diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7281 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997582789460963,
+  "eval_steps": 500,
+  "global_step": 1034,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009668842156151801,
+      "grad_norm": 19.466110229492188,
+      "learning_rate": 0.0,
+      "loss": 1.0726,
+      "step": 1
+    },
+    {
+      "epoch": 0.0019337684312303602,
+      "grad_norm": 21.088153839111328,
+      "learning_rate": 1.5625e-07,
+      "loss": 1.0254,
+      "step": 2
+    },
+    {
+      "epoch": 0.0029006526468455403,
+      "grad_norm": 19.635608673095703,
+      "learning_rate": 3.125e-07,
+      "loss": 1.043,
+      "step": 3
+    },
+    {
+      "epoch": 0.0038675368624607204,
+      "grad_norm": 19.43179702758789,
+      "learning_rate": 4.6875000000000006e-07,
+      "loss": 1.1058,
+      "step": 4
+    },
+    {
+      "epoch": 0.0048344210780759,
+      "grad_norm": 20.224998474121094,
+      "learning_rate": 6.25e-07,
+      "loss": 1.1457,
+      "step": 5
+    },
+    {
+      "epoch": 0.005801305293691081,
+      "grad_norm": 16.54555320739746,
+      "learning_rate": 7.8125e-07,
+      "loss": 0.9547,
+      "step": 6
+    },
+    {
+      "epoch": 0.00676818950930626,
+      "grad_norm": 17.746612548828125,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 1.0488,
+      "step": 7
+    },
+    {
+      "epoch": 0.007735073724921441,
+      "grad_norm": 18.536388397216797,
+      "learning_rate": 1.0937500000000001e-06,
+      "loss": 1.0446,
+      "step": 8
+    },
+    {
+      "epoch": 0.00870195794053662,
+      "grad_norm": 15.782112121582031,
+      "learning_rate": 1.25e-06,
+      "loss": 0.9769,
+      "step": 9
+    },
+    {
+      "epoch": 0.0096688421561518,
+      "grad_norm": 15.247085571289062,
+      "learning_rate": 1.40625e-06,
+      "loss": 0.976,
+      "step": 10
+    },
+    {
+      "epoch": 0.010635726371766982,
+      "grad_norm": 14.069637298583984,
+      "learning_rate": 1.5625e-06,
+      "loss": 1.037,
+      "step": 11
+    },
+    {
+      "epoch": 0.011602610587382161,
+      "grad_norm": 9.867317199707031,
+      "learning_rate": 1.71875e-06,
+      "loss": 0.9776,
+      "step": 12
+    },
+    {
+      "epoch": 0.012569494802997341,
+      "grad_norm": 9.262001991271973,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.9161,
+      "step": 13
+    },
+    {
+      "epoch": 0.01353637901861252,
+      "grad_norm": 9.465025901794434,
+      "learning_rate": 2.0312500000000002e-06,
+      "loss": 0.9926,
+      "step": 14
+    },
+    {
+      "epoch": 0.014503263234227702,
+      "grad_norm": 9.50775146484375,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 0.9811,
+      "step": 15
+    },
+    {
+      "epoch": 0.015470147449842882,
+      "grad_norm": 8.129717826843262,
+      "learning_rate": 2.3437500000000002e-06,
+      "loss": 0.9271,
+      "step": 16
+    },
+    {
+      "epoch": 0.01643703166545806,
+      "grad_norm": 9.526381492614746,
+      "learning_rate": 2.5e-06,
+      "loss": 0.896,
+      "step": 17
+    },
+    {
+      "epoch": 0.01740391588107324,
+      "grad_norm": 8.398605346679688,
+      "learning_rate": 2.65625e-06,
+      "loss": 0.6924,
+      "step": 18
+    },
+    {
+      "epoch": 0.01837080009668842,
+      "grad_norm": 9.33753776550293,
+      "learning_rate": 2.8125e-06,
+      "loss": 0.813,
+      "step": 19
+    },
+    {
+      "epoch": 0.0193376843123036,
+      "grad_norm": 9.168170928955078,
+      "learning_rate": 2.96875e-06,
+      "loss": 0.8602,
+      "step": 20
+    },
+    {
+      "epoch": 0.02030456852791878,
+      "grad_norm": 8.446990013122559,
+      "learning_rate": 3.125e-06,
+      "loss": 0.8427,
+      "step": 21
+    },
+    {
+      "epoch": 0.021271452743533963,
+      "grad_norm": 6.478088855743408,
+      "learning_rate": 3.28125e-06,
+      "loss": 0.6606,
+      "step": 22
+    },
+    {
+      "epoch": 0.022238336959149143,
+      "grad_norm": 6.63073205947876,
+      "learning_rate": 3.4375e-06,
+      "loss": 0.8054,
+      "step": 23
+    },
+    {
+      "epoch": 0.023205221174764323,
+      "grad_norm": 7.6455302238464355,
+      "learning_rate": 3.59375e-06,
+      "loss": 0.7505,
+      "step": 24
+    },
+    {
+      "epoch": 0.024172105390379502,
+      "grad_norm": 6.51140832901001,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6872,
+      "step": 25
+    },
+    {
+      "epoch": 0.025138989605994682,
+      "grad_norm": 5.949461460113525,
+      "learning_rate": 3.90625e-06,
+      "loss": 0.6455,
+      "step": 26
+    },
+    {
+      "epoch": 0.02610587382160986,
+      "grad_norm": 6.92095422744751,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 0.7218,
+      "step": 27
+    },
+    {
+      "epoch": 0.02707275803722504,
+      "grad_norm": 5.846500873565674,
+      "learning_rate": 4.21875e-06,
+      "loss": 0.7175,
+      "step": 28
+    },
+    {
+      "epoch": 0.02803964225284022,
+      "grad_norm": 5.6374430656433105,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.7276,
+      "step": 29
+    },
+    {
+      "epoch": 0.029006526468455404,
+      "grad_norm": 4.731724739074707,
+      "learning_rate": 4.53125e-06,
+      "loss": 0.6809,
+      "step": 30
+    },
+    {
+      "epoch": 0.029973410684070584,
+      "grad_norm": 5.237879753112793,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 0.702,
+      "step": 31
+    },
+    {
+      "epoch": 0.030940294899685764,
+      "grad_norm": 4.484801769256592,
+      "learning_rate": 4.84375e-06,
+      "loss": 0.6112,
+      "step": 32
+    },
+    {
+      "epoch": 0.03190717911530094,
+      "grad_norm": 4.658060073852539,
+      "learning_rate": 5e-06,
+      "loss": 0.5665,
+      "step": 33
+    },
+    {
+      "epoch": 0.03287406333091612,
+      "grad_norm": 4.812199592590332,
+      "learning_rate": 4.999987712204937e-06,
+      "loss": 0.6284,
+      "step": 34
+    },
+    {
+      "epoch": 0.0338409475465313,
+      "grad_norm": 4.881224155426025,
+      "learning_rate": 4.999950848940538e-06,
+      "loss": 0.6992,
+      "step": 35
+    },
+    {
+      "epoch": 0.03480783176214648,
+      "grad_norm": 4.776272773742676,
+      "learning_rate": 4.999889410569179e-06,
+      "loss": 0.6326,
+      "step": 36
+    },
+    {
+      "epoch": 0.03577471597776166,
+      "grad_norm": 4.670900344848633,
+      "learning_rate": 4.999803397694811e-06,
+      "loss": 0.6886,
+      "step": 37
+    },
+    {
+      "epoch": 0.03674160019337684,
+      "grad_norm": 4.662517070770264,
+      "learning_rate": 4.999692811162965e-06,
+      "loss": 0.5526,
+      "step": 38
+    },
+    {
+      "epoch": 0.03770848440899202,
+      "grad_norm": 4.519469738006592,
+      "learning_rate": 4.999557652060729e-06,
+      "loss": 0.6316,
+      "step": 39
+    },
+    {
+      "epoch": 0.0386753686246072,
+      "grad_norm": 4.677404403686523,
+      "learning_rate": 4.999397921716751e-06,
+      "loss": 0.662,
+      "step": 40
+    },
+    {
+      "epoch": 0.03964225284022238,
+      "grad_norm": 4.661558628082275,
+      "learning_rate": 4.9992136217012184e-06,
+      "loss": 0.6335,
+      "step": 41
+    },
+    {
+      "epoch": 0.04060913705583756,
+      "grad_norm": 4.582630634307861,
+      "learning_rate": 4.9990047538258416e-06,
+      "loss": 0.6392,
+      "step": 42
+    },
+    {
+      "epoch": 0.04157602127145275,
+      "grad_norm": 4.041533946990967,
+      "learning_rate": 4.998771320143843e-06,
+      "loss": 0.6174,
+      "step": 43
+    },
+    {
+      "epoch": 0.04254290548706793,
+      "grad_norm": 4.323627948760986,
+      "learning_rate": 4.99851332294993e-06,
+      "loss": 0.5823,
+      "step": 44
+    },
+    {
+      "epoch": 0.043509789702683106,
+      "grad_norm": 4.2771124839782715,
+      "learning_rate": 4.998230764780277e-06,
+      "loss": 0.6334,
+      "step": 45
+    },
+    {
+      "epoch": 0.044476673918298286,
+      "grad_norm": 4.1346001625061035,
+      "learning_rate": 4.997923648412495e-06,
+      "loss": 0.4832,
+      "step": 46
+    },
+    {
+      "epoch": 0.045443558133913466,
+      "grad_norm": 3.749467134475708,
+      "learning_rate": 4.9975919768656125e-06,
+      "loss": 0.6186,
+      "step": 47
+    },
+    {
+      "epoch": 0.046410442349528645,
+      "grad_norm": 4.082079887390137,
+      "learning_rate": 4.997235753400039e-06,
+      "loss": 0.5151,
+      "step": 48
+    },
+    {
+      "epoch": 0.047377326565143825,
+      "grad_norm": 4.149928569793701,
+      "learning_rate": 4.996854981517535e-06,
+      "loss": 0.5641,
+      "step": 49
+    },
+    {
+      "epoch": 0.048344210780759005,
+      "grad_norm": 4.360020637512207,
+      "learning_rate": 4.996449664961177e-06,
+      "loss": 0.6201,
+      "step": 50
+    },
+    {
+      "epoch": 0.049311094996374184,
+      "grad_norm": 3.918663263320923,
+      "learning_rate": 4.996019807715324e-06,
+      "loss": 0.6154,
+      "step": 51
+    },
+    {
+      "epoch": 0.050277979211989364,
+      "grad_norm": 4.024591445922852,
+      "learning_rate": 4.995565414005573e-06,
+      "loss": 0.6135,
+      "step": 52
+    },
+    {
+      "epoch": 0.051244863427604544,
+      "grad_norm": 4.135931491851807,
+      "learning_rate": 4.995086488298723e-06,
+      "loss": 0.6724,
+      "step": 53
+    },
+    {
+      "epoch": 0.05221174764321972,
+      "grad_norm": 3.9647462368011475,
+      "learning_rate": 4.994583035302724e-06,
+      "loss": 0.5477,
+      "step": 54
+    },
+    {
+      "epoch": 0.0531786318588349,
+      "grad_norm": 4.09563684463501,
+      "learning_rate": 4.994055059966641e-06,
+      "loss": 0.5471,
+      "step": 55
+    },
+    {
+      "epoch": 0.05414551607445008,
+      "grad_norm": 4.164375305175781,
+      "learning_rate": 4.993502567480592e-06,
+      "loss": 0.6487,
+      "step": 56
+    },
+    {
+      "epoch": 0.05511240029006526,
+      "grad_norm": 4.008912563323975,
+      "learning_rate": 4.992925563275714e-06,
+      "loss": 0.5567,
+      "step": 57
+    },
+    {
+      "epoch": 0.05607928450568044,
+      "grad_norm": 3.589381217956543,
+      "learning_rate": 4.99232405302409e-06,
+      "loss": 0.5749,
+      "step": 58
+    },
+    {
+      "epoch": 0.05704616872129562,
+      "grad_norm": 3.8242640495300293,
+      "learning_rate": 4.991698042638711e-06,
+      "loss": 0.6054,
+      "step": 59
+    },
+    {
+      "epoch": 0.05801305293691081,
+      "grad_norm": 4.131597995758057,
+      "learning_rate": 4.991047538273403e-06,
+      "loss": 0.5606,
+      "step": 60
+    },
+    {
+      "epoch": 0.05897993715252599,
+      "grad_norm": 3.8480031490325928,
+      "learning_rate": 4.990372546322782e-06,
+      "loss": 0.5264,
+      "step": 61
+    },
+    {
+      "epoch": 0.05994682136814117,
+      "grad_norm": 4.238224506378174,
+      "learning_rate": 4.9896730734221755e-06,
+      "loss": 0.6041,
+      "step": 62
+    },
+    {
+      "epoch": 0.06091370558375635,
+      "grad_norm": 3.8943910598754883,
+      "learning_rate": 4.988949126447567e-06,
+      "loss": 0.553,
+      "step": 63
+    },
+    {
+      "epoch": 0.06188058979937153,
+      "grad_norm": 3.774726629257202,
+      "learning_rate": 4.988200712515527e-06,
+      "loss": 0.505,
+      "step": 64
+    },
+    {
+      "epoch": 0.0628474740149867,
+      "grad_norm": 3.9384453296661377,
+      "learning_rate": 4.987427838983141e-06,
+      "loss": 0.6143,
+      "step": 65
+    },
+    {
+      "epoch": 0.06381435823060189,
+      "grad_norm": 3.9011735916137695,
+      "learning_rate": 4.986630513447939e-06,
+      "loss": 0.5535,
+      "step": 66
+    },
+    {
+      "epoch": 0.06478124244621707,
+      "grad_norm": 4.145521640777588,
+      "learning_rate": 4.985808743747817e-06,
+      "loss": 0.5506,
+      "step": 67
+    },
+    {
+      "epoch": 0.06574812666183225,
+      "grad_norm": 3.8869740962982178,
+      "learning_rate": 4.984962537960969e-06,
+      "loss": 0.593,
+      "step": 68
+    },
+    {
+      "epoch": 0.06671501087744743,
+      "grad_norm": 4.154409408569336,
+      "learning_rate": 4.984091904405793e-06,
+      "loss": 0.6245,
+      "step": 69
+    },
+    {
+      "epoch": 0.0676818950930626,
+      "grad_norm": 3.601078987121582,
+      "learning_rate": 4.983196851640825e-06,
+      "loss": 0.5135,
+      "step": 70
+    },
+    {
+      "epoch": 0.06864877930867778,
+      "grad_norm": 3.83754301071167,
+      "learning_rate": 4.9822773884646444e-06,
+      "loss": 0.5545,
+      "step": 71
+    },
+    {
+      "epoch": 0.06961566352429296,
+      "grad_norm": 3.9656333923339844,
+      "learning_rate": 4.981333523915792e-06,
+      "loss": 0.6262,
+      "step": 72
+    },
+    {
+      "epoch": 0.07058254773990814,
+      "grad_norm": 3.8404502868652344,
+      "learning_rate": 4.980365267272679e-06,
+      "loss": 0.5053,
+      "step": 73
+    },
+    {
+      "epoch": 0.07154943195552332,
+      "grad_norm": 4.082815170288086,
+      "learning_rate": 4.979372628053496e-06,
+      "loss": 0.6844,
+      "step": 74
+    },
+    {
+      "epoch": 0.0725163161711385,
+      "grad_norm": 3.7831435203552246,
+      "learning_rate": 4.97835561601612e-06,
+      "loss": 0.5896,
+      "step": 75
+    },
+    {
+      "epoch": 0.07348320038675368,
+      "grad_norm": 4.10606575012207,
+      "learning_rate": 4.977314241158021e-06,
+      "loss": 0.6269,
+      "step": 76
+    },
+    {
+      "epoch": 0.07445008460236886,
+      "grad_norm": 3.9456067085266113,
+      "learning_rate": 4.97624851371616e-06,
+      "loss": 0.5214,
+      "step": 77
+    },
+    {
+      "epoch": 0.07541696881798404,
+      "grad_norm": 3.6953811645507812,
+      "learning_rate": 4.975158444166887e-06,
+      "loss": 0.5133,
+      "step": 78
+    },
+    {
+      "epoch": 0.07638385303359922,
+      "grad_norm": 4.28406286239624,
+      "learning_rate": 4.974044043225846e-06,
+      "loss": 0.596,
+      "step": 79
+    },
+    {
+      "epoch": 0.0773507372492144,
+      "grad_norm": 4.1657819747924805,
+      "learning_rate": 4.972905321847859e-06,
+      "loss": 0.5556,
+      "step": 80
+    },
+    {
+      "epoch": 0.07831762146482958,
+      "grad_norm": 3.9782443046569824,
+      "learning_rate": 4.9717422912268265e-06,
+      "loss": 0.4779,
+      "step": 81
+    },
+    {
+      "epoch": 0.07928450568044476,
+      "grad_norm": 3.791666269302368,
+      "learning_rate": 4.970554962795615e-06,
+      "loss": 0.5185,
+      "step": 82
+    },
+    {
+      "epoch": 0.08025138989605994,
+      "grad_norm": 3.8677923679351807,
+      "learning_rate": 4.969343348225942e-06,
+      "loss": 0.6107,
+      "step": 83
+    },
+    {
+      "epoch": 0.08121827411167512,
+      "grad_norm": 4.23475980758667,
+      "learning_rate": 4.968107459428265e-06,
+      "loss": 0.6065,
+      "step": 84
+    },
+    {
+      "epoch": 0.0821851583272903,
+      "grad_norm": 3.8311767578125,
+      "learning_rate": 4.966847308551664e-06,
+      "loss": 0.5138,
+      "step": 85
+    },
+    {
+      "epoch": 0.0831520425429055,
+      "grad_norm": 4.179352283477783,
+      "learning_rate": 4.9655629079837176e-06,
+      "loss": 0.5707,
+      "step": 86
+    },
+    {
+      "epoch": 0.08411892675852067,
+      "grad_norm": 3.6656126976013184,
+      "learning_rate": 4.9642542703503874e-06,
+      "loss": 0.491,
+      "step": 87
+    },
+    {
+      "epoch": 0.08508581097413585,
+      "grad_norm": 4.082828521728516,
+      "learning_rate": 4.962921408515891e-06,
+      "loss": 0.5196,
+      "step": 88
+    },
+    {
+      "epoch": 0.08605269518975103,
+      "grad_norm": 4.420588493347168,
+      "learning_rate": 4.961564335582572e-06,
+      "loss": 0.5719,
+      "step": 89
+    },
+    {
+      "epoch": 0.08701957940536621,
+      "grad_norm": 4.108341217041016,
+      "learning_rate": 4.960183064890782e-06,
+      "loss": 0.577,
+      "step": 90
+    },
+    {
+      "epoch": 0.08798646362098139,
+      "grad_norm": 3.8085920810699463,
+      "learning_rate": 4.958777610018734e-06,
+      "loss": 0.5059,
+      "step": 91
+    },
+    {
+      "epoch": 0.08895334783659657,
+      "grad_norm": 3.7101457118988037,
+      "learning_rate": 4.957347984782384e-06,
+      "loss": 0.5393,
+      "step": 92
+    },
+    {
+      "epoch": 0.08992023205221175,
+      "grad_norm": 3.719266653060913,
+      "learning_rate": 4.955894203235285e-06,
+      "loss": 0.5582,
+      "step": 93
+    },
+    {
+      "epoch": 0.09088711626782693,
+      "grad_norm": 4.132759094238281,
+      "learning_rate": 4.954416279668451e-06,
+      "loss": 0.5649,
+      "step": 94
+    },
+    {
+      "epoch": 0.09185400048344211,
+      "grad_norm": 3.8168387413024902,
+      "learning_rate": 4.952914228610221e-06,
+      "loss": 0.5105,
+      "step": 95
+    },
+    {
+      "epoch": 0.09282088469905729,
+      "grad_norm": 4.188531875610352,
+      "learning_rate": 4.951388064826112e-06,
+      "loss": 0.61,
+      "step": 96
+    },
+    {
+      "epoch": 0.09378776891467247,
+      "grad_norm": 4.02006196975708,
+      "learning_rate": 4.949837803318672e-06,
+      "loss": 0.5134,
+      "step": 97
+    },
+    {
+      "epoch": 0.09475465313028765,
+      "grad_norm": 4.280331134796143,
+      "learning_rate": 4.9482634593273395e-06,
+      "loss": 0.5284,
+      "step": 98
+    },
+    {
+      "epoch": 0.09572153734590283,
+      "grad_norm": 3.8056821823120117,
+      "learning_rate": 4.946665048328288e-06,
+      "loss": 0.4675,
+      "step": 99
+    },
+    {
+      "epoch": 0.09668842156151801,
+      "grad_norm": 4.207078456878662,
+      "learning_rate": 4.945042586034272e-06,
+      "loss": 0.5671,
+      "step": 100
+    },
+    {
+      "epoch": 0.09765530577713319,
+      "grad_norm": 4.331418991088867,
+      "learning_rate": 4.943396088394482e-06,
+      "loss": 0.5816,
+      "step": 101
+    },
+    {
+      "epoch": 0.09862218999274837,
+      "grad_norm": 3.848921060562134,
+      "learning_rate": 4.941725571594377e-06,
+      "loss": 0.5999,
+      "step": 102
+    },
+    {
+      "epoch": 0.09958907420836355,
+      "grad_norm": 3.920262336730957,
+      "learning_rate": 4.940031052055532e-06,
+      "loss": 0.5007,
+      "step": 103
+    },
+    {
+      "epoch": 0.10055595842397873,
+      "grad_norm": 3.944211483001709,
+      "learning_rate": 4.938312546435474e-06,
+      "loss": 0.5734,
+      "step": 104
+    },
+    {
+      "epoch": 0.10152284263959391,
+      "grad_norm": 4.038331508636475,
+      "learning_rate": 4.936570071627517e-06,
+      "loss": 0.5748,
+      "step": 105
+    },
+    {
+      "epoch": 0.10248972685520909,
+      "grad_norm": 3.655327081680298,
+      "learning_rate": 4.934803644760604e-06,
+      "loss": 0.5011,
+      "step": 106
+    },
+    {
+      "epoch": 0.10345661107082427,
+      "grad_norm": 3.9986114501953125,
+      "learning_rate": 4.933013283199124e-06,
+      "loss": 0.5448,
+      "step": 107
+    },
+    {
+      "epoch": 0.10442349528643945,
+      "grad_norm": 3.777674674987793,
+      "learning_rate": 4.9311990045427555e-06,
+      "loss": 0.5184,
+      "step": 108
+    },
+    {
+      "epoch": 0.10539037950205463,
+      "grad_norm": 4.203485012054443,
+      "learning_rate": 4.929360826626286e-06,
+      "loss": 0.5776,
+      "step": 109
+    },
+    {
+      "epoch": 0.1063572637176698,
+      "grad_norm": 4.092432975769043,
+      "learning_rate": 4.927498767519437e-06,
+      "loss": 0.5099,
+      "step": 110
+    },
+    {
+      "epoch": 0.10732414793328499,
+      "grad_norm": 3.8467228412628174,
+      "learning_rate": 4.925612845526691e-06,
+      "loss": 0.5168,
+      "step": 111
+    },
+    {
+      "epoch": 0.10829103214890017,
+      "grad_norm": 3.9712438583374023,
+      "learning_rate": 4.923703079187104e-06,
+      "loss": 0.5442,
+      "step": 112
+    },
+    {
+      "epoch": 0.10925791636451535,
+      "grad_norm": 3.5273666381835938,
+      "learning_rate": 4.921769487274132e-06,
+      "loss": 0.5169,
+      "step": 113
+    },
+    {
+      "epoch": 0.11022480058013052,
+      "grad_norm": 3.956287384033203,
+      "learning_rate": 4.91981208879544e-06,
+      "loss": 0.5562,
+      "step": 114
+    },
+    {
+      "epoch": 0.1111916847957457,
+      "grad_norm": 3.581745147705078,
+      "learning_rate": 4.917830902992716e-06,
+      "loss": 0.5398,
+      "step": 115
+    },
+    {
+      "epoch": 0.11215856901136088,
+      "grad_norm": 3.948413133621216,
+      "learning_rate": 4.915825949341484e-06,
+      "loss": 0.5221,
+      "step": 116
+    },
+    {
+      "epoch": 0.11312545322697606,
+      "grad_norm": 3.879704713821411,
+      "learning_rate": 4.913797247550912e-06,
+      "loss": 0.4968,
+      "step": 117
+    },
+    {
+      "epoch": 0.11409233744259124,
+      "grad_norm": 3.907553195953369,
+      "learning_rate": 4.9117448175636184e-06,
+      "loss": 0.5767,
+      "step": 118
+    },
+    {
+      "epoch": 0.11505922165820642,
+      "grad_norm": 3.8580918312072754,
+      "learning_rate": 4.9096686795554725e-06,
+      "loss": 0.4927,
+      "step": 119
+    },
+    {
+      "epoch": 0.11602610587382162,
+      "grad_norm": 4.098156452178955,
+      "learning_rate": 4.907568853935403e-06,
+      "loss": 0.5588,
+      "step": 120
+    },
+    {
+      "epoch": 0.1169929900894368,
+      "grad_norm": 3.590182304382324,
+      "learning_rate": 4.90544536134519e-06,
+      "loss": 0.5009,
+      "step": 121
+    },
+    {
+      "epoch": 0.11795987430505198,
+      "grad_norm": 3.887967348098755,
+      "learning_rate": 4.9032982226592684e-06,
+      "loss": 0.5394,
+      "step": 122
+    },
+    {
+      "epoch": 0.11892675852066716,
+      "grad_norm": 3.6042134761810303,
+      "learning_rate": 4.901127458984516e-06,
+      "loss": 0.4072,
+      "step": 123
+    },
+    {
+      "epoch": 0.11989364273628234,
+      "grad_norm": 3.842017412185669,
+      "learning_rate": 4.898933091660054e-06,
+      "loss": 0.5181,
+      "step": 124
+    },
+    {
+      "epoch": 0.12086052695189752,
+      "grad_norm": 4.496882438659668,
+      "learning_rate": 4.8967151422570314e-06,
+      "loss": 0.5224,
+      "step": 125
+    },
+    {
+      "epoch": 0.1218274111675127,
+      "grad_norm": 3.38244366645813,
+      "learning_rate": 4.894473632578414e-06,
+      "loss": 0.5029,
+      "step": 126
+    },
+    {
+      "epoch": 0.12279429538312787,
+      "grad_norm": 3.902873992919922,
+      "learning_rate": 4.89220858465877e-06,
+      "loss": 0.5346,
+      "step": 127
+    },
+    {
+      "epoch": 0.12376117959874305,
+      "grad_norm": 4.05102014541626,
+      "learning_rate": 4.889920020764055e-06,
+      "loss": 0.5789,
+      "step": 128
+    },
+    {
+      "epoch": 0.12472806381435823,
+      "grad_norm": 3.9317867755889893,
+      "learning_rate": 4.887607963391394e-06,
+      "loss": 0.5399,
+      "step": 129
+    },
+    {
+      "epoch": 0.1256949480299734,
+      "grad_norm": 3.631277561187744,
+      "learning_rate": 4.8852724352688556e-06,
+      "loss": 0.4701,
+      "step": 130
+    },
+    {
+      "epoch": 0.12666183224558858,
