      "step": 79
+    },
+    {
+      "epoch": 0.37010264565563106,
+      "grad_norm": 0.26329922676086426,
+      "learning_rate": 8.765432098765432e-06,
+      "loss": 11.9119,
+      "step": 80
+    },
+    {
+      "epoch": 0.37472892872632646,
+      "grad_norm": 0.21817731857299805,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 12.815,
+      "step": 81
+    },
+    {
+      "epoch": 0.3793552117970218,
+      "grad_norm": 0.2743465304374695,
+      "learning_rate": 8.73456790123457e-06,
+      "loss": 12.2468,
+      "step": 82
+    },
+    {
+      "epoch": 0.3839814948677172,
+      "grad_norm": 0.2570422291755676,
+      "learning_rate": 8.719135802469136e-06,
+      "loss": 11.9801,
+      "step": 83
+    },
+    {
+      "epoch": 0.3886077779384126,
+      "grad_norm": 0.21237581968307495,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 11.7092,
+      "step": 84
+    },
+    {
+      "epoch": 0.393234061009108,
+      "grad_norm": 0.27444881200790405,
+      "learning_rate": 8.688271604938271e-06,
+      "loss": 13.3074,
+      "step": 85
+    },
+    {
+      "epoch": 0.39786034407980336,
+      "grad_norm": 0.22574838995933533,
+      "learning_rate": 8.67283950617284e-06,
+      "loss": 13.1573,
+      "step": 86
+    },
+    {
+      "epoch": 0.40248662715049877,
+      "grad_norm": 0.252755731344223,
+      "learning_rate": 8.657407407407408e-06,
+      "loss": 12.3472,
+      "step": 87
+    },
+    {
+      "epoch": 0.40711291022119417,
+      "grad_norm": 0.2729659378528595,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 12.8413,
+      "step": 88
+    },
+    {
+      "epoch": 0.41173919329188957,
+      "grad_norm": 0.21728309988975525,
+      "learning_rate": 8.626543209876543e-06,
+      "loss": 11.3523,
+      "step": 89
+    },
+    {
+      "epoch": 0.4163654763625849,
+      "grad_norm": 0.264457106590271,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 13.1175,
+      "step": 90
+    },
+    {
+      "epoch": 0.4209917594332803,
+      "grad_norm": 0.20264536142349243,
+      "learning_rate": 8.59567901234568e-06,
+      "loss": 12.8442,
+      "step": 91
+    },
+    {
+      "epoch": 0.4256180425039757,
+      "grad_norm": 0.3003804385662079,
+      "learning_rate": 8.580246913580249e-06,
+      "loss": 14.5755,
+      "step": 92
+    },
+    {
+      "epoch": 0.4302443255746711,
+      "grad_norm": 0.294029176235199,
+      "learning_rate": 8.564814814814816e-06,
+      "loss": 13.5777,
+      "step": 93
+    },
+    {
+      "epoch": 0.4348706086453665,
+      "grad_norm": 0.21502816677093506,
+      "learning_rate": 8.549382716049384e-06,
+      "loss": 11.9804,
+      "step": 94
+    },
+    {
+      "epoch": 0.4394968917160619,
+      "grad_norm": 0.21898359060287476,
+      "learning_rate": 8.53395061728395e-06,
+      "loss": 13.6382,
+      "step": 95
+    },
+    {
+      "epoch": 0.4441231747867573,
+      "grad_norm": 0.20010338723659515,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 12.4733,
+      "step": 96
+    },
+    {
+      "epoch": 0.4487494578574526,
+      "grad_norm": 0.2247803956270218,
+      "learning_rate": 8.503086419753088e-06,
+      "loss": 12.1589,
+      "step": 97
+    },
+    {
+      "epoch": 0.45337574092814803,
+      "grad_norm": 0.29326918721199036,
+      "learning_rate": 8.487654320987654e-06,
+      "loss": 12.8177,
+      "step": 98
+    },
+    {
+      "epoch": 0.45800202399884343,
+      "grad_norm": 0.18403670191764832,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 13.8527,
+      "step": 99
+    },
+    {
+      "epoch": 0.46262830706953884,
+      "grad_norm": 0.24531710147857666,
+      "learning_rate": 8.456790123456791e-06,
+      "loss": 11.9126,
+      "step": 100
+    },
+    {
+      "epoch": 0.4672545901402342,
+      "grad_norm": 0.24183672666549683,
+      "learning_rate": 8.441358024691358e-06,
+      "loss": 12.952,
+      "step": 101
+    },
+    {
+      "epoch": 0.4718808732109296,
+      "grad_norm": 0.29563647508621216,
+      "learning_rate": 8.425925925925926e-06,
+      "loss": 14.462,
+      "step": 102
+    },
+    {
+      "epoch": 0.476507156281625,
+      "grad_norm": 0.21156498789787292,
+      "learning_rate": 8.410493827160495e-06,
+      "loss": 13.4846,
+      "step": 103
+    },
+    {
+      "epoch": 0.4811334393523204,
+      "grad_norm": 0.24183927476406097,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 11.428,
+      "step": 104
+    },
+    {
+      "epoch": 0.48575972242301574,
+      "grad_norm": 0.21241174638271332,
+      "learning_rate": 8.37962962962963e-06,
+      "loss": 13.8068,
+      "step": 105
+    },
+    {
+      "epoch": 0.49038600549371114,
+      "grad_norm": 0.2149106115102768,
+      "learning_rate": 8.364197530864199e-06,
+      "loss": 12.7659,
+      "step": 106
+    },
+    {
+      "epoch": 0.49501228856440654,
+      "grad_norm": 0.21057994663715363,
+      "learning_rate": 8.348765432098767e-06,
+      "loss": 12.4651,
+      "step": 107
+    },
+    {
+      "epoch": 0.49963857163510195,
+      "grad_norm": 0.18793097138404846,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 11.7854,
+      "step": 108
+    },
+    {
+      "epoch": 0.5042648547057973,
+      "grad_norm": 0.1801634281873703,
+      "learning_rate": 8.317901234567902e-06,
+      "loss": 12.9306,
+      "step": 109
+    },
+    {
+      "epoch": 0.5088911377764928,
+      "grad_norm": 0.21523930132389069,
+      "learning_rate": 8.30246913580247e-06,
+      "loss": 11.6702,
+      "step": 110
+    },
+    {
+      "epoch": 0.5135174208471881,
+      "grad_norm": 0.22931678593158722,
+      "learning_rate": 8.287037037037037e-06,
+      "loss": 11.5856,
+      "step": 111
+    },
+    {
+      "epoch": 0.5181437039178834,
+      "grad_norm": 0.23802338540554047,
+      "learning_rate": 8.271604938271606e-06,
+      "loss": 12.7508,
+      "step": 112
+    },
+    {
+      "epoch": 0.5227699869885789,
+      "grad_norm": 0.23633763194084167,
+      "learning_rate": 8.256172839506174e-06,
+      "loss": 12.0906,
+      "step": 113
+    },
+    {
+      "epoch": 0.5273962700592743,
+      "grad_norm": 0.22553692758083344,
+      "learning_rate": 8.240740740740741e-06,
+      "loss": 12.4982,
+      "step": 114
+    },
+    {
+      "epoch": 0.5320225531299696,
+      "grad_norm": 0.2265913337469101,
+      "learning_rate": 8.22530864197531e-06,
+      "loss": 11.4945,
+      "step": 115
+    },
+    {
+      "epoch": 0.5366488362006651,
+      "grad_norm": 0.1848449558019638,
+      "learning_rate": 8.209876543209876e-06,
+      "loss": 11.866,
+      "step": 116
+    },
+    {
+      "epoch": 0.5412751192713604,
+      "grad_norm": 0.16533511877059937,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 12.3876,
+      "step": 117
+    },
+    {
+      "epoch": 0.5459014023420558,
+      "grad_norm": 0.19875134527683258,
+      "learning_rate": 8.179012345679013e-06,
+      "loss": 12.0186,
+      "step": 118
+    },
+    {
+      "epoch": 0.5505276854127512,
+      "grad_norm": 0.20982632040977478,
+      "learning_rate": 8.16358024691358e-06,
+      "loss": 13.1202,
+      "step": 119
+    },
+    {
+      "epoch": 0.5551539684834466,
+      "grad_norm": 0.2689734101295471,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 13.9764,
+      "step": 120
+    },
+    {
+      "epoch": 0.559780251554142,
+      "grad_norm": 0.14883318543434143,
+      "learning_rate": 8.132716049382717e-06,
+      "loss": 11.7656,
+      "step": 121
+    },
+    {
+      "epoch": 0.5644065346248374,
+      "grad_norm": 0.18115845322608948,
+      "learning_rate": 8.117283950617285e-06,
+      "loss": 11.8902,
+      "step": 122
+    },
+    {
+      "epoch": 0.5690328176955327,
+      "grad_norm": 0.23553551733493805,
+      "learning_rate": 8.101851851851854e-06,
+      "loss": 13.0425,
+      "step": 123
+    },
+    {
+      "epoch": 0.5736591007662282,
+      "grad_norm": 0.2110109180212021,
+      "learning_rate": 8.08641975308642e-06,
+      "loss": 13.6931,
+      "step": 124
+    },
+    {
+      "epoch": 0.5782853838369235,
+      "grad_norm": 0.2748431861400604,
+      "learning_rate": 8.070987654320989e-06,
+      "loss": 12.7852,
+      "step": 125
+    },
+    {
+      "epoch": 0.5829116669076189,
+      "grad_norm": 0.1847638040781021,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 12.9106,
+      "step": 126
+    },
+    {
+      "epoch": 0.5875379499783143,
+      "grad_norm": 0.48550117015838623,
+      "learning_rate": 8.040123456790124e-06,
+      "loss": 11.5484,
+      "step": 127
+    },
+    {
+      "epoch": 0.5921642330490097,
+      "grad_norm": 0.2579441964626312,
+      "learning_rate": 8.024691358024692e-06,
+      "loss": 11.9785,
+      "step": 128
+    },
+    {
+      "epoch": 0.5967905161197051,
+      "grad_norm": 0.19037479162216187,
+      "learning_rate": 8.00925925925926e-06,
+      "loss": 11.0466,
+      "step": 129
+    },
+    {
+      "epoch": 0.6014167991904005,
+      "grad_norm": 0.22901126742362976,
+      "learning_rate": 7.993827160493828e-06,
+      "loss": 11.1139,
+      "step": 130
+    },
+    {
+      "epoch": 0.6060430822610958,
+      "grad_norm": 0.16406911611557007,
+      "learning_rate": 7.978395061728396e-06,
+      "loss": 10.6613,
+      "step": 131
+    },
+    {
+      "epoch": 0.6106693653317913,
+      "grad_norm": 0.23492039740085602,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 11.3782,
+      "step": 132
+    },
+    {
+      "epoch": 0.6152956484024866,
+      "grad_norm": 0.21692270040512085,
+      "learning_rate": 7.947530864197531e-06,
+      "loss": 11.2788,
+      "step": 133
+    },
+    {
+      "epoch": 0.619921931473182,
+      "grad_norm": 0.17832407355308533,
+      "learning_rate": 7.9320987654321e-06,
+      "loss": 11.7617,
+      "step": 134
+    },
+    {
+      "epoch": 0.6245482145438774,
+      "grad_norm": 0.20935167372226715,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 10.9103,
+      "step": 135
+    },
+    {
+      "epoch": 0.6291744976145728,
+      "grad_norm": 0.199855774641037,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 11.5339,
+      "step": 136
+    },
+    {
+      "epoch": 0.6338007806852681,
+      "grad_norm": 0.1627349704504013,
+      "learning_rate": 7.885802469135803e-06,
+      "loss": 10.3593,
+      "step": 137
+    },
+    {
+      "epoch": 0.6384270637559636,
+      "grad_norm": 0.18615403771400452,
+      "learning_rate": 7.870370370370372e-06,
+      "loss": 12.3204,
+      "step": 138
+    },
+    {
+      "epoch": 0.6430533468266589,
+      "grad_norm": 0.17866994440555573,
+      "learning_rate": 7.854938271604939e-06,
+      "loss": 11.548,
+      "step": 139
+    },
+    {
+      "epoch": 0.6476796298973544,
+      "grad_norm": 0.22644223272800446,
+      "learning_rate": 7.839506172839507e-06,
+      "loss": 12.0289,
+      "step": 140
+    },
+    {
+      "epoch": 0.6523059129680497,
+      "grad_norm": 0.18510523438453674,
+      "learning_rate": 7.824074074074076e-06,
+      "loss": 12.1312,
+      "step": 141
+    },
+    {
+      "epoch": 0.6569321960387451,
+      "grad_norm": 0.1948799043893814,
+      "learning_rate": 7.808641975308642e-06,
+      "loss": 13.4831,
+      "step": 142
+    },
+    {
+      "epoch": 0.6615584791094405,
+      "grad_norm": 0.2084604948759079,
+      "learning_rate": 7.79320987654321e-06,
+      "loss": 11.1289,
+      "step": 143
+    },
+    {
+      "epoch": 0.6661847621801359,
+      "grad_norm": 0.194437637925148,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 11.5961,
+      "step": 144
+    },
+    {
+      "epoch": 0.6708110452508312,
+      "grad_norm": 0.2627541124820709,
+      "learning_rate": 7.762345679012346e-06,
+      "loss": 12.0198,
+      "step": 145
+    },
+    {
+      "epoch": 0.6754373283215267,
+      "grad_norm": 0.21856093406677246,
+      "learning_rate": 7.746913580246914e-06,
+      "loss": 12.5285,
+      "step": 146
+    },
+    {
+      "epoch": 0.680063611392222,
+      "grad_norm": 0.15228690207004547,
+      "learning_rate": 7.731481481481483e-06,
+      "loss": 10.8653,
+      "step": 147
+    },
+    {
+      "epoch": 0.6846898944629175,
+      "grad_norm": 0.22951403260231018,
+      "learning_rate": 7.71604938271605e-06,
+      "loss": 10.8751,
+      "step": 148
+    },
+    {
+      "epoch": 0.6893161775336129,
+      "grad_norm": 0.25641652941703796,
+      "learning_rate": 7.700617283950618e-06,
+      "loss": 11.0195,
+      "step": 149
+    },
+    {
+      "epoch": 0.6939424606043082,
+      "grad_norm": 0.17132331430912018,
+      "learning_rate": 7.685185185185185e-06,
+      "loss": 13.9163,
+      "step": 150
+    },
+    {
+      "epoch": 0.6985687436750037,
