diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25893 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3693,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008123476848090983,
+      "grad_norm": 2.6259655952453613,
+      "learning_rate": 2.702702702702703e-08,
+      "loss": 0.7542,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016246953696181965,
+      "grad_norm": 2.684053421020508,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 0.7729,
+      "step": 2
+    },
+    {
+      "epoch": 0.002437043054427295,
+      "grad_norm": 2.8883492946624756,
+      "learning_rate": 8.108108108108109e-08,
+      "loss": 0.8025,
+      "step": 3
+    },
+    {
+      "epoch": 0.003249390739236393,
+      "grad_norm": 2.7433581352233887,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.7697,
+      "step": 4
+    },
+    {
+      "epoch": 0.004061738424045491,
+      "grad_norm": 2.7655677795410156,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 0.742,
+      "step": 5
+    },
+    {
+      "epoch": 0.00487408610885459,
+      "grad_norm": 2.7857422828674316,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.7557,
+      "step": 6
+    },
+    {
+      "epoch": 0.005686433793663688,
+      "grad_norm": 2.877302646636963,
+      "learning_rate": 1.8918918918918921e-07,
+      "loss": 0.78,
+      "step": 7
+    },
+    {
+      "epoch": 0.006498781478472786,
+      "grad_norm": 2.7022807598114014,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 0.7535,
+      "step": 8
+    },
+    {
+      "epoch": 0.007311129163281885,
+      "grad_norm": 2.84698486328125,
+      "learning_rate": 2.4324324324324326e-07,
+      "loss": 0.7756,
+      "step": 9
+    },
+    {
+      "epoch": 0.008123476848090982,
+      "grad_norm": 2.643312931060791,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.7787,
+      "step": 10
+    },
+    {
+      "epoch": 0.008935824532900082,
+      "grad_norm": 2.7004501819610596,
+      "learning_rate": 2.972972972972973e-07,
+      "loss": 0.744,
+      "step": 11
+    },
+    {
+      "epoch": 0.00974817221770918,
+      "grad_norm": 2.7050511837005615,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 0.7101,
+      "step": 12
+    },
+    {
+      "epoch": 0.010560519902518278,
+      "grad_norm": 2.74552059173584,
+      "learning_rate": 3.513513513513514e-07,
+      "loss": 0.7573,
+      "step": 13
+    },
+    {
+      "epoch": 0.011372867587327376,
+      "grad_norm": 2.671955108642578,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.742,
+      "step": 14
+    },
+    {
+      "epoch": 0.012185215272136474,
+      "grad_norm": 2.6270015239715576,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 0.7645,
+      "step": 15
+    },
+    {
+      "epoch": 0.012997562956945572,
+      "grad_norm": 2.6428890228271484,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 0.7551,
+      "step": 16
+    },
+    {
+      "epoch": 0.01380991064175467,
+      "grad_norm": 2.545128345489502,
+      "learning_rate": 4.5945945945945953e-07,
+      "loss": 0.7686,
+      "step": 17
+    },
+    {
+      "epoch": 0.01462225832656377,
+      "grad_norm": 2.62091326713562,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 0.7641,
+      "step": 18
+    },
+    {
+      "epoch": 0.015434606011372868,
+      "grad_norm": 2.10481595993042,
+      "learning_rate": 5.135135135135135e-07,
+      "loss": 0.7199,
+      "step": 19
+    },
+    {
+      "epoch": 0.016246953696181964,
+      "grad_norm": 2.116853952407837,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.7531,
+      "step": 20
+    },
+    {
+      "epoch": 0.017059301380991064,
+      "grad_norm": 2.0691580772399902,
+      "learning_rate": 5.675675675675676e-07,
+      "loss": 0.7249,
+      "step": 21
+    },
+    {
+      "epoch": 0.017871649065800164,
+      "grad_norm": 2.082359790802002,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.7455,
+      "step": 22
+    },
+    {
+      "epoch": 0.01868399675060926,
+      "grad_norm": 2.0863492488861084,
+      "learning_rate": 6.216216216216217e-07,
+      "loss": 0.7403,
+      "step": 23
+    },
+    {
+      "epoch": 0.01949634443541836,
+      "grad_norm": 1.847527265548706,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.7142,
+      "step": 24
+    },
+    {
+      "epoch": 0.020308692120227456,
+      "grad_norm": 1.8557764291763306,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.6902,
+      "step": 25
+    },
+    {
+      "epoch": 0.021121039805036556,
+      "grad_norm": 1.5264045000076294,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.6794,
+      "step": 26
+    },
+    {
+      "epoch": 0.021933387489845652,
+      "grad_norm": 1.4047449827194214,
+      "learning_rate": 7.297297297297298e-07,
+      "loss": 0.7161,
+      "step": 27
+    },
+    {
+      "epoch": 0.022745735174654752,
+      "grad_norm": 1.4444795846939087,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.7223,
+      "step": 28
+    },
+    {
+      "epoch": 0.023558082859463852,
+      "grad_norm": 1.354972004890442,
+      "learning_rate": 7.837837837837839e-07,
+      "loss": 0.7231,
+      "step": 29
+    },
+    {
+      "epoch": 0.024370430544272948,
+      "grad_norm": 1.340536117553711,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.7346,
+      "step": 30
+    },
+    {
+      "epoch": 0.025182778229082048,
+      "grad_norm": 1.235442876815796,
+      "learning_rate": 8.37837837837838e-07,
+      "loss": 0.6796,
+      "step": 31
+    },
+    {
+      "epoch": 0.025995125913891144,
+      "grad_norm": 1.2592628002166748,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.6641,
+      "step": 32
+    },
+    {
+      "epoch": 0.026807473598700244,
+      "grad_norm": 1.2871164083480835,
+      "learning_rate": 8.91891891891892e-07,
+      "loss": 0.6715,
+      "step": 33
+    },
+    {
+      "epoch": 0.02761982128350934,
+      "grad_norm": 1.194066047668457,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.6662,
+      "step": 34
+    },
+    {
+      "epoch": 0.02843216896831844,
+      "grad_norm": 0.978544294834137,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.6815,
+      "step": 35
+    },
+    {
+      "epoch": 0.02924451665312754,
+      "grad_norm": 1.0026520490646362,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.6894,
+      "step": 36
+    },
+    {
+      "epoch": 0.030056864337936636,
+      "grad_norm": 0.9322858452796936,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6424,
+      "step": 37
+    },
+    {
+      "epoch": 0.030869212022745736,
+      "grad_norm": 0.9230260848999023,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.603,
+      "step": 38
+    },
+    {
+      "epoch": 0.03168155970755483,
+      "grad_norm": 0.9275475740432739,
+      "learning_rate": 1.0540540540540542e-06,
+      "loss": 0.6618,
+      "step": 39
+    },
+    {
+      "epoch": 0.03249390739236393,
+      "grad_norm": 0.9010961055755615,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.6604,
+      "step": 40
+    },
+    {
+      "epoch": 0.03330625507717303,
+      "grad_norm": 0.9825568795204163,
+      "learning_rate": 1.1081081081081083e-06,
+      "loss": 0.684,
+      "step": 41
+    },
+    {
+      "epoch": 0.03411860276198213,
+      "grad_norm": 0.8298603296279907,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.6099,
+      "step": 42
+    },
+    {
+      "epoch": 0.034930950446791224,
+      "grad_norm": 0.763951301574707,
+      "learning_rate": 1.1621621621621624e-06,
+      "loss": 0.6015,
+      "step": 43
+    },
+    {
+      "epoch": 0.03574329813160033,
+      "grad_norm": 0.7812492251396179,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.6071,
+      "step": 44
+    },
+    {
+      "epoch": 0.036555645816409424,
+      "grad_norm": 0.759466290473938,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.6505,
+      "step": 45
+    },
+    {
+      "epoch": 0.03736799350121852,
+      "grad_norm": 0.6911046504974365,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.6238,
+      "step": 46
+    },
+    {
+      "epoch": 0.038180341186027617,
+      "grad_norm": 0.7092072367668152,
+      "learning_rate": 1.2702702702702705e-06,
+      "loss": 0.6467,
+      "step": 47
+    },
+    {
+      "epoch": 0.03899268887083672,
+      "grad_norm": 0.6582714915275574,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.6037,
+      "step": 48
+    },
+    {
+      "epoch": 0.039805036555645816,
+      "grad_norm": 0.6226052045822144,
+      "learning_rate": 1.3243243243243246e-06,
+      "loss": 0.6345,
+      "step": 49
+    },
+    {
+      "epoch": 0.04061738424045491,
+      "grad_norm": 0.5834541916847229,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.5871,
+      "step": 50
+    },
+    {
+      "epoch": 0.041429731925264016,
+      "grad_norm": 0.545859694480896,
+      "learning_rate": 1.3783783783783786e-06,
+      "loss": 0.5789,
+      "step": 51
+    },
+    {
+      "epoch": 0.04224207961007311,
+      "grad_norm": 0.5138876438140869,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.6131,
+      "step": 52
+    },
+    {
+      "epoch": 0.04305442729488221,
+      "grad_norm": 0.4840790331363678,
+      "learning_rate": 1.4324324324324327e-06,
+      "loss": 0.5915,
+      "step": 53
+    },
+    {
+      "epoch": 0.043866774979691305,
+      "grad_norm": 0.47509336471557617,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.5875,
+      "step": 54
+    },
+    {
+      "epoch": 0.04467912266450041,
+      "grad_norm": 0.5023272633552551,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.6069,
+      "step": 55
+    },
+    {
+      "epoch": 0.045491470349309504,
+      "grad_norm": 0.4830177128314972,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.6035,
+      "step": 56
+    },
+    {
+      "epoch": 0.0463038180341186,
+      "grad_norm": 0.4710570275783539,
+      "learning_rate": 1.5405405405405409e-06,
+      "loss": 0.5828,
+      "step": 57
+    },
+    {
+      "epoch": 0.047116165718927704,
+      "grad_norm": 0.47543585300445557,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.5878,
+      "step": 58
+    },
+    {
+      "epoch": 0.0479285134037368,
+      "grad_norm": 0.41436460614204407,
+      "learning_rate": 1.5945945945945947e-06,
+      "loss": 0.5753,
+      "step": 59
+    },
+    {
+      "epoch": 0.048740861088545896,
+      "grad_norm": 0.420043408870697,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.5832,
+      "step": 60
+    },
+    {
+      "epoch": 0.04955320877335499,
+      "grad_norm": 0.4084452688694,
+      "learning_rate": 1.6486486486486488e-06,
+      "loss": 0.5738,
+      "step": 61
+    },
+    {
+      "epoch": 0.050365556458164096,
+      "grad_norm": 0.3670467436313629,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.5353,
+      "step": 62
+    },
+    {
+      "epoch": 0.05117790414297319,
+      "grad_norm": 0.38712573051452637,
+      "learning_rate": 1.7027027027027028e-06,
+      "loss": 0.552,
+      "step": 63
+    },
+    {
+      "epoch": 0.05199025182778229,
+      "grad_norm": 0.34750255942344666,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.5577,
+      "step": 64
+    },
+    {
+      "epoch": 0.05280259951259139,
+      "grad_norm": 0.3566388785839081,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.5481,
+      "step": 65
+    },
+    {
+      "epoch": 0.05361494719740049,
+      "grad_norm": 0.38597235083580017,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.5562,
+      "step": 66
+    },
+    {
+      "epoch": 0.054427294882209584,
+      "grad_norm": 0.32903724908828735,
+      "learning_rate": 1.810810810810811e-06,
+      "loss": 0.5469,
+      "step": 67
+    },
+    {
+      "epoch": 0.05523964256701868,
+      "grad_norm": 0.35578906536102295,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.5702,
+      "step": 68
+    },
+    {
+      "epoch": 0.056051990251827784,
+      "grad_norm": 0.3708760440349579,
+      "learning_rate": 1.864864864864865e-06,
+      "loss": 0.5881,
+      "step": 69
+    },
+    {
+      "epoch": 0.05686433793663688,
+      "grad_norm": 0.39675769209861755,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.5465,
+      "step": 70
+    },
+    {
+      "epoch": 0.057676685621445976,
+      "grad_norm": 0.3549048900604248,
+      "learning_rate": 1.918918918918919e-06,
+      "loss": 0.5444,
+      "step": 71
+    },
+    {
+      "epoch": 0.05848903330625508,
+      "grad_norm": 0.3596271872520447,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.5447,
+      "step": 72
+    },
+    {
+      "epoch": 0.059301380991064176,
+      "grad_norm": 0.33823689818382263,
+      "learning_rate": 1.9729729729729734e-06,
+      "loss": 0.5414,
+      "step": 73
+    },
+    {
+      "epoch": 0.06011372867587327,
+      "grad_norm": 0.3020276129245758,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5194,
+      "step": 74
+    },
+    {
+      "epoch": 0.060926076360682375,
+      "grad_norm": 0.32985058426856995,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.5335,
+      "step": 75
+    },
+    {
+      "epoch": 0.06173842404549147,
+      "grad_norm": 0.320764422416687,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.5491,
+      "step": 76
+    },
+    {
+      "epoch": 0.06255077173030058,
+      "grad_norm": 0.3342389166355133,
+      "learning_rate": 2.0810810810810815e-06,
+      "loss": 0.5494,
+      "step": 77
+    },
+    {
+      "epoch": 0.06336311941510966,
+      "grad_norm": 0.3533325493335724,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.5308,
+      "step": 78
+    },
+    {
+      "epoch": 0.06417546709991877,
+      "grad_norm": 0.28463953733444214,
+      "learning_rate": 2.1351351351351354e-06,
+      "loss": 0.5187,
+      "step": 79
+    },
+    {
+      "epoch": 0.06498781478472786,
+      "grad_norm": 0.31918853521347046,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.5514,
+      "step": 80
+    },
+    {
+      "epoch": 0.06580016246953696,
+      "grad_norm": 0.3060842454433441,
+      "learning_rate": 2.1891891891891897e-06,
+      "loss": 0.5126,
+      "step": 81
+    },
+    {
+      "epoch": 0.06661251015434606,
+      "grad_norm": 0.31643056869506836,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.5226,
+      "step": 82
+    },
+    {
+      "epoch": 0.06742485783915515,
+      "grad_norm": 0.3060136139392853,
+      "learning_rate": 2.2432432432432435e-06,
+      "loss": 0.5224,
+      "step": 83
+    },
+    {
+      "epoch": 0.06823720552396426,
+      "grad_norm": 0.3351191580295563,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.5307,
+      "step": 84
+    },
+    {
+      "epoch": 0.06904955320877336,
+      "grad_norm": 0.26106613874435425,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.5126,
+      "step": 85
+    },
+    {
+      "epoch": 0.06986190089358245,
+      "grad_norm": 0.30249306559562683,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.5285,
+      "step": 86
+    },
+    {
+      "epoch": 0.07067424857839155,
+      "grad_norm": 0.3043215572834015,
+      "learning_rate": 2.3513513513513517e-06,
+      "loss": 0.541,
+      "step": 87
+    },
+    {
+      "epoch": 0.07148659626320066,
+      "grad_norm": 0.26914525032043457,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.5219,
+      "step": 88
+    },
+    {
+      "epoch": 0.07229894394800974,
+      "grad_norm": 0.2997298538684845,
+      "learning_rate": 2.4054054054054055e-06,
+      "loss": 0.484,
+      "step": 89
+    },
+    {
+      "epoch": 0.07311129163281885,
+      "grad_norm": 0.29731836915016174,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.541,
+      "step": 90
+    },
+    {
+      "epoch": 0.07392363931762795,
+      "grad_norm": 0.2861558794975281,
+      "learning_rate": 2.45945945945946e-06,
+      "loss": 0.5235,
+      "step": 91
+    },
+    {
+      "epoch": 0.07473598700243704,
+      "grad_norm": 0.28016558289527893,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.5049,
+      "step": 92
+    },
+    {
+      "epoch": 0.07554833468724614,
+      "grad_norm": 0.28269246220588684,
+      "learning_rate": 2.5135135135135137e-06,
+      "loss": 0.4736,
+      "step": 93
+    },
+    {
+      "epoch": 0.07636068237205523,
+      "grad_norm": 0.2894327640533447,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.484,
+      "step": 94
+    },
+    {
+      "epoch": 0.07717303005686434,
+      "grad_norm": 0.32706964015960693,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.5046,
+      "step": 95
+    },
+    {
+      "epoch": 0.07798537774167344,
+      "grad_norm": 0.2872982919216156,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.4888,
+      "step": 96
+    },
+    {
+      "epoch": 0.07879772542648253,
+      "grad_norm": 0.25910407304763794,
+      "learning_rate": 2.621621621621622e-06,
+      "loss": 0.5285,
+      "step": 97
+    },
+    {
+      "epoch": 0.07961007311129163,
+      "grad_norm": 0.2655861973762512,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.5215,
+      "step": 98
+    },
+    {
+      "epoch": 0.08042242079610074,
+      "grad_norm": 0.25582069158554077,
+      "learning_rate": 2.6756756756756757e-06,
+      "loss": 0.5159,
+      "step": 99
+    },
+    {
+      "epoch": 0.08123476848090982,
+      "grad_norm": 0.27265170216560364,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.4695,
+      "step": 100
+    },
+    {
+      "epoch": 0.08204711616571893,
+      "grad_norm": 0.291411817073822,
+      "learning_rate": 2.72972972972973e-06,
+      "loss": 0.4973,
+      "step": 101
+    },
+    {
+      "epoch": 0.08285946385052803,
+      "grad_norm": 0.2375059723854065,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.5051,
+      "step": 102
+    },
+    {
+      "epoch": 0.08367181153533712,
+      "grad_norm": 0.26637566089630127,
+      "learning_rate": 2.783783783783784e-06,
+      "loss": 0.4905,
+      "step": 103
+    },
+    {
+      "epoch": 0.08448415922014622,
+      "grad_norm": 0.2717621326446533,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.4877,
+      "step": 104
+    },
+    {
+      "epoch": 0.08529650690495533,
+      "grad_norm": 0.24362023174762726,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.4791,
+      "step": 105
+    },
+    {
+      "epoch": 0.08610885458976442,
+      "grad_norm": 0.252564936876297,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.51,
+      "step": 106
+    },
+    {
+      "epoch": 0.08692120227457352,
+      "grad_norm": 0.23802724480628967,
+      "learning_rate": 2.891891891891892e-06,
+      "loss": 0.5131,
+      "step": 107
+    },
+    {
+      "epoch": 0.08773354995938261,
+      "grad_norm": 0.2623497545719147,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.4972,
+      "step": 108
+    },
+    {
+      "epoch": 0.08854589764419171,
+      "grad_norm": 0.27165302634239197,
+      "learning_rate": 2.9459459459459462e-06,
+      "loss": 0.513,
+      "step": 109
+    },
+    {
+      "epoch": 0.08935824532900082,
+      "grad_norm": 0.26373592019081116,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.513,
+      "step": 110
+    },
+    {
+      "epoch": 0.0901705930138099,
+      "grad_norm": 0.2797335684299469,
+      "learning_rate": 3e-06,
+      "loss": 0.4937,
+      "step": 111
+    },
+    {
+      "epoch": 0.09098294069861901,
+      "grad_norm": 0.25746583938598633,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.4915,
+      "step": 112
+    },
+    {
+      "epoch": 0.09179528838342811,
+      "grad_norm": 0.3143309950828552,
+      "learning_rate": 3.0540540540540544e-06,
+      "loss": 0.5143,
+      "step": 113
+    },
+    {
+      "epoch": 0.0926076360682372,
+      "grad_norm": 0.28754913806915283,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.5014,
+      "step": 114
+    },
+    {
+      "epoch": 0.0934199837530463,
+      "grad_norm": 0.26556989550590515,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.4454,
+      "step": 115
+    },
+    {
+      "epoch": 0.09423233143785541,
+      "grad_norm": 0.2956714928150177,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.4964,
+      "step": 116
+    },
+    {
+      "epoch": 0.0950446791226645,
+      "grad_norm": 0.3376867175102234,
+      "learning_rate": 3.1621621621621625e-06,
+      "loss": 0.5126,
+      "step": 117
+    },
+    {
+      "epoch": 0.0958570268074736,
+      "grad_norm": 0.2606317102909088,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.4885,
+      "step": 118
+    },
+    {
+      "epoch": 0.0966693744922827,
+      "grad_norm": 0.2570057213306427,
+      "learning_rate": 3.2162162162162164e-06,
+      "loss": 0.4821,
+      "step": 119
+    },
+    {
+      "epoch": 0.09748172217709179,
+      "grad_norm": 0.28612828254699707,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.4882,
+      "step": 120
+    },
+    {
+      "epoch": 0.0982940698619009,
+      "grad_norm": 0.27355316281318665,
+      "learning_rate": 3.2702702702702706e-06,
+      "loss": 0.4904,
+      "step": 121
+    },
+    {
+      "epoch": 0.09910641754670999,
+      "grad_norm": 0.26200518012046814,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.4877,
+      "step": 122
+    },
+    {
+      "epoch": 0.09991876523151909,
+      "grad_norm": 0.40036794543266296,
+      "learning_rate": 3.3243243243243245e-06,
+      "loss": 0.4778,
+      "step": 123
+    },
+    {
+      "epoch": 0.10073111291632819,
+      "grad_norm": 0.2663668394088745,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.487,
+      "step": 124
+    },
+    {
+      "epoch": 0.10154346060113728,
+      "grad_norm": 0.26201802492141724,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.4815,
+      "step": 125
+    },
+    {
+      "epoch": 0.10235580828594638,
+      "grad_norm": 0.3520504832267761,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.5029,
+      "step": 126
+    },
+    {
+      "epoch": 0.10316815597075549,
+      "grad_norm": 0.25964856147766113,
+      "learning_rate": 3.4324324324324326e-06,
+      "loss": 0.501,
+      "step": 127
+    },
+    {
+      "epoch": 0.10398050365556458,
+      "grad_norm": 0.29509663581848145,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.483,
+      "step": 128
+    },
+    {
+      "epoch": 0.10479285134037368,
+      "grad_norm": 0.2779506742954254,
+      "learning_rate": 3.4864864864864865e-06,
+      "loss": 0.4913,
+      "step": 129
+    },
+    {
+      "epoch": 0.10560519902518278,
+      "grad_norm": 0.24384281039237976,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.4579,
+      "step": 130
+    },
+    {
+      "epoch": 0.10641754670999187,
+      "grad_norm": 0.324032723903656,
+      "learning_rate": 3.5405405405405408e-06,
+      "loss": 0.478,
+      "step": 131
+    },
+    {
+      "epoch": 0.10722989439480098,
+      "grad_norm": 0.29088786244392395,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.4813,
+      "step": 132
+    },
+    {
+      "epoch": 0.10804224207961008,
+      "grad_norm": 0.28625380992889404,
+      "learning_rate": 3.5945945945945946e-06,
+      "loss": 0.5118,
+      "step": 133
+    },
+    {
+      "epoch": 0.10885458976441917,
+      "grad_norm": 0.25363340973854065,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.5141,
+      "step": 134
+    },
+    {
+      "epoch": 0.10966693744922827,
+      "grad_norm": 0.25895047187805176,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.4719,
+      "step": 135
+    },
+    {
+      "epoch": 0.11047928513403736,
+      "grad_norm": 0.2995051145553589,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.5042,
+      "step": 136
+    },
+    {
+      "epoch": 0.11129163281884646,
+      "grad_norm": 0.2758341133594513,
+      "learning_rate": 3.7027027027027028e-06,
+      "loss": 0.472,
+      "step": 137
+    },
+    {
+      "epoch": 0.11210398050365557,
+      "grad_norm": 0.2719106674194336,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.4848,
+      "step": 138
+    },
+    {
+      "epoch": 0.11291632818846466,
+      "grad_norm": 0.2795969545841217,
+      "learning_rate": 3.756756756756757e-06,
+      "loss": 0.4945,
+      "step": 139
+    },
+    {
+      "epoch": 0.11372867587327376,
+      "grad_norm": 0.28690218925476074,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.4588,
+      "step": 140
+    },
+    {
+      "epoch": 0.11454102355808286,
+      "grad_norm": 0.27972084283828735,
+      "learning_rate": 3.810810810810811e-06,
+      "loss": 0.4942,
+      "step": 141
+    },
+    {
+      "epoch": 0.11535337124289195,
+      "grad_norm": 0.24317964911460876,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.4801,
+      "step": 142
+    },
+    {
+      "epoch": 0.11616571892770106,
+      "grad_norm": 0.2486281394958496,
+      "learning_rate": 3.864864864864865e-06,
+      "loss": 0.4571,
+      "step": 143
+    },
+    {
+      "epoch": 0.11697806661251016,
+      "grad_norm": 0.2687414288520813,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.4922,
+      "step": 144
+    },
+    {
+      "epoch": 0.11779041429731925,
+      "grad_norm": 0.25073814392089844,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.467,
+      "step": 145
+    },
+    {
+      "epoch": 0.11860276198212835,
+      "grad_norm": 0.23224234580993652,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.469,
+      "step": 146
+    },
+    {
+      "epoch": 0.11941510966693746,
+      "grad_norm": 0.25092196464538574,
+      "learning_rate": 3.972972972972973e-06,
+      "loss": 0.4639,
+      "step": 147
+    },
+    {
+      "epoch": 0.12022745735174654,
+      "grad_norm": 0.25130388140678406,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.4727,
+      "step": 148
+    },
+    {
+      "epoch": 0.12103980503655565,
+      "grad_norm": 0.24852167069911957,
+      "learning_rate": 4.027027027027028e-06,
+      "loss": 0.4875,
+      "step": 149
+    },
+    {
+      "epoch": 0.12185215272136475,
+      "grad_norm": 0.30451756715774536,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.4848,
+      "step": 150
+    },
+    {
+      "epoch": 0.12266450040617384,
+      "grad_norm": 0.2702716290950775,
+      "learning_rate": 4.0810810810810815e-06,
+      "loss": 0.4669,
+      "step": 151
+    },
+    {
+      "epoch": 0.12347684809098294,
+      "grad_norm": 0.2570095360279083,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.4277,
+      "step": 152
+    },
+    {
+      "epoch": 0.12428919577579203,
+      "grad_norm": 0.2763490080833435,
+      "learning_rate": 4.135135135135135e-06,
+      "loss": 0.4559,
+      "step": 153
+    },
+    {
+      "epoch": 0.12510154346060115,
+      "grad_norm": 0.2596904933452606,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.468,
+      "step": 154
+    },
+    {
+      "epoch": 0.12591389114541024,
+      "grad_norm": 0.2649424076080322,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.4761,
+      "step": 155
+    },
+    {
+      "epoch": 0.12672623883021933,
+      "grad_norm": 0.25593119859695435,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.4609,
+      "step": 156
+    },
+    {
+      "epoch": 0.12753858651502845,
+      "grad_norm": 0.26088911294937134,
+      "learning_rate": 4.243243243243244e-06,
+      "loss": 0.4762,
+      "step": 157
+    },
+    {
+      "epoch": 0.12835093419983754,
+      "grad_norm": 0.24846747517585754,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.4779,
+      "step": 158
+    },
+    {
+      "epoch": 0.12916328188464662,
+      "grad_norm": 0.25237545371055603,
+      "learning_rate": 4.297297297297298e-06,
+      "loss": 0.4532,
+      "step": 159
+    },
+    {
+      "epoch": 0.12997562956945571,
+      "grad_norm": 0.25048646330833435,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.4727,
+      "step": 160
+    },
+    {
+      "epoch": 0.13078797725426483,
+      "grad_norm": 0.6334772706031799,
+      "learning_rate": 4.351351351351352e-06,
+      "loss": 0.4406,
+      "step": 161
+    },
+    {
+      "epoch": 0.13160032493907392,
+      "grad_norm": 0.28383150696754456,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.457,
+      "step": 162
+    },
+    {
+      "epoch": 0.132412672623883,
+      "grad_norm": 0.2885434627532959,
+      "learning_rate": 4.4054054054054054e-06,
+      "loss": 0.4627,
+      "step": 163
+    },
+    {
+      "epoch": 0.13322502030869213,
+      "grad_norm": 0.2567489743232727,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.4704,
+      "step": 164
+    },
+    {
+      "epoch": 0.13403736799350122,
+      "grad_norm": 0.25216516852378845,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.4626,
+      "step": 165
+    },
+    {
+      "epoch": 0.1348497156783103,
+      "grad_norm": 0.3566623330116272,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.4803,
+      "step": 166
+    },
+    {
+      "epoch": 0.13566206336311942,
+      "grad_norm": 0.28616055846214294,
+      "learning_rate": 4.513513513513514e-06,
+      "loss": 0.4668,
+      "step": 167
+    },
+    {
+      "epoch": 0.1364744110479285,
+      "grad_norm": 0.2680723965167999,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.4506,
+      "step": 168
+    },
+    {
+      "epoch": 0.1372867587327376,
+      "grad_norm": 0.2549089789390564,
+      "learning_rate": 4.567567567567568e-06,
+      "loss": 0.4375,
+      "step": 169
+    },
+    {
+      "epoch": 0.13809910641754672,
+      "grad_norm": 0.3143906593322754,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.462,
+      "step": 170
+    },
+    {
+      "epoch": 0.1389114541023558,
+      "grad_norm": 0.33917030692100525,
+      "learning_rate": 4.621621621621622e-06,
+      "loss": 0.4621,
+      "step": 171
+    },
+    {
+      "epoch": 0.1397238017871649,
+      "grad_norm": 0.2662394642829895,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.4824,
+      "step": 172
+    },
+    {
+      "epoch": 0.14053614947197401,
+      "grad_norm": 0.2562621533870697,
+      "learning_rate": 4.675675675675676e-06,
+      "loss": 0.4579,
+      "step": 173
+    },
+    {
+      "epoch": 0.1413484971567831,
+      "grad_norm": 0.25691378116607666,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.4303,
+      "step": 174
+    },
+    {
+      "epoch": 0.1421608448415922,
+      "grad_norm": 0.26857617497444153,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.4636,
+      "step": 175
+    },
+    {
+      "epoch": 0.1429731925264013,
+      "grad_norm": 0.26752474904060364,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.477,
+      "step": 176
+    },
+    {
+      "epoch": 0.1437855402112104,
+      "grad_norm": 0.24300289154052734,
+      "learning_rate": 4.783783783783784e-06,
+      "loss": 0.4335,
+      "step": 177
+    },
+    {
+      "epoch": 0.1445978878960195,
+      "grad_norm": 0.3101292550563812,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.4578,
+      "step": 178
+    },
+    {
+      "epoch": 0.1454102355808286,
+      "grad_norm": 0.2360691875219345,
+      "learning_rate": 4.837837837837838e-06,
+      "loss": 0.4656,
+      "step": 179
+    },
+    {
+      "epoch": 0.1462225832656377,
+      "grad_norm": 0.2245030403137207,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.4594,
+      "step": 180
+    },
+    {
+      "epoch": 0.14703493095044678,
+      "grad_norm": 0.24627269804477692,
+      "learning_rate": 4.891891891891893e-06,
+      "loss": 0.4641,
+      "step": 181
+    },
+    {
+      "epoch": 0.1478472786352559,
+      "grad_norm": 0.2386178821325302,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.4495,
+      "step": 182
+    },
+    {
+      "epoch": 0.148659626320065,
+      "grad_norm": 0.24851150810718536,
+      "learning_rate": 4.9459459459459466e-06,
+      "loss": 0.4781,
+      "step": 183
+    },
+    {
+      "epoch": 0.14947197400487408,
+      "grad_norm": 0.23232993483543396,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.441,
+      "step": 184
+    },
+    {
+      "epoch": 0.1502843216896832,
+      "grad_norm": 0.24872195720672607,
+      "learning_rate": 5e-06,
+      "loss": 0.4714,
+      "step": 185
+    },
+    {
+      "epoch": 0.1510966693744923,
+      "grad_norm": 0.248837411403656,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.4362,
+      "step": 186
+    },
+    {
+      "epoch": 0.15190901705930138,
+      "grad_norm": 0.24189503490924835,
+      "learning_rate": 5.054054054054054e-06,
+      "loss": 0.4275,
+      "step": 187
+    },
+    {
+      "epoch": 0.15272136474411047,
+      "grad_norm": 0.2560644745826721,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.4597,
+      "step": 188
+    },
+    {
+      "epoch": 0.15353371242891958,
+      "grad_norm": 0.28990936279296875,
+      "learning_rate": 5.108108108108108e-06,
+      "loss": 0.4435,
+      "step": 189
+    },
+    {
+      "epoch": 0.15434606011372867,
+      "grad_norm": 0.29622894525527954,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.4541,
+      "step": 190
+    },
+    {
+      "epoch": 0.15515840779853776,
+      "grad_norm": 0.31443700194358826,
+      "learning_rate": 5.162162162162162e-06,
+      "loss": 0.4363,
+      "step": 191
+    },
+    {
+      "epoch": 0.15597075548334688,
+      "grad_norm": 0.3032810091972351,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 0.4609,
+      "step": 192
+    },
+    {
+      "epoch": 0.15678310316815597,
+      "grad_norm": 0.25123271346092224,
+      "learning_rate": 5.216216216216217e-06,
+      "loss": 0.4736,
+      "step": 193
+    },
+    {
+      "epoch": 0.15759545085296506,
+      "grad_norm": 0.2665044367313385,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.4687,
+      "step": 194
+    },
+    {
+      "epoch": 0.15840779853777417,
+      "grad_norm": 0.2593575119972229,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.4558,
+      "step": 195
+    },
+    {
+      "epoch": 0.15922014622258326,
+      "grad_norm": 0.32702040672302246,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.4338,
+      "step": 196
+    },
+    {
+      "epoch": 0.16003249390739235,
+      "grad_norm": 0.24266457557678223,
+      "learning_rate": 5.324324324324324e-06,
+      "loss": 0.4551,
+      "step": 197
+    },
+    {
+      "epoch": 0.16084484159220147,
+      "grad_norm": 0.27933767437934875,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.4582,
+      "step": 198
+    },
+    {
+      "epoch": 0.16165718927701056,
+      "grad_norm": 0.24401335418224335,
+      "learning_rate": 5.378378378378378e-06,
+      "loss": 0.4282,
+      "step": 199
+    },
+    {
+      "epoch": 0.16246953696181965,
+      "grad_norm": 0.3174748718738556,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.4529,
+      "step": 200
+    },
+    {
+      "epoch": 0.16328188464662877,
+      "grad_norm": 0.2680133283138275,
+      "learning_rate": 5.432432432432433e-06,
+      "loss": 0.4557,
+      "step": 201
+    },
+    {
+      "epoch": 0.16409423233143786,
+      "grad_norm": 0.2866247594356537,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.4209,
+      "step": 202
+    },
+    {
+      "epoch": 0.16490658001624695,
+      "grad_norm": 0.28320467472076416,
+      "learning_rate": 5.486486486486487e-06,
+      "loss": 0.4527,
+      "step": 203
+    },
+    {
+      "epoch": 0.16571892770105606,
+      "grad_norm": 0.26222071051597595,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.4435,
+      "step": 204
+    },
+    {
+      "epoch": 0.16653127538586515,
+      "grad_norm": 0.24433837831020355,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.4245,
+      "step": 205
+    },
+    {
+      "epoch": 0.16734362307067424,
+      "grad_norm": 0.27183082699775696,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.4331,
+      "step": 206
+    },
+    {
+      "epoch": 0.16815597075548336,
+      "grad_norm": 0.2708176374435425,
+      "learning_rate": 5.5945945945945945e-06,
+      "loss": 0.4435,
+      "step": 207
+    },
+    {
+      "epoch": 0.16896831844029245,
+      "grad_norm": 0.2555078864097595,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.4578,
+      "step": 208
+    },
+    {
+      "epoch": 0.16978066612510154,
+      "grad_norm": 0.27564021944999695,
+      "learning_rate": 5.648648648648649e-06,
+      "loss": 0.4489,
+      "step": 209
+    },
+    {
+      "epoch": 0.17059301380991065,
+      "grad_norm": 0.24576057493686676,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.4378,
+      "step": 210
+    },
+    {
+      "epoch": 0.17140536149471974,
+      "grad_norm": 0.26149407029151917,
+      "learning_rate": 5.702702702702702e-06,
+      "loss": 0.4491,
+      "step": 211
+    },
+    {
+      "epoch": 0.17221770917952883,
+      "grad_norm": 0.2775116264820099,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 0.4353,
+      "step": 212
+    },
+    {
+      "epoch": 0.17303005686433795,
+      "grad_norm": 0.25006476044654846,
+      "learning_rate": 5.756756756756757e-06,
+      "loss": 0.4603,
+      "step": 213
+    },
+    {
+      "epoch": 0.17384240454914704,
+      "grad_norm": 0.2590186297893524,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.4168,
+      "step": 214
+    },
+    {
+      "epoch": 0.17465475223395613,
+      "grad_norm": 0.290622353553772,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.442,
+      "step": 215
+    },
+    {
+      "epoch": 0.17546709991876522,
+      "grad_norm": 0.2712083160877228,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.4261,
+      "step": 216
+    },
+    {
+      "epoch": 0.17627944760357434,
+      "grad_norm": 0.2958680987358093,
+      "learning_rate": 5.8648648648648655e-06,
+      "loss": 0.4409,
+      "step": 217
+    },
+    {
+      "epoch": 0.17709179528838342,
+      "grad_norm": 0.2784006595611572,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.4565,
+      "step": 218
+    },
+    {
+      "epoch": 0.17790414297319251,
+      "grad_norm": 0.29732710123062134,
+      "learning_rate": 5.9189189189189185e-06,
+      "loss": 0.4539,
+      "step": 219
+    },
+    {
+      "epoch": 0.17871649065800163,
+      "grad_norm": 0.25454479455947876,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.4339,
+      "step": 220
+    },
+    {
+      "epoch": 0.17952883834281072,
+      "grad_norm": 0.25578466057777405,
+      "learning_rate": 5.972972972972973e-06,
+      "loss": 0.4421,
+      "step": 221
+    },
+    {
+      "epoch": 0.1803411860276198,
+      "grad_norm": 0.2576488256454468,
+      "learning_rate": 6e-06,
+      "loss": 0.4453,
+      "step": 222
+    },
+    {
+      "epoch": 0.18115353371242893,
+      "grad_norm": 0.24713659286499023,
+      "learning_rate": 6.027027027027027e-06,
+      "loss": 0.46,
+      "step": 223
+    },
+    {
+      "epoch": 0.18196588139723802,
+      "grad_norm": 0.24417951703071594,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.4439,
+      "step": 224
+    },
+    {
+      "epoch": 0.1827782290820471,
+      "grad_norm": 0.24308623373508453,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.45,
+      "step": 225
+    },
+    {
+      "epoch": 0.18359057676685622,
+      "grad_norm": 0.2346763014793396,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.4284,
+      "step": 226
+    },
+    {
+      "epoch": 0.1844029244516653,
+      "grad_norm": 0.25247687101364136,
+      "learning_rate": 6.135135135135135e-06,
+      "loss": 0.4305,
+      "step": 227
+    },
+    {
+      "epoch": 0.1852152721364744,
+      "grad_norm": 0.26878121495246887,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.4375,
+      "step": 228
+    },
+    {
+      "epoch": 0.18602761982128352,
+      "grad_norm": 0.26521703600883484,
+      "learning_rate": 6.1891891891891895e-06,
+      "loss": 0.463,
+      "step": 229
+    },
+    {
+      "epoch": 0.1868399675060926,
+      "grad_norm": 0.23791255056858063,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.4306,
+      "step": 230
+    },
+    {
+      "epoch": 0.1876523151909017,
+      "grad_norm": 0.2792019844055176,
+      "learning_rate": 6.243243243243243e-06,
+      "loss": 0.4588,
+      "step": 231
+    },
+    {
+      "epoch": 0.18846466287571081,
+      "grad_norm": 0.2523825168609619,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.4166,
+      "step": 232
+    },
+    {
+      "epoch": 0.1892770105605199,
+      "grad_norm": 0.31810125708580017,
+      "learning_rate": 6.297297297297298e-06,
+      "loss": 0.4273,
+      "step": 233
+    },
+    {
+      "epoch": 0.190089358245329,
+      "grad_norm": 0.29075729846954346,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.4361,
+      "step": 234
+    },
+    {
+      "epoch": 0.1909017059301381,
+      "grad_norm": 0.29173803329467773,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.4743,
+      "step": 235
+    },
+    {
+      "epoch": 0.1917140536149472,
+      "grad_norm": 0.27029022574424744,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.4142,
+      "step": 236
+    },
+    {
+      "epoch": 0.1925264012997563,
+      "grad_norm": 0.2797292172908783,
+      "learning_rate": 6.405405405405406e-06,
+      "loss": 0.4261,
+      "step": 237
+    },
+    {
+      "epoch": 0.1933387489845654,
+      "grad_norm": 0.25351187586784363,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 0.4353,
+      "step": 238
+    },
+    {
+      "epoch": 0.1941510966693745,
+      "grad_norm": 0.3089354932308197,
+      "learning_rate": 6.45945945945946e-06,
+      "loss": 0.4336,
+      "step": 239
+    },
+    {
+      "epoch": 0.19496344435418358,
+      "grad_norm": 0.26509737968444824,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.4679,
+      "step": 240
+    },
+    {
+      "epoch": 0.1957757920389927,
+      "grad_norm": 0.2815364897251129,
+      "learning_rate": 6.513513513513514e-06,
+      "loss": 0.4327,
+      "step": 241
+    },
+    {
+      "epoch": 0.1965881397238018,
+      "grad_norm": 0.25884097814559937,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 0.4474,
+      "step": 242
+    },
+    {
+      "epoch": 0.19740048740861088,
+      "grad_norm": 0.2712419629096985,
+      "learning_rate": 6.567567567567567e-06,
+      "loss": 0.4259,
+      "step": 243
+    },
+    {
+      "epoch": 0.19821283509341997,
+      "grad_norm": 0.25983932614326477,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.4566,
+      "step": 244
+    },
+    {
+      "epoch": 0.1990251827782291,
+      "grad_norm": 0.2675042152404785,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.4472,
+      "step": 245
+    },
+    {
+      "epoch": 0.19983753046303818,
+      "grad_norm": 0.2535611689090729,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.42,
+      "step": 246
+    },
+    {
+      "epoch": 0.20064987814784727,
+      "grad_norm": 0.2994939088821411,
+      "learning_rate": 6.675675675675676e-06,
+      "loss": 0.4552,
+      "step": 247
+    },
+    {
+      "epoch": 0.20146222583265638,
+      "grad_norm": 0.27710896730422974,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.4605,
+      "step": 248
+    },
+    {
+      "epoch": 0.20227457351746547,
+      "grad_norm": 0.2915697395801544,
+      "learning_rate": 6.729729729729731e-06,
+      "loss": 0.4164,
+      "step": 249
+    },
+    {
+      "epoch": 0.20308692120227456,
+      "grad_norm": 0.2697984576225281,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.432,
+      "step": 250
+    },
+    {
+      "epoch": 0.20389926888708368,
+      "grad_norm": 0.2787509858608246,
+      "learning_rate": 6.783783783783784e-06,
+      "loss": 0.4316,
+      "step": 251
+    },
+    {
+      "epoch": 0.20471161657189277,
+      "grad_norm": 0.2442382574081421,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.4312,
+      "step": 252
+    },
+    {
+      "epoch": 0.20552396425670186,
+      "grad_norm": 0.3157908022403717,
+      "learning_rate": 6.837837837837838e-06,
+      "loss": 0.4634,
+      "step": 253
+    },
+    {
+      "epoch": 0.20633631194151097,
+      "grad_norm": 0.24259699881076813,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 0.4367,
+      "step": 254
+    },
+    {
+      "epoch": 0.20714865962632006,
+      "grad_norm": 0.2827763557434082,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.4299,
+      "step": 255
+    },
+    {
+      "epoch": 0.20796100731112915,
+      "grad_norm": 0.26174700260162354,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.4267,
+      "step": 256
+    },
+    {
+      "epoch": 0.20877335499593827,
+      "grad_norm": 0.2505721151828766,
+      "learning_rate": 6.945945945945947e-06,
+      "loss": 0.4357,
+      "step": 257
+    },
+    {
+      "epoch": 0.20958570268074736,
+      "grad_norm": 0.23815499246120453,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.458,
+      "step": 258
+    },
+    {
+      "epoch": 0.21039805036555645,
+      "grad_norm": 0.3047056794166565,
+      "learning_rate": 7e-06,
+      "loss": 0.4285,
+      "step": 259
+    },
+    {
+      "epoch": 0.21121039805036557,
+      "grad_norm": 0.25866857171058655,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.4072,
+      "step": 260
+    },
+    {
+      "epoch": 0.21202274573517466,
+      "grad_norm": 0.2684086561203003,
+      "learning_rate": 7.054054054054055e-06,
+      "loss": 0.4244,
+      "step": 261
+    },
+    {
+      "epoch": 0.21283509341998375,
+      "grad_norm": 0.2664843797683716,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.4558,
+      "step": 262
+    },
+    {
+      "epoch": 0.21364744110479286,
+      "grad_norm": 0.29981517791748047,
+      "learning_rate": 7.1081081081081085e-06,
+      "loss": 0.4447,
+      "step": 263
+    },
+    {
+      "epoch": 0.21445978878960195,
+      "grad_norm": 0.2871261537075043,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.4491,
+      "step": 264
+    },
+    {
+      "epoch": 0.21527213647441104,
+      "grad_norm": 0.26452645659446716,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.4466,
+      "step": 265
+    },
+    {
+      "epoch": 0.21608448415922016,
+      "grad_norm": 0.28123050928115845,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.4356,
+      "step": 266
+    },
+    {
+      "epoch": 0.21689683184402925,
+      "grad_norm": 0.24838213622570038,
+      "learning_rate": 7.216216216216216e-06,
+      "loss": 0.4366,
+      "step": 267
+    },
+    {
+      "epoch": 0.21770917952883834,
+      "grad_norm": 0.2503170967102051,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.4523,
+      "step": 268
+    },
+    {
+      "epoch": 0.21852152721364745,
+      "grad_norm": 0.24902306497097015,
+      "learning_rate": 7.270270270270271e-06,
+      "loss": 0.4428,
+      "step": 269
+    },
+    {
+      "epoch": 0.21933387489845654,
+      "grad_norm": 0.24608734250068665,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.4248,
+      "step": 270
+    },
+    {
+      "epoch": 0.22014622258326563,
+      "grad_norm": 0.2528620660305023,
+      "learning_rate": 7.324324324324325e-06,
+      "loss": 0.4128,
+      "step": 271
+    },
+    {
+      "epoch": 0.22095857026807472,
+      "grad_norm": 0.2830362617969513,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.4335,
+      "step": 272
+    },
+    {
+      "epoch": 0.22177091795288384,
+      "grad_norm": 0.25344064831733704,
+      "learning_rate": 7.3783783783783794e-06,
+      "loss": 0.4379,
+      "step": 273
+    },
+    {
+      "epoch": 0.22258326563769293,
+      "grad_norm": 0.2958681285381317,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.4373,
+      "step": 274
+    },
+    {
+      "epoch": 0.22339561332250202,
+      "grad_norm": 0.2359871119260788,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.4117,
+      "step": 275
+    },
+    {
+      "epoch": 0.22420796100731114,
+      "grad_norm": 0.2195477932691574,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.4269,
+      "step": 276
+    },
+    {
+      "epoch": 0.22502030869212022,
+      "grad_norm": 0.24011681973934174,
+      "learning_rate": 7.486486486486487e-06,
+      "loss": 0.4396,
+      "step": 277
+    },
+    {
+      "epoch": 0.2258326563769293,
+      "grad_norm": 0.24231268465518951,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.4412,
+      "step": 278
+    },
+    {
+      "epoch": 0.22664500406173843,
+      "grad_norm": 0.23785413801670074,
+      "learning_rate": 7.540540540540541e-06,
+      "loss": 0.4217,
+      "step": 279
+    },
+    {
+      "epoch": 0.22745735174654752,
+      "grad_norm": 0.25451719760894775,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.447,
+      "step": 280
+    },
+    {
+      "epoch": 0.2282696994313566,
+      "grad_norm": 0.24388407170772552,
+      "learning_rate": 7.594594594594596e-06,
+      "loss": 0.4605,
+      "step": 281
+    },
+    {
+      "epoch": 0.22908204711616573,
+      "grad_norm": 0.26032689213752747,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.4498,
+      "step": 282
+    },
+    {
+      "epoch": 0.22989439480097482,
+      "grad_norm": 0.23606286942958832,
+      "learning_rate": 7.648648648648649e-06,
+      "loss": 0.4001,
+      "step": 283
+    },
+    {
+      "epoch": 0.2307067424857839,
+      "grad_norm": 0.26948028802871704,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.4481,
+      "step": 284
+    },
+    {
+      "epoch": 0.23151909017059302,
+      "grad_norm": 0.2689862549304962,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.428,
+      "step": 285
+    },
+    {
+      "epoch": 0.2323314378554021,
+      "grad_norm": 0.26609501242637634,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.4389,
+      "step": 286
+    },
+    {
+      "epoch": 0.2331437855402112,
+      "grad_norm": 0.28276529908180237,
+      "learning_rate": 7.756756756756756e-06,
+      "loss": 0.4313,
+      "step": 287
+    },
+    {
+      "epoch": 0.23395613322502032,
+      "grad_norm": 0.236102893948555,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.4254,
+      "step": 288
+    },
+    {
+      "epoch": 0.2347684809098294,
+      "grad_norm": 0.284013569355011,
+      "learning_rate": 7.810810810810812e-06,
+      "loss": 0.4481,
+      "step": 289
+    },
+    {
+      "epoch": 0.2355808285946385,
+      "grad_norm": 0.26729243993759155,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.4294,
+      "step": 290
+    },
+    {
+      "epoch": 0.23639317627944761,
+      "grad_norm": 0.28469952940940857,
+      "learning_rate": 7.864864864864866e-06,
+      "loss": 0.4394,
+      "step": 291
+    },
+    {
+      "epoch": 0.2372055239642567,
+      "grad_norm": 0.2740805745124817,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.4127,
+      "step": 292
+    },
+    {
+      "epoch": 0.2380178716490658,
+      "grad_norm": 0.3010256886482239,
+      "learning_rate": 7.91891891891892e-06,
+      "loss": 0.4439,
+      "step": 293
+    },
+    {
+      "epoch": 0.2388302193338749,
+      "grad_norm": 0.2634214758872986,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.4188,
+      "step": 294
+    },
+    {
+      "epoch": 0.239642567018684,
+      "grad_norm": 0.2585866451263428,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.404,
+      "step": 295
+    },
+    {
+      "epoch": 0.2404549147034931,
+      "grad_norm": 0.2678999900817871,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.4113,
+      "step": 296
+    },
+    {
+      "epoch": 0.2412672623883022,
+      "grad_norm": 0.23457008600234985,
+      "learning_rate": 8.027027027027027e-06,
+      "loss": 0.4301,
+      "step": 297
+    },
+    {
+      "epoch": 0.2420796100731113,
+      "grad_norm": 0.2669728994369507,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.4458,
+      "step": 298
+    },
+    {
+      "epoch": 0.24289195775792038,
+      "grad_norm": 0.26931875944137573,
+      "learning_rate": 8.081081081081081e-06,
+      "loss": 0.4219,
+      "step": 299
+    },
+    {
+      "epoch": 0.2437043054427295,
+      "grad_norm": 0.265720933675766,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.4136,
+      "step": 300
+    },
+    {
+      "epoch": 0.2445166531275386,
+      "grad_norm": 0.28660666942596436,
+      "learning_rate": 8.135135135135137e-06,
+      "loss": 0.4258,
+      "step": 301
+    },
+    {
+      "epoch": 0.24532900081234768,
+      "grad_norm": 0.28213775157928467,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.431,
+      "step": 302
+    },
+    {
+      "epoch": 0.24614134849715677,
+      "grad_norm": 0.29654309153556824,
+      "learning_rate": 8.189189189189189e-06,
+      "loss": 0.4162,
+      "step": 303
+    },
+    {
+      "epoch": 0.2469536961819659,
+      "grad_norm": 0.28635480999946594,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.4191,
+      "step": 304
+    },
+    {
+      "epoch": 0.24776604386677498,
+      "grad_norm": 0.284882515668869,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.4289,
+      "step": 305
+    },
+    {
+      "epoch": 0.24857839155158407,
+      "grad_norm": 0.2642437219619751,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.4159,
+      "step": 306
+    },
+    {
+      "epoch": 0.24939073923639318,
+      "grad_norm": 0.2782119810581207,
+      "learning_rate": 8.297297297297298e-06,
+      "loss": 0.4136,
+      "step": 307
+    },
+    {
+      "epoch": 0.2502030869212023,
+      "grad_norm": 0.2893719971179962,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.4622,
+      "step": 308
+    },
+    {
+      "epoch": 0.2510154346060114,
+      "grad_norm": 0.23060378432273865,
+      "learning_rate": 8.351351351351352e-06,
+      "loss": 0.4174,
+      "step": 309
+    },
+    {
+      "epoch": 0.2518277822908205,
+      "grad_norm": 0.30577003955841064,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.4478,
+      "step": 310
+    },
+    {
+      "epoch": 0.25264012997562957,
+      "grad_norm": 0.2814001441001892,
+      "learning_rate": 8.405405405405406e-06,
+      "loss": 0.4246,
+      "step": 311
+    },
+    {
+      "epoch": 0.25345247766043866,
+      "grad_norm": 0.24841687083244324,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.4135,
+      "step": 312
+    },
+    {
+      "epoch": 0.25426482534524775,
+      "grad_norm": 0.3071884214878082,
+      "learning_rate": 8.45945945945946e-06,
+      "loss": 0.4341,
+      "step": 313
+    },
+    {
+      "epoch": 0.2550771730300569,
+      "grad_norm": 0.2592977285385132,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.3998,
+      "step": 314
+    },
+    {
+      "epoch": 0.255889520714866,
+      "grad_norm": 0.3182254731655121,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.4514,
+      "step": 315
+    },
+    {
+      "epoch": 0.25670186839967507,
+      "grad_norm": 0.24595165252685547,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.4358,
+      "step": 316
+    },
+    {
+      "epoch": 0.25751421608448416,
+      "grad_norm": 0.29478222131729126,
+      "learning_rate": 8.567567567567568e-06,
+      "loss": 0.4517,
+      "step": 317
+    },
+    {
+      "epoch": 0.25832656376929325,
+      "grad_norm": 0.2898155152797699,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.4107,
+      "step": 318
+    },
+    {
+      "epoch": 0.25913891145410234,
+      "grad_norm": 0.2626122534275055,
+      "learning_rate": 8.621621621621622e-06,
+      "loss": 0.42,
+      "step": 319
+    },
+    {
+      "epoch": 0.25995125913891143,
+      "grad_norm": 0.2655840814113617,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.4408,
+      "step": 320
+    },
+    {
+      "epoch": 0.2607636068237206,
+      "grad_norm": 0.31058090925216675,
+      "learning_rate": 8.675675675675677e-06,
+      "loss": 0.4421,
+      "step": 321
+    },
+    {
+      "epoch": 0.26157595450852966,
+      "grad_norm": 0.28188589215278625,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.4458,
+      "step": 322
+    },
+    {
+      "epoch": 0.26238830219333875,
+      "grad_norm": 0.29940065741539,
+      "learning_rate": 8.72972972972973e-06,
+      "loss": 0.4473,
+      "step": 323
+    },
+    {
+      "epoch": 0.26320064987814784,
+      "grad_norm": 0.266001433134079,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.4249,
+      "step": 324
+    },
+    {
+      "epoch": 0.26401299756295693,
+      "grad_norm": 0.26322034001350403,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 0.4066,
+      "step": 325
+    },
+    {
+      "epoch": 0.264825345247766,
+      "grad_norm": 0.3006212115287781,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.433,
+      "step": 326
+    },
+    {
+      "epoch": 0.26563769293257516,
+      "grad_norm": 0.30048561096191406,
+      "learning_rate": 8.837837837837839e-06,
+      "loss": 0.4413,
+      "step": 327
+    },
+    {
+      "epoch": 0.26645004061738425,
+      "grad_norm": 0.3031775653362274,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.4312,
+      "step": 328
+    },
+    {
+      "epoch": 0.26726238830219334,
+      "grad_norm": 0.3145838975906372,
+      "learning_rate": 8.891891891891893e-06,
+      "loss": 0.439,
+      "step": 329
+    },
+    {
+      "epoch": 0.26807473598700243,
+      "grad_norm": 0.2557166516780853,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.4333,
+      "step": 330
+    },
+    {
+      "epoch": 0.2688870836718115,
+      "grad_norm": 0.37087777256965637,
+      "learning_rate": 8.945945945945946e-06,
+      "loss": 0.4393,
+      "step": 331
+    },
+    {
+      "epoch": 0.2696994313566206,
+      "grad_norm": 0.31246912479400635,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.4348,
+      "step": 332
+    },
+    {
+      "epoch": 0.27051177904142976,
+      "grad_norm": 0.30352360010147095,
+      "learning_rate": 9e-06,
+      "loss": 0.4188,
+      "step": 333
+    },
+    {
+      "epoch": 0.27132412672623885,
+      "grad_norm": 0.25454601645469666,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.4121,
+      "step": 334
+    },
+    {
+      "epoch": 0.27213647441104794,
+      "grad_norm": 0.2911500930786133,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.399,
+      "step": 335
+    },
+    {
+      "epoch": 0.272948822095857,
+      "grad_norm": 0.3007078170776367,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.4356,
+      "step": 336
+    },
+    {
+      "epoch": 0.2737611697806661,
+      "grad_norm": 0.2928447425365448,
+      "learning_rate": 9.10810810810811e-06,
+      "loss": 0.4339,
+      "step": 337
+    },
+    {
+      "epoch": 0.2745735174654752,
+      "grad_norm": 0.2975581884384155,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.4227,
+      "step": 338
+    },
+    {
+      "epoch": 0.27538586515028435,
+      "grad_norm": 0.28905797004699707,
+      "learning_rate": 9.162162162162162e-06,
+      "loss": 0.4381,
+      "step": 339
+    },
+    {
+      "epoch": 0.27619821283509344,
+      "grad_norm": 0.2526281774044037,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.4424,
+      "step": 340
+    },
+    {
+      "epoch": 0.2770105605199025,
+      "grad_norm": 0.3195157051086426,
+      "learning_rate": 9.216216216216217e-06,
+      "loss": 0.4339,
+      "step": 341
+    },
+    {
+      "epoch": 0.2778229082047116,
+      "grad_norm": 0.30661073327064514,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.4573,
+      "step": 342
+    },
+    {
+      "epoch": 0.2786352558895207,
+      "grad_norm": 0.2697037160396576,
+      "learning_rate": 9.270270270270271e-06,
+      "loss": 0.438,
+      "step": 343
+    },
+    {
+      "epoch": 0.2794476035743298,
+      "grad_norm": 0.28349992632865906,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.4345,
+      "step": 344
+    },
+    {
+      "epoch": 0.2802599512591389,
+      "grad_norm": 0.28921619057655334,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.4209,
+      "step": 345
+    },
+    {
+      "epoch": 0.28107229894394803,
+      "grad_norm": 0.2733255624771118,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.4254,
+      "step": 346
+    },
+    {
+      "epoch": 0.2818846466287571,
+      "grad_norm": 0.27012360095977783,
+      "learning_rate": 9.378378378378379e-06,
+      "loss": 0.4555,
+      "step": 347
+    },
+    {
+      "epoch": 0.2826969943135662,
+      "grad_norm": 0.27201712131500244,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.4384,
+      "step": 348
+    },
+    {
+      "epoch": 0.2835093419983753,
+      "grad_norm": 0.27284589409828186,
+      "learning_rate": 9.432432432432433e-06,
+      "loss": 0.4079,
+      "step": 349
+    },
+    {
+      "epoch": 0.2843216896831844,
+      "grad_norm": 0.26539018750190735,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.4378,
+      "step": 350
+    },
+    {
+      "epoch": 0.2851340373679935,
+      "grad_norm": 0.2546449303627014,
+      "learning_rate": 9.486486486486487e-06,
+      "loss": 0.4195,
+      "step": 351
+    },
+    {
+      "epoch": 0.2859463850528026,
+      "grad_norm": 0.2594187259674072,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.4145,
+      "step": 352
+    },
+    {
+      "epoch": 0.2867587327376117,
+      "grad_norm": 0.28522640466690063,
+      "learning_rate": 9.540540540540542e-06,
+      "loss": 0.4218,
+      "step": 353
+    },
+    {
+      "epoch": 0.2875710804224208,
+      "grad_norm": 0.30156201124191284,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 0.4286,
+      "step": 354
+    },
+    {
+      "epoch": 0.2883834281072299,
+      "grad_norm": 0.26993638277053833,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.4003,
+      "step": 355
+    },
+    {
+      "epoch": 0.289195775792039,
+      "grad_norm": 0.3354526162147522,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 0.4415,
+      "step": 356
+    },
+    {
+      "epoch": 0.29000812347684807,
+      "grad_norm": 0.2694619596004486,
+      "learning_rate": 9.64864864864865e-06,
+      "loss": 0.418,
+      "step": 357
+    },
+    {
+      "epoch": 0.2908204711616572,
+      "grad_norm": 0.279396116733551,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 0.4026,
+      "step": 358
+    },
+    {
+      "epoch": 0.2916328188464663,
+      "grad_norm": 0.2603386342525482,
+      "learning_rate": 9.702702702702704e-06,
+      "loss": 0.4246,
+      "step": 359
+    },
+    {
+      "epoch": 0.2924451665312754,
+      "grad_norm": 0.2621866762638092,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.4041,
+      "step": 360
+    },
+    {
+      "epoch": 0.2932575142160845,
+      "grad_norm": 0.2781278192996979,
+      "learning_rate": 9.756756756756758e-06,
+      "loss": 0.4442,
+      "step": 361
+    },
+    {
+      "epoch": 0.29406986190089357,
+      "grad_norm": 0.30228888988494873,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 0.4016,
+      "step": 362
+    },
+    {
+      "epoch": 0.29488220958570266,
+      "grad_norm": 0.27156373858451843,
+      "learning_rate": 9.810810810810811e-06,
+      "loss": 0.4089,
+      "step": 363
+    },
+    {
+      "epoch": 0.2956945572705118,
+      "grad_norm": 0.27720990777015686,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 0.4024,
+      "step": 364
+    },
+    {
+      "epoch": 0.2965069049553209,
+      "grad_norm": 0.26680269837379456,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.4294,
+      "step": 365
+    },
+    {
+      "epoch": 0.29731925264013,
+      "grad_norm": 0.31633809208869934,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 0.4416,
+      "step": 366
+    },
+    {
+      "epoch": 0.2981316003249391,
+      "grad_norm": 0.32033756375312805,
+      "learning_rate": 9.91891891891892e-06,
+      "loss": 0.4251,
+      "step": 367
+    },
+    {
+      "epoch": 0.29894394800974816,
+      "grad_norm": 0.2739734947681427,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 0.4147,
+      "step": 368
+    },
+    {
+      "epoch": 0.29975629569455725,
+      "grad_norm": 0.2936696708202362,
+      "learning_rate": 9.972972972972975e-06,
+      "loss": 0.4128,
+      "step": 369
+    },
+    {
+      "epoch": 0.3005686433793664,
+      "grad_norm": 0.26244649291038513,
+      "learning_rate": 1e-05,
+      "loss": 0.4201,
+      "step": 370
+    },
+    {
+      "epoch": 0.3013809910641755,
+      "grad_norm": 0.28551116585731506,
+      "learning_rate": 9.999997765506791e-06,
+      "loss": 0.4301,
+      "step": 371
+    },
+    {
+      "epoch": 0.3021933387489846,
+      "grad_norm": 0.2799893617630005,
+      "learning_rate": 9.99999106202916e-06,
+      "loss": 0.4421,
+      "step": 372
+    },
+    {
+      "epoch": 0.30300568643379366,
+      "grad_norm": 0.2852187156677246,
+      "learning_rate": 9.9999798895731e-06,
+      "loss": 0.4539,
+      "step": 373
+    },
+    {
+      "epoch": 0.30381803411860275,
+      "grad_norm": 0.28459805250167847,
+      "learning_rate": 9.999964248148597e-06,
+      "loss": 0.3947,
+      "step": 374
+    },
+    {
+      "epoch": 0.30463038180341184,
+      "grad_norm": 0.29548877477645874,
+      "learning_rate": 9.999944137769629e-06,
+      "loss": 0.4381,
+      "step": 375
+    },
+    {
+      "epoch": 0.30544272948822093,
+      "grad_norm": 0.30635571479797363,
+      "learning_rate": 9.999919558454171e-06,
+      "loss": 0.3957,
+      "step": 376
+    },
+    {
+      "epoch": 0.3062550771730301,
+      "grad_norm": 0.26227861642837524,
+      "learning_rate": 9.999890510224195e-06,
+      "loss": 0.4019,
+      "step": 377
+    },
+    {
+      "epoch": 0.30706742485783917,
+      "grad_norm": 0.2816251516342163,
+      "learning_rate": 9.999856993105661e-06,
+      "loss": 0.4135,
+      "step": 378
+    },
+    {
+      "epoch": 0.30787977254264826,
+      "grad_norm": 0.27848535776138306,
+      "learning_rate": 9.999819007128527e-06,
+      "loss": 0.4391,
+      "step": 379
+    },
+    {
+      "epoch": 0.30869212022745735,
+      "grad_norm": 0.2927897274494171,
+      "learning_rate": 9.999776552326747e-06,
+      "loss": 0.4293,
+      "step": 380
+    },
+    {
+      "epoch": 0.30950446791226643,
+      "grad_norm": 0.28133273124694824,
+      "learning_rate": 9.999729628738263e-06,
+      "loss": 0.4286,
+      "step": 381
+    },
+    {
+      "epoch": 0.3103168155970755,
+      "grad_norm": 0.26410916447639465,
+      "learning_rate": 9.99967823640502e-06,
+      "loss": 0.4265,
+      "step": 382
+    },
+    {
+      "epoch": 0.31112916328188467,
+      "grad_norm": 0.2903502881526947,
+      "learning_rate": 9.99962237537295e-06,
+      "loss": 0.4319,
+      "step": 383
+    },
+    {
+      "epoch": 0.31194151096669376,
+      "grad_norm": 0.2803366184234619,
+      "learning_rate": 9.999562045691979e-06,
+      "loss": 0.4374,
+      "step": 384
+    },
+    {
+      "epoch": 0.31275385865150285,
+      "grad_norm": 0.28825604915618896,
+      "learning_rate": 9.999497247416035e-06,
+      "loss": 0.4377,
+      "step": 385
+    },
+    {
+      "epoch": 0.31356620633631194,
+      "grad_norm": 0.29552266001701355,
+      "learning_rate": 9.99942798060303e-06,
+      "loss": 0.4333,
+      "step": 386
+    },
+    {
+      "epoch": 0.314378554021121,
+      "grad_norm": 0.2775171995162964,
+      "learning_rate": 9.999354245314874e-06,
+      "loss": 0.4144,
+      "step": 387
+    },
+    {
+      "epoch": 0.3151909017059301,
+      "grad_norm": 0.29196983575820923,
+      "learning_rate": 9.999276041617477e-06,
+      "loss": 0.4072,
+      "step": 388
+    },
+    {
+      "epoch": 0.31600324939073926,
+      "grad_norm": 0.26133307814598083,
+      "learning_rate": 9.999193369580729e-06,
+      "loss": 0.4098,
+      "step": 389
+    },
+    {
+      "epoch": 0.31681559707554835,
+      "grad_norm": 0.26502853631973267,
+      "learning_rate": 9.999106229278531e-06,
+      "loss": 0.4383,
+      "step": 390
+    },
+    {
+      "epoch": 0.31762794476035744,
+      "grad_norm": 0.3308005630970001,
+      "learning_rate": 9.99901462078876e-06,
+      "loss": 0.4239,
+      "step": 391
+    },
+    {
+      "epoch": 0.31844029244516653,
+      "grad_norm": 0.24169456958770752,
+      "learning_rate": 9.998918544193303e-06,
+      "loss": 0.4297,
+      "step": 392
+    },
+    {
+      "epoch": 0.3192526401299756,
+      "grad_norm": 0.2756112515926361,
+      "learning_rate": 9.998817999578028e-06,
+      "loss": 0.4235,
+      "step": 393
+    },
+    {
+      "epoch": 0.3200649878147847,
+      "grad_norm": 0.28787556290626526,
+      "learning_rate": 9.998712987032803e-06,
+      "loss": 0.4329,
+      "step": 394
+    },
+    {
+      "epoch": 0.32087733549959385,
+      "grad_norm": 0.25282177329063416,
+      "learning_rate": 9.99860350665149e-06,
+      "loss": 0.4164,
+      "step": 395
+    },
+    {
+      "epoch": 0.32168968318440294,
+      "grad_norm": 0.28672826290130615,
+      "learning_rate": 9.99848955853194e-06,
+      "loss": 0.3962,
+      "step": 396
+    },
+    {
+      "epoch": 0.32250203086921203,
+      "grad_norm": 0.28003379702568054,
+      "learning_rate": 9.998371142776e-06,
+      "loss": 0.4457,
+      "step": 397
+    },
+    {
+      "epoch": 0.3233143785540211,
+      "grad_norm": 0.3006630837917328,
+      "learning_rate": 9.998248259489509e-06,
+      "loss": 0.4527,
+      "step": 398
+    },
+    {
+      "epoch": 0.3241267262388302,
+      "grad_norm": 0.2724629342556,
+      "learning_rate": 9.998120908782301e-06,
+      "loss": 0.4095,
+      "step": 399
+    },
+    {
+      "epoch": 0.3249390739236393,
+      "grad_norm": 0.2535566985607147,
+      "learning_rate": 9.9979890907682e-06,
+      "loss": 0.3924,
+      "step": 400
+    },
+    {
+      "epoch": 0.3257514216084484,
+      "grad_norm": 0.26009804010391235,
+      "learning_rate": 9.997852805565026e-06,
+      "loss": 0.4478,
+      "step": 401
+    },
+    {
+      "epoch": 0.32656376929325753,
+      "grad_norm": 0.295448899269104,
+      "learning_rate": 9.99771205329459e-06,
+      "loss": 0.4277,
+      "step": 402
+    },
+    {
+      "epoch": 0.3273761169780666,
+      "grad_norm": 0.24158211052417755,
+      "learning_rate": 9.997566834082697e-06,
+      "loss": 0.3941,
+      "step": 403
+    },
+    {
+      "epoch": 0.3281884646628757,
+      "grad_norm": 0.2603279948234558,
+      "learning_rate": 9.997417148059142e-06,
+      "loss": 0.4029,
+      "step": 404
+    },
+    {
+      "epoch": 0.3290008123476848,
+      "grad_norm": 0.2649475336074829,
+      "learning_rate": 9.997262995357714e-06,
+      "loss": 0.4269,
+      "step": 405
+    },
+    {
+      "epoch": 0.3298131600324939,
+      "grad_norm": 0.2880572974681854,
+      "learning_rate": 9.997104376116195e-06,
+      "loss": 0.4407,
+      "step": 406
+    },
+    {
+      "epoch": 0.330625507717303,
+      "grad_norm": 0.31569430232048035,
+      "learning_rate": 9.996941290476359e-06,
+      "loss": 0.4146,
+      "step": 407
+    },
+    {
+      "epoch": 0.3314378554021121,
+      "grad_norm": 0.2620488703250885,
+      "learning_rate": 9.99677373858397e-06,
+      "loss": 0.4089,
+      "step": 408
+    },
+    {
+      "epoch": 0.3322502030869212,
+      "grad_norm": 0.24172626435756683,
+      "learning_rate": 9.996601720588787e-06,
+      "loss": 0.4086,
+      "step": 409
+    },
+    {
+      "epoch": 0.3330625507717303,
+      "grad_norm": 0.26419392228126526,
+      "learning_rate": 9.996425236644558e-06,
+      "loss": 0.4184,
+      "step": 410
+    },
+    {
+      "epoch": 0.3338748984565394,
+      "grad_norm": 0.24281232059001923,
+      "learning_rate": 9.996244286909022e-06,
+      "loss": 0.4146,
+      "step": 411
+    },
+    {
+      "epoch": 0.3346872461413485,
+      "grad_norm": 0.2820408046245575,
+      "learning_rate": 9.996058871543917e-06,
+      "loss": 0.4367,
+      "step": 412
+    },
+    {
+      "epoch": 0.33549959382615757,
+      "grad_norm": 0.25832709670066833,
+      "learning_rate": 9.995868990714963e-06,
+      "loss": 0.4104,
+      "step": 413
+    },
+    {
+      "epoch": 0.3363119415109667,
+      "grad_norm": 0.2737251818180084,
+      "learning_rate": 9.995674644591873e-06,
+      "loss": 0.4205,
+      "step": 414
+    },
+    {
+      "epoch": 0.3371242891957758,
+      "grad_norm": 0.2548855245113373,
+      "learning_rate": 9.995475833348359e-06,
+      "loss": 0.4171,
+      "step": 415
+    },
+    {
+      "epoch": 0.3379366368805849,
+      "grad_norm": 0.2669034004211426,
+      "learning_rate": 9.99527255716211e-06,
+      "loss": 0.4169,
+      "step": 416
+    },
+    {
+      "epoch": 0.338748984565394,
+      "grad_norm": 0.26152005791664124,
+      "learning_rate": 9.99506481621482e-06,
+      "loss": 0.42,
+      "step": 417
+    },
+    {
+      "epoch": 0.3395613322502031,
+      "grad_norm": 0.2688899636268616,
+      "learning_rate": 9.994852610692166e-06,
+      "loss": 0.4163,
+      "step": 418
+    },
+    {
+      "epoch": 0.34037367993501216,
+      "grad_norm": 0.2710811197757721,
+      "learning_rate": 9.994635940783816e-06,
+      "loss": 0.4047,
+      "step": 419
+    },
+    {
+      "epoch": 0.3411860276198213,
+      "grad_norm": 0.2905881702899933,
+      "learning_rate": 9.994414806683429e-06,
+      "loss": 0.4132,
+      "step": 420
+    },
+    {
+      "epoch": 0.3419983753046304,
+      "grad_norm": 0.2758341431617737,
+      "learning_rate": 9.994189208588655e-06,
+      "loss": 0.398,
+      "step": 421
+    },
+    {
+      "epoch": 0.3428107229894395,
+      "grad_norm": 0.31515994668006897,
+      "learning_rate": 9.993959146701128e-06,
+      "loss": 0.4324,
+      "step": 422
+    },
+    {
+      "epoch": 0.3436230706742486,
+      "grad_norm": 0.2666127383708954,
+      "learning_rate": 9.993724621226484e-06,
+      "loss": 0.4414,
+      "step": 423
+    },
+    {
+      "epoch": 0.34443541835905767,
+      "grad_norm": 0.3440514802932739,
+      "learning_rate": 9.993485632374338e-06,
+      "loss": 0.4312,
+      "step": 424
+    },
+    {
+      "epoch": 0.34524776604386676,
+      "grad_norm": 0.2769757807254791,
+      "learning_rate": 9.993242180358298e-06,
+      "loss": 0.4259,
+      "step": 425
+    },
+    {
+      "epoch": 0.3460601137286759,
+      "grad_norm": 0.2823973298072815,
+      "learning_rate": 9.992994265395959e-06,
+      "loss": 0.4367,
+      "step": 426
+    },
+    {
+      "epoch": 0.346872461413485,
+      "grad_norm": 0.2726297080516815,
+      "learning_rate": 9.992741887708908e-06,
+      "loss": 0.4209,
+      "step": 427
+    },
+    {
+      "epoch": 0.3476848090982941,
+      "grad_norm": 0.3441031575202942,
+      "learning_rate": 9.99248504752272e-06,
+      "loss": 0.4118,
+      "step": 428
+    },
+    {
+      "epoch": 0.34849715678310317,
+      "grad_norm": 0.2610268294811249,
+      "learning_rate": 9.992223745066959e-06,
+      "loss": 0.413,
+      "step": 429
+    },
+    {
+      "epoch": 0.34930950446791226,
+      "grad_norm": 0.3228294849395752,
+      "learning_rate": 9.991957980575172e-06,
+      "loss": 0.4203,
+      "step": 430
+    },
+    {
+      "epoch": 0.35012185215272135,
+      "grad_norm": 0.2872535288333893,
+      "learning_rate": 9.991687754284904e-06,
+      "loss": 0.4201,
+      "step": 431
+    },
+    {
+      "epoch": 0.35093419983753044,
+      "grad_norm": 0.2758166491985321,
+      "learning_rate": 9.99141306643768e-06,
+      "loss": 0.4233,
+      "step": 432
+    },
+    {
+      "epoch": 0.3517465475223396,
+      "grad_norm": 0.29926180839538574,
+      "learning_rate": 9.991133917279013e-06,
+      "loss": 0.4191,
+      "step": 433
+    },
+    {
+      "epoch": 0.35255889520714867,
+      "grad_norm": 0.23357714712619781,
+      "learning_rate": 9.990850307058411e-06,
+      "loss": 0.416,
+      "step": 434
+    },
+    {
+      "epoch": 0.35337124289195776,
+      "grad_norm": 0.2523902654647827,
+      "learning_rate": 9.99056223602936e-06,
+      "loss": 0.4274,
+      "step": 435
+    },
+    {
+      "epoch": 0.35418359057676685,
+      "grad_norm": 0.2758367359638214,
+      "learning_rate": 9.990269704449338e-06,
+      "loss": 0.4182,
+      "step": 436
+    },
+    {
+      "epoch": 0.35499593826157594,
+      "grad_norm": 0.2692740261554718,
+      "learning_rate": 9.989972712579809e-06,
+      "loss": 0.431,
+      "step": 437
+    },
+    {
+      "epoch": 0.35580828594638503,
+      "grad_norm": 0.26263153553009033,
+      "learning_rate": 9.989671260686225e-06,
+      "loss": 0.4171,
+      "step": 438
+    },
+    {
+      "epoch": 0.3566206336311942,
+      "grad_norm": 0.2535237669944763,
+      "learning_rate": 9.989365349038019e-06,
+      "loss": 0.4115,
+      "step": 439
+    },
+    {
+      "epoch": 0.35743298131600326,
+      "grad_norm": 0.24168898165225983,
+      "learning_rate": 9.98905497790862e-06,
+      "loss": 0.409,
+      "step": 440
+    },
+    {
+      "epoch": 0.35824532900081235,
+      "grad_norm": 0.2753978967666626,
+      "learning_rate": 9.98874014757543e-06,
+      "loss": 0.4122,
+      "step": 441
+    },
+    {
+      "epoch": 0.35905767668562144,
+      "grad_norm": 0.2642435133457184,
+      "learning_rate": 9.98842085831985e-06,
+      "loss": 0.4017,
+      "step": 442
+    },
+    {
+      "epoch": 0.35987002437043053,
+      "grad_norm": 0.26649826765060425,
+      "learning_rate": 9.988097110427255e-06,
+      "loss": 0.4489,
+      "step": 443
+    },
+    {
+      "epoch": 0.3606823720552396,
+      "grad_norm": 0.2910461127758026,
+      "learning_rate": 9.987768904187014e-06,
+      "loss": 0.4103,
+      "step": 444
+    },
+    {
+      "epoch": 0.36149471974004876,
+      "grad_norm": 0.22695867717266083,
+      "learning_rate": 9.987436239892472e-06,
+      "loss": 0.3995,
+      "step": 445
+    },
+    {
+      "epoch": 0.36230706742485785,
+      "grad_norm": 0.2758156657218933,
+      "learning_rate": 9.987099117840969e-06,
+      "loss": 0.4041,
+      "step": 446
+    },
+    {
+      "epoch": 0.36311941510966694,
+      "grad_norm": 0.2699456214904785,
+      "learning_rate": 9.986757538333817e-06,
+      "loss": 0.4158,
+      "step": 447
+    },
+    {
+      "epoch": 0.36393176279447603,
+      "grad_norm": 0.26062092185020447,
+      "learning_rate": 9.986411501676327e-06,
+      "loss": 0.42,
+      "step": 448
+    },
+    {
+      "epoch": 0.3647441104792851,
+      "grad_norm": 0.2744119167327881,
+      "learning_rate": 9.986061008177779e-06,
+      "loss": 0.4391,
+      "step": 449
+    },
+    {
+      "epoch": 0.3655564581640942,
+      "grad_norm": 0.26661205291748047,
+      "learning_rate": 9.985706058151446e-06,
+      "loss": 0.4286,
+      "step": 450
+    },
+    {
+      "epoch": 0.36636880584890336,
+      "grad_norm": 0.2473769336938858,
+      "learning_rate": 9.985346651914581e-06,
+      "loss": 0.4103,
+      "step": 451
+    },
+    {
+      "epoch": 0.36718115353371245,
+      "grad_norm": 0.2536708414554596,
+      "learning_rate": 9.98498278978842e-06,
+      "loss": 0.4131,
+      "step": 452
+    },
+    {
+      "epoch": 0.36799350121852153,
+      "grad_norm": 0.2866707444190979,
+      "learning_rate": 9.984614472098181e-06,
+      "loss": 0.443,
+      "step": 453
+    },
+    {
+      "epoch": 0.3688058489033306,
+      "grad_norm": 0.25209885835647583,
+      "learning_rate": 9.984241699173069e-06,
+      "loss": 0.4224,
+      "step": 454
+    },
+    {
+      "epoch": 0.3696181965881397,
+      "grad_norm": 0.2892988920211792,
+      "learning_rate": 9.983864471346263e-06,
+      "loss": 0.4218,
+      "step": 455
+    },
+    {
+      "epoch": 0.3704305442729488,
+      "grad_norm": 0.269288033246994,
+      "learning_rate": 9.983482788954931e-06,
+      "loss": 0.4058,
+      "step": 456
+    },
+    {
+      "epoch": 0.37124289195775795,
+      "grad_norm": 0.27840933203697205,
+      "learning_rate": 9.983096652340219e-06,
+      "loss": 0.4123,
+      "step": 457
+    },
+    {
+      "epoch": 0.37205523964256704,
+      "grad_norm": 0.2848956286907196,
+      "learning_rate": 9.982706061847254e-06,
+      "loss": 0.4241,
+      "step": 458
+    },
+    {
+      "epoch": 0.3728675873273761,
+      "grad_norm": 0.28181788325309753,
+      "learning_rate": 9.982311017825145e-06,
+      "loss": 0.3959,
+      "step": 459
+    },
+    {
+      "epoch": 0.3736799350121852,
+      "grad_norm": 0.2536334693431854,
+      "learning_rate": 9.981911520626983e-06,
+      "loss": 0.4256,
+      "step": 460
+    },
+    {
+      "epoch": 0.3744922826969943,
+      "grad_norm": 0.26764601469039917,
+      "learning_rate": 9.981507570609837e-06,
+      "loss": 0.4203,
+      "step": 461
+    },
+    {
+      "epoch": 0.3753046303818034,
+      "grad_norm": 0.2953290045261383,
+      "learning_rate": 9.981099168134753e-06,
+      "loss": 0.4169,
+      "step": 462
+    },
+    {
+      "epoch": 0.3761169780666125,
+      "grad_norm": 0.23687197268009186,
+      "learning_rate": 9.980686313566765e-06,
+      "loss": 0.4147,
+      "step": 463
+    },
+    {
+      "epoch": 0.37692932575142163,
+      "grad_norm": 0.27557820081710815,
+      "learning_rate": 9.980269007274878e-06,
+      "loss": 0.4255,
+      "step": 464
+    },
+    {
+      "epoch": 0.3777416734362307,
+      "grad_norm": 0.26578575372695923,
+      "learning_rate": 9.97984724963208e-06,
+      "loss": 0.4266,
+      "step": 465
+    },
+    {
+      "epoch": 0.3785540211210398,
+      "grad_norm": 0.24763312935829163,
+      "learning_rate": 9.979421041015336e-06,
+      "loss": 0.4023,
+      "step": 466
+    },
+    {
+      "epoch": 0.3793663688058489,
+      "grad_norm": 0.2825430631637573,
+      "learning_rate": 9.978990381805593e-06,
+      "loss": 0.4229,
+      "step": 467
+    },
+    {
+      "epoch": 0.380178716490658,
+      "grad_norm": 0.24730287492275238,
+      "learning_rate": 9.978555272387771e-06,
+      "loss": 0.4229,
+      "step": 468
+    },
+    {
+      "epoch": 0.3809910641754671,
+      "grad_norm": 0.2744419574737549,
+      "learning_rate": 9.978115713150768e-06,
+      "loss": 0.4191,
+      "step": 469
+    },
+    {
+      "epoch": 0.3818034118602762,
+      "grad_norm": 0.26374247670173645,
+      "learning_rate": 9.977671704487465e-06,
+      "loss": 0.4234,
+      "step": 470
+    },
+    {
+      "epoch": 0.3826157595450853,
+      "grad_norm": 0.25933146476745605,
+      "learning_rate": 9.977223246794712e-06,
+      "loss": 0.4212,
+      "step": 471
+    },
+    {
+      "epoch": 0.3834281072298944,
+      "grad_norm": 0.2515645921230316,
+      "learning_rate": 9.97677034047334e-06,
+      "loss": 0.4308,
+      "step": 472
+    },
+    {
+      "epoch": 0.3842404549147035,
+      "grad_norm": 0.25708362460136414,
+      "learning_rate": 9.976312985928159e-06,
+      "loss": 0.413,
+      "step": 473
+    },
+    {
+      "epoch": 0.3850528025995126,
+      "grad_norm": 0.2593640685081482,
+      "learning_rate": 9.975851183567945e-06,