+      "grad_norm": 4.077336311340332,
+      "learning_rate": 4.882913459355233e-06,
+      "loss": 0.622,
+      "step": 131
+    },
+    {
+      "epoch": 0.12762871646120377,
+      "grad_norm": 4.419619083404541,
+      "learning_rate": 4.880531058839816e-06,
+      "loss": 0.5563,
+      "step": 132
+    },
+    {
+      "epoch": 0.12859560067681894,
+      "grad_norm": 3.4148476123809814,
+      "learning_rate": 4.878125257142165e-06,
+      "loss": 0.4959,
+      "step": 133
+    },
+    {
+      "epoch": 0.12956248489243413,
+      "grad_norm": 4.037806510925293,
+      "learning_rate": 4.8756960779118765e-06,
+      "loss": 0.5066,
+      "step": 134
+    },
+    {
+      "epoch": 0.1305293691080493,
+      "grad_norm": 3.803792715072632,
+      "learning_rate": 4.873243545028356e-06,
+      "loss": 0.5543,
+      "step": 135
+    },
+    {
+      "epoch": 0.1314962533236645,
+      "grad_norm": 3.535799503326416,
+      "learning_rate": 4.870767682600584e-06,
+      "loss": 0.465,
+      "step": 136
+    },
+    {
+      "epoch": 0.13246313753927969,
+      "grad_norm": 3.79543137550354,
+      "learning_rate": 4.868268514966869e-06,
+      "loss": 0.5385,
+      "step": 137
+    },
+    {
+      "epoch": 0.13343002175489485,
+      "grad_norm": 3.995861768722534,
+      "learning_rate": 4.86574606669462e-06,
+      "loss": 0.5965,
+      "step": 138
+    },
+    {
+      "epoch": 0.13439690597051004,
+      "grad_norm": 4.11192512512207,
+      "learning_rate": 4.8632003625800995e-06,
+      "loss": 0.5551,
+      "step": 139
+    },
+    {
+      "epoch": 0.1353637901861252,
+      "grad_norm": 4.1996259689331055,
+      "learning_rate": 4.860631427648179e-06,
+      "loss": 0.5325,
+      "step": 140
+    },
+    {
+      "epoch": 0.1363306744017404,
+      "grad_norm": 3.9488883018493652,
+      "learning_rate": 4.858039287152095e-06,
+      "loss": 0.5976,
+      "step": 141
+    },
+    {
+      "epoch": 0.13729755861735557,
+      "grad_norm": 3.8057913780212402,
+      "learning_rate": 4.855423966573202e-06,
+      "loss": 0.5458,
+      "step": 142
+    },
+    {
+      "epoch": 0.13826444283297076,
+      "grad_norm": 3.676607131958008,
+      "learning_rate": 4.852785491620716e-06,
+      "loss": 0.4622,
+      "step": 143
+    },
+    {
+      "epoch": 0.13923132704858593,
+      "grad_norm": 4.006147384643555,
+      "learning_rate": 4.850123888231472e-06,
+      "loss": 0.4659,
+      "step": 144
+    },
+    {
+      "epoch": 0.14019821126420112,
+      "grad_norm": 3.4836902618408203,
+      "learning_rate": 4.847439182569656e-06,
+      "loss": 0.5167,
+      "step": 145
+    },
+    {
+      "epoch": 0.1411650954798163,
+      "grad_norm": 4.223696231842041,
+      "learning_rate": 4.844731401026562e-06,
+      "loss": 0.5559,
+      "step": 146
+    },
+    {
+      "epoch": 0.14213197969543148,
+      "grad_norm": 4.1777262687683105,
+      "learning_rate": 4.84200057022032e-06,
+      "loss": 0.5509,
+      "step": 147
+    },
+    {
+      "epoch": 0.14309886391104665,
+      "grad_norm": 3.9227278232574463,
+      "learning_rate": 4.8392467169956405e-06,
+      "loss": 0.5256,
+      "step": 148
+    },
+    {
+      "epoch": 0.14406574812666184,
+      "grad_norm": 3.8219034671783447,
+      "learning_rate": 4.836469868423552e-06,
+      "loss": 0.4913,
+      "step": 149
+    },
+    {
+      "epoch": 0.145032632342277,
+      "grad_norm": 3.7245306968688965,
+      "learning_rate": 4.833670051801131e-06,
+      "loss": 0.4811,
+      "step": 150
+    },
+    {
+      "epoch": 0.1459995165578922,
+      "grad_norm": 3.941112756729126,
+      "learning_rate": 4.830847294651236e-06,
+      "loss": 0.5529,
+      "step": 151
+    },
+    {
+      "epoch": 0.14696640077350737,
+      "grad_norm": 3.680382013320923,
+      "learning_rate": 4.8280016247222355e-06,
+      "loss": 0.4964,
+      "step": 152
+    },
+    {
+      "epoch": 0.14793328498912256,
+      "grad_norm": 4.14690637588501,
+      "learning_rate": 4.825133069987737e-06,
+      "loss": 0.5186,
+      "step": 153
+    },
+    {
+      "epoch": 0.14890016920473773,
+      "grad_norm": 3.8385708332061768,
+      "learning_rate": 4.822241658646312e-06,
+      "loss": 0.4802,
+      "step": 154
+    },
+    {
+      "epoch": 0.14986705342035292,
+      "grad_norm": 4.05110502243042,
+      "learning_rate": 4.819327419121215e-06,
+      "loss": 0.5394,
+      "step": 155
+    },
+    {
+      "epoch": 0.15083393763596809,
+      "grad_norm": 4.225109100341797,
+      "learning_rate": 4.816390380060108e-06,
+      "loss": 0.5902,
+      "step": 156
+    },
+    {
+      "epoch": 0.15180082185158328,
+      "grad_norm": 3.542275905609131,
+      "learning_rate": 4.81343057033478e-06,
+      "loss": 0.5075,
+      "step": 157
+    },
+    {
+      "epoch": 0.15276770606719844,
+      "grad_norm": 4.327058792114258,
+      "learning_rate": 4.810448019040857e-06,
+      "loss": 0.5222,
+      "step": 158
+    },
+    {
+      "epoch": 0.15373459028281364,
+      "grad_norm": 3.977292060852051,
+      "learning_rate": 4.8074427554975235e-06,
+      "loss": 0.5177,
+      "step": 159
+    },
+    {
+      "epoch": 0.1547014744984288,
+      "grad_norm": 3.762315511703491,
+      "learning_rate": 4.804414809247231e-06,
+      "loss": 0.485,
+      "step": 160
+    },
+    {
+      "epoch": 0.155668358714044,
+      "grad_norm": 4.200908660888672,
+      "learning_rate": 4.8013642100554034e-06,
+      "loss": 0.4808,
+      "step": 161
+    },
+    {
+      "epoch": 0.15663524292965916,
+      "grad_norm": 4.091949939727783,
+      "learning_rate": 4.798290987910152e-06,
+      "loss": 0.5482,
+      "step": 162
+    },
+    {
+      "epoch": 0.15760212714527436,
+      "grad_norm": 3.7029221057891846,
+      "learning_rate": 4.795195173021976e-06,
+      "loss": 0.4768,
+      "step": 163
+    },
+    {
+      "epoch": 0.15856901136088952,
+      "grad_norm": 4.126641750335693,
+      "learning_rate": 4.792076795823466e-06,
+      "loss": 0.5219,
+      "step": 164
+    },
+    {
+      "epoch": 0.15953589557650472,
+      "grad_norm": 4.042599201202393,
+      "learning_rate": 4.7889358869690065e-06,
+      "loss": 0.4618,
+      "step": 165
+    },
+    {
+      "epoch": 0.16050277979211988,
+      "grad_norm": 3.8868730068206787,
+      "learning_rate": 4.785772477334473e-06,
+      "loss": 0.5551,
+      "step": 166
+    },
+    {
+      "epoch": 0.16146966400773508,
+      "grad_norm": 3.4996016025543213,
+      "learning_rate": 4.782586598016928e-06,
+      "loss": 0.4826,
+      "step": 167
+    },
+    {
+      "epoch": 0.16243654822335024,
+      "grad_norm": 3.912738084793091,
+      "learning_rate": 4.779378280334318e-06,
+      "loss": 0.5114,
+      "step": 168
+    },
+    {
+      "epoch": 0.16340343243896543,
+      "grad_norm": 8.732720375061035,
+      "learning_rate": 4.776147555825164e-06,
+      "loss": 0.5314,
+      "step": 169
+    },
+    {
+      "epoch": 0.1643703166545806,
+      "grad_norm": 4.118671417236328,
+      "learning_rate": 4.772894456248248e-06,
+      "loss": 0.4919,
+      "step": 170
+    },
+    {
+      "epoch": 0.1653372008701958,
+      "grad_norm": 3.8536298274993896,
+      "learning_rate": 4.769619013582309e-06,
+      "loss": 0.5183,
+      "step": 171
+    },
+    {
+      "epoch": 0.166304085085811,
+      "grad_norm": 3.6809961795806885,
+      "learning_rate": 4.766321260025721e-06,
+      "loss": 0.4654,
+      "step": 172
+    },
+    {
+      "epoch": 0.16727096930142615,
+      "grad_norm": 3.947495222091675,
+      "learning_rate": 4.7630012279961805e-06,
+      "loss": 0.45,
+      "step": 173
+    },
+    {
+      "epoch": 0.16823785351704135,
+      "grad_norm": 3.8548734188079834,
+      "learning_rate": 4.7596589501303845e-06,
+      "loss": 0.5412,
+      "step": 174
+    },
+    {
+      "epoch": 0.1692047377326565,
+      "grad_norm": 3.772038698196411,
+      "learning_rate": 4.7562944592837145e-06,
+      "loss": 0.4951,
+      "step": 175
+    },
+    {
+      "epoch": 0.1701716219482717,
+      "grad_norm": 4.423032760620117,
+      "learning_rate": 4.7529077885299095e-06,
+      "loss": 0.5264,
+      "step": 176
+    },
+    {
+      "epoch": 0.17113850616388687,
+      "grad_norm": 3.9457714557647705,
+      "learning_rate": 4.749498971160742e-06,
+      "loss": 0.5227,
+      "step": 177
+    },
+    {
+      "epoch": 0.17210539037950207,
+      "grad_norm": 3.9973223209381104,
+      "learning_rate": 4.746068040685692e-06,
+      "loss": 0.5296,
+      "step": 178
+    },
+    {
+      "epoch": 0.17307227459511723,
+      "grad_norm": 3.861050605773926,
+      "learning_rate": 4.742615030831615e-06,
+      "loss": 0.5122,
+      "step": 179
+    },
+    {
+      "epoch": 0.17403915881073242,
+      "grad_norm": 3.90736722946167,
+      "learning_rate": 4.739139975542415e-06,
+      "loss": 0.5479,
+      "step": 180
+    },
+    {
+      "epoch": 0.1750060430263476,
+      "grad_norm": 4.067862510681152,
+      "learning_rate": 4.735642908978704e-06,
+      "loss": 0.5305,
+      "step": 181
+    },
+    {
+      "epoch": 0.17597292724196278,
+      "grad_norm": 3.789336919784546,
+      "learning_rate": 4.7321238655174715e-06,
+      "loss": 0.4683,
+      "step": 182
+    },
+    {
+      "epoch": 0.17693981145757795,
+      "grad_norm": 3.6877589225769043,
+      "learning_rate": 4.728582879751746e-06,
+      "loss": 0.4879,
+      "step": 183
+    },
+    {
+      "epoch": 0.17790669567319314,
+      "grad_norm": 3.7258737087249756,
+      "learning_rate": 4.725019986490255e-06,
+      "loss": 0.4918,
+      "step": 184
+    },
+    {
+      "epoch": 0.1788735798888083,
+      "grad_norm": 3.882462501525879,
+      "learning_rate": 4.721435220757078e-06,
+      "loss": 0.5517,
+      "step": 185
+    },
+    {
+      "epoch": 0.1798404641044235,
+      "grad_norm": 4.0587568283081055,
+      "learning_rate": 4.717828617791308e-06,
+      "loss": 0.4685,
+      "step": 186
+    },
+    {
+      "epoch": 0.18080734832003867,
+      "grad_norm": 3.7173986434936523,
+      "learning_rate": 4.714200213046707e-06,
+      "loss": 0.5425,
+      "step": 187
+    },
+    {
+      "epoch": 0.18177423253565386,
+      "grad_norm": 4.385982513427734,
+      "learning_rate": 4.7105500421913465e-06,
+      "loss": 0.4914,
+      "step": 188
+    },
+    {
+      "epoch": 0.18274111675126903,
+      "grad_norm": 4.058135509490967,
+      "learning_rate": 4.706878141107269e-06,
+      "loss": 0.5249,
+      "step": 189
+    },
+    {
+      "epoch": 0.18370800096688422,
+      "grad_norm": 4.178554058074951,
+      "learning_rate": 4.703184545890129e-06,
+      "loss": 0.4403,
+      "step": 190
+    },
+    {
+      "epoch": 0.1846748851824994,
+      "grad_norm": 3.969520330429077,
+      "learning_rate": 4.699469292848839e-06,
+      "loss": 0.5016,
+      "step": 191
+    },
+    {
+      "epoch": 0.18564176939811458,
+      "grad_norm": 4.418413162231445,
+      "learning_rate": 4.695732418505214e-06,
+      "loss": 0.5239,
+      "step": 192
+    },
+    {
+      "epoch": 0.18660865361372975,
+      "grad_norm": 4.015957832336426,
+      "learning_rate": 4.691973959593609e-06,
+      "loss": 0.522,
+      "step": 193
+    },
+    {
+      "epoch": 0.18757553782934494,
+      "grad_norm": 4.055311679840088,
+      "learning_rate": 4.688193953060566e-06,
+      "loss": 0.4946,
+      "step": 194
+    },
+    {
+      "epoch": 0.1885424220449601,
+      "grad_norm": 4.2180585861206055,
+      "learning_rate": 4.6843924360644385e-06,
+      "loss": 0.54,
+      "step": 195
+    },
+    {
+      "epoch": 0.1895093062605753,
+      "grad_norm": 3.808861494064331,
+      "learning_rate": 4.680569445975036e-06,
+      "loss": 0.5254,
+      "step": 196
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 3.952702522277832,
+      "learning_rate": 4.676725020373255e-06,
+      "loss": 0.4599,
+      "step": 197
+    },
+    {
+      "epoch": 0.19144307469180566,
+      "grad_norm": 3.55468487739563,
+      "learning_rate": 4.672859197050706e-06,
+      "loss": 0.4616,
+      "step": 198
+    },
+    {
+      "epoch": 0.19240995890742082,
+      "grad_norm": 3.6720266342163086,
+      "learning_rate": 4.6689720140093445e-06,
+      "loss": 0.5532,
+      "step": 199
+    },
+    {
+      "epoch": 0.19337684312303602,
+      "grad_norm": 3.75171160697937,
+      "learning_rate": 4.665063509461098e-06,
+      "loss": 0.407,
+      "step": 200
+    },
+    {
+      "epoch": 0.19434372733865118,
+      "grad_norm": 3.524125099182129,
+      "learning_rate": 4.661133721827487e-06,
+      "loss": 0.458,
+      "step": 201
+    },
+    {
+      "epoch": 0.19531061155426638,
+      "grad_norm": 4.0818023681640625,
+      "learning_rate": 4.657182689739253e-06,
+      "loss": 0.5092,
+      "step": 202
+    },
+    {
+      "epoch": 0.19627749576988154,
+      "grad_norm": 3.552830219268799,
+      "learning_rate": 4.653210452035974e-06,
+      "loss": 0.4761,
+      "step": 203
+    },
+    {
+      "epoch": 0.19724437998549674,
+      "grad_norm": 3.4877846240997314,
+      "learning_rate": 4.649217047765685e-06,
+      "loss": 0.5329,
+      "step": 204
+    },
+    {
+      "epoch": 0.1982112642011119,
+      "grad_norm": 3.507387161254883,
+      "learning_rate": 4.645202516184492e-06,
+      "loss": 0.4731,
+      "step": 205
+    },
+    {
+      "epoch": 0.1991781484167271,
+      "grad_norm": 4.230087757110596,
+      "learning_rate": 4.641166896756188e-06,
+      "loss": 0.4747,
+      "step": 206
+    },
+    {
+      "epoch": 0.2001450326323423,
+      "grad_norm": 3.850822687149048,
+      "learning_rate": 4.6371102291518635e-06,
+      "loss": 0.5242,
+      "step": 207
+    },
+    {
+      "epoch": 0.20111191684795746,
+      "grad_norm": 3.663846969604492,
+      "learning_rate": 4.63303255324952e-06,
+      "loss": 0.4807,
+      "step": 208
+    },
+    {
+      "epoch": 0.20207880106357265,
+      "grad_norm": 3.500422477722168,
+      "learning_rate": 4.628933909133674e-06,
+      "loss": 0.4273,
+      "step": 209
+    },
+    {
+      "epoch": 0.20304568527918782,
+      "grad_norm": 3.770920753479004,
+      "learning_rate": 4.624814337094964e-06,
+      "loss": 0.5495,
+      "step": 210
+    },
+    {
+      "epoch": 0.204012569494803,
+      "grad_norm": 3.9864301681518555,
+      "learning_rate": 4.620673877629757e-06,
+      "loss": 0.5147,
+      "step": 211
+    },
+    {
+      "epoch": 0.20497945371041817,
+      "grad_norm": 4.359103679656982,
+      "learning_rate": 4.616512571439745e-06,
+      "loss": 0.4922,
+      "step": 212
+    },
+    {
+      "epoch": 0.20594633792603337,
+      "grad_norm": 3.866464376449585,
+      "learning_rate": 4.612330459431552e-06,
+      "loss": 0.4987,
+      "step": 213
+    },
+    {
+      "epoch": 0.20691322214164853,
+      "grad_norm": 4.180274963378906,
+      "learning_rate": 4.608127582716324e-06,
+      "loss": 0.5276,
+      "step": 214
+    },
+    {
+      "epoch": 0.20788010635726373,
+      "grad_norm": 3.7912492752075195,
+      "learning_rate": 4.603903982609334e-06,
+      "loss": 0.5121,
+      "step": 215
+    },
+    {
+      "epoch": 0.2088469905728789,
+      "grad_norm": 3.7710208892822266,
+      "learning_rate": 4.599659700629566e-06,
+      "loss": 0.5193,
+      "step": 216
+    },
+    {
+      "epoch": 0.2098138747884941,
+      "grad_norm": 3.5846283435821533,
+      "learning_rate": 4.595394778499314e-06,
+      "loss": 0.4799,
+      "step": 217
+    },
+    {
+      "epoch": 0.21078075900410925,
+      "grad_norm": 3.6125054359436035,
+      "learning_rate": 4.59110925814377e-06,
+      "loss": 0.5133,
+      "step": 218
+    },
+    {
+      "epoch": 0.21174764321972445,
+      "grad_norm": 4.170721530914307,
+      "learning_rate": 4.586803181690609e-06,
+      "loss": 0.5695,
+      "step": 219
+    },
+    {
+      "epoch": 0.2127145274353396,
+      "grad_norm": 3.4708595275878906,
+      "learning_rate": 4.582476591469582e-06,
+      "loss": 0.4518,
+      "step": 220
+    },
+    {
+      "epoch": 0.2136814116509548,
+      "grad_norm": 4.279057502746582,
+      "learning_rate": 4.5781295300120885e-06,
+      "loss": 0.5111,
+      "step": 221
+    },
+    {
+      "epoch": 0.21464829586656997,
+      "grad_norm": 3.40183162689209,
+      "learning_rate": 4.573762040050772e-06,
+      "loss": 0.4127,
+      "step": 222
+    },
+    {
+      "epoch": 0.21561518008218516,
+      "grad_norm": 3.459470748901367,
+      "learning_rate": 4.569374164519088e-06,
+      "loss": 0.4438,
+      "step": 223
+    },
+    {
+      "epoch": 0.21658206429780033,
+      "grad_norm": 4.130403518676758,
+      "learning_rate": 4.564965946550889e-06,
+      "loss": 0.5337,
+      "step": 224
+    },
+    {
+      "epoch": 0.21754894851341552,
+      "grad_norm": 3.649747371673584,
+      "learning_rate": 4.560537429479998e-06,
+      "loss": 0.5272,
+      "step": 225
+    },
+    {
+      "epoch": 0.2185158327290307,
+      "grad_norm": 3.7949132919311523,
+      "learning_rate": 4.556088656839785e-06,
+      "loss": 0.4954,
+      "step": 226
+    },
+    {
+      "epoch": 0.21948271694464588,
+      "grad_norm": 3.963280200958252,
+      "learning_rate": 4.5516196723627325e-06,
+      "loss": 0.5444,
+      "step": 227
+    },
+    {
+      "epoch": 0.22044960116026105,
+      "grad_norm": 3.713867425918579,
+      "learning_rate": 4.547130519980014e-06,
+      "loss": 0.4787,
+      "step": 228
+    },
+    {
+      "epoch": 0.22141648537587624,
+      "grad_norm": 3.647979736328125,
+      "learning_rate": 4.542621243821058e-06,
+      "loss": 0.4688,
+      "step": 229
+    },
+    {
+      "epoch": 0.2223833695914914,
+      "grad_norm": 4.001070499420166,
+      "learning_rate": 4.538091888213112e-06,
+      "loss": 0.4714,
+      "step": 230
+    },
+    {
+      "epoch": 0.2233502538071066,
+      "grad_norm": 4.070573806762695,
+      "learning_rate": 4.533542497680811e-06,
+      "loss": 0.5579,
+      "step": 231
+    },
+    {
+      "epoch": 0.22431713802272177,
+      "grad_norm": 3.7523491382598877,
+      "learning_rate": 4.52897311694574e-06,
+      "loss": 0.4251,
+      "step": 232
+    },
+    {
+      "epoch": 0.22528402223833696,
+      "grad_norm": 3.940523147583008,
+      "learning_rate": 4.524383790925987e-06,
+      "loss": 0.5004,
+      "step": 233
+    },
+    {
+      "epoch": 0.22625090645395213,
+      "grad_norm": 3.9226436614990234,
+      "learning_rate": 4.519774564735712e-06,
+      "loss": 0.505,
+      "step": 234
+    },
+    {
+      "epoch": 0.22721779066956732,
+      "grad_norm": 3.784249782562256,
+      "learning_rate": 4.515145483684696e-06,
+      "loss": 0.5336,
+      "step": 235
+    },
+    {
+      "epoch": 0.2281846748851825,
+      "grad_norm": 3.493046760559082,
+      "learning_rate": 4.510496593277898e-06,
+      "loss": 0.4649,
+      "step": 236
+    },
+    {
+      "epoch": 0.22915155910079768,
+      "grad_norm": 3.4909865856170654,
+      "learning_rate": 4.505827939215009e-06,
+      "loss": 0.52,
+      "step": 237
+    },
+    {
+      "epoch": 0.23011844331641285,
+      "grad_norm": 3.6906330585479736,
+      "learning_rate": 4.50113956739e-06,
+      "loss": 0.463,
+      "step": 238
+    },
+    {
+      "epoch": 0.23108532753202804,
+      "grad_norm": 3.881237268447876,
+      "learning_rate": 4.496431523890673e-06,
+      "loss": 0.5222,
+      "step": 239
+    },
+    {
+      "epoch": 0.23205221174764323,
+      "grad_norm": 3.843491554260254,
+      "learning_rate": 4.4917038549982075e-06,
+      "loss": 0.5514,
+      "step": 240
+    },
+    {
+      "epoch": 0.2330190959632584,
+      "grad_norm": 3.859520196914673,
+      "learning_rate": 4.486956607186702e-06,
+      "loss": 0.545,
+      "step": 241
+    },
+    {
+      "epoch": 0.2339859801788736,
+      "grad_norm": 3.564976215362549,
+      "learning_rate": 4.482189827122726e-06,
+      "loss": 0.4548,
+      "step": 242
+    },
+    {
+      "epoch": 0.23495286439448876,
+      "grad_norm": 3.7164249420166016,
+      "learning_rate": 4.477403561664852e-06,
+      "loss": 0.5378,
+      "step": 243
+    },
+    {
+      "epoch": 0.23591974861010395,
+      "grad_norm": 3.8178417682647705,
+      "learning_rate": 4.472597857863198e-06,
+      "loss": 0.446,
+      "step": 244
+    },
+    {
+      "epoch": 0.23688663282571912,
+      "grad_norm": 3.6128547191619873,
+      "learning_rate": 4.467772762958968e-06,
+      "loss": 0.4653,
+      "step": 245
+    },
+    {
+      "epoch": 0.2378535170413343,
+      "grad_norm": 4.037587642669678,
+      "learning_rate": 4.462928324383985e-06,
+      "loss": 0.57,
+      "step": 246
+    },
+    {
+      "epoch": 0.23882040125694948,
+      "grad_norm": 4.266468048095703,
+      "learning_rate": 4.458064589760221e-06,
+      "loss": 0.4874,
+      "step": 247
+    },
+    {
+      "epoch": 0.23978728547256467,
+      "grad_norm": 3.4381649494171143,
+      "learning_rate": 4.453181606899338e-06,
+      "loss": 0.4447,
+      "step": 248
+    },
+    {
+      "epoch": 0.24075416968817984,
+      "grad_norm": 4.215127468109131,
+      "learning_rate": 4.448279423802207e-06,
+      "loss": 0.5358,
+      "step": 249
+    },
+    {
+      "epoch": 0.24172105390379503,
+      "grad_norm": 3.6681251525878906,
+      "learning_rate": 4.443358088658448e-06,
+      "loss": 0.4914,
+      "step": 250
+    },
+    {
+      "epoch": 0.2426879381194102,
+      "grad_norm": 3.6266114711761475,
+      "learning_rate": 4.438417649845946e-06,
+      "loss": 0.5211,
+      "step": 251
+    },
+    {
+      "epoch": 0.2436548223350254,
+      "grad_norm": 4.406180381774902,
+      "learning_rate": 4.43345815593038e-06,
+      "loss": 0.5456,
+      "step": 252
+    },
+    {
+      "epoch": 0.24462170655064056,
+      "grad_norm": 3.685242176055908,
+      "learning_rate": 4.428479655664748e-06,
+      "loss": 0.4988,
+      "step": 253
+    },
+    {
+      "epoch": 0.24558859076625575,
+      "grad_norm": 3.8363120555877686,
+      "learning_rate": 4.423482197988881e-06,
+      "loss": 0.4961,
+      "step": 254
+    },
+    {
+      "epoch": 0.24655547498187091,
+      "grad_norm": 3.825303554534912,
+      "learning_rate": 4.4184658320289675e-06,
+      "loss": 0.4879,
+      "step": 255
+    },
+    {
+      "epoch": 0.2475223591974861,
+      "grad_norm": 3.914113759994507,
+      "learning_rate": 4.41343060709707e-06,
+      "loss": 0.5032,
+      "step": 256
+    },
+    {
+      "epoch": 0.24848924341310127,
+      "grad_norm": 3.840186357498169,
+      "learning_rate": 4.408376572690638e-06,
+      "loss": 0.5148,
+      "step": 257
+    },
+    {
+      "epoch": 0.24945612762871647,
+      "grad_norm": 3.9938724040985107,
+      "learning_rate": 4.403303778492023e-06,
+      "loss": 0.4698,
+      "step": 258
+    },
+    {
+      "epoch": 0.25042301184433163,
+      "grad_norm": 3.904791831970215,
+      "learning_rate": 4.3982122743679875e-06,
+      "loss": 0.5339,
+      "step": 259
+    },
+    {
+      "epoch": 0.2513898960599468,
+      "grad_norm": 3.4951677322387695,
+      "learning_rate": 4.3931021103692206e-06,