+      "grad_norm": 0.24479900300502777,
+      "learning_rate": 7.669753086419753e-06,
+      "loss": 11.4487,
+      "step": 151
+    },
+    {
+      "epoch": 0.703195026745699,
+      "grad_norm": 0.1990821659564972,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 10.6267,
+      "step": 152
+    },
+    {
+      "epoch": 0.7078213098163944,
+      "grad_norm": 0.17914509773254395,
+      "learning_rate": 7.638888888888888e-06,
+      "loss": 11.5696,
+      "step": 153
+    },
+    {
+      "epoch": 0.7124475928870898,
+      "grad_norm": 0.19086718559265137,
+      "learning_rate": 7.623456790123458e-06,
+      "loss": 12.3853,
+      "step": 154
+    },
+    {
+      "epoch": 0.7170738759577852,
+      "grad_norm": 0.1417158991098404,
+      "learning_rate": 7.608024691358026e-06,
+      "loss": 10.8805,
+      "step": 155
+    },
+    {
+      "epoch": 0.7217001590284805,
+      "grad_norm": 0.32373055815696716,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 12.586,
+      "step": 156
+    },
+    {
+      "epoch": 0.726326442099176,
+      "grad_norm": 0.1995164453983307,
+      "learning_rate": 7.577160493827161e-06,
+      "loss": 11.4977,
+      "step": 157
+    },
+    {
+      "epoch": 0.7309527251698713,
+      "grad_norm": 0.16449247300624847,
+      "learning_rate": 7.561728395061729e-06,
+      "loss": 11.8027,
+      "step": 158
+    },
+    {
+      "epoch": 0.7355790082405668,
+      "grad_norm": 0.21888476610183716,
+      "learning_rate": 7.546296296296297e-06,
+      "loss": 11.8753,
+      "step": 159
+    },
+    {
+      "epoch": 0.7402052913112621,
+      "grad_norm": 0.19433487951755524,
+      "learning_rate": 7.530864197530865e-06,
+      "loss": 11.5252,
+      "step": 160
+    },
+    {
+      "epoch": 0.7448315743819575,
+      "grad_norm": 0.19709180295467377,
+      "learning_rate": 7.515432098765433e-06,
+      "loss": 10.7059,
+      "step": 161
+    },
+    {
+      "epoch": 0.7494578574526529,
+      "grad_norm": 0.17522279918193817,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 11.2329,
+      "step": 162
+    },
+    {
+      "epoch": 0.7540841405233483,
+      "grad_norm": 0.2038147747516632,
+      "learning_rate": 7.484567901234569e-06,
+      "loss": 11.425,
+      "step": 163
+    },
+    {
+      "epoch": 0.7587104235940436,
+      "grad_norm": 0.29561126232147217,
+      "learning_rate": 7.469135802469136e-06,
+      "loss": 13.192,
+      "step": 164
+    },
+    {
+      "epoch": 0.7633367066647391,
+      "grad_norm": 0.2295159548521042,
+      "learning_rate": 7.453703703703704e-06,
+      "loss": 10.9714,
+      "step": 165
+    },
+    {
+      "epoch": 0.7679629897354344,
+      "grad_norm": 0.1554422378540039,
+      "learning_rate": 7.438271604938272e-06,
+      "loss": 10.5617,
+      "step": 166
+    },
+    {
+      "epoch": 0.7725892728061299,
+      "grad_norm": 0.19565965235233307,
+      "learning_rate": 7.42283950617284e-06,
+      "loss": 10.8574,
+      "step": 167
+    },
+    {
+      "epoch": 0.7772155558768252,
+      "grad_norm": 0.2645648419857025,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 11.4472,
+      "step": 168
+    },
+    {
+      "epoch": 0.7818418389475206,
+      "grad_norm": 0.2832958996295929,
+      "learning_rate": 7.391975308641975e-06,
+      "loss": 11.628,
+      "step": 169
+    },
+    {
+      "epoch": 0.786468122018216,
+      "grad_norm": 0.2097349464893341,
+      "learning_rate": 7.3765432098765435e-06,
+      "loss": 12.2379,
+      "step": 170
+    },
+    {
+      "epoch": 0.7910944050889114,
+      "grad_norm": 0.19168758392333984,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 10.0489,
+      "step": 171
+    },
+    {
+      "epoch": 0.7957206881596067,
+      "grad_norm": 0.22470812499523163,
+      "learning_rate": 7.34567901234568e-06,
+      "loss": 11.8545,
+      "step": 172
+    },
+    {
+      "epoch": 0.8003469712303022,
+      "grad_norm": 0.2722543179988861,
+      "learning_rate": 7.330246913580248e-06,
+      "loss": 10.4316,
+      "step": 173
+    },
+    {
+      "epoch": 0.8049732543009975,
+      "grad_norm": 0.1686207354068756,
+      "learning_rate": 7.314814814814816e-06,
+      "loss": 12.5917,
+      "step": 174
+    },
+    {
+      "epoch": 0.8095995373716929,
+      "grad_norm": 0.2805918753147125,
+      "learning_rate": 7.299382716049383e-06,
+      "loss": 10.6457,
+      "step": 175
+    },
+    {
+      "epoch": 0.8142258204423883,
+      "grad_norm": 0.28346166014671326,
+      "learning_rate": 7.283950617283952e-06,
+      "loss": 11.6686,
+      "step": 176
+    },
+    {
+      "epoch": 0.8188521035130837,
+      "grad_norm": 0.23147298395633698,
+      "learning_rate": 7.268518518518519e-06,
+      "loss": 10.316,
+      "step": 177
+    },
+    {
+      "epoch": 0.8234783865837791,
+      "grad_norm": 0.17895270884037018,
+      "learning_rate": 7.253086419753087e-06,
+      "loss": 10.5762,
+      "step": 178
+    },
+    {
+      "epoch": 0.8281046696544745,
+      "grad_norm": 0.2260911464691162,
+      "learning_rate": 7.2376543209876545e-06,
+      "loss": 10.9989,
+      "step": 179
+    },
+    {
+      "epoch": 0.8327309527251698,
+      "grad_norm": 0.2163412719964981,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 10.7737,
+      "step": 180
+    },
+    {
+      "epoch": 0.8373572357958653,
+      "grad_norm": 0.21451716125011444,
+      "learning_rate": 7.2067901234567905e-06,
+      "loss": 12.4478,
+      "step": 181
+    },
+    {
+      "epoch": 0.8419835188665606,
+      "grad_norm": 0.14181958138942719,
+      "learning_rate": 7.191358024691358e-06,
+      "loss": 11.1375,
+      "step": 182
+    },
+    {
+      "epoch": 0.846609801937256,
+      "grad_norm": 0.18365442752838135,
+      "learning_rate": 7.1759259259259266e-06,
+      "loss": 11.3768,
+      "step": 183
+    },
+    {
+      "epoch": 0.8512360850079514,
+      "grad_norm": 0.20560483634471893,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 10.6342,
+      "step": 184
+    },
+    {
+      "epoch": 0.8558623680786468,
+      "grad_norm": 0.3396760821342468,
+      "learning_rate": 7.145061728395062e-06,
+      "loss": 12.9821,
+      "step": 185
+    },
+    {
+      "epoch": 0.8604886511493423,
+      "grad_norm": 0.18169118463993073,
+      "learning_rate": 7.129629629629629e-06,
+      "loss": 11.067,
+      "step": 186
+    },
+    {
+      "epoch": 0.8651149342200376,
+      "grad_norm": 0.18200387060642242,
+      "learning_rate": 7.114197530864199e-06,
+      "loss": 10.6416,
+      "step": 187
+    },
+    {
+      "epoch": 0.869741217290733,
+      "grad_norm": 0.23793809115886688,
+      "learning_rate": 7.098765432098766e-06,
+      "loss": 12.6389,
+      "step": 188
+    },
+    {
+      "epoch": 0.8743675003614284,
+      "grad_norm": 0.6948350667953491,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 11.5131,
+      "step": 189
+    },
+    {
+      "epoch": 0.8789937834321238,
+      "grad_norm": 0.19556772708892822,
+      "learning_rate": 7.067901234567902e-06,
+      "loss": 12.9335,
+      "step": 190
+    },
+    {
+      "epoch": 0.8836200665028191,
+      "grad_norm": 0.21277347207069397,
+      "learning_rate": 7.05246913580247e-06,
+      "loss": 11.533,
+      "step": 191
+    },
+    {
+      "epoch": 0.8882463495735146,
+      "grad_norm": 0.13241365551948547,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 11.7588,
+      "step": 192
+    },
+    {
+      "epoch": 0.8928726326442099,
+      "grad_norm": 0.16065774857997894,
+      "learning_rate": 7.021604938271606e-06,
+      "loss": 10.907,
+      "step": 193
+    },
+    {
+      "epoch": 0.8974989157149053,
+      "grad_norm": 0.2425934374332428,
+      "learning_rate": 7.0061728395061735e-06,
+      "loss": 12.0332,
+      "step": 194
+    },
+    {
+      "epoch": 0.9021251987856007,
+      "grad_norm": 0.32493147253990173,
+      "learning_rate": 6.990740740740741e-06,
+      "loss": 11.0481,
+      "step": 195
+    },
+    {
+      "epoch": 0.9067514818562961,
+      "grad_norm": 0.20006415247917175,
+      "learning_rate": 6.975308641975309e-06,
+      "loss": 10.3825,
+      "step": 196
+    },
+    {
+      "epoch": 0.9113777649269915,
+      "grad_norm": 0.14633627235889435,
+      "learning_rate": 6.959876543209877e-06,
+      "loss": 11.3131,
+      "step": 197
+    },
+    {
+      "epoch": 0.9160040479976869,
+      "grad_norm": 0.1695437729358673,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 11.6805,
+      "step": 198
+    },
+    {
+      "epoch": 0.9206303310683822,
+      "grad_norm": 0.22382433712482452,
+      "learning_rate": 6.929012345679012e-06,
+      "loss": 12.4138,
+      "step": 199
+    },
+    {
+      "epoch": 0.9252566141390777,
+      "grad_norm": 0.18652208149433136,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 11.1235,
+      "step": 200
+    },
+    {
+      "epoch": 0.929882897209773,
+      "grad_norm": 0.21566657721996307,
+      "learning_rate": 6.898148148148148e-06,
+      "loss": 11.1278,
+      "step": 201
+    },
+    {
+      "epoch": 0.9345091802804684,
+      "grad_norm": 0.17630067467689514,
+      "learning_rate": 6.882716049382716e-06,
+      "loss": 12.1383,
+      "step": 202
+    },
+    {
+      "epoch": 0.9391354633511638,
+      "grad_norm": 0.22753016650676727,
+      "learning_rate": 6.867283950617285e-06,
+      "loss": 12.0065,
+      "step": 203
+    },
+    {
+      "epoch": 0.9437617464218592,
+      "grad_norm": 0.12988047301769257,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 11.9477,
+      "step": 204
+    },
+    {
+      "epoch": 0.9483880294925546,
+      "grad_norm": 0.2507847547531128,
+      "learning_rate": 6.8364197530864205e-06,
+      "loss": 11.5818,
+      "step": 205
+    },
+    {
+      "epoch": 0.95301431256325,
+      "grad_norm": 0.18030938506126404,
+      "learning_rate": 6.820987654320988e-06,
+      "loss": 13.1558,
+      "step": 206
+    },
+    {
+      "epoch": 0.9576405956339453,
+      "grad_norm": 0.21600230038166046,
+      "learning_rate": 6.8055555555555566e-06,
+      "loss": 10.7038,
+      "step": 207
+    },
+    {
+      "epoch": 0.9622668787046408,
+      "grad_norm": 0.1989142745733261,
+      "learning_rate": 6.790123456790124e-06,
+      "loss": 11.9244,
+      "step": 208
+    },
+    {
+      "epoch": 0.9668931617753361,
+      "grad_norm": 0.2517797648906708,
+      "learning_rate": 6.774691358024692e-06,
+      "loss": 11.0019,
+      "step": 209
+    },
+    {
+      "epoch": 0.9715194448460315,
+      "grad_norm": 0.24413681030273438,
+      "learning_rate": 6.75925925925926e-06,
+      "loss": 12.4322,
+      "step": 210
+    },
+    {
+      "epoch": 0.9761457279167269,
+      "grad_norm": 0.20667782425880432,
+      "learning_rate": 6.743827160493828e-06,
+      "loss": 10.6134,
+      "step": 211
+    },
+    {
+      "epoch": 0.9807720109874223,
+      "grad_norm": 0.17336109280586243,
+      "learning_rate": 6.728395061728395e-06,
+      "loss": 10.4673,
+      "step": 212
+    },
+    {
+      "epoch": 0.9853982940581176,
+      "grad_norm": 0.17548175156116486,
+      "learning_rate": 6.712962962962963e-06,
+      "loss": 10.1377,
+      "step": 213
+    },
+    {
+      "epoch": 0.9900245771288131,
+      "grad_norm": 0.1717563420534134,
+      "learning_rate": 6.6975308641975314e-06,
+      "loss": 11.5888,
+      "step": 214
+    },
+    {
+      "epoch": 0.9946508601995084,
+      "grad_norm": 0.1841057986021042,
+      "learning_rate": 6.682098765432099e-06,
+      "loss": 12.2213,
+      "step": 215
+    },
+    {
+      "epoch": 0.9992771432702039,
+      "grad_norm": 0.3000975251197815,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 11.046,
+      "step": 216
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3000975251197815,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.4199,
+      "step": 217
+    },
+    {
+      "epoch": 1.0046262830706953,
+      "grad_norm": 0.4262640178203583,
+      "learning_rate": 6.651234567901235e-06,
+      "loss": 10.5265,
+      "step": 218
+    },
+    {
+      "epoch": 1.0092525661413907,
+      "grad_norm": 0.21694016456604004,
+      "learning_rate": 6.635802469135803e-06,
+      "loss": 10.4657,
+      "step": 219
+    },
+    {
+      "epoch": 1.0138788492120863,
+      "grad_norm": 0.2662133276462555,
+      "learning_rate": 6.620370370370371e-06,
+      "loss": 10.7531,
+      "step": 220
+    },
+    {
+      "epoch": 1.0185051322827816,
+      "grad_norm": 0.4052404463291168,
+      "learning_rate": 6.60493827160494e-06,
+      "loss": 11.1598,
+      "step": 221
+    },
+    {
+      "epoch": 1.023131415353477,
+      "grad_norm": 0.23900097608566284,