+      "loss": 0.4041,
+      "step": 474
+    },
+    {
+      "epoch": 0.38586515028432167,
+      "grad_norm": 0.25679805874824524,
+      "learning_rate": 9.975384933805461e-06,
+      "loss": 0.4241,
+      "step": 475
+    },
+    {
+      "epoch": 0.3866774979691308,
+      "grad_norm": 0.24845509231090546,
+      "learning_rate": 9.974914237057435e-06,
+      "loss": 0.4225,
+      "step": 476
+    },
+    {
+      "epoch": 0.3874898456539399,
+      "grad_norm": 0.27818039059638977,
+      "learning_rate": 9.974439093744581e-06,
+      "loss": 0.44,
+      "step": 477
+    },
+    {
+      "epoch": 0.388302193338749,
+      "grad_norm": 0.2685849964618683,
+      "learning_rate": 9.973959504291574e-06,
+      "loss": 0.3852,
+      "step": 478
+    },
+    {
+      "epoch": 0.3891145410235581,
+      "grad_norm": 0.276750773191452,
+      "learning_rate": 9.973475469127074e-06,
+      "loss": 0.4059,
+      "step": 479
+    },
+    {
+      "epoch": 0.38992688870836717,
+      "grad_norm": 0.2848278880119324,
+      "learning_rate": 9.972986988683711e-06,
+      "loss": 0.4218,
+      "step": 480
+    },
+    {
+      "epoch": 0.39073923639317626,
+      "grad_norm": 0.2524794638156891,
+      "learning_rate": 9.972494063398083e-06,
+      "loss": 0.3762,
+      "step": 481
+    },
+    {
+      "epoch": 0.3915515840779854,
+      "grad_norm": 0.27972331643104553,
+      "learning_rate": 9.971996693710768e-06,
+      "loss": 0.4192,
+      "step": 482
+    },
+    {
+      "epoch": 0.3923639317627945,
+      "grad_norm": 0.31183263659477234,
+      "learning_rate": 9.971494880066315e-06,
+      "loss": 0.4364,
+      "step": 483
+    },
+    {
+      "epoch": 0.3931762794476036,
+      "grad_norm": 0.22753265500068665,
+      "learning_rate": 9.97098862291324e-06,
+      "loss": 0.4145,
+      "step": 484
+    },
+    {
+      "epoch": 0.39398862713241267,
+      "grad_norm": 0.25443321466445923,
+      "learning_rate": 9.97047792270404e-06,
+      "loss": 0.4259,
+      "step": 485
+    },
+    {
+      "epoch": 0.39480097481722176,
+      "grad_norm": 0.2565366327762604,
+      "learning_rate": 9.969962779895172e-06,
+      "loss": 0.3948,
+      "step": 486
+    },
+    {
+      "epoch": 0.39561332250203085,
+      "grad_norm": 0.2327987402677536,
+      "learning_rate": 9.96944319494707e-06,
+      "loss": 0.4011,
+      "step": 487
+    },
+    {
+      "epoch": 0.39642567018683994,
+      "grad_norm": 0.25321438908576965,
+      "learning_rate": 9.96891916832414e-06,
+      "loss": 0.4093,
+      "step": 488
+    },
+    {
+      "epoch": 0.3972380178716491,
+      "grad_norm": 0.2715853154659271,
+      "learning_rate": 9.968390700494755e-06,
+      "loss": 0.4008,
+      "step": 489
+    },
+    {
+      "epoch": 0.3980503655564582,
+      "grad_norm": 0.25635141134262085,
+      "learning_rate": 9.967857791931257e-06,
+      "loss": 0.3946,
+      "step": 490
+    },
+    {
+      "epoch": 0.39886271324126726,
+      "grad_norm": 0.2403862476348877,
+      "learning_rate": 9.967320443109958e-06,
+      "loss": 0.4063,
+      "step": 491
+    },
+    {
+      "epoch": 0.39967506092607635,
+      "grad_norm": 0.24439266324043274,
+      "learning_rate": 9.966778654511143e-06,
+      "loss": 0.4067,
+      "step": 492
+    },
+    {
+      "epoch": 0.40048740861088544,
+      "grad_norm": 0.2484285831451416,
+      "learning_rate": 9.966232426619054e-06,
+      "loss": 0.4255,
+      "step": 493
+    },
+    {
+      "epoch": 0.40129975629569453,
+      "grad_norm": 0.24619179964065552,
+      "learning_rate": 9.965681759921912e-06,
+      "loss": 0.4183,
+      "step": 494
+    },
+    {
+      "epoch": 0.4021121039805037,
+      "grad_norm": 0.24606020748615265,
+      "learning_rate": 9.965126654911904e-06,
+      "loss": 0.4294,
+      "step": 495
+    },
+    {
+      "epoch": 0.40292445166531277,
+      "grad_norm": 0.2754322588443756,
+      "learning_rate": 9.964567112085175e-06,
+      "loss": 0.4243,
+      "step": 496
+    },
+    {
+      "epoch": 0.40373679935012186,
+      "grad_norm": 0.28790491819381714,
+      "learning_rate": 9.96400313194185e-06,
+      "loss": 0.3902,
+      "step": 497
+    },
+    {
+      "epoch": 0.40454914703493094,
+      "grad_norm": 0.24758823215961456,
+      "learning_rate": 9.963434714986006e-06,
+      "loss": 0.4122,
+      "step": 498
+    },
+    {
+      "epoch": 0.40536149471974003,
+      "grad_norm": 0.24740785360336304,
+      "learning_rate": 9.962861861725698e-06,
+      "loss": 0.3921,
+      "step": 499
+    },
+    {
+      "epoch": 0.4061738424045491,
+      "grad_norm": 0.28198564052581787,
+      "learning_rate": 9.96228457267294e-06,
+      "loss": 0.4183,
+      "step": 500
+    },
+    {
+      "epoch": 0.40698619008935827,
+      "grad_norm": 0.27175086736679077,
+      "learning_rate": 9.961702848343709e-06,
+      "loss": 0.3966,
+      "step": 501
+    },
+    {
+      "epoch": 0.40779853777416736,
+      "grad_norm": 0.25431355834007263,
+      "learning_rate": 9.961116689257949e-06,
+      "loss": 0.421,
+      "step": 502
+    },
+    {
+      "epoch": 0.40861088545897645,
+      "grad_norm": 0.24923264980316162,
+      "learning_rate": 9.960526095939566e-06,
+      "loss": 0.4244,
+      "step": 503
+    },
+    {
+      "epoch": 0.40942323314378554,
+      "grad_norm": 0.2993776798248291,
+      "learning_rate": 9.959931068916436e-06,
+      "loss": 0.4094,
+      "step": 504
+    },
+    {
+      "epoch": 0.4102355808285946,
+      "grad_norm": 0.25844550132751465,
+      "learning_rate": 9.959331608720386e-06,
+      "loss": 0.4212,
+      "step": 505
+    },
+    {
+      "epoch": 0.4110479285134037,
+      "grad_norm": 0.2889086604118347,
+      "learning_rate": 9.958727715887218e-06,
+      "loss": 0.4144,
+      "step": 506
+    },
+    {
+      "epoch": 0.41186027619821286,
+      "grad_norm": 0.27981942892074585,
+      "learning_rate": 9.958119390956685e-06,
+      "loss": 0.4057,
+      "step": 507
+    },
+    {
+      "epoch": 0.41267262388302195,
+      "grad_norm": 0.2966379225254059,
+      "learning_rate": 9.95750663447251e-06,
+      "loss": 0.4214,
+      "step": 508
+    },
+    {
+      "epoch": 0.41348497156783104,
+      "grad_norm": 0.27351200580596924,
+      "learning_rate": 9.956889446982369e-06,
+      "loss": 0.4053,
+      "step": 509
+    },
+    {
+      "epoch": 0.41429731925264013,
+      "grad_norm": 0.2774636447429657,
+      "learning_rate": 9.956267829037906e-06,
+      "loss": 0.4113,
+      "step": 510
+    },
+    {
+      "epoch": 0.4151096669374492,
+      "grad_norm": 0.3242121934890747,
+      "learning_rate": 9.95564178119472e-06,
+      "loss": 0.4221,
+      "step": 511
+    },
+    {
+      "epoch": 0.4159220146222583,
+      "grad_norm": 0.3059741258621216,
+      "learning_rate": 9.95501130401237e-06,
+      "loss": 0.3939,
+      "step": 512
+    },
+    {
+      "epoch": 0.41673436230706745,
+      "grad_norm": 0.28027501702308655,
+      "learning_rate": 9.954376398054377e-06,
+      "loss": 0.4015,
+      "step": 513
+    },
+    {
+      "epoch": 0.41754670999187654,
+      "grad_norm": 0.26384809613227844,
+      "learning_rate": 9.953737063888216e-06,
+      "loss": 0.3767,
+      "step": 514
+    },
+    {
+      "epoch": 0.41835905767668563,
+      "grad_norm": 0.270491361618042,
+      "learning_rate": 9.953093302085324e-06,
+      "loss": 0.4058,
+      "step": 515
+    },
+    {
+      "epoch": 0.4191714053614947,
+      "grad_norm": 0.27940312027931213,
+      "learning_rate": 9.952445113221093e-06,
+      "loss": 0.4244,
+      "step": 516
+    },
+    {
+      "epoch": 0.4199837530463038,
+      "grad_norm": 0.274014949798584,
+      "learning_rate": 9.95179249787487e-06,
+      "loss": 0.4238,
+      "step": 517
+    },
+    {
+      "epoch": 0.4207961007311129,
+      "grad_norm": 0.26008230447769165,
+      "learning_rate": 9.951135456629966e-06,
+      "loss": 0.4053,
+      "step": 518
+    },
+    {
+      "epoch": 0.421608448415922,
+      "grad_norm": 0.2786025404930115,
+      "learning_rate": 9.950473990073637e-06,
+      "loss": 0.4139,
+      "step": 519
+    },
+    {
+      "epoch": 0.42242079610073113,
+      "grad_norm": 0.2866772711277008,
+      "learning_rate": 9.949808098797104e-06,
+      "loss": 0.4204,
+      "step": 520
+    },
+    {
+      "epoch": 0.4232331437855402,
+      "grad_norm": 0.2631932497024536,
+      "learning_rate": 9.949137783395537e-06,
+      "loss": 0.4348,
+      "step": 521
+    },
+    {
+      "epoch": 0.4240454914703493,
+      "grad_norm": 0.27093204855918884,
+      "learning_rate": 9.948463044468063e-06,
+      "loss": 0.406,
+      "step": 522
+    },
+    {
+      "epoch": 0.4248578391551584,
+      "grad_norm": 0.2539540231227875,
+      "learning_rate": 9.94778388261776e-06,
+      "loss": 0.4054,
+      "step": 523
+    },
+    {
+      "epoch": 0.4256701868399675,
+      "grad_norm": 0.25177833437919617,
+      "learning_rate": 9.947100298451663e-06,
+      "loss": 0.4094,
+      "step": 524
+    },
+    {
+      "epoch": 0.4264825345247766,
+      "grad_norm": 0.26316455006599426,
+      "learning_rate": 9.946412292580757e-06,
+      "loss": 0.4011,
+      "step": 525
+    },
+    {
+      "epoch": 0.4272948822095857,
+      "grad_norm": 0.24829813838005066,
+      "learning_rate": 9.94571986561998e-06,
+      "loss": 0.401,
+      "step": 526
+    },
+    {
+      "epoch": 0.4281072298943948,
+      "grad_norm": 0.25562816858291626,
+      "learning_rate": 9.945023018188222e-06,
+      "loss": 0.4027,
+      "step": 527
+    },
+    {
+      "epoch": 0.4289195775792039,
+      "grad_norm": 0.24581240117549896,
+      "learning_rate": 9.944321750908321e-06,
+      "loss": 0.4301,
+      "step": 528
+    },
+    {
+      "epoch": 0.429731925264013,
+      "grad_norm": 0.27080461382865906,
+      "learning_rate": 9.94361606440707e-06,
+      "loss": 0.4439,
+      "step": 529
+    },
+    {
+      "epoch": 0.4305442729488221,
+      "grad_norm": 0.2321961522102356,
+      "learning_rate": 9.94290595931521e-06,
+      "loss": 0.4011,
+      "step": 530
+    },
+    {
+      "epoch": 0.43135662063363117,
+      "grad_norm": 0.27483803033828735,
+      "learning_rate": 9.942191436267428e-06,
+      "loss": 0.4125,
+      "step": 531
+    },
+    {
+      "epoch": 0.4321689683184403,
+      "grad_norm": 0.27694573998451233,
+      "learning_rate": 9.941472495902366e-06,
+      "loss": 0.4424,
+      "step": 532
+    },
+    {
+      "epoch": 0.4329813160032494,
+      "grad_norm": 0.2753378450870514,
+      "learning_rate": 9.940749138862609e-06,
+      "loss": 0.4315,
+      "step": 533
+    },
+    {
+      "epoch": 0.4337936636880585,
+      "grad_norm": 0.2639577090740204,
+      "learning_rate": 9.940021365794692e-06,
+      "loss": 0.4057,
+      "step": 534
+    },
+    {
+      "epoch": 0.4346060113728676,
+      "grad_norm": 0.23245272040367126,
+      "learning_rate": 9.939289177349097e-06,
+      "loss": 0.4076,
+      "step": 535
+    },
+    {
+      "epoch": 0.4354183590576767,
+      "grad_norm": 0.26850783824920654,
+      "learning_rate": 9.938552574180252e-06,
+      "loss": 0.4315,
+      "step": 536
+    },
+    {
+      "epoch": 0.43623070674248576,
+      "grad_norm": 0.25538551807403564,
+      "learning_rate": 9.937811556946531e-06,
+      "loss": 0.4202,
+      "step": 537
+    },
+    {
+      "epoch": 0.4370430544272949,
+      "grad_norm": 0.27607282996177673,
+      "learning_rate": 9.937066126310253e-06,
+      "loss": 0.4331,
+      "step": 538
+    },
+    {
+      "epoch": 0.437855402112104,
+      "grad_norm": 0.26058897376060486,
+      "learning_rate": 9.936316282937682e-06,
+      "loss": 0.3951,
+      "step": 539
+    },
+    {
+      "epoch": 0.4386677497969131,
+      "grad_norm": 0.2440854012966156,
+      "learning_rate": 9.935562027499026e-06,
+      "loss": 0.4093,
+      "step": 540
+    },
+    {
+      "epoch": 0.4394800974817222,
+      "grad_norm": 0.24996644258499146,
+      "learning_rate": 9.934803360668435e-06,
+      "loss": 0.4202,
+      "step": 541
+    },
+    {
+      "epoch": 0.44029244516653127,
+      "grad_norm": 0.2503606677055359,
+      "learning_rate": 9.934040283124006e-06,
+      "loss": 0.3833,
+      "step": 542
+    },
+    {
+      "epoch": 0.44110479285134035,
+      "grad_norm": 0.23675158619880676,
+      "learning_rate": 9.933272795547773e-06,
+      "loss": 0.392,
+      "step": 543
+    },
+    {
+      "epoch": 0.44191714053614944,
+      "grad_norm": 0.27069026231765747,
+      "learning_rate": 9.932500898625716e-06,
+      "loss": 0.4321,
+      "step": 544
+    },
+    {
+      "epoch": 0.4427294882209586,
+      "grad_norm": 0.2326032668352127,
+      "learning_rate": 9.931724593047754e-06,
+      "loss": 0.4344,
+      "step": 545
+    },
+    {
+      "epoch": 0.4435418359057677,
+      "grad_norm": 0.2735949158668518,
+      "learning_rate": 9.930943879507748e-06,
+      "loss": 0.4109,
+      "step": 546
+    },
+    {
+      "epoch": 0.44435418359057677,
+      "grad_norm": 0.26179638504981995,
+      "learning_rate": 9.930158758703495e-06,
+      "loss": 0.3942,
+      "step": 547
+    },
+    {
+      "epoch": 0.44516653127538586,
+      "grad_norm": 0.24375368654727936,
+      "learning_rate": 9.929369231336735e-06,
+      "loss": 0.4019,
+      "step": 548
+    },
+    {
+      "epoch": 0.44597887896019495,
+      "grad_norm": 0.24633844196796417,
+      "learning_rate": 9.928575298113146e-06,
+      "loss": 0.3892,
+      "step": 549
+    },
+    {
+      "epoch": 0.44679122664500404,
+      "grad_norm": 0.24481603503227234,
+      "learning_rate": 9.927776959742344e-06,
+      "loss": 0.4023,
+      "step": 550
+    },
+    {
+      "epoch": 0.4476035743298132,
+      "grad_norm": 0.27662375569343567,
+      "learning_rate": 9.92697421693788e-06,
+      "loss": 0.4081,
+      "step": 551
+    },
+    {
+      "epoch": 0.44841592201462227,
+      "grad_norm": 0.2498617321252823,
+      "learning_rate": 9.926167070417243e-06,
+      "loss": 0.4064,
+      "step": 552
+    },
+    {
+      "epoch": 0.44922826969943136,
+      "grad_norm": 0.2542991042137146,
+      "learning_rate": 9.925355520901861e-06,
+      "loss": 0.3985,
+      "step": 553
+    },
+    {
+      "epoch": 0.45004061738424045,
+      "grad_norm": 0.26317328214645386,
+      "learning_rate": 9.924539569117092e-06,
+      "loss": 0.4134,
+      "step": 554
+    },
+    {
+      "epoch": 0.45085296506904954,
+      "grad_norm": 0.2610620856285095,
+      "learning_rate": 9.923719215792233e-06,
+      "loss": 0.4095,
+      "step": 555
+    },
+    {
+      "epoch": 0.4516653127538586,
+      "grad_norm": 0.27502676844596863,
+      "learning_rate": 9.922894461660513e-06,
+      "loss": 0.389,
+      "step": 556
+    },
+    {
+      "epoch": 0.4524776604386678,
+      "grad_norm": 0.23830968141555786,
+      "learning_rate": 9.922065307459096e-06,
+      "loss": 0.3977,
+      "step": 557
+    },
+    {
+      "epoch": 0.45329000812347686,
+      "grad_norm": 0.26066499948501587,
+      "learning_rate": 9.921231753929077e-06,
+      "loss": 0.4165,
+      "step": 558
+    },
+    {
+      "epoch": 0.45410235580828595,
+      "grad_norm": 0.27878984808921814,
+      "learning_rate": 9.920393801815482e-06,
+      "loss": 0.4163,
+      "step": 559
+    },
+    {
+      "epoch": 0.45491470349309504,
+      "grad_norm": 0.24139316380023956,
+      "learning_rate": 9.919551451867275e-06,
+      "loss": 0.4026,
+      "step": 560
+    },
+    {
+      "epoch": 0.45572705117790413,
+      "grad_norm": 0.2590111196041107,
+      "learning_rate": 9.91870470483734e-06,
+      "loss": 0.4095,
+      "step": 561
+    },
+    {
+      "epoch": 0.4565393988627132,
+      "grad_norm": 0.2522757053375244,
+      "learning_rate": 9.917853561482503e-06,
+      "loss": 0.4206,
+      "step": 562
+    },
+    {
+      "epoch": 0.45735174654752236,
+      "grad_norm": 0.25368309020996094,
+      "learning_rate": 9.91699802256351e-06,
+      "loss": 0.4069,
+      "step": 563
+    },
+    {
+      "epoch": 0.45816409423233145,
+      "grad_norm": 0.2389921396970749,
+      "learning_rate": 9.916138088845042e-06,
+      "loss": 0.3946,
+      "step": 564
+    },
+    {
+      "epoch": 0.45897644191714054,
+      "grad_norm": 0.2674838900566101,
+      "learning_rate": 9.915273761095701e-06,
+      "loss": 0.3991,
+      "step": 565
+    },
+    {
+      "epoch": 0.45978878960194963,
+      "grad_norm": 0.2535633444786072,
+      "learning_rate": 9.914405040088026e-06,
+      "loss": 0.4059,
+      "step": 566
+    },
+    {
+      "epoch": 0.4606011372867587,
+      "grad_norm": 0.26305484771728516,
+      "learning_rate": 9.913531926598475e-06,
+      "loss": 0.4116,
+      "step": 567
+    },
+    {
+      "epoch": 0.4614134849715678,
+      "grad_norm": 0.2504863739013672,
+      "learning_rate": 9.912654421407434e-06,
+      "loss": 0.4211,
+      "step": 568
+    },
+    {
+      "epoch": 0.46222583265637696,
+      "grad_norm": 0.24148602783679962,
+      "learning_rate": 9.911772525299214e-06,
+      "loss": 0.4216,
+      "step": 569
+    },
+    {
+      "epoch": 0.46303818034118605,
+      "grad_norm": 0.24740095436573029,
+      "learning_rate": 9.910886239062054e-06,
+      "loss": 0.4379,
+      "step": 570
+    },
+    {
+      "epoch": 0.46385052802599513,
+      "grad_norm": 0.28016653656959534,
+      "learning_rate": 9.909995563488116e-06,
+      "loss": 0.4247,
+      "step": 571
+    },
+    {
+      "epoch": 0.4646628757108042,
+      "grad_norm": 0.2479318529367447,
+      "learning_rate": 9.909100499373476e-06,
+      "loss": 0.3901,
+      "step": 572
+    },
+    {
+      "epoch": 0.4654752233956133,
+      "grad_norm": 0.2594188451766968,
+      "learning_rate": 9.908201047518146e-06,
+      "loss": 0.4055,
+      "step": 573
+    },
+    {
+      "epoch": 0.4662875710804224,
+      "grad_norm": 0.26496875286102295,
+      "learning_rate": 9.907297208726051e-06,
+      "loss": 0.4205,
+      "step": 574
+    },
+    {
+      "epoch": 0.4670999187652315,
+      "grad_norm": 0.2605689465999603,
+      "learning_rate": 9.90638898380504e-06,
+      "loss": 0.4089,
+      "step": 575
+    },
+    {
+      "epoch": 0.46791226645004064,
+      "grad_norm": 0.27707409858703613,
+      "learning_rate": 9.905476373566884e-06,
+      "loss": 0.4206,
+      "step": 576
+    },
+    {
+      "epoch": 0.4687246141348497,
+      "grad_norm": 0.25814974308013916,
+      "learning_rate": 9.90455937882727e-06,
+      "loss": 0.3971,
+      "step": 577
+    },
+    {
+      "epoch": 0.4695369618196588,
+      "grad_norm": 0.28311601281166077,
+      "learning_rate": 9.903638000405805e-06,
+      "loss": 0.4285,
+      "step": 578
+    },
+    {
+      "epoch": 0.4703493095044679,
+      "grad_norm": 0.25514543056488037,
+      "learning_rate": 9.902712239126016e-06,
+      "loss": 0.3924,
+      "step": 579
+    },
+    {
+      "epoch": 0.471161657189277,
+      "grad_norm": 0.25762978196144104,
+      "learning_rate": 9.901782095815343e-06,
+      "loss": 0.4021,
+      "step": 580
+    },
+    {
+      "epoch": 0.4719740048740861,
+      "grad_norm": 0.2764049470424652,
+      "learning_rate": 9.900847571305148e-06,
+      "loss": 0.391,
+      "step": 581
+    },
+    {
+      "epoch": 0.47278635255889523,
+      "grad_norm": 0.27375152707099915,
+      "learning_rate": 9.899908666430707e-06,
+      "loss": 0.4285,
+      "step": 582
+    },
+    {
+      "epoch": 0.4735987002437043,
+      "grad_norm": 0.24319571256637573,
+      "learning_rate": 9.898965382031208e-06,
+      "loss": 0.3897,
+      "step": 583
+    },
+    {
+      "epoch": 0.4744110479285134,
+      "grad_norm": 0.2610553801059723,
+      "learning_rate": 9.898017718949759e-06,
+      "loss": 0.3943,
+      "step": 584
+    },
+    {
+      "epoch": 0.4752233956133225,
+      "grad_norm": 0.2556527554988861,
+      "learning_rate": 9.897065678033376e-06,
+      "loss": 0.4255,
+      "step": 585
+    },
+    {
+      "epoch": 0.4760357432981316,
+      "grad_norm": 0.24279142916202545,
+      "learning_rate": 9.896109260132993e-06,
+      "loss": 0.3908,
+      "step": 586
+    },
+    {
+      "epoch": 0.4768480909829407,
+      "grad_norm": 0.2683797776699066,
+      "learning_rate": 9.895148466103451e-06,
+      "loss": 0.4337,
+      "step": 587
+    },
+    {
+      "epoch": 0.4776604386677498,
+      "grad_norm": 0.24958230555057526,
+      "learning_rate": 9.894183296803509e-06,
+      "loss": 0.4004,
+      "step": 588
+    },
+    {
+      "epoch": 0.4784727863525589,
+      "grad_norm": 0.238287553191185,
+      "learning_rate": 9.893213753095829e-06,
+      "loss": 0.4029,
+      "step": 589
+    },
+    {
+      "epoch": 0.479285134037368,
+      "grad_norm": 0.25898510217666626,
+      "learning_rate": 9.892239835846988e-06,
+      "loss": 0.404,
+      "step": 590
+    },
+    {
+      "epoch": 0.4800974817221771,
+      "grad_norm": 0.2502025067806244,
+      "learning_rate": 9.891261545927469e-06,
+      "loss": 0.4075,
+      "step": 591
+    },
+    {
+      "epoch": 0.4809098294069862,
+      "grad_norm": 0.25418946146965027,
+      "learning_rate": 9.890278884211668e-06,
+      "loss": 0.4237,
+      "step": 592
+    },
+    {
+      "epoch": 0.48172217709179527,
+      "grad_norm": 0.2660810351371765,
+      "learning_rate": 9.889291851577885e-06,
+      "loss": 0.4114,
+      "step": 593
+    },
+    {
+      "epoch": 0.4825345247766044,
+      "grad_norm": 0.26168158650398254,
+      "learning_rate": 9.888300448908324e-06,
+      "loss": 0.4126,
+      "step": 594
+    },
+    {
+      "epoch": 0.4833468724614135,
+      "grad_norm": 0.2555215656757355,
+      "learning_rate": 9.8873046770891e-06,
+      "loss": 0.4183,
+      "step": 595
+    },
+    {
+      "epoch": 0.4841592201462226,
+      "grad_norm": 0.23422159254550934,
+      "learning_rate": 9.88630453701023e-06,
+      "loss": 0.3901,
+      "step": 596
+    },
+    {
+      "epoch": 0.4849715678310317,
+      "grad_norm": 0.24864648282527924,
+      "learning_rate": 9.88530002956564e-06,
+      "loss": 0.4287,
+      "step": 597
+    },
+    {
+      "epoch": 0.48578391551584077,
+      "grad_norm": 0.23576724529266357,
+      "learning_rate": 9.884291155653151e-06,
+      "loss": 0.377,
+      "step": 598
+    },
+    {
+      "epoch": 0.48659626320064986,
+      "grad_norm": 0.25676581263542175,
+      "learning_rate": 9.883277916174496e-06,
+      "loss": 0.3984,
+      "step": 599
+    },
+    {
+      "epoch": 0.487408610885459,
+      "grad_norm": 0.2806536555290222,
+      "learning_rate": 9.882260312035304e-06,
+      "loss": 0.4044,
+      "step": 600
+    },
+    {
+      "epoch": 0.4882209585702681,
+      "grad_norm": 0.2329963892698288,
+      "learning_rate": 9.881238344145106e-06,
+      "loss": 0.3941,
+      "step": 601
+    },
+    {
+      "epoch": 0.4890333062550772,
+      "grad_norm": 0.23680846393108368,
+      "learning_rate": 9.880212013417334e-06,
+      "loss": 0.403,
+      "step": 602
+    },
+    {
+      "epoch": 0.48984565393988627,
+      "grad_norm": 0.2460140436887741,
+      "learning_rate": 9.879181320769322e-06,
+      "loss": 0.3855,
+      "step": 603
+    },
+    {
+      "epoch": 0.49065800162469536,
+      "grad_norm": 0.25164610147476196,
+      "learning_rate": 9.878146267122298e-06,
+      "loss": 0.4164,
+      "step": 604
+    },
+    {
+      "epoch": 0.49147034930950445,
+      "grad_norm": 0.2687186598777771,
+      "learning_rate": 9.877106853401392e-06,
+      "loss": 0.4071,
+      "step": 605
+    },
+    {
+      "epoch": 0.49228269699431354,
+      "grad_norm": 0.2734622359275818,
+      "learning_rate": 9.876063080535627e-06,
+      "loss": 0.4111,
+      "step": 606
+    },
+    {
+      "epoch": 0.4930950446791227,
+      "grad_norm": 0.26818451285362244,
+      "learning_rate": 9.875014949457926e-06,
+      "loss": 0.437,
+      "step": 607
+    },
+    {
+      "epoch": 0.4939073923639318,
+      "grad_norm": 0.25892549753189087,
+      "learning_rate": 9.873962461105104e-06,
+      "loss": 0.3977,
+      "step": 608
+    },
+    {
+      "epoch": 0.49471974004874086,
+      "grad_norm": 0.25586381554603577,
+      "learning_rate": 9.872905616417875e-06,
+      "loss": 0.383,
+      "step": 609
+    },
+    {
+      "epoch": 0.49553208773354995,
+      "grad_norm": 0.2599838078022003,
+      "learning_rate": 9.87184441634084e-06,
+      "loss": 0.4053,
+      "step": 610
+    },
+    {
+      "epoch": 0.49634443541835904,
+      "grad_norm": 0.2777315080165863,
+      "learning_rate": 9.870778861822502e-06,
+      "loss": 0.3984,
+      "step": 611
+    },
+    {
+      "epoch": 0.49715678310316813,
+      "grad_norm": 0.29982057213783264,
+      "learning_rate": 9.869708953815247e-06,
+      "loss": 0.403,
+      "step": 612
+    },
+    {
+      "epoch": 0.4979691307879773,
+      "grad_norm": 0.24641016125679016,
+      "learning_rate": 9.868634693275356e-06,
+      "loss": 0.4015,
+      "step": 613
+    },
+    {
+      "epoch": 0.49878147847278637,
+      "grad_norm": 0.2683729827404022,
+      "learning_rate": 9.867556081163002e-06,
+      "loss": 0.3857,
+      "step": 614
+    },
+    {
+      "epoch": 0.49959382615759546,
+      "grad_norm": 0.2607569992542267,
+      "learning_rate": 9.86647311844224e-06,
+      "loss": 0.3861,
+      "step": 615
+    },
+    {
+      "epoch": 0.5004061738424046,
+      "grad_norm": 0.2845657467842102,
+      "learning_rate": 9.865385806081028e-06,
+      "loss": 0.3897,
+      "step": 616
+    },
+    {
+      "epoch": 0.5012185215272137,
+      "grad_norm": 0.26003769040107727,
+      "learning_rate": 9.864294145051197e-06,
+      "loss": 0.4081,
+      "step": 617
+    },
+    {
+      "epoch": 0.5020308692120228,
+      "grad_norm": 0.2980784475803375,
+      "learning_rate": 9.863198136328474e-06,
+      "loss": 0.4107,
+      "step": 618
+    },
+    {
+      "epoch": 0.5028432168968319,
+      "grad_norm": 0.3123629093170166,
+      "learning_rate": 9.862097780892463e-06,
+      "loss": 0.4174,
+      "step": 619
+    },
+    {
+      "epoch": 0.503655564581641,
+      "grad_norm": 0.25948095321655273,
+      "learning_rate": 9.860993079726665e-06,
+      "loss": 0.3976,
+      "step": 620
+    },
+    {
+      "epoch": 0.50446791226645,
+      "grad_norm": 0.3188088536262512,
+      "learning_rate": 9.859884033818454e-06,
+      "loss": 0.4308,
+      "step": 621
+    },
+    {
+      "epoch": 0.5052802599512591,
+      "grad_norm": 0.28681254386901855,
+      "learning_rate": 9.858770644159098e-06,
+      "loss": 0.4095,
+      "step": 622
+    },
+    {
+      "epoch": 0.5060926076360682,
+      "grad_norm": 0.26227912306785583,
+      "learning_rate": 9.857652911743734e-06,
+      "loss": 0.409,
+      "step": 623
+    },
+    {
+      "epoch": 0.5069049553208773,
+      "grad_norm": 0.23472528159618378,
+      "learning_rate": 9.856530837571394e-06,
+      "loss": 0.4102,
+      "step": 624
+    },
+    {
+      "epoch": 0.5077173030056864,
+      "grad_norm": 0.24257154762744904,
+      "learning_rate": 9.855404422644983e-06,
+      "loss": 0.388,
+      "step": 625
+    },
+    {
+      "epoch": 0.5085296506904955,
+      "grad_norm": 0.251436710357666,
+      "learning_rate": 9.85427366797129e-06,
+      "loss": 0.4356,
+      "step": 626
+    },
+    {
+      "epoch": 0.5093419983753046,
+      "grad_norm": 0.25017985701560974,
+      "learning_rate": 9.853138574560976e-06,
+      "loss": 0.4025,
+      "step": 627
+    },
+    {
+      "epoch": 0.5101543460601138,
+      "grad_norm": 0.24616345763206482,
+      "learning_rate": 9.851999143428587e-06,
+      "loss": 0.3894,
+      "step": 628
+    },
+    {
+      "epoch": 0.5109666937449229,
+      "grad_norm": 0.2762754261493683,
+      "learning_rate": 9.850855375592543e-06,
+      "loss": 0.3883,
+      "step": 629
+    },
+    {
+      "epoch": 0.511779041429732,
+      "grad_norm": 0.2599327266216278,
+      "learning_rate": 9.84970727207514e-06,
+      "loss": 0.3897,
+      "step": 630
+    },
+    {
+      "epoch": 0.512591389114541,
+      "grad_norm": 0.25083673000335693,
+      "learning_rate": 9.848554833902551e-06,
+      "loss": 0.4238,
+      "step": 631
+    },
+    {
+      "epoch": 0.5134037367993501,
+      "grad_norm": 0.27335479855537415,
+      "learning_rate": 9.847398062104823e-06,
+      "loss": 0.4224,
+      "step": 632
+    },
+    {
+      "epoch": 0.5142160844841592,
+      "grad_norm": 0.25403791666030884,
+      "learning_rate": 9.846236957715872e-06,
+      "loss": 0.4082,
+      "step": 633
+    },
+    {
+      "epoch": 0.5150284321689683,
+      "grad_norm": 0.23988258838653564,
+      "learning_rate": 9.845071521773492e-06,
+      "loss": 0.4126,
+      "step": 634
+    },
+    {
+      "epoch": 0.5158407798537774,
+      "grad_norm": 0.2418673038482666,
+      "learning_rate": 9.843901755319347e-06,
+      "loss": 0.414,
+      "step": 635
+    },
+    {
+      "epoch": 0.5166531275385865,
+      "grad_norm": 0.25373420119285583,
+      "learning_rate": 9.842727659398971e-06,
+      "loss": 0.4078,
+      "step": 636
+    },
+    {
+      "epoch": 0.5174654752233956,
+      "grad_norm": 0.22301417589187622,
+      "learning_rate": 9.841549235061766e-06,
+      "loss": 0.3921,
+      "step": 637
+    },
+    {
+      "epoch": 0.5182778229082047,
+      "grad_norm": 0.27320346236228943,
+      "learning_rate": 9.840366483361008e-06,
+      "loss": 0.4126,
+      "step": 638
+    },
+    {
+      "epoch": 0.5190901705930138,
+      "grad_norm": 0.26155176758766174,
+      "learning_rate": 9.839179405353834e-06,
+      "loss": 0.4316,
+      "step": 639
+    },
+    {
+      "epoch": 0.5199025182778229,
+      "grad_norm": 0.2663276195526123,
+      "learning_rate": 9.837988002101253e-06,
+      "loss": 0.4113,
+      "step": 640
+    },
+    {
+      "epoch": 0.5207148659626321,
+      "grad_norm": 0.2423909455537796,
+      "learning_rate": 9.836792274668135e-06,
+      "loss": 0.3805,
+      "step": 641
+    },
+    {
+      "epoch": 0.5215272136474411,
+      "grad_norm": 0.26877593994140625,
+      "learning_rate": 9.835592224123221e-06,
+      "loss": 0.4171,
+      "step": 642
+    },
+    {
+      "epoch": 0.5223395613322502,
+      "grad_norm": 0.22546301782131195,
+      "learning_rate": 9.834387851539113e-06,
+      "loss": 0.4166,
+      "step": 643
+    },
+    {
+      "epoch": 0.5231519090170593,
+      "grad_norm": 0.26238125562667847,
+      "learning_rate": 9.833179157992274e-06,
+      "loss": 0.4172,
+      "step": 644
+    },
+    {
+      "epoch": 0.5239642567018684,
+      "grad_norm": 0.25171327590942383,
+      "learning_rate": 9.831966144563032e-06,
+      "loss": 0.4014,
+      "step": 645
+    },
+    {
+      "epoch": 0.5247766043866775,
+      "grad_norm": 0.247896209359169,
+      "learning_rate": 9.830748812335576e-06,
+      "loss": 0.4324,
+      "step": 646
+    },
+    {
+      "epoch": 0.5255889520714866,
+      "grad_norm": 0.2476065754890442,
+      "learning_rate": 9.829527162397951e-06,
+      "loss": 0.4049,
+      "step": 647
+    },
+    {
+      "epoch": 0.5264012997562957,
+      "grad_norm": 0.23791800439357758,
+      "learning_rate": 9.828301195842069e-06,
+      "loss": 0.4196,
+      "step": 648
+    },
+    {
+      "epoch": 0.5272136474411048,
+      "grad_norm": 0.24675987660884857,
+      "learning_rate": 9.827070913763693e-06,
+      "loss": 0.4175,
+      "step": 649
+    },
+    {
+      "epoch": 0.5280259951259139,
+      "grad_norm": 0.2515602111816406,
+      "learning_rate": 9.825836317262445e-06,
+      "loss": 0.3878,
+      "step": 650
+    },
+    {
+      "epoch": 0.528838342810723,
+      "grad_norm": 0.24894502758979797,
+      "learning_rate": 9.824597407441805e-06,
+      "loss": 0.4122,
+      "step": 651
+    },
+    {
+      "epoch": 0.529650690495532,
+      "grad_norm": 0.24030950665473938,
+      "learning_rate": 9.823354185409108e-06,
+      "loss": 0.402,
+      "step": 652
+    },
+    {
+      "epoch": 0.5304630381803412,
+      "grad_norm": 0.24882569909095764,
+      "learning_rate": 9.822106652275541e-06,
+      "loss": 0.4345,
+      "step": 653
+    },
+    {
+      "epoch": 0.5312753858651503,
+      "grad_norm": 0.2692911624908447,
+      "learning_rate": 9.820854809156146e-06,
+      "loss": 0.3836,
+      "step": 654
+    },
+    {
+      "epoch": 0.5320877335499594,
+      "grad_norm": 0.2759972810745239,
+      "learning_rate": 9.819598657169819e-06,
+      "loss": 0.4288,
+      "step": 655
+    },
+    {
+      "epoch": 0.5329000812347685,
+      "grad_norm": 0.24164672195911407,
+      "learning_rate": 9.818338197439304e-06,
+      "loss": 0.403,
+      "step": 656
+    },
+    {
+      "epoch": 0.5337124289195776,
+      "grad_norm": 0.28618383407592773,
+      "learning_rate": 9.817073431091194e-06,
+      "loss": 0.424,
+      "step": 657
+    },
+    {
+      "epoch": 0.5345247766043867,
+      "grad_norm": 0.28754279017448425,
+      "learning_rate": 9.815804359255937e-06,
+      "loss": 0.4003,
+      "step": 658
+    },
+    {
+      "epoch": 0.5353371242891958,
+      "grad_norm": 0.2547610104084015,
+      "learning_rate": 9.814530983067825e-06,
+      "loss": 0.4127,
+      "step": 659
+    },
+    {
+      "epoch": 0.5361494719740049,
+      "grad_norm": 0.27454274892807007,
+      "learning_rate": 9.813253303664997e-06,
+      "loss": 0.3883,
+      "step": 660
+    },
+    {
+      "epoch": 0.536961819658814,
+      "grad_norm": 0.28120797872543335,
+      "learning_rate": 9.811971322189442e-06,
+      "loss": 0.4021,
+      "step": 661
+    },
+    {
+      "epoch": 0.537774167343623,
+      "grad_norm": 0.23853729665279388,
+      "learning_rate": 9.810685039786989e-06,
+      "loss": 0.4037,
+      "step": 662
+    },
+    {
+      "epoch": 0.5385865150284321,
+      "grad_norm": 0.2646525800228119,
+      "learning_rate": 9.809394457607315e-06,
+      "loss": 0.4136,
+      "step": 663
+    },
+    {
+      "epoch": 0.5393988627132412,
+      "grad_norm": 0.2579787075519562,
+      "learning_rate": 9.808099576803937e-06,
+      "loss": 0.3988,
+      "step": 664
+    },
+    {
+      "epoch": 0.5402112103980503,
+      "grad_norm": 0.2707813084125519,
+      "learning_rate": 9.80680039853422e-06,
+      "loss": 0.394,
+      "step": 665
+    },
+    {
+      "epoch": 0.5410235580828595,
+      "grad_norm": 0.2357444316148758,
+      "learning_rate": 9.805496923959363e-06,
+      "loss": 0.3964,
+      "step": 666
+    },
+    {
+      "epoch": 0.5418359057676686,
+      "grad_norm": 0.26284006237983704,
+      "learning_rate": 9.804189154244408e-06,
+      "loss": 0.4229,
+      "step": 667
+    },
+    {
+      "epoch": 0.5426482534524777,
+      "grad_norm": 0.24147599935531616,
+      "learning_rate": 9.802877090558237e-06,
+      "loss": 0.394,
+      "step": 668
+    },
+    {
+      "epoch": 0.5434606011372868,
+      "grad_norm": 0.2943766415119171,
+      "learning_rate": 9.801560734073567e-06,
+      "loss": 0.4176,
+      "step": 669
+    },
+    {
+      "epoch": 0.5442729488220959,
+      "grad_norm": 0.24865981936454773,
+      "learning_rate": 9.800240085966957e-06,
+      "loss": 0.4168,
+      "step": 670
+    },
+    {
+      "epoch": 0.545085296506905,
+      "grad_norm": 0.2670503258705139,
+      "learning_rate": 9.798915147418795e-06,
+      "loss": 0.4088,
+      "step": 671
+    },
+    {
+      "epoch": 0.545897644191714,
+      "grad_norm": 0.28005704283714294,
+      "learning_rate": 9.797585919613311e-06,
+      "loss": 0.4248,
+      "step": 672
+    },
+    {
+      "epoch": 0.5467099918765231,
+      "grad_norm": 0.2694072425365448,
+      "learning_rate": 9.796252403738564e-06,
+      "loss": 0.4041,
+      "step": 673
+    },
+    {
+      "epoch": 0.5475223395613322,
+      "grad_norm": 0.2527916431427002,
+      "learning_rate": 9.794914600986446e-06,
+      "loss": 0.403,
+      "step": 674
+    },
+    {
+      "epoch": 0.5483346872461413,
+      "grad_norm": 0.28963831067085266,
+      "learning_rate": 9.793572512552683e-06,
+      "loss": 0.389,
+      "step": 675
+    },
+    {
+      "epoch": 0.5491470349309504,
+      "grad_norm": 0.2729482650756836,
+      "learning_rate": 9.792226139636827e-06,
+      "loss": 0.4223,
+      "step": 676
+    },
+    {
+      "epoch": 0.5499593826157595,
+      "grad_norm": 0.27206307649612427,
+      "learning_rate": 9.790875483442265e-06,
+      "loss": 0.3882,
+      "step": 677
+    },
+    {
+      "epoch": 0.5507717303005687,
+      "grad_norm": 0.3494756519794464,
+      "learning_rate": 9.78952054517621e-06,
+      "loss": 0.4265,
+      "step": 678
+    },
+    {
+      "epoch": 0.5515840779853778,
+      "grad_norm": 0.2628221809864044,
+      "learning_rate": 9.788161326049702e-06,
+      "loss": 0.3995,
+      "step": 679
+    },
+    {
+      "epoch": 0.5523964256701869,
+      "grad_norm": 0.2909046709537506,
+      "learning_rate": 9.786797827277607e-06,
+      "loss": 0.4019,
+      "step": 680
+    },
+    {
+      "epoch": 0.553208773354996,
+      "grad_norm": 0.24540294706821442,
+      "learning_rate": 9.785430050078616e-06,
+      "loss": 0.3961,
+      "step": 681
+    },
+    {
+      "epoch": 0.554021121039805,
+      "grad_norm": 0.2559117078781128,
+      "learning_rate": 9.784057995675245e-06,
+      "loss": 0.407,
+      "step": 682
+    },
+    {
+      "epoch": 0.5548334687246141,
+      "grad_norm": 0.260810911655426,
+      "learning_rate": 9.782681665293831e-06,
+      "loss": 0.4278,
+      "step": 683
+    },
+    {
+      "epoch": 0.5556458164094232,
+      "grad_norm": 0.214094877243042,
+      "learning_rate": 9.781301060164537e-06,
+      "loss": 0.3939,
+      "step": 684
+    },
+    {
+      "epoch": 0.5564581640942323,
+      "grad_norm": 0.2517912685871124,
+      "learning_rate": 9.779916181521342e-06,
+      "loss": 0.4144,
+      "step": 685
+    },
+    {
+      "epoch": 0.5572705117790414,
+      "grad_norm": 0.2638695240020752,
+      "learning_rate": 9.778527030602049e-06,
+      "loss": 0.4055,
+      "step": 686
+    },
+    {
+      "epoch": 0.5580828594638505,
+      "grad_norm": 0.244925856590271,
+      "learning_rate": 9.777133608648276e-06,
+      "loss": 0.4163,
+      "step": 687
+    },
+    {
+      "epoch": 0.5588952071486596,
+      "grad_norm": 0.2706751525402069,
+      "learning_rate": 9.775735916905458e-06,
+      "loss": 0.4033,
+      "step": 688
+    },
+    {
+      "epoch": 0.5597075548334687,
+      "grad_norm": 0.2300882339477539,
+      "learning_rate": 9.77433395662285e-06,
+      "loss": 0.4034,
+      "step": 689
+    },
+    {
+      "epoch": 0.5605199025182778,
+      "grad_norm": 0.2961938977241516,
+      "learning_rate": 9.77292772905352e-06,
+      "loss": 0.4191,
+      "step": 690
+    },
+    {
+      "epoch": 0.561332250203087,
+      "grad_norm": 0.2421022355556488,
+      "learning_rate": 9.77151723545435e-06,
+      "loss": 0.4115,
+      "step": 691
+    },
+    {
+      "epoch": 0.5621445978878961,
+      "grad_norm": 0.2442595213651657,
+      "learning_rate": 9.77010247708604e-06,
+      "loss": 0.3861,
+      "step": 692
+    },
+    {
+      "epoch": 0.5629569455727051,
+      "grad_norm": 0.24319244921207428,
+      "learning_rate": 9.768683455213089e-06,
+      "loss": 0.4018,
+      "step": 693
+    },
+    {
+      "epoch": 0.5637692932575142,
+      "grad_norm": 0.2387973517179489,
+      "learning_rate": 9.76726017110382e-06,
+      "loss": 0.396,
+      "step": 694
+    },
+    {
+      "epoch": 0.5645816409423233,
+      "grad_norm": 0.2386782169342041,
+      "learning_rate": 9.765832626030359e-06,
+      "loss": 0.3952,
+      "step": 695
+    },
+    {
+      "epoch": 0.5653939886271324,
+      "grad_norm": 0.23355117440223694,
+      "learning_rate": 9.764400821268642e-06,
+      "loss": 0.3859,
+      "step": 696
+    },
+    {
+      "epoch": 0.5662063363119415,
+      "grad_norm": 0.27135640382766724,
+      "learning_rate": 9.762964758098412e-06,
+      "loss": 0.4207,
+      "step": 697
+    },
+    {
+      "epoch": 0.5670186839967506,
+      "grad_norm": 0.2538779377937317,
+      "learning_rate": 9.761524437803221e-06,
+      "loss": 0.3895,
+      "step": 698
+    },
+    {
+      "epoch": 0.5678310316815597,
+      "grad_norm": 0.2507195472717285,
+      "learning_rate": 9.76007986167042e-06,
+      "loss": 0.4023,
+      "step": 699
+    },
+    {
+      "epoch": 0.5686433793663688,
+      "grad_norm": 0.25602036714553833,
+      "learning_rate": 9.758631030991168e-06,
+      "loss": 0.4123,
+      "step": 700
+    },
+    {
+      "epoch": 0.5694557270511779,
+      "grad_norm": 0.2566196620464325,
+      "learning_rate": 9.757177947060426e-06,
+      "loss": 0.4092,
+      "step": 701
+    },
+    {
+      "epoch": 0.570268074735987,
+      "grad_norm": 0.2596403658390045,
+      "learning_rate": 9.755720611176957e-06,
+      "loss": 0.3975,
+      "step": 702
+    },
+    {
+      "epoch": 0.5710804224207962,
+      "grad_norm": 0.2681449353694916,
+      "learning_rate": 9.754259024643324e-06,
+      "loss": 0.4008,
+      "step": 703
+    },
+    {
+      "epoch": 0.5718927701056052,
+      "grad_norm": 0.26852452754974365,
+      "learning_rate": 9.75279318876589e-06,
+      "loss": 0.4352,
+      "step": 704
+    },
+    {
+      "epoch": 0.5727051177904143,
+      "grad_norm": 0.246471107006073,
+      "learning_rate": 9.751323104854812e-06,
+      "loss": 0.3933,
+      "step": 705
+    },
+    {
+      "epoch": 0.5735174654752234,
+      "grad_norm": 0.2724572718143463,
+      "learning_rate": 9.74984877422405e-06,
+      "loss": 0.4099,
+      "step": 706
+    },
+    {
+      "epoch": 0.5743298131600325,
+      "grad_norm": 0.2941463589668274,
+      "learning_rate": 9.748370198191355e-06,
+      "loss": 0.3865,
+      "step": 707
+    },
+    {
+      "epoch": 0.5751421608448416,
+      "grad_norm": 0.2307348996400833,
+      "learning_rate": 9.746887378078276e-06,
+      "loss": 0.412,
+      "step": 708
+    },
+    {
+      "epoch": 0.5759545085296507,
+      "grad_norm": 0.282833993434906,
+      "learning_rate": 9.74540031521015e-06,
+      "loss": 0.4087,
+      "step": 709
+    },
+    {
+      "epoch": 0.5767668562144598,
+      "grad_norm": 0.25914710760116577,
+      "learning_rate": 9.743909010916115e-06,
+      "loss": 0.4262,
+      "step": 710
+    },
+    {
+      "epoch": 0.5775792038992689,
+      "grad_norm": 0.23596906661987305,
+      "learning_rate": 9.74241346652909e-06,
+      "loss": 0.3986,
+      "step": 711
+    },
+    {
+      "epoch": 0.578391551584078,
+      "grad_norm": 0.2619219124317169,
+      "learning_rate": 9.740913683385791e-06,
+      "loss": 0.4214,
+      "step": 712
+    },
+    {
+      "epoch": 0.579203899268887,
+      "grad_norm": 0.2125251144170761,
+      "learning_rate": 9.73940966282672e-06,
+      "loss": 0.4201,
+      "step": 713
+    },
+    {
+      "epoch": 0.5800162469536961,
+      "grad_norm": 0.2371375411748886,
+      "learning_rate": 9.737901406196164e-06,
+      "loss": 0.4047,
+      "step": 714
+    },
+    {
+      "epoch": 0.5808285946385053,
+      "grad_norm": 0.2688083350658417,
+      "learning_rate": 9.736388914842203e-06,
+      "loss": 0.3939,
+      "step": 715
+    },
+    {
+      "epoch": 0.5816409423233144,
+      "grad_norm": 0.25305503606796265,
+      "learning_rate": 9.734872190116695e-06,
+      "loss": 0.3987,
+      "step": 716
+    },
+    {
+      "epoch": 0.5824532900081235,
+      "grad_norm": 0.28399789333343506,
+      "learning_rate": 9.733351233375283e-06,
+      "loss": 0.4161,
+      "step": 717
+    },
+    {
+      "epoch": 0.5832656376929326,
+      "grad_norm": 0.23377855122089386,
+      "learning_rate": 9.731826045977397e-06,
+      "loss": 0.3783,
+      "step": 718
+    },
+    {
+      "epoch": 0.5840779853777417,
+      "grad_norm": 0.2532336711883545,
+      "learning_rate": 9.730296629286245e-06,
+      "loss": 0.3995,
+      "step": 719
+    },
+    {
+      "epoch": 0.5848903330625508,
+      "grad_norm": 0.29724523425102234,
+      "learning_rate": 9.728762984668813e-06,
+      "loss": 0.4503,
+      "step": 720
+    },
+    {
+      "epoch": 0.5857026807473599,
+      "grad_norm": 0.2708858847618103,
+      "learning_rate": 9.727225113495871e-06,
+      "loss": 0.3982,
+      "step": 721
+    },
+    {
+      "epoch": 0.586515028432169,
+      "grad_norm": 0.2695591449737549,
+      "learning_rate": 9.725683017141964e-06,
+      "loss": 0.4104,
+      "step": 722
+    },
+    {
+      "epoch": 0.587327376116978,
+      "grad_norm": 0.2553618848323822,
+      "learning_rate": 9.724136696985412e-06,
+      "loss": 0.4102,
+      "step": 723
+    },
+    {
+      "epoch": 0.5881397238017871,
+      "grad_norm": 0.27096158266067505,
+      "learning_rate": 9.722586154408312e-06,
+      "loss": 0.3974,
+      "step": 724
+    },
+    {
+      "epoch": 0.5889520714865962,
+      "grad_norm": 0.21913954615592957,
+      "learning_rate": 9.721031390796535e-06,
+      "loss": 0.4271,
+      "step": 725
+    },
+    {
+      "epoch": 0.5897644191714053,
+      "grad_norm": 0.24850177764892578,
+      "learning_rate": 9.719472407539725e-06,
+      "loss": 0.4121,
+      "step": 726
+    },
+    {
+      "epoch": 0.5905767668562144,
+      "grad_norm": 0.2505251467227936,
+      "learning_rate": 9.717909206031295e-06,
+      "loss": 0.4171,
+      "step": 727
+    },
+    {
+      "epoch": 0.5913891145410236,
+      "grad_norm": 0.2470681071281433,
+      "learning_rate": 9.716341787668434e-06,
+      "loss": 0.4107,
+      "step": 728
+    },
+    {
+      "epoch": 0.5922014622258327,
+      "grad_norm": 0.2643090784549713,
+      "learning_rate": 9.714770153852093e-06,
+      "loss": 0.3831,
+      "step": 729
+    },
+    {
+      "epoch": 0.5930138099106418,
+      "grad_norm": 0.26128607988357544,
+      "learning_rate": 9.713194305986995e-06,
+      "loss": 0.4055,
+      "step": 730
+    },
+    {
+      "epoch": 0.5938261575954509,
+      "grad_norm": 0.24820591509342194,
+      "learning_rate": 9.71161424548163e-06,
+      "loss": 0.4302,
+      "step": 731
+    },
+    {
+      "epoch": 0.59463850528026,
+      "grad_norm": 0.2769261300563812,
+      "learning_rate": 9.710029973748249e-06,
+      "loss": 0.3938,
+      "step": 732
+    },
+    {
+      "epoch": 0.595450852965069,
+      "grad_norm": 0.26816704869270325,
+      "learning_rate": 9.708441492202874e-06,
+      "loss": 0.3832,
+      "step": 733
+    },
+    {
+      "epoch": 0.5962632006498781,
+      "grad_norm": 0.28544846177101135,
+      "learning_rate": 9.70684880226528e-06,
+      "loss": 0.39,
+      "step": 734
+    },
+    {
+      "epoch": 0.5970755483346872,
+      "grad_norm": 0.23257094621658325,
+      "learning_rate": 9.705251905359014e-06,
+      "loss": 0.4008,
+      "step": 735
+    },
+    {
+      "epoch": 0.5978878960194963,
+      "grad_norm": 0.2571530044078827,
+      "learning_rate": 9.703650802911374e-06,
+      "loss": 0.4192,
+      "step": 736
+    },
+    {
+      "epoch": 0.5987002437043054,
+      "grad_norm": 0.25099435448646545,
+      "learning_rate": 9.702045496353422e-06,
+      "loss": 0.4,
+      "step": 737
+    },
+    {
+      "epoch": 0.5995125913891145,
+      "grad_norm": 0.25403058528900146,
+      "learning_rate": 9.700435987119981e-06,
+      "loss": 0.4213,
+      "step": 738
+    },
+    {
+      "epoch": 0.6003249390739236,
+      "grad_norm": 0.2514328956604004,
+      "learning_rate": 9.69882227664962e-06,
+      "loss": 0.3805,
+      "step": 739
+    },
+    {
+      "epoch": 0.6011372867587328,
+      "grad_norm": 0.24930667877197266,
+      "learning_rate": 9.69720436638467e-06,
+      "loss": 0.3887,
+      "step": 740
+    },
+    {
+      "epoch": 0.6019496344435419,
+      "grad_norm": 0.246470645070076,
+      "learning_rate": 9.69558225777122e-06,
+      "loss": 0.393,
+      "step": 741
+    },
+    {
+      "epoch": 0.602761982128351,
+      "grad_norm": 0.24809180200099945,
+      "learning_rate": 9.693955952259099e-06,
+      "loss": 0.3889,
+      "step": 742
+    },
+    {
+      "epoch": 0.6035743298131601,
+      "grad_norm": 0.27407053112983704,
+      "learning_rate": 9.6923254513019e-06,
+      "loss": 0.3965,
+      "step": 743
+    },
+    {
+      "epoch": 0.6043866774979691,
+      "grad_norm": 0.256459504365921,
+      "learning_rate": 9.69069075635696e-06,
+      "loss": 0.4132,
+      "step": 744
+    },
+    {
+      "epoch": 0.6051990251827782,
+      "grad_norm": 0.26402056217193604,
+      "learning_rate": 9.689051868885362e-06,
+      "loss": 0.4115,
+      "step": 745
+    },
+    {
+      "epoch": 0.6060113728675873,
+      "grad_norm": 0.24724510312080383,
+      "learning_rate": 9.68740879035194e-06,
+      "loss": 0.4242,
+      "step": 746
+    },
+    {
+      "epoch": 0.6068237205523964,
+      "grad_norm": 0.25216129422187805,
+      "learning_rate": 9.685761522225271e-06,
+      "loss": 0.4202,
+      "step": 747
+    },
+    {
+      "epoch": 0.6076360682372055,
+      "grad_norm": 0.25678226351737976,
+      "learning_rate": 9.684110065977685e-06,
+      "loss": 0.3898,
+      "step": 748
+    },
+    {
+      "epoch": 0.6084484159220146,
+      "grad_norm": 0.22425049543380737,
+      "learning_rate": 9.682454423085244e-06,
+      "loss": 0.3886,
+      "step": 749
+    },
+    {
+      "epoch": 0.6092607636068237,
+      "grad_norm": 0.2511903941631317,
+      "learning_rate": 9.680794595027761e-06,
+      "loss": 0.3999,
+      "step": 750
+    },
+    {
+      "epoch": 0.6100731112916328,
+      "grad_norm": 0.239465594291687,
+      "learning_rate": 9.67913058328878e-06,
+      "loss": 0.3921,
+      "step": 751
+    },
+    {
+      "epoch": 0.6108854589764419,
+      "grad_norm": 0.2526945173740387,
+      "learning_rate": 9.677462389355594e-06,
+      "loss": 0.4165,
+      "step": 752
+    },
+    {
+      "epoch": 0.6116978066612511,
+      "grad_norm": 0.22253625094890594,
+      "learning_rate": 9.675790014719231e-06,
+      "loss": 0.3949,
+      "step": 753
+    },
+    {
+      "epoch": 0.6125101543460602,
+      "grad_norm": 0.22744116187095642,
+      "learning_rate": 9.674113460874453e-06,
+      "loss": 0.3837,
+      "step": 754
+    },
+    {
+      "epoch": 0.6133225020308692,
+      "grad_norm": 0.27041739225387573,
+      "learning_rate": 9.67243272931976e-06,
+      "loss": 0.409,
+      "step": 755
+    },
+    {
+      "epoch": 0.6141348497156783,
+      "grad_norm": 0.256979763507843,
+      "learning_rate": 9.670747821557387e-06,
+      "loss": 0.4088,
+      "step": 756
+    },
+    {
+      "epoch": 0.6149471974004874,
+      "grad_norm": 0.23679503798484802,
+      "learning_rate": 9.669058739093296e-06,
+      "loss": 0.4136,
+      "step": 757
+    },
+    {
+      "epoch": 0.6157595450852965,
+      "grad_norm": 0.27369454503059387,
+      "learning_rate": 9.667365483437188e-06,
+      "loss": 0.4185,
+      "step": 758
+    },
+    {
+      "epoch": 0.6165718927701056,
+      "grad_norm": 0.2655172348022461,
+      "learning_rate": 9.665668056102487e-06,
+      "loss": 0.3958,
+      "step": 759
+    },
+    {
+      "epoch": 0.6173842404549147,
+      "grad_norm": 0.2676054537296295,
+      "learning_rate": 9.663966458606353e-06,
+      "loss": 0.4441,
+      "step": 760
+    },
+    {
+      "epoch": 0.6181965881397238,
+      "grad_norm": 0.24829019606113434,
+      "learning_rate": 9.662260692469664e-06,
+      "loss": 0.3899,
+      "step": 761
+    },
+    {
+      "epoch": 0.6190089358245329,
+      "grad_norm": 0.26072412729263306,
+      "learning_rate": 9.660550759217035e-06,
+      "loss": 0.3815,
+      "step": 762
+    },
+    {
+      "epoch": 0.619821283509342,
+      "grad_norm": 0.2424943447113037,
+      "learning_rate": 9.658836660376796e-06,
+      "loss": 0.3875,
+      "step": 763
+    },
+    {
+      "epoch": 0.620633631194151,
+      "grad_norm": 0.2784329652786255,
+      "learning_rate": 9.657118397481004e-06,
+      "loss": 0.3988,
+      "step": 764
+    },
+    {
+      "epoch": 0.6214459788789602,
+      "grad_norm": 0.25868117809295654,
+      "learning_rate": 9.65539597206544e-06,
+      "loss": 0.4056,
+      "step": 765
+    },
+    {
+      "epoch": 0.6222583265637693,
+      "grad_norm": 0.23047170042991638,
+      "learning_rate": 9.6536693856696e-06,
+      "loss": 0.3608,
+      "step": 766
+    },
+    {
+      "epoch": 0.6230706742485784,
+      "grad_norm": 0.25136205554008484,
+      "learning_rate": 9.651938639836705e-06,
+      "loss": 0.405,
+      "step": 767
+    },
+    {
+      "epoch": 0.6238830219333875,
+      "grad_norm": 0.2399366796016693,
+      "learning_rate": 9.650203736113689e-06,
+      "loss": 0.3993,
+      "step": 768
+    },
+    {
+      "epoch": 0.6246953696181966,
+      "grad_norm": 0.25205355882644653,
+      "learning_rate": 9.648464676051206e-06,
+      "loss": 0.4011,
+      "step": 769
+    },
+    {
+      "epoch": 0.6255077173030057,
+      "grad_norm": 0.2364083081483841,
+      "learning_rate": 9.64672146120362e-06,
+      "loss": 0.4195,
+      "step": 770
+    },
+    {
+      "epoch": 0.6263200649878148,
+      "grad_norm": 0.24087244272232056,
+      "learning_rate": 9.644974093129017e-06,
+      "loss": 0.399,
+      "step": 771
+    },
+    {
+      "epoch": 0.6271324126726239,
+      "grad_norm": 0.23569975793361664,
+      "learning_rate": 9.643222573389184e-06,
+      "loss": 0.4366,
+      "step": 772
+    },
+    {
+      "epoch": 0.627944760357433,
+      "grad_norm": 0.2397482991218567,
+      "learning_rate": 9.641466903549628e-06,
+      "loss": 0.3795,
+      "step": 773
+    },
+    {
+      "epoch": 0.628757108042242,
+      "grad_norm": 0.24417050182819366,
+      "learning_rate": 9.639707085179562e-06,
+      "loss": 0.4097,
+      "step": 774
+    },
+    {
+      "epoch": 0.6295694557270511,
+      "grad_norm": 0.23295964300632477,
+      "learning_rate": 9.637943119851905e-06,
+      "loss": 0.4201,
+      "step": 775
+    },
+    {
+      "epoch": 0.6303818034118602,
+      "grad_norm": 0.23731529712677002,
+      "learning_rate": 9.636175009143286e-06,
+      "loss": 0.381,
+      "step": 776
+    },
+    {
+      "epoch": 0.6311941510966693,
+      "grad_norm": 0.24949167668819427,
+      "learning_rate": 9.634402754634037e-06,
+      "loss": 0.4286,
+      "step": 777
+    },
+    {
+      "epoch": 0.6320064987814785,
+      "grad_norm": 0.23785102367401123,
+      "learning_rate": 9.632626357908194e-06,
+      "loss": 0.423,
+      "step": 778
+    },
+    {
+      "epoch": 0.6328188464662876,
+      "grad_norm": 0.22728615999221802,
+      "learning_rate": 9.630845820553495e-06,
+      "loss": 0.3881,
+      "step": 779
+    },
+    {
+      "epoch": 0.6336311941510967,
+      "grad_norm": 0.2315235137939453,
+      "learning_rate": 9.629061144161383e-06,
+      "loss": 0.4104,
+      "step": 780
+    },
+    {
+      "epoch": 0.6344435418359058,
+      "grad_norm": 0.23670697212219238,
+      "learning_rate": 9.627272330326993e-06,
+      "loss": 0.396,
+      "step": 781
+    },
+    {
+      "epoch": 0.6352558895207149,
+      "grad_norm": 0.26050522923469543,
+      "learning_rate": 9.625479380649163e-06,
+      "loss": 0.4148,
+      "step": 782
+    },
+    {
+      "epoch": 0.636068237205524,
+      "grad_norm": 0.2343936413526535,
+      "learning_rate": 9.623682296730427e-06,
+      "loss": 0.4079,
+      "step": 783
+    },
+    {
+      "epoch": 0.6368805848903331,
+      "grad_norm": 0.27195045351982117,
+      "learning_rate": 9.621881080177013e-06,
+      "loss": 0.3963,
+      "step": 784
+    },
+    {
+      "epoch": 0.6376929325751421,
+      "grad_norm": 0.26169589161872864,
+      "learning_rate": 9.620075732598843e-06,
+      "loss": 0.411,
+      "step": 785
+    },
+    {
+      "epoch": 0.6385052802599512,
+      "grad_norm": 0.22942736744880676,
+      "learning_rate": 9.618266255609533e-06,
+      "loss": 0.4059,
+      "step": 786
+    },
+    {
+      "epoch": 0.6393176279447603,
+      "grad_norm": 0.25635722279548645,
+      "learning_rate": 9.616452650826392e-06,
+      "loss": 0.4032,
+      "step": 787
+    },
+    {
+      "epoch": 0.6401299756295694,
+      "grad_norm": 0.21978473663330078,
+      "learning_rate": 9.614634919870407e-06,
+      "loss": 0.3923,
+      "step": 788
+    },
+    {
+      "epoch": 0.6409423233143785,
+      "grad_norm": 0.2556641101837158,
+      "learning_rate": 9.612813064366267e-06,
+      "loss": 0.3912,
+      "step": 789
+    },
+    {
+      "epoch": 0.6417546709991877,
+      "grad_norm": 0.24955183267593384,
+      "learning_rate": 9.61098708594234e-06,
+      "loss": 0.4038,
+      "step": 790
+    },
+    {
+      "epoch": 0.6425670186839968,
+      "grad_norm": 0.2555236220359802,
+      "learning_rate": 9.60915698623068e-06,
+      "loss": 0.4323,
+      "step": 791
+    },
+    {
+      "epoch": 0.6433793663688059,
+      "grad_norm": 0.2639031708240509,
+      "learning_rate": 9.607322766867026e-06,
+      "loss": 0.3945,
+      "step": 792
+    },
+    {
+      "epoch": 0.644191714053615,
+      "grad_norm": 0.2693772315979004,
+      "learning_rate": 9.605484429490796e-06,
+      "loss": 0.3953,
+      "step": 793
+    },
+    {
+      "epoch": 0.6450040617384241,
+      "grad_norm": 0.2682175934314728,
+      "learning_rate": 9.603641975745095e-06,
+      "loss": 0.3843,
+      "step": 794
+    },
+    {
+      "epoch": 0.6458164094232332,
+      "grad_norm": 0.26068294048309326,
+      "learning_rate": 9.601795407276699e-06,
+      "loss": 0.4038,
+      "step": 795
+    },
+    {
+      "epoch": 0.6466287571080422,
+      "grad_norm": 0.2643123269081116,
+      "learning_rate": 9.59994472573607e-06,
+      "loss": 0.4033,
+      "step": 796
+    },
+    {
+      "epoch": 0.6474411047928513,
+      "grad_norm": 0.27633050084114075,
+      "learning_rate": 9.598089932777339e-06,
+      "loss": 0.3943,
+      "step": 797
+    },
+    {
+      "epoch": 0.6482534524776604,
+      "grad_norm": 0.26549965143203735,
+      "learning_rate": 9.596231030058315e-06,
+      "loss": 0.3821,
+      "step": 798
+    },
+    {
+      "epoch": 0.6490658001624695,
+      "grad_norm": 0.27333611249923706,
+      "learning_rate": 9.594368019240484e-06,
+      "loss": 0.4301,
+      "step": 799
+    },
+    {
+      "epoch": 0.6498781478472786,
+      "grad_norm": 0.31025347113609314,
+      "learning_rate": 9.592500901988994e-06,
+      "loss": 0.4209,
+      "step": 800
+    },
+    {
+      "epoch": 0.6506904955320877,
+      "grad_norm": 0.25535473227500916,
+      "learning_rate": 9.590629679972673e-06,
+      "loss": 0.417,
+      "step": 801
+    },
+    {
+      "epoch": 0.6515028432168968,
+      "grad_norm": 0.26266351342201233,
+      "learning_rate": 9.588754354864014e-06,
+      "loss": 0.4067,
+      "step": 802
+    },
+    {
+      "epoch": 0.652315190901706,
+      "grad_norm": 0.23795172572135925,
+      "learning_rate": 9.586874928339176e-06,
+      "loss": 0.3912,
+      "step": 803
+    },
+    {
+      "epoch": 0.6531275385865151,
+      "grad_norm": 0.23893754184246063,
+      "learning_rate": 9.584991402077988e-06,
+      "loss": 0.4183,
+      "step": 804
+    },
+    {
+      "epoch": 0.6539398862713242,
+      "grad_norm": 0.255137175321579,
+      "learning_rate": 9.583103777763938e-06,
+      "loss": 0.4094,
+      "step": 805
+    },
+    {
+      "epoch": 0.6547522339561332,
+      "grad_norm": 0.253273606300354,
+      "learning_rate": 9.58121205708418e-06,
+      "loss": 0.3859,
+      "step": 806
+    },
+    {
+      "epoch": 0.6555645816409423,
+      "grad_norm": 0.3177814781665802,
+      "learning_rate": 9.57931624172953e-06,
+      "loss": 0.4167,
+      "step": 807
+    },
+    {
+      "epoch": 0.6563769293257514,
+      "grad_norm": 0.24421189725399017,
+      "learning_rate": 9.577416333394463e-06,
+      "loss": 0.4137,
+      "step": 808
+    },
+    {
+      "epoch": 0.6571892770105605,
+      "grad_norm": 0.25665339827537537,
+      "learning_rate": 9.575512333777108e-06,
+      "loss": 0.3932,
+      "step": 809
+    },
+    {
+      "epoch": 0.6580016246953696,
+      "grad_norm": 0.2548246383666992,
+      "learning_rate": 9.57360424457926e-06,
+      "loss": 0.407,
+      "step": 810
+    },
+    {
+      "epoch": 0.6588139723801787,
+      "grad_norm": 0.2451210618019104,
+      "learning_rate": 9.571692067506363e-06,
+      "loss": 0.3908,
+      "step": 811
+    },
+    {
+      "epoch": 0.6596263200649878,
+      "grad_norm": 0.2415851354598999,
+      "learning_rate": 9.56977580426751e-06,
+      "loss": 0.4011,
+      "step": 812
+    },
+    {
+      "epoch": 0.6604386677497969,
+      "grad_norm": 0.24131278693675995,
+      "learning_rate": 9.567855456575459e-06,
+      "loss": 0.3984,
+      "step": 813
+    },
+    {
+      "epoch": 0.661251015434606,
+      "grad_norm": 0.27100619673728943,
+      "learning_rate": 9.565931026146607e-06,
+      "loss": 0.3936,
+      "step": 814
+    },
+    {
+      "epoch": 0.6620633631194152,
+      "grad_norm": 0.28757667541503906,
+      "learning_rate": 9.564002514701006e-06,
+      "loss": 0.3884,
+      "step": 815
+    },
+    {
+      "epoch": 0.6628757108042242,
+      "grad_norm": 0.22644314169883728,
+      "learning_rate": 9.562069923962355e-06,
+      "loss": 0.4027,
+      "step": 816
+    },
+    {
+      "epoch": 0.6636880584890333,
+      "grad_norm": 0.2752091586589813,
+      "learning_rate": 9.560133255657997e-06,
+      "loss": 0.4146,
+      "step": 817
+    },
+    {
+      "epoch": 0.6645004061738424,
+      "grad_norm": 0.2774292826652527,
+      "learning_rate": 9.558192511518925e-06,
+      "loss": 0.4024,
+      "step": 818
+    },
+    {
+      "epoch": 0.6653127538586515,
+      "grad_norm": 0.273128867149353,
+      "learning_rate": 9.556247693279764e-06,
+      "loss": 0.3936,
+      "step": 819
+    },
+    {
+      "epoch": 0.6661251015434606,
+      "grad_norm": 0.2355608195066452,
+      "learning_rate": 9.554298802678793e-06,
+      "loss": 0.4176,
+      "step": 820
+    },
+    {
+      "epoch": 0.6669374492282697,
+      "grad_norm": 0.2568332254886627,
+      "learning_rate": 9.552345841457922e-06,
+      "loss": 0.4287,
+      "step": 821
+    },
+    {
+      "epoch": 0.6677497969130788,
+      "grad_norm": 0.2939576208591461,
+      "learning_rate": 9.550388811362704e-06,
+      "loss": 0.3864,
+      "step": 822
+    },
+    {
+      "epoch": 0.6685621445978879,
+      "grad_norm": 0.24154028296470642,
+      "learning_rate": 9.548427714142326e-06,
+      "loss": 0.3917,
+      "step": 823
+    },
+    {
+      "epoch": 0.669374492282697,
+      "grad_norm": 0.25152885913848877,
+      "learning_rate": 9.546462551549612e-06,
+      "loss": 0.4048,
+      "step": 824
+    },
+    {
+      "epoch": 0.670186839967506,
+      "grad_norm": 0.24856674671173096,
+      "learning_rate": 9.54449332534102e-06,
+      "loss": 0.4141,
+      "step": 825
+    },
+    {
+      "epoch": 0.6709991876523151,
+      "grad_norm": 0.2473793923854828,
+      "learning_rate": 9.542520037276636e-06,
+      "loss": 0.4159,
+      "step": 826
+    },
+    {
+      "epoch": 0.6718115353371243,
+      "grad_norm": 0.24448703229427338,
+      "learning_rate": 9.540542689120184e-06,
+      "loss": 0.4062,
+      "step": 827
+    },
+    {
+      "epoch": 0.6726238830219334,
+      "grad_norm": 0.24718528985977173,
+      "learning_rate": 9.538561282639008e-06,
+      "loss": 0.4039,
+      "step": 828
+    },
+    {
+      "epoch": 0.6734362307067425,
+      "grad_norm": 0.23815859854221344,
+      "learning_rate": 9.536575819604087e-06,
+      "loss": 0.4051,
+      "step": 829
+    },
+    {
+      "epoch": 0.6742485783915516,
+      "grad_norm": 0.21430295705795288,
+      "learning_rate": 9.534586301790021e-06,
+      "loss": 0.3927,
+      "step": 830
+    },
+    {
+      "epoch": 0.6750609260763607,
+      "grad_norm": 0.24119433760643005,
+      "learning_rate": 9.532592730975035e-06,
+      "loss": 0.4269,
+      "step": 831
+    },
+    {
+      "epoch": 0.6758732737611698,
+      "grad_norm": 0.21082492172718048,
+      "learning_rate": 9.530595108940978e-06,
+      "loss": 0.4005,
+      "step": 832
+    },
+    {
+      "epoch": 0.6766856214459789,
+      "grad_norm": 0.24085891246795654,
+      "learning_rate": 9.52859343747332e-06,
+      "loss": 0.4153,
+      "step": 833
+    },
+    {
+      "epoch": 0.677497969130788,
+      "grad_norm": 0.21763074398040771,
+      "learning_rate": 9.526587718361147e-06,
+      "loss": 0.4017,
+      "step": 834
+    },
+    {
+      "epoch": 0.6783103168155971,
+      "grad_norm": 0.2484983503818512,
+      "learning_rate": 9.524577953397167e-06,
+      "loss": 0.3983,
+      "step": 835
+    },
+    {
+      "epoch": 0.6791226645004061,
+      "grad_norm": 0.22234702110290527,
+      "learning_rate": 9.522564144377703e-06,
+      "loss": 0.392,
+      "step": 836
+    },
+    {
+      "epoch": 0.6799350121852152,
+      "grad_norm": 0.22872908413410187,
+      "learning_rate": 9.52054629310269e-06,
+      "loss": 0.3961,
+      "step": 837
+    },
+    {
+      "epoch": 0.6807473598700243,
+      "grad_norm": 0.21813155710697174,
+      "learning_rate": 9.518524401375682e-06,
+      "loss": 0.3927,
+      "step": 838
+    },
+    {
+      "epoch": 0.6815597075548334,
+      "grad_norm": 0.2366940826177597,
+      "learning_rate": 9.516498471003837e-06,
+      "loss": 0.3882,
+      "step": 839
+    },
+    {
+      "epoch": 0.6823720552396426,
+      "grad_norm": 0.24393349885940552,
+      "learning_rate": 9.514468503797927e-06,
+      "loss": 0.4236,
+      "step": 840
+    },
+    {
+      "epoch": 0.6831844029244517,
+      "grad_norm": 0.21230487525463104,
+      "learning_rate": 9.51243450157233e-06,
+      "loss": 0.4089,
+      "step": 841
+    },
+    {
+      "epoch": 0.6839967506092608,
+      "grad_norm": 0.24379970133304596,
+      "learning_rate": 9.510396466145032e-06,
+      "loss": 0.3838,
+      "step": 842
+    },
+    {
+      "epoch": 0.6848090982940699,
+      "grad_norm": 0.24387714266777039,
+      "learning_rate": 9.508354399337625e-06,
+      "loss": 0.4172,
+      "step": 843
+    },
+    {
+      "epoch": 0.685621445978879,
+      "grad_norm": 0.26364970207214355,
+      "learning_rate": 9.5063083029753e-06,
+      "loss": 0.3903,
+      "step": 844
+    },
+    {
+      "epoch": 0.6864337936636881,
+      "grad_norm": 0.22990629076957703,
+      "learning_rate": 9.504258178886858e-06,
+      "loss": 0.3851,
+      "step": 845
+    },
+    {
+      "epoch": 0.6872461413484972,
+      "grad_norm": 0.23497091233730316,
+      "learning_rate": 9.502204028904687e-06,
+      "loss": 0.4005,
+      "step": 846
+    },
+    {
+      "epoch": 0.6880584890333062,
+      "grad_norm": 0.23972159624099731,
+      "learning_rate": 9.500145854864785e-06,
+      "loss": 0.3897,
+      "step": 847
+    },
+    {
+      "epoch": 0.6888708367181153,
+      "grad_norm": 0.2384214699268341,
+      "learning_rate": 9.498083658606744e-06,
+      "loss": 0.3875,
+      "step": 848
+    },
+    {
+      "epoch": 0.6896831844029244,
+      "grad_norm": 0.24600253999233246,
+      "learning_rate": 9.496017441973747e-06,
+      "loss": 0.3864,
+      "step": 849
+    },
+    {
+      "epoch": 0.6904955320877335,
+      "grad_norm": 0.25432920455932617,
+      "learning_rate": 9.49394720681257e-06,
+      "loss": 0.4048,
+      "step": 850
+    },
+    {
+      "epoch": 0.6913078797725426,
+      "grad_norm": 0.264542818069458,
+      "learning_rate": 9.49187295497359e-06,
+      "loss": 0.4233,
+      "step": 851
+    },
+    {
+      "epoch": 0.6921202274573518,
+      "grad_norm": 0.2774144411087036,
+      "learning_rate": 9.48979468831076e-06,
+      "loss": 0.3972,
+      "step": 852
+    },
+    {
+      "epoch": 0.6929325751421609,
+      "grad_norm": 0.2689442038536072,
+      "learning_rate": 9.487712408681635e-06,
+      "loss": 0.4,
+      "step": 853
+    },
+    {
+      "epoch": 0.69374492282697,
+      "grad_norm": 0.23342889547348022,
+      "learning_rate": 9.485626117947351e-06,
+      "loss": 0.4023,
+      "step": 854
+    },
+    {
+      "epoch": 0.6945572705117791,
+      "grad_norm": 0.2716900706291199,
+      "learning_rate": 9.483535817972625e-06,
+      "loss": 0.4175,
+      "step": 855
+    },
+    {
+      "epoch": 0.6953696181965882,
+      "grad_norm": 0.25904518365859985,
+      "learning_rate": 9.481441510625765e-06,
+      "loss": 0.4174,
+      "step": 856
+    },
+    {
+      "epoch": 0.6961819658813972,
+      "grad_norm": 0.25597989559173584,
+      "learning_rate": 9.479343197778653e-06,
+      "loss": 0.4004,
+      "step": 857
+    },
+    {
+      "epoch": 0.6969943135662063,
+      "grad_norm": 0.3103913366794586,
+      "learning_rate": 9.477240881306759e-06,
+      "loss": 0.3701,
+      "step": 858
+    },
+    {
+      "epoch": 0.6978066612510154,
+      "grad_norm": 0.22577781975269318,
+      "learning_rate": 9.475134563089128e-06,
+      "loss": 0.4259,
+      "step": 859
+    },
+    {
+      "epoch": 0.6986190089358245,
+      "grad_norm": 0.3199273347854614,
+      "learning_rate": 9.47302424500838e-06,
+      "loss": 0.4108,
+      "step": 860
+    },
+    {
+      "epoch": 0.6994313566206336,
+      "grad_norm": 0.24692107737064362,
+      "learning_rate": 9.47090992895071e-06,
+      "loss": 0.4003,
+      "step": 861
+    },
+    {
+      "epoch": 0.7002437043054427,
+      "grad_norm": 0.26060476899147034,
+      "learning_rate": 9.468791616805893e-06,
+      "loss": 0.3691,
+      "step": 862
+    },
+    {
+      "epoch": 0.7010560519902518,
+      "grad_norm": 0.29198670387268066,
+      "learning_rate": 9.466669310467266e-06,
+      "loss": 0.4243,
+      "step": 863
+    },
+    {
+      "epoch": 0.7018683996750609,
+      "grad_norm": 0.252745658159256,
+      "learning_rate": 9.464543011831742e-06,
+      "loss": 0.3945,
+      "step": 864
+    },
+    {
+      "epoch": 0.7026807473598701,
+      "grad_norm": 0.29863619804382324,
+      "learning_rate": 9.462412722799801e-06,
+      "loss": 0.3996,
+      "step": 865
+    },
+    {
+      "epoch": 0.7034930950446792,
+      "grad_norm": 0.2699202597141266,
+      "learning_rate": 9.46027844527549e-06,
+      "loss": 0.408,
+      "step": 866
+    },
+    {
+      "epoch": 0.7043054427294883,
+      "grad_norm": 0.3055218756198883,
+      "learning_rate": 9.45814018116642e-06,
+      "loss": 0.4066,
+      "step": 867
+    },
+    {
+      "epoch": 0.7051177904142973,
+      "grad_norm": 0.34577521681785583,
+      "learning_rate": 9.455997932383766e-06,
+      "loss": 0.4087,
+      "step": 868
+    },
+    {
+      "epoch": 0.7059301380991064,
+      "grad_norm": 0.23180043697357178,
+      "learning_rate": 9.453851700842262e-06,
+      "loss": 0.3751,
+      "step": 869
+    },
+    {
+      "epoch": 0.7067424857839155,
+      "grad_norm": 0.43736666440963745,
+      "learning_rate": 9.451701488460207e-06,
+      "loss": 0.4031,
+      "step": 870
+    },
+    {
+      "epoch": 0.7075548334687246,
+      "grad_norm": 0.2777930200099945,
+      "learning_rate": 9.449547297159453e-06,
+      "loss": 0.4264,
+      "step": 871
+    },
+    {
+      "epoch": 0.7083671811535337,
+      "grad_norm": 0.2624521553516388,
+      "learning_rate": 9.447389128865413e-06,
+      "loss": 0.4095,
+      "step": 872
+    },
+    {
+      "epoch": 0.7091795288383428,
+      "grad_norm": 0.27372607588768005,
+      "learning_rate": 9.445226985507047e-06,
+      "loss": 0.4037,
+      "step": 873
+    },
+    {
+      "epoch": 0.7099918765231519,
+      "grad_norm": 0.2766979932785034,
+      "learning_rate": 9.44306086901688e-06,
+      "loss": 0.3935,
+      "step": 874
+    },
+    {
+      "epoch": 0.710804224207961,
+      "grad_norm": 0.2481609433889389,
+      "learning_rate": 9.440890781330974e-06,
+      "loss": 0.4139,
+      "step": 875
+    },
+    {
+      "epoch": 0.7116165718927701,
+      "grad_norm": 0.27301961183547974,
+      "learning_rate": 9.43871672438895e-06,
+      "loss": 0.3845,
+      "step": 876
+    },
+    {
+      "epoch": 0.7124289195775793,
+      "grad_norm": 0.2765979766845703,
+      "learning_rate": 9.436538700133977e-06,
+      "loss": 0.3931,
+      "step": 877
+    },
+    {
+      "epoch": 0.7132412672623883,
+      "grad_norm": 0.23905827105045319,
+      "learning_rate": 9.434356710512763e-06,
+      "loss": 0.4176,
+      "step": 878
+    },
+    {
+      "epoch": 0.7140536149471974,
+      "grad_norm": 0.2674000561237335,
+      "learning_rate": 9.432170757475566e-06,
+      "loss": 0.3757,
+      "step": 879
+    },
+    {
+      "epoch": 0.7148659626320065,
+      "grad_norm": 0.24318534135818481,
+      "learning_rate": 9.429980842976186e-06,
+      "loss": 0.3774,
+      "step": 880
+    },
+    {
+      "epoch": 0.7156783103168156,
+      "grad_norm": 0.28106003999710083,
+      "learning_rate": 9.427786968971961e-06,
+      "loss": 0.4239,
+      "step": 881
+    },
+    {
+      "epoch": 0.7164906580016247,
+      "grad_norm": 0.25018641352653503,
+      "learning_rate": 9.425589137423772e-06,
+      "loss": 0.4033,
+      "step": 882
+    },
+    {
+      "epoch": 0.7173030056864338,
+      "grad_norm": 0.24814604222774506,
+      "learning_rate": 9.423387350296032e-06,
+      "loss": 0.4142,
+      "step": 883
+    },
+    {
+      "epoch": 0.7181153533712429,
+      "grad_norm": 0.23805955052375793,
+      "learning_rate": 9.421181609556693e-06,
+      "loss": 0.4065,
+      "step": 884
+    },
+    {
+      "epoch": 0.718927701056052,
+      "grad_norm": 0.2833292782306671,
+      "learning_rate": 9.418971917177241e-06,
+      "loss": 0.393,
+      "step": 885
+    },
+    {
+      "epoch": 0.7197400487408611,
+      "grad_norm": 0.2426101565361023,
+      "learning_rate": 9.416758275132693e-06,
+      "loss": 0.4207,
+      "step": 886
+    },
+    {
+      "epoch": 0.7205523964256701,
+      "grad_norm": 0.2663332521915436,
+      "learning_rate": 9.414540685401596e-06,
+      "loss": 0.4244,
+      "step": 887
+    },
+    {
+      "epoch": 0.7213647441104792,
+      "grad_norm": 0.27566030621528625,
+      "learning_rate": 9.412319149966025e-06,
+      "loss": 0.3956,
+      "step": 888
+    },
+    {
+      "epoch": 0.7221770917952883,
+      "grad_norm": 0.23856551945209503,
+      "learning_rate": 9.410093670811582e-06,
+      "loss": 0.4167,
+      "step": 889
+    },
+    {
+      "epoch": 0.7229894394800975,
+      "grad_norm": 0.24433845281600952,
+      "learning_rate": 9.407864249927396e-06,
+      "loss": 0.4333,
+      "step": 890
+    },
+    {
+      "epoch": 0.7238017871649066,
+      "grad_norm": 0.268272340297699,
+      "learning_rate": 9.405630889306116e-06,
+      "loss": 0.3999,
+      "step": 891
+    },
+    {
+      "epoch": 0.7246141348497157,
+      "grad_norm": 0.28072091937065125,
+      "learning_rate": 9.403393590943916e-06,
+      "loss": 0.4089,
+      "step": 892
+    },
+    {
+      "epoch": 0.7254264825345248,
+      "grad_norm": 0.23813197016716003,
+      "learning_rate": 9.401152356840484e-06,
+      "loss": 0.4087,
+      "step": 893
+    },
+    {
+      "epoch": 0.7262388302193339,
+      "grad_norm": 0.2340707927942276,
+      "learning_rate": 9.39890718899903e-06,
+      "loss": 0.3908,
+      "step": 894
+    },
+    {
+      "epoch": 0.727051177904143,
+      "grad_norm": 0.2526983916759491,
+      "learning_rate": 9.39665808942628e-06,
+      "loss": 0.4181,
+      "step": 895
+    },
+    {
+      "epoch": 0.7278635255889521,
+      "grad_norm": 0.22656987607479095,
+      "learning_rate": 9.394405060132473e-06,
+      "loss": 0.3998,
+      "step": 896
+    },
+    {
+      "epoch": 0.7286758732737612,
+      "grad_norm": 0.2439814805984497,
+      "learning_rate": 9.392148103131358e-06,
+      "loss": 0.3902,
+      "step": 897
+    },
+    {
+      "epoch": 0.7294882209585702,
+      "grad_norm": 0.22937926650047302,
+      "learning_rate": 9.3898872204402e-06,
+      "loss": 0.4034,
+      "step": 898
+    },
+    {
+      "epoch": 0.7303005686433793,
+      "grad_norm": 0.22087787091732025,
+      "learning_rate": 9.38762241407977e-06,
+      "loss": 0.4057,
+      "step": 899
+    },
+    {
+      "epoch": 0.7311129163281884,
+      "grad_norm": 0.24558734893798828,
+      "learning_rate": 9.38535368607434e-06,
+      "loss": 0.4054,
+      "step": 900
+    },
+    {
+      "epoch": 0.7319252640129975,
+      "grad_norm": 0.2345886379480362,
+      "learning_rate": 9.3830810384517e-06,
+      "loss": 0.3928,
+      "step": 901
+    },
+    {
+      "epoch": 0.7327376116978067,
+      "grad_norm": 0.3032779395580292,
+      "learning_rate": 9.380804473243133e-06,
+      "loss": 0.4007,
+      "step": 902
+    },
+    {
+      "epoch": 0.7335499593826158,
+      "grad_norm": 0.22959354519844055,
+      "learning_rate": 9.378523992483429e-06,
+      "loss": 0.3876,
+      "step": 903
+    },
+    {
+      "epoch": 0.7343623070674249,
+      "grad_norm": 0.2513275146484375,
+      "learning_rate": 9.376239598210873e-06,
+      "loss": 0.3991,
+      "step": 904
+    },
+    {
+      "epoch": 0.735174654752234,
+      "grad_norm": 0.23065359890460968,
+      "learning_rate": 9.37395129246725e-06,
+      "loss": 0.3947,
+      "step": 905
+    },
+    {
+      "epoch": 0.7359870024370431,
+      "grad_norm": 0.26329517364501953,
+      "learning_rate": 9.371659077297843e-06,
+      "loss": 0.4115,
+      "step": 906
+    },
+    {
+      "epoch": 0.7367993501218522,
+      "grad_norm": 0.2607138156890869,
+      "learning_rate": 9.369362954751428e-06,
+      "loss": 0.3944,
+      "step": 907
+    },
+    {
+      "epoch": 0.7376116978066612,
+      "grad_norm": 0.2535332143306732,
+      "learning_rate": 9.367062926880273e-06,
+      "loss": 0.3868,
+      "step": 908
+    },
+    {
+      "epoch": 0.7384240454914703,
+      "grad_norm": 0.27264413237571716,
+      "learning_rate": 9.364758995740136e-06,
+      "loss": 0.3958,
+      "step": 909
+    },
+    {
+      "epoch": 0.7392363931762794,
+      "grad_norm": 0.24130773544311523,
+      "learning_rate": 9.362451163390265e-06,
+      "loss": 0.3853,
+      "step": 910
+    },
+    {
+      "epoch": 0.7400487408610885,
+      "grad_norm": 0.24737271666526794,
+      "learning_rate": 9.360139431893394e-06,
+      "loss": 0.3884,
+      "step": 911
+    },
+    {
+      "epoch": 0.7408610885458976,
+      "grad_norm": 0.24589332938194275,
+      "learning_rate": 9.35782380331574e-06,
+      "loss": 0.3894,
+      "step": 912
+    },
+    {
+      "epoch": 0.7416734362307067,
+      "grad_norm": 0.25000646710395813,
+      "learning_rate": 9.35550427972701e-06,
+      "loss": 0.399,
+      "step": 913
+    },
+    {
+      "epoch": 0.7424857839155159,
+      "grad_norm": 0.22364689409732819,
+      "learning_rate": 9.353180863200385e-06,
+      "loss": 0.3717,
+      "step": 914
+    },
+    {
+      "epoch": 0.743298131600325,
+      "grad_norm": 0.24286335706710815,
+      "learning_rate": 9.350853555812529e-06,
+      "loss": 0.3819,
+      "step": 915
+    },
+    {
+      "epoch": 0.7441104792851341,
+      "grad_norm": 0.24053864181041718,
+      "learning_rate": 9.348522359643583e-06,
+      "loss": 0.4191,
+      "step": 916
+    },
+    {
+      "epoch": 0.7449228269699432,
+      "grad_norm": 0.21675679087638855,
+      "learning_rate": 9.346187276777163e-06,
+      "loss": 0.4036,
+      "step": 917
+    },
+    {
+      "epoch": 0.7457351746547523,
+      "grad_norm": 0.25289851427078247,
+      "learning_rate": 9.34384830930036e-06,
+      "loss": 0.4196,
+      "step": 918
+    },
+    {
+      "epoch": 0.7465475223395613,
+      "grad_norm": 0.2356320023536682,
+      "learning_rate": 9.34150545930374e-06,
+      "loss": 0.3825,
+      "step": 919
+    },
+    {
+      "epoch": 0.7473598700243704,
+      "grad_norm": 0.2313036173582077,
+      "learning_rate": 9.339158728881332e-06,
+      "loss": 0.414,
+      "step": 920
+    },
+    {
+      "epoch": 0.7481722177091795,
+      "grad_norm": 0.2682669460773468,
+      "learning_rate": 9.336808120130638e-06,
+      "loss": 0.4052,
+      "step": 921
+    },
+    {
+      "epoch": 0.7489845653939886,
+      "grad_norm": 0.24153448641300201,
+      "learning_rate": 9.334453635152628e-06,
+      "loss": 0.4022,
+      "step": 922
+    },
+    {
+      "epoch": 0.7497969130787977,
+      "grad_norm": 0.22754059731960297,
+      "learning_rate": 9.332095276051729e-06,
+      "loss": 0.4024,
+      "step": 923
+    },
+    {
+      "epoch": 0.7506092607636068,
+      "grad_norm": 0.23084460198879242,
+      "learning_rate": 9.329733044935843e-06,
+      "loss": 0.3979,
+      "step": 924
+    },
+    {
+      "epoch": 0.7514216084484159,
+      "grad_norm": 0.2530255913734436,
+      "learning_rate": 9.32736694391632e-06,
+      "loss": 0.4041,
+      "step": 925
+    },
+    {
+      "epoch": 0.752233956133225,
+      "grad_norm": 0.22373178601264954,
+      "learning_rate": 9.324996975107978e-06,
+      "loss": 0.387,
+      "step": 926
+    },
+    {
+      "epoch": 0.7530463038180342,
+      "grad_norm": 0.24156294763088226,
+      "learning_rate": 9.322623140629088e-06,
+      "loss": 0.3954,
+      "step": 927
+    },
+    {
+      "epoch": 0.7538586515028433,
+      "grad_norm": 0.24491259455680847,
+      "learning_rate": 9.320245442601377e-06,
+      "loss": 0.4031,
+      "step": 928
+    },
+    {
+      "epoch": 0.7546709991876523,
+      "grad_norm": 0.2385050505399704,
+      "learning_rate": 9.317863883150022e-06,
+      "loss": 0.3896,
+      "step": 929
+    },
+    {
+      "epoch": 0.7554833468724614,
+      "grad_norm": 0.2402818650007248,
+      "learning_rate": 9.31547846440366e-06,
+      "loss": 0.4097,
+      "step": 930
+    },
+    {
+      "epoch": 0.7562956945572705,
+      "grad_norm": 0.21302340924739838,
+      "learning_rate": 9.313089188494366e-06,
+      "loss": 0.3834,
+      "step": 931
+    },
+    {
+      "epoch": 0.7571080422420796,
+      "grad_norm": 0.2159285843372345,
+      "learning_rate": 9.31069605755767e-06,
+      "loss": 0.3914,
+      "step": 932
+    },
+    {
+      "epoch": 0.7579203899268887,
+      "grad_norm": 0.2634787857532501,
+      "learning_rate": 9.30829907373255e-06,
+      "loss": 0.4031,
+      "step": 933
+    },
+    {
+      "epoch": 0.7587327376116978,
+      "grad_norm": 0.2373684048652649,
+      "learning_rate": 9.30589823916142e-06,
+      "loss": 0.3893,
+      "step": 934
+    },
+    {
+      "epoch": 0.7595450852965069,
+      "grad_norm": 0.23416855931282043,
+      "learning_rate": 9.303493555990139e-06,
+      "loss": 0.4024,
+      "step": 935
+    },
+    {
+      "epoch": 0.760357432981316,
+      "grad_norm": 0.2544425129890442,
+      "learning_rate": 9.301085026368006e-06,
+      "loss": 0.3676,
+      "step": 936