+      "loss": 0.4536,
+      "step": 260
+    },
+    {
+      "epoch": 0.252356780275562,
+      "grad_norm": 3.8062729835510254,
+      "learning_rate": 4.387973336729841e-06,
+      "loss": 0.4929,
+      "step": 261
+    },
+    {
+      "epoch": 0.25332366449117716,
+      "grad_norm": 4.109336853027344,
+      "learning_rate": 4.382826003866903e-06,
+      "loss": 0.5335,
+      "step": 262
+    },
+    {
+      "epoch": 0.25429054870679235,
+      "grad_norm": 3.5560543537139893,
+      "learning_rate": 4.377660162379904e-06,
+      "loss": 0.4673,
+      "step": 263
+    },
+    {
+      "epoch": 0.25525743292240755,
+      "grad_norm": 3.9022483825683594,
+      "learning_rate": 4.372475863050286e-06,
+      "loss": 0.5222,
+      "step": 264
+    },
+    {
+      "epoch": 0.25622431713802274,
+      "grad_norm": 3.8836610317230225,
+      "learning_rate": 4.3672731568409344e-06,
+      "loss": 0.4954,
+      "step": 265
+    },
+    {
+      "epoch": 0.2571912013536379,
+      "grad_norm": 3.6870367527008057,
+      "learning_rate": 4.36205209489568e-06,
+      "loss": 0.4124,
+      "step": 266
+    },
+    {
+      "epoch": 0.25815808556925307,
+      "grad_norm": 3.783818244934082,
+      "learning_rate": 4.3568127285387925e-06,
+      "loss": 0.4833,
+      "step": 267
+    },
+    {
+      "epoch": 0.25912496978486826,
+      "grad_norm": 3.5868349075317383,
+      "learning_rate": 4.351555109274482e-06,
+      "loss": 0.4759,
+      "step": 268
+    },
+    {
+      "epoch": 0.26009185400048346,
+      "grad_norm": 3.780407667160034,
+      "learning_rate": 4.346279288786387e-06,
+      "loss": 0.4834,
+      "step": 269
+    },
+    {
+      "epoch": 0.2610587382160986,
+      "grad_norm": 3.6964030265808105,
+      "learning_rate": 4.340985318937066e-06,
+      "loss": 0.4428,
+      "step": 270
+    },
+    {
+      "epoch": 0.2620256224317138,
+      "grad_norm": 3.7668895721435547,
+      "learning_rate": 4.3356732517674935e-06,
+      "loss": 0.5027,
+      "step": 271
+    },
+    {
+      "epoch": 0.262992506647329,
+      "grad_norm": 4.010104656219482,
+      "learning_rate": 4.330343139496545e-06,
+      "loss": 0.5497,
+      "step": 272
+    },
+    {
+      "epoch": 0.2639593908629442,
+      "grad_norm": 3.8650081157684326,
+      "learning_rate": 4.32499503452048e-06,
+      "loss": 0.5673,
+      "step": 273
+    },
+    {
+      "epoch": 0.26492627507855937,
+      "grad_norm": 3.683579444885254,
+      "learning_rate": 4.319628989412435e-06,
+      "loss": 0.501,
+      "step": 274
+    },
+    {
+      "epoch": 0.2658931592941745,
+      "grad_norm": 4.045727252960205,
+      "learning_rate": 4.314245056921899e-06,
+      "loss": 0.5478,
+      "step": 275
+    },
+    {
+      "epoch": 0.2668600435097897,
+      "grad_norm": 4.400568008422852,
+      "learning_rate": 4.308843289974198e-06,
+      "loss": 0.6053,
+      "step": 276
+    },
+    {
+      "epoch": 0.2678269277254049,
+      "grad_norm": 3.3865373134613037,
+      "learning_rate": 4.303423741669978e-06,
+      "loss": 0.4358,
+      "step": 277
+    },
+    {
+      "epoch": 0.2687938119410201,
+      "grad_norm": 3.7650463581085205,
+      "learning_rate": 4.297986465284678e-06,
+      "loss": 0.473,
+      "step": 278
+    },
+    {
+      "epoch": 0.2697606961566352,
+      "grad_norm": 3.752685546875,
+      "learning_rate": 4.292531514268008e-06,
+      "loss": 0.4686,
+      "step": 279
+    },
+    {
+      "epoch": 0.2707275803722504,
+      "grad_norm": 4.085626125335693,
+      "learning_rate": 4.287058942243424e-06,
+      "loss": 0.46,
+      "step": 280
+    },
+    {
+      "epoch": 0.2716944645878656,
+      "grad_norm": 4.068567752838135,
+      "learning_rate": 4.281568803007601e-06,
+      "loss": 0.4952,
+      "step": 281
+    },
+    {
+      "epoch": 0.2726613488034808,
+      "grad_norm": 4.001227378845215,
+      "learning_rate": 4.276061150529903e-06,
+      "loss": 0.5389,
+      "step": 282
+    },
+    {
+      "epoch": 0.27362823301909595,
+      "grad_norm": 3.950711488723755,
+      "learning_rate": 4.270536038951855e-06,
+      "loss": 0.5659,
+      "step": 283
+    },
+    {
+      "epoch": 0.27459511723471114,
+      "grad_norm": 3.8639347553253174,
+      "learning_rate": 4.2649935225866066e-06,
+      "loss": 0.541,
+      "step": 284
+    },
+    {
+      "epoch": 0.27556200145032633,
+      "grad_norm": 3.983407735824585,
+      "learning_rate": 4.259433655918404e-06,
+      "loss": 0.5081,
+      "step": 285
+    },
+    {
+      "epoch": 0.2765288856659415,
+      "grad_norm": 3.7507271766662598,
+      "learning_rate": 4.253856493602047e-06,
+      "loss": 0.4813,
+      "step": 286
+    },
+    {
+      "epoch": 0.27749576988155666,
+      "grad_norm": 3.543520927429199,
+      "learning_rate": 4.24826209046236e-06,
+      "loss": 0.4907,
+      "step": 287
+    },
+    {
+      "epoch": 0.27846265409717186,
+      "grad_norm": 3.6161043643951416,
+      "learning_rate": 4.242650501493642e-06,
+      "loss": 0.5002,
+      "step": 288
+    },
+    {
+      "epoch": 0.27942953831278705,
+      "grad_norm": 3.732840061187744,
+      "learning_rate": 4.237021781859143e-06,
+      "loss": 0.4547,
+      "step": 289
+    },
+    {
+      "epoch": 0.28039642252840224,
+      "grad_norm": 3.6558990478515625,
+      "learning_rate": 4.231375986890502e-06,
+      "loss": 0.4796,
+      "step": 290
+    },
+    {
+      "epoch": 0.2813633067440174,
+      "grad_norm": 3.682800769805908,
+      "learning_rate": 4.225713172087216e-06,
+      "loss": 0.4748,
+      "step": 291
+    },
+    {
+      "epoch": 0.2823301909596326,
+      "grad_norm": 4.0862040519714355,
+      "learning_rate": 4.220033393116093e-06,
+      "loss": 0.5725,
+      "step": 292
+    },
+    {
+      "epoch": 0.28329707517524777,
+      "grad_norm": 3.8720784187316895,
+      "learning_rate": 4.2143367058107e-06,
+      "loss": 0.5085,
+      "step": 293
+    },
+    {
+      "epoch": 0.28426395939086296,
+      "grad_norm": 3.7148096561431885,
+      "learning_rate": 4.208623166170819e-06,
+      "loss": 0.4869,
+      "step": 294
+    },
+    {
+      "epoch": 0.2852308436064781,
+      "grad_norm": 4.130632400512695,
+      "learning_rate": 4.202892830361892e-06,
+      "loss": 0.4793,
+      "step": 295
+    },
+    {
+      "epoch": 0.2861977278220933,
+      "grad_norm": 3.892319440841675,
+      "learning_rate": 4.197145754714473e-06,
+      "loss": 0.5266,
+      "step": 296
+    },
+    {
+      "epoch": 0.2871646120377085,
+      "grad_norm": 3.7769579887390137,
+      "learning_rate": 4.191381995723672e-06,
+      "loss": 0.4877,
+      "step": 297
+    },
+    {
+      "epoch": 0.2881314962533237,
+      "grad_norm": 3.810389757156372,
+      "learning_rate": 4.185601610048601e-06,
+      "loss": 0.5253,
+      "step": 298
+    },
+    {
+      "epoch": 0.2890983804689388,
+      "grad_norm": 3.692101001739502,
+      "learning_rate": 4.179804654511816e-06,
+      "loss": 0.5104,
+      "step": 299
+    },
+    {
+      "epoch": 0.290065264684554,
+      "grad_norm": 4.029324054718018,
+      "learning_rate": 4.1739911860987575e-06,
+      "loss": 0.4898,
+      "step": 300
+    },
+    {
+      "epoch": 0.2910321489001692,
+      "grad_norm": 3.8836569786071777,
+      "learning_rate": 4.168161261957192e-06,
+      "loss": 0.54,
+      "step": 301
+    },
+    {
+      "epoch": 0.2919990331157844,
+      "grad_norm": 3.8458945751190186,
+      "learning_rate": 4.162314939396652e-06,
+      "loss": 0.501,
+      "step": 302
+    },
+    {
+      "epoch": 0.29296591733139954,
+      "grad_norm": 3.7436769008636475,
+      "learning_rate": 4.1564522758878656e-06,
+      "loss": 0.5324,
+      "step": 303
+    },
+    {
+      "epoch": 0.29393280154701473,
+      "grad_norm": 3.9301185607910156,
+      "learning_rate": 4.1505733290622e-06,
+      "loss": 0.5143,
+      "step": 304
+    },
+    {
+      "epoch": 0.2948996857626299,
+      "grad_norm": 4.087481498718262,
+      "learning_rate": 4.144678156711091e-06,
+      "loss": 0.4861,
+      "step": 305
+    },
+    {
+      "epoch": 0.2958665699782451,
+      "grad_norm": 4.267450332641602,
+      "learning_rate": 4.138766816785474e-06,
+      "loss": 0.5108,
+      "step": 306
+    },
+    {
+      "epoch": 0.29683345419386026,
+      "grad_norm": 3.688645362854004,
+      "learning_rate": 4.132839367395215e-06,
+      "loss": 0.4646,
+      "step": 307
+    },
+    {
+      "epoch": 0.29780033840947545,
+      "grad_norm": 3.762014389038086,
+      "learning_rate": 4.1268958668085415e-06,
+      "loss": 0.5341,
+      "step": 308
+    },
+    {
+      "epoch": 0.29876722262509064,
+      "grad_norm": 4.061873435974121,
+      "learning_rate": 4.120936373451467e-06,
+      "loss": 0.4799,
+      "step": 309
+    },
+    {
+      "epoch": 0.29973410684070584,
+      "grad_norm": 3.6946425437927246,
+      "learning_rate": 4.1149609459072185e-06,
+      "loss": 0.4608,
+      "step": 310
+    },
+    {
+      "epoch": 0.30070099105632103,
+      "grad_norm": 3.5779032707214355,
+      "learning_rate": 4.108969642915658e-06,
+      "loss": 0.4416,
+      "step": 311
+    },
+    {
+      "epoch": 0.30166787527193617,
+      "grad_norm": 4.075043678283691,
+      "learning_rate": 4.10296252337271e-06,
+      "loss": 0.5143,
+      "step": 312
+    },
+    {
+      "epoch": 0.30263475948755136,
+      "grad_norm": 3.8208065032958984,
+      "learning_rate": 4.096939646329775e-06,
+      "loss": 0.4665,
+      "step": 313
+    },
+    {
+      "epoch": 0.30360164370316656,
+      "grad_norm": 3.35614275932312,
+      "learning_rate": 4.090901070993159e-06,
+      "loss": 0.5053,
+      "step": 314
+    },
+    {
+      "epoch": 0.30456852791878175,
+      "grad_norm": 3.444335460662842,
+      "learning_rate": 4.08484685672348e-06,
+      "loss": 0.4517,
+      "step": 315
+    },
+    {
+      "epoch": 0.3055354121343969,
+      "grad_norm": 3.4999983310699463,
+      "learning_rate": 4.078777063035096e-06,
+      "loss": 0.4485,
+      "step": 316
+    },
+    {
+      "epoch": 0.3065022963500121,
+      "grad_norm": 3.6743931770324707,
+      "learning_rate": 4.07269174959551e-06,
+      "loss": 0.4993,
+      "step": 317
+    },
+    {
+      "epoch": 0.3074691805656273,
+      "grad_norm": 4.012587547302246,
+      "learning_rate": 4.06659097622479e-06,
+      "loss": 0.4892,
+      "step": 318
+    },
+    {
+      "epoch": 0.30843606478124247,
+      "grad_norm": 3.8606793880462646,
+      "learning_rate": 4.06047480289498e-06,
+      "loss": 0.5095,
+      "step": 319
+    },
+    {
+      "epoch": 0.3094029489968576,
+      "grad_norm": 3.4969189167022705,
+      "learning_rate": 4.0543432897295055e-06,
+      "loss": 0.4492,
+      "step": 320
+    },
+    {
+      "epoch": 0.3103698332124728,
+      "grad_norm": 3.996771812438965,
+      "learning_rate": 4.0481964970025885e-06,
+      "loss": 0.5217,
+      "step": 321
+    },
+    {
+      "epoch": 0.311336717428088,
+      "grad_norm": 3.5888161659240723,
+      "learning_rate": 4.042034485138653e-06,
+      "loss": 0.5112,
+      "step": 322
+    },
+    {
+      "epoch": 0.3123036016437032,
+      "grad_norm": 3.8949966430664062,
+      "learning_rate": 4.035857314711729e-06,
+      "loss": 0.509,
+      "step": 323
+    },
+    {
+      "epoch": 0.3132704858593183,
+      "grad_norm": 3.5536141395568848,
+      "learning_rate": 4.029665046444862e-06,
+      "loss": 0.4585,
+      "step": 324
+    },
+    {
+      "epoch": 0.3142373700749335,
+      "grad_norm": 3.8758373260498047,
+      "learning_rate": 4.023457741209509e-06,
+      "loss": 0.5344,
+      "step": 325
+    },
+    {
+      "epoch": 0.3152042542905487,
+      "grad_norm": 3.6032464504241943,
+      "learning_rate": 4.0172354600249465e-06,
+      "loss": 0.4507,
+      "step": 326
+    },
+    {
+      "epoch": 0.3161711385061639,
+      "grad_norm": 3.7095227241516113,
+      "learning_rate": 4.0109982640576676e-06,
+      "loss": 0.5182,
+      "step": 327
+    },
+    {
+      "epoch": 0.31713802272177904,
+      "grad_norm": 3.854598045349121,
+      "learning_rate": 4.0047462146207795e-06,
+      "loss": 0.5601,
+      "step": 328
+    },
+    {
+      "epoch": 0.31810490693739424,
+      "grad_norm": 3.6497013568878174,
+      "learning_rate": 3.998479373173406e-06,
+      "loss": 0.4906,
+      "step": 329
+    },
+    {
+      "epoch": 0.31907179115300943,
+      "grad_norm": 3.552006483078003,
+      "learning_rate": 3.9921978013200764e-06,
+      "loss": 0.5249,
+      "step": 330
+    },
+    {
+      "epoch": 0.3200386753686246,
+      "grad_norm": 3.874676465988159,
+      "learning_rate": 3.985901560810126e-06,
+      "loss": 0.5477,
+      "step": 331
+    },
+    {
+      "epoch": 0.32100555958423976,
+      "grad_norm": 3.7521750926971436,
+      "learning_rate": 3.979590713537084e-06,
+      "loss": 0.5294,
+      "step": 332
+    },
+    {
+      "epoch": 0.32197244379985496,
+      "grad_norm": 3.6338000297546387,
+      "learning_rate": 3.973265321538069e-06,
+      "loss": 0.5001,
+      "step": 333
+    },
+    {
+      "epoch": 0.32293932801547015,
+      "grad_norm": 3.83825945854187,
+      "learning_rate": 3.966925446993177e-06,
+      "loss": 0.4977,
+      "step": 334
+    },
+    {
+      "epoch": 0.32390621223108534,
+      "grad_norm": 3.9020495414733887,
+      "learning_rate": 3.960571152224872e-06,
+      "loss": 0.4189,
+      "step": 335
+    },
+    {
+      "epoch": 0.3248730964467005,
+      "grad_norm": 3.4360294342041016,
+      "learning_rate": 3.954202499697373e-06,
+      "loss": 0.5238,
+      "step": 336
+    },
+    {
+      "epoch": 0.3258399806623157,
+      "grad_norm": 3.462728500366211,
+      "learning_rate": 3.9478195520160355e-06,
+      "loss": 0.4317,
+      "step": 337
+    },
+    {
+      "epoch": 0.32680686487793087,
+      "grad_norm": 3.646033525466919,
+      "learning_rate": 3.941422371926742e-06,
+      "loss": 0.4208,
+      "step": 338
+    },
+    {
+      "epoch": 0.32777374909354606,
+      "grad_norm": 3.4529306888580322,
+      "learning_rate": 3.935011022315284e-06,
+      "loss": 0.4753,
+      "step": 339
+    },
+    {
+      "epoch": 0.3287406333091612,
+      "grad_norm": 3.7135887145996094,
+      "learning_rate": 3.928585566206742e-06,
+      "loss": 0.405,
+      "step": 340
+    },
+    {
+      "epoch": 0.3297075175247764,
+      "grad_norm": 3.571955442428589,
+      "learning_rate": 3.922146066764863e-06,
+      "loss": 0.4933,
+      "step": 341
+    },
+    {
+      "epoch": 0.3306744017403916,
+      "grad_norm": 3.5327978134155273,
+      "learning_rate": 3.9156925872914504e-06,
+      "loss": 0.5044,
+      "step": 342
+    },
+    {
+      "epoch": 0.3316412859560068,
+      "grad_norm": 4.009077548980713,
+      "learning_rate": 3.9092251912257286e-06,
+      "loss": 0.5184,
+      "step": 343
+    },
+    {
+      "epoch": 0.332608170171622,
+      "grad_norm": 3.705810785293579,
+      "learning_rate": 3.902743942143728e-06,
+      "loss": 0.5028,
+      "step": 344
+    },
+    {
+      "epoch": 0.3335750543872371,
+      "grad_norm": 3.7308077812194824,
+      "learning_rate": 3.896248903757658e-06,
+      "loss": 0.5334,
+      "step": 345
+    },
+    {
+      "epoch": 0.3345419386028523,
+      "grad_norm": 3.8802003860473633,
+      "learning_rate": 3.889740139915279e-06,
+      "loss": 0.5459,
+      "step": 346
+    },
+    {
+      "epoch": 0.3355088228184675,
+      "grad_norm": 3.775622606277466,
+      "learning_rate": 3.883217714599273e-06,
+      "loss": 0.498,
+      "step": 347
+    },
+    {
+      "epoch": 0.3364757070340827,
+      "grad_norm": 4.032291889190674,
+      "learning_rate": 3.876681691926625e-06,
+      "loss": 0.4995,
+      "step": 348
+    },
+    {
+      "epoch": 0.33744259124969783,
+      "grad_norm": 3.9446566104888916,
+      "learning_rate": 3.870132136147977e-06,
+      "loss": 0.5006,
+      "step": 349
+    },
+    {
+      "epoch": 0.338409475465313,
+      "grad_norm": 3.741894006729126,
+      "learning_rate": 3.8635691116470095e-06,
+      "loss": 0.4801,
+      "step": 350
+    },
+    {
+      "epoch": 0.3393763596809282,
+      "grad_norm": 3.4935619831085205,
+      "learning_rate": 3.856992682939803e-06,
+      "loss": 0.4642,
+      "step": 351
+    },
+    {
+      "epoch": 0.3403432438965434,
+      "grad_norm": 3.661100149154663,
+      "learning_rate": 3.850402914674205e-06,
+      "loss": 0.5108,
+      "step": 352
+    },
+    {
+      "epoch": 0.34131012811215855,
+      "grad_norm": 3.8849480152130127,
+      "learning_rate": 3.84379987162919e-06,
+      "loss": 0.4919,
+      "step": 353
+    },
+    {
+      "epoch": 0.34227701232777374,
+      "grad_norm": 3.7999656200408936,
+      "learning_rate": 3.837183618714233e-06,
+      "loss": 0.548,
+      "step": 354
+    },
+    {
+      "epoch": 0.34324389654338894,
+      "grad_norm": 3.730701446533203,
+      "learning_rate": 3.830554220968661e-06,
+      "loss": 0.5114,
+      "step": 355
+    },
+    {
+      "epoch": 0.34421078075900413,
+      "grad_norm": 4.108263969421387,
+      "learning_rate": 3.823911743561018e-06,
+      "loss": 0.5186,
+      "step": 356
+    },
+    {
+      "epoch": 0.34517766497461927,
+      "grad_norm": 3.543729066848755,
+      "learning_rate": 3.817256251788425e-06,
+      "loss": 0.4591,
+      "step": 357
+    },
+    {
+      "epoch": 0.34614454919023446,
+      "grad_norm": 3.803205728530884,
+      "learning_rate": 3.810587811075937e-06,
+      "loss": 0.4616,
+      "step": 358
+    },
+    {
+      "epoch": 0.34711143340584966,
+      "grad_norm": 3.6811118125915527,
+      "learning_rate": 3.803906486975901e-06,
+      "loss": 0.4723,
+      "step": 359
+    },
+    {
+      "epoch": 0.34807831762146485,
+      "grad_norm": 3.6985318660736084,
+      "learning_rate": 3.7972123451673102e-06,
+      "loss": 0.4351,
+      "step": 360
+    },
+    {
+      "epoch": 0.34904520183708,
+      "grad_norm": 3.970501661300659,
+      "learning_rate": 3.790505451455158e-06,
+      "loss": 0.4424,
+      "step": 361
+    },
+    {
+      "epoch": 0.3500120860526952,
+      "grad_norm": 3.792994976043701,
+      "learning_rate": 3.783785871769793e-06,
+      "loss": 0.5227,
+      "step": 362
+    },
+    {
+      "epoch": 0.3509789702683104,
+      "grad_norm": 3.7695376873016357,
+      "learning_rate": 3.77705367216627e-06,
+      "loss": 0.5037,
+      "step": 363
+    },
+    {
+      "epoch": 0.35194585448392557,
+      "grad_norm": 3.754580020904541,
+      "learning_rate": 3.7703089188237e-06,
+      "loss": 0.5397,
+      "step": 364
+    },
+    {
+      "epoch": 0.3529127386995407,
+      "grad_norm": 3.678025245666504,
+      "learning_rate": 3.7635516780446e-06,
+      "loss": 0.4307,
+      "step": 365
+    },
+    {
+      "epoch": 0.3538796229151559,
+      "grad_norm": 3.763305187225342,
+      "learning_rate": 3.756782016254242e-06,
+      "loss": 0.5489,
+      "step": 366
+    },
+    {
+      "epoch": 0.3548465071307711,
+      "grad_norm": 3.853353261947632,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4606,
+      "step": 367
+    },
+    {
+      "epoch": 0.3558133913463863,
+      "grad_norm": 3.5719709396362305,
+      "learning_rate": 3.743205695950695e-06,
+      "loss": 0.4928,
+      "step": 368
+    },
+    {
+      "epoch": 0.3567802755620014,
+      "grad_norm": 3.763885021209717,
+      "learning_rate": 3.7363991708959386e-06,
+      "loss": 0.4669,
+      "step": 369
+    },
+    {
+      "epoch": 0.3577471597776166,
+      "grad_norm": 3.9263694286346436,
+      "learning_rate": 3.729580491745479e-06,
+      "loss": 0.5033,
+      "step": 370
+    },
+    {
+      "epoch": 0.3587140439932318,
+      "grad_norm": 3.462216377258301,
+      "learning_rate": 3.7227497255285416e-06,
+      "loss": 0.4507,
+      "step": 371
+    },
+    {
+      "epoch": 0.359680928208847,
+      "grad_norm": 3.9465153217315674,
+      "learning_rate": 3.7159069393931717e-06,
+      "loss": 0.4838,
+      "step": 372
+    },
+    {
+      "epoch": 0.36064781242446214,
+      "grad_norm": 4.065799713134766,
+      "learning_rate": 3.709052200605572e-06,
+      "loss": 0.5648,
+      "step": 373
+    },
+    {
+      "epoch": 0.36161469664007734,
+      "grad_norm": 3.8492698669433594,
+      "learning_rate": 3.702185576549441e-06,
+      "loss": 0.4143,
+      "step": 374
+    },
+    {
+      "epoch": 0.36258158085569253,
+      "grad_norm": 3.912066698074341,
+      "learning_rate": 3.6953071347253167e-06,
+      "loss": 0.4447,
+      "step": 375
+    },
+    {
+      "epoch": 0.3635484650713077,
+      "grad_norm": 3.6531989574432373,
+      "learning_rate": 3.688416942749905e-06,
+      "loss": 0.4697,
+      "step": 376
+    },
+    {
+      "epoch": 0.3645153492869229,
+      "grad_norm": 3.5612435340881348,
+      "learning_rate": 3.6815150683554187e-06,
+      "loss": 0.473,
+      "step": 377
+    },
+    {
+      "epoch": 0.36548223350253806,
+      "grad_norm": 3.904783248901367,
+      "learning_rate": 3.674601579388913e-06,
+      "loss": 0.5046,
+      "step": 378
+    },
+    {
+      "epoch": 0.36644911771815325,
+      "grad_norm": 4.0458903312683105,
+      "learning_rate": 3.6676765438116157e-06,
+      "loss": 0.5251,
+      "step": 379
+    },
+    {
+      "epoch": 0.36741600193376844,
+      "grad_norm": 3.6802172660827637,
+      "learning_rate": 3.660740029698262e-06,
+      "loss": 0.4432,
+      "step": 380
+    },
+    {
+      "epoch": 0.36838288614938364,
+      "grad_norm": 4.1497626304626465,
+      "learning_rate": 3.6537921052364223e-06,
+      "loss": 0.4888,
+      "step": 381
+    },
+    {
+      "epoch": 0.3693497703649988,
+      "grad_norm": 4.301728248596191,
+      "learning_rate": 3.646832838725835e-06,
+      "loss": 0.5565,
+      "step": 382
+    },
+    {
+      "epoch": 0.37031665458061397,
+      "grad_norm": 3.8046703338623047,
+      "learning_rate": 3.6398622985777314e-06,
+      "loss": 0.5819,
+      "step": 383
+    },
+    {
+      "epoch": 0.37128353879622916,
+      "grad_norm": 3.943380355834961,
+      "learning_rate": 3.632880553314169e-06,
+      "loss": 0.5247,
+      "step": 384
+    },
+    {
+      "epoch": 0.37225042301184436,
+      "grad_norm": 4.136211395263672,
+      "learning_rate": 3.6258876715673475e-06,
+      "loss": 0.469,
+      "step": 385
+    },
+    {
+      "epoch": 0.3732173072274595,
+      "grad_norm": 3.70596981048584,
+      "learning_rate": 3.6188837220789496e-06,
+      "loss": 0.544,
+      "step": 386
+    },
+    {
+      "epoch": 0.3741841914430747,
+      "grad_norm": 3.52099347114563,
+      "learning_rate": 3.611868773699449e-06,
+      "loss": 0.4492,
+      "step": 387
+    },
+    {
+      "epoch": 0.3751510756586899,
+      "grad_norm": 3.8500866889953613,
+      "learning_rate": 3.604842895387446e-06,
+      "loss": 0.4829,
+      "step": 388
+    },
+    {
+      "epoch": 0.3761179598743051,
+      "grad_norm": 3.67982816696167,
+      "learning_rate": 3.597806156208982e-06,
+      "loss": 0.4814,
+      "step": 389
+    },
+    {
+      "epoch": 0.3770848440899202,