+      "learning_rate": 6.589506172839507e-06,
+      "loss": 11.5257,
+      "step": 222
+    },
+    {
+      "epoch": 1.0277576984241723,
+      "grad_norm": 0.22569526731967926,
+      "learning_rate": 6.574074074074075e-06,
+      "loss": 11.5564,
+      "step": 223
+    },
+    {
+      "epoch": 1.0323839814948677,
+      "grad_norm": 0.29018473625183105,
+      "learning_rate": 6.558641975308642e-06,
+      "loss": 11.8655,
+      "step": 224
+    },
+    {
+      "epoch": 1.037010264565563,
+      "grad_norm": 0.2856806516647339,
+      "learning_rate": 6.543209876543211e-06,
+      "loss": 11.8046,
+      "step": 225
+    },
+    {
+      "epoch": 1.0416365476362586,
+      "grad_norm": 0.16691464185714722,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 12.3704,
+      "step": 226
+    },
+    {
+      "epoch": 1.046262830706954,
+      "grad_norm": 0.24121911823749542,
+      "learning_rate": 6.512345679012346e-06,
+      "loss": 10.7236,
+      "step": 227
+    },
+    {
+      "epoch": 1.0508891137776493,
+      "grad_norm": 0.205901101231575,
+      "learning_rate": 6.4969135802469145e-06,
+      "loss": 10.9316,
+      "step": 228
+    },
+    {
+      "epoch": 1.0555153968483446,
+      "grad_norm": 0.18375830352306366,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 9.422,
+      "step": 229
+    },
+    {
+      "epoch": 1.06014167991904,
+      "grad_norm": 0.2132750004529953,
+      "learning_rate": 6.46604938271605e-06,
+      "loss": 12.6065,
+      "step": 230
+    },
+    {
+      "epoch": 1.0647679629897355,
+      "grad_norm": 0.23490285873413086,
+      "learning_rate": 6.450617283950617e-06,
+      "loss": 11.2282,
+      "step": 231
+    },
+    {
+      "epoch": 1.0693942460604309,
+      "grad_norm": 0.2604386806488037,
+      "learning_rate": 6.435185185185186e-06,
+      "loss": 10.8157,
+      "step": 232
+    },
+    {
+      "epoch": 1.0740205291311262,
+      "grad_norm": 0.21887636184692383,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 10.6996,
+      "step": 233
+    },
+    {
+      "epoch": 1.0786468122018216,
+      "grad_norm": 0.16946089267730713,
+      "learning_rate": 6.404320987654321e-06,
+      "loss": 10.3805,
+      "step": 234
+    },
+    {
+      "epoch": 1.083273095272517,
+      "grad_norm": 0.20204631984233856,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 11.4941,
+      "step": 235
+    },
+    {
+      "epoch": 1.0878993783432125,
+      "grad_norm": 0.202669158577919,
+      "learning_rate": 6.373456790123458e-06,
+      "loss": 11.7944,
+      "step": 236
+    },
+    {
+      "epoch": 1.0925256614139078,
+      "grad_norm": 0.22706535458564758,
+      "learning_rate": 6.358024691358025e-06,
+      "loss": 12.6031,
+      "step": 237
+    },
+    {
+      "epoch": 1.0971519444846032,
+      "grad_norm": 0.17785371840000153,
+      "learning_rate": 6.342592592592594e-06,
+      "loss": 10.5881,
+      "step": 238
+    },
+    {
+      "epoch": 1.1017782275552985,
+      "grad_norm": 0.17540261149406433,
+      "learning_rate": 6.3271604938271615e-06,
+      "loss": 10.996,
+      "step": 239
+    },
+    {
+      "epoch": 1.1064045106259939,
+      "grad_norm": 0.22038735449314117,
+      "learning_rate": 6.311728395061729e-06,
+      "loss": 10.0293,
+      "step": 240
+    },
+    {
+      "epoch": 1.1110307936966892,
+      "grad_norm": 0.17491741478443146,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 10.0996,
+      "step": 241
+    },
+    {
+      "epoch": 1.1156570767673848,
+      "grad_norm": 0.2827455699443817,
+      "learning_rate": 6.280864197530865e-06,
+      "loss": 11.0431,
+      "step": 242
+    },
+    {
+      "epoch": 1.1202833598380801,
+      "grad_norm": 0.21907830238342285,
+      "learning_rate": 6.265432098765433e-06,
+      "loss": 10.4707,
+      "step": 243
+    },
+    {
+      "epoch": 1.1249096429087755,
+      "grad_norm": 0.19320356845855713,
+      "learning_rate": 6.25e-06,
+      "loss": 10.7269,
+      "step": 244
+    },
+    {
+      "epoch": 1.1295359259794708,
+      "grad_norm": 0.22684165835380554,
+      "learning_rate": 6.234567901234569e-06,
+      "loss": 11.3143,
+      "step": 245
+    },
+    {
+      "epoch": 1.1341622090501662,
+      "grad_norm": 0.21541574597358704,
+      "learning_rate": 6.219135802469136e-06,
+      "loss": 10.397,
+      "step": 246
+    },
+    {
+      "epoch": 1.1387884921208618,
+      "grad_norm": 0.21086207032203674,
+      "learning_rate": 6.203703703703704e-06,
+      "loss": 11.0086,
+      "step": 247
+    },
+    {
+      "epoch": 1.143414775191557,
+      "grad_norm": 0.26652148365974426,
+      "learning_rate": 6.1882716049382715e-06,
+      "loss": 11.5252,
+      "step": 248
+    },
+    {
+      "epoch": 1.1480410582622524,
+      "grad_norm": 0.1995311677455902,
+      "learning_rate": 6.17283950617284e-06,
+      "loss": 9.3802,
+      "step": 249
+    },
+    {
+      "epoch": 1.1526673413329478,
+      "grad_norm": 0.19715267419815063,
+      "learning_rate": 6.157407407407408e-06,
+      "loss": 11.6554,
+      "step": 250
+    },
+    {
+      "epoch": 1.1572936244036431,
+      "grad_norm": 0.16357477009296417,
+      "learning_rate": 6.141975308641975e-06,
+      "loss": 10.7531,
+      "step": 251
+    },
+    {
+      "epoch": 1.1619199074743385,
+      "grad_norm": 0.2201809138059616,
+      "learning_rate": 6.126543209876543e-06,
+      "loss": 11.4639,
+      "step": 252
+    },
+    {
+      "epoch": 1.166546190545034,
+      "grad_norm": 0.24760210514068604,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 10.1224,
+      "step": 253
+    },
+    {
+      "epoch": 1.1711724736157294,
+      "grad_norm": 0.18471167981624603,
+      "learning_rate": 6.09567901234568e-06,
+      "loss": 10.8261,
+      "step": 254
+    },
+    {
+      "epoch": 1.1757987566864248,
+      "grad_norm": 0.20257946848869324,
+      "learning_rate": 6.080246913580248e-06,
+      "loss": 9.8945,
+      "step": 255
+    },
+    {
+      "epoch": 1.18042503975712,
+      "grad_norm": 0.15809153020381927,
+      "learning_rate": 6.064814814814816e-06,
+      "loss": 10.4489,
+      "step": 256
+    },
+    {
+      "epoch": 1.1850513228278154,
+      "grad_norm": 0.2175232172012329,
+      "learning_rate": 6.049382716049383e-06,
+      "loss": 10.6301,
+      "step": 257
+    },
+    {
+      "epoch": 1.189677605898511,
+      "grad_norm": 0.22652743756771088,
+      "learning_rate": 6.033950617283951e-06,
+      "loss": 9.6189,
+      "step": 258
+    },
+    {
+      "epoch": 1.1943038889692064,
+      "grad_norm": 0.21281598508358002,
+      "learning_rate": 6.018518518518519e-06,
+      "loss": 10.8016,
+      "step": 259
+    },
+    {
+      "epoch": 1.1989301720399017,
+      "grad_norm": 0.20046480000019073,
+      "learning_rate": 6.003086419753087e-06,
+      "loss": 10.1304,
+      "step": 260
+    },
+    {
+      "epoch": 1.203556455110597,
+      "grad_norm": 0.19008351862430573,
+      "learning_rate": 5.9876543209876546e-06,
+      "loss": 10.6718,
+      "step": 261
+    },
+    {
+      "epoch": 1.2081827381812924,
+      "grad_norm": 0.2159290611743927,
+      "learning_rate": 5.972222222222222e-06,
+      "loss": 10.1963,
+      "step": 262
+    },
+    {
+      "epoch": 1.2128090212519878,
+      "grad_norm": 0.1703040450811386,
+      "learning_rate": 5.956790123456791e-06,
+      "loss": 10.9394,
+      "step": 263
+    },
+    {
+      "epoch": 1.2174353043226833,
+      "grad_norm": 0.20243868231773376,
+      "learning_rate": 5.941358024691358e-06,
+      "loss": 9.7358,
+      "step": 264
+    },
+    {
+      "epoch": 1.2220615873933787,
+      "grad_norm": 0.21622534096240997,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 9.2896,
+      "step": 265
+    },
+    {
+      "epoch": 1.226687870464074,
+      "grad_norm": 0.19408589601516724,
+      "learning_rate": 5.910493827160494e-06,
+      "loss": 10.0141,
+      "step": 266
+    },
+    {
+      "epoch": 1.2313141535347694,
+      "grad_norm": 0.19652226567268372,
+      "learning_rate": 5.895061728395062e-06,
+      "loss": 10.6724,
+      "step": 267
+    },
+    {
+      "epoch": 1.2359404366054647,
+      "grad_norm": 0.22394828498363495,
+      "learning_rate": 5.8796296296296295e-06,
+      "loss": 11.7472,
+      "step": 268
+    },
+    {
+      "epoch": 1.2405667196761603,
+      "grad_norm": 0.18796682357788086,
+      "learning_rate": 5.864197530864199e-06,
+      "loss": 11.3051,
+      "step": 269
+    },
+    {
+      "epoch": 1.2451930027468556,
+      "grad_norm": 0.1897086501121521,
+      "learning_rate": 5.848765432098766e-06,
+      "loss": 9.9475,
+      "step": 270
+    },
+    {
+      "epoch": 1.249819285817551,
+      "grad_norm": 0.19912078976631165,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 10.6907,
+      "step": 271
+    },
+    {
+      "epoch": 1.2544455688882463,
+      "grad_norm": 0.25889354944229126,
+      "learning_rate": 5.817901234567902e-06,
+      "loss": 9.9069,
+      "step": 272
+    },
+    {
+      "epoch": 1.2590718519589417,
+      "grad_norm": 0.19028806686401367,
+      "learning_rate": 5.80246913580247e-06,
+      "loss": 10.0748,
+      "step": 273
+    },
+    {
+      "epoch": 1.263698135029637,
+      "grad_norm": 0.1661495566368103,
+      "learning_rate": 5.787037037037038e-06,
+      "loss": 10.1342,
+      "step": 274
+    },
+    {
+      "epoch": 1.2683244181003326,
+      "grad_norm": 0.17706620693206787,
+      "learning_rate": 5.771604938271605e-06,
+      "loss": 12.4703,
+      "step": 275
+    },
+    {
+      "epoch": 1.272950701171028,
+      "grad_norm": 0.26582854986190796,
+      "learning_rate": 5.756172839506174e-06,
+      "loss": 12.6464,
+      "step": 276
+    },
+    {
+      "epoch": 1.2775769842417233,
+      "grad_norm": 0.178712397813797,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 9.6011,
+      "step": 277
+    },
+    {
+      "epoch": 1.2822032673124186,
+      "grad_norm": 0.18910570442676544,
+      "learning_rate": 5.725308641975309e-06,
+      "loss": 10.8773,
+      "step": 278
+    },
+    {
+      "epoch": 1.2868295503831142,
+      "grad_norm": 0.24754610657691956,
+      "learning_rate": 5.7098765432098764e-06,
+      "loss": 10.8981,
+      "step": 279
+    },
+    {
+      "epoch": 1.2914558334538095,
+      "grad_norm": 0.14676739275455475,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 10.1851,
+      "step": 280
+    },
+    {
+      "epoch": 1.296082116524505,
+      "grad_norm": 0.21573857963085175,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 10.2565,
+      "step": 281
+    },
+    {
+      "epoch": 1.3007083995952002,
+      "grad_norm": 0.24850860238075256,
+      "learning_rate": 5.66358024691358e-06,
+      "loss": 9.9282,
+      "step": 282
+    },
+    {
+      "epoch": 1.3053346826658956,
+      "grad_norm": 0.180882066488266,
+      "learning_rate": 5.6481481481481485e-06,
+      "loss": 11.6456,
+      "step": 283
+    },
+    {
+      "epoch": 1.309960965736591,
+      "grad_norm": 0.19305004179477692,
+      "learning_rate": 5.632716049382716e-06,
+      "loss": 10.0129,
+      "step": 284
+    },
+    {
+      "epoch": 1.3145872488072863,
+      "grad_norm": 0.18499980866909027,
+      "learning_rate": 5.617283950617285e-06,
+      "loss": 10.6136,
+      "step": 285
+    },
+    {
+      "epoch": 1.3192135318779818,
+      "grad_norm": 0.2523798942565918,
+      "learning_rate": 5.601851851851853e-06,
+      "loss": 9.1629,
+      "step": 286
+    },
+    {
+      "epoch": 1.3238398149486772,
+      "grad_norm": 0.19359612464904785,
+      "learning_rate": 5.586419753086421e-06,
+      "loss": 9.9917,
+      "step": 287
+    },
+    {
+      "epoch": 1.3284660980193725,
+      "grad_norm": 0.15306031703948975,
+      "learning_rate": 5.570987654320988e-06,
+      "loss": 9.6058,
+      "step": 288
+    },
+    {
+      "epoch": 1.333092381090068,
+      "grad_norm": 0.2070600688457489,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 10.2776,
+      "step": 289
+    },
+    {
+      "epoch": 1.3377186641607635,
+      "grad_norm": 0.1619105190038681,
+      "learning_rate": 5.540123456790124e-06,
+      "loss": 9.1465,
+      "step": 290
+    },
+    {
+      "epoch": 1.3423449472314588,
+      "grad_norm": 0.16202016174793243,
+      "learning_rate": 5.524691358024692e-06,
+      "loss": 10.0362,
+      "step": 291
+    },
+    {
+      "epoch": 1.3469712303021542,
+      "grad_norm": 0.19075003266334534,
+      "learning_rate": 5.5092592592592595e-06,
+      "loss": 10.0189,
+      "step": 292
+    },
+    {
+      "epoch": 1.3515975133728495,
+      "grad_norm": 0.16484172642230988,
+      "learning_rate": 5.493827160493828e-06,
+      "loss": 12.3743,