+    },
+    {
+      "epoch": 0.7611697806661251,
+      "grad_norm": 0.22414150834083557,
+      "learning_rate": 9.298672652447763e-06,
+      "loss": 0.3909,
+      "step": 937
+    },
+    {
+      "epoch": 0.7619821283509342,
+      "grad_norm": 0.24809353053569794,
+      "learning_rate": 9.296256436385576e-06,
+      "loss": 0.4094,
+      "step": 938
+    },
+    {
+      "epoch": 0.7627944760357434,
+      "grad_norm": 0.24038027226924896,
+      "learning_rate": 9.293836380341059e-06,
+      "loss": 0.385,
+      "step": 939
+    },
+    {
+      "epoch": 0.7636068237205524,
+      "grad_norm": 0.2751830518245697,
+      "learning_rate": 9.291412486477247e-06,
+      "loss": 0.4025,
+      "step": 940
+    },
+    {
+      "epoch": 0.7644191714053615,
+      "grad_norm": 0.25096094608306885,
+      "learning_rate": 9.28898475696061e-06,
+      "loss": 0.3958,
+      "step": 941
+    },
+    {
+      "epoch": 0.7652315190901706,
+      "grad_norm": 0.2446352243423462,
+      "learning_rate": 9.28655319396105e-06,
+      "loss": 0.411,
+      "step": 942
+    },
+    {
+      "epoch": 0.7660438667749797,
+      "grad_norm": 0.23936989903450012,
+      "learning_rate": 9.284117799651887e-06,
+      "loss": 0.3928,
+      "step": 943
+    },
+    {
+      "epoch": 0.7668562144597888,
+      "grad_norm": 0.2408171445131302,
+      "learning_rate": 9.281678576209873e-06,
+      "loss": 0.3705,
+      "step": 944
+    },
+    {
+      "epoch": 0.7676685621445979,
+      "grad_norm": 0.28500789403915405,
+      "learning_rate": 9.279235525815177e-06,
+      "loss": 0.4063,
+      "step": 945
+    },
+    {
+      "epoch": 0.768480909829407,
+      "grad_norm": 0.23443956673145294,
+      "learning_rate": 9.276788650651392e-06,
+      "loss": 0.4006,
+      "step": 946
+    },
+    {
+      "epoch": 0.7692932575142161,
+      "grad_norm": 0.2724791467189789,
+      "learning_rate": 9.274337952905528e-06,
+      "loss": 0.393,
+      "step": 947
+    },
+    {
+      "epoch": 0.7701056051990252,
+      "grad_norm": 0.2770431935787201,
+      "learning_rate": 9.271883434768012e-06,
+      "loss": 0.4347,
+      "step": 948
+    },
+    {
+      "epoch": 0.7709179528838342,
+      "grad_norm": 0.2513534724712372,
+      "learning_rate": 9.269425098432686e-06,
+      "loss": 0.4087,
+      "step": 949
+    },
+    {
+      "epoch": 0.7717303005686433,
+      "grad_norm": 0.26289039850234985,
+      "learning_rate": 9.266962946096802e-06,
+      "loss": 0.4158,
+      "step": 950
+    },
+    {
+      "epoch": 0.7725426482534524,
+      "grad_norm": 0.28713810443878174,
+      "learning_rate": 9.264496979961031e-06,
+      "loss": 0.4023,
+      "step": 951
+    },
+    {
+      "epoch": 0.7733549959382616,
+      "grad_norm": 0.24732619524002075,
+      "learning_rate": 9.26202720222944e-06,
+      "loss": 0.3816,
+      "step": 952
+    },
+    {
+      "epoch": 0.7741673436230707,
+      "grad_norm": 0.29945671558380127,
+      "learning_rate": 9.259553615109514e-06,
+      "loss": 0.4127,
+      "step": 953
+    },
+    {
+      "epoch": 0.7749796913078798,
+      "grad_norm": 0.24035924673080444,
+      "learning_rate": 9.257076220812136e-06,
+      "loss": 0.4225,
+      "step": 954
+    },
+    {
+      "epoch": 0.7757920389926889,
+      "grad_norm": 0.27464190125465393,
+      "learning_rate": 9.254595021551595e-06,
+      "loss": 0.3798,
+      "step": 955
+    },
+    {
+      "epoch": 0.776604386677498,
+      "grad_norm": 0.2515372931957245,
+      "learning_rate": 9.252110019545581e-06,
+      "loss": 0.4197,
+      "step": 956
+    },
+    {
+      "epoch": 0.7774167343623071,
+      "grad_norm": 0.24792397022247314,
+      "learning_rate": 9.249621217015182e-06,
+      "loss": 0.393,
+      "step": 957
+    },
+    {
+      "epoch": 0.7782290820471162,
+      "grad_norm": 0.2802797257900238,
+      "learning_rate": 9.247128616184881e-06,
+      "loss": 0.3962,
+      "step": 958
+    },
+    {
+      "epoch": 0.7790414297319253,
+      "grad_norm": 0.22957704961299896,
+      "learning_rate": 9.244632219282561e-06,
+      "loss": 0.3637,
+      "step": 959
+    },
+    {
+      "epoch": 0.7798537774167343,
+      "grad_norm": 0.2474144995212555,
+      "learning_rate": 9.242132028539493e-06,
+      "loss": 0.4004,
+      "step": 960
+    },
+    {
+      "epoch": 0.7806661251015434,
+      "grad_norm": 0.25579920411109924,
+      "learning_rate": 9.239628046190342e-06,
+      "loss": 0.404,
+      "step": 961
+    },
+    {
+      "epoch": 0.7814784727863525,
+      "grad_norm": 0.24421872198581696,
+      "learning_rate": 9.237120274473157e-06,
+      "loss": 0.3766,
+      "step": 962
+    },
+    {
+      "epoch": 0.7822908204711616,
+      "grad_norm": 0.25123193860054016,
+      "learning_rate": 9.23460871562938e-06,
+      "loss": 0.4081,
+      "step": 963
+    },
+    {
+      "epoch": 0.7831031681559708,
+      "grad_norm": 0.2396664023399353,
+      "learning_rate": 9.232093371903836e-06,
+      "loss": 0.3913,
+      "step": 964
+    },
+    {
+      "epoch": 0.7839155158407799,
+      "grad_norm": 0.22551880776882172,
+      "learning_rate": 9.229574245544732e-06,
+      "loss": 0.3898,
+      "step": 965
+    },
+    {
+      "epoch": 0.784727863525589,
+      "grad_norm": 0.23533564805984497,
+      "learning_rate": 9.227051338803656e-06,
+      "loss": 0.3953,
+      "step": 966
+    },
+    {
+      "epoch": 0.7855402112103981,
+      "grad_norm": 0.22905725240707397,
+      "learning_rate": 9.224524653935577e-06,
+      "loss": 0.3945,
+      "step": 967
+    },
+    {
+      "epoch": 0.7863525588952072,
+      "grad_norm": 0.2537819445133209,
+      "learning_rate": 9.221994193198835e-06,
+      "loss": 0.4128,
+      "step": 968
+    },
+    {
+      "epoch": 0.7871649065800163,
+      "grad_norm": 0.2295469045639038,
+      "learning_rate": 9.219459958855152e-06,
+      "loss": 0.3943,
+      "step": 969
+    },
+    {
+      "epoch": 0.7879772542648253,
+      "grad_norm": 0.205052450299263,
+      "learning_rate": 9.216921953169618e-06,
+      "loss": 0.3819,
+      "step": 970
+    },
+    {
+      "epoch": 0.7887896019496344,
+      "grad_norm": 0.2627088129520416,
+      "learning_rate": 9.214380178410698e-06,
+      "loss": 0.4233,
+      "step": 971
+    },
+    {
+      "epoch": 0.7896019496344435,
+      "grad_norm": 0.24686412513256073,
+      "learning_rate": 9.211834636850222e-06,
+      "loss": 0.4155,
+      "step": 972
+    },
+    {
+      "epoch": 0.7904142973192526,
+      "grad_norm": 0.23723474144935608,
+      "learning_rate": 9.209285330763388e-06,
+      "loss": 0.3785,
+      "step": 973
+    },
+    {
+      "epoch": 0.7912266450040617,
+      "grad_norm": 0.26609280705451965,
+      "learning_rate": 9.20673226242876e-06,
+      "loss": 0.3958,
+      "step": 974
+    },
+    {
+      "epoch": 0.7920389926888708,
+      "grad_norm": 0.2366759032011032,
+      "learning_rate": 9.204175434128259e-06,
+      "loss": 0.4142,
+      "step": 975
+    },
+    {
+      "epoch": 0.7928513403736799,
+      "grad_norm": 0.2277480810880661,
+      "learning_rate": 9.201614848147179e-06,
+      "loss": 0.4103,
+      "step": 976
+    },
+    {
+      "epoch": 0.7936636880584891,
+      "grad_norm": 0.2492954581975937,
+      "learning_rate": 9.199050506774159e-06,
+      "loss": 0.4147,
+      "step": 977
+    },
+    {
+      "epoch": 0.7944760357432982,
+      "grad_norm": 0.25515016913414,
+      "learning_rate": 9.196482412301202e-06,
+      "loss": 0.4337,
+      "step": 978
+    },
+    {
+      "epoch": 0.7952883834281073,
+      "grad_norm": 0.2385532706975937,
+      "learning_rate": 9.193910567023663e-06,
+      "loss": 0.411,
+      "step": 979
+    },
+    {
+      "epoch": 0.7961007311129163,
+      "grad_norm": 0.27272728085517883,
+      "learning_rate": 9.19133497324025e-06,
+      "loss": 0.3722,
+      "step": 980
+    },
+    {
+      "epoch": 0.7969130787977254,
+      "grad_norm": 0.26492252945899963,
+      "learning_rate": 9.188755633253025e-06,
+      "loss": 0.4057,
+      "step": 981
+    },
+    {
+      "epoch": 0.7977254264825345,
+      "grad_norm": 0.2597847878932953,
+      "learning_rate": 9.186172549367394e-06,
+      "loss": 0.4018,
+      "step": 982
+    },
+    {
+      "epoch": 0.7985377741673436,
+      "grad_norm": 0.2276986837387085,
+      "learning_rate": 9.183585723892107e-06,
+      "loss": 0.3907,
+      "step": 983
+    },
+    {
+      "epoch": 0.7993501218521527,
+      "grad_norm": 0.2739187777042389,
+      "learning_rate": 9.180995159139265e-06,
+      "loss": 0.3947,
+      "step": 984
+    },
+    {
+      "epoch": 0.8001624695369618,
+      "grad_norm": 0.23972639441490173,
+      "learning_rate": 9.178400857424305e-06,
+      "loss": 0.4002,
+      "step": 985
+    },
+    {
+      "epoch": 0.8009748172217709,
+      "grad_norm": 0.24927566945552826,
+      "learning_rate": 9.175802821066009e-06,
+      "loss": 0.3938,
+      "step": 986
+    },
+    {
+      "epoch": 0.80178716490658,
+      "grad_norm": 0.21761095523834229,
+      "learning_rate": 9.173201052386496e-06,
+      "loss": 0.4018,
+      "step": 987
+    },
+    {
+      "epoch": 0.8025995125913891,
+      "grad_norm": 0.24998874962329865,
+      "learning_rate": 9.170595553711216e-06,
+      "loss": 0.403,
+      "step": 988
+    },
+    {
+      "epoch": 0.8034118602761983,
+      "grad_norm": 0.22575093805789948,
+      "learning_rate": 9.167986327368958e-06,
+      "loss": 0.3857,
+      "step": 989
+    },
+    {
+      "epoch": 0.8042242079610074,
+      "grad_norm": 0.24034999310970306,
+      "learning_rate": 9.165373375691845e-06,
+      "loss": 0.4066,
+      "step": 990
+    },
+    {
+      "epoch": 0.8050365556458164,
+      "grad_norm": 0.26667869091033936,
+      "learning_rate": 9.16275670101532e-06,
+      "loss": 0.3895,
+      "step": 991
+    },
+    {
+      "epoch": 0.8058489033306255,
+      "grad_norm": 0.2402062565088272,
+      "learning_rate": 9.160136305678167e-06,
+      "loss": 0.4113,
+      "step": 992
+    },
+    {
+      "epoch": 0.8066612510154346,
+      "grad_norm": 0.2527025640010834,
+      "learning_rate": 9.15751219202248e-06,
+      "loss": 0.386,
+      "step": 993
+    },
+    {
+      "epoch": 0.8074735987002437,
+      "grad_norm": 0.22969551384449005,
+      "learning_rate": 9.15488436239369e-06,
+      "loss": 0.4029,
+      "step": 994
+    },
+    {
+      "epoch": 0.8082859463850528,
+      "grad_norm": 0.24026134610176086,
+      "learning_rate": 9.152252819140544e-06,
+      "loss": 0.4035,
+      "step": 995
+    },
+    {
+      "epoch": 0.8090982940698619,
+      "grad_norm": 0.2233525961637497,
+      "learning_rate": 9.149617564615106e-06,
+      "loss": 0.3681,
+      "step": 996
+    },
+    {
+      "epoch": 0.809910641754671,
+      "grad_norm": 0.24326178431510925,
+      "learning_rate": 9.146978601172761e-06,
+      "loss": 0.3791,
+      "step": 997
+    },
+    {
+      "epoch": 0.8107229894394801,
+      "grad_norm": 0.23843257129192352,
+      "learning_rate": 9.144335931172205e-06,
+      "loss": 0.4307,
+      "step": 998
+    },
+    {
+      "epoch": 0.8115353371242892,
+      "grad_norm": 0.24528557062149048,
+      "learning_rate": 9.141689556975453e-06,
+      "loss": 0.4112,
+      "step": 999
+    },
+    {
+      "epoch": 0.8123476848090982,
+      "grad_norm": 0.23784443736076355,
+      "learning_rate": 9.139039480947825e-06,
+      "loss": 0.396,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8131600324939074,
+      "grad_norm": 0.25523674488067627,
+      "learning_rate": 9.136385705457951e-06,
+      "loss": 0.3699,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8139723801787165,
+      "grad_norm": 0.23561649024486542,
+      "learning_rate": 9.13372823287777e-06,
+      "loss": 0.4084,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8147847278635256,
+      "grad_norm": 0.2523971199989319,
+      "learning_rate": 9.131067065582521e-06,
+      "loss": 0.4152,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8155970755483347,
+      "grad_norm": 0.2359113097190857,
+      "learning_rate": 9.128402205950751e-06,
+      "loss": 0.3894,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8164094232331438,
+      "grad_norm": 0.2240229845046997,
+      "learning_rate": 9.125733656364304e-06,
+      "loss": 0.4031,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8172217709179529,
+      "grad_norm": 0.25402912497520447,
+      "learning_rate": 9.12306141920832e-06,
+      "loss": 0.4132,
+      "step": 1006
+    },
+    {
+      "epoch": 0.818034118602762,
+      "grad_norm": 0.22948139905929565,
+      "learning_rate": 9.12038549687124e-06,
+      "loss": 0.4052,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8188464662875711,
+      "grad_norm": 0.23393946886062622,
+      "learning_rate": 9.117705891744795e-06,
+      "loss": 0.4274,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8196588139723802,
+      "grad_norm": 0.23718692362308502,
+      "learning_rate": 9.115022606224008e-06,
+      "loss": 0.4059,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8204711616571893,
+      "grad_norm": 0.22119073569774628,
+      "learning_rate": 9.112335642707196e-06,
+      "loss": 0.4011,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8212835093419983,
+      "grad_norm": 0.2506997585296631,
+      "learning_rate": 9.109645003595954e-06,
+      "loss": 0.411,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8220958570268074,
+      "grad_norm": 0.24170063436031342,
+      "learning_rate": 9.106950691295172e-06,
+      "loss": 0.4066,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8229082047116165,
+      "grad_norm": 0.2074459344148636,
+      "learning_rate": 9.104252708213018e-06,
+      "loss": 0.356,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8237205523964257,
+      "grad_norm": 0.2900426387786865,
+      "learning_rate": 9.101551056760942e-06,
+      "loss": 0.4021,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8245329000812348,
+      "grad_norm": 0.23278403282165527,
+      "learning_rate": 9.098845739353672e-06,
+      "loss": 0.3992,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8253452477660439,
+      "grad_norm": 0.23077310621738434,
+      "learning_rate": 9.096136758409212e-06,
+      "loss": 0.4266,
+      "step": 1016
+    },
+    {
+      "epoch": 0.826157595450853,
+      "grad_norm": 0.23523063957691193,
+      "learning_rate": 9.093424116348846e-06,
+      "loss": 0.4088,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8269699431356621,
+      "grad_norm": 0.23515664041042328,
+      "learning_rate": 9.090707815597124e-06,
+      "loss": 0.4081,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8277822908204712,
+      "grad_norm": 0.22891110181808472,
+      "learning_rate": 9.087987858581867e-06,
+      "loss": 0.4227,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8285946385052803,
+      "grad_norm": 0.22268006205558777,
+      "learning_rate": 9.085264247734165e-06,
+      "loss": 0.3958,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8294069861900893,
+      "grad_norm": 0.22710324823856354,
+      "learning_rate": 9.082536985488377e-06,
+      "loss": 0.3903,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8302193338748984,
+      "grad_norm": 0.20737770199775696,
+      "learning_rate": 9.07980607428212e-06,
+      "loss": 0.3913,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8310316815597075,
+      "grad_norm": 0.23445822298526764,
+      "learning_rate": 9.077071516556276e-06,
+      "loss": 0.3966,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8318440292445166,
+      "grad_norm": 0.2507627606391907,
+      "learning_rate": 9.074333314754984e-06,
+      "loss": 0.3843,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8326563769293257,
+      "grad_norm": 0.20807425677776337,
+      "learning_rate": 9.071591471325644e-06,
+      "loss": 0.3915,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8334687246141349,
+      "grad_norm": 0.23468923568725586,
+      "learning_rate": 9.068845988718906e-06,
+      "loss": 0.4023,
+      "step": 1026
+    },
+    {
+      "epoch": 0.834281072298944,
+      "grad_norm": 0.2682033181190491,
+      "learning_rate": 9.066096869388674e-06,
+      "loss": 0.3787,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8350934199837531,
+      "grad_norm": 0.22748582065105438,
+      "learning_rate": 9.063344115792108e-06,
+      "loss": 0.3864,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8359057676685622,
+      "grad_norm": 0.22402749955654144,
+      "learning_rate": 9.060587730389606e-06,
+      "loss": 0.4091,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8367181153533713,
+      "grad_norm": 0.2373138815164566,
+      "learning_rate": 9.05782771564482e-06,
+      "loss": 0.4085,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8375304630381804,
+      "grad_norm": 0.22854498028755188,
+      "learning_rate": 9.055064074024646e-06,
+      "loss": 0.4285,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8383428107229894,
+      "grad_norm": 0.2209593653678894,
+      "learning_rate": 9.052296807999216e-06,
+      "loss": 0.3879,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8391551584077985,
+      "grad_norm": 0.2887367904186249,
+      "learning_rate": 9.049525920041909e-06,
+      "loss": 0.4264,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8399675060926076,
+      "grad_norm": 0.22837641835212708,
+      "learning_rate": 9.046751412629331e-06,
+      "loss": 0.395,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8407798537774167,
+      "grad_norm": 0.26165908575057983,
+      "learning_rate": 9.043973288241334e-06,
+      "loss": 0.4018,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8415922014622258,
+      "grad_norm": 0.25582313537597656,
+      "learning_rate": 9.041191549360995e-06,
+      "loss": 0.4002,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8424045491470349,
+      "grad_norm": 0.24917741119861603,
+      "learning_rate": 9.038406198474627e-06,
+      "loss": 0.3869,
+      "step": 1037
+    },
+    {
+      "epoch": 0.843216896831844,
+      "grad_norm": 0.2176097333431244,
+      "learning_rate": 9.03561723807177e-06,
+      "loss": 0.3962,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8440292445166532,
+      "grad_norm": 0.24159877002239227,
+      "learning_rate": 9.032824670645187e-06,
+      "loss": 0.39,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8448415922014623,
+      "grad_norm": 0.2842418849468231,
+      "learning_rate": 9.030028498690866e-06,
+      "loss": 0.4279,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8456539398862714,
+      "grad_norm": 0.22591236233711243,
+      "learning_rate": 9.027228724708022e-06,
+      "loss": 0.3955,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8464662875710804,
+      "grad_norm": 0.22790557146072388,
+      "learning_rate": 9.02442535119908e-06,
+      "loss": 0.375,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8472786352558895,
+      "grad_norm": 0.5459648966789246,
+      "learning_rate": 9.021618380669693e-06,
+      "loss": 0.3736,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8480909829406986,
+      "grad_norm": 0.25459349155426025,
+      "learning_rate": 9.018807815628721e-06,
+      "loss": 0.3839,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8489033306255077,
+      "grad_norm": 0.24523377418518066,
+      "learning_rate": 9.01599365858824e-06,
+      "loss": 0.3864,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8497156783103168,
+      "grad_norm": 0.2346375584602356,
+      "learning_rate": 9.013175912063534e-06,
+      "loss": 0.3857,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8505280259951259,
+      "grad_norm": 0.22873444855213165,
+      "learning_rate": 9.0103545785731e-06,
+      "loss": 0.3881,
+      "step": 1047
+    },
+    {
+      "epoch": 0.851340373679935,
+      "grad_norm": 0.25674566626548767,
+      "learning_rate": 9.007529660638637e-06,
+      "loss": 0.3811,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8521527213647441,
+      "grad_norm": 0.25758472084999084,
+      "learning_rate": 9.004701160785047e-06,
+      "loss": 0.3985,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8529650690495532,
+      "grad_norm": 0.21978415548801422,
+      "learning_rate": 9.00186908154044e-06,
+      "loss": 0.4073,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8537774167343624,
+      "grad_norm": 0.25125598907470703,
+      "learning_rate": 8.999033425436116e-06,
+      "loss": 0.4026,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8545897644191714,
+      "grad_norm": 0.24195928871631622,
+      "learning_rate": 8.996194195006582e-06,
+      "loss": 0.392,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8554021121039805,
+      "grad_norm": 0.24349184334278107,
+      "learning_rate": 8.993351392789529e-06,
+      "loss": 0.3946,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8562144597887896,
+      "grad_norm": 0.2190394550561905,
+      "learning_rate": 8.990505021325849e-06,
+      "loss": 0.4017,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8570268074735987,
+      "grad_norm": 0.22362494468688965,
+      "learning_rate": 8.987655083159618e-06,
+      "loss": 0.4248,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8578391551584078,
+      "grad_norm": 0.2421010285615921,
+      "learning_rate": 8.984801580838109e-06,
+      "loss": 0.4115,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8586515028432169,
+      "grad_norm": 0.22442738711833954,
+      "learning_rate": 8.98194451691177e-06,
+      "loss": 0.3758,
+      "step": 1057
+    },
+    {
+      "epoch": 0.859463850528026,
+      "grad_norm": 0.23938056826591492,
+      "learning_rate": 8.979083893934236e-06,
+      "loss": 0.4036,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8602761982128351,
+      "grad_norm": 0.2265082597732544,
+      "learning_rate": 8.976219714462326e-06,
+      "loss": 0.4398,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8610885458976442,
+      "grad_norm": 0.2464844286441803,
+      "learning_rate": 8.973351981056037e-06,
+      "loss": 0.3981,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8619008935824533,
+      "grad_norm": 0.25511249899864197,
+      "learning_rate": 8.97048069627854e-06,
+      "loss": 0.3968,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8627132412672623,
+      "grad_norm": 0.21273177862167358,
+      "learning_rate": 8.967605862696183e-06,
+      "loss": 0.4136,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8635255889520714,
+      "grad_norm": 0.2335020750761032,
+      "learning_rate": 8.964727482878482e-06,
+      "loss": 0.3786,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8643379366368806,
+      "grad_norm": 0.2552463412284851,
+      "learning_rate": 8.961845559398125e-06,
+      "loss": 0.391,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8651502843216897,
+      "grad_norm": 0.22984963655471802,
+      "learning_rate": 8.95896009483097e-06,
+      "loss": 0.3989,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8659626320064988,
+      "grad_norm": 0.23188404738903046,
+      "learning_rate": 8.956071091756036e-06,
+      "loss": 0.4012,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8667749796913079,
+      "grad_norm": 0.23402655124664307,
+      "learning_rate": 8.953178552755506e-06,
+      "loss": 0.4109,
+      "step": 1067
+    },
+    {
+      "epoch": 0.867587327376117,
+      "grad_norm": 0.23716522753238678,
+      "learning_rate": 8.950282480414723e-06,
+      "loss": 0.4174,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8683996750609261,
+      "grad_norm": 0.23065772652626038,
+      "learning_rate": 8.947382877322189e-06,
+      "loss": 0.4137,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8692120227457352,
+      "grad_norm": 0.22798073291778564,
+      "learning_rate": 8.94447974606956e-06,
+      "loss": 0.3987,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8700243704305443,
+      "grad_norm": 0.21862100064754486,
+      "learning_rate": 8.941573089251652e-06,
+      "loss": 0.3901,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8708367181153533,
+      "grad_norm": 0.2147633284330368,
+      "learning_rate": 8.93866290946642e-06,
+      "loss": 0.4047,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8716490658001624,
+      "grad_norm": 0.23569658398628235,
+      "learning_rate": 8.935749209314981e-06,
+      "loss": 0.4179,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8724614134849715,
+      "grad_norm": 0.23709242045879364,
+      "learning_rate": 8.932831991401587e-06,
+      "loss": 0.4247,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8732737611697806,
+      "grad_norm": 0.2109891027212143,
+      "learning_rate": 8.929911258333644e-06,
+      "loss": 0.3968,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8740861088545898,
+      "grad_norm": 0.2774312496185303,
+      "learning_rate": 8.92698701272169e-06,
+      "loss": 0.4071,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8748984565393989,
+      "grad_norm": 0.2340877503156662,
+      "learning_rate": 8.924059257179414e-06,
+      "loss": 0.3972,
+      "step": 1077
+    },
+    {
+      "epoch": 0.875710804224208,
+      "grad_norm": 0.21561944484710693,
+      "learning_rate": 8.92112799432363e-06,
+      "loss": 0.4203,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8765231519090171,
+      "grad_norm": 0.2575530409812927,
+      "learning_rate": 8.9181932267743e-06,
+      "loss": 0.3992,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8773354995938262,
+      "grad_norm": 0.2672640085220337,
+      "learning_rate": 8.915254957154501e-06,
+      "loss": 0.4119,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8781478472786353,
+      "grad_norm": 0.2290709763765335,
+      "learning_rate": 8.91231318809046e-06,
+      "loss": 0.3979,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8789601949634444,
+      "grad_norm": 0.22523616254329681,
+      "learning_rate": 8.909367922211519e-06,
+      "loss": 0.4274,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8797725426482534,
+      "grad_norm": 0.26314398646354675,
+      "learning_rate": 8.906419162150147e-06,
+      "loss": 0.3898,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8805848903330625,
+      "grad_norm": 0.29318350553512573,
+      "learning_rate": 8.903466910541938e-06,
+      "loss": 0.4135,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8813972380178716,
+      "grad_norm": 0.2129216343164444,
+      "learning_rate": 8.900511170025609e-06,
+      "loss": 0.4006,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8822095857026807,
+      "grad_norm": 0.25386422872543335,
+      "learning_rate": 8.89755194324299e-06,
+      "loss": 0.3902,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8830219333874898,
+      "grad_norm": 0.2543341815471649,
+      "learning_rate": 8.89458923283903e-06,
+      "loss": 0.3786,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8838342810722989,
+      "grad_norm": 0.23379240930080414,
+      "learning_rate": 8.891623041461794e-06,
+      "loss": 0.3938,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8846466287571081,
+      "grad_norm": 0.22442100942134857,
+      "learning_rate": 8.888653371762454e-06,
+      "loss": 0.4129,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8854589764419172,
+      "grad_norm": 0.2562439441680908,
+      "learning_rate": 8.885680226395295e-06,
+      "loss": 0.4165,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8862713241267263,
+      "grad_norm": 0.2497139424085617,
+      "learning_rate": 8.882703608017702e-06,
+      "loss": 0.3924,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8870836718115354,
+      "grad_norm": 0.22368566691875458,
+      "learning_rate": 8.87972351929017e-06,
+      "loss": 0.3859,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8878960194963444,
+      "grad_norm": 0.29025521874427795,
+      "learning_rate": 8.876739962876298e-06,
+      "loss": 0.4153,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8887083671811535,
+      "grad_norm": 0.2524016499519348,
+      "learning_rate": 8.873752941442775e-06,
+      "loss": 0.3883,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8895207148659626,
+      "grad_norm": 0.21893171966075897,
+      "learning_rate": 8.870762457659397e-06,
+      "loss": 0.4116,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8903330625507717,
+      "grad_norm": 0.2526397407054901,
+      "learning_rate": 8.867768514199046e-06,
+      "loss": 0.4011,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8911454102355808,
+      "grad_norm": 0.27337589859962463,
+      "learning_rate": 8.864771113737705e-06,
+      "loss": 0.3956,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8919577579203899,
+      "grad_norm": 0.25696468353271484,
+      "learning_rate": 8.86177025895444e-06,
+      "loss": 0.3922,
+      "step": 1098
+    },
+    {
+      "epoch": 0.892770105605199,
+      "grad_norm": 0.23625078797340393,
+      "learning_rate": 8.858765952531407e-06,
+      "loss": 0.3872,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8935824532900081,
+      "grad_norm": 0.21721084415912628,
+      "learning_rate": 8.855758197153848e-06,
+      "loss": 0.3857,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8943948009748173,
+      "grad_norm": 0.2577703297138214,
+      "learning_rate": 8.852746995510084e-06,
+      "loss": 0.399,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8952071486596264,
+      "grad_norm": 0.24921587109565735,
+      "learning_rate": 8.849732350291522e-06,
+      "loss": 0.4054,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8960194963444355,
+      "grad_norm": 0.22682757675647736,
+      "learning_rate": 8.846714264192642e-06,
+      "loss": 0.4024,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8968318440292445,
+      "grad_norm": 0.22311890125274658,
+      "learning_rate": 8.843692739911e-06,
+      "loss": 0.3926,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8976441917140536,
+      "grad_norm": 0.2570987045764923,
+      "learning_rate": 8.84066778014723e-06,
+      "loss": 0.3897,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8984565393988627,
+      "grad_norm": 0.21885989606380463,
+      "learning_rate": 8.837639387605031e-06,
+      "loss": 0.3652,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8992688870836718,
+      "grad_norm": 0.2354506254196167,
+      "learning_rate": 8.834607564991168e-06,
+      "loss": 0.4225,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9000812347684809,
+      "grad_norm": 0.24837671220302582,
+      "learning_rate": 8.831572315015484e-06,
+      "loss": 0.3697,
+      "step": 1108
+    },
+    {
+      "epoch": 0.90089358245329,
+      "grad_norm": 0.2783718407154083,
+      "learning_rate": 8.828533640390868e-06,
+      "loss": 0.3949,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9017059301380991,
+      "grad_norm": 0.22809027135372162,
+      "learning_rate": 8.825491543833286e-06,
+      "loss": 0.4026,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9025182778229082,
+      "grad_norm": 0.23428133130073547,
+      "learning_rate": 8.822446028061752e-06,
+      "loss": 0.4026,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9033306255077173,
+      "grad_norm": 0.25569191575050354,
+      "learning_rate": 8.819397095798343e-06,
+      "loss": 0.3778,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9041429731925265,
+      "grad_norm": 0.22279872000217438,
+      "learning_rate": 8.816344749768184e-06,
+      "loss": 0.3889,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9049553208773355,
+      "grad_norm": 0.23801083862781525,
+      "learning_rate": 8.813288992699454e-06,
+      "loss": 0.382,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9057676685621446,
+      "grad_norm": 0.2438342124223709,
+      "learning_rate": 8.810229827323378e-06,
+      "loss": 0.3969,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9065800162469537,
+      "grad_norm": 0.2055959403514862,
+      "learning_rate": 8.807167256374234e-06,
+      "loss": 0.3576,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9073923639317628,
+      "grad_norm": 0.2217477262020111,
+      "learning_rate": 8.804101282589338e-06,
+      "loss": 0.4118,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9082047116165719,
+      "grad_norm": 0.24302777647972107,
+      "learning_rate": 8.801031908709047e-06,
+      "loss": 0.4039,
+      "step": 1118
+    },
+    {
+      "epoch": 0.909017059301381,
+      "grad_norm": 0.2164711207151413,
+      "learning_rate": 8.797959137476763e-06,
+      "loss": 0.3896,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9098294069861901,
+      "grad_norm": 0.22885668277740479,
+      "learning_rate": 8.794882971638916e-06,
+      "loss": 0.3978,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9106417546709992,
+      "grad_norm": 0.23751376569271088,
+      "learning_rate": 8.791803413944979e-06,
+      "loss": 0.4014,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9114541023558083,
+      "grad_norm": 0.23991163074970245,
+      "learning_rate": 8.78872046714745e-06,
+      "loss": 0.4295,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9122664500406173,
+      "grad_norm": 0.22737851738929749,
+      "learning_rate": 8.785634134001856e-06,
+      "loss": 0.3798,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9130787977254264,
+      "grad_norm": 0.22865904867649078,
+      "learning_rate": 8.782544417266758e-06,
+      "loss": 0.3828,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9138911454102355,
+      "grad_norm": 0.2549517750740051,
+      "learning_rate": 8.779451319703735e-06,
+      "loss": 0.4,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9147034930950447,
+      "grad_norm": 0.21218614280223846,
+      "learning_rate": 8.776354844077389e-06,
+      "loss": 0.3728,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9155158407798538,
+      "grad_norm": 0.23499026894569397,
+      "learning_rate": 8.77325499315534e-06,
+      "loss": 0.4311,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9163281884646629,
+      "grad_norm": 0.23233026266098022,
+      "learning_rate": 8.770151769708226e-06,
+      "loss": 0.394,
+      "step": 1128
+    },
+    {
+      "epoch": 0.917140536149472,
+      "grad_norm": 0.23237285017967224,
+      "learning_rate": 8.767045176509703e-06,
+      "loss": 0.409,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9179528838342811,
+      "grad_norm": 0.26131319999694824,
+      "learning_rate": 8.763935216336433e-06,
+      "loss": 0.3846,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9187652315190902,
+      "grad_norm": 0.23263418674468994,
+      "learning_rate": 8.76082189196809e-06,
+      "loss": 0.3875,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9195775792038993,
+      "grad_norm": 0.25382182002067566,
+      "learning_rate": 8.757705206187357e-06,
+      "loss": 0.3763,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9203899268887084,
+      "grad_norm": 0.24425232410430908,
+      "learning_rate": 8.754585161779916e-06,
+      "loss": 0.4166,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9212022745735174,
+      "grad_norm": 0.23813597857952118,
+      "learning_rate": 8.751461761534457e-06,
+      "loss": 0.3808,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9220146222583265,
+      "grad_norm": 0.25072214007377625,
+      "learning_rate": 8.748335008242667e-06,
+      "loss": 0.4271,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9228269699431356,
+      "grad_norm": 0.2326296716928482,
+      "learning_rate": 8.745204904699226e-06,
+      "loss": 0.3867,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9236393176279447,
+      "grad_norm": 0.24460066854953766,
+      "learning_rate": 8.742071453701815e-06,
+      "loss": 0.4061,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9244516653127539,
+      "grad_norm": 0.23288196325302124,
+      "learning_rate": 8.738934658051103e-06,
+      "loss": 0.4259,
+      "step": 1138
+    },
+    {
+      "epoch": 0.925264012997563,
+      "grad_norm": 0.2466251254081726,
+      "learning_rate": 8.73579452055075e-06,
+      "loss": 0.397,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9260763606823721,
+      "grad_norm": 0.23213860392570496,
+      "learning_rate": 8.732651044007402e-06,
+      "loss": 0.3762,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9268887083671812,
+      "grad_norm": 0.23878903687000275,
+      "learning_rate": 8.72950423123069e-06,
+      "loss": 0.3846,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9277010560519903,
+      "grad_norm": 0.24751131236553192,
+      "learning_rate": 8.726354085033227e-06,
+      "loss": 0.4199,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9285134037367994,
+      "grad_norm": 0.20475398004055023,
+      "learning_rate": 8.723200608230605e-06,
+      "loss": 0.3832,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9293257514216084,
+      "grad_norm": 0.22067125141620636,
+      "learning_rate": 8.72004380364139e-06,
+      "loss": 0.3904,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9301380991064175,
+      "grad_norm": 0.24281468987464905,
+      "learning_rate": 8.71688367408713e-06,
+      "loss": 0.3999,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9309504467912266,
+      "grad_norm": 0.23550941050052643,
+      "learning_rate": 8.713720222392338e-06,
+      "loss": 0.4057,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9317627944760357,
+      "grad_norm": 0.22814086079597473,
+      "learning_rate": 8.710553451384499e-06,
+      "loss": 0.3824,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9325751421608448,
+      "grad_norm": 0.21902748942375183,
+      "learning_rate": 8.707383363894064e-06,
+      "loss": 0.3946,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9333874898456539,
+      "grad_norm": 0.24229693412780762,
+      "learning_rate": 8.704209962754448e-06,
+      "loss": 0.3826,
+      "step": 1149
+    },
+    {
+      "epoch": 0.934199837530463,
+      "grad_norm": 0.24235258996486664,
+      "learning_rate": 8.70103325080203e-06,
+      "loss": 0.3883,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9350121852152722,
+      "grad_norm": 0.23495063185691833,
+      "learning_rate": 8.697853230876145e-06,
+      "loss": 0.4151,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9358245329000813,
+      "grad_norm": 0.23322738707065582,
+      "learning_rate": 8.694669905819087e-06,
+      "loss": 0.3845,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9366368805848904,
+      "grad_norm": 0.214872807264328,
+      "learning_rate": 8.691483278476103e-06,
+      "loss": 0.3852,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9374492282696995,
+      "grad_norm": 0.23536399006843567,
+      "learning_rate": 8.688293351695392e-06,
+      "loss": 0.3866,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9382615759545085,
+      "grad_norm": 0.22190436720848083,
+      "learning_rate": 8.685100128328101e-06,
+      "loss": 0.3741,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9390739236393176,
+      "grad_norm": 0.23180842399597168,
+      "learning_rate": 8.681903611228327e-06,
+      "loss": 0.3984,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9398862713241267,
+      "grad_norm": 0.22982825338840485,
+      "learning_rate": 8.678703803253103e-06,
+      "loss": 0.3939,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9406986190089358,
+      "grad_norm": 0.24492645263671875,
+      "learning_rate": 8.675500707262415e-06,
+      "loss": 0.4002,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9415109666937449,
+      "grad_norm": 0.22684451937675476,
+      "learning_rate": 8.672294326119177e-06,
+      "loss": 0.3887,
+      "step": 1159
+    },
+    {
+      "epoch": 0.942323314378554,
+      "grad_norm": 0.21713843941688538,
+      "learning_rate": 8.669084662689246e-06,
+      "loss": 0.3855,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9431356620633631,
+      "grad_norm": 0.26320168375968933,
+      "learning_rate": 8.66587171984141e-06,
+      "loss": 0.3949,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9439480097481722,
+      "grad_norm": 0.22817212343215942,
+      "learning_rate": 8.662655500447388e-06,
+      "loss": 0.4099,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9447603574329814,
+      "grad_norm": 0.23591649532318115,
+      "learning_rate": 8.65943600738183e-06,
+      "loss": 0.3954,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9455727051177905,
+      "grad_norm": 0.2221234142780304,
+      "learning_rate": 8.656213243522307e-06,
+      "loss": 0.4047,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9463850528025995,
+      "grad_norm": 0.23155957460403442,
+      "learning_rate": 8.65298721174932e-06,
+      "loss": 0.3911,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9471974004874086,
+      "grad_norm": 0.2185499370098114,
+      "learning_rate": 8.649757914946284e-06,
+      "loss": 0.3658,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9480097481722177,
+      "grad_norm": 0.22856676578521729,
+      "learning_rate": 8.646525355999538e-06,
+      "loss": 0.3988,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9488220958570268,
+      "grad_norm": 0.24251709878444672,
+      "learning_rate": 8.643289537798332e-06,
+      "loss": 0.3951,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9496344435418359,
+      "grad_norm": 0.22046338021755219,
+      "learning_rate": 8.640050463234837e-06,
+      "loss": 0.3784,
+      "step": 1169
+    },
+    {
+      "epoch": 0.950446791226645,
+      "grad_norm": 0.24068596959114075,
+      "learning_rate": 8.636808135204123e-06,
+      "loss": 0.4363,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9512591389114541,
+      "grad_norm": 0.239118292927742,
+      "learning_rate": 8.633562556604177e-06,
+      "loss": 0.4129,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9520714865962632,
+      "grad_norm": 0.2545301914215088,
+      "learning_rate": 8.630313730335888e-06,
+      "loss": 0.3887,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9528838342810723,
+      "grad_norm": 0.23855891823768616,
+      "learning_rate": 8.627061659303047e-06,
+      "loss": 0.372,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9536961819658814,
+      "grad_norm": 0.21994520723819733,
+      "learning_rate": 8.623806346412349e-06,
+      "loss": 0.3968,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9545085296506904,
+      "grad_norm": 0.25787654519081116,
+      "learning_rate": 8.62054779457338e-06,
+      "loss": 0.3917,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9553208773354996,
+      "grad_norm": 0.23166845738887787,
+      "learning_rate": 8.617286006698628e-06,
+      "loss": 0.3719,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9561332250203087,
+      "grad_norm": 0.2616572380065918,
+      "learning_rate": 8.614020985703468e-06,
+      "loss": 0.4005,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9569455727051178,
+      "grad_norm": 0.2372180074453354,
+      "learning_rate": 8.610752734506166e-06,
+      "loss": 0.4078,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9577579203899269,
+      "grad_norm": 0.22400681674480438,
+      "learning_rate": 8.60748125602788e-06,
+      "loss": 0.3866,
+      "step": 1179
+    },
+    {
+      "epoch": 0.958570268074736,
+      "grad_norm": 0.2184598296880722,
+      "learning_rate": 8.604206553192647e-06,
+      "loss": 0.3875,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9593826157595451,
+      "grad_norm": 0.22089387476444244,
+      "learning_rate": 8.600928628927385e-06,
+      "loss": 0.392,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9601949634443542,
+      "grad_norm": 0.2410588562488556,
+      "learning_rate": 8.597647486161898e-06,
+      "loss": 0.4087,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9610073111291633,
+      "grad_norm": 0.22465266287326813,
+      "learning_rate": 8.594363127828858e-06,
+      "loss": 0.3971,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9618196588139724,
+      "grad_norm": 0.25125211477279663,
+      "learning_rate": 8.591075556863818e-06,
+      "loss": 0.4188,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9626320064987814,
+      "grad_norm": 0.21878761053085327,
+      "learning_rate": 8.5877847762052e-06,
+      "loss": 0.4198,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9634443541835905,
+      "grad_norm": 0.2255849987268448,
+      "learning_rate": 8.584490788794296e-06,
+      "loss": 0.3807,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9642567018683996,
+      "grad_norm": 0.21891926229000092,
+      "learning_rate": 8.581193597575259e-06,
+      "loss": 0.4008,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9650690495532088,
+      "grad_norm": 0.24072152376174927,
+      "learning_rate": 8.577893205495113e-06,
+      "loss": 0.4131,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9658813972380179,
+      "grad_norm": 0.24198724329471588,
+      "learning_rate": 8.574589615503739e-06,
+      "loss": 0.392,
+      "step": 1189
+    },
+    {
+      "epoch": 0.966693744922827,
+      "grad_norm": 0.23036138713359833,
+      "learning_rate": 8.571282830553875e-06,
+      "loss": 0.3867,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9675060926076361,
+      "grad_norm": 0.21972692012786865,
+      "learning_rate": 8.567972853601119e-06,
+      "loss": 0.3883,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9683184402924452,
+      "grad_norm": 0.22416172921657562,
+      "learning_rate": 8.564659687603917e-06,
+      "loss": 0.3755,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9691307879772543,
+      "grad_norm": 0.25966474413871765,
+      "learning_rate": 8.56134333552357e-06,
+      "loss": 0.3947,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9699431356620634,
+      "grad_norm": 0.22933460772037506,
+      "learning_rate": 8.558023800324223e-06,
+      "loss": 0.3897,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9707554833468724,
+      "grad_norm": 0.2445192188024521,
+      "learning_rate": 8.554701084972869e-06,
+      "loss": 0.4179,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9715678310316815,
+      "grad_norm": 0.21129579842090607,
+      "learning_rate": 8.551375192439339e-06,
+      "loss": 0.4061,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9723801787164906,
+      "grad_norm": 0.24654905498027802,
+      "learning_rate": 8.54804612569631e-06,
+      "loss": 0.4042,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9731925264012997,
+      "grad_norm": 0.2261275351047516,
+      "learning_rate": 8.54471388771929e-06,
+      "loss": 0.3997,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9740048740861088,
+      "grad_norm": 0.20965610444545746,
+      "learning_rate": 8.541378481486627e-06,
+      "loss": 0.3819,
+      "step": 1199
+    },
+    {
+      "epoch": 0.974817221770918,
+      "grad_norm": 0.22704555094242096,
+      "learning_rate": 8.538039909979496e-06,
+      "loss": 0.385,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9756295694557271,
+      "grad_norm": 0.22059553861618042,
+      "learning_rate": 8.534698176181904e-06,
+      "loss": 0.3629,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9764419171405362,
+      "grad_norm": 0.21852490305900574,
+      "learning_rate": 8.531353283080683e-06,
+      "loss": 0.4057,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9772542648253453,
+      "grad_norm": 0.22532199323177338,
+      "learning_rate": 8.528005233665491e-06,
+      "loss": 0.3962,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9780666125101544,
+      "grad_norm": 0.22272399067878723,
+      "learning_rate": 8.524654030928802e-06,
+      "loss": 0.3774,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9788789601949635,
+      "grad_norm": 0.22655266523361206,
+      "learning_rate": 8.521299677865915e-06,
+      "loss": 0.3975,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9796913078797725,
+      "grad_norm": 0.21311216056346893,
+      "learning_rate": 8.517942177474943e-06,
+      "loss": 0.3898,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9805036555645816,
+      "grad_norm": 0.23059318959712982,
+      "learning_rate": 8.514581532756806e-06,
+      "loss": 0.4008,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9813160032493907,
+      "grad_norm": 0.21966615319252014,
+      "learning_rate": 8.511217746715243e-06,
+      "loss": 0.4046,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9821283509341998,
+      "grad_norm": 0.2153371125459671,
+      "learning_rate": 8.507850822356796e-06,
+      "loss": 0.3848,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9829406986190089,
+      "grad_norm": 0.21917933225631714,
+      "learning_rate": 8.50448076269081e-06,
+      "loss": 0.3916,
+      "step": 1210
+    },
+    {
+      "epoch": 0.983753046303818,
+      "grad_norm": 0.2173098921775818,
+      "learning_rate": 8.50110757072944e-06,
+      "loss": 0.3774,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9845653939886271,
+      "grad_norm": 0.2481478452682495,
+      "learning_rate": 8.497731249487632e-06,
+      "loss": 0.4028,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9853777416734363,
+      "grad_norm": 0.22047555446624756,
+      "learning_rate": 8.494351801983135e-06,
+      "loss": 0.372,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9861900893582454,
+      "grad_norm": 0.22642996907234192,
+      "learning_rate": 8.490969231236487e-06,
+      "loss": 0.3831,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9870024370430545,
+      "grad_norm": 0.24283529818058014,
+      "learning_rate": 8.487583540271024e-06,
+      "loss": 0.3808,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9878147847278635,
+      "grad_norm": 0.2092595398426056,
+      "learning_rate": 8.484194732112866e-06,
+      "loss": 0.3797,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9886271324126726,
+      "grad_norm": 0.2043076753616333,
+      "learning_rate": 8.480802809790922e-06,
+      "loss": 0.3793,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9894394800974817,
+      "grad_norm": 0.2730744779109955,
+      "learning_rate": 8.47740777633688e-06,
+      "loss": 0.402,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9902518277822908,
+      "grad_norm": 0.24871286749839783,
+      "learning_rate": 8.474009634785214e-06,
+      "loss": 0.4107,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9910641754670999,
+      "grad_norm": 0.2324652373790741,
+      "learning_rate": 8.47060838817317e-06,
+      "loss": 0.3851,
+      "step": 1220
+    },
+    {
+      "epoch": 0.991876523151909,
+      "grad_norm": 0.2355189323425293,
+      "learning_rate": 8.467204039540779e-06,
+      "loss": 0.3997,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9926888708367181,
+      "grad_norm": 0.22385317087173462,
+      "learning_rate": 8.463796591930833e-06,
+      "loss": 0.3887,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9935012185215272,
+      "grad_norm": 0.22244158387184143,
+      "learning_rate": 8.460386048388905e-06,
+      "loss": 0.3973,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9943135662063363,
+      "grad_norm": 0.23022933304309845,
+      "learning_rate": 8.456972411963323e-06,
+      "loss": 0.4122,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9951259138911455,
+      "grad_norm": 0.24528944492340088,
+      "learning_rate": 8.453555685705192e-06,
+      "loss": 0.3974,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9959382615759546,
+      "grad_norm": 0.2035081684589386,
+      "learning_rate": 8.450135872668369e-06,
+      "loss": 0.3788,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9967506092607636,
+      "grad_norm": 0.21834662556648254,
+      "learning_rate": 8.446712975909474e-06,
+      "loss": 0.3682,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9975629569455727,
+      "grad_norm": 0.24424555897712708,
+      "learning_rate": 8.443286998487884e-06,
+      "loss": 0.3893,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9983753046303818,
+      "grad_norm": 0.22875617444515228,
+      "learning_rate": 8.439857943465728e-06,
+      "loss": 0.4095,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9991876523151909,
+      "grad_norm": 0.20575514435768127,
+      "learning_rate": 8.436425813907885e-06,
+      "loss": 0.3807,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2407267689704895,
+      "learning_rate": 8.432990612881985e-06,
+      "loss": 0.4027,
+      "step": 1231
+    },
+    {
+      "epoch": 1.0008123476848092,
+      "grad_norm": 0.26612457633018494,
+      "learning_rate": 8.4295523434584e-06,
+      "loss": 0.3767,
+      "step": 1232
+    },
+    {
+      "epoch": 1.0016246953696182,
+      "grad_norm": 0.21691322326660156,
+      "learning_rate": 8.426111008710245e-06,
+      "loss": 0.3673,
+      "step": 1233
+    },
+    {
+      "epoch": 1.0024370430544274,
+      "grad_norm": 0.25480735301971436,
+      "learning_rate": 8.422666611713378e-06,
+      "loss": 0.3701,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0032493907392364,
+      "grad_norm": 0.24979840219020844,
+      "learning_rate": 8.41921915554639e-06,
+      "loss": 0.3988,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0040617384240456,
+      "grad_norm": 0.23771551251411438,
+      "learning_rate": 8.415768643290608e-06,
+      "loss": 0.3794,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0048740861088545,
+      "grad_norm": 0.252147376537323,
+      "learning_rate": 8.412315078030091e-06,
+      "loss": 0.3829,
+      "step": 1237
+    },
+    {
+      "epoch": 1.0056864337936637,
+      "grad_norm": 0.21350406110286713,
+      "learning_rate": 8.408858462851625e-06,
+      "loss": 0.3628,
+      "step": 1238
+    },
+    {
+      "epoch": 1.0064987814784727,
+      "grad_norm": 0.2362724244594574,
+      "learning_rate": 8.405398800844726e-06,
+      "loss": 0.3697,
+      "step": 1239
+    },
+    {
+      "epoch": 1.007311129163282,
+      "grad_norm": 0.2549120783805847,
+      "learning_rate": 8.401936095101629e-06,
+      "loss": 0.3565,
+      "step": 1240
+    },
+    {
+      "epoch": 1.008123476848091,
+      "grad_norm": 0.22161637246608734,
+      "learning_rate": 8.398470348717288e-06,
+      "loss": 0.3591,
+      "step": 1241
+    },
+    {
+      "epoch": 1.0089358245329,
+      "grad_norm": 0.271334707736969,
+      "learning_rate": 8.395001564789385e-06,
+      "loss": 0.399,
+      "step": 1242
+    },
+    {
+      "epoch": 1.009748172217709,
+      "grad_norm": 0.21318240463733673,
+      "learning_rate": 8.391529746418303e-06,
+      "loss": 0.3807,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0105605199025183,
+      "grad_norm": 0.242513969540596,
+      "learning_rate": 8.388054896707146e-06,
+      "loss": 0.3582,
+      "step": 1244
+    },
+    {
+      "epoch": 1.0113728675873275,
+      "grad_norm": 0.22750258445739746,
+      "learning_rate": 8.384577018761725e-06,
+      "loss": 0.366,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0121852152721365,
+      "grad_norm": 0.21878233551979065,
+      "learning_rate": 8.38109611569056e-06,
+      "loss": 0.3661,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0129975629569457,
+      "grad_norm": 0.2563244700431824,
+      "learning_rate": 8.377612190604868e-06,
+      "loss": 0.3876,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0138099106417546,
+      "grad_norm": 0.22054894268512726,
+      "learning_rate": 8.374125246618576e-06,
+      "loss": 0.3681,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0146222583265638,
+      "grad_norm": 0.21887746453285217,
+      "learning_rate": 8.370635286848304e-06,
+      "loss": 0.372,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0154346060113728,
+      "grad_norm": 0.21507126092910767,
+      "learning_rate": 8.367142314413368e-06,
+      "loss": 0.3642,
+      "step": 1250
+    },
+    {
+      "epoch": 1.016246953696182,
+      "grad_norm": 0.227680504322052,
+      "learning_rate": 8.363646332435778e-06,
+      "loss": 0.3844,
+      "step": 1251
+    },
+    {
+      "epoch": 1.017059301380991,
+      "grad_norm": 0.22613193094730377,
+      "learning_rate": 8.360147344040233e-06,
+      "loss": 0.3878,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0178716490658002,
+      "grad_norm": 0.21775279939174652,
+      "learning_rate": 8.356645352354117e-06,
+      "loss": 0.3646,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0186839967506092,
+      "grad_norm": 0.24164386093616486,
+      "learning_rate": 8.353140360507506e-06,
+      "loss": 0.3847,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0194963444354184,
+      "grad_norm": 0.2225622534751892,
+      "learning_rate": 8.349632371633146e-06,
+      "loss": 0.3882,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0203086921202273,
+      "grad_norm": 0.22485634684562683,
+      "learning_rate": 8.346121388866472e-06,
+      "loss": 0.3847,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0211210398050365,
+      "grad_norm": 0.25174346566200256,
+      "learning_rate": 8.34260741534559e-06,
+      "loss": 0.3564,
+      "step": 1257
+    },
+    {
+      "epoch": 1.0219333874898457,
+      "grad_norm": 0.2345314472913742,
+      "learning_rate": 8.339090454211278e-06,
+      "loss": 0.3582,
+      "step": 1258
+    },
+    {
+      "epoch": 1.0227457351746547,
+      "grad_norm": 0.24508541822433472,
+      "learning_rate": 8.33557050860699e-06,
+      "loss": 0.3666,
+      "step": 1259
+    },
+    {
+      "epoch": 1.023558082859464,
+      "grad_norm": 0.23955978453159332,
+      "learning_rate": 8.332047581678839e-06,
+      "loss": 0.3594,
+      "step": 1260
+    },
+    {
+      "epoch": 1.024370430544273,
+      "grad_norm": 0.24590964615345,
+      "learning_rate": 8.328521676575612e-06,
+      "loss": 0.361,
+      "step": 1261
+    },
+    {
+      "epoch": 1.025182778229082,
+      "grad_norm": 0.235018789768219,
+      "learning_rate": 8.324992796448752e-06,
+      "loss": 0.3817,
+      "step": 1262
+    },
+    {
+      "epoch": 1.025995125913891,
+      "grad_norm": 0.23992042243480682,
+      "learning_rate": 8.321460944452364e-06,
+      "loss": 0.3548,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0268074735987003,
+      "grad_norm": 0.23473818600177765,
+      "learning_rate": 8.317926123743202e-06,
+      "loss": 0.3743,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0276198212835093,
+      "grad_norm": 0.2195303589105606,
+      "learning_rate": 8.314388337480686e-06,
+      "loss": 0.3735,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0284321689683185,
+      "grad_norm": 0.24628973007202148,
+      "learning_rate": 8.310847588826876e-06,
+      "loss": 0.379,
+      "step": 1266
+    },
+    {
+      "epoch": 1.0292445166531274,
+      "grad_norm": 0.23521539568901062,
+      "learning_rate": 8.307303880946485e-06,
+      "loss": 0.3776,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0300568643379366,
+      "grad_norm": 0.20346832275390625,
+      "learning_rate": 8.303757217006869e-06,
+      "loss": 0.3593,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0308692120227458,
+      "grad_norm": 0.24549604952335358,
+      "learning_rate": 8.300207600178024e-06,
+      "loss": 0.3682,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0316815597075548,
+      "grad_norm": 0.23384250700473785,
+      "learning_rate": 8.296655033632592e-06,
+      "loss": 0.369,
+      "step": 1270
+    },
+    {
+      "epoch": 1.032493907392364,
+      "grad_norm": 0.24577541649341583,
+      "learning_rate": 8.293099520545845e-06,
+      "loss": 0.3717,
+      "step": 1271
+    },
+    {
+      "epoch": 1.033306255077173,
+      "grad_norm": 0.24165937304496765,
+      "learning_rate": 8.289541064095693e-06,
+      "loss": 0.3626,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0341186027619822,
+      "grad_norm": 0.24167566001415253,
+      "learning_rate": 8.285979667462671e-06,
+      "loss": 0.3707,
+      "step": 1273
+    },
+    {
+      "epoch": 1.0349309504467912,
+      "grad_norm": 0.2546290457248688,
+      "learning_rate": 8.28241533382995e-06,
+      "loss": 0.3744,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0357432981316004,
+      "grad_norm": 0.24485579133033752,
+      "learning_rate": 8.27884806638332e-06,
+      "loss": 0.355,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0365556458164094,
+      "grad_norm": 0.20959609746932983,
+      "learning_rate": 8.275277868311193e-06,
+      "loss": 0.3687,
+      "step": 1276
+    },
+    {
+      "epoch": 1.0373679935012186,
+      "grad_norm": 0.24303072690963745,
+      "learning_rate": 8.271704742804605e-06,
+      "loss": 0.3941,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0381803411860275,
+      "grad_norm": 0.2421116977930069,
+      "learning_rate": 8.268128693057202e-06,
+      "loss": 0.3697,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0389926888708367,
+      "grad_norm": 0.23019076883792877,
+      "learning_rate": 8.264549722265254e-06,
+      "loss": 0.3452,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0398050365556457,
+      "grad_norm": 0.24382421374320984,
+      "learning_rate": 8.260967833627628e-06,
+      "loss": 0.3554,
+      "step": 1280
+    },
+    {
+      "epoch": 1.040617384240455,
+      "grad_norm": 0.2152639627456665,
+      "learning_rate": 8.257383030345814e-06,
+      "loss": 0.3785,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0414297319252641,
+      "grad_norm": 0.21227115392684937,
+      "learning_rate": 8.253795315623894e-06,
+      "loss": 0.3749,
+      "step": 1282
+    },
+    {
+      "epoch": 1.042242079610073,
+      "grad_norm": 0.22943924367427826,
+      "learning_rate": 8.250204692668559e-06,
+      "loss": 0.3657,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0430544272948823,
+      "grad_norm": 0.23793566226959229,
+      "learning_rate": 8.246611164689097e-06,
+      "loss": 0.3889,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0438667749796913,
+      "grad_norm": 0.21289688348770142,
+      "learning_rate": 8.243014734897397e-06,
+      "loss": 0.3809,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0446791226645005,
+      "grad_norm": 0.1968437284231186,
+      "learning_rate": 8.239415406507934e-06,
+      "loss": 0.3859,
+      "step": 1286
+    },
+    {
+      "epoch": 1.0454914703493094,
+      "grad_norm": 0.22225528955459595,
+      "learning_rate": 8.235813182737782e-06,
+      "loss": 0.3568,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0463038180341186,
+      "grad_norm": 0.22756069898605347,
+      "learning_rate": 8.232208066806596e-06,
+      "loss": 0.3613,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0471161657189276,
+      "grad_norm": 0.2268378585577011,
+      "learning_rate": 8.228600061936617e-06,
+      "loss": 0.3465,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0479285134037368,
+      "grad_norm": 0.2169281393289566,
+      "learning_rate": 8.224989171352675e-06,
+      "loss": 0.3776,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0487408610885458,
+      "grad_norm": 0.2514020800590515,
+      "learning_rate": 8.221375398282172e-06,
+      "loss": 0.3868,
+      "step": 1291
+    },
+    {
+      "epoch": 1.049553208773355,
+      "grad_norm": 0.24864666163921356,
+      "learning_rate": 8.217758745955087e-06,
+      "loss": 0.3859,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0503655564581642,
+      "grad_norm": 0.2141553908586502,
+      "learning_rate": 8.214139217603975e-06,
+      "loss": 0.3784,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0511779041429732,
+      "grad_norm": 0.26216641068458557,
+      "learning_rate": 8.210516816463961e-06,
+      "loss": 0.3726,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0519902518277824,
+      "grad_norm": 0.27425530552864075,
+      "learning_rate": 8.206891545772738e-06,
+      "loss": 0.369,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0528025995125914,
+      "grad_norm": 0.2198823243379593,
+      "learning_rate": 8.203263408770562e-06,
+      "loss": 0.367,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0536149471974006,
+      "grad_norm": 0.2437698394060135,
+      "learning_rate": 8.19963240870025e-06,
+      "loss": 0.3903,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0544272948822095,
+      "grad_norm": 0.2560095489025116,
+      "learning_rate": 8.195998548807185e-06,
+      "loss": 0.3756,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0552396425670187,
+      "grad_norm": 0.22770293056964874,
+      "learning_rate": 8.192361832339299e-06,
+      "loss": 0.3603,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0560519902518277,
+      "grad_norm": 0.21887357532978058,
+      "learning_rate": 8.188722262547078e-06,
+      "loss": 0.3619,
+      "step": 1300
+    },
+    {
+      "epoch": 1.056864337936637,
+      "grad_norm": 0.25635048747062683,
+      "learning_rate": 8.185079842683558e-06,
+      "loss": 0.3887,
+      "step": 1301
+    },
+    {
+      "epoch": 1.057676685621446,
+      "grad_norm": 0.23161499202251434,
+      "learning_rate": 8.181434576004328e-06,
+      "loss": 0.3689,
+      "step": 1302
+    },
+    {
+      "epoch": 1.058489033306255,
+      "grad_norm": 0.19766703248023987,
+      "learning_rate": 8.177786465767517e-06,
+      "loss": 0.3501,
+      "step": 1303
+    },
+    {
+      "epoch": 1.059301380991064,
+      "grad_norm": 0.21386130154132843,
+      "learning_rate": 8.174135515233792e-06,
+      "loss": 0.3661,
+      "step": 1304
+    },
+    {
+      "epoch": 1.0601137286758733,
+      "grad_norm": 0.22687497735023499,
+      "learning_rate": 8.170481727666366e-06,
+      "loss": 0.3594,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0609260763606825,
+      "grad_norm": 0.23310726881027222,
+      "learning_rate": 8.166825106330985e-06,
+      "loss": 0.3727,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0617384240454915,
+      "grad_norm": 0.22907818853855133,
+      "learning_rate": 8.163165654495925e-06,
+      "loss": 0.3707,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0625507717303007,
+      "grad_norm": 0.2283329963684082,
+      "learning_rate": 8.159503375431996e-06,
+      "loss": 0.3618,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0633631194151096,
+      "grad_norm": 0.23639951646327972,
+      "learning_rate": 8.155838272412531e-06,
+      "loss": 0.3573,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0641754670999188,
+      "grad_norm": 0.2399565577507019,
+      "learning_rate": 8.152170348713392e-06,
+      "loss": 0.374,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0649878147847278,
+      "grad_norm": 0.23268310725688934,
+      "learning_rate": 8.148499607612956e-06,
+      "loss": 0.354,
+      "step": 1311
+    },
+    {
+      "epoch": 1.065800162469537,
+      "grad_norm": 0.21993541717529297,
+      "learning_rate": 8.144826052392123e-06,
+      "loss": 0.3738,
+      "step": 1312
+    },
+    {
+      "epoch": 1.066612510154346,
+      "grad_norm": 0.23509740829467773,
+      "learning_rate": 8.141149686334307e-06,
+      "loss": 0.366,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0674248578391552,
+      "grad_norm": 0.2555445730686188,
+      "learning_rate": 8.137470512725434e-06,
+      "loss": 0.3636,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0682372055239642,
+      "grad_norm": 0.25647762417793274,
+      "learning_rate": 8.133788534853939e-06,
+      "loss": 0.3703,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0690495532087734,
+      "grad_norm": 0.2428644895553589,
+      "learning_rate": 8.130103756010763e-06,
+      "loss": 0.3688,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0698619008935824,
+      "grad_norm": 0.23498383164405823,
+      "learning_rate": 8.126416179489354e-06,
+      "loss": 0.3713,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0706742485783916,
+      "grad_norm": 0.259798526763916,
+      "learning_rate": 8.122725808585654e-06,
+      "loss": 0.3788,
+      "step": 1318
+    },
+    {
+      "epoch": 1.0714865962632008,
+      "grad_norm": 0.24521780014038086,
+      "learning_rate": 8.11903264659811e-06,
+      "loss": 0.3878,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0722989439480097,
+      "grad_norm": 0.26938098669052124,
+      "learning_rate": 8.115336696827658e-06,
+      "loss": 0.3832,
+      "step": 1320
+    },
+    {
+      "epoch": 1.073111291632819,
+      "grad_norm": 0.2828563153743744,
+      "learning_rate": 8.111637962577728e-06,
+      "loss": 0.3679,
+      "step": 1321
+    },
+    {
+      "epoch": 1.073923639317628,
+      "grad_norm": 0.2225974053144455,
+      "learning_rate": 8.10793644715424e-06,
+      "loss": 0.3824,
+      "step": 1322
+    },
+    {
+      "epoch": 1.074735987002437,
+      "grad_norm": 0.2237105518579483,
+      "learning_rate": 8.104232153865597e-06,
+      "loss": 0.3628,
+      "step": 1323
+    },
+    {
+      "epoch": 1.075548334687246,
+      "grad_norm": 0.26092565059661865,
+      "learning_rate": 8.100525086022687e-06,
+      "loss": 0.3898,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0763606823720553,
+      "grad_norm": 0.2504673898220062,
+      "learning_rate": 8.096815246938879e-06,
+      "loss": 0.3767,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0771730300568643,
+      "grad_norm": 0.21380344033241272,
+      "learning_rate": 8.093102639930013e-06,
+      "loss": 0.3877,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0779853777416735,
+      "grad_norm": 0.23998433351516724,
+      "learning_rate": 8.089387268314408e-06,
+      "loss": 0.3887,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0787977254264824,
+      "grad_norm": 0.21309086680412292,
+      "learning_rate": 8.085669135412856e-06,
+      "loss": 0.3506,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0796100731112916,
+      "grad_norm": 0.22502297163009644,
+      "learning_rate": 8.081948244548615e-06,
+      "loss": 0.36,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0804224207961006,
+      "grad_norm": 0.23831559717655182,
+      "learning_rate": 8.0782245990474e-06,
+      "loss": 0.361,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0812347684809098,
+      "grad_norm": 0.21888858079910278,
+      "learning_rate": 8.074498202237402e-06,
+      "loss": 0.3823,
+      "step": 1331
+    },
+    {
+      "epoch": 1.082047116165719,
+      "grad_norm": 0.21353641152381897,
+      "learning_rate": 8.070769057449262e-06,
+      "loss": 0.3582,
+      "step": 1332
+    },
+    {
+      "epoch": 1.082859463850528,
+      "grad_norm": 0.4268950819969177,
+      "learning_rate": 8.067037168016079e-06,
+      "loss": 0.3621,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0836718115353372,
+      "grad_norm": 0.23871247470378876,
+      "learning_rate": 8.063302537273407e-06,
+      "loss": 0.3655,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0844841592201462,
+      "grad_norm": 0.21593041718006134,
+      "learning_rate": 8.059565168559247e-06,
+      "loss": 0.3999,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0852965069049554,
+      "grad_norm": 0.2377825379371643,
+      "learning_rate": 8.05582506521405e-06,
+      "loss": 0.3624,
+      "step": 1336
+    },
+    {
+      "epoch": 1.0861088545897644,
+      "grad_norm": 0.2596246004104614,
+      "learning_rate": 8.052082230580711e-06,
+      "loss": 0.3579,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0869212022745736,
+      "grad_norm": 0.2309873402118683,
+      "learning_rate": 8.048336668004564e-06,
+      "loss": 0.3776,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0877335499593825,
+      "grad_norm": 0.261382132768631,
+      "learning_rate": 8.044588380833384e-06,
+      "loss": 0.3723,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0885458976441917,
+      "grad_norm": 0.22110995650291443,
+      "learning_rate": 8.040837372417379e-06,
+      "loss": 0.3638,
+      "step": 1340
+    },
+    {
+      "epoch": 1.0893582453290007,
+      "grad_norm": 0.2678782045841217,
+      "learning_rate": 8.03708364610919e-06,
+      "loss": 0.3795,
+      "step": 1341
+    },
+    {
+      "epoch": 1.09017059301381,
+      "grad_norm": 0.26484349370002747,
+      "learning_rate": 8.033327205263888e-06,
+      "loss": 0.362,
+      "step": 1342
+    },
+    {
+      "epoch": 1.090982940698619,
+      "grad_norm": 0.27659714221954346,
+      "learning_rate": 8.02956805323897e-06,
+      "loss": 0.3631,
+      "step": 1343
+    },
+    {
+      "epoch": 1.091795288383428,
+      "grad_norm": 0.24894599616527557,
+      "learning_rate": 8.025806193394354e-06,
+      "loss": 0.3812,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0926076360682373,
+      "grad_norm": 0.2507006824016571,
+      "learning_rate": 8.022041629092382e-06,
+      "loss": 0.3536,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0934199837530463,
+      "grad_norm": 0.2553721070289612,
+      "learning_rate": 8.01827436369781e-06,
+      "loss": 0.3703,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0942323314378555,
+      "grad_norm": 0.25392505526542664,
+      "learning_rate": 8.01450440057781e-06,
+      "loss": 0.378,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0950446791226645,
+      "grad_norm": 0.2316385805606842,
+      "learning_rate": 8.010731743101967e-06,
+      "loss": 0.3626,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0958570268074737,
+      "grad_norm": 0.2432144433259964,
+      "learning_rate": 8.006956394642268e-06,
+      "loss": 0.3667,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0966693744922826,
+      "grad_norm": 0.2178601175546646,
+      "learning_rate": 8.003178358573112e-06,
+      "loss": 0.3949,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0974817221770918,
+      "grad_norm": 0.21221160888671875,
+      "learning_rate": 7.999397638271298e-06,
+      "loss": 0.3718,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0982940698619008,
+      "grad_norm": 0.22046715021133423,
+      "learning_rate": 7.99561423711602e-06,
+      "loss": 0.3789,
+      "step": 1352
+    },
+    {
+      "epoch": 1.09910641754671,
+      "grad_norm": 0.22993077337741852,
+      "learning_rate": 7.991828158488875e-06,
+      "loss": 0.3615,
+      "step": 1353
+    },
+    {
+      "epoch": 1.099918765231519,
+      "grad_norm": 0.23864513635635376,
+      "learning_rate": 7.98803940577385e-06,
+      "loss": 0.4107,
+      "step": 1354
+    },
+    {
+      "epoch": 1.1007311129163282,
+      "grad_norm": 0.2199014574289322,
+      "learning_rate": 7.984247982357317e-06,
+      "loss": 0.3717,
+      "step": 1355
+    },
+    {
+      "epoch": 1.1015434606011372,
+      "grad_norm": 0.24915708601474762,
+      "learning_rate": 7.980453891628047e-06,
+      "loss": 0.3733,
+      "step": 1356
+    },
+    {
+      "epoch": 1.1023558082859464,
+      "grad_norm": 0.22118164598941803,
+      "learning_rate": 7.976657136977181e-06,
+      "loss": 0.3585,
+      "step": 1357
+    },
+    {
+      "epoch": 1.1031681559707556,
+      "grad_norm": 0.2285040020942688,
+      "learning_rate": 7.972857721798254e-06,
+      "loss": 0.3795,
+      "step": 1358
+    },
+    {
+      "epoch": 1.1039805036555645,
+      "grad_norm": 0.2370767742395401,
+      "learning_rate": 7.96905564948717e-06,
+      "loss": 0.3921,
+      "step": 1359
+    },
+    {
+      "epoch": 1.1047928513403737,
+      "grad_norm": 0.24940864741802216,
+      "learning_rate": 7.96525092344221e-06,
+      "loss": 0.3894,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1056051990251827,
+      "grad_norm": 0.2110314816236496,
+      "learning_rate": 7.96144354706403e-06,
+      "loss": 0.3484,
+      "step": 1361
+    },
+    {
+      "epoch": 1.106417546709992,
+      "grad_norm": 0.22861716151237488,
+      "learning_rate": 7.957633523755652e-06,
+      "loss": 0.3769,
+      "step": 1362
+    },
+    {
+      "epoch": 1.107229894394801,
+      "grad_norm": 0.24666036665439606,
+      "learning_rate": 7.953820856922465e-06,
+      "loss": 0.3619,
+      "step": 1363
+    },
+    {
+      "epoch": 1.10804224207961,
+      "grad_norm": 0.20891974866390228,
+      "learning_rate": 7.95000554997222e-06,
+      "loss": 0.3786,
+      "step": 1364
+    },
+    {
+      "epoch": 1.108854589764419,
+      "grad_norm": 0.2506312131881714,
+      "learning_rate": 7.946187606315026e-06,
+      "loss": 0.383,
+      "step": 1365
+    },
+    {
+      "epoch": 1.1096669374492283,
+      "grad_norm": 0.23187628388404846,
+      "learning_rate": 7.942367029363351e-06,
+      "loss": 0.3738,
+      "step": 1366
+    },
+    {
+      "epoch": 1.1104792851340373,
+      "grad_norm": 0.24712975323200226,
+      "learning_rate": 7.938543822532019e-06,
+      "loss": 0.3753,
+      "step": 1367
+    },
+    {
+      "epoch": 1.1112916328188465,
+      "grad_norm": 0.24223390221595764,
+      "learning_rate": 7.9347179892382e-06,
+      "loss": 0.3652,
+      "step": 1368
+    },
+    {
+      "epoch": 1.1121039805036557,
+      "grad_norm": 0.23025573790073395,
+      "learning_rate": 7.930889532901416e-06,
+      "loss": 0.3725,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1129163281884646,
+      "grad_norm": 0.22451052069664001,
+      "learning_rate": 7.927058456943527e-06,
+      "loss": 0.3773,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1137286758732738,
+      "grad_norm": 0.2359568029642105,
+      "learning_rate": 7.92322476478874e-06,
+      "loss": 0.3691,
+      "step": 1371
+    },
+    {
+      "epoch": 1.1145410235580828,
+      "grad_norm": 0.24245385825634003,
+      "learning_rate": 7.919388459863597e-06,
+      "loss": 0.3683,
+      "step": 1372
+    },
+    {
+      "epoch": 1.115353371242892,
+      "grad_norm": 0.22840920090675354,
+      "learning_rate": 7.915549545596979e-06,
+      "loss": 0.3691,
+      "step": 1373
+    },
+    {
+      "epoch": 1.116165718927701,
+      "grad_norm": 0.2256499081850052,
+      "learning_rate": 7.911708025420097e-06,
+      "loss": 0.3791,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1169780666125102,
+      "grad_norm": 0.24837297201156616,
+      "learning_rate": 7.907863902766494e-06,
+      "loss": 0.3413,
+      "step": 1375
+    },
+    {
+      "epoch": 1.1177904142973192,
+      "grad_norm": 0.2394474893808365,
+      "learning_rate": 7.90401718107203e-06,
+      "loss": 0.3863,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1186027619821284,
+      "grad_norm": 0.25105607509613037,
+      "learning_rate": 7.900167863774896e-06,
+      "loss": 0.3542,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1194151096669374,
+      "grad_norm": 0.25730058550834656,
+      "learning_rate": 7.896315954315608e-06,
+      "loss": 0.3861,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1202274573517466,
+      "grad_norm": 0.2151753306388855,
+      "learning_rate": 7.892461456136984e-06,
+      "loss": 0.3658,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1210398050365558,
+      "grad_norm": 0.25783830881118774,
+      "learning_rate": 7.888604372684168e-06,
+      "loss": 0.3897,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1218521527213647,
+      "grad_norm": 0.22147522866725922,
+      "learning_rate": 7.88474470740461e-06,
+      "loss": 0.3722,
+      "step": 1381
+    },
+    {
+      "epoch": 1.122664500406174,
+      "grad_norm": 0.21444956958293915,
+      "learning_rate": 7.880882463748067e-06,
+      "loss": 0.3619,
+      "step": 1382
+    },
+    {
+      "epoch": 1.123476848090983,
+      "grad_norm": 0.23158355057239532,
+      "learning_rate": 7.877017645166606e-06,
+      "loss": 0.372,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1242891957757921,
+      "grad_norm": 0.24209995567798615,
+      "learning_rate": 7.873150255114588e-06,
+      "loss": 0.3773,
+      "step": 1384
+    },
+    {
+      "epoch": 1.125101543460601,
+      "grad_norm": 0.21470318734645844,
+      "learning_rate": 7.869280297048674e-06,
+      "loss": 0.3795,
+      "step": 1385
+    },
+    {
+      "epoch": 1.1259138911454103,
+      "grad_norm": 0.22509267926216125,
+      "learning_rate": 7.865407774427828e-06,
+      "loss": 0.3537,
+      "step": 1386
+    },
+    {
+      "epoch": 1.1267262388302193,
+      "grad_norm": 0.2624915540218353,
+      "learning_rate": 7.861532690713293e-06,
+      "loss": 0.3618,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1275385865150285,
+      "grad_norm": 0.21535582840442657,
+      "learning_rate": 7.857655049368614e-06,
+      "loss": 0.3638,
+      "step": 1388
+    },
+    {
+      "epoch": 1.1283509341998375,
+      "grad_norm": 0.24028939008712769,
+      "learning_rate": 7.853774853859612e-06,
+      "loss": 0.3652,
+      "step": 1389
+    },
+    {
+      "epoch": 1.1291632818846467,
+      "grad_norm": 0.2211923748254776,
+      "learning_rate": 7.8498921076544e-06,
+      "loss": 0.3711,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1299756295694556,
+      "grad_norm": 0.22399570047855377,
+      "learning_rate": 7.846006814223362e-06,
+      "loss": 0.3658,
+      "step": 1391
+    },
+    {
+      "epoch": 1.1307879772542648,
+      "grad_norm": 0.21122337877750397,
+      "learning_rate": 7.842118977039162e-06,
+      "loss": 0.3762,
+      "step": 1392
+    },
+    {
+      "epoch": 1.131600324939074,
+      "grad_norm": 0.22835077345371246,
+      "learning_rate": 7.838228599576743e-06,
+      "loss": 0.3556,
+      "step": 1393
+    },
+    {
+      "epoch": 1.132412672623883,
+      "grad_norm": 0.2244669646024704,
+      "learning_rate": 7.83433568531331e-06,
+      "loss": 0.3816,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1332250203086922,
+      "grad_norm": 0.21755845844745636,
+      "learning_rate": 7.830440237728341e-06,
+      "loss": 0.37,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1340373679935012,
+      "grad_norm": 0.23084872961044312,
+      "learning_rate": 7.826542260303576e-06,
+      "loss": 0.3805,
+      "step": 1396
+    },
+    {
+      "epoch": 1.1348497156783104,
+      "grad_norm": 0.21853198111057281,
+      "learning_rate": 7.822641756523015e-06,
+      "loss": 0.3851,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1356620633631194,
+      "grad_norm": 0.23380815982818604,
+      "learning_rate": 7.818738729872921e-06,