+      "grad_norm": 3.550171136856079,
+      "learning_rate": 3.5907586253368643e-06,
+      "loss": 0.4134,
+      "step": 390
+    },
+    {
+      "epoch": 0.3780517283055354,
+      "grad_norm": 3.7428512573242188,
+      "learning_rate": 3.5837003720499853e-06,
+      "loss": 0.4985,
+      "step": 391
+    },
+    {
+      "epoch": 0.3790186125211506,
+      "grad_norm": 3.5214486122131348,
+      "learning_rate": 3.57663146573264e-06,
+      "loss": 0.4259,
+      "step": 392
+    },
+    {
+      "epoch": 0.3799854967367658,
+      "grad_norm": 3.610581398010254,
+      "learning_rate": 3.569551975873847e-06,
+      "loss": 0.4875,
+      "step": 393
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 3.991929292678833,
+      "learning_rate": 3.562461972066663e-06,
+      "loss": 0.4683,
+      "step": 394
+    },
+    {
+      "epoch": 0.3819192651679961,
+      "grad_norm": 3.5541305541992188,
+      "learning_rate": 3.555361524007498e-06,
+      "loss": 0.4556,
+      "step": 395
+    },
+    {
+      "epoch": 0.3828861493836113,
+      "grad_norm": 3.4029953479766846,
+      "learning_rate": 3.5482507014954327e-06,
+      "loss": 0.5061,
+      "step": 396
+    },
+    {
+      "epoch": 0.3838530335992265,
+      "grad_norm": 3.640984058380127,
+      "learning_rate": 3.541129574431532e-06,
+      "loss": 0.434,
+      "step": 397
+    },
+    {
+      "epoch": 0.38481991781484165,
+      "grad_norm": 3.657914876937866,
+      "learning_rate": 3.5339982128181536e-06,
+      "loss": 0.4475,
+      "step": 398
+    },
+    {
+      "epoch": 0.38578680203045684,
+      "grad_norm": 4.057721138000488,
+      "learning_rate": 3.526856686758269e-06,
+      "loss": 0.4982,
+      "step": 399
+    },
+    {
+      "epoch": 0.38675368624607204,
+      "grad_norm": 3.833341121673584,
+      "learning_rate": 3.519705066454762e-06,
+      "loss": 0.5104,
+      "step": 400
+    },
+    {
+      "epoch": 0.38772057046168723,
+      "grad_norm": 3.7433853149414062,
+      "learning_rate": 3.51254342220975e-06,
+      "loss": 0.4551,
+      "step": 401
+    },
+    {
+      "epoch": 0.38868745467730237,
+      "grad_norm": 3.844796657562256,
+      "learning_rate": 3.5053718244238847e-06,
+      "loss": 0.4761,
+      "step": 402
+    },
+    {
+      "epoch": 0.38965433889291756,
+      "grad_norm": 3.6285436153411865,
+      "learning_rate": 3.4981903435956675e-06,
+      "loss": 0.4168,
+      "step": 403
+    },
+    {
+      "epoch": 0.39062122310853276,
+      "grad_norm": 3.8488385677337646,
+      "learning_rate": 3.4909990503207488e-06,
+      "loss": 0.5304,
+      "step": 404
+    },
+    {
+      "epoch": 0.39158810732414795,
+      "grad_norm": 3.696931838989258,
+      "learning_rate": 3.4837980152912393e-06,
+      "loss": 0.5355,
+      "step": 405
+    },
+    {
+      "epoch": 0.3925549915397631,
+      "grad_norm": 3.8406333923339844,
+      "learning_rate": 3.4765873092950126e-06,
+      "loss": 0.5835,
+      "step": 406
+    },
+    {
+      "epoch": 0.3935218757553783,
+      "grad_norm": 3.689175605773926,
+      "learning_rate": 3.4693670032150117e-06,
+      "loss": 0.5174,
+      "step": 407
+    },
+    {
+      "epoch": 0.3944887599709935,
+      "grad_norm": 3.8330447673797607,
+      "learning_rate": 3.4621371680285492e-06,
+      "loss": 0.5241,
+      "step": 408
+    },
+    {
+      "epoch": 0.39545564418660867,
+      "grad_norm": 3.984819173812866,
+      "learning_rate": 3.4548978748066115e-06,
+      "loss": 0.4701,
+      "step": 409
+    },
+    {
+      "epoch": 0.3964225284022238,
+      "grad_norm": 3.6191537380218506,
+      "learning_rate": 3.4476491947131596e-06,
+      "loss": 0.4835,
+      "step": 410
+    },
+    {
+      "epoch": 0.397389412617839,
+      "grad_norm": 3.8275318145751953,
+      "learning_rate": 3.440391199004431e-06,
+      "loss": 0.4989,
+      "step": 411
+    },
+    {
+      "epoch": 0.3983562968334542,
+      "grad_norm": 3.577519416809082,
+      "learning_rate": 3.4331239590282354e-06,
+      "loss": 0.4495,
+      "step": 412
+    },
+    {
+      "epoch": 0.3993231810490694,
+      "grad_norm": 3.678297281265259,
+      "learning_rate": 3.4258475462232586e-06,
+      "loss": 0.4558,
+      "step": 413
+    },
+    {
+      "epoch": 0.4002900652646846,
+      "grad_norm": 3.846923828125,
+      "learning_rate": 3.4185620321183543e-06,
+      "loss": 0.4499,
+      "step": 414
+    },
+    {
+      "epoch": 0.4012569494802997,
+      "grad_norm": 3.836149215698242,
+      "learning_rate": 3.4112674883318477e-06,
+      "loss": 0.5519,
+      "step": 415
+    },
+    {
+      "epoch": 0.4022238336959149,
+      "grad_norm": 3.9182722568511963,
+      "learning_rate": 3.403963986570825e-06,
+      "loss": 0.5123,
+      "step": 416
+    },
+    {
+      "epoch": 0.4031907179115301,
+      "grad_norm": 3.498328924179077,
+      "learning_rate": 3.3966515986304322e-06,
+      "loss": 0.4568,
+      "step": 417
+    },
+    {
+      "epoch": 0.4041576021271453,
+      "grad_norm": 3.3499412536621094,
+      "learning_rate": 3.3893303963931703e-06,
+      "loss": 0.4542,
+      "step": 418
+    },
+    {
+      "epoch": 0.40512448634276044,
+      "grad_norm": 3.920807123184204,
+      "learning_rate": 3.3820004518281835e-06,
+      "loss": 0.4842,
+      "step": 419
+    },
+    {
+      "epoch": 0.40609137055837563,
+      "grad_norm": 3.754964590072632,
+      "learning_rate": 3.37466183699056e-06,
+      "loss": 0.4783,
+      "step": 420
+    },
+    {
+      "epoch": 0.4070582547739908,
+      "grad_norm": 3.6585028171539307,
+      "learning_rate": 3.367314624020613e-06,
+      "loss": 0.4872,
+      "step": 421
+    },
+    {
+      "epoch": 0.408025138989606,
+      "grad_norm": 4.0016374588012695,
+      "learning_rate": 3.3599588851431814e-06,
+      "loss": 0.5722,
+      "step": 422
+    },
+    {
+      "epoch": 0.40899202320522116,
+      "grad_norm": 3.19661283493042,
+      "learning_rate": 3.352594692666915e-06,
+      "loss": 0.4261,
+      "step": 423
+    },
+    {
+      "epoch": 0.40995890742083635,
+      "grad_norm": 3.5285444259643555,
+      "learning_rate": 3.3452221189835642e-06,
+      "loss": 0.495,
+      "step": 424
+    },
+    {
+      "epoch": 0.41092579163645154,
+      "grad_norm": 3.6112709045410156,
+      "learning_rate": 3.337841236567268e-06,
+      "loss": 0.4101,
+      "step": 425
+    },
+    {
+      "epoch": 0.41189267585206674,
+      "grad_norm": 3.411106586456299,
+      "learning_rate": 3.3304521179738435e-06,
+      "loss": 0.4433,
+      "step": 426
+    },
+    {
+      "epoch": 0.4128595600676819,
+      "grad_norm": 3.8429133892059326,
+      "learning_rate": 3.32305483584007e-06,
+      "loss": 0.4486,
+      "step": 427
+    },
+    {
+      "epoch": 0.41382644428329707,
+      "grad_norm": 3.9495856761932373,
+      "learning_rate": 3.3156494628829778e-06,
+      "loss": 0.5468,
+      "step": 428
+    },
+    {
+      "epoch": 0.41479332849891226,
+      "grad_norm": 3.9919161796569824,
+      "learning_rate": 3.30823607189913e-06,
+      "loss": 0.5148,
+      "step": 429
+    },
+    {
+      "epoch": 0.41576021271452746,
+      "grad_norm": 3.898630380630493,
+      "learning_rate": 3.3008147357639115e-06,
+      "loss": 0.5074,
+      "step": 430
+    },
+    {
+      "epoch": 0.4167270969301426,
+      "grad_norm": 3.9545817375183105,
+      "learning_rate": 3.2933855274308067e-06,
+      "loss": 0.4954,
+      "step": 431
+    },
+    {
+      "epoch": 0.4176939811457578,
+      "grad_norm": 3.883395195007324,
+      "learning_rate": 3.2859485199306886e-06,
+      "loss": 0.5059,
+      "step": 432
+    },
+    {
+      "epoch": 0.418660865361373,
+      "grad_norm": 3.6193654537200928,
+      "learning_rate": 3.278503786371095e-06,
+      "loss": 0.5204,
+      "step": 433
+    },
+    {
+      "epoch": 0.4196277495769882,
+      "grad_norm": 4.254292964935303,
+      "learning_rate": 3.2710513999355147e-06,
+      "loss": 0.5259,
+      "step": 434
+    },
+    {
+      "epoch": 0.4205946337926033,
+      "grad_norm": 3.752901077270508,
+      "learning_rate": 3.2635914338826665e-06,
+      "loss": 0.4255,
+      "step": 435
+    },
+    {
+      "epoch": 0.4215615180082185,
+      "grad_norm": 3.523738145828247,
+      "learning_rate": 3.2561239615457764e-06,
+      "loss": 0.3845,
+      "step": 436
+    },
+    {
+      "epoch": 0.4225284022238337,
+      "grad_norm": 4.015591621398926,
+      "learning_rate": 3.2486490563318605e-06,
+      "loss": 0.4784,
+      "step": 437
+    },
+    {
+      "epoch": 0.4234952864394489,
+      "grad_norm": 3.526104211807251,
+      "learning_rate": 3.241166791721001e-06,
+      "loss": 0.4133,
+      "step": 438
+    },
+    {
+      "epoch": 0.42446217065506403,
+      "grad_norm": 3.768324375152588,
+      "learning_rate": 3.233677241265627e-06,
+      "loss": 0.5096,
+      "step": 439
+    },
+    {
+      "epoch": 0.4254290548706792,
+      "grad_norm": 3.6191327571868896,
+      "learning_rate": 3.2261804785897855e-06,
+      "loss": 0.4197,
+      "step": 440
+    },
+    {
+      "epoch": 0.4263959390862944,
+      "grad_norm": 3.9885098934173584,
+      "learning_rate": 3.218676577388424e-06,
+      "loss": 0.4907,
+      "step": 441
+    },
+    {
+      "epoch": 0.4273628233019096,
+      "grad_norm": 3.5632357597351074,
+      "learning_rate": 3.2111656114266625e-06,
+      "loss": 0.4452,
+      "step": 442
+    },
+    {
+      "epoch": 0.42832970751752475,
+      "grad_norm": 4.117615699768066,
+      "learning_rate": 3.2036476545390695e-06,
+      "loss": 0.5533,
+      "step": 443
+    },
+    {
+      "epoch": 0.42929659173313994,
+      "grad_norm": 4.148410797119141,
+      "learning_rate": 3.1961227806289358e-06,
+      "loss": 0.5051,
+      "step": 444
+    },
+    {
+      "epoch": 0.43026347594875514,
+      "grad_norm": 3.9031429290771484,
+      "learning_rate": 3.188591063667548e-06,
+      "loss": 0.5358,
+      "step": 445
+    },
+    {
+      "epoch": 0.43123036016437033,
+      "grad_norm": 3.73646879196167,
+      "learning_rate": 3.1810525776934614e-06,
+      "loss": 0.4306,
+      "step": 446
+    },
+    {
+      "epoch": 0.4321972443799855,
+      "grad_norm": 3.6302988529205322,
+      "learning_rate": 3.1735073968117743e-06,
+      "loss": 0.3704,
+      "step": 447
+    },
+    {
+      "epoch": 0.43316412859560066,
+      "grad_norm": 3.804278612136841,
+      "learning_rate": 3.165955595193393e-06,
+      "loss": 0.5157,
+      "step": 448
+    },
+    {
+      "epoch": 0.43413101281121586,
+      "grad_norm": 3.7024874687194824,
+      "learning_rate": 3.1583972470743123e-06,
+      "loss": 0.4806,
+      "step": 449
+    },
+    {
+      "epoch": 0.43509789702683105,
+      "grad_norm": 3.7965049743652344,
+      "learning_rate": 3.1508324267548775e-06,
+      "loss": 0.4658,
+      "step": 450
+    },
+    {
+      "epoch": 0.43606478124244624,
+      "grad_norm": 4.142588138580322,
+      "learning_rate": 3.1432612085990576e-06,
+      "loss": 0.5566,
+      "step": 451
+    },
+    {
+      "epoch": 0.4370316654580614,
+      "grad_norm": 3.6118361949920654,
+      "learning_rate": 3.1356836670337153e-06,
+      "loss": 0.4803,
+      "step": 452
+    },
+    {
+      "epoch": 0.4379985496736766,
+      "grad_norm": 3.7696590423583984,
+      "learning_rate": 3.1280998765478725e-06,
+      "loss": 0.4552,
+      "step": 453
+    },
+    {
+      "epoch": 0.43896543388929177,
+      "grad_norm": 3.716688632965088,
+      "learning_rate": 3.1205099116919794e-06,
+      "loss": 0.4393,
+      "step": 454
+    },
+    {
+      "epoch": 0.43993231810490696,
+      "grad_norm": 3.637057304382324,
+      "learning_rate": 3.1129138470771823e-06,
+      "loss": 0.4637,
+      "step": 455
+    },
+    {
+      "epoch": 0.4408992023205221,
+      "grad_norm": 3.727297067642212,
+      "learning_rate": 3.10531175737459e-06,
+      "loss": 0.4762,
+      "step": 456
+    },
+    {
+      "epoch": 0.4418660865361373,
+      "grad_norm": 3.2276439666748047,
+      "learning_rate": 3.0977037173145387e-06,
+      "loss": 0.4832,
+      "step": 457
+    },
+    {
+      "epoch": 0.4428329707517525,
+      "grad_norm": 4.06106424331665,
+      "learning_rate": 3.090089801685857e-06,
+      "loss": 0.5277,
+      "step": 458
+    },
+    {
+      "epoch": 0.4437998549673677,
+      "grad_norm": 3.59995698928833,
+      "learning_rate": 3.082470085335133e-06,
+      "loss": 0.4655,
+      "step": 459
+    },
+    {
+      "epoch": 0.4447667391829828,
+      "grad_norm": 3.5383338928222656,
+      "learning_rate": 3.074844643165979e-06,
+      "loss": 0.5158,
+      "step": 460
+    },
+    {
+      "epoch": 0.445733623398598,
+      "grad_norm": 3.6818606853485107,
+      "learning_rate": 3.0672135501382894e-06,
+      "loss": 0.3861,
+      "step": 461
+    },
+    {
+      "epoch": 0.4467005076142132,
+      "grad_norm": 3.7184290885925293,
+      "learning_rate": 3.05957688126751e-06,
+      "loss": 0.5045,
+      "step": 462
+    },
+    {
+      "epoch": 0.4476673918298284,
+      "grad_norm": 3.7670133113861084,
+      "learning_rate": 3.0519347116239e-06,
+      "loss": 0.4893,
+      "step": 463
+    },
+    {
+      "epoch": 0.44863427604544354,
+      "grad_norm": 3.8827385902404785,
+      "learning_rate": 3.04428711633179e-06,
+      "loss": 0.453,
+      "step": 464
+    },
+    {
+      "epoch": 0.44960116026105873,
+      "grad_norm": 3.7055163383483887,
+      "learning_rate": 3.036634170568847e-06,
+      "loss": 0.4931,
+      "step": 465
+    },
+    {
+      "epoch": 0.4505680444766739,
+      "grad_norm": 3.6650643348693848,
+      "learning_rate": 3.0289759495653344e-06,
+      "loss": 0.5074,
+      "step": 466
+    },
+    {
+      "epoch": 0.4515349286922891,
+      "grad_norm": 3.983785629272461,
+      "learning_rate": 3.021312528603371e-06,
+      "loss": 0.47,
+      "step": 467
+    },
+    {
+      "epoch": 0.45250181290790426,
+      "grad_norm": 3.402285099029541,
+      "learning_rate": 3.0136439830161968e-06,
+      "loss": 0.4588,
+      "step": 468
+    },
+    {
+      "epoch": 0.45346869712351945,
+      "grad_norm": 3.4030966758728027,
+      "learning_rate": 3.0059703881874232e-06,
+      "loss": 0.4406,
+      "step": 469
+    },
+    {
+      "epoch": 0.45443558133913464,
+      "grad_norm": 4.0612359046936035,
+      "learning_rate": 2.9982918195502985e-06,
+      "loss": 0.5245,
+      "step": 470
+    },
+    {
+      "epoch": 0.45540246555474984,
+      "grad_norm": 4.0308051109313965,
+      "learning_rate": 2.990608352586965e-06,
+      "loss": 0.5318,
+      "step": 471
+    },
+    {
+      "epoch": 0.456369349770365,
+      "grad_norm": 3.894338369369507,
+      "learning_rate": 2.9829200628277176e-06,
+      "loss": 0.469,
+      "step": 472
+    },
+    {
+      "epoch": 0.45733623398598017,
+      "grad_norm": 3.9288504123687744,
+      "learning_rate": 2.9752270258502593e-06,
+      "loss": 0.5062,
+      "step": 473
+    },
+    {
+      "epoch": 0.45830311820159536,
+      "grad_norm": 3.727022171020508,
+      "learning_rate": 2.9675293172789583e-06,
+      "loss": 0.4718,
+      "step": 474
+    },
+    {
+      "epoch": 0.45927000241721055,
+      "grad_norm": 3.7620606422424316,
+      "learning_rate": 2.959827012784108e-06,
+      "loss": 0.4797,
+      "step": 475
+    },
+    {
+      "epoch": 0.4602368866328257,
+      "grad_norm": 3.69411301612854,
+      "learning_rate": 2.952120188081179e-06,
+      "loss": 0.4773,
+      "step": 476
+    },
+    {
+      "epoch": 0.4612037708484409,
+      "grad_norm": 3.8497121334075928,
+      "learning_rate": 2.9444089189300783e-06,
+      "loss": 0.4913,
+      "step": 477
+    },
+    {
+      "epoch": 0.4621706550640561,
+      "grad_norm": 3.8784735202789307,
+      "learning_rate": 2.936693281134402e-06,
+      "loss": 0.5083,
+      "step": 478
+    },
+    {
+      "epoch": 0.4631375392796713,
+      "grad_norm": 3.7350783348083496,
+      "learning_rate": 2.92897335054069e-06,
+      "loss": 0.5107,
+      "step": 479
+    },
+    {
+      "epoch": 0.46410442349528647,
+      "grad_norm": 3.7133052349090576,
+      "learning_rate": 2.9212492030376815e-06,
+      "loss": 0.4692,
+      "step": 480
+    },
+    {
+      "epoch": 0.4650713077109016,
+      "grad_norm": 3.5680742263793945,
+      "learning_rate": 2.913520914555572e-06,
+      "loss": 0.421,
+      "step": 481
+    },
+    {
+      "epoch": 0.4660381919265168,
+      "grad_norm": 3.4685680866241455,
+      "learning_rate": 2.90578856106526e-06,
+      "loss": 0.4427,
+      "step": 482
+    },
+    {
+      "epoch": 0.467005076142132,
+      "grad_norm": 3.5732250213623047,
+      "learning_rate": 2.8980522185776065e-06,
+      "loss": 0.4766,
+      "step": 483
+    },
+    {
+      "epoch": 0.4679719603577472,
+      "grad_norm": 3.5667455196380615,
+      "learning_rate": 2.8903119631426835e-06,
+      "loss": 0.4288,
+      "step": 484
+    },
+    {
+      "epoch": 0.4689388445733623,
+      "grad_norm": 3.793536424636841,
+      "learning_rate": 2.882567870849029e-06,
+      "loss": 0.4478,
+      "step": 485
+    },
+    {
+      "epoch": 0.4699057287889775,
+      "grad_norm": 3.8462436199188232,
+      "learning_rate": 2.874820017822899e-06,
+      "loss": 0.5199,
+      "step": 486
+    },
+    {
+      "epoch": 0.4708726130045927,
+      "grad_norm": 3.7388360500335693,
+      "learning_rate": 2.8670684802275173e-06,
+      "loss": 0.5316,
+      "step": 487
+    },
+    {
+      "epoch": 0.4718394972202079,
+      "grad_norm": 3.991246223449707,
+      "learning_rate": 2.8593133342623276e-06,
+      "loss": 0.5396,
+      "step": 488
+    },
+    {
+      "epoch": 0.47280638143582304,
+      "grad_norm": 4.0928449630737305,
+      "learning_rate": 2.8515546561622464e-06,
+      "loss": 0.4738,
+      "step": 489
+    },
+    {
+      "epoch": 0.47377326565143824,
+      "grad_norm": 4.5484819412231445,
+      "learning_rate": 2.8437925221969097e-06,
+      "loss": 0.5258,
+      "step": 490
+    },
+    {
+      "epoch": 0.47474014986705343,
+      "grad_norm": 3.6725525856018066,
+      "learning_rate": 2.8360270086699274e-06,
+      "loss": 0.4532,
+      "step": 491
+    },
+    {
+      "epoch": 0.4757070340826686,
+      "grad_norm": 3.562164068222046,
+      "learning_rate": 2.8282581919181314e-06,
+      "loss": 0.4455,
+      "step": 492
+    },
+    {
+      "epoch": 0.47667391829828376,
+      "grad_norm": 3.7910706996917725,
+      "learning_rate": 2.820486148310822e-06,
+      "loss": 0.5149,
+      "step": 493
+    },
+    {
+      "epoch": 0.47764080251389895,
+      "grad_norm": 3.9550485610961914,
+      "learning_rate": 2.8127109542490238e-06,
+      "loss": 0.4868,
+      "step": 494
+    },
+    {
+      "epoch": 0.47860768672951415,
+      "grad_norm": 3.7173829078674316,
+      "learning_rate": 2.8049326861647303e-06,
+      "loss": 0.5135,
+      "step": 495
+    },
+    {
+      "epoch": 0.47957457094512934,
+      "grad_norm": 3.9203269481658936,
+      "learning_rate": 2.7971514205201515e-06,
+      "loss": 0.4916,
+      "step": 496
+    },
+    {
+      "epoch": 0.4805414551607445,
+      "grad_norm": 4.3923726081848145,
+      "learning_rate": 2.7893672338069666e-06,
+      "loss": 0.5507,
+      "step": 497
+    },
+    {
+      "epoch": 0.4815083393763597,
+      "grad_norm": 3.923211097717285,
+      "learning_rate": 2.7815802025455677e-06,
+      "loss": 0.4686,
+      "step": 498
+    },
+    {
+      "epoch": 0.48247522359197487,
+      "grad_norm": 3.6547012329101562,
+      "learning_rate": 2.7737904032843105e-06,
+      "loss": 0.4615,
+      "step": 499
+    },
+    {
+      "epoch": 0.48344210780759006,
+      "grad_norm": 4.033823013305664,
+      "learning_rate": 2.765997912598761e-06,
+      "loss": 0.5064,
+      "step": 500
+    },
+    {
+      "epoch": 0.4844089920232052,
+      "grad_norm": 3.4618935585021973,
+      "learning_rate": 2.7582028070909415e-06,
+      "loss": 0.4413,
+      "step": 501
+    },
+    {
+      "epoch": 0.4853758762388204,
+      "grad_norm": 3.9825124740600586,
+      "learning_rate": 2.750405163388579e-06,
+      "loss": 0.5646,
+      "step": 502
+    },
+    {
+      "epoch": 0.4863427604544356,
+      "grad_norm": 3.5702192783355713,
+      "learning_rate": 2.742605058144352e-06,
+      "loss": 0.455,
+      "step": 503
+    },
+    {
+      "epoch": 0.4873096446700508,
+      "grad_norm": 3.779356002807617,
+      "learning_rate": 2.7348025680351363e-06,
+      "loss": 0.432,
+      "step": 504
+    },
+    {
+      "epoch": 0.4882765288856659,
+      "grad_norm": 3.8731324672698975,
+      "learning_rate": 2.7269977697612515e-06,
+      "loss": 0.4606,
+      "step": 505
+    },
+    {
+      "epoch": 0.4892434131012811,
+      "grad_norm": 3.5303452014923096,
+      "learning_rate": 2.7191907400457068e-06,
+      "loss": 0.4369,
+      "step": 506
+    },
+    {
+      "epoch": 0.4902102973168963,
+      "grad_norm": 3.997253894805908,
+      "learning_rate": 2.7113815556334478e-06,
+      "loss": 0.4808,
+      "step": 507
+    },
+    {
+      "epoch": 0.4911771815325115,
+      "grad_norm": 3.6927413940429688,
+      "learning_rate": 2.7035702932906e-06,
+      "loss": 0.4606,
+      "step": 508
+    },
+    {
+      "epoch": 0.49214406574812664,
+      "grad_norm": 3.5967085361480713,
+      "learning_rate": 2.6957570298037156e-06,
+      "loss": 0.4379,
+      "step": 509
+    },
+    {
+      "epoch": 0.49311094996374183,
+      "grad_norm": 3.671698808670044,
+      "learning_rate": 2.6879418419790203e-06,
+      "loss": 0.4879,
+      "step": 510
+    },
+    {
+      "epoch": 0.494077834179357,
+      "grad_norm": 3.4737465381622314,
+      "learning_rate": 2.680124806641654e-06,
+      "loss": 0.4498,
+      "step": 511
+    },
+    {
+      "epoch": 0.4950447183949722,
+      "grad_norm": 3.8905839920043945,
+      "learning_rate": 2.6723060006349198e-06,
+      "loss": 0.5119,
+      "step": 512
+    },
+    {
+      "epoch": 0.4960116026105874,
+      "grad_norm": 3.9527182579040527,
+      "learning_rate": 2.664485500819527e-06,
+      "loss": 0.4407,
+      "step": 513
+    },
+    {
+      "epoch": 0.49697848682620255,
+      "grad_norm": 3.410557985305786,
+      "learning_rate": 2.656663384072834e-06,
+      "loss": 0.4549,
+      "step": 514
+    },
+    {
+      "epoch": 0.49794537104181774,
+      "grad_norm": 3.4379289150238037,
+      "learning_rate": 2.6488397272880943e-06,
+      "loss": 0.4566,
+      "step": 515
+    },
+    {
+      "epoch": 0.49891225525743294,
+      "grad_norm": 3.790123224258423,
+      "learning_rate": 2.641014607373702e-06,
+      "loss": 0.4753,
+      "step": 516
+    },
+    {
+      "epoch": 0.49987913947304813,
+      "grad_norm": 3.5022404193878174,
+      "learning_rate": 2.633188101252433e-06,
+      "loss": 0.4508,
+      "step": 517
+    },
+    {
+      "epoch": 0.5008460236886633,
+      "grad_norm": 3.726773738861084,
+      "learning_rate": 2.625360285860689e-06,
+      "loss": 0.4535,
+      "step": 518
+    },
+    {
+      "epoch": 0.5018129079042785,