+      "step": 293
+    },
+    {
+      "epoch": 1.3562237964435448,
+      "grad_norm": 0.2151799499988556,
+      "learning_rate": 5.4783950617283955e-06,
+      "loss": 9.9846,
+      "step": 294
+    },
+    {
+      "epoch": 1.3608500795142402,
+      "grad_norm": 0.20363835990428925,
+      "learning_rate": 5.462962962962963e-06,
+      "loss": 9.671,
+      "step": 295
+    },
+    {
+      "epoch": 1.3654763625849355,
+      "grad_norm": 0.15196539461612701,
+      "learning_rate": 5.447530864197531e-06,
+      "loss": 9.97,
+      "step": 296
+    },
+    {
+      "epoch": 1.3701026456556311,
+      "grad_norm": 0.17592975497245789,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 10.8506,
+      "step": 297
+    },
+    {
+      "epoch": 1.3747289287263265,
+      "grad_norm": 0.20012719929218292,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 10.4503,
+      "step": 298
+    },
+    {
+      "epoch": 1.3793552117970218,
+      "grad_norm": 0.19312864542007446,
+      "learning_rate": 5.401234567901234e-06,
+      "loss": 9.7727,
+      "step": 299
+    },
+    {
+      "epoch": 1.3839814948677172,
+      "grad_norm": 0.14697419106960297,
+      "learning_rate": 5.385802469135803e-06,
+      "loss": 11.9123,
+      "step": 300
+    },
+    {
+      "epoch": 1.3886077779384127,
+      "grad_norm": 0.15580600500106812,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 10.3442,
+      "step": 301
+    },
+    {
+      "epoch": 1.393234061009108,
+      "grad_norm": 0.19586268067359924,
+      "learning_rate": 5.354938271604939e-06,
+      "loss": 8.9558,
+      "step": 302
+    },
+    {
+      "epoch": 1.3978603440798034,
+      "grad_norm": 0.19394423067569733,
+      "learning_rate": 5.339506172839507e-06,
+      "loss": 10.8312,
+      "step": 303
+    },
+    {
+      "epoch": 1.4024866271504988,
+      "grad_norm": 0.21832065284252167,
+      "learning_rate": 5.324074074074075e-06,
+      "loss": 9.9943,
+      "step": 304
+    },
+    {
+      "epoch": 1.4071129102211941,
+      "grad_norm": 0.16751375794410706,
+      "learning_rate": 5.3086419753086425e-06,
+      "loss": 11.3979,
+      "step": 305
+    },
+    {
+      "epoch": 1.4117391932918895,
+      "grad_norm": 0.1763811856508255,
+      "learning_rate": 5.29320987654321e-06,
+      "loss": 11.175,
+      "step": 306
+    },
+    {
+      "epoch": 1.4163654763625848,
+      "grad_norm": 0.1916954666376114,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 8.8938,
+      "step": 307
+    },
+    {
+      "epoch": 1.4209917594332804,
+      "grad_norm": 0.12443282455205917,
+      "learning_rate": 5.262345679012346e-06,
+      "loss": 9.6263,
+      "step": 308
+    },
+    {
+      "epoch": 1.4256180425039757,
+      "grad_norm": 0.18852603435516357,
+      "learning_rate": 5.246913580246914e-06,
+      "loss": 10.6811,
+      "step": 309
+    },
+    {
+      "epoch": 1.430244325574671,
+      "grad_norm": 0.1854352355003357,
+      "learning_rate": 5.231481481481482e-06,
+      "loss": 9.9107,
+      "step": 310
+    },
+    {
+      "epoch": 1.4348706086453664,
+      "grad_norm": 0.2382035106420517,
+      "learning_rate": 5.21604938271605e-06,
+      "loss": 10.8001,
+      "step": 311
+    },
+    {
+      "epoch": 1.439496891716062,
+      "grad_norm": 0.18645620346069336,
+      "learning_rate": 5.200617283950617e-06,
+      "loss": 9.686,
+      "step": 312
+    },
+    {
+      "epoch": 1.4441231747867573,
+      "grad_norm": 0.1910058856010437,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 10.3407,
+      "step": 313
+    },
+    {
+      "epoch": 1.4487494578574527,
+      "grad_norm": 0.23428039252758026,
+      "learning_rate": 5.1697530864197534e-06,
+      "loss": 9.2624,
+      "step": 314
+    },
+    {
+      "epoch": 1.453375740928148,
+      "grad_norm": 0.17926917970180511,
+      "learning_rate": 5.154320987654321e-06,
+      "loss": 9.5153,
+      "step": 315
+    },
+    {
+      "epoch": 1.4580020239988434,
+      "grad_norm": 0.22386913001537323,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 10.6948,
+      "step": 316
+    },
+    {
+      "epoch": 1.4626283070695387,
+      "grad_norm": 0.20423167943954468,
+      "learning_rate": 5.123456790123458e-06,
+      "loss": 10.3966,
+      "step": 317
+    },
+    {
+      "epoch": 1.467254590140234,
+      "grad_norm": 0.15411533415317535,
+      "learning_rate": 5.1080246913580255e-06,
+      "loss": 10.2981,
+      "step": 318
+    },
+    {
+      "epoch": 1.4718808732109296,
+      "grad_norm": 0.11654549837112427,
+      "learning_rate": 5.092592592592593e-06,
+      "loss": 10.1807,
+      "step": 319
+    },
+    {
+      "epoch": 1.476507156281625,
+      "grad_norm": 0.20476582646369934,
+      "learning_rate": 5.0771604938271616e-06,
+      "loss": 9.9201,
+      "step": 320
+    },
+    {
+      "epoch": 1.4811334393523203,
+      "grad_norm": 0.19140280783176422,
+      "learning_rate": 5.061728395061729e-06,
+      "loss": 10.3292,
+      "step": 321
+    },
+    {
+      "epoch": 1.4857597224230157,
+      "grad_norm": 0.16844482719898224,
+      "learning_rate": 5.046296296296297e-06,
+      "loss": 10.0726,
+      "step": 322
+    },
+    {
+      "epoch": 1.4903860054937113,
+      "grad_norm": 0.22298437356948853,
+      "learning_rate": 5.030864197530864e-06,
+      "loss": 10.9447,
+      "step": 323
+    },
+    {
+      "epoch": 1.4950122885644066,
+      "grad_norm": 0.1775355190038681,
+      "learning_rate": 5.015432098765433e-06,
+      "loss": 12.1205,
+      "step": 324
+    },
+    {
+      "epoch": 1.499638571635102,
+      "grad_norm": 0.198349267244339,
+      "learning_rate": 5e-06,
+      "loss": 9.0277,
+      "step": 325
+    },
+    {
+      "epoch": 1.5042648547057973,
+      "grad_norm": 0.16838030517101288,
+      "learning_rate": 4.984567901234568e-06,
+      "loss": 10.3551,
+      "step": 326
+    },
+    {
+      "epoch": 1.5088911377764926,
+      "grad_norm": 0.15806855261325836,
+      "learning_rate": 4.9691358024691365e-06,
+      "loss": 10.0048,
+      "step": 327
+    },
+    {
+      "epoch": 1.513517420847188,
+      "grad_norm": 0.21915429830551147,
+      "learning_rate": 4.953703703703704e-06,
+      "loss": 10.554,
+      "step": 328
+    },
+    {
+      "epoch": 1.5181437039178833,
+      "grad_norm": 0.23262116312980652,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 10.8198,
+      "step": 329
+    },
+    {
+      "epoch": 1.522769986988579,
+      "grad_norm": 0.18927887082099915,
+      "learning_rate": 4.92283950617284e-06,
+      "loss": 10.7398,
+      "step": 330
+    },
+    {
+      "epoch": 1.5273962700592743,
+      "grad_norm": 0.138357013463974,
+      "learning_rate": 4.907407407407408e-06,
+      "loss": 9.8479,
+      "step": 331
+    },
+    {
+      "epoch": 1.5320225531299696,
+      "grad_norm": 0.20223885774612427,
+      "learning_rate": 4.891975308641976e-06,
+      "loss": 11.1796,
+      "step": 332
+    },
+    {
+      "epoch": 1.5366488362006652,
+      "grad_norm": 0.17738379538059235,
+      "learning_rate": 4.876543209876544e-06,
+      "loss": 9.1805,
+      "step": 333
+    },
+    {
+      "epoch": 1.5412751192713605,
+      "grad_norm": 0.18442702293395996,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 8.4714,
+      "step": 334
+    },
+    {
+      "epoch": 1.5459014023420559,
+      "grad_norm": 0.19169731438159943,
+      "learning_rate": 4.845679012345679e-06,
+      "loss": 9.1503,
+      "step": 335
+    },
+    {
+      "epoch": 1.5505276854127512,
+      "grad_norm": 0.19981014728546143,
+      "learning_rate": 4.830246913580247e-06,
+      "loss": 9.6427,
+      "step": 336
+    },
+    {
+      "epoch": 1.5551539684834466,
+      "grad_norm": 0.1916283518075943,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 8.8991,
+      "step": 337
+    },
+    {
+      "epoch": 1.559780251554142,
+      "grad_norm": 0.14049910008907318,
+      "learning_rate": 4.7993827160493834e-06,
+      "loss": 9.2888,
+      "step": 338
+    },
+    {
+      "epoch": 1.5644065346248373,
+      "grad_norm": 0.14010216295719147,
+      "learning_rate": 4.783950617283951e-06,
+      "loss": 10.4533,
+      "step": 339
+    },
+    {
+      "epoch": 1.5690328176955326,
+      "grad_norm": 0.1919959932565689,
+      "learning_rate": 4.768518518518519e-06,
+      "loss": 10.149,
+      "step": 340
+    },
+    {
+      "epoch": 1.5736591007662282,
+      "grad_norm": 0.1598203033208847,
+      "learning_rate": 4.753086419753087e-06,
+      "loss": 11.1529,
+      "step": 341
+    },
+    {
+      "epoch": 1.5782853838369235,
+      "grad_norm": 0.19053767621517181,
+      "learning_rate": 4.737654320987655e-06,
+      "loss": 10.5543,
+      "step": 342
+    },
+    {
+      "epoch": 1.5829116669076189,
+      "grad_norm": 0.21635524928569794,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 8.7965,
+      "step": 343
+    },
+    {
+      "epoch": 1.5875379499783144,
+      "grad_norm": 0.2088153064250946,
+      "learning_rate": 4.706790123456791e-06,
+      "loss": 9.1113,
+      "step": 344
+    },
+    {
+      "epoch": 1.5921642330490098,
+      "grad_norm": 0.1993483006954193,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 9.7291,
+      "step": 345
+    },
+    {
+      "epoch": 1.5967905161197051,
+      "grad_norm": 0.24835029244422913,
+      "learning_rate": 4.675925925925927e-06,
+      "loss": 9.5215,
+      "step": 346
+    },
+    {
+      "epoch": 1.6014167991904005,
+      "grad_norm": 0.24942326545715332,
+      "learning_rate": 4.660493827160494e-06,
+      "loss": 9.3023,
+      "step": 347
+    },
+    {
+      "epoch": 1.6060430822610958,
+      "grad_norm": 0.18829648196697235,
+      "learning_rate": 4.645061728395062e-06,
+      "loss": 9.6869,
+      "step": 348
+    },
+    {
+      "epoch": 1.6106693653317912,
+      "grad_norm": 0.17725762724876404,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 10.0154,
+      "step": 349
+    },
+    {
+      "epoch": 1.6152956484024865,
+      "grad_norm": 0.145020991563797,
+      "learning_rate": 4.614197530864198e-06,
+      "loss": 10.6015,
+      "step": 350
+    },
+    {
+      "epoch": 1.6199219314731819,
+      "grad_norm": 0.12719136476516724,
+      "learning_rate": 4.598765432098766e-06,
+      "loss": 9.9244,
+      "step": 351
+    },
+    {
+      "epoch": 1.6245482145438774,
+      "grad_norm": 0.1771680861711502,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 9.4468,
+      "step": 352
+    },
+    {
+      "epoch": 1.6291744976145728,
+      "grad_norm": 0.1684349775314331,
+      "learning_rate": 4.567901234567902e-06,
+      "loss": 11.5601,
+      "step": 353
+    },
+    {
+      "epoch": 1.6338007806852681,
+      "grad_norm": 0.1284870207309723,
+      "learning_rate": 4.55246913580247e-06,
+      "loss": 9.5658,
+      "step": 354
+    },
+    {
+      "epoch": 1.6384270637559637,
+      "grad_norm": 0.15432706475257874,
+      "learning_rate": 4.537037037037038e-06,
+      "loss": 10.2547,
+      "step": 355
+    },
+    {
+      "epoch": 1.643053346826659,
+      "grad_norm": 0.19519680738449097,
+      "learning_rate": 4.521604938271605e-06,
+      "loss": 9.2126,
+      "step": 356
+    },
+    {
+      "epoch": 1.6476796298973544,
+      "grad_norm": 0.1388140469789505,
+      "learning_rate": 4.506172839506173e-06,
+      "loss": 9.9578,
+      "step": 357
+    },
+    {
+      "epoch": 1.6523059129680497,
+      "grad_norm": 0.19714529812335968,
+      "learning_rate": 4.490740740740741e-06,
+      "loss": 11.7677,
+      "step": 358
+    },
+    {
+      "epoch": 1.656932196038745,
+      "grad_norm": 0.22792014479637146,
+      "learning_rate": 4.475308641975309e-06,
+      "loss": 9.6164,
+      "step": 359
+    },
+    {
+      "epoch": 1.6615584791094404,
+      "grad_norm": 0.1519528478384018,
+      "learning_rate": 4.4598765432098765e-06,
+      "loss": 9.4579,
+      "step": 360
+    },
+    {
+      "epoch": 1.6661847621801358,
+      "grad_norm": 0.17547698318958282,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 8.2953,
+      "step": 361
+    },
+    {
+      "epoch": 1.6708110452508311,
+      "grad_norm": 0.2348184585571289,
+      "learning_rate": 4.429012345679013e-06,
+      "loss": 9.8739,
+      "step": 362
+    },
+    {
+      "epoch": 1.6754373283215267,
+      "grad_norm": 0.16753868758678436,
+      "learning_rate": 4.413580246913581e-06,
+      "loss": 9.8006,
+      "step": 363
+    },
+    {
+      "epoch": 1.680063611392222,
+      "grad_norm": 0.16872891783714294,
+      "learning_rate": 4.398148148148149e-06,
+      "loss": 10.2161,
+      "step": 364
+    },
+    {
+      "epoch": 1.6846898944629176,