+      "loss": 0.3682,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1364744110479286,
+      "grad_norm": 0.21045514941215515,
+      "learning_rate": 7.814833183841805e-06,
+      "loss": 0.3743,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1372867587327375,
+      "grad_norm": 0.19610148668289185,
+      "learning_rate": 7.810925121920437e-06,
+      "loss": 0.3808,
+      "step": 1400
+    },
+    {
+      "epoch": 1.1380991064175467,
+      "grad_norm": 0.1989269107580185,
+      "learning_rate": 7.80701454760183e-06,
+      "loss": 0.3751,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1389114541023557,
+      "grad_norm": 0.23900388181209564,
+      "learning_rate": 7.803101464381244e-06,
+      "loss": 0.3976,
+      "step": 1402
+    },
+    {
+      "epoch": 1.139723801787165,
+      "grad_norm": 0.20955105125904083,
+      "learning_rate": 7.799185875756184e-06,
+      "loss": 0.3515,
+      "step": 1403
+    },
+    {
+      "epoch": 1.140536149471974,
+      "grad_norm": 0.2138160914182663,
+      "learning_rate": 7.795267785226392e-06,
+      "loss": 0.3805,
+      "step": 1404
+    },
+    {
+      "epoch": 1.141348497156783,
+      "grad_norm": 0.25349172949790955,
+      "learning_rate": 7.791347196293846e-06,
+      "loss": 0.3567,
+      "step": 1405
+    },
+    {
+      "epoch": 1.1421608448415923,
+      "grad_norm": 0.2161046713590622,
+      "learning_rate": 7.787424112462758e-06,
+      "loss": 0.3593,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1429731925264013,
+      "grad_norm": 0.2594912350177765,
+      "learning_rate": 7.783498537239568e-06,
+      "loss": 0.4038,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1437855402112105,
+      "grad_norm": 0.24835051596164703,
+      "learning_rate": 7.779570474132949e-06,
+      "loss": 0.3672,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1445978878960195,
+      "grad_norm": 0.19419212639331818,
+      "learning_rate": 7.775639926653789e-06,
+      "loss": 0.3616,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1454102355808287,
+      "grad_norm": 0.23529349267482758,
+      "learning_rate": 7.771706898315204e-06,
+      "loss": 0.3845,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1462225832656376,
+      "grad_norm": 0.22840334475040436,
+      "learning_rate": 7.767771392632518e-06,
+      "loss": 0.3575,
+      "step": 1411
+    },
+    {
+      "epoch": 1.1470349309504468,
+      "grad_norm": 0.22261232137680054,
+      "learning_rate": 7.763833413123281e-06,
+      "loss": 0.3808,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1478472786352558,
+      "grad_norm": 0.232461079955101,
+      "learning_rate": 7.759892963307249e-06,
+      "loss": 0.3891,
+      "step": 1413
+    },
+    {
+      "epoch": 1.148659626320065,
+      "grad_norm": 0.2366735190153122,
+      "learning_rate": 7.75595004670638e-06,
+      "loss": 0.3812,
+      "step": 1414
+    },
+    {
+      "epoch": 1.149471974004874,
+      "grad_norm": 0.21002799272537231,
+      "learning_rate": 7.752004666844849e-06,
+      "loss": 0.3604,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1502843216896832,
+      "grad_norm": 0.23796050250530243,
+      "learning_rate": 7.748056827249018e-06,
+      "loss": 0.3631,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1510966693744922,
+      "grad_norm": 0.2469610720872879,
+      "learning_rate": 7.744106531447462e-06,
+      "loss": 0.3571,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1519090170593014,
+      "grad_norm": 0.2165290266275406,
+      "learning_rate": 7.74015378297094e-06,
+      "loss": 0.3751,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1527213647441106,
+      "grad_norm": 0.2321062982082367,
+      "learning_rate": 7.736198585352411e-06,
+      "loss": 0.3909,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1535337124289196,
+      "grad_norm": 0.23364657163619995,
+      "learning_rate": 7.732240942127018e-06,
+      "loss": 0.3704,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1543460601137288,
+      "grad_norm": 0.2127617597579956,
+      "learning_rate": 7.728280856832094e-06,
+      "loss": 0.3668,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1551584077985377,
+      "grad_norm": 0.23370081186294556,
+      "learning_rate": 7.72431833300715e-06,
+      "loss": 0.3734,
+      "step": 1422
+    },
+    {
+      "epoch": 1.155970755483347,
+      "grad_norm": 0.22226890921592712,
+      "learning_rate": 7.72035337419388e-06,
+      "loss": 0.3769,
+      "step": 1423
+    },
+    {
+      "epoch": 1.156783103168156,
+      "grad_norm": 0.20126818120479584,
+      "learning_rate": 7.716385983936154e-06,
+      "loss": 0.3432,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1575954508529651,
+      "grad_norm": 0.24522702395915985,
+      "learning_rate": 7.712416165780014e-06,
+      "loss": 0.3817,
+      "step": 1425
+    },
+    {
+      "epoch": 1.158407798537774,
+      "grad_norm": 0.20912979543209076,
+      "learning_rate": 7.708443923273671e-06,
+      "loss": 0.3897,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1592201462225833,
+      "grad_norm": 0.2064291387796402,
+      "learning_rate": 7.704469259967509e-06,
+      "loss": 0.3812,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1600324939073923,
+      "grad_norm": 0.222136989235878,
+      "learning_rate": 7.700492179414067e-06,
+      "loss": 0.3678,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1608448415922015,
+      "grad_norm": 0.22645947337150574,
+      "learning_rate": 7.69651268516805e-06,
+      "loss": 0.3736,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1616571892770104,
+      "grad_norm": 0.22382651269435883,
+      "learning_rate": 7.692530780786321e-06,
+      "loss": 0.3581,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1624695369618196,
+      "grad_norm": 0.23129256069660187,
+      "learning_rate": 7.688546469827892e-06,
+      "loss": 0.3658,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1632818846466288,
+      "grad_norm": 0.20880858600139618,
+      "learning_rate": 7.684559755853932e-06,
+      "loss": 0.3539,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1640942323314378,
+      "grad_norm": 0.2030506581068039,
+      "learning_rate": 7.680570642427755e-06,
+      "loss": 0.3567,
+      "step": 1433
+    },
+    {
+      "epoch": 1.164906580016247,
+      "grad_norm": 0.2391713410615921,
+      "learning_rate": 7.676579133114819e-06,
+      "loss": 0.3906,
+      "step": 1434
+    },
+    {
+      "epoch": 1.165718927701056,
+      "grad_norm": 0.21054549515247345,
+      "learning_rate": 7.672585231482723e-06,
+      "loss": 0.3724,
+      "step": 1435
+    },
+    {
+      "epoch": 1.1665312753858652,
+      "grad_norm": 0.2624136805534363,
+      "learning_rate": 7.668588941101208e-06,
+      "loss": 0.3901,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1673436230706742,
+      "grad_norm": 0.21168768405914307,
+      "learning_rate": 7.664590265542144e-06,
+      "loss": 0.3745,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1681559707554834,
+      "grad_norm": 0.22687077522277832,
+      "learning_rate": 7.66058920837954e-06,
+      "loss": 0.3808,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1689683184402924,
+      "grad_norm": 0.2079876959323883,
+      "learning_rate": 7.65658577318953e-06,
+      "loss": 0.3643,
+      "step": 1439
+    },
+    {
+      "epoch": 1.1697806661251016,
+      "grad_norm": 0.2142096906900406,
+      "learning_rate": 7.65257996355037e-06,
+      "loss": 0.3536,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1705930138099108,
+      "grad_norm": 0.20655301213264465,
+      "learning_rate": 7.648571783042445e-06,
+      "loss": 0.3522,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1714053614947197,
+      "grad_norm": 0.20702239871025085,
+      "learning_rate": 7.644561235248254e-06,
+      "loss": 0.364,
+      "step": 1442
+    },
+    {
+      "epoch": 1.1722177091795287,
+      "grad_norm": 0.22796288132667542,
+      "learning_rate": 7.640548323752412e-06,
+      "loss": 0.3599,
+      "step": 1443
+    },
+    {
+      "epoch": 1.173030056864338,
+      "grad_norm": 0.22326219081878662,
+      "learning_rate": 7.636533052141653e-06,
+      "loss": 0.3792,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1738424045491471,
+      "grad_norm": 0.23828524351119995,
+      "learning_rate": 7.632515424004813e-06,
+      "loss": 0.3731,
+      "step": 1445
+    },
+    {
+      "epoch": 1.174654752233956,
+      "grad_norm": 0.19556228816509247,
+      "learning_rate": 7.628495442932838e-06,
+      "loss": 0.3633,
+      "step": 1446
+    },
+    {
+      "epoch": 1.1754670999187653,
+      "grad_norm": 0.231424018740654,
+      "learning_rate": 7.624473112518777e-06,
+      "loss": 0.3697,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1762794476035743,
+      "grad_norm": 0.27811962366104126,
+      "learning_rate": 7.6204484363577756e-06,
+      "loss": 0.3757,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1770917952883835,
+      "grad_norm": 0.2192411869764328,
+      "learning_rate": 7.61642141804708e-06,
+      "loss": 0.3802,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1779041429731925,
+      "grad_norm": 0.2456451952457428,
+      "learning_rate": 7.612392061186027e-06,
+      "loss": 0.3648,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1787164906580017,
+      "grad_norm": 0.22763483226299286,
+      "learning_rate": 7.608360369376047e-06,
+      "loss": 0.3707,
+      "step": 1451
+    },
+    {
+      "epoch": 1.1795288383428106,
+      "grad_norm": 0.2714596092700958,
+      "learning_rate": 7.604326346220654e-06,
+      "loss": 0.3667,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1803411860276198,
+      "grad_norm": 0.2504202723503113,
+      "learning_rate": 7.600289995325446e-06,
+      "loss": 0.3799,
+      "step": 1453
+    },
+    {
+      "epoch": 1.181153533712429,
+      "grad_norm": 0.22636272013187408,
+      "learning_rate": 7.596251320298105e-06,
+      "loss": 0.3519,
+      "step": 1454
+    },
+    {
+      "epoch": 1.181965881397238,
+      "grad_norm": 0.22676870226860046,
+      "learning_rate": 7.592210324748384e-06,
+      "loss": 0.3804,
+      "step": 1455
+    },
+    {
+      "epoch": 1.182778229082047,
+      "grad_norm": 0.2239973396062851,
+      "learning_rate": 7.588167012288117e-06,
+      "loss": 0.3435,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1835905767668562,
+      "grad_norm": 0.23214679956436157,
+      "learning_rate": 7.584121386531205e-06,
+      "loss": 0.3705,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1844029244516654,
+      "grad_norm": 0.2385970950126648,
+      "learning_rate": 7.580073451093617e-06,
+      "loss": 0.3734,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1852152721364744,
+      "grad_norm": 0.22154447436332703,
+      "learning_rate": 7.576023209593386e-06,
+      "loss": 0.3662,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1860276198212836,
+      "grad_norm": 0.2049044817686081,
+      "learning_rate": 7.571970665650607e-06,
+      "loss": 0.3828,
+      "step": 1460
+    },
+    {
+      "epoch": 1.1868399675060926,
+      "grad_norm": 0.24579772353172302,
+      "learning_rate": 7.567915822887434e-06,
+      "loss": 0.3681,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1876523151909018,
+      "grad_norm": 0.23100274801254272,
+      "learning_rate": 7.563858684928074e-06,
+      "loss": 0.3577,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1884646628757107,
+      "grad_norm": 0.22084347903728485,
+      "learning_rate": 7.559799255398785e-06,
+      "loss": 0.3693,
+      "step": 1463
+    },
+    {
+      "epoch": 1.18927701056052,
+      "grad_norm": 0.2381676733493805,
+      "learning_rate": 7.555737537927875e-06,
+      "loss": 0.3652,
+      "step": 1464
+    },
+    {
+      "epoch": 1.190089358245329,
+      "grad_norm": 0.239349365234375,
+      "learning_rate": 7.551673536145695e-06,
+      "loss": 0.373,
+      "step": 1465
+    },
+    {
+      "epoch": 1.190901705930138,
+      "grad_norm": 0.22234754264354706,
+      "learning_rate": 7.54760725368464e-06,
+      "loss": 0.3844,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1917140536149473,
+      "grad_norm": 0.2186950147151947,
+      "learning_rate": 7.543538694179142e-06,
+      "loss": 0.357,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1925264012997563,
+      "grad_norm": 0.2303467094898224,
+      "learning_rate": 7.539467861265668e-06,
+      "loss": 0.3709,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1933387489845655,
+      "grad_norm": 0.21808545291423798,
+      "learning_rate": 7.535394758582717e-06,
+      "loss": 0.3717,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1941510966693745,
+      "grad_norm": 0.2261328548192978,
+      "learning_rate": 7.531319389770818e-06,
+      "loss": 0.3883,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1949634443541837,
+      "grad_norm": 0.22057317197322845,
+      "learning_rate": 7.527241758472525e-06,
+      "loss": 0.3626,
+      "step": 1471
+    },
+    {
+      "epoch": 1.1957757920389926,
+      "grad_norm": 0.21070244908332825,
+      "learning_rate": 7.523161868332412e-06,
+      "loss": 0.3413,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1965881397238018,
+      "grad_norm": 0.20846307277679443,
+      "learning_rate": 7.519079722997076e-06,
+      "loss": 0.3953,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1974004874086108,
+      "grad_norm": 0.225921630859375,
+      "learning_rate": 7.514995326115124e-06,
+      "loss": 0.374,
+      "step": 1474
+    },
+    {
+      "epoch": 1.19821283509342,
+      "grad_norm": 0.25228631496429443,
+      "learning_rate": 7.5109086813371835e-06,
+      "loss": 0.385,
+      "step": 1475
+    },
+    {
+      "epoch": 1.199025182778229,
+      "grad_norm": 0.21774347126483917,
+      "learning_rate": 7.506819792315883e-06,
+      "loss": 0.3971,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1998375304630382,
+      "grad_norm": 0.22089850902557373,
+      "learning_rate": 7.502728662705861e-06,
+      "loss": 0.3626,
+      "step": 1477
+    },
+    {
+      "epoch": 1.2006498781478472,
+      "grad_norm": 0.22246122360229492,
+      "learning_rate": 7.498635296163758e-06,
+      "loss": 0.3754,
+      "step": 1478
+    },
+    {
+      "epoch": 1.2014622258326564,
+      "grad_norm": 0.2548845410346985,
+      "learning_rate": 7.494539696348215e-06,
+      "loss": 0.3681,
+      "step": 1479
+    },
+    {
+      "epoch": 1.2022745735174656,
+      "grad_norm": 0.20565412938594818,
+      "learning_rate": 7.490441866919869e-06,
+      "loss": 0.3733,
+      "step": 1480
+    },
+    {
+      "epoch": 1.2030869212022746,
+      "grad_norm": 0.19571135938167572,
+      "learning_rate": 7.486341811541345e-06,
+      "loss": 0.3758,
+      "step": 1481
+    },
+    {
+      "epoch": 1.2038992688870838,
+      "grad_norm": 0.22220642864704132,
+      "learning_rate": 7.482239533877264e-06,
+      "loss": 0.366,
+      "step": 1482
+    },
+    {
+      "epoch": 1.2047116165718927,
+      "grad_norm": 0.2103623002767563,
+      "learning_rate": 7.4781350375942296e-06,
+      "loss": 0.3707,
+      "step": 1483
+    },
+    {
+      "epoch": 1.205523964256702,
+      "grad_norm": 0.2133968025445938,
+      "learning_rate": 7.4740283263608305e-06,
+      "loss": 0.3646,
+      "step": 1484
+    },
+    {
+      "epoch": 1.206336311941511,
+      "grad_norm": 0.22945621609687805,
+      "learning_rate": 7.469919403847633e-06,
+      "loss": 0.3619,
+      "step": 1485
+    },
+    {
+      "epoch": 1.2071486596263201,
+      "grad_norm": 0.21983151137828827,
+      "learning_rate": 7.465808273727182e-06,
+      "loss": 0.3789,
+      "step": 1486
+    },
+    {
+      "epoch": 1.207961007311129,
+      "grad_norm": 0.233592689037323,
+      "learning_rate": 7.461694939673992e-06,
+      "loss": 0.3583,
+      "step": 1487
+    },
+    {
+      "epoch": 1.2087733549959383,
+      "grad_norm": 0.21641771495342255,
+      "learning_rate": 7.457579405364555e-06,
+      "loss": 0.3541,
+      "step": 1488
+    },
+    {
+      "epoch": 1.2095857026807473,
+      "grad_norm": 0.22718307375907898,
+      "learning_rate": 7.453461674477318e-06,
+      "loss": 0.3495,
+      "step": 1489
+    },
+    {
+      "epoch": 1.2103980503655565,
+      "grad_norm": 0.2268872708082199,
+      "learning_rate": 7.449341750692703e-06,
+      "loss": 0.3798,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2112103980503655,
+      "grad_norm": 0.20311862230300903,
+      "learning_rate": 7.445219637693082e-06,
+      "loss": 0.3747,
+      "step": 1491
+    },
+    {
+      "epoch": 1.2120227457351747,
+      "grad_norm": 0.1977260559797287,
+      "learning_rate": 7.441095339162791e-06,
+      "loss": 0.3811,
+      "step": 1492
+    },
+    {
+      "epoch": 1.2128350934199839,
+      "grad_norm": 0.21673862636089325,
+      "learning_rate": 7.436968858788118e-06,
+      "loss": 0.3624,
+      "step": 1493
+    },
+    {
+      "epoch": 1.2136474411047928,
+      "grad_norm": 0.23074494302272797,
+      "learning_rate": 7.4328402002572985e-06,
+      "loss": 0.3913,
+      "step": 1494
+    },
+    {
+      "epoch": 1.214459788789602,
+      "grad_norm": 0.20012116432189941,
+      "learning_rate": 7.428709367260516e-06,
+      "loss": 0.3701,
+      "step": 1495
+    },
+    {
+      "epoch": 1.215272136474411,
+      "grad_norm": 0.22077693045139313,
+      "learning_rate": 7.424576363489897e-06,
+      "loss": 0.4084,
+      "step": 1496
+    },
+    {
+      "epoch": 1.2160844841592202,
+      "grad_norm": 0.222623810172081,
+      "learning_rate": 7.420441192639511e-06,
+      "loss": 0.3775,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2168968318440292,
+      "grad_norm": 0.20730969309806824,
+      "learning_rate": 7.416303858405363e-06,
+      "loss": 0.3426,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2177091795288384,
+      "grad_norm": 0.23206910490989685,
+      "learning_rate": 7.412164364485388e-06,
+      "loss": 0.3823,
+      "step": 1499
+    },
+    {
+      "epoch": 1.2185215272136474,
+      "grad_norm": 0.2167363315820694,
+      "learning_rate": 7.408022714579457e-06,
+      "loss": 0.3649,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2193338748984566,
+      "grad_norm": 0.21870382130146027,
+      "learning_rate": 7.403878912389365e-06,
+      "loss": 0.3841,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2201462225832655,
+      "grad_norm": 0.26287588477134705,
+      "learning_rate": 7.3997329616188305e-06,
+      "loss": 0.4061,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2209585702680747,
+      "grad_norm": 0.22786174714565277,
+      "learning_rate": 7.395584865973493e-06,
+      "loss": 0.3798,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2217709179528837,
+      "grad_norm": 0.20315760374069214,
+      "learning_rate": 7.391434629160909e-06,
+      "loss": 0.3721,
+      "step": 1504
+    },
+    {
+      "epoch": 1.222583265637693,
+      "grad_norm": 0.22724851965904236,
+      "learning_rate": 7.38728225489055e-06,
+      "loss": 0.3723,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2233956133225021,
+      "grad_norm": 0.20715878903865814,
+      "learning_rate": 7.383127746873796e-06,
+      "loss": 0.3751,
+      "step": 1506
+    },
+    {
+      "epoch": 1.224207961007311,
+      "grad_norm": 0.21685153245925903,
+      "learning_rate": 7.378971108823933e-06,
+      "loss": 0.3669,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2250203086921203,
+      "grad_norm": 0.24110952019691467,
+      "learning_rate": 7.374812344456157e-06,
+      "loss": 0.3791,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2258326563769293,
+      "grad_norm": 0.22349266707897186,
+      "learning_rate": 7.370651457487559e-06,
+      "loss": 0.3865,
+      "step": 1509
+    },
+    {
+      "epoch": 1.2266450040617385,
+      "grad_norm": 0.2146981805562973,
+      "learning_rate": 7.366488451637126e-06,
+      "loss": 0.3641,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2274573517465475,
+      "grad_norm": 0.22227326035499573,
+      "learning_rate": 7.362323330625744e-06,
+      "loss": 0.3625,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2282696994313567,
+      "grad_norm": 0.22912243008613586,
+      "learning_rate": 7.358156098176185e-06,
+      "loss": 0.378,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2290820471161656,
+      "grad_norm": 0.21404817700386047,
+      "learning_rate": 7.353986758013112e-06,
+      "loss": 0.3646,
+      "step": 1513
+    },
+    {
+      "epoch": 1.2298943948009748,
+      "grad_norm": 0.21086253225803375,
+      "learning_rate": 7.349815313863068e-06,
+      "loss": 0.3821,
+      "step": 1514
+    },
+    {
+      "epoch": 1.2307067424857838,
+      "grad_norm": 0.22307930886745453,
+      "learning_rate": 7.3456417694544804e-06,
+      "loss": 0.3903,
+      "step": 1515
+    },
+    {
+      "epoch": 1.231519090170593,
+      "grad_norm": 0.21537169814109802,
+      "learning_rate": 7.341466128517649e-06,
+      "loss": 0.37,
+      "step": 1516
+    },
+    {
+      "epoch": 1.232331437855402,
+      "grad_norm": 0.21556614339351654,
+      "learning_rate": 7.337288394784754e-06,
+      "loss": 0.3593,
+      "step": 1517
+    },
+    {
+      "epoch": 1.2331437855402112,
+      "grad_norm": 0.20076751708984375,
+      "learning_rate": 7.3331085719898385e-06,
+      "loss": 0.3715,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2339561332250204,
+      "grad_norm": 0.21523988246917725,
+      "learning_rate": 7.32892666386882e-06,
+      "loss": 0.3438,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2347684809098294,
+      "grad_norm": 0.22590205073356628,
+      "learning_rate": 7.324742674159475e-06,
+      "loss": 0.362,
+      "step": 1520
+    },
+    {
+      "epoch": 1.2355808285946386,
+      "grad_norm": 0.24053142964839935,
+      "learning_rate": 7.3205566066014436e-06,
+      "loss": 0.3696,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2363931762794476,
+      "grad_norm": 0.22861604392528534,
+      "learning_rate": 7.316368464936219e-06,
+      "loss": 0.3757,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2372055239642568,
+      "grad_norm": 0.24482139945030212,
+      "learning_rate": 7.312178252907154e-06,
+      "loss": 0.3432,
+      "step": 1523
+    },
+    {
+      "epoch": 1.2380178716490657,
+      "grad_norm": 0.24735522270202637,
+      "learning_rate": 7.307985974259445e-06,
+      "loss": 0.3909,
+      "step": 1524
+    },
+    {
+      "epoch": 1.238830219333875,
+      "grad_norm": 0.22502003610134125,
+      "learning_rate": 7.303791632740143e-06,
+      "loss": 0.3778,
+      "step": 1525
+    },
+    {
+      "epoch": 1.239642567018684,
+      "grad_norm": 0.2182607650756836,
+      "learning_rate": 7.2995952320981356e-06,
+      "loss": 0.3453,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2404549147034931,
+      "grad_norm": 0.23144473135471344,
+      "learning_rate": 7.2953967760841585e-06,
+      "loss": 0.3676,
+      "step": 1527
+    },
+    {
+      "epoch": 1.2412672623883023,
+      "grad_norm": 0.23631858825683594,
+      "learning_rate": 7.291196268450777e-06,
+      "loss": 0.3767,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2420796100731113,
+      "grad_norm": 0.23022179305553436,
+      "learning_rate": 7.286993712952394e-06,
+      "loss": 0.3559,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2428919577579203,
+      "grad_norm": 0.21879801154136658,
+      "learning_rate": 7.282789113345242e-06,
+      "loss": 0.3849,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2437043054427295,
+      "grad_norm": 0.22631043195724487,
+      "learning_rate": 7.278582473387382e-06,
+      "loss": 0.3799,
+      "step": 1531
+    },
+    {
+      "epoch": 1.2445166531275387,
+      "grad_norm": 0.22810177505016327,
+      "learning_rate": 7.274373796838696e-06,
+      "loss": 0.398,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2453290008123477,
+      "grad_norm": 0.21939127147197723,
+      "learning_rate": 7.27016308746089e-06,
+      "loss": 0.3708,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2461413484971569,
+      "grad_norm": 0.22725021839141846,
+      "learning_rate": 7.265950349017481e-06,
+      "loss": 0.3747,
+      "step": 1534
+    },
+    {
+      "epoch": 1.2469536961819658,
+      "grad_norm": 0.22905021905899048,
+      "learning_rate": 7.261735585273804e-06,
+      "loss": 0.369,
+      "step": 1535
+    },
+    {
+      "epoch": 1.247766043866775,
+      "grad_norm": 0.21436680853366852,
+      "learning_rate": 7.257518799997007e-06,
+      "loss": 0.3856,
+      "step": 1536
+    },
+    {
+      "epoch": 1.248578391551584,
+      "grad_norm": 0.21557889878749847,
+      "learning_rate": 7.253299996956038e-06,
+      "loss": 0.3471,
+      "step": 1537
+    },
+    {
+      "epoch": 1.2493907392363932,
+      "grad_norm": 0.21934670209884644,
+      "learning_rate": 7.249079179921652e-06,
+      "loss": 0.3576,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2502030869212022,
+      "grad_norm": 0.21984748542308807,
+      "learning_rate": 7.244856352666404e-06,
+      "loss": 0.3557,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2510154346060114,
+      "grad_norm": 0.2120233029127121,
+      "learning_rate": 7.240631518964646e-06,
+      "loss": 0.3752,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2518277822908206,
+      "grad_norm": 0.21128596365451813,
+      "learning_rate": 7.236404682592523e-06,
+      "loss": 0.3564,
+      "step": 1541
+    },
+    {
+      "epoch": 1.2526401299756296,
+      "grad_norm": 0.23532924056053162,
+      "learning_rate": 7.232175847327969e-06,
+      "loss": 0.3629,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2534524776604385,
+      "grad_norm": 0.22936467826366425,
+      "learning_rate": 7.227945016950706e-06,
+      "loss": 0.3736,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2542648253452477,
+      "grad_norm": 0.23190708458423615,
+      "learning_rate": 7.223712195242238e-06,
+      "loss": 0.3776,
+      "step": 1544
+    },
+    {
+      "epoch": 1.255077173030057,
+      "grad_norm": 0.23108626902103424,
+      "learning_rate": 7.2194773859858495e-06,
+      "loss": 0.3808,
+      "step": 1545
+    },
+    {
+      "epoch": 1.255889520714866,
+      "grad_norm": 0.244134783744812,
+      "learning_rate": 7.215240592966603e-06,
+      "loss": 0.372,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2567018683996751,
+      "grad_norm": 0.23643700778484344,
+      "learning_rate": 7.2110018199713325e-06,
+      "loss": 0.3926,
+      "step": 1547
+    },
+    {
+      "epoch": 1.257514216084484,
+      "grad_norm": 0.22385773062705994,
+      "learning_rate": 7.206761070788641e-06,
+      "loss": 0.3697,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2583265637692933,
+      "grad_norm": 0.21756193041801453,
+      "learning_rate": 7.202518349208898e-06,
+      "loss": 0.3688,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2591389114541023,
+      "grad_norm": 0.22555877268314362,
+      "learning_rate": 7.1982736590242375e-06,
+      "loss": 0.3707,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2599512591389115,
+      "grad_norm": 0.2219279259443283,
+      "learning_rate": 7.194027004028552e-06,
+      "loss": 0.3598,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2607636068237205,
+      "grad_norm": 0.22899870574474335,
+      "learning_rate": 7.189778388017489e-06,
+      "loss": 0.3753,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2615759545085297,
+      "grad_norm": 0.19873790442943573,
+      "learning_rate": 7.185527814788451e-06,
+      "loss": 0.3728,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2623883021933389,
+      "grad_norm": 0.20611122250556946,
+      "learning_rate": 7.18127528814059e-06,
+      "loss": 0.35,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2632006498781478,
+      "grad_norm": 0.23903603851795197,
+      "learning_rate": 7.1770208118748e-06,
+      "loss": 0.3519,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2640129975629568,
+      "grad_norm": 0.23457713425159454,
+      "learning_rate": 7.172764389793723e-06,
+      "loss": 0.375,
+      "step": 1556
+    },
+    {
+      "epoch": 1.264825345247766,
+      "grad_norm": 0.21742449700832367,
+      "learning_rate": 7.168506025701734e-06,
+      "loss": 0.3653,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2656376929325752,
+      "grad_norm": 0.23635496199131012,
+      "learning_rate": 7.164245723404951e-06,
+      "loss": 0.3691,
+      "step": 1558
+    },
+    {
+      "epoch": 1.2664500406173842,
+      "grad_norm": 0.23171082139015198,
+      "learning_rate": 7.159983486711219e-06,
+      "loss": 0.3677,
+      "step": 1559
+    },
+    {
+      "epoch": 1.2672623883021934,
+      "grad_norm": 0.22449378669261932,
+      "learning_rate": 7.155719319430114e-06,
+      "loss": 0.3806,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2680747359870024,
+      "grad_norm": 0.21499858796596527,
+      "learning_rate": 7.151453225372938e-06,
+      "loss": 0.3801,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2688870836718116,
+      "grad_norm": 0.24271003901958466,
+      "learning_rate": 7.147185208352712e-06,
+      "loss": 0.3738,
+      "step": 1562
+    },
+    {
+      "epoch": 1.2696994313566206,
+      "grad_norm": 0.1977013647556305,
+      "learning_rate": 7.142915272184179e-06,
+      "loss": 0.3646,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2705117790414298,
+      "grad_norm": 0.2271679937839508,
+      "learning_rate": 7.138643420683795e-06,
+      "loss": 0.3673,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2713241267262387,
+      "grad_norm": 0.22604072093963623,
+      "learning_rate": 7.134369657669733e-06,
+      "loss": 0.381,
+      "step": 1565
+    },
+    {
+      "epoch": 1.272136474411048,
+      "grad_norm": 0.2143831104040146,
+      "learning_rate": 7.130093986961868e-06,
+      "loss": 0.3691,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2729488220958571,
+      "grad_norm": 0.2499876469373703,
+      "learning_rate": 7.1258164123817834e-06,
+      "loss": 0.3613,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2737611697806661,
+      "grad_norm": 0.235340416431427,
+      "learning_rate": 7.121536937752764e-06,
+      "loss": 0.3668,
+      "step": 1568
+    },
+    {
+      "epoch": 1.274573517465475,
+      "grad_norm": 0.1983313411474228,
+      "learning_rate": 7.1172555668997925e-06,
+      "loss": 0.3696,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2753858651502843,
+      "grad_norm": 0.21165205538272858,
+      "learning_rate": 7.112972303649546e-06,
+      "loss": 0.3768,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2761982128350935,
+      "grad_norm": 0.24132968485355377,
+      "learning_rate": 7.108687151830394e-06,
+      "loss": 0.3646,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2770105605199025,
+      "grad_norm": 0.2138853520154953,
+      "learning_rate": 7.104400115272393e-06,
+      "loss": 0.3695,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2778229082047117,
+      "grad_norm": 0.20121292769908905,
+      "learning_rate": 7.100111197807286e-06,
+      "loss": 0.3751,
+      "step": 1573
+    },
+    {
+      "epoch": 1.2786352558895206,
+      "grad_norm": 0.21837638318538666,
+      "learning_rate": 7.095820403268494e-06,
+      "loss": 0.39,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2794476035743299,
+      "grad_norm": 0.22090759873390198,
+      "learning_rate": 7.091527735491119e-06,
+      "loss": 0.3888,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2802599512591388,
+      "grad_norm": 0.23896671831607819,
+      "learning_rate": 7.087233198311936e-06,
+      "loss": 0.3865,
+      "step": 1576
+    },
+    {
+      "epoch": 1.281072298943948,
+      "grad_norm": 0.2100042998790741,
+      "learning_rate": 7.082936795569388e-06,
+      "loss": 0.3958,
+      "step": 1577
+    },
+    {
+      "epoch": 1.281884646628757,
+      "grad_norm": 0.21028761565685272,
+      "learning_rate": 7.078638531103592e-06,
+      "loss": 0.3928,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2826969943135662,
+      "grad_norm": 0.20510919392108917,
+      "learning_rate": 7.074338408756321e-06,
+      "loss": 0.3741,
+      "step": 1579
+    },
+    {
+      "epoch": 1.2835093419983754,
+      "grad_norm": 0.2120136022567749,
+      "learning_rate": 7.070036432371017e-06,
+      "loss": 0.3537,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2843216896831844,
+      "grad_norm": 0.232395738363266,
+      "learning_rate": 7.065732605792771e-06,
+      "loss": 0.372,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2851340373679934,
+      "grad_norm": 0.1969270408153534,
+      "learning_rate": 7.061426932868334e-06,
+      "loss": 0.3617,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2859463850528026,
+      "grad_norm": 0.22735582292079926,
+      "learning_rate": 7.057119417446102e-06,
+      "loss": 0.3905,
+      "step": 1583
+    },
+    {
+      "epoch": 1.2867587327376118,
+      "grad_norm": 0.2126646786928177,
+      "learning_rate": 7.052810063376124e-06,
+      "loss": 0.3627,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2875710804224207,
+      "grad_norm": 0.22967879474163055,
+      "learning_rate": 7.048498874510088e-06,
+      "loss": 0.3624,
+      "step": 1585
+    },
+    {
+      "epoch": 1.28838342810723,
+      "grad_norm": 0.205107182264328,
+      "learning_rate": 7.044185854701321e-06,
+      "loss": 0.3545,
+      "step": 1586
+    },
+    {
+      "epoch": 1.289195775792039,
+      "grad_norm": 0.22394341230392456,
+      "learning_rate": 7.03987100780479e-06,
+      "loss": 0.3652,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2900081234768481,
+      "grad_norm": 0.21900473535060883,
+      "learning_rate": 7.035554337677094e-06,
+      "loss": 0.3751,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2908204711616573,
+      "grad_norm": 0.22297774255275726,
+      "learning_rate": 7.03123584817646e-06,
+      "loss": 0.3455,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2916328188464663,
+      "grad_norm": 0.22928905487060547,
+      "learning_rate": 7.026915543162741e-06,
+      "loss": 0.3784,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2924451665312753,
+      "grad_norm": 0.22926761209964752,
+      "learning_rate": 7.0225934264974175e-06,
+      "loss": 0.3814,
+      "step": 1591
+    },
+    {
+      "epoch": 1.2932575142160845,
+      "grad_norm": 0.21393662691116333,
+      "learning_rate": 7.018269502043581e-06,
+      "loss": 0.3736,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2940698619008937,
+      "grad_norm": 0.20423708856105804,
+      "learning_rate": 7.013943773665947e-06,
+      "loss": 0.3873,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2948822095857027,
+      "grad_norm": 0.22244593501091003,
+      "learning_rate": 7.009616245230838e-06,
+      "loss": 0.4003,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2956945572705119,
+      "grad_norm": 0.22295843064785004,
+      "learning_rate": 7.005286920606188e-06,
+      "loss": 0.3647,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2965069049553208,
+      "grad_norm": 0.19805020093917847,
+      "learning_rate": 7.0009558036615355e-06,
+      "loss": 0.3598,
+      "step": 1596
+    },
+    {
+      "epoch": 1.29731925264013,
+      "grad_norm": 0.24980755150318146,
+      "learning_rate": 6.996622898268021e-06,
+      "loss": 0.3845,
+      "step": 1597
+    },
+    {
+      "epoch": 1.298131600324939,
+      "grad_norm": 0.23719848692417145,
+      "learning_rate": 6.992288208298383e-06,
+      "loss": 0.3705,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2989439480097482,
+      "grad_norm": 0.2210855931043625,
+      "learning_rate": 6.987951737626956e-06,
+      "loss": 0.376,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2997562956945572,
+      "grad_norm": 0.21392883360385895,
+      "learning_rate": 6.983613490129666e-06,
+      "loss": 0.3698,
+      "step": 1600
+    },
+    {
+      "epoch": 1.3005686433793664,
+      "grad_norm": 0.21564728021621704,
+      "learning_rate": 6.979273469684026e-06,
+      "loss": 0.3683,
+      "step": 1601
+    },
+    {
+      "epoch": 1.3013809910641756,
+      "grad_norm": 0.2227163165807724,
+      "learning_rate": 6.974931680169136e-06,
+      "loss": 0.3807,
+      "step": 1602
+    },
+    {
+      "epoch": 1.3021933387489846,
+      "grad_norm": 0.2266252487897873,
+      "learning_rate": 6.970588125465674e-06,
+      "loss": 0.3556,
+      "step": 1603
+    },
+    {
+      "epoch": 1.3030056864337936,
+      "grad_norm": 0.20680905878543854,
+      "learning_rate": 6.966242809455899e-06,
+      "loss": 0.3524,
+      "step": 1604
+    },
+    {
+      "epoch": 1.3038180341186028,
+      "grad_norm": 0.19485142827033997,
+      "learning_rate": 6.961895736023641e-06,
+      "loss": 0.358,
+      "step": 1605
+    },
+    {
+      "epoch": 1.304630381803412,
+      "grad_norm": 0.23355814814567566,
+      "learning_rate": 6.957546909054304e-06,
+      "loss": 0.3686,
+      "step": 1606
+    },
+    {
+      "epoch": 1.305442729488221,
+      "grad_norm": 0.2633036673069,
+      "learning_rate": 6.953196332434856e-06,
+      "loss": 0.3741,
+      "step": 1607
+    },
+    {
+      "epoch": 1.3062550771730301,
+      "grad_norm": 0.2111603319644928,
+      "learning_rate": 6.948844010053832e-06,
+      "loss": 0.3728,
+      "step": 1608
+    },
+    {
+      "epoch": 1.307067424857839,
+      "grad_norm": 0.25015655159950256,
+      "learning_rate": 6.944489945801326e-06,
+      "loss": 0.3587,
+      "step": 1609
+    },
+    {
+      "epoch": 1.3078797725426483,
+      "grad_norm": 0.22723735868930817,
+      "learning_rate": 6.940134143568987e-06,
+      "loss": 0.3708,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3086921202274573,
+      "grad_norm": 0.21782447397708893,
+      "learning_rate": 6.935776607250022e-06,
+      "loss": 0.3815,
+      "step": 1611
+    },
+    {
+      "epoch": 1.3095044679122665,
+      "grad_norm": 0.2229449599981308,
+      "learning_rate": 6.931417340739183e-06,
+      "loss": 0.3724,
+      "step": 1612
+    },
+    {
+      "epoch": 1.3103168155970755,
+      "grad_norm": 0.23954704403877258,
+      "learning_rate": 6.927056347932772e-06,
+      "loss": 0.3717,
+      "step": 1613
+    },
+    {
+      "epoch": 1.3111291632818847,
+      "grad_norm": 0.20666253566741943,
+      "learning_rate": 6.92269363272863e-06,
+      "loss": 0.3853,
+      "step": 1614
+    },
+    {
+      "epoch": 1.3119415109666939,
+      "grad_norm": 0.2357204407453537,
+      "learning_rate": 6.918329199026143e-06,
+      "loss": 0.3808,
+      "step": 1615
+    },
+    {
+      "epoch": 1.3127538586515028,
+      "grad_norm": 0.2109731137752533,
+      "learning_rate": 6.913963050726226e-06,
+      "loss": 0.3631,
+      "step": 1616
+    },
+    {
+      "epoch": 1.3135662063363118,
+      "grad_norm": 0.21741336584091187,
+      "learning_rate": 6.909595191731337e-06,
+      "loss": 0.3585,
+      "step": 1617
+    },
+    {
+      "epoch": 1.314378554021121,
+      "grad_norm": 0.22346995770931244,
+      "learning_rate": 6.905225625945451e-06,
+      "loss": 0.3625,
+      "step": 1618
+    },
+    {
+      "epoch": 1.3151909017059302,
+      "grad_norm": 0.24195951223373413,
+      "learning_rate": 6.900854357274075e-06,
+      "loss": 0.3964,
+      "step": 1619
+    },
+    {
+      "epoch": 1.3160032493907392,
+      "grad_norm": 0.20008091628551483,
+      "learning_rate": 6.896481389624239e-06,
+      "loss": 0.3779,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3168155970755484,
+      "grad_norm": 0.22882522642612457,
+      "learning_rate": 6.892106726904486e-06,
+      "loss": 0.363,
+      "step": 1621
+    },
+    {
+      "epoch": 1.3176279447603574,
+      "grad_norm": 0.2112092524766922,
+      "learning_rate": 6.887730373024881e-06,
+      "loss": 0.3791,
+      "step": 1622
+    },
+    {
+      "epoch": 1.3184402924451666,
+      "grad_norm": 0.2170286923646927,
+      "learning_rate": 6.883352331896998e-06,
+      "loss": 0.3712,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3192526401299756,
+      "grad_norm": 0.20295237004756927,
+      "learning_rate": 6.878972607433915e-06,
+      "loss": 0.3502,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3200649878147848,
+      "grad_norm": 0.21459592878818512,
+      "learning_rate": 6.8745912035502205e-06,
+      "loss": 0.3597,
+      "step": 1625
+    },
+    {
+      "epoch": 1.3208773354995937,
+      "grad_norm": 0.21338880062103271,
+      "learning_rate": 6.870208124161998e-06,
+      "loss": 0.3503,
+      "step": 1626
+    },
+    {
+      "epoch": 1.321689683184403,
+      "grad_norm": 0.21886205673217773,
+      "learning_rate": 6.8658233731868355e-06,
+      "loss": 0.356,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3225020308692121,
+      "grad_norm": 0.23136785626411438,
+      "learning_rate": 6.86143695454381e-06,
+      "loss": 0.3649,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3233143785540211,
+      "grad_norm": 0.21470150351524353,
+      "learning_rate": 6.857048872153491e-06,
+      "loss": 0.3627,
+      "step": 1629
+    },
+    {
+      "epoch": 1.32412672623883,
+      "grad_norm": 0.21059399843215942,
+      "learning_rate": 6.852659129937936e-06,
+      "loss": 0.3728,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3249390739236393,
+      "grad_norm": 0.22825659811496735,
+      "learning_rate": 6.84826773182068e-06,
+      "loss": 0.372,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3257514216084485,
+      "grad_norm": 0.21476778388023376,
+      "learning_rate": 6.843874681726747e-06,
+      "loss": 0.3713,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3265637692932575,
+      "grad_norm": 0.26189494132995605,
+      "learning_rate": 6.839479983582632e-06,
+      "loss": 0.3672,
+      "step": 1633
+    },
+    {
+      "epoch": 1.3273761169780667,
+      "grad_norm": 0.2089308500289917,
+      "learning_rate": 6.835083641316304e-06,
+      "loss": 0.382,
+      "step": 1634
+    },
+    {
+      "epoch": 1.3281884646628757,
+      "grad_norm": 0.20792783796787262,
+      "learning_rate": 6.830685658857203e-06,
+      "loss": 0.3445,
+      "step": 1635
+    },
+    {
+      "epoch": 1.3290008123476849,
+      "grad_norm": 0.22559711337089539,
+      "learning_rate": 6.826286040136232e-06,
+      "loss": 0.3645,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3298131600324938,
+      "grad_norm": 0.2342708855867386,
+      "learning_rate": 6.82188478908576e-06,
+      "loss": 0.3435,
+      "step": 1637
+    },
+    {
+      "epoch": 1.330625507717303,
+      "grad_norm": 0.21147669851779938,
+      "learning_rate": 6.817481909639611e-06,
+      "loss": 0.381,
+      "step": 1638
+    },
+    {
+      "epoch": 1.331437855402112,
+      "grad_norm": 0.21588994562625885,
+      "learning_rate": 6.813077405733068e-06,
+      "loss": 0.3672,
+      "step": 1639
+    },
+    {
+      "epoch": 1.3322502030869212,
+      "grad_norm": 0.22194498777389526,
+      "learning_rate": 6.808671281302865e-06,
+      "loss": 0.3611,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3330625507717304,
+      "grad_norm": 0.19620025157928467,
+      "learning_rate": 6.804263540287184e-06,
+      "loss": 0.3741,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3338748984565394,
+      "grad_norm": 0.2333213835954666,
+      "learning_rate": 6.799854186625651e-06,
+      "loss": 0.3779,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3346872461413484,
+      "grad_norm": 0.22805556654930115,
+      "learning_rate": 6.795443224259335e-06,
+      "loss": 0.3825,
+      "step": 1643
+    },
+    {
+      "epoch": 1.3354995938261576,
+      "grad_norm": 0.22976873815059662,
+      "learning_rate": 6.7910306571307425e-06,
+      "loss": 0.3795,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3363119415109668,
+      "grad_norm": 0.22796057164669037,
+      "learning_rate": 6.786616489183814e-06,
+      "loss": 0.3999,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3371242891957758,
+      "grad_norm": 0.2186834216117859,
+      "learning_rate": 6.78220072436392e-06,
+      "loss": 0.3914,
+      "step": 1646
+    },
+    {
+      "epoch": 1.337936636880585,
+      "grad_norm": 0.23626239597797394,
+      "learning_rate": 6.7777833666178594e-06,
+      "loss": 0.3783,
+      "step": 1647
+    },
+    {
+      "epoch": 1.338748984565394,
+      "grad_norm": 0.2400982528924942,
+      "learning_rate": 6.773364419893854e-06,
+      "loss": 0.3772,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3395613322502031,
+      "grad_norm": 0.21469588577747345,
+      "learning_rate": 6.768943888141548e-06,
+      "loss": 0.3489,
+      "step": 1649
+    },
+    {
+      "epoch": 1.340373679935012,
+      "grad_norm": 0.2091098129749298,
+      "learning_rate": 6.764521775312e-06,
+      "loss": 0.3679,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3411860276198213,
+      "grad_norm": 0.22241105139255524,
+      "learning_rate": 6.760098085357681e-06,
+      "loss": 0.3644,
+      "step": 1651
+    },
+    {
+      "epoch": 1.3419983753046303,
+      "grad_norm": 0.22590091824531555,
+      "learning_rate": 6.755672822232475e-06,
+      "loss": 0.3514,
+      "step": 1652
+    },
+    {
+      "epoch": 1.3428107229894395,
+      "grad_norm": 0.22745101153850555,
+      "learning_rate": 6.7512459898916704e-06,
+      "loss": 0.385,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3436230706742487,
+      "grad_norm": 0.20958414673805237,
+      "learning_rate": 6.746817592291957e-06,
+      "loss": 0.3786,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3444354183590577,
+      "grad_norm": 0.21807672083377838,
+      "learning_rate": 6.742387633391423e-06,
+      "loss": 0.38,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3452477660438666,
+      "grad_norm": 0.22972120344638824,
+      "learning_rate": 6.737956117149555e-06,
+      "loss": 0.3598,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3460601137286758,
+      "grad_norm": 0.21386568248271942,
+      "learning_rate": 6.733523047527231e-06,
+      "loss": 0.3533,
+      "step": 1657
+    },
+    {
+      "epoch": 1.346872461413485,
+      "grad_norm": 0.19474758207798004,
+      "learning_rate": 6.729088428486717e-06,
+      "loss": 0.362,
+      "step": 1658
+    },
+    {
+      "epoch": 1.347684809098294,
+      "grad_norm": 0.24409079551696777,
+      "learning_rate": 6.72465226399166e-06,
+      "loss": 0.3767,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3484971567831032,
+      "grad_norm": 0.20776718854904175,
+      "learning_rate": 6.720214558007094e-06,
+      "loss": 0.36,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3493095044679122,
+      "grad_norm": 0.20188705623149872,
+      "learning_rate": 6.715775314499431e-06,
+      "loss": 0.3624,
+      "step": 1661
+    },
+    {
+      "epoch": 1.3501218521527214,
+      "grad_norm": 0.2113410085439682,
+      "learning_rate": 6.711334537436451e-06,
+      "loss": 0.3947,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3509341998375304,
+      "grad_norm": 0.19894282519817352,
+      "learning_rate": 6.7068922307873104e-06,
+      "loss": 0.3532,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3517465475223396,
+      "grad_norm": 0.2061915099620819,
+      "learning_rate": 6.7024483985225295e-06,
+      "loss": 0.3713,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3525588952071486,
+      "grad_norm": 0.20251737534999847,
+      "learning_rate": 6.698003044613997e-06,
+      "loss": 0.3728,
+      "step": 1665
+    },
+    {
+      "epoch": 1.3533712428919578,
+      "grad_norm": 0.2825920283794403,
+      "learning_rate": 6.693556173034953e-06,
+      "loss": 0.3706,
+      "step": 1666
+    },
+    {
+      "epoch": 1.354183590576767,
+      "grad_norm": 0.21718469262123108,
+      "learning_rate": 6.689107787760002e-06,
+      "loss": 0.3626,
+      "step": 1667
+    },
+    {
+      "epoch": 1.354995938261576,
+      "grad_norm": 0.2175828516483307,
+      "learning_rate": 6.6846578927650985e-06,
+      "loss": 0.3641,
+      "step": 1668
+    },
+    {
+      "epoch": 1.355808285946385,
+      "grad_norm": 0.21424439549446106,
+      "learning_rate": 6.6802064920275475e-06,
+      "loss": 0.3612,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3566206336311941,
+      "grad_norm": 0.21859072148799896,
+      "learning_rate": 6.675753589525996e-06,
+      "loss": 0.3659,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3574329813160033,
+      "grad_norm": 0.21158170700073242,
+      "learning_rate": 6.671299189240439e-06,
+      "loss": 0.3654,
+      "step": 1671
+    },
+    {
+      "epoch": 1.3582453290008123,
+      "grad_norm": 0.21030624210834503,
+      "learning_rate": 6.666843295152207e-06,
+      "loss": 0.3795,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3590576766856215,
+      "grad_norm": 0.22555941343307495,
+      "learning_rate": 6.662385911243965e-06,
+      "loss": 0.3583,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3598700243704305,
+      "grad_norm": 0.1947505623102188,
+      "learning_rate": 6.657927041499711e-06,
+      "loss": 0.3482,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3606823720552397,
+      "grad_norm": 0.2315574735403061,
+      "learning_rate": 6.653466689904771e-06,
+      "loss": 0.3875,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3614947197400489,
+      "grad_norm": 0.21630355715751648,
+      "learning_rate": 6.649004860445795e-06,
+      "loss": 0.3734,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3623070674248579,
+      "grad_norm": 0.21112656593322754,
+      "learning_rate": 6.644541557110753e-06,
+      "loss": 0.3901,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3631194151096668,
+      "grad_norm": 0.24334968626499176,
+      "learning_rate": 6.640076783888934e-06,
+      "loss": 0.3933,
+      "step": 1678
+    },
+    {
+      "epoch": 1.363931762794476,
+      "grad_norm": 0.21816273033618927,
+      "learning_rate": 6.63561054477094e-06,
+      "loss": 0.3522,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3647441104792852,
+      "grad_norm": 0.1857845038175583,
+      "learning_rate": 6.631142843748686e-06,
+      "loss": 0.3564,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3655564581640942,
+      "grad_norm": 0.20670706033706665,
+      "learning_rate": 6.626673684815389e-06,
+      "loss": 0.3465,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3663688058489034,
+      "grad_norm": 0.23002231121063232,
+      "learning_rate": 6.62220307196557e-06,
+      "loss": 0.3727,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3671811535337124,
+      "grad_norm": 0.2385770082473755,
+      "learning_rate": 6.617731009195052e-06,
+      "loss": 0.368,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3679935012185216,
+      "grad_norm": 0.23323537409305573,
+      "learning_rate": 6.613257500500952e-06,
+      "loss": 0.3684,
+      "step": 1684
+    },
+    {
+      "epoch": 1.3688058489033306,
+      "grad_norm": 0.21502023935317993,
+      "learning_rate": 6.60878254988168e-06,
+      "loss": 0.3773,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3696181965881398,
+      "grad_norm": 0.22893846035003662,
+      "learning_rate": 6.6043061613369356e-06,
+      "loss": 0.388,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3704305442729487,
+      "grad_norm": 0.23281772434711456,
+      "learning_rate": 6.5998283388677005e-06,
+      "loss": 0.374,
+      "step": 1687
+    },
+    {
+      "epoch": 1.371242891957758,
+      "grad_norm": 0.2198653370141983,
+      "learning_rate": 6.595349086476244e-06,
+      "loss": 0.3525,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3720552396425671,
+      "grad_norm": 0.23118118941783905,
+      "learning_rate": 6.590868408166105e-06,
+      "loss": 0.3759,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3728675873273761,
+      "grad_norm": 0.22136609256267548,
+      "learning_rate": 6.586386307942105e-06,
+      "loss": 0.353,
+      "step": 1690
+    },
+    {
+      "epoch": 1.373679935012185,
+      "grad_norm": 0.22904831171035767,
+      "learning_rate": 6.581902789810333e-06,
+      "loss": 0.3579,
+      "step": 1691
+    },
+    {
+      "epoch": 1.3744922826969943,
+      "grad_norm": 0.2517954409122467,
+      "learning_rate": 6.577417857778142e-06,
+      "loss": 0.3772,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3753046303818035,
+      "grad_norm": 0.21787036955356598,
+      "learning_rate": 6.572931515854158e-06,
+      "loss": 0.3447,
+      "step": 1693
+    },
+    {
+      "epoch": 1.3761169780666125,
+      "grad_norm": 0.22375719249248505,
+      "learning_rate": 6.568443768048254e-06,
+      "loss": 0.3623,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3769293257514217,
+      "grad_norm": 0.226768359541893,
+      "learning_rate": 6.563954618371573e-06,
+      "loss": 0.353,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3777416734362307,
+      "grad_norm": 0.20279915630817413,
+      "learning_rate": 6.559464070836501e-06,
+      "loss": 0.387,
+      "step": 1696
+    },
+    {
+      "epoch": 1.3785540211210399,
+      "grad_norm": 0.24256636202335358,
+      "learning_rate": 6.5549721294566794e-06,
+      "loss": 0.3629,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3793663688058488,
+      "grad_norm": 0.25589874386787415,
+      "learning_rate": 6.550478798246991e-06,
+      "loss": 0.3654,
+      "step": 1698
+    },
+    {
+      "epoch": 1.380178716490658,
+      "grad_norm": 0.22566331923007965,
+      "learning_rate": 6.545984081223564e-06,
+      "loss": 0.3782,
+      "step": 1699
+    },
+    {
+      "epoch": 1.380991064175467,
+      "grad_norm": 0.20067761838436127,
+      "learning_rate": 6.541487982403767e-06,
+      "loss": 0.3733,
+      "step": 1700
+    },
+    {
+      "epoch": 1.3818034118602762,
+      "grad_norm": 0.21134236454963684,
+      "learning_rate": 6.536990505806197e-06,
+      "loss": 0.3726,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3826157595450854,
+      "grad_norm": 0.20838569104671478,
+      "learning_rate": 6.532491655450688e-06,
+      "loss": 0.3561,
+      "step": 1702
+    },
+    {
+      "epoch": 1.3834281072298944,
+      "grad_norm": 0.22978723049163818,
+      "learning_rate": 6.5279914353582996e-06,
+      "loss": 0.3494,
+      "step": 1703
+    },
+    {
+      "epoch": 1.3842404549147034,
+      "grad_norm": 0.19743117690086365,
+      "learning_rate": 6.523489849551316e-06,
+      "loss": 0.3274,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3850528025995126,
+      "grad_norm": 0.1934005618095398,
+      "learning_rate": 6.518986902053244e-06,
+      "loss": 0.3546,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3858651502843218,
+      "grad_norm": 0.22205540537834167,
+      "learning_rate": 6.514482596888807e-06,
+      "loss": 0.3561,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3866774979691308,
+      "grad_norm": 0.19166147708892822,
+      "learning_rate": 6.509976938083936e-06,
+      "loss": 0.3636,
+      "step": 1707
+    },
+    {
+      "epoch": 1.38748984565394,
+      "grad_norm": 0.2089797407388687,
+      "learning_rate": 6.505469929665781e-06,
+      "loss": 0.3582,
+      "step": 1708
+    },
+    {
+      "epoch": 1.388302193338749,
+      "grad_norm": 0.21961554884910583,
+      "learning_rate": 6.500961575662691e-06,
+      "loss": 0.3796,
+      "step": 1709
+    },
+    {
+      "epoch": 1.3891145410235581,
+      "grad_norm": 0.23188024759292603,
+      "learning_rate": 6.496451880104222e-06,
+      "loss": 0.3575,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3899268887083671,
+      "grad_norm": 0.19699709117412567,
+      "learning_rate": 6.491940847021128e-06,
+      "loss": 0.3472,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3907392363931763,
+      "grad_norm": 0.20743319392204285,
+      "learning_rate": 6.487428480445357e-06,
+      "loss": 0.3628,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3915515840779853,
+      "grad_norm": 0.2423766553401947,
+      "learning_rate": 6.48291478441005e-06,
+      "loss": 0.3645,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3923639317627945,
+      "grad_norm": 0.22434589266777039,
+      "learning_rate": 6.478399762949537e-06,
+      "loss": 0.3635,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3931762794476037,
+      "grad_norm": 0.20460708439350128,
+      "learning_rate": 6.47388342009933e-06,
+      "loss": 0.3596,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3939886271324127,
+      "grad_norm": 0.22267088294029236,
+      "learning_rate": 6.469365759896128e-06,
+      "loss": 0.3683,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3948009748172217,
+      "grad_norm": 0.22853101789951324,
+      "learning_rate": 6.464846786377801e-06,
+      "loss": 0.3718,
+      "step": 1717
+    },
+    {
+      "epoch": 1.3956133225020309,
+      "grad_norm": 0.22584837675094604,
+      "learning_rate": 6.460326503583394e-06,
+      "loss": 0.3467,
+      "step": 1718
+    },
+    {
+      "epoch": 1.39642567018684,
+      "grad_norm": 0.20108698308467865,
+      "learning_rate": 6.4558049155531255e-06,
+      "loss": 0.3353,
+      "step": 1719
+    },
+    {
+      "epoch": 1.397238017871649,
+      "grad_norm": 0.23486927151679993,
+      "learning_rate": 6.451282026328377e-06,
+      "loss": 0.3564,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3980503655564582,
+      "grad_norm": 0.22398073971271515,
+      "learning_rate": 6.4467578399516965e-06,
+      "loss": 0.3686,
+      "step": 1721
+    },
+    {
+      "epoch": 1.3988627132412672,
+      "grad_norm": 0.25351375341415405,
+      "learning_rate": 6.442232360466789e-06,
+      "loss": 0.3699,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3996750609260764,
+      "grad_norm": 0.19639666378498077,
+      "learning_rate": 6.437705591918514e-06,
+      "loss": 0.366,
+      "step": 1723
+    },
+    {
+      "epoch": 1.4004874086108854,
+      "grad_norm": 0.20924267172813416,
+      "learning_rate": 6.433177538352887e-06,
+      "loss": 0.3735,
+      "step": 1724
+    },
+    {
+      "epoch": 1.4012997562956946,
+      "grad_norm": 0.21767950057983398,
+      "learning_rate": 6.428648203817069e-06,
+      "loss": 0.3671,
+      "step": 1725
+    },
+    {
+      "epoch": 1.4021121039805036,
+      "grad_norm": 0.21599704027175903,
+      "learning_rate": 6.424117592359367e-06,
+      "loss": 0.3586,
+      "step": 1726
+    },
+    {
+      "epoch": 1.4029244516653128,
+      "grad_norm": 0.20830616354942322,
+      "learning_rate": 6.419585708029229e-06,
+      "loss": 0.3683,
+      "step": 1727
+    },
+    {
+      "epoch": 1.403736799350122,
+      "grad_norm": 0.22411523759365082,
+      "learning_rate": 6.415052554877241e-06,
+      "loss": 0.353,
+      "step": 1728
+    },
+    {
+      "epoch": 1.404549147034931,
+      "grad_norm": 0.21645550429821014,
+      "learning_rate": 6.410518136955124e-06,
+      "loss": 0.3698,
+      "step": 1729
+    },
+    {
+      "epoch": 1.40536149471974,
+      "grad_norm": 0.21562649309635162,
+      "learning_rate": 6.405982458315727e-06,
+      "loss": 0.3536,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4061738424045491,
+      "grad_norm": 0.21301081776618958,
+      "learning_rate": 6.401445523013028e-06,
+      "loss": 0.3829,
+      "step": 1731
+    },
+    {
+      "epoch": 1.4069861900893583,
+      "grad_norm": 0.21843348443508148,
+      "learning_rate": 6.396907335102127e-06,
+      "loss": 0.3725,
+      "step": 1732
+    },
+    {
+      "epoch": 1.4077985377741673,
+      "grad_norm": 0.2166627198457718,
+      "learning_rate": 6.392367898639245e-06,
+      "loss": 0.3662,
+      "step": 1733
+    },
+    {
+      "epoch": 1.4086108854589765,
+      "grad_norm": 0.21323154866695404,
+      "learning_rate": 6.387827217681717e-06,
+      "loss": 0.3728,
+      "step": 1734
+    },
+    {
+      "epoch": 1.4094232331437855,
+      "grad_norm": 0.21641981601715088,
+      "learning_rate": 6.383285296287991e-06,
+      "loss": 0.3521,
+      "step": 1735
+    },
+    {
+      "epoch": 1.4102355808285947,
+      "grad_norm": 0.19464868307113647,
+      "learning_rate": 6.378742138517624e-06,
+      "loss": 0.3504,
+      "step": 1736
+    },
+    {
+      "epoch": 1.4110479285134037,
+      "grad_norm": 0.21125568449497223,
+      "learning_rate": 6.37419774843128e-06,
+      "loss": 0.3614,
+      "step": 1737
+    },
+    {
+      "epoch": 1.4118602761982129,
+      "grad_norm": 0.20467282831668854,
+      "learning_rate": 6.3696521300907214e-06,
+      "loss": 0.3634,
+      "step": 1738
+    },
+    {
+      "epoch": 1.4126726238830218,
+      "grad_norm": 0.23788075149059296,
+      "learning_rate": 6.3651052875588085e-06,
+      "loss": 0.3447,
+      "step": 1739
+    },
+    {
+      "epoch": 1.413484971567831,
+      "grad_norm": 0.2198992520570755,
+      "learning_rate": 6.360557224899496e-06,
+      "loss": 0.3474,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4142973192526402,
+      "grad_norm": 0.24107486009597778,
+      "learning_rate": 6.356007946177833e-06,
+      "loss": 0.3923,
+      "step": 1741
+    },
+    {
+      "epoch": 1.4151096669374492,
+      "grad_norm": 0.21335986256599426,
+      "learning_rate": 6.351457455459953e-06,
+      "loss": 0.3569,
+      "step": 1742
+    },
+    {
+      "epoch": 1.4159220146222582,
+      "grad_norm": 0.2371571660041809,
+      "learning_rate": 6.346905756813069e-06,
+      "loss": 0.3678,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4167343623070674,
+      "grad_norm": 0.2299576699733734,
+      "learning_rate": 6.342352854305477e-06,
+      "loss": 0.3737,
+      "step": 1744
+    },
+    {
+      "epoch": 1.4175467099918766,
+      "grad_norm": 0.22928734123706818,
+      "learning_rate": 6.337798752006551e-06,
+      "loss": 0.359,
+      "step": 1745
+    },
+    {
+      "epoch": 1.4183590576766856,
+      "grad_norm": 0.21503311395645142,
+      "learning_rate": 6.333243453986734e-06,
+      "loss": 0.373,
+      "step": 1746
+    },
+    {
+      "epoch": 1.4191714053614948,
+      "grad_norm": 0.21958230435848236,
+      "learning_rate": 6.32868696431754e-06,
+      "loss": 0.37,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4199837530463038,
+      "grad_norm": 0.20250442624092102,
+      "learning_rate": 6.324129287071546e-06,
+      "loss": 0.3602,
+      "step": 1748
+    },
+    {
+      "epoch": 1.420796100731113,
+      "grad_norm": 0.212167888879776,
+      "learning_rate": 6.3195704263223914e-06,
+      "loss": 0.3596,
+      "step": 1749
+    },
+    {
+      "epoch": 1.421608448415922,
+      "grad_norm": 0.2204589545726776,
+      "learning_rate": 6.315010386144776e-06,
+      "loss": 0.3947,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4224207961007311,
+      "grad_norm": 0.22291171550750732,
+      "learning_rate": 6.310449170614446e-06,
+      "loss": 0.3727,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4232331437855401,
+      "grad_norm": 0.24020782113075256,
+      "learning_rate": 6.305886783808209e-06,
+      "loss": 0.3725,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4240454914703493,
+      "grad_norm": 0.19424057006835938,
+      "learning_rate": 6.301323229803913e-06,
+      "loss": 0.3643,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4248578391551585,
+      "grad_norm": 0.23812809586524963,
+      "learning_rate": 6.296758512680448e-06,
+      "loss": 0.3565,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4256701868399675,
+      "grad_norm": 0.22839517891407013,
+      "learning_rate": 6.292192636517744e-06,
+      "loss": 0.346,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4264825345247765,
+      "grad_norm": 0.21499398350715637,
+      "learning_rate": 6.287625605396774e-06,
+      "loss": 0.3635,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4272948822095857,
+      "grad_norm": 0.22419267892837524,
+      "learning_rate": 6.2830574233995356e-06,
+      "loss": 0.347,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4281072298943949,
+      "grad_norm": 0.2213687300682068,
+      "learning_rate": 6.278488094609057e-06,
+      "loss": 0.3874,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4289195775792038,
+      "grad_norm": 0.20119765400886536,
+      "learning_rate": 6.273917623109392e-06,
+      "loss": 0.3675,
+      "step": 1759
+    },
+    {
+      "epoch": 1.429731925264013,
+      "grad_norm": 0.21178485453128815,
+      "learning_rate": 6.269346012985617e-06,
+      "loss": 0.3725,
+      "step": 1760
+    },
+    {
+      "epoch": 1.430544272948822,
+      "grad_norm": 0.22064568102359772,
+      "learning_rate": 6.264773268323821e-06,
+      "loss": 0.371,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4313566206336312,
+      "grad_norm": 0.21204030513763428,
+      "learning_rate": 6.260199393211115e-06,
+      "loss": 0.3531,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4321689683184404,
+      "grad_norm": 0.23089207708835602,
+      "learning_rate": 6.255624391735614e-06,
+      "loss": 0.3668,
+      "step": 1763
+    },
+    {
+      "epoch": 1.4329813160032494,
+      "grad_norm": 0.21955296397209167,
+      "learning_rate": 6.2510482679864425e-06,
+      "loss": 0.3585,
+      "step": 1764
+    },
+    {
+      "epoch": 1.4337936636880584,
+      "grad_norm": 0.21400710940361023,
+      "learning_rate": 6.246471026053727e-06,
+      "loss": 0.3732,
+      "step": 1765
+    },
+    {
+      "epoch": 1.4346060113728676,
+      "grad_norm": 0.23580454289913177,
+      "learning_rate": 6.241892670028595e-06,
+      "loss": 0.3965,
+      "step": 1766
+    },
+    {
+      "epoch": 1.4354183590576768,
+      "grad_norm": 0.2152038812637329,
+      "learning_rate": 6.237313204003167e-06,
+      "loss": 0.3663,
+      "step": 1767
+    },
+    {
+      "epoch": 1.4362307067424858,
+      "grad_norm": 0.21602573990821838,
+      "learning_rate": 6.23273263207056e-06,
+      "loss": 0.3498,
+      "step": 1768
+    },
+    {
+      "epoch": 1.437043054427295,
+      "grad_norm": 0.2185019999742508,
+      "learning_rate": 6.228150958324872e-06,
+      "loss": 0.3391,
+      "step": 1769
+    },
+    {
+      "epoch": 1.437855402112104,
+      "grad_norm": 0.21839012205600739,
+      "learning_rate": 6.223568186861195e-06,
+      "loss": 0.3812,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4386677497969131,
+      "grad_norm": 0.19603215157985687,
+      "learning_rate": 6.218984321775596e-06,
+      "loss": 0.3696,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4394800974817221,
+      "grad_norm": 0.20903214812278748,
+      "learning_rate": 6.214399367165122e-06,
+      "loss": 0.3593,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4402924451665313,
+      "grad_norm": 0.19151747226715088,
+      "learning_rate": 6.209813327127792e-06,
+      "loss": 0.3481,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4411047928513403,
+      "grad_norm": 0.22467660903930664,
+      "learning_rate": 6.205226205762595e-06,
+      "loss": 0.3576,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4419171405361495,
+      "grad_norm": 0.21798183023929596,
+      "learning_rate": 6.20063800716949e-06,
+      "loss": 0.3703,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4427294882209587,
+      "grad_norm": 0.21683548390865326,
+      "learning_rate": 6.196048735449396e-06,
+      "loss": 0.36,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4435418359057677,
+      "grad_norm": 0.23905183374881744,
+      "learning_rate": 6.1914583947041906e-06,
+      "loss": 0.401,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4443541835905767,
+      "grad_norm": 0.2091539353132248,
+      "learning_rate": 6.186866989036708e-06,
+      "loss": 0.3769,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4451665312753859,
+      "grad_norm": 0.22964638471603394,
+      "learning_rate": 6.1822745225507355e-06,
+      "loss": 0.3768,
+      "step": 1779
+    },
+    {
+      "epoch": 1.445978878960195,
+      "grad_norm": 0.2302178144454956,
+      "learning_rate": 6.177680999351006e-06,
+      "loss": 0.3725,
+      "step": 1780
+    },
+    {
+      "epoch": 1.446791226645004,
+      "grad_norm": 0.2177625298500061,
+      "learning_rate": 6.173086423543198e-06,
+      "loss": 0.3861,
+      "step": 1781
+    },
+    {
+      "epoch": 1.4476035743298132,
+      "grad_norm": 0.23169207572937012,
+      "learning_rate": 6.168490799233931e-06,
+      "loss": 0.3628,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4484159220146222,
+      "grad_norm": 0.21119895577430725,
+      "learning_rate": 6.163894130530761e-06,
+      "loss": 0.3663,
+      "step": 1783
+    },
+    {
+      "epoch": 1.4492282696994314,
+      "grad_norm": 0.21805810928344727,
+      "learning_rate": 6.159296421542178e-06,
+      "loss": 0.3779,
+      "step": 1784
+    },
+    {
+      "epoch": 1.4500406173842404,
+      "grad_norm": 0.21920476853847504,
+      "learning_rate": 6.154697676377605e-06,
+      "loss": 0.3565,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4508529650690496,
+      "grad_norm": 0.22448565065860748,
+      "learning_rate": 6.150097899147384e-06,
+      "loss": 0.3625,
+      "step": 1786
+    },
+    {
+      "epoch": 1.4516653127538586,
+      "grad_norm": 0.22628174722194672,
+      "learning_rate": 6.1454970939627845e-06,
+      "loss": 0.3962,
+      "step": 1787
+    },
+    {
+      "epoch": 1.4524776604386678,
+      "grad_norm": 0.21023136377334595,
+      "learning_rate": 6.140895264935993e-06,
+      "loss": 0.377,
+      "step": 1788
+    },
+    {
+      "epoch": 1.453290008123477,
+      "grad_norm": 0.23886841535568237,
+      "learning_rate": 6.136292416180114e-06,
+      "loss": 0.3933,
+      "step": 1789
+    },
+    {
+      "epoch": 1.454102355808286,
+      "grad_norm": 0.22020167112350464,
+      "learning_rate": 6.131688551809161e-06,
+      "loss": 0.3548,
+      "step": 1790
+    },
+    {
+      "epoch": 1.454914703493095,
+      "grad_norm": 0.22509029507637024,
+      "learning_rate": 6.127083675938053e-06,
+      "loss": 0.3942,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4557270511779041,
+      "grad_norm": 0.20428264141082764,
+      "learning_rate": 6.122477792682616e-06,
+      "loss": 0.3665,
+      "step": 1792
+    },
+    {
+      "epoch": 1.4565393988627133,
+      "grad_norm": 0.22779758274555206,
+      "learning_rate": 6.117870906159577e-06,
+      "loss": 0.3844,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4573517465475223,
+      "grad_norm": 0.2249261885881424,
+      "learning_rate": 6.113263020486559e-06,
+      "loss": 0.3405,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4581640942323315,
+      "grad_norm": 0.19353903830051422,
+      "learning_rate": 6.1086541397820785e-06,
+      "loss": 0.3691,
+      "step": 1795
+    },
+    {
+      "epoch": 1.4589764419171405,
+      "grad_norm": 0.22219936549663544,
+      "learning_rate": 6.104044268165539e-06,
+      "loss": 0.3635,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4597887896019497,
+      "grad_norm": 0.21199151873588562,
+      "learning_rate": 6.0994334097572305e-06,
+      "loss": 0.3427,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4606011372867587,
+      "grad_norm": 0.23505902290344238,
+      "learning_rate": 6.094821568678328e-06,
+      "loss": 0.3655,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4614134849715679,
+      "grad_norm": 0.2172694206237793,
+      "learning_rate": 6.090208749050881e-06,
+      "loss": 0.3408,
+      "step": 1799
+    },
+    {
+      "epoch": 1.4622258326563768,
+      "grad_norm": 0.2226714938879013,
+      "learning_rate": 6.085594954997814e-06,
+      "loss": 0.356,
+      "step": 1800
+    },
+    {
+      "epoch": 1.463038180341186,
+      "grad_norm": 0.2276824712753296,
+      "learning_rate": 6.080980190642927e-06,
+      "loss": 0.3748,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4638505280259952,
+      "grad_norm": 0.2026449739933014,
+      "learning_rate": 6.076364460110881e-06,
+      "loss": 0.3718,
+      "step": 1802
+    },
+    {
+      "epoch": 1.4646628757108042,
+      "grad_norm": 0.20521801710128784,
+      "learning_rate": 6.071747767527204e-06,
+      "loss": 0.3718,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4654752233956132,
+      "grad_norm": 0.20978206396102905,
+      "learning_rate": 6.067130117018284e-06,
+      "loss": 0.3618,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4662875710804224,
+      "grad_norm": 0.2084031105041504,
+      "learning_rate": 6.062511512711362e-06,
+      "loss": 0.3597,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4670999187652316,
+      "grad_norm": 0.19140009582042694,
+      "learning_rate": 6.057891958734538e-06,
+      "loss": 0.3472,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4679122664500406,
+      "grad_norm": 0.21363341808319092,
+      "learning_rate": 6.053271459216755e-06,
+      "loss": 0.381,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4687246141348498,
+      "grad_norm": 0.19715535640716553,
+      "learning_rate": 6.048650018287803e-06,
+      "loss": 0.3543,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4695369618196588,
+      "grad_norm": 0.20029179751873016,
+      "learning_rate": 6.044027640078312e-06,
+      "loss": 0.3573,
+      "step": 1809
+    },
+    {
+      "epoch": 1.470349309504468,
+      "grad_norm": 0.2043846696615219,
+      "learning_rate": 6.0394043287197524e-06,
+      "loss": 0.3511,
+      "step": 1810
+    },
+    {
+      "epoch": 1.471161657189277,
+      "grad_norm": 0.20173698663711548,
+      "learning_rate": 6.034780088344427e-06,
+      "loss": 0.352,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4719740048740861,
+      "grad_norm": 0.19896847009658813,
+      "learning_rate": 6.030154923085469e-06,
+      "loss": 0.361,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4727863525588951,
+      "grad_norm": 0.23438048362731934,
+      "learning_rate": 6.025528837076841e-06,
+      "loss": 0.3801,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4735987002437043,
+      "grad_norm": 0.22181890904903412,
+      "learning_rate": 6.020901834453322e-06,
+      "loss": 0.361,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4744110479285135,
+      "grad_norm": 0.2254992127418518,
+      "learning_rate": 6.016273919350517e-06,
+      "loss": 0.3695,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4752233956133225,
+      "grad_norm": 0.20468659698963165,
+      "learning_rate": 6.011645095904845e-06,
+      "loss": 0.3643,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4760357432981315,
+      "grad_norm": 0.20557959377765656,
+      "learning_rate": 6.0070153682535315e-06,
+      "loss": 0.3682,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4768480909829407,
+      "grad_norm": 0.20261891186237335,
+      "learning_rate": 6.002384740534619e-06,
+      "loss": 0.3654,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4776604386677499,
+      "grad_norm": 0.1946525275707245,
+      "learning_rate": 5.997753216886948e-06,
+      "loss": 0.3989,
+      "step": 1819
+    },
+    {
+      "epoch": 1.4784727863525589,
+      "grad_norm": 0.21877935528755188,
+      "learning_rate": 5.993120801450162e-06,
+      "loss": 0.3553,
+      "step": 1820
+    },
+    {
+      "epoch": 1.479285134037368,
+      "grad_norm": 0.21097946166992188,
+      "learning_rate": 5.9884874983647e-06,
+      "loss": 0.3716,
+      "step": 1821
+    },
+    {
+      "epoch": 1.480097481722177,
+      "grad_norm": 0.19853568077087402,
+      "learning_rate": 5.983853311771799e-06,
+      "loss": 0.3512,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4809098294069862,
+      "grad_norm": 0.2122683972120285,
+      "learning_rate": 5.979218245813478e-06,
+      "loss": 0.3559,
+      "step": 1823
+    },
+    {
+      "epoch": 1.4817221770917952,
+      "grad_norm": 0.20788151025772095,
+      "learning_rate": 5.974582304632551e-06,
+      "loss": 0.3676,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4825345247766044,
+      "grad_norm": 0.19462397694587708,
+      "learning_rate": 5.969945492372606e-06,
+      "loss": 0.3595,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4833468724614134,
+      "grad_norm": 0.22324948012828827,
+      "learning_rate": 5.965307813178015e-06,
+      "loss": 0.4035,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4841592201462226,
+      "grad_norm": 0.2141384482383728,
+      "learning_rate": 5.960669271193922e-06,
+      "loss": 0.4034,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4849715678310318,
+      "grad_norm": 0.222634956240654,
+      "learning_rate": 5.9560298705662444e-06,
+      "loss": 0.3655,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4857839155158408,
+      "grad_norm": 0.18775755167007446,
+      "learning_rate": 5.951389615441666e-06,
+      "loss": 0.3716,
+      "step": 1829
+    },
+    {
+      "epoch": 1.4865962632006497,
+      "grad_norm": 0.2034786343574524,
+      "learning_rate": 5.9467485099676325e-06,
+      "loss": 0.3501,
+      "step": 1830
+    },
+    {
+      "epoch": 1.487408610885459,
+      "grad_norm": 0.20798316597938538,
+      "learning_rate": 5.942106558292352e-06,
+      "loss": 0.3803,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4882209585702681,
+      "grad_norm": 0.22644081711769104,
+      "learning_rate": 5.93746376456479e-06,
+      "loss": 0.3758,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4890333062550771,
+      "grad_norm": 0.19429653882980347,
+      "learning_rate": 5.932820132934661e-06,
+      "loss": 0.3379,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4898456539398863,
+      "grad_norm": 0.2251264899969101,
+      "learning_rate": 5.928175667552432e-06,
+      "loss": 0.3804,
+      "step": 1834
+    },
+    {
+      "epoch": 1.4906580016246953,
+      "grad_norm": 0.1874600052833557,
+      "learning_rate": 5.923530372569312e-06,
+      "loss": 0.3598,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4914703493095045,
+      "grad_norm": 0.22459742426872253,
+      "learning_rate": 5.9188842521372535e-06,
+      "loss": 0.3786,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4922826969943135,
+      "grad_norm": 0.19838471710681915,
+      "learning_rate": 5.914237310408947e-06,
+      "loss": 0.3532,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4930950446791227,
+      "grad_norm": 0.21161030232906342,
+      "learning_rate": 5.909589551537814e-06,
+      "loss": 0.3737,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4939073923639317,
+      "grad_norm": 0.2067354917526245,
+      "learning_rate": 5.904940979678012e-06,
+      "loss": 0.35,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4947197400487409,
+      "grad_norm": 0.20879192650318146,
+      "learning_rate": 5.90029159898442e-06,
+      "loss": 0.3414,
+      "step": 1840
+    },
+    {
+      "epoch": 1.49553208773355,
+      "grad_norm": 0.21208713948726654,
+      "learning_rate": 5.895641413612643e-06,
+      "loss": 0.3866,
+      "step": 1841
+    },
+    {
+      "epoch": 1.496344435418359,
+      "grad_norm": 0.20697559416294098,
+      "learning_rate": 5.890990427719003e-06,
+      "loss": 0.3394,
+      "step": 1842
+    },
+    {
+      "epoch": 1.497156783103168,
+      "grad_norm": 0.21676813066005707,
+      "learning_rate": 5.886338645460539e-06,
+      "loss": 0.3751,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4979691307879772,
+      "grad_norm": 0.22341342270374298,
+      "learning_rate": 5.881686070995001e-06,
+      "loss": 0.3791,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4987814784727864,
+      "grad_norm": 0.20767274498939514,
+      "learning_rate": 5.877032708480847e-06,
+      "loss": 0.3702,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4995938261575954,
+      "grad_norm": 0.21275372803211212,
+      "learning_rate": 5.872378562077241e-06,
+      "loss": 0.354,
+      "step": 1846
+    },
+    {
+      "epoch": 1.5004061738424046,
+      "grad_norm": 0.21069873869419098,
+      "learning_rate": 5.867723635944047e-06,
+      "loss": 0.3541,
+      "step": 1847
+    },
+    {
+      "epoch": 1.5012185215272136,
+      "grad_norm": 0.22311097383499146,
+      "learning_rate": 5.863067934241823e-06,
+      "loss": 0.3841,
+      "step": 1848
+    },
+    {
+      "epoch": 1.5020308692120228,
+      "grad_norm": 0.20990557968616486,
+      "learning_rate": 5.8584114611318225e-06,
+      "loss": 0.3517,
+      "step": 1849
+    },
+    {
+      "epoch": 1.502843216896832,
+      "grad_norm": 0.2012338489294052,
+      "learning_rate": 5.853754220775992e-06,
+      "loss": 0.3559,
+      "step": 1850
+    },
+    {
+      "epoch": 1.503655564581641,
+      "grad_norm": 0.20925575494766235,
+      "learning_rate": 5.849096217336956e-06,
+      "loss": 0.3547,
+      "step": 1851
+    },
+    {
+      "epoch": 1.50446791226645,
+      "grad_norm": 0.21246813237667084,
+      "learning_rate": 5.844437454978029e-06,
+      "loss": 0.3551,
+      "step": 1852
+    },
+    {
+      "epoch": 1.5052802599512591,
+      "grad_norm": 0.2141440063714981,
+      "learning_rate": 5.8397779378631955e-06,
+      "loss": 0.3735,
+      "step": 1853
+    },
+    {
+      "epoch": 1.5060926076360683,
+      "grad_norm": 0.2014346718788147,
+      "learning_rate": 5.835117670157126e-06,
+      "loss": 0.373,
+      "step": 1854
+    },
+    {
+      "epoch": 1.5069049553208773,
+      "grad_norm": 0.22243471443653107,
+      "learning_rate": 5.83045665602515e-06,
+      "loss": 0.3782,
+      "step": 1855
+    },
+    {
+      "epoch": 1.5077173030056863,
+      "grad_norm": 0.21789714694023132,
+      "learning_rate": 5.82579489963327e-06,
+      "loss": 0.365,
+      "step": 1856
+    },
+    {
+      "epoch": 1.5085296506904955,
+      "grad_norm": 0.21942900121212006,
+      "learning_rate": 5.8211324051481534e-06,
+      "loss": 0.3802,
+      "step": 1857
+    },
+    {
+      "epoch": 1.5093419983753047,
+      "grad_norm": 0.18887938559055328,
+      "learning_rate": 5.816469176737122e-06,
+      "loss": 0.3688,
+      "step": 1858
+    },
+    {
+      "epoch": 1.510154346060114,
+      "grad_norm": 0.20212872326374054,
+      "learning_rate": 5.8118052185681584e-06,