+      "grad_norm": 4.066783905029297,
+      "learning_rate": 2.617531238147744e-06,
+      "loss": 0.4777,
+      "step": 519
+    },
+    {
+      "epoch": 0.5027797921198937,
+      "grad_norm": 3.717376470565796,
+      "learning_rate": 2.6097010350749843e-06,
+      "loss": 0.4507,
+      "step": 520
+    },
+    {
+      "epoch": 0.5037466763355088,
+      "grad_norm": 4.117121696472168,
+      "learning_rate": 2.6018697536151554e-06,
+      "loss": 0.4365,
+      "step": 521
+    },
+    {
+      "epoch": 0.504713560551124,
+      "grad_norm": 3.5613620281219482,
+      "learning_rate": 2.5940374707516014e-06,
+      "loss": 0.4911,
+      "step": 522
+    },
+    {
+      "epoch": 0.5056804447667392,
+      "grad_norm": 3.5967776775360107,
+      "learning_rate": 2.5862042634775125e-06,
+      "loss": 0.4665,
+      "step": 523
+    },
+    {
+      "epoch": 0.5066473289823543,
+      "grad_norm": 3.632263660430908,
+      "learning_rate": 2.5783702087951646e-06,
+      "loss": 0.4511,
+      "step": 524
+    },
+    {
+      "epoch": 0.5076142131979695,
+      "grad_norm": 3.460890293121338,
+      "learning_rate": 2.5705353837151655e-06,
+      "loss": 0.4014,
+      "step": 525
+    },
+    {
+      "epoch": 0.5085810974135847,
+      "grad_norm": 3.7510364055633545,
+      "learning_rate": 2.5626998652556937e-06,
+      "loss": 0.5107,
+      "step": 526
+    },
+    {
+      "epoch": 0.5095479816291999,
+      "grad_norm": 3.4469544887542725,
+      "learning_rate": 2.554863730441748e-06,
+      "loss": 0.4476,
+      "step": 527
+    },
+    {
+      "epoch": 0.5105148658448151,
+      "grad_norm": 5.02686882019043,
+      "learning_rate": 2.547027056304379e-06,
+      "loss": 0.4338,
+      "step": 528
+    },
+    {
+      "epoch": 0.5114817500604303,
+      "grad_norm": 3.9072558879852295,
+      "learning_rate": 2.5391899198799475e-06,
+      "loss": 0.4529,
+      "step": 529
+    },
+    {
+      "epoch": 0.5124486342760455,
+      "grad_norm": 4.003458499908447,
+      "learning_rate": 2.5313523982093517e-06,
+      "loss": 0.4483,
+      "step": 530
+    },
+    {
+      "epoch": 0.5134155184916607,
+      "grad_norm": 3.5714404582977295,
+      "learning_rate": 2.5235145683372813e-06,
+      "loss": 0.4508,
+      "step": 531
+    },
+    {
+      "epoch": 0.5143824027072758,
+      "grad_norm": 4.432825565338135,
+      "learning_rate": 2.5156765073114535e-06,
+      "loss": 0.4886,
+      "step": 532
+    },
+    {
+      "epoch": 0.515349286922891,
+      "grad_norm": 3.702751636505127,
+      "learning_rate": 2.507838292181858e-06,
+      "loss": 0.4578,
+      "step": 533
+    },
+    {
+      "epoch": 0.5163161711385061,
+      "grad_norm": 3.845242738723755,
+      "learning_rate": 2.5e-06,
+      "loss": 0.4034,
+      "step": 534
+    },
+    {
+      "epoch": 0.5172830553541213,
+      "grad_norm": 3.649942398071289,
+      "learning_rate": 2.4921617078181425e-06,
+      "loss": 0.4518,
+      "step": 535
+    },
+    {
+      "epoch": 0.5182499395697365,
+      "grad_norm": 3.932555913925171,
+      "learning_rate": 2.484323492688547e-06,
+      "loss": 0.4826,
+      "step": 536
+    },
+    {
+      "epoch": 0.5192168237853517,
+      "grad_norm": 3.628704071044922,
+      "learning_rate": 2.47648543166272e-06,
+      "loss": 0.4381,
+      "step": 537
+    },
+    {
+      "epoch": 0.5201837080009669,
+      "grad_norm": 3.462639093399048,
+      "learning_rate": 2.4686476017906487e-06,
+      "loss": 0.4244,
+      "step": 538
+    },
+    {
+      "epoch": 0.5211505922165821,
+      "grad_norm": 3.4104435443878174,
+      "learning_rate": 2.4608100801200533e-06,
+      "loss": 0.3825,
+      "step": 539
+    },
+    {
+      "epoch": 0.5221174764321972,
+      "grad_norm": 3.861297369003296,
+      "learning_rate": 2.452972943695621e-06,
+      "loss": 0.4183,
+      "step": 540
+    },
+    {
+      "epoch": 0.5230843606478124,
+      "grad_norm": 3.5461347103118896,
+      "learning_rate": 2.445136269558254e-06,
+      "loss": 0.4373,
+      "step": 541
+    },
+    {
+      "epoch": 0.5240512448634276,
+      "grad_norm": 3.585427761077881,
+      "learning_rate": 2.4373001347443067e-06,
+      "loss": 0.4556,
+      "step": 542
+    },
+    {
+      "epoch": 0.5250181290790428,
+      "grad_norm": 3.582169771194458,
+      "learning_rate": 2.4294646162848353e-06,
+      "loss": 0.4544,
+      "step": 543
+    },
+    {
+      "epoch": 0.525985013294658,
+      "grad_norm": 3.5394082069396973,
+      "learning_rate": 2.4216297912048354e-06,
+      "loss": 0.4314,
+      "step": 544
+    },
+    {
+      "epoch": 0.5269518975102732,
+      "grad_norm": 3.7783658504486084,
+      "learning_rate": 2.413795736522489e-06,
+      "loss": 0.4848,
+      "step": 545
+    },
+    {
+      "epoch": 0.5279187817258884,
+      "grad_norm": 3.5096867084503174,
+      "learning_rate": 2.4059625292483994e-06,
+      "loss": 0.4411,
+      "step": 546
+    },
+    {
+      "epoch": 0.5288856659415035,
+      "grad_norm": 3.744575023651123,
+      "learning_rate": 2.3981302463848454e-06,
+      "loss": 0.4997,
+      "step": 547
+    },
+    {
+      "epoch": 0.5298525501571187,
+      "grad_norm": 3.5636816024780273,
+      "learning_rate": 2.390298964925016e-06,
+      "loss": 0.4279,
+      "step": 548
+    },
+    {
+      "epoch": 0.5308194343727338,
+      "grad_norm": 3.968400716781616,
+      "learning_rate": 2.3824687618522567e-06,
+      "loss": 0.4238,
+      "step": 549
+    },
+    {
+      "epoch": 0.531786318588349,
+      "grad_norm": 3.7942044734954834,
+      "learning_rate": 2.374639714139312e-06,
+      "loss": 0.4884,
+      "step": 550
+    },
+    {
+      "epoch": 0.5327532028039642,
+      "grad_norm": 3.705420732498169,
+      "learning_rate": 2.366811898747568e-06,
+      "loss": 0.4936,
+      "step": 551
+    },
+    {
+      "epoch": 0.5337200870195794,
+      "grad_norm": 3.5354163646698,
+      "learning_rate": 2.358985392626298e-06,
+      "loss": 0.4697,
+      "step": 552
+    },
+    {
+      "epoch": 0.5346869712351946,
+      "grad_norm": 3.759593963623047,
+      "learning_rate": 2.351160272711907e-06,
+      "loss": 0.4631,
+      "step": 553
+    },
+    {
+      "epoch": 0.5356538554508098,
+      "grad_norm": 3.913259506225586,
+      "learning_rate": 2.343336615927167e-06,
+      "loss": 0.471,
+      "step": 554
+    },
+    {
+      "epoch": 0.536620739666425,
+      "grad_norm": 4.187595367431641,
+      "learning_rate": 2.3355144991804736e-06,
+      "loss": 0.5586,
+      "step": 555
+    },
+    {
+      "epoch": 0.5375876238820402,
+      "grad_norm": 3.8247408866882324,
+      "learning_rate": 2.3276939993650806e-06,
+      "loss": 0.5067,
+      "step": 556
+    },
+    {
+      "epoch": 0.5385545080976553,
+      "grad_norm": 3.486445665359497,
+      "learning_rate": 2.3198751933583463e-06,
+      "loss": 0.4399,
+      "step": 557
+    },
+    {
+      "epoch": 0.5395213923132705,
+      "grad_norm": 3.785252571105957,
+      "learning_rate": 2.312058158020981e-06,
+      "loss": 0.4527,
+      "step": 558
+    },
+    {
+      "epoch": 0.5404882765288856,
+      "grad_norm": 3.7378010749816895,
+      "learning_rate": 2.304242970196285e-06,
+      "loss": 0.458,
+      "step": 559
+    },
+    {
+      "epoch": 0.5414551607445008,
+      "grad_norm": 3.6285452842712402,
+      "learning_rate": 2.2964297067094004e-06,
+      "loss": 0.4403,
+      "step": 560
+    },
+    {
+      "epoch": 0.542422044960116,
+      "grad_norm": 3.6113779544830322,
+      "learning_rate": 2.2886184443665522e-06,
+      "loss": 0.41,
+      "step": 561
+    },
+    {
+      "epoch": 0.5433889291757312,
+      "grad_norm": 3.54592227935791,
+      "learning_rate": 2.2808092599542936e-06,
+      "loss": 0.4754,
+      "step": 562
+    },
+    {
+      "epoch": 0.5443558133913464,
+      "grad_norm": 3.5506227016448975,
+      "learning_rate": 2.2730022302387493e-06,
+      "loss": 0.4601,
+      "step": 563
+    },
+    {
+      "epoch": 0.5453226976069616,
+      "grad_norm": 3.6648545265197754,
+      "learning_rate": 2.265197431964864e-06,
+      "loss": 0.4528,
+      "step": 564
+    },
+    {
+      "epoch": 0.5462895818225767,
+      "grad_norm": 3.643314838409424,
+      "learning_rate": 2.257394941855648e-06,
+      "loss": 0.4552,
+      "step": 565
+    },
+    {
+      "epoch": 0.5472564660381919,
+      "grad_norm": 3.9885194301605225,
+      "learning_rate": 2.249594836611422e-06,
+      "loss": 0.4111,
+      "step": 566
+    },
+    {
+      "epoch": 0.5482233502538071,
+      "grad_norm": 3.288057804107666,
+      "learning_rate": 2.2417971929090593e-06,
+      "loss": 0.3875,
+      "step": 567
+    },
+    {
+      "epoch": 0.5491902344694223,
+      "grad_norm": 4.048732280731201,
+      "learning_rate": 2.2340020874012393e-06,
+      "loss": 0.5485,
+      "step": 568
+    },
+    {
+      "epoch": 0.5501571186850375,
+      "grad_norm": 3.265153646469116,
+      "learning_rate": 2.2262095967156895e-06,
+      "loss": 0.4145,
+      "step": 569
+    },
+    {
+      "epoch": 0.5511240029006527,
+      "grad_norm": 3.73302960395813,
+      "learning_rate": 2.218419797454433e-06,
+      "loss": 0.5199,
+      "step": 570
+    },
+    {
+      "epoch": 0.5520908871162679,
+      "grad_norm": 3.5953993797302246,
+      "learning_rate": 2.2106327661930343e-06,
+      "loss": 0.4172,
+      "step": 571
+    },
+    {
+      "epoch": 0.553057771331883,
+      "grad_norm": 3.707054853439331,
+      "learning_rate": 2.202848579479849e-06,
+      "loss": 0.4282,
+      "step": 572
+    },
+    {
+      "epoch": 0.5540246555474981,
+      "grad_norm": 3.5350615978240967,
+      "learning_rate": 2.19506731383527e-06,
+      "loss": 0.4952,
+      "step": 573
+    },
+    {
+      "epoch": 0.5549915397631133,
+      "grad_norm": 4.000664234161377,
+      "learning_rate": 2.187289045750977e-06,
+      "loss": 0.4619,
+      "step": 574
+    },
+    {
+      "epoch": 0.5559584239787285,
+      "grad_norm": 4.087167739868164,
+      "learning_rate": 2.1795138516891786e-06,
+      "loss": 0.5084,
+      "step": 575
+    },
+    {
+      "epoch": 0.5569253081943437,
+      "grad_norm": 3.46340012550354,
+      "learning_rate": 2.17174180808187e-06,
+      "loss": 0.4312,
+      "step": 576
+    },
+    {
+      "epoch": 0.5578921924099589,
+      "grad_norm": 3.5368056297302246,
+      "learning_rate": 2.163972991330073e-06,
+      "loss": 0.5035,
+      "step": 577
+    },
+    {
+      "epoch": 0.5588590766255741,
+      "grad_norm": 3.7972068786621094,
+      "learning_rate": 2.1562074778030907e-06,
+      "loss": 0.4531,
+      "step": 578
+    },
+    {
+      "epoch": 0.5598259608411893,
+      "grad_norm": 3.9204769134521484,
+      "learning_rate": 2.148445343837755e-06,
+      "loss": 0.4403,
+      "step": 579
+    },
+    {
+      "epoch": 0.5607928450568045,
+      "grad_norm": 3.8079707622528076,
+      "learning_rate": 2.1406866657376733e-06,
+      "loss": 0.4951,
+      "step": 580
+    },
+    {
+      "epoch": 0.5617597292724196,
+      "grad_norm": 3.6291754245758057,
+      "learning_rate": 2.1329315197724835e-06,
+      "loss": 0.4237,
+      "step": 581
+    },
+    {
+      "epoch": 0.5627266134880348,
+      "grad_norm": 3.552342414855957,
+      "learning_rate": 2.1251799821771015e-06,
+      "loss": 0.4063,
+      "step": 582
+    },
+    {
+      "epoch": 0.56369349770365,
+      "grad_norm": 3.4691274166107178,
+      "learning_rate": 2.1174321291509716e-06,
+      "loss": 0.4758,
+      "step": 583
+    },
+    {
+      "epoch": 0.5646603819192652,
+      "grad_norm": 3.760558605194092,
+      "learning_rate": 2.1096880368573173e-06,
+      "loss": 0.5106,
+      "step": 584
+    },
+    {
+      "epoch": 0.5656272661348803,
+      "grad_norm": 3.534207582473755,
+      "learning_rate": 2.1019477814223943e-06,
+      "loss": 0.5242,
+      "step": 585
+    },
+    {
+      "epoch": 0.5665941503504955,
+      "grad_norm": 4.03209114074707,
+      "learning_rate": 2.0942114389347402e-06,
+      "loss": 0.4693,
+      "step": 586
+    },
+    {
+      "epoch": 0.5675610345661107,
+      "grad_norm": 3.6348366737365723,
+      "learning_rate": 2.086479085444429e-06,
+      "loss": 0.5009,
+      "step": 587
+    },
+    {
+      "epoch": 0.5685279187817259,
+      "grad_norm": 3.4231724739074707,
+      "learning_rate": 2.078750796962319e-06,
+      "loss": 0.457,
+      "step": 588
+    },
+    {
+      "epoch": 0.5694948029973411,
+      "grad_norm": 3.9665541648864746,
+      "learning_rate": 2.071026649459311e-06,
+      "loss": 0.4962,
+      "step": 589
+    },
+    {
+      "epoch": 0.5704616872129562,
+      "grad_norm": 3.6691770553588867,
+      "learning_rate": 2.0633067188655985e-06,
+      "loss": 0.4548,
+      "step": 590
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 3.683633804321289,
+      "learning_rate": 2.055591081069922e-06,
+      "loss": 0.4559,
+      "step": 591
+    },
+    {
+      "epoch": 0.5723954556441866,
+      "grad_norm": 3.508237600326538,
+      "learning_rate": 2.0478798119188216e-06,
+      "loss": 0.4858,
+      "step": 592
+    },
+    {
+      "epoch": 0.5733623398598018,
+      "grad_norm": 3.7423527240753174,
+      "learning_rate": 2.040172987215893e-06,
+      "loss": 0.4467,
+      "step": 593
+    },
+    {
+      "epoch": 0.574329224075417,
+      "grad_norm": 2.975794792175293,
+      "learning_rate": 2.032470682721042e-06,
+      "loss": 0.4223,
+      "step": 594
+    },
+    {
+      "epoch": 0.5752961082910322,
+      "grad_norm": 3.9564762115478516,
+      "learning_rate": 2.024772974149741e-06,
+      "loss": 0.5339,
+      "step": 595
+    },
+    {
+      "epoch": 0.5762629925066474,
+      "grad_norm": 3.652500629425049,
+      "learning_rate": 2.0170799371722828e-06,
+      "loss": 0.4471,
+      "step": 596
+    },
+    {
+      "epoch": 0.5772298767222626,
+      "grad_norm": 3.6535003185272217,
+      "learning_rate": 2.0093916474130354e-06,
+      "loss": 0.5303,
+      "step": 597
+    },
+    {
+      "epoch": 0.5781967609378776,
+      "grad_norm": 3.983417510986328,
+      "learning_rate": 2.0017081804497023e-06,
+      "loss": 0.4422,
+      "step": 598
+    },
+    {
+      "epoch": 0.5791636451534928,
+      "grad_norm": 3.6490795612335205,
+      "learning_rate": 1.9940296118125776e-06,
+      "loss": 0.4801,
+      "step": 599
+    },
+    {
+      "epoch": 0.580130529369108,
+      "grad_norm": 3.816997528076172,
+      "learning_rate": 1.986356016983804e-06,
+      "loss": 0.45,
+      "step": 600
+    },
+    {
+      "epoch": 0.5810974135847232,
+      "grad_norm": 3.5687811374664307,
+      "learning_rate": 1.9786874713966293e-06,
+      "loss": 0.5209,
+      "step": 601
+    },
+    {
+      "epoch": 0.5820642978003384,
+      "grad_norm": 3.970633029937744,
+      "learning_rate": 1.9710240504346664e-06,
+      "loss": 0.5093,
+      "step": 602
+    },
+    {
+      "epoch": 0.5830311820159536,
+      "grad_norm": 3.7453064918518066,
+      "learning_rate": 1.9633658294311535e-06,
+      "loss": 0.4749,
+      "step": 603
+    },
+    {
+      "epoch": 0.5839980662315688,
+      "grad_norm": 3.2988827228546143,
+      "learning_rate": 1.9557128836682107e-06,
+      "loss": 0.3784,
+      "step": 604
+    },
+    {
+      "epoch": 0.584964950447184,
+      "grad_norm": 3.251164674758911,
+      "learning_rate": 1.9480652883761007e-06,
+      "loss": 0.4132,
+      "step": 605
+    },
+    {
+      "epoch": 0.5859318346627991,
+      "grad_norm": 3.453906297683716,
+      "learning_rate": 1.9404231187324903e-06,
+      "loss": 0.4589,
+      "step": 606
+    },
+    {
+      "epoch": 0.5868987188784143,
+      "grad_norm": 3.687375783920288,
+      "learning_rate": 1.9327864498617114e-06,
+      "loss": 0.5179,
+      "step": 607
+    },
+    {
+      "epoch": 0.5878656030940295,
+      "grad_norm": 3.5259013175964355,
+      "learning_rate": 1.925155356834022e-06,
+      "loss": 0.4519,
+      "step": 608
+    },
+    {
+      "epoch": 0.5888324873096447,
+      "grad_norm": 3.542691469192505,
+      "learning_rate": 1.9175299146648672e-06,
+      "loss": 0.4736,
+      "step": 609
+    },
+    {
+      "epoch": 0.5897993715252599,
+      "grad_norm": 3.4081952571868896,
+      "learning_rate": 1.9099101983141434e-06,
+      "loss": 0.4722,
+      "step": 610
+    },
+    {
+      "epoch": 0.590766255740875,
+      "grad_norm": 3.5784544944763184,
+      "learning_rate": 1.9022962826854619e-06,
+      "loss": 0.5243,
+      "step": 611
+    },
+    {
+      "epoch": 0.5917331399564902,
+      "grad_norm": 3.597616672515869,
+      "learning_rate": 1.8946882426254103e-06,
+      "loss": 0.4564,
+      "step": 612
+    },
+    {
+      "epoch": 0.5927000241721054,
+      "grad_norm": 3.7966713905334473,
+      "learning_rate": 1.887086152922818e-06,
+      "loss": 0.5083,
+      "step": 613
+    },
+    {
+      "epoch": 0.5936669083877205,
+      "grad_norm": 3.926304817199707,
+      "learning_rate": 1.879490088308021e-06,
+      "loss": 0.4445,
+      "step": 614
+    },
+    {
+      "epoch": 0.5946337926033357,
+      "grad_norm": 3.6007909774780273,
+      "learning_rate": 1.8719001234521283e-06,
+      "loss": 0.4761,
+      "step": 615
+    },
+    {
+      "epoch": 0.5956006768189509,
+      "grad_norm": 3.7344369888305664,
+      "learning_rate": 1.8643163329662844e-06,
+      "loss": 0.4451,
+      "step": 616
+    },
+    {
+      "epoch": 0.5965675610345661,
+      "grad_norm": 3.4734723567962646,
+      "learning_rate": 1.8567387914009432e-06,
+      "loss": 0.4302,
+      "step": 617
+    },
+    {
+      "epoch": 0.5975344452501813,
+      "grad_norm": 3.6221048831939697,
+      "learning_rate": 1.849167573245123e-06,
+      "loss": 0.4603,
+      "step": 618
+    },
+    {
+      "epoch": 0.5985013294657965,
+      "grad_norm": 3.7050702571868896,
+      "learning_rate": 1.8416027529256885e-06,
+      "loss": 0.4371,
+      "step": 619
+    },
+    {
+      "epoch": 0.5994682136814117,
+      "grad_norm": 3.6579859256744385,
+      "learning_rate": 1.8340444048066075e-06,
+      "loss": 0.4887,
+      "step": 620
+    },
+    {
+      "epoch": 0.6004350978970269,
+      "grad_norm": 3.98431658744812,
+      "learning_rate": 1.8264926031882274e-06,
+      "loss": 0.474,
+      "step": 621
+    },
+    {
+      "epoch": 0.6014019821126421,
+      "grad_norm": 3.3613712787628174,
+      "learning_rate": 1.8189474223065392e-06,
+      "loss": 0.4559,
+      "step": 622
+    },
+    {
+      "epoch": 0.6023688663282571,
+      "grad_norm": 3.6685123443603516,
+      "learning_rate": 1.8114089363324525e-06,
+      "loss": 0.3974,
+      "step": 623
+    },
+    {
+      "epoch": 0.6033357505438723,
+      "grad_norm": 3.8673484325408936,
+      "learning_rate": 1.8038772193710646e-06,
+      "loss": 0.4535,
+      "step": 624
+    },
+    {
+      "epoch": 0.6043026347594875,
+      "grad_norm": 3.636211395263672,
+      "learning_rate": 1.7963523454609317e-06,
+      "loss": 0.482,
+      "step": 625
+    },
+    {
+      "epoch": 0.6052695189751027,
+      "grad_norm": 3.6469950675964355,
+      "learning_rate": 1.7888343885733381e-06,
+      "loss": 0.4359,
+      "step": 626
+    },
+    {
+      "epoch": 0.6062364031907179,
+      "grad_norm": 3.870375633239746,
+      "learning_rate": 1.7813234226115767e-06,
+      "loss": 0.4971,
+      "step": 627
+    },
+    {
+      "epoch": 0.6072032874063331,
+      "grad_norm": 3.4417426586151123,
+      "learning_rate": 1.773819521410215e-06,
+      "loss": 0.4375,
+      "step": 628
+    },
+    {
+      "epoch": 0.6081701716219483,
+      "grad_norm": 3.7465872764587402,
+      "learning_rate": 1.766322758734374e-06,
+      "loss": 0.4631,
+      "step": 629
+    },
+    {
+      "epoch": 0.6091370558375635,
+      "grad_norm": 3.620347499847412,
+      "learning_rate": 1.7588332082789995e-06,
+      "loss": 0.4412,
+      "step": 630
+    },
+    {
+      "epoch": 0.6101039400531786,
+      "grad_norm": 3.565495014190674,
+      "learning_rate": 1.75135094366814e-06,
+      "loss": 0.4535,
+      "step": 631
+    },
+    {
+      "epoch": 0.6110708242687938,
+      "grad_norm": 3.6394903659820557,
+      "learning_rate": 1.7438760384542242e-06,
+      "loss": 0.4759,
+      "step": 632
+    },
+    {
+      "epoch": 0.612037708484409,
+      "grad_norm": 4.002828598022461,
+      "learning_rate": 1.7364085661173346e-06,
+      "loss": 0.4823,
+      "step": 633
+    },
+    {
+      "epoch": 0.6130045927000242,
+      "grad_norm": 3.7082462310791016,
+      "learning_rate": 1.7289486000644857e-06,
+      "loss": 0.4105,
+      "step": 634
+    },
+    {
+      "epoch": 0.6139714769156394,
+      "grad_norm": 4.062910079956055,
+      "learning_rate": 1.721496213628906e-06,
+      "loss": 0.4587,
+      "step": 635
+    },
+    {
+      "epoch": 0.6149383611312546,
+      "grad_norm": 3.607834577560425,
+      "learning_rate": 1.7140514800693125e-06,
+      "loss": 0.4679,
+      "step": 636
+    },
+    {
+      "epoch": 0.6159052453468697,
+      "grad_norm": 3.3181495666503906,
+      "learning_rate": 1.7066144725691933e-06,
+      "loss": 0.4362,
+      "step": 637
+    },
+    {
+      "epoch": 0.6168721295624849,
+      "grad_norm": 4.290891170501709,
+      "learning_rate": 1.6991852642360898e-06,
+      "loss": 0.4636,
+      "step": 638
+    },
+    {
+      "epoch": 0.6178390137781,
+      "grad_norm": 3.491584539413452,
+      "learning_rate": 1.6917639281008703e-06,
+      "loss": 0.4388,
+      "step": 639
+    },
+    {
+      "epoch": 0.6188058979937152,
+      "grad_norm": 3.530135154724121,
+      "learning_rate": 1.6843505371170226e-06,
+      "loss": 0.4667,
+      "step": 640
+    },
+    {
+      "epoch": 0.6197727822093304,
+      "grad_norm": 3.9121077060699463,
+      "learning_rate": 1.6769451641599305e-06,
+      "loss": 0.5047,
+      "step": 641
+    },
+    {
+      "epoch": 0.6207396664249456,
+      "grad_norm": 3.8880884647369385,
+      "learning_rate": 1.6695478820261573e-06,
+      "loss": 0.4943,
+      "step": 642
+    },
+    {
+      "epoch": 0.6217065506405608,
+      "grad_norm": 3.671125888824463,
+      "learning_rate": 1.6621587634327328e-06,
+      "loss": 0.5352,
+      "step": 643
+    },
+    {
+      "epoch": 0.622673434856176,
+      "grad_norm": 3.2758569717407227,
+      "learning_rate": 1.6547778810164366e-06,
+      "loss": 0.4389,
+      "step": 644
+    },
+    {
+      "epoch": 0.6236403190717912,
+      "grad_norm": 4.217796325683594,
+      "learning_rate": 1.647405307333085e-06,
+      "loss": 0.5478,
+      "step": 645
+    },
+    {
+      "epoch": 0.6246072032874064,
+      "grad_norm": 3.5154519081115723,
+      "learning_rate": 1.6400411148568196e-06,
+      "loss": 0.394,
+      "step": 646
+    },
+    {
+      "epoch": 0.6255740875030215,
+      "grad_norm": 3.519857883453369,
+      "learning_rate": 1.6326853759793878e-06,
+      "loss": 0.4146,
+      "step": 647
+    },
+    {
+      "epoch": 0.6265409717186367,
+      "grad_norm": 3.5776000022888184,
+      "learning_rate": 1.6253381630094412e-06,
+      "loss": 0.4827,