+      "grad_norm": 0.18189671635627747,
+      "learning_rate": 4.382716049382716e-06,
+      "loss": 9.5235,
+      "step": 365
+    },
+    {
+      "epoch": 1.689316177533613,
+      "grad_norm": 0.12333816289901733,
+      "learning_rate": 4.367283950617285e-06,
+      "loss": 10.2188,
+      "step": 366
+    },
+    {
+      "epoch": 1.6939424606043083,
+      "grad_norm": 0.22701486945152283,
+      "learning_rate": 4.351851851851852e-06,
+      "loss": 10.7382,
+      "step": 367
+    },
+    {
+      "epoch": 1.6985687436750037,
+      "grad_norm": 0.20510244369506836,
+      "learning_rate": 4.33641975308642e-06,
+      "loss": 8.7558,
+      "step": 368
+    },
+    {
+      "epoch": 1.703195026745699,
+      "grad_norm": 0.1589343398809433,
+      "learning_rate": 4.3209876543209875e-06,
+      "loss": 9.924,
+      "step": 369
+    },
+    {
+      "epoch": 1.7078213098163944,
+      "grad_norm": 0.1433723270893097,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 11.0008,
+      "step": 370
+    },
+    {
+      "epoch": 1.7124475928870897,
+      "grad_norm": 0.17607755959033966,
+      "learning_rate": 4.290123456790124e-06,
+      "loss": 10.0096,
+      "step": 371
+    },
+    {
+      "epoch": 1.717073875957785,
+      "grad_norm": 0.16202040016651154,
+      "learning_rate": 4.274691358024692e-06,
+      "loss": 9.9295,
+      "step": 372
+    },
+    {
+      "epoch": 1.7217001590284804,
+      "grad_norm": 0.21179550886154175,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 8.9856,
+      "step": 373
+    },
+    {
+      "epoch": 1.726326442099176,
+      "grad_norm": 0.35698196291923523,
+      "learning_rate": 4.243827160493827e-06,
+      "loss": 9.8793,
+      "step": 374
+    },
+    {
+      "epoch": 1.7309527251698713,
+      "grad_norm": 0.18081314861774445,
+      "learning_rate": 4.228395061728396e-06,
+      "loss": 10.8514,
+      "step": 375
+    },
+    {
+      "epoch": 1.7355790082405669,
+      "grad_norm": 0.19385862350463867,
+      "learning_rate": 4.212962962962963e-06,
+      "loss": 8.6034,
+      "step": 376
+    },
+    {
+      "epoch": 1.7402052913112622,
+      "grad_norm": 0.17623740434646606,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 10.7457,
+      "step": 377
+    },
+    {
+      "epoch": 1.7448315743819576,
+      "grad_norm": 0.1551784873008728,
+      "learning_rate": 4.182098765432099e-06,
+      "loss": 9.2378,
+      "step": 378
+    },
+    {
+      "epoch": 1.749457857452653,
+      "grad_norm": 0.15923646092414856,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 10.0731,
+      "step": 379
+    },
+    {
+      "epoch": 1.7540841405233483,
+      "grad_norm": 0.16185562312602997,
+      "learning_rate": 4.151234567901235e-06,
+      "loss": 8.9162,
+      "step": 380
+    },
+    {
+      "epoch": 1.7587104235940436,
+      "grad_norm": 0.23534594476222992,
+      "learning_rate": 4.135802469135803e-06,
+      "loss": 9.3805,
+      "step": 381
+    },
+    {
+      "epoch": 1.763336706664739,
+      "grad_norm": 0.17162498831748962,
+      "learning_rate": 4.1203703703703705e-06,
+      "loss": 10.0609,
+      "step": 382
+    },
+    {
+      "epoch": 1.7679629897354343,
+      "grad_norm": 0.15775223076343536,
+      "learning_rate": 4.104938271604938e-06,
+      "loss": 9.3534,
+      "step": 383
+    },
+    {
+      "epoch": 1.7725892728061299,
+      "grad_norm": 0.2251134216785431,
+      "learning_rate": 4.0895061728395066e-06,
+      "loss": 9.746,
+      "step": 384
+    },
+    {
+      "epoch": 1.7772155558768252,
+      "grad_norm": 0.19613635540008545,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 10.42,
+      "step": 385
+    },
+    {
+      "epoch": 1.7818418389475206,
+      "grad_norm": 0.16256873309612274,
+      "learning_rate": 4.058641975308643e-06,
+      "loss": 9.4703,
+      "step": 386
+    },
+    {
+      "epoch": 1.7864681220182161,
+      "grad_norm": 0.16490335762500763,
+      "learning_rate": 4.04320987654321e-06,
+      "loss": 9.5418,
+      "step": 387
+    },
+    {
+      "epoch": 1.7910944050889115,
+      "grad_norm": 0.159736767411232,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 10.1403,
+      "step": 388
+    },
+    {
+      "epoch": 1.7957206881596068,
+      "grad_norm": 0.1494799256324768,
+      "learning_rate": 4.012345679012346e-06,
+      "loss": 9.1583,
+      "step": 389
+    },
+    {
+      "epoch": 1.8003469712303022,
+      "grad_norm": 0.19815048575401306,
+      "learning_rate": 3.996913580246914e-06,
+      "loss": 10.2705,
+      "step": 390
+    },
+    {
+      "epoch": 1.8049732543009975,
+      "grad_norm": 0.17572833597660065,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 9.4608,
+      "step": 391
+    },
+    {
+      "epoch": 1.8095995373716929,
+      "grad_norm": 0.23244017362594604,
+      "learning_rate": 3.96604938271605e-06,
+      "loss": 10.6273,
+      "step": 392
+    },
+    {
+      "epoch": 1.8142258204423882,
+      "grad_norm": 0.17120610177516937,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 10.0444,
+      "step": 393
+    },
+    {
+      "epoch": 1.8188521035130836,
+      "grad_norm": 0.27220579981803894,
+      "learning_rate": 3.935185185185186e-06,
+      "loss": 10.64,
+      "step": 394
+    },
+    {
+      "epoch": 1.8234783865837791,
+      "grad_norm": 0.1648477017879486,
+      "learning_rate": 3.9197530864197535e-06,
+      "loss": 9.6264,
+      "step": 395
+    },
+    {
+      "epoch": 1.8281046696544745,
+      "grad_norm": 0.15771055221557617,
+      "learning_rate": 3.904320987654321e-06,
+      "loss": 10.5564,
+      "step": 396
+    },
+    {
+      "epoch": 1.8327309527251698,
+      "grad_norm": 0.21713097393512726,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 10.6551,
+      "step": 397
+    },
+    {
+      "epoch": 1.8373572357958654,
+      "grad_norm": 0.17076119780540466,
+      "learning_rate": 3.873456790123457e-06,
+      "loss": 11.1632,
+      "step": 398
+    },
+    {
+      "epoch": 1.8419835188665608,
+      "grad_norm": 0.160685196518898,
+      "learning_rate": 3.858024691358025e-06,
+      "loss": 9.4666,
+      "step": 399
+    },
+    {
+      "epoch": 1.846609801937256,
+      "grad_norm": 0.17023253440856934,
+      "learning_rate": 3.842592592592592e-06,
+      "loss": 9.8118,
+      "step": 400
+    },
+    {
+      "epoch": 1.8512360850079514,
+      "grad_norm": 0.2186374068260193,
+      "learning_rate": 3.827160493827161e-06,
+      "loss": 10.4903,
+      "step": 401
+    },
+    {
+      "epoch": 1.8558623680786468,
+      "grad_norm": 0.17954406142234802,
+      "learning_rate": 3.811728395061729e-06,
+      "loss": 11.3069,
+      "step": 402
+    },
+    {
+      "epoch": 1.8604886511493421,
+      "grad_norm": 0.2778474986553192,
+      "learning_rate": 3.796296296296297e-06,
+      "loss": 10.5854,
+      "step": 403
+    },
+    {
+      "epoch": 1.8651149342200375,
+      "grad_norm": 0.27708959579467773,
+      "learning_rate": 3.7808641975308645e-06,
+      "loss": 10.2884,
+      "step": 404
+    },
+    {
+      "epoch": 1.8697412172907328,
+      "grad_norm": 0.14399868249893188,
+      "learning_rate": 3.7654320987654325e-06,
+      "loss": 10.1108,
+      "step": 405
+    },
+    {
+      "epoch": 1.8743675003614284,
+      "grad_norm": 0.19901257753372192,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 10.2892,
+      "step": 406
+    },
+    {
+      "epoch": 1.8789937834321238,
+      "grad_norm": 0.21875345706939697,
+      "learning_rate": 3.734567901234568e-06,
+      "loss": 9.0521,
+      "step": 407
+    },
+    {
+      "epoch": 1.883620066502819,
+      "grad_norm": 0.14913184940814972,
+      "learning_rate": 3.719135802469136e-06,
+      "loss": 10.5905,
+      "step": 408
+    },
+    {
+      "epoch": 1.8882463495735147,
+      "grad_norm": 0.22289720177650452,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 9.1175,
+      "step": 409
+    },
+    {
+      "epoch": 1.89287263264421,
+      "grad_norm": 0.16322648525238037,
+      "learning_rate": 3.6882716049382718e-06,
+      "loss": 10.1569,
+      "step": 410
+    },
+    {
+      "epoch": 1.8974989157149054,
+      "grad_norm": 0.13389474153518677,
+      "learning_rate": 3.67283950617284e-06,
+      "loss": 9.7421,
+      "step": 411
+    },
+    {
+      "epoch": 1.9021251987856007,
+      "grad_norm": 0.17904877662658691,
+      "learning_rate": 3.657407407407408e-06,
+      "loss": 10.306,
+      "step": 412
+    },
+    {
+      "epoch": 1.906751481856296,
+      "grad_norm": 0.22490617632865906,
+      "learning_rate": 3.641975308641976e-06,
+      "loss": 10.0275,
+      "step": 413
+    },
+    {
+      "epoch": 1.9113777649269914,
+      "grad_norm": 0.20119769871234894,
+      "learning_rate": 3.6265432098765434e-06,
+      "loss": 9.5942,
+      "step": 414
+    },
+    {
+      "epoch": 1.9160040479976868,
+      "grad_norm": 0.156976118683815,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 10.3435,
+      "step": 415
+    },
+    {
+      "epoch": 1.920630331068382,
+      "grad_norm": 0.13303594291210175,
+      "learning_rate": 3.595679012345679e-06,
+      "loss": 8.8616,
+      "step": 416
+    },
+    {
+      "epoch": 1.9252566141390777,
+      "grad_norm": 0.21186932921409607,
+      "learning_rate": 3.580246913580247e-06,
+      "loss": 9.8981,
+      "step": 417
+    },
+    {
+      "epoch": 1.929882897209773,
+      "grad_norm": 0.20669791102409363,
+      "learning_rate": 3.5648148148148147e-06,
+      "loss": 10.3689,
+      "step": 418
+    },
+    {
+      "epoch": 1.9345091802804684,
+      "grad_norm": 0.20095276832580566,
+      "learning_rate": 3.549382716049383e-06,
+      "loss": 9.136,
+      "step": 419
+    },
+    {
+      "epoch": 1.939135463351164,
+      "grad_norm": 0.16348451375961304,
+      "learning_rate": 3.533950617283951e-06,
+      "loss": 10.4419,
+      "step": 420
+    },
+    {
+      "epoch": 1.9437617464218593,
+      "grad_norm": 0.16439932584762573,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 10.0692,
+      "step": 421
+    },
+    {
+      "epoch": 1.9483880294925546,
+      "grad_norm": 0.18562501668930054,
+      "learning_rate": 3.5030864197530868e-06,
+      "loss": 10.4452,
+      "step": 422
+    },
+    {
+      "epoch": 1.95301431256325,
+      "grad_norm": 0.15668661892414093,
+      "learning_rate": 3.4876543209876544e-06,
+      "loss": 9.1397,
+      "step": 423
+    },
+    {
+      "epoch": 1.9576405956339453,
+      "grad_norm": 0.129827082157135,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 8.428,
+      "step": 424
+    },
+    {
+      "epoch": 1.9622668787046407,
+      "grad_norm": 0.14903448522090912,
+      "learning_rate": 3.4567901234567904e-06,
+      "loss": 9.0113,
+      "step": 425
+    },
+    {
+      "epoch": 1.966893161775336,
+      "grad_norm": 0.2033424973487854,
+      "learning_rate": 3.441358024691358e-06,
+      "loss": 8.6822,
+      "step": 426
+    },
+    {
+      "epoch": 1.9715194448460314,
+      "grad_norm": 0.1724868267774582,
+      "learning_rate": 3.4259259259259265e-06,
+      "loss": 9.3924,
+      "step": 427
+    },
+    {
+      "epoch": 1.976145727916727,
+      "grad_norm": 0.1788475513458252,
+      "learning_rate": 3.410493827160494e-06,
+      "loss": 7.8849,
+      "step": 428
+    },
+    {
+      "epoch": 1.9807720109874223,
+      "grad_norm": 0.13475210964679718,
+      "learning_rate": 3.395061728395062e-06,
+      "loss": 10.321,
+      "step": 429
+    },
+    {
+      "epoch": 1.9853982940581176,
+      "grad_norm": 0.15269149839878082,
+      "learning_rate": 3.37962962962963e-06,
+      "loss": 9.5632,
+      "step": 430
+    },
+    {
+      "epoch": 1.9900245771288132,
+      "grad_norm": 0.20584136247634888,
+      "learning_rate": 3.3641975308641977e-06,
+      "loss": 9.0406,
+      "step": 431
+    },
+    {
+      "epoch": 1.9946508601995085,
+      "grad_norm": 0.1267031729221344,
+      "learning_rate": 3.3487654320987657e-06,
+      "loss": 9.1269,
+      "step": 432
+    },
+    {
+      "epoch": 1.999277143270204,
+      "grad_norm": 0.20229946076869965,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 9.1644,
+      "step": 433
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.20229946076869965,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6753,
+      "step": 434
+    },
+    {
+      "epoch": 2.0046262830706953,
+      "grad_norm": 0.19674813747406006,
+      "learning_rate": 3.3179012345679013e-06,
+      "loss": 8.9472,
+      "step": 435
+    },
+    {
+      "epoch": 2.0092525661413907,
+      "grad_norm": 0.17176903784275055,
+      "learning_rate": 3.30246913580247e-06,
+      "loss": 8.1618,
+      "step": 436