+      "loss": 0.3628,
+      "step": 1859
+    },
+    {
+      "epoch": 1.5109666937449229,
+      "grad_norm": 0.20805245637893677,
+      "learning_rate": 5.807140534809896e-06,
+      "loss": 0.353,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5117790414297319,
+      "grad_norm": 0.22409535944461823,
+      "learning_rate": 5.802475129631616e-06,
+      "loss": 0.3843,
+      "step": 1861
+    },
+    {
+      "epoch": 1.512591389114541,
+      "grad_norm": 0.2166099101305008,
+      "learning_rate": 5.797809007203245e-06,
+      "loss": 0.3547,
+      "step": 1862
+    },
+    {
+      "epoch": 1.5134037367993503,
+      "grad_norm": 0.20986557006835938,
+      "learning_rate": 5.79314217169535e-06,
+      "loss": 0.3584,
+      "step": 1863
+    },
+    {
+      "epoch": 1.5142160844841592,
+      "grad_norm": 0.20937256515026093,
+      "learning_rate": 5.788474627279137e-06,
+      "loss": 0.3706,
+      "step": 1864
+    },
+    {
+      "epoch": 1.5150284321689682,
+      "grad_norm": 0.21648110449314117,
+      "learning_rate": 5.783806378126443e-06,
+      "loss": 0.353,
+      "step": 1865
+    },
+    {
+      "epoch": 1.5158407798537774,
+      "grad_norm": 0.22735030949115753,
+      "learning_rate": 5.779137428409738e-06,
+      "loss": 0.353,
+      "step": 1866
+    },
+    {
+      "epoch": 1.5166531275385866,
+      "grad_norm": 0.21764878928661346,
+      "learning_rate": 5.774467782302115e-06,
+      "loss": 0.3488,
+      "step": 1867
+    },
+    {
+      "epoch": 1.5174654752233956,
+      "grad_norm": 0.21141499280929565,
+      "learning_rate": 5.7697974439772934e-06,
+      "loss": 0.3737,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5182778229082046,
+      "grad_norm": 0.2078760862350464,
+      "learning_rate": 5.765126417609606e-06,
+      "loss": 0.3508,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5190901705930138,
+      "grad_norm": 0.20966270565986633,
+      "learning_rate": 5.760454707374004e-06,
+      "loss": 0.3593,
+      "step": 1870
+    },
+    {
+      "epoch": 1.519902518277823,
+      "grad_norm": 0.21229758858680725,
+      "learning_rate": 5.755782317446051e-06,
+      "loss": 0.3704,
+      "step": 1871
+    },
+    {
+      "epoch": 1.5207148659626322,
+      "grad_norm": 0.19520069658756256,
+      "learning_rate": 5.751109252001916e-06,
+      "loss": 0.351,
+      "step": 1872
+    },
+    {
+      "epoch": 1.5215272136474411,
+      "grad_norm": 0.21193231642246246,
+      "learning_rate": 5.746435515218373e-06,
+      "loss": 0.373,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5223395613322501,
+      "grad_norm": 0.21480803191661835,
+      "learning_rate": 5.741761111272792e-06,
+      "loss": 0.3704,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5231519090170593,
+      "grad_norm": 0.21679390966892242,
+      "learning_rate": 5.737086044343145e-06,
+      "loss": 0.3732,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5239642567018685,
+      "grad_norm": 0.2435125857591629,
+      "learning_rate": 5.732410318607993e-06,
+      "loss": 0.3746,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5247766043866775,
+      "grad_norm": 0.21069671213626862,
+      "learning_rate": 5.72773393824649e-06,
+      "loss": 0.3701,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5255889520714865,
+      "grad_norm": 0.1959526687860489,
+      "learning_rate": 5.723056907438369e-06,
+      "loss": 0.3806,
+      "step": 1878
+    },
+    {
+      "epoch": 1.5264012997562957,
+      "grad_norm": 0.20878078043460846,
+      "learning_rate": 5.718379230363948e-06,
+      "loss": 0.3891,
+      "step": 1879
+    },
+    {
+      "epoch": 1.5272136474411049,
+      "grad_norm": 0.22115765511989594,
+      "learning_rate": 5.713700911204121e-06,
+      "loss": 0.3902,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5280259951259139,
+      "grad_norm": 0.19739648699760437,
+      "learning_rate": 5.709021954140359e-06,
+      "loss": 0.3582,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5288383428107228,
+      "grad_norm": 0.21069017052650452,
+      "learning_rate": 5.704342363354701e-06,
+      "loss": 0.3788,
+      "step": 1882
+    },
+    {
+      "epoch": 1.529650690495532,
+      "grad_norm": 0.202728271484375,
+      "learning_rate": 5.6996621430297514e-06,
+      "loss": 0.3858,
+      "step": 1883
+    },
+    {
+      "epoch": 1.5304630381803412,
+      "grad_norm": 0.20673729479312897,
+      "learning_rate": 5.6949812973486785e-06,
+      "loss": 0.3723,
+      "step": 1884
+    },
+    {
+      "epoch": 1.5312753858651504,
+      "grad_norm": 0.21876247227191925,
+      "learning_rate": 5.690299830495211e-06,
+      "loss": 0.3881,
+      "step": 1885
+    },
+    {
+      "epoch": 1.5320877335499594,
+      "grad_norm": 0.19667233526706696,
+      "learning_rate": 5.685617746653629e-06,
+      "loss": 0.3537,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5329000812347684,
+      "grad_norm": 0.19297103583812714,
+      "learning_rate": 5.680935050008767e-06,
+      "loss": 0.3765,
+      "step": 1887
+    },
+    {
+      "epoch": 1.5337124289195776,
+      "grad_norm": 0.20468205213546753,
+      "learning_rate": 5.676251744746008e-06,
+      "loss": 0.3568,
+      "step": 1888
+    },
+    {
+      "epoch": 1.5345247766043868,
+      "grad_norm": 0.20642681419849396,
+      "learning_rate": 5.671567835051274e-06,
+      "loss": 0.3655,
+      "step": 1889
+    },
+    {
+      "epoch": 1.5353371242891958,
+      "grad_norm": 0.220041885972023,
+      "learning_rate": 5.666883325111035e-06,
+      "loss": 0.3728,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5361494719740048,
+      "grad_norm": 0.2034541666507721,
+      "learning_rate": 5.662198219112291e-06,
+      "loss": 0.3883,
+      "step": 1891
+    },
+    {
+      "epoch": 1.536961819658814,
+      "grad_norm": 0.19788607954978943,
+      "learning_rate": 5.657512521242579e-06,
+      "loss": 0.3771,
+      "step": 1892
+    },
+    {
+      "epoch": 1.5377741673436232,
+      "grad_norm": 0.23450106382369995,
+      "learning_rate": 5.652826235689961e-06,
+      "loss": 0.3523,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5385865150284321,
+      "grad_norm": 0.21347558498382568,
+      "learning_rate": 5.648139366643026e-06,
+      "loss": 0.3855,
+      "step": 1894
+    },
+    {
+      "epoch": 1.5393988627132411,
+      "grad_norm": 0.2051730901002884,
+      "learning_rate": 5.643451918290885e-06,
+      "loss": 0.3633,
+      "step": 1895
+    },
+    {
+      "epoch": 1.5402112103980503,
+      "grad_norm": 0.21504707634449005,
+      "learning_rate": 5.638763894823169e-06,
+      "loss": 0.3806,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5410235580828595,
+      "grad_norm": 0.21568334102630615,
+      "learning_rate": 5.634075300430017e-06,
+      "loss": 0.3526,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5418359057676687,
+      "grad_norm": 0.20783135294914246,
+      "learning_rate": 5.6293861393020855e-06,
+      "loss": 0.3424,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5426482534524777,
+      "grad_norm": 0.207193523645401,
+      "learning_rate": 5.624696415630532e-06,
+      "loss": 0.3823,
+      "step": 1899
+    },
+    {
+      "epoch": 1.5434606011372867,
+      "grad_norm": 0.2268107831478119,
+      "learning_rate": 5.620006133607017e-06,
+      "loss": 0.3629,
+      "step": 1900
+    },
+    {
+      "epoch": 1.5442729488220959,
+      "grad_norm": 0.21138052642345428,
+      "learning_rate": 5.615315297423707e-06,
+      "loss": 0.3475,
+      "step": 1901
+    },
+    {
+      "epoch": 1.545085296506905,
+      "grad_norm": 0.19991956651210785,
+      "learning_rate": 5.610623911273254e-06,
+      "loss": 0.3648,
+      "step": 1902
+    },
+    {
+      "epoch": 1.545897644191714,
+      "grad_norm": 0.21667282283306122,
+      "learning_rate": 5.605931979348807e-06,
+      "loss": 0.3507,
+      "step": 1903
+    },
+    {
+      "epoch": 1.546709991876523,
+      "grad_norm": 0.21392609179019928,
+      "learning_rate": 5.601239505844005e-06,
+      "loss": 0.3573,
+      "step": 1904
+    },
+    {
+      "epoch": 1.5475223395613322,
+      "grad_norm": 0.22831600904464722,
+      "learning_rate": 5.596546494952965e-06,
+      "loss": 0.3467,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5483346872461414,
+      "grad_norm": 0.22945600748062134,
+      "learning_rate": 5.591852950870287e-06,
+      "loss": 0.3674,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5491470349309504,
+      "grad_norm": 0.2112620323896408,
+      "learning_rate": 5.58715887779105e-06,
+      "loss": 0.3618,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5499593826157594,
+      "grad_norm": 0.25337469577789307,
+      "learning_rate": 5.582464279910802e-06,
+      "loss": 0.3704,
+      "step": 1908
+    },
+    {
+      "epoch": 1.5507717303005686,
+      "grad_norm": 0.21739469468593597,
+      "learning_rate": 5.577769161425563e-06,
+      "loss": 0.3543,
+      "step": 1909
+    },
+    {
+      "epoch": 1.5515840779853778,
+      "grad_norm": 0.23328793048858643,
+      "learning_rate": 5.573073526531818e-06,
+      "loss": 0.3899,
+      "step": 1910
+    },
+    {
+      "epoch": 1.552396425670187,
+      "grad_norm": 0.23258273303508759,
+      "learning_rate": 5.568377379426511e-06,
+      "loss": 0.3652,
+      "step": 1911
+    },
+    {
+      "epoch": 1.553208773354996,
+      "grad_norm": 0.2019314020872116,
+      "learning_rate": 5.563680724307046e-06,
+      "loss": 0.3499,
+      "step": 1912
+    },
+    {
+      "epoch": 1.554021121039805,
+      "grad_norm": 0.2076883167028427,
+      "learning_rate": 5.558983565371281e-06,
+      "loss": 0.3769,
+      "step": 1913
+    },
+    {
+      "epoch": 1.5548334687246141,
+      "grad_norm": 0.24279169738292694,
+      "learning_rate": 5.554285906817524e-06,
+      "loss": 0.3755,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5556458164094233,
+      "grad_norm": 0.21853002905845642,
+      "learning_rate": 5.549587752844529e-06,
+      "loss": 0.3545,
+      "step": 1915
+    },
+    {
+      "epoch": 1.5564581640942323,
+      "grad_norm": 0.21231620013713837,
+      "learning_rate": 5.5448891076514925e-06,
+      "loss": 0.369,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5572705117790413,
+      "grad_norm": 0.21879728138446808,
+      "learning_rate": 5.540189975438053e-06,
+      "loss": 0.3577,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5580828594638505,
+      "grad_norm": 0.22006534039974213,
+      "learning_rate": 5.53549036040428e-06,
+      "loss": 0.393,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5588952071486597,
+      "grad_norm": 0.19018791615962982,
+      "learning_rate": 5.530790266750675e-06,
+      "loss": 0.3524,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5597075548334687,
+      "grad_norm": 0.22319787740707397,
+      "learning_rate": 5.526089698678174e-06,
+      "loss": 0.3966,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5605199025182777,
+      "grad_norm": 0.21308141946792603,
+      "learning_rate": 5.521388660388129e-06,
+      "loss": 0.378,
+      "step": 1921
+    },
+    {
+      "epoch": 1.5613322502030869,
+      "grad_norm": 0.2050178050994873,
+      "learning_rate": 5.516687156082317e-06,
+      "loss": 0.3605,
+      "step": 1922
+    },
+    {
+      "epoch": 1.562144597887896,
+      "grad_norm": 0.20580431818962097,
+      "learning_rate": 5.511985189962926e-06,
+      "loss": 0.3709,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5629569455727053,
+      "grad_norm": 0.20234569907188416,
+      "learning_rate": 5.507282766232565e-06,
+      "loss": 0.364,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5637692932575142,
+      "grad_norm": 0.2178141176700592,
+      "learning_rate": 5.502579889094244e-06,
+      "loss": 0.3792,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5645816409423232,
+      "grad_norm": 0.20122039318084717,
+      "learning_rate": 5.497876562751384e-06,
+      "loss": 0.3741,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5653939886271324,
+      "grad_norm": 0.20861156284809113,
+      "learning_rate": 5.493172791407803e-06,
+      "loss": 0.3584,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5662063363119416,
+      "grad_norm": 0.22253373265266418,
+      "learning_rate": 5.488468579267721e-06,
+      "loss": 0.38,
+      "step": 1928
+    },
+    {
+      "epoch": 1.5670186839967506,
+      "grad_norm": 0.21035031974315643,
+      "learning_rate": 5.483763930535752e-06,
+      "loss": 0.3692,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5678310316815596,
+      "grad_norm": 0.2321050763130188,
+      "learning_rate": 5.479058849416894e-06,
+      "loss": 0.3509,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5686433793663688,
+      "grad_norm": 0.19157031178474426,
+      "learning_rate": 5.474353340116538e-06,
+      "loss": 0.3544,
+      "step": 1931
+    },
+    {
+      "epoch": 1.569455727051178,
+      "grad_norm": 0.20730344951152802,
+      "learning_rate": 5.469647406840456e-06,
+      "loss": 0.3729,
+      "step": 1932
+    },
+    {
+      "epoch": 1.570268074735987,
+      "grad_norm": 0.22637911140918732,
+      "learning_rate": 5.464941053794797e-06,
+      "loss": 0.3636,
+      "step": 1933
+    },
+    {
+      "epoch": 1.5710804224207962,
+      "grad_norm": 0.21629253029823303,
+      "learning_rate": 5.460234285186087e-06,
+      "loss": 0.3632,
+      "step": 1934
+    },
+    {
+      "epoch": 1.5718927701056051,
+      "grad_norm": 0.19545549154281616,
+      "learning_rate": 5.455527105221222e-06,
+      "loss": 0.3914,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5727051177904143,
+      "grad_norm": 0.22614127397537231,
+      "learning_rate": 5.450819518107468e-06,
+      "loss": 0.391,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5735174654752235,
+      "grad_norm": 0.21225355565547943,
+      "learning_rate": 5.446111528052455e-06,
+      "loss": 0.3591,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5743298131600325,
+      "grad_norm": 0.2212030589580536,
+      "learning_rate": 5.441403139264169e-06,
+      "loss": 0.3428,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5751421608448415,
+      "grad_norm": 0.21320189535617828,
+      "learning_rate": 5.436694355950955e-06,
+      "loss": 0.3839,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5759545085296507,
+      "grad_norm": 0.22996681928634644,
+      "learning_rate": 5.431985182321513e-06,
+      "loss": 0.3675,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5767668562144599,
+      "grad_norm": 0.2970804274082184,
+      "learning_rate": 5.427275622584888e-06,
+      "loss": 0.3878,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5775792038992689,
+      "grad_norm": 0.23101255297660828,
+      "learning_rate": 5.422565680950472e-06,
+      "loss": 0.3618,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5783915515840778,
+      "grad_norm": 0.2025032490491867,
+      "learning_rate": 5.417855361627998e-06,
+      "loss": 0.3663,
+      "step": 1943
+    },
+    {
+      "epoch": 1.579203899268887,
+      "grad_norm": 0.21822071075439453,
+      "learning_rate": 5.413144668827538e-06,
+      "loss": 0.359,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5800162469536962,
+      "grad_norm": 0.22706273198127747,
+      "learning_rate": 5.4084336067594945e-06,
+      "loss": 0.3458,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5808285946385054,
+      "grad_norm": 0.20122767984867096,
+      "learning_rate": 5.403722179634602e-06,
+      "loss": 0.3646,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5816409423233144,
+      "grad_norm": 0.22411894798278809,
+      "learning_rate": 5.399010391663923e-06,
+      "loss": 0.362,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5824532900081234,
+      "grad_norm": 0.22113461792469025,
+      "learning_rate": 5.3942982470588386e-06,
+      "loss": 0.3892,
+      "step": 1948
+    },
+    {
+      "epoch": 1.5832656376929326,
+      "grad_norm": 0.20014901459217072,
+      "learning_rate": 5.389585750031053e-06,
+      "loss": 0.3719,
+      "step": 1949
+    },
+    {
+      "epoch": 1.5840779853777418,
+      "grad_norm": 0.20549479126930237,
+      "learning_rate": 5.384872904792583e-06,
+      "loss": 0.3816,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5848903330625508,
+      "grad_norm": 0.44480717182159424,
+      "learning_rate": 5.380159715555755e-06,
+      "loss": 0.3743,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5857026807473598,
+      "grad_norm": 0.2574334144592285,
+      "learning_rate": 5.375446186533206e-06,
+      "loss": 0.3646,
+      "step": 1952
+    },
+    {
+      "epoch": 1.586515028432169,
+      "grad_norm": 0.21341513097286224,
+      "learning_rate": 5.370732321937876e-06,
+      "loss": 0.3642,
+      "step": 1953
+    },
+    {
+      "epoch": 1.5873273761169782,
+      "grad_norm": 0.2170240432024002,
+      "learning_rate": 5.3660181259830024e-06,
+      "loss": 0.3598,
+      "step": 1954
+    },
+    {
+      "epoch": 1.5881397238017871,
+      "grad_norm": 0.21053136885166168,
+      "learning_rate": 5.361303602882123e-06,
+      "loss": 0.3535,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5889520714865961,
+      "grad_norm": 0.20790649950504303,
+      "learning_rate": 5.356588756849065e-06,
+      "loss": 0.3715,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5897644191714053,
+      "grad_norm": 0.2106579691171646,
+      "learning_rate": 5.3518735920979436e-06,
+      "loss": 0.3593,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5905767668562145,
+      "grad_norm": 0.21671061217784882,
+      "learning_rate": 5.347158112843163e-06,
+      "loss": 0.3751,
+      "step": 1958
+    },
+    {
+      "epoch": 1.5913891145410237,
+      "grad_norm": 0.19723957777023315,
+      "learning_rate": 5.342442323299403e-06,
+      "loss": 0.3482,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5922014622258327,
+      "grad_norm": 0.2107599824666977,
+      "learning_rate": 5.3377262276816245e-06,
+      "loss": 0.3626,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5930138099106417,
+      "grad_norm": 0.1985115110874176,
+      "learning_rate": 5.3330098302050615e-06,
+      "loss": 0.3748,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5938261575954509,
+      "grad_norm": 0.20472393929958344,
+      "learning_rate": 5.328293135085215e-06,
+      "loss": 0.3676,
+      "step": 1962
+    },
+    {
+      "epoch": 1.59463850528026,
+      "grad_norm": 0.19153429567813873,
+      "learning_rate": 5.323576146537858e-06,
+      "loss": 0.3431,
+      "step": 1963
+    },
+    {
+      "epoch": 1.595450852965069,
+      "grad_norm": 0.2078780084848404,
+      "learning_rate": 5.31885886877902e-06,
+      "loss": 0.36,
+      "step": 1964
+    },
+    {
+      "epoch": 1.596263200649878,
+      "grad_norm": 0.20471395552158356,
+      "learning_rate": 5.314141306024991e-06,
+      "loss": 0.3676,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5970755483346872,
+      "grad_norm": 0.2085876315832138,
+      "learning_rate": 5.309423462492314e-06,
+      "loss": 0.3601,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5978878960194964,
+      "grad_norm": 0.22403091192245483,
+      "learning_rate": 5.30470534239779e-06,
+      "loss": 0.37,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5987002437043054,
+      "grad_norm": 0.20715373754501343,
+      "learning_rate": 5.299986949958457e-06,
+      "loss": 0.3494,
+      "step": 1968
+    },
+    {
+      "epoch": 1.5995125913891144,
+      "grad_norm": 0.207205131649971,
+      "learning_rate": 5.295268289391603e-06,
+      "loss": 0.356,
+      "step": 1969
+    },
+    {
+      "epoch": 1.6003249390739236,
+      "grad_norm": 0.27880996465682983,
+      "learning_rate": 5.290549364914754e-06,
+      "loss": 0.3693,
+      "step": 1970
+    },
+    {
+      "epoch": 1.6011372867587328,
+      "grad_norm": 0.2372736632823944,
+      "learning_rate": 5.285830180745673e-06,
+      "loss": 0.394,
+      "step": 1971
+    },
+    {
+      "epoch": 1.601949634443542,
+      "grad_norm": 0.2087928056716919,
+      "learning_rate": 5.281110741102352e-06,
+      "loss": 0.3739,
+      "step": 1972
+    },
+    {
+      "epoch": 1.602761982128351,
+      "grad_norm": 0.1995851844549179,
+      "learning_rate": 5.276391050203015e-06,
+      "loss": 0.3595,
+      "step": 1973
+    },
+    {
+      "epoch": 1.60357432981316,
+      "grad_norm": 0.2056497484445572,
+      "learning_rate": 5.2716711122661065e-06,
+      "loss": 0.358,
+      "step": 1974
+    },
+    {
+      "epoch": 1.6043866774979691,
+      "grad_norm": 0.19597133994102478,
+      "learning_rate": 5.266950931510296e-06,
+      "loss": 0.3532,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6051990251827783,
+      "grad_norm": 0.21456460654735565,
+      "learning_rate": 5.262230512154469e-06,
+      "loss": 0.378,
+      "step": 1976
+    },
+    {
+      "epoch": 1.6060113728675873,
+      "grad_norm": 0.23084308207035065,
+      "learning_rate": 5.257509858417723e-06,
+      "loss": 0.3579,
+      "step": 1977
+    },
+    {
+      "epoch": 1.6068237205523963,
+      "grad_norm": 0.21318721771240234,
+      "learning_rate": 5.252788974519364e-06,
+      "loss": 0.37,
+      "step": 1978
+    },
+    {
+      "epoch": 1.6076360682372055,
+      "grad_norm": 0.20079384744167328,
+      "learning_rate": 5.248067864678906e-06,
+      "loss": 0.3771,
+      "step": 1979
+    },
+    {
+      "epoch": 1.6084484159220147,
+      "grad_norm": 0.19270260632038116,
+      "learning_rate": 5.243346533116065e-06,
+      "loss": 0.3619,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6092607636068237,
+      "grad_norm": 0.22233016788959503,
+      "learning_rate": 5.238624984050754e-06,
+      "loss": 0.3727,
+      "step": 1981
+    },
+    {
+      "epoch": 1.6100731112916327,
+      "grad_norm": 0.21705234050750732,
+      "learning_rate": 5.233903221703079e-06,
+      "loss": 0.359,
+      "step": 1982
+    },
+    {
+      "epoch": 1.6108854589764419,
+      "grad_norm": 0.21018142998218536,
+      "learning_rate": 5.229181250293341e-06,
+      "loss": 0.3777,
+      "step": 1983
+    },
+    {
+      "epoch": 1.611697806661251,
+      "grad_norm": 0.2114810049533844,
+      "learning_rate": 5.224459074042023e-06,
+      "loss": 0.3521,
+      "step": 1984
+    },
+    {
+      "epoch": 1.6125101543460603,
+      "grad_norm": 0.22905325889587402,
+      "learning_rate": 5.219736697169795e-06,
+      "loss": 0.3795,
+      "step": 1985
+    },
+    {
+      "epoch": 1.6133225020308692,
+      "grad_norm": 0.22420945763587952,
+      "learning_rate": 5.215014123897504e-06,
+      "loss": 0.3647,
+      "step": 1986
+    },
+    {
+      "epoch": 1.6141348497156782,
+      "grad_norm": 0.20017196238040924,
+      "learning_rate": 5.210291358446173e-06,
+      "loss": 0.3598,
+      "step": 1987
+    },
+    {
+      "epoch": 1.6149471974004874,
+      "grad_norm": 0.20282283425331116,
+      "learning_rate": 5.205568405036996e-06,
+      "loss": 0.3473,
+      "step": 1988
+    },
+    {
+      "epoch": 1.6157595450852966,
+      "grad_norm": 0.22572042047977448,
+      "learning_rate": 5.200845267891338e-06,
+      "loss": 0.3656,
+      "step": 1989
+    },
+    {
+      "epoch": 1.6165718927701056,
+      "grad_norm": 0.2448638677597046,
+      "learning_rate": 5.196121951230726e-06,
+      "loss": 0.3736,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6173842404549146,
+      "grad_norm": 0.20216621458530426,
+      "learning_rate": 5.1913984592768455e-06,
+      "loss": 0.3465,
+      "step": 1991
+    },
+    {
+      "epoch": 1.6181965881397238,
+      "grad_norm": 0.1830821931362152,
+      "learning_rate": 5.186674796251543e-06,
+      "loss": 0.3651,
+      "step": 1992
+    },
+    {
+      "epoch": 1.619008935824533,
+      "grad_norm": 0.2045586109161377,
+      "learning_rate": 5.181950966376814e-06,
+      "loss": 0.3412,
+      "step": 1993
+    },
+    {
+      "epoch": 1.619821283509342,
+      "grad_norm": 0.21435225009918213,
+      "learning_rate": 5.177226973874805e-06,
+      "loss": 0.3578,
+      "step": 1994
+    },
+    {
+      "epoch": 1.620633631194151,
+      "grad_norm": 0.19829899072647095,
+      "learning_rate": 5.172502822967808e-06,
+      "loss": 0.3582,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6214459788789601,
+      "grad_norm": 0.18878455460071564,
+      "learning_rate": 5.167778517878258e-06,
+      "loss": 0.3464,
+      "step": 1996
+    },
+    {
+      "epoch": 1.6222583265637693,
+      "grad_norm": 0.20314450562000275,
+      "learning_rate": 5.163054062828724e-06,
+      "loss": 0.359,
+      "step": 1997
+    },
+    {
+      "epoch": 1.6230706742485785,
+      "grad_norm": 0.22078199684619904,
+      "learning_rate": 5.158329462041911e-06,
+      "loss": 0.379,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6238830219333875,
+      "grad_norm": 0.19584254920482635,
+      "learning_rate": 5.153604719740655e-06,
+      "loss": 0.3543,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6246953696181965,
+      "grad_norm": 0.199188232421875,
+      "learning_rate": 5.148879840147918e-06,
+      "loss": 0.3812,
+      "step": 2000
+    },
+    {
+      "epoch": 1.6255077173030057,
+      "grad_norm": 0.22061820328235626,
+      "learning_rate": 5.1441548274867834e-06,
+      "loss": 0.3753,
+      "step": 2001
+    },
+    {
+      "epoch": 1.626320064987815,
+      "grad_norm": 0.1806270033121109,
+      "learning_rate": 5.139429685980457e-06,
+      "loss": 0.3501,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6271324126726239,
+      "grad_norm": 0.19546157121658325,
+      "learning_rate": 5.134704419852253e-06,
+      "loss": 0.3771,
+      "step": 2003
+    },
+    {
+      "epoch": 1.6279447603574329,
+      "grad_norm": 0.1982298642396927,
+      "learning_rate": 5.129979033325608e-06,
+      "loss": 0.3642,
+      "step": 2004
+    },
+    {
+      "epoch": 1.628757108042242,
+      "grad_norm": 0.23276983201503754,
+      "learning_rate": 5.125253530624054e-06,
+      "loss": 0.3823,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6295694557270513,
+      "grad_norm": 0.2189137041568756,
+      "learning_rate": 5.120527915971235e-06,
+      "loss": 0.3763,
+      "step": 2006
+    },
+    {
+      "epoch": 1.6303818034118602,
+      "grad_norm": 0.25944945216178894,
+      "learning_rate": 5.115802193590893e-06,
+      "loss": 0.3883,
+      "step": 2007
+    },
+    {
+      "epoch": 1.6311941510966692,
+      "grad_norm": 0.21448108553886414,
+      "learning_rate": 5.111076367706864e-06,
+      "loss": 0.3766,
+      "step": 2008
+    },
+    {
+      "epoch": 1.6320064987814784,
+      "grad_norm": 0.1974007934331894,
+      "learning_rate": 5.10635044254308e-06,
+      "loss": 0.3532,
+      "step": 2009
+    },
+    {
+      "epoch": 1.6328188464662876,
+      "grad_norm": 0.21336163580417633,
+      "learning_rate": 5.1016244223235576e-06,
+      "loss": 0.3646,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6336311941510968,
+      "grad_norm": 0.23709414899349213,
+      "learning_rate": 5.096898311272404e-06,
+      "loss": 0.3646,
+      "step": 2011
+    },
+    {
+      "epoch": 1.6344435418359058,
+      "grad_norm": 0.20962148904800415,
+      "learning_rate": 5.0921721136138025e-06,
+      "loss": 0.3889,
+      "step": 2012
+    },
+    {
+      "epoch": 1.6352558895207148,
+      "grad_norm": 0.1968088001012802,
+      "learning_rate": 5.087445833572017e-06,
+      "loss": 0.3603,
+      "step": 2013
+    },
+    {
+      "epoch": 1.636068237205524,
+      "grad_norm": 0.20752999186515808,
+      "learning_rate": 5.082719475371382e-06,
+      "loss": 0.3583,
+      "step": 2014
+    },
+    {
+      "epoch": 1.6368805848903332,
+      "grad_norm": 0.2408829778432846,
+      "learning_rate": 5.077993043236306e-06,
+      "loss": 0.3522,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6376929325751421,
+      "grad_norm": 0.2068435400724411,
+      "learning_rate": 5.073266541391259e-06,
+      "loss": 0.3358,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6385052802599511,
+      "grad_norm": 0.20575673878192902,
+      "learning_rate": 5.068539974060776e-06,
+      "loss": 0.394,
+      "step": 2017
+    },
+    {
+      "epoch": 1.6393176279447603,
+      "grad_norm": 0.21019533276557922,
+      "learning_rate": 5.06381334546945e-06,
+      "loss": 0.3606,
+      "step": 2018
+    },
+    {
+      "epoch": 1.6401299756295695,
+      "grad_norm": 0.21029528975486755,
+      "learning_rate": 5.05908665984193e-06,
+      "loss": 0.3638,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6409423233143785,
+      "grad_norm": 0.21318206191062927,
+      "learning_rate": 5.054359921402914e-06,
+      "loss": 0.3413,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6417546709991877,
+      "grad_norm": 0.2192971110343933,
+      "learning_rate": 5.049633134377147e-06,
+      "loss": 0.3739,
+      "step": 2021
+    },
+    {
+      "epoch": 1.6425670186839967,
+      "grad_norm": 0.3250647485256195,
+      "learning_rate": 5.04490630298942e-06,
+      "loss": 0.3519,
+      "step": 2022
+    },
+    {
+      "epoch": 1.6433793663688059,
+      "grad_norm": 0.21779467165470123,
+      "learning_rate": 5.04017943146456e-06,
+      "loss": 0.3591,
+      "step": 2023
+    },
+    {
+      "epoch": 1.644191714053615,
+      "grad_norm": 0.1847766488790512,
+      "learning_rate": 5.035452524027436e-06,
+      "loss": 0.3512,
+      "step": 2024
+    },
+    {
+      "epoch": 1.645004061738424,
+      "grad_norm": 0.20778779685497284,
+      "learning_rate": 5.03072558490294e-06,
+      "loss": 0.3634,
+      "step": 2025
+    },
+    {
+      "epoch": 1.645816409423233,
+      "grad_norm": 0.22379712760448456,
+      "learning_rate": 5.0259986183160006e-06,
+      "loss": 0.379,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6466287571080422,
+      "grad_norm": 0.20117464661598206,
+      "learning_rate": 5.021271628491566e-06,
+      "loss": 0.3481,
+      "step": 2027
+    },
+    {
+      "epoch": 1.6474411047928514,
+      "grad_norm": 0.2075277715921402,
+      "learning_rate": 5.016544619654609e-06,
+      "loss": 0.3566,
+      "step": 2028
+    },
+    {
+      "epoch": 1.6482534524776604,
+      "grad_norm": 0.20679563283920288,
+      "learning_rate": 5.011817596030115e-06,
+      "loss": 0.3683,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6490658001624694,
+      "grad_norm": 0.20642119646072388,
+      "learning_rate": 5.0070905618430865e-06,
+      "loss": 0.377,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6498781478472786,
+      "grad_norm": 0.19743715226650238,
+      "learning_rate": 5.0023635213185325e-06,
+      "loss": 0.3754,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6506904955320878,
+      "grad_norm": 0.19053667783737183,
+      "learning_rate": 4.997636478681468e-06,
+      "loss": 0.3788,
+      "step": 2032
+    },
+    {
+      "epoch": 1.6515028432168968,
+      "grad_norm": 0.22911037504673004,
+      "learning_rate": 4.992909438156915e-06,
+      "loss": 0.3613,
+      "step": 2033
+    },
+    {
+      "epoch": 1.652315190901706,
+      "grad_norm": 0.21842557191848755,
+      "learning_rate": 4.988182403969886e-06,
+      "loss": 0.3558,
+      "step": 2034
+    },
+    {
+      "epoch": 1.653127538586515,
+      "grad_norm": 0.2066747546195984,
+      "learning_rate": 4.983455380345392e-06,
+      "loss": 0.3793,
+      "step": 2035
+    },
+    {
+      "epoch": 1.6539398862713242,
+      "grad_norm": 0.19180165231227875,
+      "learning_rate": 4.978728371508433e-06,
+      "loss": 0.3384,
+      "step": 2036
+    },
+    {
+      "epoch": 1.6547522339561334,
+      "grad_norm": 0.18281777203083038,
+      "learning_rate": 4.974001381684e-06,
+      "loss": 0.3648,
+      "step": 2037
+    },
+    {
+      "epoch": 1.6555645816409423,
+      "grad_norm": 0.22709514200687408,
+      "learning_rate": 4.96927441509706e-06,
+      "loss": 0.3656,
+      "step": 2038
+    },
+    {
+      "epoch": 1.6563769293257513,
+      "grad_norm": 0.19342143833637238,
+      "learning_rate": 4.964547475972565e-06,
+      "loss": 0.3637,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6571892770105605,
+      "grad_norm": 0.20271103084087372,
+      "learning_rate": 4.9598205685354404e-06,
+      "loss": 0.3764,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6580016246953697,
+      "grad_norm": 0.20335036516189575,
+      "learning_rate": 4.955093697010581e-06,
+      "loss": 0.3788,
+      "step": 2041
+    },
+    {
+      "epoch": 1.6588139723801787,
+      "grad_norm": 0.19768527150154114,
+      "learning_rate": 4.950366865622854e-06,
+      "loss": 0.3565,
+      "step": 2042
+    },
+    {
+      "epoch": 1.6596263200649877,
+      "grad_norm": 0.21353186666965485,
+      "learning_rate": 4.9456400785970875e-06,
+      "loss": 0.3845,
+      "step": 2043
+    },
+    {
+      "epoch": 1.6604386677497969,
+      "grad_norm": 0.21100345253944397,
+      "learning_rate": 4.940913340158072e-06,
+      "loss": 0.3833,
+      "step": 2044
+    },
+    {
+      "epoch": 1.661251015434606,
+      "grad_norm": 0.20696090161800385,
+      "learning_rate": 4.936186654530551e-06,
+      "loss": 0.3783,
+      "step": 2045
+    },
+    {
+      "epoch": 1.6620633631194153,
+      "grad_norm": 0.2021302580833435,
+      "learning_rate": 4.931460025939226e-06,
+      "loss": 0.3603,
+      "step": 2046
+    },
+    {
+      "epoch": 1.6628757108042242,
+      "grad_norm": 0.22418685257434845,
+      "learning_rate": 4.926733458608744e-06,
+      "loss": 0.3884,
+      "step": 2047
+    },
+    {
+      "epoch": 1.6636880584890332,
+      "grad_norm": 0.18940535187721252,
+      "learning_rate": 4.922006956763697e-06,
+      "loss": 0.3491,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6645004061738424,
+      "grad_norm": 0.20452234148979187,
+      "learning_rate": 4.917280524628619e-06,
+      "loss": 0.3702,
+      "step": 2049
+    },
+    {
+      "epoch": 1.6653127538586516,
+      "grad_norm": 0.19087113440036774,
+      "learning_rate": 4.912554166427985e-06,
+      "loss": 0.3627,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6661251015434606,
+      "grad_norm": 0.21248455345630646,
+      "learning_rate": 4.907827886386199e-06,
+      "loss": 0.3656,
+      "step": 2051
+    },
+    {
+      "epoch": 1.6669374492282696,
+      "grad_norm": 0.20654655992984772,
+      "learning_rate": 4.903101688727598e-06,
+      "loss": 0.3403,
+      "step": 2052
+    },
+    {
+      "epoch": 1.6677497969130788,
+      "grad_norm": 0.21557243168354034,
+      "learning_rate": 4.898375577676444e-06,
+      "loss": 0.3854,
+      "step": 2053
+    },
+    {
+      "epoch": 1.668562144597888,
+      "grad_norm": 0.19879405200481415,
+      "learning_rate": 4.893649557456922e-06,
+      "loss": 0.3372,
+      "step": 2054
+    },
+    {
+      "epoch": 1.669374492282697,
+      "grad_norm": 0.20041202008724213,
+      "learning_rate": 4.888923632293137e-06,
+      "loss": 0.3376,
+      "step": 2055
+    },
+    {
+      "epoch": 1.670186839967506,
+      "grad_norm": 0.19786328077316284,
+      "learning_rate": 4.884197806409109e-06,
+      "loss": 0.3462,
+      "step": 2056
+    },
+    {
+      "epoch": 1.6709991876523151,
+      "grad_norm": 0.22934700548648834,
+      "learning_rate": 4.8794720840287665e-06,
+      "loss": 0.3679,
+      "step": 2057
+    },
+    {
+      "epoch": 1.6718115353371243,
+      "grad_norm": 0.21463751792907715,
+      "learning_rate": 4.874746469375947e-06,
+      "loss": 0.3842,
+      "step": 2058
+    },
+    {
+      "epoch": 1.6726238830219335,
+      "grad_norm": 0.21204409003257751,
+      "learning_rate": 4.870020966674394e-06,
+      "loss": 0.3611,
+      "step": 2059
+    },
+    {
+      "epoch": 1.6734362307067425,
+      "grad_norm": 0.21121002733707428,
+      "learning_rate": 4.865295580147747e-06,
+      "loss": 0.3667,
+      "step": 2060
+    },
+    {
+      "epoch": 1.6742485783915515,
+      "grad_norm": 0.20097380876541138,
+      "learning_rate": 4.860570314019546e-06,
+      "loss": 0.3566,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6750609260763607,
+      "grad_norm": 0.19698511064052582,
+      "learning_rate": 4.855845172513219e-06,
+      "loss": 0.359,
+      "step": 2062
+    },
+    {
+      "epoch": 1.67587327376117,
+      "grad_norm": 0.22207877039909363,
+      "learning_rate": 4.851120159852085e-06,
+      "loss": 0.3479,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6766856214459789,
+      "grad_norm": 0.2072446048259735,
+      "learning_rate": 4.846395280259348e-06,
+      "loss": 0.345,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6774979691307879,
+      "grad_norm": 0.21473576128482819,
+      "learning_rate": 4.841670537958092e-06,
+      "loss": 0.362,
+      "step": 2065
+    },
+    {
+      "epoch": 1.678310316815597,
+      "grad_norm": 0.22624698281288147,
+      "learning_rate": 4.836945937171279e-06,
+      "loss": 0.3836,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6791226645004063,
+      "grad_norm": 0.2040955126285553,
+      "learning_rate": 4.832221482121745e-06,
+      "loss": 0.3668,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6799350121852152,
+      "grad_norm": 0.22459113597869873,
+      "learning_rate": 4.827497177032193e-06,
+      "loss": 0.3756,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6807473598700242,
+      "grad_norm": 0.21885062754154205,
+      "learning_rate": 4.822773026125197e-06,
+      "loss": 0.3482,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6815597075548334,
+      "grad_norm": 0.21381404995918274,
+      "learning_rate": 4.8180490336231865e-06,
+      "loss": 0.3783,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6823720552396426,
+      "grad_norm": 0.2126740664243698,
+      "learning_rate": 4.813325203748458e-06,
+      "loss": 0.3492,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6831844029244518,
+      "grad_norm": 0.22239290177822113,
+      "learning_rate": 4.8086015407231545e-06,
+      "loss": 0.3712,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6839967506092608,
+      "grad_norm": 0.20915022492408752,
+      "learning_rate": 4.803878048769275e-06,
+      "loss": 0.3864,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6848090982940698,
+      "grad_norm": 0.2134694904088974,
+      "learning_rate": 4.799154732108662e-06,
+      "loss": 0.3621,
+      "step": 2074
+    },
+    {
+      "epoch": 1.685621445978879,
+      "grad_norm": 0.20330357551574707,
+      "learning_rate": 4.794431594963004e-06,
+      "loss": 0.348,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6864337936636882,
+      "grad_norm": 0.19707462191581726,
+      "learning_rate": 4.789708641553828e-06,
+      "loss": 0.3762,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6872461413484972,
+      "grad_norm": 0.21146520972251892,
+      "learning_rate": 4.784985876102498e-06,
+      "loss": 0.3662,
+      "step": 2077
+    },
+    {
+      "epoch": 1.6880584890333061,
+      "grad_norm": 0.2110634297132492,
+      "learning_rate": 4.7802633028302065e-06,
+      "loss": 0.3603,
+      "step": 2078
+    },
+    {
+      "epoch": 1.6888708367181153,
+      "grad_norm": 0.21471168100833893,
+      "learning_rate": 4.7755409259579786e-06,
+      "loss": 0.3407,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6896831844029245,
+      "grad_norm": 0.19306232035160065,
+      "learning_rate": 4.7708187497066595e-06,
+      "loss": 0.355,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6904955320877335,
+      "grad_norm": 0.21591685712337494,
+      "learning_rate": 4.766096778296922e-06,
+      "loss": 0.3833,
+      "step": 2081
+    },
+    {
+      "epoch": 1.6913078797725425,
+      "grad_norm": 0.21082915365695953,
+      "learning_rate": 4.761375015949248e-06,
+      "loss": 0.3594,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6921202274573517,
+      "grad_norm": 0.22915533185005188,
+      "learning_rate": 4.756653466883937e-06,
+      "loss": 0.3543,
+      "step": 2083
+    },
+    {
+      "epoch": 1.692932575142161,
+      "grad_norm": 0.21746471524238586,
+      "learning_rate": 4.751932135321095e-06,
+      "loss": 0.3386,
+      "step": 2084
+    },
+    {
+      "epoch": 1.69374492282697,
+      "grad_norm": 0.2531956434249878,
+      "learning_rate": 4.7472110254806375e-06,
+      "loss": 0.3751,
+      "step": 2085
+    },
+    {
+      "epoch": 1.694557270511779,
+      "grad_norm": 0.21272046864032745,
+      "learning_rate": 4.742490141582279e-06,
+      "loss": 0.3724,
+      "step": 2086
+    },
+    {
+      "epoch": 1.695369618196588,
+      "grad_norm": 0.21791671216487885,
+      "learning_rate": 4.737769487845532e-06,
+      "loss": 0.4007,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6961819658813972,
+      "grad_norm": 0.1991363912820816,
+      "learning_rate": 4.733049068489704e-06,
+      "loss": 0.3553,
+      "step": 2088
+    },
+    {
+      "epoch": 1.6969943135662064,
+      "grad_norm": 0.20660898089408875,
+      "learning_rate": 4.728328887733894e-06,
+      "loss": 0.3743,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6978066612510154,
+      "grad_norm": 0.19059039652347565,
+      "learning_rate": 4.723608949796987e-06,
+      "loss": 0.3641,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6986190089358244,
+      "grad_norm": 0.20394505560398102,
+      "learning_rate": 4.7188892588976496e-06,
+      "loss": 0.361,
+      "step": 2091
+    },
+    {
+      "epoch": 1.6994313566206336,
+      "grad_norm": 0.21822911500930786,
+      "learning_rate": 4.714169819254328e-06,
+      "loss": 0.3991,
+      "step": 2092
+    },
+    {
+      "epoch": 1.7002437043054428,
+      "grad_norm": 0.2138713151216507,
+      "learning_rate": 4.709450635085246e-06,
+      "loss": 0.3531,
+      "step": 2093
+    },
+    {
+      "epoch": 1.7010560519902518,
+      "grad_norm": 0.22133687138557434,
+      "learning_rate": 4.704731710608398e-06,
+      "loss": 0.3699,
+      "step": 2094
+    },
+    {
+      "epoch": 1.7018683996750608,
+      "grad_norm": 0.21057431399822235,
+      "learning_rate": 4.700013050041544e-06,
+      "loss": 0.3898,
+      "step": 2095
+    },
+    {
+      "epoch": 1.70268074735987,
+      "grad_norm": 0.19625353813171387,
+      "learning_rate": 4.695294657602212e-06,
+      "loss": 0.3464,
+      "step": 2096
+    },
+    {
+      "epoch": 1.7034930950446792,
+      "grad_norm": 0.2128133326768875,
+      "learning_rate": 4.6905765375076865e-06,
+      "loss": 0.3671,
+      "step": 2097
+    },
+    {
+      "epoch": 1.7043054427294884,
+      "grad_norm": 0.21622449159622192,
+      "learning_rate": 4.685858693975011e-06,
+      "loss": 0.3732,
+      "step": 2098
+    },
+    {
+      "epoch": 1.7051177904142973,
+      "grad_norm": 0.2220907360315323,
+      "learning_rate": 4.681141131220982e-06,
+      "loss": 0.3763,
+      "step": 2099
+    },
+    {
+      "epoch": 1.7059301380991063,
+      "grad_norm": 0.21030853688716888,
+      "learning_rate": 4.6764238534621436e-06,
+      "loss": 0.3898,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7067424857839155,
+      "grad_norm": 0.20800566673278809,
+      "learning_rate": 4.671706864914786e-06,
+      "loss": 0.3829,
+      "step": 2101
+    },
+    {
+      "epoch": 1.7075548334687247,
+      "grad_norm": 0.18791958689689636,
+      "learning_rate": 4.666990169794942e-06,
+      "loss": 0.34,
+      "step": 2102
+    },
+    {
+      "epoch": 1.7083671811535337,
+      "grad_norm": 0.2134494036436081,
+      "learning_rate": 4.662273772318378e-06,
+      "loss": 0.3651,
+      "step": 2103
+    },
+    {
+      "epoch": 1.7091795288383427,
+      "grad_norm": 0.22413446009159088,
+      "learning_rate": 4.6575576767006e-06,
+      "loss": 0.3652,
+      "step": 2104
+    },
+    {
+      "epoch": 1.7099918765231519,
+      "grad_norm": 0.18747375905513763,
+      "learning_rate": 4.652841887156841e-06,
+      "loss": 0.3508,
+      "step": 2105
+    },
+    {
+      "epoch": 1.710804224207961,
+      "grad_norm": 0.20869235694408417,
+      "learning_rate": 4.648126407902058e-06,
+      "loss": 0.3719,
+      "step": 2106
+    },
+    {
+      "epoch": 1.71161657189277,
+      "grad_norm": 0.19136790931224823,
+      "learning_rate": 4.6434112431509376e-06,
+      "loss": 0.354,
+      "step": 2107
+    },
+    {
+      "epoch": 1.7124289195775793,
+      "grad_norm": 0.2115483433008194,
+      "learning_rate": 4.638696397117877e-06,
+      "loss": 0.3915,
+      "step": 2108
+    },
+    {
+      "epoch": 1.7132412672623882,
+      "grad_norm": 0.19347473978996277,
+      "learning_rate": 4.6339818740169975e-06,
+      "loss": 0.3609,
+      "step": 2109
+    },
+    {
+      "epoch": 1.7140536149471974,
+      "grad_norm": 0.1990123838186264,
+      "learning_rate": 4.6292676780621245e-06,
+      "loss": 0.3798,
+      "step": 2110
+    },
+    {
+      "epoch": 1.7148659626320066,
+      "grad_norm": 0.20360954105854034,
+      "learning_rate": 4.624553813466794e-06,
+      "loss": 0.3629,
+      "step": 2111
+    },
+    {
+      "epoch": 1.7156783103168156,
+      "grad_norm": 0.20038798451423645,
+      "learning_rate": 4.619840284444245e-06,
+      "loss": 0.4132,
+      "step": 2112
+    },
+    {
+      "epoch": 1.7164906580016246,
+      "grad_norm": 0.20865312218666077,
+      "learning_rate": 4.615127095207418e-06,
+      "loss": 0.3687,
+      "step": 2113
+    },
+    {
+      "epoch": 1.7173030056864338,
+      "grad_norm": 0.20283755660057068,
+      "learning_rate": 4.610414249968947e-06,
+      "loss": 0.3583,
+      "step": 2114
+    },
+    {
+      "epoch": 1.718115353371243,
+      "grad_norm": 0.18748416006565094,
+      "learning_rate": 4.605701752941161e-06,
+      "loss": 0.387,
+      "step": 2115
+    },
+    {
+      "epoch": 1.718927701056052,
+      "grad_norm": 0.21634866297245026,
+      "learning_rate": 4.600989608336078e-06,
+      "loss": 0.3495,
+      "step": 2116
+    },
+    {
+      "epoch": 1.719740048740861,
+      "grad_norm": 0.19799114763736725,
+      "learning_rate": 4.596277820365398e-06,
+      "loss": 0.3638,
+      "step": 2117
+    },
+    {
+      "epoch": 1.7205523964256701,
+      "grad_norm": 0.21327516436576843,
+      "learning_rate": 4.591566393240507e-06,
+      "loss": 0.3592,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7213647441104794,
+      "grad_norm": 0.2014748603105545,
+      "learning_rate": 4.5868553311724635e-06,
+      "loss": 0.3874,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7221770917952883,
+      "grad_norm": 0.2064102292060852,
+      "learning_rate": 4.5821446383720025e-06,
+      "loss": 0.3725,
+      "step": 2120
+    },
+    {
+      "epoch": 1.7229894394800975,
+      "grad_norm": 0.2021154910326004,
+      "learning_rate": 4.577434319049529e-06,
+      "loss": 0.3757,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7238017871649065,
+      "grad_norm": 0.18935824930667877,
+      "learning_rate": 4.572724377415114e-06,
+      "loss": 0.3527,
+      "step": 2122
+    },
+    {
+      "epoch": 1.7246141348497157,
+      "grad_norm": 0.2226516604423523,
+      "learning_rate": 4.568014817678488e-06,
+      "loss": 0.38,
+      "step": 2123
+    },
+    {
+      "epoch": 1.725426482534525,
+      "grad_norm": 0.20491117238998413,
+      "learning_rate": 4.5633056440490455e-06,
+      "loss": 0.3759,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7262388302193339,
+      "grad_norm": 0.19322121143341064,
+      "learning_rate": 4.558596860735833e-06,
+      "loss": 0.3919,
+      "step": 2125
+    },
+    {
+      "epoch": 1.7270511779041429,
+      "grad_norm": 0.1888704001903534,
+      "learning_rate": 4.553888471947546e-06,
+      "loss": 0.3393,
+      "step": 2126
+    },
+    {
+      "epoch": 1.727863525588952,
+      "grad_norm": 0.21036258339881897,
+      "learning_rate": 4.549180481892532e-06,
+      "loss": 0.3595,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7286758732737613,
+      "grad_norm": 0.19312798976898193,
+      "learning_rate": 4.544472894778779e-06,
+      "loss": 0.3569,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7294882209585702,
+      "grad_norm": 0.20159725844860077,
+      "learning_rate": 4.539765714813915e-06,
+      "loss": 0.3889,
+      "step": 2129
+    },
+    {
+      "epoch": 1.7303005686433792,
+      "grad_norm": 0.22536255419254303,
+      "learning_rate": 4.535058946205205e-06,
+      "loss": 0.3606,
+      "step": 2130
+    },
+    {
+      "epoch": 1.7311129163281884,
+      "grad_norm": 0.18967297673225403,
+      "learning_rate": 4.530352593159546e-06,
+      "loss": 0.3388,
+      "step": 2131
+    },
+    {
+      "epoch": 1.7319252640129976,
+      "grad_norm": 0.23492856323719025,
+      "learning_rate": 4.525646659883463e-06,
+      "loss": 0.4014,
+      "step": 2132
+    },
+    {
+      "epoch": 1.7327376116978068,
+      "grad_norm": 0.19237980246543884,
+      "learning_rate": 4.5209411505831066e-06,
+      "loss": 0.3517,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7335499593826158,
+      "grad_norm": 0.2035786509513855,
+      "learning_rate": 4.51623606946425e-06,
+      "loss": 0.3707,
+      "step": 2134
+    },
+    {
+      "epoch": 1.7343623070674248,
+      "grad_norm": 0.20536081492900848,
+      "learning_rate": 4.5115314207322794e-06,
+      "loss": 0.3624,
+      "step": 2135
+    },
+    {
+      "epoch": 1.735174654752234,
+      "grad_norm": 0.1823720782995224,
+      "learning_rate": 4.506827208592198e-06,
+      "loss": 0.3638,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7359870024370432,
+      "grad_norm": 0.19357626140117645,
+      "learning_rate": 4.502123437248619e-06,
+      "loss": 0.3658,
+      "step": 2137
+    },
+    {
+      "epoch": 1.7367993501218522,
+      "grad_norm": 0.1943293958902359,
+      "learning_rate": 4.497420110905758e-06,
+      "loss": 0.3775,
+      "step": 2138
+    },
+    {
+      "epoch": 1.7376116978066611,
+      "grad_norm": 0.2030443400144577,
+      "learning_rate": 4.492717233767438e-06,
+      "loss": 0.3678,
+      "step": 2139
+    },
+    {
+      "epoch": 1.7384240454914703,
+      "grad_norm": 0.2082703560590744,
+      "learning_rate": 4.4880148100370765e-06,
+      "loss": 0.4,
+      "step": 2140
+    },
+    {
+      "epoch": 1.7392363931762795,
+      "grad_norm": 0.21757914125919342,
+      "learning_rate": 4.483312843917685e-06,
+      "loss": 0.3695,
+      "step": 2141
+    },
+    {
+      "epoch": 1.7400487408610885,
+      "grad_norm": 0.20350830256938934,
+      "learning_rate": 4.4786113396118715e-06,
+      "loss": 0.3529,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7408610885458975,
+      "grad_norm": 0.2134498953819275,
+      "learning_rate": 4.473910301321827e-06,
+      "loss": 0.3623,
+      "step": 2143
+    },
+    {
+      "epoch": 1.7416734362307067,
+      "grad_norm": 0.22364667057991028,
+      "learning_rate": 4.469209733249326e-06,
+      "loss": 0.3548,
+      "step": 2144
+    },
+    {
+      "epoch": 1.742485783915516,
+      "grad_norm": 0.1926523894071579,
+      "learning_rate": 4.4645096395957235e-06,
+      "loss": 0.3579,
+      "step": 2145
+    },
+    {
+      "epoch": 1.743298131600325,
+      "grad_norm": 0.19713754951953888,
+      "learning_rate": 4.4598100245619505e-06,
+      "loss": 0.3571,
+      "step": 2146
+    },
+    {
+      "epoch": 1.744110479285134,
+      "grad_norm": 0.223915234208107,
+      "learning_rate": 4.455110892348508e-06,
+      "loss": 0.3629,
+      "step": 2147
+    },
+    {
+      "epoch": 1.744922826969943,
+      "grad_norm": 0.1889084279537201,
+      "learning_rate": 4.450412247155473e-06,
+      "loss": 0.3644,
+      "step": 2148
+    },
+    {
+      "epoch": 1.7457351746547523,
+      "grad_norm": 0.20201948285102844,
+      "learning_rate": 4.445714093182477e-06,
+      "loss": 0.3695,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7465475223395615,
+      "grad_norm": 0.20632126927375793,
+      "learning_rate": 4.441016434628719e-06,
+      "loss": 0.3678,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7473598700243704,
+      "grad_norm": 0.20967699587345123,
+      "learning_rate": 4.436319275692954e-06,
+      "loss": 0.3798,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7481722177091794,
+      "grad_norm": 0.2088678777217865,
+      "learning_rate": 4.43162262057349e-06,
+      "loss": 0.3628,
+      "step": 2152
+    },
+    {
+      "epoch": 1.7489845653939886,
+      "grad_norm": 0.20555980503559113,
+      "learning_rate": 4.426926473468183e-06,
+      "loss": 0.3639,
+      "step": 2153
+    },
+    {
+      "epoch": 1.7497969130787978,
+      "grad_norm": 0.19456538558006287,
+      "learning_rate": 4.4222308385744375e-06,
+      "loss": 0.3448,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7506092607636068,
+      "grad_norm": 0.20734603703022003,
+      "learning_rate": 4.417535720089199e-06,
+      "loss": 0.3484,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7514216084484158,
+      "grad_norm": 0.20847731828689575,
+      "learning_rate": 4.412841122208951e-06,
+      "loss": 0.3659,
+      "step": 2156
+    },
+    {
+      "epoch": 1.752233956133225,
+      "grad_norm": 0.20452938973903656,
+      "learning_rate": 4.408147049129715e-06,
+      "loss": 0.3566,
+      "step": 2157
+    },
+    {
+      "epoch": 1.7530463038180342,
+      "grad_norm": 0.22252462804317474,
+      "learning_rate": 4.403453505047037e-06,
+      "loss": 0.3648,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7538586515028434,
+      "grad_norm": 0.179750457406044,
+      "learning_rate": 4.3987604941559966e-06,
+      "loss": 0.3569,
+      "step": 2159
+    },
+    {
+      "epoch": 1.7546709991876523,
+      "grad_norm": 0.19286935031414032,
+      "learning_rate": 4.3940680206511935e-06,
+      "loss": 0.3734,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7554833468724613,
+      "grad_norm": 0.1952231228351593,
+      "learning_rate": 4.389376088726747e-06,
+      "loss": 0.3705,
+      "step": 2161
+    },
+    {
+      "epoch": 1.7562956945572705,
+      "grad_norm": 0.20499952137470245,
+      "learning_rate": 4.3846847025762944e-06,
+      "loss": 0.3755,
+      "step": 2162
+    },
+    {
+      "epoch": 1.7571080422420797,
+      "grad_norm": 0.3142097294330597,
+      "learning_rate": 4.3799938663929835e-06,
+      "loss": 0.3625,
+      "step": 2163
+    },
+    {
+      "epoch": 1.7579203899268887,
+      "grad_norm": 0.1940368264913559,
+      "learning_rate": 4.37530358436947e-06,
+      "loss": 0.3655,
+      "step": 2164
+    },
+    {
+      "epoch": 1.7587327376116977,
+      "grad_norm": 0.22935840487480164,
+      "learning_rate": 4.370613860697916e-06,
+      "loss": 0.3724,
+      "step": 2165
+    },
+    {
+      "epoch": 1.7595450852965069,
+      "grad_norm": 0.19888950884342194,
+      "learning_rate": 4.3659246995699845e-06,
+      "loss": 0.3617,
+      "step": 2166
+    },
+    {
+      "epoch": 1.760357432981316,
+      "grad_norm": 0.18897780776023865,
+      "learning_rate": 4.361236105176832e-06,
+      "loss": 0.3678,
+      "step": 2167
+    },
+    {
+      "epoch": 1.761169780666125,
+      "grad_norm": 0.19439385831356049,
+      "learning_rate": 4.356548081709116e-06,
+      "loss": 0.3546,
+      "step": 2168
+    },
+    {
+      "epoch": 1.761982128350934,
+      "grad_norm": 0.20573462545871735,
+      "learning_rate": 4.351860633356976e-06,
+      "loss": 0.3544,
+      "step": 2169
+    },
+    {
+      "epoch": 1.7627944760357432,
+      "grad_norm": 0.20261052250862122,
+      "learning_rate": 4.347173764310041e-06,
+      "loss": 0.3553,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7636068237205524,
+      "grad_norm": 0.1910543590784073,
+      "learning_rate": 4.342487478757422e-06,
+      "loss": 0.3646,
+      "step": 2171
+    },
+    {
+      "epoch": 1.7644191714053616,
+      "grad_norm": 0.20165228843688965,
+      "learning_rate": 4.3378017808877095e-06,
+      "loss": 0.3402,
+      "step": 2172
+    },
+    {
+      "epoch": 1.7652315190901706,
+      "grad_norm": 0.2050500214099884,
+      "learning_rate": 4.333116674888966e-06,
+      "loss": 0.3628,
+      "step": 2173
+    },
+    {
+      "epoch": 1.7660438667749796,
+      "grad_norm": 0.19631178677082062,
+      "learning_rate": 4.328432164948727e-06,
+      "loss": 0.3464,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7668562144597888,
+      "grad_norm": 0.20119580626487732,
+      "learning_rate": 4.323748255253995e-06,
+      "loss": 0.388,
+      "step": 2175
+    },
+    {
+      "epoch": 1.767668562144598,
+      "grad_norm": 0.19579000771045685,
+      "learning_rate": 4.319064949991236e-06,
+      "loss": 0.3478,
+      "step": 2176
+    },
+    {
+      "epoch": 1.768480909829407,
+      "grad_norm": 0.20636332035064697,
+      "learning_rate": 4.314382253346374e-06,
+      "loss": 0.3822,
+      "step": 2177
+    },
+    {
+      "epoch": 1.769293257514216,
+      "grad_norm": 0.18717509508132935,
+      "learning_rate": 4.309700169504792e-06,
+      "loss": 0.3975,
+      "step": 2178
+    },
+    {
+      "epoch": 1.7701056051990252,
+      "grad_norm": 0.17988668382167816,
+      "learning_rate": 4.305018702651323e-06,
+      "loss": 0.3362,
+      "step": 2179
+    },
+    {
+      "epoch": 1.7709179528838344,
+      "grad_norm": 0.20296888053417206,
+      "learning_rate": 4.300337856970251e-06,
+      "loss": 0.3739,
+      "step": 2180
+    },
+    {
+      "epoch": 1.7717303005686433,
+      "grad_norm": 0.18306489288806915,
+      "learning_rate": 4.295657636645302e-06,
+      "loss": 0.3438,
+      "step": 2181
+    },
+    {
+      "epoch": 1.7725426482534523,
+      "grad_norm": 0.1907651573419571,
+      "learning_rate": 4.290978045859643e-06,
+      "loss": 0.3642,
+      "step": 2182
+    },
+    {
+      "epoch": 1.7733549959382615,
+      "grad_norm": 0.2070646733045578,
+      "learning_rate": 4.2862990887958815e-06,
+      "loss": 0.3551,
+      "step": 2183
+    },
+    {
+      "epoch": 1.7741673436230707,
+      "grad_norm": 0.20375695824623108,
+      "learning_rate": 4.281620769636055e-06,
+      "loss": 0.3699,
+      "step": 2184
+    },
+    {
+      "epoch": 1.77497969130788,
+      "grad_norm": 0.20962126553058624,
+      "learning_rate": 4.276943092561632e-06,
+      "loss": 0.3693,
+      "step": 2185
+    },
+    {
+      "epoch": 1.775792038992689,
+      "grad_norm": 0.21130754053592682,
+      "learning_rate": 4.2722660617535105e-06,
+      "loss": 0.3664,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7766043866774979,
+      "grad_norm": 0.1887819468975067,
+      "learning_rate": 4.267589681392006e-06,
+      "loss": 0.3665,
+      "step": 2187
+    },
+    {
+      "epoch": 1.777416734362307,
+      "grad_norm": 0.1906622052192688,
+      "learning_rate": 4.262913955656856e-06,
+      "loss": 0.3872,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7782290820471163,
+      "grad_norm": 0.19949328899383545,
+      "learning_rate": 4.258238888727209e-06,
+      "loss": 0.3639,
+      "step": 2189
+    },
+    {
+      "epoch": 1.7790414297319253,
+      "grad_norm": 0.21867232024669647,
+      "learning_rate": 4.253564484781629e-06,
+      "loss": 0.3643,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7798537774167342,
+      "grad_norm": 0.2005881518125534,
+      "learning_rate": 4.248890747998085e-06,
+      "loss": 0.3481,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7806661251015434,
+      "grad_norm": 0.19981642067432404,
+      "learning_rate": 4.24421768255395e-06,
+      "loss": 0.3614,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7814784727863526,
+      "grad_norm": 0.2018483281135559,
+      "learning_rate": 4.2395452926259965e-06,
+      "loss": 0.3489,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7822908204711616,
+      "grad_norm": 0.21631909906864166,
+      "learning_rate": 4.234873582390395e-06,
+      "loss": 0.3601,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7831031681559708,
+      "grad_norm": 0.19199076294898987,
+      "learning_rate": 4.230202556022708e-06,
+      "loss": 0.3484,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7839155158407798,
+      "grad_norm": 0.22114524245262146,
+      "learning_rate": 4.225532217697886e-06,
+      "loss": 0.3835,
+      "step": 2196
+    },
+    {
+      "epoch": 1.784727863525589,
+      "grad_norm": 0.19773583114147186,
+      "learning_rate": 4.220862571590264e-06,
+      "loss": 0.3686,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7855402112103982,
+      "grad_norm": 0.19582776725292206,
+      "learning_rate": 4.216193621873559e-06,
+      "loss": 0.3436,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7863525588952072,
+      "grad_norm": 0.19409973919391632,
+      "learning_rate": 4.211525372720865e-06,
+      "loss": 0.3654,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7871649065800161,
+      "grad_norm": 0.17911002039909363,
+      "learning_rate": 4.206857828304651e-06,
+      "loss": 0.362,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7879772542648253,
+      "grad_norm": 0.20142251253128052,
+      "learning_rate": 4.202190992796756e-06,
+      "loss": 0.3819,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7887896019496345,
+      "grad_norm": 0.2022000104188919,
+      "learning_rate": 4.1975248703683855e-06,
+      "loss": 0.3461,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7896019496344435,
+      "grad_norm": 0.21479280292987823,
+      "learning_rate": 4.1928594651901055e-06,
+      "loss": 0.3634,
+      "step": 2203
+    },
+    {
+      "epoch": 1.7904142973192525,
+      "grad_norm": 0.19392316043376923,
+      "learning_rate": 4.188194781431843e-06,
+      "loss": 0.3552,
+      "step": 2204
+    },
+    {
+      "epoch": 1.7912266450040617,
+      "grad_norm": 0.2030022144317627,
+      "learning_rate": 4.18353082326288e-06,
+      "loss": 0.3523,
+      "step": 2205
+    },
+    {
+      "epoch": 1.792038992688871,
+      "grad_norm": 0.18990521132946014,
+      "learning_rate": 4.178867594851849e-06,
+      "loss": 0.3549,
+      "step": 2206
+    },
+    {
+      "epoch": 1.7928513403736799,
+      "grad_norm": 0.18842945992946625,
+      "learning_rate": 4.1742051003667315e-06,
+      "loss": 0.3648,
+      "step": 2207
+    },
+    {
+      "epoch": 1.793663688058489,
+      "grad_norm": 0.20738860964775085,
+      "learning_rate": 4.169543343974852e-06,
+      "loss": 0.3527,
+      "step": 2208
+    },
+    {
+      "epoch": 1.794476035743298,
+      "grad_norm": 0.20624947547912598,
+      "learning_rate": 4.164882329842876e-06,
+      "loss": 0.3506,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7952883834281073,
+      "grad_norm": 0.20762039721012115,
+      "learning_rate": 4.160222062136805e-06,
+      "loss": 0.3622,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7961007311129165,
+      "grad_norm": 0.19993452727794647,
+      "learning_rate": 4.1555625450219725e-06,
+      "loss": 0.3505,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7969130787977254,
+      "grad_norm": 0.1918487846851349,
+      "learning_rate": 4.150903782663045e-06,
+      "loss": 0.3734,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7977254264825344,
+      "grad_norm": 0.2013358473777771,
+      "learning_rate": 4.146245779224011e-06,
+      "loss": 0.3666,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7985377741673436,
+      "grad_norm": 0.20446647703647614,
+      "learning_rate": 4.141588538868179e-06,
+      "loss": 0.3646,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7993501218521528,
+      "grad_norm": 0.20169579982757568,
+      "learning_rate": 4.13693206575818e-06,
+      "loss": 0.383,
+      "step": 2215
+    },
+    {
+      "epoch": 1.8001624695369618,
+      "grad_norm": 0.18741396069526672,
+      "learning_rate": 4.132276364055956e-06,
+      "loss": 0.3642,
+      "step": 2216
+    },
+    {
+      "epoch": 1.8009748172217708,
+      "grad_norm": 0.19950363039970398,
+      "learning_rate": 4.127621437922761e-06,
+      "loss": 0.3604,
+      "step": 2217
+    },
+    {
+      "epoch": 1.80178716490658,
+      "grad_norm": 0.21952900290489197,
+      "learning_rate": 4.122967291519155e-06,
+      "loss": 0.3557,
+      "step": 2218
+    },
+    {
+      "epoch": 1.8025995125913892,
+      "grad_norm": 0.20028498768806458,
+      "learning_rate": 4.118313929005002e-06,
+      "loss": 0.3682,
+      "step": 2219
+    },
+    {
+      "epoch": 1.8034118602761984,
+      "grad_norm": 0.2083844244480133,
+      "learning_rate": 4.113661354539464e-06,
+      "loss": 0.3716,
+      "step": 2220
+    },
+    {
+      "epoch": 1.8042242079610074,
+      "grad_norm": 0.20891471207141876,
+      "learning_rate": 4.109009572280999e-06,
+      "loss": 0.3641,
+      "step": 2221
+    },
+    {
+      "epoch": 1.8050365556458163,
+      "grad_norm": 0.22006535530090332,
+      "learning_rate": 4.1043585863873595e-06,
+      "loss": 0.3591,
+      "step": 2222
+    },
+    {
+      "epoch": 1.8058489033306255,
+      "grad_norm": 0.20179177820682526,
+      "learning_rate": 4.09970840101558e-06,
+      "loss": 0.3759,
+      "step": 2223
+    },
+    {
+      "epoch": 1.8066612510154347,
+      "grad_norm": 0.18459518253803253,
+      "learning_rate": 4.095059020321988e-06,
+      "loss": 0.3754,
+      "step": 2224
+    },
+    {
+      "epoch": 1.8074735987002437,
+      "grad_norm": 0.20829787850379944,
+      "learning_rate": 4.090410448462186e-06,
+      "loss": 0.3583,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8082859463850527,
+      "grad_norm": 0.1955033540725708,
+      "learning_rate": 4.085762689591054e-06,
+      "loss": 0.3744,
+      "step": 2226
+    },
+    {
+      "epoch": 1.809098294069862,
+      "grad_norm": 0.21954289078712463,
+      "learning_rate": 4.081115747862747e-06,
+      "loss": 0.3662,
+      "step": 2227
+    },
+    {
+      "epoch": 1.809910641754671,
+      "grad_norm": 0.2096974402666092,
+      "learning_rate": 4.076469627430689e-06,
+      "loss": 0.3865,
+      "step": 2228
+    },
+    {
+      "epoch": 1.81072298943948,
+      "grad_norm": 0.18533852696418762,
+      "learning_rate": 4.071824332447568e-06,
+      "loss": 0.3734,
+      "step": 2229
+    },
+    {
+      "epoch": 1.811535337124289,
+      "grad_norm": 0.20886552333831787,
+      "learning_rate": 4.067179867065339e-06,
+      "loss": 0.3511,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8123476848090982,
+      "grad_norm": 0.21302489936351776,
+      "learning_rate": 4.0625362354352105e-06,
+      "loss": 0.3738,
+      "step": 2231
+    },
+    {
+      "epoch": 1.8131600324939074,
+      "grad_norm": 0.2042187750339508,
+      "learning_rate": 4.0578934417076486e-06,
+      "loss": 0.3917,
+      "step": 2232
+    },
+    {
+      "epoch": 1.8139723801787166,
+      "grad_norm": 0.19181464612483978,
+      "learning_rate": 4.053251490032368e-06,
+      "loss": 0.3483,
+      "step": 2233
+    },
+    {
+      "epoch": 1.8147847278635256,
+      "grad_norm": 0.21256573498249054,
+      "learning_rate": 4.048610384558335e-06,
+      "loss": 0.3739,
+      "step": 2234
+    },
+    {
+      "epoch": 1.8155970755483346,
+      "grad_norm": 0.22171242535114288,
+      "learning_rate": 4.043970129433757e-06,
+      "loss": 0.3723,
+      "step": 2235
+    },
+    {
+      "epoch": 1.8164094232331438,
+      "grad_norm": 0.21703511476516724,
+      "learning_rate": 4.039330728806079e-06,
+      "loss": 0.3818,
+      "step": 2236
+    },
+    {
+      "epoch": 1.817221770917953,
+      "grad_norm": 0.1844750940799713,
+      "learning_rate": 4.034692186821986e-06,
+      "loss": 0.346,
+      "step": 2237
+    },
+    {
+      "epoch": 1.818034118602762,
+      "grad_norm": 0.21397578716278076,
+      "learning_rate": 4.030054507627395e-06,
+      "loss": 0.3761,
+      "step": 2238
+    },
+    {
+      "epoch": 1.818846466287571,
+      "grad_norm": 0.3555406332015991,
+      "learning_rate": 4.0254176953674505e-06,
+      "loss": 0.3504,
+      "step": 2239
+    },
+    {
+      "epoch": 1.8196588139723802,
+      "grad_norm": 0.21137873828411102,
+      "learning_rate": 4.020781754186523e-06,
+      "loss": 0.3672,
+      "step": 2240
+    },
+    {
+      "epoch": 1.8204711616571894,
+      "grad_norm": 0.1992935985326767,
+      "learning_rate": 4.016146688228203e-06,
+      "loss": 0.3545,
+      "step": 2241
+    },
+    {
+      "epoch": 1.8212835093419983,
+      "grad_norm": 0.20518332719802856,
+      "learning_rate": 4.011512501635301e-06,
+      "loss": 0.3725,
+      "step": 2242
+    },
+    {
+      "epoch": 1.8220958570268073,
+      "grad_norm": 0.2120194286108017,
+      "learning_rate": 4.00687919854984e-06,
+      "loss": 0.3796,
+      "step": 2243
+    },
+    {
+      "epoch": 1.8229082047116165,
+      "grad_norm": 0.20480699837207794,
+      "learning_rate": 4.002246783113054e-06,
+      "loss": 0.3732,
+      "step": 2244
+    },
+    {
+      "epoch": 1.8237205523964257,
+      "grad_norm": 0.19998447597026825,
+      "learning_rate": 3.997615259465382e-06,
+      "loss": 0.3351,
+      "step": 2245
+    },
+    {
+      "epoch": 1.824532900081235,
+      "grad_norm": 0.219023659825325,
+      "learning_rate": 3.992984631746469e-06,
+      "loss": 0.3486,
+      "step": 2246
+    },
+    {
+      "epoch": 1.825345247766044,
+      "grad_norm": 0.2114606648683548,
+      "learning_rate": 3.988354904095157e-06,
+      "loss": 0.3769,
+      "step": 2247
+    },
+    {
+      "epoch": 1.8261575954508529,
+      "grad_norm": 0.18678084015846252,
+      "learning_rate": 3.983726080649484e-06,
+      "loss": 0.3523,
+      "step": 2248
+    },
+    {
+      "epoch": 1.826969943135662,
+      "grad_norm": 0.2012970894575119,
+      "learning_rate": 3.979098165546679e-06,
+      "loss": 0.3541,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8277822908204713,
+      "grad_norm": 0.2007148563861847,
+      "learning_rate": 3.974471162923162e-06,
+      "loss": 0.3453,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8285946385052803,
+      "grad_norm": 0.21245332062244415,
+      "learning_rate": 3.969845076914532e-06,
+      "loss": 0.3735,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8294069861900892,
+      "grad_norm": 0.2103399634361267,
+      "learning_rate": 3.965219911655575e-06,
+      "loss": 0.3793,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8302193338748984,
+      "grad_norm": 0.19292905926704407,
+      "learning_rate": 3.96059567128025e-06,
+      "loss": 0.3645,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8310316815597076,
+      "grad_norm": 0.20069080591201782,
+      "learning_rate": 3.9559723599216905e-06,
+      "loss": 0.3775,
+      "step": 2254
+    },
+    {
+      "epoch": 1.8318440292445166,
+      "grad_norm": 0.21322022378444672,
+      "learning_rate": 3.951349981712199e-06,
+      "loss": 0.3657,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8326563769293256,
+      "grad_norm": 0.23900718986988068,
+      "learning_rate": 3.946728540783247e-06,
+      "loss": 0.3641,
+      "step": 2256
+    },
+    {
+      "epoch": 1.8334687246141348,
+      "grad_norm": 0.20546041429042816,
+      "learning_rate": 3.942108041265463e-06,
+      "loss": 0.3699,
+      "step": 2257
+    },
+    {
+      "epoch": 1.834281072298944,
+      "grad_norm": 0.1953115612268448,
+      "learning_rate": 3.9374884872886394e-06,
+      "loss": 0.3504,
+      "step": 2258
+    },
+    {
+      "epoch": 1.8350934199837532,
+      "grad_norm": 0.19697800278663635,
+      "learning_rate": 3.9328698829817195e-06,
+      "loss": 0.3637,
+      "step": 2259
+    },
+    {
+      "epoch": 1.8359057676685622,
+      "grad_norm": 0.19586117565631866,
+      "learning_rate": 3.928252232472799e-06,
+      "loss": 0.3569,
+      "step": 2260
+    },
+    {
+      "epoch": 1.8367181153533712,
+      "grad_norm": 0.2261938899755478,
+      "learning_rate": 3.923635539889121e-06,
+      "loss": 0.3662,
+      "step": 2261
+    },
+    {
+      "epoch": 1.8375304630381804,
+      "grad_norm": 0.2264709770679474,
+      "learning_rate": 3.919019809357075e-06,
+      "loss": 0.3468,
+      "step": 2262
+    },
+    {
+      "epoch": 1.8383428107229896,
+      "grad_norm": 0.2169385850429535,
+      "learning_rate": 3.914405045002186e-06,
+      "loss": 0.3471,
+      "step": 2263
+    },
+    {
+      "epoch": 1.8391551584077985,
+      "grad_norm": 0.21138444542884827,
+      "learning_rate": 3.9097912509491195e-06,
+      "loss": 0.3746,
+      "step": 2264
+    },
+    {
+      "epoch": 1.8399675060926075,
+      "grad_norm": 0.18541575968265533,
+      "learning_rate": 3.905178431321672e-06,
+      "loss": 0.3717,
+      "step": 2265
+    },
+    {
+      "epoch": 1.8407798537774167,
+      "grad_norm": 0.25584298372268677,
+      "learning_rate": 3.9005665902427695e-06,
+      "loss": 0.3504,
+      "step": 2266
+    },
+    {
+      "epoch": 1.841592201462226,
+      "grad_norm": 0.21476419270038605,
+      "learning_rate": 3.895955731834461e-06,
+      "loss": 0.3711,
+      "step": 2267
+    },
+    {
+      "epoch": 1.8424045491470349,
+      "grad_norm": 0.21513549983501434,
+      "learning_rate": 3.8913458602179215e-06,
+      "loss": 0.3576,
+      "step": 2268
+    },
+    {
+      "epoch": 1.8432168968318439,
+      "grad_norm": 0.21332353353500366,
+      "learning_rate": 3.88673697951344e-06,
+      "loss": 0.3553,
+      "step": 2269
+    },
+    {
+      "epoch": 1.844029244516653,
+      "grad_norm": 0.20180805027484894,
+      "learning_rate": 3.882129093840423e-06,
+      "loss": 0.3589,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8448415922014623,
+      "grad_norm": 0.22707995772361755,
+      "learning_rate": 3.8775222073173844e-06,
+      "loss": 0.3446,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8456539398862715,
+      "grad_norm": 0.21966059505939484,
+      "learning_rate": 3.872916324061949e-06,
+      "loss": 0.3648,
+      "step": 2272
+    },
+    {
+      "epoch": 1.8464662875710804,
+      "grad_norm": 0.189620703458786,
+      "learning_rate": 3.8683114481908415e-06,
+      "loss": 0.3406,
+      "step": 2273
+    },
+    {
+      "epoch": 1.8472786352558894,
+      "grad_norm": 0.20277510583400726,
+      "learning_rate": 3.8637075838198864e-06,
+      "loss": 0.3832,
+      "step": 2274
+    },
+    {
+      "epoch": 1.8480909829406986,
+      "grad_norm": 0.20462965965270996,
+      "learning_rate": 3.859104735064008e-06,
+      "loss": 0.3553,
+      "step": 2275
+    },
+    {
+      "epoch": 1.8489033306255078,
+      "grad_norm": 0.21150478720664978,
+      "learning_rate": 3.854502906037217e-06,
+      "loss": 0.3824,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8497156783103168,
+      "grad_norm": 0.2066216617822647,
+      "learning_rate": 3.849902100852618e-06,
+      "loss": 0.3686,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8505280259951258,
+      "grad_norm": 0.21272437274456024,
+      "learning_rate": 3.845302323622398e-06,
+      "loss": 0.3837,
+      "step": 2278
+    },
+    {
+      "epoch": 1.851340373679935,
+      "grad_norm": 0.19368839263916016,
+      "learning_rate": 3.8407035784578226e-06,
+      "loss": 0.3713,
+      "step": 2279
+    },
+    {
+      "epoch": 1.8521527213647442,
+      "grad_norm": 0.21616551280021667,
+      "learning_rate": 3.8361058694692414e-06,
+      "loss": 0.3833,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8529650690495532,
+      "grad_norm": 0.18293030560016632,
+      "learning_rate": 3.83150920076607e-06,
+      "loss": 0.3667,
+      "step": 2281
+    },
+    {
+      "epoch": 1.8537774167343624,
+      "grad_norm": 0.20746789872646332,
+      "learning_rate": 3.826913576456803e-06,
+      "loss": 0.3808,
+      "step": 2282
+    },
+    {
+      "epoch": 1.8545897644191713,
+      "grad_norm": 0.17960192263126373,
+      "learning_rate": 3.822319000648995e-06,
+      "loss": 0.3719,
+      "step": 2283
+    },
+    {
+      "epoch": 1.8554021121039805,
+      "grad_norm": 0.18644222617149353,
+      "learning_rate": 3.817725477449265e-06,
+      "loss": 0.348,
+      "step": 2284
+    },
+    {
+      "epoch": 1.8562144597887897,
+      "grad_norm": 0.18533030152320862,
+      "learning_rate": 3.8131330109632925e-06,
+      "loss": 0.3722,
+      "step": 2285
+    },
+    {
+      "epoch": 1.8570268074735987,
+      "grad_norm": 0.19705617427825928,
+      "learning_rate": 3.8085416052958107e-06,
+      "loss": 0.3745,
+      "step": 2286
+    },
+    {
+      "epoch": 1.8578391551584077,
+      "grad_norm": 0.19359134137630463,
+      "learning_rate": 3.8039512645506055e-06,
+      "loss": 0.3807,
+      "step": 2287
+    },
+    {
+      "epoch": 1.858651502843217,
+      "grad_norm": 0.18668295443058014,
+      "learning_rate": 3.7993619928305113e-06,
+      "loss": 0.374,
+      "step": 2288
+    },
+    {
+      "epoch": 1.859463850528026,
+      "grad_norm": 0.188684344291687,
+      "learning_rate": 3.7947737942374065e-06,
+      "loss": 0.3495,
+      "step": 2289
+    },
+    {
+      "epoch": 1.860276198212835,
+      "grad_norm": 0.19055981934070587,
+      "learning_rate": 3.7901866728722102e-06,
+      "loss": 0.355,
+      "step": 2290
+    },
+    {
+      "epoch": 1.861088545897644,
+      "grad_norm": 0.208862766623497,
+      "learning_rate": 3.7856006328348805e-06,
+      "loss": 0.3647,
+      "step": 2291
+    },
+    {
+      "epoch": 1.8619008935824533,
+      "grad_norm": 0.19558539986610413,
+      "learning_rate": 3.7810156782244055e-06,
+      "loss": 0.3754,
+      "step": 2292
+    },
+    {
+      "epoch": 1.8627132412672625,
+      "grad_norm": 0.20988185703754425,
+      "learning_rate": 3.7764318131388067e-06,
+      "loss": 0.3599,
+      "step": 2293
+    },
+    {
+      "epoch": 1.8635255889520714,
+      "grad_norm": 0.18746952712535858,
+      "learning_rate": 3.7718490416751297e-06,
+      "loss": 0.3797,
+      "step": 2294
+    },
+    {
+      "epoch": 1.8643379366368806,
+      "grad_norm": 0.1911388486623764,
+      "learning_rate": 3.7672673679294436e-06,
+      "loss": 0.3741,
+      "step": 2295
+    },
+    {
+      "epoch": 1.8651502843216896,
+      "grad_norm": 0.19935646653175354,
+      "learning_rate": 3.762686795996835e-06,
+      "loss": 0.3748,
+      "step": 2296
+    },
+    {
+      "epoch": 1.8659626320064988,
+      "grad_norm": 0.19823622703552246,
+      "learning_rate": 3.7581073299714077e-06,
+      "loss": 0.3514,
+      "step": 2297
+    },
+    {
+      "epoch": 1.866774979691308,
+      "grad_norm": 0.19387052953243256,
+      "learning_rate": 3.7535289739462757e-06,
+      "loss": 0.3598,
+      "step": 2298
+    },
+    {
+      "epoch": 1.867587327376117,
+      "grad_norm": 0.20031635463237762,
+      "learning_rate": 3.7489517320135583e-06,
+      "loss": 0.3721,
+      "step": 2299
+    },
+    {
+      "epoch": 1.868399675060926,
+      "grad_norm": 0.1902754008769989,
+      "learning_rate": 3.7443756082643866e-06,
+      "loss": 0.3799,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8692120227457352,
+      "grad_norm": 0.2049015760421753,
+      "learning_rate": 3.739800606788886e-06,
+      "loss": 0.3829,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8700243704305444,
+      "grad_norm": 0.19624464213848114,
+      "learning_rate": 3.735226731676179e-06,
+      "loss": 0.3619,
+      "step": 2302
+    },
+    {
+      "epoch": 1.8708367181153533,
+      "grad_norm": 0.20935142040252686,
+      "learning_rate": 3.7306539870143853e-06,
+      "loss": 0.3454,
+      "step": 2303
+    },
+    {
+      "epoch": 1.8716490658001623,
+      "grad_norm": 0.1927555799484253,
+      "learning_rate": 3.7260823768906096e-06,
+      "loss": 0.3513,
+      "step": 2304
+    },
+    {
+      "epoch": 1.8724614134849715,
+      "grad_norm": 0.19506576657295227,
+      "learning_rate": 3.7215119053909444e-06,
+      "loss": 0.3662,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8732737611697807,
+      "grad_norm": 0.1970687210559845,
+      "learning_rate": 3.7169425766004653e-06,
+      "loss": 0.367,
+      "step": 2306
+    },
+    {
+      "epoch": 1.87408610885459,
+      "grad_norm": 0.1911667138338089,
+      "learning_rate": 3.7123743946032263e-06,
+      "loss": 0.3532,
+      "step": 2307
+    },
+    {
+      "epoch": 1.874898456539399,
+      "grad_norm": 0.19865334033966064,
+      "learning_rate": 3.7078073634822565e-06,
+      "loss": 0.3617,
+      "step": 2308
+    },
+    {
+      "epoch": 1.8757108042242079,
+      "grad_norm": 0.17522376775741577,
+      "learning_rate": 3.7032414873195544e-06,
+      "loss": 0.3603,
+      "step": 2309
+    },
+    {
+      "epoch": 1.876523151909017,
+      "grad_norm": 0.2054082304239273,
+      "learning_rate": 3.698676770196089e-06,
+      "loss": 0.3657,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8773354995938263,
+      "grad_norm": 0.19884231686592102,
+      "learning_rate": 3.6941132161917916e-06,
+      "loss": 0.3652,
+      "step": 2311
+    },
+    {
+      "epoch": 1.8781478472786353,
+      "grad_norm": 0.18903817236423492,
+      "learning_rate": 3.6895508293855546e-06,
+      "loss": 0.3507,
+      "step": 2312
+    },
+    {
+      "epoch": 1.8789601949634442,
+      "grad_norm": 0.19587041437625885,
+      "learning_rate": 3.6849896138552265e-06,
+      "loss": 0.3738,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8797725426482534,
+      "grad_norm": 0.17929311096668243,
+      "learning_rate": 3.68042957367761e-06,
+      "loss": 0.377,
+      "step": 2314
+    },
+    {
+      "epoch": 1.8805848903330626,
+      "grad_norm": 0.1930428296327591,
+      "learning_rate": 3.6758707129284556e-06,
+      "loss": 0.3543,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8813972380178716,
+      "grad_norm": 0.20045731961727142,
+      "learning_rate": 3.6713130356824615e-06,
+      "loss": 0.3614,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8822095857026806,
+      "grad_norm": 0.19671465456485748,
+      "learning_rate": 3.666756546013267e-06,
+      "loss": 0.3507,
+      "step": 2317
+    },
+    {
+      "epoch": 1.8830219333874898,
+      "grad_norm": 0.19636772572994232,
+      "learning_rate": 3.6622012479934505e-06,
+      "loss": 0.3737,
+      "step": 2318
+    },
+    {
+      "epoch": 1.883834281072299,
+      "grad_norm": 0.18984784185886383,