+      "step": 648
+    },
+    {
+      "epoch": 0.6275078559342518,
+      "grad_norm": 4.16077995300293,
+      "learning_rate": 1.6179995481718165e-06,
+      "loss": 0.5614,
+      "step": 649
+    },
+    {
+      "epoch": 0.628474740149867,
+      "grad_norm": 3.562154769897461,
+      "learning_rate": 1.610669603606831e-06,
+      "loss": 0.4778,
+      "step": 650
+    },
+    {
+      "epoch": 0.6294416243654822,
+      "grad_norm": 3.7697620391845703,
+      "learning_rate": 1.6033484013695688e-06,
+      "loss": 0.4292,
+      "step": 651
+    },
+    {
+      "epoch": 0.6304085085810974,
+      "grad_norm": 3.4476921558380127,
+      "learning_rate": 1.5960360134291759e-06,
+      "loss": 0.4513,
+      "step": 652
+    },
+    {
+      "epoch": 0.6313753927967126,
+      "grad_norm": 3.656102180480957,
+      "learning_rate": 1.588732511668153e-06,
+      "loss": 0.4747,
+      "step": 653
+    },
+    {
+      "epoch": 0.6323422770123278,
+      "grad_norm": 3.675884962081909,
+      "learning_rate": 1.5814379678816471e-06,
+      "loss": 0.4267,
+      "step": 654
+    },
+    {
+      "epoch": 0.633309161227943,
+      "grad_norm": 3.7877371311187744,
+      "learning_rate": 1.5741524537767427e-06,
+      "loss": 0.4905,
+      "step": 655
+    },
+    {
+      "epoch": 0.6342760454435581,
+      "grad_norm": 3.5091817378997803,
+      "learning_rate": 1.566876040971765e-06,
+      "loss": 0.4275,
+      "step": 656
+    },
+    {
+      "epoch": 0.6352429296591733,
+      "grad_norm": 3.42677903175354,
+      "learning_rate": 1.5596088009955695e-06,
+      "loss": 0.4384,
+      "step": 657
+    },
+    {
+      "epoch": 0.6362098138747885,
+      "grad_norm": 3.765990972518921,
+      "learning_rate": 1.5523508052868402e-06,
+      "loss": 0.4993,
+      "step": 658
+    },
+    {
+      "epoch": 0.6371766980904037,
+      "grad_norm": 3.590433120727539,
+      "learning_rate": 1.5451021251933895e-06,
+      "loss": 0.432,
+      "step": 659
+    },
+    {
+      "epoch": 0.6381435823060189,
+      "grad_norm": 3.3445186614990234,
+      "learning_rate": 1.5378628319714514e-06,
+      "loss": 0.4175,
+      "step": 660
+    },
+    {
+      "epoch": 0.6391104665216341,
+      "grad_norm": 3.357813596725464,
+      "learning_rate": 1.5306329967849887e-06,
+      "loss": 0.412,
+      "step": 661
+    },
+    {
+      "epoch": 0.6400773507372493,
+      "grad_norm": 4.022750377655029,
+      "learning_rate": 1.5234126907049878e-06,
+      "loss": 0.3955,
+      "step": 662
+    },
+    {
+      "epoch": 0.6410442349528644,
+      "grad_norm": 3.672391414642334,
+      "learning_rate": 1.5162019847087616e-06,
+      "loss": 0.4514,
+      "step": 663
+    },
+    {
+      "epoch": 0.6420111191684795,
+      "grad_norm": 4.018763065338135,
+      "learning_rate": 1.509000949679252e-06,
+      "loss": 0.4605,
+      "step": 664
+    },
+    {
+      "epoch": 0.6429780033840947,
+      "grad_norm": 3.6761765480041504,
+      "learning_rate": 1.5018096564043333e-06,
+      "loss": 0.4994,
+      "step": 665
+    },
+    {
+      "epoch": 0.6439448875997099,
+      "grad_norm": 3.574371576309204,
+      "learning_rate": 1.4946281755761153e-06,
+      "loss": 0.4411,
+      "step": 666
+    },
+    {
+      "epoch": 0.6449117718153251,
+      "grad_norm": 3.484593629837036,
+      "learning_rate": 1.4874565777902518e-06,
+      "loss": 0.4839,
+      "step": 667
+    },
+    {
+      "epoch": 0.6458786560309403,
+      "grad_norm": 3.7389230728149414,
+      "learning_rate": 1.4802949335452387e-06,
+      "loss": 0.4805,
+      "step": 668
+    },
+    {
+      "epoch": 0.6468455402465555,
+      "grad_norm": 3.800123691558838,
+      "learning_rate": 1.4731433132417316e-06,
+      "loss": 0.4202,
+      "step": 669
+    },
+    {
+      "epoch": 0.6478124244621707,
+      "grad_norm": 4.061914920806885,
+      "learning_rate": 1.466001787181846e-06,
+      "loss": 0.4681,
+      "step": 670
+    },
+    {
+      "epoch": 0.6487793086777859,
+      "grad_norm": 3.831731081008911,
+      "learning_rate": 1.4588704255684697e-06,
+      "loss": 0.4506,
+      "step": 671
+    },
+    {
+      "epoch": 0.649746192893401,
+      "grad_norm": 3.3978049755096436,
+      "learning_rate": 1.451749298504568e-06,
+      "loss": 0.4754,
+      "step": 672
+    },
+    {
+      "epoch": 0.6507130771090162,
+      "grad_norm": 3.271028518676758,
+      "learning_rate": 1.4446384759925024e-06,
+      "loss": 0.4375,
+      "step": 673
+    },
+    {
+      "epoch": 0.6516799613246314,
+      "grad_norm": 3.7137019634246826,
+      "learning_rate": 1.4375380279333378e-06,
+      "loss": 0.5393,
+      "step": 674
+    },
+    {
+      "epoch": 0.6526468455402465,
+      "grad_norm": 3.694547176361084,
+      "learning_rate": 1.4304480241261529e-06,
+      "loss": 0.5164,
+      "step": 675
+    },
+    {
+      "epoch": 0.6536137297558617,
+      "grad_norm": 3.5363903045654297,
+      "learning_rate": 1.4233685342673602e-06,
+      "loss": 0.447,
+      "step": 676
+    },
+    {
+      "epoch": 0.6545806139714769,
+      "grad_norm": 3.400479555130005,
+      "learning_rate": 1.4162996279500158e-06,
+      "loss": 0.4502,
+      "step": 677
+    },
+    {
+      "epoch": 0.6555474981870921,
+      "grad_norm": 3.27048397064209,
+      "learning_rate": 1.409241374663136e-06,
+      "loss": 0.4497,
+      "step": 678
+    },
+    {
+      "epoch": 0.6565143824027073,
+      "grad_norm": 3.7610373497009277,
+      "learning_rate": 1.4021938437910181e-06,
+      "loss": 0.4564,
+      "step": 679
+    },
+    {
+      "epoch": 0.6574812666183224,
+      "grad_norm": 3.84147572517395,
+      "learning_rate": 1.3951571046125546e-06,
+      "loss": 0.493,
+      "step": 680
+    },
+    {
+      "epoch": 0.6584481508339376,
+      "grad_norm": 3.6611154079437256,
+      "learning_rate": 1.388131226300552e-06,
+      "loss": 0.3995,
+      "step": 681
+    },
+    {
+      "epoch": 0.6594150350495528,
+      "grad_norm": 3.796962261199951,
+      "learning_rate": 1.3811162779210512e-06,
+      "loss": 0.5106,
+      "step": 682
+    },
+    {
+      "epoch": 0.660381919265168,
+      "grad_norm": 3.6814441680908203,
+      "learning_rate": 1.374112328432652e-06,
+      "loss": 0.522,
+      "step": 683
+    },
+    {
+      "epoch": 0.6613488034807832,
+      "grad_norm": 3.965012788772583,
+      "learning_rate": 1.3671194466858335e-06,
+      "loss": 0.4978,
+      "step": 684
+    },
+    {
+      "epoch": 0.6623156876963984,
+      "grad_norm": 3.3158814907073975,
+      "learning_rate": 1.3601377014222688e-06,
+      "loss": 0.4425,
+      "step": 685
+    },
+    {
+      "epoch": 0.6632825719120136,
+      "grad_norm": 3.8214852809906006,
+      "learning_rate": 1.3531671612741653e-06,
+      "loss": 0.5121,
+      "step": 686
+    },
+    {
+      "epoch": 0.6642494561276288,
+      "grad_norm": 3.924567222595215,
+      "learning_rate": 1.3462078947635781e-06,
+      "loss": 0.5657,
+      "step": 687
+    },
+    {
+      "epoch": 0.665216340343244,
+      "grad_norm": 3.8759005069732666,
+      "learning_rate": 1.3392599703017393e-06,
+      "loss": 0.4754,
+      "step": 688
+    },
+    {
+      "epoch": 0.666183224558859,
+      "grad_norm": 3.805741548538208,
+      "learning_rate": 1.3323234561883847e-06,
+      "loss": 0.498,
+      "step": 689
+    },
+    {
+      "epoch": 0.6671501087744742,
+      "grad_norm": 3.289720296859741,
+      "learning_rate": 1.3253984206110881e-06,
+      "loss": 0.4543,
+      "step": 690
+    },
+    {
+      "epoch": 0.6681169929900894,
+      "grad_norm": 3.9015278816223145,
+      "learning_rate": 1.318484931644582e-06,
+      "loss": 0.5214,
+      "step": 691
+    },
+    {
+      "epoch": 0.6690838772057046,
+      "grad_norm": 3.7777562141418457,
+      "learning_rate": 1.311583057250096e-06,
+      "loss": 0.4999,
+      "step": 692
+    },
+    {
+      "epoch": 0.6700507614213198,
+      "grad_norm": 3.9760212898254395,
+      "learning_rate": 1.3046928652746833e-06,
+      "loss": 0.5266,
+      "step": 693
+    },
+    {
+      "epoch": 0.671017645636935,
+      "grad_norm": 4.030066967010498,
+      "learning_rate": 1.297814423450559e-06,
+      "loss": 0.4069,
+      "step": 694
+    },
+    {
+      "epoch": 0.6719845298525502,
+      "grad_norm": 3.4443728923797607,
+      "learning_rate": 1.2909477993944286e-06,
+      "loss": 0.4446,
+      "step": 695
+    },
+    {
+      "epoch": 0.6729514140681654,
+      "grad_norm": 3.7695109844207764,
+      "learning_rate": 1.284093060606829e-06,
+      "loss": 0.514,
+      "step": 696
+    },
+    {
+      "epoch": 0.6739182982837805,
+      "grad_norm": 3.507645606994629,
+      "learning_rate": 1.2772502744714592e-06,
+      "loss": 0.4932,
+      "step": 697
+    },
+    {
+      "epoch": 0.6748851824993957,
+      "grad_norm": 3.8569910526275635,
+      "learning_rate": 1.2704195082545217e-06,
+      "loss": 0.4999,
+      "step": 698
+    },
+    {
+      "epoch": 0.6758520667150109,
+      "grad_norm": 3.4546616077423096,
+      "learning_rate": 1.2636008291040618e-06,
+      "loss": 0.3947,
+      "step": 699
+    },
+    {
+      "epoch": 0.676818950930626,
+      "grad_norm": 3.5067853927612305,
+      "learning_rate": 1.2567943040493059e-06,
+      "loss": 0.4806,
+      "step": 700
+    },
+    {
+      "epoch": 0.6777858351462412,
+      "grad_norm": 3.6664950847625732,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.4407,
+      "step": 701
+    },
+    {
+      "epoch": 0.6787527193618564,
+      "grad_norm": 3.515747308731079,
+      "learning_rate": 1.243217983745758e-06,
+      "loss": 0.4741,
+      "step": 702
+    },
+    {
+      "epoch": 0.6797196035774716,
+      "grad_norm": 3.3077144622802734,
+      "learning_rate": 1.236448321955401e-06,
+      "loss": 0.4433,
+      "step": 703
+    },
+    {
+      "epoch": 0.6806864877930868,
+      "grad_norm": 3.554198741912842,
+      "learning_rate": 1.2296910811763008e-06,
+      "loss": 0.4852,
+      "step": 704
+    },
+    {
+      "epoch": 0.6816533720087019,
+      "grad_norm": 3.428964138031006,
+      "learning_rate": 1.222946327833731e-06,
+      "loss": 0.436,
+      "step": 705
+    },
+    {
+      "epoch": 0.6826202562243171,
+      "grad_norm": 3.696312665939331,
+      "learning_rate": 1.2162141282302078e-06,
+      "loss": 0.4375,
+      "step": 706
+    },
+    {
+      "epoch": 0.6835871404399323,
+      "grad_norm": 3.7810776233673096,
+      "learning_rate": 1.2094945485448424e-06,
+      "loss": 0.5128,
+      "step": 707
+    },
+    {
+      "epoch": 0.6845540246555475,
+      "grad_norm": 3.7111852169036865,
+      "learning_rate": 1.2027876548326898e-06,
+      "loss": 0.4304,
+      "step": 708
+    },
+    {
+      "epoch": 0.6855209088711627,
+      "grad_norm": 3.6479649543762207,
+      "learning_rate": 1.196093513024099e-06,
+      "loss": 0.5147,
+      "step": 709
+    },
+    {
+      "epoch": 0.6864877930867779,
+      "grad_norm": 3.8986546993255615,
+      "learning_rate": 1.1894121889240638e-06,
+      "loss": 0.4977,
+      "step": 710
+    },
+    {
+      "epoch": 0.6874546773023931,
+      "grad_norm": 3.825770378112793,
+      "learning_rate": 1.182743748211576e-06,
+      "loss": 0.3706,
+      "step": 711
+    },
+    {
+      "epoch": 0.6884215615180083,
+      "grad_norm": 3.962735891342163,
+      "learning_rate": 1.1760882564389823e-06,
+      "loss": 0.4488,
+      "step": 712
+    },
+    {
+      "epoch": 0.6893884457336233,
+      "grad_norm": 3.5851974487304688,
+      "learning_rate": 1.1694457790313403e-06,
+      "loss": 0.4938,
+      "step": 713
+    },
+    {
+      "epoch": 0.6903553299492385,
+      "grad_norm": 3.4636547565460205,
+      "learning_rate": 1.1628163812857672e-06,
+      "loss": 0.4451,
+      "step": 714
+    },
+    {
+      "epoch": 0.6913222141648537,
+      "grad_norm": 3.505995750427246,
+      "learning_rate": 1.15620012837081e-06,
+      "loss": 0.435,
+      "step": 715
+    },
+    {
+      "epoch": 0.6922890983804689,
+      "grad_norm": 3.468104124069214,
+      "learning_rate": 1.1495970853257963e-06,
+      "loss": 0.5059,
+      "step": 716
+    },
+    {
+      "epoch": 0.6932559825960841,
+      "grad_norm": 3.6698925495147705,
+      "learning_rate": 1.1430073170601968e-06,
+      "loss": 0.4358,
+      "step": 717
+    },
+    {
+      "epoch": 0.6942228668116993,
+      "grad_norm": 3.6901962757110596,
+      "learning_rate": 1.1364308883529913e-06,
+      "loss": 0.4783,
+      "step": 718
+    },
+    {
+      "epoch": 0.6951897510273145,
+      "grad_norm": 3.789227247238159,
+      "learning_rate": 1.1298678638520247e-06,
+      "loss": 0.4928,
+      "step": 719
+    },
+    {
+      "epoch": 0.6961566352429297,
+      "grad_norm": 3.7630035877227783,
+      "learning_rate": 1.1233183080733764e-06,
+      "loss": 0.4331,
+      "step": 720
+    },
+    {
+      "epoch": 0.6971235194585449,
+      "grad_norm": 3.66995906829834,
+      "learning_rate": 1.1167822854007265e-06,
+      "loss": 0.4984,
+      "step": 721
+    },
+    {
+      "epoch": 0.69809040367416,
+      "grad_norm": 3.4906699657440186,
+      "learning_rate": 1.1102598600847222e-06,
+      "loss": 0.5019,
+      "step": 722
+    },
+    {
+      "epoch": 0.6990572878897752,
+      "grad_norm": 4.130275726318359,
+      "learning_rate": 1.1037510962423425e-06,
+      "loss": 0.4758,
+      "step": 723
+    },
+    {
+      "epoch": 0.7000241721053904,
+      "grad_norm": 3.688877820968628,
+      "learning_rate": 1.0972560578562715e-06,
+      "loss": 0.3597,
+      "step": 724
+    },
+    {
+      "epoch": 0.7009910563210056,
+      "grad_norm": 3.8207924365997314,
+      "learning_rate": 1.0907748087742716e-06,
+      "loss": 0.4928,
+      "step": 725
+    },
+    {
+      "epoch": 0.7019579405366208,
+      "grad_norm": 3.633645534515381,
+      "learning_rate": 1.0843074127085506e-06,
+      "loss": 0.4764,
+      "step": 726
+    },
+    {
+      "epoch": 0.7029248247522359,
+      "grad_norm": 4.134641170501709,
+      "learning_rate": 1.0778539332351374e-06,
+      "loss": 0.4815,
+      "step": 727
+    },
+    {
+      "epoch": 0.7038917089678511,
+      "grad_norm": 3.7726616859436035,
+      "learning_rate": 1.0714144337932588e-06,
+      "loss": 0.4984,
+      "step": 728
+    },
+    {
+      "epoch": 0.7048585931834663,
+      "grad_norm": 3.6876401901245117,
+      "learning_rate": 1.0649889776847161e-06,
+      "loss": 0.4851,
+      "step": 729
+    },
+    {
+      "epoch": 0.7058254773990814,
+      "grad_norm": 3.7998340129852295,
+      "learning_rate": 1.0585776280732585e-06,
+      "loss": 0.463,
+      "step": 730
+    },
+    {
+      "epoch": 0.7067923616146966,
+      "grad_norm": 3.6184780597686768,
+      "learning_rate": 1.0521804479839651e-06,
+      "loss": 0.4611,
+      "step": 731
+    },
+    {
+      "epoch": 0.7077592458303118,
+      "grad_norm": 3.5277678966522217,
+      "learning_rate": 1.0457975003026278e-06,
+      "loss": 0.4036,
+      "step": 732
+    },
+    {
+      "epoch": 0.708726130045927,
+      "grad_norm": 3.882558822631836,
+      "learning_rate": 1.0394288477751274e-06,
+      "loss": 0.4848,
+      "step": 733
+    },
+    {
+      "epoch": 0.7096930142615422,
+      "grad_norm": 3.571218729019165,
+      "learning_rate": 1.0330745530068235e-06,
+      "loss": 0.4654,
+      "step": 734
+    },
+    {
+      "epoch": 0.7106598984771574,
+      "grad_norm": 3.78566312789917,
+      "learning_rate": 1.0267346784619324e-06,
+      "loss": 0.3912,
+      "step": 735
+    },
+    {
+      "epoch": 0.7116267826927726,
+      "grad_norm": 3.657423496246338,
+      "learning_rate": 1.0204092864629173e-06,
+      "loss": 0.4067,
+      "step": 736
+    },
+    {
+      "epoch": 0.7125936669083878,
+      "grad_norm": 3.9299519062042236,
+      "learning_rate": 1.0140984391898744e-06,
+      "loss": 0.4697,
+      "step": 737
+    },
+    {
+      "epoch": 0.7135605511240029,
+      "grad_norm": 3.5789129734039307,
+      "learning_rate": 1.0078021986799238e-06,
+      "loss": 0.4346,
+      "step": 738
+    },
+    {
+      "epoch": 0.714527435339618,
+      "grad_norm": 3.3570632934570312,
+      "learning_rate": 1.0015206268265948e-06,
+      "loss": 0.388,
+      "step": 739
+    },
+    {
+      "epoch": 0.7154943195552332,
+      "grad_norm": 3.806309938430786,
+      "learning_rate": 9.952537853792204e-07,
+      "loss": 0.481,
+      "step": 740
+    },
+    {
+      "epoch": 0.7164612037708484,
+      "grad_norm": 4.444820880889893,
+      "learning_rate": 9.890017359423326e-07,
+      "loss": 0.4859,
+      "step": 741
+    },
+    {
+      "epoch": 0.7174280879864636,
+      "grad_norm": 3.6907238960266113,
+      "learning_rate": 9.827645399750539e-07,
+      "loss": 0.4662,
+      "step": 742
+    },
+    {
+      "epoch": 0.7183949722020788,
+      "grad_norm": 3.746755361557007,
+      "learning_rate": 9.765422587904919e-07,
+      "loss": 0.4504,
+      "step": 743
+    },
+    {
+      "epoch": 0.719361856417694,
+      "grad_norm": 3.8143019676208496,
+      "learning_rate": 9.703349535551387e-07,
+      "loss": 0.4632,
+      "step": 744
+    },
+    {
+      "epoch": 0.7203287406333092,
+      "grad_norm": 3.932626485824585,
+      "learning_rate": 9.641426852882717e-07,
+      "loss": 0.4846,
+      "step": 745
+    },
+    {
+      "epoch": 0.7212956248489243,
+      "grad_norm": 3.918102502822876,
+      "learning_rate": 9.579655148613478e-07,
+      "loss": 0.5095,
+      "step": 746
+    },
+    {
+      "epoch": 0.7222625090645395,
+      "grad_norm": 3.4765045642852783,
+      "learning_rate": 9.518035029974127e-07,
+      "loss": 0.4269,
+      "step": 747
+    },
+    {
+      "epoch": 0.7232293932801547,
+      "grad_norm": 3.6121413707733154,
+      "learning_rate": 9.456567102704961e-07,
+      "loss": 0.412,
+      "step": 748
+    },
+    {
+      "epoch": 0.7241962774957699,
+      "grad_norm": 3.615751028060913,
+      "learning_rate": 9.395251971050206e-07,
+      "loss": 0.4399,
+      "step": 749
+    },
+    {
+      "epoch": 0.7251631617113851,
+      "grad_norm": 3.4750213623046875,
+      "learning_rate": 9.334090237752095e-07,
+      "loss": 0.452,
+      "step": 750
+    },
+    {
+      "epoch": 0.7261300459270003,
+      "grad_norm": 3.536120891571045,
+      "learning_rate": 9.273082504044903e-07,
+      "loss": 0.4228,
+      "step": 751
+    },
+    {
+      "epoch": 0.7270969301426154,
+      "grad_norm": 3.8346316814422607,
+      "learning_rate": 9.21222936964905e-07,
+      "loss": 0.5377,
+      "step": 752
+    },
+    {
+      "epoch": 0.7280638143582306,
+      "grad_norm": 3.4409730434417725,
+      "learning_rate": 9.151531432765204e-07,
+      "loss": 0.3675,
+      "step": 753
+    },
+    {
+      "epoch": 0.7290306985738458,
+      "grad_norm": 3.424736261367798,
+      "learning_rate": 9.090989290068425e-07,
+      "loss": 0.4304,
+      "step": 754
+    },
+    {
+      "epoch": 0.7299975827894609,
+      "grad_norm": 3.5875778198242188,
+      "learning_rate": 9.030603536702254e-07,
+      "loss": 0.4541,
+      "step": 755
+    },
+    {
+      "epoch": 0.7309644670050761,
+      "grad_norm": 3.5809242725372314,
+      "learning_rate": 8.970374766272916e-07,
+      "loss": 0.4664,
+      "step": 756
+    },
+    {
+      "epoch": 0.7319313512206913,
+      "grad_norm": 3.7358856201171875,
+      "learning_rate": 8.910303570843423e-07,
+      "loss": 0.4463,
+      "step": 757
+    },
+    {
+      "epoch": 0.7328982354363065,
+      "grad_norm": 3.236128807067871,
+      "learning_rate": 8.850390540927825e-07,
+      "loss": 0.4242,
+      "step": 758
+    },
+    {
+      "epoch": 0.7338651196519217,
+      "grad_norm": 3.585001230239868,
+      "learning_rate": 8.790636265485333e-07,
+      "loss": 0.4475,
+      "step": 759
+    },
+    {
+      "epoch": 0.7348320038675369,
+      "grad_norm": 4.007416248321533,
+      "learning_rate": 8.731041331914591e-07,
+      "loss": 0.4996,
+      "step": 760
+    },
+    {
+      "epoch": 0.7357988880831521,
+      "grad_norm": 3.5270919799804688,
+      "learning_rate": 8.67160632604786e-07,
+      "loss": 0.412,
+      "step": 761
+    },
+    {
+      "epoch": 0.7367657722987673,
+      "grad_norm": 3.9546616077423096,
+      "learning_rate": 8.612331832145269e-07,
+      "loss": 0.4678,
+      "step": 762
+    },
+    {
+      "epoch": 0.7377326565143824,
+      "grad_norm": 3.887314558029175,
+      "learning_rate": 8.553218432889091e-07,
+      "loss": 0.4595,
+      "step": 763
+    },
+    {
+      "epoch": 0.7386995407299976,
+      "grad_norm": 3.572169303894043,
+      "learning_rate": 8.49426670937801e-07,
+      "loss": 0.3734,
+      "step": 764
+    },
+    {
+      "epoch": 0.7396664249456127,
+      "grad_norm": 3.3383076190948486,
+      "learning_rate": 8.435477241121354e-07,
+      "loss": 0.396,
+      "step": 765
+    },
+    {
+      "epoch": 0.7406333091612279,
+      "grad_norm": 3.5187559127807617,
+      "learning_rate": 8.376850606033487e-07,
+      "loss": 0.4395,
+      "step": 766
+    },
+    {
+      "epoch": 0.7416001933768431,
+      "grad_norm": 3.5995166301727295,
+      "learning_rate": 8.31838738042808e-07,
+      "loss": 0.4179,
+      "step": 767
+    },
+    {
+      "epoch": 0.7425670775924583,
+      "grad_norm": 3.620683193206787,
+      "learning_rate": 8.260088139012435e-07,
+      "loss": 0.4468,
+      "step": 768
+    },
+    {
+      "epoch": 0.7435339618080735,
+      "grad_norm": 3.868516445159912,
+      "learning_rate": 8.201953454881844e-07,
+      "loss": 0.5147,
+      "step": 769
+    },
+    {
+      "epoch": 0.7445008460236887,
+      "grad_norm": 3.560850143432617,
+      "learning_rate": 8.143983899513988e-07,
+      "loss": 0.4294,
+      "step": 770
+    },
+    {
+      "epoch": 0.7454677302393038,
+      "grad_norm": 3.773320436477661,
+      "learning_rate": 8.086180042763284e-07,
+      "loss": 0.4832,
+      "step": 771
+    },
+    {
+      "epoch": 0.746434614454919,
+      "grad_norm": 3.5478060245513916,
+      "learning_rate": 8.028542452855281e-07,
+      "loss": 0.45,
+      "step": 772
+    },
+    {
+      "epoch": 0.7474014986705342,
+      "grad_norm": 3.715156316757202,
+      "learning_rate": 7.971071696381089e-07,
+      "loss": 0.4849,
+      "step": 773
+    },
+    {
+      "epoch": 0.7483683828861494,
+      "grad_norm": 3.5130743980407715,
+      "learning_rate": 7.913768338291821e-07,
+      "loss": 0.456,
+      "step": 774
+    },
+    {
+      "epoch": 0.7493352671017646,
+      "grad_norm": 3.6129679679870605,
+      "learning_rate": 7.856632941893e-07,
+      "loss": 0.3651,
+      "step": 775
+    },
+    {
+      "epoch": 0.7503021513173798,
+      "grad_norm": 3.693694591522217,
+      "learning_rate": 7.799666068839068e-07,
+      "loss": 0.4264,
+      "step": 776
+    },
+    {
+      "epoch": 0.751269035532995,
+      "grad_norm": 3.7626357078552246,
+      "learning_rate": 7.74286827912785e-07,
+      "loss": 0.4372,
+      "step": 777
+    },
+    {
+      "epoch": 0.7522359197486101,