+    },
+    {
+      "epoch": 2.013878849212086,
+      "grad_norm": 0.21751543879508972,
+      "learning_rate": 3.2870370370370374e-06,
+      "loss": 11.298,
+      "step": 437
+    },
+    {
+      "epoch": 2.0185051322827814,
+      "grad_norm": 0.15177738666534424,
+      "learning_rate": 3.2716049382716054e-06,
+      "loss": 8.9182,
+      "step": 438
+    },
+    {
+      "epoch": 2.0231314153534767,
+      "grad_norm": 0.13782791793346405,
+      "learning_rate": 3.256172839506173e-06,
+      "loss": 9.138,
+      "step": 439
+    },
+    {
+      "epoch": 2.0277576984241725,
+      "grad_norm": 0.1871718168258667,
+      "learning_rate": 3.240740740740741e-06,
+      "loss": 9.3935,
+      "step": 440
+    },
+    {
+      "epoch": 2.032383981494868,
+      "grad_norm": 0.13550569117069244,
+      "learning_rate": 3.2253086419753086e-06,
+      "loss": 9.4065,
+      "step": 441
+    },
+    {
+      "epoch": 2.0370102645655632,
+      "grad_norm": 0.1414715200662613,
+      "learning_rate": 3.2098765432098767e-06,
+      "loss": 10.0772,
+      "step": 442
+    },
+    {
+      "epoch": 2.0416365476362586,
+      "grad_norm": 0.16839131712913513,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 8.4968,
+      "step": 443
+    },
+    {
+      "epoch": 2.046262830706954,
+      "grad_norm": 0.181600421667099,
+      "learning_rate": 3.1790123456790127e-06,
+      "loss": 8.0076,
+      "step": 444
+    },
+    {
+      "epoch": 2.0508891137776493,
+      "grad_norm": 0.1106899231672287,
+      "learning_rate": 3.1635802469135807e-06,
+      "loss": 8.5825,
+      "step": 445
+    },
+    {
+      "epoch": 2.0555153968483446,
+      "grad_norm": 0.17763055860996246,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 9.2981,
+      "step": 446
+    },
+    {
+      "epoch": 2.06014167991904,
+      "grad_norm": 0.19595400989055634,
+      "learning_rate": 3.1327160493827164e-06,
+      "loss": 10.5243,
+      "step": 447
+    },
+    {
+      "epoch": 2.0647679629897353,
+      "grad_norm": 0.18254916369915009,
+      "learning_rate": 3.1172839506172844e-06,
+      "loss": 10.9755,
+      "step": 448
+    },
+    {
+      "epoch": 2.0693942460604307,
+      "grad_norm": 0.18021439015865326,
+      "learning_rate": 3.101851851851852e-06,
+      "loss": 9.3634,
+      "step": 449
+    },
+    {
+      "epoch": 2.074020529131126,
+      "grad_norm": 0.20369423925876617,
+      "learning_rate": 3.08641975308642e-06,
+      "loss": 10.1676,
+      "step": 450
+    },
+    {
+      "epoch": 2.078646812201822,
+      "grad_norm": 0.1611659973859787,
+      "learning_rate": 3.0709876543209876e-06,
+      "loss": 10.5072,
+      "step": 451
+    },
+    {
+      "epoch": 2.083273095272517,
+      "grad_norm": 0.16623584926128387,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 8.6923,
+      "step": 452
+    },
+    {
+      "epoch": 2.0878993783432125,
+      "grad_norm": 0.20521709322929382,
+      "learning_rate": 3.040123456790124e-06,
+      "loss": 10.3398,
+      "step": 453
+    },
+    {
+      "epoch": 2.092525661413908,
+      "grad_norm": 0.150822713971138,
+      "learning_rate": 3.0246913580246917e-06,
+      "loss": 8.1734,
+      "step": 454
+    },
+    {
+      "epoch": 2.097151944484603,
+      "grad_norm": 0.14610332250595093,
+      "learning_rate": 3.0092592592592597e-06,
+      "loss": 8.0033,
+      "step": 455
+    },
+    {
+      "epoch": 2.1017782275552985,
+      "grad_norm": 0.16596634685993195,
+      "learning_rate": 2.9938271604938273e-06,
+      "loss": 9.923,
+      "step": 456
+    },
+    {
+      "epoch": 2.106404510625994,
+      "grad_norm": 0.15604960918426514,
+      "learning_rate": 2.9783950617283953e-06,
+      "loss": 9.5128,
+      "step": 457
+    },
+    {
+      "epoch": 2.1110307936966892,
+      "grad_norm": 0.1488286554813385,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 8.704,
+      "step": 458
+    },
+    {
+      "epoch": 2.1156570767673846,
+      "grad_norm": 0.1483355164527893,
+      "learning_rate": 2.947530864197531e-06,
+      "loss": 8.4275,
+      "step": 459
+    },
+    {
+      "epoch": 2.12028335983808,
+      "grad_norm": 0.2545453906059265,
+      "learning_rate": 2.9320987654320994e-06,
+      "loss": 9.332,
+      "step": 460
+    },
+    {
+      "epoch": 2.1249096429087757,
+      "grad_norm": 0.18272359669208527,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 9.4908,
+      "step": 461
+    },
+    {
+      "epoch": 2.129535925979471,
+      "grad_norm": 0.19491931796073914,
+      "learning_rate": 2.901234567901235e-06,
+      "loss": 8.5668,
+      "step": 462
+    },
+    {
+      "epoch": 2.1341622090501664,
+      "grad_norm": 0.171804741024971,
+      "learning_rate": 2.8858024691358026e-06,
+      "loss": 8.9085,
+      "step": 463
+    },
+    {
+      "epoch": 2.1387884921208618,
+      "grad_norm": 0.22477325797080994,
+      "learning_rate": 2.8703703703703706e-06,
+      "loss": 9.5643,
+      "step": 464
+    },
+    {
+      "epoch": 2.143414775191557,
+      "grad_norm": 0.2512359023094177,
+      "learning_rate": 2.8549382716049382e-06,
+      "loss": 10.6236,
+      "step": 465
+    },
+    {
+      "epoch": 2.1480410582622524,
+      "grad_norm": 0.15513500571250916,
+      "learning_rate": 2.8395061728395062e-06,
+      "loss": 9.3385,
+      "step": 466
+    },
+    {
+      "epoch": 2.152667341332948,
+      "grad_norm": 0.18065772950649261,
+      "learning_rate": 2.8240740740740743e-06,
+      "loss": 9.3942,
+      "step": 467
+    },
+    {
+      "epoch": 2.157293624403643,
+      "grad_norm": 0.19600172340869904,
+      "learning_rate": 2.8086419753086423e-06,
+      "loss": 9.0732,
+      "step": 468
+    },
+    {
+      "epoch": 2.1619199074743385,
+      "grad_norm": 0.17671746015548706,
+      "learning_rate": 2.7932098765432103e-06,
+      "loss": 10.023,
+      "step": 469
+    },
+    {
+      "epoch": 2.166546190545034,
+      "grad_norm": 0.16406480967998505,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 9.8105,
+      "step": 470
+    },
+    {
+      "epoch": 2.171172473615729,
+      "grad_norm": 0.11806418746709824,
+      "learning_rate": 2.762345679012346e-06,
+      "loss": 8.2566,
+      "step": 471
+    },
+    {
+      "epoch": 2.175798756686425,
+      "grad_norm": 0.32428812980651855,
+      "learning_rate": 2.746913580246914e-06,
+      "loss": 9.1191,
+      "step": 472
+    },
+    {
+      "epoch": 2.1804250397571203,
+      "grad_norm": 0.22312945127487183,
+      "learning_rate": 2.7314814814814816e-06,
+      "loss": 9.9265,
+      "step": 473
+    },
+    {
+      "epoch": 2.1850513228278157,
+      "grad_norm": 0.32287222146987915,
+      "learning_rate": 2.7160493827160496e-06,
+      "loss": 10.6421,
+      "step": 474
+    },
+    {
+      "epoch": 2.189677605898511,
+      "grad_norm": 0.15497097373008728,
+      "learning_rate": 2.700617283950617e-06,
+      "loss": 8.9862,
+      "step": 475
+    },
+    {
+      "epoch": 2.1943038889692064,
+      "grad_norm": 0.13556216657161713,
+      "learning_rate": 2.6851851851851856e-06,
+      "loss": 8.658,
+      "step": 476
+    },
+    {
+      "epoch": 2.1989301720399017,
+      "grad_norm": 0.17666535079479218,
+      "learning_rate": 2.6697530864197536e-06,
+      "loss": 10.2978,
+      "step": 477
+    },
+    {
+      "epoch": 2.203556455110597,
+      "grad_norm": 0.16342324018478394,
+      "learning_rate": 2.6543209876543212e-06,
+      "loss": 8.4887,
+      "step": 478
+    },
+    {
+      "epoch": 2.2081827381812924,
+      "grad_norm": 0.2034367471933365,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 7.8828,
+      "step": 479
+    },
+    {
+      "epoch": 2.2128090212519878,
+      "grad_norm": 0.23170992732048035,
+      "learning_rate": 2.623456790123457e-06,
+      "loss": 9.6533,
+      "step": 480
+    },
+    {
+      "epoch": 2.217435304322683,
+      "grad_norm": 0.16840022802352905,
+      "learning_rate": 2.608024691358025e-06,
+      "loss": 8.1141,
+      "step": 481
+    },
+    {
+      "epoch": 2.2220615873933784,
+      "grad_norm": 0.1619078814983368,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 8.6294,
+      "step": 482
+    },
+    {
+      "epoch": 2.2266878704640742,
+      "grad_norm": 0.22204962372779846,
+      "learning_rate": 2.5771604938271605e-06,
+      "loss": 10.4391,
+      "step": 483
+    },
+    {
+      "epoch": 2.2313141535347696,
+      "grad_norm": 0.1876753270626068,
+      "learning_rate": 2.561728395061729e-06,
+      "loss": 9.9287,
+      "step": 484
+    },
+    {
+      "epoch": 2.235940436605465,
+      "grad_norm": 0.18601296842098236,
+      "learning_rate": 2.5462962962962966e-06,
+      "loss": 9.3439,
+      "step": 485
+    },
+    {
+      "epoch": 2.2405667196761603,
+      "grad_norm": 0.17974725365638733,
+      "learning_rate": 2.5308641975308646e-06,
+      "loss": 9.6812,
+      "step": 486
+    },
+    {
+      "epoch": 2.2451930027468556,
+      "grad_norm": 0.1832571029663086,
+      "learning_rate": 2.515432098765432e-06,
+      "loss": 7.6883,
+      "step": 487
+    },
+    {
+      "epoch": 2.249819285817551,
+      "grad_norm": 0.18652378022670746,
+      "learning_rate": 2.5e-06,
+      "loss": 8.9935,
+      "step": 488
+    },
+    {
+      "epoch": 2.2544455688882463,
+      "grad_norm": 0.20331954956054688,
+      "learning_rate": 2.4845679012345682e-06,
+      "loss": 9.1709,
+      "step": 489
+    },
+    {
+      "epoch": 2.2590718519589417,
+      "grad_norm": 0.15439340472221375,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 9.0578,
+      "step": 490
+    },
+    {
+      "epoch": 2.263698135029637,
+      "grad_norm": 0.16174978017807007,
+      "learning_rate": 2.453703703703704e-06,
+      "loss": 9.2311,
+      "step": 491
+    },
+    {
+      "epoch": 2.2683244181003324,
+      "grad_norm": 0.20344924926757812,
+      "learning_rate": 2.438271604938272e-06,
+      "loss": 9.8727,
+      "step": 492
+    },
+    {
+      "epoch": 2.2729507011710277,
+      "grad_norm": 0.19355060160160065,
+      "learning_rate": 2.4228395061728395e-06,
+      "loss": 9.3433,
+      "step": 493
+    },
+    {
+      "epoch": 2.2775769842417235,
+      "grad_norm": 0.17609727382659912,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 8.8289,
+      "step": 494
+    },
+    {
+      "epoch": 2.282203267312419,
+      "grad_norm": 0.16068409383296967,
+      "learning_rate": 2.3919753086419755e-06,
+      "loss": 9.1044,
+      "step": 495
+    },
+    {
+      "epoch": 2.286829550383114,
+      "grad_norm": 0.15089063346385956,
+      "learning_rate": 2.3765432098765435e-06,
+      "loss": 8.8363,
+      "step": 496
+    },
+    {
+      "epoch": 2.2914558334538095,
+      "grad_norm": 0.3163600265979767,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 9.3244,
+      "step": 497
+    },
+    {
+      "epoch": 2.296082116524505,
+      "grad_norm": 0.18847279250621796,
+      "learning_rate": 2.345679012345679e-06,
+      "loss": 9.1338,
+      "step": 498
+    },
+    {
+      "epoch": 2.3007083995952002,
+      "grad_norm": 0.23115108907222748,
+      "learning_rate": 2.330246913580247e-06,
+      "loss": 9.6425,
+      "step": 499
+    },
+    {
+      "epoch": 2.3053346826658956,
+      "grad_norm": 0.15251374244689941,
+      "learning_rate": 2.314814814814815e-06,
+      "loss": 9.6591,
+      "step": 500
+    },
+    {
+      "epoch": 2.309960965736591,
+      "grad_norm": 0.1706661432981491,
+      "learning_rate": 2.299382716049383e-06,
+      "loss": 10.3708,
+      "step": 501
+    },
+    {
+      "epoch": 2.3145872488072863,
+      "grad_norm": 0.16492144763469696,
+      "learning_rate": 2.283950617283951e-06,
+      "loss": 8.8977,
+      "step": 502
+    },
+    {
+      "epoch": 2.3192135318779816,
+      "grad_norm": 0.15373440086841583,
+      "learning_rate": 2.268518518518519e-06,
+      "loss": 10.9027,
+      "step": 503
+    },
+    {
+      "epoch": 2.323839814948677,
+      "grad_norm": 0.18270155787467957,
+      "learning_rate": 2.2530864197530865e-06,
+      "loss": 9.8633,
+      "step": 504
+    },
+    {
+      "epoch": 2.3284660980193728,
+      "grad_norm": 0.22537460923194885,
+      "learning_rate": 2.2376543209876545e-06,
+      "loss": 8.6262,
+      "step": 505
+    },
+    {
+      "epoch": 2.333092381090068,
+      "grad_norm": 0.1788664162158966,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 8.3806,
+      "step": 506
+    },
+    {
+      "epoch": 2.3377186641607635,
+      "grad_norm": 0.2230851948261261,
+      "learning_rate": 2.2067901234567905e-06,
+      "loss": 9.6656,
+      "step": 507
+    },
+    {
+      "epoch": 2.342344947231459,
+      "grad_norm": 0.157254159450531,