+      "learning_rate": 3.6576471456945244e-06,
+      "loss": 0.3602,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8846466287571082,
+      "grad_norm": 0.20403878390789032,
+      "learning_rate": 3.653094243186933e-06,
+      "loss": 0.355,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8854589764419172,
+      "grad_norm": 0.208901047706604,
+      "learning_rate": 3.648542544540049e-06,
+      "loss": 0.3721,
+      "step": 2321
+    },
+    {
+      "epoch": 1.8862713241267262,
+      "grad_norm": 0.17939354479312897,
+      "learning_rate": 3.6439920538221675e-06,
+      "loss": 0.3814,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8870836718115354,
+      "grad_norm": 0.1909172534942627,
+      "learning_rate": 3.639442775100505e-06,
+      "loss": 0.3845,
+      "step": 2323
+    },
+    {
+      "epoch": 1.8878960194963446,
+      "grad_norm": 0.2058834284543991,
+      "learning_rate": 3.634894712441193e-06,
+      "loss": 0.3758,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8887083671811535,
+      "grad_norm": 0.19150476157665253,
+      "learning_rate": 3.630347869909281e-06,
+      "loss": 0.3522,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8895207148659625,
+      "grad_norm": 0.18946987390518188,
+      "learning_rate": 3.6258022515687215e-06,
+      "loss": 0.3879,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8903330625507717,
+      "grad_norm": 0.19789531826972961,
+      "learning_rate": 3.6212578614823767e-06,
+      "loss": 0.3497,
+      "step": 2327
+    },
+    {
+      "epoch": 1.891145410235581,
+      "grad_norm": 0.20082390308380127,
+      "learning_rate": 3.616714703712011e-06,
+      "loss": 0.3639,
+      "step": 2328
+    },
+    {
+      "epoch": 1.89195775792039,
+      "grad_norm": 0.2046635001897812,
+      "learning_rate": 3.6121727823182854e-06,
+      "loss": 0.3589,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8927701056051989,
+      "grad_norm": 0.20669087767601013,
+      "learning_rate": 3.6076321013607575e-06,
+      "loss": 0.3692,
+      "step": 2330
+    },
+    {
+      "epoch": 1.893582453290008,
+      "grad_norm": 0.1891566962003708,
+      "learning_rate": 3.6030926648978748e-06,
+      "loss": 0.3549,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8943948009748173,
+      "grad_norm": 0.1896549016237259,
+      "learning_rate": 3.598554476986974e-06,
+      "loss": 0.3717,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8952071486596265,
+      "grad_norm": 0.19204919040203094,
+      "learning_rate": 3.594017541684275e-06,
+      "loss": 0.3747,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8960194963444355,
+      "grad_norm": 0.20279569923877716,
+      "learning_rate": 3.589481863044878e-06,
+      "loss": 0.3641,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8968318440292444,
+      "grad_norm": 0.18637162446975708,
+      "learning_rate": 3.5849474451227607e-06,
+      "loss": 0.3586,
+      "step": 2335
+    },
+    {
+      "epoch": 1.8976441917140536,
+      "grad_norm": 0.20507194101810455,
+      "learning_rate": 3.5804142919707734e-06,
+      "loss": 0.3796,
+      "step": 2336
+    },
+    {
+      "epoch": 1.8984565393988628,
+      "grad_norm": 0.20348142087459564,
+      "learning_rate": 3.5758824076406358e-06,
+      "loss": 0.3656,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8992688870836718,
+      "grad_norm": 0.21495845913887024,
+      "learning_rate": 3.5713517961829314e-06,
+      "loss": 0.3484,
+      "step": 2338
+    },
+    {
+      "epoch": 1.9000812347684808,
+      "grad_norm": 0.20073948800563812,
+      "learning_rate": 3.5668224616471136e-06,
+      "loss": 0.3791,
+      "step": 2339
+    },
+    {
+      "epoch": 1.90089358245329,
+      "grad_norm": 0.19227367639541626,
+      "learning_rate": 3.562294408081486e-06,
+      "loss": 0.3793,
+      "step": 2340
+    },
+    {
+      "epoch": 1.9017059301380992,
+      "grad_norm": 0.20308589935302734,
+      "learning_rate": 3.5577676395332116e-06,
+      "loss": 0.3694,
+      "step": 2341
+    },
+    {
+      "epoch": 1.9025182778229082,
+      "grad_norm": 0.19665677845478058,
+      "learning_rate": 3.5532421600483035e-06,
+      "loss": 0.3615,
+      "step": 2342
+    },
+    {
+      "epoch": 1.9033306255077171,
+      "grad_norm": 0.18354636430740356,
+      "learning_rate": 3.5487179736716233e-06,
+      "loss": 0.3431,
+      "step": 2343
+    },
+    {
+      "epoch": 1.9041429731925263,
+      "grad_norm": 0.19023191928863525,
+      "learning_rate": 3.5441950844468754e-06,
+      "loss": 0.3449,
+      "step": 2344
+    },
+    {
+      "epoch": 1.9049553208773355,
+      "grad_norm": 0.19710108637809753,
+      "learning_rate": 3.539673496416607e-06,
+      "loss": 0.3579,
+      "step": 2345
+    },
+    {
+      "epoch": 1.9057676685621447,
+      "grad_norm": 0.19279316067695618,
+      "learning_rate": 3.5351532136222012e-06,
+      "loss": 0.3713,
+      "step": 2346
+    },
+    {
+      "epoch": 1.9065800162469537,
+      "grad_norm": 0.18499085307121277,
+      "learning_rate": 3.5306342401038727e-06,
+      "loss": 0.3835,
+      "step": 2347
+    },
+    {
+      "epoch": 1.9073923639317627,
+      "grad_norm": 0.20572194457054138,
+      "learning_rate": 3.52611657990067e-06,
+      "loss": 0.3482,
+      "step": 2348
+    },
+    {
+      "epoch": 1.908204711616572,
+      "grad_norm": 0.1966007649898529,
+      "learning_rate": 3.5216002370504656e-06,
+      "loss": 0.356,
+      "step": 2349
+    },
+    {
+      "epoch": 1.909017059301381,
+      "grad_norm": 0.20190007984638214,
+      "learning_rate": 3.5170852155899517e-06,
+      "loss": 0.3614,
+      "step": 2350
+    },
+    {
+      "epoch": 1.90982940698619,
+      "grad_norm": 0.19767874479293823,
+      "learning_rate": 3.512571519554645e-06,
+      "loss": 0.3806,
+      "step": 2351
+    },
+    {
+      "epoch": 1.910641754670999,
+      "grad_norm": 0.19874687492847443,
+      "learning_rate": 3.5080591529788736e-06,
+      "loss": 0.3892,
+      "step": 2352
+    },
+    {
+      "epoch": 1.9114541023558083,
+      "grad_norm": 0.2030428647994995,
+      "learning_rate": 3.5035481198957794e-06,
+      "loss": 0.3615,
+      "step": 2353
+    },
+    {
+      "epoch": 1.9122664500406175,
+      "grad_norm": 0.22884011268615723,
+      "learning_rate": 3.4990384243373105e-06,
+      "loss": 0.3635,
+      "step": 2354
+    },
+    {
+      "epoch": 1.9130787977254264,
+      "grad_norm": 0.21030786633491516,
+      "learning_rate": 3.494530070334221e-06,
+      "loss": 0.3595,
+      "step": 2355
+    },
+    {
+      "epoch": 1.9138911454102354,
+      "grad_norm": 0.2006007879972458,
+      "learning_rate": 3.490023061916065e-06,
+      "loss": 0.3594,
+      "step": 2356
+    },
+    {
+      "epoch": 1.9147034930950446,
+      "grad_norm": 0.19408449530601501,
+      "learning_rate": 3.4855174031111947e-06,
+      "loss": 0.3757,
+      "step": 2357
+    },
+    {
+      "epoch": 1.9155158407798538,
+      "grad_norm": 0.21764403581619263,
+      "learning_rate": 3.481013097946756e-06,
+      "loss": 0.3855,
+      "step": 2358
+    },
+    {
+      "epoch": 1.916328188464663,
+      "grad_norm": 0.21326786279678345,
+      "learning_rate": 3.4765101504486842e-06,
+      "loss": 0.3502,
+      "step": 2359
+    },
+    {
+      "epoch": 1.917140536149472,
+      "grad_norm": 0.1768977791070938,
+      "learning_rate": 3.4720085646417025e-06,
+      "loss": 0.3874,
+      "step": 2360
+    },
+    {
+      "epoch": 1.917952883834281,
+      "grad_norm": 0.20083840191364288,
+      "learning_rate": 3.467508344549314e-06,
+      "loss": 0.3828,
+      "step": 2361
+    },
+    {
+      "epoch": 1.9187652315190902,
+      "grad_norm": 0.20045652985572815,
+      "learning_rate": 3.4630094941938054e-06,
+      "loss": 0.359,
+      "step": 2362
+    },
+    {
+      "epoch": 1.9195775792038994,
+      "grad_norm": 0.17939534783363342,
+      "learning_rate": 3.458512017596235e-06,
+      "loss": 0.365,
+      "step": 2363
+    },
+    {
+      "epoch": 1.9203899268887084,
+      "grad_norm": 0.18722450733184814,
+      "learning_rate": 3.454015918776437e-06,
+      "loss": 0.3642,
+      "step": 2364
+    },
+    {
+      "epoch": 1.9212022745735173,
+      "grad_norm": 0.1833220273256302,
+      "learning_rate": 3.449521201753011e-06,
+      "loss": 0.3603,
+      "step": 2365
+    },
+    {
+      "epoch": 1.9220146222583265,
+      "grad_norm": 0.19760198891162872,
+      "learning_rate": 3.445027870543323e-06,
+      "loss": 0.3706,
+      "step": 2366
+    },
+    {
+      "epoch": 1.9228269699431357,
+      "grad_norm": 0.1866169422864914,
+      "learning_rate": 3.4405359291635008e-06,
+      "loss": 0.3521,
+      "step": 2367
+    },
+    {
+      "epoch": 1.9236393176279447,
+      "grad_norm": 0.21740421652793884,
+      "learning_rate": 3.436045381628429e-06,
+      "loss": 0.373,
+      "step": 2368
+    },
+    {
+      "epoch": 1.924451665312754,
+      "grad_norm": 0.18812507390975952,
+      "learning_rate": 3.431556231951747e-06,
+      "loss": 0.3867,
+      "step": 2369
+    },
+    {
+      "epoch": 1.925264012997563,
+      "grad_norm": 0.1717631071805954,
+      "learning_rate": 3.4270684841458445e-06,
+      "loss": 0.3501,
+      "step": 2370
+    },
+    {
+      "epoch": 1.926076360682372,
+      "grad_norm": 0.1901012361049652,
+      "learning_rate": 3.4225821422218585e-06,
+      "loss": 0.3605,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9268887083671813,
+      "grad_norm": 0.19323980808258057,
+      "learning_rate": 3.418097210189669e-06,
+      "loss": 0.3773,
+      "step": 2372
+    },
+    {
+      "epoch": 1.9277010560519903,
+      "grad_norm": 0.18326778709888458,
+      "learning_rate": 3.4136136920578967e-06,
+      "loss": 0.3919,
+      "step": 2373
+    },
+    {
+      "epoch": 1.9285134037367992,
+      "grad_norm": 0.1953468769788742,
+      "learning_rate": 3.4091315918338968e-06,
+      "loss": 0.3729,
+      "step": 2374
+    },
+    {
+      "epoch": 1.9293257514216084,
+      "grad_norm": 0.20474830269813538,
+      "learning_rate": 3.4046509135237593e-06,
+      "loss": 0.3712,
+      "step": 2375
+    },
+    {
+      "epoch": 1.9301380991064176,
+      "grad_norm": 0.19130025804042816,
+      "learning_rate": 3.400171661132299e-06,
+      "loss": 0.3587,
+      "step": 2376
+    },
+    {
+      "epoch": 1.9309504467912266,
+      "grad_norm": 0.1944482922554016,
+      "learning_rate": 3.395693838663065e-06,
+      "loss": 0.349,
+      "step": 2377
+    },
+    {
+      "epoch": 1.9317627944760356,
+      "grad_norm": 0.18541264533996582,
+      "learning_rate": 3.3912174501183206e-06,
+      "loss": 0.3771,
+      "step": 2378
+    },
+    {
+      "epoch": 1.9325751421608448,
+      "grad_norm": 0.18710224330425262,
+      "learning_rate": 3.3867424994990483e-06,
+      "loss": 0.3528,
+      "step": 2379
+    },
+    {
+      "epoch": 1.933387489845654,
+      "grad_norm": 0.18347835540771484,
+      "learning_rate": 3.3822689908049487e-06,
+      "loss": 0.3637,
+      "step": 2380
+    },
+    {
+      "epoch": 1.934199837530463,
+      "grad_norm": 0.2024335414171219,
+      "learning_rate": 3.37779692803443e-06,
+      "loss": 0.3433,
+      "step": 2381
+    },
+    {
+      "epoch": 1.9350121852152722,
+      "grad_norm": 0.1819867044687271,
+      "learning_rate": 3.373326315184612e-06,
+      "loss": 0.3733,
+      "step": 2382
+    },
+    {
+      "epoch": 1.9358245329000812,
+      "grad_norm": 0.18470355868339539,
+      "learning_rate": 3.368857156251314e-06,
+      "loss": 0.3606,
+      "step": 2383
+    },
+    {
+      "epoch": 1.9366368805848904,
+      "grad_norm": 0.18617497384548187,
+      "learning_rate": 3.364389455229059e-06,
+      "loss": 0.3678,
+      "step": 2384
+    },
+    {
+      "epoch": 1.9374492282696996,
+      "grad_norm": 0.18571610748767853,
+      "learning_rate": 3.359923216111067e-06,
+      "loss": 0.3689,
+      "step": 2385
+    },
+    {
+      "epoch": 1.9382615759545085,
+      "grad_norm": 0.20680467784404755,
+      "learning_rate": 3.3554584428892488e-06,
+      "loss": 0.3691,
+      "step": 2386
+    },
+    {
+      "epoch": 1.9390739236393175,
+      "grad_norm": 0.2055879384279251,
+      "learning_rate": 3.3509951395542073e-06,
+      "loss": 0.3632,
+      "step": 2387
+    },
+    {
+      "epoch": 1.9398862713241267,
+      "grad_norm": 0.1900627166032791,
+      "learning_rate": 3.3465333100952306e-06,
+      "loss": 0.3486,
+      "step": 2388
+    },
+    {
+      "epoch": 1.940698619008936,
+      "grad_norm": 0.2149014174938202,
+      "learning_rate": 3.342072958500291e-06,
+      "loss": 0.3816,
+      "step": 2389
+    },
+    {
+      "epoch": 1.941510966693745,
+      "grad_norm": 0.2362845093011856,
+      "learning_rate": 3.337614088756037e-06,
+      "loss": 0.3831,
+      "step": 2390
+    },
+    {
+      "epoch": 1.9423233143785539,
+      "grad_norm": 0.19367368519306183,
+      "learning_rate": 3.3331567048477946e-06,
+      "loss": 0.3522,
+      "step": 2391
+    },
+    {
+      "epoch": 1.943135662063363,
+      "grad_norm": 0.19858333468437195,
+      "learning_rate": 3.328700810759562e-06,
+      "loss": 0.38,
+      "step": 2392
+    },
+    {
+      "epoch": 1.9439480097481723,
+      "grad_norm": 0.20249874889850616,
+      "learning_rate": 3.324246410474006e-06,
+      "loss": 0.3524,
+      "step": 2393
+    },
+    {
+      "epoch": 1.9447603574329815,
+      "grad_norm": 0.197356179356575,
+      "learning_rate": 3.319793507972454e-06,
+      "loss": 0.3725,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9455727051177905,
+      "grad_norm": 0.18508528172969818,
+      "learning_rate": 3.3153421072349024e-06,
+      "loss": 0.3469,
+      "step": 2395
+    },
+    {
+      "epoch": 1.9463850528025994,
+      "grad_norm": 0.19767791032791138,
+      "learning_rate": 3.310892212239999e-06,
+      "loss": 0.3755,
+      "step": 2396
+    },
+    {
+      "epoch": 1.9471974004874086,
+      "grad_norm": 0.21507628262043,
+      "learning_rate": 3.306443826965049e-06,
+      "loss": 0.3533,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9480097481722178,
+      "grad_norm": 0.20103462040424347,
+      "learning_rate": 3.3019969553860056e-06,
+      "loss": 0.3501,
+      "step": 2398
+    },
+    {
+      "epoch": 1.9488220958570268,
+      "grad_norm": 0.202636256814003,
+      "learning_rate": 3.2975516014774713e-06,
+      "loss": 0.3384,
+      "step": 2399
+    },
+    {
+      "epoch": 1.9496344435418358,
+      "grad_norm": 0.19370941817760468,
+      "learning_rate": 3.2931077692126912e-06,
+      "loss": 0.3677,
+      "step": 2400
+    },
+    {
+      "epoch": 1.950446791226645,
+      "grad_norm": 0.21495641767978668,
+      "learning_rate": 3.2886654625635507e-06,
+      "loss": 0.3773,
+      "step": 2401
+    },
+    {
+      "epoch": 1.9512591389114542,
+      "grad_norm": 0.1767066866159439,
+      "learning_rate": 3.284224685500571e-06,
+      "loss": 0.3597,
+      "step": 2402
+    },
+    {
+      "epoch": 1.9520714865962632,
+      "grad_norm": 0.2057410478591919,
+      "learning_rate": 3.2797854419929064e-06,
+      "loss": 0.361,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9528838342810722,
+      "grad_norm": 0.19295302033424377,
+      "learning_rate": 3.2753477360083417e-06,
+      "loss": 0.3679,
+      "step": 2404
+    },
+    {
+      "epoch": 1.9536961819658814,
+      "grad_norm": 0.1963021606206894,
+      "learning_rate": 3.2709115715132868e-06,
+      "loss": 0.3446,
+      "step": 2405
+    },
+    {
+      "epoch": 1.9545085296506906,
+      "grad_norm": 0.1936715841293335,
+      "learning_rate": 3.2664769524727712e-06,
+      "loss": 0.3403,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9553208773354998,
+      "grad_norm": 0.19580389559268951,
+      "learning_rate": 3.2620438828504473e-06,
+      "loss": 0.3716,
+      "step": 2407
+    },
+    {
+      "epoch": 1.9561332250203087,
+      "grad_norm": 0.2160162776708603,
+      "learning_rate": 3.25761236660858e-06,
+      "loss": 0.3782,
+      "step": 2408
+    },
+    {
+      "epoch": 1.9569455727051177,
+      "grad_norm": 0.22301678359508514,
+      "learning_rate": 3.2531824077080466e-06,
+      "loss": 0.3738,
+      "step": 2409
+    },
+    {
+      "epoch": 1.957757920389927,
+      "grad_norm": 0.21052102744579315,
+      "learning_rate": 3.2487540101083325e-06,
+      "loss": 0.3596,
+      "step": 2410
+    },
+    {
+      "epoch": 1.958570268074736,
+      "grad_norm": 0.18537986278533936,
+      "learning_rate": 3.244327177767527e-06,
+      "loss": 0.37,
+      "step": 2411
+    },
+    {
+      "epoch": 1.959382615759545,
+      "grad_norm": 0.19060613214969635,
+      "learning_rate": 3.23990191464232e-06,
+      "loss": 0.3673,
+      "step": 2412
+    },
+    {
+      "epoch": 1.960194963444354,
+      "grad_norm": 0.1806173324584961,
+      "learning_rate": 3.2354782246880016e-06,
+      "loss": 0.3714,
+      "step": 2413
+    },
+    {
+      "epoch": 1.9610073111291633,
+      "grad_norm": 0.18987645208835602,
+      "learning_rate": 3.2310561118584526e-06,
+      "loss": 0.3627,
+      "step": 2414
+    },
+    {
+      "epoch": 1.9618196588139725,
+      "grad_norm": 0.18514928221702576,
+      "learning_rate": 3.226635580106147e-06,
+      "loss": 0.3561,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9626320064987814,
+      "grad_norm": 0.20279461145401,
+      "learning_rate": 3.222216633382142e-06,
+      "loss": 0.3489,
+      "step": 2416
+    },
+    {
+      "epoch": 1.9634443541835904,
+      "grad_norm": 0.20018169283866882,
+      "learning_rate": 3.2177992756360814e-06,
+      "loss": 0.3697,
+      "step": 2417
+    },
+    {
+      "epoch": 1.9642567018683996,
+      "grad_norm": 0.19038128852844238,
+      "learning_rate": 3.2133835108161875e-06,
+      "loss": 0.3658,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9650690495532088,
+      "grad_norm": 0.22191911935806274,
+      "learning_rate": 3.2089693428692587e-06,
+      "loss": 0.392,
+      "step": 2419
+    },
+    {
+      "epoch": 1.965881397238018,
+      "grad_norm": 0.18225757777690887,
+      "learning_rate": 3.204556775740665e-06,
+      "loss": 0.357,
+      "step": 2420
+    },
+    {
+      "epoch": 1.966693744922827,
+      "grad_norm": 0.18354985117912292,
+      "learning_rate": 3.200145813374349e-06,
+      "loss": 0.345,
+      "step": 2421
+    },
+    {
+      "epoch": 1.967506092607636,
+      "grad_norm": 0.19686222076416016,
+      "learning_rate": 3.195736459712817e-06,
+      "loss": 0.3321,
+      "step": 2422
+    },
+    {
+      "epoch": 1.9683184402924452,
+      "grad_norm": 0.18914003670215607,
+      "learning_rate": 3.1913287186971353e-06,
+      "loss": 0.3562,
+      "step": 2423
+    },
+    {
+      "epoch": 1.9691307879772544,
+      "grad_norm": 0.20192477107048035,
+      "learning_rate": 3.186922594266932e-06,
+      "loss": 0.3704,
+      "step": 2424
+    },
+    {
+      "epoch": 1.9699431356620634,
+      "grad_norm": 0.18320918083190918,
+      "learning_rate": 3.1825180903603903e-06,
+      "loss": 0.3786,
+      "step": 2425
+    },
+    {
+      "epoch": 1.9707554833468723,
+      "grad_norm": 0.19555409252643585,
+      "learning_rate": 3.178115210914242e-06,
+      "loss": 0.3504,
+      "step": 2426
+    },
+    {
+      "epoch": 1.9715678310316815,
+      "grad_norm": 0.21619202196598053,
+      "learning_rate": 3.1737139598637688e-06,
+      "loss": 0.3582,
+      "step": 2427
+    },
+    {
+      "epoch": 1.9723801787164907,
+      "grad_norm": 0.19229483604431152,
+      "learning_rate": 3.1693143411427984e-06,
+      "loss": 0.3798,
+      "step": 2428
+    },
+    {
+      "epoch": 1.9731925264012997,
+      "grad_norm": 0.18061791360378265,
+      "learning_rate": 3.164916358683697e-06,
+      "loss": 0.36,
+      "step": 2429
+    },
+    {
+      "epoch": 1.9740048740861087,
+      "grad_norm": 0.18871212005615234,
+      "learning_rate": 3.160520016417369e-06,
+      "loss": 0.35,
+      "step": 2430
+    },
+    {
+      "epoch": 1.974817221770918,
+      "grad_norm": 0.18965497612953186,
+      "learning_rate": 3.1561253182732544e-06,
+      "loss": 0.341,
+      "step": 2431
+    },
+    {
+      "epoch": 1.975629569455727,
+      "grad_norm": 0.2057051658630371,
+      "learning_rate": 3.1517322681793215e-06,
+      "loss": 0.3577,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9764419171405363,
+      "grad_norm": 0.17941969633102417,
+      "learning_rate": 3.1473408700620665e-06,
+      "loss": 0.3562,
+      "step": 2433
+    },
+    {
+      "epoch": 1.9772542648253453,
+      "grad_norm": 0.19227257370948792,
+      "learning_rate": 3.1429511278465104e-06,
+      "loss": 0.3576,
+      "step": 2434
+    },
+    {
+      "epoch": 1.9780666125101543,
+      "grad_norm": 0.18741358816623688,
+      "learning_rate": 3.138563045456191e-06,
+      "loss": 0.3736,
+      "step": 2435
+    },
+    {
+      "epoch": 1.9788789601949635,
+      "grad_norm": 0.18073122203350067,
+      "learning_rate": 3.1341766268131657e-06,
+      "loss": 0.357,
+      "step": 2436
+    },
+    {
+      "epoch": 1.9796913078797727,
+      "grad_norm": 0.19586725533008575,
+      "learning_rate": 3.1297918758380035e-06,
+      "loss": 0.3705,
+      "step": 2437
+    },
+    {
+      "epoch": 1.9805036555645816,
+      "grad_norm": 0.19240979850292206,
+      "learning_rate": 3.1254087964497816e-06,
+      "loss": 0.3599,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9813160032493906,
+      "grad_norm": 0.18977375328540802,
+      "learning_rate": 3.121027392566086e-06,
+      "loss": 0.3544,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9821283509341998,
+      "grad_norm": 0.193256214261055,
+      "learning_rate": 3.1166476681030035e-06,
+      "loss": 0.3675,
+      "step": 2440
+    },
+    {
+      "epoch": 1.982940698619009,
+      "grad_norm": 0.2032395601272583,
+      "learning_rate": 3.1122696269751196e-06,
+      "loss": 0.3666,
+      "step": 2441
+    },
+    {
+      "epoch": 1.983753046303818,
+      "grad_norm": 0.1855606734752655,
+      "learning_rate": 3.1078932730955152e-06,
+      "loss": 0.3691,
+      "step": 2442
+    },
+    {
+      "epoch": 1.984565393988627,
+      "grad_norm": 0.18164251744747162,
+      "learning_rate": 3.103518610375764e-06,
+      "loss": 0.3763,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9853777416734362,
+      "grad_norm": 0.1909756362438202,
+      "learning_rate": 3.099145642725927e-06,
+      "loss": 0.3504,
+      "step": 2444
+    },
+    {
+      "epoch": 1.9861900893582454,
+      "grad_norm": 0.18413370847702026,
+      "learning_rate": 3.0947743740545523e-06,
+      "loss": 0.3738,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9870024370430546,
+      "grad_norm": 0.1860595941543579,
+      "learning_rate": 3.0904048082686655e-06,
+      "loss": 0.3423,
+      "step": 2446
+    },
+    {
+      "epoch": 1.9878147847278635,
+      "grad_norm": 0.19008679687976837,
+      "learning_rate": 3.0860369492737753e-06,
+      "loss": 0.3797,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9886271324126725,
+      "grad_norm": 0.19690118730068207,
+      "learning_rate": 3.0816708009738605e-06,
+      "loss": 0.3532,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9894394800974817,
+      "grad_norm": 0.18678411841392517,
+      "learning_rate": 3.0773063672713725e-06,
+      "loss": 0.3682,
+      "step": 2449
+    },
+    {
+      "epoch": 1.990251827782291,
+      "grad_norm": 0.1980656385421753,
+      "learning_rate": 3.0729436520672318e-06,
+      "loss": 0.365,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9910641754671,
+      "grad_norm": 0.1962713897228241,
+      "learning_rate": 3.0685826592608193e-06,
+      "loss": 0.3409,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9918765231519089,
+      "grad_norm": 0.19101202487945557,
+      "learning_rate": 3.0642233927499786e-06,
+      "loss": 0.3716,
+      "step": 2452
+    },
+    {
+      "epoch": 1.992688870836718,
+      "grad_norm": 0.18097829818725586,
+      "learning_rate": 3.0598658564310122e-06,
+      "loss": 0.3631,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9935012185215273,
+      "grad_norm": 0.1840217113494873,
+      "learning_rate": 3.055510054198675e-06,
+      "loss": 0.3447,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9943135662063363,
+      "grad_norm": 0.17797893285751343,
+      "learning_rate": 3.0511559899461684e-06,
+      "loss": 0.3324,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9951259138911455,
+      "grad_norm": 0.17646688222885132,
+      "learning_rate": 3.0468036675651447e-06,
+      "loss": 0.3892,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9959382615759544,
+      "grad_norm": 0.1951950043439865,
+      "learning_rate": 3.0424530909456974e-06,
+      "loss": 0.3643,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9967506092607636,
+      "grad_norm": 0.18400615453720093,
+      "learning_rate": 3.03810426397636e-06,
+      "loss": 0.3648,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9975629569455728,
+      "grad_norm": 0.19749762117862701,
+      "learning_rate": 3.033757190544102e-06,
+      "loss": 0.3516,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9983753046303818,
+      "grad_norm": 0.18275102972984314,
+      "learning_rate": 3.029411874534327e-06,
+      "loss": 0.352,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9991876523151908,
+      "grad_norm": 0.1868658810853958,
+      "learning_rate": 3.0250683198308656e-06,
+      "loss": 0.3607,
+      "step": 2461
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2018297165632248,
+      "learning_rate": 3.020726530315975e-06,
+      "loss": 0.3358,
+      "step": 2462
+    },
+    {
+      "epoch": 2.000812347684809,
+      "grad_norm": 0.2223242223262787,
+      "learning_rate": 3.016386509870335e-06,
+      "loss": 0.3316,
+      "step": 2463
+    },
+    {
+      "epoch": 2.0016246953696184,
+      "grad_norm": 0.18743713200092316,
+      "learning_rate": 3.0120482623730452e-06,
+      "loss": 0.3574,
+      "step": 2464
+    },
+    {
+      "epoch": 2.002437043054427,
+      "grad_norm": 0.19702546298503876,
+      "learning_rate": 3.0077117917016187e-06,
+      "loss": 0.3353,
+      "step": 2465
+    },
+    {
+      "epoch": 2.0032493907392364,
+      "grad_norm": 0.19529274106025696,
+      "learning_rate": 3.00337710173198e-06,
+      "loss": 0.3494,
+      "step": 2466
+    },
+    {
+      "epoch": 2.0040617384240456,
+      "grad_norm": 0.22567792236804962,
+      "learning_rate": 2.9990441963384657e-06,
+      "loss": 0.3316,
+      "step": 2467
+    },
+    {
+      "epoch": 2.0048740861088548,
+      "grad_norm": 0.19708938896656036,
+      "learning_rate": 2.9947130793938127e-06,
+      "loss": 0.3375,
+      "step": 2468
+    },
+    {
+      "epoch": 2.0056864337936635,
+      "grad_norm": 0.20258599519729614,
+      "learning_rate": 2.9903837547691634e-06,
+      "loss": 0.3603,
+      "step": 2469
+    },
+    {
+      "epoch": 2.0064987814784727,
+      "grad_norm": 0.20922255516052246,
+      "learning_rate": 2.986056226334054e-06,
+      "loss": 0.3451,
+      "step": 2470
+    },
+    {
+      "epoch": 2.007311129163282,
+      "grad_norm": 0.18125241994857788,
+      "learning_rate": 2.9817304979564206e-06,
+      "loss": 0.3349,
+      "step": 2471
+    },
+    {
+      "epoch": 2.008123476848091,
+      "grad_norm": 0.19832035899162292,
+      "learning_rate": 2.977406573502585e-06,
+      "loss": 0.3762,
+      "step": 2472
+    },
+    {
+      "epoch": 2.0089358245329,
+      "grad_norm": 0.185808464884758,
+      "learning_rate": 2.9730844568372598e-06,
+      "loss": 0.3139,
+      "step": 2473
+    },
+    {
+      "epoch": 2.009748172217709,
+      "grad_norm": 0.17618891596794128,
+      "learning_rate": 2.968764151823542e-06,
+      "loss": 0.3374,
+      "step": 2474
+    },
+    {
+      "epoch": 2.0105605199025183,
+      "grad_norm": 0.1855766922235489,
+      "learning_rate": 2.964445662322908e-06,
+      "loss": 0.3465,
+      "step": 2475
+    },
+    {
+      "epoch": 2.0113728675873275,
+      "grad_norm": 0.18088580667972565,
+      "learning_rate": 2.960128992195211e-06,
+      "loss": 0.3278,
+      "step": 2476
+    },
+    {
+      "epoch": 2.0121852152721367,
+      "grad_norm": 0.19273820519447327,
+      "learning_rate": 2.95581414529868e-06,
+      "loss": 0.3455,
+      "step": 2477
+    },
+    {
+      "epoch": 2.0129975629569454,
+      "grad_norm": 0.18665222823619843,
+      "learning_rate": 2.951501125489914e-06,
+      "loss": 0.3467,
+      "step": 2478
+    },
+    {
+      "epoch": 2.0138099106417546,
+      "grad_norm": 0.18336251378059387,
+      "learning_rate": 2.9471899366238777e-06,
+      "loss": 0.3347,
+      "step": 2479
+    },
+    {
+      "epoch": 2.014622258326564,
+      "grad_norm": 0.20175795257091522,
+      "learning_rate": 2.9428805825538993e-06,
+      "loss": 0.3495,
+      "step": 2480
+    },
+    {
+      "epoch": 2.015434606011373,
+      "grad_norm": 0.2023724466562271,
+      "learning_rate": 2.9385730671316687e-06,
+      "loss": 0.3227,
+      "step": 2481
+    },
+    {
+      "epoch": 2.016246953696182,
+      "grad_norm": 0.18277622759342194,
+      "learning_rate": 2.934267394207231e-06,
+      "loss": 0.3131,
+      "step": 2482
+    },
+    {
+      "epoch": 2.017059301380991,
+      "grad_norm": 0.19147224724292755,
+      "learning_rate": 2.929963567628985e-06,
+      "loss": 0.3401,
+      "step": 2483
+    },
+    {
+      "epoch": 2.0178716490658,
+      "grad_norm": 0.1918727606534958,
+      "learning_rate": 2.9256615912436803e-06,
+      "loss": 0.3415,
+      "step": 2484
+    },
+    {
+      "epoch": 2.0186839967506094,
+      "grad_norm": 0.18736647069454193,
+      "learning_rate": 2.9213614688964104e-06,
+      "loss": 0.3629,
+      "step": 2485
+    },
+    {
+      "epoch": 2.019496344435418,
+      "grad_norm": 0.201188862323761,
+      "learning_rate": 2.9170632044306137e-06,
+      "loss": 0.3403,
+      "step": 2486
+    },
+    {
+      "epoch": 2.0203086921202273,
+      "grad_norm": 0.1955653876066208,
+      "learning_rate": 2.9127668016880674e-06,
+      "loss": 0.3149,
+      "step": 2487
+    },
+    {
+      "epoch": 2.0211210398050365,
+      "grad_norm": 0.18354539573192596,
+      "learning_rate": 2.9084722645088825e-06,
+      "loss": 0.3568,
+      "step": 2488
+    },
+    {
+      "epoch": 2.0219333874898457,
+      "grad_norm": 0.18689025938510895,
+      "learning_rate": 2.904179596731509e-06,
+      "loss": 0.3277,
+      "step": 2489
+    },
+    {
+      "epoch": 2.022745735174655,
+      "grad_norm": 0.19269832968711853,
+      "learning_rate": 2.8998888021927146e-06,
+      "loss": 0.3235,
+      "step": 2490
+    },
+    {
+      "epoch": 2.0235580828594637,
+      "grad_norm": 0.1811206042766571,
+      "learning_rate": 2.895599884727608e-06,
+      "loss": 0.3328,
+      "step": 2491
+    },
+    {
+      "epoch": 2.024370430544273,
+      "grad_norm": 0.18936234712600708,
+      "learning_rate": 2.8913128481696067e-06,
+      "loss": 0.3336,
+      "step": 2492
+    },
+    {
+      "epoch": 2.025182778229082,
+      "grad_norm": 0.19612517952919006,
+      "learning_rate": 2.8870276963504562e-06,
+      "loss": 0.3287,
+      "step": 2493
+    },
+    {
+      "epoch": 2.0259951259138913,
+      "grad_norm": 0.1961199939250946,
+      "learning_rate": 2.8827444331002087e-06,
+      "loss": 0.3494,
+      "step": 2494
+    },
+    {
+      "epoch": 2.0268074735987,
+      "grad_norm": 0.1858479082584381,
+      "learning_rate": 2.8784630622472376e-06,
+      "loss": 0.3484,
+      "step": 2495
+    },
+    {
+      "epoch": 2.0276198212835093,
+      "grad_norm": 0.19146062433719635,
+      "learning_rate": 2.8741835876182165e-06,
+      "loss": 0.3293,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0284321689683185,
+      "grad_norm": 0.19271667301654816,
+      "learning_rate": 2.8699060130381334e-06,
+      "loss": 0.3255,
+      "step": 2497
+    },
+    {
+      "epoch": 2.0292445166531277,
+      "grad_norm": 0.19364219903945923,
+      "learning_rate": 2.8656303423302678e-06,
+      "loss": 0.3587,
+      "step": 2498
+    },
+    {
+      "epoch": 2.0300568643379364,
+      "grad_norm": 0.18399131298065186,
+      "learning_rate": 2.861356579316204e-06,
+      "loss": 0.3276,
+      "step": 2499
+    },
+    {
+      "epoch": 2.0308692120227456,
+      "grad_norm": 0.1925853192806244,
+      "learning_rate": 2.857084727815823e-06,
+      "loss": 0.3477,
+      "step": 2500
+    },
+    {
+      "epoch": 2.031681559707555,
+      "grad_norm": 0.18710607290267944,
+      "learning_rate": 2.8528147916472895e-06,
+      "loss": 0.3081,
+      "step": 2501
+    },
+    {
+      "epoch": 2.032493907392364,
+      "grad_norm": 0.200908362865448,
+      "learning_rate": 2.848546774627064e-06,
+      "loss": 0.3369,
+      "step": 2502
+    },
+    {
+      "epoch": 2.033306255077173,
+      "grad_norm": 0.18102343380451202,
+      "learning_rate": 2.8442806805698857e-06,
+      "loss": 0.3445,
+      "step": 2503
+    },
+    {
+      "epoch": 2.034118602761982,
+      "grad_norm": 0.19128885865211487,
+      "learning_rate": 2.840016513288782e-06,
+      "loss": 0.3344,
+      "step": 2504
+    },
+    {
+      "epoch": 2.034930950446791,
+      "grad_norm": 0.20118798315525055,
+      "learning_rate": 2.8357542765950493e-06,
+      "loss": 0.3385,
+      "step": 2505
+    },
+    {
+      "epoch": 2.0357432981316004,
+      "grad_norm": 0.20980440080165863,
+      "learning_rate": 2.8314939742982673e-06,
+      "loss": 0.3223,
+      "step": 2506
+    },
+    {
+      "epoch": 2.0365556458164096,
+      "grad_norm": 0.20583616197109222,
+      "learning_rate": 2.8272356102062792e-06,
+      "loss": 0.3554,
+      "step": 2507
+    },
+    {
+      "epoch": 2.0373679935012183,
+      "grad_norm": 0.19419552385807037,
+      "learning_rate": 2.8229791881252023e-06,
+      "loss": 0.336,
+      "step": 2508
+    },
+    {
+      "epoch": 2.0381803411860275,
+      "grad_norm": 0.18173423409461975,
+      "learning_rate": 2.818724711859412e-06,
+      "loss": 0.3212,
+      "step": 2509
+    },
+    {
+      "epoch": 2.0389926888708367,
+      "grad_norm": 0.19776740670204163,
+      "learning_rate": 2.814472185211551e-06,
+      "loss": 0.3515,
+      "step": 2510
+    },
+    {
+      "epoch": 2.039805036555646,
+      "grad_norm": 0.19620846211910248,
+      "learning_rate": 2.810221611982512e-06,
+      "loss": 0.3419,
+      "step": 2511
+    },
+    {
+      "epoch": 2.0406173842404547,
+      "grad_norm": 0.2454618513584137,
+      "learning_rate": 2.805972995971451e-06,
+      "loss": 0.3378,
+      "step": 2512
+    },
+    {
+      "epoch": 2.041429731925264,
+      "grad_norm": 0.1863708347082138,
+      "learning_rate": 2.8017263409757646e-06,
+      "loss": 0.3666,
+      "step": 2513
+    },
+    {
+      "epoch": 2.042242079610073,
+      "grad_norm": 0.1914231777191162,
+      "learning_rate": 2.7974816507911047e-06,
+      "loss": 0.3396,
+      "step": 2514
+    },
+    {
+      "epoch": 2.0430544272948823,
+      "grad_norm": 0.18796302378177643,
+      "learning_rate": 2.7932389292113604e-06,
+      "loss": 0.3404,
+      "step": 2515
+    },
+    {
+      "epoch": 2.0438667749796915,
+      "grad_norm": 0.187236949801445,
+      "learning_rate": 2.788998180028668e-06,
+      "loss": 0.3421,
+      "step": 2516
+    },
+    {
+      "epoch": 2.0446791226645002,
+      "grad_norm": 0.191828191280365,
+      "learning_rate": 2.784759407033398e-06,
+      "loss": 0.3576,
+      "step": 2517
+    },
+    {
+      "epoch": 2.0454914703493094,
+      "grad_norm": 0.1981503963470459,
+      "learning_rate": 2.7805226140141505e-06,
+      "loss": 0.363,
+      "step": 2518
+    },
+    {
+      "epoch": 2.0463038180341186,
+      "grad_norm": 0.20514748990535736,
+      "learning_rate": 2.776287804757765e-06,
+      "loss": 0.3568,
+      "step": 2519
+    },
+    {
+      "epoch": 2.047116165718928,
+      "grad_norm": 0.19908776879310608,
+      "learning_rate": 2.772054983049296e-06,
+      "loss": 0.3485,
+      "step": 2520
+    },
+    {
+      "epoch": 2.0479285134037366,
+      "grad_norm": 0.19092464447021484,
+      "learning_rate": 2.7678241526720338e-06,
+      "loss": 0.333,
+      "step": 2521
+    },
+    {
+      "epoch": 2.048740861088546,
+      "grad_norm": 0.19402703642845154,
+      "learning_rate": 2.7635953174074787e-06,
+      "loss": 0.343,
+      "step": 2522
+    },
+    {
+      "epoch": 2.049553208773355,
+      "grad_norm": 0.19679027795791626,
+      "learning_rate": 2.759368481035356e-06,
+      "loss": 0.3741,
+      "step": 2523
+    },
+    {
+      "epoch": 2.050365556458164,
+      "grad_norm": 0.19457149505615234,
+      "learning_rate": 2.755143647333597e-06,
+      "loss": 0.3381,
+      "step": 2524
+    },
+    {
+      "epoch": 2.0511779041429734,
+      "grad_norm": 0.20557363331317902,
+      "learning_rate": 2.7509208200783505e-06,
+      "loss": 0.3481,
+      "step": 2525
+    },
+    {
+      "epoch": 2.051990251827782,
+      "grad_norm": 0.18139460682868958,
+      "learning_rate": 2.746700003043964e-06,
+      "loss": 0.3481,
+      "step": 2526
+    },
+    {
+      "epoch": 2.0528025995125914,
+      "grad_norm": 0.17667388916015625,
+      "learning_rate": 2.7424812000029955e-06,
+      "loss": 0.3362,
+      "step": 2527
+    },
+    {
+      "epoch": 2.0536149471974006,
+      "grad_norm": 0.20645713806152344,
+      "learning_rate": 2.738264414726196e-06,
+      "loss": 0.3572,
+      "step": 2528
+    },
+    {
+      "epoch": 2.0544272948822098,
+      "grad_norm": 0.194272980093956,
+      "learning_rate": 2.73404965098252e-06,
+      "loss": 0.3459,
+      "step": 2529
+    },
+    {
+      "epoch": 2.0552396425670185,
+      "grad_norm": 0.1812710464000702,
+      "learning_rate": 2.7298369125391112e-06,
+      "loss": 0.352,
+      "step": 2530
+    },
+    {
+      "epoch": 2.0560519902518277,
+      "grad_norm": 0.18473777174949646,
+      "learning_rate": 2.7256262031613044e-06,
+      "loss": 0.3405,
+      "step": 2531
+    },
+    {
+      "epoch": 2.056864337936637,
+      "grad_norm": 0.19355922937393188,
+      "learning_rate": 2.7214175266126184e-06,
+      "loss": 0.3619,
+      "step": 2532
+    },
+    {
+      "epoch": 2.057676685621446,
+      "grad_norm": 0.18142838776111603,
+      "learning_rate": 2.7172108866547593e-06,
+      "loss": 0.325,
+      "step": 2533
+    },
+    {
+      "epoch": 2.058489033306255,
+      "grad_norm": 0.17555084824562073,
+      "learning_rate": 2.7130062870476072e-06,
+      "loss": 0.3341,
+      "step": 2534
+    },
+    {
+      "epoch": 2.059301380991064,
+      "grad_norm": 0.17836911976337433,
+      "learning_rate": 2.7088037315492254e-06,
+      "loss": 0.3503,
+      "step": 2535
+    },
+    {
+      "epoch": 2.0601137286758733,
+      "grad_norm": 0.17953598499298096,
+      "learning_rate": 2.7046032239158428e-06,
+      "loss": 0.3434,
+      "step": 2536
+    },
+    {
+      "epoch": 2.0609260763606825,
+      "grad_norm": 0.19661524891853333,
+      "learning_rate": 2.7004047679018652e-06,
+      "loss": 0.343,
+      "step": 2537
+    },
+    {
+      "epoch": 2.0617384240454917,
+      "grad_norm": 0.2072705775499344,
+      "learning_rate": 2.6962083672598584e-06,
+      "loss": 0.3427,
+      "step": 2538
+    },
+    {
+      "epoch": 2.0625507717303004,
+      "grad_norm": 0.17464444041252136,
+      "learning_rate": 2.692014025740557e-06,
+      "loss": 0.3305,
+      "step": 2539
+    },
+    {
+      "epoch": 2.0633631194151096,
+      "grad_norm": 0.18838562071323395,
+      "learning_rate": 2.687821747092848e-06,
+      "loss": 0.3609,
+      "step": 2540
+    },
+    {
+      "epoch": 2.064175467099919,
+      "grad_norm": 0.20013700425624847,
+      "learning_rate": 2.683631535063783e-06,
+      "loss": 0.3628,
+      "step": 2541
+    },
+    {
+      "epoch": 2.064987814784728,
+      "grad_norm": 0.18039432168006897,
+      "learning_rate": 2.679443393398558e-06,
+      "loss": 0.3344,
+      "step": 2542
+    },
+    {
+      "epoch": 2.065800162469537,
+      "grad_norm": 0.19316977262496948,
+      "learning_rate": 2.6752573258405245e-06,
+      "loss": 0.3534,
+      "step": 2543
+    },
+    {
+      "epoch": 2.066612510154346,
+      "grad_norm": 0.17717351019382477,
+      "learning_rate": 2.6710733361311812e-06,
+      "loss": 0.3464,
+      "step": 2544
+    },
+    {
+      "epoch": 2.067424857839155,
+      "grad_norm": 0.18476805090904236,
+      "learning_rate": 2.6668914280101615e-06,
+      "loss": 0.3336,
+      "step": 2545
+    },
+    {
+      "epoch": 2.0682372055239644,
+      "grad_norm": 0.18112051486968994,
+      "learning_rate": 2.662711605215248e-06,
+      "loss": 0.3278,
+      "step": 2546
+    },
+    {
+      "epoch": 2.069049553208773,
+      "grad_norm": 0.17877787351608276,
+      "learning_rate": 2.658533871482351e-06,
+      "loss": 0.3264,
+      "step": 2547
+    },
+    {
+      "epoch": 2.0698619008935824,
+      "grad_norm": 0.17650224268436432,
+      "learning_rate": 2.6543582305455217e-06,
+      "loss": 0.3272,
+      "step": 2548
+    },
+    {
+      "epoch": 2.0706742485783916,
+      "grad_norm": 0.20429790019989014,
+      "learning_rate": 2.650184686136932e-06,
+      "loss": 0.3337,
+      "step": 2549
+    },
+    {
+      "epoch": 2.0714865962632008,
+      "grad_norm": 0.19406256079673767,
+      "learning_rate": 2.6460132419868893e-06,
+      "loss": 0.3588,
+      "step": 2550
+    },
+    {
+      "epoch": 2.07229894394801,
+      "grad_norm": 0.1869717389345169,
+      "learning_rate": 2.641843901823815e-06,
+      "loss": 0.3415,
+      "step": 2551
+    },
+    {
+      "epoch": 2.0731112916328187,
+      "grad_norm": 0.18785248696804047,
+      "learning_rate": 2.637676669374258e-06,
+      "loss": 0.3446,
+      "step": 2552
+    },
+    {
+      "epoch": 2.073923639317628,
+      "grad_norm": 0.17342785000801086,
+      "learning_rate": 2.6335115483628747e-06,
+      "loss": 0.3525,
+      "step": 2553
+    },
+    {
+      "epoch": 2.074735987002437,
+      "grad_norm": 0.18115387856960297,
+      "learning_rate": 2.629348542512443e-06,
+      "loss": 0.3656,
+      "step": 2554
+    },
+    {
+      "epoch": 2.0755483346872463,
+      "grad_norm": 0.21591994166374207,
+      "learning_rate": 2.625187655543844e-06,
+      "loss": 0.3298,
+      "step": 2555
+    },
+    {
+      "epoch": 2.076360682372055,
+      "grad_norm": 0.17302779853343964,
+      "learning_rate": 2.6210288911760684e-06,
+      "loss": 0.3321,
+      "step": 2556
+    },
+    {
+      "epoch": 2.0771730300568643,
+      "grad_norm": 0.19388459622859955,
+      "learning_rate": 2.616872253126207e-06,
+      "loss": 0.3499,
+      "step": 2557
+    },
+    {
+      "epoch": 2.0779853777416735,
+      "grad_norm": 0.18613405525684357,
+      "learning_rate": 2.612717745109453e-06,
+      "loss": 0.3274,
+      "step": 2558
+    },
+    {
+      "epoch": 2.0787977254264827,
+      "grad_norm": 0.17904594540596008,
+      "learning_rate": 2.6085653708390936e-06,
+      "loss": 0.3507,
+      "step": 2559
+    },
+    {
+      "epoch": 2.0796100731112914,
+      "grad_norm": 0.19837471842765808,
+      "learning_rate": 2.604415134026509e-06,
+      "loss": 0.3529,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0804224207961006,
+      "grad_norm": 0.19224460422992706,
+      "learning_rate": 2.6002670383811724e-06,
+      "loss": 0.3176,
+      "step": 2561
+    },
+    {
+      "epoch": 2.08123476848091,
+      "grad_norm": 0.19925154745578766,
+      "learning_rate": 2.596121087610636e-06,
+      "loss": 0.3281,
+      "step": 2562
+    },
+    {
+      "epoch": 2.082047116165719,
+      "grad_norm": 0.1904037892818451,
+      "learning_rate": 2.591977285420545e-06,
+      "loss": 0.3317,
+      "step": 2563
+    },
+    {
+      "epoch": 2.0828594638505282,
+      "grad_norm": 0.1858888864517212,
+      "learning_rate": 2.587835635514613e-06,
+      "loss": 0.3498,
+      "step": 2564
+    },
+    {
+      "epoch": 2.083671811535337,
+      "grad_norm": 0.18410108983516693,
+      "learning_rate": 2.5836961415946394e-06,
+      "loss": 0.3213,
+      "step": 2565
+    },
+    {
+      "epoch": 2.084484159220146,
+      "grad_norm": 0.1948295682668686,
+      "learning_rate": 2.579558807360489e-06,
+      "loss": 0.3641,
+      "step": 2566
+    },
+    {
+      "epoch": 2.0852965069049554,
+      "grad_norm": 0.19511108100414276,
+      "learning_rate": 2.5754236365101024e-06,
+      "loss": 0.3509,
+      "step": 2567
+    },
+    {
+      "epoch": 2.0861088545897646,
+      "grad_norm": 0.1823418289422989,
+      "learning_rate": 2.5712906327394836e-06,
+      "loss": 0.3444,
+      "step": 2568
+    },
+    {
+      "epoch": 2.0869212022745733,
+      "grad_norm": 0.19730766117572784,
+      "learning_rate": 2.5671597997427023e-06,
+      "loss": 0.3381,
+      "step": 2569
+    },
+    {
+      "epoch": 2.0877335499593825,
+      "grad_norm": 0.20157241821289062,
+      "learning_rate": 2.5630311412118815e-06,
+      "loss": 0.3353,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0885458976441917,
+      "grad_norm": 0.18424846231937408,
+      "learning_rate": 2.558904660837209e-06,
+      "loss": 0.3554,
+      "step": 2571
+    },
+    {
+      "epoch": 2.089358245329001,
+      "grad_norm": 0.19493387639522552,
+      "learning_rate": 2.554780362306918e-06,
+      "loss": 0.3277,
+      "step": 2572
+    },
+    {
+      "epoch": 2.0901705930138097,
+      "grad_norm": 0.17826463282108307,
+      "learning_rate": 2.5506582493072994e-06,
+      "loss": 0.3376,
+      "step": 2573
+    },
+    {
+      "epoch": 2.090982940698619,
+      "grad_norm": 0.18388231098651886,
+      "learning_rate": 2.546538325522682e-06,
+      "loss": 0.3451,
+      "step": 2574
+    },
+    {
+      "epoch": 2.091795288383428,
+      "grad_norm": 0.19287042319774628,
+      "learning_rate": 2.5424205946354475e-06,
+      "loss": 0.3461,
+      "step": 2575
+    },
+    {
+      "epoch": 2.0926076360682373,
+      "grad_norm": 0.18980476260185242,
+      "learning_rate": 2.5383050603260074e-06,
+      "loss": 0.3565,
+      "step": 2576
+    },
+    {
+      "epoch": 2.0934199837530465,
+      "grad_norm": 0.18670912086963654,
+      "learning_rate": 2.53419172627282e-06,
+      "loss": 0.3428,
+      "step": 2577
+    },
+    {
+      "epoch": 2.0942323314378553,
+      "grad_norm": 0.2061207890510559,
+      "learning_rate": 2.5300805961523677e-06,
+      "loss": 0.377,
+      "step": 2578
+    },
+    {
+      "epoch": 2.0950446791226645,
+      "grad_norm": 0.18906429409980774,
+      "learning_rate": 2.525971673639172e-06,
+      "loss": 0.3395,
+      "step": 2579
+    },
+    {
+      "epoch": 2.0958570268074737,
+      "grad_norm": 0.17904910445213318,
+      "learning_rate": 2.5218649624057717e-06,
+      "loss": 0.3345,
+      "step": 2580
+    },
+    {
+      "epoch": 2.096669374492283,
+      "grad_norm": 0.19162659347057343,
+      "learning_rate": 2.517760466122739e-06,
+      "loss": 0.3332,
+      "step": 2581
+    },
+    {
+      "epoch": 2.0974817221770916,
+      "grad_norm": 0.19197677075862885,
+      "learning_rate": 2.513658188458657e-06,
+      "loss": 0.3629,
+      "step": 2582
+    },
+    {
+      "epoch": 2.098294069861901,
+      "grad_norm": 0.19245107471942902,
+      "learning_rate": 2.5095581330801345e-06,
+      "loss": 0.3258,
+      "step": 2583
+    },
+    {
+      "epoch": 2.09910641754671,
+      "grad_norm": 0.16872963309288025,
+      "learning_rate": 2.5054603036517856e-06,
+      "loss": 0.366,
+      "step": 2584
+    },
+    {
+      "epoch": 2.099918765231519,
+      "grad_norm": 0.1753062605857849,
+      "learning_rate": 2.5013647038362437e-06,
+      "loss": 0.3366,
+      "step": 2585
+    },
+    {
+      "epoch": 2.1007311129163284,
+      "grad_norm": 0.5546891093254089,
+      "learning_rate": 2.4972713372941406e-06,
+      "loss": 0.3297,
+      "step": 2586
+    },
+    {
+      "epoch": 2.101543460601137,
+      "grad_norm": 0.18364566564559937,
+      "learning_rate": 2.4931802076841176e-06,
+      "loss": 0.3271,
+      "step": 2587
+    },
+    {
+      "epoch": 2.1023558082859464,
+      "grad_norm": 0.18113116919994354,
+      "learning_rate": 2.4890913186628178e-06,
+      "loss": 0.3365,
+      "step": 2588
+    },
+    {
+      "epoch": 2.1031681559707556,
+      "grad_norm": 0.1849890798330307,
+      "learning_rate": 2.4850046738848752e-06,
+      "loss": 0.3368,
+      "step": 2589
+    },
+    {
+      "epoch": 2.1039805036555648,
+      "grad_norm": 0.1745717078447342,
+      "learning_rate": 2.480920277002926e-06,
+      "loss": 0.3423,
+      "step": 2590
+    },
+    {
+      "epoch": 2.1047928513403735,
+      "grad_norm": 0.20073281228542328,
+      "learning_rate": 2.4768381316675883e-06,
+      "loss": 0.3607,
+      "step": 2591
+    },
+    {
+      "epoch": 2.1056051990251827,
+      "grad_norm": 0.17096105217933655,
+      "learning_rate": 2.472758241527477e-06,
+      "loss": 0.3405,
+      "step": 2592
+    },
+    {
+      "epoch": 2.106417546709992,
+      "grad_norm": 0.20449016988277435,
+      "learning_rate": 2.468680610229182e-06,
+      "loss": 0.3514,
+      "step": 2593
+    },
+    {
+      "epoch": 2.107229894394801,
+      "grad_norm": 0.19506649672985077,
+      "learning_rate": 2.4646052414172844e-06,
+      "loss": 0.3488,
+      "step": 2594
+    },
+    {
+      "epoch": 2.10804224207961,
+      "grad_norm": 0.18447697162628174,
+      "learning_rate": 2.4605321387343334e-06,
+      "loss": 0.3218,
+      "step": 2595
+    },
+    {
+      "epoch": 2.108854589764419,
+      "grad_norm": 0.1863415688276291,
+      "learning_rate": 2.4564613058208603e-06,
+      "loss": 0.3051,
+      "step": 2596
+    },
+    {
+      "epoch": 2.1096669374492283,
+      "grad_norm": 0.18662725389003754,
+      "learning_rate": 2.452392746315361e-06,
+      "loss": 0.3345,
+      "step": 2597
+    },
+    {
+      "epoch": 2.1104792851340375,
+      "grad_norm": 0.19308018684387207,
+      "learning_rate": 2.4483264638543076e-06,
+      "loss": 0.3543,
+      "step": 2598
+    },
+    {
+      "epoch": 2.1112916328188467,
+      "grad_norm": 0.19195525348186493,
+      "learning_rate": 2.4442624620721274e-06,
+      "loss": 0.3223,
+      "step": 2599
+    },
+    {
+      "epoch": 2.1121039805036554,
+      "grad_norm": 0.19035731256008148,
+      "learning_rate": 2.440200744601218e-06,
+      "loss": 0.3299,
+      "step": 2600
+    },
+    {
+      "epoch": 2.1129163281884646,
+      "grad_norm": 0.19552075862884521,
+      "learning_rate": 2.4361413150719287e-06,
+      "loss": 0.3346,
+      "step": 2601
+    },
+    {
+      "epoch": 2.113728675873274,
+      "grad_norm": 0.20039652287960052,
+      "learning_rate": 2.432084177112569e-06,
+      "loss": 0.3462,
+      "step": 2602
+    },
+    {
+      "epoch": 2.114541023558083,
+      "grad_norm": 0.18460877239704132,
+      "learning_rate": 2.4280293343493954e-06,
+      "loss": 0.3161,
+      "step": 2603
+    },
+    {
+      "epoch": 2.115353371242892,
+      "grad_norm": 0.1791103333234787,
+      "learning_rate": 2.423976790406616e-06,
+      "loss": 0.3399,
+      "step": 2604
+    },
+    {
+      "epoch": 2.116165718927701,
+      "grad_norm": 0.19011233747005463,
+      "learning_rate": 2.4199265489063846e-06,
+      "loss": 0.3334,
+      "step": 2605
+    },
+    {
+      "epoch": 2.11697806661251,
+      "grad_norm": 0.17156776785850525,
+      "learning_rate": 2.4158786134687966e-06,
+      "loss": 0.3235,
+      "step": 2606
+    },
+    {
+      "epoch": 2.1177904142973194,
+      "grad_norm": 0.19191782176494598,
+      "learning_rate": 2.4118329877118836e-06,
+      "loss": 0.3222,
+      "step": 2607
+    },
+    {
+      "epoch": 2.118602761982128,
+      "grad_norm": 0.17937642335891724,
+      "learning_rate": 2.407789675251617e-06,
+      "loss": 0.3451,
+      "step": 2608
+    },
+    {
+      "epoch": 2.1194151096669374,
+      "grad_norm": 0.18976730108261108,
+      "learning_rate": 2.403748679701896e-06,
+      "loss": 0.3305,
+      "step": 2609
+    },
+    {
+      "epoch": 2.1202274573517466,
+      "grad_norm": 0.1787577122449875,
+      "learning_rate": 2.3997100046745554e-06,
+      "loss": 0.3354,
+      "step": 2610
+    },
+    {
+      "epoch": 2.1210398050365558,
+      "grad_norm": 0.21625161170959473,
+      "learning_rate": 2.3956736537793468e-06,
+      "loss": 0.3264,
+      "step": 2611
+    },
+    {
+      "epoch": 2.121852152721365,
+      "grad_norm": 0.1897812783718109,
+      "learning_rate": 2.3916396306239525e-06,
+      "loss": 0.3173,
+      "step": 2612
+    },
+    {
+      "epoch": 2.1226645004061737,
+      "grad_norm": 0.19838854670524597,
+      "learning_rate": 2.3876079388139733e-06,
+      "loss": 0.3527,
+      "step": 2613
+    },
+    {
+      "epoch": 2.123476848090983,
+      "grad_norm": 0.19564975798130035,
+      "learning_rate": 2.38357858195292e-06,
+      "loss": 0.3323,
+      "step": 2614
+    },
+    {
+      "epoch": 2.124289195775792,
+      "grad_norm": 0.18761178851127625,
+      "learning_rate": 2.3795515636422257e-06,
+      "loss": 0.3309,
+      "step": 2615
+    },
+    {
+      "epoch": 2.1251015434606013,
+      "grad_norm": 0.20530404150485992,
+      "learning_rate": 2.3755268874812237e-06,
+      "loss": 0.3505,
+      "step": 2616
+    },
+    {
+      "epoch": 2.12591389114541,
+      "grad_norm": 0.1941995471715927,
+      "learning_rate": 2.371504557067163e-06,
+      "loss": 0.334,
+      "step": 2617
+    },
+    {
+      "epoch": 2.1267262388302193,
+      "grad_norm": 0.19202159345149994,
+      "learning_rate": 2.367484575995187e-06,
+      "loss": 0.3545,
+      "step": 2618
+    },
+    {
+      "epoch": 2.1275385865150285,
+      "grad_norm": 0.18702876567840576,
+      "learning_rate": 2.3634669478583484e-06,
+      "loss": 0.3415,
+      "step": 2619
+    },
+    {
+      "epoch": 2.1283509341998377,
+      "grad_norm": 0.20608265697956085,
+      "learning_rate": 2.3594516762475882e-06,
+      "loss": 0.3304,
+      "step": 2620
+    },
+    {
+      "epoch": 2.1291632818846464,
+      "grad_norm": 0.19020918011665344,
+      "learning_rate": 2.3554387647517497e-06,
+      "loss": 0.3549,
+      "step": 2621
+    },
+    {
+      "epoch": 2.1299756295694556,
+      "grad_norm": 0.18843194842338562,
+      "learning_rate": 2.3514282169575575e-06,
+      "loss": 0.3524,
+      "step": 2622
+    },
+    {
+      "epoch": 2.130787977254265,
+      "grad_norm": 0.19501042366027832,
+      "learning_rate": 2.3474200364496326e-06,
+      "loss": 0.3202,
+      "step": 2623
+    },
+    {
+      "epoch": 2.131600324939074,
+      "grad_norm": 0.1910451352596283,
+      "learning_rate": 2.343414226810472e-06,
+      "loss": 0.3404,
+      "step": 2624
+    },
+    {
+      "epoch": 2.1324126726238832,
+      "grad_norm": 0.18781183660030365,
+      "learning_rate": 2.3394107916204617e-06,
+      "loss": 0.3306,
+      "step": 2625
+    },
+    {
+      "epoch": 2.133225020308692,
+      "grad_norm": 0.18810205161571503,
+      "learning_rate": 2.3354097344578565e-06,
+      "loss": 0.3431,
+      "step": 2626
+    },
+    {
+      "epoch": 2.134037367993501,
+      "grad_norm": 0.1869698166847229,
+      "learning_rate": 2.3314110588987946e-06,
+      "loss": 0.3349,
+      "step": 2627
+    },
+    {
+      "epoch": 2.1348497156783104,
+      "grad_norm": 0.18929243087768555,
+      "learning_rate": 2.327414768517278e-06,
+      "loss": 0.3565,
+      "step": 2628
+    },
+    {
+      "epoch": 2.1356620633631196,
+      "grad_norm": 0.1803095042705536,
+      "learning_rate": 2.3234208668851843e-06,
+      "loss": 0.327,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1364744110479283,
+      "grad_norm": 0.17848876118659973,
+      "learning_rate": 2.3194293575722464e-06,
+      "loss": 0.3264,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1372867587327375,
+      "grad_norm": 0.2121448814868927,
+      "learning_rate": 2.3154402441460685e-06,
+      "loss": 0.3304,
+      "step": 2631
+    },
+    {
+      "epoch": 2.1380991064175467,
+      "grad_norm": 0.20267386734485626,
+      "learning_rate": 2.311453530172109e-06,
+      "loss": 0.3346,
+      "step": 2632
+    },
+    {
+      "epoch": 2.138911454102356,
+      "grad_norm": 0.19933384656906128,
+      "learning_rate": 2.3074692192136803e-06,
+      "loss": 0.3291,
+      "step": 2633
+    },
+    {
+      "epoch": 2.1397238017871647,
+      "grad_norm": 0.18126361072063446,
+      "learning_rate": 2.303487314831952e-06,
+      "loss": 0.3429,
+      "step": 2634
+    },
+    {
+      "epoch": 2.140536149471974,
+      "grad_norm": 0.19297178089618683,
+      "learning_rate": 2.2995078205859344e-06,
+      "loss": 0.3342,
+      "step": 2635
+    },
+    {
+      "epoch": 2.141348497156783,
+      "grad_norm": 0.20260757207870483,
+      "learning_rate": 2.295530740032494e-06,
+      "loss": 0.337,
+      "step": 2636
+    },
+    {
+      "epoch": 2.1421608448415923,
+      "grad_norm": 0.17624327540397644,
+      "learning_rate": 2.2915560767263296e-06,
+      "loss": 0.3349,
+      "step": 2637
+    },
+    {
+      "epoch": 2.1429731925264015,
+      "grad_norm": 0.18880315124988556,
+      "learning_rate": 2.287583834219989e-06,
+      "loss": 0.3328,
+      "step": 2638
+    },
+    {
+      "epoch": 2.1437855402112103,
+      "grad_norm": 0.21348613500595093,
+      "learning_rate": 2.2836140160638477e-06,
+      "loss": 0.3537,
+      "step": 2639
+    },
+    {
+      "epoch": 2.1445978878960195,
+      "grad_norm": 0.18454164266586304,
+      "learning_rate": 2.2796466258061223e-06,
+      "loss": 0.3298,
+      "step": 2640
+    },
+    {
+      "epoch": 2.1454102355808287,
+      "grad_norm": 0.18294665217399597,
+      "learning_rate": 2.2756816669928516e-06,
+      "loss": 0.3566,
+      "step": 2641
+    },
+    {
+      "epoch": 2.146222583265638,
+      "grad_norm": 0.19477292895317078,
+      "learning_rate": 2.2717191431679085e-06,
+      "loss": 0.3314,
+      "step": 2642
+    },
+    {
+      "epoch": 2.1470349309504466,
+      "grad_norm": 0.18652811646461487,
+      "learning_rate": 2.2677590578729815e-06,
+      "loss": 0.3393,
+      "step": 2643
+    },
+    {
+      "epoch": 2.147847278635256,
+      "grad_norm": 0.19134627282619476,
+      "learning_rate": 2.2638014146475894e-06,
+      "loss": 0.3194,
+      "step": 2644
+    },
+    {
+      "epoch": 2.148659626320065,
+      "grad_norm": 0.19269698858261108,
+      "learning_rate": 2.2598462170290598e-06,
+      "loss": 0.3392,
+      "step": 2645
+    },
+    {
+      "epoch": 2.149471974004874,
+      "grad_norm": 0.18735498189926147,
+      "learning_rate": 2.25589346855254e-06,
+      "loss": 0.3675,
+      "step": 2646
+    },
+    {
+      "epoch": 2.150284321689683,
+      "grad_norm": 0.2042829841375351,
+      "learning_rate": 2.2519431727509823e-06,
+      "loss": 0.3524,
+      "step": 2647
+    },
+    {
+      "epoch": 2.151096669374492,
+      "grad_norm": 0.18242736160755157,
+      "learning_rate": 2.2479953331551534e-06,
+      "loss": 0.3101,
+      "step": 2648
+    },
+    {
+      "epoch": 2.1519090170593014,
+      "grad_norm": 0.19145604968070984,
+      "learning_rate": 2.2440499532936196e-06,
+      "loss": 0.3111,
+      "step": 2649
+    },
+    {
+      "epoch": 2.1527213647441106,
+      "grad_norm": 0.2023184448480606,
+      "learning_rate": 2.240107036692753e-06,
+      "loss": 0.3378,
+      "step": 2650
+    },
+    {
+      "epoch": 2.1535337124289198,
+      "grad_norm": 0.18131576478481293,
+      "learning_rate": 2.2361665868767185e-06,
+      "loss": 0.3428,
+      "step": 2651
+    },
+    {
+      "epoch": 2.1543460601137285,
+      "grad_norm": 0.17908252775669098,
+      "learning_rate": 2.2322286073674832e-06,
+      "loss": 0.3337,
+      "step": 2652
+    },
+    {
+      "epoch": 2.1551584077985377,
+      "grad_norm": 0.18516302108764648,
+      "learning_rate": 2.228293101684799e-06,
+      "loss": 0.321,
+      "step": 2653
+    },
+    {
+      "epoch": 2.155970755483347,
+      "grad_norm": 0.1751285195350647,
+      "learning_rate": 2.224360073346213e-06,
+      "loss": 0.3346,
+      "step": 2654
+    },
+    {
+      "epoch": 2.156783103168156,
+      "grad_norm": 0.2096243053674698,
+      "learning_rate": 2.2204295258670524e-06,
+      "loss": 0.357,
+      "step": 2655
+    },
+    {
+      "epoch": 2.157595450852965,
+      "grad_norm": 0.19087880849838257,
+      "learning_rate": 2.2165014627604313e-06,
+      "loss": 0.341,
+      "step": 2656
+    },
+    {
+      "epoch": 2.158407798537774,
+      "grad_norm": 0.18925930559635162,
+      "learning_rate": 2.212575887537244e-06,
+      "loss": 0.3112,
+      "step": 2657
+    },
+    {
+      "epoch": 2.1592201462225833,
+      "grad_norm": 0.20250439643859863,
+      "learning_rate": 2.208652803706155e-06,
+      "loss": 0.3347,
+      "step": 2658
+    },
+    {
+      "epoch": 2.1600324939073925,
+      "grad_norm": 0.17150594294071198,
+      "learning_rate": 2.2047322147736096e-06,
+      "loss": 0.3304,
+      "step": 2659
+    },
+    {
+      "epoch": 2.1608448415922012,
+      "grad_norm": 0.19643433392047882,
+      "learning_rate": 2.2008141242438165e-06,
+      "loss": 0.3476,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1616571892770104,
+      "grad_norm": 0.175235778093338,
+      "learning_rate": 2.196898535618757e-06,
+      "loss": 0.3454,
+      "step": 2661
+    },
+    {
+      "epoch": 2.1624695369618196,
+      "grad_norm": 0.18826501071453094,
+      "learning_rate": 2.192985452398171e-06,
+      "loss": 0.3313,
+      "step": 2662
+    },
+    {
+      "epoch": 2.163281884646629,
+      "grad_norm": 0.20117543637752533,
+      "learning_rate": 2.1890748780795644e-06,
+      "loss": 0.3453,
+      "step": 2663
+    },
+    {
+      "epoch": 2.164094232331438,
+      "grad_norm": 0.19149214029312134,
+      "learning_rate": 2.1851668161581946e-06,
+      "loss": 0.3311,
+      "step": 2664
+    },
+    {
+      "epoch": 2.164906580016247,
+      "grad_norm": 0.2020440697669983,
+      "learning_rate": 2.181261270127081e-06,
+      "loss": 0.3319,
+      "step": 2665
+    },
+    {
+      "epoch": 2.165718927701056,
+      "grad_norm": 0.1966685801744461,
+      "learning_rate": 2.1773582434769854e-06,
+      "loss": 0.3442,
+      "step": 2666
+    },
+    {
+      "epoch": 2.166531275385865,
+      "grad_norm": 0.18625487387180328,
+      "learning_rate": 2.1734577396964264e-06,
+      "loss": 0.3556,
+      "step": 2667
+    },
+    {
+      "epoch": 2.1673436230706744,
+      "grad_norm": 0.18913507461547852,
+      "learning_rate": 2.1695597622716603e-06,
+      "loss": 0.3393,
+      "step": 2668
+    },
+    {
+      "epoch": 2.168155970755483,
+      "grad_norm": 0.17889167368412018,
+      "learning_rate": 2.165664314686692e-06,
+      "loss": 0.3283,
+      "step": 2669
+    },
+    {
+      "epoch": 2.1689683184402924,
+      "grad_norm": 0.18265171349048615,
+      "learning_rate": 2.1617714004232588e-06,
+      "loss": 0.338,
+      "step": 2670
+    },
+    {
+      "epoch": 2.1697806661251016,
+      "grad_norm": 0.20406004786491394,
+      "learning_rate": 2.15788102296084e-06,
+      "loss": 0.3489,
+      "step": 2671
+    },
+    {
+      "epoch": 2.1705930138099108,
+      "grad_norm": 0.18505322933197021,
+      "learning_rate": 2.153993185776641e-06,
+      "loss": 0.3229,
+      "step": 2672
+    },
+    {
+      "epoch": 2.1714053614947195,
+      "grad_norm": 0.1915331929922104,
+      "learning_rate": 2.150107892345602e-06,
+      "loss": 0.3442,
+      "step": 2673
+    },
+    {
+      "epoch": 2.1722177091795287,
+      "grad_norm": 0.1837342530488968,
+      "learning_rate": 2.14622514614039e-06,
+      "loss": 0.3402,
+      "step": 2674
+    },
+    {
+      "epoch": 2.173030056864338,
+      "grad_norm": 0.19198153913021088,
+      "learning_rate": 2.142344950631388e-06,
+      "loss": 0.3249,
+      "step": 2675
+    },
+    {
+      "epoch": 2.173842404549147,
+      "grad_norm": 0.18943563103675842,
+      "learning_rate": 2.138467309286709e-06,
+      "loss": 0.3552,
+      "step": 2676
+    },
+    {
+      "epoch": 2.1746547522339563,
+      "grad_norm": 0.2094210833311081,
+      "learning_rate": 2.1345922255721745e-06,
+      "loss": 0.3503,
+      "step": 2677
+    },
+    {
+      "epoch": 2.175467099918765,
+      "grad_norm": 0.18778784573078156,
+      "learning_rate": 2.1307197029513276e-06,
+      "loss": 0.3323,
+      "step": 2678
+    },
+    {
+      "epoch": 2.1762794476035743,
+      "grad_norm": 0.17381109297275543,
+      "learning_rate": 2.1268497448854143e-06,
+      "loss": 0.3529,
+      "step": 2679
+    },
+    {
+      "epoch": 2.1770917952883835,
+      "grad_norm": 0.18056559562683105,
+      "learning_rate": 2.1229823548333965e-06,
+      "loss": 0.3445,
+      "step": 2680
+    },
+    {
+      "epoch": 2.1779041429731927,
+      "grad_norm": 0.18377594649791718,
+      "learning_rate": 2.1191175362519316e-06,
+      "loss": 0.3145,
+      "step": 2681
+    },
+    {
+      "epoch": 2.1787164906580014,
+      "grad_norm": 0.1822587549686432,
+      "learning_rate": 2.115255292595391e-06,
+      "loss": 0.3588,
+      "step": 2682
+    },
+    {
+      "epoch": 2.1795288383428106,
+      "grad_norm": 0.17814859747886658,
+      "learning_rate": 2.111395627315832e-06,
+      "loss": 0.3138,
+      "step": 2683
+    },
+    {
+      "epoch": 2.18034118602762,
+      "grad_norm": 0.1775684356689453,
+      "learning_rate": 2.1075385438630168e-06,
+      "loss": 0.3526,
+      "step": 2684
+    },
+    {
+      "epoch": 2.181153533712429,
+      "grad_norm": 0.18389162421226501,
+      "learning_rate": 2.103684045684393e-06,
+      "loss": 0.3328,
+      "step": 2685
+    },
+    {
+      "epoch": 2.181965881397238,
+      "grad_norm": 0.19145937263965607,
+      "learning_rate": 2.0998321362251036e-06,
+      "loss": 0.3319,
+      "step": 2686
+    },
+    {
+      "epoch": 2.182778229082047,
+      "grad_norm": 0.18174660205841064,
+      "learning_rate": 2.0959828189279712e-06,
+      "loss": 0.3212,
+      "step": 2687
+    },
+    {
+      "epoch": 2.183590576766856,
+      "grad_norm": 0.17451320588588715,
+      "learning_rate": 2.0921360972335086e-06,
+      "loss": 0.3308,
+      "step": 2688
+    },
+    {
+      "epoch": 2.1844029244516654,
+      "grad_norm": 0.186851367354393,
+      "learning_rate": 2.0882919745799023e-06,
+      "loss": 0.3471,
+      "step": 2689
+    },
+    {
+      "epoch": 2.1852152721364746,
+      "grad_norm": 0.18102659285068512,
+      "learning_rate": 2.084450454403022e-06,
+      "loss": 0.324,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1860276198212834,
+      "grad_norm": 0.193750262260437,
+      "learning_rate": 2.0806115401364036e-06,
+      "loss": 0.3422,
+      "step": 2691
+    },
+    {
+      "epoch": 2.1868399675060926,
+      "grad_norm": 0.18444637954235077,
+      "learning_rate": 2.076775235211263e-06,
+      "loss": 0.3426,
+      "step": 2692
+    },
+    {
+      "epoch": 2.1876523151909018,
+      "grad_norm": 0.18047113716602325,
+      "learning_rate": 2.0729415430564748e-06,
+      "loss": 0.3387,
+      "step": 2693
+    },
+    {
+      "epoch": 2.188464662875711,
+      "grad_norm": 0.18147914111614227,
+      "learning_rate": 2.0691104670985868e-06,
+      "loss": 0.3325,
+      "step": 2694
+    },
+    {
+      "epoch": 2.1892770105605197,
+      "grad_norm": 0.21577785909175873,
+      "learning_rate": 2.0652820107618e-06,
+      "loss": 0.3475,
+      "step": 2695
+    },
+    {
+      "epoch": 2.190089358245329,
+      "grad_norm": 0.1822993904352188,
+      "learning_rate": 2.0614561774679826e-06,
+      "loss": 0.3709,
+      "step": 2696
+    },
+    {
+      "epoch": 2.190901705930138,
+      "grad_norm": 0.17449821531772614,
+      "learning_rate": 2.0576329706366494e-06,
+      "loss": 0.3528,
+      "step": 2697
+    },
+    {
+      "epoch": 2.1917140536149473,
+      "grad_norm": 0.19325679540634155,
+      "learning_rate": 2.053812393684977e-06,
+      "loss": 0.344,
+      "step": 2698
+    },
+    {
+      "epoch": 2.192526401299756,
+      "grad_norm": 0.17103978991508484,
+      "learning_rate": 2.0499944500277825e-06,
+      "loss": 0.3214,
+      "step": 2699
+    },
+    {
+      "epoch": 2.1933387489845653,
+      "grad_norm": 0.1822185516357422,
+      "learning_rate": 2.046179143077535e-06,
+      "loss": 0.3403,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1941510966693745,
+      "grad_norm": 0.19423195719718933,
+      "learning_rate": 2.0423664762443483e-06,
+      "loss": 0.3374,
+      "step": 2701
+    },
+    {
+      "epoch": 2.1949634443541837,
+      "grad_norm": 0.18714384734630585,
+      "learning_rate": 2.0385564529359696e-06,
+      "loss": 0.332,
+      "step": 2702
+    },
+    {
+      "epoch": 2.195775792038993,
+      "grad_norm": 0.18590772151947021,
+      "learning_rate": 2.034749076557791e-06,
+      "loss": 0.3483,
+      "step": 2703
+    },
+    {
+      "epoch": 2.1965881397238016,
+      "grad_norm": 0.1875942349433899,
+      "learning_rate": 2.030944350512831e-06,
+      "loss": 0.3459,
+      "step": 2704
+    },
+    {
+      "epoch": 2.197400487408611,
+      "grad_norm": 0.1738954484462738,
+      "learning_rate": 2.027142278201748e-06,
+      "loss": 0.3586,
+      "step": 2705
+    },
+    {
+      "epoch": 2.19821283509342,
+      "grad_norm": 0.18496756255626678,
+      "learning_rate": 2.023342863022819e-06,
+      "loss": 0.3502,
+      "step": 2706
+    },
+    {
+      "epoch": 2.1990251827782292,
+      "grad_norm": 0.18201375007629395,
+      "learning_rate": 2.0195461083719557e-06,
+      "loss": 0.3366,
+      "step": 2707
+    },
+    {
+      "epoch": 2.199837530463038,
+      "grad_norm": 0.18157470226287842,
+      "learning_rate": 2.015752017642684e-06,
+      "loss": 0.3497,
+      "step": 2708
+    },
+    {
+      "epoch": 2.200649878147847,
+      "grad_norm": 0.19371703267097473,
+      "learning_rate": 2.0119605942261535e-06,
+      "loss": 0.3482,
+      "step": 2709
+    },
+    {
+      "epoch": 2.2014622258326564,
+      "grad_norm": 0.1842738389968872,
+      "learning_rate": 2.0081718415111263e-06,
+      "loss": 0.3461,
+      "step": 2710
+    },
+    {
+      "epoch": 2.2022745735174656,
+      "grad_norm": 0.17922085523605347,
+      "learning_rate": 2.004385762883982e-06,
+      "loss": 0.3508,
+      "step": 2711
+    },
+    {
+      "epoch": 2.2030869212022743,
+      "grad_norm": 0.19208742678165436,
+      "learning_rate": 2.000602361728704e-06,
+      "loss": 0.3421,
+      "step": 2712
+    },
+    {
+      "epoch": 2.2038992688870835,
+      "grad_norm": 0.18030697107315063,
+      "learning_rate": 1.99682164142689e-06,
+      "loss": 0.3358,
+      "step": 2713
+    },
+    {
+      "epoch": 2.2047116165718927,
+      "grad_norm": 0.18157561123371124,
+      "learning_rate": 1.993043605357733e-06,
+      "loss": 0.3581,
+      "step": 2714
+    },
+    {
+      "epoch": 2.205523964256702,
+      "grad_norm": 0.1775515228509903,
+      "learning_rate": 1.989268256898036e-06,
+      "loss": 0.353,
+      "step": 2715
+    },
+    {
+      "epoch": 2.206336311941511,
+      "grad_norm": 0.19155611097812653,
+      "learning_rate": 1.985495599422191e-06,
+      "loss": 0.3557,
+      "step": 2716
+    },
+    {
+      "epoch": 2.20714865962632,
+      "grad_norm": 0.1961548924446106,
+      "learning_rate": 1.9817256363021907e-06,
+      "loss": 0.3436,
+      "step": 2717
+    },
+    {
+      "epoch": 2.207961007311129,
+      "grad_norm": 0.17819705605506897,
+      "learning_rate": 1.9779583709076207e-06,
+      "loss": 0.3436,
+      "step": 2718
+    },
+    {
+      "epoch": 2.2087733549959383,
+      "grad_norm": 0.19437868893146515,
+      "learning_rate": 1.9741938066056475e-06,
+      "loss": 0.3543,
+      "step": 2719
+    },
+    {
+      "epoch": 2.2095857026807475,
+      "grad_norm": 0.18450374901294708,
+      "learning_rate": 1.970431946761031e-06,
+      "loss": 0.3438,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2103980503655563,
+      "grad_norm": 0.17478664219379425,
+      "learning_rate": 1.9666727947361135e-06,
+      "loss": 0.3263,
+      "step": 2721
+    },
+    {
+      "epoch": 2.2112103980503655,
+      "grad_norm": 0.19347523152828217,
+      "learning_rate": 1.962916353890811e-06,
+      "loss": 0.3486,
+      "step": 2722
+    },
+    {
+      "epoch": 2.2120227457351747,
+      "grad_norm": 0.1820995807647705,
+      "learning_rate": 1.959162627582623e-06,
+      "loss": 0.3328,
+      "step": 2723
+    },
+    {
+      "epoch": 2.212835093419984,
+      "grad_norm": 0.17816931009292603,
+      "learning_rate": 1.955411619166617e-06,
+      "loss": 0.3363,
+      "step": 2724
+    },
+    {
+      "epoch": 2.213647441104793,
+      "grad_norm": 0.17839615046977997,
+      "learning_rate": 1.951663331995436e-06,
+      "loss": 0.3595,
+      "step": 2725
+    },
+    {
+      "epoch": 2.214459788789602,
+      "grad_norm": 0.18878445029258728,
+      "learning_rate": 1.94791776941929e-06,
+      "loss": 0.3441,
+      "step": 2726
+    },
+    {
+      "epoch": 2.215272136474411,
+      "grad_norm": 0.17879395186901093,
+      "learning_rate": 1.9441749347859503e-06,
+      "loss": 0.3316,
+      "step": 2727
+    },
+    {
+      "epoch": 2.21608448415922,
+      "grad_norm": 0.18412092328071594,
+      "learning_rate": 1.9404348314407544e-06,
+      "loss": 0.3371,
+      "step": 2728
+    },
+    {
+      "epoch": 2.2168968318440294,
+      "grad_norm": 0.19431211054325104,
+      "learning_rate": 1.9366974627265945e-06,
+      "loss": 0.3392,
+      "step": 2729
+    },
+    {
+      "epoch": 2.217709179528838,
+      "grad_norm": 0.17589940130710602,
+      "learning_rate": 1.9329628319839227e-06,
+      "loss": 0.3419,
+      "step": 2730
+    },
+    {
+      "epoch": 2.2185215272136474,
+      "grad_norm": 0.1925750970840454,
+      "learning_rate": 1.929230942550739e-06,
+      "loss": 0.3473,
+      "step": 2731
+    },
+    {
+      "epoch": 2.2193338748984566,
+      "grad_norm": 0.1965939849615097,
+      "learning_rate": 1.9255017977625995e-06,
+      "loss": 0.3244,
+      "step": 2732
+    },
+    {
+      "epoch": 2.2201462225832658,
+      "grad_norm": 0.16616679728031158,
+      "learning_rate": 1.9217754009526e-06,
+      "loss": 0.3308,
+      "step": 2733
+    },
+    {
+      "epoch": 2.2209585702680745,
+      "grad_norm": 0.17430813610553741,
+      "learning_rate": 1.9180517554513884e-06,
+      "loss": 0.3121,
+      "step": 2734
+    },
+    {
+      "epoch": 2.2217709179528837,
+      "grad_norm": 0.1847398579120636,
+      "learning_rate": 1.914330864587144e-06,
+      "loss": 0.3489,
+      "step": 2735
+    },
+    {
+      "epoch": 2.222583265637693,
+      "grad_norm": 0.18155977129936218,
+      "learning_rate": 1.910612731685593e-06,
+      "loss": 0.3255,
+      "step": 2736
+    },
+    {
+      "epoch": 2.223395613322502,
+      "grad_norm": 0.1859879046678543,
+      "learning_rate": 1.9068973600699887e-06,
+      "loss": 0.3384,
+      "step": 2737
+    },
+    {
+      "epoch": 2.2242079610073113,
+      "grad_norm": 0.1829480677843094,
+      "learning_rate": 1.9031847530611237e-06,
+      "loss": 0.3337,
+      "step": 2738
+    },
+    {
+      "epoch": 2.22502030869212,
+      "grad_norm": 0.19858740270137787,
+      "learning_rate": 1.8994749139773133e-06,
+      "loss": 0.3342,
+      "step": 2739
+    },
+    {
+      "epoch": 2.2258326563769293,
+      "grad_norm": 0.1862955093383789,
+      "learning_rate": 1.8957678461344048e-06,
+      "loss": 0.3193,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2266450040617385,
+      "grad_norm": 0.21041829884052277,
+      "learning_rate": 1.892063552845761e-06,
+      "loss": 0.3563,
+      "step": 2741
+    },
+    {
+      "epoch": 2.2274573517465477,
+      "grad_norm": 0.18352819979190826,
+      "learning_rate": 1.888362037422274e-06,
+      "loss": 0.3505,
+      "step": 2742
+    },
+    {
+      "epoch": 2.2282696994313564,
+      "grad_norm": 0.17644517123699188,
+      "learning_rate": 1.8846633031723438e-06,
+      "loss": 0.3283,
+      "step": 2743
+    },
+    {
+      "epoch": 2.2290820471161656,
+      "grad_norm": 0.1697062849998474,
+      "learning_rate": 1.8809673534018908e-06,
+      "loss": 0.3178,
+      "step": 2744
+    },
+    {
+      "epoch": 2.229894394800975,
+      "grad_norm": 0.2081773430109024,
+      "learning_rate": 1.8772741914143472e-06,
+      "loss": 0.36,
+      "step": 2745
+    },
+    {
+      "epoch": 2.230706742485784,
+      "grad_norm": 0.18180976808071136,
+      "learning_rate": 1.873583820510647e-06,
+      "loss": 0.3384,
+      "step": 2746
+    },
+    {
+      "epoch": 2.2315190901705932,
+      "grad_norm": 0.18545198440551758,
+      "learning_rate": 1.8698962439892383e-06,
+      "loss": 0.3293,
+      "step": 2747
+    },
+    {
+      "epoch": 2.232331437855402,
+      "grad_norm": 0.17745718359947205,
+      "learning_rate": 1.8662114651460617e-06,
+      "loss": 0.3427,
+      "step": 2748
+    },
+    {
+      "epoch": 2.233143785540211,
+      "grad_norm": 0.18710409104824066,
+      "learning_rate": 1.8625294872745675e-06,
+      "loss": 0.337,
+      "step": 2749
+    },
+    {
+      "epoch": 2.2339561332250204,
+      "grad_norm": 0.18384474515914917,
+      "learning_rate": 1.8588503136656938e-06,
+      "loss": 0.3553,
+      "step": 2750
+    },
+    {
+      "epoch": 2.2347684809098296,
+      "grad_norm": 0.1860503852367401,
+      "learning_rate": 1.8551739476078784e-06,
+      "loss": 0.3494,
+      "step": 2751
+    },
+    {
+      "epoch": 2.2355808285946384,
+      "grad_norm": 0.1856382191181183,
+      "learning_rate": 1.851500392387045e-06,
+      "loss": 0.3177,
+      "step": 2752
+    },
+    {
+      "epoch": 2.2363931762794476,
+      "grad_norm": 0.189076229929924,
+      "learning_rate": 1.8478296512866105e-06,
+      "loss": 0.3508,
+      "step": 2753
+    },
+    {
+      "epoch": 2.2372055239642568,
+      "grad_norm": 0.1935068815946579,
+      "learning_rate": 1.8441617275874701e-06,
+      "loss": 0.3237,
+      "step": 2754
+    },
+    {
+      "epoch": 2.238017871649066,
+      "grad_norm": 0.16717834770679474,
+      "learning_rate": 1.8404966245680068e-06,
+      "loss": 0.3444,
+      "step": 2755
+    },
+    {
+      "epoch": 2.2388302193338747,
+      "grad_norm": 0.18783438205718994,
+      "learning_rate": 1.836834345504076e-06,
+      "loss": 0.3321,
+      "step": 2756
+    },
+    {
+      "epoch": 2.239642567018684,
+      "grad_norm": 0.18179334700107574,
+      "learning_rate": 1.8331748936690175e-06,
+      "loss": 0.3407,
+      "step": 2757
+    },
+    {
+      "epoch": 2.240454914703493,
+      "grad_norm": 0.20100826025009155,
+      "learning_rate": 1.829518272333633e-06,
+      "loss": 0.3409,
+      "step": 2758
+    },
+    {
+      "epoch": 2.2412672623883023,
+      "grad_norm": 0.1757151186466217,
+      "learning_rate": 1.8258644847662087e-06,
+      "loss": 0.343,
+      "step": 2759
+    },
+    {
+      "epoch": 2.2420796100731115,
+      "grad_norm": 0.19307364523410797,
+      "learning_rate": 1.8222135342324844e-06,
+      "loss": 0.3187,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2428919577579203,
+      "grad_norm": 0.21492883563041687,
+      "learning_rate": 1.8185654239956723e-06,
+      "loss": 0.3402,
+      "step": 2761
+    },
+    {
+      "epoch": 2.2437043054427295,
+      "grad_norm": 0.19775381684303284,
+      "learning_rate": 1.8149201573164415e-06,
+      "loss": 0.3558,
+      "step": 2762
+    },
+    {
+      "epoch": 2.2445166531275387,
+      "grad_norm": 0.1957978457212448,
+      "learning_rate": 1.811277737452924e-06,
+      "loss": 0.3451,
+      "step": 2763
+    },
+    {
+      "epoch": 2.245329000812348,
+      "grad_norm": 0.1742393523454666,
+      "learning_rate": 1.807638167660702e-06,
+      "loss": 0.3343,
+      "step": 2764
+    },
+    {
+      "epoch": 2.2461413484971566,
+      "grad_norm": 0.1988954097032547,
+      "learning_rate": 1.8040014511928155e-06,
+      "loss": 0.3635,
+      "step": 2765
+    },
+    {
+      "epoch": 2.246953696181966,
+      "grad_norm": 0.2034665197134018,
+      "learning_rate": 1.8003675912997487e-06,
+      "loss": 0.3622,
+      "step": 2766
+    },
+    {
+      "epoch": 2.247766043866775,
+      "grad_norm": 0.17482174932956696,
+      "learning_rate": 1.7967365912294398e-06,
+      "loss": 0.323,
+      "step": 2767
+    },
+    {
+      "epoch": 2.2485783915515842,
+      "grad_norm": 0.19553378224372864,
+      "learning_rate": 1.7931084542272632e-06,
+      "loss": 0.3356,
+      "step": 2768
+    },
+    {
+      "epoch": 2.249390739236393,
+      "grad_norm": 0.18534177541732788,
+      "learning_rate": 1.7894831835360387e-06,
+      "loss": 0.3388,
+      "step": 2769
+    },
+    {
+      "epoch": 2.250203086921202,
+      "grad_norm": 0.18021878600120544,
+      "learning_rate": 1.7858607823960262e-06,
+      "loss": 0.3377,
+      "step": 2770
+    },
+    {
+      "epoch": 2.2510154346060114,
+      "grad_norm": 0.1973440796136856,
+      "learning_rate": 1.7822412540449141e-06,
+      "loss": 0.3114,
+      "step": 2771
+    },
+    {
+      "epoch": 2.2518277822908206,
+      "grad_norm": 0.21309779584407806,
+      "learning_rate": 1.77862460171783e-06,
+      "loss": 0.3642,
+      "step": 2772
+    },
+    {
+      "epoch": 2.25264012997563,
+      "grad_norm": 0.17069785296916962,
+      "learning_rate": 1.7750108286473255e-06,
+      "loss": 0.3427,
+      "step": 2773
+    },
+    {
+      "epoch": 2.2534524776604385,
+      "grad_norm": 0.19674034416675568,
+      "learning_rate": 1.771399938063384e-06,
+      "loss": 0.3443,
+      "step": 2774
+    },
+    {
+      "epoch": 2.2542648253452477,
+      "grad_norm": 0.18298090994358063,
+      "learning_rate": 1.7677919331934057e-06,
+      "loss": 0.3347,
+      "step": 2775
+    },
+    {
+      "epoch": 2.255077173030057,
+      "grad_norm": 0.180808886885643,
+      "learning_rate": 1.7641868172622201e-06,
+      "loss": 0.3217,
+      "step": 2776
+    },
+    {
+      "epoch": 2.255889520714866,
+      "grad_norm": 0.18136948347091675,
+      "learning_rate": 1.7605845934920663e-06,
+      "loss": 0.3434,
+      "step": 2777
+    },
+    {
+      "epoch": 2.256701868399675,