+      "grad_norm": 3.6865146160125732,
+      "learning_rate": 7.686240131094993e-07,
+      "loss": 0.514,
+      "step": 778
+    },
+    {
+      "epoch": 0.7532028039642252,
+      "grad_norm": 3.815524101257324,
+      "learning_rate": 7.629782181408574e-07,
+      "loss": 0.4652,
+      "step": 779
+    },
+    {
+      "epoch": 0.7541696881798404,
+      "grad_norm": 3.8076345920562744,
+      "learning_rate": 7.573494985063579e-07,
+      "loss": 0.479,
+      "step": 780
+    },
+    {
+      "epoch": 0.7551365723954556,
+      "grad_norm": 3.990732192993164,
+      "learning_rate": 7.517379095376418e-07,
+      "loss": 0.5049,
+      "step": 781
+    },
+    {
+      "epoch": 0.7561034566110708,
+      "grad_norm": 3.6248281002044678,
+      "learning_rate": 7.461435063979533e-07,
+      "loss": 0.4247,
+      "step": 782
+    },
+    {
+      "epoch": 0.757070340826686,
+      "grad_norm": 4.307629585266113,
+      "learning_rate": 7.405663440815968e-07,
+      "loss": 0.4989,
+      "step": 783
+    },
+    {
+      "epoch": 0.7580372250423012,
+      "grad_norm": 3.800755500793457,
+      "learning_rate": 7.350064774133936e-07,
+      "loss": 0.4744,
+      "step": 784
+    },
+    {
+      "epoch": 0.7590041092579164,
+      "grad_norm": 3.5235469341278076,
+      "learning_rate": 7.294639610481461e-07,
+      "loss": 0.4778,
+      "step": 785
+    },
+    {
+      "epoch": 0.7599709934735316,
+      "grad_norm": 3.831073760986328,
+      "learning_rate": 7.239388494700974e-07,
+      "loss": 0.4422,
+      "step": 786
+    },
+    {
+      "epoch": 0.7609378776891468,
+      "grad_norm": 3.861233949661255,
+      "learning_rate": 7.184311969924002e-07,
+      "loss": 0.465,
+      "step": 787
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 4.040560722351074,
+      "learning_rate": 7.129410577565766e-07,
+      "loss": 0.5075,
+      "step": 788
+    },
+    {
+      "epoch": 0.7628716461203771,
+      "grad_norm": 3.890537977218628,
+      "learning_rate": 7.074684857319928e-07,
+      "loss": 0.4742,
+      "step": 789
+    },
+    {
+      "epoch": 0.7638385303359922,
+      "grad_norm": 3.9048914909362793,
+      "learning_rate": 7.020135347153229e-07,
+      "loss": 0.4557,
+      "step": 790
+    },
+    {
+      "epoch": 0.7648054145516074,
+      "grad_norm": 4.083683490753174,
+      "learning_rate": 6.965762583300223e-07,
+      "loss": 0.459,
+      "step": 791
+    },
+    {
+      "epoch": 0.7657722987672226,
+      "grad_norm": 3.9095261096954346,
+      "learning_rate": 6.911567100258021e-07,
+      "loss": 0.4768,
+      "step": 792
+    },
+    {
+      "epoch": 0.7667391829828378,
+      "grad_norm": 4.056056976318359,
+      "learning_rate": 6.85754943078103e-07,
+      "loss": 0.5037,
+      "step": 793
+    },
+    {
+      "epoch": 0.767706067198453,
+      "grad_norm": 3.6234371662139893,
+      "learning_rate": 6.803710105875661e-07,
+      "loss": 0.4934,
+      "step": 794
+    },
+    {
+      "epoch": 0.7686729514140682,
+      "grad_norm": 3.658932685852051,
+      "learning_rate": 6.750049654795199e-07,
+      "loss": 0.5005,
+      "step": 795
+    },
+    {
+      "epoch": 0.7696398356296833,
+      "grad_norm": 3.7866780757904053,
+      "learning_rate": 6.696568605034559e-07,
+      "loss": 0.4743,
+      "step": 796
+    },
+    {
+      "epoch": 0.7706067198452985,
+      "grad_norm": 3.559441328048706,
+      "learning_rate": 6.643267482325061e-07,
+      "loss": 0.4083,
+      "step": 797
+    },
+    {
+      "epoch": 0.7715736040609137,
+      "grad_norm": 3.7595908641815186,
+      "learning_rate": 6.590146810629347e-07,
+      "loss": 0.4967,
+      "step": 798
+    },
+    {
+      "epoch": 0.7725404882765289,
+      "grad_norm": 3.7670183181762695,
+      "learning_rate": 6.537207112136143e-07,
+      "loss": 0.4929,
+      "step": 799
+    },
+    {
+      "epoch": 0.7735073724921441,
+      "grad_norm": 4.016094207763672,
+      "learning_rate": 6.484448907255181e-07,
+      "loss": 0.4798,
+      "step": 800
+    },
+    {
+      "epoch": 0.7744742567077593,
+      "grad_norm": 3.9730477333068848,
+      "learning_rate": 6.431872714612072e-07,
+      "loss": 0.5017,
+      "step": 801
+    },
+    {
+      "epoch": 0.7754411409233745,
+      "grad_norm": 4.174206733703613,
+      "learning_rate": 6.379479051043208e-07,
+      "loss": 0.4918,
+      "step": 802
+    },
+    {
+      "epoch": 0.7764080251389897,
+      "grad_norm": 3.839644193649292,
+      "learning_rate": 6.327268431590664e-07,
+      "loss": 0.4634,
+      "step": 803
+    },
+    {
+      "epoch": 0.7773749093546047,
+      "grad_norm": 3.615432024002075,
+      "learning_rate": 6.275241369497142e-07,
+      "loss": 0.4287,
+      "step": 804
+    },
+    {
+      "epoch": 0.7783417935702199,
+      "grad_norm": 3.912182569503784,
+      "learning_rate": 6.223398376200956e-07,
+      "loss": 0.4709,
+      "step": 805
+    },
+    {
+      "epoch": 0.7793086777858351,
+      "grad_norm": 3.828120470046997,
+      "learning_rate": 6.171739961330983e-07,
+      "loss": 0.4846,
+      "step": 806
+    },
+    {
+      "epoch": 0.7802755620014503,
+      "grad_norm": 3.653789758682251,
+      "learning_rate": 6.1202666327016e-07,
+      "loss": 0.4179,
+      "step": 807
+    },
+    {
+      "epoch": 0.7812424462170655,
+      "grad_norm": 3.5782525539398193,
+      "learning_rate": 6.068978896307795e-07,
+      "loss": 0.4177,
+      "step": 808
+    },
+    {
+      "epoch": 0.7822093304326807,
+      "grad_norm": 3.655728340148926,
+      "learning_rate": 6.017877256320132e-07,
+      "loss": 0.4135,
+      "step": 809
+    },
+    {
+      "epoch": 0.7831762146482959,
+      "grad_norm": 3.3964128494262695,
+      "learning_rate": 5.966962215079786e-07,
+      "loss": 0.3804,
+      "step": 810
+    },
+    {
+      "epoch": 0.7841430988639111,
+      "grad_norm": 3.8653347492218018,
+      "learning_rate": 5.916234273093624e-07,
+      "loss": 0.5036,
+      "step": 811
+    },
+    {
+      "epoch": 0.7851099830795262,
+      "grad_norm": 3.8347480297088623,
+      "learning_rate": 5.865693929029306e-07,
+      "loss": 0.4827,
+      "step": 812
+    },
+    {
+      "epoch": 0.7860768672951414,
+      "grad_norm": 3.4399571418762207,
+      "learning_rate": 5.815341679710327e-07,
+      "loss": 0.475,
+      "step": 813
+    },
+    {
+      "epoch": 0.7870437515107566,
+      "grad_norm": 3.6704349517822266,
+      "learning_rate": 5.765178020111201e-07,
+      "loss": 0.4524,
+      "step": 814
+    },
+    {
+      "epoch": 0.7880106357263718,
+      "grad_norm": 3.6934151649475098,
+      "learning_rate": 5.715203443352526e-07,
+      "loss": 0.507,
+      "step": 815
+    },
+    {
+      "epoch": 0.788977519941987,
+      "grad_norm": 3.8846731185913086,
+      "learning_rate": 5.665418440696202e-07,
+      "loss": 0.4698,
+      "step": 816
+    },
+    {
+      "epoch": 0.7899444041576021,
+      "grad_norm": 3.6200265884399414,
+      "learning_rate": 5.615823501540546e-07,
+      "loss": 0.4948,
+      "step": 817
+    },
+    {
+      "epoch": 0.7909112883732173,
+      "grad_norm": 3.749350070953369,
+      "learning_rate": 5.566419113415528e-07,
+      "loss": 0.4372,
+      "step": 818
+    },
+    {
+      "epoch": 0.7918781725888325,
+      "grad_norm": 4.006725788116455,
+      "learning_rate": 5.51720576197794e-07,
+      "loss": 0.4777,
+      "step": 819
+    },
+    {
+      "epoch": 0.7928450568044476,
+      "grad_norm": 3.7433784008026123,
+      "learning_rate": 5.468183931006635e-07,
+      "loss": 0.4171,
+      "step": 820
+    },
+    {
+      "epoch": 0.7938119410200628,
+      "grad_norm": 3.7018020153045654,
+      "learning_rate": 5.419354102397792e-07,
+      "loss": 0.4783,
+      "step": 821
+    },
+    {
+      "epoch": 0.794778825235678,
+      "grad_norm": 3.5116868019104004,
+      "learning_rate": 5.370716756160158e-07,
+      "loss": 0.4673,
+      "step": 822
+    },
+    {
+      "epoch": 0.7957457094512932,
+      "grad_norm": 3.450648069381714,
+      "learning_rate": 5.32227237041032e-07,
+      "loss": 0.4079,
+      "step": 823
+    },
+    {
+      "epoch": 0.7967125936669084,
+      "grad_norm": 3.5097784996032715,
+      "learning_rate": 5.274021421368019e-07,
+      "loss": 0.4189,
+      "step": 824
+    },
+    {
+      "epoch": 0.7976794778825236,
+      "grad_norm": 3.470055341720581,
+      "learning_rate": 5.22596438335149e-07,
+      "loss": 0.4872,
+      "step": 825
+    },
+    {
+      "epoch": 0.7986463620981388,
+      "grad_norm": 4.227003574371338,
+      "learning_rate": 5.178101728772741e-07,
+      "loss": 0.4725,
+      "step": 826
+    },
+    {
+      "epoch": 0.799613246313754,
+      "grad_norm": 4.068714618682861,
+      "learning_rate": 5.130433928132983e-07,
+      "loss": 0.4568,
+      "step": 827
+    },
+    {
+      "epoch": 0.8005801305293692,
+      "grad_norm": 4.06673526763916,
+      "learning_rate": 5.082961450017943e-07,
+      "loss": 0.4902,
+      "step": 828
+    },
+    {
+      "epoch": 0.8015470147449842,
+      "grad_norm": 3.6162238121032715,
+      "learning_rate": 5.035684761093273e-07,
+      "loss": 0.4429,
+      "step": 829
+    },
+    {
+      "epoch": 0.8025138989605994,
+      "grad_norm": 3.8663253784179688,
+      "learning_rate": 4.988604326099999e-07,
+      "loss": 0.4794,
+      "step": 830
+    },
+    {
+      "epoch": 0.8034807831762146,
+      "grad_norm": 3.618603467941284,
+      "learning_rate": 4.941720607849912e-07,
+      "loss": 0.4308,
+      "step": 831
+    },
+    {
+      "epoch": 0.8044476673918298,
+      "grad_norm": 3.8849265575408936,
+      "learning_rate": 4.895034067221025e-07,
+      "loss": 0.4932,
+      "step": 832
+    },
+    {
+      "epoch": 0.805414551607445,
+      "grad_norm": 3.5201117992401123,
+      "learning_rate": 4.848545163153048e-07,
+      "loss": 0.443,
+      "step": 833
+    },
+    {
+      "epoch": 0.8063814358230602,
+      "grad_norm": 3.6072754859924316,
+      "learning_rate": 4.802254352642883e-07,
+      "loss": 0.4063,
+      "step": 834
+    },
+    {
+      "epoch": 0.8073483200386754,
+      "grad_norm": 3.856928586959839,
+      "learning_rate": 4.756162090740135e-07,
+      "loss": 0.5103,
+      "step": 835
+    },
+    {
+      "epoch": 0.8083152042542906,
+      "grad_norm": 3.5394015312194824,
+      "learning_rate": 4.7102688305426106e-07,
+      "loss": 0.4409,
+      "step": 836
+    },
+    {
+      "epoch": 0.8092820884699057,
+      "grad_norm": 3.6355698108673096,
+      "learning_rate": 4.6645750231918864e-07,
+      "loss": 0.5113,
+      "step": 837
+    },
+    {
+      "epoch": 0.8102489726855209,
+      "grad_norm": 4.026984691619873,
+      "learning_rate": 4.6190811178688843e-07,
+      "loss": 0.4681,
+      "step": 838
+    },
+    {
+      "epoch": 0.8112158569011361,
+      "grad_norm": 3.748511552810669,
+      "learning_rate": 4.5737875617894225e-07,
+      "loss": 0.4808,
+      "step": 839
+    },
+    {
+      "epoch": 0.8121827411167513,
+      "grad_norm": 3.9079511165618896,
+      "learning_rate": 4.528694800199859e-07,
+      "loss": 0.4995,
+      "step": 840
+    },
+    {
+      "epoch": 0.8131496253323665,
+      "grad_norm": 3.926765203475952,
+      "learning_rate": 4.4838032763726806e-07,
+      "loss": 0.4712,
+      "step": 841
+    },
+    {
+      "epoch": 0.8141165095479816,
+      "grad_norm": 3.8391571044921875,
+      "learning_rate": 4.4391134316021545e-07,
+      "loss": 0.4801,
+      "step": 842
+    },
+    {
+      "epoch": 0.8150833937635968,
+      "grad_norm": 3.581597089767456,
+      "learning_rate": 4.394625705200012e-07,
+      "loss": 0.3686,
+      "step": 843
+    },
+    {
+      "epoch": 0.816050277979212,
+      "grad_norm": 3.869783401489258,
+      "learning_rate": 4.3503405344911125e-07,
+      "loss": 0.4921,
+      "step": 844
+    },
+    {
+      "epoch": 0.8170171621948271,
+      "grad_norm": 3.75410532951355,
+      "learning_rate": 4.3062583548091256e-07,
+      "loss": 0.4503,
+      "step": 845
+    },
+    {
+      "epoch": 0.8179840464104423,
+      "grad_norm": 3.647118330001831,
+      "learning_rate": 4.2623795994922836e-07,
+      "loss": 0.4418,
+      "step": 846
+    },
+    {
+      "epoch": 0.8189509306260575,
+      "grad_norm": 3.6203055381774902,
+      "learning_rate": 4.218704699879117e-07,
+      "loss": 0.415,
+      "step": 847
+    },
+    {
+      "epoch": 0.8199178148416727,
+      "grad_norm": 3.5551204681396484,
+      "learning_rate": 4.1752340853041973e-07,
+      "loss": 0.4879,
+      "step": 848
+    },
+    {
+      "epoch": 0.8208846990572879,
+      "grad_norm": 3.568030595779419,
+      "learning_rate": 4.1319681830939124e-07,
+      "loss": 0.4573,
+      "step": 849
+    },
+    {
+      "epoch": 0.8218515832729031,
+      "grad_norm": 3.600548028945923,
+      "learning_rate": 4.0889074185623103e-07,
+      "loss": 0.4589,
+      "step": 850
+    },
+    {
+      "epoch": 0.8228184674885183,
+      "grad_norm": 3.7248218059539795,
+      "learning_rate": 4.0460522150068684e-07,
+      "loss": 0.4588,
+      "step": 851
+    },
+    {
+      "epoch": 0.8237853517041335,
+      "grad_norm": 3.6307175159454346,
+      "learning_rate": 4.0034029937043527e-07,
+      "loss": 0.5157,
+      "step": 852
+    },
+    {
+      "epoch": 0.8247522359197486,
+      "grad_norm": 3.7944419384002686,
+      "learning_rate": 3.9609601739066664e-07,
+      "loss": 0.4348,
+      "step": 853
+    },
+    {
+      "epoch": 0.8257191201353637,
+      "grad_norm": 3.4527225494384766,
+      "learning_rate": 3.918724172836763e-07,
+      "loss": 0.5047,
+      "step": 854
+    },
+    {
+      "epoch": 0.8266860043509789,
+      "grad_norm": 3.6622514724731445,
+      "learning_rate": 3.876695405684486e-07,
+      "loss": 0.4555,
+      "step": 855
+    },
+    {
+      "epoch": 0.8276528885665941,
+      "grad_norm": 3.5915470123291016,
+      "learning_rate": 3.8348742856025495e-07,
+      "loss": 0.4388,
+      "step": 856
+    },
+    {
+      "epoch": 0.8286197727822093,
+      "grad_norm": 3.6613149642944336,
+      "learning_rate": 3.793261223702441e-07,
+      "loss": 0.4704,
+      "step": 857
+    },
+    {
+      "epoch": 0.8295866569978245,
+      "grad_norm": 3.679746389389038,
+      "learning_rate": 3.751856629050363e-07,
+      "loss": 0.4167,
+      "step": 858
+    },
+    {
+      "epoch": 0.8305535412134397,
+      "grad_norm": 3.6246674060821533,
+      "learning_rate": 3.7106609086632635e-07,
+      "loss": 0.4532,
+      "step": 859
+    },
+    {
+      "epoch": 0.8315204254290549,
+      "grad_norm": 3.590588092803955,
+      "learning_rate": 3.669674467504808e-07,
+      "loss": 0.458,
+      "step": 860
+    },
+    {
+      "epoch": 0.8324873096446701,
+      "grad_norm": 3.574864149093628,
+      "learning_rate": 3.628897708481377e-07,
+      "loss": 0.5124,
+      "step": 861
+    },
+    {
+      "epoch": 0.8334541938602852,
+      "grad_norm": 3.350720167160034,
+      "learning_rate": 3.588331032438133e-07,
+      "loss": 0.3843,
+      "step": 862
+    },
+    {
+      "epoch": 0.8344210780759004,
+      "grad_norm": 3.3735225200653076,
+      "learning_rate": 3.5479748381550855e-07,
+      "loss": 0.4444,
+      "step": 863
+    },
+    {
+      "epoch": 0.8353879622915156,
+      "grad_norm": 3.831509590148926,
+      "learning_rate": 3.5078295223431534e-07,
+      "loss": 0.493,
+      "step": 864
+    },
+    {
+      "epoch": 0.8363548465071308,
+      "grad_norm": 3.679081916809082,
+      "learning_rate": 3.4678954796402624e-07,
+      "loss": 0.4197,
+      "step": 865
+    },
+    {
+      "epoch": 0.837321730722746,
+      "grad_norm": 3.6159653663635254,
+      "learning_rate": 3.4281731026074737e-07,
+      "loss": 0.4386,
+      "step": 866
+    },
+    {
+      "epoch": 0.8382886149383612,
+      "grad_norm": 3.782343626022339,
+      "learning_rate": 3.388662781725141e-07,
+      "loss": 0.4017,
+      "step": 867
+    },
+    {
+      "epoch": 0.8392554991539763,
+      "grad_norm": 3.8932251930236816,
+      "learning_rate": 3.3493649053890325e-07,
+      "loss": 0.4875,
+      "step": 868
+    },
+    {
+      "epoch": 0.8402223833695915,
+      "grad_norm": 3.8822195529937744,
+      "learning_rate": 3.310279859906565e-07,
+      "loss": 0.4377,
+      "step": 869
+    },
+    {
+      "epoch": 0.8411892675852066,
+      "grad_norm": 3.878910779953003,
+      "learning_rate": 3.271408029492948e-07,
+      "loss": 0.3554,
+      "step": 870
+    },
+    {
+      "epoch": 0.8421561518008218,
+      "grad_norm": 3.642711877822876,
+      "learning_rate": 3.232749796267451e-07,
+      "loss": 0.4133,
+      "step": 871
+    },
+    {
+      "epoch": 0.843123036016437,
+      "grad_norm": 3.5945470333099365,
+      "learning_rate": 3.194305540249637e-07,
+      "loss": 0.46,
+      "step": 872
+    },
+    {
+      "epoch": 0.8440899202320522,
+      "grad_norm": 3.501006603240967,
+      "learning_rate": 3.1560756393556187e-07,
+      "loss": 0.4509,
+      "step": 873
+    },
+    {
+      "epoch": 0.8450568044476674,
+      "grad_norm": 3.5597221851348877,
+      "learning_rate": 3.118060469394346e-07,
+      "loss": 0.4517,
+      "step": 874
+    },
+    {
+      "epoch": 0.8460236886632826,
+      "grad_norm": 3.781863212585449,
+      "learning_rate": 3.0802604040639034e-07,
+      "loss": 0.4685,
+      "step": 875
+    },
+    {
+      "epoch": 0.8469905728788978,
+      "grad_norm": 3.9921908378601074,
+      "learning_rate": 3.042675814947868e-07,
+      "loss": 0.4825,
+      "step": 876
+    },
+    {
+      "epoch": 0.847957457094513,
+      "grad_norm": 3.7082839012145996,
+      "learning_rate": 3.0053070715116153e-07,
+      "loss": 0.3958,
+      "step": 877
+    },
+    {
+      "epoch": 0.8489243413101281,
+      "grad_norm": 4.115004062652588,
+      "learning_rate": 2.9681545410987193e-07,
+      "loss": 0.4948,
+      "step": 878
+    },
+    {
+      "epoch": 0.8498912255257433,
+      "grad_norm": 3.928007125854492,
+      "learning_rate": 2.9312185889273147e-07,
+      "loss": 0.3921,
+      "step": 879
+    },
+    {
+      "epoch": 0.8508581097413584,
+      "grad_norm": 3.7645504474639893,
+      "learning_rate": 2.894499578086543e-07,
+      "loss": 0.4806,
+      "step": 880
+    },
+    {
+      "epoch": 0.8518249939569736,
+      "grad_norm": 3.6392364501953125,
+      "learning_rate": 2.8579978695329386e-07,
+      "loss": 0.5108,
+      "step": 881
+    },
+    {
+      "epoch": 0.8527918781725888,
+      "grad_norm": 3.5847926139831543,
+      "learning_rate": 2.821713822086919e-07,
+      "loss": 0.4466,
+      "step": 882
+    },
+    {
+      "epoch": 0.853758762388204,
+      "grad_norm": 3.5183796882629395,
+      "learning_rate": 2.785647792429233e-07,
+      "loss": 0.4317,
+      "step": 883
+    },
+    {
+      "epoch": 0.8547256466038192,
+      "grad_norm": 3.4891810417175293,
+      "learning_rate": 2.7498001350974605e-07,
+      "loss": 0.4389,
+      "step": 884
+    },
+    {
+      "epoch": 0.8556925308194344,
+      "grad_norm": 3.8687262535095215,
+      "learning_rate": 2.714171202482538e-07,
+      "loss": 0.4642,
+      "step": 885
+    },
+    {
+      "epoch": 0.8566594150350495,
+      "grad_norm": 3.986570358276367,
+      "learning_rate": 2.678761344825295e-07,
+      "loss": 0.4655,
+      "step": 886
+    },
+    {
+      "epoch": 0.8576262992506647,
+      "grad_norm": 3.5800821781158447,
+      "learning_rate": 2.6435709102129727e-07,
+      "loss": 0.418,
+      "step": 887
+    },
+    {
+      "epoch": 0.8585931834662799,
+      "grad_norm": 3.695192813873291,
+      "learning_rate": 2.608600244575857e-07,
+      "loss": 0.4942,
+      "step": 888
+    },
+    {
+      "epoch": 0.8595600676818951,
+      "grad_norm": 3.764131546020508,
+      "learning_rate": 2.5738496916838524e-07,
+      "loss": 0.4745,
+      "step": 889
+    },
+    {
+      "epoch": 0.8605269518975103,
+      "grad_norm": 3.749889373779297,
+      "learning_rate": 2.5393195931430884e-07,
+      "loss": 0.4866,
+      "step": 890
+    },
+    {
+      "epoch": 0.8614938361131255,
+      "grad_norm": 3.4770333766937256,
+      "learning_rate": 2.505010288392587e-07,
+      "loss": 0.4979,
+      "step": 891
+    },
+    {
+      "epoch": 0.8624607203287407,
+      "grad_norm": 3.926590919494629,
+      "learning_rate": 2.4709221147009105e-07,
+      "loss": 0.4444,
+      "step": 892
+    },
+    {
+      "epoch": 0.8634276045443559,
+      "grad_norm": 3.7266767024993896,
+      "learning_rate": 2.4370554071628613e-07,
+      "loss": 0.4518,
+      "step": 893
+    },
+    {
+      "epoch": 0.864394488759971,
+      "grad_norm": 3.3029301166534424,
+      "learning_rate": 2.4034104986961626e-07,
+      "loss": 0.4383,
+      "step": 894
+    },
+    {
+      "epoch": 0.8653613729755861,
+      "grad_norm": 3.774355888366699,
+      "learning_rate": 2.3699877200382026e-07,
+      "loss": 0.4556,
+      "step": 895
+    },
+    {
+      "epoch": 0.8663282571912013,
+      "grad_norm": 3.818542718887329,
+      "learning_rate": 2.336787399742793e-07,
+      "loss": 0.4664,
+      "step": 896
+    },
+    {
+      "epoch": 0.8672951414068165,
+      "grad_norm": 3.992581605911255,
+      "learning_rate": 2.303809864176909e-07,
+      "loss": 0.4589,
+      "step": 897
+    },
+    {
+      "epoch": 0.8682620256224317,
+      "grad_norm": 3.4998490810394287,
+      "learning_rate": 2.271055437517519e-07,
+      "loss": 0.4027,
+      "step": 898
+    },
+    {
+      "epoch": 0.8692289098380469,
+      "grad_norm": 3.571521520614624,
+      "learning_rate": 2.2385244417483743e-07,
+      "loss": 0.4391,
+      "step": 899
+    },
+    {
+      "epoch": 0.8701957940536621,
+      "grad_norm": 3.1858935356140137,
+      "learning_rate": 2.2062171966568263e-07,
+      "loss": 0.4333,
+      "step": 900
+    },
+    {
+      "epoch": 0.8711626782692773,
+      "grad_norm": 3.4342987537384033,
+      "learning_rate": 2.174134019830726e-07,
+      "loss": 0.4417,
+      "step": 901
+    },
+    {
+      "epoch": 0.8721295624848925,
+      "grad_norm": 3.857280731201172,
+      "learning_rate": 2.14227522665528e-07,
+      "loss": 0.4099,
+      "step": 902
+    },
+    {
+      "epoch": 0.8730964467005076,
+      "grad_norm": 4.276398658752441,
+      "learning_rate": 2.1106411303099455e-07,
+      "loss": 0.4882,
+      "step": 903
+    },
+    {
+      "epoch": 0.8740633309161228,
+      "grad_norm": 3.4768717288970947,
+      "learning_rate": 2.0792320417653472e-07,
+      "loss": 0.4288,
+      "step": 904
+    },
+    {
+      "epoch": 0.875030215131738,
+      "grad_norm": 3.416771173477173,
+      "learning_rate": 2.0480482697802507e-07,
+      "loss": 0.4147,
+      "step": 905
+    },
+    {
+      "epoch": 0.8759970993473531,
+      "grad_norm": 3.6449015140533447,
+      "learning_rate": 2.017090120898485e-07,
+      "loss": 0.4636,
+      "step": 906
+    },
+    {
+      "epoch": 0.8769639835629683,
+      "grad_norm": 3.3985326290130615,
+      "learning_rate": 1.986357899445976e-07,
+      "loss": 0.4414,
+      "step": 907
+    },