+      "learning_rate": 2.191358024691358e-06,
+      "loss": 9.0586,
+      "step": 508
+    },
+    {
+      "epoch": 2.346971230302154,
+      "grad_norm": 0.21109546720981598,
+      "learning_rate": 2.175925925925926e-06,
+      "loss": 10.4577,
+      "step": 509
+    },
+    {
+      "epoch": 2.3515975133728495,
+      "grad_norm": 0.16909867525100708,
+      "learning_rate": 2.1604938271604937e-06,
+      "loss": 9.5222,
+      "step": 510
+    },
+    {
+      "epoch": 2.356223796443545,
+      "grad_norm": 0.1652510166168213,
+      "learning_rate": 2.145061728395062e-06,
+      "loss": 9.5384,
+      "step": 511
+    },
+    {
+      "epoch": 2.36085007951424,
+      "grad_norm": 0.14154984056949615,
+      "learning_rate": 2.1296296296296298e-06,
+      "loss": 9.1477,
+      "step": 512
+    },
+    {
+      "epoch": 2.3654763625849355,
+      "grad_norm": 0.18620611727237701,
+      "learning_rate": 2.114197530864198e-06,
+      "loss": 9.8372,
+      "step": 513
+    },
+    {
+      "epoch": 2.370102645655631,
+      "grad_norm": 0.17792826890945435,
+      "learning_rate": 2.0987654320987654e-06,
+      "loss": 9.4475,
+      "step": 514
+    },
+    {
+      "epoch": 2.3747289287263262,
+      "grad_norm": 0.17278362810611725,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 8.2241,
+      "step": 515
+    },
+    {
+      "epoch": 2.379355211797022,
+      "grad_norm": 0.17132383584976196,
+      "learning_rate": 2.0679012345679015e-06,
+      "loss": 10.9298,
+      "step": 516
+    },
+    {
+      "epoch": 2.3839814948677174,
+      "grad_norm": 0.15168847143650055,
+      "learning_rate": 2.052469135802469e-06,
+      "loss": 9.0072,
+      "step": 517
+    },
+    {
+      "epoch": 2.3886077779384127,
+      "grad_norm": 0.186196431517601,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 9.0923,
+      "step": 518
+    },
+    {
+      "epoch": 2.393234061009108,
+      "grad_norm": 0.12113353610038757,
+      "learning_rate": 2.021604938271605e-06,
+      "loss": 10.696,
+      "step": 519
+    },
+    {
+      "epoch": 2.3978603440798034,
+      "grad_norm": 0.25200334191322327,
+      "learning_rate": 2.006172839506173e-06,
+      "loss": 8.5775,
+      "step": 520
+    },
+    {
+      "epoch": 2.4024866271504988,
+      "grad_norm": 0.1717706322669983,
+      "learning_rate": 1.9907407407407407e-06,
+      "loss": 9.8223,
+      "step": 521
+    },
+    {
+      "epoch": 2.407112910221194,
+      "grad_norm": 0.21262796223163605,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 9.1006,
+      "step": 522
+    },
+    {
+      "epoch": 2.4117391932918895,
+      "grad_norm": 0.15240874886512756,
+      "learning_rate": 1.9598765432098768e-06,
+      "loss": 8.8133,
+      "step": 523
+    },
+    {
+      "epoch": 2.416365476362585,
+      "grad_norm": 0.20154468715190887,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 8.9349,
+      "step": 524
+    },
+    {
+      "epoch": 2.42099175943328,
+      "grad_norm": 0.2465580701828003,
+      "learning_rate": 1.9290123456790124e-06,
+      "loss": 9.7463,
+      "step": 525
+    },
+    {
+      "epoch": 2.4256180425039755,
+      "grad_norm": 0.262588769197464,
+      "learning_rate": 1.9135802469135804e-06,
+      "loss": 8.9882,
+      "step": 526
+    },
+    {
+      "epoch": 2.4302443255746713,
+      "grad_norm": 0.1886514276266098,
+      "learning_rate": 1.8981481481481484e-06,
+      "loss": 10.1046,
+      "step": 527
+    },
+    {
+      "epoch": 2.4348706086453666,
+      "grad_norm": 0.15457630157470703,
+      "learning_rate": 1.8827160493827162e-06,
+      "loss": 9.1876,
+      "step": 528
+    },
+    {
+      "epoch": 2.439496891716062,
+      "grad_norm": 0.18536067008972168,
+      "learning_rate": 1.867283950617284e-06,
+      "loss": 8.9485,
+      "step": 529
+    },
+    {
+      "epoch": 2.4441231747867573,
+      "grad_norm": 0.16247576475143433,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 9.7982,
+      "step": 530
+    },
+    {
+      "epoch": 2.4487494578574527,
+      "grad_norm": 0.13770771026611328,
+      "learning_rate": 1.83641975308642e-06,
+      "loss": 9.9136,
+      "step": 531
+    },
+    {
+      "epoch": 2.453375740928148,
+      "grad_norm": 0.1654641479253769,
+      "learning_rate": 1.820987654320988e-06,
+      "loss": 9.1883,
+      "step": 532
+    },
+    {
+      "epoch": 2.4580020239988434,
+      "grad_norm": 0.17362841963768005,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 9.0925,
+      "step": 533
+    },
+    {
+      "epoch": 2.4626283070695387,
+      "grad_norm": 0.15451891720294952,
+      "learning_rate": 1.7901234567901235e-06,
+      "loss": 9.477,
+      "step": 534
+    },
+    {
+      "epoch": 2.467254590140234,
+      "grad_norm": 0.14848832786083221,
+      "learning_rate": 1.7746913580246916e-06,
+      "loss": 10.4986,
+      "step": 535
+    },
+    {
+      "epoch": 2.4718808732109294,
+      "grad_norm": 0.18573352694511414,
+      "learning_rate": 1.7592592592592594e-06,
+      "loss": 9.2772,
+      "step": 536
+    },
+    {
+      "epoch": 2.4765071562816248,
+      "grad_norm": 0.16295011341571808,
+      "learning_rate": 1.7438271604938272e-06,
+      "loss": 10.3985,
+      "step": 537
+    },
+    {
+      "epoch": 2.4811334393523206,
+      "grad_norm": 0.1890224814414978,
+      "learning_rate": 1.7283950617283952e-06,
+      "loss": 8.1752,
+      "step": 538
+    },
+    {
+      "epoch": 2.485759722423016,
+      "grad_norm": 0.1372375637292862,
+      "learning_rate": 1.7129629629629632e-06,
+      "loss": 8.903,
+      "step": 539
+    },
+    {
+      "epoch": 2.4903860054937113,
+      "grad_norm": 0.17448656260967255,
+      "learning_rate": 1.697530864197531e-06,
+      "loss": 8.5784,
+      "step": 540
+    },
+    {
+      "epoch": 2.4950122885644066,
+      "grad_norm": 0.1701819896697998,
+      "learning_rate": 1.6820987654320989e-06,
+      "loss": 8.7201,
+      "step": 541
+    },
+    {
+      "epoch": 2.499638571635102,
+      "grad_norm": 0.14620746672153473,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 9.211,
+      "step": 542
+    },
+    {
+      "epoch": 2.5042648547057973,
+      "grad_norm": 0.16162265837192535,
+      "learning_rate": 1.651234567901235e-06,
+      "loss": 9.4963,
+      "step": 543
+    },
+    {
+      "epoch": 2.5088911377764926,
+      "grad_norm": 0.15886524319648743,
+      "learning_rate": 1.6358024691358027e-06,
+      "loss": 8.8424,
+      "step": 544
+    },
+    {
+      "epoch": 2.513517420847188,
+      "grad_norm": 0.14556364715099335,
+      "learning_rate": 1.6203703703703705e-06,
+      "loss": 7.9769,
+      "step": 545
+    },
+    {
+      "epoch": 2.5181437039178833,
+      "grad_norm": 0.21578781306743622,
+      "learning_rate": 1.6049382716049383e-06,
+      "loss": 8.9909,
+      "step": 546
+    },
+    {
+      "epoch": 2.522769986988579,
+      "grad_norm": 0.18850085139274597,
+      "learning_rate": 1.5895061728395064e-06,
+      "loss": 9.8237,
+      "step": 547
+    },
+    {
+      "epoch": 2.527396270059274,
+      "grad_norm": 0.1415654569864273,
+      "learning_rate": 1.5740740740740742e-06,
+      "loss": 8.9615,
+      "step": 548
+    },
+    {
+      "epoch": 2.53202255312997,
+      "grad_norm": 0.17242810130119324,
+      "learning_rate": 1.5586419753086422e-06,
+      "loss": 8.9936,
+      "step": 549
+    },
+    {
+      "epoch": 2.536648836200665,
+      "grad_norm": 0.20581774413585663,
+      "learning_rate": 1.54320987654321e-06,
+      "loss": 9.376,
+      "step": 550
+    },
+    {
+      "epoch": 2.5412751192713605,
+      "grad_norm": 0.17876863479614258,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 8.9055,
+      "step": 551
+    },
+    {
+      "epoch": 2.545901402342056,
+      "grad_norm": 0.1470121592283249,
+      "learning_rate": 1.5123456790123458e-06,
+      "loss": 9.3424,
+      "step": 552
+    },
+    {
+      "epoch": 2.550527685412751,
+      "grad_norm": 0.23757006227970123,
+      "learning_rate": 1.4969135802469136e-06,
+      "loss": 9.7866,
+      "step": 553
+    },
+    {
+      "epoch": 2.5551539684834466,
+      "grad_norm": 0.16553503274917603,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 8.6868,
+      "step": 554
+    },
+    {
+      "epoch": 2.559780251554142,
+      "grad_norm": 0.12771116197109222,
+      "learning_rate": 1.4660493827160497e-06,
+      "loss": 8.9301,
+      "step": 555
+    },
+    {
+      "epoch": 2.5644065346248373,
+      "grad_norm": 0.35117506980895996,
+      "learning_rate": 1.4506172839506175e-06,
+      "loss": 9.3037,
+      "step": 556
+    },
+    {
+      "epoch": 2.5690328176955326,
+      "grad_norm": 0.1245264932513237,
+      "learning_rate": 1.4351851851851853e-06,
+      "loss": 9.0822,
+      "step": 557
+    },
+    {
+      "epoch": 2.5736591007662284,
+      "grad_norm": 0.18772459030151367,
+      "learning_rate": 1.4197530864197531e-06,
+      "loss": 9.146,
+      "step": 558
+    },
+    {
+      "epoch": 2.5782853838369233,
+      "grad_norm": 0.18729011714458466,
+      "learning_rate": 1.4043209876543211e-06,
+      "loss": 8.183,
+      "step": 559
+    },
+    {
+      "epoch": 2.582911666907619,
+      "grad_norm": 0.12940698862075806,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 9.4745,
+      "step": 560
+    },
+    {
+      "epoch": 2.5875379499783144,
+      "grad_norm": 0.16963091492652893,
+      "learning_rate": 1.373456790123457e-06,
+      "loss": 8.8193,
+      "step": 561
+    },
+    {
+      "epoch": 2.59216423304901,
+      "grad_norm": 0.17672613263130188,
+      "learning_rate": 1.3580246913580248e-06,
+      "loss": 8.2093,
+      "step": 562
+    },
+    {
+      "epoch": 2.596790516119705,
+      "grad_norm": 0.24918967485427856,
+      "learning_rate": 1.3425925925925928e-06,
+      "loss": 9.2257,
+      "step": 563
+    },
+    {
+      "epoch": 2.6014167991904005,
+      "grad_norm": 0.15989799797534943,
+      "learning_rate": 1.3271604938271606e-06,
+      "loss": 9.08,
+      "step": 564
+    },
+    {
+      "epoch": 2.606043082261096,
+      "grad_norm": 0.14997775852680206,
+      "learning_rate": 1.3117283950617284e-06,
+      "loss": 9.8745,
+      "step": 565
+    },
+    {
+      "epoch": 2.610669365331791,
+      "grad_norm": 0.17529721558094025,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 8.8959,
+      "step": 566
+    },
+    {
+      "epoch": 2.6152956484024865,
+      "grad_norm": 0.16525782644748688,
+      "learning_rate": 1.2808641975308645e-06,
+      "loss": 8.2331,
+      "step": 567
+    },
+    {
+      "epoch": 2.619921931473182,
+      "grad_norm": 0.14114739000797272,
+      "learning_rate": 1.2654320987654323e-06,
+      "loss": 8.1592,
+      "step": 568
+    },
+    {
+      "epoch": 2.6245482145438777,
+      "grad_norm": 0.21027730405330658,
+      "learning_rate": 1.25e-06,
+      "loss": 9.1182,
+      "step": 569
+    },
+    {
+      "epoch": 2.6291744976145726,
+      "grad_norm": 0.13383808732032776,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 8.8642,
+      "step": 570
+    },
+    {
+      "epoch": 2.6338007806852684,
+      "grad_norm": 0.15454170107841492,
+      "learning_rate": 1.219135802469136e-06,
+      "loss": 7.1809,
+      "step": 571
+    },
+    {
+      "epoch": 2.6384270637559637,
+      "grad_norm": 0.1575554460287094,
+      "learning_rate": 1.2037037037037037e-06,
+      "loss": 8.5097,
+      "step": 572
+    },
+    {
+      "epoch": 2.643053346826659,
+      "grad_norm": 0.16009144484996796,
+      "learning_rate": 1.1882716049382718e-06,
+      "loss": 8.8008,
+      "step": 573
+    },
+    {
+      "epoch": 2.6476796298973544,
+      "grad_norm": 0.18514862656593323,
+      "learning_rate": 1.1728395061728396e-06,
+      "loss": 8.3351,
+      "step": 574
+    },
+    {
+      "epoch": 2.6523059129680497,
+      "grad_norm": 0.16186843812465668,
+      "learning_rate": 1.1574074074074076e-06,
+      "loss": 8.702,
+      "step": 575
+    },
+    {
+      "epoch": 2.656932196038745,
+      "grad_norm": 0.19223704934120178,
+      "learning_rate": 1.1419753086419754e-06,
+      "loss": 9.1931,
+      "step": 576
+    },
+    {
+      "epoch": 2.6615584791094404,
+      "grad_norm": 0.1866094172000885,
+      "learning_rate": 1.1265432098765432e-06,
+      "loss": 9.5196,
+      "step": 577
+    },
+    {
+      "epoch": 2.666184762180136,
+      "grad_norm": 0.2299501746892929,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 9.4269,
+      "step": 578
+    },
+    {
+      "epoch": 2.670811045250831,
+      "grad_norm": 0.19736136496067047,
+      "learning_rate": 1.095679012345679e-06,
+      "loss": 9.3998,