+      "grad_norm": 0.1935713291168213,
+      "learning_rate": 1.7569852651026048e-06,
+      "loss": 0.329,
+      "step": 2778
+    },
+    {
+      "epoch": 2.257514216084484,
+      "grad_norm": 0.1822848618030548,
+      "learning_rate": 1.7533888353109029e-06,
+      "loss": 0.3339,
+      "step": 2779
+    },
+    {
+      "epoch": 2.2583265637692933,
+      "grad_norm": 0.1933061182498932,
+      "learning_rate": 1.7497953073314433e-06,
+      "loss": 0.3386,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2591389114541025,
+      "grad_norm": 0.20631670951843262,
+      "learning_rate": 1.7462046843761078e-06,
+      "loss": 0.3539,
+      "step": 2781
+    },
+    {
+      "epoch": 2.2599512591389113,
+      "grad_norm": 0.18343046307563782,
+      "learning_rate": 1.7426169696541884e-06,
+      "loss": 0.3451,
+      "step": 2782
+    },
+    {
+      "epoch": 2.2607636068237205,
+      "grad_norm": 0.18953903019428253,
+      "learning_rate": 1.739032166372372e-06,
+      "loss": 0.3502,
+      "step": 2783
+    },
+    {
+      "epoch": 2.2615759545085297,
+      "grad_norm": 0.18665237724781036,
+      "learning_rate": 1.7354502777347487e-06,
+      "loss": 0.3344,
+      "step": 2784
+    },
+    {
+      "epoch": 2.262388302193339,
+      "grad_norm": 0.19272740185260773,
+      "learning_rate": 1.731871306942799e-06,
+      "loss": 0.3322,
+      "step": 2785
+    },
+    {
+      "epoch": 2.263200649878148,
+      "grad_norm": 0.19423414766788483,
+      "learning_rate": 1.7282952571953987e-06,
+      "loss": 0.3299,
+      "step": 2786
+    },
+    {
+      "epoch": 2.264012997562957,
+      "grad_norm": 0.17249779403209686,
+      "learning_rate": 1.724722131688809e-06,
+      "loss": 0.3344,
+      "step": 2787
+    },
+    {
+      "epoch": 2.264825345247766,
+      "grad_norm": 0.1780373752117157,
+      "learning_rate": 1.721151933616682e-06,
+      "loss": 0.3584,
+      "step": 2788
+    },
+    {
+      "epoch": 2.265637692932575,
+      "grad_norm": 0.18957936763763428,
+      "learning_rate": 1.7175846661700518e-06,
+      "loss": 0.3265,
+      "step": 2789
+    },
+    {
+      "epoch": 2.2664500406173844,
+      "grad_norm": 0.1805725246667862,
+      "learning_rate": 1.7140203325373294e-06,
+      "loss": 0.3466,
+      "step": 2790
+    },
+    {
+      "epoch": 2.267262388302193,
+      "grad_norm": 0.18261782824993134,
+      "learning_rate": 1.7104589359043095e-06,
+      "loss": 0.3359,
+      "step": 2791
+    },
+    {
+      "epoch": 2.2680747359870024,
+      "grad_norm": 0.1764928698539734,
+      "learning_rate": 1.706900479454156e-06,
+      "loss": 0.3331,
+      "step": 2792
+    },
+    {
+      "epoch": 2.2688870836718116,
+      "grad_norm": 0.19667404890060425,
+      "learning_rate": 1.7033449663674102e-06,
+      "loss": 0.3249,
+      "step": 2793
+    },
+    {
+      "epoch": 2.2696994313566208,
+      "grad_norm": 0.17136602103710175,
+      "learning_rate": 1.6997923998219774e-06,
+      "loss": 0.3469,
+      "step": 2794
+    },
+    {
+      "epoch": 2.2705117790414295,
+      "grad_norm": 0.18346178531646729,
+      "learning_rate": 1.6962427829931348e-06,
+      "loss": 0.3459,
+      "step": 2795
+    },
+    {
+      "epoch": 2.2713241267262387,
+      "grad_norm": 0.1869693100452423,
+      "learning_rate": 1.6926961190535152e-06,
+      "loss": 0.3467,
+      "step": 2796
+    },
+    {
+      "epoch": 2.272136474411048,
+      "grad_norm": 0.16732285916805267,
+      "learning_rate": 1.6891524111731246e-06,
+      "loss": 0.3463,
+      "step": 2797
+    },
+    {
+      "epoch": 2.272948822095857,
+      "grad_norm": 0.17263120412826538,
+      "learning_rate": 1.6856116625193135e-06,
+      "loss": 0.3482,
+      "step": 2798
+    },
+    {
+      "epoch": 2.2737611697806663,
+      "grad_norm": 0.1856747269630432,
+      "learning_rate": 1.6820738762567978e-06,
+      "loss": 0.3471,
+      "step": 2799
+    },
+    {
+      "epoch": 2.274573517465475,
+      "grad_norm": 0.21691639721393585,
+      "learning_rate": 1.678539055547637e-06,
+      "loss": 0.3321,
+      "step": 2800
+    },
+    {
+      "epoch": 2.2753858651502843,
+      "grad_norm": 0.18851470947265625,
+      "learning_rate": 1.6750072035512482e-06,
+      "loss": 0.3589,
+      "step": 2801
+    },
+    {
+      "epoch": 2.2761982128350935,
+      "grad_norm": 0.1820124387741089,
+      "learning_rate": 1.6714783234243875e-06,
+      "loss": 0.3183,
+      "step": 2802
+    },
+    {
+      "epoch": 2.2770105605199027,
+      "grad_norm": 0.1821414679288864,
+      "learning_rate": 1.6679524183211616e-06,
+      "loss": 0.3526,
+      "step": 2803
+    },
+    {
+      "epoch": 2.2778229082047114,
+      "grad_norm": 0.18364256620407104,
+      "learning_rate": 1.6644294913930115e-06,
+      "loss": 0.355,
+      "step": 2804
+    },
+    {
+      "epoch": 2.2786352558895206,
+      "grad_norm": 0.19215938448905945,
+      "learning_rate": 1.6609095457887237e-06,
+      "loss": 0.3464,
+      "step": 2805
+    },
+    {
+      "epoch": 2.27944760357433,
+      "grad_norm": 0.1848536729812622,
+      "learning_rate": 1.657392584654412e-06,
+      "loss": 0.3494,
+      "step": 2806
+    },
+    {
+      "epoch": 2.280259951259139,
+      "grad_norm": 0.17051635682582855,
+      "learning_rate": 1.6538786111335303e-06,
+      "loss": 0.3324,
+      "step": 2807
+    },
+    {
+      "epoch": 2.281072298943948,
+      "grad_norm": 0.2004299908876419,
+      "learning_rate": 1.6503676283668553e-06,
+      "loss": 0.3281,
+      "step": 2808
+    },
+    {
+      "epoch": 2.281884646628757,
+      "grad_norm": 0.20010870695114136,
+      "learning_rate": 1.6468596394924974e-06,
+      "loss": 0.3447,
+      "step": 2809
+    },
+    {
+      "epoch": 2.282696994313566,
+      "grad_norm": 0.1928035169839859,
+      "learning_rate": 1.6433546476458833e-06,
+      "loss": 0.3338,
+      "step": 2810
+    },
+    {
+      "epoch": 2.2835093419983754,
+      "grad_norm": 0.1745130568742752,
+      "learning_rate": 1.63985265595977e-06,
+      "loss": 0.3248,
+      "step": 2811
+    },
+    {
+      "epoch": 2.2843216896831846,
+      "grad_norm": 0.18135616183280945,
+      "learning_rate": 1.6363536675642233e-06,
+      "loss": 0.315,
+      "step": 2812
+    },
+    {
+      "epoch": 2.2851340373679934,
+      "grad_norm": 0.18031030893325806,
+      "learning_rate": 1.632857685586632e-06,
+      "loss": 0.3481,
+      "step": 2813
+    },
+    {
+      "epoch": 2.2859463850528026,
+      "grad_norm": 0.21053160727024078,
+      "learning_rate": 1.629364713151697e-06,
+      "loss": 0.3476,
+      "step": 2814
+    },
+    {
+      "epoch": 2.2867587327376118,
+      "grad_norm": 0.1913031041622162,
+      "learning_rate": 1.625874753381424e-06,
+      "loss": 0.3358,
+      "step": 2815
+    },
+    {
+      "epoch": 2.287571080422421,
+      "grad_norm": 0.17437809705734253,
+      "learning_rate": 1.6223878093951329e-06,
+      "loss": 0.3387,
+      "step": 2816
+    },
+    {
+      "epoch": 2.2883834281072297,
+      "grad_norm": 0.1778300553560257,
+      "learning_rate": 1.6189038843094419e-06,
+      "loss": 0.3287,
+      "step": 2817
+    },
+    {
+      "epoch": 2.289195775792039,
+      "grad_norm": 0.18271763622760773,
+      "learning_rate": 1.615422981238276e-06,
+      "loss": 0.3311,
+      "step": 2818
+    },
+    {
+      "epoch": 2.290008123476848,
+      "grad_norm": 0.21277537941932678,
+      "learning_rate": 1.6119451032928547e-06,
+      "loss": 0.3424,
+      "step": 2819
+    },
+    {
+      "epoch": 2.2908204711616573,
+      "grad_norm": 0.18759609758853912,
+      "learning_rate": 1.6084702535816987e-06,
+      "loss": 0.3548,
+      "step": 2820
+    },
+    {
+      "epoch": 2.291632818846466,
+      "grad_norm": 0.1712750345468521,
+      "learning_rate": 1.6049984352106163e-06,
+      "loss": 0.3394,
+      "step": 2821
+    },
+    {
+      "epoch": 2.2924451665312753,
+      "grad_norm": 0.17708075046539307,
+      "learning_rate": 1.6015296512827122e-06,
+      "loss": 0.3274,
+      "step": 2822
+    },
+    {
+      "epoch": 2.2932575142160845,
+      "grad_norm": 0.17528541386127472,
+      "learning_rate": 1.5980639048983726e-06,
+      "loss": 0.3408,
+      "step": 2823
+    },
+    {
+      "epoch": 2.2940698619008937,
+      "grad_norm": 0.18100440502166748,
+      "learning_rate": 1.5946011991552757e-06,
+      "loss": 0.3397,
+      "step": 2824
+    },
+    {
+      "epoch": 2.294882209585703,
+      "grad_norm": 0.18964216113090515,
+      "learning_rate": 1.5911415371483757e-06,
+      "loss": 0.3437,
+      "step": 2825
+    },
+    {
+      "epoch": 2.2956945572705116,
+      "grad_norm": 0.16908679902553558,
+      "learning_rate": 1.587684921969912e-06,
+      "loss": 0.3383,
+      "step": 2826
+    },
+    {
+      "epoch": 2.296506904955321,
+      "grad_norm": 0.18741890788078308,
+      "learning_rate": 1.5842313567093942e-06,
+      "loss": 0.3271,
+      "step": 2827
+    },
+    {
+      "epoch": 2.29731925264013,
+      "grad_norm": 0.17021752893924713,
+      "learning_rate": 1.5807808444536132e-06,
+      "loss": 0.3306,
+      "step": 2828
+    },
+    {
+      "epoch": 2.2981316003249392,
+      "grad_norm": 0.18824386596679688,
+      "learning_rate": 1.5773333882866242e-06,
+      "loss": 0.3415,
+      "step": 2829
+    },
+    {
+      "epoch": 2.298943948009748,
+      "grad_norm": 0.18836955726146698,
+      "learning_rate": 1.5738889912897575e-06,
+      "loss": 0.3444,
+      "step": 2830
+    },
+    {
+      "epoch": 2.299756295694557,
+      "grad_norm": 0.17527669668197632,
+      "learning_rate": 1.5704476565416027e-06,
+      "loss": 0.3307,
+      "step": 2831
+    },
+    {
+      "epoch": 2.3005686433793664,
+      "grad_norm": 0.18583188951015472,
+      "learning_rate": 1.5670093871180165e-06,
+      "loss": 0.3348,
+      "step": 2832
+    },
+    {
+      "epoch": 2.3013809910641756,
+      "grad_norm": 0.17782540619373322,
+      "learning_rate": 1.5635741860921166e-06,
+      "loss": 0.3535,
+      "step": 2833
+    },
+    {
+      "epoch": 2.3021933387489844,
+      "grad_norm": 0.20423337817192078,
+      "learning_rate": 1.5601420565342734e-06,
+      "loss": 0.3632,
+      "step": 2834
+    },
+    {
+      "epoch": 2.3030056864337936,
+      "grad_norm": 0.19374577701091766,
+      "learning_rate": 1.5567130015121169e-06,
+      "loss": 0.3442,
+      "step": 2835
+    },
+    {
+      "epoch": 2.3038180341186028,
+      "grad_norm": 0.18472182750701904,
+      "learning_rate": 1.5532870240905272e-06,
+      "loss": 0.3369,
+      "step": 2836
+    },
+    {
+      "epoch": 2.304630381803412,
+      "grad_norm": 0.19575855135917664,
+      "learning_rate": 1.5498641273316322e-06,
+      "loss": 0.3256,
+      "step": 2837
+    },
+    {
+      "epoch": 2.305442729488221,
+      "grad_norm": 0.18174710869789124,
+      "learning_rate": 1.5464443142948083e-06,
+      "loss": 0.3237,
+      "step": 2838
+    },
+    {
+      "epoch": 2.30625507717303,
+      "grad_norm": 0.1778310388326645,
+      "learning_rate": 1.5430275880366774e-06,
+      "loss": 0.3532,
+      "step": 2839
+    },
+    {
+      "epoch": 2.307067424857839,
+      "grad_norm": 0.24693480134010315,
+      "learning_rate": 1.5396139516110959e-06,
+      "loss": 0.337,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3078797725426483,
+      "grad_norm": 0.1802762746810913,
+      "learning_rate": 1.5362034080691673e-06,
+      "loss": 0.3306,
+      "step": 2841
+    },
+    {
+      "epoch": 2.3086921202274575,
+      "grad_norm": 0.19342151284217834,
+      "learning_rate": 1.5327959604592218e-06,
+      "loss": 0.3396,
+      "step": 2842
+    },
+    {
+      "epoch": 2.3095044679122663,
+      "grad_norm": 0.17494244873523712,
+      "learning_rate": 1.5293916118268304e-06,
+      "loss": 0.3319,
+      "step": 2843
+    },
+    {
+      "epoch": 2.3103168155970755,
+      "grad_norm": 0.17358238995075226,
+      "learning_rate": 1.5259903652147878e-06,
+      "loss": 0.3588,
+      "step": 2844
+    },
+    {
+      "epoch": 2.3111291632818847,
+      "grad_norm": 0.16715854406356812,
+      "learning_rate": 1.5225922236631225e-06,
+      "loss": 0.3427,
+      "step": 2845
+    },
+    {
+      "epoch": 2.311941510966694,
+      "grad_norm": 0.1861705183982849,
+      "learning_rate": 1.5191971902090797e-06,
+      "loss": 0.3473,
+      "step": 2846
+    },
+    {
+      "epoch": 2.3127538586515026,
+      "grad_norm": 0.19890528917312622,
+      "learning_rate": 1.515805267887135e-06,
+      "loss": 0.3292,
+      "step": 2847
+    },
+    {
+      "epoch": 2.313566206336312,
+      "grad_norm": 0.18326164782047272,
+      "learning_rate": 1.5124164597289765e-06,
+      "loss": 0.3283,
+      "step": 2848
+    },
+    {
+      "epoch": 2.314378554021121,
+      "grad_norm": 0.18717941641807556,
+      "learning_rate": 1.5090307687635141e-06,
+      "loss": 0.353,
+      "step": 2849
+    },
+    {
+      "epoch": 2.3151909017059302,
+      "grad_norm": 0.16956616938114166,
+      "learning_rate": 1.505648198016867e-06,
+      "loss": 0.3273,
+      "step": 2850
+    },
+    {
+      "epoch": 2.3160032493907394,
+      "grad_norm": 0.1749870479106903,
+      "learning_rate": 1.5022687505123702e-06,
+      "loss": 0.3252,
+      "step": 2851
+    },
+    {
+      "epoch": 2.316815597075548,
+      "grad_norm": 0.1754606068134308,
+      "learning_rate": 1.4988924292705615e-06,
+      "loss": 0.3556,
+      "step": 2852
+    },
+    {
+      "epoch": 2.3176279447603574,
+      "grad_norm": 0.1884399801492691,
+      "learning_rate": 1.4955192373091915e-06,
+      "loss": 0.3418,
+      "step": 2853
+    },
+    {
+      "epoch": 2.3184402924451666,
+      "grad_norm": 0.2050260752439499,
+      "learning_rate": 1.4921491776432062e-06,
+      "loss": 0.3459,
+      "step": 2854
+    },
+    {
+      "epoch": 2.319252640129976,
+      "grad_norm": 0.19927607476711273,
+      "learning_rate": 1.488782253284759e-06,
+      "loss": 0.3604,
+      "step": 2855
+    },
+    {
+      "epoch": 2.3200649878147845,
+      "grad_norm": 0.18859179317951202,
+      "learning_rate": 1.4854184672431948e-06,
+      "loss": 0.3314,
+      "step": 2856
+    },
+    {
+      "epoch": 2.3208773354995937,
+      "grad_norm": 0.18433798849582672,
+      "learning_rate": 1.4820578225250582e-06,
+      "loss": 0.3444,
+      "step": 2857
+    },
+    {
+      "epoch": 2.321689683184403,
+      "grad_norm": 0.18978331983089447,
+      "learning_rate": 1.4787003221340851e-06,
+      "loss": 0.339,
+      "step": 2858
+    },
+    {
+      "epoch": 2.322502030869212,
+      "grad_norm": 0.1812579482793808,
+      "learning_rate": 1.475345969071198e-06,
+      "loss": 0.3514,
+      "step": 2859
+    },
+    {
+      "epoch": 2.323314378554021,
+      "grad_norm": 0.18212780356407166,
+      "learning_rate": 1.4719947663345112e-06,
+      "loss": 0.3061,
+      "step": 2860
+    },
+    {
+      "epoch": 2.32412672623883,
+      "grad_norm": 0.17269931733608246,
+      "learning_rate": 1.468646716919318e-06,
+      "loss": 0.3216,
+      "step": 2861
+    },
+    {
+      "epoch": 2.3249390739236393,
+      "grad_norm": 0.1763554960489273,
+      "learning_rate": 1.4653018238180982e-06,
+      "loss": 0.341,
+      "step": 2862
+    },
+    {
+      "epoch": 2.3257514216084485,
+      "grad_norm": 0.1807367205619812,
+      "learning_rate": 1.461960090020505e-06,
+      "loss": 0.3496,
+      "step": 2863
+    },
+    {
+      "epoch": 2.3265637692932577,
+      "grad_norm": 0.17457567155361176,
+      "learning_rate": 1.458621518513375e-06,
+      "loss": 0.3241,
+      "step": 2864
+    },
+    {
+      "epoch": 2.3273761169780665,
+      "grad_norm": 0.18420977890491486,
+      "learning_rate": 1.455286112280711e-06,
+      "loss": 0.3601,
+      "step": 2865
+    },
+    {
+      "epoch": 2.3281884646628757,
+      "grad_norm": 0.1852586567401886,
+      "learning_rate": 1.4519538743036927e-06,
+      "loss": 0.3573,
+      "step": 2866
+    },
+    {
+      "epoch": 2.329000812347685,
+      "grad_norm": 0.18355216085910797,
+      "learning_rate": 1.4486248075606624e-06,
+      "loss": 0.3311,
+      "step": 2867
+    },
+    {
+      "epoch": 2.329813160032494,
+      "grad_norm": 0.17317213118076324,
+      "learning_rate": 1.445298915027134e-06,
+      "loss": 0.3347,
+      "step": 2868
+    },
+    {
+      "epoch": 2.330625507717303,
+      "grad_norm": 0.18310576677322388,
+      "learning_rate": 1.441976199675778e-06,
+      "loss": 0.3487,
+      "step": 2869
+    },
+    {
+      "epoch": 2.331437855402112,
+      "grad_norm": 0.18443046510219574,
+      "learning_rate": 1.4386566644764321e-06,
+      "loss": 0.3654,
+      "step": 2870
+    },
+    {
+      "epoch": 2.332250203086921,
+      "grad_norm": 0.17004066705703735,
+      "learning_rate": 1.435340312396084e-06,
+      "loss": 0.326,
+      "step": 2871
+    },
+    {
+      "epoch": 2.3330625507717304,
+      "grad_norm": 0.1767830103635788,
+      "learning_rate": 1.4320271463988822e-06,
+      "loss": 0.3431,
+      "step": 2872
+    },
+    {
+      "epoch": 2.333874898456539,
+      "grad_norm": 0.16938655078411102,
+      "learning_rate": 1.428717169446125e-06,
+      "loss": 0.3264,
+      "step": 2873
+    },
+    {
+      "epoch": 2.3346872461413484,
+      "grad_norm": 0.20696672797203064,
+      "learning_rate": 1.425410384496263e-06,
+      "loss": 0.3566,
+      "step": 2874
+    },
+    {
+      "epoch": 2.3354995938261576,
+      "grad_norm": 0.1607482135295868,
+      "learning_rate": 1.4221067945048877e-06,
+      "loss": 0.3375,
+      "step": 2875
+    },
+    {
+      "epoch": 2.3363119415109668,
+      "grad_norm": 0.18662799894809723,
+      "learning_rate": 1.4188064024247428e-06,
+      "loss": 0.3199,
+      "step": 2876
+    },
+    {
+      "epoch": 2.337124289195776,
+      "grad_norm": 0.17572836577892303,
+      "learning_rate": 1.415509211205706e-06,
+      "loss": 0.3396,
+      "step": 2877
+    },
+    {
+      "epoch": 2.3379366368805847,
+      "grad_norm": 0.20585259795188904,
+      "learning_rate": 1.4122152237948018e-06,
+      "loss": 0.3372,
+      "step": 2878
+    },
+    {
+      "epoch": 2.338748984565394,
+      "grad_norm": 0.17710377275943756,
+      "learning_rate": 1.4089244431361826e-06,
+      "loss": 0.3477,
+      "step": 2879
+    },
+    {
+      "epoch": 2.339561332250203,
+      "grad_norm": 0.16925492882728577,
+      "learning_rate": 1.405636872171144e-06,
+      "loss": 0.3204,
+      "step": 2880
+    },
+    {
+      "epoch": 2.3403736799350123,
+      "grad_norm": 0.17079217731952667,
+      "learning_rate": 1.4023525138381039e-06,
+      "loss": 0.3188,
+      "step": 2881
+    },
+    {
+      "epoch": 2.3411860276198215,
+      "grad_norm": 0.18307355046272278,
+      "learning_rate": 1.3990713710726145e-06,
+      "loss": 0.3167,
+      "step": 2882
+    },
+    {
+      "epoch": 2.3419983753046303,
+      "grad_norm": 0.1895926594734192,
+      "learning_rate": 1.395793446807354e-06,
+      "loss": 0.3667,
+      "step": 2883
+    },
+    {
+      "epoch": 2.3428107229894395,
+      "grad_norm": 0.17633506655693054,
+      "learning_rate": 1.3925187439721193e-06,
+      "loss": 0.3386,
+      "step": 2884
+    },
+    {
+      "epoch": 2.3436230706742487,
+      "grad_norm": 0.18229733407497406,
+      "learning_rate": 1.3892472654938338e-06,
+      "loss": 0.3624,
+      "step": 2885
+    },
+    {
+      "epoch": 2.3444354183590574,
+      "grad_norm": 0.18110080063343048,
+      "learning_rate": 1.385979014296533e-06,
+      "loss": 0.3227,
+      "step": 2886
+    },
+    {
+      "epoch": 2.3452477660438666,
+      "grad_norm": 0.17933672666549683,
+      "learning_rate": 1.3827139933013739e-06,
+      "loss": 0.3354,
+      "step": 2887
+    },
+    {
+      "epoch": 2.346060113728676,
+      "grad_norm": 0.18171724677085876,
+      "learning_rate": 1.3794522054266207e-06,
+      "loss": 0.3293,
+      "step": 2888
+    },
+    {
+      "epoch": 2.346872461413485,
+      "grad_norm": 0.19242213666439056,
+      "learning_rate": 1.376193653587653e-06,
+      "loss": 0.3279,
+      "step": 2889
+    },
+    {
+      "epoch": 2.3476848090982942,
+      "grad_norm": 0.23667767643928528,
+      "learning_rate": 1.3729383406969533e-06,
+      "loss": 0.3139,
+      "step": 2890
+    },
+    {
+      "epoch": 2.348497156783103,
+      "grad_norm": 0.17930817604064941,
+      "learning_rate": 1.3696862696641133e-06,
+      "loss": 0.3235,
+      "step": 2891
+    },
+    {
+      "epoch": 2.349309504467912,
+      "grad_norm": 0.17131567001342773,
+      "learning_rate": 1.366437443395824e-06,
+      "loss": 0.3206,
+      "step": 2892
+    },
+    {
+      "epoch": 2.3501218521527214,
+      "grad_norm": 0.18371903896331787,
+      "learning_rate": 1.3631918647958786e-06,
+      "loss": 0.3373,
+      "step": 2893
+    },
+    {
+      "epoch": 2.3509341998375306,
+      "grad_norm": 0.17160633206367493,
+      "learning_rate": 1.3599495367651645e-06,
+      "loss": 0.3317,
+      "step": 2894
+    },
+    {
+      "epoch": 2.35174654752234,
+      "grad_norm": 0.16599975526332855,
+      "learning_rate": 1.3567104622016685e-06,
+      "loss": 0.3238,
+      "step": 2895
+    },
+    {
+      "epoch": 2.3525588952071486,
+      "grad_norm": 0.17499838769435883,
+      "learning_rate": 1.3534746440004637e-06,
+      "loss": 0.3421,
+      "step": 2896
+    },
+    {
+      "epoch": 2.3533712428919578,
+      "grad_norm": 0.17254510521888733,
+      "learning_rate": 1.3502420850537185e-06,
+      "loss": 0.354,
+      "step": 2897
+    },
+    {
+      "epoch": 2.354183590576767,
+      "grad_norm": 0.1808159202337265,
+      "learning_rate": 1.3470127882506824e-06,
+      "loss": 0.3363,
+      "step": 2898
+    },
+    {
+      "epoch": 2.3549959382615757,
+      "grad_norm": 0.1670791208744049,
+      "learning_rate": 1.3437867564776952e-06,
+      "loss": 0.3351,
+      "step": 2899
+    },
+    {
+      "epoch": 2.355808285946385,
+      "grad_norm": 0.1665712594985962,
+      "learning_rate": 1.3405639926181718e-06,
+      "loss": 0.337,
+      "step": 2900
+    },
+    {
+      "epoch": 2.356620633631194,
+      "grad_norm": 0.17454230785369873,
+      "learning_rate": 1.337344499552612e-06,
+      "loss": 0.3418,
+      "step": 2901
+    },
+    {
+      "epoch": 2.3574329813160033,
+      "grad_norm": 0.18587443232536316,
+      "learning_rate": 1.3341282801585914e-06,
+      "loss": 0.3309,
+      "step": 2902
+    },
+    {
+      "epoch": 2.3582453290008125,
+      "grad_norm": 0.19230163097381592,
+      "learning_rate": 1.3309153373107547e-06,
+      "loss": 0.3417,
+      "step": 2903
+    },
+    {
+      "epoch": 2.3590576766856213,
+      "grad_norm": 0.19198952615261078,
+      "learning_rate": 1.3277056738808242e-06,
+      "loss": 0.347,
+      "step": 2904
+    },
+    {
+      "epoch": 2.3598700243704305,
+      "grad_norm": 0.19337230920791626,
+      "learning_rate": 1.3244992927375865e-06,
+      "loss": 0.3524,
+      "step": 2905
+    },
+    {
+      "epoch": 2.3606823720552397,
+      "grad_norm": 0.18349269032478333,
+      "learning_rate": 1.3212961967468985e-06,
+      "loss": 0.3369,
+      "step": 2906
+    },
+    {
+      "epoch": 2.361494719740049,
+      "grad_norm": 0.18029166758060455,
+      "learning_rate": 1.3180963887716758e-06,
+      "loss": 0.3477,
+      "step": 2907
+    },
+    {
+      "epoch": 2.362307067424858,
+      "grad_norm": 0.17530770599842072,
+      "learning_rate": 1.3148998716719014e-06,
+      "loss": 0.3146,
+      "step": 2908
+    },
+    {
+      "epoch": 2.363119415109667,
+      "grad_norm": 0.18205562233924866,
+      "learning_rate": 1.3117066483046098e-06,
+      "loss": 0.3317,
+      "step": 2909
+    },
+    {
+      "epoch": 2.363931762794476,
+      "grad_norm": 0.17223860323429108,
+      "learning_rate": 1.3085167215238982e-06,
+      "loss": 0.3382,
+      "step": 2910
+    },
+    {
+      "epoch": 2.3647441104792852,
+      "grad_norm": 0.17894266545772552,
+      "learning_rate": 1.3053300941809132e-06,
+      "loss": 0.3346,
+      "step": 2911
+    },
+    {
+      "epoch": 2.365556458164094,
+      "grad_norm": 0.19638913869857788,
+      "learning_rate": 1.302146769123856e-06,
+      "loss": 0.3433,
+      "step": 2912
+    },
+    {
+      "epoch": 2.366368805848903,
+      "grad_norm": 0.18420980870723724,
+      "learning_rate": 1.2989667491979702e-06,
+      "loss": 0.3347,
+      "step": 2913
+    },
+    {
+      "epoch": 2.3671811535337124,
+      "grad_norm": 0.19545131921768188,
+      "learning_rate": 1.2957900372455529e-06,
+      "loss": 0.3393,
+      "step": 2914
+    },
+    {
+      "epoch": 2.3679935012185216,
+      "grad_norm": 0.18420866131782532,
+      "learning_rate": 1.2926166361059366e-06,
+      "loss": 0.3363,
+      "step": 2915
+    },
+    {
+      "epoch": 2.368805848903331,
+      "grad_norm": 0.18952317535877228,
+      "learning_rate": 1.2894465486155028e-06,
+      "loss": 0.3625,
+      "step": 2916
+    },
+    {
+      "epoch": 2.3696181965881395,
+      "grad_norm": 0.1771267205476761,
+      "learning_rate": 1.2862797776076624e-06,
+      "loss": 0.3395,
+      "step": 2917
+    },
+    {
+      "epoch": 2.3704305442729487,
+      "grad_norm": 0.19918197393417358,
+      "learning_rate": 1.2831163259128714e-06,
+      "loss": 0.3605,
+      "step": 2918
+    },
+    {
+      "epoch": 2.371242891957758,
+      "grad_norm": 0.18986870348453522,
+      "learning_rate": 1.2799561963586104e-06,
+      "loss": 0.3331,
+      "step": 2919
+    },
+    {
+      "epoch": 2.372055239642567,
+      "grad_norm": 0.18701660633087158,
+      "learning_rate": 1.2767993917693984e-06,
+      "loss": 0.3407,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3728675873273763,
+      "grad_norm": 0.1802656650543213,
+      "learning_rate": 1.2736459149667747e-06,
+      "loss": 0.3293,
+      "step": 2921
+    },
+    {
+      "epoch": 2.373679935012185,
+      "grad_norm": 0.16924260556697845,
+      "learning_rate": 1.2704957687693114e-06,
+      "loss": 0.3362,
+      "step": 2922
+    },
+    {
+      "epoch": 2.3744922826969943,
+      "grad_norm": 0.19035495817661285,
+      "learning_rate": 1.2673489559925989e-06,
+      "loss": 0.3405,
+      "step": 2923
+    },
+    {
+      "epoch": 2.3753046303818035,
+      "grad_norm": 0.1699064075946808,
+      "learning_rate": 1.2642054794492514e-06,
+      "loss": 0.3392,
+      "step": 2924
+    },
+    {
+      "epoch": 2.3761169780666123,
+      "grad_norm": 0.17672938108444214,
+      "learning_rate": 1.2610653419488978e-06,
+      "loss": 0.3307,
+      "step": 2925
+    },
+    {
+      "epoch": 2.3769293257514215,
+      "grad_norm": 0.18439297378063202,
+      "learning_rate": 1.2579285462981855e-06,
+      "loss": 0.3402,
+      "step": 2926
+    },
+    {
+      "epoch": 2.3777416734362307,
+      "grad_norm": 0.19033187627792358,
+      "learning_rate": 1.2547950953007754e-06,
+      "loss": 0.3238,
+      "step": 2927
+    },
+    {
+      "epoch": 2.37855402112104,
+      "grad_norm": 0.17743420600891113,
+      "learning_rate": 1.2516649917573342e-06,
+      "loss": 0.3297,
+      "step": 2928
+    },
+    {
+      "epoch": 2.379366368805849,
+      "grad_norm": 0.19800128042697906,
+      "learning_rate": 1.2485382384655437e-06,
+      "loss": 0.3395,
+      "step": 2929
+    },
+    {
+      "epoch": 2.380178716490658,
+      "grad_norm": 0.17256544530391693,
+      "learning_rate": 1.2454148382200837e-06,
+      "loss": 0.3338,
+      "step": 2930
+    },
+    {
+      "epoch": 2.380991064175467,
+      "grad_norm": 0.19352488219738007,
+      "learning_rate": 1.2422947938126445e-06,
+      "loss": 0.3239,
+      "step": 2931
+    },
+    {
+      "epoch": 2.381803411860276,
+      "grad_norm": 0.1882380098104477,
+      "learning_rate": 1.2391781080319098e-06,
+      "loss": 0.3417,
+      "step": 2932
+    },
+    {
+      "epoch": 2.3826157595450854,
+      "grad_norm": 0.19563433527946472,
+      "learning_rate": 1.236064783663568e-06,
+      "loss": 0.3222,
+      "step": 2933
+    },
+    {
+      "epoch": 2.3834281072298946,
+      "grad_norm": 0.1861247569322586,
+      "learning_rate": 1.2329548234902976e-06,
+      "loss": 0.3392,
+      "step": 2934
+    },
+    {
+      "epoch": 2.3842404549147034,
+      "grad_norm": 0.1825811117887497,
+      "learning_rate": 1.2298482302917752e-06,
+      "loss": 0.3142,
+      "step": 2935
+    },
+    {
+      "epoch": 2.3850528025995126,
+      "grad_norm": 0.17453645169734955,
+      "learning_rate": 1.2267450068446617e-06,
+      "loss": 0.3524,
+      "step": 2936
+    },
+    {
+      "epoch": 2.385865150284322,
+      "grad_norm": 0.18708840012550354,
+      "learning_rate": 1.2236451559226137e-06,
+      "loss": 0.3434,
+      "step": 2937
+    },
+    {
+      "epoch": 2.386677497969131,
+      "grad_norm": 0.1807887852191925,
+      "learning_rate": 1.2205486802962656e-06,
+      "loss": 0.3357,
+      "step": 2938
+    },
+    {
+      "epoch": 2.3874898456539397,
+      "grad_norm": 0.18743029236793518,
+      "learning_rate": 1.217455582733243e-06,
+      "loss": 0.3421,
+      "step": 2939
+    },
+    {
+      "epoch": 2.388302193338749,
+      "grad_norm": 0.17689308524131775,
+      "learning_rate": 1.2143658659981444e-06,
+      "loss": 0.3241,
+      "step": 2940
+    },
+    {
+      "epoch": 2.389114541023558,
+      "grad_norm": 0.1675174981355667,
+      "learning_rate": 1.2112795328525529e-06,
+      "loss": 0.3433,
+      "step": 2941
+    },
+    {
+      "epoch": 2.3899268887083673,
+      "grad_norm": 0.1752869039773941,
+      "learning_rate": 1.2081965860550227e-06,
+      "loss": 0.3469,
+      "step": 2942
+    },
+    {
+      "epoch": 2.390739236393176,
+      "grad_norm": 0.2606005370616913,
+      "learning_rate": 1.2051170283610858e-06,
+      "loss": 0.3461,
+      "step": 2943
+    },
+    {
+      "epoch": 2.3915515840779853,
+      "grad_norm": 0.1763869822025299,
+      "learning_rate": 1.2020408625232388e-06,
+      "loss": 0.3121,
+      "step": 2944
+    },
+    {
+      "epoch": 2.3923639317627945,
+      "grad_norm": 0.19493189454078674,
+      "learning_rate": 1.198968091290953e-06,
+      "loss": 0.3334,
+      "step": 2945
+    },
+    {
+      "epoch": 2.3931762794476037,
+      "grad_norm": 0.17276941239833832,
+      "learning_rate": 1.195898717410664e-06,
+      "loss": 0.3335,
+      "step": 2946
+    },
+    {
+      "epoch": 2.393988627132413,
+      "grad_norm": 0.18116745352745056,
+      "learning_rate": 1.1928327436257669e-06,
+      "loss": 0.3139,
+      "step": 2947
+    },
+    {
+      "epoch": 2.3948009748172217,
+      "grad_norm": 0.1803610920906067,
+      "learning_rate": 1.189770172676623e-06,
+      "loss": 0.3446,
+      "step": 2948
+    },
+    {
+      "epoch": 2.395613322502031,
+      "grad_norm": 0.1594579666852951,
+      "learning_rate": 1.1867110073005483e-06,
+      "loss": 0.345,
+      "step": 2949
+    },
+    {
+      "epoch": 2.39642567018684,
+      "grad_norm": 0.1838209629058838,
+      "learning_rate": 1.1836552502318171e-06,
+      "loss": 0.3464,
+      "step": 2950
+    },
+    {
+      "epoch": 2.3972380178716493,
+      "grad_norm": 0.18979939818382263,
+      "learning_rate": 1.1806029042016581e-06,
+      "loss": 0.3299,
+      "step": 2951
+    },
+    {
+      "epoch": 2.398050365556458,
+      "grad_norm": 0.1795673668384552,
+      "learning_rate": 1.1775539719382473e-06,
+      "loss": 0.3258,
+      "step": 2952
+    },
+    {
+      "epoch": 2.398862713241267,
+      "grad_norm": 0.16197015345096588,
+      "learning_rate": 1.1745084561667136e-06,
+      "loss": 0.3534,
+      "step": 2953
+    },
+    {
+      "epoch": 2.3996750609260764,
+      "grad_norm": 0.18996970355510712,
+      "learning_rate": 1.1714663596091319e-06,
+      "loss": 0.3428,
+      "step": 2954
+    },
+    {
+      "epoch": 2.4004874086108856,
+      "grad_norm": 0.18349558115005493,
+      "learning_rate": 1.1684276849845171e-06,
+      "loss": 0.343,
+      "step": 2955
+    },
+    {
+      "epoch": 2.4012997562956944,
+      "grad_norm": 0.17613866925239563,
+      "learning_rate": 1.1653924350088319e-06,
+      "loss": 0.3289,
+      "step": 2956
+    },
+    {
+      "epoch": 2.4021121039805036,
+      "grad_norm": 0.16812676191329956,
+      "learning_rate": 1.1623606123949705e-06,
+      "loss": 0.3497,
+      "step": 2957
+    },
+    {
+      "epoch": 2.4029244516653128,
+      "grad_norm": 0.19753962755203247,
+      "learning_rate": 1.1593322198527707e-06,
+      "loss": 0.3437,
+      "step": 2958
+    },
+    {
+      "epoch": 2.403736799350122,
+      "grad_norm": 0.1808290183544159,
+      "learning_rate": 1.1563072600889992e-06,
+      "loss": 0.3345,
+      "step": 2959
+    },
+    {
+      "epoch": 2.404549147034931,
+      "grad_norm": 0.19299624860286713,
+      "learning_rate": 1.1532857358073597e-06,
+      "loss": 0.3249,
+      "step": 2960
+    },
+    {
+      "epoch": 2.40536149471974,
+      "grad_norm": 0.1924418807029724,
+      "learning_rate": 1.1502676497084786e-06,
+      "loss": 0.3452,
+      "step": 2961
+    },
+    {
+      "epoch": 2.406173842404549,
+      "grad_norm": 0.17788253724575043,
+      "learning_rate": 1.147253004489917e-06,
+      "loss": 0.3379,
+      "step": 2962
+    },
+    {
+      "epoch": 2.4069861900893583,
+      "grad_norm": 0.17544260621070862,
+      "learning_rate": 1.1442418028461537e-06,
+      "loss": 0.3711,
+      "step": 2963
+    },
+    {
+      "epoch": 2.4077985377741675,
+      "grad_norm": 0.18478776514530182,
+      "learning_rate": 1.1412340474685946e-06,
+      "loss": 0.3432,
+      "step": 2964
+    },
+    {
+      "epoch": 2.4086108854589763,
+      "grad_norm": 0.19612661004066467,
+      "learning_rate": 1.1382297410455612e-06,
+      "loss": 0.337,
+      "step": 2965
+    },
+    {
+      "epoch": 2.4094232331437855,
+      "grad_norm": 0.1720297485589981,
+      "learning_rate": 1.1352288862622968e-06,
+      "loss": 0.3257,
+      "step": 2966
+    },
+    {
+      "epoch": 2.4102355808285947,
+      "grad_norm": 0.1809699982404709,
+      "learning_rate": 1.1322314858009547e-06,
+      "loss": 0.354,
+      "step": 2967
+    },
+    {
+      "epoch": 2.411047928513404,
+      "grad_norm": 0.16778486967086792,
+      "learning_rate": 1.129237542340606e-06,
+      "loss": 0.3343,
+      "step": 2968
+    },
+    {
+      "epoch": 2.4118602761982126,
+      "grad_norm": 0.19114293158054352,
+      "learning_rate": 1.126247058557226e-06,
+      "loss": 0.3404,
+      "step": 2969
+    },
+    {
+      "epoch": 2.412672623883022,
+      "grad_norm": 0.19056615233421326,
+      "learning_rate": 1.1232600371237035e-06,
+      "loss": 0.3343,
+      "step": 2970
+    },
+    {
+      "epoch": 2.413484971567831,
+      "grad_norm": 0.18827873468399048,
+      "learning_rate": 1.1202764807098305e-06,
+      "loss": 0.3598,
+      "step": 2971
+    },
+    {
+      "epoch": 2.4142973192526402,
+      "grad_norm": 0.17923729121685028,
+      "learning_rate": 1.1172963919822988e-06,
+      "loss": 0.3418,
+      "step": 2972
+    },
+    {
+      "epoch": 2.4151096669374494,
+      "grad_norm": 0.18181850016117096,
+      "learning_rate": 1.1143197736047073e-06,
+      "loss": 0.3379,
+      "step": 2973
+    },
+    {
+      "epoch": 2.415922014622258,
+      "grad_norm": 0.18248872458934784,
+      "learning_rate": 1.1113466282375457e-06,
+      "loss": 0.3378,
+      "step": 2974
+    },
+    {
+      "epoch": 2.4167343623070674,
+      "grad_norm": 0.17363198101520538,
+      "learning_rate": 1.108376958538207e-06,
+      "loss": 0.3465,
+      "step": 2975
+    },
+    {
+      "epoch": 2.4175467099918766,
+      "grad_norm": 0.18106308579444885,
+      "learning_rate": 1.1054107671609697e-06,
+      "loss": 0.3386,
+      "step": 2976
+    },
+    {
+      "epoch": 2.418359057676686,
+      "grad_norm": 0.1792658269405365,
+      "learning_rate": 1.1024480567570123e-06,
+      "loss": 0.3452,
+      "step": 2977
+    },
+    {
+      "epoch": 2.4191714053614946,
+      "grad_norm": 0.1741252988576889,
+      "learning_rate": 1.0994888299743929e-06,
+      "loss": 0.3254,
+      "step": 2978
+    },
+    {
+      "epoch": 2.4199837530463038,
+      "grad_norm": 0.17059576511383057,
+      "learning_rate": 1.0965330894580638e-06,
+      "loss": 0.339,
+      "step": 2979
+    },
+    {
+      "epoch": 2.420796100731113,
+      "grad_norm": 0.17307350039482117,
+      "learning_rate": 1.0935808378498547e-06,
+      "loss": 0.3351,
+      "step": 2980
+    },
+    {
+      "epoch": 2.421608448415922,
+      "grad_norm": 0.17830128967761993,
+      "learning_rate": 1.0906320777884833e-06,
+      "loss": 0.3572,
+      "step": 2981
+    },
+    {
+      "epoch": 2.422420796100731,
+      "grad_norm": 0.18078364431858063,
+      "learning_rate": 1.0876868119095407e-06,
+      "loss": 0.3343,
+      "step": 2982
+    },
+    {
+      "epoch": 2.42323314378554,
+      "grad_norm": 0.17530150711536407,
+      "learning_rate": 1.0847450428454997e-06,
+      "loss": 0.3411,
+      "step": 2983
+    },
+    {
+      "epoch": 2.4240454914703493,
+      "grad_norm": 0.17664070427417755,
+      "learning_rate": 1.0818067732257032e-06,
+      "loss": 0.3498,
+      "step": 2984
+    },
+    {
+      "epoch": 2.4248578391551585,
+      "grad_norm": 0.17368198931217194,
+      "learning_rate": 1.0788720056763712e-06,
+      "loss": 0.3417,
+      "step": 2985
+    },
+    {
+      "epoch": 2.4256701868399677,
+      "grad_norm": 0.17261163890361786,
+      "learning_rate": 1.075940742820588e-06,
+      "loss": 0.3205,
+      "step": 2986
+    },
+    {
+      "epoch": 2.4264825345247765,
+      "grad_norm": 0.19903680682182312,
+      "learning_rate": 1.0730129872783102e-06,
+      "loss": 0.3554,
+      "step": 2987
+    },
+    {
+      "epoch": 2.4272948822095857,
+      "grad_norm": 0.17863957583904266,
+      "learning_rate": 1.0700887416663575e-06,
+      "loss": 0.3256,
+      "step": 2988
+    },
+    {
+      "epoch": 2.428107229894395,
+      "grad_norm": 0.17853491008281708,
+      "learning_rate": 1.0671680085984137e-06,
+      "loss": 0.3585,
+      "step": 2989
+    },
+    {
+      "epoch": 2.428919577579204,
+      "grad_norm": 0.1919107288122177,
+      "learning_rate": 1.0642507906850202e-06,
+      "loss": 0.3434,
+      "step": 2990
+    },
+    {
+      "epoch": 2.429731925264013,
+      "grad_norm": 0.181913360953331,
+      "learning_rate": 1.0613370905335801e-06,
+      "loss": 0.3307,
+      "step": 2991
+    },
+    {
+      "epoch": 2.430544272948822,
+      "grad_norm": 0.17933307588100433,
+      "learning_rate": 1.058426910748349e-06,
+      "loss": 0.3434,
+      "step": 2992
+    },
+    {
+      "epoch": 2.4313566206336312,
+      "grad_norm": 0.1801685094833374,
+      "learning_rate": 1.0555202539304398e-06,
+      "loss": 0.3295,
+      "step": 2993
+    },
+    {
+      "epoch": 2.4321689683184404,
+      "grad_norm": 0.1846536099910736,
+      "learning_rate": 1.052617122677812e-06,
+      "loss": 0.3506,
+      "step": 2994
+    },
+    {
+      "epoch": 2.432981316003249,
+      "grad_norm": 0.18310005962848663,
+      "learning_rate": 1.0497175195852777e-06,
+      "loss": 0.3634,
+      "step": 2995
+    },
+    {
+      "epoch": 2.4337936636880584,
+      "grad_norm": 0.18904468417167664,
+      "learning_rate": 1.0468214472444955e-06,
+      "loss": 0.3358,
+      "step": 2996
+    },
+    {
+      "epoch": 2.4346060113728676,
+      "grad_norm": 0.18425582349300385,
+      "learning_rate": 1.0439289082439647e-06,
+      "loss": 0.3173,
+      "step": 2997
+    },
+    {
+      "epoch": 2.435418359057677,
+      "grad_norm": 0.17741823196411133,
+      "learning_rate": 1.0410399051690306e-06,
+      "loss": 0.3244,
+      "step": 2998
+    },
+    {
+      "epoch": 2.436230706742486,
+      "grad_norm": 0.18424047529697418,
+      "learning_rate": 1.0381544406018751e-06,
+      "loss": 0.3381,
+      "step": 2999
+    },
+    {
+      "epoch": 2.4370430544272947,
+      "grad_norm": 0.1770264059305191,
+      "learning_rate": 1.0352725171215194e-06,
+      "loss": 0.3642,
+      "step": 3000
+    },
+    {
+      "epoch": 2.437855402112104,
+      "grad_norm": 0.17478081583976746,
+      "learning_rate": 1.0323941373038177e-06,
+      "loss": 0.3339,
+      "step": 3001
+    },
+    {
+      "epoch": 2.438667749796913,
+      "grad_norm": 0.18800880014896393,
+      "learning_rate": 1.0295193037214606e-06,
+      "loss": 0.3484,
+      "step": 3002
+    },
+    {
+      "epoch": 2.4394800974817223,
+      "grad_norm": 0.19301719963550568,
+      "learning_rate": 1.0266480189439631e-06,
+      "loss": 0.3484,
+      "step": 3003
+    },
+    {
+      "epoch": 2.440292445166531,
+      "grad_norm": 0.18275128304958344,
+      "learning_rate": 1.0237802855376745e-06,
+      "loss": 0.3246,
+      "step": 3004
+    },
+    {
+      "epoch": 2.4411047928513403,
+      "grad_norm": 0.19416914880275726,
+      "learning_rate": 1.0209161060657651e-06,
+      "loss": 0.3596,
+      "step": 3005
+    },
+    {
+      "epoch": 2.4419171405361495,
+      "grad_norm": 0.1789558231830597,
+      "learning_rate": 1.0180554830882333e-06,
+      "loss": 0.3342,
+      "step": 3006
+    },
+    {
+      "epoch": 2.4427294882209587,
+      "grad_norm": 0.18320375680923462,
+      "learning_rate": 1.015198419161893e-06,
+      "loss": 0.3522,
+      "step": 3007
+    },
+    {
+      "epoch": 2.4435418359057675,
+      "grad_norm": 0.17755351960659027,
+      "learning_rate": 1.0123449168403826e-06,
+      "loss": 0.354,
+      "step": 3008
+    },
+    {
+      "epoch": 2.4443541835905767,
+      "grad_norm": 0.17199918627738953,
+      "learning_rate": 1.0094949786741531e-06,
+      "loss": 0.3369,
+      "step": 3009
+    },
+    {
+      "epoch": 2.445166531275386,
+      "grad_norm": 0.17997734248638153,
+      "learning_rate": 1.0066486072104736e-06,
+      "loss": 0.3441,
+      "step": 3010
+    },
+    {
+      "epoch": 2.445978878960195,
+      "grad_norm": 0.18085622787475586,
+      "learning_rate": 1.0038058049934203e-06,
+      "loss": 0.3556,
+      "step": 3011
+    },
+    {
+      "epoch": 2.4467912266450043,
+      "grad_norm": 0.188663512468338,
+      "learning_rate": 1.0009665745638852e-06,
+      "loss": 0.3274,
+      "step": 3012
+    },
+    {
+      "epoch": 2.447603574329813,
+      "grad_norm": 0.18169471621513367,
+      "learning_rate": 9.981309184595617e-07,
+      "loss": 0.3193,
+      "step": 3013
+    },
+    {
+      "epoch": 2.448415922014622,
+      "grad_norm": 0.17063698172569275,
+      "learning_rate": 9.952988392149532e-07,
+      "loss": 0.3213,
+      "step": 3014
+    },
+    {
+      "epoch": 2.4492282696994314,
+      "grad_norm": 0.18316926062107086,
+      "learning_rate": 9.924703393613655e-07,
+      "loss": 0.3346,
+      "step": 3015
+    },
+    {
+      "epoch": 2.4500406173842406,
+      "grad_norm": 0.16675007343292236,
+      "learning_rate": 9.896454214269014e-07,
+      "loss": 0.3501,
+      "step": 3016
+    },
+    {
+      "epoch": 2.4508529650690494,
+      "grad_norm": 0.1772899180650711,
+      "learning_rate": 9.868240879364676e-07,
+      "loss": 0.3537,
+      "step": 3017
+    },
+    {
+      "epoch": 2.4516653127538586,
+      "grad_norm": 0.17640678584575653,
+      "learning_rate": 9.840063414117618e-07,
+      "loss": 0.3462,
+      "step": 3018
+    },
+    {
+      "epoch": 2.4524776604386678,
+      "grad_norm": 0.1919635683298111,
+      "learning_rate": 9.811921843712813e-07,
+      "loss": 0.3379,
+      "step": 3019
+    },
+    {
+      "epoch": 2.453290008123477,
+      "grad_norm": 0.18285124003887177,
+      "learning_rate": 9.78381619330308e-07,
+      "loss": 0.3378,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4541023558082857,
+      "grad_norm": 0.17984622716903687,
+      "learning_rate": 9.75574648800921e-07,
+      "loss": 0.3581,
+      "step": 3021
+    },
+    {
+      "epoch": 2.454914703493095,
+      "grad_norm": 0.1948661208152771,
+      "learning_rate": 9.727712752919804e-07,
+      "loss": 0.3439,
+      "step": 3022
+    },
+    {
+      "epoch": 2.455727051177904,
+      "grad_norm": 0.17914021015167236,
+      "learning_rate": 9.699715013091354e-07,
+      "loss": 0.3391,
+      "step": 3023
+    },
+    {
+      "epoch": 2.4565393988627133,
+      "grad_norm": 0.18532979488372803,
+      "learning_rate": 9.671753293548153e-07,
+      "loss": 0.3429,
+      "step": 3024
+    },
+    {
+      "epoch": 2.4573517465475225,
+      "grad_norm": 0.1757841259241104,
+      "learning_rate": 9.643827619282309e-07,
+      "loss": 0.3215,
+      "step": 3025
+    },
+    {
+      "epoch": 2.4581640942323313,
+      "grad_norm": 0.18114466965198517,
+      "learning_rate": 9.615938015253723e-07,
+      "loss": 0.3354,
+      "step": 3026
+    },
+    {
+      "epoch": 2.4589764419171405,
+      "grad_norm": 0.1770598441362381,
+      "learning_rate": 9.588084506390055e-07,
+      "loss": 0.3179,
+      "step": 3027
+    },
+    {
+      "epoch": 2.4597887896019497,
+      "grad_norm": 0.1846877485513687,
+      "learning_rate": 9.56026711758667e-07,
+      "loss": 0.3357,
+      "step": 3028
+    },
+    {
+      "epoch": 2.460601137286759,
+      "grad_norm": 0.168557271361351,
+      "learning_rate": 9.532485873706704e-07,
+      "loss": 0.3444,
+      "step": 3029
+    },
+    {
+      "epoch": 2.4614134849715676,
+      "grad_norm": 0.16589036583900452,
+      "learning_rate": 9.504740799580931e-07,
+      "loss": 0.3074,
+      "step": 3030
+    },
+    {
+      "epoch": 2.462225832656377,
+      "grad_norm": 0.18915972113609314,
+      "learning_rate": 9.477031920007845e-07,
+      "loss": 0.3553,
+      "step": 3031
+    },
+    {
+      "epoch": 2.463038180341186,
+      "grad_norm": 0.18364332616329193,
+      "learning_rate": 9.449359259753543e-07,
+      "loss": 0.3382,
+      "step": 3032
+    },
+    {
+      "epoch": 2.4638505280259952,
+      "grad_norm": 0.1853816956281662,
+      "learning_rate": 9.421722843551801e-07,
+      "loss": 0.3691,
+      "step": 3033
+    },
+    {
+      "epoch": 2.464662875710804,
+      "grad_norm": 0.17224419116973877,
+      "learning_rate": 9.394122696103947e-07,
+      "loss": 0.3369,
+      "step": 3034
+    },
+    {
+      "epoch": 2.465475223395613,
+      "grad_norm": 0.16597969830036163,
+      "learning_rate": 9.366558842078943e-07,
+      "loss": 0.3262,
+      "step": 3035
+    },
+    {
+      "epoch": 2.4662875710804224,
+      "grad_norm": 0.18060778081417084,
+      "learning_rate": 9.339031306113256e-07,
+      "loss": 0.3345,
+      "step": 3036
+    },
+    {
+      "epoch": 2.4670999187652316,
+      "grad_norm": 0.17041003704071045,
+      "learning_rate": 9.311540112810957e-07,
+      "loss": 0.3319,
+      "step": 3037
+    },
+    {
+      "epoch": 2.467912266450041,
+      "grad_norm": 0.19015489518642426,
+      "learning_rate": 9.284085286743571e-07,
+      "loss": 0.3228,
+      "step": 3038
+    },
+    {
+      "epoch": 2.4687246141348496,
+      "grad_norm": 0.16801008582115173,
+      "learning_rate": 9.256666852450158e-07,
+      "loss": 0.3226,
+      "step": 3039
+    },
+    {
+      "epoch": 2.4695369618196588,
+      "grad_norm": 0.18515458703041077,
+      "learning_rate": 9.229284834437252e-07,
+      "loss": 0.3312,
+      "step": 3040
+    },
+    {
+      "epoch": 2.470349309504468,
+      "grad_norm": 0.18913587927818298,
+      "learning_rate": 9.20193925717881e-07,
+      "loss": 0.3528,
+      "step": 3041
+    },
+    {
+      "epoch": 2.471161657189277,
+      "grad_norm": 0.1823357343673706,
+      "learning_rate": 9.174630145116248e-07,
+      "loss": 0.3388,
+      "step": 3042
+    },
+    {
+      "epoch": 2.471974004874086,
+      "grad_norm": 0.1713714599609375,
+      "learning_rate": 9.147357522658351e-07,
+      "loss": 0.3338,
+      "step": 3043
+    },
+    {
+      "epoch": 2.472786352558895,
+      "grad_norm": 0.16238875687122345,
+      "learning_rate": 9.120121414181354e-07,
+      "loss": 0.3375,
+      "step": 3044
+    },
+    {
+      "epoch": 2.4735987002437043,
+      "grad_norm": 0.18934985995292664,
+      "learning_rate": 9.092921844028779e-07,
+      "loss": 0.3327,
+      "step": 3045
+    },
+    {
+      "epoch": 2.4744110479285135,
+      "grad_norm": 0.1832655668258667,
+      "learning_rate": 9.065758836511556e-07,
+      "loss": 0.3354,
+      "step": 3046
+    },
+    {
+      "epoch": 2.4752233956133223,
+      "grad_norm": 0.19869759678840637,
+      "learning_rate": 9.038632415907883e-07,
+      "loss": 0.3237,
+      "step": 3047
+    },
+    {
+      "epoch": 2.4760357432981315,
+      "grad_norm": 0.19543825089931488,
+      "learning_rate": 9.011542606463309e-07,
+      "loss": 0.3311,
+      "step": 3048
+    },
+    {
+      "epoch": 2.4768480909829407,
+      "grad_norm": 0.18725082278251648,
+      "learning_rate": 8.984489432390597e-07,
+      "loss": 0.335,
+      "step": 3049
+    },
+    {
+      "epoch": 2.47766043866775,
+      "grad_norm": 0.18799757957458496,
+      "learning_rate": 8.957472917869836e-07,
+      "loss": 0.3393,
+      "step": 3050
+    },
+    {
+      "epoch": 2.478472786352559,
+      "grad_norm": 0.17602214217185974,
+      "learning_rate": 8.930493087048286e-07,
+      "loss": 0.3372,
+      "step": 3051
+    },
+    {
+      "epoch": 2.479285134037368,
+      "grad_norm": 0.18185590207576752,
+      "learning_rate": 8.90354996404047e-07,
+      "loss": 0.33,
+      "step": 3052
+    },
+    {
+      "epoch": 2.480097481722177,
+      "grad_norm": 0.1793457269668579,
+      "learning_rate": 8.876643572928056e-07,
+      "loss": 0.3467,
+      "step": 3053
+    },
+    {
+      "epoch": 2.4809098294069862,
+      "grad_norm": 0.17249809205532074,
+      "learning_rate": 8.84977393775992e-07,
+      "loss": 0.3214,
+      "step": 3054
+    },
+    {
+      "epoch": 2.4817221770917954,
+      "grad_norm": 0.17010989785194397,
+      "learning_rate": 8.822941082552055e-07,
+      "loss": 0.3093,
+      "step": 3055
+    },
+    {
+      "epoch": 2.4825345247766046,
+      "grad_norm": 0.1763264387845993,
+      "learning_rate": 8.796145031287612e-07,
+      "loss": 0.3416,
+      "step": 3056
+    },
+    {
+      "epoch": 2.4833468724614134,
+      "grad_norm": 0.18473419547080994,
+      "learning_rate": 8.769385807916808e-07,
+      "loss": 0.331,
+      "step": 3057
+    },
+    {
+      "epoch": 2.4841592201462226,
+      "grad_norm": 0.1846732646226883,
+      "learning_rate": 8.742663436356969e-07,
+      "loss": 0.3469,
+      "step": 3058
+    },
+    {
+      "epoch": 2.484971567831032,
+      "grad_norm": 0.17163896560668945,
+      "learning_rate": 8.715977940492504e-07,
+      "loss": 0.3476,
+      "step": 3059
+    },
+    {
+      "epoch": 2.4857839155158405,
+      "grad_norm": 0.17901067435741425,
+      "learning_rate": 8.689329344174797e-07,
+      "loss": 0.3562,
+      "step": 3060
+    },
+    {
+      "epoch": 2.4865962632006497,
+      "grad_norm": 0.18028099834918976,
+      "learning_rate": 8.662717671222326e-07,
+      "loss": 0.3433,
+      "step": 3061
+    },
+    {
+      "epoch": 2.487408610885459,
+      "grad_norm": 0.1885780245065689,
+      "learning_rate": 8.6361429454205e-07,
+      "loss": 0.3291,
+      "step": 3062
+    },
+    {
+      "epoch": 2.488220958570268,
+      "grad_norm": 0.1944161206483841,
+      "learning_rate": 8.609605190521752e-07,
+      "loss": 0.3399,
+      "step": 3063
+    },
+    {
+      "epoch": 2.4890333062550773,
+      "grad_norm": 0.16996414959430695,
+      "learning_rate": 8.583104430245476e-07,
+      "loss": 0.3294,
+      "step": 3064
+    },
+    {
+      "epoch": 2.489845653939886,
+      "grad_norm": 0.17831741273403168,
+      "learning_rate": 8.55664068827794e-07,
+      "loss": 0.3193,
+      "step": 3065
+    },
+    {
+      "epoch": 2.4906580016246953,
+      "grad_norm": 0.17355430126190186,
+      "learning_rate": 8.53021398827239e-07,
+      "loss": 0.336,
+      "step": 3066
+    },
+    {
+      "epoch": 2.4914703493095045,
+      "grad_norm": 0.18929772078990936,
+      "learning_rate": 8.503824353848944e-07,
+      "loss": 0.3206,
+      "step": 3067
+    },
+    {
+      "epoch": 2.4922826969943137,
+      "grad_norm": 0.17056943476200104,
+      "learning_rate": 8.477471808594556e-07,
+      "loss": 0.3433,
+      "step": 3068
+    },
+    {
+      "epoch": 2.493095044679123,
+      "grad_norm": 0.19628924131393433,
+      "learning_rate": 8.451156376063096e-07,
+      "loss": 0.3482,
+      "step": 3069
+    },
+    {
+      "epoch": 2.4939073923639317,
+      "grad_norm": 0.19640123844146729,
+      "learning_rate": 8.424878079775195e-07,
+      "loss": 0.3378,
+      "step": 3070
+    },
+    {
+      "epoch": 2.494719740048741,
+      "grad_norm": 0.17787081003189087,
+      "learning_rate": 8.398636943218353e-07,
+      "loss": 0.3303,
+      "step": 3071
+    },
+    {
+      "epoch": 2.49553208773355,
+      "grad_norm": 0.17414399981498718,
+      "learning_rate": 8.37243298984679e-07,
+      "loss": 0.3426,
+      "step": 3072
+    },
+    {
+      "epoch": 2.496344435418359,
+      "grad_norm": 0.18051770329475403,
+      "learning_rate": 8.346266243081563e-07,
+      "loss": 0.3752,
+      "step": 3073
+    },
+    {
+      "epoch": 2.497156783103168,
+      "grad_norm": 0.17315062880516052,
+      "learning_rate": 8.320136726310413e-07,
+      "loss": 0.3215,
+      "step": 3074
+    },
+    {
+      "epoch": 2.497969130787977,
+      "grad_norm": 0.18126212060451508,
+      "learning_rate": 8.294044462887852e-07,
+      "loss": 0.3261,
+      "step": 3075
+    },
+    {
+      "epoch": 2.4987814784727864,
+      "grad_norm": 0.19095557928085327,
+      "learning_rate": 8.267989476135052e-07,
+      "loss": 0.3306,
+      "step": 3076
+    },
+    {
+      "epoch": 2.4995938261575956,
+      "grad_norm": 0.1935988962650299,
+      "learning_rate": 8.24197178933992e-07,
+      "loss": 0.3292,
+      "step": 3077
+    },
+    {
+      "epoch": 2.5004061738424044,
+      "grad_norm": 0.17755883932113647,
+      "learning_rate": 8.215991425756959e-07,
+      "loss": 0.3278,
+      "step": 3078
+    },
+    {
+      "epoch": 2.5012185215272136,
+      "grad_norm": 0.18131813406944275,
+      "learning_rate": 8.190048408607376e-07,
+      "loss": 0.3344,
+      "step": 3079
+    },
+    {
+      "epoch": 2.502030869212023,
+      "grad_norm": 0.17563143372535706,
+      "learning_rate": 8.164142761078941e-07,
+      "loss": 0.3617,
+      "step": 3080
+    },
+    {
+      "epoch": 2.502843216896832,
+      "grad_norm": 0.18351511657238007,
+      "learning_rate": 8.138274506326088e-07,
+      "loss": 0.3428,
+      "step": 3081
+    },
+    {
+      "epoch": 2.503655564581641,
+      "grad_norm": 0.16707070171833038,
+      "learning_rate": 8.112443667469754e-07,
+      "loss": 0.3452,
+      "step": 3082
+    },
+    {
+      "epoch": 2.50446791226645,
+      "grad_norm": 0.18889297544956207,
+      "learning_rate": 8.08665026759749e-07,
+      "loss": 0.3267,
+      "step": 3083
+    },
+    {
+      "epoch": 2.505280259951259,
+      "grad_norm": 0.19080211222171783,
+      "learning_rate": 8.060894329763386e-07,
+      "loss": 0.3513,
+      "step": 3084
+    },
+    {
+      "epoch": 2.5060926076360683,
+      "grad_norm": 0.16487501561641693,
+      "learning_rate": 8.035175876987994e-07,
+      "loss": 0.3308,
+      "step": 3085
+    },
+    {
+      "epoch": 2.506904955320877,
+      "grad_norm": 0.23335939645767212,
+      "learning_rate": 8.009494932258427e-07,
+      "loss": 0.3397,
+      "step": 3086
+    },
+    {
+      "epoch": 2.5077173030056863,
+      "grad_norm": 0.18221263587474823,
+      "learning_rate": 7.983851518528224e-07,
+      "loss": 0.3595,
+      "step": 3087
+    },
+    {
+      "epoch": 2.5085296506904955,
+      "grad_norm": 0.18782977759838104,
+      "learning_rate": 7.958245658717411e-07,
+      "loss": 0.3378,
+      "step": 3088
+    },
+    {
+      "epoch": 2.5093419983753047,
+      "grad_norm": 0.1916222870349884,
+      "learning_rate": 7.932677375712423e-07,
+      "loss": 0.3411,
+      "step": 3089
+    },
+    {
+      "epoch": 2.510154346060114,
+      "grad_norm": 0.1950717270374298,
+      "learning_rate": 7.907146692366135e-07,
+      "loss": 0.3397,
+      "step": 3090
+    },
+    {
+      "epoch": 2.5109666937449227,
+      "grad_norm": 0.17090554535388947,
+      "learning_rate": 7.881653631497793e-07,
+      "loss": 0.3327,
+      "step": 3091
+    },
+    {
+      "epoch": 2.511779041429732,
+      "grad_norm": 0.20299291610717773,
+      "learning_rate": 7.856198215893035e-07,
+      "loss": 0.3357,
+      "step": 3092
+    },
+    {
+      "epoch": 2.512591389114541,
+      "grad_norm": 0.17985275387763977,
+      "learning_rate": 7.830780468303828e-07,
+      "loss": 0.309,
+      "step": 3093
+    },
+    {
+      "epoch": 2.5134037367993503,
+      "grad_norm": 0.18262885510921478,
+      "learning_rate": 7.805400411448505e-07,
+      "loss": 0.3423,
+      "step": 3094
+    },
+    {
+      "epoch": 2.5142160844841595,
+      "grad_norm": 0.16891194880008698,
+      "learning_rate": 7.78005806801167e-07,
+      "loss": 0.3737,
+      "step": 3095
+    },
+    {
+      "epoch": 2.515028432168968,
+      "grad_norm": 0.17340035736560822,
+      "learning_rate": 7.754753460644266e-07,
+      "loss": 0.3421,
+      "step": 3096
+    },
+    {
+      "epoch": 2.5158407798537774,
+      "grad_norm": 0.1823718547821045,
+      "learning_rate": 7.729486611963449e-07,
+      "loss": 0.3296,
+      "step": 3097
+    },
+    {
+      "epoch": 2.5166531275385866,
+      "grad_norm": 0.16918951272964478,
+      "learning_rate": 7.704257544552696e-07,
+      "loss": 0.3317,
+      "step": 3098
+    },
+    {
+      "epoch": 2.5174654752233954,
+      "grad_norm": 0.19431622326374054,
+      "learning_rate": 7.67906628096165e-07,
+      "loss": 0.3526,
+      "step": 3099
+    },
+    {
+      "epoch": 2.5182778229082046,
+      "grad_norm": 0.17819340527057648,
+      "learning_rate": 7.653912843706219e-07,
+      "loss": 0.3625,
+      "step": 3100
+    },
+    {
+      "epoch": 2.5190901705930138,
+      "grad_norm": 0.17150840163230896,
+      "learning_rate": 7.628797255268439e-07,
+      "loss": 0.3727,
+      "step": 3101
+    },
+    {
+      "epoch": 2.519902518277823,
+      "grad_norm": 0.17863328754901886,
+      "learning_rate": 7.603719538096604e-07,
+      "loss": 0.3477,
+      "step": 3102
+    },
+    {
+      "epoch": 2.520714865962632,
+      "grad_norm": 0.1778559535741806,
+      "learning_rate": 7.57867971460507e-07,
+      "loss": 0.3191,
+      "step": 3103
+    },
+    {
+      "epoch": 2.521527213647441,
+      "grad_norm": 0.18575048446655273,
+      "learning_rate": 7.553677807174392e-07,
+      "loss": 0.332,
+      "step": 3104
+    },
+    {
+      "epoch": 2.52233956133225,
+      "grad_norm": 0.16384240984916687,
+      "learning_rate": 7.52871383815118e-07,
+      "loss": 0.3227,
+      "step": 3105
+    },
+    {
+      "epoch": 2.5231519090170593,
+      "grad_norm": 0.16317924857139587,
+      "learning_rate": 7.503787829848191e-07,
+      "loss": 0.336,
+      "step": 3106
+    },
+    {
+      "epoch": 2.5239642567018685,
+      "grad_norm": 0.1824788898229599,
+      "learning_rate": 7.478899804544193e-07,
+      "loss": 0.3515,
+      "step": 3107
+    },
+    {
+      "epoch": 2.5247766043866777,
+      "grad_norm": 0.15330667793750763,
+      "learning_rate": 7.454049784484063e-07,
+      "loss": 0.3196,
+      "step": 3108
+    },
+    {
+      "epoch": 2.5255889520714865,
+      "grad_norm": 0.17183725535869598,
+      "learning_rate": 7.429237791878652e-07,
+      "loss": 0.3249,
+      "step": 3109
+    },
+    {
+      "epoch": 2.5264012997562957,
+      "grad_norm": 0.16835932433605194,
+      "learning_rate": 7.40446384890487e-07,
+      "loss": 0.3456,
+      "step": 3110
+    },
+    {
+      "epoch": 2.527213647441105,
+      "grad_norm": 0.18033276498317719,
+      "learning_rate": 7.379727977705614e-07,
+      "loss": 0.3328,
+      "step": 3111
+    },
+    {
+      "epoch": 2.5280259951259136,
+      "grad_norm": 0.17658378183841705,
+      "learning_rate": 7.355030200389706e-07,
+      "loss": 0.3367,
+      "step": 3112
+    },
+    {
+      "epoch": 2.528838342810723,
+      "grad_norm": 0.18689408898353577,
+      "learning_rate": 7.330370539031978e-07,
+      "loss": 0.3485,
+      "step": 3113
+    },
+    {
+      "epoch": 2.529650690495532,
+      "grad_norm": 0.16749997437000275,
+      "learning_rate": 7.305749015673153e-07,
+      "loss": 0.3349,
+      "step": 3114
+    },
+    {
+      "epoch": 2.5304630381803412,
+      "grad_norm": 0.1772633194923401,
+      "learning_rate": 7.281165652319893e-07,
+      "loss": 0.333,
+      "step": 3115
+    },
+    {
+      "epoch": 2.5312753858651504,
+      "grad_norm": 0.17680130898952484,
+      "learning_rate": 7.25662047094473e-07,
+      "loss": 0.3736,
+      "step": 3116
+    },
+    {
+      "epoch": 2.532087733549959,
+      "grad_norm": 0.16055281460285187,
+      "learning_rate": 7.232113493486099e-07,
+      "loss": 0.3183,
+      "step": 3117
+    },
+    {
+      "epoch": 2.5329000812347684,
+      "grad_norm": 0.1764497458934784,
+      "learning_rate": 7.207644741848241e-07,
+      "loss": 0.3185,
+      "step": 3118
+    },
+    {
+      "epoch": 2.5337124289195776,
+      "grad_norm": 0.1772480458021164,
+      "learning_rate": 7.183214237901287e-07,
+      "loss": 0.3638,
+      "step": 3119
+    },
+    {
+      "epoch": 2.534524776604387,
+      "grad_norm": 0.17999756336212158,
+      "learning_rate": 7.158822003481136e-07,
+      "loss": 0.3516,
+      "step": 3120
+    },
+    {
+      "epoch": 2.535337124289196,
+      "grad_norm": 0.17519594728946686,
+      "learning_rate": 7.134468060389521e-07,
+      "loss": 0.3435,
+      "step": 3121
+    },
+    {
+      "epoch": 2.5361494719740048,
+      "grad_norm": 0.19134707748889923,
+      "learning_rate": 7.110152430393902e-07,
+      "loss": 0.3241,
+      "step": 3122
+    },
+    {
+      "epoch": 2.536961819658814,
+      "grad_norm": 0.16600191593170166,
+      "learning_rate": 7.085875135227555e-07,
+      "loss": 0.3417,
+      "step": 3123
+    },
+    {
+      "epoch": 2.537774167343623,
+      "grad_norm": 0.1737103909254074,
+      "learning_rate": 7.061636196589433e-07,
+      "loss": 0.3426,
+      "step": 3124
+    },
+    {
+      "epoch": 2.538586515028432,
+      "grad_norm": 0.18079333007335663,
+      "learning_rate": 7.037435636144252e-07,
+      "loss": 0.3376,
+      "step": 3125
+    },
+    {
+      "epoch": 2.539398862713241,
+      "grad_norm": 0.16416773200035095,
+      "learning_rate": 7.013273475522392e-07,
+      "loss": 0.344,
+      "step": 3126
+    },
+    {
+      "epoch": 2.5402112103980503,
+      "grad_norm": 0.17515701055526733,
+      "learning_rate": 6.989149736319939e-07,
+      "loss": 0.3562,
+      "step": 3127
+    },
+    {
+      "epoch": 2.5410235580828595,
+      "grad_norm": 0.1666688323020935,
+      "learning_rate": 6.965064440098629e-07,
+      "loss": 0.3293,
+      "step": 3128
+    },
+    {
+      "epoch": 2.5418359057676687,
+      "grad_norm": 0.19385720789432526,
+      "learning_rate": 6.941017608385813e-07,
+      "loss": 0.3242,
+      "step": 3129
+    },
+    {
+      "epoch": 2.5426482534524775,
+      "grad_norm": 0.18202096223831177,
+      "learning_rate": 6.917009262674512e-07,
+      "loss": 0.3426,
+      "step": 3130
+    },
+    {
+      "epoch": 2.5434606011372867,
+      "grad_norm": 0.17699360847473145,
+      "learning_rate": 6.893039424423298e-07,
+      "loss": 0.344,
+      "step": 3131
+    },
+    {
+      "epoch": 2.544272948822096,
+      "grad_norm": 0.1712283045053482,
+      "learning_rate": 6.869108115056361e-07,
+      "loss": 0.3562,
+      "step": 3132
+    },
+    {
+      "epoch": 2.545085296506905,
+      "grad_norm": 0.16445420682430267,
+      "learning_rate": 6.845215355963425e-07,
+      "loss": 0.3431,
+      "step": 3133
+    },
+    {
+      "epoch": 2.5458976441917143,
+      "grad_norm": 0.17720817029476166,
+      "learning_rate": 6.821361168499796e-07,
+      "loss": 0.3449,
+      "step": 3134
+    },
+    {
+      "epoch": 2.546709991876523,
+      "grad_norm": 0.1646021455526352,
+      "learning_rate": 6.797545573986253e-07,
+      "loss": 0.3373,
+      "step": 3135
+    },
+    {
+      "epoch": 2.5475223395613322,
+      "grad_norm": 0.16465988755226135,
+      "learning_rate": 6.77376859370914e-07,
+      "loss": 0.3715,
+      "step": 3136
+    },
+    {
+      "epoch": 2.5483346872461414,
+      "grad_norm": 0.18502001464366913,
+      "learning_rate": 6.750030248920231e-07,
+      "loss": 0.3279,
+      "step": 3137
+    },
+    {
+      "epoch": 2.54914703493095,
+      "grad_norm": 0.16900311410427094,
+      "learning_rate": 6.726330560836813e-07,
+      "loss": 0.3411,
+      "step": 3138
+    },
+    {
+      "epoch": 2.5499593826157594,
+      "grad_norm": 0.1652085781097412,
+      "learning_rate": 6.702669550641589e-07,
+      "loss": 0.3311,
+      "step": 3139
+    },
+    {
+      "epoch": 2.5507717303005686,
+      "grad_norm": 0.15948505699634552,
+      "learning_rate": 6.679047239482711e-07,
+      "loss": 0.3445,
+      "step": 3140
+    },
+    {
+      "epoch": 2.551584077985378,
+      "grad_norm": 0.16511304676532745,
+      "learning_rate": 6.655463648473736e-07,
+      "loss": 0.3332,
+      "step": 3141
+    },
+    {
+      "epoch": 2.552396425670187,
+      "grad_norm": 0.18308746814727783,
+      "learning_rate": 6.631918798693632e-07,
+      "loss": 0.3332,
+      "step": 3142
+    },
+    {
+      "epoch": 2.5532087733549957,
+      "grad_norm": 0.17767249047756195,
+      "learning_rate": 6.60841271118669e-07,
+      "loss": 0.3349,
+      "step": 3143
+    },
+    {
+      "epoch": 2.554021121039805,
+      "grad_norm": 0.17968331277370453,
+      "learning_rate": 6.584945406962617e-07,
+      "loss": 0.3544,
+      "step": 3144
+    },
+    {
+      "epoch": 2.554833468724614,
+      "grad_norm": 0.1753772646188736,
+      "learning_rate": 6.561516906996395e-07,
+      "loss": 0.3348,
+      "step": 3145
+    },
+    {
+      "epoch": 2.5556458164094233,
+      "grad_norm": 0.17655889689922333,
+      "learning_rate": 6.53812723222838e-07,
+      "loss": 0.3538,
+      "step": 3146
+    },
+    {
+      "epoch": 2.5564581640942325,
+      "grad_norm": 0.1694171130657196,
+      "learning_rate": 6.514776403564183e-07,
+      "loss": 0.3434,
+      "step": 3147
+    },
+    {
+      "epoch": 2.5572705117790413,
+      "grad_norm": 0.17572705447673798,
+      "learning_rate": 6.491464441874728e-07,
+      "loss": 0.3341,
+      "step": 3148
+    },
+    {
+      "epoch": 2.5580828594638505,
+      "grad_norm": 0.17185816168785095,
+      "learning_rate": 6.468191367996163e-07,
+      "loss": 0.3378,
+      "step": 3149
+    },
+    {
+      "epoch": 2.5588952071486597,
+      "grad_norm": 0.17797580361366272,
+      "learning_rate": 6.444957202729918e-07,
+      "loss": 0.321,
+      "step": 3150
+    },
+    {
+      "epoch": 2.5597075548334685,
+      "grad_norm": 0.18337340652942657,
+      "learning_rate": 6.421761966842604e-07,
+      "loss": 0.3369,
+      "step": 3151
+    },
+    {
+      "epoch": 2.5605199025182777,
+      "grad_norm": 0.1736641675233841,
+      "learning_rate": 6.398605681066083e-07,
+      "loss": 0.3461,
+      "step": 3152
+    },
+    {
+      "epoch": 2.561332250203087,
+      "grad_norm": 0.16828618943691254,
+      "learning_rate": 6.37548836609736e-07,
+      "loss": 0.3363,
+      "step": 3153
+    },
+    {
+      "epoch": 2.562144597887896,
+      "grad_norm": 0.1754714846611023,
+      "learning_rate": 6.352410042598639e-07,
+      "loss": 0.3423,
+      "step": 3154
+    },
+    {
+      "epoch": 2.5629569455727053,
+      "grad_norm": 0.15649889409542084,
+      "learning_rate": 6.329370731197276e-07,
+      "loss": 0.3556,
+      "step": 3155
+    },
+    {
+      "epoch": 2.563769293257514,
+      "grad_norm": 0.18230096995830536,
+      "learning_rate": 6.306370452485716e-07,
+      "loss": 0.3484,
+      "step": 3156
+    },
+    {
+      "epoch": 2.564581640942323,
+      "grad_norm": 0.17224666476249695,
+      "learning_rate": 6.283409227021575e-07,
+      "loss": 0.336,
+      "step": 3157
+    },
+    {
+      "epoch": 2.5653939886271324,
+      "grad_norm": 0.1669265776872635,
+      "learning_rate": 6.260487075327504e-07,
+      "loss": 0.3441,
+      "step": 3158
+    },
+    {
+      "epoch": 2.5662063363119416,
+      "grad_norm": 0.18116922676563263,
+      "learning_rate": 6.237604017891291e-07,
+      "loss": 0.325,
+      "step": 3159
+    },
+    {
+      "epoch": 2.567018683996751,
+      "grad_norm": 0.17613795399665833,
+      "learning_rate": 6.214760075165721e-07,
+      "loss": 0.3466,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5678310316815596,
+      "grad_norm": 0.1926133632659912,
+      "learning_rate": 6.191955267568678e-07,
+      "loss": 0.3378,
+      "step": 3161
+    },
+    {
+      "epoch": 2.5686433793663688,
+      "grad_norm": 0.16265663504600525,
+      "learning_rate": 6.169189615483007e-07,
+      "loss": 0.3233,
+      "step": 3162
+    },
+    {
+      "epoch": 2.569455727051178,
+      "grad_norm": 0.18894782662391663,
+      "learning_rate": 6.146463139256609e-07,
+      "loss": 0.3127,
+      "step": 3163
+    },
+    {
+      "epoch": 2.5702680747359867,
+      "grad_norm": 0.17088234424591064,
+      "learning_rate": 6.12377585920233e-07,
+      "loss": 0.3348,
+      "step": 3164
+    },
+    {
+      "epoch": 2.5710804224207964,
+      "grad_norm": 0.1820337176322937,
+      "learning_rate": 6.101127795598016e-07,
+      "loss": 0.3309,
+      "step": 3165
+    },
+    {
+      "epoch": 2.571892770105605,
+      "grad_norm": 0.16138805449008942,
+      "learning_rate": 6.078518968686426e-07,
+      "loss": 0.3251,
+      "step": 3166
+    },
+    {
+      "epoch": 2.5727051177904143,
+      "grad_norm": 0.18208256363868713,
+      "learning_rate": 6.055949398675293e-07,
+      "loss": 0.3436,
+      "step": 3167
+    },
+    {
+      "epoch": 2.5735174654752235,
+      "grad_norm": 0.1862913966178894,
+      "learning_rate": 6.033419105737204e-07,
+      "loss": 0.3633,
+      "step": 3168
+    },
+    {
+      "epoch": 2.5743298131600323,
+      "grad_norm": 0.1726512908935547,
+      "learning_rate": 6.010928110009711e-07,
+      "loss": 0.3178,
+      "step": 3169
+    },
+    {
+      "epoch": 2.5751421608448415,
+      "grad_norm": 0.1823875904083252,
+      "learning_rate": 5.98847643159517e-07,
+      "loss": 0.3327,
+      "step": 3170
+    },
+    {
+      "epoch": 2.5759545085296507,
+      "grad_norm": 0.18314602971076965,
+      "learning_rate": 5.966064090560847e-07,
+      "loss": 0.3477,
+      "step": 3171
+    },
+    {
+      "epoch": 2.57676685621446,
+      "grad_norm": 0.17187102138996124,
+      "learning_rate": 5.943691106938843e-07,
+      "loss": 0.3307,
+      "step": 3172
+    },
+    {
+      "epoch": 2.577579203899269,
+      "grad_norm": 0.17291021347045898,
+      "learning_rate": 5.921357500726044e-07,
+      "loss": 0.3261,
+      "step": 3173
+    },
+    {
+      "epoch": 2.578391551584078,
+      "grad_norm": 0.1772870570421219,
+      "learning_rate": 5.899063291884193e-07,
+      "loss": 0.3251,
+      "step": 3174
+    },
+    {
+      "epoch": 2.579203899268887,
+      "grad_norm": 0.17599156498908997,
+      "learning_rate": 5.87680850033977e-07,
+      "loss": 0.3361,
+      "step": 3175
+    },
+    {
+      "epoch": 2.5800162469536962,
+      "grad_norm": 0.17639926075935364,
+      "learning_rate": 5.854593145984067e-07,
+      "loss": 0.3426,
+      "step": 3176
+    },
+    {
+      "epoch": 2.5808285946385054,
+      "grad_norm": 0.1676466017961502,
+      "learning_rate": 5.832417248673089e-07,
+      "loss": 0.3276,
+      "step": 3177
+    },
+    {
+      "epoch": 2.5816409423233146,
+      "grad_norm": 0.17429669201374054,
+      "learning_rate": 5.810280828227599e-07,
+      "loss": 0.351,
+      "step": 3178
+    },
+    {
+      "epoch": 2.5824532900081234,
+      "grad_norm": 0.1665806919336319,
+      "learning_rate": 5.788183904433076e-07,
+      "loss": 0.3309,
+      "step": 3179
+    },
+    {
+      "epoch": 2.5832656376929326,
+      "grad_norm": 0.16565895080566406,
+      "learning_rate": 5.766126497039698e-07,
+      "loss": 0.3332,
+      "step": 3180
+    },
+    {
+      "epoch": 2.584077985377742,
+      "grad_norm": 0.1792539656162262,
+      "learning_rate": 5.744108625762291e-07,
+      "loss": 0.3531,
+      "step": 3181
+    },
+    {
+      "epoch": 2.5848903330625506,
+      "grad_norm": 0.17965464293956757,
+      "learning_rate": 5.722130310280394e-07,
+      "loss": 0.3255,
+      "step": 3182
+    },
+    {
+      "epoch": 2.5857026807473598,
+      "grad_norm": 0.17184710502624512,
+      "learning_rate": 5.70019157023814e-07,
+      "loss": 0.3334,
+      "step": 3183
+    },
+    {
+      "epoch": 2.586515028432169,
+      "grad_norm": 0.17832615971565247,
+      "learning_rate": 5.678292425244347e-07,
+      "loss": 0.3564,
+      "step": 3184
+    },
+    {
+      "epoch": 2.587327376116978,
+      "grad_norm": 0.17243875563144684,
+      "learning_rate": 5.656432894872376e-07,
+      "loss": 0.3431,
+      "step": 3185
+    },
+    {
+      "epoch": 2.5881397238017874,
+      "grad_norm": 0.18398647010326385,
+      "learning_rate": 5.634612998660249e-07,
+      "loss": 0.3334,
+      "step": 3186
+    },
+    {
+      "epoch": 2.588952071486596,
+      "grad_norm": 0.17349156737327576,
+      "learning_rate": 5.612832756110498e-07,
+      "loss": 0.3335,
+      "step": 3187
+    },
+    {
+      "epoch": 2.5897644191714053,
+      "grad_norm": 0.16523019969463348,
+      "learning_rate": 5.591092186690278e-07,
+      "loss": 0.33,
+      "step": 3188
+    },
+    {
+      "epoch": 2.5905767668562145,
+      "grad_norm": 0.17117264866828918,
+      "learning_rate": 5.569391309831218e-07,
+      "loss": 0.3558,
+      "step": 3189
+    },
+    {
+      "epoch": 2.5913891145410237,
+      "grad_norm": 0.17711777985095978,
+      "learning_rate": 5.547730144929531e-07,
+      "loss": 0.3523,
+      "step": 3190
+    },
+    {
+      "epoch": 2.592201462225833,
+      "grad_norm": 0.17078348994255066,
+      "learning_rate": 5.526108711345884e-07,
+      "loss": 0.3177,
+      "step": 3191
+    },
+    {
+      "epoch": 2.5930138099106417,
+      "grad_norm": 0.1742868721485138,
+      "learning_rate": 5.504527028405482e-07,
+      "loss": 0.3244,
+      "step": 3192
+    },
+    {
+      "epoch": 2.593826157595451,
+      "grad_norm": 0.19203737378120422,
+      "learning_rate": 5.482985115397943e-07,
+      "loss": 0.3478,
+      "step": 3193
+    },
+    {
+      "epoch": 2.59463850528026,
+      "grad_norm": 0.16968604922294617,
+      "learning_rate": 5.461482991577399e-07,
+      "loss": 0.3323,
+      "step": 3194
+    },
+    {
+      "epoch": 2.595450852965069,
+      "grad_norm": 0.1733492761850357,
+      "learning_rate": 5.440020676162367e-07,
+      "loss": 0.3256,
+      "step": 3195
+    },
+    {
+      "epoch": 2.596263200649878,
+      "grad_norm": 0.1735830008983612,
+      "learning_rate": 5.418598188335816e-07,
+      "loss": 0.3446,
+      "step": 3196
+    },
+    {
+      "epoch": 2.5970755483346872,
+      "grad_norm": 0.18031729757785797,
+      "learning_rate": 5.397215547245116e-07,
+      "loss": 0.3353,
+      "step": 3197
+    },
+    {
+      "epoch": 2.5978878960194964,
+      "grad_norm": 0.1724662184715271,
+      "learning_rate": 5.375872772002e-07,
+      "loss": 0.3301,
+      "step": 3198
+    },
+    {
+      "epoch": 2.5987002437043056,
+      "grad_norm": 0.175658717751503,
+      "learning_rate": 5.354569881682592e-07,
+      "loss": 0.3275,
+      "step": 3199
+    },
+    {
+      "epoch": 2.5995125913891144,
+      "grad_norm": 0.16623395681381226,
+      "learning_rate": 5.33330689532735e-07,
+      "loss": 0.3411,
+      "step": 3200
+    },
+    {
+      "epoch": 2.6003249390739236,
+      "grad_norm": 0.18601422011852264,
+      "learning_rate": 5.312083831941084e-07,
+      "loss": 0.3301,
+      "step": 3201
+    },
+    {
+      "epoch": 2.601137286758733,
+      "grad_norm": 0.17198805510997772,
+      "learning_rate": 5.290900710492891e-07,
+      "loss": 0.3442,
+      "step": 3202
+    },
+    {
+      "epoch": 2.601949634443542,
+      "grad_norm": 0.1760576069355011,
+      "learning_rate": 5.269757549916216e-07,
+      "loss": 0.3408,
+      "step": 3203
+    },
+    {
+      "epoch": 2.602761982128351,
+      "grad_norm": 0.16578637063503265,
+      "learning_rate": 5.24865436910873e-07,
+      "loss": 0.3677,
+      "step": 3204
+    },
+    {
+      "epoch": 2.60357432981316,
+      "grad_norm": 0.18038761615753174,
+      "learning_rate": 5.227591186932418e-07,
+      "loss": 0.3311,
+      "step": 3205
+    },
+    {
+      "epoch": 2.604386677497969,
+      "grad_norm": 0.21716651320457458,
+      "learning_rate": 5.206568022213482e-07,
+      "loss": 0.3294,
+      "step": 3206
+    },
+    {
+      "epoch": 2.6051990251827783,
+      "grad_norm": 0.18020156025886536,
+      "learning_rate": 5.185584893742379e-07,
+      "loss": 0.3179,
+      "step": 3207
+    },
+    {
+      "epoch": 2.606011372867587,
+      "grad_norm": 0.17517174780368805,
+      "learning_rate": 5.164641820273764e-07,
+      "loss": 0.3234,
+      "step": 3208
+    },
+    {
+      "epoch": 2.6068237205523963,
+      "grad_norm": 0.1591583490371704,
+      "learning_rate": 5.143738820526517e-07,
+      "loss": 0.3362,
+      "step": 3209
+    },
+    {
+      "epoch": 2.6076360682372055,
+      "grad_norm": 0.17372798919677734,
+      "learning_rate": 5.122875913183656e-07,
+      "loss": 0.3464,
+      "step": 3210
+    },
+    {
+      "epoch": 2.6084484159220147,
+      "grad_norm": 0.16566318273544312,
+      "learning_rate": 5.102053116892414e-07,
+      "loss": 0.33,
+      "step": 3211
+    },
+    {
+      "epoch": 2.609260763606824,
+      "grad_norm": 0.17878791689872742,
+      "learning_rate": 5.081270450264125e-07,
+      "loss": 0.3631,
+      "step": 3212
+    },
+    {
+      "epoch": 2.6100731112916327,
+      "grad_norm": 0.1784702092409134,
+      "learning_rate": 5.060527931874309e-07,
+      "loss": 0.3048,
+      "step": 3213
+    },
+    {
+      "epoch": 2.610885458976442,
+      "grad_norm": 0.18168620765209198,
+      "learning_rate": 5.039825580262553e-07,
+      "loss": 0.3603,
+      "step": 3214
+    },
+    {
+      "epoch": 2.611697806661251,
+      "grad_norm": 0.18354642391204834,
+      "learning_rate": 5.019163413932565e-07,
+      "loss": 0.3387,
+      "step": 3215
+    },
+    {
+      "epoch": 2.6125101543460603,
+      "grad_norm": 0.17604851722717285,
+      "learning_rate": 4.998541451352134e-07,
+      "loss": 0.3411,
+      "step": 3216
+    },
+    {
+      "epoch": 2.6133225020308695,
+      "grad_norm": 0.16872385144233704,
+      "learning_rate": 4.977959710953128e-07,
+      "loss": 0.3291,
+      "step": 3217
+    },
+    {
+      "epoch": 2.614134849715678,
+      "grad_norm": 0.1720227599143982,
+      "learning_rate": 4.957418211131432e-07,
+      "loss": 0.3357,
+      "step": 3218
+    },
+    {
+      "epoch": 2.6149471974004874,
+      "grad_norm": 0.17956313490867615,
+      "learning_rate": 4.936916970246997e-07,
+      "loss": 0.3425,
+      "step": 3219
+    },
+    {
+      "epoch": 2.6157595450852966,
+      "grad_norm": 0.1722511202096939,
+      "learning_rate": 4.91645600662376e-07,
+      "loss": 0.3341,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6165718927701054,
+      "grad_norm": 0.1700020432472229,
+      "learning_rate": 4.896035338549693e-07,
+      "loss": 0.3354,
+      "step": 3221
+    },
+    {
+      "epoch": 2.6173842404549146,
+      "grad_norm": 0.17739498615264893,
+      "learning_rate": 4.875654984276718e-07,
+      "loss": 0.3194,
+      "step": 3222
+    },
+    {
+      "epoch": 2.618196588139724,
+      "grad_norm": 0.18680332601070404,
+      "learning_rate": 4.855314962020746e-07,
+      "loss": 0.3485,
+      "step": 3223
+    },
+    {
+      "epoch": 2.619008935824533,
+      "grad_norm": 0.18406620621681213,
+      "learning_rate": 4.835015289961642e-07,
+      "loss": 0.3575,
+      "step": 3224
+    },
+    {
+      "epoch": 2.619821283509342,
+      "grad_norm": 0.16914711892604828,
+      "learning_rate": 4.81475598624318e-07,
+      "loss": 0.3387,
+      "step": 3225
+    },
+    {
+      "epoch": 2.620633631194151,
+      "grad_norm": 0.21210598945617676,
+      "learning_rate": 4.794537068973093e-07,
+      "loss": 0.3513,
+      "step": 3226
+    },
+    {
+      "epoch": 2.62144597887896,
+      "grad_norm": 0.1974373608827591,
+      "learning_rate": 4.774358556222974e-07,
+      "loss": 0.3329,
+      "step": 3227
+    },
+    {
+      "epoch": 2.6222583265637693,
+      "grad_norm": 0.18106713891029358,
+      "learning_rate": 4.7542204660283363e-07,
+      "loss": 0.3435,
+      "step": 3228
+    },
+    {
+      "epoch": 2.6230706742485785,
+      "grad_norm": 0.17550316452980042,
+      "learning_rate": 4.7341228163885424e-07,
+      "loss": 0.3519,
+      "step": 3229
+    },
+    {
+      "epoch": 2.6238830219333877,
+      "grad_norm": 0.1688171625137329,
+      "learning_rate": 4.7140656252668226e-07,
+      "loss": 0.3394,
+      "step": 3230
+    },
+    {
+      "epoch": 2.6246953696181965,
+      "grad_norm": 0.16755716502666473,
+      "learning_rate": 4.6940489105902323e-07,
+      "loss": 0.3368,
+      "step": 3231
+    },
+    {
+      "epoch": 2.6255077173030057,
+      "grad_norm": 0.16448429226875305,
+      "learning_rate": 4.67407269024967e-07,
+      "loss": 0.3473,
+      "step": 3232
+    },
+    {
+      "epoch": 2.626320064987815,
+      "grad_norm": 0.17386433482170105,
+      "learning_rate": 4.654136982099805e-07,
+      "loss": 0.3521,
+      "step": 3233
+    },
+    {
+      "epoch": 2.6271324126726237,
+      "grad_norm": 0.1657935231924057,
+      "learning_rate": 4.634241803959144e-07,
+      "loss": 0.3386,
+      "step": 3234
+    },
+    {
+      "epoch": 2.627944760357433,
+      "grad_norm": 0.17407283186912537,
+      "learning_rate": 4.6143871736099246e-07,
+      "loss": 0.3371,
+      "step": 3235
+    },
+    {
+      "epoch": 2.628757108042242,
+      "grad_norm": 0.18489646911621094,
+      "learning_rate": 4.5945731087981783e-07,
+      "loss": 0.3461,
+      "step": 3236
+    },
+    {
+      "epoch": 2.6295694557270513,