+    {
+      "epoch": 0.8779308677785835,
+      "grad_norm": 3.730642557144165,
+      "learning_rate": 1.9558519075276976e-07,
+      "loss": 0.4355,
+      "step": 908
+    },
+    {
+      "epoch": 0.8788977519941987,
+      "grad_norm": 3.829969644546509,
+      "learning_rate": 1.9255724450247676e-07,
+      "loss": 0.491,
+      "step": 909
+    },
+    {
+      "epoch": 0.8798646362098139,
+      "grad_norm": 3.6221253871917725,
+      "learning_rate": 1.895519809591434e-07,
+      "loss": 0.4487,
+      "step": 910
+    },
+    {
+      "epoch": 0.880831520425429,
+      "grad_norm": 3.4621918201446533,
+      "learning_rate": 1.8656942966522124e-07,
+      "loss": 0.4012,
+      "step": 911
+    },
+    {
+      "epoch": 0.8817984046410442,
+      "grad_norm": 3.4926211833953857,
+      "learning_rate": 1.836096199398929e-07,
+      "loss": 0.4418,
+      "step": 912
+    },
+    {
+      "epoch": 0.8827652888566594,
+      "grad_norm": 3.554006814956665,
+      "learning_rate": 1.8067258087878597e-07,
+      "loss": 0.372,
+      "step": 913
+    },
+    {
+      "epoch": 0.8837321730722746,
+      "grad_norm": 3.646104097366333,
+      "learning_rate": 1.7775834135368818e-07,
+      "loss": 0.4249,
+      "step": 914
+    },
+    {
+      "epoch": 0.8846990572878898,
+      "grad_norm": 4.161782264709473,
+      "learning_rate": 1.748669300122627e-07,
+      "loss": 0.5372,
+      "step": 915
+    },
+    {
+      "epoch": 0.885665941503505,
+      "grad_norm": 3.7760252952575684,
+      "learning_rate": 1.7199837527776485e-07,
+      "loss": 0.4686,
+      "step": 916
+    },
+    {
+      "epoch": 0.8866328257191202,
+      "grad_norm": 3.922210693359375,
+      "learning_rate": 1.691527053487646e-07,
+      "loss": 0.5029,
+      "step": 917
+    },
+    {
+      "epoch": 0.8875997099347354,
+      "grad_norm": 3.84432315826416,
+      "learning_rate": 1.6632994819886976e-07,
+      "loss": 0.4649,
+      "step": 918
+    },
+    {
+      "epoch": 0.8885665941503504,
+      "grad_norm": 3.71447491645813,
+      "learning_rate": 1.635301315764484e-07,
+      "loss": 0.4049,
+      "step": 919
+    },
+    {
+      "epoch": 0.8895334783659656,
+      "grad_norm": 3.771190643310547,
+      "learning_rate": 1.607532830043601e-07,
+      "loss": 0.4949,
+      "step": 920
+    },
+    {
+      "epoch": 0.8905003625815808,
+      "grad_norm": 3.759092092514038,
+      "learning_rate": 1.579994297796808e-07,
+      "loss": 0.4791,
+      "step": 921
+    },
+    {
+      "epoch": 0.891467246797196,
+      "grad_norm": 3.479841470718384,
+      "learning_rate": 1.5526859897343815e-07,
+      "loss": 0.4447,
+      "step": 922
+    },
+    {
+      "epoch": 0.8924341310128112,
+      "grad_norm": 3.0901901721954346,
+      "learning_rate": 1.5256081743034336e-07,
+      "loss": 0.3394,
+      "step": 923
+    },
+    {
+      "epoch": 0.8934010152284264,
+      "grad_norm": 3.611321210861206,
+      "learning_rate": 1.4987611176852879e-07,
+      "loss": 0.4004,
+      "step": 924
+    },
+    {
+      "epoch": 0.8943678994440416,
+      "grad_norm": 3.6917665004730225,
+      "learning_rate": 1.472145083792842e-07,
+      "loss": 0.4116,
+      "step": 925
+    },
+    {
+      "epoch": 0.8953347836596568,
+      "grad_norm": 4.044205665588379,
+      "learning_rate": 1.44576033426799e-07,
+      "loss": 0.5017,
+      "step": 926
+    },
+    {
+      "epoch": 0.896301667875272,
+      "grad_norm": 3.850446939468384,
+      "learning_rate": 1.419607128479053e-07,
+      "loss": 0.5565,
+      "step": 927
+    },
+    {
+      "epoch": 0.8972685520908871,
+      "grad_norm": 3.598113775253296,
+      "learning_rate": 1.3936857235182233e-07,
+      "loss": 0.4689,
+      "step": 928
+    },
+    {
+      "epoch": 0.8982354363065023,
+      "grad_norm": 3.6002085208892822,
+      "learning_rate": 1.3679963741990127e-07,
+      "loss": 0.4742,
+      "step": 929
+    },
+    {
+      "epoch": 0.8992023205221175,
+      "grad_norm": 3.4773659706115723,
+      "learning_rate": 1.3425393330538023e-07,
+      "loss": 0.3849,
+      "step": 930
+    },
+    {
+      "epoch": 0.9001692047377327,
+      "grad_norm": 3.768770456314087,
+      "learning_rate": 1.317314850331314e-07,
+      "loss": 0.4926,
+      "step": 931
+    },
+    {
+      "epoch": 0.9011360889533478,
+      "grad_norm": 3.5853402614593506,
+      "learning_rate": 1.292323173994167e-07,
+      "loss": 0.4547,
+      "step": 932
+    },
+    {
+      "epoch": 0.902102973168963,
+      "grad_norm": 3.9394826889038086,
+      "learning_rate": 1.2675645497164352e-07,
+      "loss": 0.4413,
+      "step": 933
+    },
+    {
+      "epoch": 0.9030698573845782,
+      "grad_norm": 4.156927108764648,
+      "learning_rate": 1.2430392208812452e-07,
+      "loss": 0.5271,
+      "step": 934
+    },
+    {
+      "epoch": 0.9040367416001934,
+      "grad_norm": 3.9245412349700928,
+      "learning_rate": 1.2187474285783623e-07,
+      "loss": 0.5127,
+      "step": 935
+    },
+    {
+      "epoch": 0.9050036258158085,
+      "grad_norm": 3.750504970550537,
+      "learning_rate": 1.1946894116018404e-07,
+      "loss": 0.5002,
+      "step": 936
+    },
+    {
+      "epoch": 0.9059705100314237,
+      "grad_norm": 3.6622493267059326,
+      "learning_rate": 1.1708654064476743e-07,
+      "loss": 0.5017,
+      "step": 937
+    },
+    {
+      "epoch": 0.9069373942470389,
+      "grad_norm": 3.3695197105407715,
+      "learning_rate": 1.1472756473114493e-07,
+      "loss": 0.4487,
+      "step": 938
+    },
+    {
+      "epoch": 0.9079042784626541,
+      "grad_norm": 3.628037452697754,
+      "learning_rate": 1.1239203660860648e-07,
+      "loss": 0.4447,
+      "step": 939
+    },
+    {
+      "epoch": 0.9088711626782693,
+      "grad_norm": 3.471571922302246,
+      "learning_rate": 1.1007997923594555e-07,
+      "loss": 0.4211,
+      "step": 940
+    },
+    {
+      "epoch": 0.9098380468938845,
+      "grad_norm": 3.611616849899292,
+      "learning_rate": 1.0779141534123127e-07,
+      "loss": 0.4236,
+      "step": 941
+    },
+    {
+      "epoch": 0.9108049311094997,
+      "grad_norm": 3.4879791736602783,
+      "learning_rate": 1.0552636742158701e-07,
+      "loss": 0.4596,
+      "step": 942
+    },
+    {
+      "epoch": 0.9117718153251149,
+      "grad_norm": 3.495035409927368,
+      "learning_rate": 1.0328485774296875e-07,
+      "loss": 0.4096,
+      "step": 943
+    },
+    {
+      "epoch": 0.91273869954073,
+      "grad_norm": 3.4880497455596924,
+      "learning_rate": 1.010669083399457e-07,
+      "loss": 0.4351,
+      "step": 944
+    },
+    {
+      "epoch": 0.9137055837563451,
+      "grad_norm": 3.4250264167785645,
+      "learning_rate": 9.887254101548422e-08,
+      "loss": 0.4098,
+      "step": 945
+    },
+    {
+      "epoch": 0.9146724679719603,
+      "grad_norm": 3.6895461082458496,
+      "learning_rate": 9.670177734073221e-08,
+      "loss": 0.4395,
+      "step": 946
+    },
+    {
+      "epoch": 0.9156393521875755,
+      "grad_norm": 3.6795530319213867,
+      "learning_rate": 9.455463865481019e-08,
+      "loss": 0.398,
+      "step": 947
+    },
+    {
+      "epoch": 0.9166062364031907,
+      "grad_norm": 3.933271884918213,
+      "learning_rate": 9.243114606459741e-08,
+      "loss": 0.4955,
+      "step": 948
+    },
+    {
+      "epoch": 0.9175731206188059,
+      "grad_norm": 4.0044708251953125,
+      "learning_rate": 9.033132044452775e-08,
+      "loss": 0.4639,
+      "step": 949
+    },
+    {
+      "epoch": 0.9185400048344211,
+      "grad_norm": 3.7048990726470947,
+      "learning_rate": 8.825518243638226e-08,
+      "loss": 0.5338,
+      "step": 950
+    },
+    {
+      "epoch": 0.9195068890500363,
+      "grad_norm": 3.9390463829040527,
+      "learning_rate": 8.620275244908826e-08,
+      "loss": 0.4594,
+      "step": 951
+    },
+    {
+      "epoch": 0.9204737732656514,
+      "grad_norm": 3.997107744216919,
+      "learning_rate": 8.417405065851647e-08,
+      "loss": 0.4062,
+      "step": 952
+    },
+    {
+      "epoch": 0.9214406574812666,
+      "grad_norm": 3.6523690223693848,
+      "learning_rate": 8.216909700728498e-08,
+      "loss": 0.422,
+      "step": 953
+    },
+    {
+      "epoch": 0.9224075416968818,
+      "grad_norm": 3.660822629928589,
+      "learning_rate": 8.018791120456088e-08,
+      "loss": 0.459,
+      "step": 954
+    },
+    {
+      "epoch": 0.923374425912497,
+      "grad_norm": 4.010040760040283,
+      "learning_rate": 7.823051272586812e-08,
+      "loss": 0.5026,
+      "step": 955
+    },
+    {
+      "epoch": 0.9243413101281122,
+      "grad_norm": 3.4889373779296875,
+      "learning_rate": 7.629692081289608e-08,
+      "loss": 0.4264,
+      "step": 956
+    },
+    {
+      "epoch": 0.9253081943437274,
+      "grad_norm": 3.5010697841644287,
+      "learning_rate": 7.438715447331018e-08,
+      "loss": 0.471,
+      "step": 957
+    },
+    {
+      "epoch": 0.9262750785593425,
+      "grad_norm": 3.4076452255249023,
+      "learning_rate": 7.250123248056379e-08,
+      "loss": 0.3945,
+      "step": 958
+    },
+    {
+      "epoch": 0.9272419627749577,
+      "grad_norm": 3.8815410137176514,
+      "learning_rate": 7.063917337371495e-08,
+      "loss": 0.3895,
+      "step": 959
+    },
+    {
+      "epoch": 0.9282088469905729,
+      "grad_norm": 3.7122230529785156,
+      "learning_rate": 6.880099545724522e-08,
+      "loss": 0.3973,
+      "step": 960
+    },
+    {
+      "epoch": 0.929175731206188,
+      "grad_norm": 3.873793125152588,
+      "learning_rate": 6.698671680087643e-08,
+      "loss": 0.4093,
+      "step": 961
+    },
+    {
+      "epoch": 0.9301426154218032,
+      "grad_norm": 3.5222861766815186,
+      "learning_rate": 6.519635523939666e-08,
+      "loss": 0.4055,
+      "step": 962
+    },
+    {
+      "epoch": 0.9311094996374184,
+      "grad_norm": 3.970353841781616,
+      "learning_rate": 6.342992837248235e-08,
+      "loss": 0.4792,
+      "step": 963
+    },
+    {
+      "epoch": 0.9320763838530336,
+      "grad_norm": 3.4243791103363037,
+      "learning_rate": 6.168745356452676e-08,
+      "loss": 0.387,
+      "step": 964
+    },
+    {
+      "epoch": 0.9330432680686488,
+      "grad_norm": 3.5594351291656494,
+      "learning_rate": 5.996894794446817e-08,
+      "loss": 0.4666,
+      "step": 965
+    },
+    {
+      "epoch": 0.934010152284264,
+      "grad_norm": 3.504962205886841,
+      "learning_rate": 5.8274428405623307e-08,
+      "loss": 0.4723,
+      "step": 966
+    },
+    {
+      "epoch": 0.9349770364998792,
+      "grad_norm": 3.720414876937866,
+      "learning_rate": 5.660391160551837e-08,
+      "loss": 0.4996,
+      "step": 967
+    },
+    {
+      "epoch": 0.9359439207154944,
+      "grad_norm": 3.542029857635498,
+      "learning_rate": 5.495741396572829e-08,
+      "loss": 0.4294,
+      "step": 968
+    },
+    {
+      "epoch": 0.9369108049311095,
+      "grad_norm": 3.799464225769043,
+      "learning_rate": 5.333495167171354e-08,
+      "loss": 0.4009,
+      "step": 969
+    },
+    {
+      "epoch": 0.9378776891467246,
+      "grad_norm": 4.372755527496338,
+      "learning_rate": 5.17365406726611e-08,
+      "loss": 0.5372,
+      "step": 970
+    },
+    {
+      "epoch": 0.9388445733623398,
+      "grad_norm": 3.79471492767334,
+      "learning_rate": 5.016219668132871e-08,
+      "loss": 0.4428,
+      "step": 971
+    },
+    {
+      "epoch": 0.939811457577955,
+      "grad_norm": 3.7523727416992188,
+      "learning_rate": 4.8611935173889235e-08,
+      "loss": 0.472,
+      "step": 972
+    },
+    {
+      "epoch": 0.9407783417935702,
+      "grad_norm": 3.699140787124634,
+      "learning_rate": 4.708577138977932e-08,
+      "loss": 0.4489,
+      "step": 973
+    },
+    {
+      "epoch": 0.9417452260091854,
+      "grad_norm": 3.6899728775024414,
+      "learning_rate": 4.5583720331549563e-08,
+      "loss": 0.5004,
+      "step": 974
+    },
+    {
+      "epoch": 0.9427121102248006,
+      "grad_norm": 3.827622652053833,
+      "learning_rate": 4.410579676471571e-08,
+      "loss": 0.493,
+      "step": 975
+    },
+    {
+      "epoch": 0.9436789944404158,
+      "grad_norm": 3.3393099308013916,
+      "learning_rate": 4.265201521761603e-08,
+      "loss": 0.4271,
+      "step": 976
+    },
+    {
+      "epoch": 0.9446458786560309,
+      "grad_norm": 3.922778367996216,
+      "learning_rate": 4.1222389981265546e-08,
+      "loss": 0.4808,
+      "step": 977
+    },
+    {
+      "epoch": 0.9456127628716461,
+      "grad_norm": 3.676297903060913,
+      "learning_rate": 3.9816935109218414e-08,
+      "loss": 0.4545,
+      "step": 978
+    },
+    {
+      "epoch": 0.9465796470872613,
+      "grad_norm": 3.3293559551239014,
+      "learning_rate": 3.843566441742774e-08,
+      "loss": 0.4173,
+      "step": 979
+    },
+    {
+      "epoch": 0.9475465313028765,
+      "grad_norm": 3.5446856021881104,
+      "learning_rate": 3.7078591484110136e-08,
+      "loss": 0.4439,
+      "step": 980
+    },
+    {
+      "epoch": 0.9485134155184917,
+      "grad_norm": 3.5858993530273438,
+      "learning_rate": 3.574572964961304e-08,
+      "loss": 0.405,
+      "step": 981
+    },
+    {
+      "epoch": 0.9494802997341069,
+      "grad_norm": 3.8750436305999756,
+      "learning_rate": 3.443709201628287e-08,
+      "loss": 0.4514,
+      "step": 982
+    },
+    {
+      "epoch": 0.950447183949722,
+      "grad_norm": 3.9488742351531982,
+      "learning_rate": 3.3152691448336825e-08,
+      "loss": 0.4698,
+      "step": 983
+    },
+    {
+      "epoch": 0.9514140681653372,
+      "grad_norm": 3.8947930335998535,
+      "learning_rate": 3.189254057173491e-08,
+      "loss": 0.4765,
+      "step": 984
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 3.8969860076904297,
+      "learning_rate": 3.065665177405808e-08,
+      "loss": 0.4567,
+      "step": 985
+    },
+    {
+      "epoch": 0.9533478365965675,
+      "grad_norm": 3.5067756175994873,
+      "learning_rate": 2.9445037204385297e-08,
+      "loss": 0.4525,
+      "step": 986
+    },
+    {
+      "epoch": 0.9543147208121827,
+      "grad_norm": 3.490902900695801,
+      "learning_rate": 2.825770877317363e-08,
+      "loss": 0.4457,
+      "step": 987
+    },
+    {
+      "epoch": 0.9552816050277979,
+      "grad_norm": 3.8223989009857178,
+      "learning_rate": 2.7094678152141396e-08,
+      "loss": 0.4631,
+      "step": 988
+    },
+    {
+      "epoch": 0.9562484892434131,
+      "grad_norm": 3.704549551010132,
+      "learning_rate": 2.5955956774154633e-08,
+      "loss": 0.3899,
+      "step": 989
+    },
+    {
+      "epoch": 0.9572153734590283,
+      "grad_norm": 3.6381337642669678,
+      "learning_rate": 2.4841555833112764e-08,
+      "loss": 0.4808,
+      "step": 990
+    },
+    {
+      "epoch": 0.9581822576746435,
+      "grad_norm": 3.9116270542144775,
+      "learning_rate": 2.3751486283840884e-08,
+      "loss": 0.4986,
+      "step": 991
+    },
+    {
+      "epoch": 0.9591491418902587,
+      "grad_norm": 3.7075469493865967,
+      "learning_rate": 2.268575884197932e-08,
+      "loss": 0.4586,
+      "step": 992
+    },
+    {
+      "epoch": 0.9601160261058739,
+      "grad_norm": 3.487811326980591,
+      "learning_rate": 2.1644383983880356e-08,
+      "loss": 0.4241,
+      "step": 993
+    },
+    {
+      "epoch": 0.961082910321489,
+      "grad_norm": 3.439530372619629,
+      "learning_rate": 2.0627371946504992e-08,
+      "loss": 0.4778,
+      "step": 994
+    },
+    {
+      "epoch": 0.9620497945371042,
+      "grad_norm": 3.838062286376953,
+      "learning_rate": 1.9634732727321636e-08,
+      "loss": 0.4545,
+      "step": 995
+    },
+    {
+      "epoch": 0.9630166787527193,
+      "grad_norm": 3.6535823345184326,
+      "learning_rate": 1.866647608420813e-08,
+      "loss": 0.4729,
+      "step": 996
+    },
+    {
+      "epoch": 0.9639835629683345,
+      "grad_norm": 3.867074489593506,
+      "learning_rate": 1.7722611535355426e-08,
+      "loss": 0.4372,
+      "step": 997
+    },
+    {
+      "epoch": 0.9649504471839497,
+      "grad_norm": 3.6077470779418945,
+      "learning_rate": 1.6803148359175182e-08,
+      "loss": 0.4406,
+      "step": 998
+    },
+    {
+      "epoch": 0.9659173313995649,
+      "grad_norm": 3.9167864322662354,
+      "learning_rate": 1.5908095594207585e-08,
+      "loss": 0.4684,
+      "step": 999
+    },
+    {
+      "epoch": 0.9668842156151801,
+      "grad_norm": 4.19531774520874,
+      "learning_rate": 1.503746203903228e-08,
+      "loss": 0.4953,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9678510998307953,
+      "grad_norm": 4.0359601974487305,
+      "learning_rate": 1.4191256252182595e-08,
+      "loss": 0.4722,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9688179840464104,
+      "grad_norm": 3.364161968231201,
+      "learning_rate": 1.336948655206144e-08,
+      "loss": 0.361,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9697848682620256,
+      "grad_norm": 3.708479642868042,
+      "learning_rate": 1.2572161016858874e-08,
+      "loss": 0.4385,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9707517524776408,
+      "grad_norm": 3.9538137912750244,
+      "learning_rate": 1.179928748447301e-08,
+      "loss": 0.4849,
+      "step": 1004
+    },
+    {
+      "epoch": 0.971718636693256,
+      "grad_norm": 3.6462512016296387,
+      "learning_rate": 1.1050873552433394e-08,
+      "loss": 0.4823,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9726855209088712,
+      "grad_norm": 4.054479122161865,
+      "learning_rate": 1.0326926577825247e-08,
+      "loss": 0.5613,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9736524051244864,
+      "grad_norm": 3.732670545578003,
+      "learning_rate": 9.627453677218402e-09,
+      "loss": 0.4506,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9746192893401016,
+      "grad_norm": 3.2570137977600098,
+      "learning_rate": 8.952461726596528e-09,
+      "loss": 0.4087,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9755861735557168,
+      "grad_norm": 3.9363338947296143,
+      "learning_rate": 8.301957361289969e-09,
+      "loss": 0.4716,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9765530577713318,
+      "grad_norm": 3.6260664463043213,
+      "learning_rate": 7.675946975909953e-09,
+      "loss": 0.5033,
+      "step": 1010
+    },
+    {
+      "epoch": 0.977519941986947,
+      "grad_norm": 3.406822919845581,
+      "learning_rate": 7.074436724286704e-09,
+      "loss": 0.456,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9784868262025622,
+      "grad_norm": 3.807194948196411,
+      "learning_rate": 6.497432519407543e-09,
+      "loss": 0.4096,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9794537104181774,
+      "grad_norm": 3.6305527687072754,
+      "learning_rate": 5.944940033360269e-09,
+      "loss": 0.3987,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9804205946337926,
+      "grad_norm": 3.2959699630737305,
+      "learning_rate": 5.4169646972762615e-09,
+      "loss": 0.3819,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9813874788494078,
+      "grad_norm": 3.903257369995117,
+      "learning_rate": 4.913511701278017e-09,
+      "loss": 0.447,
+      "step": 1015
+    },
+    {
+      "epoch": 0.982354363065023,
+      "grad_norm": 3.6549253463745117,
+      "learning_rate": 4.434585994427254e-09,
+      "loss": 0.4847,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9833212472806382,
+      "grad_norm": 3.9208667278289795,
+      "learning_rate": 3.98019228467661e-09,
+      "loss": 0.4342,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9842881314962533,
+      "grad_norm": 3.740576982498169,
+      "learning_rate": 3.5503350388232936e-09,
+      "loss": 0.4214,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9852550157118685,
+      "grad_norm": 3.650252342224121,
+      "learning_rate": 3.1450184824657892e-09,
+      "loss": 0.4118,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9862218999274837,
+      "grad_norm": 3.3320415019989014,
+      "learning_rate": 2.7642465999613843e-09,
+      "loss": 0.4433,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9871887841430989,
+      "grad_norm": 3.4924774169921875,
+      "learning_rate": 2.408023134387871e-09,
+      "loss": 0.4358,
+      "step": 1021
+    },
+    {
+      "epoch": 0.988155668358714,
+      "grad_norm": 3.6019985675811768,
+      "learning_rate": 2.0763515875055207e-09,
+      "loss": 0.4682,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9891225525743292,
+      "grad_norm": 3.7121329307556152,
+      "learning_rate": 1.7692352197240525e-09,
+      "loss": 0.4277,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9900894367899444,
+      "grad_norm": 3.643587589263916,
+      "learning_rate": 1.4866770500704398e-09,
+      "loss": 0.4418,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9910563210055596,
+      "grad_norm": 3.526364803314209,
+      "learning_rate": 1.2286798561572666e-09,
+      "loss": 0.4761,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9920232052211748,
+      "grad_norm": 3.6432149410247803,
+      "learning_rate": 9.952461741585816e-10,
+      "loss": 0.4749,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9929900894367899,
+      "grad_norm": 3.4780168533325195,
+      "learning_rate": 7.863782987821422e-10,
+      "loss": 0.4458,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9939569736524051,
+      "grad_norm": 3.4899373054504395,
+      "learning_rate": 6.020782832488747e-10,
+      "loss": 0.4496,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9949238578680203,
+      "grad_norm": 3.6461098194122314,
+      "learning_rate": 4.4234793927094845e-10,
+      "loss": 0.4434,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9958907420836355,
+      "grad_norm": 3.9469432830810547,
+      "learning_rate": 3.0718883703567684e-10,
+      "loss": 0.4967,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9968576262992507,
+      "grad_norm": 3.649221658706665,
+      "learning_rate": 1.9660230518886436e-10,
+      "loss": 0.4199,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9978245105148659,
+      "grad_norm": 3.793912172317505,
+      "learning_rate": 1.1058943082203898e-10,
+      "loss": 0.4808,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9987913947304811,
+      "grad_norm": 3.6841025352478027,
+      "learning_rate": 4.915105946246002e-11,
+      "loss": 0.4407,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9997582789460963,
+      "grad_norm": 3.665886640548706,
+      "learning_rate": 1.2287795063681274e-11,
+      "loss": 0.4681,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9997582789460963,
+      "step": 1034,
+      "total_flos": 1.8974566881186284e+18,
+      "train_loss": 0.49419337980867123,
+      "train_runtime": 23447.7413,
+      "train_samples_per_second": 1.411,
+      "train_steps_per_second": 0.044
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1034,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8974566881186284e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}