+      "step": 579
+    },
+    {
+      "epoch": 2.675437328321527,
+      "grad_norm": 0.16987043619155884,
+      "learning_rate": 1.0802469135802469e-06,
+      "loss": 8.2513,
+      "step": 580
+    },
+    {
+      "epoch": 2.680063611392222,
+      "grad_norm": 0.18940360844135284,
+      "learning_rate": 1.0648148148148149e-06,
+      "loss": 7.8771,
+      "step": 581
+    },
+    {
+      "epoch": 2.6846898944629176,
+      "grad_norm": 0.16819283366203308,
+      "learning_rate": 1.0493827160493827e-06,
+      "loss": 8.9367,
+      "step": 582
+    },
+    {
+      "epoch": 2.689316177533613,
+      "grad_norm": 0.14170341193675995,
+      "learning_rate": 1.0339506172839507e-06,
+      "loss": 7.1555,
+      "step": 583
+    },
+    {
+      "epoch": 2.6939424606043083,
+      "grad_norm": 0.16584675014019012,
+      "learning_rate": 1.0185185185185185e-06,
+      "loss": 9.7405,
+      "step": 584
+    },
+    {
+      "epoch": 2.6985687436750037,
+      "grad_norm": 0.15231122076511383,
+      "learning_rate": 1.0030864197530866e-06,
+      "loss": 9.8814,
+      "step": 585
+    },
+    {
+      "epoch": 2.703195026745699,
+      "grad_norm": 0.14071619510650635,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 8.0466,
+      "step": 586
+    },
+    {
+      "epoch": 2.7078213098163944,
+      "grad_norm": 0.21130667626857758,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 10.1551,
+      "step": 587
+    },
+    {
+      "epoch": 2.7124475928870897,
+      "grad_norm": 0.14736154675483704,
+      "learning_rate": 9.567901234567902e-07,
+      "loss": 7.5648,
+      "step": 588
+    },
+    {
+      "epoch": 2.717073875957785,
+      "grad_norm": 0.18979448080062866,
+      "learning_rate": 9.413580246913581e-07,
+      "loss": 9.5863,
+      "step": 589
+    },
+    {
+      "epoch": 2.7217001590284804,
+      "grad_norm": 0.14803054928779602,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 9.2012,
+      "step": 590
+    },
+    {
+      "epoch": 2.726326442099176,
+      "grad_norm": 0.24644902348518372,
+      "learning_rate": 9.10493827160494e-07,
+      "loss": 7.9091,
+      "step": 591
+    },
+    {
+      "epoch": 2.730952725169871,
+      "grad_norm": 0.18411195278167725,
+      "learning_rate": 8.950617283950618e-07,
+      "loss": 9.0894,
+      "step": 592
+    },
+    {
+      "epoch": 2.735579008240567,
+      "grad_norm": 0.21505457162857056,
+      "learning_rate": 8.796296296296297e-07,
+      "loss": 7.4988,
+      "step": 593
+    },
+    {
+      "epoch": 2.7402052913112622,
+      "grad_norm": 0.18174538016319275,
+      "learning_rate": 8.641975308641976e-07,
+      "loss": 8.2626,
+      "step": 594
+    },
+    {
+      "epoch": 2.7448315743819576,
+      "grad_norm": 0.1982118785381317,
+      "learning_rate": 8.487654320987655e-07,
+      "loss": 8.415,
+      "step": 595
+    },
+    {
+      "epoch": 2.749457857452653,
+      "grad_norm": 0.18941333889961243,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 8.931,
+      "step": 596
+    },
+    {
+      "epoch": 2.7540841405233483,
+      "grad_norm": 0.2093527615070343,
+      "learning_rate": 8.179012345679014e-07,
+      "loss": 9.6871,
+      "step": 597
+    },
+    {
+      "epoch": 2.7587104235940436,
+      "grad_norm": 0.25371572375297546,
+      "learning_rate": 8.024691358024692e-07,
+      "loss": 9.719,
+      "step": 598
+    },
+    {
+      "epoch": 2.763336706664739,
+      "grad_norm": 0.13767634332180023,
+      "learning_rate": 7.870370370370371e-07,
+      "loss": 7.0931,
+      "step": 599
+    },
+    {
+      "epoch": 2.7679629897354343,
+      "grad_norm": 0.18252308666706085,
+      "learning_rate": 7.71604938271605e-07,
+      "loss": 9.0813,
+      "step": 600
+    },
+    {
+      "epoch": 2.7725892728061297,
+      "grad_norm": 0.13574370741844177,
+      "learning_rate": 7.561728395061729e-07,
+      "loss": 8.8218,
+      "step": 601
+    },
+    {
+      "epoch": 2.7772155558768254,
+      "grad_norm": 0.17202846705913544,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 8.954,
+      "step": 602
+    },
+    {
+      "epoch": 2.7818418389475204,
+      "grad_norm": 0.18582294881343842,
+      "learning_rate": 7.253086419753087e-07,
+      "loss": 7.7078,
+      "step": 603
+    },
+    {
+      "epoch": 2.786468122018216,
+      "grad_norm": 0.27104949951171875,
+      "learning_rate": 7.098765432098766e-07,
+      "loss": 8.3435,
+      "step": 604
+    },
+    {
+      "epoch": 2.7910944050889115,
+      "grad_norm": 0.32489752769470215,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 8.5035,
+      "step": 605
+    },
+    {
+      "epoch": 2.795720688159607,
+      "grad_norm": 0.1454378217458725,
+      "learning_rate": 6.790123456790124e-07,
+      "loss": 8.4067,
+      "step": 606
+    },
+    {
+      "epoch": 2.800346971230302,
+      "grad_norm": 0.20940132439136505,
+      "learning_rate": 6.635802469135803e-07,
+      "loss": 9.4978,
+      "step": 607
+    },
+    {
+      "epoch": 2.8049732543009975,
+      "grad_norm": 0.17158959805965424,
+      "learning_rate": 6.481481481481481e-07,
+      "loss": 9.3416,
+      "step": 608
+    },
+    {
+      "epoch": 2.809599537371693,
+      "grad_norm": 0.17591050267219543,
+      "learning_rate": 6.327160493827161e-07,
+      "loss": 8.7415,
+      "step": 609
+    },
+    {
+      "epoch": 2.8142258204423882,
+      "grad_norm": 0.18380632996559143,
+      "learning_rate": 6.17283950617284e-07,
+      "loss": 8.2376,
+      "step": 610
+    },
+    {
+      "epoch": 2.8188521035130836,
+      "grad_norm": 0.19443170726299286,
+      "learning_rate": 6.018518518518519e-07,
+      "loss": 7.562,
+      "step": 611
+    },
+    {
+      "epoch": 2.823478386583779,
+      "grad_norm": 0.24004080891609192,
+      "learning_rate": 5.864197530864198e-07,
+      "loss": 9.346,
+      "step": 612
+    },
+    {
+      "epoch": 2.8281046696544747,
+      "grad_norm": 0.19434408843517303,
+      "learning_rate": 5.709876543209877e-07,
+      "loss": 6.8662,
+      "step": 613
+    },
+    {
+      "epoch": 2.8327309527251696,
+      "grad_norm": 0.1413394808769226,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 7.7949,
+      "step": 614
+    },
+    {
+      "epoch": 2.8373572357958654,
+      "grad_norm": 0.21421702206134796,
+      "learning_rate": 5.401234567901234e-07,
+      "loss": 8.6497,
+      "step": 615
+    },
+    {
+      "epoch": 2.8419835188665608,
+      "grad_norm": 0.18243171274662018,
+      "learning_rate": 5.246913580246914e-07,
+      "loss": 9.0258,
+      "step": 616
+    },
+    {
+      "epoch": 2.846609801937256,
+      "grad_norm": 0.15553410351276398,
+      "learning_rate": 5.092592592592593e-07,
+      "loss": 9.0613,
+      "step": 617
+    },
+    {
+      "epoch": 2.8512360850079514,
+      "grad_norm": 0.1925041526556015,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 8.8279,
+      "step": 618
+    },
+    {
+      "epoch": 2.855862368078647,
+      "grad_norm": 0.25465813279151917,
+      "learning_rate": 4.783950617283951e-07,
+      "loss": 8.5684,
+      "step": 619
+    },
+    {
+      "epoch": 2.860488651149342,
+      "grad_norm": 0.17943061888217926,
+      "learning_rate": 4.6296296296296297e-07,
+      "loss": 8.9017,
+      "step": 620
+    },
+    {
+      "epoch": 2.8651149342200375,
+      "grad_norm": 0.22506891191005707,
+      "learning_rate": 4.475308641975309e-07,
+      "loss": 8.7791,
+      "step": 621
+    },
+    {
+      "epoch": 2.869741217290733,
+      "grad_norm": 0.1924242228269577,
+      "learning_rate": 4.320987654320988e-07,
+      "loss": 8.4592,
+      "step": 622
+    },
+    {
+      "epoch": 2.874367500361428,
+      "grad_norm": 0.1336522102355957,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 9.2921,
+      "step": 623
+    },
+    {
+      "epoch": 2.878993783432124,
+      "grad_norm": 0.16631512343883514,
+      "learning_rate": 4.012345679012346e-07,
+      "loss": 8.7771,
+      "step": 624
+    },
+    {
+      "epoch": 2.883620066502819,
+      "grad_norm": 0.13000570237636566,
+      "learning_rate": 3.858024691358025e-07,
+      "loss": 7.6173,
+      "step": 625
+    },
+    {
+      "epoch": 2.8882463495735147,
+      "grad_norm": 0.16806651651859283,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 8.2835,
+      "step": 626
+    },
+    {
+      "epoch": 2.89287263264421,
+      "grad_norm": 0.16624124348163605,
+      "learning_rate": 3.549382716049383e-07,
+      "loss": 9.0295,
+      "step": 627
+    },
+    {
+      "epoch": 2.8974989157149054,
+      "grad_norm": 0.18300163745880127,
+      "learning_rate": 3.395061728395062e-07,
+      "loss": 9.3673,
+      "step": 628
+    },
+    {
+      "epoch": 2.9021251987856007,
+      "grad_norm": 0.1591711938381195,
+      "learning_rate": 3.2407407407407406e-07,
+      "loss": 9.2437,
+      "step": 629
+    },
+    {
+      "epoch": 2.906751481856296,
+      "grad_norm": 0.1935225874185562,
+      "learning_rate": 3.08641975308642e-07,
+      "loss": 8.1388,
+      "step": 630
+    },
+    {
+      "epoch": 2.9113777649269914,
+      "grad_norm": 0.18095123767852783,
+      "learning_rate": 2.932098765432099e-07,
+      "loss": 8.8998,
+      "step": 631
+    },
+    {
+      "epoch": 2.9160040479976868,
+      "grad_norm": 0.15141603350639343,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 9.5553,
+      "step": 632
+    },
+    {
+      "epoch": 2.920630331068382,
+      "grad_norm": 0.1664353758096695,
+      "learning_rate": 2.623456790123457e-07,
+      "loss": 9.2671,
+      "step": 633
+    },
+    {
+      "epoch": 2.9252566141390774,
+      "grad_norm": 0.14757901430130005,
+      "learning_rate": 2.469135802469136e-07,
+      "loss": 9.3753,
+      "step": 634
+    },
+    {
+      "epoch": 2.9298828972097732,
+      "grad_norm": 0.12832246720790863,
+      "learning_rate": 2.3148148148148148e-07,
+      "loss": 7.55,
+      "step": 635
+    },
+    {
+      "epoch": 2.934509180280468,
+      "grad_norm": 0.14796899259090424,
+      "learning_rate": 2.160493827160494e-07,
+      "loss": 8.9455,
+      "step": 636
+    },
+    {
+      "epoch": 2.939135463351164,
+      "grad_norm": 0.16375280916690826,
+      "learning_rate": 2.006172839506173e-07,
+      "loss": 9.1212,
+      "step": 637
+    },
+    {
+      "epoch": 2.9437617464218593,
+      "grad_norm": 0.15986521542072296,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 8.0966,
+      "step": 638
+    },
+    {
+      "epoch": 2.9483880294925546,
+      "grad_norm": 0.14000752568244934,
+      "learning_rate": 1.697530864197531e-07,
+      "loss": 8.1283,
+      "step": 639
+    },
+    {
+      "epoch": 2.95301431256325,
+      "grad_norm": 0.10660137236118317,
+      "learning_rate": 1.54320987654321e-07,
+      "loss": 7.7013,
+      "step": 640
+    },
+    {
+      "epoch": 2.9576405956339453,
+      "grad_norm": 0.12879547476768494,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 8.2472,
+      "step": 641
+    },
+    {
+      "epoch": 2.9622668787046407,
+      "grad_norm": 0.24027042090892792,
+      "learning_rate": 1.234567901234568e-07,
+      "loss": 9.3202,
+      "step": 642
+    },
+    {
+      "epoch": 2.966893161775336,
+      "grad_norm": 0.18989317119121552,
+      "learning_rate": 1.080246913580247e-07,
+      "loss": 8.6135,
+      "step": 643
+    },
+    {
+      "epoch": 2.9715194448460314,
+      "grad_norm": 0.20127813518047333,
+      "learning_rate": 9.259259259259259e-08,
+      "loss": 8.3557,
+      "step": 644
+    },
+    {
+      "epoch": 2.9761457279167267,
+      "grad_norm": 0.20816563069820404,
+      "learning_rate": 7.71604938271605e-08,
+      "loss": 9.6997,
+      "step": 645
+    },
+    {
+      "epoch": 2.9807720109874225,
+      "grad_norm": 0.2171671837568283,
+      "learning_rate": 6.17283950617284e-08,
+      "loss": 9.6809,
+      "step": 646
+    },
+    {
+      "epoch": 2.9853982940581174,
+      "grad_norm": 0.18569743633270264,
+      "learning_rate": 4.6296296296296295e-08,
+      "loss": 8.4721,
+      "step": 647
+    },
+    {
+      "epoch": 2.990024577128813,
+      "grad_norm": 0.16961927711963654,
+      "learning_rate": 3.08641975308642e-08,
+      "loss": 9.1963,
+      "step": 648
+    },
+    {
+      "epoch": 2.990024577128813,
+      "step": 648,
+      "total_flos": 2.9427853089130414e+18,
+      "train_loss": 10.740490390930647,
+      "train_runtime": 52216.2733,
+      "train_samples_per_second": 1.59,
+      "train_steps_per_second": 0.012
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 648,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.9427853089130414e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}