+      "grad_norm": 0.178299680352211,
+      "learning_rate": 4.574799627233639e-07,
+      "loss": 0.358,
+      "step": 3237
+    },
+    {
+      "epoch": 2.6303818034118605,
+      "grad_norm": 0.1594986915588379,
+      "learning_rate": 4.5550667465898204e-07,
+      "loss": 0.3565,
+      "step": 3238
+    },
+    {
+      "epoch": 2.631194151096669,
+      "grad_norm": 0.17217151820659637,
+      "learning_rate": 4.535374484503885e-07,
+      "loss": 0.3449,
+      "step": 3239
+    },
+    {
+      "epoch": 2.6320064987814784,
+      "grad_norm": 0.17462515830993652,
+      "learning_rate": 4.515722858576743e-07,
+      "loss": 0.3451,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6328188464662876,
+      "grad_norm": 0.1841244250535965,
+      "learning_rate": 4.4961118863729725e-07,
+      "loss": 0.3348,
+      "step": 3241
+    },
+    {
+      "epoch": 2.633631194151097,
+      "grad_norm": 0.16596297919750214,
+      "learning_rate": 4.476541585420785e-07,
+      "loss": 0.3518,
+      "step": 3242
+    },
+    {
+      "epoch": 2.634443541835906,
+      "grad_norm": 0.17588528990745544,
+      "learning_rate": 4.457011973212083e-07,
+      "loss": 0.3513,
+      "step": 3243
+    },
+    {
+      "epoch": 2.6352558895207148,
+      "grad_norm": 0.1814480721950531,
+      "learning_rate": 4.437523067202365e-07,
+      "loss": 0.3331,
+      "step": 3244
+    },
+    {
+      "epoch": 2.636068237205524,
+      "grad_norm": 0.17102524638175964,
+      "learning_rate": 4.4180748848107745e-07,
+      "loss": 0.3238,
+      "step": 3245
+    },
+    {
+      "epoch": 2.636880584890333,
+      "grad_norm": 0.18116185069084167,
+      "learning_rate": 4.398667443420029e-07,
+      "loss": 0.3394,
+      "step": 3246
+    },
+    {
+      "epoch": 2.637692932575142,
+      "grad_norm": 0.16158704459667206,
+      "learning_rate": 4.379300760376465e-07,
+      "loss": 0.3441,
+      "step": 3247
+    },
+    {
+      "epoch": 2.638505280259951,
+      "grad_norm": 0.17871388792991638,
+      "learning_rate": 4.359974852989951e-07,
+      "loss": 0.3344,
+      "step": 3248
+    },
+    {
+      "epoch": 2.6393176279447603,
+      "grad_norm": 0.17713786661624908,
+      "learning_rate": 4.3406897385339544e-07,
+      "loss": 0.3318,
+      "step": 3249
+    },
+    {
+      "epoch": 2.6401299756295695,
+      "grad_norm": 0.17349779605865479,
+      "learning_rate": 4.3214454342454357e-07,
+      "loss": 0.331,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6409423233143787,
+      "grad_norm": 0.1696607619524002,
+      "learning_rate": 4.302241957324915e-07,
+      "loss": 0.3785,
+      "step": 3251
+    },
+    {
+      "epoch": 2.6417546709991875,
+      "grad_norm": 0.16446122527122498,
+      "learning_rate": 4.283079324936401e-07,
+      "loss": 0.3338,
+      "step": 3252
+    },
+    {
+      "epoch": 2.6425670186839967,
+      "grad_norm": 0.18211811780929565,
+      "learning_rate": 4.263957554207415e-07,
+      "loss": 0.3186,
+      "step": 3253
+    },
+    {
+      "epoch": 2.643379366368806,
+      "grad_norm": 0.17557822167873383,
+      "learning_rate": 4.244876662228914e-07,
+      "loss": 0.3307,
+      "step": 3254
+    },
+    {
+      "epoch": 2.644191714053615,
+      "grad_norm": 0.18266832828521729,
+      "learning_rate": 4.2258366660553885e-07,
+      "loss": 0.3296,
+      "step": 3255
+    },
+    {
+      "epoch": 2.6450040617384243,
+      "grad_norm": 0.1776798665523529,
+      "learning_rate": 4.206837582704704e-07,
+      "loss": 0.3231,
+      "step": 3256
+    },
+    {
+      "epoch": 2.645816409423233,
+      "grad_norm": 0.22492127120494843,
+      "learning_rate": 4.187879429158204e-07,
+      "loss": 0.3644,
+      "step": 3257
+    },
+    {
+      "epoch": 2.6466287571080422,
+      "grad_norm": 0.17288881540298462,
+      "learning_rate": 4.1689622223606283e-07,
+      "loss": 0.3329,
+      "step": 3258
+    },
+    {
+      "epoch": 2.6474411047928514,
+      "grad_norm": 0.1833348572254181,
+      "learning_rate": 4.1500859792201363e-07,
+      "loss": 0.3382,
+      "step": 3259
+    },
+    {
+      "epoch": 2.64825345247766,
+      "grad_norm": 0.15958520770072937,
+      "learning_rate": 4.1312507166082425e-07,
+      "loss": 0.336,
+      "step": 3260
+    },
+    {
+      "epoch": 2.6490658001624694,
+      "grad_norm": 0.16342513263225555,
+      "learning_rate": 4.1124564513598755e-07,
+      "loss": 0.3384,
+      "step": 3261
+    },
+    {
+      "epoch": 2.6498781478472786,
+      "grad_norm": 0.17694011330604553,
+      "learning_rate": 4.09370320027328e-07,
+      "loss": 0.3358,
+      "step": 3262
+    },
+    {
+      "epoch": 2.650690495532088,
+      "grad_norm": 0.16496531665325165,
+      "learning_rate": 4.0749909801100775e-07,
+      "loss": 0.3319,
+      "step": 3263
+    },
+    {
+      "epoch": 2.651502843216897,
+      "grad_norm": 0.18786412477493286,
+      "learning_rate": 4.0563198075951824e-07,
+      "loss": 0.3321,
+      "step": 3264
+    },
+    {
+      "epoch": 2.6523151909017058,
+      "grad_norm": 0.16537027060985565,
+      "learning_rate": 4.037689699416858e-07,
+      "loss": 0.3403,
+      "step": 3265
+    },
+    {
+      "epoch": 2.653127538586515,
+      "grad_norm": 0.1738942712545395,
+      "learning_rate": 4.019100672226617e-07,
+      "loss": 0.335,
+      "step": 3266
+    },
+    {
+      "epoch": 2.653939886271324,
+      "grad_norm": 0.1755458265542984,
+      "learning_rate": 4.000552742639302e-07,
+      "loss": 0.3739,
+      "step": 3267
+    },
+    {
+      "epoch": 2.6547522339561334,
+      "grad_norm": 0.1743946522474289,
+      "learning_rate": 3.982045927233008e-07,
+      "loss": 0.3568,
+      "step": 3268
+    },
+    {
+      "epoch": 2.6555645816409426,
+      "grad_norm": 0.18358829617500305,
+      "learning_rate": 3.963580242549053e-07,
+      "loss": 0.3146,
+      "step": 3269
+    },
+    {
+      "epoch": 2.6563769293257513,
+      "grad_norm": 0.16691046953201294,
+      "learning_rate": 3.9451557050920406e-07,
+      "loss": 0.3086,
+      "step": 3270
+    },
+    {
+      "epoch": 2.6571892770105605,
+      "grad_norm": 0.17378726601600647,
+      "learning_rate": 3.9267723313297535e-07,
+      "loss": 0.3445,
+      "step": 3271
+    },
+    {
+      "epoch": 2.6580016246953697,
+      "grad_norm": 0.17037120461463928,
+      "learning_rate": 3.9084301376932143e-07,
+      "loss": 0.3519,
+      "step": 3272
+    },
+    {
+      "epoch": 2.6588139723801785,
+      "grad_norm": 0.18074087798595428,
+      "learning_rate": 3.8901291405766097e-07,
+      "loss": 0.364,
+      "step": 3273
+    },
+    {
+      "epoch": 2.6596263200649877,
+      "grad_norm": 0.16735582053661346,
+      "learning_rate": 3.8718693563373376e-07,
+      "loss": 0.3518,
+      "step": 3274
+    },
+    {
+      "epoch": 2.660438667749797,
+      "grad_norm": 0.15548273921012878,
+      "learning_rate": 3.853650801295933e-07,
+      "loss": 0.3332,
+      "step": 3275
+    },
+    {
+      "epoch": 2.661251015434606,
+      "grad_norm": 0.16619595885276794,
+      "learning_rate": 3.8354734917361024e-07,
+      "loss": 0.3609,
+      "step": 3276
+    },
+    {
+      "epoch": 2.6620633631194153,
+      "grad_norm": 0.17247936129570007,
+      "learning_rate": 3.817337443904662e-07,
+      "loss": 0.3409,
+      "step": 3277
+    },
+    {
+      "epoch": 2.662875710804224,
+      "grad_norm": 0.16212189197540283,
+      "learning_rate": 3.799242674011572e-07,
+      "loss": 0.3352,
+      "step": 3278
+    },
+    {
+      "epoch": 2.6636880584890332,
+      "grad_norm": 0.1766878068447113,
+      "learning_rate": 3.7811891982298843e-07,
+      "loss": 0.3282,
+      "step": 3279
+    },
+    {
+      "epoch": 2.6645004061738424,
+      "grad_norm": 0.17718113958835602,
+      "learning_rate": 3.7631770326957475e-07,
+      "loss": 0.339,
+      "step": 3280
+    },
+    {
+      "epoch": 2.6653127538586516,
+      "grad_norm": 0.1784893274307251,
+      "learning_rate": 3.7452061935083826e-07,
+      "loss": 0.3419,
+      "step": 3281
+    },
+    {
+      "epoch": 2.666125101543461,
+      "grad_norm": 0.17668229341506958,
+      "learning_rate": 3.7272766967300864e-07,
+      "loss": 0.3409,
+      "step": 3282
+    },
+    {
+      "epoch": 2.6669374492282696,
+      "grad_norm": 0.19247283041477203,
+      "learning_rate": 3.709388558386179e-07,
+      "loss": 0.3448,
+      "step": 3283
+    },
+    {
+      "epoch": 2.667749796913079,
+      "grad_norm": 0.187876895070076,
+      "learning_rate": 3.691541794465042e-07,
+      "loss": 0.3196,
+      "step": 3284
+    },
+    {
+      "epoch": 2.668562144597888,
+      "grad_norm": 0.17168372869491577,
+      "learning_rate": 3.673736420918067e-07,
+      "loss": 0.3375,
+      "step": 3285
+    },
+    {
+      "epoch": 2.6693744922826967,
+      "grad_norm": 0.1650395393371582,
+      "learning_rate": 3.65597245365964e-07,
+      "loss": 0.3274,
+      "step": 3286
+    },
+    {
+      "epoch": 2.670186839967506,
+      "grad_norm": 0.16993574798107147,
+      "learning_rate": 3.638249908567154e-07,
+      "loss": 0.3462,
+      "step": 3287
+    },
+    {
+      "epoch": 2.670999187652315,
+      "grad_norm": 0.1902117282152176,
+      "learning_rate": 3.6205688014809615e-07,
+      "loss": 0.34,
+      "step": 3288
+    },
+    {
+      "epoch": 2.6718115353371243,
+      "grad_norm": 0.15195420384407043,
+      "learning_rate": 3.602929148204398e-07,
+      "loss": 0.3259,
+      "step": 3289
+    },
+    {
+      "epoch": 2.6726238830219335,
+      "grad_norm": 0.16937190294265747,
+      "learning_rate": 3.5853309645037306e-07,
+      "loss": 0.3324,
+      "step": 3290
+    },
+    {
+      "epoch": 2.6734362307067423,
+      "grad_norm": 0.17342647910118103,
+      "learning_rate": 3.567774266108176e-07,
+      "loss": 0.3601,
+      "step": 3291
+    },
+    {
+      "epoch": 2.6742485783915515,
+      "grad_norm": 0.16924655437469482,
+      "learning_rate": 3.550259068709844e-07,
+      "loss": 0.3343,
+      "step": 3292
+    },
+    {
+      "epoch": 2.6750609260763607,
+      "grad_norm": 0.17719601094722748,
+      "learning_rate": 3.5327853879637994e-07,
+      "loss": 0.3413,
+      "step": 3293
+    },
+    {
+      "epoch": 2.67587327376117,
+      "grad_norm": 0.16743668913841248,
+      "learning_rate": 3.5153532394879507e-07,
+      "loss": 0.3392,
+      "step": 3294
+    },
+    {
+      "epoch": 2.676685621445979,
+      "grad_norm": 0.18809351325035095,
+      "learning_rate": 3.4979626388631174e-07,
+      "loss": 0.3517,
+      "step": 3295
+    },
+    {
+      "epoch": 2.677497969130788,
+      "grad_norm": 0.1739170253276825,
+      "learning_rate": 3.4806136016329537e-07,
+      "loss": 0.3361,
+      "step": 3296
+    },
+    {
+      "epoch": 2.678310316815597,
+      "grad_norm": 0.17206750810146332,
+      "learning_rate": 3.4633061433040014e-07,
+      "loss": 0.3441,
+      "step": 3297
+    },
+    {
+      "epoch": 2.6791226645004063,
+      "grad_norm": 0.17231225967407227,
+      "learning_rate": 3.4460402793456083e-07,
+      "loss": 0.3387,
+      "step": 3298
+    },
+    {
+      "epoch": 2.679935012185215,
+      "grad_norm": 0.1686933934688568,
+      "learning_rate": 3.4288160251899625e-07,
+      "loss": 0.3304,
+      "step": 3299
+    },
+    {
+      "epoch": 2.680747359870024,
+      "grad_norm": 0.17131495475769043,
+      "learning_rate": 3.4116333962320445e-07,
+      "loss": 0.3203,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6815597075548334,
+      "grad_norm": 0.18447710573673248,
+      "learning_rate": 3.394492407829658e-07,
+      "loss": 0.3651,
+      "step": 3301
+    },
+    {
+      "epoch": 2.6823720552396426,
+      "grad_norm": 0.18501834571361542,
+      "learning_rate": 3.3773930753033535e-07,
+      "loss": 0.3266,
+      "step": 3302
+    },
+    {
+      "epoch": 2.683184402924452,
+      "grad_norm": 0.16900694370269775,
+      "learning_rate": 3.360335413936489e-07,
+      "loss": 0.3394,
+      "step": 3303
+    },
+    {
+      "epoch": 2.6839967506092606,
+      "grad_norm": 0.1805695742368698,
+      "learning_rate": 3.343319438975134e-07,
+      "loss": 0.3412,
+      "step": 3304
+    },
+    {
+      "epoch": 2.6848090982940698,
+      "grad_norm": 0.17752397060394287,
+      "learning_rate": 3.3263451656281364e-07,
+      "loss": 0.323,
+      "step": 3305
+    },
+    {
+      "epoch": 2.685621445978879,
+      "grad_norm": 0.17056338489055634,
+      "learning_rate": 3.3094126090670477e-07,
+      "loss": 0.325,
+      "step": 3306
+    },
+    {
+      "epoch": 2.686433793663688,
+      "grad_norm": 0.17152434587478638,
+      "learning_rate": 3.2925217844261503e-07,
+      "loss": 0.3435,
+      "step": 3307
+    },
+    {
+      "epoch": 2.6872461413484974,
+      "grad_norm": 0.179646298289299,
+      "learning_rate": 3.275672706802402e-07,
+      "loss": 0.3582,
+      "step": 3308
+    },
+    {
+      "epoch": 2.688058489033306,
+      "grad_norm": 0.17010807991027832,
+      "learning_rate": 3.258865391255478e-07,
+      "loss": 0.3539,
+      "step": 3309
+    },
+    {
+      "epoch": 2.6888708367181153,
+      "grad_norm": 0.19829419255256653,
+      "learning_rate": 3.242099852807695e-07,
+      "loss": 0.3377,
+      "step": 3310
+    },
+    {
+      "epoch": 2.6896831844029245,
+      "grad_norm": 0.17125728726387024,
+      "learning_rate": 3.225376106444056e-07,
+      "loss": 0.3551,
+      "step": 3311
+    },
+    {
+      "epoch": 2.6904955320877333,
+      "grad_norm": 0.1694766730070114,
+      "learning_rate": 3.2086941671122084e-07,
+      "loss": 0.3601,
+      "step": 3312
+    },
+    {
+      "epoch": 2.6913078797725425,
+      "grad_norm": 0.17748236656188965,
+      "learning_rate": 3.1920540497224095e-07,
+      "loss": 0.3182,
+      "step": 3313
+    },
+    {
+      "epoch": 2.6921202274573517,
+      "grad_norm": 0.17460255324840546,
+      "learning_rate": 3.1754557691475617e-07,
+      "loss": 0.3491,
+      "step": 3314
+    },
+    {
+      "epoch": 2.692932575142161,
+      "grad_norm": 0.18233874440193176,
+      "learning_rate": 3.158899340223154e-07,
+      "loss": 0.3454,
+      "step": 3315
+    },
+    {
+      "epoch": 2.69374492282697,
+      "grad_norm": 0.19256389141082764,
+      "learning_rate": 3.142384777747287e-07,
+      "loss": 0.3389,
+      "step": 3316
+    },
+    {
+      "epoch": 2.694557270511779,
+      "grad_norm": 0.173871248960495,
+      "learning_rate": 3.125912096480621e-07,
+      "loss": 0.3355,
+      "step": 3317
+    },
+    {
+      "epoch": 2.695369618196588,
+      "grad_norm": 0.1650243103504181,
+      "learning_rate": 3.1094813111464004e-07,
+      "loss": 0.3145,
+      "step": 3318
+    },
+    {
+      "epoch": 2.6961819658813972,
+      "grad_norm": 0.16139931976795197,
+      "learning_rate": 3.0930924364304117e-07,
+      "loss": 0.3237,
+      "step": 3319
+    },
+    {
+      "epoch": 2.6969943135662064,
+      "grad_norm": 0.17425020039081573,
+      "learning_rate": 3.076745486981003e-07,
+      "loss": 0.3316,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6978066612510156,
+      "grad_norm": 0.185983806848526,
+      "learning_rate": 3.060440477409005e-07,
+      "loss": 0.342,
+      "step": 3321
+    },
+    {
+      "epoch": 2.6986190089358244,
+      "grad_norm": 0.1678110659122467,
+      "learning_rate": 3.0441774222878194e-07,
+      "loss": 0.3346,
+      "step": 3322
+    },
+    {
+      "epoch": 2.6994313566206336,
+      "grad_norm": 0.18859970569610596,
+      "learning_rate": 3.027956336153298e-07,
+      "loss": 0.3476,
+      "step": 3323
+    },
+    {
+      "epoch": 2.700243704305443,
+      "grad_norm": 0.1779106706380844,
+      "learning_rate": 3.011777233503821e-07,
+      "loss": 0.3463,
+      "step": 3324
+    },
+    {
+      "epoch": 2.7010560519902516,
+      "grad_norm": 0.16367895901203156,
+      "learning_rate": 2.995640128800209e-07,
+      "loss": 0.328,
+      "step": 3325
+    },
+    {
+      "epoch": 2.7018683996750608,
+      "grad_norm": 0.1745673418045044,
+      "learning_rate": 2.9795450364657865e-07,
+      "loss": 0.3614,
+      "step": 3326
+    },
+    {
+      "epoch": 2.70268074735987,
+      "grad_norm": 0.17031517624855042,
+      "learning_rate": 2.963491970886273e-07,
+      "loss": 0.3306,
+      "step": 3327
+    },
+    {
+      "epoch": 2.703493095044679,
+      "grad_norm": 0.17382757365703583,
+      "learning_rate": 2.947480946409875e-07,
+      "loss": 0.3341,
+      "step": 3328
+    },
+    {
+      "epoch": 2.7043054427294884,
+      "grad_norm": 0.16126902401447296,
+      "learning_rate": 2.9315119773472067e-07,
+      "loss": 0.3368,
+      "step": 3329
+    },
+    {
+      "epoch": 2.705117790414297,
+      "grad_norm": 0.17267869412899017,
+      "learning_rate": 2.9155850779712723e-07,
+      "loss": 0.3623,
+      "step": 3330
+    },
+    {
+      "epoch": 2.7059301380991063,
+      "grad_norm": 0.1696598082780838,
+      "learning_rate": 2.899700262517502e-07,
+      "loss": 0.3231,
+      "step": 3331
+    },
+    {
+      "epoch": 2.7067424857839155,
+      "grad_norm": 0.17520111799240112,
+      "learning_rate": 2.8838575451837046e-07,
+      "loss": 0.3393,
+      "step": 3332
+    },
+    {
+      "epoch": 2.7075548334687247,
+      "grad_norm": 0.1601610779762268,
+      "learning_rate": 2.868056940130048e-07,
+      "loss": 0.3263,
+      "step": 3333
+    },
+    {
+      "epoch": 2.708367181153534,
+      "grad_norm": 0.1674559861421585,
+      "learning_rate": 2.852298461479075e-07,
+      "loss": 0.3508,
+      "step": 3334
+    },
+    {
+      "epoch": 2.7091795288383427,
+      "grad_norm": 0.17282532155513763,
+      "learning_rate": 2.8365821233156677e-07,
+      "loss": 0.3267,
+      "step": 3335
+    },
+    {
+      "epoch": 2.709991876523152,
+      "grad_norm": 0.2057839334011078,
+      "learning_rate": 2.8209079396870454e-07,
+      "loss": 0.3645,
+      "step": 3336
+    },
+    {
+      "epoch": 2.710804224207961,
+      "grad_norm": 0.16768312454223633,
+      "learning_rate": 2.8052759246027626e-07,
+      "loss": 0.3518,
+      "step": 3337
+    },
+    {
+      "epoch": 2.71161657189277,
+      "grad_norm": 0.18222440779209137,
+      "learning_rate": 2.7896860920346605e-07,
+      "loss": 0.3482,
+      "step": 3338
+    },
+    {
+      "epoch": 2.7124289195775795,
+      "grad_norm": 0.17751635611057281,
+      "learning_rate": 2.774138455916897e-07,
+      "loss": 0.3336,
+      "step": 3339
+    },
+    {
+      "epoch": 2.7132412672623882,
+      "grad_norm": 0.17399737238883972,
+      "learning_rate": 2.758633030145891e-07,
+      "loss": 0.3441,
+      "step": 3340
+    },
+    {
+      "epoch": 2.7140536149471974,
+      "grad_norm": 0.17185451090335846,
+      "learning_rate": 2.743169828580372e-07,
+      "loss": 0.3501,
+      "step": 3341
+    },
+    {
+      "epoch": 2.7148659626320066,
+      "grad_norm": 0.17467281222343445,
+      "learning_rate": 2.727748865041291e-07,
+      "loss": 0.3332,
+      "step": 3342
+    },
+    {
+      "epoch": 2.7156783103168154,
+      "grad_norm": 0.1646127849817276,
+      "learning_rate": 2.712370153311872e-07,
+      "loss": 0.3322,
+      "step": 3343
+    },
+    {
+      "epoch": 2.7164906580016246,
+      "grad_norm": 0.17819376289844513,
+      "learning_rate": 2.6970337071375584e-07,
+      "loss": 0.3258,
+      "step": 3344
+    },
+    {
+      "epoch": 2.717303005686434,
+      "grad_norm": 0.1667768657207489,
+      "learning_rate": 2.681739540226036e-07,
+      "loss": 0.3537,
+      "step": 3345
+    },
+    {
+      "epoch": 2.718115353371243,
+      "grad_norm": 0.17270100116729736,
+      "learning_rate": 2.6664876662471697e-07,
+      "loss": 0.3452,
+      "step": 3346
+    },
+    {
+      "epoch": 2.718927701056052,
+      "grad_norm": 0.1658916175365448,
+      "learning_rate": 2.6512780988330654e-07,
+      "loss": 0.333,
+      "step": 3347
+    },
+    {
+      "epoch": 2.719740048740861,
+      "grad_norm": 0.16423659026622772,
+      "learning_rate": 2.6361108515779777e-07,
+      "loss": 0.3504,
+      "step": 3348
+    },
+    {
+      "epoch": 2.72055239642567,
+      "grad_norm": 0.1589057594537735,
+      "learning_rate": 2.6209859380383673e-07,
+      "loss": 0.3392,
+      "step": 3349
+    },
+    {
+      "epoch": 2.7213647441104794,
+      "grad_norm": 0.16956721246242523,
+      "learning_rate": 2.6059033717328154e-07,
+      "loss": 0.3462,
+      "step": 3350
+    },
+    {
+      "epoch": 2.722177091795288,
+      "grad_norm": 0.192663311958313,
+      "learning_rate": 2.590863166142105e-07,
+      "loss": 0.3254,
+      "step": 3351
+    },
+    {
+      "epoch": 2.7229894394800978,
+      "grad_norm": 0.17895929515361786,
+      "learning_rate": 2.5758653347091103e-07,
+      "loss": 0.3345,
+      "step": 3352
+    },
+    {
+      "epoch": 2.7238017871649065,
+      "grad_norm": 0.17127051949501038,
+      "learning_rate": 2.560909890838864e-07,
+      "loss": 0.3282,
+      "step": 3353
+    },
+    {
+      "epoch": 2.7246141348497157,
+      "grad_norm": 0.17665603756904602,
+      "learning_rate": 2.5459968478985074e-07,
+      "loss": 0.35,
+      "step": 3354
+    },
+    {
+      "epoch": 2.725426482534525,
+      "grad_norm": 0.15274031460285187,
+      "learning_rate": 2.531126219217256e-07,
+      "loss": 0.34,
+      "step": 3355
+    },
+    {
+      "epoch": 2.7262388302193337,
+      "grad_norm": 0.1716984063386917,
+      "learning_rate": 2.5162980180864605e-07,
+      "loss": 0.3449,
+      "step": 3356
+    },
+    {
+      "epoch": 2.727051177904143,
+      "grad_norm": 0.17302429676055908,
+      "learning_rate": 2.5015122577595084e-07,
+      "loss": 0.3425,
+      "step": 3357
+    },
+    {
+      "epoch": 2.727863525588952,
+      "grad_norm": 0.1649889498949051,
+      "learning_rate": 2.48676895145189e-07,
+      "loss": 0.339,
+      "step": 3358
+    },
+    {
+      "epoch": 2.7286758732737613,
+      "grad_norm": 0.1701282411813736,
+      "learning_rate": 2.472068112341114e-07,
+      "loss": 0.3427,
+      "step": 3359
+    },
+    {
+      "epoch": 2.7294882209585705,
+      "grad_norm": 0.18499287962913513,
+      "learning_rate": 2.4574097535667696e-07,
+      "loss": 0.3367,
+      "step": 3360
+    },
+    {
+      "epoch": 2.7303005686433792,
+      "grad_norm": 0.17386947572231293,
+      "learning_rate": 2.4427938882304383e-07,
+      "loss": 0.3632,
+      "step": 3361
+    },
+    {
+      "epoch": 2.7311129163281884,
+      "grad_norm": 0.1793205887079239,
+      "learning_rate": 2.428220529395753e-07,
+      "loss": 0.3553,
+      "step": 3362
+    },
+    {
+      "epoch": 2.7319252640129976,
+      "grad_norm": 0.17017847299575806,
+      "learning_rate": 2.413689690088339e-07,
+      "loss": 0.3409,
+      "step": 3363
+    },
+    {
+      "epoch": 2.732737611697807,
+      "grad_norm": 0.17719526588916779,
+      "learning_rate": 2.399201383295824e-07,
+      "loss": 0.3492,
+      "step": 3364
+    },
+    {
+      "epoch": 2.733549959382616,
+      "grad_norm": 0.18211428821086884,
+      "learning_rate": 2.3847556219678105e-07,
+      "loss": 0.3559,
+      "step": 3365
+    },
+    {
+      "epoch": 2.734362307067425,
+      "grad_norm": 0.16936899721622467,
+      "learning_rate": 2.370352419015892e-07,
+      "loss": 0.3439,
+      "step": 3366
+    },
+    {
+      "epoch": 2.735174654752234,
+      "grad_norm": 0.16700021922588348,
+      "learning_rate": 2.3559917873135996e-07,
+      "loss": 0.3327,
+      "step": 3367
+    },
+    {
+      "epoch": 2.735987002437043,
+      "grad_norm": 0.1747794896364212,
+      "learning_rate": 2.3416737396964373e-07,
+      "loss": 0.3163,
+      "step": 3368
+    },
+    {
+      "epoch": 2.736799350121852,
+      "grad_norm": 0.1975403130054474,
+      "learning_rate": 2.3273982889618186e-07,
+      "loss": 0.3581,
+      "step": 3369
+    },
+    {
+      "epoch": 2.737611697806661,
+      "grad_norm": 0.1789219230413437,
+      "learning_rate": 2.3131654478691258e-07,
+      "loss": 0.3433,
+      "step": 3370
+    },
+    {
+      "epoch": 2.7384240454914703,
+      "grad_norm": 0.17237426340579987,
+      "learning_rate": 2.2989752291396218e-07,
+      "loss": 0.3317,
+      "step": 3371
+    },
+    {
+      "epoch": 2.7392363931762795,
+      "grad_norm": 0.1727220118045807,
+      "learning_rate": 2.2848276454564943e-07,
+      "loss": 0.3318,
+      "step": 3372
+    },
+    {
+      "epoch": 2.7400487408610887,
+      "grad_norm": 0.18127276003360748,
+      "learning_rate": 2.270722709464801e-07,
+      "loss": 0.3457,
+      "step": 3373
+    },
+    {
+      "epoch": 2.7408610885458975,
+      "grad_norm": 0.1693258285522461,
+      "learning_rate": 2.2566604337715127e-07,
+      "loss": 0.324,
+      "step": 3374
+    },
+    {
+      "epoch": 2.7416734362307067,
+      "grad_norm": 0.16555818915367126,
+      "learning_rate": 2.242640830945436e-07,
+      "loss": 0.3561,
+      "step": 3375
+    },
+    {
+      "epoch": 2.742485783915516,
+      "grad_norm": 0.17436178028583527,
+      "learning_rate": 2.228663913517265e-07,
+      "loss": 0.3596,
+      "step": 3376
+    },
+    {
+      "epoch": 2.743298131600325,
+      "grad_norm": 0.17412234842777252,
+      "learning_rate": 2.2147296939795226e-07,
+      "loss": 0.3382,
+      "step": 3377
+    },
+    {
+      "epoch": 2.7441104792851343,
+      "grad_norm": 0.17754004895687103,
+      "learning_rate": 2.2008381847865911e-07,
+      "loss": 0.3478,
+      "step": 3378
+    },
+    {
+      "epoch": 2.744922826969943,
+      "grad_norm": 0.17008231580257416,
+      "learning_rate": 2.1869893983546442e-07,
+      "loss": 0.3373,
+      "step": 3379
+    },
+    {
+      "epoch": 2.7457351746547523,
+      "grad_norm": 0.19424036145210266,
+      "learning_rate": 2.173183347061697e-07,
+      "loss": 0.341,
+      "step": 3380
+    },
+    {
+      "epoch": 2.7465475223395615,
+      "grad_norm": 0.17514194548130035,
+      "learning_rate": 2.1594200432475675e-07,
+      "loss": 0.341,
+      "step": 3381
+    },
+    {
+      "epoch": 2.74735987002437,
+      "grad_norm": 0.1712406724691391,
+      "learning_rate": 2.145699499213849e-07,
+      "loss": 0.3413,
+      "step": 3382
+    },
+    {
+      "epoch": 2.7481722177091794,
+      "grad_norm": 0.1732463389635086,
+      "learning_rate": 2.132021727223943e-07,
+      "loss": 0.3379,
+      "step": 3383
+    },
+    {
+      "epoch": 2.7489845653939886,
+      "grad_norm": 0.16992147266864777,
+      "learning_rate": 2.1183867395029812e-07,
+      "loss": 0.3505,
+      "step": 3384
+    },
+    {
+      "epoch": 2.749796913078798,
+      "grad_norm": 0.1826762557029724,
+      "learning_rate": 2.1047945482378985e-07,
+      "loss": 0.3392,
+      "step": 3385
+    },
+    {
+      "epoch": 2.750609260763607,
+      "grad_norm": 0.16906915605068207,
+      "learning_rate": 2.091245165577349e-07,
+      "loss": 0.341,
+      "step": 3386
+    },
+    {
+      "epoch": 2.7514216084484158,
+      "grad_norm": 0.17001153528690338,
+      "learning_rate": 2.0777386036317404e-07,
+      "loss": 0.3392,
+      "step": 3387
+    },
+    {
+      "epoch": 2.752233956133225,
+      "grad_norm": 0.1595863252878189,
+      "learning_rate": 2.0642748744731878e-07,
+      "loss": 0.3097,
+      "step": 3388
+    },
+    {
+      "epoch": 2.753046303818034,
+      "grad_norm": 0.15934428572654724,
+      "learning_rate": 2.050853990135554e-07,
+      "loss": 0.3346,
+      "step": 3389
+    },
+    {
+      "epoch": 2.7538586515028434,
+      "grad_norm": 0.1680375188589096,
+      "learning_rate": 2.0374759626143714e-07,
+      "loss": 0.3514,
+      "step": 3390
+    },
+    {
+      "epoch": 2.7546709991876526,
+      "grad_norm": 0.16608387231826782,
+      "learning_rate": 2.0241408038668976e-07,
+      "loss": 0.3249,
+      "step": 3391
+    },
+    {
+      "epoch": 2.7554833468724613,
+      "grad_norm": 0.16605138778686523,
+      "learning_rate": 2.0108485258120536e-07,
+      "loss": 0.3405,
+      "step": 3392
+    },
+    {
+      "epoch": 2.7562956945572705,
+      "grad_norm": 0.17663165926933289,
+      "learning_rate": 1.9975991403304529e-07,
+      "loss": 0.3176,
+      "step": 3393
+    },
+    {
+      "epoch": 2.7571080422420797,
+      "grad_norm": 0.20068401098251343,
+      "learning_rate": 1.9843926592643392e-07,
+      "loss": 0.348,
+      "step": 3394
+    },
+    {
+      "epoch": 2.7579203899268885,
+      "grad_norm": 0.17273786664009094,
+      "learning_rate": 1.9712290944176537e-07,
+      "loss": 0.3505,
+      "step": 3395
+    },
+    {
+      "epoch": 2.7587327376116977,
+      "grad_norm": 0.16676771640777588,
+      "learning_rate": 1.9581084575559352e-07,
+      "loss": 0.3393,
+      "step": 3396
+    },
+    {
+      "epoch": 2.759545085296507,
+      "grad_norm": 0.1717795729637146,
+      "learning_rate": 1.94503076040638e-07,
+      "loss": 0.3488,
+      "step": 3397
+    },
+    {
+      "epoch": 2.760357432981316,
+      "grad_norm": 0.17079754173755646,
+      "learning_rate": 1.9319960146578055e-07,
+      "loss": 0.3505,
+      "step": 3398
+    },
+    {
+      "epoch": 2.7611697806661253,
+      "grad_norm": 0.17131462693214417,
+      "learning_rate": 1.919004231960625e-07,
+      "loss": 0.3175,
+      "step": 3399
+    },
+    {
+      "epoch": 2.761982128350934,
+      "grad_norm": 0.18056423962116241,
+      "learning_rate": 1.9060554239268613e-07,
+      "loss": 0.3377,
+      "step": 3400
+    },
+    {
+      "epoch": 2.7627944760357432,
+      "grad_norm": 0.17499056458473206,
+      "learning_rate": 1.8931496021301178e-07,
+      "loss": 0.341,
+      "step": 3401
+    },
+    {
+      "epoch": 2.7636068237205524,
+      "grad_norm": 0.1658184826374054,
+      "learning_rate": 1.8802867781055943e-07,
+      "loss": 0.3166,
+      "step": 3402
+    },
+    {
+      "epoch": 2.7644191714053616,
+      "grad_norm": 0.17809660732746124,
+      "learning_rate": 1.8674669633500342e-07,
+      "loss": 0.3387,
+      "step": 3403
+    },
+    {
+      "epoch": 2.765231519090171,
+      "grad_norm": 0.1794024407863617,
+      "learning_rate": 1.8546901693217655e-07,
+      "loss": 0.3344,
+      "step": 3404
+    },
+    {
+      "epoch": 2.7660438667749796,
+      "grad_norm": 0.1777004450559616,
+      "learning_rate": 1.8419564074406427e-07,
+      "loss": 0.3558,
+      "step": 3405
+    },
+    {
+      "epoch": 2.766856214459789,
+      "grad_norm": 0.1698988527059555,
+      "learning_rate": 1.8292656890880722e-07,
+      "loss": 0.343,
+      "step": 3406
+    },
+    {
+      "epoch": 2.767668562144598,
+      "grad_norm": 0.16787488758563995,
+      "learning_rate": 1.8166180256069753e-07,
+      "loss": 0.3231,
+      "step": 3407
+    },
+    {
+      "epoch": 2.7684809098294068,
+      "grad_norm": 0.16986756026744843,
+      "learning_rate": 1.8040134283018207e-07,
+      "loss": 0.3428,
+      "step": 3408
+    },
+    {
+      "epoch": 2.769293257514216,
+      "grad_norm": 0.15704527497291565,
+      "learning_rate": 1.7914519084385407e-07,
+      "loss": 0.3352,
+      "step": 3409
+    },
+    {
+      "epoch": 2.770105605199025,
+      "grad_norm": 0.1737799197435379,
+      "learning_rate": 1.7789334772445987e-07,
+      "loss": 0.3412,
+      "step": 3410
+    },
+    {
+      "epoch": 2.7709179528838344,
+      "grad_norm": 0.1854993999004364,
+      "learning_rate": 1.7664581459089335e-07,
+      "loss": 0.3415,
+      "step": 3411
+    },
+    {
+      "epoch": 2.7717303005686436,
+      "grad_norm": 0.16324488818645477,
+      "learning_rate": 1.754025925581959e-07,
+      "loss": 0.3446,
+      "step": 3412
+    },
+    {
+      "epoch": 2.7725426482534523,
+      "grad_norm": 0.18013517558574677,
+      "learning_rate": 1.7416368273755646e-07,
+      "loss": 0.352,
+      "step": 3413
+    },
+    {
+      "epoch": 2.7733549959382615,
+      "grad_norm": 0.19459854066371918,
+      "learning_rate": 1.7292908623630866e-07,
+      "loss": 0.3344,
+      "step": 3414
+    },
+    {
+      "epoch": 2.7741673436230707,
+      "grad_norm": 0.1738632470369339,
+      "learning_rate": 1.7169880415793206e-07,
+      "loss": 0.3536,
+      "step": 3415
+    },
+    {
+      "epoch": 2.77497969130788,
+      "grad_norm": 0.1701546162366867,
+      "learning_rate": 1.7047283760204925e-07,
+      "loss": 0.3412,
+      "step": 3416
+    },
+    {
+      "epoch": 2.775792038992689,
+      "grad_norm": 0.18425250053405762,
+      "learning_rate": 1.6925118766442538e-07,
+      "loss": 0.3369,
+      "step": 3417
+    },
+    {
+      "epoch": 2.776604386677498,
+      "grad_norm": 0.18728294968605042,
+      "learning_rate": 1.6803385543696872e-07,
+      "loss": 0.3387,
+      "step": 3418
+    },
+    {
+      "epoch": 2.777416734362307,
+      "grad_norm": 0.18155620992183685,
+      "learning_rate": 1.6682084200772663e-07,
+      "loss": 0.326,
+      "step": 3419
+    },
+    {
+      "epoch": 2.7782290820471163,
+      "grad_norm": 0.1819823831319809,
+      "learning_rate": 1.65612148460888e-07,
+      "loss": 0.3317,
+      "step": 3420
+    },
+    {
+      "epoch": 2.779041429731925,
+      "grad_norm": 0.18677079677581787,
+      "learning_rate": 1.6440777587677924e-07,
+      "loss": 0.3573,
+      "step": 3421
+    },
+    {
+      "epoch": 2.7798537774167342,
+      "grad_norm": 0.20521953701972961,
+      "learning_rate": 1.6320772533186536e-07,
+      "loss": 0.3428,
+      "step": 3422
+    },
+    {
+      "epoch": 2.7806661251015434,
+      "grad_norm": 0.165541410446167,
+      "learning_rate": 1.6201199789874833e-07,
+      "loss": 0.3346,
+      "step": 3423
+    },
+    {
+      "epoch": 2.7814784727863526,
+      "grad_norm": 0.16370461881160736,
+      "learning_rate": 1.6082059464616662e-07,
+      "loss": 0.3589,
+      "step": 3424
+    },
+    {
+      "epoch": 2.782290820471162,
+      "grad_norm": 0.16716153919696808,
+      "learning_rate": 1.5963351663899284e-07,
+      "loss": 0.3396,
+      "step": 3425
+    },
+    {
+      "epoch": 2.7831031681559706,
+      "grad_norm": 0.1835746169090271,
+      "learning_rate": 1.5845076493823331e-07,
+      "loss": 0.3385,
+      "step": 3426
+    },
+    {
+      "epoch": 2.78391551584078,
+      "grad_norm": 0.16560205817222595,
+      "learning_rate": 1.572723406010296e-07,
+      "loss": 0.3344,
+      "step": 3427
+    },
+    {
+      "epoch": 2.784727863525589,
+      "grad_norm": 0.18238085508346558,
+      "learning_rate": 1.5609824468065306e-07,
+      "loss": 0.3329,
+      "step": 3428
+    },
+    {
+      "epoch": 2.785540211210398,
+      "grad_norm": 0.17051947116851807,
+      "learning_rate": 1.549284782265087e-07,
+      "loss": 0.3408,
+      "step": 3429
+    },
+    {
+      "epoch": 2.7863525588952074,
+      "grad_norm": 0.16310739517211914,
+      "learning_rate": 1.537630422841291e-07,
+      "loss": 0.3301,
+      "step": 3430
+    },
+    {
+      "epoch": 2.787164906580016,
+      "grad_norm": 0.17702160775661469,
+      "learning_rate": 1.526019378951793e-07,
+      "loss": 0.3508,
+      "step": 3431
+    },
+    {
+      "epoch": 2.7879772542648253,
+      "grad_norm": 0.1876174360513687,
+      "learning_rate": 1.5144516609744975e-07,
+      "loss": 0.3376,
+      "step": 3432
+    },
+    {
+      "epoch": 2.7887896019496345,
+      "grad_norm": 0.17953813076019287,
+      "learning_rate": 1.5029272792486116e-07,
+      "loss": 0.3408,
+      "step": 3433
+    },
+    {
+      "epoch": 2.7896019496344433,
+      "grad_norm": 0.19048582017421722,
+      "learning_rate": 1.4914462440745847e-07,
+      "loss": 0.3493,
+      "step": 3434
+    },
+    {
+      "epoch": 2.7904142973192525,
+      "grad_norm": 0.1747557669878006,
+      "learning_rate": 1.480008565714147e-07,
+      "loss": 0.3454,
+      "step": 3435
+    },
+    {
+      "epoch": 2.7912266450040617,
+      "grad_norm": 0.1843181848526001,
+      "learning_rate": 1.4686142543902548e-07,
+      "loss": 0.3375,
+      "step": 3436
+    },
+    {
+      "epoch": 2.792038992688871,
+      "grad_norm": 0.1811114102602005,
+      "learning_rate": 1.4572633202871222e-07,
+      "loss": 0.3356,
+      "step": 3437
+    },
+    {
+      "epoch": 2.79285134037368,
+      "grad_norm": 0.1699988692998886,
+      "learning_rate": 1.4459557735501671e-07,
+      "loss": 0.3512,
+      "step": 3438
+    },
+    {
+      "epoch": 2.793663688058489,
+      "grad_norm": 0.17693445086479187,
+      "learning_rate": 1.4346916242860664e-07,
+      "loss": 0.3235,
+      "step": 3439
+    },
+    {
+      "epoch": 2.794476035743298,
+      "grad_norm": 0.16166749596595764,
+      "learning_rate": 1.4234708825626664e-07,
+      "loss": 0.34,
+      "step": 3440
+    },
+    {
+      "epoch": 2.7952883834281073,
+      "grad_norm": 0.17706045508384705,
+      "learning_rate": 1.4122935584090447e-07,
+      "loss": 0.3309,
+      "step": 3441
+    },
+    {
+      "epoch": 2.7961007311129165,
+      "grad_norm": 0.15944308042526245,
+      "learning_rate": 1.4011596618154655e-07,
+      "loss": 0.3503,
+      "step": 3442
+    },
+    {
+      "epoch": 2.7969130787977257,
+      "grad_norm": 0.18206733465194702,
+      "learning_rate": 1.3900692027333628e-07,
+      "loss": 0.3257,
+      "step": 3443
+    },
+    {
+      "epoch": 2.7977254264825344,
+      "grad_norm": 0.18252019584178925,
+      "learning_rate": 1.37902219107538e-07,
+      "loss": 0.3557,
+      "step": 3444
+    },
+    {
+      "epoch": 2.7985377741673436,
+      "grad_norm": 0.1714627593755722,
+      "learning_rate": 1.3680186367152848e-07,
+      "loss": 0.3515,
+      "step": 3445
+    },
+    {
+      "epoch": 2.799350121852153,
+      "grad_norm": 0.16286300122737885,
+      "learning_rate": 1.3570585494880328e-07,
+      "loss": 0.3434,
+      "step": 3446
+    },
+    {
+      "epoch": 2.8001624695369616,
+      "grad_norm": 0.17028257250785828,
+      "learning_rate": 1.3461419391897268e-07,
+      "loss": 0.3438,
+      "step": 3447
+    },
+    {
+      "epoch": 2.8009748172217708,
+      "grad_norm": 0.17940400540828705,
+      "learning_rate": 1.3352688155775896e-07,
+      "loss": 0.3403,
+      "step": 3448
+    },
+    {
+      "epoch": 2.80178716490658,
+      "grad_norm": 0.17609933018684387,
+      "learning_rate": 1.3244391883700025e-07,
+      "loss": 0.3476,
+      "step": 3449
+    },
+    {
+      "epoch": 2.802599512591389,
+      "grad_norm": 0.1769014149904251,
+      "learning_rate": 1.3136530672464564e-07,
+      "loss": 0.3602,
+      "step": 3450
+    },
+    {
+      "epoch": 2.8034118602761984,
+      "grad_norm": 0.1628442108631134,
+      "learning_rate": 1.3029104618475397e-07,
+      "loss": 0.3392,
+      "step": 3451
+    },
+    {
+      "epoch": 2.804224207961007,
+      "grad_norm": 0.164116770029068,
+      "learning_rate": 1.292211381774988e-07,
+      "loss": 0.3283,
+      "step": 3452
+    },
+    {
+      "epoch": 2.8050365556458163,
+      "grad_norm": 0.16578687727451324,
+      "learning_rate": 1.2815558365915916e-07,
+      "loss": 0.3226,
+      "step": 3453
+    },
+    {
+      "epoch": 2.8058489033306255,
+      "grad_norm": 0.16905303299427032,
+      "learning_rate": 1.2709438358212644e-07,
+      "loss": 0.3232,
+      "step": 3454
+    },
+    {
+      "epoch": 2.8066612510154347,
+      "grad_norm": 0.1690799593925476,
+      "learning_rate": 1.2603753889489645e-07,
+      "loss": 0.3613,
+      "step": 3455
+    },
+    {
+      "epoch": 2.807473598700244,
+      "grad_norm": 0.1783326268196106,
+      "learning_rate": 1.2498505054207578e-07,
+      "loss": 0.3362,
+      "step": 3456
+    },
+    {
+      "epoch": 2.8082859463850527,
+      "grad_norm": 0.16798356175422668,
+      "learning_rate": 1.2393691946437414e-07,
+      "loss": 0.3364,
+      "step": 3457
+    },
+    {
+      "epoch": 2.809098294069862,
+      "grad_norm": 0.17550289630889893,
+      "learning_rate": 1.228931465986094e-07,
+      "loss": 0.3658,
+      "step": 3458
+    },
+    {
+      "epoch": 2.809910641754671,
+      "grad_norm": 0.1718290150165558,
+      "learning_rate": 1.2185373287770254e-07,
+      "loss": 0.3489,
+      "step": 3459
+    },
+    {
+      "epoch": 2.81072298943948,
+      "grad_norm": 0.15857884287834167,
+      "learning_rate": 1.2081867923067925e-07,
+      "loss": 0.3405,
+      "step": 3460
+    },
+    {
+      "epoch": 2.811535337124289,
+      "grad_norm": 0.17132586240768433,
+      "learning_rate": 1.1978798658266678e-07,
+      "loss": 0.3525,
+      "step": 3461
+    },
+    {
+      "epoch": 2.8123476848090982,
+      "grad_norm": 0.17074228823184967,
+      "learning_rate": 1.1876165585489596e-07,
+      "loss": 0.3683,
+      "step": 3462
+    },
+    {
+      "epoch": 2.8131600324939074,
+      "grad_norm": 0.1732131987810135,
+      "learning_rate": 1.1773968796469748e-07,
+      "loss": 0.3531,
+      "step": 3463
+    },
+    {
+      "epoch": 2.8139723801787166,
+      "grad_norm": 0.163502037525177,
+      "learning_rate": 1.1672208382550509e-07,
+      "loss": 0.339,
+      "step": 3464
+    },
+    {
+      "epoch": 2.8147847278635254,
+      "grad_norm": 0.18306565284729004,
+      "learning_rate": 1.1570884434684959e-07,
+      "loss": 0.3177,
+      "step": 3465
+    },
+    {
+      "epoch": 2.8155970755483346,
+      "grad_norm": 0.1676400899887085,
+      "learning_rate": 1.1469997043436154e-07,
+      "loss": 0.3262,
+      "step": 3466
+    },
+    {
+      "epoch": 2.816409423233144,
+      "grad_norm": 0.18637576699256897,
+      "learning_rate": 1.1369546298977018e-07,
+      "loss": 0.311,
+      "step": 3467
+    },
+    {
+      "epoch": 2.817221770917953,
+      "grad_norm": 0.17321625351905823,
+      "learning_rate": 1.1269532291090124e-07,
+      "loss": 0.3616,
+      "step": 3468
+    },
+    {
+      "epoch": 2.818034118602762,
+      "grad_norm": 0.1734827160835266,
+      "learning_rate": 1.1169955109167741e-07,
+      "loss": 0.3296,
+      "step": 3469
+    },
+    {
+      "epoch": 2.818846466287571,
+      "grad_norm": 0.17088426649570465,
+      "learning_rate": 1.1070814842211675e-07,
+      "loss": 0.3413,
+      "step": 3470
+    },
+    {
+      "epoch": 2.81965881397238,
+      "grad_norm": 0.17042440176010132,
+      "learning_rate": 1.0972111578833266e-07,
+      "loss": 0.3454,
+      "step": 3471
+    },
+    {
+      "epoch": 2.8204711616571894,
+      "grad_norm": 0.17211608588695526,
+      "learning_rate": 1.0873845407253114e-07,
+      "loss": 0.3352,
+      "step": 3472
+    },
+    {
+      "epoch": 2.821283509341998,
+      "grad_norm": 0.17747366428375244,
+      "learning_rate": 1.0776016415301405e-07,
+      "loss": 0.338,
+      "step": 3473
+    },
+    {
+      "epoch": 2.8220958570268073,
+      "grad_norm": 0.16217266023159027,
+      "learning_rate": 1.067862469041725e-07,
+      "loss": 0.3386,
+      "step": 3474
+    },
+    {
+      "epoch": 2.8229082047116165,
+      "grad_norm": 0.162852942943573,
+      "learning_rate": 1.0581670319649295e-07,
+      "loss": 0.3311,
+      "step": 3475
+    },
+    {
+      "epoch": 2.8237205523964257,
+      "grad_norm": 0.1789432317018509,
+      "learning_rate": 1.0485153389654945e-07,
+      "loss": 0.3459,
+      "step": 3476
+    },
+    {
+      "epoch": 2.824532900081235,
+      "grad_norm": 0.16955526173114777,
+      "learning_rate": 1.038907398670086e-07,
+      "loss": 0.3564,
+      "step": 3477
+    },
+    {
+      "epoch": 2.8253452477660437,
+      "grad_norm": 0.17996205389499664,
+      "learning_rate": 1.0293432196662512e-07,
+      "loss": 0.3352,
+      "step": 3478
+    },
+    {
+      "epoch": 2.826157595450853,
+      "grad_norm": 0.17074958980083466,
+      "learning_rate": 1.01982281050243e-07,
+      "loss": 0.3467,
+      "step": 3479
+    },
+    {
+      "epoch": 2.826969943135662,
+      "grad_norm": 0.1742672324180603,
+      "learning_rate": 1.0103461796879265e-07,
+      "loss": 0.3207,
+      "step": 3480
+    },
+    {
+      "epoch": 2.8277822908204713,
+      "grad_norm": 0.17090949416160583,
+      "learning_rate": 1.000913335692949e-07,
+      "loss": 0.3345,
+      "step": 3481
+    },
+    {
+      "epoch": 2.8285946385052805,
+      "grad_norm": 0.17967191338539124,
+      "learning_rate": 9.915242869485309e-08,
+      "loss": 0.3575,
+      "step": 3482
+    },
+    {
+      "epoch": 2.8294069861900892,
+      "grad_norm": 0.18289880454540253,
+      "learning_rate": 9.821790418465815e-08,
+      "loss": 0.341,
+      "step": 3483
+    },
+    {
+      "epoch": 2.8302193338748984,
+      "grad_norm": 0.17416805028915405,
+      "learning_rate": 9.728776087398583e-08,
+      "loss": 0.3359,
+      "step": 3484
+    },
+    {
+      "epoch": 2.8310316815597076,
+      "grad_norm": 0.17138482630252838,
+      "learning_rate": 9.636199959419557e-08,
+      "loss": 0.3361,
+      "step": 3485
+    },
+    {
+      "epoch": 2.8318440292445164,
+      "grad_norm": 0.17991484701633453,
+      "learning_rate": 9.544062117273045e-08,
+      "loss": 0.3322,
+      "step": 3486
+    },
+    {
+      "epoch": 2.8326563769293256,
+      "grad_norm": 0.17254002392292023,
+      "learning_rate": 9.452362643311619e-08,
+      "loss": 0.341,
+      "step": 3487
+    },
+    {
+      "epoch": 2.833468724614135,
+      "grad_norm": 0.1662832796573639,
+      "learning_rate": 9.361101619495994e-08,
+      "loss": 0.3283,
+      "step": 3488
+    },
+    {
+      "epoch": 2.834281072298944,
+      "grad_norm": 0.1698969453573227,
+      "learning_rate": 9.270279127394977e-08,
+      "loss": 0.3288,
+      "step": 3489
+    },
+    {
+      "epoch": 2.835093419983753,
+      "grad_norm": 0.16670358180999756,
+      "learning_rate": 9.179895248185522e-08,
+      "loss": 0.3531,
+      "step": 3490
+    },
+    {
+      "epoch": 2.835905767668562,
+      "grad_norm": 0.16833537817001343,
+      "learning_rate": 9.089950062652508e-08,
+      "loss": 0.3496,
+      "step": 3491
+    },
+    {
+      "epoch": 2.836718115353371,
+      "grad_norm": 0.17753861844539642,
+      "learning_rate": 9.000443651188628e-08,
+      "loss": 0.3523,
+      "step": 3492
+    },
+    {
+      "epoch": 2.8375304630381804,
+      "grad_norm": 0.1767495721578598,
+      "learning_rate": 8.911376093794555e-08,
+      "loss": 0.341,
+      "step": 3493
+    },
+    {
+      "epoch": 2.8383428107229896,
+      "grad_norm": 0.17570336163043976,
+      "learning_rate": 8.822747470078552e-08,
+      "loss": 0.3126,
+      "step": 3494
+    },
+    {
+      "epoch": 2.8391551584077988,
+      "grad_norm": 0.18217268586158752,
+      "learning_rate": 8.734557859256698e-08,
+      "loss": 0.3522,
+      "step": 3495
+    },
+    {
+      "epoch": 2.8399675060926075,
+      "grad_norm": 0.17785193026065826,
+      "learning_rate": 8.646807340152607e-08,
+      "loss": 0.3345,
+      "step": 3496
+    },
+    {
+      "epoch": 2.8407798537774167,
+      "grad_norm": 0.1636761724948883,
+      "learning_rate": 8.55949599119743e-08,
+      "loss": 0.3256,
+      "step": 3497
+    },
+    {
+      "epoch": 2.841592201462226,
+      "grad_norm": 0.17912082374095917,
+      "learning_rate": 8.47262389042991e-08,
+      "loss": 0.3243,
+      "step": 3498
+    },
+    {
+      "epoch": 2.8424045491470347,
+      "grad_norm": 0.17411912977695465,
+      "learning_rate": 8.386191115495934e-08,
+      "loss": 0.3452,
+      "step": 3499
+    },
+    {
+      "epoch": 2.843216896831844,
+      "grad_norm": 0.16836050152778625,
+      "learning_rate": 8.300197743649041e-08,
+      "loss": 0.3412,
+      "step": 3500
+    },
+    {
+      "epoch": 2.844029244516653,
+      "grad_norm": 0.16881278157234192,
+      "learning_rate": 8.214643851749748e-08,
+      "loss": 0.345,
+      "step": 3501
+    },
+    {
+      "epoch": 2.8448415922014623,
+      "grad_norm": 0.17016053199768066,
+      "learning_rate": 8.129529516266e-08,
+      "loss": 0.3405,
+      "step": 3502
+    },
+    {
+      "epoch": 2.8456539398862715,
+      "grad_norm": 0.18227164447307587,
+      "learning_rate": 8.044854813272662e-08,
+      "loss": 0.3287,
+      "step": 3503
+    },
+    {
+      "epoch": 2.8464662875710802,
+      "grad_norm": 0.1802210807800293,
+      "learning_rate": 7.960619818451865e-08,
+      "loss": 0.3405,
+      "step": 3504
+    },
+    {
+      "epoch": 2.8472786352558894,
+      "grad_norm": 0.1780749261379242,
+      "learning_rate": 7.876824607092437e-08,
+      "loss": 0.3463,
+      "step": 3505
+    },
+    {
+      "epoch": 2.8480909829406986,
+      "grad_norm": 0.1702185571193695,
+      "learning_rate": 7.793469254090524e-08,
+      "loss": 0.3318,
+      "step": 3506
+    },
+    {
+      "epoch": 2.848903330625508,
+      "grad_norm": 0.1623818576335907,
+      "learning_rate": 7.710553833948753e-08,
+      "loss": 0.3393,
+      "step": 3507
+    },
+    {
+      "epoch": 2.849715678310317,
+      "grad_norm": 0.1746123880147934,
+      "learning_rate": 7.628078420776786e-08,
+      "loss": 0.3502,
+      "step": 3508
+    },
+    {
+      "epoch": 2.850528025995126,
+      "grad_norm": 0.17239214479923248,
+      "learning_rate": 7.546043088290878e-08,
+      "loss": 0.3121,
+      "step": 3509
+    },
+    {
+      "epoch": 2.851340373679935,
+      "grad_norm": 0.18538153171539307,
+      "learning_rate": 7.464447909814043e-08,
+      "loss": 0.3205,
+      "step": 3510
+    },
+    {
+      "epoch": 2.852152721364744,
+      "grad_norm": 0.1748277246952057,
+      "learning_rate": 7.383292958275778e-08,
+      "loss": 0.3372,
+      "step": 3511
+    },
+    {
+      "epoch": 2.852965069049553,
+      "grad_norm": 0.16428975760936737,
+      "learning_rate": 7.302578306212116e-08,
+      "loss": 0.3293,
+      "step": 3512
+    },
+    {
+      "epoch": 2.8537774167343626,
+      "grad_norm": 0.17461180686950684,
+      "learning_rate": 7.222304025765736e-08,
+      "loss": 0.3306,
+      "step": 3513
+    },
+    {
+      "epoch": 2.8545897644191713,
+      "grad_norm": 0.17125719785690308,
+      "learning_rate": 7.142470188685413e-08,
+      "loss": 0.3268,
+      "step": 3514
+    },
+    {
+      "epoch": 2.8554021121039805,
+      "grad_norm": 0.160209521651268,
+      "learning_rate": 7.063076866326568e-08,
+      "loss": 0.3262,
+      "step": 3515
+    },
+    {
+      "epoch": 2.8562144597887897,
+      "grad_norm": 0.16147373616695404,
+      "learning_rate": 6.984124129650605e-08,
+      "loss": 0.3548,
+      "step": 3516
+    },
+    {
+      "epoch": 2.8570268074735985,
+      "grad_norm": 0.16544464230537415,
+      "learning_rate": 6.905612049225352e-08,
+      "loss": 0.3495,
+      "step": 3517
+    },
+    {
+      "epoch": 2.8578391551584077,
+      "grad_norm": 0.1813347041606903,
+      "learning_rate": 6.827540695224622e-08,
+      "loss": 0.3306,
+      "step": 3518
+    },
+    {
+      "epoch": 2.858651502843217,
+      "grad_norm": 0.18975712358951569,
+      "learning_rate": 6.749910137428484e-08,
+      "loss": 0.3525,
+      "step": 3519
+    },
+    {
+      "epoch": 2.859463850528026,
+      "grad_norm": 0.15720954537391663,
+      "learning_rate": 6.672720445222769e-08,
+      "loss": 0.3232,
+      "step": 3520
+    },
+    {
+      "epoch": 2.8602761982128353,
+      "grad_norm": 0.1732877790927887,
+      "learning_rate": 6.595971687599567e-08,
+      "loss": 0.3696,
+      "step": 3521
+    },
+    {
+      "epoch": 2.861088545897644,
+      "grad_norm": 0.17921675741672516,
+      "learning_rate": 6.519663933156562e-08,
+      "loss": 0.3339,
+      "step": 3522
+    },
+    {
+      "epoch": 2.8619008935824533,
+      "grad_norm": 0.16972674429416656,
+      "learning_rate": 6.443797250097528e-08,
+      "loss": 0.3409,
+      "step": 3523
+    },
+    {
+      "epoch": 2.8627132412672625,
+      "grad_norm": 0.1683843582868576,
+      "learning_rate": 6.368371706231891e-08,
+      "loss": 0.3533,
+      "step": 3524
+    },
+    {
+      "epoch": 2.863525588952071,
+      "grad_norm": 0.1704019159078598,
+      "learning_rate": 6.29338736897478e-08,
+      "loss": 0.3295,
+      "step": 3525
+    },
+    {
+      "epoch": 2.864337936636881,
+      "grad_norm": 0.17854823172092438,
+      "learning_rate": 6.218844305346916e-08,
+      "loss": 0.347,
+      "step": 3526
+    },
+    {
+      "epoch": 2.8651502843216896,
+      "grad_norm": 0.1683601588010788,
+      "learning_rate": 6.144742581974839e-08,
+      "loss": 0.325,
+      "step": 3527
+    },
+    {
+      "epoch": 2.865962632006499,
+      "grad_norm": 0.1620054990053177,
+      "learning_rate": 6.071082265090344e-08,
+      "loss": 0.3514,
+      "step": 3528
+    },
+    {
+      "epoch": 2.866774979691308,
+      "grad_norm": 0.17255538702011108,
+      "learning_rate": 5.99786342053088e-08,
+      "loss": 0.3649,
+      "step": 3529
+    },
+    {
+      "epoch": 2.8675873273761168,
+      "grad_norm": 0.17407040297985077,
+      "learning_rate": 5.925086113739209e-08,
+      "loss": 0.3382,
+      "step": 3530
+    },
+    {
+      "epoch": 2.868399675060926,
+      "grad_norm": 0.15634657442569733,
+      "learning_rate": 5.852750409763575e-08,
+      "loss": 0.3303,
+      "step": 3531
+    },
+    {
+      "epoch": 2.869212022745735,
+      "grad_norm": 0.1811855137348175,
+      "learning_rate": 5.7808563732573155e-08,
+      "loss": 0.3567,
+      "step": 3532
+    },
+    {
+      "epoch": 2.8700243704305444,
+      "grad_norm": 0.16500124335289001,
+      "learning_rate": 5.709404068479196e-08,
+      "loss": 0.3389,
+      "step": 3533
+    },
+    {
+      "epoch": 2.8708367181153536,
+      "grad_norm": 0.16591687500476837,
+      "learning_rate": 5.638393559293076e-08,
+      "loss": 0.3291,
+      "step": 3534
+    },
+    {
+      "epoch": 2.8716490658001623,
+      "grad_norm": 0.1704830676317215,
+      "learning_rate": 5.5678249091679624e-08,
+      "loss": 0.3566,
+      "step": 3535
+    },
+    {
+      "epoch": 2.8724614134849715,
+      "grad_norm": 0.16660542786121368,
+      "learning_rate": 5.4976981811779575e-08,
+      "loss": 0.3373,
+      "step": 3536
+    },
+    {
+      "epoch": 2.8732737611697807,
+      "grad_norm": 0.16853873431682587,
+      "learning_rate": 5.4280134380020336e-08,
+      "loss": 0.3495,
+      "step": 3537
+    },
+    {
+      "epoch": 2.87408610885459,
+      "grad_norm": 0.16803507506847382,
+      "learning_rate": 5.358770741924368e-08,
+      "loss": 0.3329,
+      "step": 3538
+    },
+    {
+      "epoch": 2.874898456539399,
+      "grad_norm": 0.16132017970085144,
+      "learning_rate": 5.28997015483379e-08,
+      "loss": 0.3309,
+      "step": 3539
+    },
+    {
+      "epoch": 2.875710804224208,
+      "grad_norm": 0.17590539157390594,
+      "learning_rate": 5.2216117382241084e-08,
+      "loss": 0.3542,
+      "step": 3540
+    },
+    {
+      "epoch": 2.876523151909017,
+      "grad_norm": 0.17157350480556488,
+      "learning_rate": 5.1536955531938937e-08,
+      "loss": 0.3496,
+      "step": 3541
+    },
+    {
+      "epoch": 2.8773354995938263,
+      "grad_norm": 0.1736210137605667,
+      "learning_rate": 5.086221660446422e-08,
+      "loss": 0.3327,
+      "step": 3542
+    },
+    {
+      "epoch": 2.878147847278635,
+      "grad_norm": 0.18695068359375,
+      "learning_rate": 5.0191901202897295e-08,
+      "loss": 0.3381,
+      "step": 3543
+    },
+    {
+      "epoch": 2.8789601949634442,
+      "grad_norm": 0.17251946032047272,
+      "learning_rate": 4.952600992636392e-08,
+      "loss": 0.354,
+      "step": 3544
+    },
+    {
+      "epoch": 2.8797725426482534,
+      "grad_norm": 0.1889452487230301,
+      "learning_rate": 4.8864543370035764e-08,
+      "loss": 0.3403,
+      "step": 3545
+    },
+    {
+      "epoch": 2.8805848903330626,
+      "grad_norm": 0.15615011751651764,
+      "learning_rate": 4.820750212513048e-08,
+      "loss": 0.3508,
+      "step": 3546
+    },
+    {
+      "epoch": 2.881397238017872,
+      "grad_norm": 0.1875838041305542,
+      "learning_rate": 4.755488677890829e-08,
+      "loss": 0.3421,
+      "step": 3547
+    },
+    {
+      "epoch": 2.8822095857026806,
+      "grad_norm": 0.18914175033569336,
+      "learning_rate": 4.6906697914677055e-08,
+      "loss": 0.331,
+      "step": 3548
+    },
+    {
+      "epoch": 2.88302193338749,
+      "grad_norm": 0.15523317456245422,
+      "learning_rate": 4.626293611178445e-08,
+      "loss": 0.3364,
+      "step": 3549
+    },
+    {
+      "epoch": 2.883834281072299,
+      "grad_norm": 0.18995723128318787,
+      "learning_rate": 4.5623601945624096e-08,
+      "loss": 0.3407,
+      "step": 3550
+    },
+    {
+      "epoch": 2.884646628757108,
+      "grad_norm": 0.1719774305820465,
+      "learning_rate": 4.498869598763056e-08,
+      "loss": 0.3665,
+      "step": 3551
+    },
+    {
+      "epoch": 2.8854589764419174,
+      "grad_norm": 0.18686708807945251,
+      "learning_rate": 4.435821880528157e-08,
+      "loss": 0.3154,
+      "step": 3552
+    },
+    {
+      "epoch": 2.886271324126726,
+      "grad_norm": 0.1659349799156189,
+      "learning_rate": 4.373217096209526e-08,
+      "loss": 0.3405,
+      "step": 3553
+    },
+    {
+      "epoch": 2.8870836718115354,
+      "grad_norm": 0.1958494335412979,
+      "learning_rate": 4.3110553017631786e-08,
+      "loss": 0.3324,
+      "step": 3554
+    },
+    {
+      "epoch": 2.8878960194963446,
+      "grad_norm": 0.1634817272424698,
+      "learning_rate": 4.249336552749172e-08,
+      "loss": 0.323,
+      "step": 3555
+    },
+    {
+      "epoch": 2.8887083671811533,
+      "grad_norm": 0.17413552105426788,
+      "learning_rate": 4.188060904331548e-08,
+      "loss": 0.3255,
+      "step": 3556
+    },
+    {
+      "epoch": 2.8895207148659625,
+      "grad_norm": 0.16219204664230347,
+      "learning_rate": 4.127228411278383e-08,
+      "loss": 0.3453,
+      "step": 3557
+    },
+    {
+      "epoch": 2.8903330625507717,
+      "grad_norm": 0.17060883343219757,
+      "learning_rate": 4.0668391279614085e-08,
+      "loss": 0.3416,
+      "step": 3558
+    },
+    {
+      "epoch": 2.891145410235581,
+      "grad_norm": 0.16695643961429596,
+      "learning_rate": 4.0068931083566134e-08,
+      "loss": 0.3281,
+      "step": 3559
+    },
+    {
+      "epoch": 2.89195775792039,
+      "grad_norm": 0.17889586091041565,
+      "learning_rate": 3.947390406043472e-08,
+      "loss": 0.3328,
+      "step": 3560
+    },
+    {
+      "epoch": 2.892770105605199,
+      "grad_norm": 0.16611623764038086,
+      "learning_rate": 3.8883310742053316e-08,
+      "loss": 0.3231,
+      "step": 3561
+    },
+    {
+      "epoch": 2.893582453290008,
+      "grad_norm": 0.17451776564121246,
+      "learning_rate": 3.8297151656293e-08,
+      "loss": 0.3379,
+      "step": 3562
+    },
+    {
+      "epoch": 2.8943948009748173,
+      "grad_norm": 0.16730843484401703,
+      "learning_rate": 3.771542732706135e-08,
+      "loss": 0.3222,
+      "step": 3563
+    },
+    {
+      "epoch": 2.8952071486596265,
+      "grad_norm": 0.17202986776828766,
+      "learning_rate": 3.713813827430135e-08,
+      "loss": 0.3397,
+      "step": 3564
+    },
+    {
+      "epoch": 2.8960194963444357,
+      "grad_norm": 0.1633976697921753,
+      "learning_rate": 3.656528501399359e-08,
+      "loss": 0.3621,
+      "step": 3565
+    },
+    {
+      "epoch": 2.8968318440292444,
+      "grad_norm": 0.1854783147573471,
+      "learning_rate": 3.599686805815128e-08,
+      "loss": 0.3342,
+      "step": 3566
+    },
+    {
+      "epoch": 2.8976441917140536,
+      "grad_norm": 0.17665822803974152,
+      "learning_rate": 3.543288791482469e-08,
+      "loss": 0.3158,
+      "step": 3567
+    },
+    {
+      "epoch": 2.898456539398863,
+      "grad_norm": 0.17450326681137085,
+      "learning_rate": 3.4873345088097255e-08,
+      "loss": 0.332,
+      "step": 3568
+    },
+    {
+      "epoch": 2.8992688870836716,
+      "grad_norm": 0.17787764966487885,
+      "learning_rate": 3.4318240078087816e-08,
+      "loss": 0.3136,
+      "step": 3569
+    },
+    {
+      "epoch": 2.900081234768481,
+      "grad_norm": 0.15900519490242004,
+      "learning_rate": 3.3767573380946717e-08,
+      "loss": 0.3258,
+      "step": 3570
+    },
+    {
+      "epoch": 2.90089358245329,
+      "grad_norm": 0.1793006956577301,
+      "learning_rate": 3.3221345488859134e-08,
+      "loss": 0.3191,
+      "step": 3571
+    },
+    {
+      "epoch": 2.901705930138099,
+      "grad_norm": 0.17321613430976868,
+      "learning_rate": 3.267955689004176e-08,
+      "loss": 0.3269,
+      "step": 3572
+    },
+    {
+      "epoch": 2.9025182778229084,
+      "grad_norm": 0.1684640645980835,
+      "learning_rate": 3.214220806874335e-08,
+      "loss": 0.3231,
+      "step": 3573
+    },
+    {
+      "epoch": 2.903330625507717,
+      "grad_norm": 0.1858198195695877,
+      "learning_rate": 3.1609299505245275e-08,
+      "loss": 0.3309,
+      "step": 3574
+    },
+    {
+      "epoch": 2.9041429731925263,
+      "grad_norm": 0.16869331896305084,
+      "learning_rate": 3.1080831675859845e-08,
+      "loss": 0.3363,
+      "step": 3575
+    },
+    {
+      "epoch": 2.9049553208773355,
+      "grad_norm": 0.17246854305267334,
+      "learning_rate": 3.055680505292979e-08,
+      "loss": 0.3565,
+      "step": 3576
+    },
+    {
+      "epoch": 2.9057676685621447,
+      "grad_norm": 0.17026278376579285,
+      "learning_rate": 3.0037220104829344e-08,
+      "loss": 0.3454,
+      "step": 3577
+    },
+    {
+      "epoch": 2.906580016246954,
+      "grad_norm": 0.1871359646320343,
+      "learning_rate": 2.952207729596146e-08,
+      "loss": 0.3384,
+      "step": 3578
+    },
+    {
+      "epoch": 2.9073923639317627,
+      "grad_norm": 0.17827601730823517,
+      "learning_rate": 2.9011377086759496e-08,
+      "loss": 0.3321,
+      "step": 3579
+    },
+    {
+      "epoch": 2.908204711616572,
+      "grad_norm": 0.1715182214975357,
+      "learning_rate": 2.85051199336861e-08,
+      "loss": 0.3355,
+      "step": 3580
+    },
+    {
+      "epoch": 2.909017059301381,
+      "grad_norm": 0.17041844129562378,
+      "learning_rate": 2.800330628923209e-08,
+      "loss": 0.323,
+      "step": 3581
+    },
+    {
+      "epoch": 2.90982940698619,
+      "grad_norm": 0.17716741561889648,
+      "learning_rate": 2.7505936601918137e-08,
+      "loss": 0.341,
+      "step": 3582
+    },
+    {
+      "epoch": 2.910641754670999,
+      "grad_norm": 0.18312935531139374,
+      "learning_rate": 2.701301131629086e-08,
+      "loss": 0.3483,
+      "step": 3583
+    },
+    {
+      "epoch": 2.9114541023558083,
+      "grad_norm": 0.1649809330701828,
+      "learning_rate": 2.652453087292617e-08,
+      "loss": 0.3401,
+      "step": 3584
+    },
+    {
+      "epoch": 2.9122664500406175,
+      "grad_norm": 0.1680380403995514,
+      "learning_rate": 2.6040495708425928e-08,
+      "loss": 0.3278,
+      "step": 3585
+    },
+    {
+      "epoch": 2.9130787977254267,
+      "grad_norm": 0.17580777406692505,
+      "learning_rate": 2.5560906255420737e-08,
+      "loss": 0.3534,
+      "step": 3586
+    },
+    {
+      "epoch": 2.9138911454102354,
+      "grad_norm": 0.1739831566810608,
+      "learning_rate": 2.5085762942564927e-08,
+      "loss": 0.3264,
+      "step": 3587
+    },
+    {
+      "epoch": 2.9147034930950446,
+      "grad_norm": 0.2028699815273285,
+      "learning_rate": 2.4615066194541015e-08,
+      "loss": 0.3325,
+      "step": 3588
+    },
+    {
+      "epoch": 2.915515840779854,
+      "grad_norm": 0.17960543930530548,
+      "learning_rate": 2.41488164320558e-08,
+      "loss": 0.3359,
+      "step": 3589
+    },
+    {
+      "epoch": 2.916328188464663,
+      "grad_norm": 0.18402814865112305,
+      "learning_rate": 2.368701407184315e-08,
+      "loss": 0.3391,
+      "step": 3590
+    },
+    {
+      "epoch": 2.917140536149472,
+      "grad_norm": 0.1731589436531067,
+      "learning_rate": 2.322965952666012e-08,
+      "loss": 0.3615,
+      "step": 3591
+    },
+    {
+      "epoch": 2.917952883834281,
+      "grad_norm": 0.17914262413978577,
+      "learning_rate": 2.277675320528916e-08,
+      "loss": 0.3642,
+      "step": 3592
+    },
+    {
+      "epoch": 2.91876523151909,
+      "grad_norm": 0.15858109295368195,
+      "learning_rate": 2.2328295512535902e-08,
+      "loss": 0.3266,
+      "step": 3593
+    },
+    {
+      "epoch": 2.9195775792038994,
+      "grad_norm": 0.16807782649993896,
+      "learning_rate": 2.188428684923194e-08,
+      "loss": 0.3454,
+      "step": 3594
+    },
+    {
+      "epoch": 2.920389926888708,
+      "grad_norm": 0.17134982347488403,
+      "learning_rate": 2.144472761222982e-08,
+      "loss": 0.3401,
+      "step": 3595
+    },
+    {
+      "epoch": 2.9212022745735173,
+      "grad_norm": 0.16261695325374603,
+      "learning_rate": 2.10096181944075e-08,
+      "loss": 0.3652,
+      "step": 3596
+    },
+    {
+      "epoch": 2.9220146222583265,
+      "grad_norm": 0.18075568974018097,
+      "learning_rate": 2.0578958984663888e-08,
+      "loss": 0.3353,
+      "step": 3597
+    },
+    {
+      "epoch": 2.9228269699431357,
+      "grad_norm": 0.17909198999404907,
+      "learning_rate": 2.015275036792108e-08,
+      "loss": 0.3325,
+      "step": 3598
+    },
+    {
+      "epoch": 2.923639317627945,
+      "grad_norm": 0.1711125671863556,
+      "learning_rate": 1.973099272512269e-08,
+      "loss": 0.3434,
+      "step": 3599
+    },
+    {
+      "epoch": 2.9244516653127537,
+      "grad_norm": 0.1665475219488144,
+      "learning_rate": 1.9313686433236057e-08,
+      "loss": 0.3181,
+      "step": 3600
+    },
+    {
+      "epoch": 2.925264012997563,
+      "grad_norm": 0.15880082547664642,
+      "learning_rate": 1.8900831865246717e-08,
+      "loss": 0.3346,
+      "step": 3601
+    },
+    {
+      "epoch": 2.926076360682372,
+      "grad_norm": 0.17674149572849274,
+      "learning_rate": 1.849242939016449e-08,
+      "loss": 0.3447,
+      "step": 3602
+    },
+    {
+      "epoch": 2.9268887083671813,
+      "grad_norm": 0.1760532706975937,
+      "learning_rate": 1.8088479373016833e-08,
+      "loss": 0.3261,
+      "step": 3603
+    },
+    {
+      "epoch": 2.9277010560519905,
+      "grad_norm": 0.17778144776821136,
+      "learning_rate": 1.768898217485493e-08,
+      "loss": 0.3435,
+      "step": 3604
+    },
+    {
+      "epoch": 2.9285134037367992,
+      "grad_norm": 0.16828177869319916,
+      "learning_rate": 1.7293938152746493e-08,
+      "loss": 0.3385,
+      "step": 3605
+    },
+    {
+      "epoch": 2.9293257514216084,
+      "grad_norm": 0.16665509343147278,
+      "learning_rate": 1.6903347659781856e-08,
+      "loss": 0.3407,
+      "step": 3606
+    },
+    {
+      "epoch": 2.9301380991064176,
+      "grad_norm": 0.16861803829669952,
+      "learning_rate": 1.6517211045069537e-08,
+      "loss": 0.3233,
+      "step": 3607
+    },
+    {
+      "epoch": 2.9309504467912264,
+      "grad_norm": 0.17334215342998505,
+      "learning_rate": 1.6135528653737353e-08,
+      "loss": 0.3642,
+      "step": 3608
+    },
+    {
+      "epoch": 2.9317627944760356,
+      "grad_norm": 0.18384909629821777,
+      "learning_rate": 1.575830082693186e-08,
+      "loss": 0.3382,
+      "step": 3609
+    },
+    {
+      "epoch": 2.932575142160845,
+      "grad_norm": 0.18216487765312195,
+      "learning_rate": 1.5385527901818907e-08,
+      "loss": 0.3545,
+      "step": 3610
+    },
+    {
+      "epoch": 2.933387489845654,
+      "grad_norm": 0.1944405436515808,
+      "learning_rate": 1.5017210211580868e-08,
+      "loss": 0.3405,
+      "step": 3611
+    },
+    {
+      "epoch": 2.934199837530463,
+      "grad_norm": 0.17205077409744263,
+      "learning_rate": 1.4653348085419961e-08,
+      "loss": 0.3561,
+      "step": 3612
+    },
+    {
+      "epoch": 2.935012185215272,
+      "grad_norm": 0.16744673252105713,
+      "learning_rate": 1.4293941848554926e-08,
+      "loss": 0.3388,
+      "step": 3613
+    },
+    {
+      "epoch": 2.935824532900081,
+      "grad_norm": 0.171042338013649,
+      "learning_rate": 1.3938991822222137e-08,
+      "loss": 0.34,
+      "step": 3614
+    },
+    {
+      "epoch": 2.9366368805848904,
+      "grad_norm": 0.17146341502666473,
+      "learning_rate": 1.3588498323674482e-08,
+      "loss": 0.3491,
+      "step": 3615
+    },
+    {
+      "epoch": 2.9374492282696996,
+      "grad_norm": 0.16417255997657776,
+      "learning_rate": 1.3242461666182483e-08,
+      "loss": 0.3302,
+      "step": 3616
+    },
+    {
+      "epoch": 2.9382615759545088,
+      "grad_norm": 0.17271371185779572,
+      "learning_rate": 1.2900882159033178e-08,
+      "loss": 0.3351,
+      "step": 3617
+    },
+    {
+      "epoch": 2.9390739236393175,
+      "grad_norm": 0.17015333473682404,
+      "learning_rate": 1.2563760107528466e-08,
+      "loss": 0.3482,
+      "step": 3618
+    },
+    {
+      "epoch": 2.9398862713241267,
+      "grad_norm": 0.1642562597990036,
+      "learning_rate": 1.223109581298787e-08,
+      "loss": 0.3457,
+      "step": 3619
+    },
+    {
+      "epoch": 2.940698619008936,
+      "grad_norm": 0.17732499539852142,
+      "learning_rate": 1.1902889572745213e-08,
+      "loss": 0.3453,
+      "step": 3620
+    },
+    {
+      "epoch": 2.9415109666937447,
+      "grad_norm": 0.17006877064704895,
+      "learning_rate": 1.1579141680150841e-08,
+      "loss": 0.3321,
+      "step": 3621
+    },
+    {
+      "epoch": 2.942323314378554,
+      "grad_norm": 0.1719018518924713,
+      "learning_rate": 1.125985242456995e-08,
+      "loss": 0.3394,
+      "step": 3622
+    },
+    {
+      "epoch": 2.943135662063363,
+      "grad_norm": 0.17922943830490112,
+      "learning_rate": 1.094502209138204e-08,
+      "loss": 0.3414,
+      "step": 3623
+    },
+    {
+      "epoch": 2.9439480097481723,
+      "grad_norm": 0.1786583960056305,
+      "learning_rate": 1.0634650961981462e-08,
+      "loss": 0.3343,
+      "step": 3624
+    },
+    {
+      "epoch": 2.9447603574329815,
+      "grad_norm": 0.1797739565372467,
+      "learning_rate": 1.0328739313776869e-08,
+      "loss": 0.3378,
+      "step": 3625
+    },
+    {
+      "epoch": 2.9455727051177902,
+      "grad_norm": 0.16388891637325287,
+      "learning_rate": 1.0027287420192322e-08,
+      "loss": 0.3311,
+      "step": 3626
+    },
+    {
+      "epoch": 2.9463850528025994,
+      "grad_norm": 0.17406275868415833,
+      "learning_rate": 9.730295550663959e-09,
+      "loss": 0.3336,
+      "step": 3627
+    },
+    {
+      "epoch": 2.9471974004874086,
+      "grad_norm": 0.17846065759658813,
+      "learning_rate": 9.437763970642223e-09,
+      "loss": 0.3649,
+      "step": 3628
+    },
+    {
+      "epoch": 2.948009748172218,
+      "grad_norm": 0.17392204701900482,
+      "learning_rate": 9.149692941590738e-09,
+      "loss": 0.3329,
+      "step": 3629
+    },
+    {
+      "epoch": 2.948822095857027,
+      "grad_norm": 0.19088295102119446,
+      "learning_rate": 8.866082720987989e-09,
+      "loss": 0.3444,
+      "step": 3630
+    },
+    {
+      "epoch": 2.949634443541836,
+      "grad_norm": 0.16360828280448914,
+      "learning_rate": 8.58693356232232e-09,
+      "loss": 0.3352,
+      "step": 3631
+    },
+    {
+      "epoch": 2.950446791226645,
+      "grad_norm": 0.17218966782093048,
+      "learning_rate": 8.312245715097478e-09,
+      "loss": 0.3416,
+      "step": 3632
+    },
+    {
+      "epoch": 2.951259138911454,
+      "grad_norm": 0.1636769026517868,
+      "learning_rate": 8.042019424828739e-09,
+      "loss": 0.3357,
+      "step": 3633
+    },
+    {
+      "epoch": 2.952071486596263,
+      "grad_norm": 0.16861622035503387,
+      "learning_rate": 7.776254933043458e-09,
+      "loss": 0.3038,
+      "step": 3634
+    },
+    {
+      "epoch": 2.952883834281072,
+      "grad_norm": 0.1798638254404068,
+      "learning_rate": 7.514952477281067e-09,
+      "loss": 0.3327,
+      "step": 3635
+    },
+    {
+      "epoch": 2.9536961819658814,
+      "grad_norm": 0.18052923679351807,
+      "learning_rate": 7.258112291093078e-09,
+      "loss": 0.3602,
+      "step": 3636
+    },
+    {
+      "epoch": 2.9545085296506906,
+      "grad_norm": 0.1719256043434143,
+      "learning_rate": 7.005734604041969e-09,
+      "loss": 0.3516,
+      "step": 3637
+    },
+    {
+      "epoch": 2.9553208773354998,
+      "grad_norm": 0.1846006214618683,
+      "learning_rate": 6.757819641703966e-09,
+      "loss": 0.3758,
+      "step": 3638
+    },
+    {
+      "epoch": 2.9561332250203085,
+      "grad_norm": 0.18285855650901794,
+      "learning_rate": 6.514367625662932e-09,
+      "loss": 0.332,
+      "step": 3639
+    },
+    {
+      "epoch": 2.9569455727051177,
+      "grad_norm": 0.16396565735340118,
+      "learning_rate": 6.275378773516472e-09,
+      "loss": 0.3265,
+      "step": 3640
+    },
+    {
+      "epoch": 2.957757920389927,
+      "grad_norm": 0.1745087057352066,
+      "learning_rate": 6.040853298872051e-09,
+      "loss": 0.3513,
+      "step": 3641
+    },
+    {
+      "epoch": 2.958570268074736,
+      "grad_norm": 0.16682590544223785,
+      "learning_rate": 5.8107914113475494e-09,
+      "loss": 0.3379,
+      "step": 3642
+    },
+    {
+      "epoch": 2.9593826157595453,
+      "grad_norm": 0.15551526844501495,
+      "learning_rate": 5.585193316572368e-09,
+      "loss": 0.3015,
+      "step": 3643
+    },
+    {
+      "epoch": 2.960194963444354,
+      "grad_norm": 0.17867189645767212,
+      "learning_rate": 5.364059216185213e-09,
+      "loss": 0.3492,
+      "step": 3644
+    },
+    {
+      "epoch": 2.9610073111291633,
+      "grad_norm": 0.16583143174648285,
+      "learning_rate": 5.147389307834649e-09,
+      "loss": 0.3329,
+      "step": 3645
+    },
+    {
+      "epoch": 2.9618196588139725,
+      "grad_norm": 0.18984027206897736,
+      "learning_rate": 4.935183785180209e-09,
+      "loss": 0.3447,
+      "step": 3646
+    },
+    {
+      "epoch": 2.9626320064987812,
+      "grad_norm": 0.17806944251060486,
+      "learning_rate": 4.727442837890173e-09,
+      "loss": 0.3478,
+      "step": 3647
+    },
+    {
+      "epoch": 2.9634443541835904,
+      "grad_norm": 0.1923036128282547,
+      "learning_rate": 4.5241666516432355e-09,
+      "loss": 0.3176,
+      "step": 3648
+    },
+    {
+      "epoch": 2.9642567018683996,
+      "grad_norm": 0.17094267904758453,
+      "learning_rate": 4.325355408127396e-09,
+      "loss": 0.3376,
+      "step": 3649
+    },
+    {
+      "epoch": 2.965069049553209,
+      "grad_norm": 0.18617306649684906,
+      "learning_rate": 4.131009285038845e-09,
+      "loss": 0.3385,
+      "step": 3650
+    },
+    {
+      "epoch": 2.965881397238018,
+      "grad_norm": 0.20705363154411316,
+      "learning_rate": 3.941128456083631e-09,
+      "loss": 0.3422,
+      "step": 3651
+    },
+    {
+      "epoch": 2.966693744922827,
+      "grad_norm": 0.1735122799873352,
+      "learning_rate": 3.755713090977664e-09,
+      "loss": 0.3746,
+      "step": 3652
+    },
+    {
+      "epoch": 2.967506092607636,
+      "grad_norm": 0.16184407472610474,
+      "learning_rate": 3.5747633554433782e-09,
+      "loss": 0.3314,
+      "step": 3653
+    },
+    {
+      "epoch": 2.968318440292445,
+      "grad_norm": 0.16914601624011993,
+      "learning_rate": 3.3982794112141783e-09,
+      "loss": 0.3365,
+      "step": 3654
+    },
+    {
+      "epoch": 2.9691307879772544,
+      "grad_norm": 0.17131902277469635,
+      "learning_rate": 3.226261416030552e-09,
+      "loss": 0.3483,
+      "step": 3655
+    },
+    {
+      "epoch": 2.9699431356620636,
+      "grad_norm": 0.16713456809520721,
+      "learning_rate": 3.0587095236417342e-09,
+      "loss": 0.3328,
+      "step": 3656
+    },
+    {
+      "epoch": 2.9707554833468723,
+      "grad_norm": 0.16102290153503418,
+      "learning_rate": 2.8956238838051544e-09,
+      "loss": 0.3201,
+      "step": 3657
+    },
+    {
+      "epoch": 2.9715678310316815,
+      "grad_norm": 0.18010598421096802,
+      "learning_rate": 2.737004642286434e-09,
+      "loss": 0.3344,
+      "step": 3658
+    },
+    {
+      "epoch": 2.9723801787164907,
+      "grad_norm": 0.1792420893907547,
+      "learning_rate": 2.5828519408593877e-09,
+      "loss": 0.3321,
+      "step": 3659
+    },
+    {
+      "epoch": 2.9731925264012995,
+      "grad_norm": 0.17092587053775787,
+      "learning_rate": 2.4331659173038038e-09,
+      "loss": 0.3194,
+      "step": 3660
+    },
+    {
+      "epoch": 2.9740048740861087,
+      "grad_norm": 0.17841510474681854,
+      "learning_rate": 2.2879467054104378e-09,
+      "loss": 0.3256,
+      "step": 3661
+    },
+    {
+      "epoch": 2.974817221770918,
+      "grad_norm": 0.17252080142498016,
+      "learning_rate": 2.1471944349749086e-09,
+      "loss": 0.3365,
+      "step": 3662
+    },
+    {
+      "epoch": 2.975629569455727,
+      "grad_norm": 0.15820114314556122,
+      "learning_rate": 2.0109092318010283e-09,
+      "loss": 0.3378,
+      "step": 3663
+    },
+    {
+      "epoch": 2.9764419171405363,
+      "grad_norm": 0.1723087579011917,
+      "learning_rate": 1.8790912177008015e-09,
+      "loss": 0.3389,
+      "step": 3664
+    },
+    {
+      "epoch": 2.977254264825345,
+      "grad_norm": 0.1630985587835312,
+      "learning_rate": 1.7517405104922057e-09,
+      "loss": 0.3347,
+      "step": 3665
+    },
+    {
+      "epoch": 2.9780666125101543,
+      "grad_norm": 0.16742774844169617,
+      "learning_rate": 1.6288572240014123e-09,
+      "loss": 0.3299,
+      "step": 3666
+    },
+    {
+      "epoch": 2.9788789601949635,
+      "grad_norm": 0.18413236737251282,
+      "learning_rate": 1.510441468061119e-09,
+      "loss": 0.3431,
+      "step": 3667
+    },
+    {
+      "epoch": 2.9796913078797727,
+      "grad_norm": 0.16789467632770538,
+      "learning_rate": 1.3964933485105526e-09,
+      "loss": 0.3421,
+      "step": 3668
+    },
+    {
+      "epoch": 2.980503655564582,
+      "grad_norm": 0.1612119823694229,
+      "learning_rate": 1.287012967196577e-09,
+      "loss": 0.3225,
+      "step": 3669
+    },
+    {
+      "epoch": 2.9813160032493906,
+      "grad_norm": 0.16701161861419678,
+      "learning_rate": 1.1820004219725845e-09,
+      "loss": 0.3317,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9821283509342,
+      "grad_norm": 0.16738298535346985,
+      "learning_rate": 1.0814558066979397e-09,
+      "loss": 0.3452,
+      "step": 3671
+    },
+    {
+      "epoch": 2.982940698619009,
+      "grad_norm": 0.17423883080482483,
+      "learning_rate": 9.853792112402005e-10,
+      "loss": 0.3433,
+      "step": 3672
+    },
+    {
+      "epoch": 2.9837530463038178,
+      "grad_norm": 0.16599829494953156,
+      "learning_rate": 8.937707214712321e-10,
+      "loss": 0.3259,
+      "step": 3673
+    },
+    {
+      "epoch": 2.984565393988627,
+      "grad_norm": 0.1561869978904724,
+      "learning_rate": 8.066304192710927e-10,
+      "loss": 0.3358,
+      "step": 3674
+    },
+    {
+      "epoch": 2.985377741673436,
+      "grad_norm": 0.17110835015773773,
+      "learning_rate": 7.239583825252583e-10,
+      "loss": 0.3353,
+      "step": 3675
+    },
+    {
+      "epoch": 2.9861900893582454,
+      "grad_norm": 0.16126969456672668,
+      "learning_rate": 6.457546851262875e-10,
+      "loss": 0.3313,
+      "step": 3676
+    },
+    {
+      "epoch": 2.9870024370430546,
+      "grad_norm": 0.17735671997070312,
+      "learning_rate": 5.720193969716015e-10,
+      "loss": 0.3264,
+      "step": 3677
+    },
+    {
+      "epoch": 2.9878147847278633,
+      "grad_norm": 0.1813308745622635,
+      "learning_rate": 5.027525839662594e-10,
+      "loss": 0.3519,
+      "step": 3678
+    },
+    {
+      "epoch": 2.9886271324126725,
+      "grad_norm": 0.17318227887153625,
+      "learning_rate": 4.3795430802073825e-10,
+      "loss": 0.3543,
+      "step": 3679
+    },
+    {
+      "epoch": 2.9894394800974817,
+      "grad_norm": 0.1736849993467331,
+      "learning_rate": 3.7762462705093204e-10,
+      "loss": 0.3513,
+      "step": 3680
+    },
+    {
+      "epoch": 2.990251827782291,
+      "grad_norm": 0.16233214735984802,
+      "learning_rate": 3.2176359498037325e-10,
+      "loss": 0.3183,
+      "step": 3681
+    },
+    {
+      "epoch": 2.9910641754671,
+      "grad_norm": 0.17749440670013428,
+      "learning_rate": 2.703712617363463e-10,
+      "loss": 0.3357,
+      "step": 3682
+    },
+    {
+      "epoch": 2.991876523151909,
+      "grad_norm": 0.16298575699329376,
+      "learning_rate": 2.234476732537738e-10,
+      "loss": 0.3393,
+      "step": 3683
+    },
+    {
+      "epoch": 2.992688870836718,
+      "grad_norm": 0.18048207461833954,
+      "learning_rate": 1.8099287147355095e-10,
+      "loss": 0.3179,
+      "step": 3684
+    },
+    {
+      "epoch": 2.9935012185215273,
+      "grad_norm": 0.17113125324249268,
+      "learning_rate": 1.4300689434032512e-10,
+      "loss": 0.3674,
+      "step": 3685
+    },
+    {
+      "epoch": 2.994313566206336,
+      "grad_norm": 0.17127206921577454,
+      "learning_rate": 1.0948977580638176e-10,
+      "loss": 0.3391,
+      "step": 3686
+    },
+    {
+      "epoch": 2.9951259138911457,
+      "grad_norm": 0.1717831790447235,
+      "learning_rate": 8.044154582942388e-11,
+      "loss": 0.328,
+      "step": 3687
+    },
+    {
+      "epoch": 2.9959382615759544,
+      "grad_norm": 0.18055877089500427,
+      "learning_rate": 5.586223037257199e-11,
+      "loss": 0.3256,
+      "step": 3688
+    },
+    {
+      "epoch": 2.9967506092607636,
+      "grad_norm": 0.18643145263195038,
+      "learning_rate": 3.575185140436421e-11,
+      "loss": 0.3248,
+      "step": 3689
+    },
+    {
+      "epoch": 2.997562956945573,
+      "grad_norm": 0.18154817819595337,
+      "learning_rate": 2.0110426900421532e-11,
+      "loss": 0.3523,
+      "step": 3690
+    },
+    {
+      "epoch": 2.9983753046303816,
+      "grad_norm": 0.17546379566192627,
+      "learning_rate": 8.937970840117161e-12,
+      "loss": 0.3477,
+      "step": 3691
+    },
+    {
+      "epoch": 2.999187652315191,
+      "grad_norm": 0.16674262285232544,
+      "learning_rate": 2.2344932093520955e-12,
+      "loss": 0.3494,
+      "step": 3692
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.16627278923988342,
+      "learning_rate": 0.0,
+      "loss": 0.3643,
+      "step": 3693
+    },
+    {
+      "epoch": 3.0,
+      "step": 3693,
+      "total_flos": 1.6530581665087488e+16,
+      "train_loss": 0.3794273254576331,
+      "train_runtime": 265798.511,
+      "train_samples_per_second": 1.334,
+      "train_steps_per_second": 0.014
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3693,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6530581665087488e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}